252 12 12MB
English Pages 380 [407] Year 2022
Mill’s Principle of Utility
Value Inquiry Book Series Founding Editor Robert Ginsberg Managing Editor J.D. Mininger
volume 366
Philosophy, Literature, and Politics Edited by J.D. Mininger (lcc International University)
The titles published in this series are listed at brill.com/vibs and brill.com/plp
Mill’s Principle of Utility Origins, Proof, and Implications revised and enlarged edition
By
Necip Fikri Alican
LEIDEN | BOSTON
Revised and enlarged edition of Mill’s Principle of Utility: A Defense of John Stuart Mill’s Notorious Proof. Amsterdam: Editions Rodopi b.v., 1994. Cover illustration: The Miriam and Ira D. Wallach Division of Art, Prints and Photographs: Print Collection, The New York Public Library. “John Stuart Mill, M.P.” The New York Public Library Digital Collections. 1861–1880. https://digitalcollections.nypl.org/items/510d47da-2877-a3d9-e040-e00a18064a99 The Library of Congress Cataloging-in-Publication Data is available online at https://catalog.loc.gov lc record available at https://lccn.loc.gov/2021055542
Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface. issn 0929-8 436 isbn 978-9 0-0 4-5 0387-8 (hardback) isbn 978-9 0-0 4-5 0395-3 (e-book) Copyright 2022 by Necip Fikri Alican. Published by Koninklijke Brill nv, Leiden, The Netherlands. Koninklijke Brill nv incorporates the imprints Brill, Brill Nijhoff, Brill Hotei, Brill Schöningh, Brill Fink, Brill mentis, Vandenhoeck & Ruprecht, Böhlau Verlag and V&R Unipress. Koninklijke Brill nv reserves the right to protect this publication against unauthorized use. Requests for re-use and/or translations must be addressed to Koninklijke Brill nv via brill.com or copyright.com. This book is printed on acid-free paper and produced in a sustainable manner.
To my wife Banu Beste Başol Alican
∵
Contents Foreword xi Preface xviii Acknowledgments xxiii A Note on References xxv Introduction 1
part 1 Mill’s Principle of Utility and Scholarly Reactions to His Proof: Exegetical and Historical Background 1 Classical Utilitarianism before John Stuart Mill The Legacy of Jeremy Bentham 13 1.1 History of Utilitarianism: Sources and References 14 1.2 Bentham as an Anchor for Historical Insight 30 1.3 Origins and Development of Utilitarian Nomenclature 39 1.4 Bentham’s Debt to Predecessors and Contemporaries 53 1.5 Patterns of Indirect Inspiration and Transmission 74 1.6 Bentham’s Own Terminological Predilections 77 2 The Nature and Function of Mill’s Principle of Utility 92 2.1 What Is Bentham’s Principle of Utility? 96 2.2 What Is Mill’s Principle of Utility? 98 2.3 The Multiple Functions of Mill’s Principle of Utility 106 2.3.1 PU1: Theory of Value 109 2.3.2 PU2: Theory of Obligation 109 2.3.3 PU3: Theory of Justification 111 2.4 Mill’s Apparent Definitions of the Principle of Utility 112 2.5 The Primary Function of Mill’s Principle of Utility 115 2.6 The Structure of Mill’s Proof of the Principle of Utility 118 3 The Historical Reception of Mill’s Proof of the Principle of Utility 122 3.1 John Grote: An Examination of the Utilitarian Philosophy 127 3.2 Henry Sidgwick: The Methods of Ethics 134 3.3 Francis Herbert Bradley: Ethical Studies 137
viii Contents
3.4 William Ritchie Sorley: The Ethics of Naturalism: A Criticism 141 3.5 John Dewey: Outlines of a Critical Theory of Ethics 144 3.6 John Dewey and James Hayden Tufts: Ethics 146 3.7 John Stuart Mackenzie: A Manual of Ethics 148 3.8 George Edward Moore: Principia Ethica 152
part 2 The Alleged Fallacies in Mill’s Proof of the Principle of Utility: Analysis and Response 4 The Alleged Fallacy of Equivocation in Mill’s Proof 161 4.1 Charges against Mill 161 4.2 Analysis of the Charges 163 4.2.1 False Discount Factor in Critical Evaluation 164 4.2.2 Multiple Interpretations of “Desirable” 165 4.2.3 Alternative Formulations of Mill’s Argument 167 4.3 Response to the Charges 169 4.3.1 The Principle of Charity in Critical Evaluation 170 4.3.2 The Methodology behind Mill’s Proof 173 4.3.3 Desires as Evidence of Desirability 178 4.3.4 Distinction between Means and Ends 181 The Alleged Fallacy of Composition in Mill’s Proof 185 5 5.1 Charges against Mill 186 5.2 Analysis of the Charges 193 5.2.1 First Scenario 194 5.2.2 Second Scenario 198 5.2.3 Third Scenario 200 5.2.4 Fourth Scenario 202 5.3 Response to the Charges 208 5.3.1 Deconstruction of the Fallacy of Composition 208 5.3.2 Mill’s Conception of the General Happiness 216 5.3.3 Mill’s Conception of the Aggregate of All Persons 223 5.4 Critical Summary of the Response 228 The Alleged Naturalistic Fallacy in Mill’s Proof 232 6 6.1 Moore’s Broad Construal of the Naturalistic Fallacy 233 6.2 The Naturalistic Fallacy Anchored to the Good 239 6.3 The Alleged Naturalistic Fallacy in Mill’s Proof 244
Contents
part 3 Mill’s Proof of the Principle of Utility: Reconstruction and Implications 7 Reconstruction of Mill’s Proof of the Principle of Utility 255 7.1 First Part of the Proof 256 7.1.1 The Logical Role of Emphasis on the General Happiness 259 7.1.2 The Moral Implications of Emphasis on the General Happiness 263 7.2 Second Part of the Proof 267 Implications of Mill’s Proof of the Principle of Utility 280 8 8.1 Implications for a Theory of Moral Obligation 281 8.2 Directions for Further Research 291 8.2.1 Act Utilitarianism vs. Rule Utilitarianism 292 8.2.2 Actual, Intended, and Foreseeable Consequences 334 8.2.3 Total and Average Happiness 337 Works Cited 345 Index 375
ix
Foreword An adequate grasp of the origins, substance, and ingredients of anything, whatever it may be, illuminates the essential nature, structure, and characteristics that define that thing as an individual entity with a unique identity. This is true whether the thing in question is physical or mental, natural or artificial, simple or complex. The light of this knowledge enables the mind to comprehend any object of inquiry in the fullness of its being and truth. We may characterize this type of comprehension as holistic, because the subject gains epistemic access not just through the compositional structure of the object’s internal features but also through the causal efficacy of its external relations. The internal dimension corresponds to the capacities or powers that inhere in the object, while the external dimension concerns the effect, impact, or influence it has on other objects. We know the object in the fullness of its being and truth only when we comprehend its essential nature and principal function. I know this cat, for example, not merely through my acquaintance with its anatomy, which tells me how it differs in that regard from other animals, including other cats, but also through my understanding of its tendencies and capabilities. Comprehensive insight of this sort is made possible by a metaphysical process whereby the object derives its being and nature from the source that gives rise to it. This derivation is a complex and dynamic process: complex because a multiplicity of factors contribute to the generation of the object; dynamic because, whether it is viewed in parts or as a whole, reality is constantly changing. Despite the constancy of change, however, the object retains its identity as its succeeding state of being preserves (acquires and assimilates) the essential nature and structure of its preceding state of being. The kind or scope of the process in which this retention takes place depends on the conditions under which the change takes place. But regardless of the kind, complexity, or magnitude of these conditions, the change is always a progressive process. It is indeed a procession. We may reasonably say that, insofar as it is a constitutive element of the fabric of reality, any object, natural or artificial, is always an object-in-the-making. It is endlessly developing, disintegrating, and becoming different than it was just a moment ago. The basis of this process is the inherent capacity of the object to transfer its essential features or elements from the preceding to the succeeding state of its being. When we construe an object as constantly changing, forever turning into something else without losing its identity, we imply that it is essentially a temporal or historical reality. Its present state of being sums up its life history from its inception onward. The life of a human being, or the history of an object, is a
xii Foreword slice of time as well as a trajectory in time. Every new experience contributes to the emergence of a new state of being, determined by and fashioned out of the one hosted in and by the preceding moment. What it is now, in the specious present, is the result of the experiences it has undergone since the moment it came into being. A human being is its history, and its history is the creative synthesis of its experiences. It is a generally accepted truism that we know a person through her deeds, not through what she thinks, feels, or imagines she is (although what she thinks, feels, or imagines she is may be true). The essence of the human “I,” that is to say, whatever it is that breathes life into the cogito, is neither a physical entity nor a mental phenomenon. Nor can it be reduced to one or the other. It is a dynamic reality revealing itself in the unity of its thoughts, feelings, and actions. It is its deeds, more than anything else, and its deeds are its experiences. Although it is constantly changing, its identity endures. Now, suppose that a philosopher with a psychological bent of mind wants to get to know me as a human being that exists in the world at this particular time. How can she come to know me? She may examine my body, as a start, but I am not my body. My body is merely an aspect, attribute, component, or constituent of the person she wants to know. The “I” that presides over my life is grounded in my body, and it governs the development of my individual identity, but in ontological terms, it is uniquely different from my body. Broadly speaking, the “I” that interacts with itself, with others, and with its environment is different from the body with which it is associated. But what, then, could this curious philosopher possibly know about me? She cannot enter into my mind, or even peek into it, but she can ask me questions whose answers may reveal something about me. Knowledge attained in this way is not knowledge by acquaintance, yet it can still be reliable and truthful. The knowledge she acquires through dialogue will reflect a dynamic reality that has long been growing and developing in various different ways, naturally reaching its latest stage of development during the corresponding interview with the inquiring philosopher. The trajectory that was my being during this stretch of time will no longer exist during the interview. What exists at that point will be the synthesis of the essential nature and structure of my historical being. Significant events, whether they be actions, activities, adventures, achievements, or projects, serve as the building blocks of this existential synthesis. Our knowledge of any human being thus comes from the deeds that make up her life. Do we not identify people—farmers, soldiers, teachers, lawyers, engineers—by what they do and what they achieve? Such knowledge is adequate inasmuch as it is derived from an analytical, critical, and objective examination of the natural, social, and cultural environment in which the subject
Foreword
xiii
lives her life. I assume here that a human being is an integral part of the natural and social environment providing the material conditions of her growth and development as an existential reality. The kind of environment in which the individual flourishes plays a significant role in the kind of person that she becomes. Do we not come to know people in general through the kind of life that they lead and the conditions under which they thrive? Does the psychoanalyst not probe the early history of her patient in an effort to understand her present situation? Why indeed do philosophers such as Plato, Spinoza, Hegel, and Whitehead, to name a few of the brightest minds in the history of philosophy, inquire into the source, origin, cause, or initial state of the universe? Was the supreme passion of a cosmologist such as Stephen Hawking not directed at the source of the cosmos as he tried to understand its nature, structure, purpose, and processes? Was this passion any different from the mission of Thales to discover the arche underlying the existence and meaning of the universe? Do counselors, lawyers, parents, and teachers—basically any authority figure seeking a solution to a problem—not begin with an exploration of the cause of the problem? Can we solve any problem rationally if we do not know the facts that make up the whole and the factors that cause it? I do not assume, in this series of questions, that discovering the source of a problem is sufficient for solving it. No, I assume simply that the discovery illuminates the nature and structure of the problem and that such illumination is essential for contemplating and formulating a solution. The book you are about to read meets this condition competently and admirably. Keeping in mind that knowledge of the source of an object illuminates its essential nature, and that the light of this knowledge enables the mind to comprehend it in the fullness of its being and truth, we can now ask, what is the source of the philosophical work? What is the initial state from which it originates? Again, what is the stuff out of which it is made or fashioned? Perhaps I can be more specific to some benefit: If the essence of philosophical activity is thinking, accordingly, if the philosopher is a thinker, and as a thinker she is a seeker of knowledge, what does she think about and what does she seek to know? We can say that she constructs ideas, propositions, arguments, theories, and systems of thought. But what does she communicate in these and similar constructions? What is their stuff? I would not be mistaken if I say that, as a thinker, the philosopher is a seeker of knowledge. Otherwise, the ancient claim that she is a lover of wisdom would be either spurious or misleading. But what kind of knowledge does she love and seek? She loves all types of knowledge, yet she cannot seek all of them. It has been clear for some time now, perhaps since the turn of the previous century, that identifying and verifying the facts that make up the scheme of
xiv Foreword nature is the task of the empirical scientist, while uncovering and interpreting the meaning of these facts, and consequently of the universe as a whole, is the task of the philosopher. Let me state at once that the realm of meaning is the realm of human values: goodness, truth, beauty, and their derivatives. Each of these concepts stands for a type or category of value. For example, goodness includes values such as justice, courage, love, and toleration; truth includes values such as wisdom, erudition, deliberation, and understanding; and beauty includes values such as grace, elegance, sublimity, and tragedy. These values exist as ideals—as schemas or plans of action. They are “objects” of desire and aspiration. We prize them because they originate as responses to the essential needs of human nature. Such values come to life as goals and problems, typically expressed in the form of questions. They revolve specifically around three pivotal questions: How should we live? How should we love? How should we die? Whether directly or indirectly, an answer to at least one of these questions underlies our every pursuit in terms of what we desire, what we hope for, and what we enjoy, in short, in connection with our quest for happiness. If I am to express this point more succinctly, I can say that these values center on, and are founded upon, the category of importance: That which is valuable is important, and therefore worth seeking, because it creates a deep feeling of satisfaction, fulfillment, completion, and inner growth. However, it is not enough to say that human values arise as a response to the basic needs of human nature. It is, moreover, critically important to know the dynamics that underlie the mode of existence of human nature as well as the conditions under which it exists and flourishes. We are temporal beings. We do not choose our existence. Our humanity is not given as a ready-made reality but as a potentiality awaiting realization under certain conditions in a certain environment. We do not live in a friendly habitat. Life is neither convenient nor meaningful. It can certainly become convenient or meaningful, sometimes even both, but the point is that neither convenience nor meaning is a natural part of life on earth. We typically have to create either one, or at least actively seek it out, rather than accidentally stumbling upon it. We do not even know why we exist. We understand the mechanics, of course, of how we come into being, but we do not know why we are here. Why do we exist at all? Why, in fact, does the universe exist? The realm of nature is a realm of brute fact, wholly devoid of value, yet it exhibits some type of order, a rational and comprehensible order, perhaps even a cosmic order pointing to the possibility of an ultimate force or creator. Suppose that such a cosmic creator exists. Why would it create this and not some other cosmic order? Why would it create anything at all? Why indeed is there something rather than nothing? Can we really answer such
Foreword
xv
questions unless we proceed with an adequate understanding of the source of the scheme of nature and the place of humanity in that scheme? The purpose of the preceding remarks is not to advance a theory of value, nor to answer any of the questions it raises, but only to emphasize that, while the task of the scientist is to discover the facts of nature, the task of the philosopher is to understand the facts that make up the realm of meaning, namely human values. This realm provides the datum of philosophical inquiry. I do not exaggerate if I say that the most important element of the schools and systems of thought that constitute the world of philosophy is their approach to value: ethics, aesthetics, and social and political theory, all drawing on a common account of the meaning of existence in general and the existence of humanity in particular. This claim is based on the fundamental assumption that any theory of goodness, justice, or beauty presupposes a certain understanding of human nature and the universe. Can the metaphysician theorize about the purpose or meaning of existence without some understanding of the universe? What is the source or basis of this understanding if not the knowledge provided by the scientist? Can the ethicist theorize about the nature of goodness without some understanding of human nature? Now, if the task of the scientist is to discover the facts of nature, and the task of the philosopher is to uncover the meaning of these facts, what is the task of the artist? What is the datum of her reflection? When she creates a work of art, what is it that she creates? Creation is an activity in which a new reality comes into being. What, then, does the artist bring into being? It is clear that she does not bring into being the corresponding medium of expression, for example, the words, the sounds, or the marble that she uses. She does not create any of those things; she simply shapes and combines them in a certain way. Our greatest clue for an answer is in the works of art standing the test of time and persisting as monuments of the human spirit. We know that masterpieces in the realm of music, painting, sculpture, film, dance, theater, and literature are all uniquely valuable. But what makes them so? Why do we prize them so much that we end up erecting majestic buildings in their honor—museums, art centers, galleries, opera houses, dance halls, theaters—the way the ancient Greeks and Romans built temples for their gods? The ancients knew why they honored their gods. Why do we, indeed why should we, honor the arts? Are our museums, art centers, galleries, opera houses, and so on, temples of Beauty? I think the majority of artists and philosophers would agree with me if I proposed that human meaning is not only the datum of reflection in artistic creation but also the stuff out of which any work of art is created. If we were to cast an investigative look at artistic masterpieces in various cultures, whether in the West or in the East, we would invariably find that they revolve around
xvi Foreword the values of goodness, truth, and beauty, together with their derivatives, and more directly around the questions of life, love, and death. The fountain of inspiration motivating the artist and the philosopher alike is one and the same fountain. Accordingly, the stuff out of which both the work of art and the work of philosophy are made is meaning. Did Plato not ban the artist from the ideal state because he viewed art as imitation and therefore as lacking in reality and truth? The realm of values nourishing the mind of the artist and the philosopher as a wealth of potentiality, on the one hand, and the way that people understand and assimilate these values in their endeavor to fulfill themselves as individuals, on the other, is the source of both art and philosophy. But what, then, is the creative difference between the philosopher and the artist? I tend to think that the difference is in their means and modes of articulating and communicating meaning. All human expression and communication is symbolic in nature. The philosopher thinks and communicates her insight, knowledge, and understanding by means of concepts, which are units of meaning expressed in words. Every word and every syntactical formation of words—be it a sentence, a paragraph, or a line of reasoning—is a way of shaping meaning in accordance with established rules, conventions, and practices. The philosopher thus communicates meaning conceptually in the medium of thinking. The artist, in contrast, communicates meaning symbolically in the medium of imagination, thinking as she does in terms of images representing her insight, knowledge, and understanding. An image can be visible, audible, literal, allegorical, iconic, dynamic, or static. The painter thinks in terms of lines and colors, the sculptor in terms of marble or bronze, the musician in terms of sounds, the dancer in terms of movement, the novelist in terms of depiction, and the dramatist in terms of action. The philosopher analyzes, argues, demonstrates, clarifies, and evaluates. The artist, on the other hand, presents an image serving as a luminous insight into a slice of human meaning. I do not deduce, ponder, or contemplate the truth that the image reveals. I perceive it directly in the fullness of its being. The artistic image does not have interior or exterior dimensions. We see it and we feel it as a reality in itself. For example, I do not think (of or about) alienation when I read Camus’s The Stranger. I simply see it and feel it directly in the work. Yet I do think (of or about) alienation when I read Sartre’s Being and Nothingness. Our acquaintance with truth and reality is not just a matter of comprehension but also a process of perception through the mind’s eye. I proposed at the beginning that an adequate understanding of the source of any object, natural or artificial, illuminates its essential nature and structure, and that the light of this knowledge enables the mind to comprehend it in the fullness of its being and truth. I can now add another proposal: A work of philosophy can be artistic and a work of art can be philosophical. The
Foreword
xvii
reasonableness of this assertion is based on the fact that both philosophy and art aim at and originate from the same source: reflection on a dimension of human meaning, either a human value or a spectrum of such values. The philosopher articulates the content of her reflection in one symbolic form, namely concepts, while the artist articulates the content of her reflection in another symbolic form, namely musical, dynamical, or theatrical form. This does not mean that every work of philosophy is artistic and that every work of art is philosophical. It means only that it is, in principle, possible for a philosophical work to be artistic and for artwork to be philosophical. Some of the greatest works of literature, which are consequently magnificent works of art, for example, Tolstoy’s The Death of Ivan Ilych or Melville’s Moby Dick, are indeed philosophical, and some of the greatest works of philosophy, for example, Plato’s Republic or Spinoza’s Ethics, are truly artistic. This assertion is based on the twofold assumption that artwork can communicate philosophical insight and that philosophical work can possess aesthetic qualities, including beauty, charm, elegance, harmony, majesty, and wisdom. It may seem strange to say that a conceptual construct, such as a philosophical idea, analysis, theory, or line of reasoning, can possess aesthetic qualities, but it is quite natural that it should, and rather delightful when it does. Is the novel, the poem, the play, or the short story not a conceptual construct? Do we not ascribe aesthetic qualities to literary works? Does the novel as a world of meaning not inhere in a story as a conceptual construct? Again, does the work of art qua art not come to life in the aesthetic experience as a spiritual object? The reality of art, pace Plato, who rejected art but was himself an artist, easily rivals that of the mountains, rivers, trees, and animals that populate the scheme of nature. I shall not comment on the philosophical and artistic dimensions of the book that will soon enough be unfolding before your eyes, but please allow me to conclude my testimony with the following remark: When a work of philosophy is also a work of art, it is no longer strictly philosophy or purely art. The concept becomes an image as the image becomes a concept. Truth streams into the mind on the wings of beauty. And when truth and beauty thus come together in a human creation, they blend into each other so that truth becomes beautiful and beauty truthful. This kind of merger is one of the mind’s highest aspirations, and the resulting union, one of its greatest achievements. That is what awaits readers in Mill’s Principle of Utility. Michael H. Mitias Professor Emeritus of Philosophy Millsaps College
Preface This book is a critical analysis of the third paragraph of the fourth chapter of John Stuart Mill’s Utilitarianism. That is not all that it is, to be sure, but that is indeed the center of attention. Such a precise description may constitute an exaggeration of the level of specificity to be found here, perhaps even a caricature of the intensity of concentration likely to be seen or tolerated anywhere. And it may consequently remain too vague to accommodate the time-honored intellectual ritual of inspecting the preface for signs of promise in the rest of the book. But it does provide the most accurate orientation toward what to expect between the covers. To elaborate, then, this book is a defense of Mill’s proof of the principle of utility. Its coverage is not, strictly speaking, limited to a single paragraph in Utilitarianism. It also includes several other paragraphs in that essay, as well as some in Mill’s other works, not to mention a few others by other people. However, since the proof is laid out in its entirety in the fourth chapter of Utilitarianism, never to be repeated, revised, or revisited elsewhere, it would not be grossly misleading to describe the present volume as a commentary on the fourth chapter of Mill’s Utilitarianism. And considering that proponents and opponents alike tend to confine the proof to the third paragraph of that chapter, it would not be entirely hyperbolic to say that this book is largely about the third paragraph of the fourth chapter of Mill’s Utilitarianism. The last thing I wrote on a subject of such limited scope was my first term paper in college. It was a report on the planetary ring system of Saturn, bearing a descriptive title communicating the subject matter: “The Rings of Saturn.” I wrote that piece rather reluctantly and turned it in as the final assignment in English 101, the basic course introducing college students to formal writing. The requirement was to prepare a research paper of standard length on any topic so long as the plan met with the instructor’s prior approval. My reluctance was grounded in a profound frustration with the approval mechanism. And it grew in proportion to the instructor’s relentless opposition to my various proposals revolving around a central theme that otherwise seemed perfectly reasonable to me. My first choice, a paper on the universe, was rejected out of hand, despite an ambitious yet feasible plan accompanied by a promising outline, both of which were concerned specifically with what is in the universe rather than with what the universe is, why it exists at all, or how it actually works. The instructor, viewing the assignment as the first step of an initiation process in the art and craft of writing, did not seem to have high expectations regarding
Preface
xix
content and was pushing instead for good craftsmanship with as narrow a focus as possible. I, on the other hand, was willing to accept tips on how best to say what I had to say, but not on what I had to say. I may have had a few things to learn about good writing and proper composition, but the pedagogical process would surely be the same, I thought, regardless of the subject matter. Moreover, I was convinced that the topic of my choice was already focused, exactly as instructed, on a sensible project of manageable proportions. I was not, after all, contemplating a study of the universe in the broadest sense, requiring the aid of philosophy in combination with the entire spectrum of the physical sciences. I was interested merely in celestial objects, including groups and systems of such, in “space, the final frontier,” as it was called on my favorite show on television. I had prudently ruled out research into anything that was not already verified to be in the universe. I knew my limitations. I was not out to make discoveries. I just wanted to report the facts. Having developed a fascination with astronomy and space exploration ever since the first episode of Star Trek in its original run, having watched the live broadcast of the Apollo 11 mission promising to turn the moon into our own backyard, at least in my own mind, and having gained such vast experience while still in elementary school, I felt well prepared, as a freshman in college, to write about the universe. Yet my confidence and my enthusiasm far exceeded the instructor’s expectations, which were aligned instead with my skills and my wisdom, or rather, with my obvious lack of both. Gently urged to reconsider my focus, I was ready, by way of compromise, to settle for coverage of our own galaxy, the Milky Way. Asked immediately to try again, I made further concessions, offering to restrict my study to our solar system. As it turned out, that, too, was unacceptable. I eventually ended up with a project concerning only one feature of only one planet. I quickly lost interest and completed the assignment with apathy and resentment rather than with the enthusiasm I had originally cultivated. My indignation was anchored to my conviction that, compared to a study of the rings of Saturn, a study of the universe as a whole held far stronger potential for far greater relevance to far more people. Although a paper of ten double-spaced pages on the universe at large would have had to be cast at a level of generality avoiding any discussion of meaningful details, such as the rings of Saturn, a sacrifice of that magnitude seemed insignificant to me, especially in comparison with the insufferable prospect of studying the rings of Saturn while leaving out everything else in the universe. I was well aware that a report on the rings of Saturn would support our understanding of the planet, a better understanding of which would enhance our knowledge of the solar system, more extensive and reliable information
xx Preface on which would expand what we knew about our galaxy, and so on, with the implication that I would indeed be taking part thereby in an ongoing study of the universe. That, however, was not satisfactory. I wanted to be engaged in that process directly at the universal level. I felt both entitled and qualified to skip all the minor details better left to those who had not yet mastered the big picture. Several years later, Stephen Hawking proved the feasibility of my original plan, when he managed to produce a popular book on the universe without ever mentioning the rings of Saturn, the polar ice caps of Mars, that red spot on Jupiter, or any of the other countless details making up the whole. Notwithstanding Hawking’s predilection for devoting too much space to time and too little time to space—a prejudice coloring his comparably popular sequels as well—I still believe that a good report on the universe would appeal to a greater number of people than would a good one on the rings of Saturn. Yet I no longer believe that the quantitative difference would necessarily make it more valuable. This confession may seem oddly out of place in a positive commentary on utilitarianism, but the confession in question doubles as a recognition that value measured along utilitarian lines depends as much on the consequences as it does on the number of people affected by those consequences. Universal and specific studies need not be compared in terms of merit, but even when they are so compared, each one can be valuable in its own right. A common measure of value in that sense is the strength of the contribution made to the field. The proper comparison is not between generalist and specialist studies, as if they were mutual substitutes, but between each approach and its own alternatives. A scholarly work must first meet the minimally acceptable standards of adequacy relevant to any study conceived in a formal academic context and carried out with a combination of facts and arguments. Its unique contribution to the state of scholarship peculiar to its domain carries and confers additional value, independently of the extent to which it satisfies the general conditions that make any such study technically adequate. Academic studies are often said to be produced in isolation, which is an allusion to the final stage of production where the author is alone with a keyboard and monitor. Yet any academic project is actually an outgrowth of endless interaction with the repository of knowledge already available on the subject. It is contemplated and developed in reaction to existing work, just as it is subsequently evaluated in reference to such work. No matter how good a scholarly study may be in itself, it becomes all the more valuable if it fills an unmet need or offers greater satisfaction of one that has so far been met only partially.
Preface
xxi
The most memorable expression of this scholarly symbiosis is Alfred North Whitehead’s reflection on the whole of Western intellectual thought as a unitary process. Drawing on what he found to be its safest general characterization, Whitehead famously described the European philosophical tradition as a series of footnotes to Plato. This would have been a more felicitous generalization had Plato’s most celebrated pupil not been so quick to establish a competing system of comparable influence. Nevertheless, the emphasis on Plato is not so much on the value or validity of his specific ideas as it is on the significance of his general contribution to the germination, institution, and orientation of philosophical thought as a mode of inquiry. As for sheer brilliance and accomplishment, a historical assessment of that sort might have worked just as well if it were cast in terms of a sequence of meditations anticipating Descartes, or a succession of prolegomena preceding Kant, instead of, or in addition to, a series of footnotes following Plato. With some sacrifice in modesty, even a protracted preface to Whitehead himself would have been a fairly meaningful metaphor. Regardless of his true intentions, however, Whitehead’s original statement represents the strongest praise imaginable for anyone in any field of study. And the proposition rings true even after the exaggeration is filtered out to recognize other pivotal figures in the history of Western thought. The present volume, then, is my footnote to John Stuart Mill’s Utilitarianism. Or from a different perspective, still in reference to Whitehead’s metaphor, it is a demonstration of the significance of one of Mill’s footnotes to Plato. I was as eager to undertake this project as I was reluctant to write about the rings of Saturn. This may seem inconsistent, both logically and psychologically, if Mill’s proof is to ethics what the rings of Saturn are to the universe. Yet the decades going by between the composition of “The Rings of Saturn” and the publication of Mill’s Principle of Utility confirm that any such inconsistency is, in fact, a reflection of the author’s gradual development of an appreciation for specialized research and analysis. The analogy, of course, is not perfect, given that there is no particular measure of the relative importance of the apparent analogues. Perhaps the universal setting corresponding to Mill’s proof should be taken as the entire field of philosophy instead of just as moral philosophy. Or it might plausibly be expanded even further to include the full range of scientific inquiry. Or it might best be restricted to the history of philosophy, or possibly to the history of ethical thought. While it is indeed difficult to choose between size and significance in placing the objects of comparison in their respective settings, none of that really matters, because the story on which the analogy is based is true. It is an autobiographical anecdote rather than a literary tool employed to embellish or
xxii Preface emphasize a relevant similitude, which could have otherwise been expressed in a single sentence in the first place. However that may be, Mill’s proof is more important to ethical theory, at least in the sense of enriching it rather than merely taking part in it, than the rings of Saturn are to the universe. Fortunately, in more or less the same proportion, what I have to say about Mill’s proof is more interesting than what I had to say about the rings of Saturn. This might not have been a good reason to offer a book on the subject if what I had to say about either one were not very interesting to begin with, but I believe what I have to say about Mill’s proof enhances what has been said to date on the subject. This still might not have been a sufficient motivation or justification for writing a book if the subject itself were not philosophically important, but the timeline of bibliography on the proof confirms the importance of the subject. Secondary literature dealing with or related to the proof starts in Mill’s own lifetime and continues without interruption to the present day. The longevity of powerful reactions by prominent scholars, most of them opposing the proof as a logical and philosophical abomination, has been instrumental in convincing me of the need for critical intervention and comprehensive reconsideration. As for what is actually in the book, the introduction that follows the front matter offers a more traditional overview, providing a properly informative breakdown of the various goals and strategies as well as the particular contents, in contrast to the playfully allegorical expression of the same concerns serving the same end in the present preface.
Acknowledgments I am indebted first and foremost to Gary E. Varner for fruitful discussions on various aspects of Mill’s ethics, particularly on his proof of the principle of utility, which helped shape my initial thoughts on a comprehensive analysis and defense, eventually leading to the inception of this project. I am also grateful to Carl P. Wellman, Larry May, and Roger F. Gibson Jr. for their generosity with their time and effort in supporting the subsequent course of development. Wellman was instrumental in the evolution of the central thesis and main arguments, which steadily benefited from his felicitous interventions with a comprehensive command of the material and meticulous attention to detail. May was generously forthcoming with ideas drawing on his proven expertise in herd mentality, group morality, and collective responsibility, both stimulating and facilitating my investigation of Mill’s position on the ontological and moral status of groups. Gibson was indispensable in an advisory capacity on a broader scale and more permanent basis, not only in connection with this particular project but also in contribution to my overall progress as a professional philosopher. My earlier intellectual debt is to some of the best teachers anyone has ever had: Feridun Baydar, Michael H. Mitias, and Robert E. Bergmark. Baydar was the primary impetus behind my early education, also playing a central role in the establishment of my character, especially in my formative years. Fully transcending his official role and contractual responsibilities as a teacher of mathematics, he provided profound insight on a diverse range of subjects through years of Socratic discussion, all the while demonstrating a philosophy of life with a wit and wisdom that has always been my ambition to emulate. He was easily the greatest inspiration in my youth and his memory continues to be a guiding light in my life. Mitias and Bergmark, in turn, were responsible for my formal introduction to philosophy in college. They supervised my first encounter with the philosophical classics and equipped me with the tools of the trade for recognizing and analyzing philosophical problems. Both as my teacher and as my mentor, in either case as the personification of Aristotelian excellence, Mitias ignited and nurtured my passion for the study of philosophy, immediately influencing my major in college and ultimately motivating my transition to graduate school. As my friend—friendship being a relationship he values above everything else, so much so as to publish a compelling monograph on the subject—Mitias later traveled halfway around the world to stand with me as the best man at my wedding. He remains, to this day, a definitive benchmark for my endeavors as a philosopher and my journey as a human being.
xxiv Acknowledgments Friends in general represent the richest resource available in our philosophical adventures. Philosophy is, after all, neither theoretical nor experimental but dialectical. And dialectic best reveals its pedagogical potential when conducted with the natural spontaneity of an unscripted exchange of ideas with the people we deem most worthy of our regard. A scholarly contribution to ethical theory is particularly responsive to, if not outright dependent on, the personal relationships comprising its driving force. It is indeed ethical discourse, more than any other intellectual activity, that thrives on intimate interaction with one’s friends, who thus serve not only as dialectical partners but also as a collective moral compass. Our greatest debt thereby is typically to those with whom we disagree the most. Whether in agreement or in disagreement, my most valuable learning opportunities in moral deliberation, gradually refining my predilections in ethical theory, have come from discussions with Timothy L. Anderson, William Allen Andrews, Frederick Scott Bauer, Mary Frances (Weir) Billups, Wesley Haas Blacksher, James Arnette Bobo, Michael Howard Brunson, David Sylvester Butler Jr., William Jolley Carr iii, Collin Creswell Cope, Ned Mims French ii, Philip Walter Gaines, Jeffrey Ernest Good, Nancy Kincade (Williams) Green, Stuart Byron Green, John Douglas Hermann, William Thomas Hetrick, Albert Anne Labasse, Cecile Elizabeth (Williams) Leggett, John Clifford Leggett, Paul Owen Martin, Stephen Kelly Martin, William Whitfield McKinley Jr., Tara Lyn McPherson, Jon Garraway Nance, John Thomas Ray iii, Frederick Joseph Rein Jr., Charles Allen Scarboro, David Marcus Wilkerson, and Benjamin Ray Wynne. Learning opportunities become all the more effective when they are close at hand and available on demand. With existential experience constituting an essential ingredient in personal growth, our principal prospects for any kind of enlightenment are, as it happens, conveniently located at home, which is where most of us spend most of our time. It has accordingly been my great fortune to share my life with a veritable fountain of inspiration, my wife, Banu Beste Başol Alican. Regular exposure to the mystical blend of prophetic intuition and sagely wisdom animating her visionary spirit has been the single most important factor in my ongoing development both as a person and as a scholar. Banu also shares creative responsibility for the conception and contemplation of the present volume—a revised and enlarged edition of the original—and consequently for the transformation of the initial effort into a better concept and product. It is through her electrifying originality as a critic and her infectious optimism as an advocate that everything worthwhile in my life inevitably finds its true potential.
newgenprepdf
A Note on References References to John Stuart Mill’s works are to the standard critical edition: Collected Works of John Stuart Mill. Thirty-three volumes. General editors: Francis Ethelbert Louis Priestley and John Mercel Robson. Toronto (University of Toronto Press) and London (Routledge and Kegan Paul), 1963–1991. Page references identify the Collected Works edition as “CW,” followed by volume and page numbers in Arabic numerals separated by a colon. References to Utilitarianism add greater detail to accommodate readers following any of the popular editions. The CW pagination in that case is preceded by a reference to Utilitarianism, denoted as “U,” followed by chapter and paragraph numbers, again in Arabic numerals separated by a colon. References to A System of Logic are also documented in greater detail in reflection of the organizational divisions Mill himself provides in that work: “L” signifies the work, followed by book, chapter, and section numbers, all in Arabic numerals with a colon separator followed by a period separator. Then comes the CW pagination. Mill’s works that are most relevant to the present volume appear as follows in the Collected Works edition: Auguste Comte and Positivism (CW10:261–368); Autobiography (CW1:1– 290); “Coleridge” (CW10:117– 163); Letters (CW12–17 and CW32); On Liberty (CW18:213–310); “Remarks on Bentham’s Philosophy” (CW10:3–18); “Sedgwick’s Discourse” (CW10:31–74); A System of Logic (CW7–8); Three Essays on Religion (CW10:369–489); “Use and Abuse of Political Terms” (CW18:1–13); Utilitarianism (CW10:203–259); “Whewell on Moral Philosophy” (CW10:165–201); “The Works of Jeremy Bentham” (CW10:75–115). References to sources other than Mill’s Collected Works follow the author- date system of citation. Footnotes are reserved for information or argumentation that would otherwise disrupt the natural progression of the main text. Unless specified otherwise, quotations omit any and all footnote reference markers (whether asterisks or numerals or letters) present in the original.
Introduction John Stuart Mill was a prolific writer. He loved to write, he wrote a lot, and he covered practically everything. While he often got carried away in a stream of consciousness as he took pen to paper, his meaning remained accessible and his expression eloquent. A clear testament to his love of writing is an unsolicited twenty-page commentary he once saw fit to contribute as a footnote to someone else’s work he was editing for publication. The work was his father’s, posthumously undergoing a second edition at the time, with the footnote in question representing the longest of several dozen such contributions by the son, many of them longer than this paragraph.1 The Collected Works edition of his own output spans thirty-three volumes, each of them roughly the size of a brick, together taking up twice as much shelf space as the full edition of the Oxford English Dictionary. The corpus of his published writings penetrates so many fields with such proficiency that expertise on Mill requires a team of specialists in different areas, difficult to cover by a single scholar working alone. He wrote with great insight and considerable authority on anything that was of interest to him, coming to be known by different people for different achievements. His vocation and orientation as a philosopher made him a jack of all trades, like any other philosopher of the period, but his specialization at the very highest levels of several distinct areas of expertise made him a master of those trades as well, unlike most philosophers in any period. He was indeed not merely a philosopher but truly also a logician, historian, economist, moralist, sociologist, and political theorist. This makes any reasonable claim to being a Mill scholar only a partial claim in connection with just one of his myriad interests and manifold demonstrations of brilliance. As a full claim, it would be a rare mark of distinction for the student, much like how Aristotle might qualify as a Plato scholar. Most of Mill’s works made significant contributions to their respective fields, at least at the time when they first came out, but typically also in the long run, some perhaps on a permanent basis. Those that have decisively stood the test of time, going by the continuity in general readership, include 1 Mill organized his essay-length editorial commentary in two sections and appended it as the fifty-eighth footnote to the second volume of the second edition of his father’s Analysis of the Phenomena of the Human Mind (London: Longmans, Green, Reader, and Dyer, 1869, vol. 2, 307–326, n. 58). This edition was published forty years after the original (London: Baldwin and Cradock, 1829) and a third of a century after the death of James Mill (1773–1836).
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_002
2 Introduction his Autobiography, On Liberty, and Utilitarianism. His Autobiography, like that of anyone else important enough to have one published, appeals to a wide audience, but it holds the greatest interest for educators and psychologists. It is best remembered for its striking account of Mill’s education, providing evidence of the advantages of home schooling, while illustrating the downside of a successful outcome. It thus documents the depth and breadth of his intelligence as a child, and the limits of his father’s understanding of child psychology, despite a reputation for expertise in both psychology and the philosophy of mind. On Liberty is probably Mill’s most widely respected work. It has become a pillar of the liberal arts curriculum, securing a firm place in the syllabi of undergraduate and graduate classes in philosophy, sociology, and political science, occasionally even finding its way into the personal reading lists of law students in pursuit of theoretical grounding for their core studies and professional development. Utilitarianism, the shortest of the three works and the most controversial of Mill’s writings, represents his attempt to popularize the school of thought on which he was reared and toward the promotion of which he provided the greatest impetus of anyone outside Jeremy Bentham and his own father. It still stands today as the most frequently consulted text in and on classical utilitarianism. Yet the proof offered there for the principle of utility remains the most frequently faulted contribution to utilitarian theory after Bentham’s felicific calculus. What emerges as curious, even upon a cursory glance, as exemplified by the preceding ruminations, is that, among the thirty-three volumes of his Collected Works, Mill devoted only fifty-seven pages to a standalone discussion of the doctrine representing his most passionate conviction, and only twelve paragraphs, constituting the second shortest chapter in the corresponding book, to a proof of its first principle. He wrote more in just footnotes to his father’s book. While the number of words in an essay is certainly no indication of either its quality or its impact, the length of Utilitarianism seriously underrepresents Mill’s interest in the subject. The most plausible explanation for the brevity of Utilitarianism, an explanation for which there is explicit textual evidence, both in that work and elsewhere, is that Mill considered the underlying theory patently sound, easily comprehensible, and inherently persuasive, at least for those who were not brainwashed by gratuitous opposition and misguided criticism. He found the platform in need of little help beyond effective publicity through exposition, clarification, and propagation in print. Ironically, the stock of secondary literature on Mill’s Utilitarianism in any library that carries his Collected Works outweighs the entirety of that collection if not also the combined commentary on the rest of its contents. This is mostly because not everyone has found the theory as compelling as Mill obviously
Introduction
3
did. However, it is also at least partly because a good many people have found his approach even less convincing than the theory itself. As a matter of fact, some of the harshest critics of Mill are among the strongest supporters of utilitarianism. Regardless of what they may think of the theory, those who object to Mill’s articulation and promotion of it, if they have nothing else in common, are generally united in denouncing his proof of the principle of utility. The common ground bringing them together is logic. The proof has a reputation for being not merely ineffective but logically defective as well, so much so as to remain beyond repair or redemption. A logical flaw in a central argument leaves hardly any room for debate on the merits of a philosophical proof, just as a miscalculation in a key equation leaves none in a mathematical proof. A logical error, especially one in the form of a fallacy, whether formal or informal, is quite possibly the most embarrassing defect to be found in a philosophical presentation, something on the order of an erroneous sum in a mathematical demonstration. One such error is humiliating enough, but several in a row can ruin the perpetrator’s reputation, particularly if they are all concentrated in a short stretch of text in the same proof. Indeed, the potential for embarrassment in Mill’s case is compounded by allegations of multiple transgressions of the rules of logic, to say nothing of the bounds of common sense, in connection with his proof of the principle of utility. Philosophers are habitually and painstakingly careful about how they express themselves, even when they are jotting down a family recipe for pecan pie, but especially when they are grooming the final draft of a philosophical essay. When a philosopher is portrayed as proceeding in a state of utter confusion in the course of defending his considered opinion on matters of serious personal and professional interest, the portrait is to be suspected of misrepresentation before the philosopher is of misology or incompetence. Mill himself said something to that effect in a similar context, though not in response to a critique of his own proof, which he probably never imagined could attract such ridicule: A man must now learn, by experience, what once came almost by nature to those who had any faculty of seeing; to look upon all things with a benevolent, but upon great men and their works with a reverential spirit; rather to seek in them for what he may learn from them, than for opportunities of shewing what they might have learned from him; to give such men the benefit of every possibility of their having spoken with a rational meaning; not easily or hastily to persuade himself that men like Plato, and Locke, and Rousseau, and Bentham, gave themselves a world
4 Introduction of trouble in running after something which they thought was a reality, but which he Mr. A. B. can clearly see to be an unsubstantial phantom; to exhaust every other hypothesis, before supposing himself wiser than they; and even then to examine, with good will and without prejudice, if their error do not contain some germ of truth; and if any conclusion, such as a philosopher can adopt, may even yet be built upon the foundation on which they, it may be, have reared nothing but an edifice of sand. mill CW18:72
It is in this spirit that the present volume approaches Mill’s proof of the principle of utility. The approach is analytic and interpretive as well as defensive. The mode of treatment varies with the methodological objective. Coverage is analytic throughout the book but particularly so in isolating and studying the steps of the proof piecemeal as extracts from the narrow context of the fourth chapter of Utilitarianism. The interpretive aspects, in turn, come out most vividly in a reconstruction and examination of the proof within the broader context of Mill’s general philosophical convictions, including not only his moral philosophy but also his logic, epistemology, and philosophy of mind. The defensive elements, for their own part, are anchored to and clustered around specific responses on behalf of Mill to historically popular and currently unresolved charges against the proof. The aim is not to glorify utilitarianism as a normative ethical theory. The book is concerned only with Mill’s utilitarianism, that is, with his universalistic ethical hedonism, and specifically with his proof of the principle of utility. The overarching purpose guiding the overall effort is to show that Mill proceeds intelligibly and systematically in pursuing a well-defined project in the fourth chapter of Utilitarianism and that he succeeds in what he describes, promises, and carries out there as a proof of the principle of utility. Any such attempt would be in vain without a definitive response to specific objections to the proof.3 Despite being polemical and ill-conceived in origin, 2 This passage (CW18:7) is from “Use and Abuse of Political Terms” (1832), Mill’s review of George Cornewall Lewis’s Remarks on the Use and Abuse of Some Political Terms (London: Fellowes, 1832). The review was originally published in Tait’s Edinburgh Magazine 1 (May 1832): 164–172. The complete text is available in CW18:1–13. 3 All objections to Mill’s proof can be traced back to the early reception of Utilitarianism. The constituents of that reception relevant to this study consist of the following works in chronological order, which is the order of their treatment in chapter 3 of the present volume: John Grote’s An Examination of the Utilitarian Philosophy (1870); Henry Sidgwick’s The Methods of Ethics (1874–1907; see also his essay of 1879); Francis Herbert Bradley’s Ethical Studies (1876/ 1927); William Ritchie Sorley’s The Ethics of Naturalism (1885/1904); John Dewey’s Outlines of a Critical Theory of Ethics (1891); John Dewey and James Hayden Tufts’s Ethics (1908/1932);
Introduction
5
traditional objections have been set forth competently by eminent philosophers, each of them contributing either deliberately or unwittingly to the gradual development of a parody of the proof amusing the philosophical community for generations. Unfortunately for Mill, and in all fairness to his critics, some of the objections are prima facie legitimate challenges to crucial points in Mill’s apparent line of reasoning. Most are variations on the theme that the proof breaks down under its own weight, its collapse being due almost entirely to shoddy workmanship manifested as logical errors and conceptual confusions, leaving little factual or philosophical demolition work for critics. The main objections taken up here, devoting a separate chapter to each, are the charges that Mill commits the fallacy of equivocation, the fallacy of composition, and the naturalistic fallacy in the course of his proof. These are not the only mistakes a utilitarian doctrine or demonstration might make, and therefore must avoid, as if classical utilitarianism would have been embraced universally had it not been for the inability of its chief advocate to put together a few words without committing a major fallacy. Critical debates continue to probe and challenge prevailing reconstructions and contemporary reconceptions of utilitarianism, regularly revamped to avoid commonly recognized weaknesses in its original formulation. Traditional objections to Mill’s Utilitarianism thus extend well beyond his notorious proof. Among the objections to which Mill does not reply, be it in Utilitarianism or elsewhere, at least two stand out as philosophically interesting even today. One of those, anchored to his exposition of utilitarianism in the second chapter of Utilitarianism, concerns his distinction between higher and lower pleasures in response to charges that utilitarianism is “pig philosophy.”4 The other one involves his efforts in the fifth chapter, the longest in the book, to incorporate interdependent conceptions of justice and moral rights into the principle of utility. Both of these areas of concern, along with various other matters of import, remain outside the scope of this book. The focus here is specifically on the proof. It is naturally also on the principle itself, including its historical origins
John Stuart Mackenzie’s A Manual of Ethics (1893–1929); George Edward Moore’s Principia Ethica (1903). 4 Mill’s reaction to the reputation of utilitarianism as “pig philosophy” is at least partly inspired by Thomas Carlyle’s (1850) critique of Jesuitism in an essay called “Jesuitism,” which comes with strongly worded opposition in a section titled “Pig Philosophy” (Carlyle 1850, 268–270). While Carlyle’s piece is not positioned against utilitarianism, instead targeting Jesuitism just as the title indicates, the rhetoric there is nevertheless consistent with the general alignment at the time against utilitarianism.
6 Introduction and its philosophical implications, but only insofar as such considerations are necessary for an accurate understanding and proper evaluation of the proof. The justification for such specialization is that nothing Mill ever wrote has been both misunderstood and criticized as widely and as persistently as his proof of the principle of utility. Commentators tend to disagree not only on the success of the proof but also on its aims, steps, and structure. Many of them come together on what conclusion Mill intends to prove, but they part ways in identifying and interpreting the main arguments and thereby in determining the overall structure. Some of them even differ on what conclusion he intends to prove, because they differ more fundamentally on what his principle of utility is. Some regard the principle as a doctrine of intrinsic value, others as a standard of moral obligation, yet others as a decision procedure for ethical justification or a guide for moral deliberation. The greatest incentive for intervention on behalf of Mill, however, is that his critics accuse him not merely of producing an unsuccessful proof but of delivering one infested with elementary logical blunders and serious conceptual confusions. From the opposite perspective, Mill has not been without defenders either. Although philosophically significant support for the proof was sparse for a full century following the publication of Utilitarianism, advocates of Mill have been defending him regularly and competently since the middle of the twentieth century. Yet even his defenders disagree in their interpretations of the proof. And they often focus on one or another but not on all of the fallacies, defects, and deficiencies supposedly permeating the inherent reasoning. Moreover, their contributions have been in the form of journal articles and anthology chapters, with no book-length treatment focusing exclusively on the proof. That would not matter, of course, if the contributions on record were sufficient in combination, but they are not. Despite a good stock of secondary literature growing in a scattered point- and-counterpoint fashion, existing journal articles and anthology chapters on the subject are neither severally nor jointly sufficient to provide the kind of defense that would do justice to Mill’s proof of the principle of utility.5 That is a 5 Even when skipping the authoritative portion of the early reception (covered in chapter 3), and even after excluding commentaries that are primarily or exclusively on the naturalistic fallacy (covered in chapter 6), the remaining body of secondary literature on Mill’s proof is still enormous: Ronald Field Atkinson (1957); John. M. Baker (1980); David Owen Brink (2013, 113– 134); Donald George Brown (1973); George A. Clark (1959); Stephen Cohen (1990); Neil Cooper (1969); John Patrick Day (1964); Guy Fletcher (2013); Robert K. Fullinwider (1972); Andrew Thomas Fyfe (2011); Nicholas Griffin (1972); Pepita Haezrahi (1949); Everett W. Hall (1949); S. J. Heans (1992); Robert W. Hoag (1987; 1992); Terence Henry Irwin (2009, 415–418); Hardy Jones (1978; 1979); John Kleinig (1970); Norman Kretzmann (1958); Berel Lang and Gary Stahl
Introduction
7
gap that the present initiative fills with a comprehensive defense. Its principal contribution is an interpretation of Mill’s proof in its entirety, absolved of the three commonly alleged fallacies, among other accusations, yet faithful to the text at all points. The basic motivation for contemplating the possibility and practicality of an interpretation exonerating Mill, at least from a logical point of view, is nothing other than the principle of charity in critical interpretation. Giving the author the benefit of the doubt is not merely a matter of professional courtesy but one of intellectual integrity as well. A charitable reading is particularly appropriate for a critique of Mill, who was the author of the most authoritative text on logic in the nineteenth century, prior to the mathematization of the field through the advent of symbolic logic, as developed by the likes of George Boole, Gottlob Frege, Giuseppe Peano, Bertrand Russell, and Alfred North Whitehead. Granted, logical prowess and philosophical reputation may not always be enough to keep an author from advancing fallacious arguments, but Mill’s crimes against reason are all supposed to be located in a single paragraph of standard size, which indicates a level of incompetence that would be difficult for anyone to achieve, making it a highly unlikely prospect for a master logician. At the other extreme, a sympathetic predisposition can be just as misguided as a predatory attack. Any commentary following a beeline to the most charitable interpretation of each controversial point is more suitable for propaganda than for critical analysis. Blind devotion to empathy or sympathy with respect to the parts of the proof under critical scrutiny runs the risk of explicating those parts with mutually incompatible interpretations, thus clearing the obstacles attached to the separate steps without establishing their convergence into a coherent whole. Zealous trial and error might eventually strike a
(1969); W. H. Long (1967); Christopher Macleod (2013a; 2013b; 2014); Maurice Mandelbaum (1968); Joseph Margolis (1967); John Marshall (1973); Tim Mawson (2002); Francis Stewart McNeilly (1958); Dale E. Miller (2004; 2010b, 31–53; 2014a, 369–373; 2019); Elijah Millgram (2000); Dorothy Mitchell (1970); Shia Moser (1963); George Nakhnikian (1951); Francesco Orsi (2021); Ingmar Persson (2000); Richard H. Popkin (1950); Anthony Meredith Quinton (1973, 58–71); David Daiches Raphael (1955; 1994); Dennis Rohatyn (1971; 1975); Bertrand Russell (1945, 773–782, especially 778–780); Alan James Ryan (1970, 187–212; 1974, 116–119); Geoffrey Sayre-McCord (2001); Jerome Borges Schneewind (1976); James Seth (1896; 1908); Ching-Lai (Qinglai) Sheng (1991, 215–223); John Skorupski (1989, 283–336, especially 285–288); David Soles (1998); Gordon W. Spence (1968); Robert Scott Stewart (1993); Michael Stocker (1969); Avrum Stroll (1965); James Sutherland (1886); Grenville Wall (1982); Carl P. Wellman (1959); Spencer K. Wertz (1971); Henry Robison West (1972; 1982; 2004, 118–145; 2006; 2017); Fred Wilson (1982; 1983); Dan Yim (2008); Peter Zinkernagel (1952).
8 Introduction balance between a defense of parts of the proof and a representation of the whole, but a patchwork rendition of that sort would be fundamentally misleading even if it were to simulate such a balance as an afterthought. Any argument can be reformulated to remove problems, especially those that are merely apparent, but also any that are real. All it takes is the right assumptions, clarifications, and qualifications. The question is whether the proposed assumptions, clarifications, and qualifications are justifiable. The authenticity of any such interpretation demands that the corresponding principles of reconstruction reflect Mill’s own views, perhaps by contextual implication, but preferably through textual confirmation. Ideally, the reformulation will continue to represent his original argument, essentially as it is, but with greater clarity. Nearly as acceptable, it might produce an argument that Mill himself did not advance but clearly would have endorsed. Otherwise, it becomes an altogether different argument, one which he should have employed instead of the one that he actually did. Taking the last route would be tantamount to offering an alternative proof of the principle of utility, with no credit to Mill, except in having espoused a tenable doctrine better defended by someone else. The intention here is not to develop a line of reasoning that Mill should have followed toward the kind of proof that would have succeeded where his own failed. On the contrary, the motivating conviction is that Mill was already doing a good job on his own, arguably the best that could have been done in the relevant context. The intention, rather, is to clarify what that was and to demonstrate why it was at least acceptable if not also admirable. The pursuit thus proceeds with Mill’s own arguments to show that his thought process does not suffer from the problems commonly associated with it, and furthermore, that its suspect parts and aspects work well together as a coherent and successful proof. Progress toward that end supports a natural division of the project into three stages making up the three parts of the book: (1) an exegetical and historical background introducing Mill’s principle of utility and documenting the scholarly reception of his proof; (2) an analysis and response to the traditional charges against various parts and aspects of the proof; (3) a reconstruction of the proof as a unitary whole, together with implications for further consideration. The three parts are organized in eight chapters. The first chapter reviews salient formulations of the principle of utility prior to Mill, focusing particularly on Bentham, both as the most relevant source of inspiration for Mill and as the most passionate exponent of classical utilitarianism. The second chapter, setting up the dialectical groundwork, introduces Mill’s principle of
Introduction
9
utility, accompanied by a preliminary outline of his proof. The third chapter, building on the groundwork already established, reports the strongest objections to the proof in its academic reception. The next three chapters, pursuing a purely defensive agenda, examine the proof in stages, fallacy by fallacy, focusing separately and respectively on the fallacy of equivocation, the fallacy of composition, and the naturalistic fallacy. The seventh chapter, as the positive counterpart of the three defensive chapters preceding it, offers a holistic interpretation of the proof in its entirety, drawing on and putting together the separate components of the defensive strategy unfolding up to that point, and thereby exposing qualities in Mill’s reasoning beyond the mere absence of glaring problems. The eighth and final chapter concludes with a discussion of the broader role of the proof in Mill’s ethical system as well as its implications for the viability of utilitarianism in general. Given that this book is essentially a commentary on someone else’s book, namely Mill’s, its value is largely parasitic on the value of the original. While a commentary can easily be vacuous even where the original is brilliant, it can hardly be useful unless the original is indeed valuable. A question lurking behind the derivative value in this case places the original under suspicion while suggesting a contingent existence for the commentary: If Mill’s proof is so good, why does it require a whole book in its defense? The question is tempting but misplaced. It is no different from any other inquiry into any other difficulty. The following problem, for example, is quick to come to mind through a comparable question: If Kant’s transcendental deduction of the categories is so good, why does it have to be sugar-coated by a handful of specialists who then spoon-feed it to everyone else, including the rest of the philosophical community? Rarely would an ordinary philosopher attempt to read the Critique of Pure Reason without the aid of one of the many user manuals issued by Kant scholars. Yet the value of Kant’s work is not diminished by the need for numerous commentaries in explanation of it. Just as Kant’s deduction might be valid and worthwhile even though it has to be explained and justified by experts, so too could Mill’s proof be acceptable and impressive despite requiring a book in its defense. The critical question might best be turned around, then, to correct the false implication of an inverse correlation between the value of the original and the need for a commentary. Only then could one properly respond that Mill’s proof is so good that it merits a whole book in its defense. Bad theories or arguments do not, especially in the long run, attract as much praise, criticism, or even indifferent exposition as the good ones. Nevertheless, the question is serious enough to deserve a response that is not merely a critique of the question itself. The answer is that the proof
10 Introduction originally fell into the wrong hands, struggling thereafter to recover from gross and chronic misinterpretation. The parts laying out the logical foundation were phrased too economically to satisfy analytic inquisition, while the parts representing the rhetorical flourishes dominated the presentation. This is mainly because the corresponding arguments, as well as the entire essay in which they were advanced, were composed with a broad audience in mind. In the final analysis, the motivating force behind defending Mill in this respect is not so much that the proof is absolutely flawless as it is that careless interpretation and willful misrepresentation, which have together shaped the main lines of its scholarly reception to date, can point to flaws that are not actually there. This is not to say that the proof is or is not valuable in itself, but that its value, not to mention its meaning, remains obscured in a cloud of negativity. The ultimate aim of this book is to free it from that cloud.
pa rt 1 Mill’s Principle of Utility and Scholarly Reactions to His Proof: Exegetical and Historical Background
∵
c hapter 1
Classical Utilitarianism before John Stuart Mill The Legacy of Jeremy Bentham
This chapter is a historical prologue to subsequent chapters. Its inherent aim is to reconstruct the philosophical background relevant to a proper interpretation of John Stuart Mill’s proof of the principle of utility within the parameters of the ideology and vocabulary he inherited from predecessors. The aim determines the approach. The main methodological vehicle is an examination of the emergence and development of classical utilitarianism with particular attention to the conceptual and terminological dynamics in place immediately prior to Mill. As anything immediately prior to Mill is invariably associated with Bentham, whether in the domain of moral discourse or within the scope of social and political theory, the analysis here revolves around the formative forces contributing to Bentham’s position on utilitarianism. The chapter is organized around six sections. The first section takes stock of the sources and references required for combing through the history of utilitarianism coincident with the specific aims of the chapter. The second section documents Bentham’s favorite terms and expressions in their original context as a point of reference for historical orientation and further investigation. The third section traces the origins and development of utilitarian nomenclature, including ancillary and peripheral contributions as well as central trends and critical iterations. The fourth section appraises Bentham’s intellectual debt to predecessors and contemporaries in terms of ideology and terminology. The fifth section explores the network of indirect inspiration and transmission complementing the forces directly responsible for shaping Bentham’s perspective. The sixth section examines the structure and implications of prominent patterns of development in Bentham’s own terminological predilections. The central focus of the chapter is on utilitarian nomenclature, which is a common thread running through the individual sections. While this emphasis is a natural consequence of the supreme value Bentham habitually placed on proper terminological choices, it also serves as an inventory of the terminological heritage available to Mill, and therefore as a dialectical prelude to the remainder of the present volume. We would hardly be in a position to follow what Mill was trying to prove, let alone being able to determine whether he was successful, without first understanding the conception, intention, and
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_003
14
Chapter 1
evolution of the relevant vocabulary, which is best appreciated in the light of historical records in primary sources. 1.1
History of Utilitarianism: Sources and References
Sources on the history of utilitarianism are practically inexhaustible, especially if we include surveys of the broader history of moral, social, and political philosophy, most of which offer some degree of acquaintance with the emergence and development of utilitarianism.1 Depending on the objective, however, studies concerned specifically with the history of utilitarianism, either in its entirety or with respect to any of its distinct periods, might prove more useful in this context. Examples include, but are not limited to, the following works listed in chronological order: – Leslie Stephen: The English Utilitarians (1900); – Élie Halévy: La Formation du Radicalisme Philosophique (1901–1904); – Ernest Albee: A History of English Utilitarianism (1902); – John Petrov Plamenatz: The English Utilitarians (1949); – Anthony Meredith Quinton: Utilitarian Ethics (1973); – Jerome Borges Schneewind: Sidgwick’s Ethics and Victorian Moral Philosophy (1977); – Frederick Rosen: Classical Utilitarianism from Hume to Mill (2003); – Ben Eggleston and Dale E. Miller (eds.): The Cambridge Companion to Utilitarianism (2014); – Bart Schultz: The Happiness Philosophers: The Lives and Works of the Great Utilitarians (2017); – Samuel Hollander: A History of Utilitarian Ethics: Studies in Private Motivation and Distributive Justice, 1700–1875 (2019). The list is restricted, for the sake of brevity, to monographs and other book- length treatments. Most of the entries are fairly comprehensive in their historical range. A couple of them may appear to stand out from the rest, but they are 1 A prime example of a broader survey of the history of ethics with commensurate coverage of utilitarianism is Terence Henry Irwin’s The Development of Ethics: A Historical and Critical Study. Vol. 1: From Socrates to the Reformation (2007). Vol. 2: From Suarez to Rousseau (2008). Vol. 3: From Kant to Rawls (2009). Oxford: Oxford University Press. As the chronology of the individual titles may readily indicate, coverage of utilitarianism is divided between the last two volumes, with the last one naturally coinciding with more of its history. John Stuart Mill is in c hapters 79 (pp. 364–397) and 80 (pp. 398–425) of volume 3. Bentham does not have a chapter of his own, though he does receive adequate attention, with his name coming up wherever one might expect, especially in connection with Mill.
Classical Utilitarianism before John Stuart Mill
15
no less relevant or useful than any of the other items on the list. For example, the contribution by Schneewind (1977) is primarily on Sidgwick, but it covers enough of the background to constitute a general history in its own right. Likewise, the entry by Eggleston and Miller (2014) is not so much a historical study as it is a general companion, but it is no less ambitious in its historical coverage, which makes up nearly a third of the book.2 As for sources on the history of utilitarianism with a primary or exclusive focus on Bentham, that is, from the perspective of Bentham as the central element, the most significant works, still not exhaustive of the relevant literature, consist of the following: – Jeremy Bentham: “Article on Utilitarianism” (1829/1983); – Thomas Perronet Thompson: “ ‘Greatest Happiness’ Principle” (1829); – John Bowring: “History of the Greatest-Happiness Principle” (1834); – Henry Sidgwick: “Bentham and Benthamism in Politics and Ethics” (1877); – Robert Shackleton: “The Greatest Happiness of the Greatest Number: The History of Bentham’s Phrase” (1972). Full bibliographical details for both lists are available in the “Works Cited” section at the end. A look at the evidentiary context is a greater priority at this point. Acquaintance with the most compelling sources, including their mutual publication history, in addition to their individual contents, will enable us to draw more reliable conclusions with greater confidence. In terms of documentary value, it is the second list, the one focusing specifically on Bentham, that best supports the overarching aim of the present chapter to explore the strongest influences on John Stuart Mill. The first three entries on that list are particularly important in that regard because they are, in essence, separate reports of a single account, with the first one constituting the groundwork and providing the narrative for the next two. Understanding the nature and structure of their relationship will help appreciate the strength of the collective insight they provide into Bentham’s place in the history of 2 The Cambridge Companion to Utilitarianism, edited by Ben Eggleston and Dale E. Miller (2014), offers substantial insight into the historical background in addition to explicating contemporary issues in theory and application. Each of the first five chapters is drafted by a prominent scholar covering a different period in the history of utilitarianism. The combined result is a concise and comprehensive coverage of the emergence and development of utilitarianism as a school of thought. The historical chapters are as follows in the order in which they appear in the book: (1) Colin Heydt: “Utilitarianism before Bentham” (pp. 16– 37). (2) James E. Crimmins: “Bentham and Utilitarianism in the Early Nineteenth Century” (pp. 38–60). (3) Henry Robison West: “Mill and Utilitarianism in the Mid-Nineteenth Century” (pp. 61–80). (4) Roger Crisp: “Sidgwick and Utilitarianism in the Late Nineteenth Century” (pp. 81–102). (5) Krister Bykvist: “Utilitarianism in the Twentieth Century” (pp. 103–124).
16
Chapter 1
utilitarianism, as well as his own position on the history of utilitarianism, consequently illuminating John Stuart Mill’s role in the fate of Bentham’s intellectual legacy. The textual relationship in play is that of a common source, the manuscript of Bentham (1829), never published in his lifetime, informing both of the other two works. A brief look at the details will help bring out the connections. Bentham’s “Article on Utilitarianism” (1829) is an exegetical defense and historical survey of utilitarianism, originally drafted in two separate versions, a longer one and a shorter one, together serving the same purpose. Both versions of the article (1829) were intended specifically to facilitate the anonymous response of The Westminster Review, or more precisely, of its coeditor and joint owner, Thomas Perronet Thompson (1829), to an anonymous critique of James Mill’s Essays on Government, Jurisprudence, Liberty of the Press, and Law of Nations (1823/1828), by Thomas Babington Macaulay (1829) in The Edinburgh Review.3 Yet only parts of Bentham’s (1829) manuscript actually made it into print, mainly in the form of anonymous excerpts within the body of the editorial prepared by Perronet Thompson (1829) as a response to Macaulay’s critique of James Mill. Substantial chunks of Bentham’s text also became available afterwards through a combination of direct quotation and paraphrased narration in Bentham’s memoirs and other biographical material, all compiled and edited by his friend, editor, and literary executor, John Bowring. Bowring’s editorial efforts included a survey of the history of the greatest happiness principle, appearing at the very end of the first volume of Bentham’s Deontology (1834), which was itself a patchwork of notes supported by testimony, published posthumously under the supervision of Bowring, who was also responsible for the composition of the actual manuscript, at least in the sense of creating a unitary text and publishable draft from bits and pieces of original material. It is in this survey by Bowring that the contents of Bentham’s article made their
3 Thomas Babington Macaulay’s (1829) critique of James Mill was published anonymously as “Mill’s Essay on Government /Utilitarian Logic and Politics” (running headers), Article 7 of The Edinburgh Review 49:97 (March 1829): 159–189. Thomas Perronet Thompson’s (1829) response was published anonymously as “ ‘Greatest Happiness’ Principle” (running headers), Article 16 of The Westminster Review 11:21 (July 1829): 254–268. At the time, Perronet Thompson had just recently become coeditor and joint owner of The Westminster Review, serving in an editorial capacity between January 1829 and January 1836 (together with John Bowring, who had editorial responsibilities between January 1824 and January 1836), while holding a proprietary stake between January 1829 and July 1832 (together with Jeremy Bentham, who claimed ownership between January 1824 and April 1832).
Classical Utilitarianism before John Stuart Mill
17
fullest appearance in print during the relevant period, though coming a couple of years after Bentham’s death. The original script of both versions of Bentham’s article (1829), as well as the text of his Deontology (1834), was later recovered in full from the corresponding manuscripts and reprinted as part of Deontology, Together with A Table of the Springs of Action and the Article on Utilitarianism (1983), a compilation edited by Amnon Goldworth and published as a volume in The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Perronet Thompson and Bowring, each one independently of the other, had evidently assisted Bentham in the production of the original manuscript of his “Article on Utilitarianism” (1829), subsequently drawing liberally on the text of the final draft as they developed their own expositions. Perronet Thompson (1829) seems to have focused on certain portions of the long version, while Bowring (1834) clearly followed both versions, paraphrasing long stretches of text in addition to quoting lengthy passages.4 The significance of the original manuscript of Bentham (1829) as a direct source for the editorial piece by Perronet Thompson (1829) and for the historical survey by Bowring (1834) makes authorship an open question in the latter two works. Any uncertainty or hesitation in that regard is exacerbated by the fact that the editorial by Perronet Thompson was published anonymously and that the survey by Bowring was published under Bentham’s name as part of his Deontology. Given that neither Perronet Thompson nor Bowring was credited as the author of their respective contributions, and given furthermore that both publications were based on the text of Bentham in the first place, it would not be unreasonable to credit Bentham as the author of those two works as well as his own article. Yet it is also reasonable to give credit where it is due, even if some of it must be shared with a silent collaborator, namely Bentham. This is especially important in the case of Perronet Thompson, who was directly responsible for the anonymous editorial in response to Macaulay, but it is no less relevant in the case of Bowring, who served both as an editor of Bentham and as a biographer of Bentham, his recognition as an author being particularly appropriate in the latter role if not in the former.5 The circumstances 4 Details of derivative and parallel development are available in Bentham (1829/1983, xxxiii– xxxvi), Sidgwick (1877, 647, 650–651, 651–652), and Shackleton (1972, 1462–1463, 1481–1482). Amnon Goldworth, for one, reports in his editorial introduction to Bentham (1829/1983, xxxv) that pp. 258–260 of Perronet Thompson’s text (1829) were “loosely based” on sections 2–15 of Bentham’s text (1829/1983) and that pp. 267–268 of Perronet Thompson’s text (1829) were based on sections 17–21 and 54–58 of Bentham’s text (1829/1983). 5 Any objection that Bowring’s “History of the Greatest-Happiness Principle” (1834) is not his own work but Bentham’s, and that it should therefore be credited to Bentham instead of
18
Chapter 1
being what they are, citing separate authors for the separate works will, if nothing else, provide greater convenience in reference. A related matter in achieving a balance between precision and convenience in citation arises in connection with Bentham’s publications that were subsequently recovered either more fully or more accurately from the original manuscripts. Such is the state of most of his ethical works, including not only his Deontology (1834) and his “Article on Utilitarianism” (1829), both of which were edited from the original manuscripts and reprinted in a combined critical edition (1983), as mentioned above, but also A Table of the Springs of Action (1817), which was included in the same collection (1983). Quotations from these three works are best indexed to the critical edition, yet each one is better referenced in its historical context. Accordingly, distinguishing between the separate works in the compilation, while continuing to use the author-date system of citation, requires the employment of two dates, the first indicating the date of the original publication, different in each case, and the second, following a slash, indicating the date of the edition quoted (1983). Thus: Bentham 1817/1983 (A Table of the Springs of Action); Bentham 1829/1983 (“Article on Utilitarianism”); Bentham 1834/1983 (Deontology). Fuller documentation in this manner will prove more convenient, or at least more distinctive, especially for readers who are familiar with the original publication dates, compared to the standard convention of adding letters to the date in common (Bentham 1983a; 1983b; 1983c). A comparable case can be made both for A Fragment on Government (1776) and for An Introduction to the Principles of Morals and Legislation (1789), each of which is available in a standard critical edition, that of the former published in 1977 and that of the latter in 1970. Yet the original publications, both belonging to Bentham’s early period and both affording ample time for modification while undergoing relatively little alteration between initial publication and the second edition, which came out in 1823 in either case, do not reflect any difficulties or discrepancies on the order of those seen in the posthumously published Deontology (1834) or the privately circulated “Article on Utilitarianism” (1829).6 Given the philosophical significance of these two works, together with Bowring, would be redundant with the fact that the piece is already attributed to Bentham (1834), having been published as part of his Deontology (1834). I acknowledge the fact and concede the point, departing from the proper reference only for the sake of a convenient distinction between the original source in Bentham (1829/1983) and the arrangement and elaboration in Bowring (1834). 6 The textual maturity and stability of An Introduction to the Principles of Morals and Legislation is even stronger than its official publication dates might seem to indicate, given that the work was printed privately in 1780 and circulated exclusively for a decade prior to formal
Classical Utilitarianism before John Stuart Mill
19
the familiarity of their historical context, it is perfectly acceptable to quote from the original editions, thus preserving the proper chronology of references within the author-date system of citation. That being said, neither one of the two works is relevant to the history of utilitarianism, certainly not in an expository capacity, though they themselves have a prominent position in the history of utilitarianism, specifically as developmental milestones in the movement, as discussed later. While Bentham’s chief publications in ethical theory and moral philosophy are An Introduction to the Principles of Morals and Legislation (1789), A Table of the Springs of Action (1817/1983), and Deontology (1834/1983), his only work specifically on the historical and methodological development of utilitarianism is his “Article on Utilitarianism” (1829/1983). The latter therefore constitutes the main source of information for the present chapter, at least insofar as Bentham’s personal perspective on the history of utilitarianism is concerned, which can fruitfully be supplemented by scholarly analysis and philosophical insight by others, whether in agreement or in amendment. Bentham’s account of the history of utilitarianism includes an assessment of his own place in the ongoing process as well as an overview of previous and parallel efforts. Both the long version (pp. 289–318) and the short version (pp. 320–328) of his “Article on Utilitarianism” (1829/1983) start with the contributions of predecessors and contemporaries, gradually building up to what Bentham viewed as his completion, perfection, and application of a viable yet rudimentary foundation already in place. Despite considerable repetition, overall coverage is far richer for its delivery through two complementary albeit overlapping versions than it would have been with just one or the other. The long version (pp. 289–318) proceeds as follows in terms of subject matter: – Bentham’s predecessors and contemporaries (pp. 289–292); – Bentham’s own contributions (pp. 292–294, 310–311); – Bentham’s opposition to Locke (pp. 293, 298–299, 314–316); – essence and significance of nomenclature (pp. 296–300, 302–303, 309–310). The short version (pp. 320–328) follows a similar arrangement: – Bentham’s predecessors and contemporaries (pp. 321–326); – Bentham’s own contributions (pp. 326–328); – Bentham’s opposition to Paley (p. 328).
publication in 1789, with ample opportunities for correction in between, not to mention the revision opportunities available for a third of a century prior to the second edition, eventually published in 1823.
20
Chapter 1
The common denominator between the two versions is a structural observation concerning the dialectical development of utilitarianism. Bentham (1829/1983) regards and relates the history of utilitarianism as a fragmentary process unfolding with the successive implementation of momentous contributions as opposed to a linear path proceeding with smooth increments of accumulated wisdom. While this perspective is primarily about the development of utilitarianism as a school of thought, both in moral philosophy and in social and political theory, it naturally also holds for its practical instantiation as a revolutionary movement in support of democratic reform, especially in the hands of the philosophical radicals collectively facilitating the inception of classical liberalism. As for the conceptual and methodological background, Bentham recognizes the following steps and stages as the formative components of philosophical progress in the corresponding historical process:7 – the recognition of utility as a useful concept with practical and theoretical applications; – the appeal to utility in demarcating good from bad and right from wrong; – the association of utility with happiness; – the elucidation of happiness in terms of pleasure and pain; – the classification of pleasures and pains on a grand scale, specifically toward the establishment of a definitive and comprehensive source of appeal, through the institution and employment of utilitarian decision-making in moral, social, and political questions; – the promotion of utility, or happiness (pleasure net of pain), as the arbiter of good and bad and the standard of right and wrong, in both private matters and public affairs. Bentham is, on the whole, as generous in assigning credit for contributions to the development of utilitarianism as he is eager to take credit for that same process, though typically with respect to decidedly different aspects or portions of it. For example, he readily and explicitly acknowledges the input of both his predecessors and his contemporaries in association with the first three stages of development, but he clearly finds his own work foundational and indispensable relative to the last three stages, while remaining receptive, even there, to parallel if inadequate attempts by others.
7 See Bentham (1829/1983), Bowring (1834), and Perronet Thompson (1829) for Bentham’s identification of these steps and stages in the development of utilitarianism. Note how he associates each step or stage specifically with the publication of at least one seminal work, either by himself or by others, sometimes citing one of each.
Classical Utilitarianism before John Stuart Mill
21
The highlights of Bentham’s survey of utilitarian milestones in the works of his predecessors and contemporaries, not all of them discussed in a favorable light, can be enumerated as follows:8 – Horace: Satires; – Phaedrus: Fables; – Aristotle: Posterior Analytics (71a–100b); – Francis Bacon: Novum Organum (1620); – John Locke: An Essay Concerning Human Understanding (1690); – David Hume: Essays (1741–1742);9 – David Hartley: Observations on Man, His Frame, His Duty and His Expectations (1749); – Claude Adrien Helvétius: De l’Esprit (1758); – Cesare Beccaria: Dei Delitti e delle Pene (1764/1767); – Joseph Priestley: An Essay on the First Principles of Government (1768); – William Paley: The Principles of Moral and Political Philosophy (1785). Horace, Phaedrus, and Aristotle come up merely as forerunners, commendable at most for recognizing the relevance of utility as a conceptual tool for valuation, evaluation, and judgment, not necessarily in moral matters, but in the course of life in general. Locke10 and Paley11 attract greater attention, but only in the context of opposition from Bentham, with some acknowledgment and obvious appreciation of Locke’s contributions to philosophy outside the 8 9
10
11
The list is compiled in correlation with the views of Bentham as articulated by Bentham (1829/1983, 289–292, 321–326, 328), Bowring (1834, 290–300, 304–311, 316–318), and Perronet Thompson (1829, 258–261). While Bentham’s references to Hume are always simply to Hume’s Essays, using nothing but the short title, his persistent albeit hesitant provision of 1742 as the date of publication, typically qualified as being true to the best of his recollection (Bentham 1829/1983, 289, 322; Bowring 1834, 29; Perronet Thompson 1829, 258), indicates that the references are to Hume’s Essays, Moral and Political (1741–1742), published in two volumes. Amnon Goldworth (Bentham 1829/1983) warns, however, that the content referenced by Bentham shows that he was actually working either with the first edition of An Enquiry Concerning the Principles of Morals (1751) or with Essays and Treatises on Several Subjects (1753), where An Enquiry Concerning the Principles of Morals constitutes the third volume of the compilation (Bentham 1829/1983, 290, n. 1, 322, n. 6, 323, n. 1, 350, n. 4). Locke (1690) receives ample attention from Bentham, most of it in forceful opposition, though with fair acknowledgment of his contributions outside ethics and social and political philosophy: Bentham (1829/1983, 293, 298–299, 314–316, 322); Bowring (1834, 291, 304–310); Perronet Thompson (1829, 259). Paley (1785) is invoked in the short version, but not in the long version, of Bentham’s “Article on Utilitarianism” (1829/1983, 328). He is also mentioned in Bowring’s “History of the Greatest-Happiness Principle” (1834, 310–311) but not in Perronet Thompson’s “ ‘Greatest Happiness’ Principle” (1829).
22
Chapter 1
domain of moral, social, and political philosophy (but none at all for those within that domain). That leaves just over half the list as truly consequential contributions to the development of utilitarianism prior to Bentham. Indeed, both in his “Article on Utilitarianism” and elsewhere, the main sources of inspiration Bentham acknowledges in connection with the conceptual, methodological, and terminological background of utilitarianism are Bacon (1620),12 Hume (1739–1740; 1751; 1753),13 Hartley (1749),14 Helvétius (1758),15 Beccaria (1764/1767),16 and Priestley (1768).17 He professes an intellectual debt to all six with respect to the basic philosophical outlook in utilitarianism, while singling out Hume, Beccaria, and Priestley as his primary sources for the corresponding terminology, though with apparent confusion and outright hesitation between the precise influence of the latter two, as discussed later (see section 1.3 below). Five of the six are examined in greater detail, some more than others, over the course of this chapter, as their respective contributions become relevant to the theme under discussion, first in regard to utilitarian terminology, next in relation to utilitarian philosophy. Not belonging to either category, Bacon nevertheless deserves some attention here. Among the six consequential influences, Bacon is the only one associated with ancillary insight and inspiration rather than with direct development
12
Bacon (1620) is discussed in the long version, but not in the short version, of Bentham’s “Article on Utilitarianism” (1829/1983, 294–295, 310–311). He is also mentioned in Bowring’s “History of the Greatest- Happiness Principle” (1834, 313, 316– 318) and in Perronet Thompson’s “ ‘Greatest Happiness’ Principle” (1829, 260). 13 Hume (1741–1742) enjoys a prominent place in all three of the sources directly relevant to Bentham’s perspective on utilitarian history: Bentham (1829/1983, 289–290, 292, 299, 322–324); Bowring (1834, 291–295); Perronet Thompson (1829, 258). 14 Hartley (1749) comes up in all three of the relevant sources, albeit relatively briefly in each one: Bentham (1829/1983, 290–291, 324); Bowring (1834, 295); Perronet Thompson (1829, 258). 15 Helvétius (1758) appears in Bentham’s biographical fragments (see Bentham 1843, vol. 10, 27, 54, 70–71), among other places, in addition to the three sources considered here: Bentham (1829/1983, 290–291, 299, 325, 327); Bowring (1834, 295–297, 314); Perronet Thompson (1829, 258). 16 Beccaria’s (1764/1767) influence comes out in Bentham’s memoirs and correspondence (see Bentham 1843, vol. 10, 54, 142, 296; cf. Shackleton 1972, 1465–1466) instead of in any of the three works ventilating his perspective on the main developmental forces in utilitarianism. His name does not come up at all in Bentham’s “Article on Utilitarianism” (1829/1983), Bowring’s “History of the Greatest-Happiness Principle” (1834), or Perronet Thompson’s “ ‘Greatest Happiness’ Principle” (1829). 17 Priestley (1768) receives considerable attention in connection with terminological issues: Bentham (1829/1983, 290–293, 325–326); Bowring (1834, 295, 298–300); Perronet Thompson (1829, 258–259).
Classical Utilitarianism before John Stuart Mill
23
within the framework of utilitarianism as a work in progress. His influence manifests itself in two ways. The first is his emphasis on experimentation, originally in Novum Organum (1620), commonly accepted as the inception of the scientific method and invoked by Bentham as the impetus for his own focus on observation, specifically on the observation of human nature, given the inherent barriers to actual experimentation in the moral, social, and political sphere.18 Bacon’s dictum, fiat experimentum (“Experimentalize!”), adapted by Bentham as fiat observatio (“Observe!”), had a profound influence on Bentham, far exceeding the merely rhetorical convenience of an otherwise superfluous catchphrase (Bentham 1829/1983, 294–295; Bowring 1834, 316–318; Perronet Thompson 1829, 260). Systematic attention to observation, reports Bentham, is what reaffirmed his adoption of happiness as the standard of utility, and of pleasure and pain as the measure of happiness, encouraging him to set up the felicific calculus as the methodological apparatus for all his work as a moralist, activist, reformist, and jurist. The second manifestation of Bacon’s influence on Bentham is his tree of knowledge, originally appearing in his taxonomic initiative, The Advancement of Learning (1605), and later becoming the inspiration for Bentham’s attempt to explore the connections between the arts and sciences in terms of their propensity to produce happiness. To be more specific, Bentham’s advancement of happiness as a pedagogical, ideological, and operational point of reference unifying the arts and sciences, which is the central theme of his Chrestomathia (1816), was inspired by Bacon’s (1605) paradigmatic tree of knowledge, the taxonomic model of all learning in the metaphor of a tree branching out in various directions from a central trunk, with further methodological developments coming through the efforts of Encyclopédistes such as Jean-Baptiste le Rond d’Alembert (in collaboration with Denis Diderot) and Ephraim Chambers (Bentham 1829/1983, 310–311; Bowring 1834, 313; Shackleton 1972, 1476). Bentham’s own contribution to the development of utilitarianism comes through a mixture of moral, social, political, and legal tracts and treatises collectively underscoring the complex historical significance of Bentham as a moral philosopher, social commentator, political reformer, and, perhaps above all, a jurist par excellence. That being so, the works relevant to utilitarianism 18 Bacon’s Novum Organum (1620) is a revolutionary tract instituting the scientific method, or at the very least anticipating it, in opposition to the Aristotelian tradition still in currency at the time. The novelty asserted in the title is an allusion to the author’s intended improvement upon Aristotle’s Organon, a collection of the latter’s works on logic: Categories; On Interpretation; Prior Analytics; Posterior Analytics; Topics; Sophistical Refutations.
24
Chapter 1
in his intellectual output are not limited to his publications in moral philosophy. Bentham (1829/1983) himself considers his intellectual contribution to utilitarianism to be concentrated in the following five books, annotated below with respect to the relevant contribution of each, though not necessarily in explication of the central thesis or overall theme of each:19 – A Fragment on Government (1776) associates utility with happiness, specifically through the elucidation of “the principle of utility” (a term adopted from Hume) in terms of “the greatest happiness of the greatest number” (a phrase inspired by Beccaria, later recalled erroneously as Priestley). – An Introduction to the Principles of Morals and Legislation (1789) defines happiness in terms of pleasure net of pain, together with a detailed inventory of pleasures and pains as motives, building on the premise, or observation, that pleasures and pains are the fundamental constituents of cognitive content and sentient experience, and therefore, the principal determinants of human motivation. – Chrestomathia (1816) identifies and employs happiness as a jointly indicative organizational standard in all branches of the arts and sciences, effectively illuminating their mutual relationships while facilitating their categorization on the basis of the contribution of each one to happiness as a common frame of reference. – A Table of the Springs of Action (1817/1983) classifies desires and interests with respect to their relationship with motives, while illustrating the connection of each one with pleasures and pains, together with a supplementary analysis of “other psychological entities, such as affections, passions, virtues, vices, moral good, moral evil, etc.” (Bentham 1829/1983, 327). – Codification Proposal (1822) elaborates on the previously established links between utility and happiness, and between either one of those and pleasure and pain, while persistently qualifying “the greatest happiness” as pertaining to “the greatest number,” a refinement Bentham later mistakenly recollects (Bentham 1829/1983, 327; Bowring 1834, 319) as having been introduced here for the first time, whereas its inception actually coincides with A Fragment on Government (1776), as discussed in section 1.6 below.
19
See Bentham (1829/1983, 292–294, 310–311, 326–328), Bowring (1834, 302–304, 311–319), and Perronet Thompson (1829, 259–261) for Bentham’s personal appraisal of the contribution of each of his works to the development of utilitarianism. Competing assessments and complementary observations are both conceivable and available, but these three references are the most relevant sources, as they are the only ones that come directly from Bentham, the latter two being grounded in the first one.
Classical Utilitarianism before John Stuart Mill
25
Bentham’s estimation of the creative forces behind the emergence and development of utilitarianism can be supplemented by a host of other perspectives on the same process. While the assessment of any standpoint benefits from a consideration of alternatives, including those that are contradictory as well as those that are complementary, Bentham’s position in this instance is not so much an opinion to be evaluated as it is a chronicle to be assimilated. He does have certain biases regarding what is important and what is not, which can and must be tempered with a dispassionate review of the facts, and he does assign himself a singularly important role in the development of utilitarianism, which can perhaps be disputed, albeit not very fairly or reasonably, but his historical sketch probably requires, if anything at all, further elaboration rather than substantive correction. Hardly anyone denies Bentham’s role in the history of utilitarianism, least of all his critics, the staunchest of whom oppose him not for his understanding of utilitarianism, nor even for his subscription to utilitarianism, but specifically for his contributions to utilitarianism. As for complementary sources, the two obvious choices for consultation are the closely related pieces by Perronet Thompson (1829) and Bowring (1834). Given the fairly detailed introductory discussion above, including the exclusive emphasis on the relationship between these two works and Bentham’s manuscript (1829), a brief look at the nature and contents of these derivative contributions will suffice at this point. Perronet Thompson’s “ ‘Greatest Happiness’ Principle” (1829) is a defense of utilitarianism, particularly of its fundamental principles, as well as an overview of its development, even though it was conceived specifically as an editorial response representing The Westminster Review in opposition to a critique of James Mill by Macaulay (1829) in The Edinburgh Review. Although the piece is not concerned directly with the history of utilitarianism, it is useful in that capacity because it, much like Bowring’s “History of the Greatest-Happiness Principle” (1834), draws on the text of Bentham’s “Article on Utilitarianism” (1829/1983), the latter of which was, in fact, prepared as a reference manual of sorts, first used to facilitate the editorial response to Macaulay. Perronet Thompson’s approach can be broken down as follows (1829): – review of the charges in Macaulay’s critique of James Mill (pp. 254–258); – historical overview of utilitarianism (pp. 258–261); – Bentham’s predecessors and contemporaries (p. 258); – Bentham’s own contributions (pp. 259–261); – Bacon’s influence on Bentham (p. 260); – response to the relevant charges against utilitarianism (pp. 261–267); – essence and significance of nomenclature (pp. 267–268).
26
Chapter 1
The penultimate item on the list, the editorial response to the charges leveled against utilitarianism in Macaulay’s critique of the work of James Mill, is an expository sketch of utilitarianism, including a discussion of its superiority over competing schools of thought, especially over social and political theories grounded in the social contact tradition. It is also a defense of some of the more visible public virtues of the utilitarian movement, such as its advocacy of representative government and its promotion of universal suffrage, both a matter of contention at the time. The expository and defensive portions together serve as the intended response to the critique of James Mill, and that response, in turn, serves as a vindication of utilitarianism itself against the corresponding charges. The overall effort now stands as an informative source on the development of classical utilitarianism with a close look at Bentham’s role in that process. Bowring’s “History of the Greatest-Happiness Principle” (1834), much like the editorial coverage of Perronet Thompson (1829), is a descriptive account of the nature of utilitarianism, interlaced with a historical survey of its emergence and development. There are two significant differences between the two. The first is that Bowring’s appraisal is not concerned with the views or works of James Mill, nor with any particular attack upon them, whereas Perronet Thompson’s defense is indeed indexed to that issue. The second is that, even though both are grounded in Bentham’s “Article on Utilitarianism” (1829/1983), Bowring’s text follows Bentham’s manuscript more fully, if not more closely, than does the piece by Perronet Thompson. The second difference follows from the first. Bowring’s text remains closer to Bentham’s, covering more of its contents, precisely because it was not burdened with formulating a response to Macaulay’s critique of James Mill, a polemical assault directly motivating the defensive initiative by Perronet Thompson. Bowring was, in fact, entirely at liberty to duplicate whatever he found useful in Bentham (1829), without the slightest hesitation or limitation, given that the survey he was preparing was to be published under Bentham’s name anyway, specifically as part of the latter’s posthumously published Deontology (1834). He therefore drew liberally on the text of Bentham, employing long quotations as well as parallel narration, with some differences and some elaboration, all sanctioned in advance by Bentham, though more through the blanket delegation of full editorial authority than through the direct and separate approval of each editorial decision. The survey is positioned physically as a forty-five page addendum at the end of the first volume of Bentham’s Deontology (1834), edited by Bowring without professing any personal input other than the arrangement and articulation of the notes, testimony, and memoirs of Bentham.
Classical Utilitarianism before John Stuart Mill
27
Since Bowring’s goal was nothing more than the dissemination of Bentham’s latest views on ethics, including the publication of the manuscript, Bentham’s (1829/1983) and Bowring’s (1834) texts are so similar that distinguishing between them is nearly impossible in places. As a matter of fact, the original by Bentham (1829/1983) refers to Bentham in the third person, as if the reference were to someone else, which was a deliberate choice by Bentham in the preparation of a grounding narrative to be used as source material in derivative accounts, including the anonymous editorial by Perronet Thompson (1829), whereas the historical survey by Bowring (1834) presents long quotations with Bentham speaking in the first person, converted from the original in the third person, which then makes Bowring’s version look like the original and Bentham’s the biographical commentary, at least in the parts in question. The topical breakdown of Bowring’s treatment is as follows (1834): – Bentham’s predecessors and contemporaries (pp. 290–300); – Bentham’s own contributions (pp. 302–304, 311–319); – Bentham’s opposition to Locke (pp. 304–310); – Bentham’s opposition to Paley (pp. 310–311); – Bacon’s influence on Bentham (pp. 316–318); – essence and significance of nomenclature (pp. 320–323, 328–331). Moving on to sources outside the domain of those anchored directly to Bentham’s “Article on Utilitarianism” (1829/1983), one would find the literature practically inexhaustible, though further removed from Bentham, if not from classical utilitarianism in general. Coverage of even just the most useful entries there would be a task for another kind of book, but a brief look at a couple of more of the items already listed above would be helpful here, one a highly learned commentary by Henry Sidgwick (1877), the other a tightly focused study by Robert Shackleton (1972). Sidgwick’s (1877) perspective on Bentham’s contributions to utilitarianism is important not only because he rejected Bentham’s approach to utilitarianism, thus precluding the possibility of an exaggerated view of the latter’s achievements, but also because Sidgwick himself was a prominent figure in the subsequent development of classical utilitarianism. Shackleton (1972) is more important as a commentator than as a contributor, given the propaedeutic nature of his investigative account of utilitarian nomenclature and terminology. Unlike the other works examined so far, Sidgwick’s “Bentham and Benthamism in Politics and Ethics” (1877) is neither an overview of utilitarianism nor a survey of its history. As the title may readily indicate, it is instead a biographical appraisal of the life, character, career, and philosophy of Bentham. That, of course, makes it a natural resource for a glimpse into the development of classical utilitarianism as well. It is even more closely concerned with
28
Chapter 1
Bentham than are the other works examined here, including the article by Bentham himself. Sidgwick introduces his assessment as an attempt to make up for what he considers inadequate attention to Bentham in Leslie Stephen’s History of English Thought in the Eighteenth Century (1876). Stephen indeed devotes little attention to Bentham in the work in question, but only because he associates Bentham more with the nineteenth century than with the eighteenth, which is precisely what Sidgwick takes issue with in the first place. For what it is worth, Stephen dedicates considerable attention and devotes enormous space to Bentham more than three decades later in The English Utilitarians (1900), which Sidgwick naturally would have had no way of knowing at the time he was reacting to the earlier work (1876), and no chance to read later, or ever at all, because he died in the same year as the publication of the later work (1900). Sidgwick succeeds admirably in bringing out Bentham’s contributions to the development of utilitarianism as well as in illustrating the personal prejudices coloring his endeavors and the ideological conflicts frustrating his plans toward greater accomplishments. While he demonstrates obvious ideological biases against Bentham’s approach to moral, social, and political issues, his analysis remains accurate enough to serve as a complementary perspective opposite those of Perronet Thompson, Bowring, and indeed Bentham himself in their supremely positive estimations of Bentham as a moral, social, and political philosopher, apparently without any faults or shortcomings. The highlights of Sidgwick’s (1877) study can be summarized as follows in keeping with the presentation format followed thus far with the other entries: – characteristic merits and notable defects of Bentham as a scholar (p. 628); – development of Bentham’s career and reputation (pp. 629–630); – originality and importance of Bentham as a man of letters (p. 631); – approach and style of Bentham as cold, pedantic, and prudential, with excessive attention to detail, especially in social and political matters (pp. 631–634); – remarkable diversity of Bentham’s interests, all revolving around the theme of utility (pp. 634–637); – Bentham’s preoccupation with promoting his design for a Panopticon (Inspection House), an institutional complex with a central rotunda allowing the observation of occupants without their awareness, suitable for implementation in any institutional setting but developed and promoted specifically as a penitentiary (pp. 634–635); – dual aspect of Bentham’s utilitarianism in combining psychological egoism with universalistic ethical hedonism through the external
Classical Utilitarianism before John Stuart Mill
29
intermediation of legislative controls: “The moralist must have recourse to legislation” (pp. 638–641, 647–652); – Bentham’s participation in radical reform movements, together with his work as a jurist and codifier, appreciated abroad, particularly in France and Russia, while held back domestically for two decades (1791–1811) in connection with the fate of his pet project, the Panopticon, eventually rejected by King George iii (pp. 642–647); – pressing and persistent emphasis of Bentham on limiting sovereign authority through popular controls in addition to internal checks and balances (pp. 644–647); – dependence of Bentham’s moral theory on legislation in order to secure the requisite coincidence between interests and duties, given the dual aspect of his utilitarianism as a combination of psychological egoism and universalistic ethical hedonism (pp. 647–652; cf. 638–641); – normative role of utilitarianism in “judgments of approbation and disapprobation” and “as a system of distributing praise and blame,” with the primary function of the utilitarian moralist being “to apply this standard to the particulars of human life, so as to determine by it the different special virtues or rules of duty” (pp. 647–652); – professional and philosophical development of John Stuart Mill through the modification of Bentham’s utilitarianism with a combination of “Comtian sociology, Associational psychology, and Neo-Baconian logic” (p. 652). Shackleton’s contribution, “The Greatest Happiness of the Greatest Number: The History of Bentham’s Phrase” (1972), differs in at least two respects from the rest of the sources under consideration here. First, it is a far more recent contribution than any of the others, coming as it does nearly a full century after the biographical assessment of Sidgwick, with the benefit of hindsight relative to later developments. Second, it investigates the history of utilitarianism with an exclusive focus on the terminology of the movement, in contrast to the more typical initiative to explore the underlying philosophy or ideology, thereby complementing rather than competing with all such sources. Despite coming much later than the others, at least compared to the ones examined here, it constitutes a seminal study of the emergence and development of classical utilitarianism, as it stands out from the rest both thematically and methodologically. Shackleton works with the original manuscript of Bentham’s “Article on Utilitarianism” (1829), which also informs the historical accounts of Perronet Thompson (1829) and Bowring (1834).20 His analysis is consistent with all three 20
Robert Shackleton (1972) recognizes Bentham as the author of Bowring’s “History of the Greatest-Happiness Principle” (1834), while identifying Bowring as the author of Perronet
30
Chapter 1
sources serving as more or less direct records of Bentham’s perspective. He is thus able to illuminate Bentham’s role in the development of utilitarianism, especially in terms of the attendant terminology, by drawing insightful historical and philosophical connections between the contributors and contributions commonly acknowledged to have inspired Bentham. The thematic details of his research can be broken down as follows (pp. 1461–1482): – Bentham’s deference to Priestley (pp. 1463–1465); – Bentham’s debt to Helvétius and Beccaria (pp. 1465–1466); – Hutcheson’s contribution to utilitarian terminology (pp. 1466–1471); – Beccaria’s role in propagating Hutcheson’s terminology (pp. 1471–1474); – Bentham’s experiments with terminology (pp. 1474–1482); – terminological hiatus in Bentham (pp. 1476–1482). 1.2
Bentham as an Anchor for Historical Insight
Bentham is a good anchor for historical insight into utilitarianism, not just from the vantage point of John Stuart Mill, but also as a pivotal benchmark for utilitarianism in general. Exploring the origins and development of utilitarianism from the perspective of Bentham is not so much a restriction of the scope of consideration as it is a consolidation of the aims of investigation. It is not, at any rate, a deviation from the actual course of events, nor worse, a distortion of their historical significance. It is not even a compression of the chronology, given that there is no particular point of origin or path of progress, either for the central ideology or for the corresponding terminology, which then makes developments relative to Bentham not only perfectly representative but also supremely relevant. It is both desirable and advisable, therefore, if not outright
Thompson’s “ ‘Greatest Happiness’ Principle” (1829). Attributing authorship to Bentham instead of Bowring is quite reasonable in the former case (Shackleton 1972, 1462), where the piece was published under Bentham’s name anyway, with Bowring himself claiming no more than editorial credit, but attributing authorship to Bowring instead of Perronet Thompson is not entirely convincing in the latter case (Shackleton 1972, 1481), where the original publication was anonymous, thus pointing simply to the editor of the corresponding journal as its author, given the editorial nature and tone of the contribution. Shackleton (1972, 1481) indeed takes Bowring to have been the editor in charge, while Amnon Goldworth (Bentham 1829/1983, xxxiv–xxxv) reports that the editor was actually Perronet Thompson. As a matter of fact, both are correct, at least in their assessment of editorship, as Bowring and Perronet Thompson were serving as coeditors at the time, with Bowring holding office between January 1824 and January 1836, and Perronet Thompson joining him between January 1829 and January 1836.
Classical Utilitarianism before John Stuart Mill
31
necessary, to keep the coverage firmly anchored to Bentham, especially where the aim is to establish the interpretive groundwork for Mill. The absence of a definitive origin for utilitarianism is a natural reflection of its ontological character and existential structure rather than an accidental consequence of gaps or uncertainties in the available records. The problem is not that we are not able to trace the relevant history accurately but that there is no particular path to follow to the exclusion of others. There are instead several such paths, each with an independent chronology, jointly representing various features and dimensions of utilitarianism as a school of thought. This is because utilitarianism is not a simple concept or single outlook but a complex construct bringing together separate philosophical perspectives, including psychological egoism, ethical hedonism, and moral universalism, a combination commonly described as the marriage of psychological egoism with universalistic ethical hedonism. The complexity of philosophical forces making up utilitarianism, while emphasizing the distributive and collective significance of its integral components, draws attention to ancillary developments and peripheral influences as they contribute to the evolution of the unitary platform. Adequate acquaintance with any philosophical system requires complete familiarity with the proper nomenclature. The best insight into classical utilitarianism is likewise through its own vocabulary, which was a tremendously important consideration for Bentham, which, in turn, makes it equally important for the study of Mill, who had no qualms about borrowing what he found in Bentham, proudly citing him as a source both in terminology and in ideology, with obvious modifications to the latter but hardly any to the former. The most important terms and expressions in that regard are “the principle of utility,” “the greatest happiness principle,” and “the greatest happiness of the greatest number.” These constructions represent staples of utilitarian nomenclature, at least from a historical perspective, while exemplifying the most familiar fixtures in the vocabulary of Bentham. The terminological predilections of Bentham varied over time, changing course more than once and rather abruptly and radically each time, as discussed in detail in the final section of this chapter (section 1.6). Yet even though the motivational dynamics shaping his inclinations are better appreciated in cognizance of the larger history of utilitarian nomenclature, which is why developmental details are left to the end of the chapter, the terminology itself is eminently relevant at this point. The name that Bentham originally favored for the “fundamental axiom” of moral discourse within the framework of utilitarianism was “the principle of utility,” later supplemented, and at times supplanted, by “the greatest happiness principle.” Both were elucidated as the appraisal of morality in terms of the production of happiness,
32
Chapter 1
which was enthusiastically articulated and elaborated as “the greatest happiness of the greatest number” in several of Bentham’s works, especially later in his career. All three expressions, together with conceptual correlates and experimental variations, can be explored in context through most of his ethical works and found in abundance in a good many of his other publications. The principle of utility, both the doctrine and the designation, makes its debut as early as A Fragment on Government (1776), but it finds a much better venue for articulation, explication and ventilation in An Introduction to the Principles of Morals and Legislation (1789). The entirety of the first chapter of the latter work (“Of the Principle of Utility”) is dedicated to the systematic exposition of the standard of morality envisaged by Bentham (1789, 1–6). The very first paragraph there opens with a rhetorically gifted description of the principle of utility in terms of a compulsory submission to the hedonistic reality awaiting at the intersection of the course of nature and the standard of morality: nature has placed mankind under the governance of two sovereign masters, pain and pleasure. It is for them alone to point out what we ought to do, as well as to determine what we shall do. On the one hand the standard of right and wrong, on the other the chain of causes and effects, are fastened to their throne. They govern us in all we do, in all we say, in all we think: every effort we can make to throw off our subjection, will serve but to demonstrate and confirm it. In words a man may pretend to abjure their empire: but in reality he will remain subject to it all the while. The principle of utility recognises this subjection, and assumes it for the foundation of that system, the object of which is to rear the fabric of felicity by the hands of reason and of law. Systems which attempt to question it, deal in sounds instead of sense, in caprice instead of reason, in darkness instead of light. bentham 1789, 1
This naturalistic account of the necessity and rationality of a utilitarian approach to morality is followed immediately by a definition of its fundamental principle, using the felicific formula still invoked today as a standard model, though often with modifications in accordance with conceptual departures, ideological deviations, and philosophical developments: The principle of utility is the foundation of the present work: it will be proper therefore at the outset to give an explicit and determinate account of what is meant by it. By the principle of utility is meant that principle
Classical Utilitarianism before John Stuart Mill
33
which approves or disapproves of every action whatsoever, according to the tendency which it appears to have to augment or diminish the happiness of the party whose interest is in question: or, what is the same thing in other words, to promote or to oppose that happiness. I say of every action whatsoever; and therefore not only of every action of a private individual, but of every measure of government. bentham 1789, 2
This introductory definition of the principle in strictly hedonistic terms quickly blends into a fuller articulation of the notion of utility with reference to the production of advantage, broadly construed, together with the correlative prevention of disadvantage, thereby anticipating the desire-satisfaction models later developed as a refinement of the classical version of the doctrine: By utility is meant that property in any object, whereby it tends to produce benefit, advantage, pleasure, good, or happiness (all this in the present case comes to the same thing), or (what comes again to the same thing) to prevent the happening of mischief, pain, evil, or unhappiness to the party whose interest is considered: if that party be the community in general, then the happiness of the community: if a particular individual, then the happiness of that individual. bentham 1789, 2
The remainder of the chapter, and indeed the rest of the book, then goes on to discuss the principle in greater detail, including the underlying aims and intended uses, as well as the main rivals, leading alternatives, and common objections. Regular references to alternative names for the principle come in later publications. Yet the ensuing expansion of the terminology is announced and explained in a footnote included in subsequent editions of both A Fragment on Government (1776) and An Introduction to the Principles of Morals and Legislation (1789), each coming out with its respective second edition in the same year (1823). The note in the former is more explanatory than the one in the latter and, in fact, includes the entirety of the text of the latter almost verbatim as part of its own content. The main purpose of the relevant note in the second edition of A Fragment on Government (1823) is to introduce “the greatest happiness principle” and “the greatest felicity principle” as alternative names for the designation “the principle of utility.” The introduction comes with an explanation, doubling as justification, that the alternatives are superior to the original because they immediately and permanently bring out the connection between utility and
34
Chapter 1
happiness, which is lost in a reference to utility alone. The point of the explanation is that the singular focus of the principle of utility on the greatest happiness of everyone concerned as the only standard of morality is best served by a name that actually reflects that focus, which then validates both “the greatest happiness principle” and “the greatest felicity principle” as appropriate choices. The note opens as follows: To this denomination [“the principle of utility”], has of late been added, or substituted, the greatest happiness or greatest felicity principle: this for shortness, instead of saying at length, that principle which states the greatest happiness of all those whose interest is in question, as being the right and proper, and only right and proper and universally desirable, end of human action: of human action in every situation; and, in particular, in that of a functionary, or set of functionaries, exercising the powers of Government. The word utility does not so clearly point to the ideas of pleasure and pain, as the words happiness and felicity do; nor does it lead us to the consideration of the number of the interests affected; so the number, as being the circumstance which contributes, in the largest proportion, to the formation of the standard here in question; the standard of right and wrong, by which alone the propriety of human conduct, in every situation, can with propriety be tried. bentham 1823, A Fragment on Government, 45–4621
This clarification does not represent the first time the expanded terminology is invoked or employed, a development discussed in detail in section 1.6 below, but it does represent the first explanation offered for the expansion itself. The remainder of the note is a rhetorical response to a polemical critique of the principle of utility in an entirely different context. That is the portion that incorporates the text of the corresponding note in the second edition of An Introduction to the Principles of Morals and Legislation (1823).22 While 21
22
This is the opening paragraph of a footnote (note l) added to the first sentence of section 48 at the end of the first chapter of Bentham’s A Fragment on Government (1776, 62) in its second edition (1823, 45–48). More than half the text of the note consists of the repetition (with slight stylistic modifications) of an addendum Bentham had recently (12 July 1822) introduced to a footnote (note c) to the first sentence of section 13 of the first chapter of An Introduction to the Principles of Morals and Legislation (1789, 5) in its second edition (1823, 7–9). This is a reference to a footnote (note c) to the first sentence of section 13 of the first chapter of Bentham’s An Introduction to the Principles of Morals and Legislation (1789, 5) as expanded with an addendum (12 July 1822) in the second edition of that work (1823, 7–9).
Classical Utilitarianism before John Stuart Mill
35
Bentham’s response to the polemical attack constitutes a digression in the flow of either work, which is why it is presented in a footnote in both places, his response nevertheless remains relevant to the explanation quoted above, as it goes on to elaborate on the intended distribution of the happiness to be maximized in accordance with the principle of utility. Repeated references to “the greatest happiness of the greatest number” become especially significant, since the expression had already been introduced in the preface to A Fragment on Government (1776) in its original edition, though indeed remaining confined there to the preface.23 While burying the matter in footnotes may not seem like the most effective means of communication, the notes in either case are rather lengthy, each spanning a few pages, and Bentham certainly makes up for any deficiency they may represent in publicity by passionately promulgating the terminological expansion in subsequent publications (see section 1.6 below). To consider just his ethical works, both A Table of the Springs of Action (1817/1983) and Deontology (1834/1983), for example, make use of the expanded vocabulary in abundance. The marginalia in A Table of the Springs of Action are especially rich in references to “the greatest happiness of the greatest number.” The following are just a few examples from the 815 marginal notes accompanying the work:24 231. Utilitarianism—Utilitarian philosophy—Utilitarian principle of utility: ‘Act according to the greatest happiness of the greatest number.’ bentham 1817/1983, 25
415. In its censorial sense, in so far as a man takes for his guide the principle of utility, the greatest happiness of the greatest number is what his conduct tends to, by definition. bentham 1817/1983, 41
23 24
Other passages from Bentham’s A Fragment on Government (1776) that are relevant to the present discussion include: pp. i–ii, xlv–xlvi, 62, 149–150, 158–159, 170–171. The marginalia recovered from Bentham’s original manuscripts and incorporated into the 1983 edition of A Table of the Springs of Action (1817/1983) are numbered sequentially and categorized under five headings: “Introduction” (## 1–51, pp. 5–9); “Explanations” (## 52–68, pp. 10–11); “Observations” (## 69–202, pp. 11–23); “Added Observations” (## 203– 653, pp. 23–59); “Uses” (## 654–815, pp. 59–73). Citations in the main text above provide item numbers in addition to page numbers, because item numbers enhance specificity, given that there are multiple entries on each page. Marginal notes that are most relevant to the discussion here, including the ones already quoted, are the following (citing item numbers alone): ## 231, 362, 366, 415, 512, 540, 600, 608, 655, 656, 658, 664, 665, 666, 667, 668, 669, 678, 679, 680, 696, 697, 698:2, 726, 729, 730, 735, 785, 786, 787.
36
Chapter 1
540. The end of the Utilitarian, the greatest happiness of the greatest number—the pleasure of sympathy and reputation from his exertions. bentham 1817/1983, 51
600. The principle of the greatest happiness to the greatest numbers is opposite to selfishness. bentham 1817/1983, 56
608. Answer. The end proposed by the Utilitarian is the greatest happiness of the greatest number, not the temporary gratification of any. In this maxim is no selfishness. The charge of selfishness only arises from the enunciation of the principle. bentham 1817/1983, 57
655. Principle of Utility: 1. In its censorial sense, it holds up the greatest happiness of the greatest number as the only universally desirable end. bentham 1817/1983, 59–60
658. Utilitarianism states as the only proper end in view of the moralist and legislator the greatest happiness of the greatest number. bentham 1817/1983, 60
697. Principle of utility, what. 1. Indication of what ought to be, it indicates as the only universally desirable object and end, greatest happiness of greatest number. Bentham 1817/1983, 62
Complementing the references scattered throughout the marginalia of A Table of the Springs of Action (1817/1983), Bentham’s Deontology (1834/1983) opens with a general overview (pp. 121–124), followed immediately by an expository section dedicated to “Definitions, Explanations and Arrangements” (pp. 124– 129).25 The exposition establishes up front, in the very process of defining Bentham’s conception of deontology, that moral philosophy is all about the pursuit and promotion of happiness:
25 Bentham’s Deontology (1834/1983) may be consulted at length in connection with the present discussion. Some of the most relevant passages are on pp. 121–124, 124–129, 130– 134, 147–150, 150–153, 166, 187–190, 192, 213.
Classical Utilitarianism before John Stuart Mill
37
Deontology, or Ethics (taken in the largest sense of the word), is that branch of art and science which has for its object the learning and shewing for the information of each individual, by what means the net amount of his happiness may be made as large as possible; of each in so far as it is dependent on his own conduct: the happiness of each individual separately being considered, and thereby that of every individual among those whose happiness is on this occasion an object of regard. bentham 1834/1983, 124–125
This early albeit implicit emphasis on the maximization and distribution of happiness later becomes explicit as Bentham goes on to expand on the intended coverage by expressly invoking “the greatest happiness of the greatest number” throughout the book: Take any religion, meaning by any religion the notion or notions of any human person or persons on the subject of religion—if in any of its precepts it be irreconcilable to the principle of utility, what is the consequence? That the religion is false. For of the falsity of any religion there can not be plainer nor more conclusive evidence than the fact, supposing it proved, that in any point it is repugnant to that principle which for the test of propriety in every act takes its conduciveness or repugnancy to the greatest known happiness of the greatest number of mankind. bentham 1834/1983, 166
According to you, the principle which in the case of every action establishes as the ground and measure of its propriety its conduciveness or repugnance to the greatest happiness of the greatest number is the sure and only true principle. bentham 1834/1983, 192
But if it be not, then if the principle of utility—if the greatest happiness of the greatest number—be the true and only standard, and conduciveness or opposition to that end the only test of right and wrong, then in any other case, except where in one or other of these two shapes or both together enjoyment is productive of preponderant mischief, neither can abstinence have any claim to the name and praise of virtue, nor enjoyment be with justice considered as subject to the reproach of vice. bentham 1834/1983, 213
38
Chapter 1
The expansion of Bentham’s philosophical and rhetorical vocabulary, from “the principle of utility” as the standard designation, to various formulations invoking “happiness,” figures prominently in his “Article on Utilitarianism” (1829/1983) as well. The short version of the article begins with a reference to the new terminology, announcing the transition in the subtitle of the article, or in the very first sentence at the latest, depending on one’s perspective on the organizational structure: “History of the principle originally styled the principle of utility, now at last the greatest happiness principle” (Bentham 1829/1983, 320). The first three paragraphs of the short version, immediately following the subtitle, or the opening sentence, present a particularly useful overview of the principle in terms of its aims, uses, and intentions: 1. The character in which, for so many years past, Mr. Bentham has been putting this principle to use is that of an instrument of direction: or, say, in more familiar language, a direction for pointing out the path most proper to be pursued, on every occasion, in public as well as private life; to be pursued by every individual, whether acting in his private and individual capacity, as a member of the whole community, for his own benefit alone, or in his public capacity acting for the benefit of others in the character of a member of the governing part of the community, in the exercise of the powers belonging to him in that same character. 2. It is the oracle to which, on every occasion, he applies for instruction: and the several arrangements which in relation to the several parts of the field of thought and action it suggests, together with the reasons from which those same proposed arrangements derive their explanation and support, constitute its responses. 3. To this purpose it renders service to him in three distinguishable capacities: 1. as end in view; 2. as a storehouse of means employable for the attainment of that end; 3. as a storehouse furnishing motives by the force of which, on the several occasions, men may be induced to act in ways conducive to that end. bentham 1829/1983, 320
While these details are common and peculiar to utilitarianism in any formulation, the reason that Bentham was eager to introduce “the greatest happiness principle” as an alternative name for “the principle of utility” is that he recognized the potential of the former to capture and communicate the utmost concern of utilitarianism, namely the greatest happiness of the greatest number, which was a common ideal shared with the collective consciousness evolving at the time in Western social and political systems. The paradigm of the greatest
Classical Utilitarianism before John Stuart Mill
39
happiness of the greatest number provided such a natural if loose fit with the central ideals of the European Enlightenment that the reference was already on its way to becoming a slogan among philosophers and other intellectuals active during the eighteenth and nineteenth centuries.26 The expression presented a natural fit in the sense that it seemed to take into account everyone’s happiness, especially if read simply as the greatest amount of happiness altogether, among the greatest number of people possible, regardless of who they might be, hence presumably without discrimination. Yet the expression also constituted a loose fit because it was vague, disconcertingly so in remaining open to interpretation, whether by proponents or by critics, with respect to the distribution intended as a consequence of the happiness maximized. To elaborate, making a sufficient number of people adequately happy could conceivably be taken to compensate for making any number of people utterly miserable, no matter the reason or the circumstances, so long as there was a net increase in aggregate happiness. Even the mere possibility of such a reading came to be a source of profound frustration for Bentham, as discussed in section 1.6 below, because the otherwise innocuous formula could then be appropriated for the justification of great social injustice, at the very least in protecting, preserving, and promoting the good of the prevailing majority at the expense of any and all minorities. For better or worse, though, once the terminology caught on, it remained in currency throughout the Age of Reason, retaining its popularity at least through the end of the nineteenth century. Its rhetorical force continues to hold attraction today, though the phrase itself is no longer employed in serious scholarship, especially not by professional ethicists, except when discussing the historical context or illustrating the qualifications required to remove the inherent ambiguities and to clarify the intended meaning, which can still support different interpretations. 1.3
Origins and Development of Utilitarian Nomenclature
Bentham was responsible more for the institution and longevity of utilitarian nomenclature than for the invention or formulation of the corresponding terms. He had a prominent role in developing utilitarian terminology and in bringing the prevailing and emerging vocabulary into common use in moral, 26
Some of the leading figures, all discussed in due course in the main text, include the following: Cesare Beccaria (1738–1794); Jeremy Bentham (1748–1832); Claude Adrien Helvétius (1715–1771); Francis Hutcheson (1694–1746); James Mill (1773–1836); John Stuart Mill (1806–1873); William Paley (1743–1805); Joseph Priestley (1733–1804).
40
Chapter 1
social, and political discourse. He was not, however, the first person to use the leading designations and expressions in that context, nor in any other context, and certainly not the first to do so in print.27 Nor did he ever assert, affirm, or imply that he had personally come up with either the terms or the notions, save for his various terminological experiments that never actually became a part of the established vocabulary, not even his own.28 Quite the contrary, he was eager to credit the relevant sources, at least his own sources, though he was torn between two in particular as his actual inspiration for the expression “the greatest happiness of the greatest number.” He seems confident in certain places that he had originally encountered the phrase in Joseph Priestley, while intimating elsewhere that it may have instead been Cesare Beccaria. This does not mean that either one of the two was, in fact, the first to have used the term in print, just that Bentham thought, to the best of his recollection, that one or the other was the first that he himself had seen in print. To be more specific, Bentham’s earliest reference to Priestley’s terminological influence is in the second (18 October 1820) of his open letters to the Spanish people (addressing “Spaniards!”), where he explicitly albeit incidentally names Priestley as his inspiration for the expression “the greatest happiness of the greatest number”: “a phrase for which, upwards of fifty years ago, I became indebted to a pamphlet of Dr. Priestley’s” (Bentham 1821, 24; reprinted in 1843, vol. 2, 288). This acknowledgment, though made in passing in a letter with entirely different concerns, is confirmed in detail in all three of the organically related original sources documenting Bentham’s perspective on the emergence and development of utilitarianism. Bentham’s “Article on Utilitarianism” (1829/ 1983), to begin with the most relevant source, notes the connection in both
27
28
Bentham’s derivation of “the greatest happiness principle” from “the greatest happiness of the greatest number” may reasonably be adduced as an example of his creative involvement in the development of utilitarian nomenclature, but his contribution in that instance is not so much the invention of a new term as it is the abbreviation of an expression already in use. The point, at any rate, is not that Bentham was not creative in any sense but that he was most creative in the development and application of what had already been created. Bentham’s experimentation with terminological alternatives to “the principle of utility” and “the greatest happiness principle” includes close variations with a parallel structure, such as “the felicity-maximization principle” (Bentham 1829/1983, 302; Bowring 1834, 321), as well as conceptual or ideological analogues, such as “eudaimonology” and “felicitarianism” (Bowring 1834, 320).
Classical Utilitarianism before John Stuart Mill
41
versions, the long one and the short one. The long version credits Priestley as follows: Between the years 1762 and 1769 came out a pamphlet of Dr. Priestley’s, written as usual with him currente calamo and without any precise method predetermined, but containing at the close of it, it is believed in the very last page, in so many words the phrase ‘the greatest happiness of the greatest number’, and this was stated in the character of a principle constituting not only a rational foundation, but the only rational foundation, of all enactments in legislation and all rules and precepts destined for the direction of human conduct in private life. bentham 1829/1983, 291
The mode of acknowledgment here indicates that Bentham values the solitary expression more than the accompanying analysis, as he credits Priestley without hesitation but with implicit surprise at his discovery as if the latter had stumbled upon a felicitous discovery without appreciating its importance. The very next page shows just how much Bentham himself appreciates the discovery and, in fact, regards it as a momentous revelation: Be this as it may, it was by that pamphlet and this phrase in it [“the greatest happiness of the greatest number”] that his [Bentham’s] principles on the subject of morality, public and private together, were determined. It was from that pamphlet and that page of it that he drew that phrase, the words and import of which have by his writings been so widely diffused over the civilized world. At sight of it he cried out as it were in an inward ecstacy [sic] like Archimedes on the discovery of the fundamental principle of Hydrostatics, Ευρηκα. bentham 1829/1983, 292
The systematic references to Bentham in the third person, as if the article had been written by someone else, reflect Bentham’s original motivation for the piece as an institutional manual for proponents of utilitarianism, particularly for philosophical radicals, thereby providing both factual information and talking points, all presented with anonymized authorship and suitable for duplication either in part or in whole. The short version of the article elaborates on Bentham’s debt to Priestley, supplying details absent in the long version:
42
Chapter 1
Epoch the sixth. Year 1768: date of Priestley’s tract entitled Essay on Government. In the concluding page of it, if memory is not deceitful, on the character of the only proper end of government, appears in italics the phrase ‘the greatest happiness of the greatest number’. Of this phrase, the good effect is the substituting to the equivocal word ‘utility’ the unequivocal phrase of which happiness is the principal and sole characteristic ingredient. In the change herein consists the whole of the improvement at this time made. bentham 1829/1983, 325–326
Here we get both the title and the publication date of the corresponding work, which the long version identifies only as a “pamphlet.” We also get a sense of why Bentham finds the contribution valuable, namely because of its emphasis on happiness, which is otherwise buried under talk of utility, especially in the absence of a deliberate effort to make the connection and explain the association between utility and happiness. Perronet Thompson’s “ ‘Greatest Happiness’ Principle” (1829), published shortly after the completion of Bentham’s manuscript (1829/1983) and based largely on the ideas presented and promoted there, if not on the exact words through which they are articulated, reverts to the generic designation of “pamphlet” in reference to the corresponding work of Priestley, which is identified by name and classified as a “tract” in the short version of Bentham’s article despite repeated references to it as a “pamphlet” in the long version: In the year 1768 appeared a pamphlet of Dr. Priestley’s, written, as was his custom, in a hasty manner, and with little precise method; but containing in one of its pages the express phrase ‘the greatest happiness of the greatest number.’ And this was represented as a principle containing the only rational foundation of rules for human conduct. perronet thompson 1829, 258
The reception and estimation of Priestley’s contribution, as described here by Perronet Thompson, is consistent with the original description in Bentham. Even the enthusiasm of Bentham comes across in almost exactly the same words in the delivery of Perronet Thompson: In the same year this pamphlet fell into the hands of Mr. Bentham at Oxford; he being at that time not quite twenty-one years of age. Like Archimedes on the discovery of the principle of hydrostatics, he exclaimed Ευρηκα, and from that page of that pamphlet, was drawn the
Classical Utilitarianism before John Stuart Mill
43
phrase [“the greatest happiness of the greatest number”], the import of which it has been the object of his subsequent writings to diffuse. perronet thompson 1829, 258–259
Bowring’s “History of the Greatest-Happiness Principle” (1834), following Bentham’s manuscript (1829/1983) even more closely and duplicating passages even more precisely than the piece by Perronet Thompson (1829), describes and quotes Bentham as identifying Priestley as his source and inspiration for the expression “the greatest happiness of the greatest number” (Bowring 1834, 298–300). The descriptive portion, narrated in Bowring’s voice, properly supplies the title and publication year for Priestley’s publication, making sure to summarize and explain why the new expression represented such exciting progress in utilitarian nomenclature: Dr Priestley published his Essay on Government in 1768. He there introduced, in italics, as the only reasonable and proper object of government, ‘the greatest happiness of the greatest number.’ It was a great improvement upon the word utility. It represented the principal end, the capital, the characteristic ingredient. It took possession, by a single phrase, of every thing that had hitherto been done. It went, in fact, beyond all notions that had preceded it. It exhibited not only happiness, but it made that happiness diffusive; it associated it with the majority, with the many. bowring 1834, 298
Drawing on autobiographical elements in the original manuscript of Bentham (1829/1983, 291–292), Bowring goes on in great detail over the next few pages (1834, 298–300) to relate the circumstances of exactly how Bentham came to be acquainted with the work of Priestley. The account culminates in the quoted portion of the presentation, where Bowring switches to the generic term “pamphlet” in reference to Priestley’s work, despite having already identified it by name. Here is how Bowring puts the matter in Bentham’s words: This year, 1768, was the latest of all the years in which this pamphlet could have come into my hands. Be this as it may, it was by that pamphlet, and this phrase in it [“the greatest happiness of the greatest number”], that my principles on the subject of morality, public and private together, were determined. It was from that pamphlet and that page of it, that I drew the phrase, the words and import of which have been so widely diffused over the civilized world. At the sight of it, I cried out, as it were in
44
Chapter 1
an inward ecstasy, like Archimedes on the discovery of the fundamental principle of hydrostatics, Ευρηκα. bowring 1834, 300
The apparent discrepancy, perhaps just a curiosity, in Bentham’s recollection of the source both as a “pamphlet” and as a “tract,” combined with Bowring’s reference to it as a “pamphlet,” particularly when quoting Bentham, despite both of them also invoking the work by its proper title and publication year, to say nothing of Perronet Thompson’s identification of it simply as a “pamphlet,” is not an obstacle to getting at the truth of the matter. A pamphlet can certainly serve as the medium for a tract, and any tract, in turn, can be published as a pamphlet. So, favoring one designation over the other is not really a problem. Nor does the persistent employment of either designation instead of the actual title present a difficulty, so long as the title is given at least once in the same context. One problem, however, is that the book is too long to be a pamphlet. The full title of Priestley’s work, cited in its short form both by Bentham (1829/ 1983, 325–326) and by Bowring (1834, 298), is An Essay on the First Principles of Government; and on the Nature of Political, Civil, and Religious Liberty (1768). Spanning three-hundred pages, it is not a pamphlet by any measure, certainly not in terms of length. The real discrepancy, though, is that the exact phrase, “the greatest happiness of the greatest number,” is nowhere to be found in the source cited, nor in any of Priestley’s other works, as scholars specializing in eighteenth-century philosophy have been noticing from the nineteenth century onward.29 The closest that Priestley ever comes to the phrase inspiring Bentham is in a couple of places in his Essay on the First Principles of Government: Let us then, my fellow citizens, consider the business of government with these enlarged views, and trace some of the fundamental principles of it, by an attention to what is most conducive to the happiness of mankind at present, and most favourable to the increase of this happiness in futurity. priestley 1768, 8–9; cf. 1771, 5
29
A good example of early scholarly interest in the origins of the phrase “the greatest happiness of the greatest number,” particularly with respect to where Bentham may have first encountered the expression, is available, among other places, in a series of brief discussion notes in the nineteenth century in various issues of Notes and Queries: F. Adams (1897); William Comyns Beaumont (1897); Bibliothecary (1879); Charles Edward Doble (1897); James Augustus Henry Murray (1897a; 1897b); T. Wilson (1897).
Classical Utilitarianism before John Stuart Mill
45
The good and happiness of the members, that is the majority of the members of any state, is the great standard by which every thing relating to that state must finally be determined. priestley 1768, 17; cf. 1771, 13
The absence of an exact match between the terms is not very important. Close analogues and comparable expressions, all in some way identifying happiness as the greatest good, were already prevalent in the literature. Any one of them could have inspired the formulation popularized as “the greatest happiness of the greatest number.” Such an experience could have prompted the same formulation by multiple authors on different occasions, all presumably inspired by inexact correlates not necessarily coming from the same source. In other words, the history of the expression need not have unfolded as an unbroken chain of verbatim transmission dating back to a single point of origin. The words of Priestley quoted above, for example, are expressive enough, even without further input or prodding, to lead a thoughtful reader to come up with, or at least to end up with, “the greatest happiness of the greatest number” as a respectable aim in moral, social, or political contexts. The notion itself could conceivably have occurred to more than one person independently of others. And articulating that notion in a forceful manner with an eloquent delivery through a memorable slogan could certainly have seemed like a good idea to them all, again independently of one another. Unfortunately, spontaneous creation in multiple sources can hardly be confirmed with any certainty. It is much easier to trace a straight line of historical usage in connection with a single formulation than to ascertain who all might have come up with the term independently. There can therefore be no uniquely valid explanation of the absence of an exact match between what Priestley actually said and what Bentham thought Priestley had said. One may be tempted to attribute the discrepancy to Bentham’s fading recollection of the actual source, having heard it at the youthful age of twenty, “not being at that time arrived at the age of twenty- one” (Bentham 1829/1983, 292; Bowring 1834, 299; cf. Perronet Thompson 1829, 259). Yet that would be inconsistent with such a “eureka” moment of discovery as Bentham reports almost immediately afterwards (Bentham 1829/1983, 292; Bowring 1834, 300; Perronet Thompson 1829, 259) with an obviously irrepressible enthusiasm that should have been sufficient to preserve his memory of the source. The absence of an exact match in Priestley explains why Bentham brings Beccaria into the fold, professing uncertainty regarding where he might have first encountered the expression, admitting that it need not have been Priestley
46
Chapter 1
(1768) after all, and speculating that it might instead have been Beccaria (1764), most likely in English translation (1767). His acknowledgment of Beccaria is not in any of Benham’s works recognizing Priestley, certainly not in any of the three sources cited above in that connection, nor in any of his other essays and treatises. Beccaria, in fact, is not mentioned at all in Bentham’s “Article on Utilitarianism” (1829/1983). Nor does he ever come up in Bowring’s “History of the Greatest-Happiness Principle” (1834) or in Perronet Thompson’s “ ‘Greatest Happiness’ Principle” (1829). But his influence is expressly acknowledged by Bentham elsewhere, mainly in memoirs and correspondence.30 A particularly useful source in that regard is a commonplace book kept by Bentham during the 1781–1785 period, which was published by Bowring in 1843 in the form of partial selections combined in a single volume with other miscellaneous writings, including conversations and correspondence as well as additional autobiographical details (Bentham 1843, vol. 10, 54, 142, 296).31 The clearest indication of the possible influence of Beccaria’s terminology on Bentham’s vocabulary is in a parenthetical insertion in one of the entries in this commonplace book: Priestley was the first (unless it was Beccaria) who taught my lips to pronounce this sacred truth:—That the greatest happiness of the greatest number is the foundation of morals and legislation. bentham 1843, vol. 10, 142
30
31
Relevant references to Beccaria in Bentham’s correspondence include at least one letter to Voltaire (1776) and another to John Foster (1778), both brought to light by Robert Shackleton (1972). The letters expressly acknowledge an intellectual debt to Beccaria, together with a comparable debt to Helvétius, though the debt is not expressly specified in either case as being terminological in nature: “In a letter to Voltaire, in November 1776, he writes: I have built solely on the foundation of utility, laid as it is by Helvétius. Beccaria has been lucerna pedibus, or if you please manibus, meis. In a letter to John Foster of April or May 1778 he again acknowledges Helvétius and Beccaria (ii.99)” (Shackleton 1972, 1465–1466). A commonplace book is a personal compendium of brief and random notes, typically for future reference and repeated consultation, in essence, a notebook or scrapbook for recording ideas and compiling information of interest to the author, a medium that was particularly popular in the nineteenth century, though fairly common throughout history. The one cited here is actually an indeterminate number of such notebooks, screened for material of interest by Bowring, who then released the corresponding miscellanea, under the heading “Extracts from Commonplace Book,” in dated clusters organized in accordance with subject matter. The 1781–1785 cluster constitutes part of the sixth chapter of the tenth volume of The Works of Jeremy Bentham (pp. 141–147), the first compilation of Bentham’s writings, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843 (Edinburgh: William Tait. London: William Simpkin and Richard Marshall).
Classical Utilitarianism before John Stuart Mill
47
The reference here to Beccaria must be to An Essay on Crimes and Punishments (London: John Almon, 1767), the first English translation of his Dei Delitti e delle Pene (Livorno: Marco Coltellini, 1764), as that represents the only opportunity Bentham would have had to observe the phrase in Beccaria in 1768, the benchmark year for Bentham’s momentous encounter with the terminology (Bentham 1829/1983, 292, 325; Bowring 1834, 298–300; Perronet Thompson 1829, 258): If we look into history we shall find, that laws, which are, or ought to be, conventions between men in a state of freedom, have been, for the most part, the work of the passions of a few, or the consequences of a fortuitous, or temporary necessity; not dictated by a cool examiner of human nature, who knew how to collect in one point, the actions of a multitude, and had this only end in view, the greatest happiness of the greatest number. Happy are those few nations, who have not waited, till the slow succession of human vicissitudes, should, from the extremity of evil, produce a transition to good; but, by prudent laws, have facilitated the progress from one to the other! And how great are the obligations due from mankind to that philosopher, who, from the obscurity of his closet, had the courage to scatter among the multitude, the seeds of useful truths, so long unfruitful! beccaria 1767, 2
Adding an editorial note to Bentham’s reference to Beccaria, on the same page as the one quoted above from Bentham’s commonplace book covering the 1781–1785 period, Bowring gives a more literal translation of the original phrase in Italian: The expression is used by Beccaria in the Introduction to his Essay on Crimes and Punishments, where he condemns the laws made by passion and ignorance as not having—questo punto in vista, La massima felicità divisa nel maggior mumero (this end in view,—The greatest happiness divided among the greatest number.)—The italics are the author’s. bentham 1843, vol. 10, 142, editorial note; both the italics and the comment about the italics are in the original
The fact that the only English translation available to Bentham in 1768, or at least the only one available to him in print at that time, uses the conventional “of” (as rendered in the Beccaria passage quoted above) instead of the literal “divided among” (as reported by Bowring) raises the question whether the
48
Chapter 1
initial acquaintance of Bentham and Bowring with the text of Beccaria was through an unofficial translation, or perhaps even through the original edition, with each one relying separately on his own ability, or on that of a helpful associate, to render some or all of the Italian into English.32 However one might be inclined to translate it, Beccaria’s formulation is much closer than Priestley’s, if not identical with Bentham’s. Yet no matter which version Bentham had actually seen first, neither one was the first overall to make it into print as the terminological origin of what was soon to become the favorite mantra of the intelligentsia, particularly of scholars engaged in moral, social, or political philosophy, including just about everyone writing on political economy at the time. The distinction of being the first in print belongs to Francis Hutcheson, who introduced the relevant terminology in An Inquiry into the Original of Our Ideas of Beauty and Virtue, published anonymously in 1725, more than forty years before either Priestley or Beccaria came out with anything that could possibly have inspired Bentham. The book is composed of two treatises, the first largely on beauty, and hence in the field of aesthetics, the second mainly on virtue, and thereby within the domain of ethics. Hutcheson uses the expression “the greatest happiness of the greatest number,” or rather a close variation as “the greatest happiness for the greatest numbers,” in section 3.8 of his second treatise, namely “An Inquiry Concerning the Original of Our Ideas of Virtue or Moral Good” (1725, 164): In comparing the moral Qualitys of Actions, in order to regulate our Election among various Actions propos’d, or to find which of them has the greatest moral Excellency, we are led by our moral Sense of Virtue thus to judge, that in equal Degrees of Happiness, expected to proceed from the Action, the Virtue is in proportion to the Number of Persons to whom the Happiness shall extend: And here the Dignity, or moral Importance of Persons, may compensate Numbers; and in equal Numbers, the Virtue is as the Quantity of the Happiness, or natural Good; or that the Virtue is in a compound Ratio of the Quantity of Good, and Number of Enjoyers: And
32
Bentham was fluent in French, boasted a working knowledge of Italian, Spanish, and German, and even had some acquaintance with Russian and Swedish (see Bentham 1843, vol. 10, 193). This is to say nothing of his command of Greek and Latin, as expected of any respectable scholar of the period, though Whewell discounts Bentham’s grasp of classical languages, together with his knowledge of philosophical classics (see Whewell 1852, 190; 1862, 205). This appears to be what is meant by the “ignorant contempt for the past” that Bentham is later accused of by Sidgwick (1877, 628).
Classical Utilitarianism before John Stuart Mill
49
in the same manner, the moral Evil, or Vice, is as the Degree of Misery, and Number of Sufferers; so that, that Action is best, which accomplishes the greatest Happiness for the greatest Numbers; and that, worst, which, in like manner, occasions Misery. hutcheson 1725, 163–164
He then elucidates the meaning of the expression as he proceeds to elaborate on the moral implications of the corresponding assessment: From the two last Observations, we may see what Actions our moral Sense would most recommend to our Election, as the most perfectly Virtuous: viz. such as appear to have the most universal unlimited Tendency to the greatest and most extensive Happiness of all the rational Agents, to whom our Influence can extend. hutcheson 1725, 165
Hutcheson’s formulation is an almost exact match with Bentham’s, the difference being entirely negligible and nearly unnoticeable. Yet Hutcheson is not the end of the line, especially if one is willing to ignore greater terminological variation so long as the meaning is preserved. Hutcheson himself makes no claim to having invented the phrase in question, which he uses only once, as illustrated in the first passage and elaborated in the second. But he also makes no reference to external sources, either with respect to phrasing or with respect to meaning. Regardless of how he came to formulate his version, the general notion extends further back in the history of philosophy. There is, in fact, some indication, at least through circumstantial evidence, that Hutcheson may have been inspired by others, if not expressly borrowing from them. One such source of inspiration, or of outright origination, depending on the interpretation, is Gottfried Wilhelm Leibniz, whose name comes up every so often in this context, perhaps more than it should, owing to dedicated historical scholarship by Joachim Hruschka (1991). Reviving the details of Leibniz’s contribution to the relevant literature, Hruschka demonstrates that the fundamental insight motivating the explication of the principle of utility as the greatest happiness principle, particularly with explicit reference to the greatest happiness of the greatest number, was articulated by Leibniz (1700) well before Hutcheson (1725), with at least one full generation of scholarship separating the two. Here is Hruschka’s translation of what Leibniz says in Latin:
50
Chapter 1
To act in accordance with supreme reason is to act in such a manner that the greatest quantity of good available is obtained for the greatest multitude possible and that as much felicity is diffused as the reason of things can bear. leibniz 1700, 378, translated by hruschka 1991, 166
The passage is from an anonymous review by Leibniz (1700) of a pamphlet originating in a doctoral dissertation by Samuel von Cocceji (1699) at the University of Frankfurt an der Oder.33 The review is in the form of critical observations on the thesis that God is the only true and sufficient source of natural law, where the remark of Leibniz concerning the maximization of good and the distribution of happiness constitutes a tangential point, asserted but not defended, an obiter dictum, as Hruschka puts it (1991, 168, 169, 170). Although the passage is effectively a casual remark made in passing in the twelfth observation in the “Observationes” of Leibniz (1700, 378), Hruschka (1991, 172–173) locates the same idea expressed in similar terms in a slightly earlier paper by Leibniz, thereby suggesting that the resulting dictum was the culmination of serious reflection on the subject matter. The combination of rigorous research and plausible inferences in Hruschka’s investigation places Leibniz ahead of Hutcheson by a quarter of a century. But ahead in what? An isolated remark, even one of profound insight grounded in prior reflection, hardly establishes Leibniz as the fountainhead of utilitarianism. Nor does Hruschka claim that it does. His claim is merely that the contribution of Leibniz was a milestone in the development of utilitarianism in general and a direct resource for Hutcheson in particular. Yet this relatively modest claim has somehow turned into a scholarly magnet for exaggeration by those invoking Hruschka’s assessment of the historical and philosophical relevance of Leibniz, who sometimes ends up in first place, not just ahead of Hutcheson but ahead of everyone else
33
The pamphlet comprising Samuel von Cocceji’s dissertation—defended with his father Heinrich von Cocceji (a professor of law at the same school) presiding over the evaluation committee—was published as Disputatio Juridica Inauguralis de Principio Juris Naturalis Unico, Vero, et Adaequato [Inaugural Juridical Disputation on the Only True and Adequate Starting Point of Natural Law] in Frankfurt an der Oder: Christoph Andreas Zeitler, 1699. The review of Leibniz appeared in print the following year as “Observationes de Principio Juris” in a monthly periodical dedicated to recent scholarship and emerging trends in philosophy, including abstracts and excerpts from new and notable publications, together with commentaries on selected works: Monathlicher Auszug aus Allerhand Neu-Herausgegebenen/Nützlichen und Artigen Büchern 4 (July 1700): 371–382.
Classical Utilitarianism before John Stuart Mill
51
as well.34 There can be very little doubt that Leibniz precedes Hutcheson but even less that he does not precede absolutely everyone. Leibniz may deserve some recognition for anticipating the core principle of classical utilitarianism, but he is not the singular source or original inspiration giving rise to all subsequent efforts. What Leibniz succeeds in capturing of the spirit of utilitarianism is not even articulated in the standard formulation as “the greatest happiness of the greatest number.” The terminology of Leibniz is certainly not as close to that of Hutcheson, especially not as a catchphrase with the same aphoristic flavor, as the terminology of Hutcheson is to that of Beccaria or Bentham. Leibniz does, of course, say more or less the same thing in different words, which establishes a philosophical precedence, but the terminological origin still seems to rest with Hutcheson, or at least not with Leibniz. This is not to say that the wording is more important than the meaning, or even that the wording is just as important as the meaning, but simply that the wording of Leibniz is not in the same line of succession as the rest of the examples considered above. Hruschka readily admits that the match is in ideology and not in terminology. That is what he is attempting to demonstrate anyway. He finds it sufficient to establish meaningful anticipation, where Leibniz predates Hutcheson in recognizing the importance of both maximization and distribution as essential components in any moral insight grounded in happiness or utility as the standard of right and wrong, or rather, as the standard of action in accordance with supreme reason, and in combining those components as the correlative dimensions of a single principle, or “dictum,” as Hruschka has it. That is a perfectly reasonable assessment in itself. Yet it opens the door to comparable precedents in the history of philosophy, where others may have anticipated Leibniz with the same standards of similarity and clarity as Leibniz seems to have anticipated Hutcheson. Leibniz is not, after all, a utilitarian in the standard sense. Any utilitarian tendencies he exhibits are inherently metaphysical and specifically theological in origin. The maximization and distribution of happiness become joint standards of right and wrong only in
34
See, for example, Wolfgang Leidhold (2004), whose introduction to a critical edition of Hutcheson’s An Inquiry into the Original of Our Ideas of Beauty and Virtue (1725) identifies Leibniz as the first person to employ an identifiable version of the expression “the greatest happiness of the greatest number.” Citing Hruschka (1991), he asserts outright that “the formula was first used by Gottfried Wilhelm Leibniz” (Leidhold 2004, x, n. 2). Note that “the formula,” as it were, is not quite the same in Leibniz as it is in Hutcheson, and that, even if it were the same, it would establish only that Hutcheson was not the first rather than that Leibniz was.
52
Chapter 1
compliance with the dictates of reason pursuant to the divine perfection of the world. Recall that “God is the monarch of the most perfect republic composed of all the spirits, and the happiness of this city of God is his principle purpose” (Leibniz 1686, §36 [=1908, 60]). While it may not be fair to Leibniz, nor even accurate in itself, to discount the strength of the correlation between Leibniz and Hutcheson, it also does not seem altogether reasonable to position the work of Hutcheson as entirely derivative relative that of Leibniz, which is how Hruschka puts the matter in evaluating the evidence. For all intents and purposes, Hruschka (1991) ends up portraying Hutcheson as a habitual plagiarist. He does not make a direct accusation in that regard, apparently doing little more than reporting research findings, but the facts with which he leaves readers explicitly include two counts of plagiarism: “That Hutcheson adopted Leibniz’s principle without any reference to his source is not surprising. He did the same with the universalization principle which he took from Johann Balthasar Wernher and Jean Barbeyrac” (Hruschka 1991, 172, n. 34).35 The possibility of an influence cannot be denied, but the reality of it need not be admitted without question. It is particularly difficult to accept that Leibniz was perfectly capable of anticipating the fundamental principle of utilitarianism before any of the actual utilitarians, and capable furthermore of articulating that principle competently in the process of fleshing out an entirely different idea in an altogether different area, while Hutcheson was not capable of comparable innovation in his own field of expertise without direct inspiration by an obscure remark made in passing by Leibniz. Even if Hutcheson had somehow come across the review in question, he would not have been likely to recognize it as the work of Leibniz, whose authorship was not made public until more than a full decade after the publication of Hutcheson’s Inquiry.36 Any match would have still been plagiarism, if Hutcheson had indeed adopted the original without acknowledgment, but it would have then been the plagiarism of a peripheral remark in a classical language in an anonymous contribution to a foreign periodical. The isolated remark was admittedly relevant to what Hutcheson was doing, despite being tangential in its original context, but 35
36
Details of the charge concerning the universalization principle are available in Hruschka (1992) and Raynor (1987), both sources being referenced in Hruschka (1991), specifically in the remainder of the footnote quoted in part in the main text above (Hruschka 1991, 172, n. 34). The mention of Hruschka (1992) in Hruschka (1991) is naturally in the form of a reference to an article that was then forthcoming. The anonymity of Leibniz was evidently preserved throughout the period between the publication of his review (1700) and the appearance of Hutcheson’s Inquiry (1725). According to Hruschka (1991, 169, n. 20), the authorship of Leibniz did not become public
Classical Utilitarianism before John Stuart Mill
53
it was never very likely to meet his eyes, or to cross his desk, thus making it a remote prospect either for research or for plagiarism. 1.4
Bentham’s Debt to Predecessors and Contemporaries
To extend the survey in the preceding section, particularly the considerations at the end, we can note that Bentham does not seem to owe anything either to Hutcheson or to Leibniz in connection with utilitarian terminology and philosophy. For one thing, he mentions neither one as a source worth noting; for another, he does mention various others. While the absence of acknowledgment is not always indicative of the absence of inspiration, it is indeed significant in this case, because we know that Bentham was both prepared and inclined to acknowledge his sources (see section 1.3). To start with the more obvious of the two, Hutcheson’s articulation of utilitarian ideology in terms of “the greatest happiness for the greatest numbers” is so close to Bentham’s formulation as “the greatest happiness of the greatest number” that Bentham, had he seen it in print, would have been sure to cite the former when introducing the latter. He must have been unaware of the passage altogether. He would have otherwise acknowledged the inspiration, lest he be accused of stealing the idea. What makes this a reasonable conclusion is common sense more than anything else. The point is not that Bentham was an honest man who would have never tried to pass off as his own anything that was not his own. He may well have been just that honest. The real point, however, is that no serious scholar would knowingly proceed with such a close adaptation of such a distinctive expression without full acknowledgment, at least for fear of being exposed as a plagiarist by anyone acquainted with both the original and the adaptation. Given that Bentham was in the habit of openly and eagerly expressing his intellectual debts, persistently so to Priestley, despite being wrong about it, and occasionally even to Beccaria, despite being hazy about it, he certainly would have acknowledged the same debt in connection with Hutcheson, had he believed he actually had any such debt to Hutcheson. As for Leibniz, even if he had somehow influenced Hutcheson, it is doubtful that he ever had any effect on Bentham, who does not even acknowledge Hutcheson as a source, even though the latter’s formulation is much closer to his own, and in the same language to boot. The difference is not just in the knowledge until 1737, which places it twelve years after the emergence of the first edition of Hutcheson’s Inquiry (1725) and only one year before the release of the same work in its fourth edition (1738), the last in Hutcheson’s lifetime.
54
Chapter 1
accessibility or availability of the works in question but also in their relevance to Bentham’s interests. Bentham simply had no good reason to follow Leibniz in the course of his own development and promotion of utilitarianism. Recall the theological foundation of hedonistic inclinations and utilitarian insights in Leibniz. Any and all emphasis on happiness comes directly from God: “The happy and prosperous condition of his empire which consists in the greatest possible felicity of its inhabitants, becomes supreme among his laws” (Leibniz 1686, §36 [=1908, 62]). But all that attention to happiness is, in fact, nothing more than a corollary of the divine perfection of the world: And if the dominant principle in the existence of the physical world is the decree to give it the greatest possible perfection, the primary purpose in the moral world or in the city of God which constitutes the noblest part of the universe ought to be to extend the greatest happiness possible. leibniz 1686, §36 (=1908, 62)
Leibniz could hardly have been a benchmark for Bentham, whose notion of happiness is entirely secular and decidedly naturalistic, defined as it is in terms of pleasure and pain, pure and simple. This is because Bentham’s perspective is shaped by sheer quantity in either case, regardless of the kind, or the source, of the pleasure and pain involved, and without any external qualification or justification, divine or otherwise. As summarized succinctly by Sidgwick, Bentham’s conception of good is “purged of all mystical elements, and reduced to the positive, palpable, empirical, definitely quantitative notion of ‘maximum balance of pleasure over pain’ ” (Sidgwick 1877, 638). Direct observations and general impressions concerning the absence of a significant connection between Bentham and either Hutcheson or Leibniz are supported by external evidence as well. The point is not just that Bentham (1829/1983) himself says nothing about a debt to Hutcheson or Leibniz but also that neither his contemporaries nor subsequent generations have ever reported a relevant connection in their surveys and discussions of utilitarian terminology prior to Bentham. Those who have looked into the matter include, in addition to John Bowring (1834) and Thomas Perronet Thompson (1829), James Reddie (1840), Henry Sidgwick (1877), and William Whewell (1852/1862), a fair mix of people, some close to Bentham, some opposed to Bentham. Bowring (1834) and Perronet Thompson (1829) do not even mention Hutcheson or Leibniz in their respective studies of the emergence and development of utilitarianism. Following Bentham (1829/1983), they both trace the history of the greatest happiness principle all the way back to Horace, naming Phaedrus and Aristotle among those showing early insight without
Classical Utilitarianism before John Stuart Mill
55
full awareness or proper articulation, and associating Locke and Hume, particularly the latter, with the germination of utilitarianism as a philosophical position subsequently developed through the efforts of Hartley, Priestley, and Helvétius, the last of which they identify as a particularly important inspiration for Bentham (Bowring 1834, 287–296; Perronet Thompson 1829, 258–261).37 Paley’s name also comes up, but only in Bowring (1834, 310–311; cf. Bentham 1829/1983, 328), and even there, only in passing and strictly in an unfavorable light, thus minimizing his position in the development of utilitarianism. There is no reference, through all of this, either to Hutcheson or to Leibniz. Reddie (1840) likewise finds no connection between Bentham and either Hutcheson or Leibniz. Himself a Scottish advocate, Reddie was an adherent of Bentham and a contemporary of James Mill. He was also the author of Inquiries Elementary and Historical in the Science of Law (1840). It is there that he contemplates the origins of the greatest happiness principle (Reddie 1840, 12–16, 68–72). Although he invokes and discusses various formulations of the principle predating Bentham’s version, he does not mention Hutcheson or Leibniz in that regard. As a matter of fact, with respect to Hutcheson, it is not just that Reddie does not consider him among the inventors or supporters of the greatest happiness principle in any formulation but more so that he places him entirely outside that camp: Thus, in modern times, the philosophers, who found morality on the sentiments of approbation or disapprobation excited in the mind of man by the contemplation of human actions, without any immediate reference to their consequences, whether they ascribe these sentiments to the intellectual faculties, or practical reason, such as Cudworth, Clarke, and Kant; or to conscience, or a separate moral faculty, such as Cumberland, Hutcheson, Shaftesbury, Reid, and Dugald Stewart, appear to hold, that the general utility of actions cannot, in any point of view, be admitted as a criterion of their morality, because the simultaneous or immediate perception of that utility does not form an essential element, or even a palpable constituent part of our moral sentiments. reddie 1840, 12
37
Bowring (1834, 304–310) and Perronet Thompson (1829, 259), both taking their cue from Bentham (1829/1983, 293, 298–299, 314–316, 322), find more to criticize than to celebrate in Locke. The intellectual legacy of Locke is acknowledged but not in any detail (see Bentham 1829/1983, 298; Bowring 1834, 291). This is because Locke’s most valuable contributions are not considered relevant in the immediate context, while his most relevant contributions are not considered valuable in any context—or even acceptable, tolerable, or reasonable in any way.
56
Chapter 1
Mentioning Hutcheson only once, and Leibniz not at all, Reddie (1840, 70–71) names Hume alone as the inspiration for Bentham’s principle of utility, though he does not claim that Hume was, in fact, the only one.38 Sidgwick (1877, 638), for his own part, considers Helvétius rather than Hume to have had the greatest influence on Bentham, as discussed further below, while he says nothing about either Hutcheson or Leibniz in that respect. Sidgwick’s inquiry into the matter is in an article devoted entirely to Bentham, “Bentham and Benthamism in Politics and Ethics” (1877), cited earlier in the first section of the present chapter. His primary aim is to demonstrate the enduring relevance of Bentham and Benthamism in England close to half a century after Bentham’s death. While he is no advocate of Benthamism, on the contrary, representing quite an adversary, his assessment is dispassionate and his coverage comprehensive. As a utilitarian himself, Sidgwick is quick to recognize Bentham’s achievements as well as his shortcomings. His focus is not restricted to individual works, nor even to specific ideas, as he freely comments on whatever he finds good or bad in Bentham, pausing even to note “his exaggerated reliance on his own method, his ignorant contempt for the past, and his intolerant misinterpretation of all that opposed him in the present” (Sidgwick 1877, 628). For all his opposition, though, Sidgwick finds no fault in Bentham with respect to his acknowledgment of the contributions of his predecessors and contemporaries. Whewell (1852/1862) is the only commentator, among those considered here, to criticize Bentham for an excessive inclination to claim or feign ownership of the principle of utility. Like Sidgwick, Whewell is opposed to Bentham and Benthamism. He devotes considerable space and attention to both in his Lectures on the History of Moral Philosophy in England (1852/1862).39 Despite extensive discussion, however, he makes no reference to Hutcheson, except once in passing, and none at all to Leibniz, thus indicating that he found neither one to have inspired Bentham in any meaningful way.40 He specifically 38
39
40
Reddie (1840, 70–71) names Hume as Bentham’s inspiration for the principle of utility, which is accurate insofar as Hume is indeed one of Bentham’s expressly acknowledged sources of inspiration in that regard, but Reddie’s reference (“Bentham’s Works, vol. i. p. 268”) is erroneous, as it corresponds to Bentham’s appeal to Hume as an opponent of social contract theory rather than as a proponent of utilitarianism. Whewell’s coverage is thorough enough to make room for all manner of criticism, including the inadequacy of Bentham’s classical training, contending that he “scarcely ever made any reference to Greek or Latin without showing some extraordinary ignorance” (1852, 190; 1862, 205). Whewell never mentions Hutcheson when discussing Bentham (Whewell 1852, 188– 265; 1862, 203–276), except once in passing (Whewell 1852, 205; 1862, 219). Nor does he
Classical Utilitarianism before John Stuart Mill
57
names Priestley as an inspiration for Bentham (Whewell 1852, 190; 1862, 205), though he rightly indicates that the relevant sources of inspiration are distributed widely among the scholarly contributions animating and reflecting the spirit of the times, with Bentham generally inclined to give credit where it is due: It is evident that there was in the general current of literature and thought at that time a set towards such doctrines and such expressions; and indeed Bentham himself pointed out other previous writers in whom expressions and thoughts very similar occur. whewell 1852, 190; 1862, 205
Despite his recognition of Bentham’s tendency to credit his predecessors and contemporaries with various contributions to utilitarian thought, both in terms of terminology and with respect to philosophy, Whewell curiously accuses Bentham of pretending to have come up with the principle of utility himself, or more specifically, of “constantly” talking of himself as “the discoverer of the principle” (1852, 190; 1862, 205). His main objection, however, is not that Bentham appropriates the ideas of others but that he tends to dismiss them out of hand. More precisely, he finds Bentham adequately prepared to acknowledge the contributions of others and their influence upon his own work but too ready to dismiss the views of rivals without due consideration, in fact, treating them with the opposite attitude: gross oversimplification and outright misrepresentation (Whewell 1852, 202–207; 1862, 216–221). It is in this connection that Hutcheson’s name comes up, along with several others, plus an allusion to a multitude of people presumably too numerous to specify but apparently possible to exemplify: “Shaftesbury, Hutcheson, Hume, Beattie, Price, Clarke, Wollaston, and many others” (Whewell 1852, 205; 1862, 219). These are the alleged victims of Bentham with respect to what Whewell describes as “the extravagant unfairness to adversaries which was habitual in him” (1852, 202; 1862, 216). The assessment of his contemporaries and successors confirms that Bentham was generally found to demonstrate the intellectual integrity commonly expected of all scholars in the acknowledgment of relevant sources. Even Whewell, who is critical of Bentham’s approach to others, faults him not for representing their ideas as his own but for rejecting theirs all too
mention Bentham when discussing Hutcheson (Whewell 1852, 92–97; 1862, 114–119). As for Leibniz, Whewell does not mention him at all in any capacity in any context.
58
Chapter 1
easily. None of the most relevant commentators finds any association between Bentham and either Hutcheson or Leibniz, thus enhancing the evidentiary value of Bentham’s silence in that regard and undermining any suspicions or speculations that either one of them might have influenced his approach to utilitarianism in any way or degree. The absence of scholarly, testimonial, and autobiographical references by Bentham to a source thought to have influenced his work may not seem like much to go by in denying a possible or actual connection. Yet Bentham’s silence becomes far more relevant, and that much more persuasive, in the light of his personal experience with plagiarism as documented by his response to pressing concerns brought to his attention in that regard. The full details are available in the correspondence between Bentham and his friend George Wilson, where the discussion reveals a sensitivity to plagiarism as a serious issue to be avoided both as a perpetrator and as a victim. Wilson broaches the subject in a letter drafted in the first days of the autumn of 1786, expressing great concern that a recent book by William Paley, The Principles of Moral and Political Philosophy (1785), bears significant similarities to An Introduction to the Principles of Morals and Legislation (1789), Bentham’s magnum opus, printed and circulated privately in 1780 and published in 1789: There is a Mr Paley, a parson and archdeacon of Carlisle, who has written a book called Principles of Moral and Political Philosophy, in quarto, and it has gone through two editions, with prodigious applause. It is founded entirely on utility, or, as he chooses to call it, the will of God, as declared by expediency, to which he adds, as a supplement, the revealed will of God. But notwithstanding this, and some weak places, particularly as to oaths and subscriptions, where he is hampered by his profession and his past conduct, it is a capital book, and by much the best that has been written on the subject in this country. Almost everything he says about morals, government, and our own constitution, is sound, practical, and free from commonplace. He has got many of your notions about punishment, which I always thought the most important of your discoveries; and I could almost suspect, if it were possible, that he had read your introduction; and I very much fear, that, if you ever do publish on these subjects, you may be charged with stealing from him what you have honestly invented with the sweat of your own brow. wilson 1786, 163–164, excerpt from letter to Bentham, dated 24 September 1786
The “introduction” Wilson invokes here is not the introductory chapter of the book but the book itself, namely the text of what later became An Introduction
Classical Utilitarianism before John Stuart Mill
59
to the Principles of Morals and Legislation, which was originally intended as part of a standalone introduction to a comprehensive penal code in multiple volumes.41 Wilson’s thoughts evidently run immediately to the possibility of Bentham being accused of plagiarism upon the publication of the material then in private circulation, even though Bentham’s work was printed five years before Paley’s book was published, which is why he wonders, at the same time, whether Paley might have somehow become acquainted with Bentham’s views while still articulating his own. The latter possibility, however, is no more likely than the former, as Paley’s book (1785) originates in his lectures at Cambridge, dating back at least ten years, thus placing its origin five years before the printing of Bentham’s manuscript (1780). Bentham, in turn, seems untroubled by either possibility, as evidenced both by his response to Wilson and by the lack of any particular urgency on his part to publish his own book, which eventually came out four years after Paley’s. His response in the early winter of the same year reveals no apprehension or precaution in reaction to the developments: I had ordered horses for England, to take triumphant possession of the throne of Legislation, but finding it full of Mr Paley, I ordered them back into the stable. Since then, I have been tormenting myself to no purpose, to find out some blind alley in the career of fame, which Mr Paley’s magnanimity may have disdained. bentham 1786, 165, excerpt from letter to Wilson, dated 19–30 December 1786
The reason why Bentham’s response shows no indignation over the reported coincidence between Paley’s views and his own is not that he does not take plagiarism seriously but that he does not take the coincidence as plagiarism. The printed version of Bentham’s manuscript was close enough in content to the published version that its publication could have been carried out immediately upon Wilson’s warning, or shortly afterwards anyway, had Bentham felt at risk either as a potential victim of plagiarism or as a future suspect of it. His reaction suggests that he found the coincidence natural, as might be expected in any conceptual breakthrough or philosophical trend spearheaded simultaneously by a multitude of scholars working independently as pioneers in the 41
Bentham’s response to George Wilson shows that Bentham himself employed different working titles, where “introduction” would indeed have served as the most reasonable and convenient shorthand, especially during a period when the absence of a definitive title inspired expressions such as “my Introduction to a Penal Code, alias, Principles of Legislation, alias, I don’t know what besides” (Bentham 1786, 165, excerpt from letter to Wilson, dated 19–30 December 1786).
60
Chapter 1
relevant field. Bentham does not seem to have suspected illicit prepublication acquaintance with his work by Paley, nor to have worried that he himself might be suspected of copying the work of Paley. Yet his early lack of concern with competition from Paley, as evidenced in his correspondence on the matter (1786), seems to have gradually grown into a curious and wanton denigration of contributions by Paley, as witnessed in Bentham’s “Article on Utilitarianism” (1829/1983): In the year 1785 came out for the first time the work of the Reverend Dr William Paley entitled Elements of Moral and Political Philosophy. But, in the catalogue of the works constitutive of so many epochs in the history of the principle in question, this work has not been inserted, because though, to wit by the name of the principle of utility, mention is made of this same principle, yet, as to any indication made of the relation borne by the idea attached to it to the ideas attached to the words ‘pain’ and ‘pleasure’, no such elucidation does it give. The eyes for which the work was designed were those of the youth belonging to the University of Cambridge in one of the Colleges of which he was at that time Tutor: in that meridian eyes were not strong enough, nor was it his desire that they should be strong enough, to endure true lights in such a field. In black and white, professed and self-declared advocate of insincerity and subornation of insincerity in the shape of subscription to articles—over a bottle, self-avowed lover of corruption, rich enough to keep an equipage but not ‘to keep a conscience’—after publishing editions upon editions of that same work and filling both universities with them to saturation, he departed this life in the year 1805, knowing better all that time [than] to know any thing of the works in and by which that of which mention has here been made had even then been done towards putting to its use that all-beneficent principle, or to know that any such person as the author was in existence. bentham 1829/1983, 328; original interpolation
The sentiments expressed here by Bentham (1829/1983) are fully echoed in the parallel treatment by Bowring (1834), where the part about Paley opens with an even more incisive focus on Bentham’s reservations, while the remainder includes the same material presented in the same tone: It was in 1785 that Paley published his Elements of Moral and Political Philosophy. He mentions the principle of utility, but seems to have no idea
Classical Utilitarianism before John Stuart Mill
61
of its bearing upon happiness. And if he had any such idea, he was the last man to give expression to it. bowring 1834, 310
Any real insight in Bentham’s and Bowring’s assessments of Paley’s work is obscured by vitriol, as they both offer scathing criticism, actually the very same diatribe, as if there were no redeeming quality either to the man or to his work. The reason for Bentham’s early lack of concern with competition from Paley was probably his preoccupation with the more serious threat of competition from the opponents of utilitarianism, leaving him little time or cause to worry about someone defending the same position as himself albeit in a different way, from a different perspective, with a different agenda. He very likely saw Paley more as an ally he did not need than as an enemy he could not beat. That seems to have been the full extent of his reception until later when he became more familiar with the participation of Paley in moral and political discourse from a utilitarian perspective with a theological foundation. Paley then seems to have become a corroborating example of what Whewell (1852, 190–191, 197–201, 202; 1862, 205, 212–215, 216) describes as unfair treatment of adversaries at the hands of Bentham. Wilson does not seem to have ever been as comfortable as Bentham was early on, refusing from the beginning to let go of his suspicions regarding Paley’s early acquaintance with Bentham’s even earlier work, as is evident in his continuing concerns expressed two years after the original discussion. He brings up the same subject in a letter to Bentham in the autumn of 1788: I have often been tempted to think that Paley had either seen your Introduction, or conversed with somebody that was intimate with you. There are many things in his book so like you, and so out of the common road, that they cannot be the production of the same person who wrote other things in the same book which are really puerile. wilson 1788, 195, excerpt from letter to Bentham, dated 30 November 178842
42
The similarity, and hence any cause for concern, seems to be at its greatest where Paley turns to an examination of “utility” and to “the consideration of general consequences” in chapters 6 and 8, respectively, of the second book of The Principles of Moral and Political Philosophy (Paley 1785), particularly in comparison with chapter 12 of An Introduction to the Principles of Morals and Legislation (Bentham 1789). See the editorial note on p. 195, that is, the original footnote to the passage quoted here as “Wilson 1788,” which comes from the first volume of The Works of Jeremy Bentham (Bentham 1843).
62
Chapter 1
It is interesting to note how, in the estimation of Wilson, Paley’s work goes from being a “capital book” representing “the best that has been written on the subject” (1786, 163–164) to one that stands out more with its puerile ideas (1788, 195). However that may be, Bentham’s book was published in 1789, shortly after receipt of Wilson’s second letter, making it tempting to think that he was finally convinced by Wilson. While this is certainly plausible, or at least conceivable, it is also possible that the book, already nine years in the waiting after the original private printing, was about to be published anyway. Bentham himself mentions in the preface that the circulation of the printed version kept growing over time as the manuscript was transmitted from its intended recipients in Bentham’s inner circle to outsiders gaining access through loans, inheritance, and other common transactions and interactions. This still does not confirm, however, that he was worried about plagiarism, whether by Paley or by anyone else, given that the manuscript had a formal date of printing (1780) that could easily be established as predating the publication of Paley’s book (1785). What all this shows, more than anything, is an attitude. The discussion between Wilson and Bentham is so clear and lasts so long, owing to Wilson’s persistence, that we can rest assured that they both considered plagiarism a serious matter. Bentham would surely have acknowledged any debts he may have had to Paley, Hutcheson, or Leibniz, or to anyone else, just as he had done with Priestley, Beccaria, and Hume, among numerous others mentioned in comparable roles of influence. While the discussion so far has been anchored to terminology, largely in connection with the expression “the greatest happiness of the greatest number” and partly in reference to the title “the principle of utility,” Bentham was also open to acknowledging where he got his ideas, independently of how they were expressed. Even Priestley and Beccaria, to whom he felt indebted mainly in terminological matters, also had an obvious influence on him in philosophical terms, as often acknowledged by Bentham himself. Indirect evidence of this is already available in his erroneous recognition of Priestley as his source for the expression “the greatest happiness of the greatest number,” nowhere to be found in the published works of Priestley, which demonstrates that Bentham felt Priestley’s philosophical presence strongly enough to imagine the terminological match in question. The same sort of general influence is true of Beccaria as well. Yet neither Bentham’s known sources nor his express gratitude is limited to a couple of chance encounters with the random works of predecessors and contemporaries. Hume, Hartley, and Helvétius stand out among others as having had a particularly profound influence on Bentham, who regarded their contributions as formative milestones in the history of utilitarianism.
Classical Utilitarianism before John Stuart Mill
63
A brief look at Bentham’s estimation of each will provide the proper orientation for the remainder of the discussion. Hume (1741–1742) occupies a prominent place in Bentham’s (1829/1983) reflections on the history of utilitarianism, including the parallel accounts by Bowring (1834) and Perronet Thompson (1829), where the most positive references to his influence can best be summarized as the establishment of the notion of utility as the foundation of moral discourse and the introduction of the term “the principle of utility” as the first formal and systematic reference to the grounding doctrine of utilitarian philosophy.43 Hartley (1749) enters into the spotlight of historical development for explicating happiness in terms of pleasure and pain, accompanied by a detailed classification of pleasures and pains in accordance with type, while demonstrating keen insight into their various effects on rational choice, intentional action, and human behavior, but he also receives some critical attention for failing to appreciate the full implications of such connections for moral agency, allegedly remaining oblivious even to the presumably obvious association between happiness and utility.44 Helvétius (1758) gets credit specifically, emphatically, and enthusiastically for elucidating utility in terms of happiness, and happiness in terms of pleasure and pain, thereby advancing utilitarian theory further than Hartley by both drawing out and fleshing out the full implications of the constituents of sentience for moral discourse, though without producing as comprehensive a classification of pleasures and pains as Hartley.45 Bentham held Hume and Helvétius in especially high esteem as contributors to utilitarianism. Hume typically comes up together with Helvétius, and Helvétius together with Hartley, in each case with the works of the latter as improvements upon the contributions of the former. Even so, Bentham clearly favors Hume and Helvétius over Hartley. He credits Hume both as a methodological pioneer and as his personal source in terminological matters, while reserving an even more special place for Helvétius, particularly as an inspiration for his basic orientation in utilitarianism. To appreciate the importance Bentham attaches to Hume, despite recognizing serious shortcomings even in his most valuable contributions, we need only note the precedence he assigns to him in the history of utilitarianism. He 43 44 45
Hume’s role in the development of utilitarianism is covered in Bentham (1829/1983, 289– 290, 292, 299, 322–324), Bowring (1834, 291–295), and Perronet Thompson (1829, 258). Hartley’s impact on utilitarian theory is discussed in Bentham (1829/1983, 290–291, 324), Bowring (1834, 295), and Perronet Thompson (1829, 258). Helvétius’s contributions to utilitarianism are sketched in Bentham (1829/1983, 290–291, 299, 325, 327), Bowring (1834, 295–297, 314), and Perronet Thompson (1829, 258).
64
Chapter 1
begins his “Article on Utilitarianism” (1829/1983) with Hume as the opening reference. To be more specific, after introducing the long version of the article as a historical survey of utilitarian terminology, Bentham immediately credits Hume with inventing the title, “the principle of utility,” or at least with being the first to use it in print: In and by means of the locution of phrase ‘principle of utility’, the signification of the word ‘utility’, with its conjugates, as the logicians say, ‘Utilitarian’ and ‘Utilitarianism’, has on this occasion amongst others attracted so much of public attention and become an object of such importance that a short history of its adventures, so to speak, may perhaps be not altogether unacceptable to our readers. The first work in which it ever made its appearance in the character of a subject of discussion was the work entitled Essays, by David Hume, date of the first edition, as far as recollection serves, the year 1742 or thereabouts. In that work it is spoken of as the name of a principle which might be considered as the foundation or corner-stone of one of the systems of morals at that time known and embraced by a philosophical sect, the moral sense being the denomination of a different system and that in truth a very widely different one. bentham 1829/1983, 289–290
He then goes on to acknowledge exactly where he first saw and adopted Hume’s term “the principle of utility” for use in his own work, which is as early as his first publication, A Fragment on Government (Bentham 1776): In the year 1776 came out his [Bentham’s] first publication entitled A Fragment on Government. On this occasion borrowing the phrase [“the principle of utility”] from David Hume, regarding himself as amply warranted in making use of it by a philosopher the most eminent of his day, the idea of happiness being in his mind constantly connected with that of utility, and not suspecting that it could fail of being so in any other, this was the denomination he on that occasion employed, that phrase [“the greatest happiness of the greatest number”] of Priestley’s, by which so strong an impression had been made upon him, not having presented itself to his view as having ever been employed in giving denomination to a principle. bentham 1829/1983, 292
Classical Utilitarianism before John Stuart Mill
65
He also makes sure to repeat the same acknowledgment in summary form in the short version of the article (Bentham 1829/1983): Epoch the third. Year 1742: year in which David Hume’s Essays made their first appearance. Contained in this year was the first moment at which it came out in the garb and with the style and note of a principle. bentham 1829/1983, 322
Bentham’s acknowledgment in both places, not to mention various others, comes with the explicit caveat that Hume’s role in the development of utilitarianism was one of foundational insight rather than comprehensive formulation, or even clear and complete anticipation. He is careful to recognize the limitations of Hume’s utilitarianism, including its overall vagueness as well as its particular failure to associate utility with happiness. His dual focus on Hume’s contributions and shortcomings finds its way into the two related accounts grounded in his “Article on Utilitarianism” (Bentham 1829/1983). The first one, the editorial by Perronet Thompson (1829), succinctly captures both aspects of Bentham’s estimation of Hume: The first time the phrase of ‘the principle of utility’ was brought decidedly into notice, was in the ‘Essays, by David Hume,’ published about the year 1742. In that work it is mentioned as the name of a principle which might be made the foundation of a system of morals, in opposition to a system then in vogue, which was founded on what was called the ‘moral sense.’ The ideas, however, there attached to it, are vague, and defective in practical application. perronet thompson 1829, 258
The second one, the historical survey by Bowring (1834), while going into greater detail than Perronet Thompson (1829) and virtually duplicating the entire content of the original, ultimately reflects the same combination of praise and censure: Hume’s Essays recognised Utility as a principle. He employed the word with much indistinctness, sometimes representing the idea of usefulness—usefulness considered as conduciveness to an end, no matter what; sometimes as synonymous with conduciveness to happiness as an end. On no occasion does he intimate, that the idea of happiness is to be inseparably connected with the idea of utility. bowring 1834, 291–292
66
Chapter 1
Hume, however, did this great service. He pointed out utility as the foundation or cornerstone of a system of morals. bowring 1834, 294
Bentham (1828/1843) was already asserting the importance and influence of Hume before the organically related historical accounts began to come out, for example, when he was preparing the second edition of A Fragment on Government (1776/1823): Under the name of the principle of utility, (for that was the name adopted from David Hume), the Fragment set up, as above, the greatest happiness principle in the character of the standard of right and wrong in the field of Morality in general, and of Government in particular. bentham 1828/1843, 242
The quotation is from a preface drafted by Bentham for the second edition of the book in question (A Fragment on Government, 1823) but eventually left out of that edition and printed instead, though not published, simply as “Historical Preface Intended for the Second Edition” (1828).46 As he makes clear at every turn, it is not just the name of the principle that he gets from Hume but also the approach that the principle recommends for the philosophical distinction between right and wrong in moral, social, and political discourse. The methodological shortcomings Bentham (1829/1983) notes in Hume’s utilitarianism constitute his segue into Hartley and Helvétius, whom he regards as bringing the conceptual and philosophical development of utilitarianism more into line with the classical paradigm eventually perfected by Bentham himself. While Hartley (1749) gets credit for producing the most extensive classification of pleasures and pains, far more advanced than anything conceived by Helvétius, he does not rank anywhere near as high as Helvétius in Bentham’s estimation of the creative forces in the history of utilitarianism. Helvétius (1758) receives incomparable praise for drawing foundational connections between human nature and moral behavior that anticipate Bentham’s own appeal to psychological theory in elaboration and justification of moral philosophy.
46
The preface was reprinted as part of the first volume of The Works of Jeremy Bentham (pp. 240–259). See Bentham (1828/1843) for full publication details concerning the preface, Bentham (1843) for the same details concerning the compilation.
Classical Utilitarianism before John Stuart Mill
67
The relevance of the insights in Helvétius is why his name comes up after every credit extended to Hume for his own innovations, which remain rudimentary by comparison, that is, before elaboration through Helvétius. Bentham’s “Article on Utilitarianism” (1829/1983), for example, turns to Helvétius right after opening with a tribute to Hume. The transition comes in the third paragraph of the long version: Not long before or after this time [the publication of Hume’s Essays] came out in French the then, and still, so highly celebrated work of Helvetius entitled Sur l’esprit [De l’Esprit]. In this work, a commencement was made of the application of the principle of utility to practical uses. To the direction of human conduct, in the ordinary course of life, a connection was formed between the idea attached to the word ‘happiness’, and again between the idea attached to the word ‘happiness’ and the ideas respectively attached to the words ‘pleasure’ and ‘pain’. To several pleasures which as yet had been without a ‘local habitation and a name’, both those appendages were now and thus attributed: attached to the words ‘utility’ and ‘principle of utility’ were now ideas in abundance, ideas which could not but be continually present and familiar to the most inattentive, unobservant and scantily-instructed minds. bentham 1829/1983, 290
The short version is even less restrained in its praise for the genius of Helvétius and for the value of his work (Bentham 1829/1983): Epoch the fifth. Year 1758: date of the first publication of the work of Helvetius entitled de l’Esprit: a word to which, unhappily, no clear idea stands attached and of which no equivalent is to be found in the English language. Important is the service for which morals and legislation stand indebted to this work: but to give in any small number of words any totally correct and complete conception of the virtues of that service is scarcely possible. The light it spreads, on the field of this branch of art and science, is to that steady light which would be diffused over it by a regular institute or say didactic treatise, like what the meridian sun sheds over a place when bursting forth one moment from behind a cloud it hides itself the next moment behind another, is to that comparatively pale but regular and steady system of illumination afforded to a street by two constantly lighted rows of lamps. bentham 1829/1983, 325
68
Chapter 1
The strength of Bentham’s praise for Helvétius is fully reflected in the enthusiasm of his acknowledgment of a direct influence on his own work (Bentham 1829/1983): To this work [De l’Esprit by Helvétius] Mr. Bentham has often been heard to say that he stands indebted for no small part of the ardour of his desire to render his labours useful to mankind upon the largest scale, and for the energies he finds to persevere in them, and for the hope and belief that they would not be altogether fruitless. bentham 1829/1983, 325
Helvétius naturally receives comparable attention from both Perronet Thompson (1829) and Bowring (1834). Perronet Thompson’s editorial (1829), actually conceived with a more specific end in mind than historical exposition, as discussed above, nevertheless recognizes the role of Helvétius in the development of utilitarianism: Nearly at the same time [as “Essays, by David Hume”] appeared in French the celebrated work of Helvetius ‘Sur l’Esprit’ [De l’Esprit]. In this a commencement was made, of the application of the principle to practical use. A connection was established between the ideas attached to the word ‘happiness,’ and those attached to the words ‘pleasure’ and ‘pain;’ by which a great advance was made in the developement of the meaning of the terms ‘utility’ and ‘principle of utility.’ perronet thompson 1829, 258
Bowring’s survey of the history of utilitarianism (1834), dedicated exclusively to the process in question and focusing more closely on the original manuscript of Bentham (1829/1983), allocates greater room to the role of Helvétius, both with respect to the development of utilitarianism and in connection with the personal enlightenment of Bentham, as illustrated, respectively, in the following two excerpts: Helvetius wrote, in 1758, his work De l’Esprit, a title for which no adequate translation has been found in our language; the word, unfortunately, having no English equivalent. Great indeed was the contribution which that book brought to the science of morals and legislation; but it would be most difficult, in a few sentences, or even pages, to convey a correct estimate of all it did, and all it left to do. bowring 1834, 295–296
Classical Utilitarianism before John Stuart Mill
69
To that book [De l’Esprit], however, Mr Bentham has often been heard to say he stood indebted for no small portion of the zeal and ardor with which he advocated his happiness-producing theory. It was from thence he took encouragement, flattering his efforts with the assurance that they would not be useless. It was there he learned to persevere, in the conviction that his power would strengthen, and his field of usefulness extend. bowring 1834, 296
The works revolving around Bentham’s (1829/1983) account of utilitarian history are not the only examples of his assessment of Helvétius. They are merely reflections in retrospect, drawing out the implications of an incomparable influence that began to shape Bentham’s thoughts during the formative years of his life. Bentham reportedly discovered very early that he, like Helvétius, had a “genius” for legislation (Bentham 1843, vol. 10, 27, cf. 54, 70–71). Here is how Bowring puts the matter, as his biographer, in relating Bentham’s reflections in 1768, at the age of twenty, looking back on a question occupying his mind from 1755 onward, that is, ever since childhood: The question, “What is genius?” haunted young Bentham for many years. No distinct conception could be attached to it; but, at the age of twenty, Helvetius’ book, De l’Esprit, having fallen into his hands, it occurred him that Genius was a noun-conjugate, derived from the verb gigne, of which the perfect tense was genui, and the sense became sufficiently indicated. Horace’s line, “Seit genius, natale comes qui temperat astrum,” did not bring any solution of the difficulty. But, to discover that genius meant invention or production, was no small matter; and the discovery acted powerfully on Bentham’s mind. “Have I a genius for anything? What can I produce?” That was the first inquiry he made of himself. Then came another: “What of all earthly pursuits is the most important?” Legislation, was the answer Helvetius gave. “Have I a genius for legislation?” Again and again was the question put to himself. He turned it over in his thoughts: he sought every symptom he could discover in his natural disposition or acquired habits. “And have I indeed a genius for legislation? I gave myself the answer, fearfully and tremblingly—Yes!” bentham’s memoirs from 1768, narrated by Bowring: bentham 1843, vol. 10, 27
The anecdote, particularly where it turns to direct quotation, recounts what Bentham says almost word for word (“almost in Bentham’s words”) as Bowring assures us. The influence of Helvétius, however, evidently extended beyond
70
Chapter 1
a shared affinity, or common propensity, for legislation. Reflecting on his experiences from the very next year, 1769, Bentham reports, in his own words, that Helvétius had a direct influence on him in connection with the principle of utility, going so far as to name the French philosopher as his single most important source and inspiration: The year 1769 was to me a most interesting year. I was, I remember, reading Montesquieu, when the Archbishop of York called on me, to solicit my vote for Jenkinson and Hay. Prodigiously courteous was his grace; though I was only half dressed, and was busy, too, on chemistry, evaporating urine in order to obtain phosphorus. The ignorant mother of Chamberlain Clarke laughed at me, but laughed in vain. I was beginning to get gleams of practical philosophy. Montesquieu, Barrington, Beccaria, and Helvetius, but most of all Helvetius, set me on the principle of utility. When I had sketched a few vague notions on the subject, I looked delighted at my work. I remember asking myself—Would I take £500 for that sheet of paper? Poor as I was, I answered myself—No! that I would not. bentham’s memoirs from 1769: bentham 1843, vol. 10, 54
The strength of the connection reported by Bentham is confirmed by Bowring, who writes that De l’Esprit (1758), a highly controversial study of psychological egoism, was indeed Bentham’s favorite book at one time: In the visits which Bentham paid to the country with his father and stepmother, and which were frequent at this period [during 1770], he usually walked behind them, alone, reading; and his favourite book was “Helvetius de l’Esprit.” bentham’s memoirs from 1770, narrated by Bowring: bentham 1843, vol. 10, 54
This is consistent with Bowring’s earlier identification of De l’Esprit as an influential source in Bentham’s development and formulation of utilitarianism, for which “he stood indebted” to Helvétius, as noted and quoted above, “for no small portion of the zeal and ardor with which he advocated his happiness- producing theory” (Bowring 1834, 296). Bentham himself leaves no room for doubt concerning the role of Helvétius in his development as a utilitarian, particularly in the early years of his scholarly and professional life, as he outlines the details of his debt to Helvétius, whose influence he remembers together with that of Locke:
Classical Utilitarianism before John Stuart Mill
71
A digest of the Laws is a work that could not have been executed with advantage before Locke and Helvetius had written: the first establishing a test of perspicuity for ideas; the latter establishing a standard of rectitude for actions. The idea annexed to a word is a perspicuous one, when the simple ideas included under it are assignable. This is what we owe to Locke. A sort of action is a right one, when the tendency of it is to augment the mass of happiness in the community. This is what we are indebted for to Helvetius. The matter of the Law is to be governed by Helvetius. For the form and expression of it we must resort to Locke. From Locke it must receive the ruling principle of its form—from Helvetius of its matter. By the principles laid down by Locke it must be governed, inasmuch as it is a discourse; by those of Helvetius, inasmuch as it is a discourse from authority, predicting punishment for some modes of conduct, and reward for others. bentham’s memoirs from 1773–1774 (“Sundry Memoranda of Bentham Made in 1773–74”): bentham 1843, vol. 10, 70–71
It is a testament to Bentham’s intellectual integrity that he takes the time to acknowledge the various contributions of Locke, despite opposing him in matters that are more relevant in the course of discussion, most notably including Locke’s advancement of the social contract tradition.47 Bentham’s sense of professional responsibility actually extends as far back as impressions from his childhood, where we encounter a record of his earliest debt, one that precedes even his acquaintance with Helvétius. He reports his fondest memories in that regard of Fénelon’s Les Aventures de Télémaque (1699/1717), having read the novel at the age of six or seven as his first source of inspiration for the principle of utility, though it is not the protagonist, Telemachus, but one of his antagonists, a competitor in a contest, who voices what at the time strikes Bentham as an approximation of the greatest happiness principle (Bentham 1843, vol. 1, 10). Helvétius, of course, was a more influential source than Fénelon, despite the latter’s hold on Bentham at a more impressionable age.
47
Relevant coverage of Locke includes, in addition to the preceding block quotation in the main text above, Bentham’s “Article on Utilitarianism” (1829/1983, 293, 298–299, 314–316, 322), Bowring’s “History of the Greatest-Happiness Principle” (1834, 291, 304–310), and Perronet Thompson’s “ ‘Greatest Happiness’ Principle” (1829, 259). Favorable opinions are limited largely to p. 298 of Bentham (1829/1983) and p. 291 of Bowring (1834).
72
Chapter 1
The extent of Bentham’s debt to Helvétius is fully appreciated by Sidgwick (1877), who identifies the connection as the most important influence on Bentham’s development as a philosopher and radical reformer: Something of the same blending of contraries is found in Helvetius; and he perhaps, rather than Hume, should be taken as the intellectual progenitor of Bentham. In Helvetius, however, though utilitarianism is passing out of the critical and explanatory phase in which we find it in Hume, into the practical and reforming phase, the transition is not yet complete. Still the premises of Bentham are all clearly given by Helvetius; and the task which the former took up is that which the latter clearly marks out for the moralist. Indeed, if we imagine the effect of L’Esprit on the mind of an eager young law-student, we seem to have the whole intellectual career of Bentham implicitly contained in a “pensée de jeunesse.” sidgwick 1877, 638
Bentham’s early exposure to Helvétius evidently stayed with him to the end, shaping his synthesis of the twin pillars supporting his own approach to utilitarianism, a combination of psychological egoism and (universalistic) ethical hedonism, the former a matter of fact grounded in observation, the latter a normative corollary supported by extrapolation, though Bentham himself never claimed to have demonstrated the validity of the connection beyond what he personally found to be a satisfactory explanation:48 Helvetius puts with a highly effective simplicity, from which Hume was precluded by his more subtle and complex psychological analysis, these two doctrines: first, that every human being “en tout temps, en tout lieu” seeks his own interest, and judges of things and persons according as they promote it; and secondly, that, as the public is made up of individuals, the qualities that naturally and normally gain public esteem and are called virtues are those useful to the public. sidgwick 1877, 638–639
48
There is nothing in Bentham approaching the obvious resolve in the systematic initiative of John Stuart Mill to “prove” the validity of the theoretical link between psychological egoism and ethical hedonism. Bentham was content to acknowledge the connection as an intuitive position, evidently finding it a hopeless cause to try to bridge the gap between personal interest and universal benefit beyond a reasonable doubt, though he himself clearly found the connection reasonable. As a matter of fact, in the course of the first
Classical Utilitarianism before John Stuart Mill
73
As suggested in the first section, and documented afterwards with examples, Bentham’s greatest muses in the development of utilitarianism were Bacon, Beccaria, Hartley, Helvétius, Hume, and Priestley, some more in terminology, others more in philosophy, but all gratefully acknowledged. Despite his claims, often bordering on immodesty, to have vastly improved and presumably perfected every aspect of practically everything inspiring him in the works of his predecessors, his own greatest achievement was in systematizing and applying what was already available to him in a long chain of conceptual and methodological contributions waiting to be consolidated and elaborated in a meaningful way. This also seems to be Sidgwick’s considered opinion of Bentham’s place in the history of utilitarianism: Bentham’s originality and importance lay not in his verbal adoption of utility as an end and standard of right political action, but in his real exclusion of any other standard; in the definiteness with which he conceived the “general good;” the clearness and precision with which he analysed it into its empirically ascertainable constituents; the exhaustive and methodical consistency with which he applied this one standard to all departments of practice; and the rigour with which he kept its application free from all alien elements. sidgwick 1877, 631
chapter (section 11) of An Introduction to the Principles of Morals and Legislation (1789), Bentham explicitly denies both the possibility and the necessity of proving the principle of utility in the standard sense (“direct proof”): “Has the rectitude of this principle been ever formally contested? It should seem that it had, by those who have not known what they have been meaning. Is it susceptible of any direct proof? It should seem not: for that which is used to prove every thing else, cannot itself be proved: a chain of proofs must have their commencement somewhere. To give such proof is as impossible as it is needless” (Bentham 1789, 4). This admission is what Sidgwick seems to have in mind where he oversimplifies Bentham’s position with rhetorical undertones albeit without gross distortion: “Psychologically analysed, common morality appears as a simple result of common selfishness. ‘Each man likes and approves what he thinks useful to him; the public (which is merely an aggregate of individuals) likes and praises what it thinks useful to the public; that is the whole account of virtue.’ How, on this theory, men’s moral judgments come to agree as much as they actually do is not sufficiently explained; and in any case there is no rational transition possible from this psychological theory to the ethical principle that ‘the standard of rectitude for all actions’ is ‘public utility.’ Nor does Bentham really maintain that there is: when he is pressed, he explains frankly that his first principle is really his individual sentiment; that, in fact, he aims at the general happiness because he happens to prefer it” (Sidgwick 1877, 648; cf. 1886, 232–233).
74
Chapter 1
The gist of this assessment is confirmed, with the addition of a few specific references, by John Hill Burton, one of the editors of the first volume of The Works of Jeremy Bentham, where Burton (1843, 5–83) contributes a long introduction: But it was not to the announcement of his first principle that Bentham trusted for its adoption, but to the influence it would have on the minds of his readers when they studied the forms in which he brought it out in detail. And this brings us to examine the extent to which the author lays claim to the merit of originality. It was not the principle itself, that constituted his discovery, but his rigid adherence to it in all his expositions—his never losing sight of it, in what he did himself or called upon others to do. He did not say that the world had hitherto been ignorant of such a principle; he found the theory of utility to a certain extent promulgated by Hume, and references to the “greatest happiness” in the works of Beccaria and of Priestley; while something like the Utilitarian Principle is announced at the commencement of the Nicomachean Ethics. He found indeed that it was at the root of all systems of religion and morality; that all codes of law were more or less founded upon it; and that it was, in all places and at all times, an unseen and unacknowledged guide to human action. burton 1843, 19–20
This is a fair summary, consistent with the opinions and assessments of various others likewise giving the matter serious thought. Without denying any of Bentham’s substantive contributions, all summarized in the first section, one can fairly conclude that he left his mark in the history of philosophy both through his comprehensive command of utilitarian doctrine, including all its implications and ramifications, and through his passionate promotion and rigorous application of it in moral, social, political, and legal contexts. That is at least partly why he never hesitated to give credit where it was due, though obviously not shrinking from taking off a few points either where it seemed appropriate and necessary. 1.5
Patterns of Indirect Inspiration and Transmission
Direct influence must not be confused with indirect inspiration or transmission. The adoption of ideas and terms from specific works is an entirely different matter from the formulation of thoughts and narratives on the basis of general familiarity with the relevant scholarship. What this means is that the
Classical Utilitarianism before John Stuart Mill
75
question of inspiration is not limited to immediate or comprehensive acquaintance with any particular contribution by any given person. Both ideology and terminology have a way of traveling from mind to mind with or without direct exposure. The dissemination of good ideas expressed in memorable terms is unstoppable, though the corresponding path is difficult to predict in advance, or to trace in retrospect, given that the transfer of ideas from one node to the next is not always neatly structured or clearly documented. This was true even before the advent of electronic or digital communication, where new ideas in print would later reach foreign audiences in translation, consequently inspiring local contributions, which, in turn, would promote and perpetuate the source material, with either the original or a translation or an elaboration, or perhaps several of them in succession, subsequently reaching the next country in translation, and any element in that chain of transmission eventually making its way back to the point of origin. It is therefore conceivable as a possibility, though not demonstrable as a fact, that Bentham may have been inspired by Hutcheson, or even by Leibniz, without being aware of it himself. A graphic illustration of just such a scenario is available in the work of Robert Shackleton (1972), who traces the path of the expression “the greatest happiness of the greatest number” from Hutcheson, where it presumably originates, to Bentham, where it famously culminates. The transmission envisaged proceeds in three languages through three centers of activity, as “the greatest happiness for the greatest numbers” in the English formulation by Hutcheson (1725) becomes “un plus grand bonheur à un plus grand nombre de personnes” in its French rendition by Marc Antoine Eidous (1749), inspiring “la massima felicità divisa nel maggior mumero” in its Italian elaboration by Beccaria (1764), the latter being rendered as “the greatest happiness of the greatest number” in an anonymous English translation (1767), which then starts circulating as the version encountered by Bentham in 1768 as he later recollects with some hesitation (Bentham 1843, vol. 10, 142). This is merely a skeletal sketch of the sequence of events meticulously documented by Shackleton (1972, 1466–1473), providing a plethora of names, publications, and locations pertinent to the network of activity responsible for the propagation of this rallying cry of utilitarian ideology. The aphorism evidently originated simply as a metaethical assessment, gradually taking on practical and normative characteristics, and eventually turning into a social and political slogan, particularly through the efforts of radicals in pursuit of democratic reform throughout the monarchies of the eighteenth century. Hutcheson himself was expressing a strictly mathematical relationship grounded in scientific evidence and philosophical deliberation as he spoke of “the greatest happiness for the greatest numbers” in An Inquiry
76
Chapter 1
into the Original of Our Ideas of Beauty and Virtue (1725, 163–164), the full title of which terminates in an explicit designation of the work as an “Attempt to Introduce a Mathematical Calculation in Subjects of Morality.” The continental transmission of the phrase, as well as the correlative reception of the host publication, evidently started out similarly with a metaethical framework, later expanding into moral, social, and political dimensions as it began to attract the attention of the leading scholars of the period, including the prominent philosophes of the European Enlightenment. The journey of the phrase from the British Isles to continental Europe, and back again to Britain, combines various channels of direct and indirect transmission.49 Hutcheson’s work (1725) initially met with a more favorable reception in Italy than in France, until the French translation by Eidous (1749), the first in any language, boosted familiarity in France, where it attracted the attention of the Encyclopédistes, most notably Denis Diderot, Charles Pinot Duclos, and Louis de Jaucourt, subsequently reaching the Physiocrats through Pierre-Paul Mercier de La Rivière (cf. Shackleton 1972, 1469–1470). In the meantime, the French translation made its way into Italy, where the influence of the book, together with the visibility of the phrase, immediately doubled when the English and French editions became available simultaneously. The two countries thus became the most likely locations for exposure to Hutcheson apart from his immediate circle. Growing acquaintance with the formula of “the greatest happiness of the greatest number,” following its availability in two languages in two countries outside Britain, coincided with the propagation of reformation ideals and revolutionary sentiments stoked by a stream of contributions by the likes of Montesquieu, Rousseau, and Helvétius. Robust changes in the conduct and quality of social and political philosophy, combined with increasing attention to equality and liberty at the level of the individual, had their metaethical counterparts in the proliferation of mathematically oriented continental studies in happiness, particularly in Italy and France, where such work was already burgeoning by the middle of the eighteenth century independently of Hutcheson (cf. Shackleton 1972, 1468). It is this auspicious combination of developments, not to mention his personal acquaintance with the work of 49
A general reference to the British Isles is probably the most specific designation possible for a place of origin for the expression in question, provided, of course, that it really does originate with Francis Hutcheson (1694–1746) as suspected. Hutcheson was born in Ireland to Scottish parents, moved to Scotland for his education, returned to Ireland for an academic career, and published An Inquiry into the Original of Our Ideas of Beauty and Virtue (1725) in London.
Classical Utilitarianism before John Stuart Mill
77
Hutcheson, presumably through its French edition, that ultimately inspired Beccaria to produce his own tract, Dei Delitti e delle Pene (1764), where the utilitarian maxim of “the greatest happiness of the greatest number” found its first Italian expression. The balance of terminological exposure between Italy and France was restored as Beccaria’s book, proving immediately and immensely popular throughout Europe, was translated into French by André Morellet (1766), whereupon the phrase soon to start serving as the alpha and omega of utilitarian vocabulary became available in three languages in both countries. That final step from Italian into French, where the phrase then became available in two local renditions, “un plus grand bonheur à un plus grand nombre de personnes” (1749) in the translation from Hutcheson (1725), and “la plus grande félicité du plus grand nombre” (1766) in the translation from Beccaria (1764), turned out to be the last stop before Bentham. The English translation of Beccaria came a year after the French, as an anonymous translator drawing simultaneously on the Italian original and the French translation produced the first English edition (1767), which still stands as the most likely candidate for where Bentham first encountered the fateful expression. Shackleton’s (1972) own summary of the process is as follows: The migration of the phrase has now been traced: starting with the Scottish philosopher [Hutcheson] in 1725, it has passed with variations of language and in shade of meaning into France [French translation of Hutcheson] and then into Italy [French translation of Hutcheson], from Italy [text of Beccaria] back in Morellet’s version into France [French translation of Beccaria], and thence again into England [English translation of Beccaria]. In the course of this journey it has gathered fame and momentum until it is ready to be an efficient propagandist weapon in the hands of Bentham, as well as being able to serve so abstract a concept as the felicific calculus. But this is not the end of its journey. It remains to study its fate in the hands of Bentham. shackleton 1972, 1473; cf. 1466–1473
1.6
Bentham’s Own Terminological Predilections
Regardless of where he got his ideas and no matter where he picked up the corresponding terminology, both classical utilitarianism and the associated vocabulary came to be identified more with Bentham than with anyone else.
78
Chapter 1
With respect to ideology, his application of utilitarian doctrine is so consuming, so comprehensive, and so consistent that he tends to come across as if he alone were responsible for its every detail, ostensibly having created and developed the underlying philosophy with a personal stake in the corresponding movement. With respect to terminology, he uses the terms “the principle of utility” and “the greatest happiness principle,” as well as the expression “the greatest happiness of the greatest number,” with such regularity and such rhetorical force—albeit in separate preferential patterns often competing with each other, though sometimes instead complementing one another—that hardly anyone other than historians of philosophy are moved to explore earlier sources. Although this dual association of utilitarianism with Bentham has become something of a legend, a fixed record habitually repeated without reflection as if he were the solitary mastermind behind every concept and expression, the man himself had a mercurial relationship with the prevailing nomenclature if not also with the supporting thoughts and propositions. He attached great importance to proper terminological choices, firmly convinced that the best ideas were only as good as the words used to express them (Bentham 1829/ 1983, 302, 309–310; Bowring 1834, 319–321, 328–330; Perronet Thompson 1829, 267–268). His own survey of the emergence and development of utilitarianism is, after all, primarily a terminological history, not a philosophical or ideological study, as he promptly and clearly announces in the opening paragraphs of both versions of his “Article on Utilitarianism” (Bentham 1829/1983, 289, 320–321). That article, in turn, provides not only the groundwork but, in some places, also the actual content, of two others following its lead. First, Perronet Thompson (1829, 267–268) takes his cue directly from the terminological focus of Bentham’s approach, as he concludes his editorial on utilitarianism with a sketch of the evolution of Bentham’s terminological predilections. Later, Bowring (1834) follows suit in his choice of title (“History of the Greatest- Happiness Principle”) for his account drawing and elaborating on Bentham’s manuscript (1829/1983). It is no wonder, then, that nomenclature is a central concern in all three works. Bentham (1829/1983) introduces the subject as follows: Little aware are people in general of what importance the business of nomenclature (nomenclature [as] an instrument in the hand of Logic) is in the plantation of new ideas and dissemination of already-rooted ones: how extensive, serious and indisputable the importance of it may be in practice; how considerable the benefit sometimes capable of flowing
Classical Utilitarianism before John Stuart Mill
79
from the possession of an apt appellation; how considerable the inconvenience, the evil, capable of growing out of the want of it. bentham 1829/1983, 302; original interpolation
Perronet Thompson (1829) then puts the same point, including the underlying rationale, to good rhetorical use in the service of an emphatic and evocative conclusion to his already polemical editorial: Those who object indiscriminately to inventions in nomenclature, are either ignorant of their power, or jealous of their effect. It is, however, very difficult to combine the three ideas of ‘greatest, aggregate, and happiness,’ in a single word. What friendly efforts have failed to effect, the scorner has perhaps supplied. That name is best, which most strongly excites in the minds of friends and enemies, the impressions designed to be conveyed. It has always been permitted, to learn from an opponent. The Utilitarians shall abandon foreign titles, and ‘the sacred language of the Benthamites’ be all that shall be heard of by posterity. perronet thompson 1829, 268
Bowring (1834) subsequently brings out more of the personal dimension as he reveals the full extent of the value Bentham placed on nomenclature: It is scarcely out of place, by the way, to state here, in answer to those who have so frequently animadverted on Mr Bentham’s unusual terms, that there is no topic on which his mind was more habitually occupied than in the search of fit terms to convey his ideas. No man was ever more impressed with the importance of appropriate nomenclature, as the necessary instrument for logical reasoning, for introducing and disseminating correct ideas. bowring 1834, 321
All three make it abundantly clear that a good idea expressed badly was, in Bentham’s estimation, hardly an improvement over an inferior idea presented persuasively. That conviction was, in fact, the driving force behind Bentham’s lifelong experimentation with terminology. Although his zealous pursuit of the best way of expressing the finer points of utilitarian doctrine often proved useful, producing memorable designations still in use today, or rather, popularizing the best of those that were already circulating in print, the extended operation was ultimately an exercise in futility, given that Bentham was, in essence, trying to capture a complex notion, namely the full meaning and
80
Chapter 1
intent of a philosophical principle, in a single term or expression, preferably one that was both accurate and catchy, much like a slogan. He was always looking for a better formulation, and if possible, the perfect formulation, even after he was certain that he had found it. Bentham was effectively on a permanent mission throughout his career to determine the best words to articulate his best thoughts. That mission took him through several distinct stages. He started out with the term “the principle of utility,” borrowed from Hume, but he gradually became disillusioned with the inherent absence there of any reference to happiness, either as the measure of utility or as the content of utility. He therefore replaced it with the phrase “the greatest happiness of the greatest number,” evidently inspired by Beccaria. Yet he subsequently recognized the weakness of that expression in reconciling maximization with distribution in the pursuit and promotion of happiness. He eventually settled on “the greatest happiness principle,” which was not so much a newly created substitute as it was a perennial favorite already in use alongside both of the other two, not to mention comparable variations of all three. These are only the main avenues of an intricate and relentless search for perfection. The most remarkable feature of Bentham’s never-ending quest is the equal resolve and enthusiasm with which he was inclined to adopt and later abandon any turn of phrase as the best or worst possible means of articulating utilitarian philosophy and ideology.50 Alternating periods of fascination and disenchantment in his experience with the expression “the greatest happiness of the greatest number” are especially noteworthy in that regard. Having originally encountered the phrase in 1768, evidently through Beccaria recollected erroneously as Priestley, with a palpably euphoric sense of enlightenment in any event, Bentham employed the expression as early as the preface to his first publication, A Fragment on Government (1776), where he introduced it as a “fundamental axiom”: “It is the greatest happiness of the greatest number that is the measure of right and wrong” (Bentham 1776, ii). Yet despite such a strong inauguration, he surprisingly made no other reference to the expression, either there or elsewhere, for a very long time. The
50
Note the decisive tone of Bentham’s announcement that he had abandoned the designation “the principle of utility” in favor of the alternative “the greatest happiness principle,” a resolution communicated in a letter dated 6 April 1822, where he goes on to contemplate a comparable transition in French: “ ‘Principle of utility—is dead and gone. Greatest happiness principle has succeeded to it: that is to say in English. What should hinder Principe du plus grand bonheur from succeeding it in French? Think of the principe de moindre action—there you have a match for it’ (ms.10.129)” (Shackleton 1972, 1478).
Classical Utilitarianism before John Stuart Mill
81
only other time it turns up in that work is several decades later in its second edition (1823, 45–48), where Bentham added a lengthy footnote (note l) to the first sentence of section 48 of the first chapter, as noted in section 1.2 of the present chapter. The term’s occurrence there, twice in the same note, is actually derivative, and explicitly acknowledged as such, coming as it did from an addendum Bentham had then recently (12 July 1822) introduced to a footnote (note c) to the first sentence of section 13 of the first chapter of An Introduction to the Principles of Morals and Legislation (1789), at the time also in its second edition (1823, 7–9). These are the only occurrences in these two works, commonly accepted as Bentham’s greatest. The same expression can be found in abundance elsewhere, but only rather late in his career, much like the second editions of the two works cited (1823). Another example is Bentham’s Codification Proposal (1822), coming out the year before the other two, and opening immediately with a reference to “the greatest happiness of the greatest number,” a set of words invoked verbatim for a total of thirty-eight times, including once on the title page, seven times in the table of contents, and thirty times throughout the text, amounting to an average of nearly once every two pages. And then there are those works, a great many, where the expression is nowhere to be found at all, despite the supreme relevance of the subject matter. The difference, as it tuns out, is not contextual but temporal. Bentham’s usage, in other words, varies in accordance with when he wrote a piece, not with what he wrote in it. The Codification Proposal (1822) happens to be the first publication where the slogan appears after its debut in the first edition of A Fragment on Government (1776), pointing to a hiatus just a few years short of half a century. The presence and prevalence of the phrase “the greatest happiness of the greatest number” is so dominant in the Codification Proposal (1822) that Bentham himself, as mentioned in section 1.1 above, later remembers that work erroneously as the first time he had ever used the expression in print (Bentham 1829/1983, 327; Bowring 1834, 319). He seems, at that point, to have forgotten the prominent position he had already assigned to the expression in his very first publication, more than four decades earlier. It remains a curiosity why he would have had such a positive experience with the expression, and such an early start to its philosophical and rhetorical employment, only to forget that he had ever used it in print, until so late in his life and career. Various explanations of a philosophically reasonable and historically illuminating nature are available, though none of them is entirely satisfactory on its own. Standing out among the rest are those by David Lyons (1973) and
82
Chapter 1
Robert Shackleton (1972), both subsequently evaluated with additional insight by James Henderson Burns (2005). Lyons (1973, 19–34) identifies the terminological anomaly in Bentham, specifically the absence of the trademark expression from An Introduction to the Principles of Morals and Legislation (1789), with what may be described as a less universalistic approach or application, where Bentham is concerned with the interests of a community decidedly smaller than the whole of humanity, the scope of consideration being correlated with the “actions directed” in the corresponding context rather than by the “interests affected” in that context.51 He finds this restricted focus to be grounded in Bentham’s distinction between private ethics, concerning the individual as a moral agent with only personal powers and responsibilities, and public ethics, concerning a community of individuals with various institutional powers and responsibilities defining their ontological and moral status as a community (Lyons 1973, 30). This makes Bentham’s subsequent restoration of the term a matter of his gravitation toward a more universalistic form of utilitarianism in a more public domain where the greatest happiness of the greatest number would have indeed been the primary concern. Shackleton (1972) associates the end of the same hiatus with Bentham’s growing advocacy of radical reform in response to social unrest in Britain in the aftermath of the Napoleonic Wars (1800–1815).52 This is not so much a competing interpretation as it is a complementary one. A transition from private to public affairs in Bentham’s utilitarianism, and thereby in his application of its fundamental principle, may well have been why he resurrected the possibly more universal formulation of “the greatest happiness of the greatest number,” just as Lyons claims. Yet that very transition could conceivably have been inspired by Bentham’s growing involvement in political radicalism, just as Shackleton suggests. One incidental difference, however, is that Shackleton (1972, 1480) specifically credits James Mill (1820) with inspiring Bentham to
51
52
The explanation David Lyons offers for why the expression “the greatest happiness of the greatest number” does not occur anywhere in An Introduction to the Principles of Morals and Legislation (1789) is that “it [the expression] does not represent his [Bentham’s] views about conflicts of interest when he wrote that book,” where “any universalistic connotations that the phrase may have” remain “foreign to that work” (1973, 24). The terminological hiatus, it must be noted, applies exclusively to the expression “the greatest happiness of the greatest number,” whose analogues and alternatives, including “the principle of utility” and “the greatest happiness principle,” were never abandoned either permanently or decisively.
Classical Utilitarianism before John Stuart Mill
83
revive the expression as a slogan, particularly as a rhetorical tool to make a methodological point in promotion of an ideological agenda.53 Burns (2005) comments at length, and in the main favorably, on both interpretations, though with some reservations in each case. He sees the course of the relevant terminology as a natural consequence of Bentham’s lifelong search for the best way of expressing his position on the felicific calculus, that is, on the quantification of happiness as the standard of right and wrong. According to Burns, the return of the expression “the greatest happiness of the greatest number” to Bentham’s published writings after such a long absence is a reflection of his growing interest in the principles of effective, efficient, and equitable organization in the public sphere, both in terms of administration and in terms of legislation, in contrast to anything of major concern in the private sphere. Hence, much like Lyons (1973), Burns (2005) recognizes a methodological difference between private and public ethics in the philosophy of Bentham. Yet unlike Burns, he does not take that difference to be correlated in any meaningful way with the universalism of Bentham’s approach at the time, or with the strategic or contextual relevance of “his views about conflicts of interest” (cf. Lyons 1973, 24), instead holding the principle of utility, under any name, to be universalistic in the same sense and to the same degree in both the private sphere and the public sphere. On the other hand, just like Shackleton (1972), he attributes the revival of “the greatest happiness of the greatest number” as a formal designation to Bentham’s evolving interest and involvement in social and political reform and to his correlative commitment to radical democracy. Unlike Shackleton, however, Burns does not credit James Mill as a catalyst for the change in Bentham, noting that Bentham’s unpublished works preceding the alleged inspiration by Mill (1820) were already replete with a return to the expression in question. Each of these interpretations captures an important aspect of the whole story. They all work well as an explanation of why Bentham, or anyone else in his place, would have been slow to adopt the expression, but not so much as an explanation of why Bentham was, in fact, rather quick to embrace it, even quicker to drop it, relatively slow to pick it back up, and subsequently eager to retire it for good. What is to be explained here is not a single process of gradual development but discrete periods of intense appreciation, abrupt 53
The inspiration suggested is in the form of leadership by example, with James Mill using the expression “the greatest happiness of the greatest number” in one of his own publications, an essay on government, known both as “Government” and as “Essay on Government” (1820).
84
Chapter 1
discontinuation, prolonged avoidance, passionate restoration, and final termination, in that order. The binary distinction between private and public ethics is simply not enough to account for the entire history of terminological calibration in utilitarian nomenclature under Bentham, as if he were frantically updating his moral vocabulary, mostly going back and forth between familiar alternatives, while apparently randomly expanding or contracting the segment of the population whose happiness was to serve as a benchmark. As documented above with references and quotations (Bentham 1821, 24; 1829/1983, 290–293, 325–326; 1843, vol. 2, 288; 1843, vol. 10, 142; Bowring 1834, 295, 298–300; Perronet Thompson 1829, 258–259), Bentham was fascinated with the phrase “the greatest happiness of the greatest number” from his formative years onward, even before making any headway of consequence in his long and productive career, but he quickly and inexplicably abandoned the designation, still during the early days of his career, only to revive it fully and enthusiastically in his maturity, to say nothing of the reversal of his affinities and convictions yet again after that. To focus more specifically on the portion relevant to the three interpretations considered above, the expression disappeared from Bentham’s vocabulary, or at least from his published writings, for a period of forty-six years, after making an ambitious debut in 1776 as a “fundamental axiom” of the distinction between right and wrong, never to be seen again until its revival in 1822 as a catchphrase repeated ad nauseam for rhetorical effect.54 The three explanations examined in this connection make a strong case for the progressive proliferation of the formula in Bentham’s works, as if the phrase grew on him along a linear path parallel to the development of his interests and inclinations, but that scenario does not go far toward explaining the paradoxical combination of fascination and avoidance preceding the subsequent revival elucidated as a gradual transition. What is at least as interesting as where Bentham originally encountered the phrase “the greatest happiness of the greatest number,” or how he first came to use it with any regularity, is why he gave it up with a calculated and conspicuous effort at avoidance, after introducing it in his very first publication as the definitive statement of utilitarianism, and why he later revived it with manifest confidence and enthusiasm, only to abandon it again just as surely, and, in fact, permanently, following a period of renewed fascination. Whatever we may think of the various explanations for the central absence of the phrase from Bentham’s works, we do not have to speculate about the remainder of the 54
See Robert Shackleton’s “The Greatest Happiness of the Greatest Number: The History of Bentham’s Phrase” (1972) for an analysis of distribution patterns pertaining to the term’s occurrence in Bentham’s published works.
Classical Utilitarianism before John Stuart Mill
85
process leading up to its eventual termination. We are fortunate in that regard to have an explanation in Bentham’s (1829/1983) own words, reproduced in condensed form by Perronet Thompson (1829/1983) and reiterated in different words by Bowring (1834), both drawing on the same material. The main problem forcing Bentham to dispense with the expression, as clearly explained in all three accounts, is its inherent potential for misinterpretation, or outright abuse, as a rhetorical device for ignoring or undermining the interests of minorities for the sake of promoting those of the majority: Greatest happiness of the greatest number. Some years have now elapsed since, upon a closer scrutiny, reason, altogether incontestable, was found for discarding this appendage. On the surface, additional clearness and correctness [was] given to the idea: at the bottom, the opposite qualities. Be the community in question what it may, divide it into two unequal parts, call one of them the majority, the other the minority, lay out of the account the feelings of the minority, include in the account no feelings but those of the majority, the result you will find is that to the aggregate stock of the happiness of the community, loss, not profit, is the result of the operation. Of this proposition the truth will be the more palpable the greater the ratio of the number of the minority to that of the majority: in other words, the less the difference between the two unequal parts: and suppose the condivident parts equal, the quantity of the error will then be at its maximum. bentham 1829/1983, 309; original interpolation; cf. bowring 1834, 328; perronet thompson 1829, 267–268
Distributive justice emerges as the operative consideration for renouncing the terminological benchmark once embraced as the best possible expression of the promotion of happiness as the spirit of utilitarianism. The gist of Bentham’s concern is with the inability of the corresponding expression to indicate a serious devotion to equality or fairness, instead remaining open to misconstrual as a reference to the maximization of aggregate happiness come what may, particularly at the expense of minority interests. Ironically, he is firmly convinced that maximizing happiness is actually, given certain psychological presuppositions, a matter of ensuring an optimal distribution, as he goes on to argue mathematically in the very next paragraph: 54. Greatest happiness of the greatest number. Some years have now elapsed since, upon a closer scrutiny, reason, altogether incontestable, was found for discarding this appendage. On the surface, additional
86
Chapter 1
clearness and correctness [was] given to the idea: at the bottom, the opposite qualities. Be the community in question what it may, divide it into two unequal parts, call one of them the majority, the other the minority, lay out of the account the feelings of the minority, include in the account no feelings but those of the majority, the result you will find is that to the aggregate stock of the happiness of the community, loss, not profit, is the result of the operation. Of this proposition the truth will be the more palpable the greater the ratio of the number of the minority to that of the majority: in other words, the less the difference between the two unequal parts: and suppose the condivident parts equal, the quantity of the error will then be at its maximum. 55. Number of the majority, suppose, 2001: number of the minority, 2000. Suppose, in the first place, the stock of happiness in such sort divided that by every one of the 4001 an equal portion of happiness shall be possessed. Take now from every one of the 2000 his share of happiness, and divide it anyhow among the 2001: instead of augmentation, vast is the diminution you will find to be the result. The feelings of the minority being by the supposition laid entirely out of the account (for such in the enlarged form is the import of the proposition), the vacuum thus left may, instead of remaining a vacuum, be filled with unhappiness, positive suffering—magnitude, intensity and duration taken together, the greatest which it is in the power of human nature to endure. 56. Take from your 2000 and give to your 2001 all the happiness you find your 2000 in possession of: insert, in the room of the happiness you have taken out, unhappiness in as large a quantity as the receptacle will contain. To the aggregate amount of the happiness possessed by the 4001 taken together, will the result be net profit? On the contrary, the whole profit will have given place to loss. How so? Because so it is that, such is the nature of the receptacle, the quantity of unhappiness it is capable of containing during any given portion of time is greater than the quantity of happiness. 57. At the outset, place your 4001 in a state of perfect equality in respect of the means, or say instruments, of happiness—and in particular power and opulence: every one of them in a state of equal liberty, every one independent of every other, every one of them possessing an equal portion of money and money’s worth: in this state it is that you find them. Taking in hand now your 2000, reduce them to a state of slavery, and, no matter in what proportion, of the slaves thus constituted divide the whole number with such their property among your 2001. The operation
Classical Utilitarianism before John Stuart Mill
87
performed, of the happiness of the whole number, 4001, will an augmentation be the result? The question answers itself. bentham 1829/1983, 309–310; original interpolation; cf. bowring 1834, 328–330; perronet thompson 1829, 267–268
The demonstration is intended to illustrate why aggregate happiness can only be maximized through a fair distribution. The reasoning is that maximization is constrained by distribution through the law of diminishing marginal returns, which dictates that aggregate happiness is more sensitive to decreases in personal happiness than it is to increases in personal happiness. This means that any redistribution of happiness, where a fixed amount is transferred from one individual to another, is not just liable to but also certain to reduce aggregate happiness, given that such a transfer cannot enhance the happiness it is added to as much as it diminishes the happiness it is subtracted from. Bentham’s extended thought experiment, or at least the portion of it quoted above, becomes more plausible, or perhaps just more conceivable, as it proceeds from one stage to the next. To be precise, it makes more sense at the end than it does at the beginning. The first paragraph (54) is introductory. It sets up the basic scenario of a disparity in happiness in the relevant community. The next two paragraphs (55 and 56) are exploratory. They illustrate the effects of a redistribution of happiness on aggregate happiness. Both of these are problematic because they work with a redistribution of the wrong thing, happiness itself, but the fourth paragraph (57) remedies the problem by switching to a redistribution of the means to happiness. The reason why the middle two paragraphs (55 and 56) do not work is that redistributing existing happiness is not the kind of thing that could possibly alter the aggregate amount of happiness. Any given quantity of happiness, regardless of the direction of transfer, would have the same presence in the total no matter where it is in the distribution.55 One cannot transfer the exact 55
It may be tempting to think of the same amount of happiness as being worth less for someone who is already sufficiently happy than for someone who is not, but the temptation would be grounded in nothing more than a conceptual confusion. It would be to think of the same amount of happiness as adding a smaller amount of happiness to the existing happiness of the person who is already sufficiently happy than to the existing happiness of the person who is not. There is no other way for the same amount of happiness to be regarded, even mistakenly, as being “worth less” for one person than it is for another. This is not to say that Bentham was confused about the matter. He may have simply been careless in his initial choice of words in his eagerness to set up what he evidently thought was an incontestable demonstration. What he ends up with in the second and third paragraphs (55 and 56) does not work, because a disparity in personal happiness (a
88
Chapter 1
same thing from one place to another in the exact same amount and end up with more or less of it in the system. The scenario makes sense only when indexed to the means to happiness and not directly to happiness itself. The fourth paragraph (57) indeed captures the relevant sense, while adopting a workable index, as it invokes “equality in respect of the means, or say instruments, of happiness—and in particular power and opulence” (Bentham 1829/ 1983, 310; cf. Bowring 1834, 329–330). Bentham seems satisfied that the formula works well if applied as intended, focusing on the distribution, which somehow takes care of the maximization. Although he does not quite say that optimizing the distribution of happiness maximizes the amount of happiness, or that an even distribution of happiness maximizes aggregate happiness, he clearly assigns priority to distribution over maximization rather than the other way around. What worries him is the plasticity of the expression, which remains open to interpretation and prone to misinterpretation. The phrase “the greatest happiness of the greatest number” appears to have struck Bentham, at a critical juncture in his career, as an imprecise and therefore risky way of articulating his considered opinion on the maximization of happiness with the requisite attention to the distribution of happiness. Bentham thereby seems to have noticed, however indistinctly, an intrinsic problem in setting up two competing criteria and aiming for their joint resolution, or simultaneous satisfaction, as in the maximization of aggregate happiness together with the optimization of its distribution.56 His thought
56
disparity in happiness between a happy person and an unhappy person, or between any persons or parties where some are happier than others) is not in any way relevant to the transfer of happiness itself, which would naturally increase the happiness of the happy person no more or less than it decreases the happiness of the unhappy person. The law of diminishing marginal returns otherwise in play here is about the incremental effect on personal happiness of adding more of whatever it was that caused that happiness in the first place, not about the fixed effect of directly injecting more happiness. The first clear articulation of this problem is in Theory of Games and Economic Behavior (1944) by John von Neumann and Oskar Morgenstern: “A particularly striking expression of the popular misunderstanding about this pseudo-maximum problem is the famous statement according to which the purpose of social effort is the ‘greatest possible good for the greatest possible number.’ A guiding principle cannot be formulated by the requirement of maximizing two (or more) functions at once. Such a principle, taken literally, is self-contradictory. (In general one function will have no maximum where the other function has one.) It is no better than saying, e.g., that a firm should obtain maximum prices at maximum turnover, or a maximum revenue at minimum outlay. If some order of importance of these principles or some weighted average is meant, this should be stated. However, in the situation of the participants in a social economy nothing of that sort is intended, but all maxima are desired at once by various participants” (von Neumann and Morgenstern 1944, 11; paragraph break in the original omitted here).
Classical Utilitarianism before John Stuart Mill
89
experiment is more convincing as an illustration of why maximizing the total is not feasible without optimizing the distribution than it is as a demonstration that optimizing the distribution somehow takes care of maximizing the total. Be that as it may, he does not push the thought experiment very far, deciding to abandon the expression altogether instead of insisting that the mathematical solution contemplated, that is, the quantitative analysis presented, uniquely satisfies the formula of “the greatest happiness of the greatest number” in a way that clarifies its meaning and demonstrates its possibility beyond a reasonable doubt. His reservations in that regard reflect a correction hinted at earlier in the same work, specifically in a passage (already quoted in part in a different context earlier in the present chapter) where Bentham describes himself, in the third person, as developing misgivings upon further consideration, following a period of infatuation with the expression: At sight of it [the phrase “the greatest happiness of the greatest number”] he cried out as it were in an inward ecstacy [sic] like Archimedes on the discovery of the fundamental principle of Hydrostatics, Ευρηκα. Little did he think of the correction which within these few years on a closer scrutiny he found himself under the necessity of applying to it. bentham 1829/1983, 292; cf. bowring 1834, 300; perronet thompson 1829, 259
Bentham was obviously frustrated with terminology throughout his career, first with the inscrutability of references to “the principle of utility,” remaining vague at best in the absence of elaboration, thus prompting the adoption of “the greatest happiness of the greatest number,” which nevertheless produced comparable frustration owing to its precarious reconciliation of maximization and distribution, gradually encouraging strategic simplification as “the greatest happiness principle,” thereafter used together with “the principle of utility,” as needed, to make sure that any references to utility were understood in connection with happiness. The clearest account of the entire process is in Perronet Thompson (1829), who skips some of the twists and turns to focus only on the most relevant highlights, all coming from Bentham’s manuscript (1829/1983) but here brought together in a concise overview not to be found either in Bentham or in Bowring (1834): But as all knowledge is only the accumulation of improvements, the very title of the principle in question was found susceptible of progressive melioration. Its first name, ‘the Principle of Utility,’ was defective in as much as it did not express the nature and extent of the utility
90
Chapter 1
intended; and the same objection extended to the terms ‘Utilitarian’ and ‘Utilitarianism.’ It may be useful to a thief to steal; but it is useful to the community at large, that men should not steal; and it was this last utility, and not the first, which was intended, but not expressed. Its next denomination was, the principle of ‘the Greatest Happiness of the Greatest Number.’ This was erroneous by superfluity; and was in fact attempting to say the same thing twice instead of once. Though nothing in the writing of any of the proposers supported such a construction, it was liable to be represented as maintaining, that if, for example, a nation was composed of a million of black men and a million and one of white, the white were justified in sacrificing as much as they pleased of the happiness of the million, for the sake of any increase that might be made to the happiness of the million and one. The latest improvement, therefore, of the philosopher, whose long life has been dedicated to the diffusion of the principle,—and of which the present Article has to boast of being the announcement and the organ,—is to dismiss the superfluous ‘greatest number,’ and declare that the just object of politics and morals, is simply ‘the greatest happiness.’ In this manner the magnificent proposition emerges clearly, and disentangled from its accessary. And the accessary proposition is, that the greatest aggregate of happiness must always include the happiness of the greatest number. For the greatest number must always be composed of those who individually possess a comparatively small portion of the good things of life; and if any thing is taken from one of these to give to the others, it is plain that what he loses in happiness, is greater than what the others gain. It is the mathematical assertion, that a quantity x is greater in comparison of a small quantity it is taken from, than of a large one it is added to. It is the avowal that half- a-crown is of more consequence to the porter that loses it, than to the Duke of Bedford who should chance to find it;—that a chief portion of the baseness of the rich man who seized the poor’s ewe lamb, consisted in taking what caused so much pater pain to the sufferer, than happiness to the receiver. perronet thompson 1829, 267–268
A nagging question in connection with Bentham’s terminological preferences is the persistence of residual references to “the greatest happiness of the greatest number” after the expression was unambiguously repudiated as a vehicle of explanation or elaboration in the service of utilitarian principles and standards. Why, for example, would Bentham still be talking about “the greatest happiness of the greatest number” in his Deontology (1834/1983, 166,
Classical Utilitarianism before John Stuart Mill
91
192, 213)—a posthumously published work, which should ipso facto contain his latest thoughts on the matter—if he had abandoned the expression earlier? The question is not as difficult as it may seem, despite the anomaly it appears to uncover. Given that the retirement of the expression is not mere speculation or interpretation but an established fact documented in Bentham’s own words, and given furthermore that the residual references are not an illusion or accident but repeated references expressed in unmistakable terms, as in the three passages cited in block quotations in section 1.2 above, to consider just a few examples (Bentham 1834/1983, 166, 192, 213), the evidence may seem to point to one final change of opinion on the part of Bentham prior to his death. Yet the scenario of a reversal of opinion is not consistent with the evidence. It fails on at least two counts: first in ignoring the special circumstances of the publication of Deontology, second in misconstruing the nature of Bentham’s apparent aversion to the offending expression. With respect to the first consideration, we must recall that the posthumous publication of Deontology was not a matter of printing and binding a finished manuscript, but an intricate process of bringing together scattered notes, not all of which may have been composed after Bentham abandoned the expression. It would not have been surprising, therefore, for some of his earlier references to “the greatest happiness of the greatest number” to creep into the final draft, which was not only edited by Bowring rather than by Bentham but also, in a very real sense, produced by Bowring rather than by Bentham, especially when interpreting “production” as “creation from existing matter.” With respect to the second consideration, what Bentham actually renounced was not his original commitment to “the greatest happiness of the greatest number” as an ideal, but rather his practice of employing the corresponding words as a formal expression or elucidation of the principle of utility, or of the greatest happiness principle. While he admittedly did come to think of that expression as wanting in clarity, he continued to consider the underlying sentiment perfectly reasonable in its proper interpretation as he took care to explain in the passages quoted above. His restraint was simply a matter of avoiding being misunderstood, regarding which there would have been no danger in his personal notes, save for an editorial oversight (or perhaps just reverent abstinence from undue editorial intervention) subsequently perpetuating a decisively abandoned habit. Bentham’s use of these terms interchangeably is why John Stuart Mill, as his disciple, switches freely between “the principle of utility” and “the greatest happiness principle” without any attempt to explain the variation except through occasional references to Bentham. He understandably refrains, in the meantime, from ever invoking the greatest happiness of the greatest number.
c hapter 2
The Nature and Function of Mill’s Principle of Utility This chapter lays out the essential nature, primary properties, and implicit functions of Mill’s principle of utility. Its aim, in short, is to establish what that principle is and what it is for. The reason why such a basic task, a presumably simple one at that, serves a genuine need and merits a separate chapter is partly that Mill himself never defines the principle of utility and partly that his various references and discussions point to decidedly different definitions. The purpose of this chapter is to examine Mill’s ostensibly definitive remarks in order to uncover and isolate his considered opinion on the matter and thus to arrive at a conclusive definition of the principle. It is important to proceed with the correct formulation from the outset because the principle of utility constitutes the unit of analysis throughout the extended analysis in subsequent chapters. Readers may find it more convenient, therefore, to start straightaway with the definition embraced in the present volume, subsequently adjudicating the gradual exposure of that choice to critical scrutiny in this chapter, instead of discovering the favored definition only upon the completion of the corresponding investigation. The position adopted and defended here, upon a careful consideration of the facts and a thorough comparison of the alternatives, all laid out in due course, is that Mill’s principle of utility, despite having various functions, is, in essence, the principle that happiness (pleasure and the absence of pain) is the only thing that is desirable as an end in itself.1 1 The parenthetical annotation explicating happiness as “pleasure and the absence of pain” in the main text above is not an interpretation of Mill’s intent but a confirmation of his very words. The clarification is straight from Utilitarianism: “By happiness is intended pleasure, and the absence of pain; by unhappiness, pain, and the privation of pleasure” (U2:2 CW10:210). What Mill means by the absence of pain is not an absolute exemption or protection from pain any more than what he means by the presence of pleasure is the pervasive persistence or total domination of pleasure. The sense of “pleasure and the absence of pain” is simply a positive balance of pleasure over pain: “If by happiness be meant a continuity of highly pleasurable excitement, it is evident enough that this is impossible. A state of exalted pleasure lasts only moments, or in some cases, and with some intermissions, hours or days, and is the occasional brilliant flash of enjoyment, not its permanent and steady flame. Of this the philosophers who have taught that happiness is the end of life were as fully aware as those who taunt them. The happiness which they meant was not a life of rapture; but
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_004
Nature and Function of Mill’s Principle of Utility
93
Philosophically insignificant terminological variations are all equally acceptable, whereby happiness is the only thing that is desirable for its own sake, the only thing that is valuable in its own right, the only thing that is intrinsically good, and so on.2 The main rival to this position in any of its instantiations is the alternative that Mill’s principle of utility is the principle of morality that actions are right insofar as they tend to promote happiness, wrong insofar as they tend to do the opposite. This alternative draws strength from widespread agreement, at least among moral philosophers, that what makes a normative ethical theory utilitarian is not that it is hedonistic but that it is consequentialist.3 The common ground in contemporary utilitarian theories is not the conviction that happiness, or anything else in particular, is intrinsically good, and certainly not that happiness is the only thing that is intrinsically good, but that the moral status of actions is a function of their consequences, whatever the standard of evaluation may be. A utilitarian platform, on that interpretation, need not be anchored to happiness or pleasure. It may take any or several or all of various other ends as desirable in themselves, just as it may instead interpret the ultimate good generically as the satisfaction of desires, preferences, or interests. The implicit elasticity of the value dimension may even make it seem as if any principle serving as the definitive and foundational expression of utilitarianism would moments of such, in an existence made up of few and transitory pains, many and various pleasures, with a decided predominance of the active over the passive, and having as the foundation of the whole, not to expect more from life than it is capable of bestowing” (U2:12 CW10:215). 2 Such variations follow from Mill’s tendency to use the terms “desirable,” “valuable,” and “good” interchangeably in certain contexts. This is neither a novelty nor a problem. He does treat the terms as synonyms, but only where they actually work as synonyms, which is entirely consistent with standard lexical and semantic conventions. He certainly does not take them to be synonymous in every conceivable case (interchangeable salva veritate). He might, for example, say indiscriminately either that freedom is desirable, or that freedom is valuable, or that freedom is good. But he would speak of “good tidings” (or “glad tidings”) rather than “desirable tidings” or “valuable tidings,” although the meaning is the same. And he would bid folks a “good evening” rather than a “desirable evening” or a “valuable evening,” although the sentiment is the same. He would, in general, use all three terms in their colloquial senses, discriminating among them only where formal or informal conventions dictate a choice of one over the others. 3 The terms “utilitarianism” and “consequentialism” have long been practically synonymous in the terminology of contemporary philosophers. Fred R. Berger, for example, warns readers early on that he is “using the term utilitarianism in the broad manner that has become standard among some philosophers, namely, to designate any moral theory that takes consequences (of acts, rules, and so on) as the criterion of right and wrong” (1984, 5).
94
Chapter 2
be, and indeed should be, formulated primarily, if not solely, in terms of moral obligation, and thereby in reference to the moral evaluation and justification of actions.4 However that may be, a consequentialist focus alone does not adequately capture how Mill conceives of the principle of utility. He is a consequentialist, to be sure, but he regards consequentialism as the only sensible choice for a normative ethical theory, where utilitarianism then stands out with its emphasis on the promotion of happiness as the standard of distinction between right and wrong, because of its inherently clear and patently valid distinction between good and bad as the arbiter of value. This is confirmed, among other places, in the following passage from Mill’s (1838) review of the first few parts of a compilation then in progress of Bentham’s works (The Works of Jeremy Bentham): That the morality of actions depends on the consequences which they tend to produce, is the doctrine of rational persons of all schools; that the good or evil of those consequences is measured solely by pleasure or pain, is all of the doctrine of the school of utility, which is peculiar to it. mill CW10:111
Hence, even if reference to happiness as the standard of moral evaluation is an incidental characteristic of utilitarian theories, where the common denominator is moral appraisal in terms of consequences, there can be no doubt that emphasis on happiness as the only thing that is desirable as an end in itself is an essential characteristic of Mill’s utilitarianism. The fact that he repeatedly calls the principle of utility the greatest happiness principle is a direct reflection of the primacy of the association between happiness and utilitarianism.5
4 This is a tempting alternative supported by prima facie evidence in primary sources. The following statement, quoted and analyzed later in its fuller context, is just one example: “The creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle, holds that actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” (U2:2 CW10:210). This is arguably the strongest indication against taking the principle of utility to be first and foremost about the summum bonum. Yet the passage quoted is not a definition of the principle of utility but an elucidation of its application as the supreme principle of morality. Mill is here elaborating on the moral evaluation of actions in accordance with the creed holding “the greatest happiness principle” to be the foundation of morals, which makes his statement a corollary of the principle rather than a definition of it. 5 Mill refers to “the principle of utility” as “the greatest happiness principle” on four separate occasions in Utilitarianism: once in the first chapter (U1:4 CW10:207), twice in the second chapter (U2:2 CW10:210; U2:10 CW10:214), and once in the fifth chapter (U5:36 CW10:257).
Nature and Function of Mill’s Principle of Utility
95
The formal definition proposed above in terms of happiness is, in fact, little more than a paraphrase of Mill’s own words in Utilitarianism: The utilitarian doctrine is, that happiness is desirable, and the only thing desirable, as an end; all other things being only desirable as means to that end. mill U4:2 CW10:234
Readers who readily agree with my identification of this hedonistic value proposition as Mill’s true conception of the principle of utility may even find it pointless to consider an alternative definition and misguided to accept one. Those who disagree, on the other hand, might find it unorthodox, to say the least, to define the principle of utility in terms of intrinsic value as opposed to moral obligation or ethical justification. There are good reasons for interpreting the principle in any one of these ways, and perhaps also in some others, but the evidence ultimately favors a definition in terms of the intrinsic value of happiness. The remainder of the chapter is devoted to the establishment of this position. The first section, drawing on the conclusions of the previous chapter, considers the matter from the standpoint of Bentham in order to delineate the interpretive paradigm available to Mill as he develops his own conception. The second section considers the contextual difficulties in identifying, or recognizing, and therefore in defining, Mill’s principle of utility. The third section distinguishes between three different functions of the principle, often invoked as three different interpretations, each taken to support a different definition. The fourth section examines relevant passages in primary sources, mostly in Utilitarianism, that support one or another of the three interpretations. The fifth section shows that the primary function of Mill’s principle of utility is to establish the utilitarian theory of value as its most basic truth, thereby confirming the hedonistic affirmation that happiness is the only thing that is desirable as an end in itself. The sixth section presents a preliminary outline of Mill’s proof, mapping out its course, paragraph by paragraph, in the fourth chapter of Utilitarianism.
The first is in reference to Bentham’s usage of the term, but the subsequent repetitions confirm that Mill had adopted the terminology as his own.
96 2.1
Chapter 2
What Is Bentham’s Principle of Utility?
A promising approach to Mill’s principle of utility is to start with Bentham’s principle of utility. Any familiarity with Bentham affords some insight into Mill. Of course, we cannot reasonably assume, especially not without due consideration and diligent investigation, that Mill’s conception of the principle is precisely the same as Bentham’s understanding of the principle. But it is, after all, basically the same principle regardless of individual perspectives and interpretive tendencies. It can only help to observe the similarities and differences between how its most prominent proponents conceived of the principle, with such a comparison proving particularly useful where there are any questions as to exactly how either of them actually did. Bentham conceives of the principle of utility, under any name and in any formulation, as a normative guide to action, that is, to moral action, thus as a principle of distinction between right and wrong. He obviously also regards it as the foundational expression of the utilitarian theory of value, hence as a standard of distinction between good and bad, but the primary function he assigns to it is in the context of moral obligation and ethical justification. A detailed discussion of the textual evidence and the historical background is available in the previous chapter (sections 1.2 and 1.6 there being particularly relevant here). The principle of utility immediately and unequivocally makes its debut as the standard of morality as early as Bentham’s preface to A Fragment on Government (1776), where it claims the limelight as an axiomatic truth (“fundamental axiom”): “It is the greatest happiness of the greatest number that is the measure of right and wrong” (p. ii). This skeletal formulation finds its clearest expression and fullest elaboration in An Introduction to the Principles of Morals and Legislation (1789): “The principle of utility is the foundation of the present work: it will be proper therefore at the outset to give an explicit and determinate account of what is meant by it. By the principle of utility is meant that principle which approves or disapproves of every action whatsoever, according to the tendency which it appears to have to augment or diminish the happiness of the party whose interest is in question: or, what is the same thing in other words, to promote or to oppose that happiness” (p. 2). The essential nature and main function of the principle as the normative benchmark of moral obligation and the methodological touchstone of ethical justification is reaffirmed in both versions of Bentham’s “Article on Utilitarianism” (1829/1983). The long version singles it out as “the only principle the observance of which affords any promise of being conducive to the maximum of the quantity of happiness in the community to which application
Nature and Function of Mill’s Principle of Utility
97
is made of it” and as “the only self-proffered guide to human conduct in every walk, public as well as private, of human conduct” (p. 312). The short version describes it as “an instrument of direction: or, say, in more familiar language, a direction for pointing out the path most proper to be pursued, on every occasion, in public as well as private life; to be pursued by every individual, whether acting in his private and individual capacity, as a member of the whole community, for his own benefit alone, or in his public capacity acting for the benefit of others in the character of a member of the governing part of the community, in the exercise of the powers belonging to him in that same character” (p. 320). The same version goes on to distinguish between three separate uses, or “three distinguishable capacities”: “1. as end in view; 2. as a storehouse of means employable for the attainment of that end; 3. as a storehouse furnishing motives by the force of which, on the several occasions, men may be induced to act in ways conducive to that end” (p. 320). The features and functions of the principle explicated in Bentham’s various works are corroborated in his posthumously published Deontology (1834/1983), which describes it as (1) “that principle which for the test of propriety in every act takes its conduciveness or repugnancy to the greatest known happiness of the greatest number of mankind” (p. 166), (2) “the standard of right and wrong” grounded in “the property of producing pleasure or preventing pain” (p. 188), and (3) “the principle which in the case of every action establishes as the ground and measure of its propriety its conduciveness or repugnance to the greatest happiness of the greatest number” (p. 192). The book thus declares the principle of utility “the true and only standard, and conduciveness or opposition to that end the only test of right and wrong” (p. 213). These references are complemented by correlative expressions among the marginalia in A Table of the Springs of Action (Bentham 1817/1983), most notably including the following: ## 231, 415, 512, 655, 658, 697.6 To quote only a couple of these entries, the first and last one together cover the overall message: “231. Utilitarianism—Utilitarian philosophy—Utilitarian principle of utility: ‘Act according to the greatest happiness of the greatest number’ ” (p. 25). “697. Principle of utility, what. 1. Indication of what ought to be, it indicates as the only universally desirable object and end, greatest happiness of greatest number” (p. 62). The combined evidence shows that Bentham duly recognizes the principle of utility as the fundamental expression of the utilitarian theory of value, while 6 The numbers provided here for marginalia are item numbers rather than page numbers. There are multiple entries on each page, which makes item numbers more distinctive, and therefore more convenient, compared to page numbers.
98
Chapter 2
focusing mostly on its role as the central paradigm of moral obligation and the ultimate criterion of ethical justification. This approach is the opposite of the emphasis in Mill, at least in his Utilitarianism, and particularly in his proof of the principle of utility, as demonstrated over the next few sections. 2.2
What Is Mill’s Principle of Utility?
What is Mill’s principle of utility? This is an easy question for college students, a difficult one for moral philosophers. Nearly everyone exposed to ethical theory in college knows the answer. The question often comes up in introductory philosophy classes, invariably so in ethics courses. The response would seem to be a matter of looking it up in Utilitarianism in the manner of looking up a definition in the dictionary. The question, on the other hand, also happens to be the title of an illuminating essay by Donald George Brown (1973), who examines fifteen different formulations of Mill’s principle of utility before settling on one of them as the definitive version.7 The mere fact that Brown’s essay fills a void in the literature bears out the difficulty of defining the principle. Instructors of elementary courses in philosophy and ethics regularly turn to Bentham and Mill, more often specifically to Mill’s Utilitarianism, in introducing classical utilitarianism as a normative ethical theory. The goal of an initial lecture on utilitarianism is typically to introduce the principle of utility as the theoretical core that helps distinguish right from wrong, often in corroboration of prevailing precepts of common morality, be it through the direct evaluation of actions or through the construction and application of derivative moral rules. The center stage in such a presentation is usually reserved for the notion of the greatest happiness of the greatest number (see sections 1.2 and 1.6 of the previous chapter) as the moral yardstick for evaluating moral problems, determining moral obligations, and assessing the morality of actions. Attentive students gather from the lecture that Mill’s principle of utility states that we ought to act so as to promote the greatest happiness of the greatest number. These students, who then read Utilitarianism, are frustrated to find that Mill himself does not define or express the principle in the exact terms used in class. A basic source of such frustration is that Mill frequently and indiscriminately uses several different terms and expressions, not to mention elegant variations of each, to refer to the principle in question:
7 Donald George Brown settles on the definition that “happiness is the only thing desirable as an end” (1973, 1, 5, 9, 12).
Nature and Function of Mill’s Principle of Utility
99
– “the principle of utility” (U1:4 CW10:207; U3:3 CW10:228; U4:9 CW10:237; U5:36n CW10:258); – “the greatest happiness principle” (U1:4 CW10:207; U2:2 CW10:210; U2:10 CW10:214; U5:36 CW10:257); – “the theory of utility” (U2:1 CW10:209); – “the doctrine of utility” (U2:22 CW10:222); – “the utilitarian doctrine” (U4:2 CW10:234). A derivative source of frustration is that the different terms Mill uses are not technical locutions that mark pertinent distinctions. Nor are the various explications he offers correlated with any one expression more than any other. It turns out that the principle in question should be called either “the principle of utility” or “the greatest happiness principle” since these are the terms Mill uses most often and since he uses them interchangeably in close textual proximity, usually within a single paragraph and sometimes within the same sentence (U1:4 CW10:207). Apparently, his favorite is “the principle of utility,” which is the name he assigns to the principle in the title of the third and fourth chapters of Utilitarianism (CW10:227, 234). Despite the absence of consensus on a definitive formulation of Mill’s principle of utility, scholars generally agree that he takes happiness to be the highest good, defines “happiness” as “pleasure and the absence of pain,” and sets up the greatest happiness of the greatest number as the standard of morality. Mill himself would very likely insist on adding several caveats to these points of scholarly agreement on the fundamental elements of the principle of utility. The first one on his list would probably be his distinction between higher pleasures, corresponding to cognitive and emotional experiences, and lower pleasures, corresponding to bodily sensations (U2:3–8 CW10:210–213). Devising a hypothetical empirical test to elucidate and justify this distinction, Mill submits that competent judges, basically ordinary people with experience of both higher and lower pleasures, would always prefer the higher over the lower. He holds, in fact, that competent judges would, given a choice, pick any amount of a higher pleasure, hence even the slightest such experience, over any amount of a lower pleasure, thus precluding the possibility of a conflict between quality and quantity. Early criticisms of Mill’s qualitative hedonism converge around the apparent inconsistency of holding both that pleasure net of pain is the only thing desirable in itself and that pleasures differ in quality as well as quantity.8 The 8 The most prominent scholars among Mill’s early critics addressing his distinction between higher and lower pleasures include: Francis Herbert Bradley (1876, 105–111; 1927, 116–122); Thomas Hill Green (1883, 168–174); George Edward Moore (1903, §§ 3:47–48, pp. 129–132); Henry Sidgwick (1874, 76–79, 104, 109, 114–117; 1907, 93–95, 121, 125–130); William Ritchie Sorley (1885, 58–60; 1904, 61–63).
100
Chapter 2
basic objection is that, if equal amounts of pleasure are not equally desirable, valuable, or good, then pleasure is not the only thing that is intrinsically desirable, valuable, or good. The rationale typically adduced for the disparity is that, if the quality of the pleasure makes a difference, then the quantity of it cannot be all that matters, which is the same as saying, so the objection goes, that pleasure itself is not all that matters. Contemporary accounts focus on a greater variety of problems than this traditional objection.9 They examine not just the conceptual viability of Mill’s distinction between kinds of pleasure but also the methodological propriety of his appeal to competent judges to establish that distinction. After more than a full century and a half, we are nowhere near an agreement on whether the distinction works, though no one denies, at least not reasonably so, that it is an integral part of Mill’s principle of utility. The distinction may not be explicit in the core formulation of the principle, even if we can agree on what that formulation might be, but it must certainly be considered implicit in any formulation, even if we cannot agree on anything else. Other caveats of importance Mill might be inclined to add to a working definition of the principle of utility would naturally include provisions that he introduces in his own discussion of utilitarianism: – We need not calculate the effects of actions on absolutely everyone but on everyone immediately involved in the relevant context and likely to be directly affected by the action(s) in question (U2:19 CW10:219–220). – Utilitarianism does not require the prediction and calculation of the consequences of each particular action prior to performing that action (U2:24 CW10:224–225). – Principles of justice as well as a fundamental respect for moral rights are implicitly inherent in the principle of utility (U5:25 CW10:250–251; U5:32–38 CW10:255–259).10 9
10
Later commentators (as opposed to early critics) on Mill’s distinction between higher and lower pleasures include: Elizabeth S. Anderson (1991); Gustaf Arrhenius and Wlodek Rabinowicz (2005); Fred R. Berger (1984, 31, 37–40); David Owen Brink (1992; 2013, 46– 78); Richard N. Bronaugh (1974); Roger Crisp and Morten Kringelbach (2018); Norman O. Dahl (1973); Wendy Donner (1991, 37–65); Lanny Ebenstein (1985); Benjamin Gibbs (1986); Steven D. Hales (2007, 97–109); Michael Hauskeller (2011); Robert W. Hoag (1992); Terence Henry Irwin (2009, 399–404); Roderick T. Long (1992); Rex Martin (1972); Dale E. Miller (2010b, 54–70); Anthony Meredith Quinton (1973, 39–43); Jonathan Riley (1993; 1999; 2003; 2008a; 2008b; 2009; 2014); Jesper Ryberg (2002); Steve F. Sapontzis (2014); Ben Saunders (2010); Geoffrey Scarre (1997); Kristin Schaupp (2013); Christoph Schmidt–Petri (2003; 2006); John Skorupski (1989, 303–307); Mark Philip Strasser (1991, 1–22); Henry Robison West (1976b; 2004, 48–73). The objection that utilitarianism cannot accommodate justice and moral rights was one of the major criticisms of the theory in Mill’s day, and it continues to be one of the most
Nature and Function of Mill’s Principle of Utility
101
Mill would thus qualify a central and rudimentary formulation of the principle of utility with his responses to popular objections at the time, most of which he examines in the second chapter of Utilitarianism, subsequently adding other pertinent clarifications of the kind he thinks “might be written under the principle of utility as an explanatory commentary” (U5:36 CW10:257). A full exposition and complete analysis of the principle of utility requires, in addition to the preceding points of agreement and the corresponding qualifications, the clarification of issues that Mill does not address explicitly even though they seem pertinent to an understanding of utilitarianism. One such issue that may come up is whether the principle of utility treats the happiness in question as the greatest happiness of the greatest number. The reason why this can be a problem is that “the greatest happiness of the greatest number” is a vague concept, which Bentham, as its most fervent advocate, eventually abandoned for want of clarity (see section 1.6 of chapter 1), evidently influencing Mill’s predilections as well, judging by the complete absence of any reference to the expression throughout Utilitarianism. One may reasonably object that the concept of “the general happiness” employed by Mill is just as ill-defined as that of “the greatest happiness of the greatest number” abandoned by Bentham and avoided by Mill. One may even object, though not without being a little unreasonable, or at least somewhat presumptuous, that “the general happiness” is nothing but “the greatest happiness of the greatest number” in disguise. While the precise nature of Mill’s “general happiness” is indeed a legitimate concern, receiving ample attention in chapter 5 of the present volume and coming up at length in various other places, the question of its
important objections today, at least against some versions of utilitarianism. The basic objection is that the promotion of happiness is not necessarily consistent either with the demands of justice or with the requirements of moral rights. Mill is quite sensitive to this issue, attempting eagerly to demonstrate that an implicit sense of respect both for justice and for moral rights is an integral part of the principle of utility (U5:25 CW10:250–251; U5:32–38 CW10:255–259). The question, however, is whether the attempt is satisfactory, in other words, whether utilitarianism really does make sufficient allowances for justice and moral rights. The following is a partial list of commentators who examine this question: Fred R. Berger (1984, 123–225, 289–292); John Gray (1981; 1983, 63–69; 1991); Jonathan Harrison (1975); Terence Henry Irwin (2009, 418–422); David Lyons (1978); Anthony Meredith Quinton (1973, 71–81); Alan James Ryan (1970, 213–231); Mark Philip Strasser (1991, 223–272).
102
Chapter 2
relationship, if any, with “the greatest happiness of the greatest number” is a moot issue, because neither concept has an established meaning suitable for comparison and contrast. It is fruitless to object that a highly vague concept is identical with another highly vague concept, at least one of which happens to be problematic, and a waste of time to try to prove in response that neither one has anything to do with the other. The possibility of their identity, affinity, or congruity cannot be determined or demonstrated beyond a reasonable doubt, any more than it can be ruled out with any degree of confidence. The question of commensurability can therefore be safely set aside, both from the perspective of Bentham, who expressly repudiates “the greatest happiness of the greatest number” (Bentham 1829/1983, 309–311; cf. Bowring 1834, 328–330; Perronet Thompson 1829, 267–268), settling instead on more meaningful formulations, such as “the greatest happiness of all those whose interest is in question” (Bentham 1823, 45), “the aggregate stock of the happiness of the community” (Bentham 1829/1983, 309), “the greatest happiness of all” (Bentham 1843, vol. 9, 6),11 or simply “the greatest happiness,” and from the perspective of Mill, who never even mentions “the greatest happiness of the greatest number,” working instead with “the general happiness.” Although chapter 5, as intimated above, explicates Mill’s conception of the general happiness insofar as it constitutes an integral part of his proof of the principle of utility, additional questions may arise in connection with any other aspect of the general happiness or in reference to the principle of utility itself. Several such questions are answered in chapter 8, which is dedicated in part (section 8.2) to some of the more important issues on which Mill does not express a clear opinion or adopt an obvious position. Significant points of inquiry requiring clarification include, but are not limited to, the question whether the promotion of happiness is the standard of moral evaluation and justification at the level of actions or at the level of moral rules and principles (subsection 8.2.1); the question whether the morally relevant consequences of actions are their actual consequences, intended consequences, or foreseeable consequences (subsection 8.2.2); and the question whether the general happiness is a simple sum or an arithmetic mean or something altogether different (subsection 8.2.3). Another issue that often comes up in discussions of Mill’s moral philosophy is the question of supererogation.12 The question, to be more specific, is 11 12
The reference here is to the posthumously published edition of Constitutional Code (Bentham 1820–1832) appearing in The Works of Jeremy Bentham (Bentham 1843, vol. 9, 6). The relevant literature on supererogation is vast. The pioneers of discussion specifically in connection with Mill include: Donald George Brown (1982); David Copp (1979); Douglas
Nature and Function of Mill’s Principle of Utility
103
whether or not the principle of utility requires the performance of moral acts above and beyond the call of duty. The short answer is that it does not, simply because it cannot. The long answer is that, if the principle of utility is the principle of distinction between right and wrong, and thereby the ultimate court of appeal in moral obligation, then the principle does not, and, in fact, cannot, require the performance of supererogatory acts, because whatever it does require would, for that very reason, be one’s duty, which would automatically preclude the possibility of its requiring anything above and beyond the call of duty. Yet neither the short answer nor the long answer is entirely satisfactory. Supererogation in the present context is about whether the principle of utility either holds or implies that the performance of an action is morally obligatory if performing that action produces the greatest balance of pleasure over pain among the possibilities open to the agent, the alternative being that the performance of an action is morally permissible insofar as performing that action produces a greater balance of pleasure over pain than does not performing that action, regardless of what other courses of action are open to the agent, including any with superior consequences in terms of the promotion of happiness. The discussion of supererogation in Mill’s moral theory is thus motivated by a quest to determine whether Mill is a maximizing utilitarian, who embraces the greatest possible happiness as a compelling standard of moral action, or a satisficing utilitarian, who prefers conduciveness to happiness as a sufficient standard of demarcation between morally permissible and impermissible actions. The problem is that the maximization of happiness as a strict obligation leaves no room for a range of morally permissible alternatives. If it is morally obligatory to maximize happiness in every situation that is morally charged, then it is morally impermissible not to maximize happiness in any such situation. To be clear, there is no question where Mill stands on the matter of supererogation. He acknowledges, beyond any doubt and leaving no room for interpretation, that there are some acts that are not morally obligatory even though they are morally laudatory. The question, rather, is whether the nature of the principle of utility nevertheless commits Mill to designating any act as morally obligatory so long as it promises a net increase in happiness, where any conflict between competing alternatives is resolved strictly in favor of the act promising the greater return on happiness.
Poole Dryer (1969; 1979); Roger Hancock (1975); David Lyons (1976; 1978; 1979); Leonard Wayne Sumner (1979); James Opie Urmson (1958).
104
Chapter 2
As for Mill’s explicit position, he asserts, in no uncertain terms, that supererogatory acts are not required: “There are other things, on the contrary, which we wish that people should do, which we like or admire them for doing, perhaps dislike or despise them for not doing, but yet admit that they are not bound to do; it is not a case of moral obligation” (U5:14 CW10:246). He clearly and distinctly recognizes the difference between what we must do and what we may do: “We say that it would be right to do so and so, or merely that it would be desirable or laudable” (U5:14 CW10:246). While both of these quotations are from a single paragraph in the fifth chapter of Utilitarianism, they are not isolated remarks. The same distinction comes up elsewhere, for example, in Auguste Comte and Positivism (1865), demonstrating that Mill is both consistent and persistent in his acknowledgment of supererogation as a separate moral category: “There is a standard of altruism to which all should be required to come up, and a degree beyond it which is not obligatory, but meritorious” (CW10:337). Even his personal correspondence attests to his insight into the matter, as illustrated in a letter to George Grote (Letter 525: 10 January 1862): “recognising the merit, though not the duty, of making still greater sacrifices of our own less good to the greater good of others, than the general conditions of human happiness render it expedient to prescribe” (CW15:762). The nagging question, however, is whether the model of moral evaluation and justification associated with the principle of utility is as sensitive to supererogation as Mill himself seems to be. Not all such questions, nor any number of others in the course of discussion, pick out actual or potential problems that are overlooked by Mill. Most reflect matters that are not neglected by Mill but placed on his doorstep in retrospect. The distinction between act and rule utilitarianism is a prime example. Successive iterations of utilitarianism, especially the more popular refinements, whether or not they actually qualify as improvements upon classical utilitarianism, tend to find their way back into Bentham or Mill through claims that the developments in question are either implicit in or consistent with their works. The ongoing quest to formulate the most plausible version of utilitarianism thus runs in tandem with parallel efforts to locate some anticipation of each development in utilitarian theory in the corpus of Bentham or Mill. An inordinate amount of scholarly reflection in this respect is indeed directed at determining whether Mill is an act utilitarian or a rule utilitarian, as the central issue is reshaped periodically through adjustments in the continuing debate over which of these two versions of utilitarianism is more tenable, not to mention discussions examining whether the two versions amount to the same thing in the final analysis. Strictly speaking, Mill is neither an act utilitarian nor a rule utilitarian, since he would have otherwise noted the difference
Nature and Function of Mill’s Principle of Utility
105
and adopted a position, even in the absence of the current terminology for the corresponding distinction. Any effort to place him in one camp or the other is a matter of digging up passages that are more consistent with one approach than with the other, while pretending that passages suggesting the opposite do not matter, either because they are irrelevant or because they are misunderstood. Yet the debate has become such an established part of the literature on utilitarianism that it requires at least acknowledgment, and perhaps some basic orientation, making it useful if not necessary to comment on the general highlights (see subsection 8.2.1). Another development in utilitarian theory that Mill apparently does not anticipate in any verifiable way is preference utilitarianism. Commentators have long been attempting to show that Mill’s universalistic ethical hedonism is open to, or at least consistent with, desire-satisfaction models of preference utilitarianism (Barrow 1991; Brown 1973; Zinkernagel 1952). Such attempts are usually grounded in the premise that Mill would take whatever satisfies desires to be pleasurable in some way and therefore to be conducive to happiness. He seems open to this interpretation where he takes “happiness” and “interests” as practically interchangeable references to the unit of morality: “the happiness, or (as speaking practically it may be called) the interest, of every individual” (U2:18 CW10:218). The apparent synonymy is also consistent with his conception of happiness, which is general enough to support desire-satisfaction models of preference utilitarianism, given that he considers all other goods to be either a means to happiness or a part of happiness (U4:4–8 CW10:234–237). Yet it is all too easy to find anachronistic evidence in the absence of an explicit contradiction, which is hardly convincing as a conclusive demonstration. The most promising and most cumbersome approach to understanding Mill’s principle of utility is to examine his every comment in Utilitarianism that helps explicate the principle of utility. The main problem with that approach is the difficulty of distinguishing between a concise statement of the principle that is not open to glaring difficulties and a full explication of it with pertinent qualifications and appendages, even if such qualifications and appendages are limited to those that Mill himself provides. A sufficiently promising and less cumbersome approach is to deduce the essence of Mill’s principle of utility from his proof of the principle of utility. If Mill is out to prove the principle of utility, then he must reveal or indicate what it is that he will attempt to prove before he actually proves it. Or he must, at the very least, disclose what he thinks he has proven after he has proven it. He may even be expected to do both, as much as he revels in writing. It is helpful to keep in mind, however, that he might conceive of the principle of utility as having multiple features and functions, only one of which requires proof in
106
Chapter 2
any sense, which is liable therefore to render conceptual inferences from the proof incomplete. Another approach is to examine secondary sources among commentators purporting to have studied and overcome the obstacles to a precise definition of Mill’s principle of utility. These are usually Mill scholars who have already done the legwork involved in the first two approaches. A parasitic survey of that sort might be useful as a supplement to other approaches, but it cannot stand alone as if understanding Mill’s principle of utility were a matter of taking a poll among the readers of Mill, or even among the scholars of Mill. 2.3
The Multiple Functions of Mill’s Principle of Utility
A major obstacle to reliable insight into Mill’s principle of utility is that Mill himself does not offer anything like a proper definition of it. He does say quite a bit about it, but he does not say exactly what it is. An even greater obstacle is that Mill’s discussion of the principle seems to support jointly inconsistent definitions of it. For example, the principle of utility appears to be a theory of moral obligation and a theory of ethical justification in the famous passage where Mill articulates “the creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle” (U2:2 CW10:210), while it comes across as a theory of value toward the end of the same paragraph where he discusses “the theory of life on which this theory of morality is grounded” (U2:2 CW10:210). As a matter of fact, Mill’s most concise expository paragraph on “the Greatest Happiness Principle” imputes all three roles to the principle of utility, as it holds that “the end of human action, is necessarily also the standard of morality” (U2:10 CW10:214). It is regrettable that the words, “This is the principle of utility,” followed immediately by a colon and a definition, are nowhere to be found in Mill’s published writings. They should have been available at least in Utilitarianism, ideally in the long chapter titled “What Utilitarianism Is.” The absence of a formal definition may seem like a trivial worry. A formal definition may even appear to be an overly demanding expectation from a work preceding the development of analytic philosophy. However, in the absence of such a definition, it remains unclear whether Mill’s principle of utility is a theory of value, a theory of moral obligation, or a theory of ethical justification (of actions and of derivative moral principles or secondary moral rules designed to guide actions). The principle obviously serves all three ends, but it would help to know, or to work out, which of its functions is primary, what the pecking order is, and how they are related. We may think of Mill’s principle of utility (pu) as a
Nature and Function of Mill’s Principle of Utility
107
general truth (or “fundamental axiom” as it was put by Bentham 1776, ii) at the intersection of three theories, or theoretical areas of concern, common and peculiar to moral discourse:
(PU1) Theory of Value: Happiness is the only thing that is desirable as an end in itself, good for its own sake, valuable in its own right, and so on, with “desirable,” “good,” and “valuable” used interchangeably in reference to intrinsic value. (PU2) Theory of Obligation: We ought to act so as to promote the greatest happiness of the greatest number of everyone concerned in the relevant context. (PU3) Theory of Justification: Actions and rules of action, including common moral codes and derivative moral principles, are justified insofar as they tend to promote happiness, unjustified insofar as they tend to do the opposite.
Synoptic and interpretive systematization is naturally incomplete as a process, and always open to objection as a conclusion, particularly with respect to the position purported to be summarized and interpreted. The breakdown above is no exception. Although it would defeat the purpose of the systematization to include copious details with a view toward completeness, or to anticipate all objections in an effort to prevent them from arising in the first place, some clarification should prove useful in avoiding a potential misunderstanding concerning each of the three functions envisaged. First, an explanation is in order with respect to terminology: While PU2, designated as a “theory of obligation,” is a theory of moral obligation, and PU3, designated as a “theory of justification,” is a theory of ethical justification, PU1, designated as a “theory of value,” is not specifically a theory of moral value. Moral value is associated strictly with virtue and vice, whereas PU1 is about value in general and intrinsic value in particular. PU1 may even be considered a “theory of nonmoral value,” following William Klaas Frankena (1963), or better yet, a “general theory of value,” following Ralph Barton Perry (1926), but it cannot accurately be called a “theory of moral value,” at least not exclusively so. Second, it must be noted that a deliberate omission in PU2 renders it incomplete as a statement of Mill’s theory of moral obligation. A complete statement of that would have to include, or otherwise deal with, Mill’s conviction that actions are morally right or wrong depending on whether their performance merits the punishment of the performing agent in some way or other (U5:14–15 CW10:246–247). Both the promotion of happiness and the desert of punishment are essential ingredients in Mill’s theory of moral obligation. PU2 ignores
108
Chapter 2
the idea of penal sanction because the aim here is to examine the principle of utility in its core formulations, independently of appendages, corollaries, and auxiliary hypotheses. Third, an apparent problem with PU3 is that it seems to imply or contain PU2 (though PU2 does not seem to imply or contain PU3), thus blurring the difference between these two functions, aspects, or renditions of the principle of utility. Put simply, if the promotion of happiness is the standard of ethical justification, and moral obligation is grounded in the moral evaluation of actions, which is a kind of ethical justification, then the promotion of happiness is also the standard of moral obligation. This is not a problem in itself, but it does make it difficult to identify passages in Mill’s writings that fall under PU2 but not under PU3, or that fall under PU3 but not under PU2, and consequently to distinguish between passages that fall under PU2 and those that fall under PU3. All three functions of Mill’s principle of utility come up immediately in Utilitarianism. The very first sentence of the book introduces the impetus for the entire work as “the decision of the controversy respecting the criterion of right and wrong” (U1:1 CW10:205). Since the criterion of right and wrong is the domain of PU3, ethical justification, it is also, at least by implication, the domain of PU2, moral obligation. The second sentence of the book identifies the question “concerning the foundation of morality” with “the question concerning the summum bonum” (U1:1 CW10:205). The question concerning the summum bonum is naturally the domain of PU1, value. The first two sentences of Utilitarianism, then, jointly announce that the work concerns value, obligation, and justification in moral discourse.13 Even if this interpretation is correct, however, these two sentences do not confirm that the principle of utility has three separate functions as the first principle of Mill’s theories of value, obligation, and justification in ethics. They indicate only that Mill is deeply concerned with all three subjects. Yet there is further evidence that indeed establishes these three dimensions as distinct functions of the principle of utility.
13
Not all commentators agree with this interpretation of the first two sentences of Utilitarianism. Some maintain, on the contrary, that Mill’s apparently random juxtaposition of value, obligation, and justification in the first two sentences of Utilitarianism (and elsewhere) shows that he was confused from the beginning with respect to the actual differences and connections. Fred R. Berger (1984, 112–114), for one, without attributing any confusion to Mill, holds that he does not distinguish between value and obligation, ultimately because, for Mill, “the value judgment and the principle of conduct would be two aspects of one principle” (1984, 114).
Nature and Function of Mill’s Principle of Utility
109
2.3.1 PU1: Theory of Value PU1 is at the forefront wherever Mill invokes the principle of utility in the context of ultimate ends or first principles, typically while elaborating on utilitarianism specifically as a theory of value. The fourth chapter of Utilitarianism, for example, asserts early on that questions about ends are questions about the good, thus emphasizing the concern of the principle of utility with questions of value: “Questions about ends are, in other words, questions what things are desirable” (U4:2 CW10:234). The basic connection suggested here by the desirability of ends is confirmed in A System of Logic, where first principles are identified with value: “Every art has one first principle, or general major premise, not borrowed from science; that which enunciates the object aimed at, and affirms it to be a desirable object” (L6:12.6 CW8:949). While the latter quotation does not expressly mention the principle of utility, the corresponding discussion does build up to it, including a footnote (n. 65) explicitly directing readers to Utilitarianism for further discussion and for vindication of the principle itself (L6:12.7 CW8:951): Without attempting in this place to justify my opinion, or even to define the kind of justification which it admits of, I merely declare my conviction, that the general principle to which all rules of practice ought to conform, and the test by which they should be tried, is that of conduciveness to the happiness of mankind, or rather, of all sentient beings: in other words, that the promotion of happiness is the ultimate principle of Teleology. mill L6:12.7 CW8:951
This is actually evidence of PU3, as discussed below, and not of PU1, but the point here is that the earlier reference to PU1 (L6:12.6 CW8:949), where the principle of utility is not positively identified, is indeed confirmation of the role of the principle of utility in value theory. References and quotations can be multiplied indefinitely, but there can be no doubt, no reasonable doubt anyway, that the principle of utility represents the utilitarian theory of value. 2.3.2 PU2: Theory of Obligation PU2 dominates the narrative where the principle of utility comes up as the relevant “standard” of morality, or as the central element in the corresponding “theory” of morality, whereby the principle emerges as the cornerstone of the utilitarian theory of obligation as distinct from, and in addition to, its role in the utilitarian theory of value.
110
Chapter 2
Even the relatively short first chapter of Utilitarianism affords strong evidence of a theory of moral obligation. One example is where the “principles of morals” are explicated as principles of “right and wrong” in the course of contrasting the “intuitive” and “inductive” schools of ethics in terms of the methods each one uses to explain and justify the principles each one espouses (U1:3 CW10:206). Another is where “the non-existence of an acknowledged first principle” in ethical theory is said to be mitigated by the fact that “the principle of utility, or as Bentham latterly called it, the greatest happiness principle,” serves as a moral guide even for those who are “unwilling to acknowledge it as the fundamental principle of morality, and the source of moral obligation” (U1:4 CW10:207). The second chapter of Utilitarianism contains unmistakable signs of a theory of moral obligation, beginning with the definitive declaration that “actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” (U2:2 CW10:210). This distinction between right and wrong is subsequently identified as the basis of moral obligation, with a reference to “Utility or Happiness, considered as the directive rule of human conduct” (U2:9 CW10:213). Corroborating evidence is available throughout the same chapter, for example, where the utilitarian “standard of morality” is discussed in terms of “the rules and precepts for human conduct” (U2:10 CW10:214), where the discussion turns to “the happiness which forms the utilitarian standard of what is right in conduct” (U2:18 CW10:218), or where the common objection crops up that the utilitarian standard is too high because it is “too much to require that people shall always act from the inducement of promoting the general interests of society” (U2:19 CW10:219). But the strongest emphasis on moral obligation comes in the last paragraph of this long chapter, where Mill affirms plainly that “utility is the ultimate source of moral obligations” (U2:25 CW10:226). With all the evidence clustered there, it is perhaps worth reiterating that the title of the second chapter is “What Utilitarianism Is” (U2 CW10:209). PU2 is the center of attention in various other places as well, including the opening of the fifth and final chapter with a direct reference to “the doctrine that Utility or Happiness is the criterion of right and wrong” (U5:1 CW10:240). What is most telling in that regard, however, is not a particular quotation from Utilitarianism but a thematic consideration with respect to the nature and structure of that work. If the principle of utility were merely the theory of value that happiness is the only thing that is desirable as an end in itself, then the matter of establishing sanctions for such a theory would have been a moot
Nature and Function of Mill’s Principle of Utility
111
issue, and thus, the entire third chapter, a superfluous project. What could it possibly mean to have sanctions for a theory of what is intrinsically desirable, valuable, or good? It is precisely because the principle of utility is a theory of obligation (as well as a theory of value) that the third chapter becomes necessary. 2.3.3 PU3: Theory of Justification PU3 receives its fair share of emphasis in Utilitarianism wherever the principle of utility is brought to bear on the possibility and process of ethical justification. The typical terminological clue in this respect is a reference to the principle of utility as some sort of test of right and wrong. Such emphasis can be found as early as the first chapter of the book, specifically in the midst of a discussion of methodology and metatheory in ethics, where Mill claims that a “test of right and wrong must be the means, one would think, of ascertaining what is right or wrong, and not a consequence of having already ascertained it” (U1:2 CW10:206). The same emphasis is present in the second chapter, which opens with an immediate affirmation of “utility as the test of right and wrong” (U2:1 CW10:209).14 PU3 also receives some attention in A System of Logic, where Mill declares outright, as mentioned and quoted above, that “the general principle to which all rules of practice ought to conform, and the test by which they should be tried, is that of conduciveness to the happiness of mankind, or rather, of all sentient beings” (L6:12.7 CW8:951). The next paragraph leaves no room for doubt that the principle of utility is the conceptual foundation of a theory of ethical justification: “I do not mean to assert that the promotion of happiness should be itself the end of all actions, or even of all rules of action. It is the justification, and ought to be the controller, of all ends, but is not itself the sole end” (L6:12.7 CW8:952).
14
The following passages from the second chapter of Utilitarianism lend additional support to the observation that the principle of utility functions as a theory of ethical justification: (1) utilitarianism involves making a “judgment respecting the rightness or wrongness of an action” (U2:20 CW10:221); (2) utilitarianism calls for deciding “an action to be good or bad” (U2:20 CW10:221); (3) utilitarianism assesses “the morality of actions, as measured by the utilitarian standard” (U2:21 CW10:221). Some or all of these may also be invoked in confirmation, or at least in support, of the function of the principle of utility as a theory of moral obligation.
112 2.4
Chapter 2
Mill’s Apparent Definitions of the Principle of Utility
The cluster of words that comes closest to making further speculation unnecessary regarding the correct definition of the principle of utility may be found in a passage well worth repeating, in the second paragraph of the fourth chapter of Utilitarianism, where Mill describes the corresponding doctrine as follows: The utilitarian doctrine is, that happiness is desirable, and the only thing desirable, as an end; all other things being only desirable as means to that end. mill U4:2 CW10:234
This is indeed the definition advocated in the present volume, but it is only fair to consider alternatives. Although the sentence begins with the words “The utilitarian doctrine is,” which suggests that Mill is defining the principle of utility, or at least that he is disclosing his conception of that principle, the matter is not that easy to lay to rest. One may be tempted to object that “the utilitarian doctrine” and “the principle of utility” are not necessarily the same thing. That is true, of course, but it is hardly the cause of the difficulty here. The problem is not that the quotation above is not a definition, nor that it is not a definition of the principle of utility, nor that Mill does not intend it as such a definition. The problem, rather, is that the assertion does not represent the whole story. If we were to take Mill’s every comment that resembles a definition of the principle of utility to be his actual definition of the principle of utility, we would end up with a series of mutually inconsistent definitions, each with a valid claim to authenticity. Another definition of sorts seems to be implicit, for example, in the following passage of the second chapter of Utilitarianism: Those who know anything about the matter are aware that every writer, from Epicurus to Bentham, who maintained the theory of utility, meant by it, not something to be contradistinguished from pleasure, but pleasure itself, together with the exemption from pain. mill U2:1 CW10:209
Yet the principle of utility, or rather “the theory of utility,” as the passage above has it, surely does not mean “pleasure itself, together with the exemption from pain.” That is not a definition of any kind of principle, let alone the principle of utility. A principle might state that “pleasure itself, together with the exemption from pain,” is a compelling moral consideration, that “pleasure itself, together with the exemption from pain,” is the highest good, or even that
Nature and Function of Mill’s Principle of Utility
113
“pleasure itself, together with the exemption from pain,” is irrelevant to moral evaluation, but never simply “pleasure itself, together with the exemption from pain.” This is not so much an objection to Mill regarding the clarity of his exposition as it is an appeal to the reader not to take every turn of phrase indicative of a definition to be an actual definition.15 Another bogus definition is available in the following remark in the fifth chapter of Utilitarianism: “What is the principle of utility, if it be not that ‘happiness’ and ‘desirable’ are synonymous terms?” (U5:36n CW10:258). This is obviously not to be taken at face value as a definition. The principle of utility is not a thesaurus entry asserting the lexical equivalence of “happiness” and “desirable” as synonymous terms. Granted, the last two passages are taken out of context, but the point is that Mill makes many comments in passing that resemble a definition of the principle of utility, though many of these do not express his considered opinion on the nature and content of the actual principle. Nevertheless, three passages in Utilitarianism are particularly useful in explicating the principle of utility and filling in the part of the story left incomplete in the presumably definitive statement quoted at the beginning of this chapter and again at the beginning of this section: “The utilitarian doctrine is, that happiness is desirable, and the only thing desirable, as an end; all other things being only desirable as means to that end” (U4:2 CW10:234). The first of the three supplementary passages, the second paragraph of the second chapter, brings out the rudiments of a theory of obligation alongside the familiar theory of value: [PU2: Theory of Obligation] The creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle, holds that actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness. By happiness is intended pleasure, and the absence of pain; by unhappiness, pain, and the privation of pleasure. To give a clear view of the moral standard set up by 15
As a matter of fact, the passage in question concerns the proper definition of utility itself rather than any sort of definition of the principle of utility: “Those who know anything about the matter are aware that every writer, from Epicurus to Bentham, who maintained the theory of utility, meant by it [meant by ‘utility’], not something to be contradistinguished from pleasure, but pleasure itself, together with the exemption from pain” (U2:1 CW10:209). The context of the discussion makes it clear that the word “it” here refers to “utility,” not to “the theory of utility.” Note especially that the statement occurs in the midst of Mill’s discussion concerning misinterpretations of the meaning of the word “utility.”
114
Chapter 2
the theory, much more requires to be said; in particular, what things it includes in the ideas of pain and pleasure; and to what extent this is left an open question. [PU1: Theory of Value] But these supplementary explanations do not affect the theory of life on which this theory of morality is grounded—namely, that pleasure, and freedom from pain, are the only things desirable as ends; and that all desirable things (which are as numerous in the utilitarian as in any other scheme) are desirable either for the pleasure inherent in themselves, or as means to the promotion of pleasure and the prevention of pain. mill U2:2 CW10:210
This passage reveals two aspects of the principle of utility (“or the Greatest Happiness Principle”), describing the principle in terms of both value and obligation. First, it spells out the utilitarian theory of obligation as the “moral standard” or “theory of morality” that “actions are right in proportion as they tend to promote happiness.” Next, it explains the underlying theory of value in terms of the “theory of life” that “pleasure, and freedom from pain, are the only things desirable as ends.” The second passage, corresponding to the tenth paragraph of the same chapter, further distinguishes Mill’s theory of value from his theory of obligation: [PU1: Theory of Value] According to the Greatest Happiness Principle, as above explained, the ultimate end, with reference to and for the sake of which all other things are desirable (whether we are considering our own good or that of other people), is an existence exempt as far as possible from pain, and as rich as possible in enjoyments, both in point of quantity and quality; the test of quality, and the rule for measuring it against quantity, being the preference felt by those who, in their opportunities of experience, to which must be added their habits of self-consciousness and self-observation, are best furnished with the means of comparison. [PU2: Theory of Obligation] This, being, according to the utilitarian opinion, the end of human action, is necessarily also the standard of morality; which may accordingly be defined, the rules and precepts for human conduct, by the observance of which an existence such as has been described might be, to the greatest extent possible, secured to all mankind; and not to them only, but, so far as the nature of things admits, to the whole sentient creation. mill U2:10 CW10:214
Nature and Function of Mill’s Principle of Utility
115
This passage distinguishes the utilitarian theory of value, introduced as the “ultimate end,” from the utilitarian theory of obligation, introduced as the “standard of morality.” The ultimate end is happiness. The standard of morality, accordingly, consists of rules and precepts for human conduct that guide the pursuit of happiness. The third passage corroborates, mostly by implication, all three functions of the principle of utility as discussed in the preceding section: We have now, then, an answer to the question, of what sort of proof the principle of utility is susceptible. If the opinion which I have now stated is psychologically true—if human nature is so constituted as to desire nothing which is not either a part of happiness or a means of happiness, we can have no other proof, and we require no other, that these are the only things desirable. If so, [PU1: Theory of Value] happiness is the sole end of human action, and [PU3: Theory of Justification] the promotion of it the test by which to judge of all human conduct; from whence it necessarily follows that [PU2: Theory of Obligation] it must be the criterion of morality, since a part is included in the whole. mill U4:9 CW10:237
The primary importance of this otherwise complex passage for our immediate purposes is that it brings everything together in a brief yet complete overview of the principle of utility: – PU1 (Theory of Value): Happiness is the sole end of human action, that is, the only thing desirable as an end in itself. – PU2 (Theory of Obligation): Happiness is the criterion of morality, and ipso facto, the source of moral obligation. – PU3 (Theory of Justification): The promotion of happiness is the standard of ethical justification and moral evaluation. The secondary importance of the passage is that it attests to the primacy of value, particularly over obligation and justification, as it states all three functions of the principle of utility in terms of happiness. The role of value in Mill’s ethical system is elucidated further in the next section. 2.5
The Primary Function of Mill’s Principle of Utility
Mill’s theory of value singles out happiness as the highest good, or as the only thing desirable as an end, which then shapes both his theory of moral obligation and his theory of ethical justification. From the very beginning of
116
Chapter 2
Utilitarianism to the very end, Mill identifies and endorses that which is good as taking precedence over that which is right, not only in his hedonistic teleology but in any ethical system. He indicates early on that his theory of moral obligation and his theory of ethical justification are grounded in his theory of value, and he suggests that this is how any ethical system ought to be organized: “All action is for the sake of some end, and rules of action, it seems natural to suppose, must take their whole character and colour from the end to which they are subservient” (U1:2 CW10:206). The most compelling evidence relevant to the matter under consideration is that Mill takes the principle of utility as a theory of value wherever he discusses utilitarianism with respect to a possible proof. The proof in question neither claims nor pretends to identify what is true or false, right or wrong, or justified or unjustified. It instead purports to identify and verify what is intrinsically desirable, valuable, or good (U1:5–6 CW10:207–208; U4:1–12 CW10:234–239). This comes across clearly even as early as the first chapter of Utilitarianism, where the proof yet to come is discussed exclusively in terms of value: “questions of ultimate ends”; “whatever can be proved to be good”; “something admitted to be good without proof”; “what proof is it possible to give that pleasure is good”; “a comprehensive formula, including all things which are in themselves good” (U1:5–6 CW10:207–208). The “comprehensive formula” of the first chapter of Utilitarianism anticipates the “ingredients of happiness” in the fourth chapter. The first chapter stipulates “that there is a comprehensive formula, including all things which are in themselves good, and that what ever else is good, is not so as an end, but as a mean[s]” (U1:5 CW10:208). The fourth chapter breaks down this “comprehensive formula” as it reveals that the “ingredients of happiness are very various, and each of them is desirable in itself, and not merely when considered as swelling an aggregate,” adding that these ingredients “are desired and desirable in and for themselves; besides being means, they are a part of the end” (U4:5 CW10:235). The comprehensive formula and the ingredients of happiness indeed turn out to be directly relevant to the proof of the principle of utility: “If human nature is so constituted as to desire nothing which is not either a part of happiness or a means of happiness, we can have no other proof, and we require no other, that these are the only things desirable” (U4:9 CW10:237). Notwithstanding the multiple functions of the principle of utility, Mill makes it clear that his proof of the principle concerns its role in value theory. This is obvious throughout the fourth chapter of Utilitarianism, both toward the beginning of the chapter, where Mill announces what he will prove, and toward the end of the chapter, where he sums up what he has proven. The preamble to his proof states that “the utilitarian doctrine is, that happiness is
Nature and Function of Mill’s Principle of Utility
117
desirable, and the only thing desirable, as an end; all other things being only desirable as means to that end” (U4:2 CW10:234). Mill then asks in direct reference to this conception of the principle of utility: “What ought to be required of this doctrine—what conditions is it requisite that the doctrine should fulfil— to make good its claim to be believed?” (U4:2 CW10:234). He thus makes it clear that the proof concerns the principle of utility as a theory of value, and more specifically, that he intends to prove that “happiness is desirable, and the only thing desirable, as an end” (U4:2 CW10:234). This is recapitulated toward the end of the same chapter, where he associates the principle of utility with “the doctrine that nothing is a good to human beings but in so far as it is either itself pleasurable, or a means of attaining pleasure or averting pain,” adding that, “if this doctrine be true, the principle of utility is proved” (U4:11–12 CW10:239). In the final analysis, then, proving the principle of utility, whatever may be the fullest and most accurate statement of that principle, amounts to proving that happiness is the only thing desirable as an end (U4:1–12 CW10:234–239). Donald George Brown (1973; 1982) and Alan James Ryan (1970), to cite a couple of examples from the seminal literature, agree that Mill’s principle of utility is the culmination of a theory of value rather than one of moral obligation or ethical justification.16 They both take the principle to state that happiness is the only thing desirable as an end (including philosophically insignificant semantic variations of this formulation). They maintain, however, that morality is only one of the applications of Mill’s principle of utility, and that the principle concerns his “Art of Life” in all of its domains (Brown 1982, 27, 36–38; Ryan 1970, xviii–xx, 187–212, 213–218, 233–241; cf. Ryan 1965; 1974, 104–106; 2014). Both Brown and Ryan appeal to passages in the last chapter of A System of Logic (L6:12.1–8 CW8:943–952), which indeed provides strong evidence in support of this interpretation. Mill divides the “Art of Life” into three branches (L6:12.6 CW8:949): Morality (the Right); Prudence or Policy (the Expedient); Æsthetics (the Beautiful or Noble). He holds that we need a general principle or standard (exactly one such principle or standard) for the “Art of Life” or the “practice of life,” and that such a principle would serve equally well as the ultimate principle of Morality, Prudence, and Æsthetics (L6:12.7 CW8:951).17 16 17
Among various other contributions to the relevant literature, those that may be consulted with great benefit include a historical analysis by Antis Loizides (2013) and a collection of essays edited by Ben Eggleston, Dale E. Miller, and David Weinstein (2010). Mill develops a comparable account in the following passage in “The Works of Jeremy Bentham” (1838): “Every human action has three aspects: its moral aspect, or that of its right and wrong; its æsthetic aspect, or that of its beauty; its sympathetic aspect, or that of its loveableness. The first addresses itself to our reason and conscience; the second to our imagination; the third to our human fellow-feeling. According to the first, we approve
118
Chapter 2
A footnote (n. 65) to the relevant passages in A System of Logic, as mentioned in section 2.3 above, refers readers to Utilitarianism for a discussion and vindication of the general principle of the “Art of Life” or the “practice of life” (L6:12.7 CW8:951). The corresponding passages in Utilitarianism then explain the principle of utility in terms of a “theory of life,” which reflects the general principle of the “Art of Life” or “practice of life” in A System of Logic. The “theory of life” in Utilitarianism is that “pleasure, and freedom from pain, are the only things desirable as ends” (U2:2 CW10:210). The general principle of the “Art of Life” or “practice of life” in A System of Logic is “that of conduciveness to the happiness of mankind, or rather, of all sentient beings” (L6:12.7 CW8:951). Brown and Ryan are surely right that Mill conceives of the principle of utility as the highest general principle of the “Art of Life” and that he holds that principle to be equally pertinent and applicable to Morality, Prudence, and Æsthetics. However, regardless of its various uses and applications, the principle of utility, even in A System of Logic, is nothing if not a theory of value holding happiness to be the only thing that is desirable as an end in itself (L6:12.7– 8 CW8:951–952). This is indeed the highest general principle of Mill’s theory of value. 2.6
The Structure of Mill’s Proof of the Principle of Utility
Mill’s express purpose in the fourth chapter of Utilitarianism, and hence his ultimate goal in his proof of the principle of utility, is to demonstrate “that happiness is desirable, and the only thing desirable, as an end; all other things being only desirable as means to that end” (U4:2 CW10:234). In essence, then, Mill’s proof is an attempt to establish a theory of value, namely ethical hedonism. The principle of utility may involve much more than this, but its proof
or disapprove; according to the second, we admire or despise; according to the third, we love, pity, or dislike. The morality of an action depends on its foreseeable consequences; its beauty, and its loveableness, or the reverse, depend on the qualities which it is evidence of” (CW10:112). This passage explains what Mill means in his proof of the principle of utility, where he speaks of a part’s being included in the whole: “If so, happiness is the sole end of human action, and the promotion of it the test by which to judge of all human conduct; from whence it necessarily follows that it must be the criterion of morality, since a part is included in the whole” (U4:9 CW10:237). The promotion of happiness is not just the criterion of morality but the criterion of morality, aesthetics, and beauty, as well as a distinctive test for the evaluation of actions in all three dimensions.
Nature and Function of Mill’s Principle of Utility
119
in the fourth chapter of Utilitarianism does not. In other words, although the principle of utility has several functions, its function as a theory of value is the exclusive subject of the proof in Utilitarianism. Given that Mill’s ultimate goal is to prove that happiness is the only thing that is desirable as an end in itself, the most sensible general reading of the fourth chapter is that Mill attempts to establish first that happiness is desirable as an end in itself and second that nothing else is desirable as an end in itself (U4:2 CW10:234). The text suggests that, as far as Mill is concerned, these are two distinct parts of the proof. As a first approximation, then, the structure of the proof is as follows:
Part 1: Happiness is desirable as an end in itself. Part 2: Nothing other than happiness is desirable as an end in itself. Result: Happiness is the only thing desirable as an end in itself.
Not all commentators agree that this is the basic structure of the proof. Some claim that the proof covers more substance than what these two parts represent. Others confirm the substantial coverage of the two parts, but find more than one natural break in the proof, thus analyzing the material in more than two parts. The greatest barrier to consensus on the structure of the proof is disagreement concerning its location in the fourth chapter. Some scholars contend that the proof is contained within one or two paragraphs, whereas others hold that it occupies the bulk of the fourth chapter. Early critics proceed as if the proof were contained wholly within the third paragraph of the fourth chapter. Even today, the role of the maligned third paragraph is controversial. Some argue that it is merely preliminary to the proof, whereas others insist that it is an integral part of the proof. To complicate matters further, the third paragraph is not the only focus of attention. Ronald Field Atkinson (1957, 161), for one, maintains that the proof is contained in the ninth and tenth paragraphs. Despite the variety of approaches, the simple two-part structure presented above is a good starting point for a deconstruction and analysis of the proof. Adjustments can be made along the way. Given widespread disagreement, it is best to proceed by locating the two parts of the proof in the text itself. Here is a skeletal outline of the fourth chapter of Utilitarianism in terms of its individual paragraphs and the relationship between the contents of each paragraph and Mill’s proof of the principle of utility:
120 1 2 3 4 5 6 7 8 9 10 11 12
Chapter 2
Paragraphs of Chapter 4 of Utilitarianism in Relation to the Proof Discussion of Methodology Thesis: Happiness is the only thing desirable as an end in itself. Part 1: Thesis: Happiness is desirable as an end in itself. Part 2: Thesis: Nothing else is desirable as an end in itself. continues continues continues continues continues concludes Objection Considered: Desire vs. Will Closing Remarks
The first two paragraphs are preliminary. The proof begins with the third paragraph and ends with the tenth. The first part of the proof is contained in the third paragraph. The second part extends from the fourth paragraph to the tenth. Textual evidence shows that the third paragraph neither contains the whole proof nor is merely preliminary to the actual proof. Toward the end of that paragraph, Mill is convinced that he has established “that happiness is a good” (U4:3 CW10:234). At the end of the same paragraph, he reiterates this conviction more emphatically: “Happiness has made out its title as one of the ends of conduct, and consequently one of the criteria of morality” (U4:3 CW10:234). He then devotes the next seven paragraphs to proving that nothing other than happiness is desirable as an end in itself. The difference in length between the two parts of the proof is not surprising, considering that the relatively weak claim of the first part, that happiness is desirable as an end in itself, stands in stark contrast to the incredibly strong claim of the second part, that nothing other than happiness is desirable as an end in itself. The difference in length, though unsurprising, proves troublesome for the first part of the proof, where Mill offers several apparently controversial claims in support of the otherwise uncontroversial conclusion that happiness is desirable as an end in itself. Two elements of the proof, both in the first part, have attracted the most attention. The first is Mill’s conviction that “the sole evidence it is possible to produce that anything is desirable, is that people do actually desire it.” The second is his inference that “each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons” (U4:3 CW10:234). Mill would have deprived introductory logic textbooks of some of
Nature and Function of Mill’s Principle of Utility
121
their favorite examples of fallacies, had he simply assumed the uncontroversial conclusion of the first part of the proof, that “happiness is desirable as an end in itself,” as a premise in the second part of the proof, or had he even briefly elaborated on the sense in which desires are evidence of desirability, and explained the sense in which the general happiness is desirable for, or is a good to, the aggregate of all persons. Ironically, Mill’s apparently enigmatic defense of the uncontroversial conclusion that happiness is desirable as an end in itself has forever kept the harshest and sharpest critics intensely focused on the first part of the proof, largely ignoring the second part, which establishes the far more controversial conclusion that nothing other than happiness is desirable as an end in itself. It is now too late, and therefore utterly unhelpful, to assume on Mill’s behalf that the uncontroversial conclusion of the first part of the proof is indeed true, and to let the proof stand or fall with the cogency of the second part. Quite the contrary, due to the history of attacks on the first part of the proof, an adequate interpretation of the whole proof, not to mention a thorough defense of it, cannot help but cover its two parts in inverse proportion to the length of Mill’s exposition of them.
c hapter 3
The Historical Reception of Mill’s Proof of the Principle of Utility This chapter is an exegetical survey of historically influential reactions to Mill’s proof of the principle of utility. Well before it secured a permanent place in elementary logic textbooks as an example of fallacies to be avoided, Mill’s proof attracted the contempt and ridicule of several generations of prominent philosophers. The cumulative effect was to undermine utilitarianism itself by reducing its most memorable exposition and defense to a string of logical fallacies and conceptual confusions. This may not be the standard view today, at least not among serious scholars, but the tradition is strong enough to merit careful consideration. The aim of this chapter is to meet that need through an analysis of the critical forces responsible for the lasting notoriety of Mill’s proof as a paradigm of bad reasoning. The motivation behind the chapter is one that is common to all studies of this kind: to put an ongoing discussion into historical perspective. The exegetical foundation built on the original sources examined here informs the reconstruction and evaluation process in the second part of the book. This attests to the authenticity of the overarching effort in the present volume as a response to actual allegations by notable experts. The effort in question is the detection and reflection of the most sensible and challenging formulations of specific objections that must be met in order to vindicate the proof. While it may always be sufficient to address the strongest conceivable versions of all possible objections, such an attempt becomes all the more relevant if the objections under scrutiny turn out to be actual charges. The ones examined here confirm that just about every objection we have on record today originates with commentators born in the same century as Mill and with critiques published within fifty years of the appearance of the proof in print. Given the nature of the charges, the vindication required is not so much about showing that Mill’s proof establishes the principle of utility beyond the slightest doubt as it is about showing that it does not commit the fallacies traditionally and persistently attributed to it. Vindication in a fuller sense is, no doubt, also important. Absolving a proof of fallacious reasoning does no more for that proof than clearing a speech of grammatical errors does for that speech. Timber without termites, water without cholera, breakfast without salmonella—whatever it is, it may not be very good, even if it is not shockingly or
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_005
Historical Reception of Mill’s Proof
123
shamefully bad. But when the charge is indeed that it is shockingly and shamefully bad, that is certainly the place to start, recalling that Francis Herbert Bradley (1876), for one, is “ashamed to have to examine such reasoning” (p. 104) and to have to warn readers repeatedly against its deceptive appeal (p. 112). Comprehensive vindication comes later. The virtues and implications of the proof, for example, come out in chapters seven and eight, respectively. But the scholarly outrage takes precedence. This chapter provides the groundwork for the next three chapters to absolve the proof of egregious mistakes and misconceptions, the removal of which then liberates what is left for constructive and reconstructive evaluation in the final two chapters. The scope of the discussion here is limited to the logical fallacies commonly attributed to the proof throughout its critical history. The naturalistic fallacy, conceived by Moore, falls outside that scope, as it is not a logical fallacy but a category mistake, if it is indeed a mistake at all. The charge itself, however, marks a milestone in metaethics and cannot responsibly be ignored. It gets plenty of attention in chapter 6, which is devoted in its entirety to the naturalistic fallacy.1 Mill allegedly commits several logical fallacies in his proof as well as the naturalistic fallacy. The logical fallacies in question are those of equivocation, composition, and division. The last of these now has more of a historical interest as a charge against Mill, but the other two accusations are still in currency. The allegations, whether historical or current, all arise in response to the third paragraph of the fourth chapter of Utilitarianism. Here, then, for ease of reference, is what Mill says in that paragraph: The only proof capable of being given that an object is visible, is that people actually see it. The only proof that a sound is audible, is that people hear it: and so of the other sources of our experience. In like manner, I apprehend, the sole evidence it is possible to produce that anything is desirable, is that people do actually desire it. If the end which the utilitarian doctrine proposes to itself were not, in theory and in practice, acknowledged to be an end, nothing could ever convince any person that it was so. No reason can be given why the general happiness is desirable, 1 As discussed in chapter 6 of the present volume, George Edward Moore finds Mill’s proof to contain two separate mistakes, neither of which constitutes a logical fallacy. First, the proof allegedly identifies “good” with “desired,” thereby confusing something’s being good with its being desired. Second, it presumably attempts to establish that happiness is intrinsically good, and subsequently that happiness is the only thing that is intrinsically good, whereas claims specifying what things are intrinsically good are not capable of proof or disproof.
124
Chapter 3
except that each person, so far as he believes it to be attainable, desires his own happiness. This, however, being a fact, we have not only all the proof which the case admits of, but all which it is possible to require, that happiness is a good: that each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons. Happiness has made out its title as one of the ends of conduct, and consequently one of the criteria of morality. mill U4:3 CW10:234
The entire passage is a source of critical interest, but it is not uncommon to find certain sentences from this paragraph extracted, either in full or in part, and presented in introductory logic textbooks as examples of informal fallacies.2 Three popular choices in this respect are the following statements: – “The sole evidence it is possible to produce that anything is desirable, is that people do actually desire it.” – “No reason can be given why the general happiness is desirable, except that each person, so far as he believes it to be attainable, desires his own happiness.” – “Each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons.” The first statement is often taken as the major premise of an implicit argument committing the fallacy of equivocation through illicit semantic variation on the word “desirable.” The argument presumably runs along the following lines: Whatever is desired is desirable (desire +able =able to be desired); happiness is desired; therefore, happiness is desirable (good). The argument is thus held to equivocate on the term “desirable” by using it in a descriptive sense in the major premise, to refer to what can be desired, or what is able to be desired, while using it in a normative sense in the conclusion, to refer to what is worthy of desire, or what ought to be desired. Leaving the second statement for last, we may note that the third statement is widely considered an example of the fallacy of composition, either with respect to the switch from personal happiness to general happiness or with respect to the switch from individual persons to the aggregate of all persons or with respect to both. As for the second statement, the typical treatment makes it the most interesting portion of 2 References are too numerous for comprehensive coverage. A representative list spanning a period of well over a century includes: Beth Black (2012, 68); Irving M. Copi, Carl Cohen, and Kenneth McMahon (2014, 153, 154); James Edwin Creighton (1898, 365); Robert Malcolm Murray and Nebojsa Kujundzic (2005, 437); Nicholas Rescher (1964, 77); Walter Sinnott- Armstrong and Robert Fogelin (2015, 296–297); William Henry Werkmeister (1957, 35).
Historical Reception of Mill’s Proof
125
the proof, if not the most amusing, because this single statement supposedly contains both fallacies at once. It is often said to contain the fallacy of equivocation in the move from desires to desirability and the fallacy of composition in the move from each person’s happiness to the general happiness. All in all, then, Mill is taken to commit two different logical fallacies in two different sentences in the same paragraph, one in each sentence, and he is taken to introduce a third sentence between those two to bring together a combination of the same two fallacies in one statement—all in the third paragraph of the fourth chapter of Utilitarianism. This is to say nothing of the fact that the naturalistic fallacy allegedly infects the same paragraph, invalidates the entire proof, and pervades the whole work in general (Moore 1903, §§ 3:39– 44, pp. 116–126).3 It is no surprise that, alluding to the infamous paragraph and quoting from it liberally, one critic goes so far as to mock Mill by pointing to the difficulty of collecting so many fallacies in such a short space (Mackenzie 1893, 103). Two authoritative sources on early scholarly reactions to Mill’s Utilitarianism are Jerome Borges Schneewind’s (1976) thematically broad coverage of critiques during the first fifteen years following its publication and James Seth’s (1908) earlier and thematically narrower study of a later period of criticism directed specifically at Mill’s proof of the principle of utility. Both are eminently useful toward developing a full and healthy appreciation of the reception of Mill. This chapter supplements the contributions they make with an annotated documentation of specific objections to Mill’s proof that have proven to be the most enduring. Schneewind (1976) covers the following categories of commentary: the possibility of deriving moral concepts from nonmoral concepts; the proof of the principle of utility; the distinction between higher and lower pleasures; the reconciliation of rules of common sense morality and utilitarian ethics. He reports that Utilitarianism had attracted little attention in print during the first seven or eight years but that critical reaction to the work during the first fifteen years was almost entirely negative (Schneewind 1976, 39–41). His most salient observation concerning the early reception of Utilitarianism is that “almost every criticism that has since been developed was made in at least rudimentary, and usually in fairly sophisticated, form during this early period, while a number of criticisms were made that have since dropped out of sight” (Schneewind 1976, 41). 3 References to Moore’s Principia Ethica (1903) give chapter, section, and where appropriate, paragraph numbers, all relevant to any edition or impression, followed by page numbers for the 1993 edition.
126
Chapter 3
Historically, the most common objections to Mill’s proof have portrayed it as failing several times over through the logical fallacies of equivocation, composition, and division, as well as committing the supralogical naturalistic fallacy. The most devastating objections continue to be those that attribute a logical fallacy to the proof. In the light of Schneewind’s survey, it seems possible that Mill might have been aware of such charges, at least in their essentials, between the publication of Utilitarianism in 1861 and his death in 1873. However, the strongest and clearest charges began to appear, for the most part, after Mill’s death. The most influential and still memorable commentaries on the proof can be found in the following monographs (thus ignoring journal articles or discussion notes):4 – John Grote: An Examination of the Utilitarian Philosophy (1870); – Henry Sidgwick: The Methods of Ethics (7 editions: 1874–1907); – Francis Herbert Bradley: Ethical Studies (2 editions: 1876/1927); – William Ritchie Sorley: The Ethics of Naturalism (2 editions: 1885/1904); – John Dewey: Outlines of a Critical Theory of Ethics (1891); – John Dewey and James Hayden Tufts: Ethics (2 editions: 1908/1932); – John Stuart Mackenzie: A Manual of Ethics (6 editions: 1893–1929); – George Edward Moore: Principia Ethica (1903). Hardly any charges concerning the logical status of Mill’s proof have been advanced that were not first covered by the works listed above. Seth (1908) discusses and challenges the central criticism in most of these. Three of them, however, are excluded from his seminal defense of Mill in that regard: Grote’s An Examination of the Utilitarian Philosophy (1870), Moore’s Principia Ethica (1903), and Dewey and Tufts’s Ethics (1908). His omission of Grote’s Examination suggests that Seth may have found the corresponding critique less worrisome than the ones he chose to tackle. His omission of Moore’s Principia Ethica is
4 The most relevant passages are as follows: (1) John Grote: An Examination of the Utilitarian Philosophy (1870, 58–78, especially 63–72). (2) Henry Sidgwick: The Methods of Ethics (seven editions: 1874–1907; cf. 1874, 364–366; 1907, 386–389); see also “The Establishment of Ethical First Principles” (1879). (3) Francis Herbert Bradley: Ethical Studies (two editions: 1876/ 1927; cf. 1876, 101–105; 1927, 111–116). (4) William Ritchie Sorley: The Ethics of Naturalism (two editions: 1885/1904; cf. 1885, 57–75; reprinted with minor emendations in 1904, 59–78). (5) John Dewey: Outlines of a Critical Theory of Ethics (1891, 53–59, 59–67). (6) John Dewey and James Hayden Tufts: Ethics (two editions: 1908/1932; 1908, 286–305, especially 290–293; 1932, 205–215). (7) John Stuart Mackenzie: A Manual of Ethics (six editions: 1893–1929; see especially 1893, 98–99, 112–113, 263; reprinted either with no alterations or with minor emendations in the editions of 1894, 98–99, 112–113, 263; 1897, 213–214, 229–230, 394; 1900/1901, 213–214, 229–230, 261; 1915, 215–216, 231–232, 287; 1929, 169, 182, 248–249). (8) George Edward Moore: Principia Ethica (1903, §§ 3:39–44, pp. 116–126, § 3:62, pp. 155–156).
Historical Reception of Mill’s Proof
127
somewhat surprising, considering its widespread popularity, but the explanation there might be the temporal proximity of the publication of Seth’s and Moore’s works, even though a difference of five years may seem substantial by modern standards. Seth’s omission of Dewey and Tufts’s Ethics is certainly understandable, and more plausibly attributable to time constraints, since his article and that book were published in the same year. The remainder of this chapter lays out the criticism directed against Mill’s proof in each of the eight works listed above. The order of presentation is chronological with respect to the date of publication of the first edition of each book. The only exception is that Dewey’s two books are kept together even though other entries chronologically belong in between. Page references are normally to the first edition, except when tracking the evolution of a particular objection, in which case all relevant editions are documented. In any event, the edition in question is clearly identified. The entries examined here present mutual overlap as they are all immediate reactions to a short passage of text: the third paragraph of the fourth chapter of Utilitarianism. The resulting redundancy, however, is fair exchange for a full grasp of the contribution each critic makes to the ongoing debate over the controversial passage. Each one brings a special perspective discussed in the corresponding section of this chapter. The presentation proceeds with documentation of original sources, particularly in the form of block quotations, accompanied by exegetical and critical commentary. The emphasis at this point, even when combined with open disagreement, is more on clarifying objections than on defending Mill. Generally, the clearer the direct quotation, the shorter the accompanying discussion. 3.1
John Grote: An Examination of the Utilitarian Philosophy
Grote is the first of Mill’s early critics taken up here, just as he was the first among them to see his opposition in print. Unlike many of those coming after him, or at least more so than any of them, Grote’s approach is philosophical rather than logical. That is to say, his reaction to the proof is more about its philosophical aims and material implications than it is about its formal structure. As for the commonly alleged logical fallacies, Grote appears as sensitive as anyone to an apparent equivocation on desirability and to an apparent fallacy of composition in the move from the desirability of each person’s happiness for that person to the desirability of the general happiness for the aggregate of all persons. Yet the fault he finds in each is not so much logical as it is empirical and rational. His critique of the corresponding passages brings out
128
Chapter 3
what he thinks are the conceptual and theoretical difficulties there as opposed to focusing solely on the structure and neglecting the content. This is in large part because Grote takes Mill’s qualification of the so-called proof seriously, heeding the explicit warning that ultimate ends are not susceptible of proof in the ordinary sense (U1:5–6 CW10:207–208; U4:1 CW10:234; U4:9 CW10:237) and thereby avoiding the kind of critique suitable for a proof in the ordinary sense: I come now to Mr Mill’s proof of utilitarianism. I am not much concerned with the logical conclusiveness of it. Mr Mill admits that what he says will most likely appear merely ‘obvious,’ and yet is not ‘proof in the ordinary meaning of the term:’ in fact the subject does not admit of it. But it is important to observe the manner of thinking which the proof involves, and what it is that is proved. grote 1870, 63
Grote’s main concern is that Mill assumes too much, taking for granted most of what he sets out to prove in any sense whatsoever. Mill’s worst mistake, according to Grote, is that he is too quick to present the good as the ultimate question of morality, ignoring the right except insofar as it is assumed, without argument or evidence, to be derivative on the good. Mill supposedly thereby elevates the summum bonum above the summum faciendum without justification. Grote’s point is not that we cannot derive the right from the good, nor even that deriving the right from the good is not the only way, or the best way, to establish or ascertain what is right. His point, rather, is that any such position is not to be postulated but proven. He thus accuses Mill of proving the obvious and assuming the rest: Mr Mill says, ‘From the dawn of philosophy the question concerning the summum bonum has been accounted the main problem in speculative thought.’ He is doing his part to solve it. But surely he cannot mean that it is solved by the laying down, as a supposed fact of observation, that what men really desire is that which is pleasant to them. Is the doubtfulness which has hitherto attended the question, and which observation has now at last put an end to, the doubtfulness whether men really do this? Mr Mill has to prove that ‘happiness,’ as the ideal summum bonum of man, is the one thing which ought to regulate his conduct (as he calls it, the sole criterion of morality): this is not a thing that any observation can prove, and it is quite a vain proceeding to set observation, as Mr Mill
Historical Reception of Mill’s Proof
129
does, to warrant a truism, and then to say that in doing so it proves a point entirely different. grote 1870, 64
Grote interprets Mill’s position that “actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” (U2:2 CW10:210) as leaving no room for duty, except either as a means to happiness or as a part of happiness. But this is a simplistic reading grounded in an unwarranted reduction and incomplete information. First, as Mill makes clear in a review article, “The Works of Jeremy Bentham” (1838), already quoted above, near the beginning of chapter 2, he takes consequentialism to be a feature of any sensible school of morality and not just of utilitarianism: That the morality of actions depends on the consequences which they tend to produce, is the doctrine of rational persons of all schools; that the good or evil of those consequences is measured solely by pleasure or pain, is all of the doctrine of the school of utility, which is peculiar to it. mill CW10:111
Second, as Mill explains in the fifth chapter of Utilitarianism, the identification of happiness as the summum bonum does not diminish the significance of duty as the summum faciendum: We do not call anything wrong, unless we mean to imply that a person ought to be punished in some way or other for doing it; if not by law, by the opinion of his fellow creatures; if not by opinion, by the reproaches of his own conscience. This seems the real turning point of the distinction between morality and simple expediency. It is a part of the notion of Duty in every one of its forms, that a person may rightfully be compelled to fulfil it. Duty is a thing which may be exacted from a person, as one exacts a debt. Unless we think that it might be exacted from him, we do not call it his duty. mill U5:14 CW10:246
Be that as it may, Grote objects that Mill’s emphasis on the good comes at the expense of his attention to the right: All reasonable action is action to an end or for a purpose: such is the idea of reason as applied to action: but the end or purpose need not necessarily be something to be attained or gained in the way of possession or
130
Chapter 3
enjoyment, which is what is implied in the phrase summum bonum; it may be something to be done. grote 1870, 66
The difference, argues Grote, is too important to be buried in an assumption reducing the right to the good: This is the idea of action being right or wrong, as distinguished from the idea of it as better or worse, more or less desirable. This is the idea of the summum jus, the faciendum, the notion of duty, under which the moral question may in some circumstances present itself to us, rather than in the idea of the summum bonum, the acquirendum, the notion of happiness. grote 1870, 67
While this is about what the proof leaves out, Grote also has reservations about what it accomplishes. He finds Mill to exaggerate the role of evidentiary insight into the desirability of happiness. He agrees that the evidence, and therefore the insight, really is there, but he argues that Mill overplays the importance of that relationship. He does not deny that the desirability of happiness, despite the vagueness of both terms (“desirability” and “happiness”), is open to observation in some sense, and to some degree, through actual desires for happiness. But he takes this to be a trivial conclusion, easily granted instead of proven, and not at all conclusive, perhaps not even altogether relevant, whether granted or proven: What Mr Mill proves, in the place where he considers that he is proving utilitarianism to be the real and only moral philosophy, (so far as anything of the sort is capable of proof,) seems to me to be only that men desire happiness or what is pleasant, or, in other words, that it is happiness that is desirable. Now this is what no one doubts and what needs no proving, as indeed Mr Mill’s proof of it is simple enough, consisting of hardly more than statement of it: the various terms here used, independently of the following them out into details and particulars, may be considered as all meaning the same thing: the τὸ ἀγαθόν, or what is a good to us, is simply the desired and desirable: in speaking of the need, for morality, of knowing what is the summum bonum, Mr Mill had already assumed all he proves here. In fact he had assumed more. For though he may prove that happiness is all that men desire, he does not prove that it is all that they think about, or that nothing but what they desire is of importance
Historical Reception of Mill’s Proof
131
to them. As I have said, it is a thing which may very well suggest itself to people, and I believe sometimes does, that there may be work, business, duty, whatever we may call it, for them to do independent of conscious effort after any happiness, and Mr Mill has not proved that utilitarianism even in this rudimentary form is the only moral philosophy, or that the summum bonum is all that men need think of, till he has proved not only, as he does, that men desire happiness, and nothing else but happiness, but also that it is nothing else but what they desire that they need take any moral account of. grote 1870, 68–69
Grote’s disapproval of what the proof leaves out is so strong that even his thoughts about what the proof accomplishes tend to come full circle to what it leaves out. But there is more to his misgivings. He also anticipates the scholarly opposition Mill attracts in the second part of the proof. This is where the argument proceeds from the desirability of each person’s happiness for that person to the desirability of the general happiness for the aggregate of all persons: But Mr Mill seems to consider that he has proved that, in the same natural manner in which a man’s happiness is an end to him, the aggregate happiness is an end to each individual of the aggregate. Mr Mill in other places, as we have seen, shows most admirably how it may become so: but if his proof here had held good, there would have been no need to show this; what I have called his ‘societarianism’ would have been superfluous. grote 1870, 70–71
In keeping with the general thrust and overall orientation of his critique, Grote’s objection in this particular passage to Mill’s compositional inference in the proof is conceptual rather than logical. But he separately makes it clear that he objects to the logic as well: The general interest and the action for that on the one side are not like the individual interest and the action for it on the other, a single object commending itself to a single will. grote 1870, 71
This observation goes to the heart of the matter. But the passage quoted immediately before that shows Grote describing a move that is not actually in the text. Nowhere in the proof does Mill say that the general happiness is a good to each person. It may be tempting, and possibly even reasonable, to assume that
132
Chapter 3
this is what we are to make of what Mill does say. This would be to assume, in other words, that “the desirability of the general happiness for the aggregate of all persons” resolves into “the desirability of the general happiness for each person.” The reason to assume this would presumably be that there is no other sense to be made of what may or may not be said of the aggregate of all persons in that regard, which must then be taken distributively rather than collectively in order to be assigned a good of its own. Yet all we can condemn in that case is the assumption and not the actual statement. There may be other conceivable and reasonable interpretations of the relationship (“the desirability of the general happiness for the aggregate of all persons”), duly explored in the chapter dedicated to the alleged fallacy of composition (chapter 5 below). The most rigorous analysis of Mill’s compositional inference in all of Grote’s discussion comes in a footnote supplied by Grote’s editor, Joseph Bickersteth Mayor: Mr Mill’s argument is really an instance of the ‘fallacy of composition,’ in which the word all is used at one time distributively, at another time collectively. Thus: each human being A, B, C, &c. naturally desires his own happiness; but A, B, C, &c. make up all human beings, and the happiness of A, B, C, &c. makes up the happiness of all human beings; therefore every human being naturally desires the happiness of all human beings. Taking it out of the abstract the proposition becomes still more glaringly untenable. Two men place their happiness in the exclusive possession of the same thing, a third places his happiness in the positive unhappiness of one who, he thinks, has wronged him. Thus the resultant of several (or all) men’s individual happiness might well be the general unhappiness. grote 1870, 70, n. 1
Mayor goes on to supplement his analysis with an analogy: The fact is, this is an attempt on the part of the utilitarians to extend to morality the principle, true under certain limitations in political economy, that the public wealth is best promoted by each man’s aiming at his own private wealth and occupying himself exclusively with that. grote 1870, 70, n. 1
Mayor evidently sees himself as providing a real service in exposing and explicating the logical problems in Mill’s switch from parts to whole. His contribution is indeed useful even if it is not entirely accurate. But Grote himself is not in any way oblivious to the suspicious compositional structure of the inference
Historical Reception of Mill’s Proof
133
in question. This is confirmed, among other places, in the last passage quoted above from Grote’s own text, where he expressly notes the operative difference between “individual interest” and “general interest” (1870, 71) as part of his objection to the reasoning there. Mayor, of course, leaves no doubt as to precisely what the logical error might be as well as just how serious it might be. That makes it useful to have his note in addition to Grote’s commentary. However, Mayor multiplies Grote’s hasty generalizations as he expands the latter’s interpretive liberties with Mill’s text. Where Grote assumes that Mill considers the general happiness a good to (desirable for) each person—in declaring it to be a good to (desirable for) the aggregate of all persons—Mayor assumes in addition that the inference proceeds with desires as opposed to goods or desirability. Mayor thus takes Mill to be inferring that each person desires the general happiness, presumably from the observation that each person desires his own happiness, whereas Mill does not even say that anyone at all, let alone everyone, desires the general happiness. Mill’s compositional inference, as discussed below in chapter 5, remains on the level of the desirable, not on that of the desired. Grote’s assessment of the inference is at the correct level in that regard, as it works either with goods or with desirability, and not with desires, but it is otherwise too quick, whether accurate or not, to associate the general happiness with each person: ‘Each person’s happiness,’ Mr Mill rightly says, ‘is a good to each,’ and he draws from this a conclusion which seems to me of very little significance: the real point of morals, which utilitarianism evades, is the knowing how to meet any one who concludes thus, Since then it is my happiness that is the good to me, it is not the general happiness that is so, and there is no reason that I at least should act for that. The more a man’s particular happiness appears a good to him, the more it is likely to engross his action, and the less he is likely to think of the happiness of the aggregate. grote 1870, 72
Nevertheless, despite any misinterpretation, Grote brings out much of what is at stake philosophically if not also logically in Mill’s proof. To pick just one example, even if Mill did not mean that the general happiness is a good to each person, in making it out to be a good to the aggregate of all persons, we may press him, with good reason, for what he did mean, for it is not obvious. What Mill meant here and elsewhere in the proof is precisely the question taken up in the following chapters. In the remainder of this one, the views of other
134
Chapter 3
critics stand to supplement the insight provided by Grote, particularly as they tend to focus more (or more strictly) on the logic of the matter. 3.2
Henry Sidgwick: The Methods of Ethics
Sidgwick is among the earliest of the eminent commentators on Mill’s utilitarianism and the second earliest among the ones examined here. He does not accuse Mill explicitly of committing any logical fallacies in the proof.5 However, he definitely anticipates, and perhaps shapes, the tendency of critics to attribute the fallacy of division and the fallacy of composition (one or the other or both at once) to the parts-whole inference concluding with the desirability of the general happiness for the aggregate of all persons. Sidgwick is eager to show that the proof would fail even if desires were conclusive evidence of desirability. He thus assumes, for the sake of argument, that actual desires for the general happiness confirm the desirability of the general happiness. He then objects that Mill’s inference fails to establish the desirability of the general happiness because it does not show that anyone actually desires the general happiness. More perspicuously, he grants the move from desires to desirability on a personal level, admitting thereby that each person’s desire for that person’s happiness establishes the desirability of each person’s happiness for the person whose happiness it is, but he objects that there is no such desire corresponding to the desirability of the general happiness. Here, in his own words, specifically in The Methods of Ethics, is Sidgwick’s critique of the inference: Now it must be borne in mind that it is as a “standard of right and wrong,” or “directive rule of conduct,” that the utilitarian principle is put forward by Mill. Hence, in giving as a statement of this principle that “the general happiness is desirable,” he must be understood to mean (and his whole treatise shews that he does mean) that it is what each individual ought to desire, or at least to aim at realizing in action. The proof he offers of this is, that each one actually does desire his own happiness. But it may surely be objected that the natural immediate conclusion from this, on Mill’s own method, is that Own Happiness, not Universal Happiness, is what each one ought to desire: the argument leads primarily to the principle of
5 Henry Sidgwick has his own “Proof of Utilitarianism,” specifically in chapter 2 of book 4 of The Methods of Ethics (1874, 388–393; 1907, 418–422).
Historical Reception of Mill’s Proof
135
Egoistic instead of Universalistic Hedonism. And I can conceive no possible way of meeting this objection, except by exhibiting (in substantially the same manner as Clarke and Kant exhibit it) the necessary universality of the ultimate end, as recognised by Reason: by shewing that the fact “that I am I” cannot make my happiness intrinsically more desirable, more fit to be accepted by my reason as the standard of right and wrong in conduct, than the happiness of any other person. sidgwick 1874, 365
Sidgwick’s opposition grows stronger between the first and seventh editions of The Methods of Ethics. The preface to the sixth edition (1901, xv–xxi) reveals his increasing dissatisfaction with the conflict he finds between Mill’s psychological hedonism (the theory that each person seeks his or her own happiness) and Mill’s ethical hedonism (the theory that everyone ought to seek the general happiness). Years of reflection lend themselves to a more robust formulation of the same assessment in the seventh and final edition (1907): Now, as we have seen, it is as a “standard of right and wrong,” or “directive rule of conduct,” that the utilitarian principle is put forward by Mill: hence, in giving as a statement of this principle that “the general happiness is desirable,” he must be understood to mean (and his whole treatise shows that he does mean) that it is what each individual ought to desire, or at least in the stricter sense of “ought” to aim at realising in action. But this proposition is not established by Mill’s reasoning, even if we grant that what is actually desired may be legitimately inferred to be in this sense desirable. For an aggregate of actual desires, each directed towards a different part of the general happiness, does not constitute an actual desire for the general happiness, existing in any individual; and Mill would certainly not contend that a desire which does not exist in any individual can possibly exist in an aggregate of individuals. There being therefore no actual desire—so far as this reasoning goes—for the general happiness, the proposition that the general happiness is desirable cannot be in this way established. sidgwick 1907, 388
The main difference between the two versions, other than clarity, is that the first seems at most to tiptoe around the possibility of a fallacy of division, whereas the second confidently points out prima facie illicit moves associated both with the fallacy of division and with the fallacy of composition.
136
Chapter 3
As he indicates at the outset in either version (and reaffirms in between in a discussion note of 1879), Sidgwick also has reservations concerning the legitimacy of Mill’s move from desires to desirability, usually associated with equivocation on the term “desirable.” Having resolved, for the sake of argument, to ignore potential problems in that respect, he invokes them anyway as he proceeds to isolate what he finds problematic about the parts-whole inference. He objects that an aggregate of desires is not a desire, and therefore that a desire for the general happiness does not exist, or rather, that it does not come into being, just because each part of the general happiness is desired by the person whose happiness it is. Sidgwick thus concludes that Mill fails to prove that the general happiness is desirable, simply because he fails to show that the general happiness is desired. Even if this were Mill’s argument (though it is not), the objection would not be decisive.6 Sidgwick grants that each person’s happiness is desired by the person whose happiness it is, and he assumes that the general happiness is the aggregate happiness. The problem with the objection that follows is that the general happiness can reasonably be said to be desired in some sense even if the collective it represents is not desired in its entirety by every single person. Imagine a box of assorted chocolates, each piece being desired by someone or another but not by everyone, and each person desiring one piece or another but not every piece. The box of chocolates, then, is indeed desired albeit by different persons. (Imagine, if you like, that someone desires the box as well, independently of what is in it.) It may be objected that what is required is an actual desire for the collective representing the general happiness, and furthermore that a plurality of desires for its constituent parts is not the same as a desire for the whole. It is admittedly not the same. But is it different in a relevant way? And even if it
6 See chapter 5 below for a demonstration of why Mill’s inference does not, pace Sidgwick’s reading, depend on everyone in unison desiring the general happiness, nor even on anyone at all desiring the general happiness. That possibility is neither required nor denied. It is simply not a part of the argument. The question is about the desirability of the general happiness, which is not inferred from its being desired by any number of people, each of whom does, on the other hand, desire his or her own personal happiness, which Mill finds sufficiently indicative for the generalization he makes. Sidgwick loses track of Mill’s train of thought somewhere along the way. Although the present chapter is historical in orientation, it is helpful even at this point, not just with respect to Sidgwick but in regard to all the critics considered here, to point out potential flaws in objections, even if they do not adequately capture Mill’s argument. After all, misdirected objections, no less than any that may actually be on target, have contributed to the development and future direction of the body of criticism surrounding Mill’s proof.
Historical Reception of Mill’s Proof
137
is, why suppose that no one at all desires the general happiness? What if one or two particularly affectionate and compassionate people do? There are at least a few of those in every community. Or must everyone desire the general happiness for the inference to work? Just how many people would indeed have to desire it for the general happiness to count as something that is desired, whether or not it can reasonably be concluded, on that basis alone, to be desirable as well? If one person is sufficient, then Sidgwick’s objection is in trouble. If comprehensive engagement is required, then Mill’s inference is in trouble. Anything in between, especially around the middle, is more difficult to judge. As it is, Sidgwick exaggerates the conditions for the desirability of the general happiness, by making it contingent upon each person’s desiring the happiness of every other person. He leaves us with the scenario of everyone desiring the happiness of everyone, including oneself as well as others, as if Mill would have missed the possibility of a conflict between the desires of one person and the desires of another. More importantly, Sidgwick misconstrues the logic of the actual inference. He does not consider the alternative that Mill might be arguing directly from the desirability of each part of the general happiness, and not from any desires for each part, to the desirability of the general happiness, in other words, that Mill’s move from parts to whole is entirely on the level of the desirable rather than proceeding from that of the desired. This may or may not absolve Mill of the fallacy of composition, a task left for chapter 5, but it is an alternative that does not seem to occur to Sidgwick at all. Put simply, Sidgwick ignores the possibility that Mill completes his move from the desired to the desirable on the level of the individual, after which he moves from what is desirable for each person to what is desirable for the aggregate of all persons. 3.3
Francis Herbert Bradley: Ethical Studies
Bradley is the most antagonistic of the early critics of Mill’s proof of the principle of utility. Framing his objections in a hostile tone throughout his review, he is surpassed by no others in that respect. The animosity of his delivery at nearly every step of the way contributes to a disparaging critique overall: This is not a good theoretical deduction, but it is the generation of the Utilitarian monster, and of that we must say that its heart is in the right place, but the brain is wanting. bradley 1876, 104
138
Chapter 3
A footnote to this assessment offers the following analysis: It is monstrous to argue thus:—“Because (1) on psychological grounds it is certain that we can desire nothing but our own private pleasure; because (2) on some other grounds something else (whatever it may be), something not my feeling of pleasure, something other than my private self, is desired and desirable; therefore (3) this something else which is desired and desirable is the pleasure of others, since, by (1), only pleasure can be desired.” If we argue in this way, we may as well go a little further to—“(4) and therefore we can and do desire something not our own private pleasure, and therefore (1) is false, and therefore the whole argument disappears, since it is upon (1) that the whole rests.” I am ashamed to have to examine such reasoning, but it is necessary to do so, since it is common enough. Is it not palpable at first sight, that (1) and (2) are absolutely incompatible, that each contradicts the other flatly? You must choose between them, and, whichever you choose, the proof of Utilitarianism goes, because that springs from the unnatural conjunction of both. bradley 1876, 104, n. 1; cf. 1927, 114–115, n. 2
It is not easy to keep up with Bradley’s tendency to go back and forth between what can be desired, what is desired, and what is desirable, especially in this passage, which doubles the terminological confusion typically attributed to Mill. This being so, it is difficult in places to tell whether Bradley is paraphrasing that confusion or adding his own. Perhaps he is doing a little of both. Nevertheless, a charitable reading of Bradley demands that we lay all the blame at Mill’s doorstep, at least until we can make out exactly what Bradley is saying. Even so, he would have done better to avoid such a free association of words, especially since his objection here depends neither on the adoption nor on the rejection of the apparent confusion. Notwithstanding the misleading terminology, a straightforward reading of Bradley’s reasoning shows his evaluation to be on the level of the desired rather than on that of the desirable. He argues that (1) and (2) are mutually inconsistent, that Mill’s proof requires a conjunction of (1) and (2), and therefore that it involves a self-contradiction. Bradley’s (1) asserts that Mill is a psychological egoist. Bradley’s (2) is vague, but if it is to contradict his (1), then it presumably suggests that Mill is not a psychological egoist. This is what places his evaluation on the level of the desired. His objection turns out to be that Mill proceeds with psychological egoism as a premise, which he then contradicts where he extends psychological hedonism to include other people’s pleasures.
Historical Reception of Mill’s Proof
139
An even more charitable reading of Bradley, considering the admixture of the desirable in his critique, is that his insight into potential problems in Mill’s proof goes deeper than the simple objection that Mill fudges on the range of psychological hedonism. Bradley’s main point on that reading is that Mill fails to reconcile egoistic and universalistic ethical hedonism because Mill’s psychological egoism precludes universalistic ethical hedonism and confines him to egoistic ethical hedonism. This portion of Bradley’s critique draws attention to even more complex potential problems in the proof. One such problem is the possibility that Mill fails to advance beyond the egoistic to the universalistic level in psychological as well as ethical hedonism. A related but separate potential problem is that Mill fails to move from the psychological to the ethical domain at the egoistic as well as the universalistic level. More accurately, the problems Bradley sees here are not that Mill fails to make the moves in question, but that he makes them even though they are not logically warranted. Even if, as it is generally recognized, Mill does not advocate egoism, psychological or ethical, it can be asked how he proceeds from (universalistic) psychological hedonism to (universalistic) ethical hedonism. Bradley’s overall critique of the proof also includes the parts-whole inference usually associated with the fallacy of composition. In this matter, Bradley, like Sidgwick, ignores the possibility that Mill’s inference is limited to the realm of the desirable. The difference between the two critics in this respect is that Sidgwick takes Mill’s move from parts to whole to extend from the desired to the desirable, whereas Bradley restricts the same move to the level of the desired: Whether our ‘great modern logician’ thought that by this he had proved that the happiness of all was desirable for each, I will not undertake to say. He either meant to prove this, or has proved what he started with, viz. that each desires his own pleasure. And yet there is a certain plausibility about it. If many pigs are fed at one trough, each desires his own food, and somehow as a consequence does seem to desire the food of all; and by parity of reasoning it should follow that each pig, desiring his own pleasure, desires also the pleasure of all. But as this scarcely seems conformable to experience, I suppose there must be something wrong with the argument, and so likewise with the argument of our philosopher. bradley 1876, 103; cf. 1927, 113
Bradley includes a footnote with this passage to explain exactly what he thinks “must be” wrong “with the argument of our philosopher.” He there presents a
140
Chapter 3
false dilemma depicting our “great modern logician” as working either with an empty tautology or with a non sequitur (the latter of which seems to commit the fallacy of composition): Either Mill meant to argue, ‘Because everybody desires his own pleasure, therefore everybody desires his own pleasure’; or ‘Because everybody desires his own pleasure, therefore everybody desires the pleasure of everybody else.’ Disciples may take their choice. To us it matters not which interpretation be correct. In the one case Mill has proved his point by a pitiable sophism; in the other he has not proved any point at all. bradley 1876, 103, n. 1; cf. 1927, 113–114, n. 3
Bradley’s criticism of Mill’s parts-whole inference is significant in that he finds fault not only with the inference itself but also with an amendment Mill offers in explanation and justification of the inference. The amendment is in a letter from Mill to Henry Jones (Letter 1257: 13 June 1868), where he explains his generalization in the third paragraph of the fourth chapter of Utilitarianism as follows: As to the sentence you quote from my “Utilitarianism”; when I said that the general happiness is a good to the aggregate of all persons I did not mean that every human being’s happiness is a good to every other human being; though I think, in a good state of society & education it would be so. I merely meant in this particular sentence to argue that since A’s happiness is a good, B’s a good, C’s a good, &c., the sum of all these goods must be a good. mill CW16:1414
This is an important clarification. Interested readers may skip ahead to chapter 5 (section 5.2) for a detailed analysis of its implications. The point of bringing it up here is merely to affirm Bradley’s acknowledgment of the letter, which is not mentioned by any of Mill’s other early critics, at least not by the ones covered here. To be specific, in the second edition of his Ethical Studies, Bradley adds a footnote quoting the last sentence of the letter, thereby ostensibly accounting for Mill’s explanatory amendment to the original inference (1927, 113, n. 1). Yet Bradley is not at all impressed with Mill’s explanation: But, surely, in this explanation (even if we accept it) the ‘since’ implies a difference between the ‘sum’ and the several goods, and that the ‘sum’
Historical Reception of Mill’s Proof
141
is a good to the ‘aggregate’ of persons as distinguished from the several persons. If we suppose Mill to be thinking only of the ‘ascetic’ we may take him to argue thus: ‘There must be in an aggregate at least what there is in the individuals; therefore pleasure, as an end, belongs to the aggregate.’ This is correct, but it ignores the question ‘belongs how?’ And here Egoistic Hedonism comes in and wrecks the desired consequence. bradley 1927, 113, n. 1
Bradley’s contention, then, seems to be that the psychology implicit in the argument is not compatible with the morality embraced on that basis. That must be what he means by egoistic hedonism wrecking the desired consequence. That, in turn, is the articulation of a philosophical disagreement, not the discovery of a “monstrous” argument. Bradley does not explicitly reject the logic of the explanation in the letter. He just finds it inadequate to vindicate the original inference. He denies that each person’s happiness being a good to the person whose happiness it is establishes that the general happiness is a good to the aggregate of all persons. But his opposition is grounded in the psychological backdrop for the scenario. He simply does not agree that the general happiness can possibly be a good to the aggregate of all persons in a context where each person desires only his or her own happiness. As a matter of fact, Bradley’s examination in the main text (where the footnote originates), proceeding as it does with the example of multiple pigs being fed at a single trough (1876, 103; 1927, 113), remains on the level of the desired rather than on that of the desirable, despite the fact that Mill’s explanation in the letter makes no reference to desires. It seems, therefore, that Bradley merely acknowledges the letter without serious consideration. Regardless of whether the letter absolves Mill of the fallacy of composition, it affords greater insight into Mill’s meaning than Bradley brings out. That difference is preserved and explored in chapter 5 below (see section 5.2). 3.4
William Ritchie Sorley: The Ethics of Naturalism: A Criticism
Sorley aims at the usual targets in Mill’s proof: (1) the alleged confusion between desires and desirability motivating an illicit shift from the desired to the desirable, and (2) the purportedly misunderstood and misconstrued difference between personal and general happiness as a good. Announcing a special sensitivity to Mill’s warning that the attempt must not be interpreted or
142
Chapter 3
evaluated as proof in the ordinary sense, he takes Mill’s primary purpose to be to establish the universal as opposed to individual relevance of the summum bonum in utilitarian methodology: The formula of utilitarianism cannot be expressed as the conclusion of a syllogism or of an inductive inference. It seems rather to have been arrived at by the production—or the recognition—of a sympathetic or “altruistic” sentiment, which was made to yield a general principle for the guidance of conduct. This process involves two steps, which are consecutive and complementary, although the positions they connect are not necessarily related. The first step is to overcome the selfish principle of action in the individual; the second to generalise it, and obtain a principle for the non-selfish action that results. Mill seems to be the only recent writer who, in making this transition, adheres strictly to the psychological hedonism distinctive of his school. sorley 1885, 62; cf. 1904, 65–66
Much like Grote, Sorley refrains from dissecting the proof as if it were a deductive argument. His opposition thus shows a broadly philosophical as opposed to strictly logical orientation. But despite a strong grasp of Mill’s project, together with an explicit acknowledgment of the kind of proof it constitutes, Sorley concludes that the execution suffers from some of the same conceptual and logical problems others associate with the proof. This is not to say that he simply repeats the prevailing objections, just that he seems mostly to agree with them. Otherwise, he does provide his own insight and emphasis: In respect of his main contention, that utilitarianism is a theory of beneficence, and not of prudence or of selfishness, Mill emphasised even more strongly than Bentham had done the distinction between the egoism which seeks its own things, and the utilitarianism according to which everybody counts for one, and nobody for more than one. But when he attempted to connect this doctrine logically with the psychological postulates of his school, he committed a double error. In the first place, he confused the purely psychological question of the motives that influence human conduct with the ethical question of the end to which conduct ought to be directed; and, in the second place, he disregarded the difference of end there may be for society as a collective whole, and for each member of the society individually. “There is in reality,” he says, “nothing desired except happiness;” and this psychological theory is too hastily identified with the ethical principle that happiness alone is desirable, or
Historical Reception of Mill’s Proof
143
what ought to be desired and pursued. Moreover, “no reason,” he says, “can be given why the general happiness is desirable, except that each person, so far as he believes it to be attainable, desires his own happiness.” And this admission, which seems as good as saying that no reason at all can be given why the individual should desire the general happiness, is only held to be a sufficient reason for it, through assuming that what is good for all as an aggregate is good for each member of the aggregate: “That each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons.” sorley 1885, 60–61; cf. 1904, 63–64
With respect to Mill’s reasoning in grounding the desirability of happiness in desires for happiness, Sorley’s reaction is not that Mill equivocates on “desirability” but that he moves from desires to desirability without sufficient evidence. There is, therefore, a problem in the perceived logic, but the problem is not in the form of a logical fallacy. The problem, according to Sorley, is that Mill “too hastily identified” the psychological theory recognizing happiness as the only thing that is desired with the ethical theory recognizing happiness as the only thing that is desirable. Yet Sorley does not go on, as many others do, to add insult to injury. He does not attribute the haste in question to a misunderstanding regarding the meaning of “desirability.” Instead, he leaves Mill with an unjustified ethical principle associated too eagerly with a psychological principle presumed too hastily to provide adequate support for it. Sorley’s main target is the next step in the proof, the relationship between parts and whole. He enriches the standard critique of this part of the proof, usually taken to commit the fallacy of composition, with the suggestion that Mill commits the fallacy of division (be it instead of or in addition to the fallacy of composition). While this possibility regarding the corresponding argument concerns Sidgwick and Dewey as well, it seems to be the only thing that concerns Sorley. Sorley attributes Mill’s compositional inference to an initiative to justify “why the individual should desire the general happiness” (1885, 61). It is not entirely clear whether he intends “should” in a predictive and descriptive sense to denote what is to be anticipated naturally by rational expectations, or in a normative and prescriptive sense to denote what is to be required categorically by moral considerations (or by prudential ones). But it is at least charitable, if not otherwise plausible, to read Sorley as intending the normative sense of “should,” and thus, unlike Sidgwick and Bradley, as recognizing the possibility that Mill is reasoning about the desirable and not the desired.
144
Chapter 3
Sorley’s interpretation differs from most others, though not so much from Sidgwick’s and Dewey’s, in objecting that Mill commits the fallacy of division here, thus deviating from the tendency to attribute the fallacy of composition to the same part of the proof. This is evident in his interpretation that Mill aims to justify “why the individual should desire the general happiness” and in his claim that Mill assumes that “what is good for all as an aggregate is good for each member of the aggregate” (Sorley 1885, 61). This is not to say that Sorley thinks Mill commits the fallacy of division instead of the fallacy of composition. He seems to believe that Mill commits at least the fallacy of division whether or not he also commits the fallacy of composition. This is how and why Sorley’s interest in the passage is different from the critical norm. Unlike Sidgwick and Bradley, Sorley is not concerned with whether each person’s desiring his own happiness establishes the desirability of the general happiness for the aggregate of all persons. That would be the place to bring up charges of the fallacy of composition. Sorley passes on that. He supposes that, in order to prove universalistic ethical hedonism, Mill must show that the general happiness is desirable for each person (that each person “should” desire the general happiness), not that the general happiness is desirable for the aggregate of all persons. Hence, Sorley does not openly challenge Mill’s reasoning in deriving the desirability of the general happiness for the aggregate of all persons from the desirability of personal happiness for the person whose happiness it is. He instead objects that Mill cannot logically infer the desirability of the general happiness for each person, without committing the fallacy of division, whether or not he can arrive at the desirability of the general happiness for the aggregate of all persons, without committing the fallacy of composition. The reason why Mill cannot make this deduction, Sorley contends, is that to do so would be to commit the fallacy of division through the assumption that “what is good for all as an aggregate is good for each member of the aggregate” (1885, 61). 3.5
John Dewey: Outlines of a Critical Theory of Ethics
Dewey goes through various parts of Mill’s proof in varying levels of detail. He does this both in his personal outlines of ethical theory (1891) and in his collaboration with Tufts (1908/1932). He disagrees with a great deal but speaks of only one fallacy. Like some of the others before him, Dewey accuses Mill of committing the fallacy of division. One difference, though, is that Dewey presents the charge directly and explains it clearly. What remains largely implicit in Sidgwick and Sorley becomes explicit in Dewey:
Historical Reception of Mill’s Proof
145
But does it follow that because the happiness of A is an end to A, the happiness of B an end to B, and the happiness of C an end to C, that, therefore, the happiness of B and C is an end to A? There is obviously no connection between the premises and the supposed conclusion. And there appears to be, as Mill puts it, only an account of the ambiguity of his last clause, “the general happiness a good to the aggregate of all persons.” The good of A and B and C may be a good to the aggregate (A +B + C), but what universalistic hedonism requires is that the aggregate good of A +B +C, be a good to A and to B and to C taken separately—a very different proposition. Mill is guilty of the fallacy known logically as the fallacy of division—arguing from a collective whole to the distributed units. Because all men want to be happy, it hardly follows that every man wants all to be happy. There is, accordingly, no direct road from individualistic hedonism—private pleasure—to universalistic—general pleasure. dewey 1891, 55–56
Dewey also recognizes that Mill’s generalization is entirely about the desirable and not at all about the desired. That is why he avoids references to desires in his analysis of the argument. In fact, given the tendency to associate the term “desirable” with its root word “desire,” Dewey avoids referring to the desirable altogether, opting to speak instead of the Millian equivalent, good. Dewey’s terminology is so similar to the way Mill phrases his explanation to Henry Jones that it might appear reasonable to suppose that Dewey was aware of the corresponding letter (Letter 1257: 13 June 1868; CW16:1414). However, Dewey’s first sentence in the passage above makes this doubtful because he questions precisely what Mill denies in the letter to be doing in the proof. Curiously, for all his opposition, Dewey seems willing to grant not only that “A and B and C may be a good to the aggregate (A +B +C)” but also that Mill may legitimately infer this from his observation that “the happiness of A is an end to A, the happiness of B an end to B, and the happiness of C an end to C.” He thus offers no resistance to Mill’s move from the personal level to the general and aggregate levels, but he imagines and protests a subsequent move in the opposite direction. In all fairness to Dewey, that subsequent move is not imagined without cause, as Mill himself provides no directions for parsing the amorphous desirability of the general happiness for the aggregate of all persons. The intellectual motivation to seek further clarification seems both natural and justified, as does the original clarification itself. This does not make it correct, but it provides a workable alternative where what is correct is already in question.
146
Chapter 3
Nevertheless, we are still left with the curiosity that Dewey’s rendition allows the fallacy of composition while censuring the fallacy of division where both are presumably present in the same context. 3.6
John Dewey and James Hayden Tufts: Ethics
The chain of reasoning in the passage quoted above from Dewey (1891) temporarily finds its way into the collaboration between Dewey and Tufts on ethics, appearing in different words in the first edition (1908), though not at all in the second edition (1932): It clearly does not follow that because the good of A and B and C, etc., is collectively, or aggregately, a good to A and B and C, etc., that therefore the good of A and B and C, etc., or of anybody beyond A himself, is regarded as a good by A—especially when the original premise is that A seeks his own good. Because all men want to be happy themselves, it hardly follows that each wants all to be so. It does follow, perhaps, that that would be the reasonable thing to want. If each man desires happiness for himself, to an outside spectator looking at the matter in the cold light of intelligence, there might be no reason why the happiness of one should be any more precious or desirable than that of another. From a mathematical standpoint, the mere fact that the individual knows he wants happiness, and knows that others are like himself, that they too are individuals who want happiness, might commit each individual, theoretically, to the necessity of regarding the happiness of every other as equally sacred with his own. But the difficulty is that there is no chance, upon the hedonistic psychology of desire, for this rational conviction to get in its work, even if it be intellectually entertained. The intellectual perception and the mechanism of human motivation remain opposed. Mill’s statement, in other words, puts the problem which hedonistic utilitarianism has to solve. dewey and tufts 1908, 290–291; passage deleted in the 1932 edition
A noticeable difference between the analysis in Dewey’s earlier outline (1891) and the discussion in the joint production with Tufts (1908), other than the different wording, is the disappearance of the mention of any fallacy. This may not be all that telling, however, as the two critiques amount to the same thing. A fallacy charge in one can reasonably be assumed to be implicit in the other. Dewey and Tufts (1908/1932) also protest the analogic transition from the desired to the desirable in the first part of Mill’s proof. Their dissatisfaction
Historical Reception of Mill’s Proof
147
with that move seems to grow stronger over time. In the first edition of their collaboration, they quote the relevant portion of the proof without strong criticism, recounting Mill’s inductive inference regarding desirability as “axiomatic and not susceptible of proof” (1908, 264). The worst that they then make of the move is to describe Mill as proceeding with the “self-evident character of happiness as the all-desirable” (1908, 265) as part of his broader “axiomatic identification of good with happiness” (1908, 264). In the second edition, they still do not accuse Mill explicitly of a specific logical fallacy or conceptual confusion, but their opposition becomes more conspicuous and more formidable: Again, the ending “able” has two meanings in different words. It signifies “capable of being seen,” when it occurs in the word “visible.” But in other words, it signifies that which is fit, proper, as in the words “enjoyable,” “lovable.” “Desirable” signifies not that which is capable of being desired (experience shows that about everything has been desired by some one at some time) but that which in the eye of impartial thought should be desired. It is true, of course, that it would be foolish to set up anything as the end of desire, or as desirable, which is not actually desired or capable of being desired. But it would be equally stupid to assume that what should be desired can be determined by a mere examination of what men do desire, until a critical examination of the reasonableness of things desired has taken place. dewey and tufts 1932, 206
Despite drawing the familiar distinction between that which is “capable of being desired” and that which “should be desired,” Dewey and Tufts are not accusing Mill here of equivocating on the term “desirable.” Nor are they accusing him of confusing psychological questions with ethical ones. They are instead emphasizing the importance of a thorough critical examination before accepting what is desired as the foundation of what is desirable. The strong wording they use (as in “foolish” and “stupid”) may seem to overshoot the typical mark of emphasis, but emphasis is indeed all they are after. Note that their use of the adjective “stupid,” introducing a point concerning Mill’s perspective, is balanced by their use of the adjective “foolish,” representing the opposite perspective, that is, the outright rejection of Mill’s perspective. There is no particular attack on Mill here, at least no disproportionate treatment, only a strong but balanced emphasis on how tricky the attempted move can be and how much care it therefore requires. Dewey and Tufts acknowledge that what is desired constitutes some sort of evidence for what is desirable. What they deny is simply that there is a
148
Chapter 3
compelling connection, or even a reliable one, between the desired and the desirable. They are well aware that Mill himself does not find the connection logically necessary or causally determinate. The point they are pressing is that the burden of proof is still on Mill even if such questions do not admit of proof in the ordinary sense (U1:5–6 CW10:207–208; U4:1 CW10:234; U4:9 CW10:237). 3.7
John Stuart Mackenzie: A Manual of Ethics
Mackenzie stands out among the early critics of Mill’s proof with the depth of his analysis, the clarity of his exposition, and the accuracy of his understanding of what is at stake. It is unfortunate that students of Mill are referred almost exclusively to Moore, and to some extent to Sidgwick, for a first-hand account of historical reactions to the proof. The critiques examined in this chapter are all worthwhile, to be sure, but Mackenzie’s contribution is indispensable. Like most critics, Mackenzie focuses on the two familiar aspects of Mill’s proof. His discussion of the alleged equivocation on “desirable” is comparable to that of others but with an edge in articulation and argumentation. The real strength of his opposition, however, is in his analysis of the portion of the proof allegedly committing the fallacy of composition. An examination of both is in order for a fruitful comparison with other commentaries. Mackenzie offers the following appraisal of Mill’s analogy between sensations and desires in moving from visibility and audibility to desirability (U4:3 CW10:234): It is here assumed that the meaning of the word “desirable” is analogous to that of “visible” and “audible.” But “visible” means “able to be seen,” and “audible” means “able to be heard”; whereas “desirable” does not usually mean “able to be desired.” When we say that anything is desirable, we do not usually mean merely that it is able to be desired. There is scarcely anything that is not able to be desired. What we mean is rather that it is reasonably to be desired, or that it ought to be desired. When the Hedonist says that pleasure is the only thing that is desirable, he means that it is the only thing that ought to be desired. But the form of the word “desirable” seems to have misled several writers into the notion that they ought to show also that pleasure is the only thing that is capable of being desired. The latter view is that of psychological Hedonism, which seems clearly to be unsound. The former is that of ethical Hedonism, which we have still to examine. mackenzie 1893, 98–99; cf. 1894, 98–99; 1897, 213–214; 1900/1901, 213–214; 1915, 215–216; 1929, 169
Historical Reception of Mill’s Proof
149
The reading here is that Mill misconstrues the desirability in question as he equivocates on the term “desirable” in attempting to ground ethical hedonism in psychological hedonism. Mackenzie takes Mill to be moving from the premise that “pleasure is the only thing that is capable of being desired” to the conclusion that “it is the only thing that ought to be desired,” wherein the alleged equivocation on “desirable” occurs because Mill represents both the premise and the conclusion with the statement that “pleasure is the only thing that is desirable.” Hence, it is only through a corruption of the meaning of desirability that pleasure as the sole psychological motivation for action becomes pleasure as the ultimate standard of value. Mackenzie does not present the apparent equivocation as a deliberate manipulation or obfuscation on the part of Mill to fake the reasoning in an otherwise unworkable proof. He merely notes that the equivocation is there. He provides a footnote identifying the confusion between the two senses of “desirable” as the “ ‘Fallacy of Figure of Speech’ (figuræ dictionis)” (Mackenzie 1893, 99, n. 1). This is consistent with the common charge that the fallacy involved is that of equivocation. Mackenzie is claiming, in effect, that Mill equivocates on “desirable” through the fallacy of figure of speech, drawing a similarity of meaning between certain words from a similarity of construction between those words, pertaining here to the superficial structural similarity between “visible” (or “audible”) and “desirable.” Mackenzie’s main contribution to the secondary literature on Mill’s proof is in his deconstruction of the inference taken to contain the fallacy of composition. Most critics who accuse Mill of committing the fallacy of composition here tend to focus either on the move from individual persons to the aggregate of all persons as the unit of reference or on the move from personal happiness to the general happiness as the standard of value. Mackenzie fully appreciates the complexity of Mill’s parts-whole inference. At the level of the subject, he detects (or suspects) a move from individual persons to the aggregate of all persons. At the level of the object, he detects (or suspects) an additional move from personal happiness to the general happiness. Here is what he says about the whole: The fallacy is that which is known in logic as “the fallacy of composition.” It is inferred that because my pleasures are a good to me, yours to you, his to him, and so on, therefore my pleasures +your pleasures +his pleasures are a good to me +you +him. It is forgotten that neither the pleasures nor the persons are capable of being made into an aggregate. It is as if we should argue that because each one of a hundred soldiers is six feet high, therefore the whole company is six hundred feet
150
Chapter 3
high. The answer is that this would be the case if the soldiers stood on one another’s heads. And similarly Mill’s argument would hold good if the minds of all human beings were to be rolled into one, so as to form an aggregate. But as it is, “the aggregate of all persons” is nobody, and consequently nothing can be a good to him. A good must be a good for somebody. mackenzie 1893, 104; cf. 1894, 104; 1897, 219–220; 1900/1901, 219–220; 1915, 221–222; 1929, 173–174
Mackenzie thus advances two related but distinct objections corresponding to the two related but distinct moves in Mill’s parts-whole inference. He argues that the general happiness cannot be a good to the aggregate of all persons, because neither the generalization nor the aggregation works. The general happiness, a conventional construct, is not itself anyone’s happiness, and the aggregate of all persons is not itself a person. To be precise, Mackenzie speaks of pleasure, not of happiness. Although Mill clearly presents happiness as a function of pleasure and pain (U2:2 CW10:210), Mackenzie makes a special effort to keep the discussion anchored to pleasure, avoiding any mention of happiness after an initial reference to the “general happiness,” subsequently reduced to a sum of pleasures.7 This makes Mill’s inference seem sillier than it might appear under any other interpretation. Mackenzie goes out of his way to fortify this impression by explicating the problem as a failure of the felicific calculus: Pleasures cannot be Summed.—It follows from this that there cannot be any calculus of pleasures—i. e. that the values of pleasures cannot be quantitatively estimated. For there can be no quantitative estimate of things that are not homogeneous. But, indeed, even apart from this consideration, there seems to be a certain confusion in the Hedonistic idea that we ought to aim at a greatest sum of pleasures. If pleasure is the one thing that is desirable, it is clear that a sum of pleasures cannot be desirable; for a sum of pleasures is not pleasure. We are apt to think that a sum of pleasures is pleasure, just as a sum of numbers is a number. But this is evidently not the case. A sum of pleasures is not pleasure, any more than a sum of men is a man. For pleasures, like men, cannot be added to
7 Hastings Rashdall (1899), a contemporary of Mackenzie, may be consulted profitably for a critical appraisal of conceptual problems in the aggregation of pleasures, with his study now holding historical interest as well as providing philosophical insight.
Historical Reception of Mill’s Proof
151
one another. Consequently, if pleasure is the only thing that is desirable, a sum of pleasures cannot possibly be desirable. mackenzie 1893, 112–113; cf. 1894, 112–113; 1897, 229–230; 1900/1901, 229–230; 1915, 231–232; 1929, 182
Much of Mackenzie’s discussion is sensible, most of it compelling, and all of it clear. But charity is missing, perhaps along with some fairness. Charity is optional, of course, but fairness is not. Even if everything Mackenzie says about the felicific calculus were true, it would still appear oddly recalcitrant of him to refuse to make any more of Mill’s conception of the general happiness than to define it as a sum of pleasures. The reason for this interpretive conflict is that Mackenzie otherwise demonstrates substantial sensitivity to the intricacies of the relationship between happiness and pleasure (1893, 110, 113–114, 115–116, 263). This is plain to see, for instance, in a passage revealing his keen insight into what is meant by the general happiness: If the theory of Universalistic Hedonism is accepted, and if this theory is made to rest on the basis of Psychological Hedonism, it becomes important to consider the motives by which the individual is led to seek the general happiness. His primary desire, according to this view, is for his own greatest happiness; and he can be led to seek the general happiness only by being led to see that the conduct which leads to “the greatest happiness of the greatest number” is in the long run identical with that which leads to his own greatest happiness. mackenzie 1893, 263; cf. 1894, 263; 1897, 394; 1900/1901, 261; 1915, 287; 1929, 248–249
Nevertheless, Mackenzie stands alone among Mill’s early critics in scrutinizing the parts-whole inference, holding it to the highest standard, and advancing the strongest criticism against it. For Mill’s inference to hold, both the general happiness and the aggregate of all persons must be compositionally viable, but neither one is, argues Mackenzie, because neither pleasures nor persons can be summed. Mackenzie thus charges Mill with two counts of the fallacy of composition in one argument. This constitutes the basic model for discussion in the chapter dedicated in the present volume to the alleged fallacy of composition in Mill’s proof (chapter 5).
152 3.8
Chapter 3
George Edward Moore: Principia Ethica
Moore’s critique is the most popular of the historical reactions to Mill’s proof of the principle of utility. One reason for this is that, independently of any connection with Mill, Moore’s Principia Ethica is generally acknowledged to be a greater contribution to ethical theory than are any of the other works considered here. Another reason is that, with respect to Mill’s proof, Moore is widely known for exposing a category mistake he calls the “naturalistic fallacy,” in addition to recognizing the usual logical fallacies. The naturalistic fallacy is the subject matter of chapter 6 below. However, it is entirely possible, and at this point useful, to isolate from Moore’s discussion of the naturalistic fallacy, passages in which he illustrates Mill’s alleged equivocation on “desirable” and those in which he discusses purported problems with Mill’s inference concerning the general happiness. To begin with the charge of equivocation, consider Moore’s following discussion of the fallacy he detects in what he appropriately describes as the first step in Mill’s proof, the initial move from the desired to the desirable: Well, the fallacy in this step is so obvious, that it is quite wonderful how Mill failed to see it. The fact is that ‘desirable’ does not mean ‘able to be desired’ as ‘visible’ means ‘able to be seen.’ The desirable means simply what ought to be desired or deserves to be desired; just as the detestable means not what can be but what ought to be detested and the damnable what deserves to be damned. … Well, then, the first step by which Mill has attempted to establish his Hedonism is simply fallacious. He has attempted to establish the identity of the good with the desired, by confusing the proper sense of ‘desirable,’ in which it denotes that which it is good to desire, with the sense which it would bear if it were analogous to such words as ‘visible.’ If ‘desirable’ is to be identical with ‘good,’ then it must bear one sense; and if it is to be identical with ‘desired,’ then it must bear quite another sense. And yet to Mill’s contention that the desired is necessarily good, it is quite essential that these two senses of ‘desirable’ should be the same. If he holds they are the same, then he has contradicted himself elsewhere; if he holds they are not the same, then the first step in his proof of Hedonism is absolutely worthless. moore 1903, § 3:40.4, pp. 118–119
What exactly is the problem here? Is it that Mill thinks “desirable” means “able to be desired,” or is it that he thinks “desirable” means “desired,” or is it that he thinks “desirable” means “good to desire”? Evidently, the last one is not a
Historical Reception of Mill’s Proof
153
problem, as Moore considers “good to desire” to be interchangeable both with “ought to be desired” and with “deserves to be desired,” all three of which he takes to capture the proper sense of “desirable.” Yet Moore’s analysis becomes tortuous as he contrasts “the proper sense of ‘desirable,’ in which it denotes that which it is good to desire” with two mistaken interpretations of “desirable”: one in which “desirable” is taken to mean “able to be desired,” the other in which “desirable” is taken to mean “desired.” This, too, is not a problem, if we grant Moore the freedom of expression he never grants Mill, because Moore obviously finds “desired” and “able to be desired” mutually relevant in this context, where the contrast he intends is a broader one between the descriptive and the normative. Then again, if Moore’s choice of words here were scrutinized as vigilantly as that of Mill, we could, both with accuracy and with fairness, charge him with conceptual confusions in the same piece in which he sets out to expose conceptual confusions in Mill. It is a matter of fact, not of opinion or interpretation, that Mill could not have confused the proper sense of “desirable” both with “able to be desired” and with “desired” at the same time, not even if he were the simpleton Moore makes him out to be. He could conceivably have been confused in one way or the other, to be fair, but not in both at once. Yet Moore repeatedly and freely switches between the two. He even reaffirms the same loose accusation later when recapitulating his objections to Mill’s argument (Moore 1903, § 3:44.3–4, pp. 124–125): – “First of all, he [Mill] takes ‘the desirable,’ which he uses as a synonym for ‘the good,’ to mean what can be desired” (§ 3:44.3). – “In Mill’s case, good is thus supposed to mean simply what is desired” (§ 3:44.3). – “Mill’s first argument then is that, because good means desired, therefore the desired is good” (§ 3:44.4). What is, in fact, desired would ipso facto have to be something that can be desired. And everything that can be desired very likely ends up being desired, given that there are so many of us with such variation in our likes and dislikes as to leave nothing undesired that can be desired. So, why not allow Moore to switch freely between the two expressions? Because they do not express the same thing. The point of highlighting the discrepancy is not to get Moore back for holding Mill’s prose to such exacting standards but to remind Moore of his own position as one of the pioneers of analytic philosophy and of his habitual insistence on prefacing any discussion with a clarification of the meaning of the terms involved. Granted, demonstrating problems in Moore’s critique does not solve any in Mill’s proof. Moore could have been careless or wrong, or both, without Mill
154
Chapter 3
being careful or right. But there is something in the way that Moore’s critique is equivocal (sloppy but not fallacious) that could and should have alerted him (but evidently did not) to the possibility that Mill’s understanding of desirability was not equivocal (fallacious). What makes it ultimately acceptable for Moore to go back and forth between “desired” and “can be desired” is that what is desired is (at least) a subset of what can be desired. We can be certain that something can be desired if that thing actually is desired. Might not Mill, then, be working with reasoning that is comparable in some way, even if not in that way, and admittedly not in every way? Perhaps all that Mill is saying is that something’s being desired is indicative of its being desirable in the proper sense. He need not be equivocating on desirability while pursuing the good. Nor would he have to take such an indication to be conclusive evidence. He need only find it useful. That would be enough to explain why he appeals to it. And if Mill had thought that desires were the only evidence of desirability—not necessarily the only evidence conceivable but the only evidence available—it would then have been natural for him to proceed on that basis, even if he did not consider it proof in the ordinary sense, and, in fact, especially if he did not consider it proof in the ordinary sense. This alternative interpretation may not keep Mill from being wrong about the matter, but it does keep him from appearing utterly incompetent in arguing about it. Mill could still be wrong, but he would then be wrong without equivocating on such a familiar word as “desirable.” It is one thing to be wrong about something, quite another to be bad at reasoning about it. Moore neglects the possibility that Mill is working with what Mill himself considers a merely indicative evidentiary relationship between desires and desirability in the proper sense, as opposed to deducing what can be desired from what is desired while thinking instead that he had thereby demonstrated what it is good to desire. The reason why Moore neglects this possibility might be that he is sidetracked with the conviction that Mill is trying to “establish the identity of the good with the desired” in order to show that “the desired is necessarily good” (Moore 1903, § 3:41.1, p. 119). Much of this is taken up in detail in chapter 4, in examining the alleged equivocation on “desirable,” and some of it again in chapter 6, in discussing the naturalistic fallacy. The motivation for providing a preview of sorts at this point is to avoid leaving the historical survey without even a hint of an alternative in Mill’s favor. This is especially important in connection with Moore, whose critique of Mill’s proof has long been overrated due to the reputation of his overall contribution to philosophy. Next, consider Moore’s assessment of Mill’s inference in the move from personal to general happiness. Moore does not accuse Mill here, at least not
Historical Reception of Mill’s Proof
155
explicitly, of committing a logical fallacy, such as composition or division. Yet he does find a logical problem in the transition from the desirability of each person’s happiness to the desirability of the general happiness. Here is Moore’s summary and evaluation of the corresponding passage: Mill’s argument is as follows: A man desires his own happiness; therefore his own happiness is desirable. Further: A man desires nothing but his own happiness; therefore his own happiness is alone desirable. We have next to remember, that everybody, according to Mill, so desires his own happiness: and then it will follow that everybody’s happiness is alone desirable. And this is simply a contradiction in terms. Just consider what it means. Each man’s happiness is the only thing desirable: several different things are each of them the only thing desirable. This is the fundamental contradiction of Egoism. In order to think that what his arguments tend to prove is not Egoism but Utilitarianism, Mill must think that he can infer from the proposition ‘Each man’s happiness is his own good,’ the proposition ‘The happiness of all is the good of all’; whereas in fact, if we understand what ‘his own good’ means, it is plain that the latter can only be inferred from ‘The happiness of all is the good of each.’ Naturalistic Hedonism, then, logically leads only to Egoism. Of course, a Naturalist might hold that what we aimed at was simply ‘pleasure’ not our own pleasure; and that, always assuming the naturalistic fallacy, would give an unobjectionable ground for Utilitarianism. But more commonly he will hold that it is his own pleasure he desires, or at least will confuse this with the other; and then he must logically be led to adopt Egoism and not Utilitarianism. moore 1903, § 3:62, p. 156
Moore claims that Mill’s proof collapses in self-contradiction before establishing the conclusion of the first part, that happiness is desirable as an end in itself. His primary objection is that it is contradictory to hold each of several different things, mutually incompatible or irreconcilable ones at that, to be the only thing desirable. It is indeed patently contradictory for entirely different things to be unique in the very same way, which is what we have in every single thing under consideration being the only thing desirable. But Moore is providing instantiations, on Mill’s behalf, of an inescapable general formula for contradiction. On this reasoning, it is also contradictory to hold each of several things to be the only thing desired, which is a problem that must arise before the contradiction concerning the several things each of which is supposed to be the only thing desirable. However, Moore passes over the level of
156
Chapter 3
the desired, where each of several people’s happiness is the only thing desired, and focuses on the level of the desirable, where each of several people’s happiness is the only thing desirable. If Mill’s proof fails where it reaches what Moore claims to be the fundamental contradiction of ethical egoism, then, by the same reasoning, it must fail even before that, because it starts with what might be called the fundamental contradiction of psychological egoism. On the level of the desired, Mill claims that each person desires his or her own happiness, without either affirming or denying, in addition, that anyone might desire anyone else’s happiness. That particular question is left open in the passage to which Moore refers (U4:3 CW10:234). Mill instead offers the observation that each person desires his or her own happiness, as evidence that each person’s happiness is desirable (specifically for that person but not necessarily only for that person), which he then proposes as evidence that the general happiness is desirable. The crux of the matter is that happiness, whether it is one’s own happiness or someone else’s, is a desirable sort of thing (and, in fact, the only sort of thing that is desirable as an end in itself). On the level of the desirable, then, Mill is generalizing from personal to general happiness, both desirable, without contradiction, as the same kind of thing, namely as happiness. Indeed, if happiness is desirable, that is, if happiness is a good, why should it be a good only at the personal level and not at the general level? The apparent contradiction in this context arises from Moore’s overeager anticipation of Mill’s conclusion for the entire proof: Happiness is the only thing desirable as an end in itself. Moore, it seems, contrasts the theses of the two parts of Mill’s proof. He criticizes the thesis of the first part of the proof, that happiness is desirable as an end in itself, which is supported by the claim that each person’s happiness is desirable for that person, by opposing that thesis to the thesis of the second part of the proof, that nothing other than happiness is desirable as an end in itself. Both theses are on a general as opposed to personal level. However, the juxtaposition of the two theses, catalyzed by the claim supporting the first thesis, somehow generates the illicit synthesis that each person’s happiness is the only thing desirable as an end in itself. Moore himself acknowledges an alternative interpretation. He considers the possibility that, throughout the proof, Mill’s assessment of happiness as the greatest good concerns the general happiness and not personal happiness. He admits that, interpreted in this way, the proof is not overturned by the contradiction he initially describes, but he objects that the proof then proceeds only at the expense of committing what he calls the “naturalistic fallacy,” examined in detail in chapter 6 below. The next two chapters absolve Mill of the two logical fallacies traditionally and still attributed to his proof of the principle of utility: equivocation and
Historical Reception of Mill’s Proof
157
composition. Chapter 4 shows that he is not guilty of equivocation in exploring desirability through desires, not even in taking the latter as the only evidence available for the former. Chapter 5 shows that he is not guilty of committing the fallacy of composition in moving from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons. The fallacy of division, on the other hand, neither merits nor gets a separate chapter, because it becomes incidentally but sufficiently clear in the chapter on the fallacy of composition that, if the interpretation proposed for that portion of Mill’s proof is correct, not only does the proof not commit the fallacy of composition, but it also does not commit the fallacy of division. The way that Mill is said to commit the fallacy of division, by inferring either that the general happiness is desired by each person or that the general happiness is desirable for each person (thus exhausting the possibilities at the personal level), makes it a step in the process of committing the fallacy of composition, that is, in concluding that the general happiness is desirable for the aggregate of all persons. This is not the only way to commit the fallacy of composition in the context of the proof, but there is no other way to commit the fallacy of division there. That is why a whole chapter is dedicated to the fallacy of composition and none to the fallacy of division.8 8 Note that the charge that Mill commits the fallacy of division in his proof seems to be one of the objections Jerome Borges Schneewind reports to have eventually dropped out of the corpus of secondary literature (1976, 41).
pa rt 2 The Alleged Fallacies in Mill’s Proof of the Principle of Utility: Analysis and Response
∵
c hapter 4
The Alleged Fallacy of Equivocation in Mill’s Proof The aim of this chapter is to absolve Mill of charges that he equivocates on the word “desirable” in the process of attempting to prove the principle of utility in the third paragraph of the fourth chapter of Utilitarianism. It is organized in three sections. The first section elucidates the charges against Mill, illustrating how and where in the proof he is alleged to commit the fallacy. The second section analyzes the nature of the charges through alternative formulations of the portion of the proof allegedly equivocating on the meaning of desirability. The third section responds to the charges, illuminating the true methodology behind the proof and thereby vindicating Mill’s appeal to desires as evidence of desirability. 4.1
Charges against Mill
The allegation of equivocation is grounded in Mill’s “analogic argument” for the desirability of happiness. This is where he asserts in quick succession, apparently in the same spirit and with the same intent, that the only proof possible for anything’s visibility is its being seen, the only proof possible for anything’s audibility is its being heard, and the only evidence available for anything’s desirability is its being desired (U4:3 CW10:234). Here is the analogy in Mill’s own words: The only proof capable of being given that an object is visible, is that people actually see it. The only proof that a sound is audible, is that people hear it: and so of the other sources of our experience. In like manner, I apprehend, the sole evidence it is possible to produce that anything is desirable, is that people do actually desire it. mill U4:3 CW10:2341
1 Critics often gloss over the switch from “proof” to “evidence” as Mill moves from visibility and audibility to desirability. As discussed in the main text below (subsection 4.3.3), Mill takes sights and sounds as proofs of visibility and audibility, but he regards desires as nothing more than evidence of desirability. Note the correlative switch from the only proof possible (“only proof capable of being given”) to the only evidence available (“sole evidence it is possible to produce”), where the former invokes necessary and sufficient conditions for a conclusive demonstration, while the latter simply acknowledges what we have to work with as evidence.
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_006
162
Chapter 4
Critics detect three implicit arguments in this passage that are identical in logical form: – If something is seen, then it is visible; people see x; therefore, x is visible. – If something is heard, then it is audible; people hear x; therefore, x is audible. – If something is desired, then it is desirable; people desire x; therefore, x is desirable. These arguments are all comparable in form, but the third one is different in meaning. The word “visible” refers to what can be seen, what is able to be seen, what is capable of being seen, what it is possible to see, and so on. Likewise, the word “audible” refers to what can be heard, what is able to be heard, what is capable of being heard, and what it is possible to hear. Yet the word “desirable” does not refer to what can be desired, what is able to be desired, what is capable of being desired, or what it is possible to desire. It is generally taken instead to refer to what is worthy of desire, what is fit to be desired, what it is good to desire, and what ought to be desired. While colloquial usage suggests that the same adjectival suffix (“able” or “ible”) affects these three words differently, Mill speaks of desirability in the same breath as visibility and audibility, which then makes it seem as if he uses the same suffix uniformly to indicate capacity, ability, or possibility with respect to vision, hearing, and desire. Based on this evidence, critics can reasonably charge Mill with giving a strange new meaning to the word “desirable,” construing it as denoting that which can be desired, that which is able to be desired, that which is capable of being desired, that which it is possible to desire, and so on with any alternative phrasing indicative of possibility. This would not be logically problematic if Mill were inclined to preserve the strange new meaning throughout the argument. If he were to take “desirable” to mean something like “can be desired,” for example, then substitution of “happiness” for “x” in the third argument above would produce the following instantiation: If something is desired, then it can be desired; people desire happiness; therefore, happiness can be desired. This is a valid and sound argument establishing the philosophically unexciting conclusion that it is possible to desire happiness. Nobody would bother to advance such an argument. It serves no practical purpose. An obvious alternative is that Mill might be using “desirable” consistently in its colloquial sense throughout the argument, never intending the strange new meaning. To illustrate, if Mill takes “desirable” to mean something like “worthy of desire,” then substitution of “happiness” for “x” in the third argument above produces the following instantiation: If something is desired,
Alleged Fallacy of Equivocation
163
then it is worthy of desire; people desire happiness; therefore, happiness is worthy of desire. This is also a valid argument, but it is not a sound one, because not everything people desire is actually worthy of desire. Similarly, if Mill takes “desirable” to mean something like “ought to be desired,” then substitution of “happiness” for “x” in the same argument produces the following instantiation: If something is desired, then it ought to be desired; people desire happiness; therefore, happiness ought to be desired. Again, this is a valid argument, but it is not a sound one, because not everything everyone desires ought really to be desired. Already, it is both possible and reasonable to grant the validity of Mill’s argument, while accusing him of advancing either a pointless but sound argument or an ambitious but unsound one. If Mill is interpreted as consistently taking “desirable” to mean something like “can be desired,” then his argument is valid and sound but not particularly useful. If he is interpreted as consistently taking “desirable” to mean something like “worthy of desire” or “ought to be desired,” then his argument is valid but not sound. Critics invariably follow an altogether different route. Presenting a valid argument with various problems is bad enough, but advancing an invalid argument containing a logical fallacy is even worse, especially for a reputable logician and philosopher. This is not to say that critics deliberately seek out the kind of critique that might best embarrass Mill, but that is indeed the kind of critique they end up offering. They do not read the analogic argument as preserving the meaning of “desirable” from start to finish. They maintain instead that Mill uses “desirable” in a descriptive sense in the major premise to indicate what can be desired, while using it in a normative sense in the conclusion to recommend that which is worthy of desire, or to prescribe that which ought to be desired. The objection, then, is that Mill uses the term “desirable” in two different senses within a single argument, thereby committing the fallacy of equivocation. 4.2
Analysis of the Charges
This section explores the charges against Mill in three stages. The first stage introduces and warns against a couple of common mistakes in critical evaluations of Mill’s proof. The second stage undertakes a reduction, for the sake of convenience, in the number of phrases representing meaningful but largely inconsequential variations on correct and incorrect interpretations of the word “desirable.” The third stage develops alternative formulations of the portion of the proof in which Mill allegedly equivocates on desirability.
164
Chapter 4
4.2.1 False Discount Factor in Critical Evaluation Commentators on Mill’s proof commonly invoke two familiar facts about his Utilitarianism. It is practically standard procedure in either attacking or defending the proof to recognize both that Utilitarianism is a popular work intended for a general audience and that Mill disavows proof in the ordinary sense of the term. These two considerations set the proof up to appear unambitious and unpromising to begin with, thus encouraging critics to dismiss the attempt in summary fashion, while inspiring defenders to make convenient excuses for Mill and to prepare emergency exits for themselves at points of conflict between the text and their interpretations. Before proceeding with an examination of the charges themselves, it will be helpful to show that these two appeals, though they both represent indisputable matters of fact about the text, are not relevant to a logical evaluation of the proof. The first consideration, that Utilitarianism is a popular work, is grounded in the fact that Mill originally wrote the piece as a series of articles directed at a wide audience and intended to have mass appeal.2 His motivation was to educate the general public on utilitarian philosophy, to correct common misconceptions and deliberate distortions of utilitarianism, and to address objections that were popular at the time. Critics and defenders alike tend nowadays to take the original purpose and intended audience of Utilitarianism as signs that the work is logically less rigorous than what is to be expected of a philosophical treatise. Critics relish this consideration as a point in their favor, that Mill’s predilection for elegant variation in literary style, often at the cost of conceptual clarity and philosophical consistency, generates irreconcilable interpretations. Defenders appeal to the same consideration in order to discount the importance or relevance of a passage that may be at odds with their own interpretations, typically attributing any inconsistency to the difficulty of accounting for everything Mill says in Utilitarianism. Yet this appeal to the original purpose and intended audience of Utilitarianism is not really as decisive or as compelling as it may seem. Nor does it even support the discount factor commonly weaved into an evaluation of the proof. A shift in Mill’s intended audience, say, a shift from an assembly of
2 A letter from Mill to George Grote (Letter 525: 10 January 1862) affords some insight into how Mill himself conceived of the work: “I am very glad that you like the papers on Utilitarianism so much. I am not more sanguine than you are about their converting opponents. The most that writing of that sort can be expected to do, is to place the doctrine in a better light, and prevent the other side having everything their own way, and triumphing in their moral and metaphysical superiority as they have done for the last half century and as they do in France still more than in England” (CW15:763).
Alleged Fallacy of Equivocation
165
scholars to the general public, is not accompanied by a transformation of Mill himself from an eminent philosopher into an incompetent demagogue who cannot think coherently or write clearly. By the same token, at the receiving end of the shift in audience, members of the general public are not impressed more by inconsistency and contradiction than by clear and cogent argumentation. Mill’s Utilitarianism might indeed be concerned with elegance in style and presentation, but that concern never degenerates into a poetic license to ignore the rules of logic, especially not where a proof of sorts for the principle of utility takes up an entire chapter. The second consideration, that Mill disavows proof in the ordinary sense, originates in his repeated warnings that he is not offering a proof in the standard sense of the term (U1:5–6 CW10:207–208; U4:1 CW10:234; U4:9 CW10:237). Any effort to clarify the kind of proof actually offered requires a proper understanding not just of what is denied but also of what is not (see subsection 4.3.2 below). The denial of ordinary proof does not make logical appraisal superfluous as if the end product were mere rhetoric or sheer poetry. It certainly does not follow that exposing fallacies, if there are any, would not be a legitimate concern in a critique of the proof. The attempt may not constitute, or even resemble, proof in the ordinary sense, but it does present relevant rationale in support of clear conclusions (U1:5 CW10:207–208). The point is that, if Mill advances any arguments toward any sort of proof, then any one of those arguments, and conceivably even all of them, might well be fallacious. And if he does not, that is, if he offers no arguments whatsoever, then so much the worse for his proof. The proof, in fact, proceeds with systematic reasoning, perhaps with lacunas to be filled in, the entirety of which can quite reasonably be reformulated, if need be, through deductive or inductive arguments, or any combination thereof, at least as a heuristic device for comprehension and evaluation. To be blunt, the scholarly tendency to spurn logical analysis, because the proof itself is not a formal proof, is an overly restrictive interpretation of the provision regarding the kind of proof that is possible and appropriate. 4.2.2 Multiple Interpretations of “Desirable” The exposition in section 4.1 summarizes the charges against Mill using numerous variations on both correct and incorrect conceptions of desirability in order to illustrate the variety of expressions Mill’s critics use in discussing his analogic argument. Having already observed the diversity of possibilities, we need not let the pedantic realism of comprehensive coverage detain us any further. We can easily and sensibly settle on a single expression to stand
166
Chapter 4
for the actual meaning of the word “desirable” and another to stand for the misconstrual. Mill supposedly equivocates on the word where he infers a conclusion containing the standard normative sense of desirability from premises containing an idiosyncratic descriptive sense. The descriptive sense refers to that which can be desired, that which is able to be desired, that which is capable of being desired, that which it is possible to desire, and so on, all of which mean roughly the same thing. We may, without consequence, pick any one of these, and ignore the rest, as the standard meaning under the idiosyncratic misconstrual. Let us proceed with “can be desired,” admitting that any of the others would do just as well. The normative sense is not as easy to settle. The alternatives there are not as close to synonymy as those in the descriptive sense. The normative sense can refer to that which is worthy of desire, that which is fit to be desired, that which it is good to desire, or that which ought to be desired, not all of which mean the same thing. The first three alternatives are close enough, but the ought-construction stands out. Given the similarity between the first three expressions, the contrast is essentially between “worthy of desire” and “ought to be desired.” Which one best captures the proper meaning of “desirable”? Reducing the normative, and thus proper, sense of “desirable” to “ought to be desired” would require rendering “Happiness is desirable” as “Happiness ought to be desired.” But is there really a moral obligation to desire happiness, or for that matter, to desire anything at all? Can there possibly be such an obligation? Are desires morally debatable and sanctionable in the manner of actions? Or is it morally permissible to desire anything whatsoever so long as one does not act on desires for morally censurable or reprehensible things? It seems that an “ought” would be excessive when invoked in regulation of desires. We cannot control our desires, emotions, feelings, and mental states to any extent approaching the degree of control we have over our actions. The Commandment “Thou shalt not covet thy neighbor’s wife,” for example, is not the kind of moral imperative that people can readily assimilate into their psychological profile, or their thought patterns, as a compelling moral obligation, especially if the neighbor in question happens to be Lars Schmidt, Aron Lindström, or Roberto Rossellini, each of whom was at one time married to Ingrid Bergman. For one thing, desires are largely beyond our control, except perhaps for our partial success in suppressing some of them once they arise, but that is not the same as being able to prevent any of them from arising in the first place. For another, desires for the neighbor’s wife are harmless in themselves, so long as neither Ms. Bergman nor any of her spouses in succession is
Alleged Fallacy of Equivocation
167
made aware of them, and quite possibly still harmless even then. The ought- construction, therefore, does not appear to be the way to go. Mill himself proceeds as if he were indifferent between the two alternatives for the normative sense of desirability. In A System of Logic, for example, he uses “desirable” to mean both “ought to be desired” and “worthy of desire” (L6:12.6–7 CW8:949–951). His greatest critic, Moore, also appears indifferent, at least in Principia Ethica, between the same two normative senses, so long as he can show that Mill equivocates on desirability between one or the other of the two normative senses and the idiosyncratic descriptive sense (Moore 1903, § 3:40.4, pp. 118–119). The Oxford English Dictionary, however, is not indifferent. It gives the primary meaning of “desirable” as “worthy to be desired.” This is close enough to Mill’s “worthy of desire” that we can proceed with the latter without fear of contradicting the most authoritative dictionary of the English language. The standard normative sense is indeed better expressed by the phrase “worthy of desire” than it is by the phrase “ought to be desired,” because the meaning of the word “ought” is vague even where the universe of discourse is restricted to the ethical realm, and because the ought-construction becomes utterly counterintuitive upon clarification, as it invokes moral obligation, which is not apposite to the context. Mill’s proof is concerned exclusively with value, at least in the portion subjected to the charge of equivocation, and that concern is best served by the “worthy of desire” interpretation. That is therefore the only alternative used throughout the remainder of the discussion here. 4.2.3 Alternative Formulations of Mill’s Argument Since the allegation of equivocation is grounded in a supposed ambiguity in Mill’s usage of the word “desirable,” it will be helpful to examine all combinations of the senses intended in the premises and conclusion of the chain of reasoning suspected of equivocation. Consider the following four arguments as interpretations of the allegedly equivocal inference, reformulated to replace the word “desirable” with the phrases “can be desired” and “worthy of desire” where appropriate: S(a)
(P1) (P2) (C)
If something is desired, then it can be desired. Happiness is desired. Happiness can be desired.
S(b)
(P1) (P2) (C)
If something is desired, then it is worthy of desire. Happiness is desired. Happiness can be desired.
168
Chapter 4
S(c)
(P1) (P2) (C)
If something is desired, then it is worthy of desire. Happiness is desired. Happiness is worthy of desire.
S(d)
(P1) (P2) (C)
If something is desired, then it can be desired. Happiness is desired. Happiness is worthy of desire.
None of these arguments represents Mill’s entire proof. Each one is a reformulation of the portion where he supposedly commits the fallacy of equivocation. Nor do these arguments fully capture every nuance of the relevant part of the proof, though each one indeed represents a self-contained and well-defined portion of it. While the relevant part of the proof is not a deductive argument, it can usefully be sketched in the manner above as a first approximation. The alternatives work well as minimalist interpretations of the inference in question, probably with missing lemmas, as each reformulation has a skeletal framework sufficient for them collectively to display what can be made of the objection that Mill obscures the difference between what can be desired and what is worthy of desire. The logical status of these arguments is rather obvious, but a logical appraisal will be helpful nonetheless. Although it is safe to assume that Mill and his critics, together with anyone studying the conflict between them, would require no assistance with the logic, it is, after all, an elementary logical fallacy that Mill is accused of, which then makes it useful to identify interpretations under which the argument is fallacious, and to determine which, if any, of these can reasonably be attributed to Mill. Here, then, is a brief analysis of the logical status of each of the four arguments: S(a) is valid by modus ponens and sound because something’s being desired leaves no room for doubt whether it can be desired. S(b) is invalid because it commits the fallacy of four terms through the substitutions unpacking the equivocation on desirability. S(c) is valid by modus ponens, but it is not sound, because its major premise is false, given that something’s being desired does not establish that it is worthy of desire. S(d) is invalid because it commits the fallacy of four terms through the corresponding substitutions for the alleged equivocation. To summarize: S(a) is valid and sound; S(b) is invalid; S(c) is valid but not sound; S(d) is invalid. Critics take S(d) as the proper interpretation. However, note that something is wrong with each formulation, even with the valid ones, including the one that is both valid and sound. Interpreted as S(a), the argument is logically acceptable but entirely trivial. Interpreted as S(b), S(c), or S(d), it does not
Alleged Fallacy of Equivocation
169
work at all, being either invalid, and fallaciously so at that, or valid but not sound. Why is it, then, that critics insist on S(d), to the exclusion of S(a), S(b), and S(c), as the proper interpretation of Mill’s argument? The answer is two-fold. First, Mill is anxious to prove that happiness is worthy of desire, not that it can be desired. Since it is of no use to Mill, and quite trivial in general, to show that happiness can be desired, critics know that he is arguing in the conclusion for what is worthy of desire. This limits the choices to S(c) and S(d). Second, Mill introduces his argument immediately following claims concerning “sight and visibility” and “sound and audibility.” Hence, it seems as if he supposes the suffix “able” to qualify the word “desire” in the major premise to refer to what can be desired. Critics thus take Mill to be equivocating on “desirable” between the major premise and the conclusion. S(a) and S(b) drop from critical consideration because no one reads Mill as trying to demonstrate that happiness can be desired. S(c) and S(d) are both good candidates for further consideration. S(d) might appear to be the appropriate subject of focus because it is the only one of the four reformulations to capture the equivocation in question. A direct response to the charges would indeed have to deal with S(d) to show that and why Mill does not equivocate in that argument. The truth of the matter, however, is not so much that Mill does not equivocate in that argument as it is that he does not even make that argument. An equally promising response to the charges is to focus on S(c) to show that and why it is a sound argument. The choice is between confronting the equivocation charges in S(d), which is not true to the text, and denying that equivocation was ever at issue by proceeding with S(c). The decision here is to proceed with S(c), at least as a heuristic orientation toward Mill’s intended argument, since he evidently never intended anything like S(d), as argued below, though he also did not intend S(c) in its deductive formulation above. 4.3
Response to the Charges
The response proceeds in four steps. The first step invokes the principle of charity in support of the rejection of S(d) and the adoption of S(c) as an interpretation of Mill’s argument without equivocation on desirability. The second step, in preparation for the next two, illustrates the methodological platform supporting the entire proof. The third step lays out the evidentiary relationship Mill finds between desires and desirability. The fourth step introduces and explores a distinction between means and ends in the evidentiary relationship between desires and desirability. The overall aim is to demonstrate
170
Chapter 4
that Mill’s appeal to desires as evidence of desirability, coupled with a distinction between means and ends, supports something like S(c), though not in its deductive formulation above. 4.3.1 The Principle of Charity in Critical Evaluation The most natural response to the accusation of equivocation is that it is entirely uncharitable to saddle a great philosopher with an outlandishly stupid argument. An alternative reading of the passage in which Mill allegedly equivocates on the word “desirable” has much to recommend it for that reason alone. Among any number of interpretations, where one involves fallacious reasoning, the appropriate reading for an eminent philosopher is one that absolves him of the fallacy, provided that there are no compelling reasons to suspect that he really did intend the most preposterous formulation possible. This is particularly important with respect to Mill, who was a reputable logician of the nineteenth century with an influential text on logic. As a matter of fact, the following passage in that text, A System of Logic, explicitly recognizes a difference between the general conceptual possibilities that Mill is accused of either confusing or conflating in his alleged equivocation on desirability: A proposition of which the predicate is expressed by the words ought or should be, is generically different from one which is expressed by is, or will be. mill L6:12.6 CW8:949
Critics will retort, no doubt, that if Mill was such a great logician, he should have not just noted the relevant difference in his logic book but also recognized that same difference in his moral discourse, especially in the course of his sensational proof. On the other hand, the point is not merely that Mill in particular, as a proficient logician, was not likely to confuse something’s desirability (or its source of oughtness) with the possibility of its being desired, but more so that no one at all, philosopher or otherwise, is ever likely to do so. Even if in the haste of a spontaneous stream of consciousness, inspired perhaps by free association with visibility and audibility, Mill had initially confused the desirability of happiness with the possibility of desiring it, he would have been only too eager to correct that mistake upon a second glance, especially if the confusion were specifically pointed out to him. It is quite telling, then, that he did not revise the third paragraph of the fourth chapter of Utilitarianism in the twelve-year period between its initial appearance in Fraser’s Magazine in 1861 and his death in 1873, a period in which he produced four editions of Utilitarianism in the form of a book (1863; 1864; 1867; 1871). Even if Mill himself
Alleged Fallacy of Equivocation
171
had not noticed any problems with the argument as it went through four editions on top of the original serial published in monthly installments, it seems that astute critics would have had ample time to react to whatever problems they thought infested that passage. Interestingly enough, Mill had an excellent opportunity for clarification and reformulation as he was warned repeatedly by his friend and translator Theodor Gomperz. Seven years after the publication of Utilitarianism, as Gomperz was preparing a German translation of the text, he wrote to Mill on 26 March 1868, complaining of difficulties in the third paragraph.3 Here is what Gomperz says in the letter: Let me conclude by expressing my regret that you did not in the later editions of the Utilitarianism remove the stumbling block (to any reader and more especially to a translator) pp. 51–52 1st ed. (audible, visible— desirable) which when pointed out to you by me, you said you would remove. Your argument looks like a verbal quibble, far as it is from being one and has besides to me the serious disadvantage of being utterly untranslatable. mill CW10:cxxvi; editor’s bracketed insertions omitted4
Although Gomperz does not, at least not in this letter, explain exactly what he thinks is wrong with the argument, it is clear that he finds the move from “visible” and “audible” to “desirable” difficult to follow, or to explain, and therefore also to translate. Furthermore, it is evident in Gomperz’s reference to what Mill had promised but failed to do in “later editions” that the letter quoted above was not Gomperz’s first attempt to persuade Mill to rewrite the passage in question. The fact that Gomperz quotes from the first edition of Utilitarianism (1863) does not quite settle the date of his initial warning to Mill. Judging by the date of Gomperz’s letter (1868) and by his reference to unrevised editions as “later editions” in the plural, at least two unrevised editions of Utilitarianism would have had to be published between Gomperz’s first warning and his subsequent letter. Those would have been the second edition in 1864 and the third in 1867. 3 John Mercel Robson (CW10:cxxvi), in his “Textual Introduction” to Utilitarianism, mistakenly cites the date of Theodor Gomperz’s letter as 18 March 1868. Francis E. Mineka (CW16:1391, n.1, n.7), the editor of Mill’s letters, confirms, in two footnotes to Mill’s reply to Gomperz, that the correct date is 26 March 1868. Adelaide Weinberg (1963, 51, n. 169) also gives the date as 26 March 1868 in a footnote immediately preceding her citation of the letter. 4 Adelaide Weinberg (1963, 51–52) quotes the same passage as above.
172
Chapter 4
It follows, therefore, that Gomperz must have warned Mill at least as early as the publication of the first edition in book form in 1863, and possibly as early as the original publication in Fraser’s Magazine in 1861. Mill evidently had at least a full decade and perhaps as much as twelve years to act upon the initial warning. It is something of a surprise, then, that Mill appears remarkably relaxed about the matter in his following response to Gomperz (Letter 1227: 23 April 1868): With regard to the passage you mention in the Utilitarianism I have not had time regularly to rewrite the book, & it had escaped my memory that you thought that argument apparently though not really fallacious which proves to me the necessity of, at least, further explanation and development. I beg that in the translation you will kindly reserve that passage to yourself, & will remove the stumbling block, by expressing the real argument in such terms as you think will express it best. mill CW16:13915
Although Mill indeed might not have had “time regularly to rewrite the book,” Gomperz’s letter (1868) precedes the publication of the fourth edition of Utilitarianism (1871) by three years and Mill’s death (1873) by five years, which, in either case, still leaves sufficient time to rewrite at least the paragraph in question. Unfortunately, a revision does not occur, not even in the German edition, as Gomperz states humbly in his response on 11 May 1868 that he declines the privilege to treat the apparently problematic passage with such editorial freedom (Weinberg 1963, 53). While Gomperz’s complaint in the letter may not be as explicit as the charges advanced after Mill’s death, it is a sufficiently clear anticipation of the actual charges to have directed Mill’s attention to potential problems. Had Mill taken the “stumbling block” mentioned by Gomperz as seriously as did the critics who later embraced it, he would have made “time regularly to rewrite the book,” or at least to reword the relevant passage in a clear manner, leaving no need for interpretation, and thus, no room for misinterpretation. Considering that Mill produced one more edition of Utilitarianism after the letter from Gomperz, his later reluctance to revise the passage in the fourth edition cannot be attributed to time constraints alone. His cavalier attitude in delegating undue editorial authority to Gomperz on a passage so crucially
5 John Mercel Robson (CW10:cxxvi) and Adelaide Weinberg (1963, 52–53) both quote the same passage as above.
Alleged Fallacy of Equivocation
173
important to his proof of the principle of utility, combined with his apparent contentment with the idea of leaving to posterity a revised version of the passage only in German, suggests that he did not anticipate any serious misunderstanding in any language. At any rate, the correspondence between Mill and Gomperz shows that Mill was warned about potential problems in the third paragraph but persisted in preserving the wording of the allegedly fallacious inference in subsequent editions of Utilitarianism. There are only two alternatives: Either Mill was so dense as to publish the same fallacious argument five times, despite repeated warnings, or he had good reasons to believe that the inference worked just fine in the first place. 4.3.2 The Methodology behind Mill’s Proof Proper insight into Mill’s methodology begins with a consideration “of what sort of proof the principle of utility is susceptible,” which recalls the title of the fourth chapter of Utilitarianism (see also: U4:9 CW10:237). What makes this a good place to start is not just that Mill himself starts there but also that what he says in that regard requires careful consideration to avoid serious misunderstanding. Mill’s preliminary reflections on the sort of proof apposite to the principle of utility tend to shape the expectations of readers regarding the forthcoming proof, consequently influencing their interpretations of the actual proof. While this is not a problem in itself, it can indeed lead to interpretive difficulties because Mill articulates his reflections vaguely enough to generate widely varying expectations regarding both the nature and the strength of the anticipated proof. Mill’s methodological prelude to the proof is a series of potentially misleading positive and negative remarks on the kind of proof that would be relevant in that context. It is important to identify his basic message without exaggerating the accompanying rhetoric. His preliminary remarks on the subject of proof are concentrated in the last two paragraphs of the first chapter and the first paragraph of the fourth chapter of Utilitarianism. The discussion there can be misleading in two ways. First, many of the positive remarks describing the appropriate approach are vague generalities that can describe the most rigorous formal proof as well as loosely connected platitudes of sophistry. Second, the negative remarks describing the inappropriate sort of proof come close to excluding rational discourse altogether. The positive remarks are too general to secure any degree of uniformity in the expectations of readers regarding the proof, or even to help them form any particular expectations at all. Observe, for example, the “larger meaning of the word proof,” introduced as “equivalent to proof”: To prove something, in
174
Chapter 4
the larger meaning, is to present considerations “capable of determining the intellect either to give or withhold its assent to the doctrine” (U1:5 CW10:207– 208). This is entirely too vague. The fliers I find tucked under my windshield wipers may well be sufficient to convince me that Jane Doe is the best candidate for county commissioner, or that God loves me and wants me to donate generously to the First Baptist Church around the corner, or that the best pizza in town is at Sal & Mookie’s New York Pizza & Ice Cream Joint. The next paragraph goes on to elaborate on this larger meaning of proof in terms of “rational grounds” for “accepting or rejecting the utilitarian formula” and of “philosophical grounds” for “assenting to the utilitarian standard” (U1:6 CW10:208). The reference to rational and philosophical grounds adds nothing to the clarity of the intended methodology. Mill’s positive remarks on the kind of proof that would be appropriate under the circumstances remain vague enough, on the whole, to apply equally well to Kant’s transcendental deduction of the categories and the promotional efforts of the marketing team at Sal & Mookie’s. The negative remarks, on the other hand, consist of various caveats and qualifications collectively appearing restrictive enough to exclude all rational discourse from the kind of proof apposite to the principle of utility. The first chapter, for example, warns that the relevant proof is neither “direct proof” nor “proof in the ordinary and popular meaning of the term” nor “what is commonly understood by proof” (U1:5 CW10:207–208). The fourth chapter reiterates that it is not “proof, in the ordinary acceptation of the term,” adding that it is not even “proof by reasoning” (U4:1 CW10:234). What, then, if not proof by reasoning? Persuasion through propaganda? No, not at all. Judging by the final outcome, the proof is entirely respectable, both logically and philosophically, but the strong disclaimers keep expectations too low for anything better than mere rhetoric. The repudiation of “proof by reasoning” is not as restrictive as it may seem. As a caveat, it is loosely expressed, not literally exclusive. It does not rule out reasoning per se. The intention is not to exclude “proof by reasoning at all” but “proof by reasoning alone.” The provision is simply that the principle of utility cannot be established through a priori reasoning, an approach which the intuitionist rivals of Mill were advocating as the proof procedure required for ultimate ends and first principles (U1:3 CW10:206–207). Mill’s delivery, however, is so theatrical as to lead even one of his defenders to misquote the first sentence of the fourth chapter by interjecting a bogus “strict” in qualification of “proof”: “It has already been remarked, that questions of ultimate ends do not admit of strict proof, in the ordinary acceptation of the term” (Hall 1949, 6). Mill does reject “direct proof” (U1:5 CW10:207), to be sure, but not “strict proof” as Everett W. Hall has it here. The point, of course, is
Alleged Fallacy of Equivocation
175
not that Hall’s citation is erroneous but that Mill’s disclaimers are misleading. The apparent opposition to this or that type of proof is so overwhelming that it inspires Hall to view the attempt as if no proof at all were being offered.6 There is no telling what “strict proof” might mean, but all that Mill intends to deny in his rejection of “direct proof” is a particular method of testing hypotheses. Put simply, no candidate for an ultimate end, or for a first principle, can be verified, established, or corroborated through its logical or ontological relationship with a higher end or prior principle. No such demonstration is possible, because no higher or prior source of appeal is available for ultimate ends or first principles (U1:5 CW10:207–208). The preliminary remarks examined thus far, both the positive ones and the negative ones, are at best nebulous if not altogether confusing. The expectations they generate fall short of proof in any sense, amounting to hardly anything better than rhetorical persuasion. The summation of the proof by Hall, as a sympathetic commentator, is a good example of the effect of such low expectations: In summary, the argument of chapter iv of Mill’s Utilitarianism is extremely simple and (in the main) sensible. To an empiricist who eschews all intuitive self-evidence, no ethical first principle can be strictly proved. All that one can do is to present considerations that will lead honest and reasonable people to accept such a principle. hall 1949, 11
This will not do. Hall underestimates Mill’s methodology. Acceptance by honest and reasonable people promises nothing more than belief and assent. Yet in direct contradiction of this reading, Mill makes it perfectly clear in A System of Logic that “belief is not proof, and does not dispense with the necessity of proof” (L3:21.1 CW7:563–564). He elucidates the difference as follows: By evidence is not meant anything and everything which produces belief. There are many things which generate belief besides evidence. A mere strong association of ideas often causes a belief so intense as to be 6 The tendency of Mill’s negative remarks to mislead his audience into underestimating the strength of the proof can be observed clearly in Everett W. Hall’s (1949) reception of the corresponding instructions and in his perception of the appropriate methodology: “In this situation Mill makes use of two considerations, both of which he got from Bentham, not to prove the principle of utility but to make it acceptable to reasonable men” (p. 8). “He is not (if the reader will tolerate another reiteration) trying to prove anything” (p. 9).
176
Chapter 4
unshakeable by experience or argument. Evidence is not that which the mind does or must yield to, but that which it ought to yield to, namely, that, by yielding to which, its belief is kept conformable to fact. There is no appeal from the human faculties generally, but there is an appeal from one human faculty to another; from the judging faculty, to those which take cognizance of fact, the faculties of sense and consciousness. The legitimacy of this appeal is admitted whenever it is allowed that our judgments ought to be conformable to fact. mill L3:21.1 CW7:564
The methodology Mill contemplates at the outset and executes throughout the proof is nothing less than the empirical methods of physical science, combining reason with both observation and experimentation. In contrast to the intuitive school of ethics, which holds that “the principles of morals are evident à priori, requiring nothing to command assent, except that the meaning of the terms be understood,” Mill is at pains to show that the principles of morals are “questions of observation and experience” (U1:3 CW10:206). The sort of proof applicable to the principle of utility is a “direct appeal to the faculties which judge of fact—namely, our senses, and our internal consciousness” (U4:1 CW10:234). Proving the principle of utility turns out to be “a question of fact and experience, dependent, like all similar questions, upon evidence,” where the answer to the corresponding question “can only be determined by practised self-consciousness and self-observation, assisted by observation of others” (U4:10 CW10:237–238). Mill reaffirms the empirical nature of the proof in question as he explains in a footnote toward the end of Utilitarianism that the approach draws on a combination of two paradigmatic benchmarks: (1) “the laws of human nature and the universal conditions of human life” and (2) “generalizations from specific experience” (U5:36n CW10:258). He asserts that “in ethics, as in all other branches of scientific study, the consilience of the results of both these processes, each corroborating and verifying the other, is requisite to give to any general proposition the kind and degree of evidence which constitutes scientific proof” (U5:36n CW10:258). Further confirmation of the scientific character of the sort of proof applicable to the principle of utility is available in A System of Logic, where Mill examines the logic of the moral sciences in his introduction to the sixth book: In substance, whatever can be done in a work like this for the Logic of the Moral Sciences, has been or ought to have been accomplished in the five preceding Books; to which the present can be only a kind of supplement
Alleged Fallacy of Equivocation
177
or appendix, since the methods of investigation applicable to moral and social science must have been already described, if I have succeeded in enumerating and characterizing those of science in general. mill L6:1.2 CW8:835
The common denominator in Mill’s general remarks on methodology, whether one inspects Utilitarianism or A System of Logic, though preferably consulting both, is the commensurability of ethics and science with respect to confirmation, demonstration, and justification. The standard espoused in either case is evidentiary reasoning. To return to the part of the proof supposedly equivocating on desirability, the key to an accurate understanding is to treat the inference as an empirical appeal directing our attention to the nature of the phenomena under consideration as opposed to a verbal analogy reducing the attempt to a play on words. The aim of the argument is to establish the desirability of happiness as an end in itself, which represents the conclusion of the first part of the proof. The strategy employed toward the achievement of that end is an empirical appeal to desires as evidence of desirability: “The sole evidence it is possible to produce that anything is desirable, is that people do actually desire it” (U4:3 CW10:234). The argument, roughly speaking, is that everyone desires happiness for its own sake, and not for the sake of any other end, which is indicative of the desirability of happiness as an end in itself. The best approach to the first part of the proof, then, without specifically addressing every controversy it has generated in the secondary literature, is to examine the goal, procedure, and substance of that part of the proof. This may indeed be a good way to approach almost any kind of argument, but it is particularly apt under the circumstances. This is because Mill presents the first part of the proof in seven sentences jointly leaving a mark in moral literature as the most extensive collection of logical fallacies, conceptual confusions, and philosophical mistakes ever compressed into a single paragraph. Understanding the goal, procedure, and substance is essential to an accurate interpretation of that paragraph and thereby of the first part of the proof. Mill’s goal is what we might nowadays call “epistemic justification” of the claim that happiness is desirable as an end in itself. The procedure he follows is the collection of empirical evidence sufficient for epistemic justification of that claim. The substance is the effective logical, dialectical, and rhetorical utilization of the observation that everyone, in fact, desires happiness for the sake of happiness itself, and not for the sake of any other end, which, as it turns out, is the only evidence available. The first part of the proof, then, can be summarized in its entirety as follows: Mill submits the observation
178
Chapter 4
that happiness is desired for its own sake as empirical evidence sufficient for epistemic justification of the position that happiness is desirable as an end in itself. The goal, procedure, and substance fit together rather well to form a compelling demonstration. The next two subsections elaborate on that fit in greater detail. 4.3.3 Desires as Evidence of Desirability Mill’s goal is not to explain why happiness is desirable as an end in itself but to prove that it is, in fact, desirable as an end in itself. The relationship between the desired and the desirable is not logical or ontological but epistemological and psychological. The point is not that desires confer value but that the desirable evokes desires. Mill is concerned with epistemic justification and psychological investigation rather than logical deduction or ontological explanation. He thus focuses on how we can possibly know that something is desirable, not on why it might inherently be desirable in the first place. As for how we can know, he suggests taking a look at the evidence, in this case, actual desires. One clue as to Mill’s intentions is the following sentence, which is methodologically the most important and least familiar sentence in the much-criticized paragraph: “If the end which the utilitarian doctrine proposes to itself were not, in theory and in practice, acknowledged to be an end, nothing could ever convince any person that it was so” (U4:3 CW10:234). What this means is that the only way to confirm the desirability of anything is to verify the existence of actual desires for that thing. How indeed could anything be recognized as desirable, unless it were, in fact, desired? To know the good is, in this case, to want the good. Even if something might conceivably be good without anyone being aware of its existence, or of its goodness, nothing can be known to be good unless it is so noted and thus noticed. This is quite different from an interest theory of value where what is good is any object of any interest or desire. Mill does not claim that something is desirable if and only if it is desired. Nor does he hold that either of the two conditionals that can be extracted from this biconditional is true. He suggests neither that if something is desired, then it is desirable, nor that if something is desirable, then it is desired. He does maintain, however, that if something is desired, then it might well be desirable, and that what happens to be desirable is quite likely to be desired. He thus proceeds on the grounds that a critical survey of things that are desired is a legitimate approach to the project in hand. The procedure followed is the collection of empirical evidence pertinent to the goal. Mill discloses the appropriate procedure in the opening words of the chapter on the proof:
Alleged Fallacy of Equivocation
179
it has already been remarked, that questions of ultimate ends do not admit of proof, in the ordinary acceptation of the term. To be incapable of proof by reasoning is common to all first principles; to the first premises of our knowledge, as well as to those of our conduct. But the former, being matters of fact, may be the subject of a direct appeal to the faculties which judge of fact—namely, our senses, and our internal consciousness. Can an appeal be made to the same faculties on questions of practical ends? Or by what other faculty is cognizance taken of them? mill U4:1 CW10:234
The questions are largely rhetorical. Mill embraces empirical reasoning for philosophical inquiry regardless of the subject matter, thus making no exceptions for moral discourse. The proof begins with a comparison of the evidentiary relationship between the seen and the visible, and between the heard and the audible, to the evidentiary relationship between the desired and the desirable: The only proof capable of being given that an object is visible, is that people actually see it. The only proof that a sound is audible, is that people hear it: and so of the other sources of our experience. In like manner, I apprehend, the sole evidence it is possible to produce that anything is desirable, is that people do actually desire it. mill U4:3 CW10:234
This is Mill’s analogic argument as intimated in the beginning of the present chapter. The analogy is between the relation of sensory perception to observation sentences and the relation of mental states to value judgments. The argument, or more accurately, the suggestion, is that both relations play a comparable evidentiary role in justification. Desires provide evidence of what is desirable, just as vision and hearing provide evidence of what is visible and what is audible. This comparison of evidentiary relationships between visibility, audibility, and desirability, on the one hand, and what people see, hear, and desire, on the other, is Mill’s implicit confirmation that “a direct appeal to the faculties which judge of fact—namely, our senses, and our internal consciousness” is methodologically useful in deciding “practical ends” as well as “matters of fact” (U4:1 CW10:234). His explicit confirmation comes later in the proof, where he states that “we have evidently arrived at a question of fact and experience, dependent, like all similar questions, upon evidence. It can only be determined by practised self-consciousness and self-observation, assisted by observation of others” (U4:10 CW10:237).
180
Chapter 4
Mill’s analogic argument is not an attempt to establish perfect parallels between the visible, the audible, and the desirable. No doubt, “desirable” would be the correct answer to a multiple-choice question asking: “Which word does not belong in the group: ‘visible,’ ‘audible,’ ‘desirable’?” It would be the correct answer because “visible” means “can be seen,” and “audible” means “can be heard,” but “desirable” does not mean “can be desired.” Expanding on the hypothetical multiple-choice question, “visible” and “audible” belong in a group of words such as “accessible,” “breakable,” and “detectable,” while “desirable” belongs in a different group, together with words such as “adorable,” “preferable,” and “remarkable.” The difference, however, does not point to a failed attempt by Mill to group together words with analogous meanings. His own choice of words is quite deliberate. He does not intend a verbal analogy. On the contrary, he wants to contrast “desirable” with “visible” and “audible” in exactly the sense in which “desirable” is the odd one out in the hypothetical question. The aim is to demonstrate that the method of investigation is the same even when the subject of investigation is not. More specifically, the aim is to illustrate that the method of investigation requires empirical evidence of desirability, which is an axiological attribute, just as it requires empirical evidence of visibility and audibility, which are physical properties. Desires are evidence of desirability, then, insofar as physical manifestations of unobservable phenomena are admissible as evidence. This is exactly the kind of analogy Mill wants. His choice of the words “visible” and “audible” to accompany “desirable” is neither careless nor misleading. Even so, an important methodological difference between proving desirability and proving visibility and audibility is the fallibility of the corresponding evidence. That is why the analogic argument invokes sights and sounds as “proofs” of visibility and audibility, while introducing desires as “evidence” of desirability. The difference is that “proof” refers to a logically rigorous formal procedure for establishing the conclusion of an argument, whereas “evidence” refers to considerations that support a premise or position without necessarily constituting proof or confirmation (L3:20.3 CW7:559–561). Mill switches his terminology from “proof” to “evidence” in the analogic argument because he does not take desires to constitute conclusive evidence of desirability, while he does consider sights and sounds to constitute conclusive evidence of visibility and audibility. Desires are obviously evidence of desirability, but just how strong is that evidentiary relationship? Sporadic, eccentric, or pathological desires for a particular thing are not conclusive evidence of the desirability of that thing, just as auditory or visual hallucinations during a psychedelic experience are not conclusive evidence of the actual audibility or visibility of whatever it is that
Alleged Fallacy of Equivocation
181
is thought to be heard or seen. If desires are evidence of desirability, then, as a general rule, the greater the number of people desiring any particular thing, the stronger the evidence for the desirability of that thing. This does not mean that things are more or less desirable in proportion to the number of people desiring them. It means that the preliminary estimate of something’s desirability is more or less reliable depending on the number of people desiring it. Mill claims that everyone desires happiness: “Each person, so far as he believes it to be attainable, desires his own happiness” (U4:3 CW10:234).7 It follows that, insofar as desires are the acid test of desirability, the preliminary report of the desirability of happiness is as reliable as possible. Thus, happiness passes the initial screening for desirability. The next step is just as important. 4.3.4 Distinction between Means and Ends Are desires really a reliable test of desirability? Setting aside the pedantic question of how many people must desire something for it to count as desirable, let us ask the incisive one of whether desires are evidentially relevant to desirability in any significant way? Might not something’s being desired turn out to be a misleading indicator of its desirability even where it is the only evidence available for its desirability. To illustrate, I might desire a particular pair of shoes for playing basketball, only to find out later, sometime after I purchase them, that they are a product of poor craftsmanship and inferior materials, and therefore not at all desirable. Among other things, they lack adequate arch and ankle support for safe and effective jumps, while also failing to provide sufficient traction to grip the floor as tightly as necessary for quick moves and sudden stops. The basic scenario is that I start out with a desire for the shoes as a means for playing basketball, yet they do not prove to be desirable as a means to that end. The reason that the shoes are not desirable, then, is that they fail to serve the end for which they are desired as a means. 7 Norman Kretzmann (1958) offers an alternative reading of the significance of the number of people who actually desire something as an indication of the desirability of that particular thing. Claiming that “in the formulae about the desired and the desirable the key words are ‘people’ and ‘actually’ ” (p. 251), he argues that Mill uses “people” to refer to “the normal observer,” not to “all people” (pp. 252–253). Kretzmann’s aim is to isolate the evidentiary relationship between the desired and the desirable from the context of happiness, and thereby to develop a general account of the evidentiary relationship between the desired and the desirable regardless of the object of desire. He interprets that relationship as follows: “If anything is desired in such a way as to occasion some overt reaction on the part of the normal desirer, and that reaction proves to have been normal for the thing in question, then that thing is desirable” (pp. 253–254).
182
Chapter 4
On the other hand, even if they turn out to be satisfactory as a means for playing basketball, I might nevertheless decide that the shoes are not desirable after all, because they rate low as a status symbol compared to the more expensive brands my friends are wearing. The reason that my shoes are not desirable, in that case, even though they satisfy the end for which they were originally desired as a means, is that the end has changed. I started out by desiring shoes that are appropriate for playing basketball, but I later came to desire shoes that are recognized as a status symbol as well as being suitable for playing basketball. Finally, suppose that I acknowledge both ends from the beginning, desiring a pair of shoes that are just right for playing basketball and highly regarded as a status symbol. So, I purchase a pair of the flashiest shoes available by the most prestigious brand in the business. They are a widely recognized status symbol, especially among my friends, in addition to being effective instruments to complement my natural talents, or rather, to compensate for my deficiency in that area. Suppose, however, that I eventually lose interest in basketball, and that my flamboyant shoes are not suitable for casual wear, so that I no longer have any use for them. The shoes are then no longer desirable, because even though they continue to satisfy the ends for which they were originally desired as a means, and even though the ends by which they are judged are still the ends for which they were originally desired, those ends have themselves ceased to be important. These examples illustrate different ways in which an object of desire might not turn out to be desirable. What is common to all three cases is that the object in question is desired not for its own sake but for the sake of its anticipated contribution to an end beyond itself. This suggests that the strength of the evidentiary relationship between desires and desirability rests on a distinction between means and ends.8 The discrepancy between means and ends in the examples is manifested in the different ways that desires might fail to predict desirability. The object of desire may turn out not to be desirable because it fails to satisfy the end for which it is desired as a means, or because it comes to be evaluated by a different end than the one for which it is originally desired as a means, or because the end for which it is desired ceases to be important altogether. These different failures share a common explanation: If something is desired as a means to an end other than itself, then it is not necessarily 8 This is not about privileged insight into the precise nature of desires as mental states or cognitive content. Even the most elementary inquiry reveals different candidates as targets of desire. There are obvious differences between desiring something, desiring to use that thing, or even simply to possess it, and desiring to experience the pleasure anticipated in using or
Alleged Fallacy of Equivocation
183
desirable in itself, because the desire for that thing is subject to revision with respect to the end for which that particular thing is desired as a means.9 In contrast, desires for things for their own sakes, and not for the sake of any other end, are not subject to revision, at least not in this manner, because such desires are independent of ends other than their specific objects of desire. The observation that people desire a particular object, then, is not reliable as evidence of the desirability of that object unless the object is desired for itself. Mill’s general understanding of the evidentiary relationship between desires and desirability can be elucidated as follows: If the only evidence that anything is desirable is that people desire it, and if people desire some things as ends in themselves, and others as means to ends beyond themselves, then the evidence that a particular thing is desirable is stronger if people desire that thing for itself and weaker if they desire it for the sake of an end other than itself. The gist of the distinction is that desires for things as means to other things are evidence of instrumental value, while desires for things as ends in themselves are evidence of intrinsic value. In the light of the preceding discussion, the evidentiary relationship between desires and desirability can be characterized in a single formula: If the only evidence that anything is desirable is that people desire it, then the evidence that a particular thing is desirable is stronger, the greater the number of people who desire that thing, and stronger still, the more they desire it for its own sake rather than for the sake of some other end. Mill indeed finds the possessing that thing. It is indeed important from the perspective of the philosophy of mind to determine precisely what it is that we desire when we desire something. The focus of discussion in the main text above, however, is a contrast between “desiring as a means” and “desiring as an end.” That distinction cuts across any range of alternatives for the actual target of desire. A distinction between means and ends remains relevant whether I desire the basketball shoes, or to own or use the basketball shoes, or to experience the pleasure anticipated in owning or using the shoes. 9 Several other philosophers take the distinction between means and ends as the key to interpreting Mill’s conviction that desires are evidence of desirability. Carl P. Wellman (1959), for one, maintains that Mill invokes the evidentiary relationship between the desired and the desirable at the level of ends rather than means: “That is, desires for means may be criticized in terms of the ends to which these means lead. The corollary would seem to be that desires for ends are incorrigible. Here in the dichotomy of ends and means lies the key to Mill’s argument” (1959, 271). Wellman reconstructs Mill’s argument with an explicit “for itself” in qualification of “desired,” and “for its own sake” in qualification of “desirable” (1959, 271– 272). George A. Clark (1959, 653) likewise inserts “as an end” in parentheses in his quotations of two statements from relevant passages in the fourth chapter of Utilitarianism. Maurice Mandelbaum (1968, 228), for his own part, interjects “as an ultimate end” in italics to qualify “desirable” in his block quotation of the first three sentences of the third paragraph of the fourth chapter.
184
Chapter 4
evidentiary relationship between desires and desirability as strong as possible in the context of happiness. He contends not merely that people desire happiness but that everyone desires happiness, and moreover, that everyone desires happiness for the sake of happiness itself and not for the sake of any other end. The portion of his proof of the principle of utility in which he allegedly equivocates on “desirable” can reasonably be reconstructed as follows:
(P1) The only evidence available that anything is desirable is that people do actually desire it. (P2) The only evidence available that anything is desirable as an end in itself is that people desire that thing for its own sake and not for the sake of any other end to which it might be a means. (P3) The only evidence available that happiness is desirable as an end in itself is that people desire happiness for its own sake and not as a means to something else. (P4) Everyone desires happiness, and everyone desires happiness for its own sake, not as a means to something else. (C) Happiness is desirable as an end in itself.
The combination of analysis and response in this chapter is not just an attempt to absolve Mill of charges that he equivocates on the word “desirable” in his proof of the principle of utility. It is also an exposition and interpretation of the first part of that proof. Several aspects of the exegetical function of this chapter are relevant to some of the chapters that follow. Two are worth noting here. First, the methodology of Mill’s proof, the appeal to desires as evidence of desirability, and the distinction between means and ends, together form the basis of the demonstration in the sixth chapter of the present volume that Mill does not commit the naturalistic fallacy. Second, the same considerations also form the basis of the discussion of the first part of the proof, in the seventh chapter, which demonstrates how the bits and pieces of the defense offered in this book fit together as a coherent interpretation of the entire proof.
c hapter 5
The Alleged Fallacy of Composition in Mill’s Proof This chapter addresses allegations that Mill commits the fallacy of composition in his proof of the principle of utility. It proceeds with three substantive sections followed by a critical summary recapitulating the central elements of the defense. The first section introduces the charges against Mill through the standard definition of the fallacy of composition as a logical error committed in arguments containing an illicit inference from the parts of a whole to the whole itself. It thus illustrates how and where Mill is alleged to commit the fallacy. The aim of this section is to expose the intricate nature of Mill’s inference, which critics often fail to capture in their cursory accusations, and thereby to demonstrate that the typical critical reading of the argument is ambiguous with respect to the meaning of key terms, which then makes the standard definition of the fallacy a poor fit with the actual reasoning behind the argument. The second section develops four scenarios representing objections that are common as well as those that are conceivable but not common. These objections are grounded in the historical charges examined in the third chapter, combined with rational speculation on how any critic must interpret Mill’s argument in order to construe it as committing the fallacy of composition. The goal in each scenario is to set up an unambiguous way of charging Mill with the fallacy, thereby providing a lucid but not necessarily correct interpretation of the argument. The intention at this point is not to refute the charges but to determine the most plausible interpretation of the inference among a variety of tempting reconstructions that make it out to commit the fallacy of composition. The third section presents a response to the charges in three steps. The first step draws on the relevant literature in informal logic to clarify the nature of the fallacy of composition, with a view to showing that not every inference from parts to whole can reasonably be censured as a bad argument in terms of the inherent logic. It also cites passages from A System of Logic, where Mill himself defines and discusses the fallacy, which confirms that he understands how it works and what is at stake. The second step explicates the elusive notion of the general happiness, arguing that, although Mill does not define the term, textual evidence suggests that he uses it distributively in reference to the personal happiness of everyone in the relevant community. The third step shows that Mill is a reductive individualist in his outlook on social groups, conceiving
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_007
186
Chapter 5
of them as little more than their individual members recognized all at once but still as individuals, which then makes his conception of the aggregate of all persons the perfect structural complement for his notion of the general happiness. The overall conclusion is that Mill’s understanding of the ontological nature of both the general happiness and the aggregate of all persons, together with the mathematical symmetry of the relationship he envisages between them, indicates that he does not commit the fallacy of composition in moving from the desirability of each person’s happiness for that person to the desirability of the general happiness for the aggregate of all persons. This conclusion is supported by critical comparisons of both valid and fallacious arguments that conform to the indicative parts-whole form of the fallacy, with the comparative results confirming that Mill’s own inference is perfectly consistent with the valid ones and quite unlike the fallacious ones. 5.1
Charges against Mill
The fallacy of composition is a logically illicit inference from the attributes, properties, or qualities of the parts of a whole to the attributes, properties, or qualities of the corresponding whole. That is the standard definition, or more accurately, a generic definition representing the typical case. The offending inference is the deduction of a conclusion about the whole from details about its parts. Hence, for the sake of minimizing repetition, except where the context requires articulation of the full version, the definition can be shortened to “a logically illicit inference from the attributes of the parts of a whole to the attributes of the whole itself,” which can be shortened further to “a logically illicit inference from the parts of a whole to the whole itself.” The full version of the definition may be assumed, whenever a shorter one is employed, unless expressly stated otherwise. Much in the same spirit, abbreviated references, such as “the parts-whole inference (or move),” or “the inference (or move) from parts to whole,” may also be used as shorthand for the full definition. Mill allegedly commits the fallacy of composition in one or both of the following two statements in the third paragraph of the fourth chapter of Utilitarianism (U4:3 CW10:234): – “No reason can be given why the general happiness is desirable, except that each person, so far as he believes it to be attainable, desires his own happiness.” – “Each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons.”
Alleged Fallacy of Composition
187
The fallacy can be attributed to either statement in isolation, but it can be observed more readily in the second one than in the first. This is because the first statement, but not the second, exhibits additional problems complicating matters, at least upon initial consideration. The first statement infers the desirability of the general happiness from each person’s desires for his own happiness. This amounts to an inference not just from the parts in question to the corresponding whole but also from an empirical observation to a normative judgment. The two moves, simultaneous though they may be, are mutually independent and must therefore be kept separate. While Mill does assert the desirability of the general happiness, he does not claim that anyone at all, let alone everyone, actually desires the general happiness. He completes his move from the desired to the desirable on the level of the individual person, after which he moves from that which is desirable for each person to that which is desirable for the aggregate of all persons. These discrete moves can be distinguished, or connected, with a lemma missing in Mill’s explicit statement. The chain of reasoning proceeds from each person’s desiring his own happiness to the desirability of each person’s happiness (for the person whose happiness it is) to the desirability of the general happiness (for the aggregate of all persons). The first transition is from fact to value, or from observation to evaluation, while the second is from parts to whole. Mill’s appeal to desires as evidence of desirability is the topic of discussion in chapter 4 of the present volume. The focus in this chapter is on the move from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons, which is where the fallacy of composition presumably occurs, or rather, one place where it supposedly occurs. The second statement taken in isolation is a straightforward and perspicuous instantiation of the fallacy of composition. Critics therefore tend to focus more on the second statement than on the first. The common formulation of the accusation, then, is that Mill commits the fallacy of composition in inferring the conclusion that the general happiness is a good to the aggregate of all persons from the premise that each person’s happiness is a good to the person whose happiness it is. Nevertheless, both of the statements quoted above appear to conflate the personal happiness of each person and the general happiness of the aggregate of all persons. Hence, Mill is charged in both of them with arguing that the sum of what is a good to A, and what is a good to B, and what is a good to C, individually, must be a good to A and B and C, collectively. Note the terminological difference between the two statements quoted above. As mentioned in the beginning of chapter 2, Mill tends to use the words “desirable” and “good” interchangeably in many contexts. He means the same
188
Chapter 5
thing in identifying something as good and in claiming it to be desirable.1 The suspect argument can be expressed in a single formulation paraphrased from the two explicit statements: Each person’s happiness is desirable for (is a good to) the person whose happiness it is; therefore, the general happiness is desirable for (is a good to) the aggregate of all persons. Paraphrased in this way, and taken in isolation, the argument does appear to commit the fallacy of composition in moving from the parts of a whole to the whole itself. However, two points in the corresponding charge merit further examination. The first point is that it is not entirely clear what parts, what whole, and what attributes are involved in the allegation that this piece of reasoning commits the fallacy of composition. Given the standard definition of the fallacy, introduced in the beginning of this section, and given the summary of Mill’s argument, paraphrased in the preceding paragraph, it is not clear whether Mill is accused of moving from parts to whole with respect to the general happiness or with respect to the aggregate of all persons or with respect to both. Nor is it clear what attributes range over the moves, that is to say, what attributes Mill carries over from the parts to the whole in either case. In order to determine if and how the argument commits the fallacy of composition, it is best to proceed by recognizing both wholes: the general happiness and the aggregate of all persons. If the general happiness and the aggregate of all persons are indeed composite wholes, as per the nature of the allegation, then the relevant parts of the general happiness must, at the very least, be the personal happiness of each person (ignoring any mereological details concerning the dynamics of how the parts may happen to be related or combined in the constitution of the whole), and the corresponding parts of the aggregate of all persons must be each of the individual persons. It is prima facie fallacious to move from an observation concerning each person’s happiness to a conclusion about the general happiness. It is additionally and separately fallacious, in the same way and to the same degree, to move from an observation concerning each individual person to a conclusion about the aggregate of all persons. Which move does Mill make? Does he make both? What attributes underlie the move(s)? The second point in need of clarification is the nature of the fallacy of composition itself. What exactly is wrong with reasoning from the attributes of the parts of a whole to the attributes of the corresponding whole? Does this description constitute, or represent, a detection mechanism that exposes a logical fallacy no matter the content? Or is the inference from parts to whole 1 Mill employs the following terms and expressions interchangeably: “good” and “desirable”; “good to” and “desirable for”; “a good” and “a desirable thing” (or “something desirable”); “a good to” and “a desirable thing for” (or “something desirable for”).
Alleged Fallacy of Composition
189
sometimes fallacious, sometimes not, depending on the context? The reason to question the mere form of the fallacy as a categorical guide to the fallacy is that not every inference in conformity with the specified form seems to be a bad argument. The following argument, for example, follows the designated form but seems to make a reasonable inference: “Each word in this sentence is in English; therefore, this sentence is in English.”2 Compare this to a verbally similar and formally identical argument which clearly does commit the fallacy of composition: “Each word in this sentence is in the Oxford English Dictionary; therefore, this sentence is in the Oxford English Dictionary.” Common sense alone suggests that the first inference works and the second one does not. Even the apparent difference is sufficiently motivating to explore the nature of the fallacy of composition in more detail as part of an effort to identify the logically relevant difference, if there really is one, between examples of each kind. The ultimate goal, of course, would be to determine whether Mill’s argument is more like the first example or more like the second. These two points are the subject matter of the following two sections. The next section (section 5.2) illustrates and analyzes various logical mistakes, each of which is supposed to represent the fallacy Mill allegedly commits where he moves from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons. The section after that (section 5.3) elaborates on the definition of the fallacy of composition in an effort to uncover all the relevant logical characteristics that together determine whether an argument that seems to commit the fallacy of composition really does commit that fallacy. The remainder 2 A possible objection might be to produce a series of words in English that do not qualify as a sentence in English, because they are not combined properly: “Chicken the white is time in you.” One response to such an objection is that the original sentence refers to itself and only to itself: “Each word in this sentence is in English; therefore, this sentence is in English.” The self-reference here makes the original sentence unique, so that no other sentence, let alone every sentence, is identified by the original sentence as a sentence in English, just because its every word is in English. In other words, the inference in the original sentence is not the same as the inference in the following variation: “If each word in any given sentence is in English, then that sentence is in English.” This confirms that “Chicken the white is time in you” does not constitute a counterexample to the original example. Another response is that a sentence is not just any collection of words but a cohesive and structured combination of words that makes a statement. If a series of words is not meaningful as a collection, then that series of words is not a sentence, and, therefore, not a counterexample to the original example. Should there be any dispute over what qualifies as a sentence, the entire problem can easily be avoided by reformulating the original example to replace the word “sentence” with the word “argument”: “Each word in this argument is in English; therefore, this argument is in English.”
190
Chapter 5
of the present section is dedicated to a preliminary exposition of what critics mean when they accuse Mill of committing the fallacy. The aim is to enumerate the most common critical approaches. From the initial premise that each person’s happiness is desirable for the person whose happiness it is, one might infer, legitimately or not, any of the following: (1) each person’s happiness is desirable for every other person; (2) the general happiness is desirable for each person; (3) each person’s happiness is desirable for the aggregate all persons. On the other hand, Mill’s explicit inference from the same premise concludes that (4) the general happiness is desirable for the aggregate of all persons. None of the first three inferences ends with that conclusion. Hence, none of the first three inferences represents the entire argument in which Mill is accused of committing the fallacy of composition. The problem, however, is that any of the first three inferences, each of which seems erroneous, could conceivably constitute a portion of Mill’s argument. The next section demonstrates that the first three inferences are indeed erroneous, but that none of them represents any part of Mill’s argument, which moves directly from the premise to the conclusion, without any intermediate steps. Even before that demonstration, however, it will be useful, particularly at this point, to consider the problem in greater detail. The main problem is that the reasoning that gets Mill from the initial premise, that each person’s happiness is desirable for the person whose happiness it is, to the explicit conclusion, that the general happiness is desirable for the aggregate of all persons, appears to be a composite argument with two inherent inferences from parts to whole. It will be illuminating, then, at least as a first approximation, to analyze the suspect argument through two parallel inferences taken as its components. One inference moves from personal happiness to the general happiness, while predicating desirability as an attribute of happiness as the subject. The other inference moves from each person to the aggregate of all persons, while predicating happiness as an attribute of people as the subject. Interpreting the two component inferences as instances of the fallacy of composition is a matter of construing the “general happiness” as the sum of all particular happinesses (whether or not that amounts to anything in particular) and the “aggregate of all persons” as the sum of all individual persons (whether or not that amounts to anything in particular). With these terms understood as stipulated, each inference clearly commits the fallacy of composition, especially in the absence of contextual qualifications that may otherwise be relevant. One inference moves from the desirability of each particular happiness to the desirability of the sum of all particular happinesses, while the other inference moves from that which is desirable for each person to that which is desirable for the sum of all persons.
Alleged Fallacy of Composition
191
Put simply, an adequate analysis of Mill’s proof in terms of the fallacy of composition must deal with potential problems in two separate moves, one in summing up particular happinesses, the other in summing up individual persons. Most critics tend to note problems with one or the other but not with both of those moves. John Stuart Mackenzie is unique in that respect among the early critics of Mill. His critique is comprehensive in terms of the moves involved in the parts-whole inference. Quoted in full in chapter 3 (section 3.7), here again is Mackenzie’s analysis of “the fallacy involved in the inference that ‘the general happiness is a good to the aggregate of all persons’ ” (1893, 104): The fallacy is that which is known in logic as “the fallacy of composition.” It is inferred that because my pleasures are a good to me, yours to you, his to him, and so on, therefore my pleasures +your pleasures +his pleasures are a good to me +you +him. It is forgotten that neither the pleasures nor the persons are capable of being made into an aggregate. mackenzie 1893, 104
The gist of Mackenzie’s extended criticism of this portion of the proof is that the general happiness and the aggregate of all persons are both illusory concepts. It is a mistake to add up particular pleasures, as if the sum itself would be a pleasure, thus constituting a personal happiness belonging to anyone in particular, and it is an additional mistake to add up individual persons, as if the sum itself would be a person, perhaps a larger or better person, or even just another person. Concerning the first mistake, Mackenzie summarizes his objection as follows: “If pleasure is the one thing that is desirable, it is clear that a sum of pleasures cannot be desirable; for a sum of pleasures is not pleasure” (1893, 112). Concerning the second mistake, he objects that “ ‘the aggregate of all persons’ is nobody, and consequently nothing can be a good to him. A good must be a good for somebody” (1893, 104). Mackenzie’s astute observations compound the difficulties facing Mill in the already controversial passage about the general happiness. It seems that Mill commits the fallacy of composition twice in one argument, or in other words, that he commits the fallacy of composition in two different ways, or in two different directions, in the same argument. One direction is where desirability, initially attributed to each person’s happiness, is consequently attributed to the general happiness, whereby desirability ends up as an attribute predicated of happiness distributively in the premises and collectively in the conclusion. The other direction is where happiness is identified as a good, first for each person, and then for the aggregate of all persons, whereby happiness ends up
192
Chapter 5
being a good associated with persons distributively in the premises and collectively in the conclusion.3 Even if Mill himself never actually took a single step in either direction, which is what the present chapter aims to demonstrate, the critical suspicion that he did requires serious consideration of how he could have done so, had he been inclined to do so, which is to explore how anyone could do so, that is to say, how it could be done at all. Covering all the relevant possibilities will help identify and refute the strongest form of the objection that Mill commits the fallacy of composition. The process requires interpreting the original argument as an enthymeme. Consider the following four reconstructions:
First Reconstruction (1) Each person’s happiness is desirable for the person whose happiness it is; therefore, each person’s happiness is desirable for every other person. (2) Each person’s happiness is desirable for every other person; therefore, the general happiness is desirable for the aggregate of all persons. Second Reconstruction (1) Each person’s happiness is desirable for the person whose happiness it is; therefore, the general happiness is desirable for each person. (2) The general happiness is desirable for each person; therefore, the general happiness is desirable for the aggregate of all persons. Third Reconstruction (1) Each person’s happiness is desirable for the person whose happiness it is; therefore, each person’s happiness is desirable for the aggregate of all persons. (2) Each person’s happiness is desirable for the aggregate of all persons; therefore, the general happiness is desirable for the aggregate of all persons.
3 This is the same as saying: The other direction is where the desirability of happiness is identified first with each person and then with the aggregate of all persons, whereby the desirability of happiness ends up being associated with persons distributively in the premises and collectively in the conclusion.
Alleged Fallacy of Composition
193
Fourth Reconstruction Each person’s happiness is desirable for the person whose happiness it is; therefore, the general happiness is desirable for the aggregate of all persons.
The first reconstruction, representing a historically common misinterpretation of Mill’s argument, assembles a composite argument from component inferences that are different from the two that were introduced above. The second and third reconstructions combine the two familiar component inferences in reverse order. The fourth reconstruction, serving more as a paraphrase of Mill’s actual words than as an interpretation of them, does not follow the enthymematic structure of the other examples. All four reconstructions start with the same premise and end with the same conclusion. The divergence comes in the intermediate step, which is different in each of the first three reconstructions and altogether absent in the fourth. The challenge is to determine which intermediate step, if any, joins Mill’s initial premise that each person’s happiness is desirable for the person whose happiness it is with his explicit conclusion that the general happiness is desirable for the aggregate of all persons. The next section shows that none of the intermediate steps in the first three reconstructions is faithful to the text. The point, of course, is to demonstrate that none of the intermediate steps is implicit in the original argument, since it is already obvious that none of them is explicit there. The negative effort in this process of elimination is followed by a positive approach to develop the fourth reconstruction as an instantiation that exhibits the fallacy of composition without the misconstrual of the first three scenarios. As the key to the question whether Mill commits the fallacy of composition appears to be in the intermediate steps producing different interpretations of how he might be doing so, the next section simplifies the reconstructions above into scenarios representing only the intermediate steps (where applicable). It thus focuses on just the intermediate steps in the first three reconstructions and the entirety of the argument in the fourth reconstruction, where there is no intermediate step. 5.2
Analysis of the Charges
The highlight of the previous section is various reconstructions of different interpretations of a brief segment of a single paragraph. That short passage attracts excessive attention in the literature because Mill leaves his main points
194
Chapter 5
undeveloped and takes excursions into ramifications that are neither as confusing nor as controversial. The danger in any attempt to make up for what Mil neglects in that regard is not so much in elaboration at the cost of brevity but in brevity at the cost of clarity. Hence, at the risk of repetition, but for the sake of clarity, as well as for convenience in reference, here is a list of scenarios representing the intermediate step in Mill’s inference from the initial premise to the explicit conclusion:
(S1) Each person’s happiness is desirable for the person whose happiness it is; therefore, each person’s happiness is desirable for every other person. (S2) Each person’s happiness is desirable for the person whose happiness it is; therefore, the general happiness is desirable for each person. (S3) Each person’s happiness is desirable for the person whose happiness it is; therefore, each person’s happiness is desirable for the aggregate of all persons. (S4) Each person’s happiness is desirable for the person whose happiness it is; therefore, the general happiness is desirable for the aggregate of all persons.
These scenarios illustrate different ways of starting at the personal level and ending up at the social level, be it through an interpersonal dimension or through a collective reference. They thus represent different ways of accusing Mill of committing the fallacy of composition, as they jointly cover the most tempting and unambiguous conclusions that do not follow from the premise that each person’s happiness is a good to the person whose happiness it is. The fourth scenario, merely paraphrasing Mill’s actual argument, represents the strongest version of the charge, because it remains open to the possibility that the offending aggregation occurs twice, once for happiness, once for persons. 5.2.1 First Scenario The first scenario interprets the intermediate step in Mill’s argument as follows: Each person’s happiness is desirable for the person whose happiness it is; therefore, each person’s happiness is desirable for every other person. This inference is obviously problematic but the underlying reasoning is obscure
Alleged Fallacy of Composition
195
enough to conceal whether it involves the fallacy of composition or some other mistake. Yet no matter why the inference in the intermediate step here fails, the transition to Mill’s explicit conclusion is a clear candidate for the fallacy of composition: Each person’s happiness is desirable for every other person; therefore, the general happiness is desirable for the aggregate of all persons. This scenario is sometimes taken as a clarification of the reasoning behind Mill’s argument rather than as an intermediate step in that argument. On that interpretation, Mill means to conclude nothing other than that each person’s happiness is desirable for every other person, where he infers that the general happiness is desirable for the aggregate of all persons. He thus argues, in essence, that each person’s happiness is desirable for the person whose happiness it is, and therefore that each person’s happiness is desirable for every other person. Whether as an intermediate step or as a complete interpretation, the first scenario takes Mill to infer the desirability of each person’s happiness for every other person from the desirability of each person’s happiness for the person whose happiness it is. This inference extracts such a strong claim from such an innocuous truism that the logical gap is difficult to attribute uniquely to any particular fallacy to the exclusion of other problems. Not every bad argument neatly exemplifies a logical fallacy, of course, but this one seems to exhibit characteristics of both the fallacy of composition and the fallacy of converse accident. It is difficult to pin down the mistake as the fallacy of composition, because the conclusion that each person’s happiness is desirable for every other person makes no reference to a collective whole, though it does seem to refer distributively to every part of the general happiness and to every member of the aggregate of all persons. The distributive reference in the conclusion suggests that the mistake might be the fallacy of converse accident, but this is difficult to show as well, because while the conclusion does indeed refer both to every happiness and to every person, the premise likewise refers to every happiness and to every person. The problem is easier to describe than it is to label. The inference relies on a misleading universal quantification of the subject and predicate throughout the argument to conceal a hasty generalization, namely from the desirability of each person’s happiness for that person, hence from a private phenomenon at the personal level, to the desirability of each person’s happiness for every other person, hence to a social phenomenon at the interpersonal level. The premise that each person’s happiness is desirable for the person whose happiness it is describes an internal albeit universal state of affairs that is true of everyone, so that, in each case, exactly one person’s happiness is desirable for exactly one person. The
196
Chapter 5
problem in the inference is the transition from a private relationship between one person and that same person’s happiness to a social relationship between each person and every other person’s happiness. The personal desirability of each person’s happiness turns inexplicably into the interpersonal desirability of everyone’s happiness. The first scenario thus construes Mill as either concealing or ignoring the obvious fact of nature, at least of human nature but also of our existential circumstances, that what is desirable for one person is not necessarily desirable for another. For example, having eggs and bacon for breakfast may be desirable for me but not for my vegan friends; taking a swing at the boxing champion may be desirable for the official challenger but not for an ordinary spectator; driving exceptionally fast may be desirable for paramedics transporting a patient but not for tourists enjoying the scenery. The emerging objection is that, insofar as what is desirable for one person is not necessarily desirable for another, the desirability of each person’s happiness for the person whose happiness it is does not support the desirability of each person’s happiness for every other person. The first scenario, however, is not faithful to the text, given that the actual argument neither concludes nor presupposes that each person’s happiness is desirable for every other person. Mill’s argument does not in any way rely on the fungibility of goods and persons. While Mill does infer the desirability of the general happiness for the aggregate of all persons from the desirability of each person’s happiness for the person whose happiness it is, he makes no reference, either in the inference itself or in support of it, to the desirability of each person’s happiness for every other person. Of course, Mill’s explicit endorsement of interconnections between goods and persons, particularly in the form of an interdependence between them, is not necessary to accuse him of holding such a position.4 His words can conceivably be taken, not entirely unreasonably in the absence of further information, as implying a belief in the desirability of each person’s happiness for every other person. Since the extensions of “each” and “every” are contained in the extensions of “general” and “aggregate,” presumably even in the context of Mill’s argument, it is not immediately clear whether or not Mill is guilty of making the illicit inference attributed to him in the first scenario.
4 What is meant by the interdependence of goods and persons is not the dependence of goods on persons, or of persons on goods, but the universal desirability of any good for any person, so that each person’s good is a good to every person.
Alleged Fallacy of Composition
197
The truth might have remained forever in obscurity had it not been for a letter from Mill to Henry Jones (Letter 1257: 13 June 1868) in response to the latter’s inquiry into this very question. Already quoted in chapter 3 (section 3.3) of the present volume, here again is Mill’s answer in his own words: As to the sentence you quote from my “Utilitarianism”; when I said that the general happiness is a good to the aggregate of all persons I did not mean that every human being’s happiness is a good to every other human being; though I think, in a good state of society & education it would be so. I merely meant in this particular sentence to argue that since A’s happiness is a good, B’s a good, C’s a good, &c., the sum of all these goods must be a good. mill CW16:14145
Mill’s response confirms that he does not confuse the desirability of the general happiness for the aggregate of all persons with the desirability of each person’s happiness for every other person.6 He neither presupposes nor concludes that what is a good to one person is a good to others. Yet even though the letter discredits the interpretation that Mill thinks each person’s happiness is a good to every other person, it clarifies and supports the interpretation that he commits the fallacy of composition, albeit that he commits it without supposing that each person’s happiness is a good to every other person. It is ironic that the very letter clearing the argument of the difficulties associated with the first scenario nevertheless presents it as a prime example of the fallacy of composition, worthy of demonstration in introductory logic textbooks.
5 Hugh Samuel Roger Elliot (1910, vol. 2, 115–116) also reproduces this passage, along with the entire letter of which it is a part, in his edition of The Letters of John Stuart Mill. While he does not identify the recipient of the letter, he claims that all but the last paragraph (cited in its entirety in the main text above) was written by Mill’s stepdaughter Helen Taylor. 6 Francis Herbert Bradley (1927, 113, n. 1), Everett W. Hall (1949, 9), David Daiches Raphael (1955, 349), Carl P. Wellman (1959, 273), Spencer K. Wertz (1971, 425–426), and Henry Robison West (1972, 257; 1982, 30; 2004, 141; 2006, 176; 2017, 338) also cite this paragraph of Mill’s letter to Jones as evidence that Mill does not commit himself, at least in the third paragraph of the fourth chapter of Utilitarianism, to a belief that each person’s happiness is a good to every person. They are not alone in doing so. Yet it is both difficult and pointless to acknowledge all Mill scholars who incorporate this letter in their commentaries. The ones cited here mention the letter prior to the publication of Mill’s letters in the standard critical edition of his Collected Works (1972)—or in the case of West (1972), in the same year, at least in his first contribution—where a detailed index enables convenient access to any letter of Mill pertinent to any study of Mill.
198
Chapter 5
Mill’s clarification that “since A’s happiness is a good, B’s a good, C’s a good, &c., the sum of all these goods must be a good” (CW16:1414) seems to exemplify the fallacy even more clearly and more convincingly than the argument it is meant to clarify. The logical problem surviving Mill’s response to Jones can be explained as follows: The fact that each of several things is individually a good does not mean that all of those things together must collectively be a good. Even if every single one of the pills in a bottle of medicine can alleviate pain, all of them together might just kill the patient. What are we to make of Mill’s persistence in summing up the personal happiness of different individuals? He already comes across in the third paragraph of the fourth chapter of Utilitarianism as someone who cannot string together a few words without committing several elementary fallacies. As if this were not bad enough, when Jones raises a question concerning the very sentence in which Mill allegedly commits the fallacy of composition, let alone noticing a problem, Mill confidently restates his position in words that approximate the form of the fallacy at least as closely as before if not more so. It is difficult to blame critics for opposing Mill in terms approaching mockery. Especially in the context of the Jones letter, which seems to do as much harm as it does good, it appears not only that Mill commits the fallacy of composition but also that he commits it repeatedly and that he would not even recognize the fallacy if it were pointed out to him. The response developed in section 5.3 shows that this is not an accurate picture of Mill. The first order of business, however, is to complete the reconstructions in progress. Mill’s explanation to Jones serves as a reminder in what follows to avoid any formulations holding or implying that what is a good to one is therefore a good to others. 5.2.2 Second Scenario The second scenario reconstructs Mill’s argument without assuming the interdependence of goods and persons. It construes Mill as switching from a distributive to collective reference to “happiness” while maintaining a distributive reference to “persons.” It thus takes him to be reasoning from the desirability of each person’s happiness (for the person whose happiness it is) to the desirability of the general happiness (for each person), or what is the same thing, from each person’s happiness (as a good) to the general happiness (as a good). The point of this scenario is to isolate and expose the problem in adding up particular happinesses by anchoring the general happiness to each individual person instead of allowing an aggregation of individual persons as well. The intermediate step in Mill’s argument works as follows under this scenario: Each person’s happiness is desirable for the person whose happiness
Alleged Fallacy of Composition
199
it is; therefore, the general happiness is desirable for each person. The same argument can be restated as follows in the notation of the Jones letter: a is a good to A, and b is a good to B, and c is a good to C; therefore, a + b + c is a good to A and B and C individually. However, this scenario heeds Mill’s warning in the Jones letter, and grants that Mill does not assume any of the following relationships: that a is a good to B and C, in addition to being a good to A; that b is a good to A and C, in addition to being a good to B; and that c is a good to A and B, in addition to being a good to C. It instead takes Mill to maintain that a + b + c is a good to A, regardless of the relation of b and c to A, because a is a good to A; that a + b + c is a good to B, regardless of the relation of a and c to B, because b is a good to B; and that a + b + c is a good to C, regardless of the relation of a and b to C, because c is a good to C. Hence, a + b + c is a good to A and B and C individually, simply because a is a good to A, and b is a good to B, and c is a good to C. The second scenario frees Mill of a commitment to the interdependence of goods and persons, but it still confirms his commitment of the fallacy of composition. The argument here is that the general happiness is desirable for each person because a particular part of the general happiness is desirable for each person. More perspicuously, the general happiness is common to every person insofar as a discrete part of the general happiness is unique to each person. The problem is in the exception introduced by the term “regardless” in the formulation in the preceding paragraph. It is hasty to focus on the desirability of any particular part of a whole for any given person, to the exclusion of the desirability of other parts of that same whole for that same person, and it is wrong to assume that not every part of a whole is relevant to determining whether the whole that is composed of those parts is desirable for any given person. The second scenario ignores the possibility that a whole can contain some parts that are desirable and others that are undesirable for any given person. Suppose that a is a part of the general happiness (a + b + c) and that a is a good to A. According to the reconstruction of Mill’s argument in the second scenario, a + b + c is a good to A for the sole reason that a (which is a part of a + b + c) is a good to A. However, consider the following complication in this population sample: a is insulin, which is a good to A, who is a diabetic, and b is chocolate, which is a good to B, who is a chocoholic. Suppose that this complication comes with the assumption that the chocolate is not sugar-free, so that it is bad for the diabetic, and furthermore, that the chocoholic does not suffer from a metabolic disorder or any other medical condition, so that the chocolate is indeed a good to the chocoholic. The complication and the assumptions show, for one thing, that a + b + c cannot be a good to A, because b (chocolate) is not a good to A (the diabetic).
200
Chapter 5
The second scenario thus takes Mill to commits the fallacy of composition in the intermediate step, which supposes that a + b + c is a good to A and B and C individually, simply because a is a good to A, and b is a good to B, and c is a good to C. Mill’s argument, as interpreted in this scenario, fails at this point, where it commits the fallacy of composition, and need not be followed in detail to its conclusion, on the way to which it commits the fallacy of composition once again. For the sake of completeness, though, the final step toward Mill’s explicit conclusion can be sketched as follows: If a + b + c is a good to A and B and C individually, then a + b + c is a good to A and B and C collectively; a + b + c is a good to A and B and C individually; therefore, a + b + c is a good to A + B + C. 5.2.3 Third Scenario The third scenario is similar to the second in reconstructing Mill’s argument without the assumption of interdependence between goods and persons. Detecting or postulating a switch from distributive to collective reference to “persons,” combined with a constant mode of distributive reference to “happiness,” this scenario takes Mill to be reasoning from something’s (happiness) being desirable for each person to that same thing’s (happiness) being desirable for the aggregate of all persons. The point of the scenario is to isolate and expose the problem in adding up individual persons by specifically testing the desirability of each particular happiness for the aggregate of all persons instead of allowing, in addition, an aggregation of particular happinesses into the general happiness. The third scenario thus interprets the intermediate step in Mill’s argument as follows: Each person’s happiness is desirable for the person whose happiness it is; therefore, each person’s happiness is desirable for the aggregate of all persons. This can be rendered as follows in the notation of the Jones letter: a is a good to A, and b is a good to B, and c is a good to C; therefore, a and b and c are severally goods to A + B + C. The third scenario, like the second, grants that Mill does not assume the following relationships: that a is a good to B and C, in addition to being a good to A; that b is a good to A and C, in addition to being a good to B; and that c is a good to A and B, in addition to being a good to C. It instead takes Mill to maintain that a is a good to A + B + C, regardless of the relation of a to B and C, because a is a good to A; that b is a good to A + B + C, regardless of the relation of b to A and C, because b is a good to B; and that c is a good to A + B + C, regardless of the relation of c to A and B, because c is a good to C. Hence, a and b and c are severally goods to A + B + C, simply because a is a good to A, and b is a good to B, and c is a good to C.
Alleged Fallacy of Composition
201
The interpretation here is similar to the one in the second scenario, but it represents the flip side of the coin. Whereas the second scenario takes the general happiness to be a good to each person, the third scenario takes each person’s happiness to be a good to the aggregate of all persons. The problem here is also similar to the one in the second scenario. Suppose that A is a part of the aggregate of all persons (A + B + C) and that a is a good to A. According to the reconstruction of Mill’s argument in the third scenario, a is a good to A + B + C for the sole reason that a is a good to A (which is a part of A + B + C). The problem is again apparent in the employment of the term “regardless” in the reconstruction of the argument. Suppose again that a is insulin, which is a good to A, who is a diabetic, and that b is chocolate, which is a good to B, who is a chocoholic. It follows that a cannot be a good to A + B + C, because a (insulin) is not a good to B (the chocoholic). The problem here is in the implicit assumption that something’s being a good to a part of a whole precludes its not being so with respect to other parts of the same whole. The third scenario, then, ignores the possibility that the same thing can be a good to one person but not so to another, possibly even being beneficial to one but harmful to another, or desirable for one but undesirable for another. The third scenario, like the ones preceding it, takes Mill to commit the fallacy of composition in the intermediate step, which supposes, in this particular case, that a and b and c are severally goods to A + B + C, simply because a is a good to A, and b is a good to B, and c is a good to C. The argument fails at this point, as it commits the fallacy of composition, but it goes on immediately afterwards to commit the same fallacy again: If a and b and c are severally goods to A + B + C, then a and b and c are collectively goods to A + B + C; a and b and c are severally goods to A + B + C; therefore, a + b + c is a good to A + B + C. The second and third scenarios demonstrate two different ways of reconstructing Mill’s argument as committing the fallacy of composition. Yet Mill does not make or intend either of the inferences attributed to him in these scenarios as an intermediate step. He infers neither that the general happiness is a good to each person, nor that each person’s happiness is a good to the aggregate of all persons. He instead moves directly from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons. Both subjects, happiness and persons, are generalized. The move from parts to whole is on both counts. This leads to the fourth scenario.
202
Chapter 5
5.2.4 Fourth Scenario The fourth scenario takes Mill’s complete argument as it is, in other words, without speculation on intermediate steps, because there are no intermediate steps: Each person’s happiness is desirable for the person whose happiness it is; therefore, the general happiness is desirable for the aggregate of all persons. The reasoning behind the argument is that the general happiness is desirable for the aggregate of all persons because each of the several parts of the general happiness is desirable for one or another member of the aggregate of all persons and because there is no part of the general happiness that is not desirable for at least one member of the aggregate of all persons. In terms of the familiar notation, a + b + c is a good to A + B + C, because a is a good to A, and b is a good to B, and c is a good to C. If this argument commits the fallacy of composition, then the underlying problem is a combination of the problems in the second and third scenarios. Granted, Mill claims neither that a + b + c is a good to A and B and C individually, as presented in the second scenario, nor that a and b and c are severally goods to A + B + C, as presented in the third scenario. However, even if Mill’s actual argument is not a combination, or a full sweep, of the intermediate steps in the second and third scenarios, those two scenarios could conceivably serve jointly as an explanation of how the argument nevertheless commits the fallacy of composition in one fell swoop without intermediate steps. The argument need not explicitly invoke the intermediate steps in the second and third scenarios in order to share the problems inherent in those steps. Hence, moving directly from the premise that a is a good to A, and b is a good to B, and c is a good to C, to the conclusion that a + b + c is a good to A + B + C, without presupposing, either with the second scenario, that a + b + c is a good to A and B and C individually, or with the third scenario, that a and b and c are severally goods to A + B + C, does not automatically absolve Mill’s argument of the fallacy of composition. This is merely a description of the problem, not a solution to it. But it is a better description than the formulations in the second and third scenarios, because those distort Mill’s conception of the general happiness and his notion of the aggregate of all persons, thus making it necessary to develop the fourth scenario along different lines. The first three scenarios describe a simple universe comprising three items (a, b, c) and three persons (A, B, C). The thought experiments that sketch the inferences in the second and third scenarios pick out a subset of items and persons. The subsequently emerging complication comprises two items and two persons where each item is a good to one person but not to the other. The second and third scenarios take this complication to be a self-evident demonstration of the failure of Mill’s inference. The second scenario takes it to show
Alleged Fallacy of Composition
203
that the two items together are not a good to either of the two persons individually. The third scenario takes it to show that neither of the two items is a good to the two persons together. If the second and third scenarios serve as a joint explanation of how Mill’s explicit argument commits the fallacy of composition, then the explanation of the problem in Mill’s argument is that the two items together are not a good to the two persons together, because each item is a good to one person but not to the other. However, the presumably problematic claim that the two items together constitute a good to the two persons together does not come with the requirement that each person must share the other person’s good. Mill’s position, then, cannot be relegated to the inferences in the second and third scenarios, because he explicitly rejects the interdependence of goods and persons in the first scenario, which turns out to provide the fundamental mechanics of the thought experiments in the second and third scenarios. The first scenario emerges as the foundation of the latter two in the sense that the diabetic and chocoholic analogies of the latter two scenarios draw on the interdependence of goods and persons introduced in the first scenario. Consider, once again, the analogies presented as complications: A is a diabetic, B is a chocoholic, a is insulin, and b is chocolate. Hence, a is a good to A, but not to B, while b is a good to B, but not to A. The second and third scenarios evaluate Mill’s argument in intermediate steps that fail by definition. The second scenario objects that a + b is not a good to A and B individually, because a is not a good to B, and b is not a good to A. The third scenario objects that a and b are not severally goods to A + B, because a is not a good to B, and b is not a good to A. If Mill’s complete argument, which does not explicitly take the intermediate steps in the second and third scenarios, nevertheless shares the problems inherent in those intermediate steps, then the objection that emerges from the second and third scenarios to Mill’s complete argument is that a + b is not a good to A + B, because a is not a good to B, and b is not a good to A. The objection, then, is that a + b is not a good to A + B, because it is not true that each person’s happiness is a good to every other person. If this is the grounds for the objection, then the second and third scenarios collapse into the first scenario, where goods and persons are taken to be interdependent. The common denominator of the objections in the first three scenarios turns out to be that the general happiness cannot be a good to the aggregate of all persons unless each person’s happiness is a good to every other person, and that this condition is not satisfied on any interpretation of Mill’s premise that each person’s happiness is a good to the person whose happiness it is. Mill, in contrast, explicitly denies the interdependence of goods and persons. The second and third scenarios distort his conception of the general
204
Chapter 5
happiness. His general happiness is not a bundle of particular objects of desire, such as insulin and chocolate and so forth. Mill argues that “each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons” (U4:3 CW10:234). This is not the same as arguing that all objects of desire known to make any person happy constitute a bundle of goods for the person they make happy, and therefore that the sum of all such bundles constitute a good for the aggregate of all persons. The definition of a as insulin and b as chocolate misconstrues Mill to be reasoning about the specific constituents of each person’s happiness. Although the general happiness does presuppose each person’s happiness, it does not presuppose any particular means to or parts of each person’s happiness. Thus, in any hedonistic conception of happiness, a represents A’s feeling pleasure, and b represents B’s feeling pleasure, so that a + b represents A’s feeling pleasure simultaneously with B’s feeling pleasure. This general formula is free of the difficulties in the previously introduced complications. The apparent problem in the thought experiment concerning the diabetic’s insulin and the chocoholic’s chocolate is that chocolate is bad for the diabetic and insulin is bad for the chocoholic, and therefore that the two items together cannot be a good to the two persons together. However, it is not necessary for each person to share or appreciate the other person’s good in order for the package as a whole to be a good to the persons as a group. It is not as if the chocolate were to be shoved down the diabetic’s throat and the insulin injected into the chocoholic’s body. The correct instantiation of the general formula requires a reconfiguration of the specifications of the thought experiment as follows: Define a as insulin in A’s body rather than as insulin, and define b as chocolate in B’s mouth rather than as chocolate. In that case, a + b means insulin in A’s body and chocolate in B’s mouth. With the complication reformulated along these lines, the supposedly illicit inference in the thought experiment concludes not simply that insulin and chocolate together is a good to the diabetic and chocoholic together but specifically that insulin in the diabetic’s body together with chocolate in the chocoholic’s mouth is a good to the diabetic and chocoholic together. The thought experiment must therefore be reconfigured and reinterpreted specifically as a challenge to the inference from the premise that the diabetic’s happiness is a good to the diabetic and the chocoholic’s happiness is a good to the chocoholic to the conclusion that the diabetic’s happiness together with the chocoholic’s happiness is a good to the diabetic together with the chocoholic. The experiment originally starts out as an ontological challenge to the private goods of individual persons adding up to a common good for the community as a whole, where the goods in question are itemized as specific
Alleged Fallacy of Composition
205
things. However, the proposed material complication that insulin and chocolate together cannot be a good to the diabetic and chocoholic together, falsely assuming that anything that is not a good to any given person in the aggregate is somehow forced upon that person, or must in any way be associated with that person, turns out to be a manifestation of the more general and strictly existential difficulty that the diabetic’s happiness together with the chocoholic’s happiness is not a practical possibility, or in short, that not everyone can be happy together. Aligning the specifications of the thought experiment with Mill’s conception of the general happiness and with his understanding of the aggregate of all persons transforms the original challenge concerning the desirability of specific goods into a different objection concerning the possibility of mutual happiness. The fourth scenario must be developed in conformity with the fact that Mill is concerned with happiness itself rather than with the specific constituents or ingredients of happiness. Consider the transition from the desirability of each person’s happiness to the desirability of the general happiness in the population sample consisting of Alice and Betty and Carol. To acknowledge the desirability of each person’s happiness in this population sample is to affirm that it is good that Alice be happy and it is good that Betty be happy and it is good that Carol be happy. To acknowledge the desirability of the general happiness in this population sample is to affirm that it is good that Alice and Betty and Carol be happy. But what exactly does that mean? Mill’s inference from the desirability of each person’s happiness to the desirability of the general happiness comes across as a rather fuzzy transition from the proposition that “it is good that Alice be happy and it is good that Betty be happy and it is good that Carol be happy” to the ambiguous conclusion that “it is good that Alice and Betty and Carol be happy.” The ambiguity in this transition is the source of the controversy whether Mill commits the fallacy of composition. The key to resolving the ambiguity lies in figuring out whether Mill means (1) that “it is good that Alice and Betty and Carol all be happy together” or (2) that “it is good that Alice and Betty and Carol each be happy individually”? The typical objection holds that Mill intends the first meaning and thereby commits the fallacy of composition. However, note that the apparent dilemma does not really leave Mill with anywhere safe to go from the initial proposition that “it is good that Alice be happy and it is good that Betty be happy and it is good that Carol be happy.” If a transition to the proposition that “it is good that Alice and Betty and Carol all be happy together” commits the fallacy of composition, then a transition to the proposition that “it is good that Alice and Betty and Carol each be happy individually,” perhaps in the same spirit, commits the fallacy of division.
206
Chapter 5
At any rate, the interpretation of Mill’s argument in the fourth scenario is the strongest form of the charge that Mill commits the fallacy of composition, at least in the sense of capturing the most comprehensive sense of that charge: If it is good that Alice be happy, and it is good that Betty be happy, and it is good that Carol be happy, then it is good that Alice and Betty and Carol all be happy. On the other hand, to object that they are not or cannot all be happy, for whatever reason (psychological, sociological, economic, etc.), says nothing against the position that it is indeed good that they all be happy. To be sure, various contingencies might practically guarantee that Alice and Betty and Carol cannot all be happy. They might, for example, be competing for an extremely scarce economic resource, which only one of the three can have, but without which none of the three can be happy. Or they might happen to hate one another so much that none of them is psychologically capable of being happy unless the other two are miserable. In either case, the fact that Alice and Betty and Carol are not or cannot all be happy does not contradict the conclusion that it is good that Alice and Betty and Carol all be happy. Given that it is not logically contradictory for Alice and Betty and Carol to be happy all at once, it can be true both that it is good that Alice and Betty and Carol all be happy and that Alice and Betty and Carol are not or cannot all be happy, due to psychological, sociological, or economic contingencies, which complicate their personal and practical relationships with one another and which thereby preclude their mutual happiness as a group. The obstacle to a clear understanding of the general happiness in this case is the failure to separate two different questions with respect to the general happiness that are evidently separate with respect to personal happiness. The question whether it is good that everyone be happy is different from the question whether everyone actually is or can be happy, just as the question whether it is good that a particular person be happy is different from the question whether that particular person actually is or can be happy. Practical conditions, whether they be psychological, sociological, or economic, admittedly tend to affect the happiness of individual persons. Yet practical contingencies that tend to detract from any person’s happiness, perhaps even preventing it altogether, do not affect the value, goodness, or desirability of that person’s happiness. Psychological serenity, social status, and economic prosperity might all be necessary conditions of Alice’s happiness, but they are not in any sense conditions of the goodness or desirability of Alice’s happiness. The state of her happiness is not relevant to the value of her happiness. Her poverty and
Alleged Fallacy of Composition
207
her depression will surely detract from her happiness, but not from the goodness or desirability of her happiness. Alice’s happiness is valuable, good, and desirable even if Alice is unhappy. Value and reality have no more to do with each other at the level of the general happiness than they do at the level of personal happiness. The value and possibility of the general happiness are just as distinct and independent as the value and possibility of any given person’s happiness. Just as it is good that Alice be happy, even if Alice is not or cannot be happy, so too is it good that Alice and Betty and Carol all be happy, even if Alice and Betty and Carol are not or cannot all be happy. The proposition that it is good that Alice and Betty and Carol all be happy is not contradicted by the proposition that Alice and Betty and Carol are not or cannot all be happy, any more than the proposition that it is good that Alice be happy is contradicted by the proposition that Alice is not or cannot be happy.7 The intention behind the preceding scenarios is not to refute the charge that Mill commits the fallacy of composition but to explore the possibility of making sense of that charge without distorting Mill’s argument. These scenarios show that the charge must observe at least three characteristics of Mill’s argument in order to remain faithful to the text: (1) The argument does not presuppose or imply that each person’s happiness is a good to every other person. (2) The argument deals with happiness, both personal and general, but not with any particular means to or conditions of happiness at any level. (3) The argument concludes that the general happiness is a good to the aggregate of all persons but not necessarily that the individual members of the aggregate of all persons are indeed all happy together or can all easily or even possibly be happy together. The next step, complementing the preceding scenarios, is to investigate the possibility of making sense of Mill’s argument without committing the fallacy of composition. 7 A related question is whether the general happiness includes the happiness of those whose happiness depends on patently immoral contingencies. Does the general happiness include, say, the happiness of a serial killer? And does the goodness or desirability of the general happiness imply the goodness or desirability of a serial killer’s happiness? Is it still good that Alice and Betty and Carol all be happy even if the happiness of Alice lies in murdering the other two? Questions of this kind are misdirected if they are directed at the general level, because such questions are asked and answered at the individual level. Is it good that the serial killer be happy, if killing people, from time to time, is a necessary condition of his or her happiness? Of course not, but once this is established at the individual level, it does not pose an additional problem at the general or collective level.
208 5.3
Chapter 5
Response to the Charges
The gist of the response is that conformity with the logical form of the fallacy of composition is indicative rather than definitive, thus presenting a good reason to investigate whether the argument in question actually commits the fallacy under consideration, but not to conclude automatically that it does. Mill’s argument seems indeed to fit the definition, but it turns out not to commit the fallacy. The full response is organized in three parts. The first part elaborates on the standard definition guiding the analysis up to this point. The aim is to show that, even though every inference that commits the fallacy of composition fits the parts-whole argument form, not every argument conforming to that form actually commits the fallacy. The basic strategy is to consult authoritative sources in informal logic in search of the most reliable definition possible in an effort to assess the accuracy of the definition used here. A survey of the literature reveals the difficulty of formulating a precise definition that picks out all and only those arguments that commit the fallacy. Contemporary accounts are hardly ahead of Mill’s own understanding of the fallacy of composition in A System of Logic. The definitive parts-whole form may be a sufficient guide for identifying candidates for the fallacy, but a formal fit is not itself sufficient to convict suspect arguments without an examination of the contents and context. A case-by-case approach is the only sure way to determine whether a candidate for the fallacy actually commits the fallacy. The second and third parts flesh out Mill’s conceptions of the general happiness and the aggregate of all persons, respectively, to clarify the key terms in his inference from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons. The second part shows that, although Utilitarianism does not provide direct or explicit confirmation of what Mill means by “general happiness,” textual evidence from his other works supports the interpretation that he uses the term to refer distributively to the personal happiness of each person. The third part illustrates what Mill means by “aggregate of all persons.” Drawing on evidence that he subscribes to reductive individualism in his approach to social groups, it demonstrates that Mill holds a social group to be nothing more than the individual members it comprises, and therefore that he uses the expression “aggregate of all persons” to refer distributively to individual persons albeit to all of them at once. 5.3.1 Deconstruction of the Fallacy of Composition The standard definition of the fallacy of composition introduced at the outset (section 5.1) and employed throughout this chapter is useful because every
Alleged Fallacy of Composition
209
instance of that fallacy conforms to the form of the parts-whole inference described. Yet the same definition is inaccurate to the extent that not every argument that conforms to the specified form actually commits the fallacy of composition. The question is whether it is possible to improve the utility or remedy the inaccuracy of the definition. Here, for convenience in reference, is a restatement of the definition: The fallacy of composition is a logically illicit inference from the attributes, properties, or qualities of the parts of a whole to the attributes, properties, or qualities of the corresponding whole. Common variations on this widely accepted definition tend to expand on the generic notion of parts to include other relationships indicative of a mereological association with the whole. They thus interpret the fallacy as a move (1) from the attributes of the parts of a whole to the attributes of the corresponding whole, (2) from the attributes of the elements of a collection to the attributes of the corresponding collection, or (3) from the attributes of the members of a class to the attributes of the corresponding class. The first version is the standard account of the fallacy, but all three interpretations illustrate meaningful variations that philosophers and logicians tend to distinguish as different ways of making the same logical mistake. Charles Leonard Hamblin (1970, 21), for one, claims that “we sometimes need to distinguish physical collections, like piles of sand, from functional collections, like football teams, and these in turn from conceptual collections, like the totality of butterflies.” Irving M. Copi, Carl Cohen, and Kenneth McMahon (2014, 149–150) take the first variation to illustrate one way of committing the fallacy of composition, and the other two together to illustrate another way, finding the latter two similar enough to group together. They characterize the first way as a move from parts to whole, the second as a move from members or elements to the corresponding collection or totality. Mill’s argument need not be analyzed in terms of all three variations, one after the other, which would needlessly complicate the discussion, especially since the illicit transitivity of attributes he allegedly embraces occurs both in generalizing happiness and in aggregating persons. He takes an attribute (goodness or desirability) of each particular happiness to be an attribute (goodness or desirability) of the general happiness as well, while also taking an attribute of each individual person (namely what is a good to them) to be an attribute of the aggregate of all persons (namely what is a good to it) as well. Any assignment of the two moves to one of the variations on the standard definition of the fallacy would be unreliable prior to a complete analysis of the argument and superfluous following a complete analysis. To distinguish between the general happiness and the aggregate of all persons with respect to whether they comprise parts of a whole, elements of a collection, or members
210
Chapter 5
of a class would be as premature as distinguishing between them with respect to whether they are physical, functional, or conceptual collections. Either distinction would also beg the question of the separateness of the two moves. Even if they seem to fit different classifications in this scheme, the two moves are inextricably intertwined, perhaps even constituting two aspects of one move. The safest route is to proceed with the generic “parts-whole” account as the core definition, modifying it in due course to exclude arguments that fit the parts-whole form but do not commit the fallacy. While the core definition covers all instances of the fallacy, it is not limited to fallacious arguments. The argument that “each entry in this dictionary is short; therefore, this dictionary is short,” fits the parts-whole form while failing miserably as an inference for any respectable dictionary. Yet the argument that “each entry in this dictionary contains words; therefore, this dictionary contains words,” also fits the specified form but works rather well for any dictionary. This suggests that it can be misleading to take the argument form alone as the defining characteristic of the fallacy of composition, because some instantiations of that form turn out to be acceptable arguments, the second one above being both valid and sound. The validity and soundness of arguments conforming to the parts-whole form depend on the particular parts, whole, and attributes making up the subject matter of the inference. It is difficult, however, to specify precisely what it is about the individual context that determines the acceptability of an argument that fits the form. Here is another example of a valid and sound argument: Each square of this chessboard is made of wood; therefore, this chessboard is made of wood. Here is another example of a fallacious one: Each square of this chessboard is adjacent to at least two others; therefore, this chessboard is adjacent to at least two others (which does not follow at all, regardless of whether the conclusion is interpreted as referring to other squares or to other chessboards). A distinguishing factor of fallaciousness in the dictionary and chessboard examples seems to be that fallacious arguments that fit the parts-whole form employ either relative terms such as “short” or relational ones such as “adjacent.” If these are indeed decisive distinctions, then the core definition can be expanded as follows in an effort to exclude good arguments: The fallacy of composition is a logically illicit inference from the attributes, properties, or qualities of the parts of a whole to the attributes, properties, or qualities of the corresponding whole, where the word designating the attribute, property, or quality in question is either a relative term or a relational term. Is this really any better? It does seem better, at least at first glance, because it accurately describes a pattern of ambiguity relevant to the preceding examples, which proceed either with relative terms or with relational terms, but
Alleged Fallacy of Composition
211
those examples are not exhaustive of the possibilities. William Leonard Rowe (1962) maintains that any expansion or modification of the core definition through amendments of the sort described above typically produces a worse definition. He finds such modifications defective in two respects. He argues that the amendments introduced fail to reduce the number of valid arguments captured by the core definition, while at the same time impairing the core definition’s ability to capture all fallacious instances. Invoking strength as a classic example of a relative quality, Rowe presents the following inference as an illustration of a valid argument conforming to the expanded definition: “All the parts of this chain are strong; therefore, this chain is strong” (1962, 91). This argument makes a parts-whole inference with a relative term designating the attribute in question, but it does not commit the fallacy of composition. The validity of the argument, suggests Rowe, is evident in the common adage that “a chain is no stronger than its weakest link” (1962, 91). This shows that the expansion of the core definition is superfluous because the core definition alone would have sufficed to characterize the argument incorrectly as committing the fallacy of composition. It is not helpful to modify the definition to make the same mistake. As for an example of an argument that commits the fallacy of composition but escapes the expanded definition, Rowe finds what he needs in plane geometry: “All the parts of this figure are triangular; therefore, this figure is triangular” (1962, 91). In this case, the core definition depicts the argument as fallacious but the expanded version rejects the argument as an instance of the fallacy because “triangular” is neither a relative term nor a relational one. This shows that such amendments to the core definition exacerbate problems because not every instance of the fallacy of composition involves an ambiguity where relative or relational terms are used in one sense in the premises and in a different sense in the conclusion. Unsuccessful or unproductive results in both directions lead Rowe to conclude that “it is clear that we can have no formal or general characterization of the fallacy of composition” (1962, 92). This is both because the core definition already captures valid arguments along with the fallacious ones and because amendments that seem likely to shut the door on any and all valid arguments tend instead to open it even wider to let out some of the fallacious ones.8 8 Yehoshua Bar-Hillel (1964) proposes an amendment to William Leonard Rowe’s (1962) account of the fallacy of composition. He agrees with Rowe’s analysis but finds his terminology inadequate and therefore inaccurate. He considers it misleading to claim that some inferences conforming to the parts-whole form are valid. Finding the claim ambiguous, he recommends a distinction between arguments that are valid on the basis of logic alone and
212
Chapter 5
Despite Rowe’s pessimism, it might still be possible to improve upon the definition of the fallacy to make it more perspicuous without compromising its ability to identify fallacious arguments or impairing its ability to exclude good arguments. Consider the following amendment to the core definition: The fallacy of composition is a logically illicit inference from the attributes, properties, or qualities of the parts of a whole to the attributes, properties, or qualities of the corresponding whole, where the attribute, property, or quality in question is predicated of the subject distributively in the premises and collectively in the conclusion.9 This variation is not susceptible to Rowe’s counterexample from geometry, since triangularity is indeed predicated of the subject distributively in the premises and collectively in the conclusion, even though his counterexample successfully demonstrates that triangularity is not used ambiguously by virtue of relativity or relationality. The proposed definition is not just an improvement, and, in fact, the best we can achieve in that regard, but also an adequate account in itself, and the closest we can get to Mill’s own understanding of the fallacy of composition, particularly as illustrated in A System of Logic (L1:2.3 CW7:27–28; L5:7.1 CW8:818–819). Before proceeding to Mill, however, it will be useful to consider an objection. The prevalence of false positives when attempting to identify fallacious arguments through the standard definition suggests that the only reasonable approach is to parse the problem on a case-by-case basis. That conclusion, in turn, is based on the assumption that validity and fallaciousness are mutually exclusive: that a valid argument cannot be fallacious and that a fallacious one cannot be valid. That assumption, however, is by no means universally accepted. Not everyone holds invalidity to be a condition of fallaciousness. Richard Cole (1965), for example, raises the following objection in opposition to Rowe’s contention that some inferences of the parts-whole form are valid:
those that are valid on the basis of logic supplemented by a set of meaning postulates. He notes that no inferences of the parts-whole type are valid on the basis of logic alone, but that some are indeed valid on the basis of logic supplemented by meaning postulates. He supports Rowe’s overall conclusion, however, as he agrees that we should reserve the term “fallacy” for arguments that are not even valid on the basis of logic supplemented by meaning postulates. 9 Willard Van Orman Quine (1951, 189) endorses this addendum to the core definition, noting that the distinction between distributive and collective predication is typically drawn to resolve the fallacies of composition and division. He adds that this distinction in traditional logic anticipates the further distinction in quantification theory between class inclusion and class membership.
Alleged Fallacy of Composition
213
The point in calling a fallacious argument fallacious is that the argument fails to demonstrate the conclusion from its premises, not that the conclusion’s contradictory is consistent with the premises. … It is too strong to require of every fallacious argument form that the conclusion must be false when the stated premises are true. cole 1965, 432–433
Granting that some arguments in conformity with the parts-whole form are indeed valid, Cole suggests that even valid arguments that fit the specified form of the fallacy still commit the fallacy. His objection to Rowe’s conception of the fallacy of composition finds some support among other philosophers and logicians. Many commentators agree that validity and soundness are not the only marks of a good argument, and further, that an argument need not violate what is required of a good, cogent, or acceptable argument in order to be considered a fallacy.10 Some argue positively that fallaciousness and validity are not mutually exclusive, in other words, that fallacious arguments can be valid, and vice versa.11 However, resolving the question whether fallaciousness and validity (or fallaciousness and soundness) can be mutually consistent is not necessary to make sense of Mill’s argument. Whether or not fallaciousness and validity (or fallaciousness and soundness) are compatible, there will always be some inferences of the parts-whole form that are perfectly acceptable arguments, as in the examples above concerning the dictionary and the chessboard, as well as Rowe’s example of the links in a chain. The debate over the mutual consistency of fallaciousness and validity does not change the situation at the level of specific examples. The position of the present chapter is not merely that some compositional inferences are valid arguments, or sound arguments, but that they are perfectly acceptable arguments that do not fail just because they move from the attributes of the parts of a whole to the attributes of that whole. The question whether they can be both valid and fallacious (or sound and fallacious) becomes irrelevant where the contention is that they work perfectly well—a claim that is not restricted to validity or soundness, though obviously including both.
10 11
Harry J. Gensler (1989, 328–359) and Howard Kahane (1971/2014), among others, list additional requirements for good, cogent, or acceptable arguments, adding that an argument need not violate all good-making characteristics in order to be considered a fallacy. Charles J. Abaté (1979, 262–266), Howard Kahane (1971/2014; 1980), Douglas N. Walton (1982; 1987; 1989), and John Woods and Douglas N. Walton (1976, 52–54) deny that fallaciousness and validity are mutually exclusive.
214
Chapter 5
Validity may not preclude fallaciousness, but form alone cannot establish the latter. That would reduce all fallacies to formal fallacies. The fallacy of composition, on the other hand, is an informal fallacy. The rationale behind it is that it is wrong to argue that what is true of the parts is therefore true of the whole. The point, however, is not that what is true of the parts cannot be true of the whole, but that what is true of the parts might not be true of the whole. While it is clearly a mistake to assume that what is true of the parts must therefore be true of the whole, it is no better, or safer, to assume that what is true of the parts is not or cannot be true of the whole.12 That is why the form of the fallacy of composition, as reflected in various iterations of the core definition developed here, helps to identify all candidates for the fallacy of composition but fails to weed out any of the ones that actually work. Mill devotes the entire fifth book, nearly a hundred pages in all, of A System of Logic to the discussion of fallacies (L5 CW8:733–830). He classifies the fallacies of composition and division as types of the “fallacy of ambiguous terms,” which, in turn, is one of the categories under “fallacies of confusion.” Here is how he introduces the subject: One not unusual form of the Fallacy of Ambiguous Terms, is known technically as the Fallacy of Composition and Division: when the same term is collective in the premises, distributive in the conclusion, or vice versâ: or when the middle term is collective in one premise, distributive in the other. mill L5:7.1 CW8:818
This long and awkward sentence, with its nested colons and various clauses, makes for a confusing introduction at best. The first clause makes it seem as if Mill thinks there is a single fallacy called the “Fallacy of Composition and Division.” The second clause demonstrates that he distinguishes between the two fallacies but not how he distinguishes between them. The order in which he gives the names of the fallacies is the opposite of the order in which he explains the fallacies. And the confusion is certainly not fixed by the appended
12
W. Ward Fearnside and William B. Holther (1959, 27–28) and Charles Leonard Hamblin (1970, 19) discuss the fallacy of composition with emphasis on the importance of recognizing that it is just as wrong to assume that what is true of the parts is not true of the whole (without ruling the possibility out as a matter of fact) as it is to assume that what is true of the parts is therefore true of the whole (without confirming both as a matter of fact).
Alleged Fallacy of Composition
215
“vice versâ,” which merely reverses the direction of a sequence that is already out of order. Fortunately, what follows this muddled introduction is indeed coherent, as Mill proceeds to draw heavily upon Richard Whately’s Elements of Logic (1848) throughout the remainder of his discussion.13 Having given the joint account above, he proceeds with Whately’s (1848, 196) characterization of the fallacy of composition: There is no fallacy more common, or more likely to deceive, than the one now before us. The form in which it is most usually employed is to establish some truth, separately, concerning each single member of a certain class, and thence to infer the same of the whole collectively. mill L5:7.1 CW8:818–819
Continuing to borrow from Whately (1848, 198), Mill also quotes the latter’s explanation of why and how people tend to commit the fallacy of composition: This is a Fallacy with which men are extremely apt to deceive themselves; for when a multitude of particulars are presented to the mind, many are too weak or too indolent to take a comprehensive view of them, but confine their attention to each single point, by turns; and then decide, infer, and act, accordingly: e.g. the imprudent spendthrift, finding that he is able to afford this, or that, or the other expense, forgets that all of them together will ruin him. mill L5:7.1 CW8:819
With Whately’s explanation and example setting up the basics, Mill introduces two of his own examples of the fallacy to elaborate on the mechanics: The debauchee destroys his health by successive acts of intemperance, because no one of those acts would be of itself sufficient to do him any serious harm. A sick person reasons with himself, “one, and another, and another, of my symptoms, do not prove that I have a fatal disease;” and practically concludes that all taken together do not prove it. mill L5:7.1 CW8:819
13
John Mercel Robson, the editor of Mill’s A System of Logic in the University of Toronto edition, reports that page references to Richard Whately’s Elements of Logic are to the ninth edition published in 1848. Robson adds that Mill, too, used the ninth edition pagination in his references to Whately in all editions of A System of Logic following the third (CW8:1233).
216
Chapter 5
Both through external references and through original commentary, Mill’s coverage of the fallacy of composition in A System of Logic leaves no room for doubt that he understands the error in question and that he would not be likely to commit it himself except perhaps through a careless mistake. This is not to say that, because Mill knows the fallacy, he would therefore not commit the fallacy. However, it seems reasonable to think that, because Mill knows the fallacy, if an instance of it in his own writings were pointed out to him, he would consequently reformulate the piece to avoid the fallacy. If either or both of the two allegedly fallacious statements under consideration here (U4:3 CW10:234) indeed commit the fallacy of composition, then it is strange that Mill does not immediately remedy the problem upon receiving an explicit warning in the letter from Henry Jones (cf. Mill’s reply in Letter 1257: 13 June 1868; CW16:1414). He instead responds with an explanation that seems to commit the same fallacy even more clearly than the inference it is intended to explain. Mill evidently had sufficient time for any manner of revision, but saw nothing at all in need of correction, nor even in need of clarification. His letter in response to Jones is dated 13 June 1868, which is three years before the publication of the fourth edition (1871) of Utilitarianism (1861), the last in Mill’s lifetime. The response is articulated in such conformity with the form of the fallacy of composition that the original query must have been about the very possibility of committing that fallacy. Even if Jones had not asked outright whether the inference in question commits the fallacy of composition, it is hard to believe that Mill would have failed to notice whether it did or did not, regardless of precisely how Jones worded the question. The fact that Mill preserved the passage, word for word, in the next edition of Utilitarianism shows that he must have had good reasons for thinking that the inference worked well despite any appearances to the contrary. Since he cites no reasons at all, however, his proof is left with two prima facie instances of the fallacy, which calls for a better defense than that Mill understands the nature of the fallacy of composition. That better defense is available through Mill’s conception of the general happiness and his understanding of the aggregate of all persons, explored respectively in the next two subsections. 5.3.2 Mill’s Conception of the General Happiness Mill’s conception of the general happiness is elusive because he does not define the concept, his understanding of which therefore remains open to interpretation and misinterpretation. Several popular misinterpretations are, in fact, so
Alleged Fallacy of Composition
217
convincing, at least on the surface, that it is best to start with what the general happiness is not and then to gather textual evidence to establish what it is. Ethical theorists, social and political philosophers, and economists alike discuss happiness, or social utility, in terms of various measures facilitating calculation and comparison. It is tempting, in the absence of a definition by Mill, to associate the general happiness with one or another of the measures in popular use. Most misunderstandings arise from a confusion of the general happiness with conventional benchmarks of collective happiness, such as total happiness and average happiness. The total happiness in any population is the sum of the happinesses of the individual persons in that population. The average happiness is the total happiness divided by the number of persons. The general happiness, in contrast to both, is simply the personal happiness of each and every individual in the aggregate of all persons. The most common misinterpretation in this respect is to equate the general happiness with the total happiness. Mill’s language may itself be vague enough to be misleading in certain passages but not so strongly as to point straight to aggregation.14 The essential difference, at any rate, is that the general happiness is not, whereas the total happiness is, a measure of the amount of happiness. The term “general happiness” refers distributively to the personal happiness of each individual, but it does not, in addition, refer to the total amount of happiness enjoyed by them as a group, any more than it refers to the arithmetic mean or standard deviation of the happiness in that population. In other words, the term “general happiness” is a distributive shorthand reference to the personal happiness of each and every person making up the aggregate of all persons, as if to enumerate them all, one by one, referring to Alice’s happiness, Betty’s happiness, Carol’s happiness, and so on. It is not, either in addition or instead, a reference to any measure of the happiness in question. The general happiness is a strictly qualitative concept. Quantitative measures of happiness, however, are not the only tempting sources of misinterpretation. Rousseau’s (1762) notion of the general will, for example, represents another model of misinterpretation for Mill’s conception of the general happiness. A comparison between Mill’s conception of the general happiness and Rousseau’s notion of the general will practically suggests 14
The following two passages from the second chapter of Utilitarianism might, for example, leave the mistaken impression that the general happiness is a measure of quantity at the collective level, whether as a total or as an average: the utilitarian “standard is not the agent’s own greatest happiness, but the greatest amount of happiness altogether” (U2:9 CW10:213); “the happiness which forms the utilitarian standard of what is right in conduct, is not the agent’s own happiness, but that of all concerned” (U2:18 CW10:218).
218
Chapter 5
itself, given that both Mill and Rousseau generalize from an attribute that is common to a multitude of persons severally. While Mill is silent on the composition of the general happiness, Rousseau provides detailed instructions on the composition of the general will. Mill’s silence invites the assumption that the general happiness is likewise composed of particular units of happiness, for which a prototype is readily available in Rousseau’s famous model of composition for the general will (see section 8.1 of chapter 8 below, including the footnotes, for further elaboration on the comparison). The temptation in the comparison is to interpret the general happiness as the common element in the happiness of everyone, especially if the general happiness is to be a good to the aggregate of all persons, without each person’s happiness being a good to every other person. However, this interpretation suggests an entirely different concept, which might perhaps be called the “common happiness,” taken as the regional intersection of everyone’s happiness, that is to say, the area of overlap between one person’s happiness and every other person’s happiness. The common happiness of the aggregate of all persons would then be a synthesis of the greatest happiness of each person compatible with the greatest happiness of every other person in the aggregate of all persons. This notion of common happiness, a heuristic construct inspired by Rousseau’s general will, cannot serve, and will not work, as an interpretive framework for Mill’s general happiness. It is, in fact, entirely alien to Mill’s intent. First, the common happiness is such an intricate abstraction that, had Mill devised and employed such a construct, he would have surely explained it. Mill’s methods of experimental inquiry constitute a set of analytic tools eminently suitable for explicating an abstraction such as the common happiness by systematically isolating and extracting it from the personal happiness of each individual in the relevant population: method of agreement; method of difference; joint method of agreement and difference; method of residues; and method of concomitant variations (L3:8–9 CW7:388–433). Second, and more important, Rousseau’s general will is the epitome of the philosophical attempt to reconcile individual freedom and sociopolitical unity with roughly equal emphasis on each, whereas Mill’s general happiness heavily favors individualism over unity in the absence of a similar reconciliation. It would have been repugnant to Mill to reduce the standard of value, obligation, and justification in ethics to the mutually compatible portion of the happiness of everyone in the community. He was unequivocally opposed to such an insipid and oppressive unification of individuals. Social harmony to the detriment of individuality is, in fact, one of Mill’s principal disagreements with Auguste Comte’s approach to sociology. Note his following summary, in
Alleged Fallacy of Composition
219
Auguste Comte and Positivism (1865), of Comte’s demand for social unity at the expense of individuality: To this theme he [Comte] continually returns, and argues that this unity or harmony among all the elements of our life is not consistent with the predominance of the personal propensities, since these drag us in different directions; it can only result from the subordination of them all to the social feelings, which may be made to act in a uniform direction by a common system of convictions, and which differ from the personal inclinations in this, that we all naturally encourage them in one another, while, on the contrary, social life is a perpetual restraint upon the selfish propensities. mill CW10:337
Claiming that the “fons errorum in M. Comte’s later speculations is this inordinate demand for ‘unity’ and ‘systematization’ ” (CW10:337), Mill presents his objection as follows: It never seems to enter into his [Comte’s] conceptions that any one could object ab initio, and ask, why this universal systematizing, systematizing, systematizing? Why is it necessary that all human life should point but to one object, and be cultivated into a system of means to a single end? May it not be the fact that mankind, who after all are made up of single human beings, obtain a greater sum of happiness when each pursues his own, under the rules and conditions required by the good of the rest, than when each makes the good of the rest his only object, and allows himself no personal pleasures not indispensable to the preservation of his faculties? mill CW10:337
To be sure, Mill’s ideal society is one where the spirit of solidarity eliminates conflict among individuals, but Mill himself repudiates any restriction of individuality in the course of progress toward that ideal. The most relevant source in that regard is On Liberty, where the opening paragraph of the third chapter emphatically identifies the restriction of individuality as a restriction of happiness: It is desirable, in short, that in things which do not primarily concern others, individuality should assert itself. Where, not the person’s own character, but the traditions or customs of other people are the rule of conduct,
220
Chapter 5
there is wanting one of the principle ingredients of human happiness, and quite the chief ingredient of individual and social progress. mill CW18:261
The same chapter of the same book later reaffirms the correlative restriction, leaving no room for doubt as to the intimate connection Mill finds between individuality and happiness: Such are the differences among human beings in their sources of pleasure, their susceptibilities of pain, and the operation on them of different physical and moral agencies, that unless there is a corresponding diversity in their modes of life, they neither obtain their fair share of happiness, nor grow up to the mental, moral, and aesthetic stature of which their nature is capable. mill CW18:270
Notwithstanding the temptation to use familiar mathematical or sociopolitical models to explain Mill’s elusive notion, the general happiness of a population is neither the quantity of happiness, be it as a total or as an average, nor the mutually compatible portion of everyone’s happiness. The general happiness is the personal happiness of each and every individual, but it is not the sum, mean, median, merger, fusion, or synthesis of their happiness. The term distributively denotes the personal happiness of everyone in the relevant population, without implying anything about how they stand in relation to each other or how an increase or decrease in one person’s happiness might affect another person’s happiness. It does not imply that the personal happiness of each person can be added to those of others, or multiplied with them, or synthesized like chemical elements in a compound, or combined to produce something better than its constituents in the manner of ingredients in a recipe for one’s favorite dish. Mill explicitly warns readers, in the portion of the proof where he discusses the possible means to and parts of happiness, that each of the various ingredients of happiness is desirable in itself, and that such ingredients are not “to be looked upon as means to a collective something termed happiness” (U4:5 CW10:235). At that stage in the proof, his specific concern is with the nature of happiness itself, not with the relationship between personal happinesses and the general happiness. However, that concern is the same concern regardless of the level of abstraction. The generic warning can easily be extended to the general happiness: Each of the various ingredients of the general happiness is
Alleged Fallacy of Composition
221
desirable in itself, and such ingredients are not to be looked upon as means to a collective something termed general happiness. In order to bring out Mill’s conception of the general happiness without conflating it with this or that measure of the amount of happiness, or with an altogether different sociopolitical concept, it is best to avoid references to any sort of aggregation or combination of everyone’s happiness. The proper approach is, again, in A System of Logic. What is just as pertinent to the proof, and perhaps more helpful than Mill’s explicit discussion of the fallacy of composition, is his discussion of “names” in the first book of A System of Logic (L1:2.1–8 CW7:24–45). His distinctions there (L1:2.3 CW7:27–29) between general and individual names, and between general and collective names, show a more coherent understanding of the fallacy of composition than do the passages on fallacies in the fifth book (L5:7.1 CW8:818–819). Moreover, his discussion of “names” can be imported more readily and directly to clarify relevant passages in Utilitarianism than can his explicit discussion of the fallacy of composition. Here is the distinction he makes between general and individual names: A general name is familiarly defined, a name which is capable of being truly affirmed, in the same sense, of each of an indefinite number of things. An individual or singular name is a name which is only capable of being truly affirmed, in the same sense, of one thing. mill L1:2.3 CW7:28
And here is the distinction he makes between general and collective names: It is necessary to distinguish general from collective names. A general name is one which can be predicated of each individual of a multitude; a collective name cannot be predicated of each separately, but only of all taken together. mill L1:2.3 CW7:28
These two passages have favorable implications for Mill’s discussion of happiness in Utilitarianism. His distinctions between general and individual names, and between general and collective names, suggest that Mill would take the terms “general happiness” and “collective happiness” to name two entirely different concepts. Especially in the light of the distinction between general and collective names, it follows that the “general happiness” refers distributively to the personal happiness of each person, whereas the “collective happiness,” had Mill used such a term, would have referred collectively to the happiness of all persons taken together.
222
Chapter 5
The implication of this distinction for Mill’s discussion of happiness, and for his proof of the principle of utility, is that both are concerned with the “general happiness,” not with anything called the “collective happiness.”15 Reconsider the portion of his proof where Mill allegedly commits the fallacy of composition: No reason can be given why the general happiness is desirable, except that each person, so far as he believes it to be attainable, desires his own happiness. … each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons. mill U4:3 CW10:234; added emphasis
Note that Mill refers twice to the “general happiness,” and not once to a “collective happiness.” Given that he uses the term “general happiness” to refer distributively to the personal happiness of each person, he is not switching from distributive to collective predication of the attribute “desirability” of the subject “happiness” in moving from the desirability of each person’s happiness to the desirability of the general happiness.16 This alone does not fully absolve Mill of the fallacy of composition. The move from what is a good to each person to what is a good to the aggregate of all persons is still a candidate for the fallacy. It is not clear in Mill’s discussion of “names” in A System of Logic whether he considers “aggregate” to be a general 15
16
One exception is the following sentence in the second chapter of Utilitarianism: “The only self-renunciation which it [the utilitarian morality] applauds, is devotion to the happiness, or to some of the means of happiness, of others; either of mankind collectively, or of individuals within the limits imposed by the collective interests of mankind” (U2:17 CW10:218). Mill’s references here to the happiness “of mankind collectively” and to “the collective interests of mankind” contradict the position taken above that he is concerned strictly with the “general happiness,” never invoking or intending anything that might be called the “collective happiness.” While these are indeed counterexamples, they are also anomalies, perhaps instances of elegant variation, as the next paragraph and the one after that both revert to the “general good” (U2:18 CW10:218–219) and the “general interests of society” (U2:19 CW10:219). At any rate, stray references to the happiness “of mankind collectively” and to “the collective interests of mankind” in the second chapter of Utilitarianism do not overturn the position that, in the relevant portion of his proof in the third paragraph of the fourth chapter of Utilitarianism, Mill uses the term “general happiness” in a distributive sense to denote each person’s happiness. Spencer K. Wertz (1971, 422–423) also notes that Mill employs the adjective “general” for the distributive modification of its subject, thus using the expression “general happiness” as a technical term referring severally to each person’s happiness. This is one of Wertz’s central arguments in his attempt to absolve Mill’s proof of the alleged fallacy of composition.
Alleged Fallacy of Composition
223
name or a collective one.17 Therefore, it is not yet clear that he does not switch from distributive to collective predication of the attribute “happiness” relative to the subject “persons” in moving from what is a good to each person to what is a good to the aggregate of all persons. 5.3.3 Mill’s Conception of the Aggregate of All Persons Mill’s understanding of the aggregate of all persons is even less accessible than his understanding of the general happiness. Worse still is the fact that any misinterpretation of his conception of the general happiness is likely to influence the corresponding interpretation of his conception of the aggregate of all persons. Although the evidence adduced above from A System of Logic (L1:2.1–8 CW7:24–45), that Mill uses the term “general happiness” in a distributive sense, does not prove that he also uses the term “aggregate of all persons” in a distributive sense, it does provide some support for that conclusion. Common sense suggests that it would be strange to use the “general happiness” in a distributive sense but the “aggregate of all persons” in a collective sense, while claiming that the former is a good to the latter, having arrived at either construct through a generalization from the individual level. The alternative is especially unlikely in the light of the central place Mill reserves for the individual in his social and political philosophy. 17
Spencer K. Wertz (1971, 428–429) pursues the appeal to Mill’s discussion of “names” in A System of Logic further than I do in the main text above. He maintains that Mill’s discussion of “names” shows that he uses “the aggregate of all persons,” as well as “the general happiness,” in a distributive sense. The evidence he adduces in support of this claim is that Mill uses the definite article to indicate individual names and definite descriptions. Yet “the” is such a common word that it is difficult to distinguish cases where Mill might be using it to make a point, from those where he uses it simply to observe grammar and syntax. Just as important, Wertz’s evidence seems to support the opposite of the thesis he intends to establish. If Mill’s usage of the definite article corroborates Wertz’s interpretation that “the aggregate of all persons” is an individual name, then “the aggregate of all persons” is not a general name but a collective one. My disagreement with Wertz is that it is inconclusive in A System of Logic whether “the aggregate of all persons” is a general name, whereas the same work provides evidence that “the aggregate of all persons” is an individual name. Mill claims, for example, in distinguishing between general and individual names, that “ ‘the 76th regiment of foot in the British army,’ which is a collective name, is not a general but an individual name” (L1:2.3 CW7:28). On that basis, “the aggregate of all persons” is more like “the 76th regiment of foot in the British army” and less like “the general happiness.” Hence, I disagree with Wertz, not because I do not believe that Mill uses “the aggregate of all persons” in a distributive sense, which I do believe, but because I find that how Mill uses “the aggregate of all persons” is not conclusive in the passages Wertz cites from A System of Logic.
224
Chapter 5
The evidence itself happens to follow common sense. In addition to clarifying Mill’s conception of the general happiness, the passages quoted from Auguste Comte and Positivism (CW10:337) and On Liberty (CW18:261, 270) in the previous subsection (5.3.2) indicate that the most plausible interpretation of Mill’s conception of the aggregate of all persons is one that reflects his strong sense of individualism and his intense aversion to thinking of people as an impersonal collective. The theme of individualism, which cuts across many of Mill’s works and permeates his reasoning about society, points to a distributive interpretation of his understanding of the aggregate of all persons.18 The most straightforward evidence for this interpretation is, again, in A System of Logic (L6:6.2–7.1 CW8:877–879), where Mill identifies the only sense in which happiness is attributable to a group to be a type of reductive individualism wherein the happiness of the group is the happiness of the individuals in that group. It follows from there that the general happiness is indeed desirable for the correlative group, which happens to be the aggregate of all persons, if each person’s happiness is desirable for the person whose happiness it is.19 The point is not merely that interpreting Mill’s inference in the context of reductive individualism helps absolve his argument of the fallacy of
18
19
Another instance of Mill’s emphasis on the individual is a parenthetical reference in a letter to Thomas Carlyle (Letter 95: 12 January 1834), which indicates that Mill takes the good of the species to be the good of its several units. Here is what he says in the relevant portion of the letter: “Though I hold the good of the species (or rather of its several units) to be the ultimate end, (which is the alpha & omega of my utilitarianism) I believe with the fullest Belief that this end can in no other way be forwarded but by the means you speak of, namely by each taking for his exclusive aim the development of what is best in himself” (CW12:207–208). His parenthetical amendment here to “the good of the species,” fleshing it out as the good of “its several units,” provides additional support for the interpretation that, where he attributes the general happiness to the aggregate of all persons, he refers distributively to the personal happiness of each person. Mill’s letter to Carlyle is reproduced in its entirety in The Letters of John Stuart Mill (1910, vol. 1, 87–93) as well as in CW12:204–209. The question whether “all” or “most” or “many” or “some” or “few” of the members of a group must be happy in order for the group to be happy is an interesting one. Obviously, “all” is not necessary, “few” is not sufficient, and “some” and “many” are not specific enough to judge, while “most” is not much better in terms of clarity, as it can refer to anything between “half” and “all.” “Half-plus-one-person” and “fifty-one percent,” though not necessarily the same thing, are both open to slippery slope objections, whether or not such objections are worthwhile. While the question is interesting, however, a definitive answer is not necessary. The number of happy members needed to declare the group happy is beside the point. What is important is that the group is happy only insofar as its members are happy. The distinction is not a quantitative one but a qualitative one. Groups are not happy or sad; their members are.
Alleged Fallacy of Composition
225
composition but also that Mill actually subscribes to reductive individualism. His following words in A System of Logic provide some corroboration of his inclination and serve as a segue into further discussion: All phenomena of society are phenomena of human nature, generated by the action of outward circumstances upon masses of human beings: and if, therefore, the phenomena of human thought, feeling, and action, are subject to fixed laws, the phenomena of society cannot but conform to fixed laws, the consequence of the preceding. mill L6:6.2 CW8:877
The groundwork established in this passage is developed over the next few pages of the same work into a complete and compelling confirmation of Mill’s espousal of reductive individualism in his social outlook and sociological orientation: The laws of the phenomena of society are, and can be, nothing but the laws of the actions and passions of human beings united together in the social state. Men, however, in a state of society, are still men; their actions and passions are obedient to the laws of individual human nature. Men are not, when brought together, converted into another kind of substance, with different properties; as hydrogen and oxygen are different from water, or as hydrogen, oxygen, carbon, and azote, are different from nerves, muscles, and tendons. Human beings in society have no properties but those which are derived from, and may be resolved into, the laws of the nature of individual man. In social phenomena the Composition of Causes is the universal law. mill L6:7.1 CW8:879
Mill’s categorical endorsement of reductive individualism is hardly the standard paradigm in contemporary sociology. However, the goal here is not to vindicate Mill’s reductionism but to show that he uses the term “aggregate of all persons” in a distributive sense in his proof of the principle of utility. The more rigorous and more comprehensive we take Mill’s reductive individualism to be, the less viable his sociological perspective may appear from a current perspective. On the other hand, the more rigorous and more comprehensive his conviction is, the more likely it is that he is referring severally to individual persons where he takes the general happiness to be a good to the aggregate of all persons.20 20
Moreover, this suggests, with respect to Mill’s distinction between general and collective names in A System of Logic (L1:2.3 CW7:28), that the “aggregate of all persons” is indeed
226
Chapter 5
Regardless of the viability of Mill’s reductive individualism as a general characteristic of group theory, including the corresponding ontology, psychology, sociology, and morality, the position does seem to make sense with reference to the desirability of the general happiness for the aggregate of all persons. The reduction in question provides a rather plausible interpretation, arguably the most plausible interpretation conceivable, of what it means to attribute happiness to a group of people, even if it is not the best or only approach to analyzing other characteristics of groups. What else would the happiness of a group of people be, if not the happiness of its members, so that a group is happy only insofar as its members are happy? Happiness is a good to the aggregate of all persons in the same sense and for the same reason that it is a good to individual persons. This is not to deny that the very same reasoning may fail to account for various other aspects of the nature and dynamics of groups. From a psychological perspective, for example, herd mentality and crowd behavior are fundamentally different from personal inclinations and actions, at least in the sense that people tend to behave differently when acting as a member of a group than when acting alone. From a sociological perspective, the goals, norms, and standards of a group are typically different from the goals, norms, and standards of its individual members. From a moral perspective, collective responsibility transcends personal responsibility, making us collectively responsible, say, for the state of the environment, even if none of us alone can make a difference, and even if some of us do not even care one way or the other. Happiness, however, differs from these other considerations grounded in psychology, sociology, and morality. Happiness cannot plausibly be attributed to a group in any sense above and beyond the happiness of everyone in the group, whereas goals, norms, standards, inclinations, behavior patterns, and even duties and responsibilities make sense at the level of the group as well as at the level of the individual. The role of reductive individualism in Mill’s proof of the principle of utility can be illustrated more clearly through examples. Consider Mill’s argument in a general name and not a collective one. In this context, the phrase “aggregate of all persons” seems to be a general name for the class of all persons. The corroborating evidence for this is in the same passage of A System of Logic, where Mill explains the relationship between general names and classes as follows: “It is not unusual, by way of explaining what is meant by a general name, to say that it is the name of a class. But this, though a convenient mode of expression for some purposes, is objectionable as a definition, since it explains the clearer of two things by the more obscure. It would be more logical to reverse the proposition, and turn it into a definition of the word class: ‘A class is the indefinite multitude of individuals denoted by a general name’ ” (L1:2.3 CW7:28).
Alleged Fallacy of Composition
227
relation to two other arguments that are identical in form to the inference in which he allegedly commits the fallacy of composition (fc):
(FC1) No reason can be given why Professor Moriarty’s class is small, except that each student in Professor Moriarty’s class is small. (FC2) No reason can be given why Professor Moriarty’s class is failing Philosophy 500, except that each student in Professor Moriarty’s class is failing Philosophy 500. (FC3) “No reason can be given why the general happiness is desirable, except that each person, so far as he believes it to be attainable, desires his own happiness” (U4:3 CW10:234).
FC1 commits the fallacy of composition because the sense of size that is attributable to a class is different from the sense of size that is attributable to students. The size of the class is a function of the number of students, while the size of the students is a function of, let us say, their height (ignoring, for the sake of simplicity, other standards of size for human beings). Suppose that a class consisting of fewer than five students is a small class and that a student under five feet tall is a small student. Suppose further that Professor Moriarty’s class, a graduate seminar in the philosophy of mathematics, consists of three students each of whom is over six feet tall. This means that Moriarty’s class is small, while none of his students is small, thus constituting a counterexample to the argument that “no reason can be given why Professor Moriarty’s class is small, except that each student in Professor Moriarty’s class is small.” These students, tall though they may be, happen to be failing Moriarty’s course, which is why the class is failing the course, as stipulated in FC2. It is in this sense, and in this sense alone, that Moriarty’s class is failing Philosophy 500. No counterexample is available in this case, as it was in FC1, for the argument that “no reason can be given why Professor Moriarty’s class is failing Philosophy 500, except that each student in Professor Moriarty’s class is failing Philosophy 500.” The difference between the two arguments, of course, is not simply that the former admits of counterexamples whereas the latter does not. The availability of counterexamples merely points to the essential difference: The fallacious FC1 has the same term predicated of the subject distributively in the premises and collectively in the conclusion, whereas the valid FC2 contains no such switch from distributive to collective predication. FC1 depicts a case where the property in question can be attributed to a particular group of entities independently of each member of the group, while FC2 depicts a case where the relevant property cannot be attributed to the group independently of each
228
Chapter 5
member. There is a sense in which the class can be small without any of the students registered in that class being small, but the only sense in which the class can be failing is that the students registered in that class are failing. Any class might be large or small as a unitary whole, but no class will pass or fail as a unitary whole. This is because classes can be large or small both distributively and collectively, but they can pass or fail only distributively. Mill’s own inference, FC3, is closer in meaning to the valid FC2 than it is to the fallacious FC1. The aggregate of all persons can be happy only distributively and not collectively. Happiness, like passing and failing, is not a property attributable to a group in any sense above and beyond the members of the group. The only sense in which the general happiness is desirable for the aggregate of all persons is that the personal happiness of each individual is desirable for the person whose happiness it is and therefore that happiness in general is desirable for everyone in question. Conversely, then, it makes perfect sense to claim that “each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons” (U4:3 CW10:234). The foregoing account of the fallacy of composition, combined with the subsequent interpretations of the general happiness and the aggregate of all persons, suggests that Mill does not commit the fallacy of composition in his proof of the principle of utility. The pithy objection that “a sum of pleasures is not pleasure, any more than a sum of men is a man” (Mackenzie 1893, 113), while both clever and valid, does not capture the intent of Mill’s argument. 5.4
Critical Summary of the Response
The goal of this chapter has been to show that Mill does not commit the fallacy of composition in his proof of the principle of utility. The suspect argument can be paraphrased as follows: Each person’s happiness is desirable for (is a good to) the person whose happiness it is; therefore, the general happiness is desirable for (is a good to) the aggregate of all persons. The fallacy of composition, in turn, can be defined, and here has been defined, as a logically illicit inference from the attributes, properties, or qualities of the parts of a whole to the attributes, properties, or qualities of the corresponding whole, where the word that refers to the attribute, property, or quality in question is predicated of the subject distributively in the premises and collectively in the conclusion. The goal, then, has been to show that Mill’s argument paraphrased as above does not commit the fallacy defined as above. The defense, put simply, is that Mill’s argument does not commit the
Alleged Fallacy of Composition
229
fallacy of composition, because it does not switch from distributive to collective predication in the course of moving from the attributes of the parts of a whole to the attributes of the whole itself. The defense points to an interesting question concerning the connection between the textual foundation of the charge and the motivation of Mill’s critics: Why do so many critics accuse Mill of committing the fallacy of composition if all it takes to absolve him of the fallacy is to take a closer look at his argument? The defense here is, after all, based on nothing more than a careful inspection of the contents and context of the argument to see whether or not it really works. As simple a task as this is, and as compelling a conclusion as its proper performance brings, why would critics keep condemning the argument? The most obvious answer is that the suspect argument does fit the basic parts-whole form, whereupon most commentators evidently feel comfortable enough to invoke the fallacy straightaway, seeing no need to investigate further. Another reason, however, is that Mill’s own narrative is indicative of a belief that what is a good to one person actually is and certainly ought to be a good to others. Despite his explicit disclaimer in response to Henry Jones (Letter 1257: 13 June 1868; CW16:1414), Mill peppers the third chapter of Utilitarianism with signs of commitment to a harmonious sentiment of benevolence, camaraderie, and cooperation among the masses. The intensity of his conviction is perhaps most explicit in the following passage: But there is this basis of powerful natural sentiment; and this it is which, when once the general happiness is recognised as the ethical standard, will constitute the strength of the utilitarian morality. This firm foundation is that of the social feelings of mankind; the desire to be in unity with our fellow creatures, which is already a powerful principle in human nature, and happily one of those which tend to become stronger, even without express inculcation, from the influences of advancing civilization. The social state is at once so natural, so necessary, and so habitual to man, that, except in some unusual circumstances or by an effort of voluntary abstraction, he never conceives himself otherwise than as a member of a body; and this association is riveted more and more, as mankind are further removed from the state of savage independence. Any condition, therefore, which is essential to a state of society, becomes more and more an inseparable part of every person’s conception of the state of things which he is born into, and which is the destiny of a human being. Now, society between human beings, except in the relation of master and slave, is manifestly impossible on any other footing than that the
230
Chapter 5
interests of all are to be consulted. Society between equals can only exist on the understanding that the interests of all are to be regarded equally. mill U3:10 CW10:231
The passage as a whole indicates that Mill is optimistic about the natural progression of humankind toward the ideal society described. The last two sentences emphasize that some degree of approximation to the ideal society, whose foundation requires at least mutual respect for one another’s interests, is necessary for social justice in the real world. This is a recurring theme in Mill’s writings. The following passage in “Coleridge” (1840), for example, intimates that a strong feeling of common interest among the members of a community is an essential condition of the sociopolitical stability of that community: The third essential condition of stability in political society, is a strong and active principle of cohesion among the members of the same community or state. We need scarcely say that we do not mean nationality in the vulgar sense of the term; a senseless antipathy to foreigners; an indifference to the general welfare of the human race, or an unjust preference of the supposed interests of our own country; a cherishing of bad peculiarities because they are national; or a refusal to adopt what has been found good by other countries. We mean a principle of sympathy, not of hostility; of union, not of separation. We mean a feeling of common interest among those who live under the same government, and are contained within the same natural or historical boundaries. We mean, that one part of the community do not consider themselves as foreigners with regard to another part; that they set a value on their connexion; feel that they are one people, that their lot is cast together, that evil to any of their fellow-countrymen is evil to themselves; and do not desire selfishly to free themselves from their share of any common inconvenience by severing the connexion. mill CW10:134–135
Mill holds that one’s desire for one’s own happiness can coexist with one’s desire for everyone else’s happiness. He maintains that the interdependence of goods and persons is the mark of an ideal society, perhaps even an essential characteristic of it, and that such interconnections ought to be at least recognized and preferably accommodated in the sociopolitical setting prevailing in practice. Presupposing the interdependence of goods and persons generates embarrassing complications for the corresponding reconstruction of Mill’s argument, as demonstrated in and through the four scenarios in section 5.2,
Alleged Fallacy of Composition
231
but that presupposition can be discarded on the authority of the disclaimer in the Jones letter. As for Mill’s actual argument, a proper appraisal requires recognizing that his faith in humanity and his positive sentiments concerning the progressive actualization of social ideals, and thereby of an ideal society, constitute the somewhat speculative subject matter of the third chapter of Utilitarianism, not to mention that of several of his other works, none of which can legitimately implicate his entirely independent argument in the fourth chapter of Utilitarianism. Yet critics tend to assume, or can at least be tempted to assume, that the interdependence of goods and persons is implicit in Mill’s inference in the fourth chapter, because they know that Mill explicitly endorses that interdependence in the third chapter. To be sure, this is to be expected from a sequential reading of Utilitarianism, which contains nothing to discount the assumption. Mill’s critics, especially his early critics, can hardly be expected to account for a disclaimer in an obscure letter, which was not even publicly available until nearly half a century after the publication of Utilitarianism (first appearing anonymously in the 1910 collection of letters edited by Elliot, then by name in the 1972 installment of the Collected Works). On the other hand, it is no longer unreasonable to expect critics to take note of Mill’s warning in the Jones letter, and it was never unreasonable to expect anyone at all to study and analyze a purportedly suspicious and possibly fallacious argument before actually condemning it. At any rate, the connection between the textual foundation of the charges and the motivation of critics is that critics project Mill’s discussion of social harmony, together with his optimism regarding progress toward ideal society, into his proof of the principle of utility, where it does not belong. It is not necessary to agree with Mill, namely on the notion that each person’s happiness is desirable for every other person, in order to vindicate his suspect argument, which neither invokes nor implies nor requires that premise. Even the harshest critics can fairly accuse Mill at most of wrongly believing in the desirability of each person’s happiness for every other person, not of building his allegedly fallacious argument on that belief as its foundation. It is not just possible but arguably also reasonable to reject Mill’s conviction that, in an ideal society, people ought to find no conflict between their own interests and the interests of others, but such a disagreement in belief does not commit Mill to fallacious reasoning in the third paragraph of the fourth chapter of Utilitarianism.
c hapter 6
The Alleged Naturalistic Fallacy in Mill’s Proof This chapter examines George Edward Moore’s conception of the naturalistic fallacy, in an effort to identify, as precisely as possible, the mistake or mistakes he attributes to Mill, in accusing him of committing that fallacy in his proof of the principle of utility.1 The approach is limited in three respects. First, the focus is on Moore’s conception of the naturalistic fallacy, and on interpretations of his conception, but not on contemporary reformulations or reconceptions that go beyond Moore.2 Second, the aim is neither to show that the naturalistic fallacy is not really a mistake nor to challenge Moore’s understanding of what counts as natural or as good.3 The aim, rather, is to grant uncritically 1 The following is a sampling of the relevant literature: Gordon Park Baker and Peter Michael Stephan Hacker (1966); Thomas Baldwin (1990); Bernard H. Baumrin (1968); William H. Bruening (1971); Elmer H. Duncan (1970); James Fieser (1993); William Klaas Frankena (1939); David P. Gauthier (1967); George R. Geiger (1949); Richard Joyce (2001; 2006); Jerrold J. Katz (1990); Philip Kitcher (2011); Aurel Kolnai (1980); Casimir Lewy (1965); L. M. Loring (1967); Neil R. Luebke (1970); George Nakhnikian (1963); Arthur N. Prior (1949, 1–12, 95– 107); Alan James Ryan (1966); John R. Searle (1964; 1969); Mary Warnock (1978, 1–28); Carl P. Wellman (1961, 45–54); Darryl F. Wright (1994). 2 Katz (1990), for example, sets out to reformulate the naturalistic fallacy to avoid what he regards as shortcomings in Moore’s original conception. He is concerned more with linguistic and logical naturalism, specifically in opposition to both, than with ethical naturalism. He thus attacks Wittgenstein’s deflationary naturalism and Quine’s naturalized epistemology instead of rehashing Mill’s supposed mistakes. 3 Frankena (1939), among others, holds that the naturalistic fallacy is not a self-evident mistake even in the loosest sense of the word “fallacy.” He maintains that exposing the naturalistic fallacy in an argument does not automatically refute or invalidate that argument as if a logical fallacy had been discovered in it. He argues that an argument confirmed to commit the naturalistic fallacy must, in addition, be shown to be a bad argument specifically for that reason, and not simply declared to be so to validate that reason. He thus reduces Moore’s naturalistic fallacy to a type of disagreement between “intuitionists” and “definists (naturalistic or metaphysical)” (1939, 472). According to Frankena, the issue “is one of inspection or intuition, and concerns the awareness or discernment of qualities and relations” (1939, 475). Moore himself may not have come down on either side of this issue as described, because he regards himself as an intuitionist only in the sense of denying that propositions specifying what is intrinsically good are incapable of proof. See the preface to the first edition of Principia Ethica (1903): “I imply nothing whatever as to the manner or origin of our cognition of them. Still less do I imply (as most Intuitionists have done) that any proposition whatever is true, because we cognise it in a particular way or by the exercise of any particular faculty” (preface: paragraph 6). However, Moore seems to shift from his unusual construal of intuitionism to Frankena’s orthodox interpretation, where he affirms that “we are all aware of a
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_008
Alleged Naturalistic Fallacy
233
that the naturalistic fallacy is a mistake and to determine whether Mill indeed makes that mistake.4 Third, coverage does not include other problems Moore attributes to Mill’s proof. Moore is, after all, an accomplished analyst raising some objection or other to each step of the proof (1903, §§ 3:39–44, pp. 116–126, § 3:62, pp. 155–156), including objections from a purely logical point of view, as discussed in chapter 3 (section 3.8) of the present volume. The concern of this chapter, on the other hand, is specifically with the mistake he calls the naturalistic fallacy. 6.1
Moore’s Broad Construal of the Naturalistic Fallacy
Mill is not the only one Moore accuses of committing the naturalistic fallacy. There are countless others, an indefinite many, to go by Moore’s count. Yet among the numerous other alleged perpetrators of the fallacy, Mill receives a privileged position and special attention in the index to Principia Ethica, where he alone is honored with a cross-reference to the entry for the fallacy itself. In all fairness, though, Moore claims that the naturalistic fallacy is committed in almost every book on ethics (1903, § 1:12, p. 65). Many thinkers from Aristotle onward, including Bentham and Mill and excepting only Sidgwick, get their fair share of criticism.5 Limiting the coverage to Mill, as done in this chapter, both permits and requires a narrow focus on Moore’s otherwise broad discussion of the fallacy. Any attempt to determine whether Mill actually commits the naturalistic fallacy best begins by ascertaining what exactly Moore means by the fallacy. certain simple quality, which (and not anything else) is what we mainly mean by the term ‘good’ ” (§ 2:24.3, p. 90). 4 Duncan (1970) follows a similar approach to examine whether “anyone” has committed the naturalistic fallacy. He agrees with Moore that the fallacy is indeed a mistake, but he denies that any of the prime suspects (Mill, Perry, Sharp) is guilty of having actually committed it. 5 Moore identifies the following perpetrators of the naturalistic fallacy, listed here in alphabetical order, with section and page references corresponding to their appearance in Principia Ethica (1903): Aristotle (§ 5:106.2, p. 225); Bentham (§ 1:14.2–4, pp. 69–71); Hobbes (§ 3:59.1, p. 148); Kant (§ 4:67.1, p. 164, §§ 4:75–76, pp. 177–179); Leibniz (§ 4:73.2, p. 175); Mill (§ 1:14.2, p. 69, §§ 3:39–44, pp. 116–126, §§ 3:62–65, pp. 155–160); Rousseau (§ 2:27.2, pp. 93–94); Spencer (§§ 2:29–33, pp. 97–106); Spinoza (§ 4:67.1, p. 164); the Stoics in general (§ 4:67.1, p. 164). The list seems to be representative rather than exhaustive, given that hardly anyone could reasonably be exempted, with Moore himself singling out Sidgwick as the only person who, to Moore’s knowledge, “clearly recognised and stated” that “good” is indefinable (§ 1:14.1, p. 69) and who “clearly recognised that by ‘good’ we do mean something unanalysable” (§ 3:36.1, p. 111).
234
Chapter 6
What Moore means, on the other hand, is not entirely clear in Principia Ethica, our best source for figuring it out. Any attempt at clarification is bound to repeat the efforts of previous scholars. No appraisal, therefore, can afford to ignore them. This section proceeds with a brief exposition of three accounts which together cover the entire range of interpretations of Moore’s conception of the naturalistic fallacy in the secondary literature: Carl P. Wellman (1961); William Klaas Frankena (1939); Mary Warnock (1978, 1–28).6 These three commentators offer significantly different yet jointly consistent interpretations. The task is to show that they are indeed severally correct and mutually compatible. Moore’s conception of the naturalistic fallacy is almost as elusive as Mill’s conception of the principle of utility. In Moore’s case, however, the problem is not the absence of a definition. On the contrary, Moore ends up with a profusion of definitions and descriptions, which point to a variety of interpretations, some of which are closely related and some of which are decidedly different. His characterization can sometimes be so vague as to suggest that committing the naturalistic fallacy is as simple a matter as contradicting Bishop Butler’s maxim that “everything is what it is, and not another thing.”7 Then again, it also tends to be clear enough at times to suggest that committing the naturalistic fallacy is all about attempting specifically to define the word “good.”8 The goal here is to uncover his considered opinion. 6
7 8
The exclusive focus here on Wellman (1961), Frankena (1939), and Warnock (1978, 1–28) should not be taken to suggest that the other contributions noted in the beginning of the chapter are not adequate or important. The fact of the matter is that the relevant entries are simply too numerous for an incidental literature survey. These three happen to be both suitable and sufficient for illustrating Moore’s conception of the naturalistic fallacy. The rest may indeed be consulted for further study, typically with good results and often to great advantage. Among early commentators, Prior (1949, 1–12, 95–107) combines an exposition of the logic behind the naturalistic fallacy with a survey of its historical background. Among later commentators, Baldwin (1990) contributes a detailed and enlightening discussion of the fallacy to the volume on Moore in Routledge’s Arguments of the Philosophers series. Among more recent commentators, Joyce (2001; 2006) and Kitcher (2011) offer contemporary reassessments in the context of ethical naturalism and moral evolution (the evolution of morality and ethics). Joyce, in particular, expresses puzzlement regarding why ethicists “seem so fearfully mesmerized by it [the naturalistic fallacy]” (2001, 153), confessing, in fact, that he has “never understood why William Frankena’s sensible 1939 article—‘The Naturalistic Fallacy,’ Mind 48, pp. 464–77—did not put an end to the whole business” (2001, 153, n. 32). Moore is evidently impressed with Bishop Butler’s maxim that “everything is what it is, and not another thing,” as he quotes it opposite the title page of Principia Ethica (1903) as well as in the final chapter of the book (§ 6:123, p. 254). The definition attempt that concerns Moore (1903) is what he considers “the most important sense of ‘definition’ ”: “that in which a definition states what are the parts which invariably compose a certain whole” (§ 1:10.1, p. 61). This is the subject matter of the next section, which examines Moore’s “most important sense of ‘definition’ ” and his
Alleged Naturalistic Fallacy
235
Wellman (1961) provides the most thorough coverage of the vagaries of Moore’s conception of the naturalistic fallacy. He illustrates, with specific references to passages in Principia Ethica, that Moore conceives of the fallacy in terms of at least seven different mistakes he attributes to naturalists (1961, 45–47):9 (1) committing a logical error, probably the fallacy of four terms, in using a key word in both an ethical and nonethical sense in the same argument;10 (2) defining the indefinable; (3) identifying two distinct characteristics or confusing two words with different meanings; (4) reducing the a priori to the empirical; (5) reducing the synthetic to the analytic; (6) confusing the nonnatural with the natural; (7) reducing the ethical to the nonethical. Observing that Moore favors “defining the indefinable” (2) as the main sense of the fallacy (1961, 48), Wellman objects that the text itself points to “the reduction of the ethical to the nonethical” (7) as the definitive interpretation (1961, 49). The plausibility of any interpretation of Moore’s naturalistic fallacy seems to be a function of the level of generality or specificity at which the fallacy is examined. This does not mean that any interpretation or definition will do. On the contrary, there is a correct interpretation and definition at each level of generality and specificity. At the most general level, hence at a level general enough to cover every mistake Moore associates with the fallacy, the correct definition is along the lines of Wellman’s third formulation, that the fallacy is a matter of identifying two distinct characteristics or of confusing two words with different meanings. At the most specific level, exposing Moore’s ultimate motivation behind introducing the fallacy, and thereby bringing out his deepest concern in its application, the correct definition is that the fallacy is the kind of mistake described at the general level but made exclusively with respect to the concept good or the word “good.”11
9 10
11
contention that “good” is indefinable in that sense (§§ 1:5–12, pp. 57–69). But it is worth noting here that this “most important sense” is indeed the sense intended throughout this chapter in any and all references to defining “good,” or to the possibility of defining “good,” in the context of Mill’s proof or Moore’s objections. Nakhnikian (1963) closely follows Wellman’s (1961) lead in exemplification and classification as he enumerates six different mistakes corresponding to Moore’s (1903) various formulations of the naturalistic fallacy. Wellman’s (1961, 45) first interpretation of the naturalistic fallacy, namely as the logical fallacy of four terms, is precisely the fallacy discussed in chapters 3 and 4 of the present volume as Mill’s alleged equivocation on “desirable.” The coverage of Moore in chapter 3 is limited to his purely logical objections to Mill, independently of his additional objection that Mill commits the naturalistic fallacy as well. But Wellman is right that the logical error, too, is one of the ways in which Moore conceives of the naturalistic fallacy. Moore is not explicit as to whether the naturalistic fallacy concerns the concept good or the word “good” or both. While he does discuss the fallacy in both respects, it is not clear
236
Chapter 6
Between the most general and most specific levels of analysis, Wellman’s reading that the central interpretation of the naturalistic fallacy is the reduction of the ethical to the nonethical goes to the heart of the matter. On the other hand, it is just as reasonable to defend a broader construal of the fallacy to explain a greater number of instantiations of its practical application, or to defend a narrower construal to illustrate a particular application. Frankena’s (1939) account of the fallacy, for example, represents the highest level of generality at which the fallacy can sensibly be interpreted. At the other extreme, Warnock (1978, 1–28) presents a strong case for taking the fallacy at the specific level of the good. It will be helpful to consider each one in turn. Frankena approaches Moore’s conception of the naturalistic fallacy in terms of a violation of the “intuitionist bifurcation” that “ethical characteristics are different in kind from non-ethical ones,” which implies that “ethical characteristics are not definable in terms of non-ethical ones,” which, in turn, implies that “ethical propositions are not deducible from non-ethical ones” (1939, 467). To commit the naturalistic fallacy in terms of the intuitionist bifurcation is to contradict any one of these statements, which is always to contradict at least the first one. Frankena thus begins his analysis with the point Wellman later identifies as the central interpretation of the fallacy. He concludes that the naturalistic fallacy turns out to be a species of the “definist fallacy,” which he defines as “the process of confusing or identifying two properties, of defining one property by another, or of substituting one property for another” (1939, 471). He notes emphatically that the fallacy involved is “always simply that two properties are being treated as one, and it is irrelevant, if it be the case, that one of them is natural or non-ethical and the other non-natural or ethical” (Frankena 1939, 471). Warnock maintains, for her own part, that the mistake involved in the naturalistic fallacy is, in the final analysis, the attempt to define the word “good.” She adds that it is not essential to Moore’s conception of the fallacy that the attempted definition should reduce a nonnatural object to a natural one (1978, 3–4). To be sure, it is widely acknowledged that Moore’s construal of the fallacy is inextricably intertwined with his conception of the good.12 Warnock’s
12
whether he prefers one sense over the other. Any attempt to settle this question would have to deal with Moore’s tendency to blur the distinction between the concept and the word with his indiscriminate use of single quotation marks, which he tends to wrap haphazardly around the four letters with little regard for use versus mention. Lewy (1965) provides a useful study of an incomplete manuscript of what Moore intended as his preface to a largely revised second edition of Principia Ethica, which was eventually reprinted with only minor corrections in 1922. Lewy’s research is especially illuminating with respect to Moore’s later reflections on his original account of the naturalistic fallacy.
Alleged Naturalistic Fallacy
237
acknowledgment is even more specific in that her account is anchored to the definition of the word “good.” The fallacy involves not just any mistaken opinion concerning the good but specifically an attempt to define the word “good,” which is always a misdirected effort inasmuch as the definition is impossible, at least according to Moore. There is a common thread running through Wellman’s, Frankena’s, and Warnock’s interpretations: It is not essential to the naturalistic fallacy, despite its name, that it should involve a confusion of the natural and the nonnatural.13 Wellman’s account works well because his considered opinion on the central interpretation of the fallacy accounts for nearly every case with which Moore is concerned. The reduction of the ethical to the nonethical covers Moore’s objection both to the naturalists, who reduce the nonnatural to the natural, and to the metaphysicians, who make not that mistake but another.14
13
14
He finds that Moore ultimately holds the naturalistic fallacy to be a matter of contradicting the fact that good is not identical with any other property or quality, regardless of whether the target of the mistaken identification is analyzable and regardless of whether it is natural or metaphysical. He reports further that, according to Moore, although the naturalistic fallacy is typically committed in an inference confusing good with something else, the fallacy can also be attributed to a proposition affirming the mistaken identification and even to the confusion itself. Prime evidence for this is that Moore (1903) finds the metaphysical school of ethics guilty of committing the naturalistic fallacy without confusing the natural and the nonnatural (§§ 4:66–85, pp. 161–191). According to Moore, the metaphysicians commit the naturalistic fallacy in the sense that they relegate the nonnatural to a realm of supersensible existence where that which cannot be perceived to exist is somehow conceived to exist (§§ 4:66–67, pp. 161–166). This charge presupposes a distinction between the nonnatural and the supersensible, or else, Moore’s objection to the metaphysicians becomes a mere disagreement on what things exist. Moore himself does not distinguish between the words “nonnatural” and “supersensible,” but it may be useful to draw such a distinction, which may or may not be implicit in his discussion. Generally speaking, “nonnatural” and “supersensible” both refer to that which is not natural, does not have temporal dimensions, and cannot be perceived by the senses. However, in order to bring out the ontological contrast between Moore and the metaphysicians, it would be useful to draw an intensional distinction between “nonnatural” and “supersensible,” which are extensionally equivalent words: While “nonnatural” denotes that which lacks temporal dimensions and connotes that which does not exist at all, “supersensible” denotes that which lacks temporal dimensions but connotes that which somehow exists, though not in time. There is no telling whether Moore would have accepted this distinction, but it is clear that the alternative is to fudge on what it means to exist. Moore (1903) applauds the contribution that the metaphysicians make to epistemology in urging that knowledge is not limited to that which is natural, but he objects that they make an unjustified ontological commitment correlative to their epistemological venture. Specifically, he accuses the metaphysicians of making a groundless ontological commitment in asserting propositions such as “This is good.” The proper interpretation of
238
Chapter 6
Frankena’s account has much to recommend it because the high level of generality at which he examines the fallacy makes his interpretation the only one to accommodate Moore’s contention that “the naturalistic fallacy has been quite as commonly committed with regard to beauty as with regard to good: its use has introduced as many errors into Aesthetics as into Ethics” (1903, § 6:121.2, p. 249). Warnock’s characterization finds credence in Moore’s incessant concern with the good throughout Principia Ethica, and her interpretation becomes particularly compelling in Moore’s early emphasis on the definition of “good” (1903, §§ 1:1–23, pp. 53–88; especially §§ 1:3–13, pp. 55–69). These three interpretations, different though they may be, are mutually complementary rather than contradictory. None of them constitutes a better overall interpretation than the others, because each one satisfies a different level of generality with respect to Moore’s various discussions and applications of the fallacy in Principia Ethica. Each one thereby provides an accurate definition in an interpretation that is faithful to the text at its own level of generality. The aim of this chapter, however, is to remain faithful to the text specifically at the level corresponding to the allegation that Mill commits the naturalistic fallacy. Moore charges Mill, as well as other naturalistic hedonists, with committing the naturalistic fallacy either in holding directly that “good” means “pleasure” or in implying that “pleasure” is somehow involved in the definition of “good” (§ 3:36.1, p. 111). It is best, therefore, to proceed with Warnock’s account, in order to stay as close as possible to what is most relevant to the charges against Mill. The next section examines the relationship between the naturalistic fallacy and the good, while the one after that investigates the role of attempted definitions in that relationship.
such a proposition is not “This existing thing is good” (which comes with an unwarranted assumption of the existence of at least the referent) but “This thing would be good if it existed” or “It would be good for this thing to exist” (§§ 4:69–70, pp. 169–171). Moore objects, even more forcefully, that when the metaphysicians assert a proposition such as “This is good,” they mistakenly take themselves to be ontologically committed not only to the existence of that particular thing but also to the existence of the good that is attributed to that thing (§ 4:73, pp. 174–176). His emphasis with respect to the naturalistic fallacy is, again, on the good, which he thinks lacks temporal dimensions and therefore does not exist at all, whereas the metaphysicians hold that it exists somehow, though not in time.
Alleged Naturalistic Fallacy
6.2
239
The Naturalistic Fallacy Anchored to the Good
Warnock is right to emphasize both the good and its definition. Moore himself claims that “this question, how ‘good’ is to be defined, is the most fundamental question in all Ethics” (1903, § 1:5.2, p. 57). Although the text supports different interpretations of the naturalistic fallacy, including at least the seven formulations Wellman reports, the strongest evidence ties the fallacy to the good. Moore’s various discussions of the fallacy, together with his numerous practical applications, indeed imply all the interpretations from Wellman, Frankena, and Warnock. Yet the operative connection between his definition of the fallacy and his understanding of the good rises above all other interpretations. There is nothing implicit about that association. Moore nearly always defines the naturalistic fallacy in terms of the good. That is to say, as a general rule, wherever he explains what it is that he calls, has called, or will call the naturalistic fallacy, the explanation is in terms of the good. The instances in which the definition is in these terms to the exclusion of any other are both telling and compelling. Seven passages in particular, all from Principia Ethica, jointly corroborate this observation. Each passage serves as a definition of the naturalistic fallacy (nf) pertinent to the scope of the present focus on Mill. The passages in question are as follows, each one preceded by a formal definition reconstructed from Moore’s actual words in the corresponding passage. nf definition #1: The naturalistic fallacy is the mistake of identifying the unique and simple object of thought denoted by the word “good” with any other object of thought: ‘Good,’ then, denotes one unique simple object of thought among innumerable others; but this object has very commonly been identified with some other—a fallacy which may be called ‘the naturalistic fallacy.’ moore 1903, toc 1:B10, pp. 39–40
nf definition #2: The naturalistic fallacy is the mistake of thinking that to name properties other than goodness that belong to all things which are good is to define “good.” It is also the mistake of thinking that those other properties are identical with goodness. Both mistakes originate in a confusion of goodness with another property even where that other property is common and peculiar to all things that are in fact good: It may be true that all things which are good are also something else, just as it is true that all things which are yellow produce a certain kind
240
Chapter 6
of vibration in the light. And it is a fact, that Ethics aims at discovering what are those other properties belonging to all things which are good. But far too many philosophers have thought that when they named those other properties they were actually defining good; that these properties, in fact, were simply not ‘other,’ but absolutely and entirely the same with goodness. This view I propose to call the ‘naturalistic fallacy’ and of it I shall now endeavour to dispose. moore 1903, § 1:10.3, p. 62
nf definition #3: The naturalistic fallacy is the mistake of confusing good, which is not a natural object, with a natural object. This is a specific version of the general mistake, which may still be called the “naturalistic fallacy,” of confusing a nonnatural object with a natural object. This, in turn, is a specific version of the more general mistake, which may no longer properly be called the “naturalistic fallacy,” of confusing any two objects: When a man confuses two natural objects with one another, defining the one by the other, if for instance, he confuses himself, who is one natural object, with ‘pleased’ or with ‘pleasure’ which are others, then there is no reason to call the fallacy naturalistic. But if he confuses ‘good,’ which is not in the same sense a natural object, with any natural object whatever, then there is a reason for calling that a naturalistic fallacy; its being made with regard to ‘good’ marks it as something quite specific, and this specific mistake deserves a name because it is so common. moore 1903, § 1:12, p. 65
nf definition #4: The naturalistic fallacy is the mistake of affirming any proposition contrary to the contention that the word “good” means nothing but a certain simple quality of which we are all aware: I propose, therefore, to discuss certain theories of what is good in itself, which are based on the naturalistic fallacy, in the sense that the commission of this fallacy has been the main cause of their wide acceptance. The discussion will be designed both (1) further to illustrate the fact that the naturalistic fallacy is a fallacy, or, in other words, that we are all aware of a certain simple quality, which (and not anything else) is what we mainly mean by the term ‘good’; and (2) to shew that not one, but many different things, possess this property. moore 1903, § 2:24.3, p. 90
Alleged Naturalistic Fallacy
241
nf definition #5: The naturalistic fallacy is the mistake of identifying the simple notion intended by the word “good” with some other notion: In this chapter I have begun the criticism of certain ethical views, which seem to owe their influence mainly to the naturalistic fallacy—the fallacy which consists in identifying the simple notion which we mean by ‘good’ with some other notion. moore 1903, § 2:35, p. 109
nf definition #6: The naturalistic fallacy is the failure to distinguish the unique and indefinable quality invoked by the word “good”: Hedonism appears in the main to be a form of Naturalistic Ethics: in other words, that pleasure has been so generally held to be the sole good, is almost entirely due to the fact that it has seemed to be somehow involved in the definition of ‘good’—to be pointed out by the very meaning of the word. If this is so, then the prevalence of Hedonism has been mainly due to what I have called the naturalistic fallacy—the failure to distinguish clearly that unique and indefinable quality which we mean by good. moore 1903, § 3:36.1, p. 111
nf definition #7: The naturalistic fallacy is the mistake of thinking that the word “good” can be defined in terms of natural qualities: In this argument the naturalistic fallacy is plainly involved. That fallacy, I explained, consists in the contention that good means nothing but some simple or complex notion, that can be defined in terms of natural qualities. moore 1903, § 3:44.3, p. 125
The significance of these passages is that they are all we need to understand Moore’s conception of the naturalistic fallacy in the context of the present chapter. Even if the excerpts do not themselves constitute proper definitions, they are such explicit indicators of what Moore means by the “naturalistic fallacy” that a definition on his behalf can readily be constructed without twisting his words around through creative editorial input. Each of the seven passages portrays the naturalistic fallacy as a mistake made with respect to the concept good or the word “good.” The second, third, sixth, and seventh passages emphasize attempts to define the word “good.”
242
Chapter 6
The next step is to examine Moore’s contention that the word “good” cannot be defined, supported by his open-question argument in defense of that contention. According to Moore, “the most important sense of ‘definition’ is that in which a definition states what are the parts which invariably compose a certain whole” (1903, § 1:10.1, p. 61). He maintains that “good” cannot be defined in this sense because “it is simple” and thus “has no parts” (1903, § 1:10.1, p. 61). Of course, in denying that “good” can be defined, Moore is not claiming that dictionaries do not carry a corresponding entry for the word, or that they do but they ought not to, or that they do and they may but they never get it right. Moore’s question “What is good?” is not a request for a definition in the standard sense. It is an inquiry into “the nature of that object or idea” for which the word “good” is generally used to stand (Moore 1903, § 1:6.1, p. 58). He wants to know the property we attribute to things we think are good: What is it that we are saying about each of the various things we find good? When his inquiry is thus understood, Moore’s immediate answer to the question “What is good?” makes perfect sense: “Good is good, and that is the end of the matter” (1903, § 1:6.2, p. 58). Moore later elaborates, still without a definition, that good is a “simple, indefinable, unanalysable object of thought” (1903, § 1:15, p. 72). The open-question argument is Moore’s proof of the indefinability of “good” (1903, § 1:13, pp. 66–69). The argument is supposed to prove that “good” cannot be defined, because with respect to any proposed definition, it is always an open question whether the definiens denotes something that is good. What this means is that it is not merely redundant but meaningful and significant to ask of any proposed definition of “good” whether the referent corresponding to the definiens is good. For example, if the proposed definition states that “good” means “happiness,” it still makes sense to ask whether happiness is good, for this is a meaningful and significant inquiry seeking additional information. To object that the answer is affirmative, even if it is indeed conclusively affirmative, and to protest therefore that happiness is obviously good, misses the point of Moore’s argument. Moore does not claim or imply that the question (in this case, “Is happiness good?”) cannot be answered conclusively, be it in the affirmative or in the negative. His point is simply that the question can be asked in earnest for information that has not yet been provided. That is what makes it an open question whether the definiens in any proposed definition of “good” refers to something that is in fact good. Moore’s conception of the good, along with his contention that “good” cannot be defined, suggests that the naturalistic fallacy is the same kind of mistake as those made by the dialectical partners of Plato’s Socrates in confusing a given concept with any number of other things in the extension of the term denoting that concept. The difficulty is reminiscent of Plato’s discussions of “piety” in
Alleged Naturalistic Fallacy
243
Euthyphro, “virtue” in Meno, “courage” in Laches, and “friendship” in Lysis, to cite a few examples. The similarity is in the pedagogical orientation of Moore and Plato in discussing their respective topics and in enumerating the failed attempts of others to define the terms denoting the concepts under discussion. Even Moore’s refutation of what he regards as attempted definitions of “good” and his contention that “good” cannot be defined have their counterparts in Plato’s tendency, especially in the “early” dialogues, to use the Socratic elenchus to comb through various definitions terminating in a Socratic aporia.15 The aporia Moore introduces and concludes with in Principia Ethica is two- fold: (1) “good” cannot be defined (1903, § 1:13, pp. 66–69); (2) propositions purporting to identify that which is good in itself, that which is intrinsically good, or that which ought to exist for its own sake are incapable of proof or disproof (see the preface to the first edition of Principia Ethica). Moore takes the open- question argument to support both of these points. More accurately, he takes that argument to prove the indefinability of “good,” which, in turn, establishes the impossibility of proving or disproving propositions purporting to identify things that are intrinsically good.
15
The analogy between Moore and Plato can arguably be extended beyond the aporetic dialogues to include Plato’s so-called middle period, where the Socratic quest for definitions is supplemented, and sometimes supplanted, with a preoccupation with Forms. After all, Moore’s conception of good as a unique, simple, noncomposite, indefinable, and unanalyzable object of thought is not too far to reach from the Platonic Good. It will not do to press this point very far, though, for it is a juxtaposition of Moore’s moral epistemology and Plato’s moral ontology. Moore’s (1903) wholesale repudiation of attempts to ground an ethical system in a metaphysical platform, in contrast to how Plato’s ethics is anchored to and ruled by his ontology, may suggest that the analogy is a bit of a stretch (§§ 4:66–85, pp. 161–191; especially § 4:67, pp. 164–166). To be precise, Moore objects to grounding a doctrine of value in a metaphysical system, but he does not deny the connection between theories of moral obligation and metaphysical systems. At any rate, it is interesting that Moore refers to Plato in several places (§ 3:52.2–16, pp. 139–140, § 3:59.3, p. 150, § 4:66.2, p. 162, § 5:107, p. 227, § 6:120.2, p. 248) without once including him in the company of “metaphysical ethicists” who commit the naturalistic fallacy: for example, Kant (§ 4:67.1, p. 164, § 4:73.2, p. 175, §§ 4:75–76, pp. 177–179), Leibniz (§ 4:73.2, p. 175), and Spinoza (§ 4:67.1, p. 164). Moore is so uncomfortable with affirming that anything exists unless it has temporal dimensions (he does not require spatial dimensions) that he goes so far as to affirm that “goodness” does not “exist at all” (§ 4:66.2, p. 161). If it were not for Moore’s reluctance to countenance the supersensible, he might have embraced more of Plato explicitly. This is not intended in the loose sense that, if he were not British, he might have been Greek. I am not disappointed that Moore does not believe in unicorns or leprechauns, but I think he would have been more consistent to make room in his ontology for good, in addition to “what we mean by the adjective ‘good’ ” and “things or qualities which are good” (§ 4:66.2, p. 161).
244 6.3
Chapter 6
The Alleged Naturalistic Fallacy in Mill’s Proof
These two aspects of Moore’s aporia correspond to two charges at the intersection of his understanding of what it means to commit the naturalistic fallacy and his objections to Mill’s proof of the principle of utility.16 The first charge is in terms of the definition of words and the conflation of concepts: Moore holds that it is wrong to define the word “good” in the “most important sense” of “definition” (see section 6.2 above) “in which a definition states what are the parts which invariably compose a certain whole” (1903, § 1:10.1, p. 61), or to confuse and identify the concept (or property) good with some other concept (or property), whereas Mill supposedly defines “good” as “desired” and confuses and identifies something’s being good with its being desired. The second charge is in terms of proving propositions: It is wrong to try to prove or disprove what things are intrinsically good, says Moore, whereas Mill indeed attempts to prove that happiness is intrinsically good (and subsequently even that happiness is the only thing that is intrinsically good). The two charges are closely related. With respect to the first charge, Mill allegedly commits the naturalistic fallacy because he presupposes or implies that “good” means “desired,” at least insofar as he proceeds as if something’s being good were the same as its being desired. With respect to the second charge, Mill allegedly commits the same fallacy because, observing that people do desire happiness, he infers that happiness is intrinsically good. The charge in terms of definitions and concepts requires clarification. While Moore accuses Mill of committing the naturalistic fallacy in taking “good” to mean “desired,” he does not object to Mill’s using “good” and “desirable” synonymously, so long as the word “desirable” is understood in its proper meaning, 16
Exploring the intersection between Moore’s (1903) conception of the naturalistic fallacy and his objections to Mill’s proof may not seem like a straightforward approach to the topic, but the apparently direct route of examining what Moore says specifically about the naturalistic fallacy in Mill’s proof turns out to have even greater detours, obstacles, and vagaries. The gist of Moore’s accusation is that Mill commits the naturalistic fallacy in inferring the good or the desirable from the desired. He elucidates this accusation in three separate passages that can be paraphrased to his benefit as follows: Mill commits the naturalistic fallacy (1) in arguing that “good” means “desirable,” that what is desirable can be determined only by finding out what is desired, and, therefore, that “good” means “desired” (§ 3:40.3–4, p. 118); (2) in attempting to establish the identity of the good with the desired, by confusing the proper sense of “desirable,” in which it denotes that which it is good to desire, with the sense which it would bear if it denoted that which can be desired (§ 3:41.1, p. 119); (3) in using “the good” and “the desirable” synonymously, and taking “the desirable” to refer to what can be desired, which, in turn, is established by what is desired (§ 3:44.1–3, pp. 124–125).
Alleged Naturalistic Fallacy
245
expressed by any of the following phrases: “ought to be desired”; “deserves to be desired”; “good to desire” (1903, § 3:40.4, pp. 118–119). The list can reasonably be expanded, with Moore’s implicit endorsement, to include similar phrases, such as “worthy of desire” and “fit to be desired.” Moore allows switching between “good” and “desirable” (in its proper meaning as a value term) because that is not an attempt to define “good” in the “most important sense” of definition. He considers it otherwise acceptable to express the meaning of one word in other words, to adopt the standard sense in which people normally use a word, and thus to follow prevailing conventions as established by custom (1903, § 1:6.1, p. 58). Hence, Mill allegedly commits the naturalistic fallacy not because he takes “good” to mean “desirable” but because he then proceeds to treat “desirable” as if it meant “desired” and thereby conflates the good and the desired (1903, § 3:40.3–4, pp. 118–119, § 3:44.3, pp. 124–125). Moore evidently believes that Mill affirms a logical or ontological relationship between desires and desirability, in addition to an evidentiary relationship of an epistemological and psychological nature, in the first part of the proof, where he appeals to the observation that happiness is desired, as the only evidence available that happiness is desirable. This deviant procedure seems to be what Moore has in mind where he summarily charges Mill with committing the naturalistic fallacy in defining “good” as “desired” and in confusing and identifying something’s being good with its being desired (1903, § 3:40.4, p. 119). Moore subsequently unpacks the charge to reveal an even greater number of errors and confusions than initially indicated (1903, § 3:40.4, p. 119). This longer version includes an admixture of the notion expressed by the phrases “can be desired” and “able to be desired” as an additional source of confusion in Mill’s misidentification of the good with the desired. The long version likewise recognizes no fault with Mill’s usage of “good” and “desirable” as synonyms but alleges that he mistakenly treats “desirable” not only as if it meant “desired” but also as if it meant “can be desired” or “able to be desired.” This allegation is directly related to the discussion of Moore in chapter 3 (section 3.8) of the present volume, specifically to the inconsistency identified there in Moore’s interpretation of Mill as taking “desirable” in two different mistaken senses before drawing an illicit conclusion in terms of the proper meaning of the word and thereby equivocating on what it means for anything to be desirable. The goal in the relevant portion of that chapter was to describe the logical fallacy Moore finds in Mill’s move from the desired to the desirable. The fallacy in question was an equivocation on “desirable” in the direction of “can be desired,” but the first block quotation there from Principia Ethica (1903, § 3:40.4, pp. 118–119) exposed the alternative that the equivocation might instead be in the direction of “desired.” While that is indeed an inconsistency
246
Chapter 6
in Moore’s critique, it is also a step in his argument. His main objection from the perspective of the naturalistic fallacy is the misidentification of the desired and the desirable, while the matter of equivocation on “desirable” in the direction of “can be desired” constitutes an intermediate step in his critique of that part of the proof. Yet it was premature at that point, that is, in chapter 3, with the naturalistic fallacy yet to be covered, to illustrate the precise position of the inconsistency in the overall structure of Moore’s critique. The place to do that is here. Here, then, is the anatomy of the discrepancy between Moore’s two accusations. Recall the generic form of the argument introduced at the beginning of chapter 3: Whatever is desired is desirable; happiness is desired; therefore, happiness is desirable. Moore correctly recognizes that Mill is trying to establish the conclusion that happiness is desirable, in the proper meaning of the word “desirable,” expressed by phrases such as “ought to be desired,” “deserves to be desired,” “good to desire,” “worthy of desire,” “fit to be desired,” and so on (1903, § 3:40.4, pp. 118–119). Suppose that “desirable” stands for “worthy of desire.” The conclusion of the generic argument, then, is that happiness is worthy of desire. Given this fixed conclusion, Mill might equivocate on “desirable” in two different ways. On the one hand, if he takes “desirable” to mean “can be desired” in the major premise, the generic argument becomes: Whatever is desired can be desired; happiness is desired; therefore, happiness is worthy of desire. In this case, the argument equivocates on “desirable” by taking it to mean “can be desired” in the major premise and “worthy of desire” in the conclusion.17 On the other hand, if he takes “desirable” to mean “desired” in the major premise, the generic argument becomes: Whatever is desired is desired; happiness is desired; therefore, happiness is worthy of desire. In this case, the argument equivocates on “desirable” by taking it to mean “desired” in the major premise and “worthy of desire” in the conclusion. The problem is that Moore’s criticism of Mill’s move from desires to desirability implies two different interpretations of the language in the same argument, alleging first that Mill equivocates on “desirable” because he takes it to mean both “can be desired” and “worthy of desire,” and second that he equivocates on “desirable” because he takes it to mean both “desired” and “worthy of desire.” Each of Moore’s objections is prima facie reasonable independently of 17
A complementary alternative in terms of Wellman’s (1961, 45) first entry in his list of Moore’s formulations of the naturalistic fallacy would be to read the argument as committing the fallacy of four terms (see section 6.1 above). The subject term in this alternative is “happiness,” and the middle term “desired,” while the predicate term, “desirable,” becomes “can be desired” in the major premise and “worthy of desire” in the conclusion.
Alleged Naturalistic Fallacy
247
the other, but they are grounded in mutually exclusive (but not jointly exhaustive) interpretations of the same argument. While Moore stands behind both allegations, it is difficult to see how they could possibly apply simultaneously to a single argument. If Mill were to take “desirable” to mean “desired,” “can be desired,” and “worthy of desire,” all in the course of a single argument, he would then be guilty of equivocating on “desirable” twice in the same argument. That would be a fascinating logical error, which might perhaps be called “the fallacy of five terms.” The inconsistency here is grounded in Moore’s projection of his own concern with lexical meaning and conceptual analysis into Mill’s proof of the principle of utility. His own methodological habits thus inspire Moore to evaluate Mill’s proof as a series of misdirected attempts at establishing the synonymy of certain words and the identity of the concepts denoted by those words. The inconsistency surfaces where Moore saddles Mill with the naturalistic fallacy while also attempting to find fault with his analogy between visibility, audibility, and desirability (Moore 1903, § 3:40.4, pp. 118–119). Accusing Mill of committing the naturalistic fallacy in proceeding as if “desirable” meant “desired,” Moore probably finds it incredible that anyone with any sense could actually think that “desirable” meant “desired.” He therefore capitalizes on Mill’s statements concerning that which is visible, audible, and desirable, respectively, to suggest that Mill is misled by his own statements into the confusion that all words ending in “able” or “ible” concern capacity or ability, and thereby into thinking that “desirable” means “can be desired” or “able to be desired.” It presumably does not seem as incredible for someone to be misled, specifically by analogic conditioning, into taking “desirable” to mean “can be desired” as it does for anyone to think that “desirable” means “desired.” This, of course, is a misunderstanding of Mill’s analogy, the correct interpretation of which is in chapter 4 of the present volume, addressing the alleged fallacy of equivocation. Nevertheless, Moore’s understanding, while incorrect, points to an explanation of the inconsistency between the two mistaken senses of “desirable” he attributes to Mill. He interprets Mill as committing the naturalistic fallacy in grounding the good in desires, or rather in grounding the sole evidence for the good in desires, which Moore takes as a commitment to the equivalence of the good and the desired, which is the sense in which he ascribes to Mill the position that “good” means “desired.” He also interprets Mill as mistakenly taking “desirable” to mean “can be desired” and confusing desirability with being able to be desired. Mill himself claims neither that “good” means “desired” nor that being good is the same as being desired. Nor does he appear to be confused in the least about desirability. The best we can make of the naturalistic fallacy, while at the same time accommodating the
248
Chapter 6
equivocation Moore detects between that which is worthy of desire and that which is able to be desired, is to read the corresponding fault or difficulty as an implicit commitment by Mill, through certain ramifications of his reasoning in the proof, to the position that “good” means “desired” and that being good is identical with being desired. The root of the problem in Moore’s evaluation of Mill’s proof is that Moore is prejudiced in his interpretation of Mill’s sentences containing key terms such as “good,” “desirable,” “desired,” “happiness,” and “pleasure.” He carries on as if Mill were using the present tense of the verb “to be” either to introduce the definiens of a definition or to establish an identity relationship between concepts denoted by the subject and the predicate. If Mill were to write “happiness is good,” for example, Moore would be convinced that Mill intended, in utter confusion, either that “ ‘happiness’ means ‘good’ ” or that “happiness and good are identical.” This is both unfair and inconsistent in the light of Moore’s own acknowledgment of the validity of such assertions: No difficulty need be found in my saying that ‘pleasure is good’ and yet not meaning that ‘pleasure’ is the same thing as ‘good,’ that pleasure means good, and that good means pleasure. moore 1903, § 1:12, p. 65
Contradicting his own provision, Moore finds precisely this difficulty, which he otherwise says need not be found, in Mill’s claim that happiness is good. He does not allow for the possibility that Mill might be claiming that happiness is good, without meaning that happiness is the same thing as good. Ironically, Moore exposes the following line of response, which Mill himself would have been eager to follow, had he been confronted with the difficulty attributed to his argument: Why, if good is good and indefinable, should I be held to deny that pleasure is good? Is there any difficulty in holding both to be true at once? On the contrary, there is no meaning in saying that pleasure is good, unless good is something different from pleasure. moore 1903, § 1:12, p. 66
With respect to Moore’s first charge, identified at the beginning of this section in terms of the synonymy of words and the conflation of concepts, Mill clearly does not commit the naturalistic fallacy. He neither presupposes nor implies that “good” and “desired” and “happiness” are synonymous in Moore’s “most important sense” of definition. Nor does he otherwise attempt to establish that
Alleged Naturalistic Fallacy
249
good and being desired and happiness are identical concepts. Moore admittedly lays out reasonable rules of interpretation concerning the synonymy of words and the conflation of concepts, but even the strictest enforcement of Moore’s rules with respect to the least charitable interpretation of Mill’s reasoning does not show that Mill commits the naturalistic fallacy in this sense, barring a gross distortion of the relevant portion of Mill’s argument in the proof. As for Moore’s second charge, concerning the proof of propositions identifying things that are intrinsically good, Moore and Mill are actually in agreement at a fundamental level with respect to ethical justification and methodology, but Moore is not willing to take the agreement seriously. Moore claims that ethical propositions purporting to identify that which is intrinsically good (that which is good in itself, that which ought to exist for its own sake, etc.) are incapable of proof or disproof (see the preface to the first edition of Principia Ethica). This is exactly what Mill claims emphatically and repeatedly in his warnings that such propositions are not capable of direct proof, or proof in the ordinary sense (U1:5–6 CW10:207–208; U4:1 CW10:234; U4:9 CW10:237). Moore acknowledges Mill’s warnings but accuses him of merely pretending not to offer a direct proof (1903, § 3:45.1, p. 126). Immediately after this accusation, Moore quotes Mill’s statement that “considerations may be presented capable of determining the intellect either to give or withhold its assent to the doctrine” (U1:5 CW10:208; Moore 1903, § 3:45.1, p. 126). The quotation comes at the turning point of Moore’s examination of hedonism, specifically where he moves from Mill to Sidgwick in his analysis. At that point, in direct reference to the paradigmatic “considerations” invoked as a benchmark by Mill, Moore asserts that “it is such considerations that Professor Sidgwick presents, and such also that I shall try to present for the opposite view” (1903, § 3:45.1, p. 126). That is to say, Moore and Sidgwick, both being aware that direct proof of what is intrinsically good is impossible, resolve to determine each other’s intellect and, one hopes, the reader’s intellect as well. Mill, in contrast, presumably ignores his own warnings, vitiates his proposed methodology, and proceeds to deduce that which is intrinsically good directly from premises that are patently analytic, most of them being definitions, tautologies, and identity statements. Moore borrows Mill’s methodology not only to describe the difference and disagreement between Sidgwick and himself but also to justify his own positive account in the last chapter of Principia Ethica, where he proposes to identify what things are intrinsically good. He claims that, in order to determine what things have intrinsic value, “it is necessary to consider what things are such that, if they existed by themselves, in absolute isolation, we should yet judge their existence to be good” (Moore 1903, § 6:112, p. 236). He reports, as
250
Chapter 6
a result of his thought experiment, that what is intrinsically good is a state of consciousness corresponding to the pleasures of social intercourse or the enjoyment of beautiful objects (Moore 1903, § 6:113.1, p. 237). His justification is as follows: No one, probably, who has asked himself the question, has ever doubted that personal affection and the appreciation of what is beautiful in Art or Nature, are good in themselves; nor, if we consider strictly what things are worth having purely for their own sakes, does it appear probable that any one will think that anything else has nearly so great a value as the things which are included under these two heads. moore 1903, § 6:113.1, p. 237
Moore’s appeal to the fact, or perhaps the mere likelihood (cf. “probably” and “probable”), that no one doubts the intrinsic value of personal affection and aesthetic enjoyment, shares the methodological basis of Mill’s claim, that nothing can convince people that happiness is desirable, unless they themselves find it desirable (U4:3 CW10:234). Moore’s thought experiment obviously makes his argument different from Mill’s argument. An important similarity, however, is that the same basic strategy shapes their individual methodologies as both philosophers present arguments in support of what they think is intrinsically good. Fully agreeing that direct proof is not possible, they both engage in some other sort of proof. Moore, of course, denies that he is offering any proof at all, because propositions of the kind under consideration are not susceptible of proof or disproof, whereas he accuses Mill of offering some sort of proof if not direct proof. Yet Moore’s characterization of Mill’s proof applies equally well to his own positive arguments. It is one thing to claim that certain propositions are not capable of proof, entirely another to refrain from trying to prove them (1903, § 3:45.1, p. 126). Moore ends up proving his own thesis no less than Mill proves his. Proving a proposition versus avoiding discussion of it are not the only relevant alternatives. One might easily convince oneself, and perhaps even others, that a certain proposition is true, without actually proving that proposition. I might insist, for example, that Ingrid Bergman is God, because Lauren Bacall told me so in a dream, or better yet, in a vision, but neither appeal qualifies as proof, not as a very good one anyway. On the other hand, there is hardly any difference between proving a proposition and producing in support of it a cogent and relevant argument, deductive or inductive, that violates none of the generally accepted rules of good reasoning. Mill and Moore, however, do see a difference, at least under certain circumstances. They both deny that the
Alleged Naturalistic Fallacy
251
truth of a proposition purporting to identify that which is intrinsically good can be deduced from the truth of any other proposition, or set of propositions, even if the other propositions fit together as a cogent and relevant argument that violates none of the generally accepted rules of good reasoning. This, however, is too specific a position to demarcate “proof” from “nonproof” in general. Mill’s crude distinction between “some sort of proof” and “direct proof” better reflects the intention behind the fundamental methodological position on which the two philosophers agree. It is in this sense that both Mill’s and Moore’s arguments qualify as proofs. Neither one offers direct proof for what is intrinsically good, but each one does offer some sort of proof. Moore’s disavowal of any proof whatsoever turns out to be an exaggerated version of Mill’s disavowal of direct proof. What Moore does in the last chapter of Principia Ethica is not methodologically different in any relevant respect from what Mill does in the fourth chapter of Utilitarianism. But Moore would sooner admit that he himself inadvertently commits the naturalistic fallacy than that Mill does not commit it at all. While it is entirely possible that both philosophers commit the naturalistic fallacy in just the way described, the fallacy then becomes a mark of cogent argumentation in ethical theory rather than a mistake to be avoided in rational discourse. The truth of the matter is that Mill is not guilty of the naturalistic fallacy in either of the two senses at the intersection of Moore’s understanding of the naturalistic fallacy and his objections to Mill’s proof of the principle of utility. First, from the perspective of the synonymy of words and the confusion or conflation of concepts, Mill would be guilty of the naturalistic fallacy if he were to define “good” as “desired” or as “happiness,” and likewise, if he were to confuse something’s being good with its being desired or with happiness. But he does not commit the fallacy in that sense, because he holds neither that “good” means “desired” or “happiness” nor that something’s being good is the same as its being desired or its producing or promoting happiness. Second, with respect to proving propositions purporting to identify that which is intrinsically good, the naturalistic fallacy is committed where one offers a direct proof of such propositions, but it is otherwise acceptable to present cogent and relevant arguments that violate none of the generally accepted rules of good reasoning. Mill does not commit the naturalistic fallacy in that sense either, because he does not offer a direct proof for his contention that happiness is intrinsically good. This conclusion does not address the question whether Mill might reasonably be held to commit the naturalistic fallacy in a modified version of the first perspective. Even if he does not argue that “good” means “desired,” or confuse something’s being good or desirable with its being desired, he clearly asserts
252
Chapter 6
that desires are evidence of desirability, and, in fact, that they are the only evidence available for desirability. Might he, then, be committing the naturalistic fallacy in a more fundamental sense, namely in the process of grounding value in natural phenomena? Yes, he might be, but any opposition to his doing so would then have to be classified as a mere disagreement, for it would certainly not qualify as the detection of an actual mistake. This chapter has not covered weaker interpretations of that sort, because its explicit aim from the outset has been to grant uncritically that the naturalistic fallacy is a mistake and to show that Mill does not make that mistake in his proof. A more general reformulation of the fallacy as the tendency to ground value in natural phenomena is not a mistake in any of the senses that the more relevant formulations of the fallacy are. Then again, this chapter has also not been concerned with a vindication of ethical naturalism. One may still reasonably ask whether Mill is really able to get anywhere with desires as the sole evidence of desirability. But that question is no longer specifically about the naturalistic fallacy. The answer, or one answer, at any rate, is in chapter 4 of the present volume, which explains and defends the way Mill infers the desirability of happiness from the observation that people do desire happiness.
pa rt 3 Mill’s Proof of the Principle of Utility: Reconstruction and Implications
∵
c hapter 7
Reconstruction of Mill’s Proof of the Principle of Utility This chapter is the culmination of the foregoing defense of various aspects of Mill’s proof of the principle of utility against the traditional objections examined up to this point. It elucidates the proof in its entirety, specifically from a constructive and holistic perspective, in contrast to the defensive focus and partial coverage of the previous chapters. The aim is to show how the self-contained parts and stages of the proof, discussed in the strictly defensive chapters, fit together as a coherent interpretation of the initiative as a whole. Mill’s explicit purpose in the fourth chapter of Utilitarianism, and hence his ultimate objective in his proof of the principle of utility, is to establish the “utilitarian doctrine” that “happiness is desirable, and the only thing desirable, as an end; all other things being only desirable as means to that end” (U4:2 CW10:234). His attempt to “prove” this “doctrine” is a matter of establishing two distinct theses constituting two separate parts of the proof. The first thesis is that happiness is desirable as an end in itself. The second thesis, building on the first, is that nothing other than happiness is desirable as an end in itself. The net result constituting the overall conclusion of the proof is that happiness is the only thing that is desirable as an end in itself. Here is a map of the fourth chapter of Utilitarianism identifying precisely where each paragraph belongs in Mill’s proof of the principle of utility:
1 2 3 4 5 6 7 8 9 10 11 12
Paragraphs of Chapter 4 of Utilitarianism in Relation to the Proof Discussion of Methodology Thesis: Happiness is the only thing desirable as an end in itself. Part 1: Thesis: Happiness is desirable as an end in itself. Part 2: Thesis: Nothing else is desirable as an end in itself. continues continues continues continues continues concludes Objection Considered: Desire vs. Will Closing Remarks
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_009
256
Chapter 7
This is the same skeletal outline provided toward the end of chapter 2 of the present volume (see section 2.6). As indicated in that chapter, the first two paragraphs of the fourth chapter of Utilitarianism consist of introductory remarks that are preliminary to the proof, which actually begins with the third paragraph and ends with the tenth. The first part of the proof is contained entirely within the third paragraph. The second part spans paragraphs four through ten. The remainder of this chapter is organized in two sections corresponding to the two parts of Mill’s proof. The first section examines his attempt to prove that happiness is desirable as an end in itself. The second section takes up his attempt to prove that nothing other than happiness is desirable as an end in itself. 7.1
First Part of the Proof
The first part of the proof, especially insofar as substance is concerned, has already been covered in the course of discussion throughout the preceding chapters concerning the historical reception of the proof in the philosophical community and the associated allegations concerning logical transgressions. To be more specific, complete albeit piecemeal coverage of the first part of the proof is available in the foregoing examination of two aspects of it attracting the most attention in the secondary literature. The first is Mill’s conviction that desires are evidence of desirability and his expression of that conviction in methodological terms: “The sole evidence it is possible to produce that anything is desirable, is that people do actually desire it” (U4:3 CW10:234). The second is his conviction that the general happiness is desirable for, or is a good to, the aggregate of all persons and his articulation of that conviction as the conclusion of the argument that “each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons” (U4:3 CW10:234). The first focus of attention was the subject matter of chapter 4, which explained why Mill’s appeal to desires as evidence of desirability, coupled with his distinction between means and ends in that evidentiary relationship, constitutes a legitimate and successful approach to establishing the conclusion of the first part of the proof, namely that happiness is desirable as an end in itself. Demonstrating on these grounds that Mill’s reasoning does not involve the traditionally and frequently alleged equivocation on desirability, that chapter was later complemented by chapter 6, which showed, based on the same considerations, that Mill’s argument in the relevant portion of
Reconstruction of Mill’s Proof
257
the proof does not commit the naturalistic fallacy either. The second focus of attention was the subject matter of chapter 5, which absolved Mill of the fallacy of composition in his inference from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons, though it did not discuss at length the role of his emphasis on the general happiness in the overall structure of the proof. It is in this sense, then, that the first part of the proof has already been covered to a large extent, rather rigorously from a defensive perspective but also with significant constructive development, in the chapters on traditional objections to the proof. Yet the presentation so far has proceeded step by step, thereby lacking a proper or natural sense of unity. What remains to be shown is how the two aspects of the first part of the proof work together to establish the conclusion of that part of the proof: that happiness is desirable as an end in itself. It will be useful to begin with a summary of the methodological role of Mill’s conviction that “the sole evidence it is possible to produce that anything is desirable, is that people do actually desire it” (U4:3 CW10:234). As discussed in chapter 4, the gist of the methodology guiding the entire proof is to submit the observation that everyone desires happiness for the sake of happiness itself, and not for the sake of any other end, as empirical evidence sufficient for epistemic justification of the claim that happiness is desirable as an end in itself. The appeal to desires as evidence of desirability then becomes sufficient evidence of desirability in the distinction between means and ends. Desires for things as means are subject to revision with respect to the ends for which such things are desired, whereby they point at best to the instrumental value of the things in question. Desires for things as ends in themselves, on the other hand, are not subject to such revision, thereby pointing to the intrinsic value of those things. A subtle but critical clarification is in order. The operative distinction between means and ends in the evidentiary relationship between desires and desirability, and the only logically meaningful distinction to draw in that regard, is not simply a distinction between desires for things as means and desires for them as ends, but a deeper one between desires for things as means and desires for them as ultimate ends. The point of this “deeper distinction” is to recognize that the very same thing can be desired both as a means and as an end in two different relationships or from two different perspectives. It is only ultimate ends that represent the termination of any chain of means-and-ends relationships. To be perfectly clear, and in the interest of avoiding any misunderstanding in the course of discussion, note that the
258
Chapter 7
expression “desires for ultimate ends” is equivalent both to “desires for things as ends in themselves (and not as means to any other end)” and to “desires for things for their own sakes (and not for the sake of any other end).” Either one of the latter two formulations should be understood as spelling out the former. The mere fact that happiness is desired does not prove that happiness is desirable, but the fact that happiness is desired for its own sake (and not for the sake of any other end) does indeed point to the intrinsic value of happiness. With the distinction thus clarified, the first part of Mill’s proof of the principle of utility can be outlined as follows:
Outline of the First Part of the Proof (P1) The only evidence available that anything is desirable is that people do actually desire it. (P2) The only evidence available that anything is desirable as an end in itself is that people desire that thing for its own sake and not for the sake of any other end to which it might be a means. (P3) The only evidence available that happiness is desirable as an end in itself is that people desire happiness for its own sake and not as a means to something else. (P4) Everyone desires happiness, and everyone desires happiness for its own sake, not as a means to something else. (C) Happiness is desirable as an end in itself.
Note a difference between the apparent relevance of the two aspects of the proof identified above as a focus of critical attention. The first focus of attention, Mill’s appeal to the evidentiary relationship between desires and desirability, coupled with his distinction between means and ends, makes a clear contribution to the first part of the proof, while the second focus of attention, Mill’s emphasis on the general happiness, makes no comparable contribution to that part of the proof. The outline above seems to suggest that Mill’s appeal to desires as evidence of desirability, if successful, suffices to establish the conclusion of the first part of the proof, without any need of support from his apparently compositional inference concerning the general happiness, whether or not that argument is successful as well. The next task toward a complete exposition of the first part of the proof is to illuminate the role of Mill’s emphasis on the general happiness in the overall structure of the proof. Mill’s inference from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the
Reconstruction of Mill’s Proof
259
aggregate of all persons plays an explanatory role that is largely external to the logic of the proof. Yet it is nevertheless crucial for an accurate understanding of the proof. The role of the apparently compositional inference concerning the general happiness is to establish that everyone’s happiness is equally desirable because everyone desires happiness, and thereby to distinguish Mill’s universalistic ethical hedonism from both egoistic ethical hedonism and altruistic ethical hedonism. Although the premise that everyone’s happiness is equally desirable makes no logical contribution toward establishing the thesis that happiness is desirable as an end in itself, it makes a substantial contribution toward understanding the overall conclusion of the proof of the principle of utility as a theory of value, namely as the theory that happiness is the only thing desirable as an end in itself. It not only clarifies whose happiness it is that is desirable but also makes that clarification an integral part of the proof as opposed to leaving it as an afterthought that may or may not be consistent with the proof itself. The remainder of this section brings out the logical and moral significance of the argument in which Mill allegedly commits the fallacy of composition. Its logical significance concerns the role of the apparently compositional inference about the general happiness in the overall structure of the proof. Its moral significance concerns the practical implications of Mill’s emphasis on the general happiness for his ethical theory. 7.1.1 The Logical Role of Emphasis on the General Happiness The proper defense of Mill’s apparently compositional inference, as demonstrated in chapter 5 of the present volume, is based on the fact that Mill does not switch from distributive to collective predication, either with respect to happiness or with respect to persons, in the course of the corresponding argument. Let us contemplate a potential problem: Doesn’t the interpretation that the reference to key words remains distributive throughout the inference undermine the point of drawing that inference in the first place? Recall the suspect argument: Each person’s happiness is desirable for the person whose happiness it is; therefore, the general happiness is desirable for the aggregate of all persons. The defensive reading that the reference to happiness and persons remains distributive throughout this argument may indeed seem to make the conclusion redundant and the inference vacuous. If there is a genuine inference involved in the argument, then it presumably means one thing to claim that each person’s happiness is desirable for the person whose happiness it is and quite another to claim that the general happiness is desirable for the aggregate of all persons. However, if the reference both to happiness and to persons remains distributive throughout the argument, then the two
260
Chapter 7
claims are identical and the argument is superficial. On the other hand, if Mill switches from distributive to collective predication in the course of the argument, either with respect to happiness or with respect to persons, then he commits the fallacy of composition. The emerging problem is the possibility that, if Mill’s inference contains no move from parts to whole, and no switch from distributive to collective predication, then it avoids the fallacy of composition, but only because it is an empty tautology rather than a genuine inference. In order to qualify as a genuine inference, it must make the move and the switch under consideration. Otherwise, the conclusion repeats the premise in different words that together mean the same thing as the premise. With consistently distributive predication, the conclusion that the general happiness is desirable for the aggregate of all persons means the same thing as the premise that each person’s happiness is desirable for the person whose happiness it is. This observation culminates in a crippling dilemma, so the objection goes, such that either the argument contains a genuine inference that commits the fallacy of composition or it does not contain a genuine inference and turns out instead to be an empty tautology. This is to claim that Mill’s “general happiness” passage contains either a fallacy or a tautology but not a sound argument that makes a substantive contribution to the proof. As tempting as this line of criticism may seem, the observations adduced in support of it substantiate a false dilemma at best. The premise and the conclusion do not, as envisaged by the objection, mean the same thing, one merely repeating the other in different words. If they did, the inference would not be erroneous, just tautologous. Conversely, if the passage in question were objectionable strictly because of the apparent deduction there, then giving up the offending deduction should result in a definitive solution to the problem. Yet it is not at all likely to do so. Consider a thought experiment to illustrate the point, or at least to appreciate it. Suppose that Mill had used the phrase “in other words” instead of the word “therefore” to connect his premise that each person’s happiness is desirable for the person whose happiness it is with his conclusion that the general happiness is desirable for the aggregate of all persons. This would not, I submit, have silenced his critics. Quite the contrary, it would have united them behind the same accusation, with the same vigor, but with a different explanation. They would have still accused Mill of committing the fallacy of composition, while adding, presumably as an explanatory note on methodology, that he was committing that fallacy specifically in the process of confusing two entirely different propositions as opposed to inferring one directly from the other.
Reconstruction of Mill’s Proof
261
Since what is at stake here is the success of Mill’s proof as a whole, the critical question is whether the compositional inference plays a significant enough role in the proof to undermine its overall success. Is it possible to vindicate Mill’s proof as a whole by admitting that he commits the fallacy of composition in his inference, or alternatively by admitting that his inference turns out to be a tautology, and showing instead, in either case, that the inference is not an essential part of the proof, which succeeds independently of such a mistake in the “general happiness” passage? What exactly is the role of that passage in the proof? The role of the “general happiness” passage is more exegetical than logical. The compositional inference there is an explanatory narrative illuminating the proof rather than an embedded argument advancing the proof. Aiming to emphasize the importance of individuals and individuality in the ideological framework of utilitarianism, where everyone’s happiness counts the same, Mill maintains a distributive mode of reference both to happiness and to persons throughout his inference from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons. This is to treat personal happiness as a common and generic phenomenon as opposed to a distinctive or privileged one. The point of the inference is to establish the fungibility of happiness as a utilitarian paradigm: Each person’s happiness counts equally because, insofar as happiness is the standard of moral evaluation, one person’s happiness is just as desirable as another’s. Mill could have gotten the same point across more clearly, no doubt, in words less suspect of committing an elementary logical fallacy. The reason why he does not bother to spell out his reasoning any further in the grossly misunderstood passage in the fourth chapter of Utilitarianism is probably that he believes that he has already made his position clear enough throughout the second chapter and in the latter part of the third chapter. The following passage in the second chapter of Utilitarianism is a typical example of Mill’s repeated emphasis on the equality of individuals under the utilitarian standard: I must again repeat, what the assailants of utilitarianism seldom have the justice to acknowledge, that the happiness which forms the utilitarian standard of what is right in conduct, is not the agent’s own happiness, but that of all concerned. As between his own happiness and that of others, utilitarianism requires him to be as strictly impartial as a disinterested and benevolent spectator. mill U2:18 CW10:218
262
Chapter 7
With respect to the structure of the proof, the inference in the “general happiness” passage is Mill’s attempt to incorporate the theme of social justice into the proof itself. As he indicates in the following passage in the fifth chapter of Utilitarianism, he holds that justice is not something external to the principle of utility but a requirement implicit in that very principle: It [a conception of social justice] is involved in the very meaning of Utility, or the Greatest-Happiness Principle. That principle is a mere form of words without rational signification, unless one person’s happiness, supposed equal in degree (with the proper allowance made for kind), is counted for exactly as much as another’s. Those conditions being supplied, Bentham’s dictum, “everybody to count for one, nobody for more than one,” might be written under the principle of utility as an explanatory commentary. mill U5:36 CW10:257
This emphasis on the value of the individual, and thereby on social equality, comes across clearly throughout Utilitarianism as well as in Mill’s other works. Nevertheless, he wants to make sure that the proof itself accounts for social justice. Otherwise, an appeal or reference to equality in connection with the principle of utility might be mistaken for a presupposition or afterthought external to the principle itself. Any such clause concerning equality might then be taken as a makeshift solution to a serious problem. The following passage, again in the fifth chapter of Utilitarianism, supports this interpretation: This implication, in the first principle of the utilitarian scheme, of perfect impartiality between persons, is regarded by Mr. Herbert Spencer (in his Social Statics) as a disproof of the pretensions of utility to be a sufficient guide to right; since (he says) the principle of utility presupposes the anterior principle, that everybody has an equal right to happiness. It may be more correctly described as supposing that equal amounts of happiness are equally desirable, whether felt by the same or by different persons. This, however, is not a presupposition; not a premise needful to support the principle of utility, but the very principle itself; for what is the principle of utility, if it be not that “happiness” and “desirable” are synonymous terms? If there is any anterior principle implied, it can be no other than this, that the truths of arithmetic are applicable to the valuation of happiness, as of all other measurable quantities. mill U5:36n CW10:257–258
Reconstruction of Mill’s Proof
263
The cumulative evidence suggests that Mill’s inference from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons plays a strategic role in the proof and has a special methodological function there. With respect to the internal structure of the proof, the notorious inference marks a point of transition in Mill’s extended argument from egoistic to universalistic ethical hedonism (while also ruling out altruistic ethical hedonism). With respect to his moral theory in general, the same inference marks the point of connection between Mill’s theories of value and obligation. To recapitulate, the original question whether the success of Mill’s proof is independent of the viability of his compositional inference turns on two implicit questions. First, would the proof still work even if the inference actually committed the fallacy of composition? Second, would the proof still work even if the inference turned out to be an empty tautology rather than a genuine argument? The response in either case is negative. Even if it serves only in an auxiliary and explanatory capacity, the inference is the logical instrument by which Mill switches from egoistic ethical hedonism to universalistic ethical hedonism. If the logical instrument through which he accomplishes that switch is defective, then the proof remains trapped on the level of egoism. 7.1.2 The Moral Implications of Emphasis on the General Happiness There is also a significant moral dimension to the position that Mill’s reference to happiness and persons remains distributive throughout his inference from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons. If the phrase “general happiness” denotes each person’s happiness distributively, and the phrase “aggregate of all persons” denotes each person severally, then the question arises why anyone ought to care about the happiness of others. The principal of utility can hardly be morally relevant if it concerns individuals in isolation but not in interaction. Even if happiness is indeed desirable as an end in itself, and even if everyone’s happiness counts the same, it may reasonably be asked why people ought to consider one another’s happiness in making moral decisions and whether they actually do. Not only does Mill believe from a moral perspective that people ought to care about one another’s happiness, he also believes as a matter of fact that people do care about one another’s happiness. It is helpful to consider his causal explanation for why people care about one another’s happiness in order to illuminate his moral explanation for why they ought to do so. That explanation is in terms of psychological and sociological factors naturally motivating people to respect one another’s pursuit of happiness.
264
Chapter 7
The psychological foundation of the relevant etiology is in naturalistic terms. Mill anchors the generation of moral feelings to an emphatic impulse in human psychology in the following passage in “Sedgwick’s Discourse” (1835): The idea of the pain of another is naturally painful; the idea of the pleasure of another is naturally pleasurable. From this fact in our natural constitution, all our affections both of love and aversion towards human beings, in so far as they are different from those we entertain towards mere inanimate objects which are pleasant or disagreeable to us, are held, by the best teachers of the theory of utility, to originate. In this, the unselfish part of our nature, lies a foundation, even independently of inculcation from without, for the generation of moral feelings. mill CW10:60
Thus, moral feelings are naturally grounded in the psychological fact that we like pleasure and dislike pain, whether in our own experience or in that of others. Mill elaborates on this explanation in the following passage in “Utility of Religion” in his Three Essays on Religion (1874): A morality grounded on large and wise views of the good of the whole, neither sacrificing the individual to the aggregate nor the aggregate to the individual, but giving to duty on the one hand and to freedom and spontaneity on the other their proper province, would derive its power in the superior natures from sympathy and benevolence and the passion for ideal excellence: in the inferior, from the same feelings cultivated up to the measure of their capacity, with the superadded force of shame. mill CW10:421
This passage shows that the causally independent and self-sufficient psychological foundation of moral feelings, in the form of a natural tendency to consider other people’s happiness, is reinforced by additional psychological forces, such as sympathy, benevolence, and shame. Mill both explains and expands on this relationship as follows in a footnote (the entirety of which spans twenty pages in two sections) to a passage in his father’s Analysis of the Phenomena of the Human Mind (1869): This disinterested love and hatred of actions, generated by the association of praise or blame with them, constitute, in the author’s [James Mill’s] opinion, the feelings of moral approbation and disapprobation, which the majority of psychologists have thought it necessary to refer
Reconstruction of Mill’s Proof
265
to an original and ultimate principle of our nature. Mr. Bain [coeditor with John Stuart Mill], in the preceding note, makes in this theory a correction, to which the author himself would probably not have objected, namely, that the mere idea of a pain or pleasure, by whomsoever felt, is intrinsically painful or pleasurable, and when raised in the mind with intensity is capable of becoming a stimulus to action, independent, not merely of expected consequences to ourselves, but of any reference whatever to Self; so that care for others is, in an admissible sense, as much an ultimate fact of our nature, as care for ourselves; though one which greatly needs strengthening by the concurrent force of the manifold associations insisted on in the author’s text. Though this of Mr. Bain is rather an account of disinterested Sympathy, than of the moral feeling, it is undoubtedly true that the foundation of the moral feeling is the adoption of the pleasures and pains of others as our own: whether this takes place by the natural force of sympathy, or by the association which has grown up in our mind between our own good or evil and theirs. The moral feeling rests upon this identification of the feelings of others with our own, but is not the same thing with it. To constitute the moral feeling, not only must the good of others have become in itself a pleasure to us, and their suffering a pain, but this pleasure or pain must be associated with our own acts as producing it, and must in this manner have become a motive, prompting us to the one sort of acts, and restraining us from the other sort. And this is, in brief, the author’s theory of the Moral Sentiments. james mill 1869, vol. 2, 308–309, n. 58
The sociological foundation of Mill’s causal explanation is his belief that a stable sociopolitical existence, by its very nature, requires some sort of harmony and encourages serious cooperation, and obversely, that it minimizes interpersonal conflicts and discourages harmful competition (U3:10–11 CW10:231–233). Mill holds that people in society “grow up unable to conceive as possible to them a state of total disregard of other people’s interests” (U3:10 CW10:231). He claims that the individual “comes, as though instinctively, to be conscious of himself as a being who of course pays regard to others” and that the “good of others becomes to him a thing naturally and necessarily to be attended to, like any of the physical conditions of our existence” (U3:10 CW10:232): Now, whatever amount of this feeling a person has, he is urged by the strongest motives both of interest and of sympathy to demonstrate it, and to the utmost of his power encourage it in others; and even if he has none of it himself, he is as greatly interested as any one else that others
266
Chapter 7
should have it. Consequently, the smallest germs of the feeling are laid hold of and nourished by the contagion of sympathy and the influences of education; and a complete web of corroborative association is woven round it, by the powerful agency of the external sanctions. mill U3:10 CW10:232
It may be asked, specifically of the principle of utility, why people ought to consider one another’s happiness in making moral decisions. The quick answer is that Mill’s conception of social justice is implicit in the principle of utility, and therefore that the principle requires people to consider one another’s happiness in making moral decisions. This is not, however, an adequate response. It is not enough to assert that the principle of utility requires giving equal weight to each person’s happiness, or that Mill is a universalistic ethical hedonist, not an egoistic one. This will not take us very far as an answer, because the question does not in any way deny or contradict the fact that concern for others is required by the principle of utility or prescribed by Mill. Rather, the question simply asks what in the utilitarian scheme makes it morally compelling for individuals to pay regard to the happiness of other individuals. Mill’s real answer applies equally well to any moral scheme and not solely to a utilitarian one: The deeply-rooted conception which every individual even now has of himself as a social being, tends to make him feel it one of his natural wants that there should be harmony between his feelings and aims and those of his fellow creatures. If differences of opinion and of mental culture make it impossible for him to share many of their actual feelings— perhaps make him denounce and defy those feelings—he still needs to be conscious that his real aim and theirs do not conflict; that he is not opposing himself to what they really wish for, namely, their own good, but is, on the contrary, promoting it. This feeling in most individuals is much inferior in strength to their selfish feelings, and is often wanting altogether. But to those who have it, it possesses all the characters of a natural feeling. It does not present itself to their minds as a superstition of education, or a law despotically imposed by the power of society, but as an attribute which it would not be well for them to be without. This conviction is the ultimate sanction of the greatest-happiness morality. This it is which makes any mind, of well-developed feelings, work with, and not against, the outward motives to care for others, afforded by what I have called the external sanctions; and when those sanctions are wanting, or act in an opposite direction, constitutes in itself a powerful
Reconstruction of Mill’s Proof
267
internal binding force, in proportion to the sensitiveness and thoughtfulness of the character; since few but those whose mind is a moral blank, could bear to lay out their course of life on the plan of paying no regard to others except so far as their own private interest compels. mill U3:11 CW10:233
Mill’s answer, in the final analysis, is that what compels people morally, to act morally, is to be looked for not in their subscription to any particular normative ethical theory but in their tendency to be morally alive and ethically conscious in the first place. Normative ethical theories are not morally therapeutic with respect to people who are not decent folks to begin with. Such individuals are no more likely, especially if they lack a conscience altogether, to respond to one normative ethical theory than they are to another. Mill’s principle of utility cannot convince people to consider or promote the happiness of others any more than Kant’s categorical imperative can convince them to act from a sense of duty, to make universalizable choices, or to treat humanity as an end, never merely as a means. Why ought people to consider the happiness of other people in making moral decisions? Because what they would otherwise be making is not moral decisions but just decisions. If this response is not sufficient, then the real question is why one ought to be moral at all, which is not the unique burden of a utilitarian theory to answer, and which, in fact, transcends comparisons between normative ethical theories. 7.2
Second Part of the Proof
Mill’s goal in the second part of the proof is to demonstrate that nothing other than happiness is desirable as an end in itself, and thereby that happiness is the only thing that is desirable as an end in itself. The two parts of the proof are mutually consistent in terms of Mill’s appeal to the evidentiary relationship between desires and desirability, including the distinction between means and ends in that relationship. His strategy in the second part of the proof is grounded in his approach to the first part: If “the sole evidence it is possible to produce that anything is desirable, is that people do actually desire it” (U4:3 CW10:234), then the sole evidence it is possible to produce that something is the only thing that is desirable must likewise be that it is the only thing that people do actually desire. Of course, the sole evidence does not, for that reason alone, that is, for the lack of further evidence, become sufficient evidence, but the point is that what we have here is still an appeal to desires as evidence of desirability.
268
Chapter 7
Although the appeal to desires is still in effect, and still the main strategy, the presentation is different. In the first part of the proof, Mill presents the appeal entirely as a matter of fact, namely as the fact that people do actually desire happiness for its own sake. In the second part, he cannot submit outright, simply as a matter of fact, that happiness just so happens to be the only thing that people desire for its own sake. He must explain and justify that claim. His explanation, stated briefly, is that happiness is the only thing that people desire for its own sake, because whatever else people desire, they desire it either for the sake of the pleasure inherent in it, and thus as a means to happiness, or for its own sake, but thereby as a part of happiness. His justification, put simply, is an appeal to a principle of association, a psychological principle governing the transformational relationship between desires for things as means and desires for them as ends. The strategy determines the structure. The second part of the proof covers paragraphs four through ten. A natural break occurs between paragraphs eight and nine. First, in paragraphs four through eight, Mill argues for the intermediate conclusion that nothing other than happiness is desired for its own sake. Next, in paragraphs nine and ten, he argues for the intended conclusion that nothing other than happiness is desirable as an end in itself. He announces the forthcoming break in the second part of the proof immediately upon completing the first part: It is “necessary to show, not only that people desire happiness, but that they never desire anything else” (U4:4 CW10:234). He proceeds by identifying and eliminating considerations against happiness as the only thing that is desired and desirable for its own sake. He first notes that many things which, in common language, are not identified with happiness are desired for their own sakes (U4:4 CW10:234–235). He then admits that happiness has various ingredients that are desired and desirable for their own sakes albeit as parts of happiness (U4:5 CW10:235). Next, using various examples, he introduces a principle of association that governs the transmutation of means into ends and parts (U4:6–7 CW10:235–237). By the eighth paragraph, he is convinced that he has reached the intermediate conclusion that nothing other than happiness is desired for its own sake (U4:8 CW10:237). Given this intermediate conclusion, the next step toward the intended conclusion is the familiar appeal to desires as the only evidence for desirability. If happiness is the only thing that is desired for its own sake, then happiness is the only thing that is desirable as an end in itself (U4:9–10 CW10:237–238). Three elements are essential to understanding Mill’s strategy as well as the substance and structure of the second part of his proof of the principle of utility: his conviction that desires are the only evidence for desirability (U4:3 CW10:234); his conception of happiness as having various ingredients (U4:5
Reconstruction of Mill’s Proof
269
CW10:235); his principle of association between means and ends (U4:6–7 CW10:235–237). In paragraph eight, Mill’s focus on desires as the only evidence available for desirability leads him to draw a weaker intermediate conclusion than what is otherwise warranted by the combined evidence he amasses through his conception of happiness as having various ingredients and his principle of association. His hand is stronger at that point than the use he makes of it. The first aspect of its strength is that his elucidation of the conception of happiness as having various ingredients covers both desires for and the desirability of the various ingredients of happiness: They are “desired and desirable in and for themselves” (U4:5 CW10:235). The second aspect is that his principle of association applies not just to the desired but also to the desirable.1 To be more precise, it applies both to that which people desire for its own sake and to that which they find desirable as an end in itself. By the eighth paragraph, then, Mill is in a position to declare happiness the sole intrinsic good as well as the unique end object of people’s desires. Instead, he cautiously derives, “from the preceding considerations,” the weaker, intermediate conclusion that “there is in reality nothing desired except happiness” (U4:8 CW10:237). The reason why Mill states his intermediate conclusion in terms of desires alone is that he wants to create an opportunity for subsequently reiterating that desires are the only evidence for desirability. The purpose of the break in the second part of the proof is to emphasize that what is desired is still the operative clue to 1 Most scholars do not recognize that Mill’s principle of association applies to the desirable as well as the desired, and thus to goods as well as desires. The text bears this interpretation, however, both in Mill’s formal statements of the principle and in his examples of its application. Where he applies the principle of association to money, for example, not only does Mill contrast original with acquired desires for money, he also discusses what is originally “desirable about money” (and the original “worth” of money) versus its transformation into a “principal ingredient of the individual’s conception of happiness” (U4:6 CW10:235– 236). Desirability likewise becomes the basis of comparison where he examines the associational transformation of power and fame, asserting that the “amount of immediate pleasure annexed” to power and fame has “the semblance of being naturally inherent in them” (U4:6 CW10:236). The most telling passage of all is where he identifies the principle of association as a “provision of nature” that transforms things that are “originally indifferent” into valuable “sources of pleasure” (U4:6 CW10:236). This is where he adds that “virtue, according to the utilitarian conception is a good of this description” in the sense that “through the association thus formed, it may be felt a good in itself” (U4:7 CW10:236). He anticipates his various “ingredients of happiness,” which are “desired and desirable in and for themselves” (U4:5 CW10:235), in the first chapter of Utilitarianism, where he sets up a “comprehensive formula” as the goal of the forthcoming proof: “If, then, it is asserted that there is a comprehensive formula, including all things which are in themselves good, and that what ever else is good, is not so as an end, but as a mean[s], the formula may be accepted or rejected, but is not a subject of what is commonly understood by proof” (U1:5 CW10:208).
270
Chapter 7
what is desirable, and to play down the role of the indefinitely inclusive conception of happiness.2 Mill is at pains to stress that the justification of happiness as the only thing desirable as an end in itself does not turn on an axiomatic stipulation arbitrarily giving happiness an organic constitution encompassing all other goods by definition. He thus tries to make it clear that the second part of the proof proceeds with a psychological analysis of desires and not with a conceptual analysis of the ontology of happiness as a moral construct.3 Mill wants it understood that he is working with an empirical principle of human psychology. He does not want to be misinterpreted as offering a pet definition of happiness or a crafty deconstruction of what it means to be happy in the relevant sense. He explains his reasoning as follows:
2 Mill makes it sufficiently clear, even prior to the break in the second part of the proof, that he derives his inclusive conception of happiness from empirical (psychological) fact rather than sterile intuition. For example, with respect to the chief ingredient of happiness, virtue, he states that utilitarian moralists “recognise as a psychological fact the possibility of its being, to the individual, a good in itself, without looking to any end beyond it” (U4:6 CW10:235). These utilitarian moralists hold that, otherwise, “the mind is not in a right state” (U4:6 CW10:235). 3 Notwithstanding Mill’s strategic interest in emphasizing the break in the structure of the second part of the proof, the essential substance of that part of the proof can be represented without the break as follows: (P1) It might seem that happiness is not the only thing that is desired for its own sake. After all, people “do desire things which, in common language, are decidedly distinguished from happiness” (U4:4 CW10:235). Admittedly, the “ingredients of happiness are very various, and each of them is desirable in itself”; they are, in fact, “desired and desirable in and for themselves” (U4:5 CW10:235). (P2) Indeed, then, quite a few things are desired for their own sakes and desirable as ends in themselves but only upon the transmutation of desires for those things as a means to happiness into desires for them as ends in themselves, which are then desired and desirable as a part of happiness. (P3) As a matter of psychological fact, with the exception of happiness itself, all things that are desired for their own sakes and desirable as ends in themselves are originally desired and desirable only for the sake of the pleasure inherent in them and thus only as a means to happiness. If things that are originally desired and desirable only as a means to happiness, through extended association with happiness as the end they serve, eventually come to be desired for their own sakes and desirable as ends in themselves and not merely for the sake of the pleasure inherent in them, they are then and thereafter desired and desirable as a part of happiness (U4:6–7 CW10:235–237). (P4) As a psychological principle, then, with the exception of happiness itself, whatever is desired and desirable is originally desired and desirable as a means to happiness, and whatever is desired for its own sake and desirable as end in itself is desired and desirable as a part of happiness. (P5) In this sense, with the exception of happiness itself, whatever is desired and desirable is desired and desirable either as a means to happiness or as a part of happiness. (C) Therefore, happiness is the only thing that is desired for its own sake and desirable as an end in itself.
Reconstruction of Mill’s Proof
271
We have now, then, an answer to the question, of what sort of proof the principle of utility is susceptible. If the opinion which I have now stated is psychologically true—if human nature is so constituted as to desire nothing which is not either a part of happiness or a means of happiness, we can have no other proof, and we require no other, that these are the only things desirable. mill U4:9 CW10:237
The inference is thus anchored to psychological hedonism. We desire nothing other than happiness simply because, happiness being what it is and sentient creatures being what they are, that is how things work in the realm of nature. It is important to note, however, that the appeal to psychology in this passage continues to build on the principle of association between means and ends rather than proceeding with a direct, arbitrary, and radical affirmation of psychological hedonism. The reason why we desire nothing independently of happiness is that whatever we desire is either a means to happiness or else becomes a part of happiness through the association of means and ends. The second part of Mill’s proof of the principle of utility can be outlined as follows:
Outline of the Second Part of the Proof (P1) It might seem as if happiness were not the only thing that is desired and desirable for its own sake. After all, people “do desire things which, in common language, are decidedly distinguished from happiness” (U4:4 CW10:235). (P2) Indeed, quite a few things are desired for their own sakes, but only upon the transmutation of desires for those things as a means to happiness into desires for them as ends in themselves, which then become a part of happiness. (P3) As a matter of psychological fact, with the exception of happiness itself, all things that are desired for their own sakes are originally desired for the sake of the pleasure inherent in them and thereby only as a means to happiness. If things that are thus desired as a means to happiness, through extended association with happiness as the end they serve, eventually come to be desired for their own sakes and not merely for the sake of the pleasure inherent in them, they are then and thereafter desired as a part of happiness (U4:6–7 CW10:235–237).
272
Chapter 7
(P4) As a psychological principle, then, with the exception of happiness itself, whatever is desired is originally desired as a means to happiness, and whatever is desired for its own sake is desired as a part of happiness. (P5) In this sense, with the exception of happiness itself, whatever is desired is desired either as a means to happiness or as a part of happiness. (C1) Therefore, nothing other than happiness is desired for its own sake. (C2) Therefore, nothing other than happiness is desirable as an end in itself. (C3) Therefore, happiness is the only thing that is desirable as an end in itself.
The second part of Mill’s proof has traditionally not attracted as much criticism as the first part, mainly because critics rarely ran out of objections to the first part and saw little need to do further damage in the second. Two outstanding candidates for critical attention in the second part of the proof are the conception of happiness as having various ingredients and the principle of association supporting that conception. Here is Mill’s exposition of the sense in which happiness has various ingredients: The ingredients of happiness are very various, and each of them is desirable in itself, and not merely when considered as swelling an aggregate. The principle of utility does not mean that any given pleasure, as music, for instance, or any given exemption from pain, as for example health, are to be looked upon as means to a collective something termed happiness, and to be desired on that account. They are desired and desirable in and for themselves; besides being means, they are a part of the end. mill U4:5 CW10:235
Mill supports this conception of happiness with the psychological principle that the various ingredients of happiness are originally desired only as a means to happiness, and that, through their association with the end (happiness) for the sake of which they are originally desired, they eventually come to be desired and desirable as ends in themselves albeit as a part of happiness (U4:6– 7 CW10:235–237). He cites and discusses various examples of such ingredients, including music, health, money, power, fame, and virtue (U4:5–7 CW10:235– 237). The following passage demonstrates how he draws out the implications of the transformation of desires for such things as means to happiness into desires for them for their own sakes but as parts of happiness:
Reconstruction of Mill’s Proof
273
In these cases the means have become a part of the end, and a more important part of it than any of the things which they are means to. What was once desired as an instrument for the attainment of happiness, has come to be desired for its own sake. In being desired for its own sake it is, however, desired as part of happiness. The person is made, or thinks he would be made, happy by its mere possession; and is made unhappy by failure to obtain it. The desire of it is not a different thing from the desire of happiness, any more than the love of music, or the desire of health. They are included in happiness. They are some of the elements of which the desire of happiness is made up. Happiness is not an abstract idea, but a concrete whole; and these are some of its parts. mill U4:6 CW10:236
Mill’s manner of exposition becomes problematic here because he switches carelessly and repeatedly between references to parts of happiness and references to parts of the desire for happiness. He thereby risks confusing and misleading even careful readers and sympathetic commentators, and he certainly does not guard his language against uncharitable interpretations. At the most uncharitable extreme, his discussion of money as an example of one of the various ingredients of happiness leaves him open to the playful objection that physical objects cannot be parts of mental states. As a matter of fact, citing this very example, George Edward Moore objects, polemically rather than playfully, that money in the form of metal coins and paper bank notes cannot physically exist in the mind as a part of happiness (1903, § 3:43, pp. 123–124). To be sure, money stands out among Mill’s examples of the ingredients of happiness (music; health; money; power; fame; virtue). Yet he neither asserts nor implies that money itself, in the form of metal coins and paper bank notes, is a part of happiness. He holds merely that the desire for money is one “of the elements of which the desire of happiness is made up” (U4:6 CW10:236). Unfortunately, he is not consistent in his extended exposition of this relationship, and an uncharitable critic might indeed press the issue. On the other hand, Mill is not proposing anything untenable. He is on firm ground here. His interpretation of the relationship between happiness and its ingredients in terms of desires is not a makeshift solution to potential problems with the example of money as an ingredient of happiness. His emphasis on the parts of the desire for happiness points to the interpretation underlying all ingredients of happiness: Desires for them are included in the desire for happiness. Although Mill appeals to the principle of association in order to explain the connection between happiness and its ingredients in terms of desires, he does not otherwise propose that the principle either reflects or establishes that
274
Chapter 7
connection as an independent fact of the matter in nature.4 He does not infer that happiness and its ingredients are connected in objective reality simply because their ideas are connected in the mind through desires. He even warns against making such an unwarranted inferential leap in the following passage in A System of Logic: The habit of philosophical analysis, (of which it is the surest effect to enable the mind to command, instead of being commanded by, the laws of the merely passive part of its own nature,) by showing to us that things are not necessarily connected in fact because their ideas are connected in our minds, is able to loosen innumerable associations which reign despotically over the undisciplined or early-prejudiced mind. mill L3:21.1 CW7:565
The principle of association governs the transformation of desires for things as means to happiness into desires for them as ends in themselves albeit as parts of happiness. The principle thus states merely that desires for these things become parts of the desire for happiness, not that the things themselves which are the objects of desire are transformed into parts of happiness. The ingredients of happiness are not physical parts of happiness. They are not, for example, spatial parts of happiness as links are parts of a chain. Nor are they temporal parts of happiness as hours are parts of a day. The reason why the ingredients of happiness are not spatial or temporal parts of happiness is that they do not stake an exclusive claim to a unique segment or aspect of happiness. They are not even necessary or sufficient conditions of happiness in the standard sense of what it means to be a condition. First, they are not severally necessary conditions insofar as happiness is possible without money, without music, and so on, though probably not possible in the complete absence of all the ingredients commonly associated with it. Second, they are not severally sufficient conditions insofar as happiness does not necessarily accompany money, music, and so on, though it is indeed common in the presence of an
4 On the other hand, Mill believes that the principle of association establishes a different but related connection as an empirical fact, namely as a fact of nature. He claims that the transformation by which things that are originally desired for the sake of happiness, and desirable as a means to happiness, come to be desired and desirable as ends in themselves, is a “provision of nature” (U4:6 CW10:236). Nevertheless, the subject of the provision of nature here is not the connection between desires for happiness and desires for the ingredients of happiness but the transformation of desires for things as means to happiness into desires for them as ends in themselves.
Reconstruction of Mill’s Proof
275
abundance of such ingredients. The ingredients of happiness are probably jointly necessary conditions of happiness, but it is not clear whether they are also jointly sufficient conditions of happiness. Mill explains that the sense in which anything, physical or nonphysical, is an ingredient of happiness is that “the person is made, or thinks he would be made, happy by its mere possession; and is made unhappy by failure to obtain it” (U4:6 CW10:236). It follows from this that, even though happiness has various specific ingredients, it has only one kind of ingredient: experience. What is desired is always an experience, though always a particular experience, never just experience in general. Thus, the experience of money is a part of happiness but money itself is not; the experience of music is a part of happiness but music itself is not; and so on for the rest of the ingredients of happiness. A more plausible objection to the second part of the proof may be to deny the strength or validity of the principle of association serving as the centerpiece of that part of the proof. It makes little sense to reject the principle altogether, given that the orthodox interpretation of the transmutation of means into ends is a widely accepted principle of moral psychology.5 However, it might appear as if Mill’s version of the principle could be relaxed a bit to recognize that at least some things that are desired and desirable as ends in themselves are not desired and desirable as parts of happiness. It may be objected, for example, that knowledge is intrinsically good in its own right, independently of any relation to happiness. The objection, however, cannot reasonably be just that people tend to seek knowledge without expecting the pursuit to generate pleasure, for Mill has an answer in that case: We tend to will out of habit what we no longer desire for pleasure (U4:11 CW10:238–239).6 The objection, rather, 5 Willard Van Orman Quine (1978), for one, not only agrees that the association of means and ends can lead to the means being valued as ends but also claims that the transmutation of means into ends is the very basis of our moral training in which good behavior is reinforced and inculcated through rewards and penalties (cf. Alican 2021, 6–8). Quine, of course, does not add, as Mill does, that if the means ever come to be valued as ends in themselves, they then become valued as a part of happiness. 6 A System of Logic provides a similar account of things that are desirable in themselves yet desired without any anticipation of pleasure: “As we proceed in the formation of habits, and become accustomed to will a particular act or a particular course of conduct because it is pleasurable, we at last continue to will it without any reference to its being pleasurable. Although, from some change in us or in our circumstances, we have ceased to find any pleasure in the action, or perhaps to anticipate any pleasure as the consequence of it, we still continue to desire the action, and consequently to do it” (L6:2.4 CW8:842). The contribution of habit to value is illuminated further in the fourth chapter of Utilitarianism: “That which is the result of habit affords no presumption of being intrinsically good; and there would be no reason for wishing that the purpose of virtue should become independent of pleasure and
276
Chapter 7
must be that people tend to seek knowledge, sometimes simply in the form of information, even in instances where they are aware that its attainment might be unpleasant or outright painful. For example, a student might want to know the result of a test, a gambler the result of a bet, and a patient the result of medical scans, even if the outcome might be unfavorable in each case, which is bound to be unpleasant and possibly even painful, at least in psychological terms. Yet even if people in such circumstances typically do desire the knowledge in question, despite the risk of some kind and degree of pain, the example reflects a narrow construal of Mill’s conception of happiness as the ultimate end. Mill’s ultimate end is not simply pleasure and the absence of pain, even though that is the Benthamite core of the definition of happiness. Immediately after offering that definition, Mill adds that “to give a clear view of the moral standard set up by the theory, much more requires to be said” (U2:2 CW10:210). If the ultimate end were happiness conceived simply as a surplus of pleasure over pain, what more could possibly need to be said? No doubt, Mill adds his disclaimer to the core definition partly because he thinks it needs to be said, among other things, that happiness concerns the quality as well as the quantity of pleasures and pains. More importantly, though, the disclaimer reflects the insufficiency of the core definition to express his belief that the ultimate end is a sentient existence abundant in the experience of pleasure, not one that is completely immersed in it, that is, not one that is composed entirely of particular encounters with pleasure and categorically free of pain. The key terms here are “existence” and “experience.” Acknowledging the nature and role of “existence” helps emphasize the temporal dimension of the ultimate end, happiness, which Mill understands precisely as a happy existence. Acknowledging the function of “experience” and connecting it with “pleasure” helps guard against the objection that what is intrinsically good cannot be pleasure but the experience of pleasure.7 Happiness is indeed nothing pain, were it not that the influence of the pleasurable and painful associations which prompt to virtue is not sufficiently to be depended on for unerring constancy of action until it has acquired the support of habit” (U4:11 CW10:239). 7 George Edward Moore (1903) entertains such an objection but admits that hedonists have a legitimate defense: “Still it may be said that, even if consciousness of pleasure, and not pleasure alone, is the sole good, this conclusion is not very damaging to Hedonism. It may be said that Hedonists have always meant by pleasure the consciousness of pleasure, though they have not been at pains to say so; and this, I think is, in the main, true” (§ 3:53.2, p. 142). On the other hand, one may refuse to recognize the objection altogether, on the grounds that happiness presupposes experience or consciousness, as do pleasures and pains. A piece of chocolate that is otherwise pleasant, for example, might not be so pleasant in the mouth of someone in a coma. Likewise, the dentist’s drill, which can otherwise make one scream
Reconstruction of Mill’s Proof
277
but a sentient existence with reasonable access to pleasure and a fair share of its experience. This makes happiness an ongoing process, a work in progress, in contrast to pleasures and pains, which are momentary or transitory events. The difference is not just a matter of degree, where happiness is simply a persistent pleasure, thus rendering a happy existence an enduring experience of pleasure. A pleasure, no matter how long it lasts, is not happiness. Pleasure is attributable to discrete experiences, one at a time, whereas happiness is grounded in an existence composed of numerous experiences, most of which are pleasant, but some of which may be painful. A happy existence incorporates both pleasures and pains. It is consistent with both kinds of experience, so long as the painful kind does not dominate. Yet whereas a happy existence (happiness) is fully compatible with the experience of pain, a pleasant experience (pleasure) is not at all compatible with the experience of pain. The first clear indication that existence and experience are essential to Mill’s conception of the ultimate end is in his recapitulation of the utilitarian doctrine following his disclaimer concerning the core definition: According to the Greatest Happiness Principle, as above explained, the ultimate end, with reference to and for the sake of which all other things are desirable (whether we are considering our own good or that of other people), is an existence exempt as far as possible from pain, and as rich as possible in enjoyments, both in point of quantity and quality; the test of quality, and the rule for measuring it against quantity, being the preference felt by those who, in their opportunities of experience, to which must be added their habits of self-consciousness and self-observation, are best furnished with the means of comparison. mill U2:10 CW10:214
Mill eventually assimilates the function of existence and experience in his conception of the ultimate end into the definition of “happiness” itself. This enables him to distance utilitarianism from the cold and prudential felicific calculus in its Benthamite origins without giving up happiness as the ultimate end. He continues to embrace hedonism outright, as he eschews any claim to originality or revision, and shares his broad conception of happiness with in agony, is not at all painful in the numb mouth of a tranquilized patient. Moore, however, although he admits that hedonists have a legitimate response to the objection, does not consider the objection and response superfluous, because he denies that pleasures and pains presuppose experience or consciousness so strictly as to rule out the possibility of unexperienced or unconscious pleasures and pains (§ 3:53.1–2, pp. 141–142).
278
Chapter 7
“the philosophers who have taught that happiness is the end of life” (U2:12 CW10:215): The happiness which they meant was not a life of rapture; but moments of such, in an existence made up of few and transitory pains, many and various pleasures, with a decided predominance of the active over the passive, and having as the foundation of the whole, not to expect more from life than it is capable of bestowing. A life thus composed, to those who have been fortunate enough to obtain it, has always appeared worthy of the name of happiness. And such an existence is even now the lot of many, during some considerable portion of their lives. mill U2:12 CW10:215
It is consistent with this conception of happiness that the pursuit of knowledge for its own sake is a part of the happy life even if such a pursuit can lead to occasional encounters with pain. What makes anything an ingredient of happiness is not that the experience of that thing is necessarily, or always, a pleasant one. Likewise, an ingredient of happiness does not lose its claim to that title in the event that a particular experience of it turns out to be painful. Hence, the fact that the acquisition of knowledge can sometimes work out to the detriment of one’s immediate state of happiness does not mean that knowledge is not a part of the happy life. The same holds for other goods, or values, that are commonly acknowledged to be desirable in themselves but not generally acknowledged thereby to be desired and desirable as parts of happiness. The fact that virtue, justice, and beauty, to name a few such goods or values, are desired and desirable as ends in themselves does not preclude the possibility that they are desired and desirable as parts of happiness. Mill is wary of the mistaken tendency to think that being desired and desirable as an end and being desired and desirable as a part of happiness are mutually exclusive possibilities. A sentient existence devoid of virtue, justice, and beauty cannot be characterized, on the whole, as a happy existence. Ironically, his broadly inclusive conception of happiness makes it difficult to defend Mill where he asks, “what is the principle of utility, if it be not that ‘happiness’ and ‘desirable’ are synonymous terms?” (U5:36n CW10:258). Of course, he does not really hold that “happiness” and “desirable” are synonymous terms. He does not even hold that “happiness” and “desirable as an end in itself” are synonymous terms. His rhetorical question is just a carelessly stated hyperbolic summary of his conviction that happiness is the only thing that is desired
Reconstruction of Mill’s Proof
279
and desirable as an end in itself, because whatever is desired and desirable as an end in itself is desired and desirable as a part of happiness. Even if all this is granted, however, what we are left with in the end is a conveniently fluid notion of happiness that may appear suspicious in its infinite capacity to absorb all that is desired and desirable. The greater the domain of relevance negotiated for happiness, the less remarkable the result that happiness is the only thing that is desired and desirable as an end in itself. One may reasonably wonder, in that case, whether Mill proves too much, arguably a more difficult question to answer than whether the proof actually works. Carl P. Wellman articulates this possibility rather well: “As the denotation of the word is broadened, hedonism becomes at once more plausible and less significant. Mill has come close to making hedonism entirely plausible” (1959, 275). Mill can always respond, though, and rightly so in my opinion, that he is merely observing, invoking, and describing the inherently inclusive nature of happiness as a natural phenomenon, rather than inventing a definition of it, or developing a conception of it, that artificially broadens the meaning of the word.
c hapter 8
Implications of Mill’s Proof of the Principle of Utility This chapter provides provisional answers to a general question suggested by the culmination and conclusion of the discussion in progress throughout the preceding chapters: What are we to do with Mill’s principle of utility, assuming that it is valid and binding? The principle of utility states that happiness is the only thing that is desirable as an end in itself. Suppose that Mill has indeed proven this principle. The question, then, is this: How is it that this principle tells us what we ought to do? This is to ask both what it tells us to do and where it gets its authority. The first section explains Mill’s justification for taking the principle of utility as the standard of obligation. The second section outlines directions for further research toward a complete understanding of the principle and its role in ethical theory. The first section concerns Mill’s justification for grounding his theory of obligation in the theory of value he establishes in the proof. The aim is not to present a comprehensive exposition or to provide an infallible defense but to show the philosophical connection between Mill’s theories of value and obligation. The fact that Mill’s proof of the principle of utility concerns only value and not obligation makes it all the more important to investigate his justification for developing a theory of obligation from his theory of value. This is important because a value proposition such as “Happiness is the only thing desirable as an end in itself” does not automatically entail a moral imperative such as “So act as to promote happiness in any moral context” or “So act as to promote happiness whenever and wherever possible.” Nevertheless, having proved the former, Mill finds the latter too obvious to require justification. As established in the previous chapter, Mill’s inference from the desirability of each person’s happiness for the person whose happiness it is to the desirability of the general happiness for the aggregate of all persons contributes to his proof of the principle of utility in an explanatory capacity to establish the universalistic character of his particular brand of ethical hedonism. However, this apparently compositional inference also plays a functional role, external to the proof, as a point of connection between his theories of value and obligation. The inference concerning the
© NECİP FİKRİ ALİCAN, 2022 | DOI:10.1163/9789004503953_010
Implications of Mill’s Proof
281
general happiness contributes directly to the rationale for a theory of obligation grounded in the proof, specifically for the very theory Mill grounds in the proof. The second section concerns possibilities for further insight into the principle of utility, including preliminary indications of where and how such insight may best be sought. If Mill’s principle of utility is the first principle of a plausible normative ethical theory, then it is worthwhile to clarify details relevant to a complete understanding of the principle and its application. In other words, if happiness is the only thing that is desirable as an end in itself, and if the promotion of happiness is the standard of evaluation for determining the morality of actions and practices, then it is useful to develop guidelines for the application of that standard. Numerous questions may arise. Three in particular stand out. First, is the moral status of actions a direct function of their promotion of happiness, or is it instead a function of their conformity with moral rules, general compliance with which tends to promote happiness over the long run (subsection 8.2.1)? Second, are the morally relevant effects of actions to be contemplated in terms of their actual consequences, intended consequences, or foreseeable consequences (subsection 8.2.2)? Third, given that the standard of happiness underlying the principle of utility is the happiness of everyone concerned, is that standard to be calculated as total happiness or as average happiness (subsection 8.2.3)? Each of these questions receives exegetical and critical attention below, where all three are discussed with a view to articulating the highlights without any pretense to providing definitive answers. 8.1
Implications for a Theory of Moral Obligation
This is a good place to step back from the operational details of the proof to review Mill’s general accomplishments in the corresponding process. The fourth chapter of Utilitarianism opens with a strictly methodological discussion of the sort of proof appropriate for the principle of utility, blending fairly quickly into a delivery of the actual proof. The conclusion of the proof is that happiness is the only thing that is desirable as an end in itself. Here is a reformulation of the previous outline (see chapters 2 and 7 above) of the fourth chapter of Utilitarianism, this time displaying intricacies in the structure of the proof as well as revealing more of the subject matter of the individual paragraphs:
282 1 2 3 4 5 6 7 8 9 10 11 12
Chapter 8
Paragraphs of Chapter 4 of Utilitarianism in Relation to the Proof Discussion of Methodology Thesis: Happiness is the only thing desirable as an end in itself. Part 1: Thesis: Happiness is desirable as an end in itself. Part 2: Step 1: Nothing but happiness is desired as an end in itself. Ingredients of Happiness. Principle of the transmutation of means into ends and parts. Examples of the transmutation of means into ends and parts. Conclusion (thesis established) Step 2: Nothing but happiness is desirable as an end in itself. Conclusion (thesis established) Objection Considered: Desire vs. Will Closing Remarks
Mill’s accomplishment here is to have proven ethical hedonism. As submitted and corroborated in chapter 2 of the present volume, the proof aims at value alone, not at duty or obligation. However, even though the proof itself is not designed to establish a theory of obligation, it clears the grounds for constructing a particular theory of obligation on the theory of value thereby defended. The “general happiness” passage in the first part of the proof supports the logical groundwork by establishing the universalistic character of Mill’s ethical hedonism. This section demonstrates how Mill’s theory of obligation follows from his proof of the principle of utility as a theory of value. A key consideration is the relevance of the “general happiness” passage to the connection between Mill’s theories of value and obligation. While the aim is to trace the implications of the proof for a correlative theory of obligation, the attempt is nowhere near a complete exposition of Mill’s theory of obligation. A complete account would have to resolve certain issues beyond the coverage here. A crucial step toward full coverage would be to show that the theory is internally consistent despite offering two decidedly different criteria for distinguishing between right and wrong in moral matters: the promotion of happiness and the desert of punishment. The second chapter of Utilitarianism introduces the promotion of happiness as the standard of right and wrong: “The creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle, holds that actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” (U2:2 CW10:210). The fifth chapter of Utilitarianism presents the desert of punishment for the very same distinction: “We do not
Implications of Mill’s Proof
283
call anything wrong, unless we mean to imply that a person ought to be punished in some way or other for doing it; if not by law, by the opinion of his fellow creatures; if not by opinion, by the reproaches of his own conscience” (U5:14 CW10:246). Unfortunately, Mill does not, whereas his serious readers do, address the question that naturally comes to mind with these apparently competing criteria: Are actions morally right or wrong depending on whether their performance promotes happiness, or are they morally right or wrong depending on whether the performing agent deserves to be punished? If the promotion of happiness were indeed the operative moral standard, why would the desert of punishment come up as an additional consideration? Would not a failure to promote happiness, that being a necessary and sufficient condition, automatically determine the desert of punishment? This inherent ambiguity is the sense in which the focus of the present section is only on half the story with respect to Mill’s theory of moral obligation, that is, insofar as it passes over the question of “paradigm primacy” and considers the promotion of happiness independently of the desert of punishment.1 The question comes up again later, specifically in the context of act versus rule utilitarianism as the proper interpretation of Mill’s moral theory (see subsection 8.2.1), but it need not hold up the present discussion further.
1 A complete account of even half the story involves questions that divide scholars on the precise formulation of Mill’s theory of obligation but do not concern us here. For example, the promotion of happiness as the standard of moral obligation can be made a more discriminating test of right and wrong by demarcating degrees of promotion which mark the boundaries between morally obligatory, permissible, and impermissible actions. Are we obligated to maximize happiness, or merely to increase happiness, or perhaps just not to decrease it? In other words, does the standard imply supererogation, or are we permitted not to maximize happiness where it is possible to maximize it, or again from the less restrictive perspective, are we permitted not to increase happiness where it is possible to increase it? These questions are taken up briefly in chapter 2 of the present volume (see section 2.2). It would be useful to specify, in addition, whether actions are right insofar as they promote happiness, or insofar as they are performed in obedience to moral rules, or moral principles, which, if generally practiced, would consequently promote happiness in the long run (see subsection 8.2.1). And perhaps most important of all, it is necessary to explain how Mill’s theory of obligation accommodates rights and whether it leaves room for an adequate theory of justice. A complete account must deal with all of these issues, most of which are the standard burden of any theory that purports to be utilitarian and none of which promises a definitive consensus or tolerates a makeshift solution.
284
Chapter 8
Once Mill establishes that happiness is the only thing that is desirable as an end in itself, it never occurs to him to prove, in addition, that actions are right or wrong in accordance with whether they tend to promote happiness. This is because “direct proof,” or what is the same thing, “proof in the ordinary and popular meaning of the term,” is readily available in that regard: “Whatever can be proved to be good, must be so by being shown to be a means to something admitted to be good without proof” (U1:5 CW10:207–208). Since happiness is indeed something admitted to be good without proof, actions can quite certainly and most reasonably be evaluated in terms of their tendency to promote happiness. It might be tempting to object that happiness is not something admitted to be good without proof, given that Mill devotes the entire fourth chapter of Utilitarianism to proving that happiness is good. Yet what Mill means here by “something admitted to be good without proof” is “something admitted to be good without direct proof.” The reason why he offers a proof of the intrinsic goodness of happiness, specifically “in the larger meaning of the word proof,” is precisely that happiness is something admitted to be good without direct proof (U1:5 CW10:208). Of course, “ultimate ends are not amenable to direct proof,” because they are ends in themselves and cannot be shown to be a means to something else (U1:5 CW10:207). Ultimate ends are, in fact, themselves the standard of measurement in the sort of direct proof by which the goodness of other things is evaluated. Things that are not ultimate ends, however, are amenable to direct proof. Mill knows that, once he has established happiness as the only thing that is intrinsically good, hence as the ultimate end, direct proof is then available for evaluating the goodness of whatever else may be in question. If things that are not ultimate ends can be proved to be good by being shown to be a means to an ultimate end, and if happiness is the sole ultimate end, then it follows that actions are right insofar as they tend to promote happiness. This explains why Mill does not offer any sort of proof for the legitimacy of the utilitarian standard for evaluating the morality of actions. He finds the legitimacy of that standard too obvious to require proof.2 And that is why he does not give even a direct proof of the morality of actions that promote happiness. Here
2 Mill finds the connection between value, on the one hand, and obligation and justification, on the other, so obvious that he casually weaves in the connection at various points in the course of his proof of the principle of utility, especially in intermediate conclusions and summary statements. Here are a couple of examples: “Happiness has made out its title as one of the ends of conduct, and consequently one of the criteria of morality” (U4:3 CW10:234). “If so, happiness is the sole end of human action, and the promotion of it the test by which to judge of all human conduct; from whence it necessarily follows that it must be the criterion of morality, since a part is included in the whole” (U4:9 CW10:237).
Implications of Mill’s Proof
285
is a direct proof reconstructed on Mill’s behalf, from his own words, brought together strictly as a thought experiment for illustration purposes:
(P1) “All action is for the sake of some end, and rules of action, it seems natural to suppose, must take their whole character and colour from the end to which they are subservient” (U1:2 CW10:206). (P2) “Questions about ends are, in other words, questions what things are desirable” (U4:2 CW10:234). (P3) “Happiness is desirable, and the only thing desirable, as an end; all other things being only desirable as means to that end” (U4:2 CW10:234). (C) “Actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” (U2:2 CW10:210).
It is often tempting to fill in the blanks on behalf of past masters. This can be a fruitful endeavor so long as the blanks are filled in with the master’s own convictions and so much the better if they are filled in with his own words. In this case, Mill’s own words show that he does not lack justification for the standard of moral obligation he espouses. Insofar as actions aim at the attainment of some end, and happiness is the ultimate end, the test of right and wrong for actions is the promotion of happiness, which is, therefore, the standard of moral obligation. It may be objected that this direct proof offered on Mill’s behalf equivocates on the term “end,” which Mill tends to use both in a descriptive sense to refer to what is desired and in a normative sense to refer to what is desirable. The objection, then, would be that Mill uses the descriptive sense of “end” in the first premise to designate intended consequences, or states of affairs, and the normative sense of “end” in the other premises to designate the value of such consequences or states of affairs. In the first premise, for example, the claim that “all action is for the sake of some end” means that actions are performed in order to produce particular consequences, or to bring about certain states of affairs, whereas, in the second premise, the expression “questions about ends” refers to the value of such consequences or states of affairs. Yet while Mill does tend in general to use the word “end” in different ways, this does not affect the statements quoted above as premises of a possible direct proof. What makes the claims descriptive or normative there is not disguised in the word “end.” The distinction instead turns on the words that qualify “end,” together with the context in which the claims are made. Note that the expression “for the sake of” accompanies “end” in (P1), while the term “desirable” accompanies “end” in (P2) and (P3).
286
Chapter 8
At any rate, the point of producing a representative direct proof is not to conceal the distinction between descriptive and normative claims, but on the contrary, to expose that distinction in order to show that they both play a role in direct proof. The first premise of the direct proof comes from a passage where Mill contrasts the methods of normative ethics and empirical science in terms of the relationship between the general theory and particular truths in each (U1:2 CW10:205–206). In ethics, the general theory precedes the particular truths, whereas in science, the particular truths precede the general theory. The methodology of direct proof is a combination of these two methods: A general theory of value precedes particular truths, which precede a theory of moral obligation. Mill explains this as follows in A System of Logic: The relation in which the rules of art stand to doctrines of science may be thus characterized. The art proposes to itself an end to be attained, defines the end, and hands it over to the science. The science receives it, considers it as a phenomenon or effect to be studied, and having investigated its causes and conditions, sends it back to art with a theorem of the combinations of circumstances by which it could be produced. Art then examines these combinations of circumstances, and according as any of them are or are not in human power, pronounces the end attainable or not. The only one of the premises, therefore, which Art supplies, is the original major premise, which asserts that the attainment of the given end is desirable. Science then lends to Art the proposition (obtained by a series of inductions or of deductions) that the performance of certain actions will attain the end. From these premises Art concludes that the performance of these actions is desirable, and finding it also practicable, converts the theorem into a rule or precept. mill L6:12.2 CW8:944–945
Mill’s explanation suggests that the direct proof constructed above works as follows: If questions about value are “questions what things are desirable” (from P2), then actions are right insofar as they tend to produce consequences or states of affairs that are desirable, wrong as they tend to produce consequences or states of affairs that are undesirable. If happiness is the only thing desirable as an end in itself, all other things being desirable only as a means to happiness (from P3), then the promotion of happiness is the only standard of evaluation for the desirability of other things, because other things are desirable only insofar as they are a means to happiness (some of them eventually becoming a part of happiness). If the promotion of happiness is the only standard for evaluating the desirability of other things, then the promotion of happiness is
Implications of Mill’s Proof
287
the only standard of evaluation for the desirability of consequences or states of affairs brought about by actions. Therefore, actions are right insofar as they tend to produce consequences or states of affairs that promote happiness, wrong as they tend to produce consequences or states of affairs that promote the reverse of happiness. The promotion of happiness, then, is the test of right and wrong. But in order to make the standard of moral obligation more perspicuous, it is helpful to specify whose happiness it is that is to be promoted. Options include everyone’s happiness, the agent’s own happiness, or everyone’s happiness other than that of the agent. Mill insists repeatedly throughout the second chapter of Utilitarianism that happiness as the standard of moral obligation is the happiness of everyone concerned (U2:18 CW10:218).3 It is not enough, however, to insist that this is so with respect to moral obligation. Not only must Mill distinguish his universalistic ethical hedonism from egoistic ethical hedonism and altruistic ethical hedonism, he must also show that the distinction is consistent with, if not inherent in, his standard of value. If happiness as his standard of value determines the promotion of happiness as his standard of obligation, then he must demonstrate the mutual consistency of those standards. The temptation to suspect inconsistency originates in the first part of the proof, where Mill speaks of desires for and the desirability of happiness in connection with the moral agent. How can universalistic ethical hedonism follow from each person’s desiring his own happiness, especially if this establishes merely that each person’s happiness is desirable for the person whose happiness it is? The connection Mill offers is that each person’s happiness is desirable for the person whose happiness it is, and therefore, that the general happiness is desirable for the aggregate of all persons (U4:3 CW10:234). This often misunderstood inference is the key, inherent in the proof itself, that unlocks the universalistic character of Mill’s version of ethical hedonism. It is thereby
3 The universalistic character of the ethical hedonism espoused is hard to miss throughout Utilitarianism. An impeccably clear articulation of universalism awaits in the second chapter, where Mill explains that the “directive rule of human conduct,” according to the utilitarian standard, “is not the agent’s own greatest happiness, but the greatest amount of happiness altogether” (U2:9 CW10:213). The very next paragraph speaks of “the rules and precepts for human conduct, by the observance of which an existence such as has been described might be, to the greatest extent possible, secured to all mankind; and not to them only, but, so far as the nature of things admits, to the whole sentient creation” (U2:10 CW10:214). Arguably the clearest example of all is again in the second chapter: “The happiness which forms the utilitarian standard of what is right in conduct, is not the agent’s own happiness, but that of all concerned” (U2:18 CW10:218).
288
Chapter 8
the point of connection and the mark of consistency between his standards of value and obligation. The proper interpretation of the “general happiness” passage rests on two factors. The first is that Mill posits the relationship between the general happiness and the aggregate of all persons on the level of the desirable, not on that of the desired. He does not claim that anyone at all, let alone every single person, desires the general happiness. The second is that Mill’s reference, both to happiness and to persons, remains distributive throughout the passage in which he affirms the general happiness to be desirable for the aggregate of all persons. He does not consider the general happiness to be some sort of composite construct, a patchwork of sorts, synthesized from everyone’s personal happiness melded into a unitary whole.4 Nor does he conceive of the aggregate of all persons as a sociopolitical giant with a collective sentience assembled from every person’s consciousness rolled into an organic entity constituting a person of its own.5 To pursue the analogy further, Mill does not hold that such a patchwork of happiness is desirable as an actual good in its own right, be it for any particular individual or for a sociopolitical giant representing each and every individual. Nor does he claim or suggest anywhere in the proof that each person’s
4 Mill’s conception of the “general happiness” can be contrasted both with Rousseau’s (1762) conception of the “will of all” and with Rousseau’s conception of the “general will.” Rousseau describes the will of all as the sum of all individual wills, while elucidating the general will as the common element synthesized from the will of all after discounting everything that remains outside the region of overlap between individual wills. Mill’s general happiness is neither like Rousseau’s will of all, which requires aggregation, nor like Rousseau’s general will, which implies distillation. The reference to happiness is both distributive and comprehensive in Mill’s general happiness, which is therefore neither a sum nor a synthesis. 5 Mill’s conception of the aggregate of all persons can be contrasted with Hobbes’s (1651) conception of the leviathan (the metaphor, not the book). Mill’s aggregate of all persons is not like Hobbes’s leviathan, which is an artificial person created by the unity of real persons who contract out of the state of nature and into civil society. For a more graphic contrast, recall the famous picture on the cover page of the Head edition of Leviathan (perhaps more commonly known today as the picture on the cover of the Penguin Books edition). It is the image of a giant person composed of many small persons. That picture symbolizes the unity of the contractors in Hobbes’s social contract. Mill’s aggregate of all persons is not at all analogous to what that picture is intended to represent. If one’s understanding of Mill’s aggregate of all persons conjures up in one’s mind an image like the artificial person pictured on the cover of Leviathan, then one has an incorrect understanding of both the notion and the expression, much in the manner of John Stuart Mackenzie’s (1893, 104) interpretation where the leading example is “soldiers [standing] on one another’s heads,” reducing Mill’s aggregate to an absurdity as if “the minds of all human beings were to be rolled into one” (see section 3.7 in chapter 3 of the present volume).
Implications of Mill’s Proof
289
happiness is desirable for the fictional sociopolitical giant. The fact that he does not make either of these connections might be obvious to the reader if only because it is ludicrous that anyone at all should reason in this manner. However, as documented in chapter 3 of the present volume, Mill’s innocence of such absurdity has not always been so obvious to his critics.6 The correct interpretation is in chapter 5, which is devoted exclusively to refuting charges against the inference concerning the general happiness. The explanatory role of that apparently compositional inference is to indicate that everyone’s happiness is equally desirable, and thereby to distinguish Mill’s universalistic ethical hedonism from egoistic and altruistic ethical hedonism. The explanatory role, then, is to emphasize whose happiness it is that is desirable, not for whom it is that happiness is desirable. Granted, each person’s happiness is desirable for the person whose happiness it is, and possibly even only for that person, but the point is that if each person’s happiness is desirable as an end in itself, then “one person’s happiness, supposed equal in degree (with the proper allowance made for kind), is counted for exactly as much as another’s” (U5:36 CW10:257). The underlying explanation for this is that “equal amounts of happiness are equally desirable, whether felt by the same or by different persons,” because “the truths of arithmetic are applicable to the valuation of happiness, as of all other measurable quantities” (U5:36n CW10:257–258; editor’s bracketed insertions omitted). Mill’s proof of the principle of utility as a theory of value and his construction of a theory of obligation on the conclusion of that proof can be summarized as follows: The first part of the proof shows both that happiness is desirable as an end in itself and that everyone’s happiness is equally desirable (counts equally). The second part shows that nothing other than happiness is desired for its own sake or desirable as an end in itself. The overall conclusion of the proof is that happiness is the only thing that is desirable as an end in itself. This conclusion, together with Mill’s conviction that obligation is grounded in value, leads him to espouse the standard of obligation that we ought to act so as to promote the greatest happiness of everyone concerned. Nevertheless, it is entirely possible to grant that Mill consistently develops a theory of obligation from his theory of value, while denying that obligation really is grounded in value. This would be to affirm either that value and obligation are completely independent or that value alone does not determine obligation. The underlying objection here is that the tendency of actions to produce 6 Recall that Francis Herbert Bradley (1876, 103) likens the general happiness to pig slop in a single trough where multiple pigs are fed at once, while John Stuart Mackenzie (1893, 104) compares the aggregate of all persons to the mind of all human beings rolled into one.
290
Chapter 8
good or bad consequences, no matter what standard of good is employed, is not sufficient to determine whether actions are right or wrong. This poses a challenge that is not unique to Mill’s principle of utility but one that concerns any teleological theory of obligation. Of course, the generality of the challenge does not exempt Mill from having to meet the challenge. What, then, if not the production of good consequences, would be the standard of demarcation between morally right and wrong actions? A popular alternative is a deontological theory of obligation. Perhaps actions are morally right or wrong depending on whether they are done from a sense of duty, determined by a universalization, generalization, or reciprocity criterion, and unconditioned by motives, intentions, or consequences. If so, then duty is the test of right and wrong, and we ought to do what it is our duty to do, for no reason other than that it is our duty to do it. Mill would respond that “a test of right and wrong must be the means, one would think, of ascertaining what is right or wrong, and not a consequence of having already ascertained it” (U1:2 CW10:206). He would admit that our sense of duty distinguishes right from wrong, but he would have no use or respect for the circular explanation, or justification, that this is because it is right to do what it is our duty to do and wrong to proceed in contradiction of that. To be sure, this response addresses only one alternative to a teleological theory of obligation. It does not meet the general challenge to establish the promotion of good as the only grounds of moral obligation. However, Mill himself was indeed concerned with only one alternative to a teleological theory of moral obligation: ethical intuitionism. The following passage in the seventh chapter of his Autobiography best expresses his passionate rejection of intuitionism: The notion that truths external to the mind may be known by intuition or consciousness, independently of observation and experience, is, I am persuaded, in these times, the great intellectual support of false doctrines and bad institutions. By the aid of this theory, every inveterate belief and every intense feeling, of which the origin is not remembered, is enabled to dispense with the obligation of justifying itself by reason, and is erected into its own all-sufficient voucher and justification. There never was such an instrument devised for consecrating all deep seated prejudices. mill CW1:233
Mill advocates consequentialism as a theory of obligation with just as much conviction as he advocates universalistic ethical hedonism as a theory of value. As a matter of fact, he seems quite willing to argue for a teleological theory of obligation independently of a hedonistic theory of value. He maintains that
Implications of Mill’s Proof
291
the morality of actions is a function of their tendency to produce good consequences, regardless of how those consequences are to be measured, though he never hesitates to add that they are best measured in terms of happiness. He expresses his affirmation of consequentialism as the best theory of obligation in the following passage in “The Works of Jeremy Bentham” (1838), already quoted in part in chapter 2 and chapter 3 (section 3.1): Whether happiness be or be not the end to which morality should be referred—that it be referred to an end of some sort, and not left in the dominion of vague feeling or inexplicable internal conviction, that it be made a matter of reason and calculation, and not merely of sentiment, is essential to the very idea of moral philosophy; is, in fact, what renders argument or discussion on moral questions possible. That the morality of actions depends on the consequences which they tend to produce, is the doctrine of rational persons of all schools; that the good or evil of those consequences is measured solely by pleasure or pain, is all of the doctrine of the school of utility, which is peculiar to it. mill CW10:111
On the other hand, if the question is why a teleological theory of obligation is better than any alternative whatsoever, deontological or otherwise, hardly any answer can satisfy everyone. Presumably, nothing short of a critical contribution to the ongoing debate would be satisfactory. Probing the potential for a definitive answer in Mill’s works might prove interesting and fruitful in general, but it is not necessary in the context of the present chapter, nor even possible in the space of the present volume. 8.2
Directions for Further Research
Provided that Mill’s proof of the principle of utility is successful, and provided that it supports the theory of obligation he espouses, it is all the more worthwhile to work out the details of the principle of utility in order to determine whether it represents a useful and comprehensive ethical theory that can serve as a guide to moral reflection and judgment. The bare principle conceals various implications that require further thought for a complete understanding of the underlying theory and its application. This section initiates a preliminary discussion of avenues for consideration and deliberation toward the clarification of relevant details in the nature and application of the principle. It is organized around three parts corresponding to three sets of distinctions commonly
292
Chapter 8
drawn in any exposition of utilitarianism: act and rule utilitarianism; actual, intended, and foreseeable consequences; total and average happiness. The first part (subsection 8.2.1) explores the difference between act and rule utilitarianism in the context of Mill’s moral theory, focusing specifically on the plausibility of drawing that distinction in connection with his principle of utility. The second part (subsection 8.2.2) compares the actual, intended, and foreseeable consequences of actions, exposes the morally significant considerations underlying that distinction, and determines whether Mill takes the morally relevant consequences of actions to be their actual consequences, intended consequences, or foreseeable consequences. The third part (subsection 8.2.3) examines the difference between total and average happiness, questions the value of that distinction in moral theory in general, and denies its relevance to Mill’s principle of utility in particular. The approach is different in each case. The first issue requires a close inspection of the primary sources as well as an exegetical survey of the secondary literature. The second issue demands a careful and constructive study of the primary sources. The third issue calls for a critical analysis with a speculative orientation. 8.2.1 Act Utilitarianism vs. Rule Utilitarianism Commentators have long been divided over the question whether Mill’s moral theory is best interpreted in terms of act utilitarianism or in terms of rule utilitarianism. The basic difference, though there are various nuances, is the strategic role assigned to the principle of utility in the decision procedure employed for moral evaluation and justification. Both approaches invoke the principle of utility in that capacity, but act utilitarianism does so directly, while rule utilitarianism does so indirectly. Act utilitarianism makes moral justification a matter of evaluating the consequences of actions directly in reference to their promotion of happiness, that is, strictly as a function of their immediate tendency to promote happiness, whether to the highest degree possible or to a level unsurpassed by alternatives. Rule utilitarianism, in contrast, favors the evaluation of actions in terms of their conformity with moral rules whose general or widespread adoption throughout the relevant community tends to promote happiness in the long run, again through the production of at least as much happiness as alternate possibilities. Put simply, the focus in act utilitarianism is on the specific consequences of performing a particular action, while the focus in rule utilitarianism is on the general consequences of following a moral rule under which such actions are sanctioned. The scholarly debate over which approach better describes Mill’s position, if either one actually does so at all, goes hand in hand with the philosophical
Implications of Mill’s Proof
293
controversy regarding which one is the better approach in the first place. The distinction itself blossomed into a controversy in the second half of the twentieth century. Its emergence was not so much the result of a momentous discovery of two different possibilities within the justification mechanism in classical utilitarianism as it was the culmination of a progressive and collective development of insight into the plausibility of utilitarian justification. A gradual shift in focus from actions to rules inspired the development (by various hands rather than by any one person in particular) of rule utilitarianism as a novel interpretation, presented and widely accepted as an alternative to act utilitarianism, which then came to be associated with classical utilitarianism as the received view. Although the methodological developments shaping the distinction can be traced back indefinitely throughout the history of philosophy, typically in rudimentary anticipation in different contexts, the present state of the corresponding debate, together with the prevailing conceptions of both perspectives, dates back to the middle of the twentieth century. Anything earlier than that is no longer an attempt at a distinction within utilitarianism but an episode of experimentation with philosophical generalization as a guide to moral evaluation and justification.7 Just about any serious attention or devotion to moral rules, particularly in connection with generalization and universalization, may conceivably represent a forerunner to the distinction between act utilitarianism and rule utilitarianism, at least as a catalyst for the emergence of the latter. All it takes to find the relevant themes in past masters is to look for thought experiments in generalization, where even the categorical imperative of Immanuel Kant can reasonably be construed as an inspiration for considerations leading to the development of rule utilitarianism. An association to that effect is indeed what prompts Richard Mervyn Hare (1993) to ask just such a question: “Could Kant Have Been a Utilitarian?” Kant was no such thing, of course, as duly noted in a rejoinder by Jens Timmermann (2005): “Why Kant Could Not Have Been a Utilitarian.” Yet the question remains divisive. Henry Roy Forbes Harrod (1936, 147–155), for one, considers the analogy
7 Studies of philosophical generalization in moral evaluation and justification are often invoked in explanation of the methodological motivation behind the development of act and rule utilitarianism as competing approaches. Some of the most significant contributions to the literature are as follows: Richard Mervyn Hare’s “Universalisability” (1955) and Moral Thinking: Its Levels, Methods, and Point (1981, 107–116); Henry Roy Forbes Harrod’s “Utilitarianism Revised” (1936); Betsy Carol Postow’s “Generalized Act Utilitarianism” (1977); George Marcus Singer’s “Generalization in Ethics” (1955) and Generalization in Ethics (1961, 13–33, 34–60, 61–95, 139–177); Alan Ker Stout’s “But Suppose Everyone Did the Same” (1954).
294
Chapter 8
with Kant not only tenable but also practical, drawing on it liberally in his own analysis of the nature of moral obligation in utilitarian methodology, whereas David Lyons (1965, 8) finds the connection tenuous at best and the association vacuous in essence.8 The point, however, is not whether Kant was a utilitarian by any stretch of the imagination, but that the question is useful in sorting out the philosophical considerations behind the central concept, while at the same time bringing its evolution into historical perspective. The exercise is valuable in itself regardless of whether Kant was a utilitarian in any sense. Another fairly typical historical reference is to George Berkeley as the first major philosopher to articulate the fundamental insight behind rule utilitarianism if not the actual mechanics common and peculiar to it as an alternative.9 Consider the following passage from his Passive Obedience (1712): In framing the general laws of nature, it is granted we must be entirely guided by the public good of mankind, but not in the ordinary moral actions of our lives. Such a rule, if universally observed, hath, from the nature of things, a necessary fitness to promote the general well-being of mankind: therefore it is a law of nature. This is good reasoning. But if we should say, such an action doth in this instance produce much good and no harm to mankind; therefore it is lawful: this were wrong. The rule is framed with respect to the good of mankind; but our practice must be always shaped immediately by the rule. They who think the public good of a nation to be the sole measure of the obedience due to the civil power seem not to have considered this distinction. berkeley 1712, 31:15–25; 1953, 34
The detection of rule utilitarian tendencies in Berkeley (1712, 31), whether in this particular passage or in the general direction of the accompanying dialectical development, is certainly plausible, but it is hardly more informative or significant than the discovery of the original formulation of the principle
8 John Jamieson Carswell Smart (1973, 9–12, 83) considers the moral philosophy of Kant a relevant subject of contemplation in any effort to appreciate the development of utilitarian thought, which he illustrates in his own exposition of the bifurcation of utilitarianism, initially into act utilitarianism and rule utilitarianism, followed by the introduction of other divisions. 9 References to George Berkeley as a prominent figure, if not as the earliest forerunner, in the development of rule utilitarianism include: Richard Booker Brandt (1983, 38; 1988, 341, 346, 352; 1989, 94; 1995, 65, 77–78); Ben Eggleston (2017, 359, 363); Matti Häyry and Heta Häyry (1994); Colin Heydt (2014, 22–23); Brad Hooker (2006, 233; 2013a, 238; 2013b, 490–491); Joseph Kupfer (1974); Dale E. Miller (2010b, 96; 2014b, 147); Paul J. Olscamp (1969, 47–84).
Implications of Mill’s Proof
295
of utility in the work of Gottfried Wilhelm Leibniz (1700, 378), as discussed in chapter 1 of the present volume.10 Either connection is of great historical interest, to be sure, but then again, either one is of purely historical interest. It is hard to take either Berkeley or Leibniz seriously as a utilitarian in the standard sense. Neither one represents a significant benchmark in the development of classical utilitarianism, especially with Berkeley reminding us “that nothing is a law merely because it conduceth to the public good, but because it is decreed by the will of God, which alone can give the sanction of a law of nature to any precept” (1712, 31:1–4; 1953, 34), and with Leibniz reasoning that, “if the dominant principle in the existence of the physical world is the decree to give it the greatest possible perfection, the primary purpose in the moral world or in the city of God which constitutes the noblest part of the universe ought to be to extend the greatest happiness possible” (1686, §36 [=1908, 62]). The creative forces behind the distinction between act utilitarianism and rule utilitarianism are difficult to identify and appraise with any precision. The historical division is obviously the outcome of conceptual inspiration, ideological motivation, and philosophical reflection. But what is obvious about it goes only so far. The safest general assessment that can be made is that the distinction is the outgrowth of discussion, disagreement, and innovation in a collective effort comprising a series of influential contributions by a range of original thinkers.11 10
11
The possibility of Gottfried Wilhelm Leibniz providing the inspiration for the principle of utility, or at least anticipating its classical formulation, is discussed in section 1.3 of chapter 1 in the light of the evidence adduced in Joachim Hruschka’s “The Greatest Happiness Principle and Other Early German Anticipations of Utilitarian Theory” (1991). The distinction between act utilitarianism and rule utilitarianism is the cumulative outcome of discrete initiatives in various contributions. The relevant literature can hardly be exhausted in a footnote. A partial and subjective impression will have to suffice. The most prominent participants in the process can be listed as follows, with full bibliographic details provided in the “Works Cited” section at the end: (1) Richard Booker Brandt: Ethical Theory: The Problems of Normative and Critical Ethics (1959, 380–406, cf. 253–258, 296–298, 380–391, 391–396, 396–400, 400–406, 429–432, 442–451, 474–479, 489–505); “Toward a Credible Form of Utilitarianism” (1963); “Some Merits of One Form of Rule Utilitarianism” (1967); A Theory of the Good and the Right (1979, 163–182, 183–199, 266–285, 286–305); “The Real & Alleged Problems of Utilitarianism” (1983); “Fairness to Indirect Optimific Theories in Ethics” (1988); “Morality and Its Critics” (1989); Morality, Utilitarianism, and Rights (1992, 111–136, 137–157, 158–175); “Conscience (Rule) Utilitarianism and the Criminal Law” (1995); Facts, Values, and Morality (1996, 145–155, 156–159, 199–221). (2) Richard Mervyn Hare: “Universalisability” (1955); Freedom and Reason (1963, 112–136); “Principles” (1973); “Ethical Theory and Utilitarianism” (1976); Moral Thinking: Its Levels, Methods, and Point (1981, 107–116); “Could Kant Have Been a Utilitarian?” (1993). (3) Jonathan Harrison: “Utilitarianism, Universalisation, and Our Duty to Be Just” (1953); “The Expedient, the Right and the Just in Mill’s Utilitarianism” (1975); “Rule Utilitarianism and
296
Chapter 8
The development of the associated terminology, on the other hand, is well documented. The terms “act utilitarianism” and “rule utilitarianism” originate with Richard Booker Brandt (1959, 253, 396), who introduced them to distinguish between actions and rules as the proper methodological vehicle for moral evaluation and justification in utilitarianism.12 The same distinction was developed independently, and, in fact, a little earlier, by John Jamieson Carswell Smart (1956), who used different terms to explicate the same two approaches. The terms Smart employed were “extreme utilitarianism,” corresponding to “act utilitarianism,” and “restricted utilitarianism,” corresponding to “rule utilitarianism.” Brandt held rule utilitarianism to be the more plausible approach, while Smart espoused extreme utilitarianism, hence the opposite of Brandt’s choice, primarily because he found that extreme utilitarianism was not substantially different from the restricted utilitarianism intended to supersede it. As for nomenclature, Brandt’s terminology proved more appealing and intuitive, quickly replacing Smart’s less perspicuous terms in the secondary literature, with Smart himself subsequently professing a preference for the appellations “act utilitarianism” and “rule utilitarianism” (Smart and Williams 1973, 10, n. 1). Neither Brandt nor Smart, however, was the first in print to consider the matter specifically in relation to Mill. The difference between act utilitarianism and rule utilitarianism was taken up in connection with Mill’s moral theory prior to the participation of either contributor in the general debate, prior even to the introduction and standardization of any terminological
12
Cumulative-Effect Utilitarianism” (1979). (4) John C. Harsanyi: “Morality and the Theory of Rational Behavior” (1977a); “Rule Utilitarianism and Decision Theory” (1977b). (5) David Lyons: Forms and Limits of Utilitarianism (1965, 62–118). (6) John David Mabbott: “Moral Rules” (1953); “Interpretations of Mill’s ‘Utilitarianism’ ” (1956); An Introduction to Ethics (1966, 15–21, 22–31, 32–50). (7) John Rawls: “Two Concepts of Rules” (1955). (8) George Marcus Singer: “Generalization in Ethics” (1955); Generalization in Ethics (1961, 96–138, 178–216). (9) John Jamieson Carswell Smart: “Extreme and Restricted Utilitarianism” (1956); Utilitarianism: For and Against (1973, 3–74), the latter representing a collaboration with Bernard Arthur Owen Williams, where the first part consists of a revision of Smart’s monograph: An Outline of a System of Utilitarian Ethics (1961). (10) James Opie Urmson: “The Interpretation of the Moral Philosophy of J. S. Mill” (1953). The terminology originates specifically in Brandt’s Ethical Theory: The Problems of Normative and Critical Ethics (1959). Chapter 15, titled “Moral Obligation and Welfare” (pp. 380–406), offers an exposition, evaluation, and comparison of act utilitarianism and rule utilitarianism, together with alternatives and critical commentary. It consists of four sections: Section 15.1: “Act-Utilitarianism: Its Force and Problems” (pp. 380–391). Section 15.2: “The Formalist Theory Stated” (pp. 391–396). Section 15.3: “The Compromise: Rule-Utilitarianism” (pp. 396–400). Section 15.4: “Formalism or Rule- Utilitarianism” (pp. 400–406).
Implications of Mill’s Proof
297
preferences for the distinction. James Opie Urmson (1953) was the first to contemplate and articulate how best to interpret Mill in the context of what later came to be distinguished as act versus rule utilitarianism. He classified Mill’s approach under rule utilitarianism, placing emphasis on its incompatibility with act utilitarianism, though his actual presentation did not employ those terms, as his discussion preceded consensus on a standard terminology for the distinction. John Rawls (1955) and John David Mabbott (1956) were quick to respond to Urmson, maintaining that Mill scarcely noticed the significance of the difference. Rawls (1955) introduced an even deeper distinction as he differentiated between “summary rules” and “practice rules,” arguing incidentally albeit forcefully that Mill subscribed to a “summary rules” conception, which concealed the importance of demarcating the justification of rules or practices from the justification of actions falling under those rules or practices. Mabbott (1956) argued, more specifically with respect to Mill and more directly in response to Urmson, that if an act-versus-rule distinction can be found in Mill’s moral theory at all, textual evidence places him on the side of act utilitarianism. Donald George Brown (1974), Brian Cupples (1972), and Maurice Mandelbaum (1968), to name a few others, were also among the early and influential participants in the evolving reaction to Urmson, each one promptly defending the position that Mill was an act utilitarian. Developments in the general discussion following the standardization of the terminological distinction have centered on efforts to fine-tune the conceptual distinction and on critical comparisons of the relative merits of the two theories. Two of the strongest and most provocative reactions to the initial popularity of rule utilitarianism came from Smart (1961/1973) and Lyons (1965), both mentioned above. Smart and Lyons held separately, and with different ends in view, that rule utilitarianism collapses into act utilitarianism. Smart, being a utilitarian, was concerned more with defending act utilitarianism, whereas Lyons, who opposed utilitarianism, was concerned more with attacking rule utilitarianism. Smart and Williams (1973, 151–155) provide a useful annotated bibliography of early scholarship on the development of the basic distinction. The question whether act utilitarianism and rule utilitarianism ultimately amount to the same thing is about the possibility that rule utilitarianism, modified as required for plausibility, might thereby collapse into act utilitarianism, a problem also known as the extensional equivalence of act and rule utilitarianism. The reason to modify rule utilitarianism in any way, though especially for plausibility, would be to account for a contingency where following an otherwise perfectly respectable moral rule, such as the widely acknowledged one we all presumably follow against lying, would lead to
298
Chapter 8
disastrous consequences. A standard objection is that modifying a rule to avoid disastrous consequences, as in a murder facilitated by slavishly telling the truth to a raging psychopath inquiring as to the whereabouts of a dear friend, is the first step toward a slippery slope leading to further modifications, not just to avoid disastrous consequences but also to achieve favorable outcomes, even slightly more favorable ones, eventually culminating in a rule that is recursively modified into nothing other than the solitary goal of act utilitarianism to promote happiness. The matter is far from settled, with various additional distinctions being drawn on an ongoing basis either in elucidation or in extension of the primary one between act and rule utilitarianism.13 All such developments in the general debate have naturally been finding their way into the interpretation of Mill’s moral philosophy in the context of act versus rule utilitarianism. While the possibility of rule utilitarianism collapsing into act utilitarianism has never been specifically about Mill, any reasonable assessment of the question would obviously be relevant to an estimation of where Mill’s approach belongs in the distinction between the two perspectives, given that it cannot reasonably be placed in either category in the absence of a real difference between the two. The extensional equivalence envisaged for the competing perspectives, however, is merely one example from a stream of developments in response to conceptual and methodological innovations. The placement of Mill in that broader context has been sensitive to discrete developments without being entirely parasitic upon them. The main impetus behind the general tendency has come from the intricacies of textual interpretation, which is indeed where we must direct our efforts for an appreciation of the difficulties.
13
Contributions to the debate concerning the extensional equivalence of act and rule utilitarianism include the following: (1) Richard Booker Brandt: “Toward a Credible Form of Utilitarianism” (1963). (2) Baruch A. Brody: “The Equivalence of Act and Rule Utilitarianism” (1967). (3) Gertrude Ezorsky: “A Defense of Rule Utilitarianism against David Lyons Who Insists on Tieing It to Act Utilitarianism, Plus a Brand New Way of Checking out General Utilitarian Properties” (1968). (4) Allan F. Gibbard: “Rule- Utilitarianism: Merely an Illusory Alternative?” (1965). (5) Richard Mervyn Hare: Freedom and Reason (1963, 112– 136, especially 130– 136). (6) Gregory S. Kavka: “Extensional Equivalence and Utilitarian Generalization” (1975). (7) David Lyons: Forms and Limits of Utilitarianism (1965, 62–118). (8) John Jamieson Carswell Smart: “Extreme and Restricted Utilitarianism” (1956); Utilitarianism: For and Against (1973, 3–74). (9) Jordan Howard Sobel: “Rule Utilitarianism” (1968). (10) George B. Wall: “More on the Equivalence of Act and Rule Utilitarianism” (1971).
Implications of Mill’s Proof
299
Mill himself is, strictly speaking, neither an act utilitarian nor a rule utilitarian. His relevant contributions precede the formal distinction by roughly a century. The question, rather, is how best to read his work in the context of such refinements. Yet that debate has been hopeless from the beginning. This is not because Mill says nothing in support of either interpretation, but because he says plenty in corroboration of each one, which would have been confusing enough as it is, had he not also gone on to say quite a bit in contradiction of both positions, which then positively precludes the possibility of a definitive answer. As things stand, primary sources provide compelling evidence, at least on the face of it, for the construal of Mill’s position as act utilitarianism, combined with equally strong evidence for the classification of his position under rule utilitarianism, in addition to a plethora of circumstantial evidence against either interpretation, not to mention rather damning evidence contradicting both renditions at once. Mill’s works are pregnant with interpretive possibilities ready to be pursued in any direction. A review of the most telling passages will help orient the discussion, while demonstrating that and illustrating why any position on the matter is easy to support merely by consulting Mill’s works, and just as easy to undermine simply by flipping through the pages to find the opposite view. Act utilitarianism is a good place to start, given its current association with the received view prior to the bifurcation. One need look no further than Utilitarianism for an apparently perfect match between Mill’s position and act utilitarianism. Most of the evidence is clustered in the second chapter, which is dedicated to articulating what utilitarianism is, while also attempting to remove popular misconceptions. The first thing to note is that the chapter opens with an introduction of “utility as the test of right and wrong” (U2:1 CW10:209), which is significant in light of the fact that the evaluation of actions by their production of happiness, or of utility, hence in reference to the principle of utility itself, is precisely what distinguishes act utilitarianism from rule utilitarianism. One may be inclined to object that the passage makes no reference to actions, mentioning nothing more than a test of right and wrong, whereupon drawing a conclusion exclusive to actions may be reading too much into either the text or the context. While the objection would be valid for that particular passage in isolation, the remainder of the chapter, no less than the rest of the book, makes up for any ambiguity in that regard, leaving no doubt as to the intended subject of the corresponding test of right and wrong. There is no shortage of references to “Utility or Happiness, considered as the directive rule of human conduct” (U2:9 CW10:213), to “happiness” as “the rational purpose of human life and action” (U2:11 CW10:214), and even more perspicuously to “the happiness which forms the utilitarian standard of
300
Chapter 8
what is right in conduct” (U2:18 CW10:218). These and similar references confirm that actions are indeed the subject of discussion wherever Mill declares “utility to be the test of morality” (U2:24 CW10:224) or affirms that “Utility or Happiness is the criterion of right and wrong” (U5:1 CW10:240). Relevant passages are practically inexhaustible, but the evidence for act utilitarianism is at its strongest in the following three statements: The creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle, holds that actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness. mill U2:2 CW10:210
If no more be meant by the objection than that many utilitarians look on the morality of actions, as measured by the utilitarian standard, with too exclusive a regard, and do not lay sufficient stress upon the other beauties of character which go towards making a human being loveable or admirable, this may be admitted. mill U2:21 CW10:221
If so, happiness is the sole end of human action, and the promotion of it the test by which to judge of all human conduct; from whence it necessarily follows that it must be the criterion of morality, since a part is included in the whole. mill U4:9 CW10:237
The wording of the selections is so clear as to require no comment. This is not to say, however, that the evidence is conclusive. One could always object, not entirely unreasonably, that none of this constitutes sufficient evidence in favor of act utilitarianism, given that there is no mention anywhere of evaluating the consequences of actions directly in reference to the principle of utility. Talk of measuring the morality of actions by the utilitarian standard, one might thus object, could conceivably be a shorthand reference to the process of conducting the same appraisal through the conformity of such actions with moral rules that satisfy the utilitarian standard. This would be to object, in other words, that the last three block quotations and any comparable references, while perfectly consistent with act utilitarianism, do not necessarily contradict rule utilitarianism, because to evaluate actions in terms of their conformity with rules that tend to promote utility, and thus tend to produce happiness, is indeed to test actions by their promotion of utility, or by their production of
Implications of Mill’s Proof
301
happiness. Doing the former, so the objection would presumably continue, is one way, indeed the proper way, of doing the latter. There is no doubt that the evidence in favor of act utilitarianism is not conclusive, but there is also no need to try so hard to spin any of the evidence that clearly favors act utilitarianism. One would do much better to quote passages that actually contradict any inclination on the part of Mill to evaluate the consequences of actions directly in reference to the principle of utility. A prime example would be his discussion of rules in the second chapter of Utilitarianism, where he speaks with obvious contempt regarding the “endeavour to test each individual action directly by the first principle” (U2:24 CW10:224). The “first principle” here refers specifically but not exclusively to the principle of utility. The statement itself is a general assessment relevant to any first principle, but any reference to a “first principle” naturally includes the principle of utility, which Mill considers the first principle of morality, and not just of morality but also of prudence and of aesthetics broadly construed.14 As discussed in chapter 2 (section 2.5) of the present volume, these three areas of rational reflection, namely morality, prudence, and aesthetics, make up what Mill considers the “Art of Life” (L6:12.6 CW8:949), where the principle of utility serves as the first principle, that is, the fundamental principle, of the corresponding “theory of life” (U2:2 CW10:210).15 Hence, the contemptuous reference to testing “each individual action directly by the first principle” (U2:24 CW10:224) is clearly intended in opposition to the evaluation of actions through a direct appeal to the principle of utility, or in other words, 14
15
A representative list of relevant references in Utilitarianism to “first principles,” including correlative references to “fundamental principles,” the latter invoked either instead of or in addition to the former, where both are obviously used interchangeably, can be compiled as follows: U1:2 CW10:205; U1:3 CW10:206 (also mentioning “fundamental principles”); U1:4 CW10:207 (also mentioning “fundamental principles”); U2:24 CW10:224; U2:24 CW10:225 (instead mentioning “fundamental principles”); U2:25 CW10:226; U4:1 CW10:234; U5:36 CW10:257. It is toward the end of A System of Logic that Mill describes the “Art of Life” as having “three departments,” which he enumerates as “Morality, Prudence or Policy, and Æsthetics,” subsequently explicating them, respectively, as “the Right, the Expedient, and the Beautiful or Noble, in human conduct and works” (L6:12.6 CW8:949). The organic whole he identifies there as the “Art of Life,” including its three components, is directly relevant to the “theory of life” he later discusses in the second chapter of Utilitarianism (U2:2–2:4 CW10:210–211): “But these supplementary explanations do not affect the theory of life on which this theory of morality is grounded—namely, that pleasure, and freedom from pain, are the only things desirable as ends; and that all desirable things (which are as numerous in the utilitarian as in any other scheme) are desirable either for the pleasure inherent in themselves, or as means to the promotion of pleasure and the prevention of pain” (U2:2 CW10:210).
302
Chapter 8
in opposition to the evaluation of each action strictly in terms of its own production of happiness.16 Notwithstanding Mill’s explicit position on any inclination to test each individual action directly by the first principle, the passage actually happens to be a controversial one that can arguably be accommodated within a framework of act utilitarianism, so long as the alternative to testing each action by the first principle is interpreted not as pointing instead to a requirement of conformity with moral rules or principles but as anchoring the very test in question to the tendencies of comparable actions, hence to a cluster or family of actions. It is not quite fair, of course, nor at all helpful, to insist both that the evidence cannot be said to favor act utilitarianism, unless it confirms the evaluation of actions directly in reference to the principle of utility, and that the evidence cannot be said to contradict act utilitarianism, even where it repudiates the evaluation of actions directly in reference to the principle of utility. Yet conflicting positions of this sort need not, and, in fact, do not, come from the same side in the longstanding debate, where incompatible stipulations and awkward disagreements are par for the course in the dialectical development of the secondary literature. All in all, the evidence against act utilitarianism is no weaker than the evidence for act utilitarianism. The evidence in opposition is at its strongest in the following three passages, two from Utilitarianism, one from A System of Logic: In the case of abstinences indeed—of things which people forbear to do, from moral considerations, though the consequences in the particular case might be beneficial—it would be unworthy of an intelligent agent not to be consciously aware that the action is of a class which, if practised
16
The following passage, among any number of others listed in the note immediately preceding the previous one, shows that there can be no reasonable doubt that Mill regards the principle of utility as the first principle, or the fundamental principle, of morality: “Although the non-existence of an acknowledged first principle has made ethics not so much a guide as a consecration of men’s actual sentiments, still, as men’s sentiments, both of favour and of aversion, are greatly influenced by what they suppose to be the effects of things upon their happiness, the principle of utility, or as Bentham latterly called it, the greatest happiness principle, has had a large share in forming the moral doctrines even of those who most scornfully reject its authority. Nor is there any school of thought which refuses to admit that the influence of actions on happiness is a most material and even predominant consideration in many of the details of morals, however unwilling to acknowledge it as the fundamental principle of morality, and the source of moral obligation” (U1:4 CW10:207).
Implications of Mill’s Proof
303
generally, would be generally injurious, and that this is the ground of the obligation to abstain from it. mill U2:19 CW10:220
We must remember that only in these cases of conflict between secondary principles is it requisite that first principles should be appealed to. There is no case of moral obligation in which some secondary principle is not involved; and if only one, there can seldom be any real doubt which one it is, in the mind of any person by whom the principle itself is recognised. mill U2:25 CW10:225–226
I do not mean to assert that the promotion of happiness should be itself the end of all actions, or even of all rules of action. It is the justification, and ought to be the controller, of all ends, but is not itself the sole end. There are many virtuous actions, and even virtuous modes of action (though the cases are, I think, less frequent than is often supposed) by which happiness in the particular instance is sacrificed, more pain being produced than pleasure. mill L6:12.7 CW8:952
Mill apparently found not just the promotion of happiness in particular but also the production of optimific consequences in general methodologically inadequate for proper moral appraisal. His reservations on the matter seem to have been already well established in his “Remarks on Bentham’s Philosophy” (1833), fully three decades before the publication of Utilitarianism in book form: Now, the great fault I have to find with Mr. Bentham as a moral philosopher, and the source of the chief part of the temporary mischief which in that character, along with a vastly greater amount of permanent good, he must be allowed to have produced, is this: that he has practically, to a very great extent, confounded the principle of Utility with the principle of specific consequences, and has habitually made up his estimate of the approbation or blame due to a particular kind of action, from a calculation solely of the consequences to which that very action, if practised generally, would itself lead. He has largely exemplified, and contributed very widely to diffuse, a tone of thinking, according to which any kind of action or any habit, which in its own specific consequences cannot be proved to be necessarily or probably productive of unhappiness to the
304
Chapter 8
agent himself or to others, is supposed to be fully justified; and any disapprobation or aversion entertained towards the individual by reason of it, is set down from that time forward as prejudice and superstition. It is not considered (at least, not habitually considered,) whether the act or habit in question, though not in itself necessarily pernicious, may not form part of a character essentially pernicious, or at least essentially deficient in some quality eminently conducive to the “greatest happiness.” To apply such a standard as this, would indeed often require a much deeper insight into the formation of character, and knowledge of the internal workings of human nature, than Mr. Bentham possessed. But, in a greater or less degree, he, and every one else, judges by this standard: even those who are warped, by some partial view, into the omission of all such elements from their general speculations. mill CW10:7–8
This passage clearly contradicts act utilitarianism, ostensibly in the process of delivering a mixed message on the whole. Its full implications are not easy to sort out. While its opposition to act utilitarianism is plain to see, it can plausibly be read in opposition to rule utilitarianism as well, if not also in opposition to utilitarianism in general.17 This is because what Mill proceeds to do, immediately after repudiating act utilitarianism, is not to embrace rule utilitarianism as an alternative, but to introduce altogether different considerations 17
It may be helpful to note that the year of publication (1833) is not a significant interpretive factor here. Mill’s objections in his “Remarks on Bentham’s Philosophy” cannot be reduced to the whimsical misgivings of a young man of twenty-seven, later coming to see the truth but only in his maturity. This is not the last time Mill questions Bentham’s methods or conclusions, nor the only place he appears to undermine utilitarian ideology or to contradict consequentialist methodology. One need only browse through his Autobiography (CW1:116, 117) and his correspondence (CW12:81, 170, 204–205, 207, 312) to witness Mill’s doubts about the extent of his own association with and commitment to utilitarianism. Although he never actually considered himself estranged entirely either from the outlook or from the movement (CW1:184, 185), it is little wonder that he inspired Christopher Miles Coope, for example, to ask, and countless others, no doubt, to wonder: “Was Mill a Utilitarian?” (1998). This is not a trick question. Nor is it a rhetorical one. It is a perfectly meaningful question, as demonstrated by, and, in fact, as motivated by, the prevalence of jarring passages in Mill, such as the one quoted above from “Remarks on Bentham’s Philosophy” (CW10:7–8). It is not in any way comparable to the rhetorical question asked in jest in popular culture: “Is the Pope Catholic?” Nor is it even unique in the literature as a type of question or mode of inquiry. Its scholarly analogues in print are readily instantiated by the inclination of David Lyons (1971) and Max Hocutt (2005) to ask separately: “Was Bentham a Utilitarian?” An even more dramatic example is available outside utilitarianism in the independent initiatives of Mark Ralkowski (2007),
Implications of Mill’s Proof
305
pertaining to the role of character in moral discourse.18 This is not to say that rule utilitarianism is not compatible with the moral relevance of character but that Mill’s first thoughts run in another direction at that critical juncture. His immediate inclination upon dismissing the importance of direct consequences, or “specific consequences” as he calls them here, is to question the relevance of consequences altogether. The “great fault” Mill attributes to “Bentham as a moral philosopher” is not just excessive attention to “specific consequences” but also exclusive faith in utilitarian generalization: “calculation solely of the consequences to which that very action, if practised generally, would itself lead” (CW10:8). Mill does not simply steer us away from consequentialist evaluation, he steers us toward a competing standard requiring “much deeper insight into the formation of character, and knowledge of the internal workings of human nature” (CW10:8). To be perfectly clear, no amount of emphasis on character undercuts any requirement for compliance with moral rules, which, in turn, could conceivably still be formulated and evaluated in accordance with their tendency to promote happiness, which is never abandoned as the ultimate value in all facets of life. The only thing that is certain in the passage is its rejection of act utilitarianism as a decision procedure in moral philosophy. Further evidence against act utilitarianism is available in much of the text in favor of rule utilitarianism, though the reverse is also true, such that any evidence supporting act utilitarianism is likely to undermine the relevance of rule utilitarianism. The two positions are not jointly exhaustive, but they are mutually exclusive, unless one considers the existence of hybrid approaches
18
Konrad Rokstad (2011), and Lloyd P. Gerson (2013) to ask: “Was Plato a Platonist?” These are all legitimate questions, but none of them can be answered in passing in a footnote. The original sources in each case may be consulted profitably toward that end, though with fair warning not to expect a conclusive answer, especially not one establishing the revisionary perspective, for if Bentham and Mill were not utilitarians, and Plato not a Platonist, hardly anyone would ever qualify as what one appeared to be. The entirety of the fifth chapter of the sixth book of A System of Logic is devoted to Mill’s study of character as a philosophical concept: “Of Ethology, or the Science of the Formation of Character” (L6:5.1–6 CW8:861–874). Janice Carlisle’s John Stuart Mill and the Writing of Character (1991) offers a comprehensive analysis of the subject, including Mill’s insight into his own character, as well as Mill’s reflections on the significance of character as a philosophical concept. Other commentaries exclusively on Mill’s ethology include Terence Ball’s “The Formation of Character: Mill’s ‘Ethology’ Reconsidered” (2000) and Nicholas Capaldi’s “Mill’s Forgotten Science of Ethology” (1973). Broader studies devoting critical attention to the matter of ethology in Mill include Yuichiro Kawana’s Logic and Society: The Political Thought of John Stuart Mill, 1827–1848 (2018, 127–152) and Frederick Rosen’s Mill (2013, 72–94).
306
Chapter 8
an indication that they are not mutually exclusive either.19 However that may be, the positive evidence for rule utilitarianism is both strong and plentiful, in many cases doubling as evidence against act utilitarianism, or from the opposite perspective, borrowing from and building upon the evidence against act utilitarianism. Relevant passages include the last of the three block quotations cited above against act utilitarianism, which is an excerpt from A System of Logic (L6:12.7 CW8:952), where Mill separates both right action and moral obligation from the promotion of happiness, as he acknowledges the possibility that an action may be right and obligatory regardless of its production of happiness, sometimes even despite its production of unhappiness. It may be objected that the same passage works against rule utilitarianism, given that it includes “rules of action” alongside particular actions, yet the “rules of action” invoked there are not “rules of morality,” that is, “moral principles,” but merely “rules of thumb” estimating and summarizing the general tendency in actions of a particular kind. The difference between rules of morality and rules of thumb is another controversial matter in the interpretation of Mill’s moral theory. While the difference between genuine rules of morality (or moral principles) and generalizations of the sort relevant to rules of thumb (or rules of action) may be clear in theory, it is not at all clear whether Mill’s discussion of rules, particularly in the second chapter of Utilitarianism (U2:24 CW10:224–225), is limited to rules of thumb or applicable in addition to rules of morality. This keeps an otherwise relevant analysis of rules from being conclusive confirmation of either act utilitarianism or rule utilitarianism, as the same passage invokes rules that are comparable to calculations in the Nautical Almanac, which is consistent with rules of thumb, and thereby with act utilitarianism, while also embracing rules of morality, which is indicative of rule utilitarianism. The paragraph in question begins as a response to what Mill describes as the common objection that the utilitarian standard is impractical and unworkable because there is not always enough time to contemplate and calculate the possible and probable consequences of actions. Mill’s response is that life is not an incoherent series of fragmentary experiences where we rediscover how the world works at every step of the way in the course of our existence 19
Richard Mervyn Hare (1981), for example, describes his “two-level utilitarianism” as combining the strengths of both approaches without incorporating any of their weaknesses: “Much of the controversy about act-utilitarianism and rule-utilitarianism has been conducted in terms which ignore the difference between the critical and intuitive levels of moral thinking. Once the levels are distinguished, a form of utilitarianism becomes available which combines the merits of both varieties” (1981, 43).
Implications of Mill’s Proof
307
as moral agents, never extrapolating from past experience and never forming generalizations as to what we may reasonably expect from life. His immediate response remains perfectly consistent with act utilitarianism for the most part. But his subsequent discussion takes a turn toward rule utilitarianism as Mill proceeds from “rules of thumb” (“learning by experience the tendencies of actions”) to “rules of morality” (U2:24 CW10:224). This is where his purely prudential analogies concerning prior acquaintance and cumulative experience with the Old and New Testaments, or with the Nautical Almanac, turn into distinctively ethical considerations incorporating references to the “rules of morality” and the “received code of ethics”: There is no difficulty in proving any ethical standard whatever to work ill, if we suppose universal idiocy to be conjoined with it; but on any hypothesis short of that, mankind must by this time have acquired positive beliefs as to the effects of some actions on their happiness; and the beliefs which have thus come down are the rules of morality for the multitude, and for the philosopher until he has succeeded in finding better. That philosophers might easily do this, even now, on many subjects; that the received code of ethics is by no means of divine right; and that mankind have still much to learn as to the effects of actions on the general happiness, I admit, or rather, earnestly maintain. The corollaries from the principle of utility, like the precepts of every practical art, admit of indefinite improvement, and, in a progressive state of the human mind, their improvement is perpetually going on. But to consider the rules of morality as improvable, is one thing; to pass over the intermediate generalizations entirely, and endeavour to test each individual action directly by the first principle, is another. mill U2:24 CW10:224
A transition of this sort from prudential to ethical considerations can easily and reasonably be interpreted in favor of rule utilitarianism. One could very well start with generalizations from the consequences of actions and develop rules of morality in accordance with the collective results. The problem, however, is that the initial transition from the prudential to the ethical realm is followed immediately afterwards by a return to a purely prudential perspective as Mill concludes the same paragraph with talk of the Nautical Almanac. Any distinction between rules of thumb and rules of morality is hopelessly blurred in the passage, with Mill switching back and forth, and apparently indiscriminately, between clues in favor of act utilitarianism and those in favor of rule utilitarianism.
308
Chapter 8
Be that as it may, the evidence in favor of rule utilitarianism is not limited to the discussion of rules in the second chapter of Utilitarianism. Relevant passages both include and extend beyond that work. Among an assortment of examples, the most revealing are the following three passages, the first two from Utilitarianism, the last from “Whewell on Moral Philosophy” (1852), Mill’s review of William Whewell’s Lectures on the History of Moral Philosophy in England (1852) and Elements of Morality (1845): The intuitive, no less than what may be termed the inductive, school of ethics, insists on the necessity of general laws. They both agree that the morality of an individual action is not a question of direct perception, but of the application of a law to an individual case. mill U1:3 CW10:206
This [“this” =“happiness” =“an existence exempt as far as possible from pain, and as rich as possible in enjoyments, both in point of quantity and quality”], being, according to the utilitarian opinion, the end of human action, is necessarily also the standard of morality; which may accordingly be defined, the rules and precepts for human conduct, by the observance of which an existence such as has been described might be, to the greatest extent possible, secured to all mankind; and not to them only, but, so far as the nature of things admits, to the whole sentient creation. mill U2:10 CW10:214
Dr. Whewell supposes his self-deceiving utilitarian to be very little master of his own principles. If the effect of a “solitary act upon the whole scheme of human action and habit” is small, the addition which the accompanying pleasure makes to the general mass of human happiness is small likewise. So small, in the great majority of cases, are both, that we have no scales to weigh them against each other, taken singly. We must look at them multiplied, and in large masses. The portion of the tendencies of an action which belong to it not individually, but as a violation of a general rule, are as certain and as calculable as any other consequences; only they must be examined not in the individual case, but in classes of cases. mill CW10:181
The evidence against rule utilitarianism is just as plentiful, and just as strong, as the evidence for rule utilitarianism, much like the symmetry of the evidentiary balance in act utilitarianism. As already mentioned, the evidence
Implications of Mill’s Proof
309
for one position often constitutes evidence against the other. As it so happens, the passages quoted above in favor of act utilitarianism are indeed of such relevance that they may reasonably be invoked, one and all, as evidence against rule utilitarianism. And the evidence in question is not limited to Utilitarianism. While the examples adduced above in favor of act utilitarianism are, in fact, all from Utilitarianism, they can easily be supplemented by various references from outside that work. The most relevant passages in this case are from A System of Logic, where Mill repeatedly speaks out against slavish obedience to rules: As the judge has laws for his guidance, so the legislator has rules, and maxims of policy; but it would be a manifest error to suppose that the legislator is bound by these maxims in the same manner as the judge is bound by the laws, and that all he has to do is to argue down from them to the particular case, as the judge does from the laws. The legislator is bound to take into consideration the reasons or grounds of the maxim; the judge has nothing to do with those of the law, except so far as a consideration of them may throw light upon the intention of the law- maker, where his words have left it doubtful. To the judge, the rule, once positively ascertained, is final; but the legislator, or other practitioner, who goes by rules rather than by their reasons, like the old-fashioned German tacticians who were vanquished by Napoleon, or the physician who preferred that his patients should die by rule rather than recover contrary to it, is rightly judged to be a mere pedant, and the slave of his formulas. mill L6:12.2 CW8:944
By a wise practitioner, therefore, rules of conduct will only be considered as provisional. Being made for the most numerous cases, or for those of most ordinary occurrence, they point out the manner in which it will be least perilous to act, where time or means do not exist for analysing the actual circumstances of the case, or where we cannot trust our judgment in estimating them. But they do not at all supersede the propriety of going through (when circumstances permit) the scientific process requisite for framing a rule from the data of the particular case before us. mill L6:12.3 CW8:946
The error is therefore apparent, of those who would deduce the line of conduct proper to particular cases, from supposed universal practical
310
Chapter 8
maxims; overlooking the necessity of constantly referring back to the principles of the speculative science, in order to be sure of attaining even the specific end which the rules have in view. How much greater still, then, must the error be, of setting up such unbending principles, not merely as universal rules for attaining a given end, but as rules of conduct generally; without regard to the possibility, not only that some modifying cause may prevent the attainment of the given end by the means which the rule prescribes, but that success itself may conflict with some other end, which may possibly chance to be more desirable. mill L6:12.4 CW8:946
Yet none of these passages is conclusive either. The difficulty with trusting the extended discussion exemplified here, particularly as a decisive statement against rule utilitarianism, is the familiar ambiguity between rules of thumb and rules of morality. It is not quite clear whether the discussion concerns nothing more than the possibility of exceptions to rules of action, which would be quite consistent with both act utilitarianism and rule utilitarianism, hence not at all contradicting rule utilitarianism, or whether it extends to rules of morality as well, presumably indicating that this or that moral rule may be ignored where doing so would result in greater utility than otherwise, which would indeed contradict rule utilitarianism. The general tenor of the discussion seems to point to exceptions to prudential rules, but the difficulty of ruling out its additional relevance to moral rules leaves lingering suspicions concerning a possible rejection of rule utilitarianism. Mill clearly does not want us to “die by rule” (L6:12.2 CW8:944), no matter how the rule in question may best be classified. While it is important to determine what kind of rule is intended, it can certainly be no better to die by a moral rule than it is to die by a prudential one. We already know that we may, and, in fact, that we must, consult the principle of utility directly whenever any rules of morality come into conflict with one another. This is where Mill says, as quoted above in its fuller context, that “only in these cases of conflict between secondary principles is it requisite that first principles should be appealed to” (U2:25 CW10:226). It is not the appeal itself, but its mandatory employment, that is restricted to cases of conflict. The wording imposes no limitations on usage. It merely specifies when the appeal becomes obligatory, leaving open when, if ever, it becomes unacceptable, though there may admittedly be valid reasons not to use it all the time. On its own, the resolution envisaged in cases of conflict reduces the ultimate decision to a matter of which option is more productive of utility, that is, of
Implications of Mill’s Proof
311
happiness. Is it any wonder, then, that anyone should be tempted to ask why we may not ignore a moral rule whenever we find that ignoring it would produce greater happiness than following it? There may be a perfectly satisfactory answer. There is always such an answer, it seems, to every question in this debate. And that is just the problem. The scholarly literature, taken up in greater detail below, is a spectacular display of creative answers to difficult questions, with everyone remaining on equal footing in a context where any number of quotations from primary sources can quickly be rounded up to address any critical concern in the secondary literature, ironically with an equal number of equally relevant counterparts confirming the exact opposite. That is why the existing stock of literature keeps growing with new contributions reconfirming old conclusions with fresh evidence, or failing that, with fresh perspectives on the prevailing evidence. Sometimes one consideration rises above all others in popularity, becoming an academic favorite, if only for a limited time, as it is repeated religiously as decisively ending the debate in favor of one position or the other. The best example of such an appeal, one in favor of act utilitarianism, is a letter from Mill to the logician John Venn (Letter 1717A: 14 April 1872). The relevant portion of the letter reads as follows: I agree with you that the right way of testing actions by their consequences, is to test them by the natural consequences of the particular action, and not by those which would follow if every one did the same. But, for the most part, the consideration of what would happen if every one did the same, is the only means we have of discovering the tendency of the act in the particular case. In your example from Austria, it is only by considering what would happen if everybody evaded his share of taxation, that we perceive the mischievous tendency of anybody’s doing so. And that this mischievous tendency overbalances (unless in very extreme cases) the private good obtained by the breach of a moral rule, is obvious if we take into consideration the importance, to the general good, of the feeling of security, or certainty; which is impaired, not only by every known actual violation of good rules, but by the belief that such violations ever occur. mill CW17:1881–1882
Published in 1972, one hundred years after its composition, this letter quickly albeit temporarily tipped the scales in favor of act utilitarianism as the proper
312
Chapter 8
interpretation of Mill’s moral theory.20 Like any other surge of fresh insight into familiar territory, the discovery grew old as new paradigms emerged while original obstacles remained as divisive as ever. The first sentence of the portion of the letter quoted above is indeed as close as one could reasonably hope to get to explicit confirmation of act utilitarianism. The reason why it failed to convert everyone, especially on a permanent basis, is that the letter basically reiterates what is already covered to a great extent, if not on an exhaustive scale, in proper publications, for example, in Utilitarianism. While it may perhaps be said to emphasize or flesh out some of the relevant details better than ever before, there is actually very little in the letter in terms of evidentiary value that is not available elsewhere, whether as material ready to be quoted, or as implications begging to be noticed, or as inferences waiting to be drawn. Anyone inclined to embrace Mill’s letter to Venn as proof of his act utilitarianism should have been perfectly satisfied well in advance with prior evidence to the same effect in Utilitarianism, as exemplified particularly well by the three block quotations cited in favor of act utilitarianism in the beginning of this subsection. The letter introduces hardly anything that cannot be inferred from the affirmation that “actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” (U2:2 CW10:210), or from the correlative reference to “the morality of actions, as measured by the utilitarian standard” (U2:21 CW10:221), or from the subsequent declaration that “happiness is the sole end of human action, and the promotion of it the test by which to judge of all human conduct” (U4:9 CW10:237). As for anyone inclined to reject the authority of the evidence in the letter to Venn, or that of the parallel narrative in Utilitarianism, nothing short of an exclusive interview with the author himself can possibly be satisfactory. Evidence brought in from outside the corpus of Mill’s regular publications is not a practice limited to the case for act utilitarianism. The case for rule utilitarianism may be defended or attacked with just as wide a base of primary 20
Mill’s letter to John Venn (Letter 1717A: 14 April 1872) was first published in 1972 in the Collected Works of John Stuart Mill (CW17:1881–1882). It was not available in the 1910 edition of Mill’s letters compiled by Hugh Samuel Roger Elliot. Donald George Brown (1974) was the first to introduce the letter in a discussion note as conclusive evidence that the moral philosophy of Mill is best understood within the framework of act utilitarianism. Commentators invoking Mill’s letter to Venn in their analyses include: Fred R. Berger (1979, 125–134); David Owen Brink (2013, 96–97); Donald George Brown (1974, 68; 1982, 34; 2010, 37); Christopher Miles Coope (1998, 60); Roger Crisp (1997, 104, 117); Ben Eggleston (2017, 363–364); Ben Eggleston and Dale E. Miller (2008, 153, 156–157); Alan E. Fuchs (2006, 143–144); Dale E. Miller (2017, 689–690); Piers Norris Turner (2015b, 728, 733); Henry Robison West (1976a, 77; 2004, 75, 93).
Implications of Mill’s Proof
313
sources as that for act utilitarianism. For every piece of writing supporting act utilitarianism, indeed for every letter as well, one may find a valid counterpart supporting rule utilitarianism. A case in point is a letter from Mill to the barrister Henry Samuel Brandreth (Letter 1028: 9 February 1867). The relevant portion of the letter reads like an explanation in promotion of rule utilitarianism: Your question respecting the obligation of veracity on the utilitarian view of ethics seems, if I understand it rightly, to proceed on a misapprehension of the utilitarian standard. The test of right on the happiness principle is not the pleasure of doing the act which is declared to be right, but the pleasurable or painful consequences to mankind which would follow if such acts were done; & these, in the case you put, could not be enunciated in any general rule, because they depend on varying circumstances. There are cases in which martyrdom is a useless self sacrifice, & a sacrifice of other means of doing real good. There are other cases in which the importance of it to the good of mankind is so great as to make it a positive duty, like the act of a soldier who gives his life in the performance of what is assigned to him. There are cases again where without being so necessary as to be, on the utilitarian ground, an absolute duty, it is yet so useful as to constitute an act of virtue, which then ought to receive the praise & honours of heroism. The duty of truth as a positive duty is also to be considered on the ground of whether more good or harm would follow to mankind in general if it were generally disregarded and not merely whether good or harm would follow in a particular case. mill CW16:1234
Mill’s letter to Brandreth appears to do for rule utilitarianism what his letter to Venn does for act utilitarianism. Yet just like the Venn letter, though with less fanfare, it has failed to sway scholarly opinion toward a permanent consensus in its favor. The reason why the letter to Brandreth falls short of the qualities required for mass conversion is again comparable to the case of the letter to Venn: articulate reiteration in repetition of what is already available elsewhere. The problem with the Brandreth letter is that its reference to rules affords no greater clarity than the same discussion in the second chapter of Utilitarianism, where it is difficult to sort out whether the extended argument culminates in rules of morality for ethical decisions or in rules of thumb for prudential matters. The tension between the two letters supporting apparently opposite perspectives is a direct reflection of the nature of the evidence available in primary sources in general. There is enough evidence both for and against either
314
Chapter 8
version of utilitarianism to keep scholars busy for generations in their contributions to the ongoing debate. All of the evidence in support of either view is naturally in support of one view or the other, not both at once, and most of the evidence against either view is indeed against one view or the other, not both at once. That is why there is so much material to keep commentators satisfied and productive regardless of their viewpoint. Yet there is a certain development in the course of discussion in Utilitarianism that contradicts both act utilitarianism and rule utilitarianism at the same time. This is the well-known and widely cited passage in the fifth and final chapter of the work, where Mill takes up the connection between justice and morality, focusing, in the process, on the difference between morality and simple expediency, a difference which he associates with the idea of penal sanction. The paragraph is worth quoting in its entirety:21 The above is, I think, a true account, as far as it goes, of the origin and progressive growth of the idea of justice. But we must observe, that it contains, as yet, nothing to distinguish that obligation from moral obligation in general. For the truth is, that the idea of penal sanction, which is the essence of law, enters not only into the conception of injustice, but into that of any kind of wrong. We do not call anything wrong, unless we mean to imply that a person ought to be punished in some way or other for doing it; if not by law, by the opinion of his fellow creatures; if not by opinion, by the reproaches of his own conscience. This seems the real turning point of the distinction between morality and simple expediency. It is a part of the notion of Duty in every one of its forms, that a person may rightfully be compelled to fulfil it. Duty is a thing which may be exacted from a person, as one exacts a debt. Unless we think that it might be exacted from him, we do not call it his duty. Reasons of prudence, or the interest of other people, may militate against actually exacting it; but 21
The paragraph in question is widely referenced by commentators on the moral philosophy of Mill. It is the fourteenth paragraph of the fifth chapter of Utilitarianism (U5:14 CW10:246). Already famous as the strongest statement of the correlation Mill finds between morally wrong action and the idea of penal sanction, the paragraph in question, or rather the typical reference to that paragraph, has turned into something of a slogan in the form of “Mill U5:14,” or simply “Mill 5:14,” somewhat like the “John 3:16” signs that tend to crop up wherever particular people congregate, especially at sporting events in the United States. A representative example would be Dale E. Miller, who abbreviates the reference as “v.14” in one contribution (Miller 2010a) and as “V14” in another (Miller 2017). The title of the latter piece is worth noting in that regard: “Mill’s Act-Utilitarian Interpreters on Utilitarianism Chapter v Paragraph 14” (Miller 2017).
Implications of Mill’s Proof
315
the person himself, it is clearly understood, would not be entitled to complain. There are other things, on the contrary, which we wish that people should do, which we like or admire them for doing, perhaps dislike or despise them for not doing, but yet admit that they are not bound to do; it is not a case of moral obligation; we do not blame them, that is, we do not think that they are proper objects of punishment. How we come by these ideas of deserving and not deserving punishment, will appear, perhaps, in the sequel; but I think there is no doubt that this distinction lies at the bottom of the notions of right and wrong; that we call any conduct wrong, or employ, instead, some other term of dislike or disparagement, according as we think that the person ought, or ought not, to be punished for it; and we say that it would be right to do so and so, or merely that it would be desirable or laudable, according as we would wish to see the person whom it concerns, compelled, or only persuaded and exhorted, to act in that manner. mill U5:14 CW10:246
To elaborate on the terminology, expediency is nothing more than utility, a reference to the promotion of happiness, wherein actions are expedient insofar as they promote happiness, preferably the general happiness, or at least the happiness of the relevant community, which is to say, everyone in a position to be affected by the consequences of the action in question. This is not simply a deduction from the context but a summary of Mill’s explicit position. Mill makes an effort to distinguish his employment of the term from connotations in common parlance, identifying the expedient as “the same thing with the useful” (U2:23 CW10:223) and rendering “expediency” as “importance, to the common interest of mankind” (U5:6 CW10:242).22 He considers that which is expedient, as well as that which is right, to be informed by the principle of utility, with the right adding the idea of penal sanction to the felicific calculus. Expediency is the mark of prudence, which, together with morality and aesthetics, makes up the three branches of Mill’s conception of the “Art of
22
The meaning Mill assigns to the term “expedient” may also be corroborated in consultation of an authoritative interpretation in the scholarly literature: “Mill has a technical term which serves as the term of positive appraisal of action by the balance of its utilities: ‘expedient’ (see [Utilitarianism] 223, 241, 242, 248–249, etc.; Ev. ed., 21, 39, 40, 47–49, etc.). Tidying up a little, we can say that doing something is expedient when it results in more utility than not doing it, inexpedient when it results in less utility than not doing it, and maximally expedient when it results in more utility than any alternative open to the agent” (Brown 1982, 30).
316
Chapter 8
Life” (as mentioned above in the process of explicating Mill’s conception of the principle of utility as a “first principle”): “These general premises, together with the principal conclusions which may be deduced from them, form (or rather might form) a body of doctrine, which is properly the Art of Life, in its three departments, Morality, Prudence or Policy, and Æsthetics; the Right, the Expedient, and the Beautiful or Noble, in human conduct and works” (L6:12.6 CW8:949). To return to the main topic under discussion, we may note that the block quotation above, from the fifth chapter of Utilitarianism, opens up interpretive possibilities on multiple levels in connection with different but related questions. The passage is obviously of prime importance both for insight into Mill’s theory of morality in general and for a proper understanding of his conception of justice in particular. But it is also quite telling in the context of the present debate between act utilitarianism and rule utilitarianism as a viable interpretation of his approach to the morality of actions. The most natural reading of the paragraph is that it contradicts both act utilitarianism and rule utilitarianism in the most basic formulation of each perspective. This is because it confirms that the promotion of happiness, regardless of how it is done, whether through the careful contemplation of each act or in general obedience to rules, is not sufficient to distinguish between right and wrong, which always requires the propriety of punishment, including not just legal sanction, but also the opinion of one’s associates and compatriots, as well as the reproaches of one’s own conscience. What this means is that the felicific calculus cannot be the sole grounds for moral appraisal or justification, that it cannot serve as the operative etiological or justificatory framework, though it might work well as an indicative methodological convenience facilitating supplementary calculations in an auxiliary capacity. The significance of the paragraph is that it makes the desert of punishment a necessary condition of wrong action, which thereby precludes the promotion of happiness as the deciding factor in morality, which then rules out simple expediency, or optimific results, as the principle of distinction between right and wrong, regardless of whether one favors act utilitarianism or rule utilitarianism. Nothing else can be a sufficient condition where something has already been identified as a necessary condition. And the desert of punishment is indeed embraced as a necessary condition, which then rules out the promotion of happiness as a sufficient condition. One may conceivably wonder, of course, whether the reverse is true as well, perhaps even whether the reverse is true instead, that is, whether all the prior emphasis on the promotion of happiness does not rule out the desert of punishment as a formal criterion, either as a necessary condition or as a sufficient
Implications of Mill’s Proof
317
one. This is to ask, in other words, which of the two apparent criteria for moral evaluation and justification trumps the other in the event of a conflict between the two. The truth of the matter is that the language Mill uses is not as rigorous as all that. There is no explicit endorsement of a necessary condition ipso facto ruling out something else as a sufficient condition. What we have instead is a potential conflict between two different perspectives for insight into moral evaluation and justification, and thereby into moral obligation, unfortunately with no apparent resolution one way or the other. Despite the absence of a clear resolution, however, the presence of the conflict is itself a strike against both act utilitarianism and rule utilitarianism as the proper interpretation of Mill’s moral philosophy. The desert of punishment may or may not trump the promotion of happiness as the ultimate criterion of moral evaluation and justification, the determination of which may require further evaluation, if not additional evidence, but the emphasis on the desert of punishment does count as prima facie evidence against both perspectives as the intended or underlying methodological framework for moral evaluation and justification. It is difficult to square the formal function of the desert of punishment with either act utilitarianism or rule utilitarianism. Yet that is what commentators favoring either interpretation attempt to do in support of their position.23 The attempt alone is a testament to the plasticity of the evidence available, if not to the futility of the process initiated, or to the banality of the conclusion established. The entire debate boils down to interpretive gaps and textual ambiguities, which create and perpetuate irreconcilable differences as to what is meant in any particular passage, especially in discussions otherwise poised to be decisive either in favor of or in contradiction of one position or the other. The greatest obstacle to accepting act utilitarianism as the proper interpretation of Mill’s moral theory is the ever-present possibility of a patently valid moral obligation to do something else where an alternative promises greater direct returns on happiness. The greatest suspicion against rule utilitarianism in the same context is the nagging question of why—if the ultimate arbiter of conflict between moral rules or principles is the net balance of happiness promised
23
Philosophical initiatives to accommodate Mill’s notion of the desert of punishment within the greater context of an interpretive effort otherwise undermined by the notion include the attempts of David Owen Brink (2013), Roger Crisp (1997), and Piers Norris Turner (2015a; 2015b), among others, to account for the relevant passage within the context of act utilitarianism. A critical analysis of all three contributions is provided by Dale E. Miller (2017), who instead favors rule utilitarianism as the best fit for Mill’s moral theory (see also: Miller 2010a; 2010b; 2014a; 2014b).
318
Chapter 8
by compliance with those rules or principles—we may not ignore any rule or principle where doing so would produce more happiness for those affected by the action under consideration. And the crucial passages relevant to both but supportive of neither exclusively of the other, owing to ambiguities in the text in each direction, are those corresponding to the significance of rules, where it is difficult to say what kind of rules are meant (U2:24 CW10:224–225). The typical response by anyone inclined to find either act utilitarianism or rule utilitarianism in the moral philosophy of Mill, despite all the interpretive difficulties associated with each perspective, is that Mill’s approach is not the standard, traditional, or conventional version of act utilitarianism or rule utilitarianism but a modified or complex version of one or the other that somehow escapes all the difficulties attributed to the standard model in either case. As there is no end to the modifications that can be introduced to meet objections and challenges, so too is there no end to the possibilities in precisely what kind of utilitarian Mill may conceivably be said to be, the question remaining wide open with act utilitarianism and rule utilitarianism both admitting of meaningful variation. But the further we depart from the letter of the text, the closer we come to reinventing the corresponding message as opposed to understanding or interpreting it. Although this part of the present chapter is concerned primarily with demonstrating the malleability of the evidentiary context, together with the consequent fragility of positive interpretations, that aim itself admittedly brings no greater clarity to the matter than any of the constructive alternatives. Nor is it supported any better by the evidence it considers inadequate to draw reliable conclusions in the first place, the inadequacy being grounded, here too, in quantitative plenitude combined with qualitative uncertainty. It would be intellectually irresponsible, if not professionally disrespectful, to dismiss the positive approaches out of hand, as if they were inherently incompetent or altogether useless. On the contrary, it is precisely because the vast majority of constructive contributions represent the height of competence in scholarly research, rational analysis, and interpretive insight that the present state of the controversy promises no hope of a mutually satisfactory solution. There is no substitute for a comprehensive survey of the literature to appreciate the growing complexity of utilitarianism as an overarching framework for the interpretation of Mill’s moral philosophy. Only exhaustive coverage can capture the proliferation of essential differences and subtle distinctions, both in act utilitarianism and in rule utilitarianism, though also in other approaches, in the ongoing effort to demonstrate their relevance to Mill’s position. Given the natural limitations standing in the way of a comprehensive survey of the literature in a limited undertaking devoted largely to a critical evaluation, a
Implications of Mill’s Proof
319
viable alternative may be to select and discuss one example at length, preferably picking a highly representative or particularly influential contribution, followed by a number of brief yet broad suggestions for further reflection. Regarding the choice of a sample contribution for closer scrutiny, there is no better piece to consider than “The Interpretation of the Moral Philosophy of J. S. Mill,” a seminal article by James Opie Urmson (1953), who as intimated in the beginning of this subsection, became the first person to recommend rule utilitarianism as the proper interpretation of Mill’s moral theory, even before the standard terminology was in place and the corresponding debate was in progress. Urmson’s construal of Mill as a rule utilitarian is specifically a reaction to what he identifies, rightly for the time, as the received view of Mill as an act utilitarian. With the terminology yet to be invented, he never actually calls Mill a rule utilitarian, nor denies that Mill is an act utilitarian. His objection is simply that Mill is widely and erroneously read as making moral justification contingent upon the evaluation of actions directly in reference to their promotion of the ultimate end, namely happiness (Urmson 1953, 34). His central thesis in opposition is that Mill instead makes justification a matter of conformity with moral rules that promote the ultimate end in the manner envisaged in reference to actions (Urmson 1953, 35).24 His interpretation of Mill’s moral theory runs as follows:
24
I shall now set out in a set of propositions what I take to be in fact Mill’s view and substantiate them afterwards from the text. This will obscure the subtleties but will make clearer the main lines of interpretation. A. A particular action is justified as being right by showing that it is in accord with some moral rule. It is shown to be wrong by showing that it [transgresses] some moral rule. B. A moral rule is shown to be correct by showing that the recognition of that rule promotes the ultimate end. C. Moral rules can be justified only in regard to matters in which the general welfare is more than negligibly affected.
Moral rules, of course, do not themselves promote any end, which is why Urmson explicitly states that it is the observance of such rules that is at issue here: “Mill thought that a moral rule is shown to be correct by showing that the recognition of that rule promotes the ultimate end” (Urmson 1953, 38). He does not specify whether the rules in question are actual or ideal moral codes, because his pioneering contribution precedes that derivative distinction, just as it precedes the primary distinction between act and rule utilitarianism. Yet the context favors an interpretation in terms of ideal moral codes.
320
Chapter 8
D. Where no moral rule is applicable the question of the rightness or wrongness of particular acts does not arise, though the worth of the actions can be estimated in other ways. As a terminological point it should be mentioned that where the phrase ‘moral rule’ occurs above Mill uses the phrase ‘secondary principle’ more generally, though he sometimes says ‘moral law’. By these terms, whichever is preferred, Mill is referring to such precepts as ‘Keep promises’, ‘Do no murder’, or ‘Tell no lies’. A list of which Mill approves is to be found in On Liberty (p. 135). urmson 1953, 35
Urmson acknowledges that scholars favoring the received view, hence those reading Mill’s theory from the perspective of what we now classify as act utilitarianism, can and do make room for rules in their interpretation, without giving up act utilitarianism as the central paradigm: “We may, it might be admitted, on Mill’s view sometimes act, by rule of thumb or in a hurry, without actually raising this question; but the actual justification, if there is one, must be directly in terms of consequences, including the consequences of the example that we have set” (Urmson 1953, 38). His objection, therefore, is not that the received view cannot accommodate rules, but that it relegates the place of rules in Mill’s moral theory to the estimation of consequences, whereas Mill himself goes a step further to incorporate them into the corresponding model of ethical justification: “For Mill moral rules are not merely rules of thumb which aid the unreflective man in making up his mind, but an essential part of moral reasoning” (Urmson 1953, 36). The strategy Urmson follows to substantiate his interpretation is sensible but not successful. He offers a balanced combination of positive and negative evidence and argumentation. He thus proceeds with citations from primary sources in corroboration of his own thesis, combined with a concerted effort to undermine the authority of passages most frequently cited in favor of the received view, primarily by demonstrating that passages in the latter category can be, and in many cases must be, interpreted in favor of rule utilitarianism, or against the received view, at any rate.25 The reason to doubt whether he succeeds is not that the passages he cites in either category are inadequate or irrelevant, nor that Urmson himself says anything unreasonable in interpretation of the passages in question, but that the main 25
Urmson cites four passages in corroboration of his thesis, quoting each one at length, and thereby advocating rule utilitarianism as the proper interpretation of Mill’s moral theory: U1:3 CW10:206; U2:24 CW10:224–225; U2:25 CW10:225–226; U5:14 CW10:246.
Implications of Mill’s Proof
321
alternative always remains just as viable. There is more than enough evidence in primary sources to support either side of the debate and therefore not enough evidence to refute or undermine the opposite side. The inherent stalemate is further strengthened by the enormous latitude available for spinning any piece of evidence as supporting either of the two sides, or as contradicting either one, as the case may be, with even the most popular piece of evidence for one perspective liable to be appropriated in service of the opposite perspective. A powerful illustration of flexibility in interpretation is how Urmson (1953, 37) hijacks, under the banner of rule utilitarianism, the most widely cited passage in confirmation of act utilitarianism as the dominant inclination in Mill’s moral theory: “The creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle, holds that actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” (U2:2 CW10:210). Admitting that this assertion “could be taken as a loose and inaccurate statement of the received view, if the general argument required it,” Urmson (1953, 37) nevertheless places the intended meaning within the domain of rule utilitarianism. He submits that the qualification of “actions” with “tendencies,” twice in the same sentence, suggests that Mill was thinking of action-types rather than action-tokens: “Mill can well be interpreted here as regarding moral rules as forbidding or enjoining types of action, in fact as making the point that the right moral rules are the ones which promote the ultimate end” (Urmson 1953, 37). This is, of course, conceivable. Mill repeatedly calls attention to the tendencies of actions to promote happiness, just as Urmson suggests, instead of focusing on the direct consequences of actions in production of happiness. Yet Urmson’s point is not merely that an alternative reading of this sort is conceivable, but that this very alternative is warranted, if not compelling, in the light of Mill’s subsequent appeal to the “tendencies of actions,” later in the same chapter, in his discussion of the place of rules in utilitarian reasoning (U2:24 CW10:224–225). Another striking example of textual evidence that can be interpreted either way is the passage often invoked in opposition to the classification of Mill’s moral theory under either act utilitarianism or rule utilitarianism, which Urmson curiously brings up in corroboration of his reading of Mill from the perspective of rule utilitarianism. This is the famous passage in the fifth chapter of Utilitarianism (U5:14 CW10:246), already quoted in full earlier in the present subsection, where Mill introduces the idea of penal sanction as a requirement for the morality of actions that works independently of the production of happiness either by the actions contemplated or by the rules under which they are sanctioned. Yet Urmson rather inexplicably declares the
322
Chapter 8
passage perfectly consistent with rule utilitarianism, as he claims, without argumentation or explanation, that “Mill here makes it quite clear that in his view right and wrong are derived from moral rules” (Urmson 1953, 37). What Mill actually makes clear there, however, as his considered opinion on the matter, is not that right and wrong are derived from moral rules but that right and wrong are grounded in the idea of penal sanction, that is, in the desert of punishment. One may conceivably, arguably, and perhaps even reasonably infer Urmson’s interpretation concerning rules, as a generalization from Mill’s assertion concerning penal sanction, but that does not mean that the former is, as Urmson indicates, either Mill’s intention or his conclusion in the corresponding passage. Urmson treats the passage as proving too much. He even goes on to quote Mill’s explanatory summary in the next paragraph of the same chapter of Utilitarianism (U5:15 CW10:246–248) as if it were a confirmation of his own overinterpretation: “The applicability of moral rules is, says Mill, ‘the characteristic difference which marks off, not justice, but morality in general, from the remaining provinces of Expediency and Worthiness’ ” (Urmson 1953, 38, quoting U5:15 CW10:246–247). This is not what Mill says at all. Urmson prefaces his direct quotation here with a spurious reference to “The applicability of moral rules,” adding the expression, “says Mill,” as if that were what Mill himself had said in the previous paragraph, in other words, as if the applicability of moral rules were what Mill had asserted in the previous paragraph as the difference between morality and expediency. What we actually have in the text in place of “The applicability of moral rules” is a solitary “This” as the first word of the paragraph referring back to the one in which the distinction is drawn between morality and expediency. The best fit for Mill’s demonstrative pronoun here is “the idea of penal sanction,” which he had just introduced as the distinction between morality and expediency. Urmson’s substitution of “The applicability of moral rules” for Mill’s “This” goes beyond what is in the text. It may or may not be a valid inference, or an accurate interpretation, but it is indeed either an inference or an interpretation, quite possibly both, rather than the original statement and definite position that it appears to be in Urmson’s rendition. The passages Urmson cites in favor of his own interpretation, as well as those he expropriates from the opposite camp and reinterprets to serve his own purposes, demonstrate that the debate between act utilitarianism and rule utilitarianism as the proper interpretation of Mill’s moral theory is a hermeneutic quagmire of interpretive possibilities. Any passage can, with some imagination and a talent for reading between the lines, be made to support just about any conclusion or position, typically already formulated in advance. Sifting through the evidence to justify a pet theory, as opposed to formulating
Implications of Mill’s Proof
323
a proper theory based on the evidence, becomes the standard approach where there is an abundance of evidence but none of it is conclusive or compelling. The point, therefore, is not that Urmson is wrong, but that it is practically impossible to prove whether he is right or wrong. His reading of passages is always thoughtful, if not plausible, at least apparently so, never being outright unreasonable. Yet his conclusions contradict equally reasonable interpretations in equally respectable considerations. The question whether Mill’s moral theory represents act utilitarianism or rule utilitarianism, if either at all, has never ceased to command attention following its inaugural treatment by Urmson (1953). It is difficult, pace Urmson, to speak of a received view, standard interpretation, or scholarly consensus permanently shaping the state of scholarship throughout its history, though it is indeed common to find references to a received view reflecting a trend current during any particular period, which may or may not be valid at any other time. Even a cursory survey of the relevant literature shows that expert opinion has always been divided, not just over the question whether act utilitarianism or rule utilitarianism constitutes a better fit for Mill’s position, but just as strongly over the logically prior matter of whether either one is a suitable paradigm for Mill’s approach in the first place. My own impression of the literature, drawing on a statistical sample of thirty contributors active during various stretches of the history of the debate, is that the issue has long been and still continues to be highly polarized with fierce opposition between the two camps. There have indeed been some periods where one view or the other was more dominant, but both the overall breakdown and the balance at present show an even distribution. To be more specific, among the thirty cases examined here, Mill’s position is classified as act utilitarianism in nine, as rule utilitarianism in nine, and as neither one or the other in twelve. The last category represents not those who cannot make up their mind but those who maintain that Mill is neither an act utilitarian nor a rule utilitarian. The nine contributors supporting act utilitarianism (au) as the proper interpretation of Mill’s moral theory are as follows: (AU1) David Owen Brink: Mill’s Progressive Principles (2013, 79– 112). (AU2) Donald George Brown: “Mill on Liberty and Morality” (1972); “What Is Mill’s Principle of Utility?” (1973); “Mill’s Act-Utilitarianism” (1974); “Mill on Harm to Others’ Interests” (1978); “Mill’s Criterion of Wrong Conduct” (1982); “Mill’s Moral Theory: Ongoing Revisionism” (2010). (AU3) Roger Crisp: Routledge Philosophy Guidebook to Mill on Utilitarianism (1997, 95–133). (AU4) Brian Cupples: “A Defence of the Received Interpretation of J. S. Mill” (1972). (AU5) John David Mabbott: “Interpretations of Mill’s ‘Utilitarianism’ ” (1956). (AU6) Maurice
324
Chapter 8
Mandelbaum: “Two Moot Issues in Mill’s Utilitarianism” (1968). (AU7) Mark Philip Strasser: The Moral Philosophy of John Stuart Mill (1991, 23–53). (AU8) Leonard Wayne Sumner: “Mill’s Theory of Rights” (2006). (AU9) Piers Norris Turner: “Punishment and Discretion in Mill’s Utilitarianism” (2015a); “Rules and Right in Mill” (2015b). The nine contributors favoring rule utilitarianism (ru) as the best fit for Mill’s approach can be summarized as follows: (RU1) Richard Booker Brandt: “Some Merits of One Form of Rule Utilitarianism” (1967); “The Real & Alleged Problems of Utilitarianism” (1983); “Conscience (Rule) Utilitarianism and the Criminal Law” (1995). (RU2) Wendy Donner: “Utilitarianism: Morality, Justice, and the Art of Life” (2009).26 (RU3) Ben Eggleston and Dale E. Miller: “India House Utilitarianism: A First Look” (2007); “Mill’s Misleading Moral Mathematics” (2008).27 (RU4) Alan E. Fuchs: “Mill’s Theory of Morally Correct Action” (2006). (RU5) David Lyons: “Mill’s Theory of Morality” (1976); “Human Rights and the 26
27
What Wendy Donner claims is not that Mill’s views come out clearly and distinctly on the side of rule utilitarianism but that, given all the commonly recognized methodological refinements in the basic distinction between act and rule utilitarianism, Mill’s overall position can reasonably be construed as being more consistent with rule utilitarianism, specifically with the “actual code” variety, than with any other division or interpretation (2009, 52–54). Her careful and qualified classification of Mill in this regard, almost as if the decision were forced upon her despite her convictions, and against her better judgment, is consistent with her earlier denial that Mill belongs in either category (Donner 1998, 279–280). Although Ben Eggleston and Dale E. Miller (2007; 2008) together classify Mill’s moral philosophy under rule utilitarianism, as discussed further in the main text below, their individual contributions to the relevant literature reveal somewhat different albeit complementary concerns. Eggleston (2010; 2014; 2017) comes across more as an exponent of the issues, while Miller (2010a; 2010b; 2010c; 2014a; 2014b; 2017) stands out more as a proponent of rule utilitarianism (for interpreting Mill). This is certainly not a rigid distinction. Each scholar contributes to progress in either direction, but they also exhibit certain differences in corroboration of the basic contrast, with their work generally completing or complementing each other rather than either repeating or contradicting each other. Eggleston admits, for example, that Mill could reasonably be classified as an act utilitarian, specifically as a “sum-ranking welfarist act-consequentialist” (2014, 133–135), while elsewhere exploring the possibility of reconciling the portrait of Mill as a rule utilitarian with Mill’s own misgivings regarding blind devotion to rules, commonly known as the “rule-worship objection” or the “incoherence objection” (2010). The best example of Eggleston’s dispassionate approach to the issues, however, is probably his appraisal of the distinction between act and rule utilitarianism in the context of an effort to uncover the precise nature of Mill’s moral standard, which doubles as the title of his work: “Mill’s Moral Standard” (2017). In contrast to Eggleston’s generally more balanced perspective, Miller’s (2010a; 2010b; 2010c; 2014a; 2014b; 2017) independent work is directed more positively toward demonstrating and defending a methodological fit between Mill’s position and some type of rule utilitarianism, typically a sophisticated version with pertinent
Implications of Mill’s Proof
325
General Welfare” (1977); “Mill’s Theory of Justice” (1978); “Liberty and Harm to Others” (1979); “Benevolence and Justice in Mill” (1982). (RU6) Rex Martin: “Two Concepts of Rule Utilitarianism: The Case of Mill” (2007); “Two Concepts of Rule Utilitarianism” (2008); “Mill’s Rule Utilitarianism in Context” (2010). (RU7) Dale E. Miller: “Brown on Mill’s Moral Theory: A Critical Response” (2010a); J. S. Mill: Moral, Social and Political Thought (2010b, 71–78, 79–110); “Mill, Rule Utilitarianism, and the Incoherence Objection” (2010c); “John Stuart Mill’s Moral, Social, and Political Philosophy” (2014a); “Rule Utilitarianism” (2014b); “Mill’s Act-Utilitarian Interpreters on Utilitarianism Chapter v Paragraph 14” (2017).28 (RU8) Richard B. Miller: “Actual Rule Utilitarianism” (2009). (RU9) James Opie Urmson: “The Interpretation of the Moral Philosophy of J. S. Mill” (1953). The twelve commentators finding neither act utilitarianism nor rule utilitarianism (nn) a good match for Mill’s position are as follows: (NN1) John M. Baker: “Utilitarianism and ‘Secondary Principles’ ” (1971). (NN2) Fred R. Berger: “John Stuart Mill on Justice and Fairness” (1979); Happiness, Justice, and Freedom: The Moral and Political Philosophy of John Stuart Mill (1984, 64– 120).29 (NN3) David Copp: “The Iterated-Utilitarianism of J. S. Mill” (1979). (NN4) Douglas Poole Dryer: “Mill’s Utilitarianism” (1969); “Justice, Liberty, and the Principle of Utility in Mill” (1979). (NN5) Rem B. Edwards: “The Principle of Utility and Mill’s Minimizing Utilitarianism” (1986). (NN6) Gerald Gaus: “Mill’s Theory of Moral Rules” (1980). (NN7) John Gray: “John Stuart Mill: Traditional and Revisionist Interpretations” (1979); Mill on Liberty: A Defence (1983, 9–14, 28–42); “Indirect Utility and Fundamental Rights” (1984). (NN8) Jonathan Harrison: “The Expedient, the Right and the Just in Mill’s Utilitarianism” (1975); “Rule Utilitarianism and Cumulative- Effect Utilitarianism” (1979). (NN9) Kenneth Einar Himma: “The Interpretation of Mill’s Utilitarianism”
28
29
qualifications. Note incidentally that Miller (2010c), like Eggleston (2010), denies that the “rule-worship objection” is decisive against the depiction of Mill as a rule utilitarian. It may be objected that Dale E. Miller is counted twice: once with Ben Eggleston (RU3) and once on his own (RU7). This is true, but it does not inflate the total for the category (ru), given that Eggleston and Miller as a team (RU3) are counted as a single entry among the advocates of rule utilitarianism as providing the best match for Mill’s moral theory. Fred R. Berger’s place on the list is neither as certain nor as safe as those of the others. While he acknowledges that “there is not conclusive evidence either way” (1979, 123), and likewise that “Mill’s theory was neither an act-nor a rule-utilitarian theory as those terms are strictly defined” (1984, 65), he also maintains that “the weight of evidence from Mill’s writings, though not conclusive, favors the claim that he held something very like the act-utilitarian view” (1979, 124), later elaborated as an “ ‘act-consequence’ interpretation” (1984, 66).
326
Chapter 8
(1998).30 (NN10) Daniel Jacobson: “J. S. Mill and the Diversity of Utilitarianism” (2003). (NN11) Jonathan Riley: “Mill’s Extraordinary Utilitarian Moral Theory” (2010). (NN12) Henry Robison West: “Mill’s Moral Conservatism” (1976a); An Introduction to Mill’s Utilitarian Ethics (2004, 74–95). This breakdown is not even close to being definitive. It consists of elementary extrapolations from personal research as opposed to statistical data from a scientific study, but the pattern is nevertheless indicative of the state of scholarship, especially over the long run, but also at present. What is interesting is not just that scholars favoring act utilitarianism and those favoring rule utilitarianism are evenly divided, at least within the parameters of this informal survey, but also that those in either category are decisively outnumbered by scholars rejecting both models as the proper interpretation of Mill. The general spirit of dissent and disagreement is mirrored both between and within the individual groups, as contributors exhibit significant interpretive differences in at least some of the essentials and most of the details. The category of rule utilitarianism, for example, can be divided further into subdivisions, depending partly on the nature of the rules and partly on the methodological role envisaged for them. Wendy Donner (2009) and Richard B. Miller (2009), for their part, classify Mill’s approach as the “actual code” type of rule utilitarianism, where the relevant rules are interpreted as actual moral codes (representing conventional morality) as opposed to ideal moral codes. Richard Booker Brandt (1967; 1983; 1995) and Alan E. Fuchs (2006), on the other hand, and arguably also David Lyons (1976; 1977; 1978; 1979; 1982), construe Mill’s position as the “ideal code” variety of rule utilitarianism, where rules are interpreted as ideal moral codes rather than being taken as the actual moral codes prevailing at present in the relevant community.31 These are the 30
31
Kenneth Einar Himma (1998) identifies Mill with what he calls the “utilitarian generalization” approach (crediting David Hargraves Hodgson 1967, 5, with the concept, though not with its attribution to Mill), a third option, differing from act utilitarianism in denying that the immediate outcome is all that matters, and differing from rule utilitarianism in denying that moral rules are binding. He describes the alternative as follows: “According to utilitarian generalization, the source of the moral obligation to perform an act resides entirely in its tendency to promote utility; thus, conformity/failure-of-conformity to a rule is not any part of what makes an act right/wrong. Acts are right (or wrong) because they tend (or do not tend) to promote utility and not because they conform (or fail to conform) to an applicable moral rule” (Himma 1998, 457). While the various opinions and arguments of David Lyons on the moral philosophy of Mill tend to be broadly consistent with rule utilitarianism, particularly with the “ideal code” variety, Lyons himself never really embraces the connection (1976, 111–119; 1977, 119; 1978, 11), except on occasion and to a limited degree, always with careful qualifications (1976, 115; 1977, 122; 1978, 8, 10).
Implications of Mill’s Proof
327
main subdivisions under the general heading of rule utilitarianism. Among the remaining contributors in the same category, Ben Eggleston and Dale E. Miller introduce “India House utilitarianism” (2007) as a type of rule utilitarianism that is neither of the “actual code” variety nor of the “ideal code” variety; Rex Martin (2007; 2008; 2010) favors “indirect utilitarianism” in much the same way, positioning it against both “actual code” and “ideal code” alternatives; and James Opie Urmson (1953) does not embrace any particular subdivision by name but can reasonably be grouped together with “ideal code” advocates by inclination. Even the subdivisions within a group, however, are not quite sufficient to capture the full extent of the various points of agreement and disagreement among the contributors. Nothing short of a comprehensive survey of the literature can reveal all the relevant issues. While full coverage of each contribution is not a realistic option here, a closer look at a few examples may help expose some of the subtleties in the dialectical foundation of the ongoing debate in the scholarly community. The considered opinions of John Gray and Rex Martin, for example, constitute a perfect illustration of the limitations of the present classification scheme as a heuristic approximation grounded in a mixture of objective facts (in the form of explicit details provided by each author) and subjective impressions (of the sort liable to be formulated by any reader). Their placement in separate groups naturally suggests significant disagreement, but their position on Mill’s moral philosophy is actually much closer than the classification above might seem to indicate. This is because they both pursue indirect utilitarianism as the best exegetical paradigm for Mill’s moral philosophy. How is it, then, that they belong to different groups in the basic division between act utilitarianism and rule utilitarianism? Only a closer inspection of their views can justify their assignment to separate groups. Gray interprets Mill’s position as a type of indirect utilitarianism, which he presents as an alternative to both act utilitarianism and rule utilitarianism (1979; 1983, 9–14, 28–42; 1984). His opposition to either one as the proper interpretation of Mill precedes his development of indirect utilitarianism as an alternative (or at least his employment of the term as a designation of that alternative): “Mill’s principle of utility, like the principle of expediency which it entails, does not mention either acts or rules, and, in fact, applies to things apart from acts and rules. … Mill’s moral theory, in short, is not accurately described in the traditional terms of act-and rule-utilitarianism” (Gray 1979, 14). He subsequently continues to oppose both when he comes to explicate his perspective in terms of indirect utilitarianism:
328
Chapter 8
I shall argue myself for the view that Mill’s position cannot be captured in any modern distinction between ‘act’ and ‘rule’ variants of utilitarianism. Mill is best interpreted as holding to a version of indirect utilitarianism wherein the Principle of Utility cannot have direct application either to individual acts or to social rules because such application is in general, and in many cases necessarily, self-defeating. gray 1983, 12
He summarizes the nature of his opposition as follows: The upshot of my argument so far is that I attribute to Mill a species of utilitarian moral theory, distinct from act-and rule-utilitarianism, which (following some recent writers) I shall call indirect utilitarianism. What are the distinctive features of Mill’s indirect utilitarianism? I suggest they are two: first, that neither the general happiness nor the agent’s own happiness is to be the object of direct pursuit; and second, that utility, in conjunction with its action-guiding corollary, expediency, serves as a principle of evaluation of whole systems of precepts of art, among which moral codes have central (but not exclusive) interest. gray 1983, 38–39
While Gray (1979; 1983; 1984) takes Mill’s indirect utilitarianism to place him outside the division between act and rule utilitarianism, Martin (2007; 2008; 2010) views the very same evidence, or at least various comparable considerations, as placing him somewhere in the camp of rule utilitarianism. Martin’s classification of Mill’s position as rule utilitarianism runs parallel to his consideration of the kind of rule utilitarianism espoused by Mill, as opposed to emerging as the result of any prior effort to negotiate the choice between act and rule utilitarianism as the traditional alternatives. Already convinced that Mill’s approach to moral philosophy is best understood as a type of rule utilitarianism, he proceeds directly with an examination of its precise formulation and classification in that category. His assessment unfolds in a series of essays—“Two Concepts of Rule Utilitarianism: The Case of Mill” (2007), “Two Concepts of Rule Utilitarianism” (2008), and “Mill’s Rule Utilitarianism in Context” (2010)—originating in a paper delivered in 2006 (first before the International Society for Utilitarian Studies in London and subsequently before the Southwest Philosophical Society in Nashville). The common denominator in all three publications, as well as the oral presentations preceding them, is a manifold of criteria sufficiently consistent with indirect utilitarianism to classify Mill’s position as such, despite falling short
Implications of Mill’s Proof
329
of a perfect and comprehensive match across the board. The unifying strategy is an evaluation of the correspondence between relevant passages in Mill and the basic methodology in rule utilitarianism in terms of three variables—(1) “the grounding of rules,” (2) “the allowed complexity of rules,” and (3) “the conflict of rules”—demonstrating that Mill comes out as an indirect utilitarian in connection with the first two indicators and as an “ideal code” type of rule utilitarian with respect to the last one. While Martin compares and contrasts his conception of indirect utilitarianism specifically with the “ideal code” variety of rule utilitarianism (which he calls “ideal-rules utilitarianism”), he does not consider it to be the same therefore as the “actual code” variety of rule utilitarianism (which he calls “actual-rule utilitarianism”). Quite the contrary, he is careful to highlight the difference: It has been suggested that the view I’ve been calling ‘indirect’ utilitarianism is in fact better called ‘actual-rule utilitarianism’. I think not; the rules in the account I’ve offered are social rules (many of them established rules) that are justifiable on utilitarian grounds. They are not simply actual rules; some established rules (perhaps a good number) are not justifiable, and some could never be revised to the point where they are justifiable (e.g., rules condoning ritual human sacrifice or establishing chattel slavery). martin 2008, 238, n. 25; cf. 244–245, including n. 33
Notwithstanding their express agreement on indirect utilitarianism as providing the best insight into Mill’s moral philosophy, Gray and Martin do not reach the same conclusions on where Mill actually belongs in the methodological continuum between act utilitarianism and rule utilitarianism. Gray (1983, 12, 38–9) explicates and defends indirect utilitarianism as an alternative to both act utilitarianism and rule utilitarianism (see also: Gray 1979, 14), whereas Martin (2007; 2008; 2010) construes indirect utilitarianism as a type of rule utilitarianism, though distinguishing it both from the “actual code” variety and from the “ideal code” variety (see especially: Martin 2008, 232, 235–237, 238, n. 25, 244–245, n. 33). Martin, however, would not agree with this comparison. He himself finds closer parallels, perhaps even a perfect match, between his views and those of Gray: “some of the main theorists I’ve been citing (in particular, John Gray and David Lyons) describe their own views as indirect utilitarian and tend to use the term ‘indirect’ so as to accord with the ideal type that I’ve laid out” (Martin 2008, 245, n. 33). A further subtlety in Martin is that, despite presenting indirect utilitarianism as a competing alternative for the “ideal code” variety of rule
330
Chapter 8
utilitarianism, he still considers it a type of “ideal” utilitarianism, just not of the “ideal code” or “ideal rules” variety: “The point is that the two profiles I have outlined constitute ideal types” (Martin 2008, 244). He thus presents himself as “developing this ideal type version of indirect utilitarianism” (Martin 2008, 245, n. 33), which is where he positions Mill’s moral theory, while at the same time emphatically rejecting the standard “ideal code” interpretation of Mill’s utilitarianism. Another example of the subtleties in the proper placement of Mill in the spectrum of positions between act utilitarianism and rule utilitarianism is the work of Ben Eggleston and Dale E. Miller. Both in separate contributions and in collaboration, Eggleston and Miller have exhausted the parameters of investigation, at least in terms of the relevant primary sources, while thereby defining the sphere of interpretation. Their work is thus indispensable, both severally and jointly, for a comprehensive command of the literature and an accurate understanding of the issues. Classifying Mill’s approach to ethics under rule utilitarianism (2007; 2008), specifically under a sophisticated version of their own invention (2007), Eggleston and Miller are careful to emphasize the importance of distinguishing between Mill’s theory of morality and his theory of prudence in any effort to identify and explicate his precise stake in utilitarianism as a school of thought (2008). Their thoughtful distinctions make their classification of Mill as a rule utilitarian especially worthy of attention. To elaborate, Eggleston and Miller find Mill’s moral philosophy consistent with a particular iteration of rule utilitarianism they call “India House utilitarianism” (2007). While their interpretation is indeed a form of rule utilitarianism, it is not just a conceptual alternative to act utilitarianism but also a methodological substitute for both the “actual code” variety and the “ideal code” variety of rule utilitarianism. Their model requires neither slavish obedience to the actual rules in the relevant community (since conventional morality is not entirely a matter of what is right from an objective perspective) nor strict conformity with the ideal rules apposite to the circumstances (since moral agents do not know for a fact what rule or rules would best promote happiness in the long run). They accordingly explicate the central condition of moral action under “India House utilitarianism” as follows: “An act is right if and only if it is not forbidden by the code of rules the agent is justified in believing to be the one, of those she can reasonably be expected to be aware of, whose general acceptance would produce the most happiness” (Eggleston and Miller 2007, 42). This is a distinctive interpretation of rule utilitarianism, differing from most “actual code” renditions in its denial that conventional morality is infallible, given that actual moral codes do not necessarily result in the maximization
Implications of Mill’s Proof
331
of happiness, while differing from typical “ideal code” examples in its refusal to admit any such code as ideal in an absolute or permanent sense, recognizing on the contrary that moral rules and principles are always susceptible of improvement, indefinitely so as Mill himself insists (U2:24 CW10:224). The alternative embraced, not just to one or the other, but to both at once, makes morality a matter of compliance with the prevailing moral rules and principles which, the moral agent is justified in believing, would maximize happiness in the long run, if adopted and observed by everyone, or at least by a sufficiently large portion of the community. This selective emphasis on the morally worthy core of conventional morality, with an eye on further improvements as a standing possibility, points to a dynamic process of moral appraisal, forever open to reconsideration and modification in response to ongoing developments in moral progress, thus representing a reflective equilibrium of sorts through moral enlightenment. The effective outcome is to welcome actual moral codes whose unanimous or widespread observance can reasonably be expected to maximize happiness in the long run, without rejecting what would otherwise qualify as ideal moral codes save for their capacity and prospects for improvement. Expanding on their overall position in a subsequent paper, Eggleston and Miller distinguish their version of rule utilitarianism, that is, their “India House utilitarianism” (2007), which they associate with Mill’s theory of morality, from a hybrid version which they associate with Mill’s theory of prudence (2008). To be more specific, they construe Mill as a rule utilitarian in his theory of morality (2007) but as a tendency utilitarian in his theory of prudence (2008), grounding the former in Mill’s association of the principle of moral distinction (between right and wrong) with binding moral rules instituted and sustained in accordance with social expediency, while defining the latter as a combination of act and rule utilitarianism where the central element is the “tendency of an action” calculated as a function of both its “particular consequences” and its “type consequences.” This useful distinction further emphasizes the difficulty of specifying just what kind of utilitarian Mill may reasonably be said to be. He is, according to Eggleston and Miller, a rule utilitarian of a highly specialized kind in one respect (his theory of morality), qualifying as an even more intricate kind of utilitarian in yet another respect (his theory of prudence), where he draws on the characteristics of both act utilitarianism and rule utilitarianism, combining the specific consequences of each token act with the expected consequences of the type of act that it is (cf. 2008, 157–158). Arguably the greatest testament to the plasticity of the various definitions and distinctions emerging in connection with the difference between act utilitarianism and rule utilitarianism is Ben Eggleston’s (2017) later identification of
332
Chapter 8
their brand of rule utilitarianism, namely “India House utilitarianism” (2007), with a version of the “actual code” variety of rule utilitarianism. This association comes in a broad overview of Mill’s moral philosophy in a contribution to a companion to Mill, where Eggleston cites the paradigm of “India House utilitarianism,” introduced in his original collaboration with Miller (2007), as an example of “views resembling” actual‐code rule utilitarianism (2017, 365). While this may appear to contradict the original article, where the position in question is clearly presented as an alternative both to the “actual code” variety and to the “ideal code” variety of rule utilitarianism, the contradiction is merely apparent and not real. This is because the actual‐code version of rule utilitarianism Eggleston (2017) claims to be resembled by “India House utilitarianism” is not the basic “actual code” variety indiscriminately embracing all the rules and principles prevailing in conventional morality but a modified version adopting only a subset of such rules and principles coinciding with the rational expectations of competent moral judges for the maximization of happiness: “This view, which can be called actual‐code rule utilitarianism, implies that the rightness or wrongness of an act depends on actual rules rather than ideal rules, but not all actual rules have moral force—only those that promote happiness do” (Eggleston 2017, 365). This is such a refined and morally enlightened version of actual‐code rule utilitarianism that the suggested resemblance is not only accurate but also consistent with the original presentation in opposition to the standard conception, or basic version, of the “actual code” variety of rule utilitarianism (Eggleston and Miller 2007). Rule utilitarianism is not the only category with meaningful differences in interpretation. The category of act utilitarianism shows just as much differentiation and no less variation, while the neutral category favoring neither act nor rule utilitarianism shows barely any uniformity outside the common rejection of those two options. The collective track record in interpretation and classification thus stands as an impressive display of scholarship and creativity. Yet it is inherently and demonstrably closer to art than to science in the continuum defining the evidentiary context relevant to the moral philosophy of Mill. The reason why we have so many apparently reasonable interpretations in mutual opposition is that the body of Mill’s published writings is replete with apparently reasonable statements in mutual opposition. Any conflict in primary sources is difficult enough to sort out as it is, without the additional burden of having to determine which of the two approaches coming later presents a better fit with the original, whereupon the task becomes all the more demanding, and the measure of success that much more nebulous. This is especially so where the later inventions under consideration are not the only alternatives available, making it unnecessary to follow either approach but
Implications of Mill’s Proof
333
tempting to embrace one or the other. These problems are complicated further where some of the primary sources both favor and oppose the later developments, depending largely on the work consulted and the passage quoted, with a few of them even contradicting both perspectives at once, as illustrated to a certain extent above. Contextual uncertainties support the reception of Mill as just about any kind of utilitarian, ranging from an act utilitarian of the maximizing or satisficing variety, to a rule utilitarian of the actual-code or ideal-code variety, if not including something altogether different with an equally exotic label. All such portrayals are grounded in developments representing an anachronistic and esoteric perspective relative to the original. There is, of course, nothing wrong with anachronism where emerging paradigms help elucidate past contributions. The danger, rather, is in setting out to demonstrate that a particular contribution is a perfect match for a later paradigm, which is all too easy to demonstrate, and nearly impossible to falsify, where the contribution is open to interpretation and the contributor is forever silent.32 That seems to be the driving force behind the controversy over whether Mill’s moral theory is best interpreted in terms of act utilitarianism or in terms of rule utilitarianism. It would be difficult to summarize recent interpretations and prevailing classifications any better than Samuel Scolnicov does in his following assessment: I’ve heard old wizards say that arguably the principal requisite for successful tea-leaf reading is having a rough idea of what you want to find. As in other cases of circular hermeneutics, here too you can do little more than point at a pattern which seems to you close to your subject and cry out, ‘Look: can you see it?’ Other patterns are discernible, of course, some more relevant, some less, and any diviner must be thankful to whomever complains about his cryptic utterances, which he idiosyncratically took for evident, or opens his eyes to see what he missed while being too engrossed in his own concerns. This is not to imply that anything goes; even tea leaves arrange themselves so that in any given cup some 32
Relevant sources on methodological propriety in research and analysis in the history of philosophy may well be inexhaustible, even where the scope of consideration is limited strictly to sources on formulating hypotheses and drawing conclusions. A few of the particularly valuable contributions include John Dunn’s “The Identity of the History of Ideas” (1968), Quentin Skinner’s “Meaning and Understanding in the History of Ideas” (1969), and Adrian Blau’s “Extended Meaning and Understanding in the History of Ideas” (2019). Douglas G. Long’s “ ‘Utility’ and the ‘Utility Principle’: Hume, Smith, Bentham, Mill” (1990) constitutes a good example of a study of Mill, among others, with an approach that is sensitive to methodological constraints and compliant with corresponding guidelines.
334
Chapter 8
patterns are just not there. And scholarly disagreement over which are and which are not is what keeps the old mantic art alive. scolnicov 1994, 149
This appraisal by Scolnicov actually has nothing to do with Mill. It is specifically about a point of contention in Plato, but it happens to apply equally well to the matter in hand. We need not go as far back as Plato to witness the art of interpretation guided largely by preconception. That approach is not peculiar to the study of any particular philosopher, but it has become something of a standard procedure with respect to Mill, most conspicuously in the matter of act utilitarianism versus rule utilitarianism, but also in association with his moral philosophy in general. 8.2.2 Actual, Intended, and Foreseeable Consequences The actual, intended, and foreseeable consequences of actions play an important role in the context of moral evaluation. At a very basic level, they can be distinguished roughly as follows: (1) Actual consequences are the states of affairs that moral agents produce as a result of their deliberate actions; (2) intended consequences are the states of affairs that moral agents want, aim, and expect to produce as a result of their deliberate actions; (3) foreseeable consequences are the states of affairs that moral agents would and should be able to predict, strictly in their capacity as ordinary people with common sense and typical experiences, as the likely result of their deliberate actions. Mill holds that the morality of actions is a function of their intended and foreseeable consequences, not of their actual consequences. While he is sensitive to the difference between actual consequences, on the one hand, and intended and foreseeable consequences, on the other, he does not find a morally significant difference between intended and foreseeable consequences, because he believes that intentions are relevant to moral evaluation only insofar as they are developed with rational foresight of their consequences. He conflates intended and foreseeable consequences in order to preclude superficial moral excuses to ignore foreseeable consequences during the development of intentions. Mill prefers to speak of tendencies rather than consequences wherever he discusses actions from a methodological perspective. He explains the motivation for this preference in the following passage in A System of Logic, where he claims that any descriptive account of cause-and-effect relations in scientific explanation should be expressed in terms of the tendencies of actions rather than in those of their actual consequences:
Implications of Mill’s Proof
335
These facts are correctly indicated by the expression tendency. All laws of causation, in consequence of their liability to be counteracted, require to be stated in words affirmative of tendencies only, and not of actual results. In those sciences of causation which have an accurate nomenclature, there are special words which signify a tendency to the particular effect with which the science is conversant; thus pressure, in mechanics, is synonymous with tendency to motion, and forces are not reasoned on as causing actual motion, but as exerting pressure. A similar improvement in terminology would be very salutary in many other branches of science. mill L3:10.5 CW7:445
A similar terminological improvement indeed finds its way into Mill’s ethical theory in one of the most widely cited passages in Utilitarianism, already quoted in the previous subsection in association with the matter of act utilitarianism versus rule utilitarianism (in addition to its various other appearances throughout the present volume): The creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle, holds that actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness. mill U2:2 CW10:210
The repeated emphasis here on tendencies suggests that Mill is not concerned with actual consequences in moral evaluation. He does, however, hold that actual consequences are foreseeable unless they are accidental. The following passage in “Sedgwick’s Discourse” (1835), for example, shows that he refuses to accept excuses for failures to account for actual consequences unless they happen to be accidental consequences: Who ever said that it was necessary to foresee all the consequences of each individual action, “as they go down into the countless ages of coming time?” Some of the consequences of an action are accidental; others are its natural result, according to the known laws of the universe. The former, for the most part, cannot be foreseen; but the whole course of human life is founded upon the fact that the latter can. mill CW10:63
336
Chapter 8
The reason why Mill does not draw a moral distinction between intended and foreseeable consequences is that he considers the morally significant intentions behind actions to be those that are developed with rational expectations and common experiences, and thereby with some awareness of the tendencies of actions and with practical foresight of their possible consequences. “The Works of Jeremy Bentham” (1838), for example, shows Mill associating the morality of actions both with their tendencies and with their foreseeable consequences: “The morality of actions depends on the consequences which they tend to produce” (CW10:111); “The morality of an action depends on its foreseeable consequences” (CW10:112). While he does not hold moral agents responsible for the actual consequences of their actions unless they are intended, he also does not excuse them from having to consider the tendencies of actions in developing their intentions. He does not require a perfect match between actual and intended consequences, but he holds moral agents to a certain standard of foresight in their intentions, thus minimizing the region of moral excuse manifested as a discrepancy between consequences that actions tend to produce and consequences that agents intend to produce. Mill’s best known discussion of the moral significance of intentions is probably the famous footnote in the second chapter of Utilitarianism where he distinguishes between intentions and motives: The morality of the action depends entirely upon the intention—that is, upon what the agent wills to do. But the motive, that is, the feeling which makes him will so to do, when it makes no difference in the act, makes none in the morality: though it makes a great difference in our moral estimation of the agent, especially if it indicates a good or a bad habitual disposition—a bent of character from which useful, or from which hurtful actions are likely to arise. mill U2:19n CW10:220; editor’s bracketed insertions omitted
This passage ties the morality of actions to the intentions underlying actions, and the morality of agents to the motives underlying intentions, but it does not go very far toward linking the intended and foreseeable consequences of actions. Mill establishes the latter link in his editorial footnote to chapter twenty-five of the second volume of his father’s Analysis of the Phenomena of the Human Mind (1869): Intention, when we are said to intend the consequences of our actions, means the foresight, or expectation of those consequences; which is a totally different thing from desiring them. The particular consequences
Implications of Mill’s Proof
337
in question, though foreseen may be disagreeable to us: the act may be done for the sake of other consequences. Intention, and motive, are two very different things. But it is the intention, that is, the foresight of consequences, which constitutes the moral rightness or wrongness of the act. james mill 1869, vol. 2, 401, n. 69
From another perspective, which is at best implicit in Mill’s discussion, the difference between actual consequences, on the one hand, and intended and foreseeable consequences, on the other, is that actual consequences are relevant to the moral evaluation of past actions for future reference, whereas intended and foreseeable consequences are relevant to the moral evaluation of actions currently being contemplated for future performance (U2:24 CW10:224–225). In terms of this distinction, the principle of utility helps both retrospective and prospective moral evaluations of actions. The retrospective evaluation of actions concerns the utility of actual consequences, confirming in hindsight whether a particular action that has been performed really ought to have been performed after all. The prospective evaluation of actions concerns the utility of intended and foreseeable consequences, indicating whether a particular action that has not yet been performed indeed ought to be performed. More specifically, the retrospective evaluation of actions involves assessing the utility of the actual consequences of those actions, and helps determine the tendency of actions of a certain class or type to produce consequences of a certain class or type, which, in turn, helps develop rational expectations and moral guidelines for various practical situations. The rational expectations and moral guidelines enable moral agents to determine the morality of actions they are contemplating for imminent performance, without having to appeal directly to the principle of utility in each case. In this way, the retrospective evaluation of actions in terms of their actual consequences assists the prospective evaluation of actions. 8.2.3 Total and Average Happiness Given that happiness is a function of pleasure and pain, and given further that pleasures and pains are fungible, quantifiable, and inversely correlated, the amount of happiness in a population can be measured in a number of different ways. Two popular measures are total happiness and average happiness. Total happiness is the sum of the pleasures and pains of everyone in the relevant population. Average happiness is the arithmetic mean, that is, the sum of the pleasures and pains of everyone divided by the number of persons. The question, then, is whether Mill advocates the promotion of total happiness or the promotion of average happiness. This is related to the question whether
338
Chapter 8
his conception of the general happiness is best interpreted in terms of total happiness or average happiness. The questions are superficial, not just because Mill does not express a preference between these two measures of the amount of happiness, nor just because the general happiness is not a quantitative measure but a qualitative concept, but also because actions and practices almost always affect total and average happiness in the same direction, and therefore because total and average happiness need not be taken as competing or mutually exclusive measures of a population’s happiness content, the promotion or maximization of which can be measured more reliably by monitoring changes in both than it can by focusing on one measure to the exclusion of the other. The common rationale behind distinguishing between total and average happiness is that the two measures tend to reflect different perspectives of the amount of happiness in a population, population sample, or possible world. Here are three examples: First, compare a world in which ten people each enjoy ten units of happiness with a world in which fifty people each enjoy ten units of happiness. Total happiness is greater in the more populous world, while average happiness is the same in both worlds. Second, compare a world in which ten people each enjoy ten units of happiness with a world in which one person enjoys a hundred units of happiness. Total happiness is the same in both worlds, while average happiness is greater in the less populous world. Third, compare a world in which ten people each enjoy ten units of happiness with a world in which a hundred people each enjoy one unit of happiness. Total happiness is the same in both worlds, while average happiness is greater in the less populous world. The point of such examples is to invoke our moral sensibilities in evaluating and comparing implications of the promotion of happiness, ostensibly revealing whether it is better to focus on the total or on the average. In the first example, the promotion of total happiness favors the more populous world, whereas the promotion of average happiness is indifferent between the two worlds. Our moral sensibilities in that scenario presumably favor the more populous world. They are thereby in line with the kind of utilitarianism that espouses the promotion of total happiness (cf. Smart and Williams 1973, 27–28). In each of the latter two examples, the promotion of total happiness is indifferent between the two worlds, whereas the promotion of average happiness favors the less populous world. It is not clear what our moral sensibilities are supposed to indicate in these two examples. The math suggests that the less populous world is better than the more populous world, given that average happiness is greater in the less populous world, while total happiness is the same in both
Implications of Mill’s Proof
339
worlds. But felicific calculus alone is not enough to decide whether it is better to have a world with more people or fewer people. The salient difference between the two worlds in each of the three examples is the number of people in the population. It seems that the basis of our moral sensibilities in all three cases is that people are valuable in themselves, and therefore that it is intrinsically better that people should exist than that they should not exist, and that, within reasonable limits, it is better that more people should exist than that fewer people should exist. If the number of people is indeed the relevant point of contrast, then the primary benefit of studying the distinction between the two measures of happiness is the corresponding opportunity to evaluate their different implications for population control. Smart, for one, speculates that the distinction between total and average happiness might have important implications for the ethics of birth control, but he does not elaborate on this, except to suggest that the promotion of total happiness does not lend itself to arguments for birth control as readily as does the promotion of average happiness (Smart and Williams 1973, 28). Brandt likewise prefers average over total happiness in his version of utilitarianism because he believes that total happiness has embarrassing consequences for problems in population control (1992, 125, n. 11). As a matter of fact, both measures of happiness have embarrassing implications for population control. Smart and Brandt are probably thinking about a basic contrast between the promotion of total happiness, which favors reproduction insofar as it is reasonable to expect future generations to be at least minimally happy, and the promotion of average happiness, which favors birth control insofar as it is reasonable to expect future generations to be no happier than the present population. However, on any interpretation beyond this simple scenario, both the promotion of total happiness and the promotion of average happiness have absurd implications for the morality of population control policies. The promotion of total happiness requires reproduction and prohibits birth control, while the promotion of average happiness requires birth control and prohibits reproduction. First, if future generations can be expected to be at least minimally happy, then everything else remaining the same, it is not just that the promotion of total happiness requires reproduction and prohibits birth control but that it requires people to reproduce as frequently as possible, which means that it requires women to be pregnant at all times, except during the time between one birth and the next conception. Second, if future generations can be expected to be no happier than the present population, again with everything else remaining the same, the promotion of average happiness prohibits reproduction and requires birth control, especially favoring vasectomies, hysterectomies, and abortions, and probably holding at least one
340
Chapter 8
of those measures to be categorically and universally mandatory. This holds unless each birth can reasonably or perhaps definitively be expected to produce a person whose happiness exceeds the average happiness in the world prior to that birth. Moreover, both the promotion of total happiness and the promotion of average happiness, on a strict enough interpretation, support an unorthodox method of population control: killing people. The death of an unhappy person eliminates more pains than pleasures in the constitution of aggregate happiness in a population, which is composed of the pleasures and pains of every person in the population. The death of an unhappy person, therefore, if everything else remains the same, raises both total happiness and average happiness. It raises total happiness because the amount of pleasure lost upon the death of an unhappy person is less than the amount of pain lost upon the death of that person. It raises average happiness for the very same reason plus the reduction in the number of people representing the divisor. As for happy people, each time one of them dies, total happiness falls, whereas average happiness can rise, unless the dead person used to enjoy a level of happiness exceeding the average happiness in the world prior to that death. These relationships between the death of persons and the amount of happiness in a population have monstrous implications both for the promotion of total happiness and for the promotion of average happiness, but they have a greater number of monstrous implications for the promotion of average happiness. First, on the presupposition that utilitarianism condones killing people for an increase in overall happiness, and ignoring all the relevant and inevitable psychological and sociological effects of such killing, both the promotion of total happiness and the promotion of average happiness favor killing all people who are unhappy (even if everyone in the population is unhappy). Second, with the same restrictions and qualifications, the promotion of total happiness discourages killing anyone who is happy, whereas the promotion of average happiness favors killing everyone except the happiest person in the world, or leaving room for ties in that title, it favors killing everyone except the happiest persons in the world. To be sure, neither the promotion of total happiness nor the promotion of average happiness requires such monstrous measures in population control except in thought experiments comparing total and average happiness independently of any and all variables that would otherwise affect the circumstances of their comparison. Extraordinary presuppositions, such as the fantastic assumption that utilitarianism would condone killing people for an increase in overall happiness, are helpful in isolating differences between total and average happiness that are inherent in the two concepts from differences
Implications of Mill’s Proof
341
that are contingent in them. The point is not that the promotion of total or average happiness would in fact require such absurd measures in population control but that one is hardly better than the other in terms of their implications for population control. Mill does not express a preference between total and average happiness as competing measures of the amount of happiness in a population. Yet given that almost any remark on the amount of happiness in a population, even if the remark is intended to be neutral between total and average happiness, lends itself more readily to an interpretation in terms of total happiness than to one in terms of average happiness, Mill might seem to be leaning toward total happiness, since his typical remark is something like “the greatest amount of happiness altogether” (U2:9 CW10:213) or “the sum total of happiness” (U2:17 CW10:218). This does not mean, however, that he is blind to the distribution of happiness. He affirms that the principle of utility “is a mere form of words without rational signification, unless one person’s happiness, supposed equal in degree (with the proper allowance made for kind), is counted for exactly as much as another’s,” insisting emphatically that “those conditions being supplied, Bentham’s dictum, ‘everybody to count for one, nobody for more than one,’ might be written under the principle of utility as an explanatory commentary” (U5:36 CW10:257). This position, namely that each person’s happiness counts just as much as another’s, confirms conclusively that Mill is sensitive to the distribution of happiness. This egalitarian outlook, on the other hand, is not evidence of a commitment to the opposite extreme, where existing happiness is redistributed until it is equally distributed, after which any additional happiness in the future is equally distributed no matter the circumstances. An equal right to happiness is not the same as a right to equal happiness. Mill maintains that a reliable estimate of the effects of actions or practices on the amount of happiness in a population should account for the happiness of everyone involved in the situation. His “Utility of Religion” (CW10:403–428), in his Three Essays on Religion (1874), describes utilitarianism as a “morality grounded on large and wise views of the good of the whole, neither sacrificing the individual to the aggregate nor the aggregate to the individual” (CW10:421). Moreover, even if it were possible to associate Mill’s moral theory with a perspective favoring total happiness over average happiness, this would not commit Mill to any particular position concerning the distribution of happiness. The fact that total and average happiness represent different perspectives of the amount of happiness in any given population does not mean that the two measures reflect different patterns of change in the happiness of any given population. Changes in total and average happiness are mathematically
342
Chapter 8
correlated because the average happiness of a population is just the total happiness of the population divided by the number of persons in that population. Raising total happiness is therefore an effective way of raising average happiness, and raising average happiness is an effective way of raising total happiness. It is sometimes thought that the promotion of total happiness is blind to the distribution of happiness, whereas the promotion of average happiness is sensitive to the distribution of happiness, and therefore that the promotion of average happiness is a more equitable goal than the promotion of total happiness. However, given that average happiness is nothing other than total happiness divided by the number of persons, changes in a population’s happiness are reflected in the same way in both total and average happiness, unless the changes include a change in the number of people in the population. Whatever affects total happiness affects average happiness in the same direction, except when counteracted by an independent and opposing change in the number of people in the population. Whatever raises total happiness, everything else remaining the same, also raises average happiness, and whatever lowers total happiness, given the same caveat, also lowers average happiness. This includes a change in the distribution of happiness. For example, compare a world of ten persons each of whom enjoys ten units of happiness with a world of ten persons one of whom enjoys a hundred units of happiness while nine enjoy none. The two worlds, in this case, have the same total happiness, as well as the same average happiness, but our moral sensibilities suggest that the world with the even distribution of happiness is better than the world with the uneven distribution of happiness, if only because an even distribution seems more just than an uneven distribution, ignoring otherwise relevant differences in need and desert. The distinction between total and average happiness is irrelevant in this case because neither total happiness nor average happiness measures the distribution of happiness. In a fixed population, a rise or fall in total happiness raises or lowers average happiness in the same direction, and a rise or fall in average happiness raises or lowers total happiness in the same direction. A redistribution of happiness is not reflected in average happiness any more than it is reflected in total happiness. First, redistributing existing happiness cannot alone affect total happiness, unless the redistribution itself becomes a new source of happiness, or of unhappiness, unavailable prior to the redistribution of existing happiness, in which case both total and average happiness would rise or fall in tandem. Second, redistributing existing happiness cannot alone affect average happiness, because the redistribution itself cannot affect the number of people in the population, unless the redistribution makes some people so miserable as
Implications of Mill’s Proof
343
to kill themselves, in which case both total and average happiness would rise. To be sure, redistributing existing happiness might affect the population indirectly by inspiring people either to reproduce or to kill people, including themselves, but barring such developments, the redistribution itself cannot affect either total or average happiness. The question is whether it is better to promote (or maximize) total happiness or to promote (or maximize) average happiness. However, given that felicific calculations are possible, and given that the goal is to promote (or maximize) the amount of happiness in a population, several other measures are available and just as useful as total and average happiness. For example, if total happiness is indeed an inadequate measure of a population’s happiness, which is better measured, instead or in addition, by average happiness, then there is all the more reason to consider all measures of central tendency (e.g., mean, median, mode) and all measures of dispersion (e.g., range, variance, standard deviation) in the distribution of happiness in a population. The most common measures of central tendency in a population distribution are the mean, median, and mode. The mean happiness in a population, or the arithmetic mean of a population distribution with respect to happiness, is the sum of every person’s happiness divided by the number of persons in the population. The median happiness in a population, or the median of a population distribution with respect to happiness, is the midpoint which divides the population into two halves in terms of each person’s happiness, such that half the people in the population enjoy a level of personal happiness higher than the median, while the other half enjoy a level of personal happiness lower than the median, which can be interpolated if necessary. The mode happiness in a population, or the mode of a population distribution with respect to happiness, is the level of personal happiness that is experienced by the greatest number of people in the population, which means that a population’s distribution of happiness can be bimodal, or trimodal, and so on, up to the point where everyone is equally happy, in which case there is said to be no mode. The most common measures of dispersion in a population distribution are the range, variance, and standard deviation. The range of a population’s distribution of happiness is the difference between the greatest personal happiness experienced by anyone in the population and the greatest personal unhappiness experienced by anyone in the population. The variance of a population’s distribution of happiness is the sum of the squares of the differences between each person’s happiness and the mean happiness, divided by the number of persons in the population. The standard deviation of a population’s distribution of happiness is the square root of the variance of that population’s distribution of happiness.
344
Chapter 8
Measures of central tendency can help monitor changes in the amount of happiness in a population. Actions that tend to increase the amount of happiness tend to raise the mean, median, and mode of the distribution of happiness. Actions that tend to decrease the amount of happiness tend to lower the mean, median, and mode of the distribution of happiness. Measures of dispersion can help monitor changes in the distribution of happiness in a population. Actions that tend to distribute happiness evenly across a population tend to lower the range, variance, and standard deviation of the distribution of happiness. Actions that tend to distribute happiness unevenly across a population tend to raise the range, variance, and standard deviation of the distribution of happiness. Hence, if the goal is to promote the growth of happiness, while simultaneously controlling (either changing or preserving) the distribution of happiness, progress can be assessed by monitoring changes in three avenues of measurement: changes in the total happiness in the relevant population; changes in the central tendency indicators of the distribution of happiness; changes in the dispersion indicators of the distribution of happiness. Otherwise, the importance of the distinction between total and average happiness is highly overrated because, for all intents and purposes, changes in the two measures run parallel to each other. Insofar as the goal of a commercial enterprise is to maximize profits, corporate net income and earnings per share are equally reliable measures of progress. Likewise, if the goal of a human being, or moral agent, is to promote happiness, total happiness and average happiness are equally reliable measures of progress. As for its equitable distribution, neither one is relevant to progress toward that end, which is still a qualitative matter at the mercy of human interaction, just as it was back in Mill’s day.
Works Cited Abaté, Charles J. 1979. “Fallaciousness and Invalidity.” Philosophy & Rhetoric 12:4 (Fall): 262–266. Adams, F. 1897. “The Greatest Happiness of the Greatest Number.” Notes and Queries S8:xi 281 (15 May 1897): 392. Albee, Ernest. 1902. A History of English Utilitarianism. London: Swan Sonnenschein. Alican, Necip Fikri. 2021. Quine on Ethics: The Gavagai of Moral Discourse. Newcastle upon Tyne: Cambridge Scholars Publishing. Anderson, Elizabeth S. 1991. “John Stuart Mill and Experiments in Living.” Ethics 102:1 (October): 4–26. Aristotle. 1831–1870. Aristotelis Opera. Five volumes. Greek text of the complete works of Aristotle, prepared by the Prussian Academy of Sciences, under the general editorship of Immanuel Bekker. Berlin: Georg Reimer (De Gruyter). Aristotle. 1984. The Complete Works of Aristotle: The Revised Oxford Translation. Two volumes. Edited by Jonathan Barnes. Princeton: Princeton University Press. Arrhenius, Gustaf, and Wlodek Rabinowicz. 2005. “Millian Superiorities.” Utilitas 17:2 (July): 127–146. Atkinson, Ronald Field. 1957. “J. S. Mill’s ‘Proof’ of the Principle of Utility.” Philosophy 32:121 (April): 158–167. Bacon, Francis. 1605. The Advancement of Learning. Originally published as The Tvvoo Bookes of Francis Bacon: Of the Proficience and Aduancement of Learning, Divine and Humane. London: Henrie Tomes. Bacon, Francis. 1620. Novum Organum. Originally published as the second of six parts of the author’s Instauratio Magna. London (Londini): Joannem Billium Typographum Regium, 1620. The host volume subsequently came to be known as Novum Organum, which was the only part of the whole to be fleshed out in the original, the rest remaining in outline form. The second edition was published as Novum Organum Scientiarum. Leiden (Lvgdvni Batavorvm): Adrianum Wijngaerde and Franciscum Moiardum, 1645. Baker, Gordon Park, and Peter Michael Stephan Hacker. 1966. “Rules, Definitions, and the Naturalistic Fallacy.” American Philosophical Quarterly 3:4 (October): 299–305. Baker, John M. 1971. “Utilitarianism and ‘Secondary Principles.’ ” The Philosophical Quarterly 21:82 (January): 69–71. Baker, John. M. 1980. “Mill’s Captivating ‘Proof’ and the Foundations of Ethics.” Social Theory and Practice 6:3 (Fall): 299–309. Baldwin, Thomas. 1990. “The Naturalistic Fallacy.” Chapter 3 of his G. E. Moore (pp. 66–110). New York: Routledge, Chapman, and Hall.
346
Works Cited
Ball, Terence. 2000. “The Formation of Character: Mill’s ‘Ethology’ Reconsidered.” Polity 33:1 (Autumn): 25–48. Bar-Hillel, Yehoshua. 1964. “More on the Fallacy of Composition.” Mind (New Series) 73:289 (January): 125–126. Barrow, Robin. 1991. Utilitarianism: A Contemporary Statement. Hants: Edward Elgar Publishing. Baumrin, Bernard H. 1968. “Is There a Naturalistic Fallacy?” American Philosophical Quarterly 5:2 (April): 79–89. Beaumont, William Comyns. 1897. “The Greatest Happiness of the Greatest Number.” Notes and Queries S8:xi 281 (15 May 1897): 392. Beccaria, Cesare (Cesare Bonesana di Beccaria, Marchese di Gualdrasco e di Villareggio). 1764/1767. Dei Delitti e delle Pene [On Crimes and Punishments]. Livorno: Marco Coltellini. Published anonymously in 1764. Numerous editions and reprints. Widely translated. First English translation was an anonymous one, An Essay on Crimes and Punishments (London: John Almon, 1767), published together with the translation of an anonymous French commentary by Voltaire (“par un Avocat du Province”): “Commentaire sur le Livre des Délits et des Peines” (Geneva: Frères Cramer, 1766), the latter of which also came to be included in French editions of Beccaria. Bentham, Jeremy. 1776. A Fragment on Government. Published anonymously as A Fragment on Government; Being an Examination of What Is Delivered, on the Subject of Government in General, in the Introduction to Sir William Blackstone’s Commentaries: With a Preface, in Which Is Given a Critique on the Work at Large. London: John Thomas Payne, 1776. Second edition (“second edition, enlarged”) published in the author’s own name as A Fragment on Government; or, A Comment on the Commentaries: Being an Examination of What Is Delivered on the Subject of Government in General, in the Introduction to Sir William Blackstone’s Commentaries: With a Preface, in Which Is Given a Critique on the Work at Large. London: Effingham William Wilson and William Pickering, 1823. Reprinted as part of the first volume of The Works of Jeremy Bentham (pp. 221–295), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty- two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Reprinted with the title A Comment on the Commentaries and A Fragment on Government, as part of The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Edited by James Henderson Burns and Herbert Lionel Adolphus Hart. London: University of London, Athlone Press, 1977. The Collected Works edition was later reissued in Oxford: Clarendon Press, 2008. Bentham, Jeremy. 1781–1785. Commonplace Book. A notebook of brief and random notes for future use and reference, in essence, a scrapbook for recording ideas and
Works Cited
347
information of interest to the author, a medium particularly popular in the nineteenth century, though common throughout history. The one cited here is actually an indeterminate number of such notebooks, screened for material of interest by John Bowring, Bentham’s friend, editor, and literary executor, who released such miscellanea under the heading “Extracts from Commonplace Book” in dated clusters broken down in accordance with subject matter. The 1781–1785 cluster constitutes part of the sixth chapter of the tenth volume of The Works of Jeremy Bentham (pp. 141–147), the first compilation of Bentham’s writings. The fourth chapter of the same volume contains the 1770–1780 cluster in two parts: pp. 72–77 covering the 1774–1775 period; pp. 84–85 corresponding to the year 1776. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Bentham, Jeremy. 1789. An Introduction to the Principles of Morals and Legislation. Printed privately in 1780 and published nine years later in London: John Thomas Payne and Son, 1789. Second edition (“a new edition, corrected by the author”) published in two volumes in London: William Pickering and Effingham William Wilson, 1823. The first volume contains chapters 1–12, the second volume contains chapters 13–17. Reprinted as part of the first volume of The Works of Jeremy Bentham (pp. 1–154), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Reprinted as part of The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Edited by James Henderson Burns and Herbert Lionel Adolphus Hart. London: University of London, Athlone Press, 1970. The Collected Works edition was reprinted in paperback, with a new introduction by Frederick Rosen and an interpretive essay by Herbert Lionel Adolphus Hart, in Oxford: Clarendon Press, 1996. Bentham, Jeremy. 1816. Chrestomathia. Originally published as Chrestomathia: Being a Collection of Papers, Explanatory of the Design of an Institution, Proposed to Be Set on Foot, Under the Name of the Chrestomathic Day School, or Chrestomathic School, for the Extension of the New System of Instruction to the Higher Branches of Learning, for the Use of the Middling and Higher Ranks in Life. London: John Thomas Payne and Henry Foss (Pall-Mall) and Rowland Hunter (St. Paul’s Church-Yard), 1816. Reprinted as part of the eighth volume of The Works of Jeremy Bentham (pp. 1–191), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Reprinted as a standalone volume in The Collected Works of Jeremy Bentham, the
348
Works Cited
standard critical edition of Bentham’s writings. Edited by Martin John Smith and Wyndham Headley Burston. Oxford: Clarendon Press, 1984. Bentham, Jeremy. 1817/1983. A Table of the Springs of Action. Portions printed in 1815 (London: Richard and Arthur Taylor) and published in 1817 (London: Rowland Hunter) with the title: A Table of the Springs of Action: Shewing the Several Species of Pleasures and Pains, of Which Man’s Nature Is Susceptible: Together with the Several Species of Interests, Desires, and Motives, Respectively Corresponding to Them: And the Several Sets of Appellatives, Neutral, Eulogistic and Dyslogistic, by Which Each Species of Motive Is Wont to Be Designated: To Which Are Added Explanatory Notes and Observations. Recovered from the original manuscript and reprinted as part of Deontology, Together with A Table of the Springs of Action and the Article on Utilitarianism (Oxford: Clarendon Press, 1983), a compilation edited by Amnon Goldworth and published as a volume in The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Bentham, Jeremy. 1820–1832. Constitutional Code. Planned as three volumes, and drafted between 1820 and 1832, the work remained incomplete upon Bentham’s death in 1832. The first volume was printed in 1827, and published in 1830, with the title Constitutional Code; for the Use of All Nations and All Governments Professing Liberal Opinions. London: Robert Heward (“Printed for the author, and published by Robert Heward, at the office of The Westminster Review”), 1830. The volume was subsequently combined with additional material, originally intended by Bentham as an “Introductory Dissertation” preceding the main text, but posthumously organized instead as Book 1 of the work, followed by the original publication as Book 2, all edited by Ricard Doane from Bentham’s manuscripts, and published together as vol. 9 of The Works of Jeremy Bentham (pp. 221–295), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. The original version was reprinted as part of The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Edited by Frederick Rosen and James Henderson Burns. Oxford: Clarendon Press, 1983. Bentham, Jeremy. 1821. On the Liberty of the Press, and Public Discussion. London: William Hone, 1821. A tract consisting of four letters from Bentham to the Spanish people (“Spaniards!”) encouraging them to support the freedom of the press, particularly against recent pressure from the government and the judiciary, and to resist a draft law intended to curtail certain civil liberties, including that of public assembly and political rally. The first letter (7 October 1820) is on the freedom of the press (“On the Liberty of the Press—the Approaching Eight Months’ Sleep of the Cortes— and the Exclusion of Experience from the Succeeding Cortes”), the other three (18 October 1820), on the right to assembly (“On the Liberty of Public Discussion in
Works Cited
349
Free Meetings”). Reprinted as part of the second volume of The Works of Jeremy Bentham (pp. 275–297), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Reprinted with the title On the Liberty of the Press, and Public Discussion, and Other Legal and Political Writings for Spain and Portugal, as part of The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Edited by Catherine Pease-Watkin and Philip Schofield. Oxford: Clarendon Press, 2012. Bentham, Jeremy. 1822. Codification Proposal, Addressed by Jeremy Bentham to All Nations Professing Liberal Opinions; or, Idea of a Proposed All-Comprehensive Body of Law, with an Accompaniment of Reasons, Applying All along to the Several Proposed Arrangements: These Reasons Being Expressive of the Considerations, by Which the Several Arrangements Have Been Presented, As Being, in a Higher Degree Than Any Other, Conducive to the Greatest Happiness of the Greatest Number, of the Individuals of Whom the Community in Question Is Composed: Including Observations Respecting the Hands, by Which the Original Draught, of Work of the Sort in Question, May, with Most Advantage, Be Composed: Also, Intimation, from the Author, to the Competent Authorities in the Several Nations and Political States, Expressive of His Desire and Readiness to Draw Up, for Their Use Respectively, the Original Draught of a Body of Law, Such As the above Proposed. London: John McCreery. Reprinted as part of the fourth volume of The Works of Jeremy Bentham (pp. 535–594), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Reprinted as part of Legislator of the World: Writings on Codification, Law, and Education, a volume in The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Edited by Philip Schofield and Jonathan Harris. Oxford: Clarendon Press, 1999. Bentham, Jeremy. 1825. The Rationale of Reward. London: John and H. L. Hunt. Reprinted as part of the second volume of The Works of Jeremy Bentham (pp. 189–266), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty- two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Bentham, Jeremy. 1828/1843. “Historical Preface Intended for the Second Edition.” A preface drafted by Bentham for the second edition of A Fragment on Government (1823) but eventually left out of that edition. It was subsequently printed, though not published, in 1828. Reprinted as part of the first volume of The Works of Jeremy Bentham (pp. 240–259), the first compilation of Bentham’s writings. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued
350
Works Cited
in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Bentham, Jeremy. 1829/1983. “Article on Utilitarianism.” Drafted in 1829 in two versions, one longer, the other shorter, neither one appearing in print, except in the form of anonymous excerpts or direct quotations, until 1983, when the original material was recovered in full from the corresponding manuscripts and reprinted as part of Deontology, Together with A Table of the Springs of Action and the Article on Utilitarianism (Oxford: Clarendon Press, 1983), a compilation edited by Amnon Goldworth and published as a volume in The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Bentham, Jeremy. 1834/1983. Deontology; or, The Science of Morality: In Which the Harmony and Co-incidence of Duty and Self-interest, Virtue and Felicity, Prudence and Benevolence, Are Explained and Exemplified. From the MSS. of Jeremy Bentham. Two volumes. Vol. 1: Theory of the Social Science and History of the Greatest- Happiness Principle. Vol. 2: Practice of the Social Science. Arranged and edited by John Bowring. London: Longman, Rees, Orme, Browne, Green, and Longman. Edinburgh: William Tait, 1834. Recovered from the original manuscript and reprinted as part of Deontology, Together with A Table of the Springs of Action and the Article on Utilitarianism (Oxford: Clarendon Press, 1983), a compilation edited by Amnon Goldworth and published as a volume in The Collected Works of Jeremy Bentham, the standard critical edition of Bentham’s writings. Bentham, Jeremy. 1843. The Works of Jeremy Bentham. The first compilation of Bentham’s writings. Edited by John Bowring, published in twenty-two installments between 1838 and 1843, and reissued in eleven volumes in 1843. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Bentham, Jeremy. 1968–present. The Collected Works of Jeremy Bentham. The standard critical edition of Bentham’s writings. Projected to be seventy volumes upon completion of the initiative under the supervision of the Bentham Committee of University College London. Eight volumes, five devoted exclusively to Bentham’s correspondence, and three to his works in jurisprudence, published in London by Athlone Press between 1968 and 1981. Remainder of the series being published by Oxford University Press. General editors: James Henderson Burns (1961–1979); John Rowland Dinwiddy (1977–1983); Frederick Rosen (1983–1994); Frederick Rosen and Philip Schofield (1995–2003); Philip Schofield (2003–present). London: University of London, Athlone Press, 1968–1981. Oxford: Oxford University Press, 1981–present. Berger, Fred R. 1979. “John Stuart Mill on Justice and Fairness.” In New Essays on John Stuart Mill and Utilitarianism. Edited by Wesley E. Cooper, Kai Nielsen, and Steven C. Patten. Canadian Journal of Philosophy (Supplementary Vol. 5): 115–136. Berger, Fred R. 1984. Happiness, Justice, and Freedom: The Moral and Political Philosophy of John Stuart Mill. Berkeley and Los Angeles: University of California Press.
Works Cited
351
Berkeley, George. 1712. Passive Obedience, or the Christian Doctrine of Not Resisting the Supreme Power, Proved and Vindicated upon the Principles of the Law of Nature. In a Discourse Delivered at the College Chapel. Dublin: Jeremy Pepyat. Reprinted in The Works of George Berkeley, Bishop of Cloyne, vol. 6 (pp. 17–46), edited by Arthur Aston Luce and Thomas Edmund Jessop. Edinburgh: Thomas Nelson and Sons, 1953. Bibliothecary. 1879. “The Greatest Happiness of the Greatest Number.” Notes and Queries S5:xi 274 (29 March 1879): 247. Black, Beth (ed.). 2012. An A to Z of Critical Thinking. London and New York: Continuum International Publishing Group. Blau, Adrian. 2019. “Extended Meaning and Understanding in the History of Ideas.” History and Theory 58:3 (September): 342–359. Bowring, John. 1834. “History of the Greatest-Happiness Principle.” Last part of the first volume (pp. 287–331) of Jeremy Bentham’s Deontology; or, The Science of Morality: In Which the Harmony and Co-incidence of Duty and Self-interest, Virtue and Felicity, Prudence and Benevolence, Are Explained and Exemplified. Two volumes. Vol. 1: Theory of the Social Science. And History of the Greatest-Happiness Principle. Vol. 2: Practice of the Social Science. Arranged and edited by John Bowring. London: Longman, Rees, Orme, Browne, Green, and Longman. Edinburgh: William Tait, 1834. Bradley, Francis Herbert. 1876. Ethical Studies. New York: G. E. Stechert and Company. London: Henry S. King and Company. Second (“revised”) edition published posthumously in 1927 with additional notes by the author and reprinted in 1988 with an introduction by Richard Wollheim. Oxford: Clarendon Press. Brandt, Richard Booker. 1959. Ethical Theory: The Problems of Normative and Critical Ethics. Englewood Cliffs: Prentice-Hall. Brandt, Richard Booker. 1963. “Toward a Credible Form of Utilitarianism.” Chapter 4 of Morality and the Language of Conduct (pp. 107–143). Edited by Hector-Neri Castañeda and George Nakhnikian. Detroit: Wayne State University Press. Brandt, Richard Booker. 1967. “Some Merits of One Form of Rule Utilitarianism.” In University of Colorado Studies: Series in Philosophy, No. 3 (pp. 39– 65). Boulder: University of Colorado Press. Brandt, Richard Booker. 1979. A Theory of the Good and the Right. Oxford: Clarendon Press. Brandt, Richard Booker. 1983. “The Real & Alleged Problems of Utilitarianism.” The Hastings Center Report 13:2 (April): 37–43. Brandt, Richard Booker. 1988. “Fairness to Indirect Optimific Theories in Ethics.” Ethics 98:2 (January): 341–360. Brandt, Richard Booker. 1989. “Morality and Its Critics.” American Philosophical Quarterly 26:2 (April): 89–100. Brandt, Richard Booker. 1992. Morality, Utilitarianism, and Rights. Cambridge: Cambridge University Press.
352
Works Cited
Brandt, Richard Booker. 1995. “Conscience (Rule) Utilitarianism and the Criminal Law.” Law and Philosophy 14:1 (February): 65–89. Brandt, Richard Booker. 1996. Facts, Values, and Morality. Cambridge: Cambridge University Press. Brink, David Owen. 1992. “Mill’s Deliberative Utilitarianism.” Philosophy & Public Affairs 21:1 (Winter): 67–103. Brink, David Owen. 2013. Mill’s Progressive Principles. Oxford: Clarendon Press. Brody, Baruch A. 1967. “The Equivalence of Act and Rule Utilitarianism.” Philosophical Studies 18:6 (December): 81–87. Bronaugh, Richard N. 1974. “The Utility of Quality: An Understanding of Mill.” Canadian Journal of Philosophy 4:2 (December): 317–325. Brown, Donald George. 1972. “Mill on Liberty and Morality.” The Philosophical Review 81:2 (April): 133–158. Brown, Donald George. 1973. “What Is Mill’s Principle of Utility?” Canadian Journal of Philosophy 3:1 (September): 1–12. Brown, Donald George. 1974. “Mill’s Act-Utilitarianism.” The Philosophical Quarterly 24:94 (January): 67–68. Brown, Donald George. 1978. “Mill on Harm to Others’ Interests.” Political Studies 26:3 (September): 395–399. Brown, Donald George. 1982. “Mill’s Criterion of Wrong Conduct.” Dialogue 21:1 (March): 27–44. Brown, Donald George. 2010. “Mill’s Moral Theory: Ongoing Revisionism.” Politics, Philosophy & Economics 9:1 (February): 5–45. Bruening, William H. 1971. “Moore and ‘Is-Ought.’ ” Ethics 81:2 (January): 143–149. Burns, James Henderson. 2005. “Happiness and Utility: Jeremy Bentham’s Equation.” Utilitas 17:1 (March): 46–61. Burton, John Hill. 1843. “Introduction.” The editorial introduction to the first volume of The Works of Jeremy Bentham (pp. 5–83). The first compilation of Bentham’s writings. Edited by John Bowring. Edinburgh: William Tait. London: William Simpkin and Richard Marshall. Published in twenty-two installments between 1838 and 1843. Reissued in eleven volumes in 1843. Capaldi, Nicholas. 1973. “Mill’s Forgotten Science of Ethology.” Social Theory and Practice 2:4 (Fall): 409–420. Carlisle, Janice. 1991. John Stuart Mill and the Writing of Character. Athens, Georgia: University of Georgia Press. Carlyle, Thomas. 1850. “Jesuitism” (1 August 1850). Eighth and final essay (“pamphlet”) in his Latter-Day Pamphlets (pp. 249–286). London: Chapman and Hall. Clark, George A. 1959. “Mill’s ‘Notorious Analogy.’” The Journal of Philosophy 56:15 (16 July 1959): 652–656.
Works Cited
353
Cocceji, Samuel von. 1699. Disputatio Juridica Inauguralis de Principio Juris Naturalis Unico, Vero, et Adaequato [Inaugural Juridical Disputation on the Only True and Adequate Starting Point of Natural Law]. Frankfurt an der Oder: Christoph Andreas Zeitler. Cohen, Stephen. 1990. “Proof and Sanction in Mill’s Utilitarianism.” History of Philosophy Quarterly 7:4 (October): 475–487. Cole, Richard. “1965. A Note on Informal Fallacies.” Mind (New Series) 74:295 (July): 432–433. Coope, Christopher Miles. 1998. “Was Mill a Utilitarian?” Utilitas 10:1 (March): 33–67. Cooper, Neil. 1969. “Mill’s ‘Proof’ of the Principle of Utility.” Mind (New Series) 78:310 (April): 278–279. Copi, Irving M., Carl Cohen, and Kenneth McMahon. 2014. Introduction to Logic. Fourteenth edition. Harlow, Essex: Pearson Education. Copp, David. 1979. “The Iterated-Utilitarianism of J. S. Mill.” In New Essays on John Stuart Mill and Utilitarianism. Edited by Wesley E. Cooper, Kai Nielsen, and Steven C. Patten. Canadian Journal of Philosophy (Supplementary Vol. 5): 75–98. Creighton, James Edwin. 1898. An Introductory Logic. New York: Macmillan Company. Crisp, Roger. 1997. Routledge Philosophy Guidebook to Mill on Utilitarianism. London: Routledge. Crisp, Roger, and Morten Kringelbach. 2018. “Higher and Lower Pleasures Revisited: Evidence from Neuroscience.” Neuroethics 11:2 (July): 211–215. Cupples, Brian. 1972. “A Defence of the Received Interpretation of J. S. Mill.” Australasian Journal of Philosophy 50:2 (August): 131–137. Dahl, Norman O. 1973. “Is Mill’s Hedonism Inconsistent?” In Studies in Ethics (pp. 37– 54). Edited by Nicholas Rescher. American Philosophical Quarterly Monograph Series 7. Oxford: Basil Blackwell. Day, John Patrick. 1964. “John Stuart Mill.” Chapter 19 of A Critical History of Western Philosophy (pp. 341–364). Edited by Daniel John O’Connor. New York: The Free Press. London: Collier-Macmillan. Dewey, John. 1891. Outlines of a Critical Theory of Ethics. Ann Arbor: Register Publishing Company. Reprinted as part of John Dewey: The Early Works, 1882–1898. Volume 3: 1889–1892 (pp. 237–388). Edited by Jo Ann Boydston. Carbondale: Southern Illinois University Press, 1969. Dewey, John, and James Hayden Tufts. 1908. Ethics. New York: Henry Holt and Company. London: G. Bell and Sons. Second, “revised” edition, 1932. Second edition reprinted as part of John Dewey: The Later Works, 1925–1953. Volume 7: 1932. Edited by Jo Ann Boydston. Carbondale: Southern Illinois University Press, 1985. Doble, Charles Edward. 1897. “The Greatest Happiness of the Greatest Number.” Notes and Queries S8:xi 281 (15 May 1897): 392.
354
Works Cited
Donner, Wendy. 1991. The Liberal Self: John Stuart Mill’s Moral and Political Philosophy. Ithaca: Cornell University Press. Donner, Wendy. 1998. “Mill’s Utilitarianism.” Chapter 7 of The Cambridge Companion to Mill (pp. 255– 292). Edited by John Skorupski. Cambridge: Cambridge University Press. Donner, Wendy. 2009. “Utilitarianism: Morality, Justice, and the Art of Life.” Chapter 3 of Mill (pp. 33–55) by Wendy Donner and Richard A. Fumerton. Blackwell Great Minds. Oxford: Wiley-Blackwell. Dryer, Douglas Poole. 1969. “Mill’s Utilitarianism.” Introductory essay on Mill’s conception of utilitarianism, following the general introduction by Francis Ethelbert Louis Priestley, in Essays on Ethics, Religion and Society (pp. lxiii–cxiii), vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press. Dryer, Douglas Poole. 1979. “Justice, Liberty, and the Principle of Utility in Mill.” In New Essays on John Stuart Mill and Utilitarianism. Edited by Wesley E. Cooper, Kai Nielsen, and Steven C. Patten. Canadian Journal of Philosophy (Supplementary Vol. 5): 63–73. Duncan, Elmer H. 1970. “Has Anyone Committed the Naturalistic Fallacy?” The Southern Journal of Philosophy 8:1 (Spring): 49–55. Reprinted in The Southwestern Journal of Philosophy 1:1/2 (Spring–Summer 1970): 40–46. Dunn, John. 1968. “The Identity of the History of Ideas.” Philosophy 43:164 (April): 85–104. Ebenstein, Lanny. 1985. “Mill’s Theory of Utility.” Philosophy 60:234 (October): 539–543. Edwards, Rem B. 1986. “The Principle of Utility and Mill’s Minimizing Utilitarianism.” The Journal of Value Inquiry 20:2 (June): 125–136. Eggleston, Ben. 2010. “Rules and Their Reasons: Mill on Morality and Instrumental Rationality.” Chapter 3 of John Stuart Mill and the Art of Life (pp. 71–93). Edited by Ben Eggleston, Dale E. Miller, and David Weinstein. Oxford: Oxford University Press. Eggleston, Ben. 2014. “Act Utilitarianism.” Chapter 6 of The Cambridge Companion to Utilitarianism (pp. 125– 145). Edited by Ben Eggleston and Dale E. Miller. Cambridge: Cambridge University Press. Eggleston, Ben. 2017. “Mill’s Moral Standard.” Chapter 24 of A Companion to Mill (pp. 358–373). Edited by Christopher Macleod and Dale E. Miller. Oxford: John Wiley and Sons. Eggleston, Ben, and Dale E. Miller. 2007. “India House Utilitarianism: A First Look.” Southwest Philosophy Review 23:1 (January): 39–47. Eggleston, Ben, and Dale E. Miller. 2008. “Mill’s Misleading Moral Mathematics.” Southwest Philosophy Review 24:1 (January): 153–161. Eggleston, Ben, and Dale E. Miller (eds.). 2014. The Cambridge Companion to Utilitarianism. Cambridge: Cambridge University Press.
Works Cited
355
Eggleston, Ben, Dale E. Miller, and David Weinstein (eds.). 2010. John Stuart Mill and the Art of Life. Oxford: Oxford University Press. Ezorsky, Gertrude. 1968. “A Defense of Rule Utilitarianism against David Lyons Who Insists on Tieing It to Act Utilitarianism, Plus a Brand New Way of Checking Out General Utilitarian Properties.” The Journal of Philosophy 65:18 (19 September 1968): 533–544. Fearnside, W. Ward, and William B. Holther. 1959. Fallacy: The Counterfeit of Argument. Englewood Cliffs: Prentice-Hall. Fénelon, François de Salignac de la Mothe, Archbishop of Cambrai (Archevéque-Duc de Cambrai). 1699/1717. Les Aventures de Télémaque, Fils d’Ulysse. Published anonymously in The Hague: Adrian Moetjens, 1699. Reissued posthumously in a revised edition in Paris: Chez Florentin Delaulne, 1717. Fieser, James. 1993. “Moore, Spencer, and the Naturalistic Fallacy.” History of Philosophy Quarterly 10:3 (July): 271–276. Fletcher, Guy. 2013. “A Millian Objection to Reasons as Evidence.” Utilitas 25:3 (September): 417–420. Frankena, William Klaas. 1939. “The Naturalistic Fallacy.” Mind (New Series) 48:192 (October): 464–477. Frankena, William Klaas. 1963. Ethics. Englewood Cliffs: Prentice-Hall. Fuchs, Alan E. 2006. “Mill’s Theory of Morally Correct Action.” Chapter 5 of The Blackwell Guide to Mill’s Utilitarianism (pp. 139–158). Edited by Henry Robison West. Oxford: Blackwell Publishing. Fullinwider, Robert K. 1972. “On Mill’s Analogy between Visible and Desirable.” The Southern Journal of Philosophy 10:1 (Spring): 17–22. Fyfe, Andrew Thomas. 2011. “Mill’s Proof of Utilitarianism.” Chapter 57 of Just the Arguments: 100 of the Most Important Arguments in Western Philosophy (pp. 223– 228). Edited by Michael Bruce and Steven Barbone. Oxford: Blackwell Publishing. Gaus, Gerald. 1980. “Mill’s Theory of Moral Rules.” Australasian Journal of Philosophy 58:3 (September): 265–279. Gauthier, David P. 1967. “Moore’s Naturalistic Fallacy.” American Philosophical Quarterly 4:4 (October): 315–320. Geiger, George R. 1949. “A Note on the Naturalistic Fallacy.” Philosophy of Science 16:4 (October): 336–342. Gensler, Harry J. 1989. Logic: Analyzing and Appraising Arguments. Englewood Cliffs: Prentice-Hall. Gerson, Lloyd P. 2013. “Was Plato a Platonist?” Chapter 1 of his From Plato to Platonism (pp. 3–33). Ithaca: Cornell University Press. Gibbard, Allan F. 1965. “Rule- Utilitarianism: Merely an Illusory Alternative?” Australasian Journal of Philosophy 43:2 (August): 211–220. Gibbs, Benjamin. 1986. “Higher and Lower Pleasures.” Philosophy 61:235 (January): 31–59.
356
Works Cited
Gray, John. 1979. “John Stuart Mill: Traditional and Revisionist Interpretations.” Literature of Liberty 2:2 (April–June): 7–37. Gray, John. 1981. “John Stuart Mill on Liberty, Utility, and Rights.” Chapter 5 of Human Rights. Edited by James Roland Pennock and John William Chapman. Nomos (Yearbook of the American Society for Political and Legal Philosophy) 23: 80–116. New York and London: New York University Press. Gray, John. 1983. Mill on Liberty: A Defence. London: Routledge. Second edition (same pagination as the first except for the addition of a new preface and a postscript), 1996. Gray, John. 1984. “Indirect Utility and Fundamental Rights.” Social Philosophy and Policy 1:2 (Spring): 73–91. Gray, John. 1991. “Introduction” to John Stuart Mill: On Liberty and Other Essays. Edited with an introduction and explanatory notes by John Gray and G. W. Smith. Oxford: Oxford University Press, vii–xxx. Green, Thomas Hill. 1883. Prolegomena to Ethics. Edited by Andrew Cecil Bradley. Published posthumously. Oxford: Clarendon Press. Griffin, Nicholas. 1972. “A Note on Mr. Cooper’s Reconstruction of Mill’s ‘Proof.’ ” Mind (New Series) 81:321 (January): 142–143. Grote, John. 1870. An Examination of the Utilitarian Philosophy. Edited by Joseph Bickersteth Mayor. Cambridge: Deighton, Bell, and Company. London: Bell and Daldy. Haezrahi, Pepita. 1949. “The Desired and the Desirable.” Analysis 10:2 (December): 40–48. Hales, Steven D. 2007. “Mill v. Miller, or Higher and Lower Pleasures.” Chapter 7 of Beer and Philosophy: The Unexamined Beer Isn’t Worth Drinking (pp. 97–109). Edited by Steven D. Hales. Oxford: Blackwell Publishing. Halévy, Élie. 1901–1904. La Formation du Radicalisme Philosophique. Three volumes. Vol. 1: La Jeunesse de Bentham (1901). Vol. 2: L’Évolution de la Doctrine Utilitaire de 1789 à 1815 (1901). Vol. 3: Le Radicalisme Philosophique (1904). Paris: Félix Alcan. Translated into English by Mary Morris and published in a single volume as The Growth of Philosophic Radicalism. London: Faber and Gwyer, 1928. Second (“new”) edition of the English translation in London: Faber and Faber, 1934. Reprinted in 1949. Reprinted, with corrections, in 1952. French edition reprinted in Paris: Presses Universitaires de France, 1995. Hall, Everett W. 1949. “The ‘Proof’ of Utility in Bentham and Mill.” Ethics 60:1 (October): 1–18. Hamblin, Charles Leonard. 1970. Fallacies. London: Methuen and Company. Hancock, Roger. 1975. “Mill, Saints and Heroes.” The Mill News Letter 10:1 (Winter): 13–15. Hare, Richard Mervyn. 1955. “Universalisability.” Proceedings of the Aristotelian Society (New Series) 55: 295–312. Hare, Richard Mervyn. 1963. Freedom and Reason. Oxford: Oxford University Press.
Works Cited
357
Hare, Richard Mervyn. 1973. “Principles.” Proceedings of the Aristotelian Society (New Series) 73: 1–18. Hare, Richard Mervyn. 1976. “Ethical Theory and Utilitarianism.” In Contemporary British Philosophy: Personal Statements. Fourth Series (pp. 113–131). Edited by Hywel David Lewis. London: George Allen and Unwin. Hare, Richard Mervyn. 1981. Moral Thinking: Its Levels, Methods, and Point. Oxford: Clarendon Press. Hare, Richard Mervyn. 1993. “Could Kant Have Been a Utilitarian?” Utilitas 5:1 (May): 1–16. Harrison, Jonathan. 1953. “Utilitarianism, Universalisation, and Our Duty to Be Just.” Proceedings of the Aristotelian Society (New Series) 53: 105–134. Harrison, Jonathan. 1975. “The Expedient, the Right and the Just in Mill’s Utilitarianism.” In New Essays in the History of Philosophy. Edited by Terence Penelhum and Roger A. Shiner. Canadian Journal of Philosophy (Supplementary Vol. 1): 93–107. Harrison, Jonathan. 1979. “Rule Utilitarianism and Cumulative-Effect Utilitarianism.” In New Essays on John Stuart Mill and Utilitarianism. Edited by Wesley E. Cooper, Kai Nielsen, and Steven C. Patten. Canadian Journal of Philosophy (Supplementary Vol. 5): 21–45. Harrod, Henry Roy Forbes. 1936. “Utilitarianism Revised.” Mind (New Series): 45:178 (April): 137–156. Harsanyi, John C. 1977a. “Morality and the Theory of Rational Behavior.” Social Research (Rationality, Choice, and Morality) 44:4 (Winter): 623–656. Harsanyi, John C. 1977b. “Rule Utilitarianism and Decision Theory.” Erkenntnis 11:1 (January): 25–53. Hartley, David. 1749. Observations on Man, His Frame, His Duty and His Expectations. Two volumes. London: Printed by Samuel Richardson for James Leake and William Frederick. Sold by Charles Hitch and Stephen Austen. Hauskeller, Michael. 2011. “No Philosophy for Swine: John Stuart Mill on the Quality of Pleasures.” Utilitas 23:4 (December): 428–446. Häyry, Matti, and Heta Häyry. 1994. “Obedience to Rules and Berkeley’s Theological Utilitarianism.” Utilitas 6:2 (November): 233–242. Heans, S. J. 1992. “Was Mill a Moral Scientist?” Philosophy 67:259 (January): 81–101. Helvétius, Claude Adrien. 1758. De l’Esprit. Paris: Chez Durand, Libraire. Translated into English as De l’Esprit: Or, Essays on the Mind, and Its Several Faculties. London: Printed for the translator and sold by Dodsley and Company; Millar, Nourse, and Tonson; Hitch and Hawes; Fletcher (Oxford); Merril (Cambridge); and J. Coote, 1759. Himma, Kenneth Einar. 1998. “The Interpretation of Mill’s Utilitarianism.” History of Philosophy Quarterly 15:4 (October): 455–473. Hoag, Robert W. 1987. “Mill’s Conception of Happiness as an Inclusive End.” Journal of the History of Philosophy 25:3 (July): 417–431.
358
Works Cited
Hoag, Robert W. 1992. “J. S. Mill’s Language of Pleasures.” Utilitas 4:2 (November): 247–278. Hobbes, Thomas. 1651. Leviathan, or, The Matter, Form, and Power of a Common-wealth, Ecclesiastical and Civil. London: Andrew Crooke. Hocutt, Max. 2005. “Was Bentham a Utilitarian?” Canadian Journal of Political Science 38:3 (September): 697–717. Hodgson, David Hargraves. 1967. Consequences of Utilitarianism: A Study in Normative Ethics and Legal Theory. Oxford: Clarendon Press. Hollander, Samuel. 2019. A History of Utilitarian Ethics: Studies in Private Motivation and Distributive Justice, 1700–1875. London: Routledge. Hooker, Brad. 2006. “Right, Wrong, and Rule-Consequentialism.” Chapter 11 of The Blackwell Guide to Mill’s Utilitarianism (pp. 233–269). Edited by Henry Robison West. Oxford: Blackwell Publishing. Hooker, Brad. 2013a. “Rule Consequentialism.” Chapter 11 of The Blackwell Guide to Ethical Theory (pp. 238–260). Second edition. Edited by La Hugh Follette and Ingmar Persson. Oxford: Blackwell Publishing. Hooker, Brad. 2013b. “Rule Utilitarianism.” Entry in The Bloomsbury Encyclopedia of Utilitarianism (pp. 490–495). Edited by James E. Crimmins. London: Bloomsbury Academic. Hruschka, Joachim. 1991. “The Greatest Happiness Principle and Other Early German Anticipations of Utilitarian Theory.” Utilitas 3:2 (November): 165–177. Hruschka, Joachim. 1992. “Universalization and Related Principles.” arsp: Archiv für Rechts-und Sozialphilosophie /Archives for Philosophy of Law and Social Philosophy 78:3: 289–300. Hume, David. 1741–1742. Essays, Moral and Political. Two volumes. Edinburgh: Printed by Robert Fleming and Alexander Alison for Alexander Kincaid. Hume, David. 1751. An Enquiry Concerning the Principles of Morals. London: Andrew Millar. Hume, David. 1753. Essays and Treatises on Several Subjects. Four volumes. Vol. 1: Essays, Moral and Political. Vol. 2: Philosophical Essays Concerning Human Understanding. Vol. 3: An Enquiry Concerning the Principles of Morals. Vol. 4: Political Discourses. London: Andrew Millar. Edinburgh: Alexander Kincaid and Alexander Donaldson. Numerous editions and reprints. Hutcheson, Francis. 1725. An Inquiry into the Original of Our Ideas of Beauty and Virtue. Published anonymously with the full title: An Inquiry into the Original of Our Ideas of Beauty and Virtue, in Two Treatises, in Which the Principles of the Late Earl of Shaftesbury Are Explain’d and Defended, against the Author of the Fable of the Bees [Bernard de Mandeville], and the Ideas of Moral Good and Evil Are Establish’d, According to the Sentiments of the Ancient Moralists, with an Attempt to Introduce a Mathematical Calculation in Subjects of Morality. Treatise 1: “An Inquiry Concerning Beauty, Order, &c. [Harmony, Design]” (pp. 1–97). Treatise 2: “An Inquiry Concerning
Works Cited
359
the Original of Our Ideas of Virtue or Moral Good” (pp. 99–276). London: Printed by John Darby for William and John Smith, 1725. Irwin, Terence Henry. 2007–2009. The Development of Ethics: A Historical and Critical Study. Vol. 1: From Socrates to the Reformation (2007). Vol. 2: From Suarez to Rousseau (2008). Vol. 3: From Kant to Rawls (2009). Oxford: Oxford University Press. Jacobson, Daniel. 2003. “J. S. Mill and the Diversity of Utilitarianism.” Philosophers’ Imprint 3:2 (June): 1–18. Jones, Hardy. 1978. “Mill’s Argument for the Principle of Utility.” Philosophy and Phenomenological Research 38:3 (March): 338–354. Jones, Hardy. 1979. “Are Fundamental Moral Principles Incapable of Proof?” Metaphilosophy 10:2 (April): 153–160. Joyce, Richard. 2001. “Evolutionary Ethical Naturalism and the Naturalistic Fallacy.” Chapter 6.3 of The Myth of Morality (pp. 153– 157). Cambridge: Cambridge University Press. Joyce, Richard. 2006. “The Naturalistic Fallacy.” Chapter 5.2 of The Evolution of Morality (pp. 146–152). Cambridge: mit Press. Kahane, Howard. 1971. Logic and Contemporary Rhetoric: The Use of Reason in Everyday Life. Belmont: Wadsworth Publishing Company. Numerous editions and reprints. Twelfth edition published posthumously in 2014 with Nancy Cavender as coauthor. Kahane, Howard. 1980. “The Nature and Classification of Fallacies.” In Informal Logic: The First International Symposium (pp. 31–39). Edited by J. Anthony Blair and Ralph H. Johnson. Inverness: Edgepress. Katz, Jerrold J. 1990. “The Naturalistic Fallacy.” Chapter 7 of The Metaphysics of Meaning (pp. 235–290). Cambridge: mit Press. Kavka, Gregory S. 1975. “Extensional Equivalence and Utilitarian Generalization.” Theoria 41:3 (December): 125–147. Kawana, Yuichiro. 2018. Logic and Society: The Political Thought of John Stuart Mill, 1827–1848. New York: Palgrave Macmillan. Kitcher, Philip. 2011. “Naturalistic Fallacies?” Chapter 7 of The Ethical Project (pp. 253– 282). Cambridge: Harvard University Press. Kleinig, John. 1970. “The Fourth Chapter of Mill’s Utilitarianism.” Australasian Journal of Philosophy 48:2 (August): 197–205. Kolnai, Aurel. 1980. “The Ghost of the Naturalistic Fallacy.” Philosophy 55: 211 (January): 5–16. Kretzmann, Norman. 1958. “Desire as Proof of Desirability.” The Philosophical Quarterly 8:32 (July): 246–258. Kupfer, Joseph. 1974. “Universalization in Berkeley’s Rule—Utilitarianism.” Revue Internationale de Philosophie 28:110 (4): 511–531. Lang, Berel, and Gary Stahl. 1969. “Mill’s ‘Howlers’ and the Logic of Naturalism.” Philosophy and Phenomenological Research 29:4 (June): 562–574.
360
Works Cited
Leibniz, Gottfried Wilhelm. 1686. Discours de Métaphysique. Translated into English by George R. Montgomery, with an introduction by Paul Janet, as part of Discourse on Metaphysics, Correspondence with Arnauld, and Monadology. Chicago: Open Court Publishing Company, 1908. Leibniz, Gottfried Wilhelm. 1700. “Observationes de Principio Juris.” Monathlicher Auszug aus Allerhand Neu-Herausgegebenen/Nützlichen und Artigen Büchern 4 (July 1700): 371–382. Leidhold, Wolfgang. 2004. “Introduction.” The introduction to a critical edition of Francis Hutcheson’s An Inquiry into the Original of Our Ideas of Beauty and Virtue (London: Printed by John Darby for William and John Smith, 1725). Reprint edited with an introduction and critical apparatus by Wolfgang Leidhold. Indianapolis: Liberty Fund, 2004, ix–xviii. Lewy, Casimir. 1965. “G. E. Moore on the Naturalistic Fallacy.” Annual philosophical lecture sponsored by the Dawes Hicks Trust Fund and delivered before the British Academy on 11 November 1964. Proceedings of the British Academy 50 (1965): 251– 262. Reprinted in Studies in the Philosophy of Thought and Action (pp. 134–146), selected and edited by Peter Frederick Strawson (Oxford: Oxford University Press, 1968), and in G. E. Moore: Essays in Retrospect (pp. 292–303), edited by Alice Ambrose and Morris Lazerowitz (London: George Allen and Unwin, 1970). Locke, John. 1690. An Essay Concerning Human [“Humane”] Understanding. London: Thomas Basset. Loizides, Antis. 2013. “The Socratic Origins of John Stuart Mill’s ‘Art of Life.’ ” Chapter 3 of John Stuart Mill: A British Socrates (pp. 75–96). Edited by Kyriakos N. Demetriou and Antis Loizides. New York: Palgrave Macmillan. Long, Douglas G. 1990. “ ‘Utility’ and the ‘Utility Principle’: Hume, Smith, Bentham, Mill.” Utilitas 2:1 (May): 12–39. Long, Roderick T. 1992. “Mill’s Higher Pleasures and the Choice of Character.” Utilitas 4:2 (November): 279–297. Long, W. H. 1967. “The Legend of Mill’s ‘Proofs.’ ” The Southern Journal of Philosophy 5:1 (Spring): 36–47. Loring, L. M. 1967. “Moore’s Criticism of Mill.” Ratio 9:1 (June): 84–90. Luebke, Neil R. 1970. “Frankena on the Naturalistic Fallacy.” Journal of Thought 5:4 (October): 262–273. Lyons, David. 1965. Forms and Limits of Utilitarianism. Oxford: Clarendon Press. Lyons, David. 1971. “Was Bentham a Utilitarian?” Royal Institute of Philosophy Lectures 5 (March): 196–221. Lyons, David. 1973. In the Interest of the Governed: A Study of Bentham’s Philosophy of Utility and Law. Oxford: Clarendon Press. Lyons, David. 1976. “Mill’s Theory of Morality.” Noûs 10:2 (May): 101–120.
Works Cited
361
Lyons, David. 1977. “Human Rights and the General Welfare.” Philosophy &Public Affairs 6:2 (Winter): 113–129. Lyons, David. 1978. “Mill’s Theory of Justice.” In Values and Morals: Essays in Honor of William Frankena, Charles Stevenson, and Richard Brandt (pp. 1–20). Edited by Alvin I. Goldman and Jaegwon Kim. Dordrecht: D. Reidel Publishing Company (Springer). Lyons, David. 1979. “Liberty and Harm to Others.” In New Essays on John Stuart Mill and Utilitarianism. Edited by Wesley E. Cooper, Kai Nielsen, and Steven C. Patten. Canadian Journal of Philosophy (Supplementary Vol. 5): 1–19. Lyons, David. 1982. “Benevolence and Justice in Mill.” Chapter 3 of The Limits of Utilitarianism (pp. 42–70). Edited by Harlan B. Miller and William H. Williams. Minneapolis: University of Minnesota Press. Mabbott, John David. 1953. “Moral Rules.” Annual philosophical lecture sponsored by the Henriette Hertz Trust Fund and delivered before the British Academy on 18 February 1953. Proceedings of the British Academy 37 (1953): 97–118. Mabbott, John David. 1956. “Interpretations of Mill’s ‘Utilitarianism.’ ” The Philosophical Quarterly 6:23 (April): 115–120. Mabbott, John David. 1966. An Introduction to Ethics. London: Hutchinson and Company. Macaulay, Thomas Babington. 1829. “Mill’s Essay on Government /Utilitarian Logic and Politics.” Running headers on alternate pages (even/odd respectively), either header, or both, often cited as the title of the corresponding article, published anonymously as a critical review of James Mill’s Essays on Government, Jurisprudence, Liberty of the Press, and Law of Nations (1823/1828), with the article bearing the actual title: “Essays on Government, Jurisprudence, the Liberty of the Press, Prisons and Prison Discipline, Colonies, the Law of Nations, and Education. By james mill, Esq. author of the History of British India. Reprinted by permission from the Supplement to the Encyclopædia Britannica. (Not for sale.) London, 1828.” Published as Article 7 of The Edinburgh Review 49:97 (March 1829): 159–189. Mackenzie, John Stuart. 1893. A Manual of Ethics. London: W. B. Clive and Company, University Correspondence College Press. Numerous editions and reprints. Second edition, 1894. Third edition, 1897. Fourth edition, 1900/1901. Fifth edition, 1915. Sixth edition, 1929. First two editions published with the subtitle “Designed for the Use of Students.” Subtitle dropped permanently starting with the third edition (“revised, enlarged, and in part rewritten”), published in London (W. B. Clive) and New York (Hinds and Noble). Fourth edition released with a shorter (June 1900) and longer preface (February 1901), corresponding, respectively, to the 1900 and 1901 impressions. Fifth edition undertaken primarily to account for the growing influence of Moore following the publication of Principia Ethica (1903). Sixth edition significantly shorter despite covering a substantial amount of new material.
362
Works Cited
Macleod, Christopher. 2013a. “Mill, Intuitions, and Normativity.” Utilitas 25:1 (March): 46–65. Macleod, Christopher. 2013b. “Was Mill a Non-Cognitivist?” The Southern Journal of Philosophy 51:2 (June): 206–223. Macleod, Christopher. 2014. “Mill on the Epistemology of Reasons: A Comparison with Kant.” Chapter 7 of Mill’s A System of Logic: Critical Appraisals (pp. 151–169). Edited by Antis Loizides. New York: Routledge. Mandelbaum, Maurice. 1968. “Two Moot Issues in Mill’s Utilitarianism.” In Mill: A Collection of Critical Essays (pp. 206–233). Edited by Jerome Borges Schneewind. London: Macmillan and Company. Margolis, Joseph. 1967. “Mill’s Utilitarianism Again.” Australasian Journal of Philosophy 45:2 (August): 179–184. Marshall, John. 1973. “The Proof of Utility and Equity in Mill’s Utilitarianism.” Canadian Journal of Philosophy 3:1 (September): 13–26. Martin, Rex. 1972. “A Defence of Mill’s Qualitative Hedonism.” Philosophy 47:180 (April): 140–151. Martin, Rex. 2007. “Two Concepts of Rule Utilitarianism: The Case of Mill.” Southwest Philosophy Review 23:1 (January): 49–58. Martin, Rex. 2008. “Two Concepts of Rule Utilitarianism.” Journal of Moral Philosophy 5:2: 227–255. Martin, Rex. 2010. “Mill’s Rule Utilitarianism in Context.” Chapter 1 of John Stuart Mill and the Art of Life (pp. 21–43). Edited by Ben Eggleston, Dale E. Miller, and David Weinstein. Oxford: Oxford University Press. Mawson, Tim. 2002. “Mill’s Proof.” Philosophy 77:301 (July): 375–405. McNeilly, Francis Stewart. 1958. “Pre-Moral Appraisals.” The Philosophical Quarterly 8:31 (April): 97–111. Mill, James. 1823. Essays on Government, Jurisprudence, Liberty of the Press, and Law of Nations. London: J. Innes. Reprinted in 1825 and 1828. A collection of articles originally appearing separately in various issues of the Supplement to the fourth, fifth, and sixth editions of the Encyclopedia Britannica, a series of half-volume supplements published between 1815 and 1824, under the editorial supervision of MacVey Napier. Mill, James. 1829. Analysis of the Phenomena of the Human Mind. Two volumes. London: Baldwin and Cradock. Second edition published posthumously, with illustrative and critical notes by Alexander Bain, Andrew Findlater, and George Grote, edited with additional notes by John Stuart Mill. London: Longmans, Green, Reader, and Dyer, 1869. Mill, John Stuart. 1812– 1873. The Letters of John Stuart Mill. Two compilations. The first was published in two volumes and edited by Hugh Samuel Roger Elliot. New York: Longmans, Green, and Company, 1910. The standard critical
Works Cited
363
edition was published in seven volumes in the Collected Works of John Stuart Mill. Toronto: University of Toronto Press. Vols. 12–13 (1963): The Earlier Letters of John Stuart Mill, 1812–1848. Edited by Francis E. Mineka. Vols. 14–17 (1972): The Later Letters of John Stuart Mill, 1849–1873. Edited by Francis E. Mineka and Dwight N. Lindley. Vol. 32 (1991): Additional Letters of John Stuart Mill. Edited by Marion Filipiuk, Michael Laine, and John Mercel Robson. Mill, John Stuart. 1832. “Use and Abuse of Political Terms.” Review of George Cornewall Lewis’s Remarks on the Use and Abuse of Some Political Terms (London: Fellowes, 1832). Originally published in Tait’s Edinburgh Magazine 1 (May 1832): 164–172. Also available in Essays on Politics and Society (pp. 1–13) as part of vol. 18 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1977. Mill, John Stuart. 1833. “Remarks on Bentham’s Philosophy.” Originally published anonymously as an appendix to the second of two volumes of Edward Lytton Bulwer’s England and the English (pp. 321–344). London: Bentley, 1833. Also available in Essays on Ethics, Religion and Society (pp. 3–18) as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1835. “Sedgwick’s Discourse.” Review of Adam Sedgwick’s Discourse on the Studies of the University (third edition, London: John W. Parker, 1834). Sedgwick’s Discourse was originally published in 1833. Mill’s review was originally published as “Professor Sedgwick’s Discourse—State of Philosophy in England / A Discourse on the Studies of the University” in London Review 1 (April 1835): 94– 135. Reprinted as “Professor Sedgwick’s Discourse on the Studies of the University of Cambridge” in Dissertations and Discussions 1 (1867): 95–159. Also available as “Sedgwick’s Discourse” in Essays on Ethics, Religion and Society (pp. 31–74) as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1838. “The Works of Jeremy Bentham.” Review of the first four parts of Jeremy Bentham’s collected works, the first-ever compilation, still in progress at the time of the review: The Works of Jeremy Bentham (Edinburgh: William Tait. London: William Simpkin and Richard Marshall), edited by John Bowring, published in twenty-two installments between 1838 and 1843, reissued in eleven volumes in 1843. The review originally appeared as “The Works of Jeremy Bentham: Now First Collected, under the Superintendence of his Executor, John Bowring. Parts I to IV.” London and Westminster Review 31:2/61 (August 1838): 467–506. Subsequently released as a standalone volume titled An Estimate of Bentham’s Philosophy. London: C. Reynell, 1838. Reprinted in Dissertations and Discussions 1 (1867): 330– 392. Also available as “Bentham” in Essays on Ethics, Religion and Society (pp. 75–115)
364
Works Cited
as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1840. “Coleridge.” Originally published as a review of multiple works by Samuel Taylor Coleridge in London and Westminster Review 33:2/65 (March 1840): 257–302. Reprinted in Dissertations and Discussions 1 (1867): 393– 466. Also available in Essays on Ethics, Religion and Society (pp. 117–163) as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1843. A System of Logic, Ratiocinative and Inductive, Being a Connected View of the Principles of Evidence and the Methods of Scientific Investigation. London: Parker, Son, and Bourn, 1843. Numerous editions and reprints. Eighth and final edition in London: Longmans, Green, Reader, and Dyer, 1872. Eighth edition also available as vols. 7 and 8 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1974. Mill, John Stuart. 1852. “Whewell on Moral Philosophy.” Review of William Whewell’s Lectures on the History of Moral Philosophy in England (London: John W. Parker and Son. Cambridge: John Deighton, 1852) and Elements of Morality (London: John W. Parker, 1845). Originally published in Westminster and Foreign Quarterly Review 58 (October 1852): 349–385. Reprinted in Dissertations and Discussions 2 (1867): 450– 509. Also available in Essays on Ethics, Religion and Society (pp. 165–201) as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1859. Dissertations and Discussions: Political, Philosophical, and Historical. Originally published in two volumes in London: Parker, Son, and Bourn, 1859. Second edition published in three volumes in London: Longmans, Green, Reader, and Dyer, 1867. Third edition published posthumously in four volumes (“reprinted chiefly from the Edinburgh and Westminster reviews”) and edited by Helen Taylor. London: Longmans, Green, Reader, and Dyer, 1875. Mill, John Stuart. 1859. On Liberty. Originally published in London: John W. Parker and Son, 1859. Second edition, by the same publisher, in the same year. Third edition in London: Longman, Green, Longman, Roberts, and Green, 1864. Fourth edition in London: Longmans, Green, Reader, and Dyer, 1869. Fourth edition, the last in Mill’s lifetime, also available in Essays on Politics and Society (pp. 213–310) as part of vol. 18 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1977. Mill, John Stuart. 1861. Utilitarianism. Originally published as an essay in three parts in Fraser’s Magazine for Town and Country: chapters 1–2 in Fraser’s Magazine 64:382 (October 1861): 391–406; chapters 3–4 in Fraser’s Magazine 64:383 (November 1861): 525–534; chapter 5 in Fraser’s Magazine 64:384 (December 1861): 659–673. First edition in book form in London: Parker, Son, and Bourn, 1863. Several editions
Works Cited
365
and numerous reprints. Fourth edition (London: Longmans, Green, Reader, and Dyer, 1871), the last in Mill’s lifetime, also available in Essays on Ethics, Religion and Society (pp. 203–259) as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1865. Auguste Comte and Positivism. Originally published as two separate essays: “The Positive Philosophy of Auguste Comte” in Westminster and Foreign Quarterly Review 83 (April 1865): 339–405 and “Later Speculations of Auguste Comte” in Westminster and Foreign Quarterly Review 84 (July 1865): 1–42. First edition in book form originally published in London: Trübner, 1865. Second and final edition in London: Trübner, 1866. Second edition also available as “Auguste Comte and Positivism” in Essays on Ethics, Religion and Society (pp. 261–368) as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1873. Autobiography. Published posthumously in eight volumes and edited by Helen Taylor. London: Longmans, Green, Reader, and Dyer, 1873. Numerous reprints. Reprinted as Autobiography of John Stuart Mill Published for the First Time without Alterations or Omissions from the Original Manuscript in the Possession of Columbia University. New York: Columbia University Press, 1924. Subsequently reprinted as John Stuart Mill: Autobiography. Edited with an introduction and notes by Jack Stillinger. Boston: Houghton Mifflin Company, 1969. Also available in Autobiography and Literary Essays (pp. 1–290) as part of vol. 1 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson and Jack Stillinger. Toronto: University of Toronto Press, 1981. Mill, John Stuart. 1874. Three Essays on Religion. Published posthumously and edited by Helen Taylor. London: Longmans, Green, Reader, and Dyer, 1874. Also available in Essays on Ethics, Religion and Society (pp. 369–489) as part of vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press, 1969. Mill, John Stuart. 1963–1991. Collected Works of John Stuart Mill. Thirty-three volumes. General editors: Francis Ethelbert Louis Priestley and John Mercel Robson. Toronto (University of Toronto Press) and London (Routledge and Kegan Paul): 1963–1991. Miller, Dale E. 2004. “On Millgram on Mill.” Utilitas 16:1 (March): 96–108. Miller, Dale E. 2010a. “Brown on Mill’s Moral Theory: A Critical Response.” Politics, Philosophy & Economics 9:1 (February): 47–66. Miller, Dale E. 2010b. J. S. Mill: Moral, Social and Political Thought. Cambridge: Polity Press. Miller, Dale E. 2010c. “Mill, Rule Utilitarianism, and the Incoherence Objection.” Chapter 4 of John Stuart Mill and the Art of Life (pp. 94–116). Edited by Ben Eggleston, Dale E. Miller, and David Weinstein. Oxford: Oxford University Press.
366
Works Cited
Miller, Dale E. 2014a. “John Stuart Mill’s Moral, Social, and Political Philosophy.” Chapter 18 of The Oxford Handbook of British Philosophy in the Nineteenth Century (pp. 365–382). Edited by William J. Mander. Oxford: Oxford University Press. Miller, Dale E. 2014b. “Rule Utilitarianism.” Chapter 7 of The Cambridge Companion 165). Edited by Ben Eggleston and Dale E. Miller. to Utilitarianism (pp. 146– Cambridge: Cambridge University Press. Miller, Dale E. 2017. “Mill’s Act-Utilitarian Interpreters on Utilitarianism Chapter V Paragraph 14.” Canadian Journal of Philosophy 47:5: 674–693. Miller, Dale E. 2019. “Mill’s Proof of the Principle of Utility.” 1000-Word Philosophy: An Introductory Anthology (3 September 2019), https://1000wordphilosophy.com/2019/ 09/03/mills-proof-of-the-principle-of-utility/. Miller, Richard B. 2009. “Actual Rule Utilitarianism.” The Journal of Philosophy 106:1 (January): 5–28. Millgram, Elijah. 2000. “Mill’s Proof of the Principle of Utility.” Ethics 110:2 (January): 282–310. Reprinted as chapter 2 of his Ethics Done Right: Practical Reasoning as a Foundation for Moral Theory (pp. 56–88). Cambridge: Cambridge University Press. Mitchell, Dorothy. 1970. “Mill’s Theory of Value.” Theoria 36:2 (August): 100–115. Moore, George Edward. 1903. Principia Ethica. Cambridge: Cambridge University Press. Reprinted in 1922 with no alterations beyond the correction of “a few misprints and grammatical mistakes” as reported by the author in an addendum to the preface. A “revised edition” with an introduction by Thomas Baldwin was published in 1993, together with Moore’s incomplete “Preface to the Second Edition” and two essays: “Free Will” (1912) and “The Conception of Intrinsic Value” (1922). References in the present volume use the pagination of the 1993 edition, preceded by chapter and section numbers for the convenience of readers consulting any other edition. Moser, Shia. 1963. “A Comment on Mill’s Argument for Utilitarianism.” Inquiry 6:1/ 4: 308–318. Murray, James Augustus Henry. 1897a. “The Greatest Happiness of the Greatest Number.” Notes and Queries S8:xi 279 (1 May 1897): 347. Murray, James Augustus Henry. 1897b. “The Greatest Happiness of the Greatest Number.” Notes and Queries S8:xi 281 (15 May 1897): 392. Murray, Robert Malcolm, and Nebojsa Kujundzic. 2005. Critical Reflection: A Textbook for Critical Thinking. Montreal and Kingston: McGill-Queen’s University Press. Nakhnikian, George. 1951. “Value and Obligation in Mill.” Ethics 62:1 (October): 33–40. Nakhnikian, George. 1963. “On the Naturalistic Fallacy.” Chapter 5 of Morality and the Language of Conduct (pp. 145–158). Edited by Hector-Neri Castañeda and George Nakhnikian. Detroit: Wayne State University Press. Olscamp, Paul J. 1969. The Moral Philosophy of George Berkeley. Archives Internationales D’Histoire Des Idées /International Archives of the History of Ideas, No. 33. The Hague: Martinus Nijhoff.
Works Cited
367
Orsi, Francesco. 2021. “Mill’s Proof and the Guise of the Good.” Philosophical Explorations: An International Journal for the Philosophy of Mind and Action 24:1: 93–105. Paley, William. 1785. The Principles of Moral and Political Philosophy. London: Robert Faulder. Perronet Thompson, Thomas. 1829. “ ‘Greatest Happiness’ Principle.” Running header, often cited as the title, of an anonymous article by Thomas Perronet Thompson, with anonymous excerpts from Jeremy Bentham’s “Article on Utilitarianism” (Bentham 1829/1983), in response to an anonymous critique of James Mill’s Essays on Government, Jurisprudence, Liberty of the Press, and Law of Nations (1823/1828), by Thomas Babington Macaulay (The Edinburgh Review 49:97 [March 1829]: 159– 189), with the responding article bearing the actual title: “Edinburgh Review, No. XCVII: Article on Mill’s Essays on Government, &c.” Published as Article 16 of The Westminster Review 11:21 (July 1829): 254–268. Perry, Ralph Barton. 1926. General Theory of Value: Its Meaning and General Principles Construed in Terms of Interest. New York: Longmans, Green, and Company. Persson, Ingmar. 2000. “Mill’s Derivation of the Intrinsic Desirability of Pleasure.” History of Philosophy Quarterly 17:3 (July): 297–310. Plamenatz, John Petrov. 1949. The English Utilitarians. Second edition, 1958. Oxford: Basil Blackwell. Plato. 1900–1907. Platonis Opera. Five volumes. Critical texts with introductions and notes. Edited by John Burnet. Oxford Classical Texts. Oxford: Oxford University Press. Reprinted 1967–1968. New edition under way since 1995. Plato. 1997. Complete Works. English translations by various hands. Edited, with an introduction and notes, by John Madison Cooper. Indianapolis: Hackett Publishing Company. Popkin, Richard H. 1950. “A Note on the ‘Proof’ of Utility in J. S. Mill.” Ethics 61:1 (October): 66–68. Postow, Betsy Carol. 1977. “Generalized Act Utilitarianism.” Analysis 37:2 (January): 49–52. Priestley, Joseph. 1768. An Essay on the First Principles of Government; and on the Nature of Political, Civil, and Religious Liberty. Dublin: James Williams, 1768. Second edition (“corrected and enlarged”) published as An Essay on the First Principles of Government, and on the Nature of Political, Civil, and Religious Liberty, Including Remarks on Dr. Brown’s Code of Education, and on Dr. Balguy’s Sermon on Church Authority. London: Joseph Johnson, 1771. Prior, Arthur N. 1949. Logic and the Basis of Ethics. Oxford: Oxford University Press. Quine, Willard Van Orman. 1951. Mathematical Logic. Revised edition. Cambridge: Harvard University Press.
368
Works Cited
Quine, Willard Van Orman. 1978. “On the Nature of Moral Values.” In Values and Morals: Essays in Honor of William Frankena, Charles Stevenson, and Richard Brandt (pp. 37–45). Edited by Alvin I. Goldman and Jaegwon Kim. Dordrecht: D. Reidel Publishing Company (Springer). Reprinted in Critical Inquiry 5:3 (Spring 1979): 471–480. Also available as essay 6 in Quine’s Theories and Things (pp. 55–66). Cambridge: Harvard University Press, 1981. References in the present volume give the pagination of the 1978 edition. Quinton, Anthony Meredith. 1973. Utilitarian Ethics. London: Macmillan Press. Ralkowski, Mark. 2007. “Was Plato a Platonist?” Ancient Philosophy 27:2 (Fall): 391–402. Raphael, David Daiches. 1955. “Fallacies in and about Mill’s Utilitarianism.” Philosophy 30:115 (October): 344–357. Raphael, David Daiches. 1994. “J. S. Mill’s Proof of the Principle of Utility.” Utilitas 6:1 (May): 55–63. Rashdall, Hastings. 1899. “Can There Be a Sum of Pleasures?” Mind (New Series) 8:31 (July): 357–382. Rawls, John. 1955. “Two Concepts of Rules.” The Philosophical Review 64:1 (January): 3–32. Raynor, David R. 1987. “Hutcheson’s Defence against a Charge of Plagiarism.” Eighteenth- Century Ireland /Iris an dá Chultúr 2: 177–181. Reddie, James. 1840. Inquiries Elementary and Historical in the Science of Law. London: Longman, Orme, Brown, Green, and Longmans. Edinburgh: William Blackwood and Sons. Glasgow: John Smith and Son. Rescher, Nicholas. 1964. Introduction to Logic. New York: St. Martin’s Press. Riley, Jonathan. 1993. “On Quantities and Qualities of Pleasure.” Utilitas 5:2 (November): 291–300. Riley, Jonathan. 1999. “Is Qualitative Hedonism Incoherent?” Utilitas 11:3 (November): 347–358. Riley, Jonathan. 2003. “Interpreting Mill’s Qualitative Hedonism.” The Philosophical Quarterly 53:212 (July): 410–418. Riley, Jonathan. 2008a. “Millian Qualitative Superiorities and Utilitarianism, Part I.” Utilitas 20:3 (September): 257–278. Riley, Jonathan. 2008b. “What Are Millian Qualitative Superiorities?” Prolegomena: Journal of Philosophy 7:1: 61–79. Riley, Jonathan. 2009. “Millian Qualitative Superiorities and Utilitarianism, Part II.” Utilitas 21:2 (June): 127–143. Riley, Jonathan. 2010. “Mill’s Extraordinary Utilitarian Moral Theory.” Politics, Philosophy & Economics 9:1 (February): 67–116. Riley, Jonathan. 2014. “Different Kinds of Pleasure.” Chapter 8 of Mill’s A System of Logic: Critical Appraisals (pp. 170–191). Edited by Antis Loizides. New York: Routledge. Robson, John Mercel. 1969. “Textual Introduction” to Essays on Ethics, Religion and Society (pp. cxv–cxxxix), vol. 10 of the Collected Works of John Stuart Mill. Edited by John Mercel Robson. Toronto: University of Toronto Press.
Works Cited
369
Rohatyn, Dennis A. 1971. “Hall and Mill’s Proof.” The Southwestern Journal of Philosophy 2:3 (Winter): 113–118. Rohatyn, Dennis A. 1975. “How to Prove the Principle of Utility.” Chapter 3 of his Naturalism and Deontology: An Essay on the Problems of Ethics (pp. 22–65). The Hague and Paris: Mouton and Company. Rokstad, Konrad. 2011. “Was Plato a Platonist?” Phenomenology/Ontopoiesis Retrieving Geo-Cosmic Horizons of Antiquity (pp. 15–22). Edited by Anna-Teresa Tymieniecka. Analecta Husserliana (The Yearbook of Phenomenological Research) No. 110. Dordrecht: Springer. Rosen, Frederick. 2003. Classical Utilitarianism from Hume to Mill. London: Routledge. Rosen, Frederick. 2013. Mill. Founders of Modern Political and Social Thought. Oxford: Oxford University Press. Rousseau, Jean-Jacques. 1762. Du Contrat Social, ou, Principes du Droit Politique. Amsterdam: Marc Michel Rey. Rowe, William Leonard. 1962. “The Fallacy of Composition.” Mind (New Series) 71:281 (January): 87–92. Russell, Bertrand Arthur William. 1945. “The Utilitarians.” Chapter 26 of A History of Western Philosophy, and Its Connection with Political and Social Circumstances from the Earliest Times to the Present Day (pp. 773–782). New York: Simon and Schuster. Appears on pp. 801–809 of the British edition, coming out the very next year in London: George Allen and Unwin, 1946. Ryan, Alan James. 1965. “John Stuart Mill’s Art of Living.” The Listener 74 (21 October 1965): 620–621. Reprinted in J. S. Mill On Liberty in Focus (pp. 162–168). Edited by John Gray and G. W. Smith. London: Routledge, 1991. Ryan, Alan James. 1966. “Mill and the Naturalistic Fallacy.” Mind (New Series) 75:299 (July): 422–425. Ryan, Alan James. 1970. The Philosophy of John Stuart Mill. London: Macmillan and Company. Ryan, Alan James. 1974. J. S. Mill. London: Routledge and Kegan Paul. Reissued as part of Routledge Revivals. Abingdon: Routledge, 2016. Ryan, Alan James. 2014. “A System of Logic and the ‘Art of Life.’ ” Chapter 11 of Mill’s A System of Logic: Critical Appraisals (pp. 246–261). Edited by Antis Loizides. New York: Routledge. Ryberg, Jesper. 2002. “Higher and Lower Pleasures—Doubts on Justification.” Ethical Theory and Moral Practice 5:4 (December): 415–429. Sapontzis, Steve F. 2014. “In Defense of the Pig.” Journal of Animal Ethics 4:1 (Spring): 5–17. Saunders, Ben. 2010. “J. S. Mill’s Conception of Utility.” Utilitas 22:1 (March): 52–69. Sayre-McCord, Geoffrey. 2001. “Mill’s ‘Proof’ of the Principle of Utility: A More Than Half-Hearted Defense.” Social Philosophy and Policy 18:2 (Spring): 330–360.
370
Works Cited
Scarre, Geoffrey. 1997. “Donner and Riley on Qualitative Hedonism.” Utilitas 9:3 (November): 351–360. Schaupp, Kristin. 2013. “Books before Chocolate? The Insufficiency of Mill’s Evidence for Higher Pleasures.” Utilitas 25:2 (June): 266–276. Schmidt–Petri, Christoph. 2003. “Mill on Quality and Quantity.” Philosophical Quarterly 53:210 (January): 102–104. Schmidt– Petri, Christoph. 2006. “On an Interpretation of Mill’s Qualitative Utilitarianism.” Prolegomena: Journal of Philosophy 5:2: 165–171. Schneewind, Jerome Borges. 1976. “Concerning Some Criticisms of Mill’s Utilitarianism, 1861–76.” In James and John Stuart Mill: Papers of the Centenary Conference (pp. 35–54). Edited by John Mercel Robson and Michael Laine. Toronto: University of Toronto Press. Schneewind, Jerome Borges. 1977. Sidgwick’s Ethics and Victorian Moral Philosophy. Oxford: Clarendon Press. Schultz, Bart. 2017. The Happiness Philosophers: The Lives and Works of the Great Utilitarians. Princeton: Princeton University Press. Scolnicov, Samuel. 1994. “Socrates, Plato and the Development of Reason: A Rejoinder to Professor Sichel.” Studies in Philosophy and Education 13:2 (June): 149–156. Searle, John R. 1964. “How to Derive ‘Ought’ from ‘Is.’ ” The Philosophical Review 73:1 (January): 43–58. Searle, John R. 1969. “The Naturalistic Fallacy Fallacy.” Chapter 6.1 of his Speech Acts: An Essay in the Philosophy of Language (pp. 132–136). Cambridge: Cambridge University Press. Seth, James. 1896. “Is Pleasure the Summum Bonum?” Ethics 6:4 (July): 409–424. Seth, James. 1908. “The Alleged Fallacies in Mill’s ‘Utilitarianism.’ ” The Philosophical Review 17:5 (September): 469–488. Shackleton, Robert. 1972. “The Greatest Happiness of the Greatest Number: The History of Bentham’s Phrase.” Studies on Voltaire and the Eighteenth Century 90: 1461–1482. Sheng, Ching-Lai (Qinglai). 1991. A New Approach to Utilitarianism: A Unified Utilitarian Theory and Its Application to Distributive Justice. Dordrecht: Kluwer Academic Publishers. Sidgwick, Henry. 1874. The Methods of Ethics. London: Macmillan and Company. Numerous editions and reprints. Sixth (1901) and seventh (1907) editions published posthumously, the sixth with the personal revisions and further instructions of the author, the seventh (and final) with minor editorial emendations. Sidgwick, Henry. 1877. “Bentham and Benthamism in Politics and Ethics.” The Fortnightly Review (New Series) 21:125 (1 May 1877): 627–652. Sidgwick, Henry. 1879. “The Establishment of Ethical First Principles.” Mind 4:13 (January): 106–111.
Works Cited
371
Sidgwick, Henry. 1886. Outlines of the History of Ethics. London: Macmillan and Company. Numerous editions and reprints. Singer, Marcus George. 1955. “Generalization in Ethics.” Mind (New Series): 64:255 (July): 361–375. Singer, Marcus George. 1961. Generalization in Ethics: An Essay in the Logic of Ethics, with the Rudiments of a System of Moral Philosophy. New York: Alfred A. Knopf. Sinnott-Armstrong, Walter, and Robert Fogelin. 2015. Understanding Arguments: An Introduction to Informal Logic. Ninth edition. Stamford: Cengage Learning. Skinner, Quentin. 1969. “Meaning and Understanding in the History of Ideas.” History and Theory 8:1 (February): 3–53. A “much abbreviated and extensively revised version” bearing the same title was reprinted as c hapter 4 of his Visions of Politics. Volume 1: Regarding Method (pp. 57–89). Cambridge: Cambridge University Press, 2002. Skorupski, John. 1989. John Stuart Mill. New York: Routledge, Chapman, and Hall. Smart, John Jamieson Carswell. 1956. “Extreme and Restricted Utilitarianism.” The Philosophical Quarterly 6:25 (October): 344–354. Reprinted in Theories of Ethics (pp. 171–183), edited by Philippa Foot (Oxford: Oxford University Press, 1967), and in Contemporary Utilitarianism (pp. 99–115), edited by Michael D. Bayles (New York: Anchor Books, 1968). Smart himself reports that he prefers the latter two versions over the original, which he says “contains bad errors” (Smart and Williams 1973, 10, n. 1). Smart, John Jamieson Carswell, and Bernard Arthur Owen Williams. 1973. Utilitarianism: For and Against. Cambridge: Cambridge University Press. The first part of the book is a revision of Smart’s monograph: An Outline of a System of Utilitarian Ethics (Melbourne: Melbourne University Press, 1961). Sobel, Jordan Howard. 1968. “Rule Utilitarianism.” Australasian Journal of Philosophy 46:2 (August): 146–165. Soles, David. 1998. “Yet Another Paper on Mill’s Proof.” Southwest Philosophy Review 14:2 (July): 29–44. Sorley, William Ritchie. 1885. On the Ethics of Naturalism. Edinburgh and London: William Blackwood and Sons. Second (“revised”) edition published in 1904 as The Ethics of Naturalism: A Criticism. Spence, Gordon W. 1968. “The Psychology behind J. S. Mill’s ‘Proof.’ ” Philosophy 43:163 (January): 18–28. Stephen, Leslie. 1876. History of English Thought in the Eighteenth Century. Two volumes. London: Smith, Elder, and Company. Second edition, 1881. Stephen, Leslie. 1900. The English Utilitarians. Three volumes. Vol. 1: Jeremy Bentham. Vol. 2: James Mill. Vol. 3: John Stuart Mill. London: Duckworth and Company. Stewart, Robert Scott. 1993. “Art for Argument’s Sake: Saving Mill from the Fallacy of Composition.” The Journal of Value Inquiry 27:3/4 (December): 443–453.
372
Works Cited
Stocker, Michael. 1969. “Mill on Desire and Desirability.” Journal of the History of Philosophy 7:2 (April): 199–201. Stout, Alan Ker. 1954. “But Suppose Everyone Did the Same.” Australasian Journal of Philosophy 32:1 (May): 1–29. Strasser, Mark Philip. 1991. The Moral Philosophy of John Stuart Mill: Toward Modifications of Contemporary Utilitarianism. Wakefield: Longwood Academic Press. Stroll, Avrum. 1965. “Mill’s Fallacy.” Dialogue 3:4 (March): 385–404. Sumner, Leonard Wayne. 1979. “The Good and the Right.” In New Essays on John Stuart Mill and Utilitarianism. Edited by Wesley E. Cooper, Kai Nielsen, and Steven C. Patten. Canadian Journal of Philosophy (Supplementary Vol. 5): 99–114. Sumner, Leonard Wayne. 2006. “Mill’s Theory of Rights.” Chapter 8 of The Blackwell Guide to Mill’s Utilitarianism (pp. 184– 198). Edited by Henry Robison West. Oxford: Blackwell Publishing. Sutherland, James. 1886. “An Alleged Gap in Mill’s Utilitarianism.” Mind 11:44 (October): 597–599. Timmermann, Jens. 2005. “Why Kant Could Not Have Been a Utilitarian.” Utilitas 17:3 (November): 243–264. Turner, Piers Norris. 2015a. “Punishment and Discretion in Mill’s Utilitarianism.” Utilitas 27:2 (June): 165–178. Turner, Piers Norris. 2015b. “Rules and Right in Mill.” Journal of the History of Philosophy 53:4 (October): 723–745. Urmson, James Opie. 1953. “The Interpretation of the Moral Philosophy of J. S. Mill.” The Philosophical Quarterly 3:10 (January): 33–39. Urmson, James Opie. 1958. “Saints and Heroes.” In Essays in Moral Philosophy (pp. 198– 216). Edited by Abraham Irving Melden. Seattle: University of Washington Press. von Neumann, John, and Oskar Morgenstern. 1944. Theory of Games and Economic Behavior. Princeton: Princeton University Press. Second edition, 1947. Third edition, 1953. Wall, George B. 1971. “More on the Equivalence of Act and Rule Utilitarianism.” Philosophical Studies 22:5/6 (December): 91–95. Wall, Grenville. 1982. “Mill on Happiness as an End.” Philosophy 57:222 (October): 537–541. Walton, Douglas N. 1982. Topical Relevance in Argumentation. Amsterdam: John Benjamins Publishing Company. Walton, Douglas N. 1987. Informal Fallacies: Towards a Theory of Argument Criticisms. Amsterdam: John Benjamins Publishing Company. Walton, Douglas N. 1989. Informal Logic: A Handbook for Critical Argumentation. Cambridge: Cambridge University Press. Warnock, Mary. 1978. Ethics since 1900. Third edition. Originally published in 1960. Oxford: Oxford University Press.
Works Cited
373
Weinberg, Adelaide. 1963. Theodor Gomperz and John Stuart Mill. Travaux de Droit, d’Économie, de Sociologie et de Sciences Politiques. Vol. 16. Geneva: Librairie Droz. Wellman, Carl P. 1959. “A Reinterpretation of Mill’s Proof.” Ethics 69:4 (July): 268–276. Wellman, Carl P. 1961. The Language of Ethics. Cambridge: Harvard University Press. Werkmeister, William Henry. 1957. An Introduction to Critical Thinking: A Beginner’s Text in Logic. Revised edition. Lincoln: Johnsen Publishing Company. Wertz, Spencer K. 1971. “Composition and Mill’s Utilitarian Principle.” The Personalist 52:3 (Summer): 417–431. West, Henry Robison. 1972. “Reconstructing Mill’s ‘Proof’ of the Principle of Utility.” Mind (New Series): 81:322 (April): 256–257. West, Henry Robison. 1976a. “Mill’s Moral Conservatism.” Midwest Studies in Philosophy 1 (Studies in the History of Philosophy): 71–80. West, Henry Robison. 1976b. “Mill’s Qualitative Hedonism.” Philosophy 51:195 (January): 97–101. West, Henry Robison. 1982. “Mill’s ‘Proof’ of the Principle of Utility.” Chapter 1 of The Limits of Utilitarianism (pp. 23–34). Edited by Harlan B. Miller and William Hatton Williams. Minneapolis: University of Minnesota Press. West, Henry Robison. 2004. An Introduction to Mill’s Utilitarian Ethics. Cambridge: Cambridge University Press. West, Henry Robison. 2006. “Mill’s ‘Proof’ of the Principle of Utility.” Chapter 7 of The Blackwell Guide to Mill’s Utilitarianism (pp. 174–183). Oxford: Blackwell Publishing. West, Henry Robison. 2017. “The Proof.” Chapter 22 of A Companion to Mill (pp. 328– 341). Edited by Christopher Macleod and Dale E. Miller. Oxford: John Wiley and Sons. Whately, Richard. 1848. Elements of Logic. Ninth edition. London: Edward Lumley. Whewell, William. 1852. Lectures on the History of Moral Philosophy in England. London: John W. Parker and Son. Cambridge: John Deighton. Second edition (“new edition with additional lectures”) published in two parts with separate pagination in a single volume. Cambridge: Deighton, Bell, and Company. London: Bell and Daldy, 1862. Wilson, Fred. 1982. “Mill’s Proof That Happiness Is the Criterion of Morality.” Journal of Business Ethics 1:1 (February): 59–72. Wilson, Fred. 1983. “Mill’s ‘Proof’ of Utility and the Composition of Causes.” Journal of Business Ethics 2:2 (May): 135–155. Wilson, George. 1786. Letter to Jeremy Bentham (24 September 1786) reproduced in The Works of Jeremy Bentham (1843, vol. 10, 163–164). Wilson, George. 1788. Letter to Jeremy Bentham (30 November 1788) reproduced in The Works of Jeremy Bentham (1843, vol. 10, 194–195). Wilson, T. 1897. “The Greatest Happiness of the Greatest Number.” Notes and Queries S8:xi 281 (15 May 1897): 392.
374
Works Cited
Woods, John, and Douglas N. Walton. 1976. “Fallaciousness without Invalidity?” Philosophy & Rhetoric 9:1 (Winter): 52–54. Wright, Darryl F. 1994. “Diagnosing the Naturalistic Fallacy: Principia Ethica Revisited.” The Southern Journal of Philosophy 32:4 (Winter): 465–482. Yim, Dan. 2008. “The Logic and Mill’s Infamous Proof in Utilitarianism.” British Journal for the History of Philosophy 16:4: 773–788. Zinkernagel, Peter. 1952. “Revaluation of J. S. Mill’s Ethical Proof.” Theoria 18:1/2 (April): 70–77.
Index Berger, Fred R. 93n3, 100n9, 101n10, 108n13, Abaté, Charles J. 213n11 312n20, 325, 325n29 act utilitarianism. See under utilitarianism actual consequences. See under consequences Bergman, Ingrid 166–167, 250 Adams, F. 44n29 Berkeley, George 294–295 Albee, Ernest 14 Black, Beth 124n2 Alican, Necip Fikri 275n5 Blau, Adrian 333n32 Anderson, Elizabeth S. 100n9 Boole, George 7 Aristotle 21, 23n18, 54, 233, 233n5 Bowring, John 15–30, 43–48, 54–70, 78– Arrhenius, Gustaf 100n9 91, 102 Atkinson, Ronald Field 6n5, 119 Bradley, Francis Herbert 4n3, 99n8, 123, 126, average happiness. See under happiness 126n4, 137–141, 143, 144, 197n6, 289n6 Brandreth, Henry Samuel 313 Bacon, Francis 21–23, 25, 27, 73 Brandt, Richard Booker 294n9, 295n11, 296, 296n12, 324, 326, 339 Baker, Gordon Park 232n1 Brink, David Owen 6n5, 100n9, 312n20, Baker, John M. 6n5, 325 317n23, 323 Baldwin, Thomas 232n1, 234n6 Bronaugh, Richard N. 100n9 Ball, Terence 305n18 Brown, Donald George 6n5, 98, 98n7, Bar-Hillel, Yehoshua 211n8 102n12, 105, 117–118, 297, 312n20, Barrington, Daines 70 Barrow, Robin 105 315n22, 323 Baumrin, Bernard H. 232n1 Bruening, William H. 232n1 Beattie, James 57 Burns, James Henderson 81–83 Beaumont, William Comyns 44n29 Burton, John Hill 74 Beccaria, Cesare 21, 22, 22n16, 24, 30, 39n26, Butler, Joseph 234, 234n7 40, 45–48, 51, 53, 62, 70, 73, 74, 75, 76– Bykvist, Krister 15n2 77, 80 Capaldi, Nicholas 305n18 Bentham, Jeremy 98, 101, 102, 102n11, 104, Carlisle, Janice 305n18 107, 110, 112, 113n15, 142, 175n6, 233, Carlyle, Thomas 5n4, 224n18 233n5, 262, 276, 277, 302n16, 304n17, Chambers, Ephraim 23 305n17, 341 Clark, George A. 6n5, 183n9 Benthamite sources on the history of Clarke, Chamberlain 70 utilitarianism 14–30 Clarke, Samuel 55, 57, 135 Bentham as an anchor for historical Cocceji, Heinrich von 50n33 insight into utilitarianism 30–39 Cocceji, Samuel von 50, 50n33 origins and development of utilitarian Cohen, Carl 124n2, 209 nomenclature 39–53 Cohen, Stephen 6n5 Bentham’s debt to predecessors and Cole, Richard 212–213 contemporaries 53–74 patterns of indirect inspiration and Comte, Auguste 104, 218–219 transmission 74–77 consequences actual, intended, and Bentham’s own terminological foreseeable 334–337 predilections 77–91 Bentham’s principle of utility 96–98 consequentialism 93–94, 129, 290–291, John Stuart Mill’s criticism of 303–305 Bentham 303–305 Coope, Christopher Miles 304n17, 312n20
376 Index Cooper, Neil 6n5 Copi, Irving M. 124n2, 209 Copp, David 102n12, 325 Creighton, James Edwin 124n2 Crimmins, James E. 15n2 Crisp, Roger. 15n2, 100n9, 312n20, 317n23, 323 Cudworth, Ralph 55 Cumberland, Richard 55 Cupples, Brian 297, 323 d’Alembert, Jean-Baptiste le Rond 23 Dahl, Norman O. 100n9 Day, John Patrick 6n5 Dewey, John 4n3, 126, 126n4, 127, 143, 144– 146, 146–148 Diderot, Denis 23, 76 Doble, Charles Edward 44n29 Donner, Wendy 100n9, 324, 324n26, 326 Dryer, Douglas Poole 103n12, 325 Duclos, Charles Pinot 76 Duncan, Elmer H. 232n1, 233n4 Dunn, John 333n32 Ebenstein, Lanny 100n9 Edwards, Rem B. 325 Eggleston, Ben 14, 15, 15n2, 117n16, 294n9, 312n20, 324–327, 330–332 egoism ethical egoism 156 psychological egoism 28–29, 70, 72, 72n48, 138–139, 156 Eidous, Marc Antoine 75, 76 epistemic justification 177–181, 257 ethical egoism 156 ethical hedonism 28–29, 31, 72, 135, 139, 144, 148–149, 259, 263, 281–291 ethical intuitionism 174, 236, 290 ethical justification 96–98, 106–108, 111, 115– 116, 249, 320 fallacies (alleged in Mill’s proof) fallacy of composition 124–125, 127–137, 139–141, 143–144, 146, 148–151, 155, 156– 157, 185–231 fallacy of division 134–135, 143–146, 157, 205, 214
fallacy of equivocation 124–125, 127, 136, 148–149, 152–154, 156–157, 161–184 fallacy of four terms 168, 235, 246n17 naturalistic fallacy 123, 125, 126, 152, 154, 155, 156, 232–252 Fearnside, W. Ward 214n12 Fénelon, François de Salignac de la Mothe 71 Fieser, James 232n1 Fletcher, Guy 6n5 Fogelin, Robert 124n2 foreseeable consequences. See under consequences Foster, John 46n30 Frankena, William Klaas 107, 234–239 Frege, Friedrich Ludwig Gottlob 7 Fuchs, Alan E. 312n20, 324, 326 Fullinwider, Robert K. 6n5 Fyfe, Andrew Thomas 6n5 Gaus, Gerald 325 Gauthier, David P. 232n1 Geiger, George R. 232n1 general happiness 216–223, 259–267 Gensler, Harry J. 213n10 George iii, King of England 29 Gerson, Lloyd P. 305n17 Gibbs, Benjamin 100n9 Goldworth, Amnon 17, 17n4, 21n9, 30n20 Gomperz, Theodor 171–173 Gray, John 101n10, 325–330 greatest happiness principle. See under principle of utility Green, Thomas Hill 99n8 Griffin, Nicholas 6n5 Grote, George 104, 164n2 Grote, John 4n3, 126, 126n4, 127–134, 142 Hacker, Peter Michael Stephan 232n1 Haezrahi, Pepita 6n5 Hales, Steven D. 100n9 Halévy, Élie 14 Hall, Everett W. 174–175, 175n6, 197n6 Hamblin, Charles Leonard 209, 214n12 Hancock, Roger 103n12 happiness distinguished from pleasure 276–278 general happiness 216–223, 259–267
377
Index parts and ingredients of happiness 116, 219–221, 267–279 promotion and maximization of happiness 337–344 total vs. average happiness 217–220, 337–344 Hare, Richard Mervyn 293, 293n7, 295n11, 298n13, 306n19 Harrison, Jonathan 101n10, 295n11, 325 Harrod, Henry Roy Forbes 293, 293n7 Harsanyi, John C. 296n11 Hartley, David 21, 22, 22n14, 55, 62–63, 66, 73 Hauskeller, Michael 100n9 Häyry, Heta 294n9 Häyry, Matti 294n9 Heans, S. J. 6n5 hedonism ethical hedonism 28–29, 31, 72, 72n48, 135, 139, 144, 148–149, 259, 263, 281–291 psychological hedonism 135, 138–139, 142, 148–151, 271 Helvétius, Claude Adrien 21, 22, 22n15, 30, 39n26, 46n30, 55, 56, 62–73, 76 Heydt, Colin 15n2, 294n9 higher vs. lower pleasures. See under pleasure Himma, Kenneth Einar 325, 326n30 Hoag, Robert W. 6n5, 100n9 Hobbes, Thomas 233n5, 288n5 Hocutt, Max 304n17 Hodgson, David Hargraves 326n30 Hollander, Samuel 14 Holther, William B. 214n12 Hooker, Brad 294n9 Horace 21, 54, 69 Hruschka, Joachim 49–52, 295n10 Hume, David 21, 21n9, 22, 22n13, 24, 55, 56, 56n38, 57, 62–74, 80 Hutcheson, Francis 30, 39n26, 48–58, 62, 74–77
Irwin, Terence Henry 6n5, 14n1, 100n9, 101n10 Jacobson, Daniel 326 Jaucourt, Louis de 76 Jones, Hardy 6n5 Jones, Henry 140, 145, 197–198, 199, 200, 216, 229, 231 Joyce, Richard 232n1, 234n6 justice justice and Bentham’s principle of utility 39, 85–89 justice and Mill’s principle of utility 5, 100, 101n10, 230, 261–262, 266, 278, 283n1, 314–316 justification epistemic justification 177–181, 257 ethical justification 6, 95, 96–98, 106– 108, 111, 115–116, 249, 320 Kahane, Howard 213n10, 213n11 Kant, Immanuel 9, 55, 135, 174, 233n5, 243n15, 267, 293–294 Katz, Jerrold J. 232n1, 232n2 Kavka, Gregory S. 298n13 Kawana, Yuichiro 305n18 Kitcher, Philip 232n1, 234n6 Kleinig, John 6n5 Kolnai, Aurel 232n1 Kretzmann, Norman 6n5, 181n7 Kringelbach, Morten 100n9 Kujundzic, Nebojsa 124n2 Kupfer, Joseph 294n9
Lang, Berel 6n5 Leibniz, Gottfried Wilhelm 49–58, 62, 75, 233n5, 243n15, 294–295 Leidhold, Wolfgang 51n34 Lewis, George Cornewall 4n2 Lewy, Casimir 232n1, 236n12 Lindström, Aron 166 Locke, John 3, 19, 21, 21n10, 27, 55, 55n37, 70–71, 71n47 ingredients of happiness. See under happiness intended consequences. See under Loizides, Antis 117n16 consequences Long, Douglas G. 333n32 intuitionism Long, Roderick T. 100n9 ethical intuitionism 174, 232n3, 233n3, Long, W. H. 7n5 236, 270n2, 290 Loring, L. M. 232n1
378 Index Luebke, Neil R. 232n1 Lyons, David 81–83, 101n10, 103n12, 294, 296n11, 297, 298n13, 304n17, 324, 326, 326n31, 329 Mabbott, John David 296n11, 297, 323 Macaulay, Thomas Babington 16, 16n3, 17, 25, 26 Mackenzie, John Stuart 5n3, 125, 126, 126n4, 148–151, 191–192, 228, 288n5, 289n6 Macleod, Christopher 7n5 Mandelbaum, Maurice 7n5, 183n9, 297, 324 Margolis, Joseph 7n5 Marshall, John 7n5 Martin, Rex 100n9, 325–330 Mawson, Tim 7n5 McMahon, Kenneth 124n2, 209 McNeilly, Francis Stewart 7n5 Mercier de La Rivière, Pierre-Paul 76 Mill, James 16, 16n3, 25–26, 39n26, 55, 82, 83, 83n53, 264–265, 336–337 Mill, John Stuart letters (from Mill to recipient) Henry Samuel Brandreth (Letter 1028, 9 February 1867) 313 Thomas Carlyle (Letter 95, 12 January 1834) 224n18 Theodor Gomperz (Letter 1227, 23 April 1868) 172 George Grote (Letter 525, 10 January 1862) 104, 164n2 Henry Jones (Letter 1257, 13 June 1868) 140, 145, 197, 216, 229 John Venn (Letter 1717A, 14 April 1872) 311 works (quotations) Auguste Comte and Positivism (1865) 104, 219 Autobiography (1873) 290 “Coleridge” (1840) 230 On Liberty (1859) 219–220 “Remarks on Bentham’s Philosophy” (1833) 303–304 “Sedgwick’s Discourse” (1835) 264, 335 A System of Logic (1843) 109, 111, 117, 118, 167, 170, 175–177, 214, 215, 221, 223n17, 225, 226n20, 274, 275n6, 286, 301n15, 303, 305n18, 309–310, 316, 335
Three Essays on Religion (1874) 264, 341 “Use and Abuse of Political Terms” (1832) 3–4 Utilitarianism (1861) 92n1, 93n1, 94n4, 94n5, 95, 99, 104, 108, 109, 110, 111, 111n14, 112, 113–114, 113n15, 115, 116–117, 118, 118n17, 120, 123–124, 129, 161, 173–174, 176, 177, 178, 179, 181, 186, 204, 217n14, 220, 222, 222n15, 228, 229–230, 249, 255, 256, 257, 261, 262, 265–267, 268, 269, 269n1, 270n2, 270n3, 271, 272, 273, 274n4, 275, 276, 276n6, 277, 278, 282–283, 284, 284n2, 285, 287n3, 289, 290, 299– 300, 301, 301n14, 301n15, 302–303, 302n16, 307, 308, 310, 312, 314–315, 321, 322, 335, 336, 341 “Whewell on Moral Philosophy” (1852) 308 “The Works of Jeremy Bentham” (1838) 94, 118n17, 129, 291, 336 Miller, Dale E. 7n5, 14, 15, 15n2, 100n9, 117n16, 294n9, 312n20, 314n21, 317n23, 324–327, 330–332 Miller, Richard B. 325, 326 Millgram, Elijah 7n5 Mineka, Francis E. 171n3 Mitchell, Dorothy 7n5 Montesquieu (Charles-Louis de Secondat, Baron de La Brède et de Montesquieu) 70, 76 Moore, George Edward 5n3, 99n8, 123, 123n1, 125, 125n3, 126, 126n4, 127, 148, 152–156, 167, 232–252, 273, 276n7 moral obligation 96–98, 102–104, 106–108, 109–111, 113–116, 166–167, 280–291, 314–318 moral rights 5, 100, 101n10 Morellet, André 77 Morgenstern, Oskar 88n56 Moser, Shia 7n5 Murray, James Augustus Henry 44n29 Murray, Robert Malcolm 124n2 Nakhnikian, George 7n5, 232n1, 235n9 naturalistic fallacy 123, 152, 232–252
Index obligation moral obligation 6, 95, 96–98, 102–104, 106–108, 109–111, 113–116, 166–167, 263, 280–291, 302n16, 303, 306, 314–318 Olscamp, Paul J. 294n9 Orsi, Francesco 7n5 Paley, William 19, 21, 21n11, 27, 39n26, 55, 58–62 parts of happiness. See under happiness Peano, Giuseppe 7 Perronet Thompson, Thomas 15–30, 42–47, 54, 55, 55n37, 63, 63n43, 63n44, 63n45, 65, 68, 71n47, 78–79, 84–90, 102 Perry, Ralph Barton 107, 233n4 Persson, Ingmar 7n5 Phaedrus 21, 54 Plamenatz, John Petrov 14 Plato 3, 242–243, 243n15, 305n17, 334 pleasure function in happiness 276–278 quality vs. quantity 5, 99–100, 125 Popkin, Richard H. 7n5 Postow, Betsy Carol 293n7 Price, Richard 57 Priestley, Joseph 21, 22, 24, 30, 39n26, 40–48, 53, 55, 57, 62, 64, 73, 74, 80 principle of association 268–276 principle of charity 7, 169–173 principle of utility Bentham’s principle of utility 13– 91, 96–98 Mill’s principle of utility 92–121 Mill’s proof of the principle of utility 255–279, See also proof Prior, Arthur N. 232n1, 234n6 proof Mill’s proof of the principle of utility 255–279 textual map of Mill’s proof 120, 255, 282 structural outline of Mill’s proof 119, 258, 271–272 methodology behind Mill’s proof 120, 173–178, 249–251, 255, 256–257, 282 evidentiary value of desires 175–181, 256–257 implications of Mill’s proof 280–344 psychological egoism 28–29, 72, 138–139, 156
379 psychological hedonism 135, 138–139, 142, 148–151, 271 quality vs. quantity of pleasure. See under pleasure Quine, Willard Van Orman 212n9, 232n2, 275n5 Quinton, Anthony Meredith 7n5, 14, 100n9, 101n10 Rabinowicz, Wlodek 100n9 Ralkowski, Mark 304n17 Raphael, David Daiches 7n5, 197n6 Rashdall, Hastings 150n7 Rawls, John 296n11, 297 Raynor, David R. 52n35 Reddie, James 54–56 reductive individualism 185, 208, 223–228 Reid, Thomas 55 Rescher, Nicholas 124n2 rights. See moral rights Riley, Jonathan 100n9, 326 Robson, John Mercel 171n3, 172n5, 215n13 Rohatyn, Dennis A. 7n5 Rokstad, Konrad 305n17 Rosen, Frederick 14, 305n18 Rossellini, Roberto 166 Rousseau, Jean-Jacques 3, 76, 217–218, 233n5, 288n4 Rowe, William Leonard 210–213 rule utilitarianism. See under utilitarianism Russell, Bertrand Arthur William 7, 7n5 Ryan, Alan James 7n5, 101n10, 117–118, 232n1 Ryberg, Jesper 100n9 Sapontzis, Steve F. 100n9 Saunders, Ben 100n9 Sayre-McCord, Geoffrey 7n5 Scarre, Geoffrey 100n9 Schaupp, Kristin 100n9 Schmidt, Lars 166 Schmidt–Petri, Christoph 100n9 Schneewind, Jerome Borges 7n5, 14, 15, 125– 126, 157n8 Schultz, Bart 14 Scolnicov, Samuel 333–334 Searle, John R. 232n1 Seth, James 7n5, 125, 126, 127
380 Shackleton, Robert 15, 17n4, 22n16, 23, 27, 29–30, 29n20, 46n30, 75–77, 80n50, 81–83, 84n54 Shaftesbury (Anthony Ashley Cooper, the third Earl of Shaftesbury) 55, 57 Sharp, Frank Chapman 233n4 Sheng, Ching-Lai (Qinglai) 7n5 Sidgwick, Henry 4n3, 15, 17n4, 27–29, 48n32, 54–56, 72–73, 99n8, 126, 134–137, 139, 143, 144, 148, 233, 249 Singer, Marcus George 293n7, 296n11 Sinnott-Armstrong, Walter 124n2 Skinner, Quentin 333n32 Skorupski, John 7n5, 100n9 Smart, John Jamieson Carswell 294n8, 296– 297, 296n11, 298n13, 338, 339 Sobel, Jordan Howard 298n13 Soles, David 7n5 Sorley, William Ritchie 4n3, 99n8, 126, 126n4, 141–144 Spence, Gordon W. 7n5 Spencer, Herbert 233n5, 262 Spinoza, Baruch 233n5, 243n15 Stahl, Gary 7n5 Stephen, Leslie 14, 28 Stewart, Dugald 55 Stewart, Robert Scott 7n5 Stocker, Michael 7n5 Stout, Alan Ker 293n7 Strasser, Mark Philip 100n9, 101n10, 324 Stroll, Avrum 7n5 Sumner, Leonard Wayne 103n12, 324 supererogation 102–104, 283n1 Sutherland, James 7n5 Timmermann, Jens 293 total happiness. See under happiness Tufts, James Hayden 4n3, 126, 126n4, 144, 146–148 Turner, Piers Norris 312n20, 317n23, 324 Urmson, James Opie 103n12, 296n11, 297, 319–323, 325, 327 utilitarianism act vs. rule utilitarianism 104–105, 283, 292–334
Index history of utilitarianism 13–91 Benthamite sources on the history of utilitarianism 14–30 Bentham as an anchor for historical insight into utilitarianism 30–39 origins and development of utilitarian nomenclature 39–53 Bentham’s debt to predecessors and contemporaries 53–74 patterns of indirect inspiration and transmission 74–77 Bentham’s own terminological predilections 77–91 value Mill’s theory of value 95, 106–108, 109, 113–119, 167, 256–259, 263, 280–291 Venn, John 311–313 Voltaire (François-Marie d’Arouet) 46n30 von Neumann, John 88n56 Wall, George B. 298n13 Wall, Grenville 7n5 Walton, Douglas N. 213n11 Warnock, Mary 232n1, 234–239 Weinberg, Adelaide 171n3, 171n4, 172, 172n5 Weinstein, David 117n16 Wellman, Carl P. 7n5, 183n9, 197n6, 232n1, 234–239, 246n17, 279 Werkmeister, William Henry 124n2 Wertz, Spencer K. 7n5, 197n6, 222n16, 223n17 West, Henry Robison 7n5, 15n2, 100n9, 197n6, 312n20, 326 Whately, Richard 215 Whewell, William 48n32, 56–58, 61, 308 Whitehead, Alfred North 7 Wilson, Fred 7n5 Wilson, George 58–62 Wilson, T. 44n29 Wollaston, Francis John Hyde 57 Woods, John 213n11 Wright, Darryl F. 232n1 Yim, Dan 7n5 Zinkernagel, Peter 7n5, 105