204 104 2MB
English Pages 384 [392] Year 2022
R e a son in Nat u r e
REASON IN NATURE New Essays on Themes from John McDowell ✣ ✣ ✣ ✣ ✣
Edited by
Matthew Boyle and
Evgenia Mylonaki
Harvard University Press Cambridge, Massachusetts London, England 2022
Copyright © 2022 by the President and Fellows of Harvard College All rights reserved Printed in the United States of America First printing Jacket photograph: Mike Hill © Getty Images Jacket design: Annamarie McMahon Why 9780674287686 (EPUB) 9780674287679 (PDF) The Library of Congress has cataloged the printed edition as follows: Names: Boyle, Matthew, 1972– editor. | Mylonaki, Evgenia, 1978– editor. Title: Reason in nature : new essays on themes from John McDowell / edited by Matthew Boyle and Evgenia Mylonaki. Description: Cambridge, Massachusetts : Harvard University Press, 2022. | Includes bibliographical references and index. Identifiers: LCCN 2022006672 | ISBN 9780674241046 (hardcover) Subjects: LCSH: McDowell, John, 1942– | Philosophy of mind. | Naturalism. | Reason. Classification: LCC B1647.M144 R43 2022 | DDC 128—dc23/eng/20220512 LC record available at https://lccn.loc.gov/2022006672
For our families: Rachel, Sylvia, and Tobias; Kostas and Markos
Contents
Introduction
1
Matthew Boyle and Evgenia Mylonaki
I. Nature and “Second Nature” 1. Skepticism and Quietism about Meaning and Normativity
19
Hannah Ginsborg
2. Forms of Nature: “First,” “Second,” “Living,” “Rational,” and “Phronetic”
40
Michael Thompson
II. Reason in Perception and Action 3. The Rational Role of Perceptual Content
83
Matthew Boyle
4. Resolute Disjunctivism
111
James Conant
5. Control and Knowledge in Action: Developing Some Themes from McDowell Markos Valaris
153
viii C o ntents
6. Naturalism in the Philosophy of Action 171 Jennifer Hornsby
III. Consequences for Metaphysics 7. Perceiving the World
193
Sebastian Rödl
8. Seeing the World: Moral Difficulty and Drama
217
Evgenia Mylonaki
IV. Historical Precedents 9. See the Right Thing: “Paternal” Reason, Love, and Phronêsis 243 Jennifer Whiting
10. Self-Consciousness and the Idea of Bildung: Hegel’s Radicalization of Kant
285
Andrea Kern
11. The Idealism in German Idealism
309
Robert Pippin
Notes 329 Acknowledgments
375
Contributors 377 Index 379
R e a son in Nat u r e
Introduction ✣
Matthew Boyle and Evgenia Mylonaki
J
ohn McDowell is widely recognized as one of the most influential anglophone philosophers of the last half century. He has made seminal contributions to a remarkably wide range of fields: to the philosophy of mind and perception, as the most prominent advocate of a “conceptualist” view of perception and as one of the subtlest thinkers about the significance of “content externalism” for our understanding of mind; to epistemology, as one of the first and most powerf ul advocates of the “disjunctivist” conception of how perception warrants belief; to the philosophy of language, as one of the founders of the “neo-Fregean” program for reinterpreting key ideas from Frege in light of post-Tarskian semantics; and to ethics, as one of the leading figures in the contemporary revival of Aristotelian ethics and as a powerf ul critic of rule-based ethical theories, of arguments for ethical subjectivism and “anti-realism,” and of forms of motivational and justificatory “internalism” that threaten to rule out the possibility of encounters with hitherto unrecognized forms of value. He has also made groundbreaking contributions to the study of the history of philosophy, particularly to our understanding of Plato, Aristotle, Kant, Hegel, and Wittgenstein. Each of these interventions has had a deep and 1
2
REA SON IN NATURE
lasting impact, transforming the terms in which philosophical issues are discussed and giving rise to debates that continue to the present day. McDowell’s work in all t hese areas has received extensive discussion, but the present volume grows out of our conviction that this discussion has not always been as broadly or as deeply conceived as it o ught to be.1 Much of McDowell’s work has taken the form of essays intervening in ongoing debates. Even his landmark Mind and World presents itself, in significant part, as an intervention in a specialized debate in the philosophy of perception, and has been received as such.2 As a result, much of the conversation generated by McDowell’s work has focused on its significance for specific debates, and as some of these debates have died down, as academic debates are apt to do, his contributions, though still hailed as groundbreaking, have at times been treated as surpassed. The present volume seeks to make a case that there is a dimension of McDowell’s work that is detachable from the specific controversies to which it originally belonged, and whose importance has not yet been fully appreciated. Our aim is thus to urge a renewed and more open-ended reckoning with McDowell’s work, one that considers how its central themes transcend both their original context and the limitations of current discussions on some of these issues. Although the topics on which he has written are diverse, there is an evident common thread running through McDowell’s work, namely, its overarching concern with the transformative significance that reason has for human lives, and with the question of how this significance can be acknowledged without simply disregarding the fact that we are no more than natural beings, whose capacities for free thought and action must be understood as rooted in our animal nature. McDowell’s distinctive perspective on this topic is aptly expressed in a brief statement he wrote to accompany his picture in a volume of photographs of philosophers by Stephen Pyke: My main concerns in philosophy centre on the effects of a metaphysical outlook into which we easily fall, at the point in the history of thought that we occupy. This outlook might be called naturalism or scientism. I believe it tends t owards a distortion of our thinking about the place of mind in the world: the damaging effects show up not only in metaphysics itself, but also (for instance) in reflection about language, and in the philosophy
I ntr o ducti o n
3
of value and action. The task of philosophy, as I see it, is to undo such distortions.3 The themes mentioned in this remark are present throughout McDowell’s work: the idea that many of the deepest and most contentious disputes in philosophy arise out of difficulties about understanding the relationship between the first-person perspective of rational beings on their own lives and the third-person or “sideways on” perspective on human beings adopted in natural science; the thought that these disputes are not inevitable, but the product of an exaggerated, “scientistic” conception of what a scientific outlook on human beings requires; and the conviction that it is the task of philosophy to show us how to understand ourselves as rational and yet natural beings without placing these two aspects of ourselves in opposition. The aim of the present volume is to draw out this unifying perspective underlying McDowell’s diverse contributions to philosophy and to consider how ideas that grow out of his work might contribute to ongoing debates in twenty-first-century philosophy. The contributors represent a mix of longtime interlocutors, former students, and other contemporary philosophers who have taken up themes from McDowell’s work. The topics of the essays are diverse, but the volume is unified by a shared concern with the place of reason in nature and by the conviction that it is worth returning to McDowell’s writings on t hese topics, not just in the context of the discussions to which they originally contributed, but in a forward-looking spirit, with a view to identifying questions that remain open and thoughts that bear on recent and developing debates. The purpose of this introduction is briefly to introduce some main themes of McDowell’s work on this topic, as a way of framing the contributions of the eleven essays that follow.
1. Additive vs. Transformative Conceptions of Rationality A long philosophical tradition, dating back to the ancient Greeks, defines human beings as rational animals. This definition contains a problem, however, that any sympathetic account of human cognition must resolve: What is the relationship between our rationality (that is, our capacity to reflect on what to think and how to act) and our animal nature (specifically,
4
REA SON IN NATURE
the capacities to perceive and act upon our environment that we share with nonrational animals, and that provide our rational powers with material on which to reflect)? Rationality is often figured as the capacity to “step back” from the forces that dominate the lives of nonrational animals—t heir instinctive tendencies to react to what they perceive and to pursue what they desire—a nd to take a critical perspective on t hese forces and tendencies.4 But how exactly should we think of the relationship between our brute tendencies to react to what we perceive and pursue what we desire on the one hand and our rational capacity to reflect on reasons on the other hand? Does each of us really have, as it were, a “mere animal” living within them that yearns to leap ahead to belief or action, but which their rational powers must discipline? Or if this “Platonic” picture is too crude, what shape would a more satisfactory one take?5 Many contemporary accounts of h uman cognition and action, although subtler and more sophisticated than the one just sketched, are committed to some form of the “Platonic” picture. They hold that our human capacities for cognition and action are founded on capacities for perception and voluntary activity of essentially the same kind as those possessed by nonrational animals, and that our special capacity for rational reflection supervenes on these as a further, distinct power. One example of such an additive conception of h uman cognition, which figures prominently in McDowell’s own discussion, is Gareth Evans’s account of human perception. According to Evans, We arrive at conscious perceptual experience when sensory input is not only connected to behavioral dispositions . . . —perhaps in some phylogenetically more ancient part of the brain—but also serves as the input to a thinking, concept-applying, and reasoning system; so that the subject’s thoughts, plans, and deliberations are also systematically dependent on the informational properties of the input . . . Of course the thoughts are not epiphenomena; what a conscious subject does depends critically upon his thoughts, and so there must be links between the thinking and concept-applying system, on the one hand, and behavior, on the other . . . Further, the intelligibility of the system I have described depends on there being a harmony between the thoughts and the behavior to which a given sensory state gives rise.6
I ntr o ducti o n
5
What makes Evans’s account of “conscious perceptual experience” count as additive is that he conceives of such experience as the product of the linking of two distinct systems, a primordial system that connects sensory input with behavioral dispositions and a reasoning system that supervises this connection and intervenes when the primordial response is problematic. The primordial system is, in effect, the “lower animal” in us, whereas the reasoning system represents our rationality, and Evans’s way of conceiving of the relation between t hese two commits him to the idea that the addition of rationality does not alter the primordial system itself, although it alters the context in which it operates. One import ant theme in McDowell’s work is the rejection of this additive conception of the relationship between the capacities we share with other animals and our rationality, which distinguishes us from them. Commenting on Evans, he writes that if we share perception with mere animals, then of course we have something in common with them. Now there is a temptation to think it must be possible to isolate what we have in common with them by stripping off what is special about us, so as to arrive at a residue that we can recognize as what figures in the perceptual lives of mere animals . . . But it is not compulsory to attempt to accommodate the combination of something in common and a striking difference in this factorizing way: to suppose our perceptual lives include a core that we can recognize in the perceptual life of a mere animal, and an extra ingredient in addition . . . Instead we can say that we have what mere animals have, perceptual sensitivity to features of our environment, but we have it in a special form.7 On the alternative McDowell proposes here, our rationality does make an essential difference to our perceptual sensitivity to our environment: it gives us a “special form” of such sensitivity, one whose operations are themselves informed by our rationality. If this is right, then our rational capacity to learn about our environment through perception cannot be accounted for in the additive style. For if our human perceptual sensitivity to our environment cannot be explained without reference to our rationality, then our rationality must be thought of as transforming our perceptual powers themselves, not just the context in which they operate. Hence this conception of the relationship between our rationality and our
6
REA SON IN NATURE
perceptual capacities can be called transformative. On the generalized transformative view, our human cognitive capacities in general—not just our perceptual capacities, but also our capacities for motivation, bodily action, and so on—are themselves transformed by our rationality.8 Many of the essays in this volume engage with this transformative idea, w hether by exploring its consequences, criticizing McDowell’s development of it, or arguing for its extension to new domains. Two views shared by many of the authors in the volume are that (1) the interest of this aspect of McDowell’s standpoint has not yet been fully appreciated, and (2) we can question the details of McDowell’s development of the transformative idea while affirming the point of principle that underlies it.
2. Nature and “Second Nature” One central theme in McDowell’s work has been that a main obstacle to understanding how our rationality transforms our animal nature is a “constriction” that our idea of nature has undergone under the influence of modern natural science.9 While agreeing that a sane philosophy of mind must not represent the human mind as something supernatural, he has argued that we need not accept the widespread view that “naturalism” requires us to understand minds as operating according to laws of the kind studied in the natural sciences (roughly: laws that hold without exception and that recognize no distinction between cases where t hings go well and cases where they go badly). A significant part of McDowell’s writing has been devoted to criticizing constricted conceptions of what a “naturalistic” understanding of the human mind must involve, and to exploring what shape a more liberal naturalism might take. The essays in Part I of the present volume, “Nature and ‘Second Nature,’ ” take up t hese topics, approaching McDowell’s views in both a critical and a constructive spirit. One question that arises for a transformative conception of rationality concerns how to conceive of the dawn of full rationality that comes with the mastery of a first language and the initiation into a h uman culture. For while all of us are born h uman beings, it seems that none of us are born as fully rational animals: we seem rather to become fully capable of exercising our rationality in learning our first language(s) and being initiated into a
I ntr o ducti o n
7
more general understanding of the possibilities of a h uman life embodied in one or more particular h uman cultures. In the context of a transformative conception of rationality, however, this observation raises a problem: if the presence of rationality categorically transforms our capacities for perception and action, how can we understand the process by which we pass from the initial, less-than-f ully rational state in which we all presumably begin as infants to the later state in which they are fully informed by our rationality? This cannot be an instantaneous, magical transformation: it must take place gradually, through processes of a recognizably natural kind. How, then, can we understand it? Hannah Ginsborg’s essay “Skepticism and Quietism about Meaning and Normativity” takes up one aspect of this question, namely, how, while respecting the spirit of McDowell’s transformative conception of reason, we might conceive of a child’s acquisition of their first language. She approaches this question via Wittgenstein’s famous remarks on rule following in his Philosophical Investigations.10 In a series of influential essays on these remarks, McDowell has argued that, although Wittgenstein’s discussion might initially seem to pose a paradox about how rule following is possible, the moral of the discussion is ultimately that this paradox dissolves once we give up a dubious and dispensable conception of what grasping a rule must consist in.11 Against this, Ginsborg argues that there is a dimension of the rule-following problem that McDowell’s “quietist” response does not address, namely, how a child can learn to produce behavior that is in accord not with their understanding of a rule, but simply with previous rule-following behavior—their own and that of others. This is, in effect, a problem about the very possibility of learning a first language, and Ginsborg argues that it can be addressed only by recognizing a kind of “primitive normativity” more basic than the normativity of rule following. She also argues that the recognition of such normativity is consistent with a deeper appreciation of the spirit of McDowell’s transformativist standpoint. Michael Thompson’s “Forms of Nature: ‘First,’ ‘Second,’ ‘Living,’ ‘Rational,’ and ‘Phronetic’ ” takes up the question of how the presence of reason transforms h uman nature from a different perspective, via a consideration of the role that facts about our nature as h uman beings can play in our ethical reflection. McDowell’s work has played a major role in
8
REA SON IN NATURE
the revival of an Aristotelian approach to ethics, which places the idea of the possession of various virtues, rather than the grasp of a framework of general rules, at the center of ethical theory.12 However, he has criticized another aspect of (what is widely thought to be) the classical Aristotelian approach to ethics, namely, the idea that the grounds on which certain traits count as virtues for us can be explained by appeal to facts about the nature of the biological kind, Homo sapiens.13 According to McDowell, reference to merely biological facts about our nature in an account of why a virtuous agent should act in a certain way would import into our ethical reflection considerations that are of the wrong kind to engage our capacity for reasoned deliberation about how to live. Thompson replies, on behalf of the classical Aristotelian approach, that this criticism takes the sense in which our rationality transforms our biological nature insufficiently seriously. Our rationality does not merely appear in the form of a “second nature” that arrives with our acquisition of a first language and a human culture—as some of McDowell’s remarks on this topic can suggest.14 It belongs essentially to our first, biological nature to live our lives through the exercise of reason, and this implies—so Thompson argues—that the crucial aspects of our h uman nature w ill be known to us not as mere observed regularities about the biological species Homo sapiens, but as facts an understanding of which is internal to our capacity for rational deliberation itself. On the resulting view, we might say, our rationality “transforms” our animal nature, not in the sense that we begin our lives as mere animals, whose animality is then transformed by the attainment of full rationality. Rather, the animality of human beings is always a distinctively rational animality, and although it is true that we come to fully actualize our nature as rational animals in learning a first language and being introduced into ah uman culture, t here is no point in our development at which our cognitive and active capacities are “merely animal” capacities. W hether fully developed or still incipient, they are always the capacities of a rational animal, and must be understood as such. So, when we say that our rationality “transforms” our animality, this can mean only that our kind of animal capacities are essentially rational and thus differ in form from the cognitive capacities of a nonrational animal—not that there is some actual transition, in the life of a human being, between having capacities of the one kind and having capacities of the other.15
I ntr o ducti o n
9
3. Reason in Perception and Action The essays in Part II, “Reason in Perception and Action,” turn from a consideration of the general relationship between h uman reason and human nature to more specific questions about how the presence of rationality transforms the capacities for perception and action that we share with nonrational animals. McDowell famously argues that the perceptual experience of rational animals must have “conceptual content,” and thus must be structurally diff erent from the perception of nonrational animals.16 This claim has been widely discussed and frequently criticized, and McDowell himself has made important revisions to his original formulation of it.17 The essays in the second part of the volume explore the motivations for this idea, what is crucial and what is dispensable in it, and how analogous points might be formulated not just for the case of perception, but also for human action. In “The Rational Role of Perceptual Content,” Matthew Boyle considers one line of criticism of McDowell’s conceptualist view of perception, namely, the objection, influentially pressed by Charles Travis and Bill Brewer, that perception should not be conceived as having “content” at all, and so a fortiori should not be conceived as having specifically conceptual content. McDowell has responded to this concern by modifying certain specific commitments of his original position (that the content of perception must be expressible in a proposition, that it must include no element of which the perceiving subject does not yet possess an articulate concept), while reaffirming his basic thesis that we should conceive of perception as presenting us with particu lar contents, whose availability depends on our possession of the kinds of fundamental conceptual capacities characteristic of a rational animal. Boyle argues that this is insufficient: the critics of perceptual content are right to think that the role of perception in supplying us with knowledge is simply to present us with objects about which we can make judgments, not, so to speak, to recommend particu lar judgments to us. Nevertheless, Boyle argues, this point need not threaten what he suggests is McDowell’s fundamental idea: that h uman perceptual capacities make possible a distinctive form of openness to the world in virtue of the fact that we are rational. This idea, which McDowell takes from Kant,
10
REA SON IN NATURE
admits of a development that makes no appeal to the notion of perceptual content. In “Resolute Disjunctivism,” James Conant considers a distinct but related McDowellian claim about perception: that we should conceive of the operations of our perceptual capacities “disjunctively,” as issuing e ither in veridical, knowledge-conducive perceptions of objects in our environment or else in mere semblances of perception, which cannot supply us with knowledge. This “disjunctivist” conception of perception stands opposed to a view on which what perception really presents to us is just a perceptual appearance of things being thus-a nd-so, whose veridicality is always a further question. In a series of influential papers, McDowell has argued that this kind of “highest common factor” conception of perception, although seemingly motivated by the observation that our senses can deceive us, is in fact unwarranted and incapable of explaining the possibility of perceptual knowledge.18 Conant argues, however, that McDowell’s defense of disjunctivism is flawed by an insufficiently consistent application of the idea that perception is fundamentally a capacity for knowledge. If this is the fundamental way to characterize our capacity to perceive, then McDowell is right to insist that veridical perception and perceptual illusion are not simply two disjuncts on a par, but rather the proper and defining kind of operation of our perceptual capacities, on the one hand, and a defective and inessential kind of operation, on the other. But, Conant argues contra McDowell, if we are to resolutely characterize our power to perceive as a capacity for knowledge, then we should be disjunctivists not only about the case of veridical perception and perceptual illusion, but also about the case of veridical perception which is reflectively recognizable and veridical perception which is not—such as are cases in which our perceptual capacities operate normally but we are not in a position to know their deliverances to be veridical. McDowell has argued that this can happen in what he calls “reflectively unfavorable circumstances” (that is, circumstances in which our perceptual capacities are in fact functioning normally, but we have some extraneous reason to doubt that this is so). Conant argues, however, that to the extent that such cases are possible, they must be understood as cases of the defective operation of our perceptual capacities themselves. Thus, from a diff erent angle, Conant’s paper approaches the thought that our rational capacity for knowledge inextricably informs our perceptual capacities themselves.
I ntr o ducti o n
11
The remaining two papers in this part consider the application of McDowell’s transformative conception of rationality to the case of physical action. In Mind and World, McDowell suggested that the idea he had developed in the case of perception ought to extend to the case of action, and in a series of recent papers he has sought to elaborate on this suggestion, arguing that not just m ental operations like deliberation and the framing of intention, but tangible h uman actions themselves must be understood as exercises of our rationality.19 In “Control and Knowledge in Action: Developing Some Themes from McDowell,” Markos Valaris seeks to develop and defend this McDowellian idea by focusing on the phenomenon of agential “control” over bodily actions. On a widespread view, although our rational capacities have a crucial role in framing the intentions on which we act, the actual execution of our bodily actions is ensured by “subpersonal” mechanisms whose operation takes place without rational supervision. Hence, according to this view, our agential control of our own unfolding actions depends on how reliable t hese unconscious mechanisms are in implementing the specifications provided by our minds. Valaris argues, however, that physical action can itself be an instance of problem solving, and so an expression of intellectual activity in its own right. Accordingly, we can conceive of agential control in a way that relates it to knowledge: you exercise agential control over your actions to the extent that your actions express your own knowledge of how to achieve your ends. In this way, Valaris suggests, we can secure, in the case of action, a version of the McDowellian thought that our rationality informs the operations of our animal powers for engaging with our environment, rather than merely exercising a kind of supervisory control over these powers. In “Naturalism in the Philosophy of Action,” Jennifer Hornsby takes up the same theme while giving closer consideration to McDowell’s own development of it in “Some Remarks on Intention in Action.”20 Hornsby endorses McDowell’s fundamental idea that a liberal naturalism should recognize human actions as themselves operations of our capacity for rational thought, rather than as mere physical effects that mental operations produce “in the world.” But she also argues that McDowell has in fact been too concessive to a restrictive conception of what h uman action must consist in. In particu lar, she criticizes (1) his idea that an intention can itself be conceived of as a kind of “continuant” that appears first as a mere plan to act in a certain way and l ater in the form of concrete
12
REA SON IN NATURE
bodily action of the relevant sort, and (2) his related suggestion that when such an intention does assume the form of a bodily action, it does so by governing a concrete “activity” that can be regarded as a kind of present existence. These ideas, Hornsby argues, are vestiges of a conception of real existence in nature to which McDowell should not subscribe, since they are not required for a sound understanding of h uman action, and in fact prevent us from appreciating the full scope of rational practical activity. Dropping these commitments, she suggests, would make possible a deeper appreciation of the idea that drives McDowell’s own work in action theory: that human agency is inextricably both a rational and a natur al phenomenon.
4. Consequences for Metaphysics The essays in Part III, “Consequences for Metaphysics,” consider the implications of McDowell’s transformativism for our conception of the objects of human knowledge and human practical activity. In Mind and World, McDowell strikingly suggests that a recognition of how rationality informs human perception and action has consequences, not just for our conception of these activities themselves, but also for our conception of the world with which they are engaged.21 If this is right, then reflection on the theme of reason in nature should have consequences, not only for our conception of human cognition and action, but also for topics in metaphysics. The essays in this part explore this idea, again in both a critical and a constructive spirit. In “Perceiving the World,” Sebastian Rödl argues that, if we take seriously McDowell’s fundamental idea that a philosophical account of perception must be given not “from sideways on,” but rather “from within” (that is, through an articulation of the understanding of perception implicitly possessed by any self-conscious subject who perceives), then we must query his suggestion that we can find a place for the human mind in the natural world simply by articulating a more liberal naturalism that does not conceive of nature as exclusively a realm of exceptionless, nonnormative laws. The basic obstacle to locating the human mind in the world, Rödl argues, is not an overly restrictive conception of nature, but
I ntr o ducti o n
13
the fact that our minds are understood, in our own self-consciousness, to know the world as a w hole, and thus not merely to stand in some describable relation to objects in the world. It follows that h uman perception, inasmuch as it is a mode of knowing, cannot be a mere relation to things in the world; it must at the same time involve a kind of openness to the world as a whole, and Rödl argues that a recognition of this point problematizes the very idea of locating our subjectivity in the world. It is this world- oriented aspect of h uman understanding, Rödl suggests, that is the real source of difficulties about the “relation” of mind and world, so these difficulties arise from McDowell’s own principle—which Rödl endorses— that a philosophical account of human cognition must proceed “from within.” The idea that human knowing involves an orientation toward the world as a whole is also a theme in Evgenia Mylonaki’s “Seeing the World: Moral Difficulty and Drama.” Mylonaki’s paper takes its departure from McDowell’s rejection of what he calls the “blueprint model” of virtuous practical deliberation, according to which a virtuous agent must possess a general, nontrivial conception of what it is to act well that is describable without reference to the choices they make in particular circumstances—a kind of general “blueprint” for virtuous action that they can apply to par ticu lar circumstances simply by bringing it to bear on a nonnormative description of the situation they confront.22 An import ant strand in McDowell’s work on ethics has centered on the criticism of this idea and the defense of the opposing thought, which he finds in Aristotle, that virtue might consist not fundamentally in the ability to apply a general rule to particular cases, but rather in a kind of quasiperceptual ability to see what action is called for in particu lar circumstances. Mylonaki endorses this idea, but argues that its full appreciation requires a recognition of how one’s openness to changing perceptions of particu lar people and circumstances essentially involves an openness to changing perception of our world as a whole. Using Henry James’s The Portrait of a Lady as a case study, Mylonaki brings out how, on a rich description of moral experience, having a clear view of the particulars in sound moral judgment may on occasion come to no less than the total collapse of one’s view of the world as a w hole. She uses this case to argue that an enriched McDowellian moral philosophy must be sensitive to this world-involving dimension of ethical experience.
14
REA SON IN NATURE
5. Historical Precedents The fourth and final part of the volume, “Historical Precedents,” examines some of McDowell’s most import ant engagements with the history of philosophy. It is characteristic of McDowell’s work in the history of philosophy that it treats historical figures neither as topics of merely scholarly interest nor as resources to be mined for ideas that bear on con temporary debates. While holding to a high standard of historical scholarship, he manages to treat t hese figures as vital interlocutors engaged with perennial questions—interlocutors whose perspectives on these questions are relevant to us precisely b ecause their views are not shaped by the same intellectual milieu and sense of the obvious that form our own thinking. The essays in Part IV consider McDowell’s readings of historical figures in this spirit, focusing in particular on the basis he finds in Aristotle, Kant, and Hegel for crucial aspects of his conception of how our reason informs our nature. In this way, t hese essays serve to deepen the consideration of the themes of the other three parts of the volume, while also demonstrating the ongoing relevance of McDowell’s work to debates in the history of philosophy. In “See the Right T hing: ‘Paternal’ Reason, Love, and Phronêsis,” Jennifer Whiting offers an account of the textual and philosophical basis in Aristotle for four key features of McDowell’s account of ethical virtue: (1) his identification of virtue with the kind of good condition of the practical intellect that Aristotle calls phronêsis (commonly translated as “practical wisdom”), (2) his conception of phronêsis as a capacity that is si multaneously cognitive and desiderative, (3) his association of phronêsis with a kind of perceptual capacity distinctive of a virtuous agent, and (4) his suggestion that what distinguishes a genuinely virtuous agent from a merely “continent” one is that, in the former but not the latter, sources of motivation that militate against virtuous action are “silenced.” Each of these features of McDowell’s view has been a source of controversy, and although McDowell claims an Aristotelian provenance for them, he has not always offered an extensive textual justification for his reading. Whiting argues that t here is in fact a powerf ul textual case to be made for each of these four ideas, and that the ideas fit together to form a picture on which the desiderative life of a h uman being constitutes a kind of matter
I ntr o ducti o n
15
that is s haped, in the course of our moral development, by the form of our practical intellect. In this way, McDowell’s account of moral development exhibits a hylomorphic structure characteristic of Aristotle’s approach to many topics; and by the same token, this structure provides a theoretical framework through which to understand his idea that reason “transforms” our animal nature. Andrea Kern’s “Self-Consciousness and the Idea of Bildung: Hegel’s Radicalization of Kant” turns to McDowell’s reading of Hegel’s critique of Kant, but the question of the role of upbringing and education in giving rational form to our animal nature, which was central to Whiting’s essay, also turns out to figure importantly in this diff erent context.23 In Mind and World, McDowell suggested that Kant was prevented from giving a fully satisfying account of the way in which our rational nature informs our perceptual capacities by his lack of a robust notion of second nature, which would have allowed him to explain how our animal capacities for perception come, in the course of our development, to be informed by what the German philosophical tradition calls Bildung (that is, by upbringing and education, conceived broadly as a kind of introduction into a specific human culture).24 Moreover, McDowell goes on to suggest, in a brief remark, that if Kant’s insights into the relation between our rational and our sensory capacities were paired with a robust account of the role of second nature in shaping our cognition, this would amount to a Hegelian conception of human mindedness.25 Kern disputes the claim that McDowell’s “naturalism of second nature” is a truly Hegelian position, and she argues that Hegel would in fact reject the idea that our animal nature must become informed by reason through Bildung. He would reject this, not because it is false to suggest that Bildung plays a crucial role in the full actualization of our rational capacities, but because, seen from a wider perspective, the very existence of Bildung is an expression of the way in which our (specifically h uman) animal nature is informed from the outset by reason. Hence Bildung cannot explain, but must rather itself be explained by, our distinctively rational animality. Finally, in “The Idealism in German Idealism,” Robert Pippin reflects on his decades-long exchange with McDowell over the interpretation of Hegel. Pippin begins from a fundamental point of agreement: his deep sympathy with McDowell’s idea that the crux of Hegel’s idealism can be captured in the thought that “there is no ontological gap between . . . the
16
REA SON IN NATURE
sort of thing one can think, and the sort of thing that can be the case.”26 McDowell has focused on the application of this idea to the realm of empirical cognition (that is, to our knowledge of the world we confront through our senses). Pippin argues, however, that an even more fundamental application of the Hegelian doctrine of the identity of the rational and the actual comes in the realm of “pure thought” that Hegel calls “logic.” Drawing on his recent discussion in Hegel’s Realm of Shadows, Pippin argues that, in Hegel’s Science of Logic, we encounter cases of concepts supplying themselves with objects that are adequate to them, and that we can only understand Hegel’s broader idealism against the background of this fundamental case of the necessary harmony between concepts and their objects.27 Thus, if McDowell’s reflections on the theme of reason in nature are to constitute—as he himself suggests—a kind of prolegomenon to the reading of Hegel, they must be enriched by a consideration of Hegel’s treatment of these logical topics.28 ✣ ✣ ✣
This introduction marks a few threads r unning through the essays that follow, but readers must find their own paths through this material, and make their own judgments about its unity. We hope, at any rate, that these essays, both through their variety and through their interrelations, serve to bring out the richness and enduring vitality of McDowell’s work.
✣
Pa rt I
✣
NAT U R E A N D “SECON D NAT U R E”
Chapter One ✣
Skepticism and Quietism about Meaning and Normativity ✣
Hannah Ginsborg
1 In a body of work going back to the 1970s, John McDowell has developed a powerf ul, persuasive and historically informed conception of human language and thought and of their relation to the world. He has developed this conception in part through what he has called a “quietist” reading of the later Wittgenstein, and in particular of Wittgenstein’s rule-following considerations.1 This reading runs counter to commonly accepted interpretations of Wittgenstein as raising and addressing philosophical problems about the possibility of meaning, rule-following, and intentionality. As McDowell sees it, the point of the rule-following considerations—a point which he himself endorses—is not to question the possibility of meaning or rule-following, but rather to undermine the defective ways of thinking which lead us to find them problematic. McDowell criticizes in particular the reading of Wittgenstein offered by Saul Kripke, who interprets the rule-following considerations as incorporating 19
20
N ature and “ S ec o nd N ature ”
a skeptical argument to the effect that t here is no such t hing as meaning.2 One of the morals which McDowell draws from Wittgenstein is that the skeptical problem described by Kripke is illusory, arising from a mistaken assumption about meaning and understanding. My aim in this paper is to challenge McDowell’s quietism about meaning, at least in the context of his disagreement with Kripke, by arguing that he fails to appreciate the full force of the difficulty which Kripke reads Wittgenstein as raising. 3 B ecause of this, I argue, the considerations he raises to support quietism about meaning are inadequate to defuse Kripke’s skeptical problem. I go on to present an alternative approach to the problem, one which draws on McDowell’s approach in invoking a kind of quietism, but on which the quietism applies not at the level of linguistic meaning, but rather at the level of the more primitive normative attitudes on which the possibility of linguistic meaning depends. I conclude that, while quietism is at some point inevitable in our attempts to make sense of rule-following and meaning, McDowell invokes it too soon. Kripke’s skeptic raises a genuine problem about how linguistic meaning is possible, and that problem can be solved, even though the solution depends on appeal to normative attitudes which cannot in turn be called into question.4
2 I begin by reviewing the skeptical problem presented by Kripke, since, as we shall see, much depends on how it is understood. Kripke introduces it by imagining a scenario in which, having never before added numbers greater than 57, I respond to the question “What is 68 + 57?” by saying “125,” only to be confronted by a skeptic who challenges my confidence in what he calls the “metalinguistic” correctness of my answer. He proposes that, if I am to accord with my previous usage of “+,” I should instead say “5.” The skeptic’s ground for this challenge is the hypothesis that, in my past uses of “+,” or the “plus” sign, I meant not addition but quaddition, where x quus y is x plus y for values of x and y less than 57, and otherw ise 5. In order to justify the correctness of “125” rather than “5” in the light of my previous uses, I must rule out the skeptical hypothesis that I previously meant quaddition rather than addition. As Kripke sees it, this requires my
S kepticism and Q uietism
21
identifying “some fact about my past usage that can be cited to refute [the hypothesis].”5 If I am unable to cite a fact which, as he puts it, “establishes that I meant plus rather than quus,” then I cannot rule out the skeptical hypothesis and my saying “125” is revealed to be an “unjustified leap in the dark.”6 Kripke emphasizes that although the skeptical argument may seem epistemological, challenging my knowledge as to w hether I meant plus or quus, it is in fact metaphysical, leading to the conclusion not just that I do not know what I meant by “plus” but that “there was no fact about me that constituted my having meant plus rather than quus.” 7 This conclusion generalizes from past to present and future meaning, and from the meaning of “plus” or “+” to the meaning of linguistic expressions generally, leading to the conclusion that there is no such t hing as meaning.8 Having laid out the general framework of the skeptical argument, Kripke fills it out by considering and rejecting various responses I might give to the skeptic, most of which take the form of proposing candidate facts in which my meaning addition might consist, and which I might accordingly cite in order to “establish” that I meant addition rather than quaddition. T hese putative facts include my having in mind a set of general instructions for the use of “plus,” or a definition of “plus”; my being disposed to give the sum in answer to “plus questions”; and my having in mind an introspectable quale or mental image associated with the addition function.9 Kripke also considers a response to the skeptic, which he characterizes as “desperate,” on which my meaning addition by “plus” is simply a primitive, sui generis, fact about me.10 Kripke provides multiple lines of argument to undermine these various responses, of which I w ill highlight two. One, aimed against the dispositional proposal, is that dispositional accounts of meaning fail to do justice to the idea that meaning is normative: my meaning addition by “plus” is supposed to justify my responding to “68 + 57” with “125,” but the claim that I was disposed to say “125” cannot play that justificatory role. The other, aimed against both the “instructions” or “definition” proposal and the “introspectable quale” proposal, is that anything I have in my mind associated with the expression “+” can be interpreted in such a way as to require me to respond “5” rather than “125” in order to be faithful to it. I might claim to have had in mind a set of directions for answering “x + y” questions which call for me to assemble a heap of x objects and another heap of y objects, and then to “count” the number of objects in both heaps taken
22
N ature and “ S ec o nd N ature ”
together; but the skeptic can maintain that I meant quus rather than plus by proposing that the word “count” is to be interpreted in a quus- like way. And similar considerations rule out qualia or mental images, which are no less open to multiple interpretations than internalized linguistic expressions such as “count the items in the heap.” Here is a passage quoted by McDowell as summarizing this part of the argument: “No matter what is in my mind at a given time, I am f ree in the f uture to interpret it in diff erent ways.”11 McDowell endorses the idea that meaning is normative,12 and so agrees with Kripke on the unacceptability of the dispositional proposal. His view that the skeptical problem is illusory is based on a challenge to the assumption he sees as underlying the second line of argument, in partic ular the claim just quoted, that whatever is in my mind is open to multiple interpretations. The heart of the skeptical problem, as he sees it, is puzzlement about the notion of normative accord which both he and Kripke see as built into the ideas of meaning and understanding: the notion that our understanding of an expression can be such that a piece of behavior can either fit or fail to fit it. The puzzlement is about how there can be something in my mind corresponding to the expression “+” which makes it the case that “125” rather than “5” fits my understanding of the query “68 + 57.” We tend to assume that any such thing must be “normatively inert,”13 something which cannot itself determine “125” as according with it, but must instead be supplemented by an interpretation. Given that assumption, the attempt to find something which constitutes our understanding of “+” leads into regress: whatever we propose as an interpretation in turn comes to seem normatively inert, and so to call for further interpretation.14 McDowell’s response to the puzzlement is to reject the assumption that items in the mind cannot in themselves determine behavior as correct or incorrect. He does so in terms of Wittgenstein’s response, at Philosophical Investigations §201, to the threat of a regress of interpretations. What the regress shows is not the skeptical conclusion that t here is no such t hing as accordance with a rule (or with one’s understanding of an expression), but rather that, in Wittgenstein’s words, “there is a way of grasping a rule which is not an interpretation.”15 The idea that t here can be understanding without interpretation allows me to meet the skeptic’s demand for a fact in which my having meant addition consists by citing, simply, the fact that I meant addition. This
S kepticism and Q uietism
23
amounts in effect to my insisting on the response which Kripke rejects as “desperate” and “mysterious,” namely that my meaning addition is a primitive, sui generis state.16 This response might seem easy, but according to McDowell the task of dislodging the assumption leading to the problem, and thereby showing that there can be understanding without interpretation, takes considerable philosophical work.17 Some of that work, he thinks, can be performed by reflecting on Wittgenstein’s example of a signpost.18 A signpost might at first appear to just “stand t here,” a tapered board on a post without normative significance.19 So we might be tempted to think that, in order to understand it as pointing in some one direction rather than another, we need to put an interpretation on it, for example, to say to ourselves “it says to go right.” But by “insisting on a bit of common sense” about signposts, we come to see that, for someone initiated into a practice of following signposts, understanding the signpost does not require interpreting it.20 In following an ordinary signpost, with its tapered end, say, at the right, we unreflectively recognize it as pointing to the right, that is, as something with which only g oing right can accord. No interpretation is needed to rule out, for example, that it is telling us to go left.21 The moral, while not spelled out explicitly by McDowell, is that something analogous holds of linguistic signs and of their mental correlates, such as the words “plus” or “sum,” or the thought that I meant plus or the sum. We might suppose that an answer to Kripke’s skeptic along the lines of “I had in mind that I should give the sum” is vulnerable to the regress b ecause all I could have had in mind to determine the correctness of “125” rather than, say, “5,” would have been the inert sign “sum” conceived of as just standing there like a board on a post. But the example of the signpost undermines the temptation to embark on the regress by helping us to see that, for someone initiated into a practice of using signs for arithmetical functions, the expressions “+” and “sum”—whether spoken or in the mind—do not stand in need of interpretation in order for us to understand them as calling for one specific response rather than another. It is by reminding ourselves of this kind of common sense about things like signposts, according to McDowell, that we can rid ourselves of the assumption which makes the “sui generis” response to the skeptic seem desperate and mysterious, and so leads us to think that Kripke raises a genuine problem about the possibility of rule-following and meaning.
24
N ature and “ S ec o nd N ature ”
3 I will argue in this section and the next that McDowell underestimates the force of Kripke’s skeptical problem, with the consequence that his response fails to address it adequately. As noted in the previous section, his approach assumes that the heart of the problem is puzzlement about the idea that our understanding of an expression can be such that a piece of behavior, say, uttering “125” in response to “68 + 57,” can fit or fail to fit it. He goes on to argue, plausibly enough, that this kind of puzzlement disappears if our conception of the facts in which meaning and understanding consist is broad enough to accommodate items which require no interpretation in order for pieces of behavior to accord or fail to accord with them. But I believe that Kripke’s problem arises from a different and more fundamental kind of puzzlement. To put the difference crudely: what is puzzling is the possibility of accord not between a piece of behavior and the understanding of an expression, but between a piece of behavior and previous pieces of behavior involving the expression. In other words, the puzzlement is about how my use of a sign at a given time can accord not with what I mean or meant by the sign, or how I understand or understood it, but with previous uses of the sign, which could be my own, as in Kripke’s way of generating the problem, but equally well those of others, such as my parents or teachers. As we will see in the next section, it concerns how I can recognize my use of a sign as correct in the light of previous uses—and so how I can, in Wittgenstein’s terms, “know how to go on” in my use of a sign—if, as is the case for a language-learner, I am not yet in a position to recognize those previous uses as bearing a specific meaning.22 Kripke’s way of developing the problem does indeed involve the challenge to cite a fact in which my meaning addition consists, but once we recognize that the challenge is motivated not by self-standing puzzlement about the possibility of meaning and understanding, but by a more basic puzzlement about how one piece of behavior can be correct in the light of previous examples of behavior, we see that it cannot be addressed simply by letting go of the assumption that all understanding involves interpretation. To begin to see the contrast between these two kinds of puzzlement, let us recall that what the skeptic initially challenges is my confidence,
S kepticism and Q uietism
25
not that I meant addition by “+,” but rather that (if I am to accord with my previous usage of “+”) I should now respond to “68 + 57” with “125.”23 The hypothesis that I previously meant quaddition is introduced as a ground for that challenge. If I cannot rule it out by citing a fact which “establishes” that I meant addition, or in which my having meant addition consists, then my confidence in the correctness of “125” is shown to be unwarranted. At least at the outset, then, the skeptic appears to be offering an epistemological challenge, and the challenge is aimed at my knowledge that I ought to say “125.”24 But, as we saw, Kripke goes on to draw a metaphysical conclusion: namely, that t here is no fact of my having meant addition (and, generalizing, no fact of anyone’s ever having meant or meaning anything). Now McDowell, like most readers of Kripke, sees Kripke’s route to that metaphysical conclusion as leading directly from my supposed failure to cite a fact in which my having meant addition consists, to the nonexistence of any such fact. Viewed in this way, the argument in its essentials is metaphysical rather than epistemological. It boils down to the thought that, since it is not possible to reduce facts about meaning to more basic facts (for example about dispositions, qualia, or mental images), we have to conclude that there is no such thing as meaning. In the terms I used earlier to characterize McDowell’s view of it, the argument is an expression of puzzlement about how there can be such a thing as meaning, given that meaning has to be the kind of thing with which behavior can accord, and that such t hings as dispositions, qualia, and m ental images do not appear to fit that bill. But this understanding of the argument is unsatisfying, both because it makes Kripke vulnerable to a charge of unargued reductionism, and because it does not do justice to the epistemological elements of the argument which I just highlighted.25 The challenge to my confidence in the correctness of “125,” on this reading, plays no essential role in reaching the skeptical conclusion. It illustrates a constraint on possible candidates for the fact of meaning addition (in that any such fact must determine that I should respond to “68 + 57” with “125”), but it does not in itself constitute a step in the skeptical argument. The alternative reading which I propose offers a more complex, but I believe more convincing, route from my failure to cite a fact in which my having meant addition consists to the conclusion that t here is no such fact. As suggested in the previous paragraph, that failure amounts to a failure to “establish” that I meant addition, and is thus an epistemological failure;
26
N ature and “ S ec o nd N ature ”
the upshot is the epistemological conclusion that I cannot know that I should say “125,” or, more picturesquely put, that “125” is a leap in the dark. It is this conclusion which in turn leads to the metaphysical claim that there is no such thing as meaning. If all my uses of language are leaps in the dark, then, as Kripke puts it, meaning “vanishes into thin air.”26 While Kripke is not explicit on this point, I take the thought underlying this last move to be that the meaningful use of language depends on our being in a position to know, in each use of an expression, that we are g oing on in a way which accords with previous uses of it. If, for all I know, each use of a sign is completely arbitrary in the light of previous uses—if I have no reason to suppose that saying “125” in response to “68 + 57?” fits the sequence of previous responses to “+” questions any better than “5”—then I do not count as understanding that expression or meaning anything by it.27 The skeptical hypothesis that “+” in those previous uses meant quaddition is intended to undermine the possibility of facts of meaning by undermining my confidence that “125” fits previous uses of “+.” It is in order to restore that confidence, and hence to avert the threat of meaning’s “vanishing into thin air,” that I have to rule out the skeptical hypothesis by citing a fact which “establishes” that I meant addition rather than quaddition. On this reading of the argument, the skeptical challenge cannot be rebutted simply by claiming that I meant addition rather than quaddition. For the skeptic calls on me not merely to cite a fact in which my having meant addition consists, but to cite a fact which allows me to justify my claim that I meant addition, and so prevent the skeptical hypothesis from undermining my knowledge that I ought to say “125.” In this dialectical context it is not enough just to insist, without grounds, that the skeptical hypothesis is false. The dialectical situation is somewhat analogous to that of the First Meditation, where Descartes’s knowledge that he is sitting by the fire with a piece of paper in his hand is challenged by the skeptical hypothesis that he is dreaming. To defend the particu lar instance of knowledge targeted by the skeptic—and, by extension, since this is a “best case,” all his knowledge of outer t hings—he has to find grounds to reject the hypothesis. The simple insistence that he is awake rather than dreaming does not count as an answer to the skeptic. The same is true in this case, where the hypothesis that I meant quaddition is introduced as a way of motivating doubt about the particular instance of knowledge which the skeptic is actually targeting, namely that I o ught to say “125.” Of course,
S kepticism and Q uietism
27
I can reject the hypothesis outright, as Descartes might reject outright the hypothesis that he is dreaming, but that is to refuse to take the skeptical challenge seriously, not to answer it. Now the situation would be diff erent if, as McDowell supposes, the skeptic’s challenge to cite a fact in which my having meant addition consists w ere motivated simply by puzzlement about how meaning states are possible, given that they have to be the kind of thing with which pieces of behavior can accord or fail to accord. It would then be reasonable to suspect, with McDowell, that the ground of this puzzlement was a misconception about what is required for this accordance to be possible, a misconception which prevented the skeptic from seeing that the state of my meaning addition can, unproblematically, meet the requirement. In that case, it would be perfectly in order for me to respond to the skeptic’s challenge by citing the fact that I meant addition. But I have suggested that the motivation for the skeptic’s challenge is deeper and more complex. The skeptic’s denial that there is such a thing as meaning stems not from puzzlement about meaning as such, but from a more specific puzzlement about how I can know that any one of my uses of a sign is correct in the light of previous uses, given that the previous uses are apparently compatible with any number of hypotheses about what was meant in those uses. Even once the misconception has been removed, that puzzlement remains. I can accept, at least provisionally, that a state like meaning addition can be the kind of t hing which determines uses as correct or incorrect, and still doubt, in light of the consistency of my previous uses with the hypothesis that what I meant was quaddition, whether saying “125” in response to “68 + 57?” accords with my previous uses of the plus sign. And that doubt threatens, a fter all, to undermine the possibility of meaning.
4 To see more clearly how my view of the skeptical problem diverges from McDowell’s, it w ill help to look more closely at the move which is made, on my reading of the skeptical argument, from the subject’s not knowing that “125” accords with her previous uses of “+” to the disappearance of meaning. I said that the move, while not spelled out by Kripke, is motivated by the thought that, to use an expression meaningfully, I must take my
28
N ature and “ S ec o nd N ature ”
uses of it to accord with previous uses. That thought, I want to suggest, can in turn be motivated by the further thought that, in order to make sense of the possibility of linguistic meaning, we need to accommodate the possibility of language-learning, where learning a language is understood as including learning what the expressions of the language mean. If a linguistic expression like “+” or “table” is to have a meaning, it must be pos sible for potential users of the expression to come to know what it means, and this is something which can happen only if they can learn how to go on appropriately from previous uses of the expression which they have observed or been shown. In many cases this learning can take place through verbal explanation, but some expressions at least must be learned simply through example: the child hears uses of “table” in contexts where tables are salient, or is shown uses of “+” in very s imple arithmetic equations, and at some point comes to know how to use the expression herself in ways which accord with or fit t hose previous uses. Now we might put this by saying that she comes to grasp the meaning that was expressed in t hose e arlier uses so that she is in a position to use the expression herself in a way which fits that previously grasped meaning. But that way of describing what happens in learning such an expression is not obligatory. We can describe the situation more simply (and, as I s hall argue in the next section, more accurately) by saying that she comes to know how to go on appropriately from previous uses, which is to say, to use the expression in a way which she can recognize as according with previous uses. On this conception of language-learning, it is essential that the learner come to be able to recognize, for each of her (correct) novel uses of an expression, that she is going on appropriately from the uses she has previously been shown. The kind of learning which issues in knowledge of meaning has to involve the recognition of correct or incorrect uses, where “correct” and “incorrect” uses are understood as t hose which do or do not fit the examples from which one is supposedly learning. If the “learning” consisted in nothing but the formation in the learner of a disposition to use the expressions in a certain way—say, exposure to standard uses of “+” and “table” engendering in the learner a disposition to give the sum in response to s imple “+” questions or to say “table” when suitably prompted in the presence of a t able—then the process would not count as learning the meaning of the expression. We could say that the “learner” was acquiring a capacity to go on from the examples in a way which we find appropriate,
S kepticism and Q uietism
29
but we could not say that she was coming to understand the expression or learning what it meant. For the process to count as learning the meaning of an expression, the learner must be capable of seeing a normative connection between the new uses she makes of the expression, and the uses she has previously been shown. In other words, if each new use of “table” for her is a “leap in the dark” from the uses she has been shown—if she is incapable of seeing, for example, that her present use of “table” for a novel table under the Eiffel Tower accords with earlier uses of “table” for tables in the Luxembourg gardens or in the kitchen at home—she does not count as learning the meaning of “t able” even if she is acquiring a disposition to use “table” in ways which her teachers regard as appropriate. As I see the argument, then, meaning “vanishes into thin air” for Kripke because the skeptic challenges my entitlement to regard “125” as according with my previous uses of “+” (broadly construed as including those I observed as well as those I made myself) and hence to think of myself as having learned the meaning of “plus.” In other words, the skeptical argument undermines the possibility of linguistic meaning by undermining the idea, essential to the idea of linguistic meaning, that there can be such a thing as language mastery. In raising the question whether I should say “125” given my past uses of “+,” the skeptic puts me in a position of a language-learner who, like the pupil of Philosophical Investigations §185, has to demonstrate his developing mastery of a linguistic expression (for Wittgenstein “add 2”) by going on correctly from previous examples of its use.28 I may now be confident that I know what “+” means, but that confidence depends on my confidence that I was successful in learning how to go on correctly from the examples I was given. The quaddition hypothesis is supposed to reveal that latter confidence as unwarranted: since I have no reason to believe that I and my teachers meant quaddition rather than addition when we used “+” previously, I have no reason to believe that I am g oing on correctly from those previous uses. The doubt generalizes to call into question that anyone can ever learn the meaning of a linguistic expression from a finite series of examples, and in this way undermines the possibility of language- learning, and hence linguistic meaning, überhaupt.29 This understanding of the argument in terms of learning helps us see more clearly how McDowell’s approach misses the force of Kripke’s problem. McDowell thinks that the problem arises only if we are u nder the
30
N ature and “ S ec o nd N ature ”
misconception that the signs of language or their m ental correlates are “normatively inert”: not the kind of thing which can itself, without interpretation, determine a given use as correct or incorrect. Once that misconception is removed, I can reply to the skeptic’s challenge to my knowledge about what I previously meant by “+” as I used it in the past by pointing out that I know I meant addition. I can know this because, among other things, I can remember thinking such t hings as now I should add when given a “+” problem. The skeptic’s regress-inducing counter of “How do you know you weren’t thinking now I should quadd?” is invalidated once we see that what I had in mind, perhaps the words “I should add,” need no interpretation to be understood as meaning that I should add and not quadd. Even more straightforwardly, I know that I meant addition because that is what “+” manifestly means, and the skeptic has given me no reason to suppose that I was, in the past, somehow imposing a meaning on “+” which it manifestly does not bear. But, as McDowell makes clear, the knowledge to which I appeal in this response to the skeptic is available to me only b ecause I have already been initiated into the language to which “+” belongs, and in particular into the correct use of “+.” It is “shared command of a language” which “equips us to know one another’s meaning without needing to arrive at that knowledge by interpretation”: if I do not share a language with my past self then I am not, in McDowell’s terms, “equip[ped] . . . to hear [her] meaning in [her] words” and so not in a position, to know, without interpretation, that I previously meant addition.30 So the proposed reply is not available if we understand the skeptical problem as turning on my capacity to become initiated into the use of “+” by learning how to go on from previous examples. The skeptical challenge, on this understanding, puts me in the position of a language-learner, someone for whom “+” does not yet bear any intrinsic meaning. Asked, in the light of the quaddition hypothesis, how I know I o ught to go on from previous uses of “+” by responding to “68 + 57” with “125,” I cannot simply appeal to my supposed knowledge of what “+” means, b ecause that is just what has now been called into question. That appeal would be adequate if, as McDowell thinks, the skeptic’s challenge arises from failure to recognize that mastery of a language can give me perceptual knowledge of meanings, knowledge which is not vulnerable to the threat of regress. But in fact the skeptic’s challenge arises from a question about how we can acquire such knowl-
S kepticism and Q uietism
31
edge: how we can get into a position from which a sign like “+” is revealed not as “normatively inert” but as, manifestly, meaning addition. A symptom of the shortcoming I have identified in McDowell’s approach is his use of the signpost example to dislodge the assumption that understanding requires interpretation. Part of the appeal of the example is that it is very difficult, assuming that we recognize a signpost as meaningful at all, to see it as bearing any other meaning than the meaning that it in fact bears. It is indeed possible to see a right-tapered signpost either as a mere board on the post, “just standing t here,” or as serving some non-signpost-related purpose (to hang laundry on, say, or for children to swing from). But as long as we see it as showing the way, we have no choice but to see it as showing that we are to go to the right. It is telling that, when McDowell considers whether a signpost could be understood as pointing in the opposite direction, he does so by imagining how Martians might use things looking like our signposts. We have to invoke Martians or other aliens to make sense of a subject’s understanding a right-tapering signpost as pointing left because we recognize how unnatural it would be for any human being, or community of human beings, to do so. Trying to imagine a human being who naturally sees a right-tapering signpost as pointing left is akin to imagining the more extreme case, mentioned by Wittgenstein in Philosophical Investigations §185, of someone who naturally responds to a pointing gesture by looking in the direction from fingertip to wrist. This makes signposts very good examples for illustrating the possibility of understanding without interpretation. The felt need for interpretation arises only when a sign is perceived to be, in itself, neutral among a wide range of possible meanings. That does not happen with a properly designed signpost. If a signpost is recognized by us as pointing at all, it is recognized as pointing in the direction from post to tapered end.31 But this very feature of signposts prevents the moral which McDowell draws from the signpost example—that there can be understanding without interpretation—from carrying over to the signs of language and their mental correlates. For the signs of language are different from signposts in that we can recognize that the function of a sign is to tell us something without recognizing what it tells us. This difference is familiar to anyone who has traveled in a country where she does not understand the language. If I do not know any French, I w ill not know w hether “Sortie,” seen on a right-tapering placard, is the name of a place or has a descriptive
32
N ature and “ S ec o nd N ature ”
meaning, but I w ill be in no doubt that if I want to go to wherever or what ever “Sortie” is, I should go to the right. 32 This means that, contrary to what McDowell claims, linguistic signs are different from signposts with respect to how readily they can “lapse into normative inertness” and thus engender the temptation to embark on the regress of interpretation.33 Even the signs of our own language can be seen by us, without recourse to the imagined perspective of Martians, as bearers of multiple meanings, as in the 2009 case of the Dutch advertising slogan “Mama, die, die, die . . .” (Mama, that one, that one, that one . . .). 34 It is this capacity for bearing multiple meanings, not shared by signposts, which Kripke’s skeptic exploits in questioning my confidence that I previously meant addition by “+.” This last point is related to my understanding of the skeptical argument as turning on the subject’s entitlement to believe that responding to “68 + 57” with “125” fits previous observed uses of “+” and hence that she has been successful in learning the meaning of “+” from the examples she had been shown. The argument is effective for linguistic signs because, at least in some cases, we can master them only by learning how to go on from examples, something which often takes considerable training, for example in the case of color words. But our understanding of signposts does not depend on this kind of ostensive learning. Although we may have to be taught in general what a signpost is for, we do not have to be trained on multiple examples of right-tapering and left-tapering signposts in order to be able to determine, for a given signpost, w hether it means “go right” or “go left.” Rather, we can rely on the perceived similarity of a signpost to a pointing gesture, the understanding of which requires no training.35 This difference is obscured in McDowell’s discussion. Following Wittgenstein, he speaks of “training” in the use of signposts, suggesting an analogy between how we learn to understand a signpost and the ostensive learning through which we come to understand linguistic signs.36 Relatedly, he speaks of our being “party to . . . conventions” for the use of signposts, suggesting that it is a m atter of convention that a right-tapering signpost 37 points to the right. But although there are conventional aspects to the use of some directional signs (for example that, in many countries, an upward-pointing arrow indicates that one should go straight ahead rather than up), and t hese need e ither to be explained or learned from example, the fact that a signpost is to be followed in the direction from post to
S kepticism and Q uietism
33
tapered end is not one of them. So the undeniable fact that we typically understand signposts without needing to interpret them does not touch the problem about the meaningfulness of linguistic signs which is revealed by Kripke’s skeptic.
5 I have been arguing that McDowell fails to appreciate the full force of Kripke’s skeptical problem, and that, as a result, his quietism about meaning, at least in connection with that problem, is unjustified. In order to get the measure of the problem, we have to recognize that the skeptical attack on meaning proceeds by way of an attack on the knowledge— essential to the possibility of learning the meaning of an expression— that, in one’s individual uses of an expression, one is g oing on appropriately from the previous uses one has undertaken or observed. Once that is recognized, we see that the skepticism cannot be defused by pointing out that someone who has grasped the meaning of the expression can recognize immediately what the expression meant in those previous uses. Kripke’s skeptic can allow that, in principle, a competent language-user can understand the expressions of her language, and the significance of the mental states in which her understanding is realized, without need for interpretation. The question remains of how an individual can know that her present use of an expression accords with previous uses, given the skeptical possibility that the expression in those previous uses bears a meaning diff erent from the one she now thinks it does, and hence that she is not a fter all, at least with respect to that expression, a competent user of the language. It is this kind of question which Kripke’s skeptic raises in asking how I know that, if I am to accord with previous uses of “+,” I should now say “125” in response to “68 + 57.” In effect, he is calling into question whether I am, at least in respect to the expression “+,” a competent language-user: that is, someone who knows how to go on correctly from previous uses of that expression. I now want to claim that the skeptic’s question can be answered. The key is to reject an unexamined assumption underlying the skeptical dialectic: that for my present use to accord with previous uses of an expression, it must accord with what was previously meant in those uses.38
34
N ature and “ S ec o nd N ature ”
Kripke takes for granted that, if I now want to remain consistent with my earlier uses of “+,” or in other words, to go on appropriately from those past uses, I must use “+” with the same meaning as that with which I used it previously.39 It follows, given that assumption, that if I meant quaddition in those previous uses, conformity to t hose uses requires that I respond to “68 + 57” with “5,” not “125.” But the assumption is unwarranted. The question of which response to “68 + 57” accords with my previous responses to “+” questions, or more generally with my previous uses of “+,” is distinct from the question of which response accords with what I meant by “+” when I gave those responses. I can allow to the skeptic, at least provisionally, that I meant quaddition in my e arlier uses of “+,” so that saying “125” fails to accord with the meaning I then associated with “+,” and still maintain that “125” accords with my previous uses. The idea that, regardless of what I meant, “125” accords with my previous uses can be brought out especially clearly if, as I have been proposing, we think of the subject in the skeptical scenario as if she were a language-learner who is in the course of learning the meaning of “+” from examples of its use. The possibility of successful learning depends on her being able to come to see, for instance, that the correct way to go on from a series of prior uses which includes responding to “18 + 7” with “25,” “28 + 17” with “45,” “38 + 27” with “65,” “48 + 37” with “85,” and “58 + 47” with “105” is to respond to “68 + 57” with “125,” independently of any assumption about what was meant in those prior uses. If, once she has come to see this, someone convinces her that the users in those cases all meant quaddition, she can still insist that responding to “68 + 57” with “125” is the appropriate way to go on from the examples she was given, considered in abstraction from what the users meant by “+” when they responded to the “+” questions as they did. Saying “125” in response to “68 + 57” counts as g oing on correctly from the sequence of prior uses, even if it yields a false answer to the question “68 + 57” when the question is understood as the previous users understood it. On this approach to the skeptical problem, the skeptical conclusion is avoided not by rejecting the skeptical hypothesis, but by pointing out that it does not lead to the disappearance of meaning. Recall that, on the interpretation of the argument I have offered, meaning “vanishes into thin air” b ecause my saying “125” in response to “68 + 57” is seemingly revealed as an unjustified leap in the dark. T here can be no meaning if each use of
S kepticism and Q uietism
35
an expression is perceived by the user as arbitrary in the light of previous uses. But now note that the ground for concluding that “125” is a leap in the dark is my inability to rule out the skeptical hypothesis that I meant quaddition. The apparently “insane” suggestion that I should say “5” gets its seeming force from the skeptical hypothesis: if that hypothesis had not been presented, then I would have dismissed the suggestion, and the skeptical dialectic would have been nipped in the bud. Now a solution on the lines that McDowell proposes would have me insist against the skeptic that I meant addition rather than quaddition. But the solution I am proposing is to allow the hypothesis, at least provisionally, but to point out that it is irrelevant to the question of how, if I am to accord with my previous uses, I should respond to “68 + 57.” There are now no grounds for claiming that “125” is a leap in the dark, and so no grounds for doubting that my use of “+” is meaningful. Once the skeptical challenge has been disposed of in this way, I can go back and deny the skeptical hypothesis. Since I no longer have grounds for doubting my own linguistic competence, I can be secure in my conviction that I now mean addition by “+,” and—in the absence of grounds for believing that I have undergone a dramatic change in my understanding of “+”—I have every right to maintain that I meant addition in the past as well. But it is important to see that my being able to take this common-sense line with respect to the skeptical hypothesis depends on my having already defused the skeptical threat to the possibility of meaning. It is only because I can be confident, independently of any considerations of what I meant by “+” in the past, that I am now g oing on appropriately in my use of “+,” that I can resist the skeptic’s attempt to cast doubt on the obvious fact that I now mean, and meant, addition. What if I am now asked what grounds I have for taking “125” to fit my previous uses of “+,” independently of what I meant by it? It is open to me at this point to appeal to features of those previous uses in virtue of which “125” is appropriate. For example, I can list the uses mentioned two paragraphs above, and point out that if we look at the tens place for each answer successively, we see the sequence “2, 4, 6, 8, 10,” and if we look at the units place, we see the sequence “5, 5, 5, 5, 5.” As long as my interlocutor sees “68 + 57” as according with the previous series of questions, she should see that “125” accords with the previous series of answers. But what if my interlocutor asks me for grounds for, say, taking “12” to be the appropriate
36
N ature and “ S ec o nd N ature ”
continuation of “2, 4, 6, 8, 10,” or “5” the appropriate continuation of “5, 5, 5, 5, 5”? H ere there is nothing I can appeal to by way of further justification. I can say, of course, that if we could not be confident of the correctness of these ways of g oing on, then t here could be no learning from examples: we could never come to grasp the meaning of expressions like “add two” or “same number,” let alone “+,” if we could not rely on our intuitions about how to continue sequences of the kind presented. But that might provoke a skeptical rejoinder: so much the worse for the possibility of learning! If I cannot justify the correctness of “125” in the light of my previous uses of “+,” then, this skeptic w ill say, t here can be no learning the use of “+” or of any other expression from examples, and, at least if I am right to think that the possibility of such learning is essential to the possibility of linguistic meaning, we have to accept Kripke’s skeptical conclusion a fter all. However, this skeptical challenge is diff erent from, and considerably weaker than, the one offered by Kripke’s skeptic. Kripke’s skeptic does not simply demand a justification for the correctness of “125” in the light of my previous uses, and then conclude, from my inability to provide a justification, that “125” is a leap in the dark. Rather, in raising the possibility that I meant quaddition, he provides an argument for questioning my confidence in the correctness of “125.” And on the face of it, the argument is a good one. It turns out to be insufficient only once we reject one of its premises, namely, that the correctness of “125” in the light of my previous uses depends on its accordance with the previous meaning of “+.” By contrast, the challenge just described is not motivated by any argument. Rather, it seems to arise from nothing more than a general puzzlement about how it is possible for a finite sequence of uses to be recognized as making appropriate any one continuation rather than any other, and so how it is possible to learn the use of an expression from a finite set of examples. But there is no reason why we should find this philosophically puzzling. It is simply a pervasive feature of human experience that we find it appropriate to go on in certain ways from the behavior exhibited by teachers and peers, and, absent any specific ground for doubt (such as that raised by Kripke’s skeptic) there is no reason to question our confidence in the legitimacy of our ways of g oing on. My approach to this weaker skeptical challenge is comparable to McDowell’s approach to the stronger skeptical challenge mounted by Kripke. As we saw, McDowell rejects Kripke’s skeptical challenge to lin-
S kepticism and Q uietism
37
guistic meaning because he sees it as insufficiently motivated. On his reading, it rests simply on puzzlement about something which he takes, in fact, to be quite unmysterious: the possibility of linguistic and m ental items (like “+” and the thought “I meant plus”) which can be meaningful without interpretation. Although he thinks it takes work to remove the puzzlement, for example, drawing our attention to “common sense” about signposts, he does not think of the skeptical challenge as a genuine problem. I am adopting a similar quietism, but not with respect to the skeptical problem about linguistic meaning, which I take to be a genuine problem for which Kripke has provided an apparently convincing rationale. Rather, I am being quietist about the supposed problem about the legitimacy of the normative attitudes on which, I have suggested, language-learning and hence linguistic meaning depend. Once we have cleared away the mistaken assumption underlying Kripke’s skeptical problem about meaning, namely, that one’s present use of an expression can accord with previous uses only by according with the meaning they bear, we can see that there is no problem, per se, with the idea of someone’s recognizing a normative fit between her response of “125” in response to “68 + 57” and the previous responses which have been given to “+” questions. The question of what justifies her in taking “125” rather than “5” to fit the previous responses— or, if the dialectic has proceeded as suggested in the previous paragraph, of what justifies her in taking “12” to fit the sequence “2, 4, 6, 8, 10” or “5” to fit the sequence “5, 5, 5, 5, 5”—can be rejected, on this quietist approach, as not adequately motivated. Ultimately, then, I am sympathetic to an approach which, like McDowell’s, recognizes that some philosophical questions are best addressed by pointing out that they lack motivation and so are not good questions. Where I disagree with McDowell, though, is on the point at which we need to take a quietist approach to the philosophical questions which arise when we consider the phenomenon of human language use. McDowell’s quietist approach starts relatively early on, when we ask, with Kripke’s skeptic, how linguistic meaning and understanding are possible. It makes use of the fact that, as long as we do not seek to understand linguistic meaning from what McDowell calls a “sideways-on” perspective—that is, in a way which attempts to abstract from the knowledge we have in virtue of our participation in the practices which constitute language use—there is no problem about its possibility.40 We simply see and hear what people
38
N ature and “ S ec o nd N ature ”
are saying and doing when they use linguistic signs. But McDowell does not recognize that Kripke’s skeptical question about the possibility of meaning is motivated by a further question about how we can become initiated into linguistic practices, and so get into a position from which we can unproblematically recognize signs as bearing the meanings they do. In order to take that question seriously, we need to think our way into the perspective of someone for whom the signs of our own language (say, En glish) do not manifest their meanings immediately, but must be learned: either by having their meaning explained to us, as in the case of someone who already knows another language, or by being shown examples of their use, as in the case of a child’s earliest language-learning. This does require taking a kind of sideways-on view, and, in the case of the child, a quite radical one. But although it is sideways-on with respect to the use of Eng lish, and in the case of the child, with respect to the use of language altogether, it is not sideways-on with respect to the norm-governed practices in which we participate simply by virtue of being h uman beings, and which make the acquisition of language possible. It takes for granted our recognition of the normative significance of the gestures, facial expressions, and other forms of behavior that are involved in the teaching of language: for example, pointing gestures (with the hand, or, in some cultures, with the chin), smiles, frowns, and the kind of physical intervention (in some cultures, moving a child’s hand to encourage her to wave bye-bye) which is intended to get her to imitate some piece of adult be havior. It also takes for granted our recognition of certain behavior as fitting or failing to fit sequences of previous behavior: for example, in the case of a very young child, the appropriateness of covering and uncovering her eyes during an impromptu peekaboo game, or of responding to an adult repeatedly clapping his hands by clapping her own hands; and in the case of a more sophisticated child, the appropriateness of continuing the sequence “0, 2, 4, 6, 8. . . . 1000” with “1002,” or of continuing the “sequence” of answers to “+” questions by answering “68 + 57?” with “125.” 41 Now, an even more radical sideways-on view would abstract even from the knowledge we have simply as human beings, attempting to take the perspective of imagined Martians who (to elaborate on McDowell’s example, cited in section 4 above) see t hings like our signposts as pointing from tapered end to post: either b ecause they see things like our pointing gestures as pointing in the opposite direction from ours, or b ecause they
S kepticism and Q uietism
39
see a right-tapered signpost as more similar to the gesture of pointing to the left than to the gesture of pointing to the right. Such Martians might have facial expressions which look like smiles but which they see as calling for a violent attack, and, if they have things which look like our numerals and arithmetical signs, they might find it appropriate to continue what looks like our “0, 2, 4, 6, 8. . . . 1000” sequence with what looks like “1004” or to respond to what looks like our “68 + 57” with what looks like our “5.” The seeming conceivability of this Martian perspective might lead us to question our grounds for taking our own responses, in the examples given in the previous paragraph, to be correct. In virtue of what, we might ask, is it appropriate to respond as we do to the behavior we observe in o thers? How could we justify our ways of going on to a Martian, or to an imagined human with Martian-like behavior, who insists (or appears to insist) that the appropriate continuation of “0, 2, 4, 6, 8. . . . 1000” is “1004”? It is with respect to these questions that I think that quietism is warranted. While we can examine human behavior from a standpoint outside language, we cannot do so from a standpoint outside more basic human practices and the normative attitudes which inform them. Our standpoint as human beings simply allows us to recognize, in the examples I have given and in countless others like them, the appropriateness of the responses which we are, in fact, naturally inclined to make; and the fact that there is no justification to be given in response to an i magined Martian challenge does not show that this recognition is illusory. My approach to these questions is not only in the spirit of, but largely inspired by, McDowell’s quietist approach to skepticism about meaning. My point of contention with McDowell is just that, in failing to recognize the possibility of a viewpoint outside language but still within h uman practices, he endorses quietism too soon.
Chapter Two ✣
Forms of Nature “First,” “Second,” “Living,” “Rational,” and “Phronetic” ✣
Michael Thompson
I
want to consider the prospects for a certain kind of theory of practical reason and ethics generally—namely a normative “naturalism,” a naturalism of first nature, a doctrine, as we might say, of practical goodness as natural goodness. I w ill speak of such a doctrine sometimes simply as “practical naturalism,” but we might call it “naïve Aristotelianism.” It is a naïve Aristotelianism, I say, because it stands opposed to the sophisticated naturalism of “second nature” that has been occasionally propounded by John McDowell; it is an Aristotelianism because it seems to have been accepted, in all its naïveté, by Aristotle. I do not know if any such theory is true, but I would like to find out if any might be. In these remarks, I w ill attempt to prepare the ground for the formulation of such an account mostly by bringing out presuppositions that block the way to it. I w ill give a number of what I suppose are equivalent definitions or characterizations of this sort of practical naturalism in the remarks that follow, but we may provisionally say that on such an account the concept 40
F o rms o f N ature
41
h uman is in some sense the central concept of practical philosophy as we attempt it. Here it is crucial that the word “human” is in a certain way put on a level with words like “Norway rat” and “coastal redwood,” and that the concept of a human is distinguished from similar imaginary concepts like those of a Martian or one of the other sorts of extraterrestrial reasoning animal that appear in science fiction stories and in philosophy. It expresses something of which two arms and two legs and a certain sort of nervous system are characteristic. The word “human” can be said to express a specific nature which individual material things can bear, one which came to be many tens of thousands of years ago on one planet by a process of Darwinian evolution. A certain concept, human, to which we can attach all those predicates, is, on the practical naturalist account I am considering, at the same time the central concept of practical philosophy, as we would write it. Thus the concept of the specifically h uman has the same position in a practical naturalism that the concept of a rational being or a “person” has in Kant’s system. The concepts of a person or rational being, as Kant understands them, would cover Martians and even angels quite as well as you or me. Where Kant makes use of a conception of practical reason in general, or of pure practical reason, the same in content for all who can be said to reason practically at all, and where he speaks of the w ill in general, a naïve Aristotelian makes use instead of a conception of a specifically human power of practical reason and the specifically human w ill. A naïve Aristotelianism, or practical naturalism, has lately been propounded by Philippa Foot in her book Natural Goodness.1 Such a theory was also, I think, maintained by Karl Marx throughout his c areer, though it was never made fully articulate because with time his peculiar obsessions led him elsewhere. This claim about Marx involves highly contentious interpretative issues, so, though I w ill mention the young Marx a bit later, I w ill mostly stick to Foot’s Natural Goodness as an illustration of the sort of view I am proposing to make possible. (I mention Foot and Marx as the philosophically sophisticated modern proponents of such a view; there are others less sophisticated.) The trouble is that Foot seems not to have made it clear that her theory is not subject to the criticisms that John McDowell, among others, has advanced, and to a number of other objections that spring immediately to mind in the contemporary thought environment. The core of all these
42
N ature and “ S ec o nd N ature ”
objections, to put the matter crudely, is that any such naturalism w ill express a sort of reductive empiricism perhaps coupled with an alarming and idiotic moral conservatism. Foot has not done enough work, I think, to break down the extreme alienation from the item expressed in the concept human, on the part of its contemporary bearers. Let us remember that in formulating his sophisticated Aristotelian naturalism, McDowell suggests that we can only see our way to a satisfying nonreductive naturalism in practical philosophy if we first combat a “constriction that the concept of a nature is liable to undergo in our thinking.”2 McDowell’s remark about a “constriction” is indeed the beginning of wisdom on the subject; the contemporary allergy to what I suppose are authentically Aristotelian ideas does I think arise from a wrong or narrow conception of nature or the nature of a thing. The trouble, McDowell thinks, is that on the ordinary understanding of nature a nature cannot be something a thing comes to bear, or to bear fully, by learning and habituation. It is, he thinks, only if we realize that some natures are so- called second natures that we can see how their full instantiation can be a matter of habituation, learning, acquisition. And it is only by reclaiming a conception of nature that includes these second natures that we can give a proper place to the concept of nature in the philosophy of the practical. But study of the text w ill show that McDowell’s second natures are basically practices which individuals come to bear or acquire (not to put too fine a point on it); they are cultures, or Bildungen, as he says, or bits and pieces of them—but considered as internalized by individual subjects and configuring their thoughts and perceptions and feelings. To speak of such a thing as a nature is all well and good. But the sort of nature in question is not the sort expressed by the word “human”; this nature McDowell, I think, discounts as a mere first nature. It is something the possession of which might make possible the possession of nature in the sense used in his practical naturalism, namely second nature, but it is not that nature. I don’t pretend properly to comprehend McDowell’s thoughts on the matter, but my belief is that we can turn his own formula against what appears the most natural reading of him. McDowell speaks of a constriction that the concept of nature undergoes in our thinking, and proposes to free us from it. But perhaps we can speak of a certain constriction that the concept of a first nature has undergone in McDowell’s thinking, and propose to free him from it. He is, as we might say, construing the concept
F o rms o f N ature
43
first nature precisely as his opponent would. McDowell is not prepared to “let” his opponent “have” the concepts of nature and naturalism and thus to dispense with them in practical philosophy. But it seems he is “letting” his opponent “have” the concept of first nature. This accounts for the downplaying of the concept h uman in his theory, and the tendency that I at least find in it for this to become the name for a sort of infrastructure or base upon which arises a superstructure of second nature, which is what is really interesting in practical philosophy. The concepts of second nature, practice, culture, habituation, and so on are of course of decisive importance in practical philosophy; my difficulty is with the role of concepts of first nature in his account. We must indeed make some such formal distinction as McDowell expresses in the terms “first vs. second nature.” But we must also make a certain formal distinction among first natures. The break with vulgar bald naturalisms does not come or does not simply come with an expansion of the concept of a nature that would permit recognition of second natures alongside first; it must come with an expansion of the concept of a first nature that would permit it to cover all that is really contained in such a concept as human.
1. Practical Cognition as the Core of the Difficulty It seems to me that the core difficulties that keep us from accepting a naïve Aristotelianism or a practical naturalism or a natural goodness theory, or whatever you want to call it, arise basically from moral epistemology, or from practical epistemology more generally. What is expressed by the uman, considered as something to put alongside Norway rat, sago concept h palm, and Martian, seems to everyone to have the wrong relation to our knowledge to count as anything relevant to fundamental ethical theory or the philosophy of the practical generally. It seems it must be something alien and outside and external, something given to us empirically. This is why Kant is so emphatic about dispensing with (what I am calling) the concept human within practical philosophy; it is something alien, impure, empirical; to introduce it into our principles would be to sully them with empirical pollution; we must replace this dirty concept with the pure concepts of a rational being in general or of a person, and address our groundworks and critiques to all such beings, aiming for a translation into
44
N ature and “ S ec o nd N ature ”
Martian languages as well as non-German human ones. This is Kant’s “favorite idea,” as Schopenhauer says mockingly, “—he never tires of repeating it.”3 The point about the translations Kant is envisaging for his Grundlegung provides a good way to reformulate the distinction between a Kantian and a practical naturalist view: Aristotle, on a naïve Aristotelian reading, is aiming to write a book that can be translated into any human language, and w ill serve the same function among speakers of any of them. He (I think) denies the possibility of an Ethics for men and Martians alike. Kant is emphatically writing for Martians as well as men. The aspect of Kant’s thinking that is in question is deeply entrenched, though outside the practical philosophy scholars inevitably finesse it. The break with the concept human, and its replacement by something more abstract, is crucial to his development and to the form of his mature rhe toric. Martians and Venusians are already discussed at absurd length, and their mental strengths compared, in Kant’s early Universal Natural History. The mysterious “we” and “our” of the Critique of Pure Reason evidently cover not just all Germans, nor all humans, but neither, on the other hand, is he speaking to all possible discursive reasoners (though they might take on some of its propositions, of course). “We” are, I think, all the finite rational beings that any one of us meet with in experience, all who are faced with this same connected system of nature. So the first critique is addressed to the Martians and Venusians of the Universal Natural History—as is, he thinks, a textbook of Euclidean geometry—but not bearers of other possible forms of intuition; and remarks like “space is the form of our outer intuition” have a difficult kind of circularity to them, since they may be unpacked as meaning something like “space is the form of outer intuition of those who are among the intelligent intuiters of things in space.”
2. Our Difficulty Pertains to the Character of Phronêsis I said that our difficulty in making room for naïve Aristotelianism arises within practical epistemology. To overcome the difficulty w ill thus first involve developing the concept of practical knowledge and to some extent the correlative concepts of self-k nowledge and self-consciousness. And it
F o rms o f N ature
45
ill also involve, of course, developing the concept of a first nature, and the w specific form of first-nature concept that we employ in framing the concept human. A certain narrow conception of practical knowledge w ill be expanded, and a certain narrow concept of first nature w ill be expanded; in the end we w ill see that a certain form of practical knowledge we might hope to possess w ill be knowledge of a certain sort of first nature, namely the first nature we bear, the one expressed in the word “human.” I just said again that the decisive impediment to naïve Aristotelianism arises in moral or practical epistemology, as we might call it. For it seems that the naïve Aristotelian is fated to give what David Wiggins called an “unconvincing speaking part to facts about h uman nature” in the account and justification of ethical knowledge and judgment.4 Wiggins’s remark about the threat of giving an implausible speaking part to facts about our nature is more than once quoted by McDowell with approval. Though Wiggins and McDowell are speaking as sympathizers with Aristotle, the form of their objection to naïve Aristotelianism is of a piece with that posed by Kant. The use of the word “facts” suggests, at least in contemporary Eng lish, that we have to do with a content of empirical knowledge— knowledge, as Frege says, of what is given to us as something alien and without, through the medium of the senses. The naïve Aristotelian is supposed to go “out” to the object, and discover its properties, and then somehow give them a speaking part in his exposition of the first principles of practical reflection. It looks like the naïve practical naturalist is introducing a scheme of practical inference of the form “Men dance, dancing is something that belongs to h uman nature, dancing is what is natural to them—so I’ll dance too.” It does seem plain that if this is right, then the theory is very silly, and not because of the silly content I have entered into the syllogism. On such an account virtue w ill be coming to let something alien rule one’s practical thoughts. As a dispute among would-be Aristotelians, then, the dispute pertains to the character of intellectual virtue or excellence, and thus of what we would have to call some kind of knowledge—and it pertains in particular to practical wisdom, phronêsis. I’m not sure, but Aristotle seems to think of phronêsis as a rather refined state involving a kind of articulateness and reflectiveness about practical m atters such as might arise from attending his lectures—so we might speak rather of a more rustic practical understanding—a perhaps inarticulate knowing of what r eally matters, or of
46
N ature and “ S ec o nd N ature ”
how to live—taking this to be the more generic intellectual virtue found even in unreflectively virtuous p eople. This practical understanding is the intellectual, thinking, judging aspect of the sort of thing Aristotle thinks his hearers w ill bring to the auditorium, if he’s going to get anywhere with them. It should be noticed that Aristotelian critics of naïve Aristotelianism like Wiggins and McDowell take their start from Aristotle’s discussion of practical understanding in Book VI of the Nicomachean Ethics. This sublime discussion of the character of choice and deliberation, of practical thought and knowledge and so forth, seems to leave no room for what we would expect from what I am calling a naïve reading. Aristotle nowhere drags anything we might call “human nature” into the picture as material for practical reflection; his picture of deliberation does not find a place for thoughts about the human in general; he does not lend it an “implausible speaking part.”
3. The Object of Phronêsis How might we formulate naïve Aristotelianism in connection with phronêsis or practical wisdom or practical understanding? There is a section heading in Heidegger’s series of lectures on Nicomachean Ethics, Book VI—a series of lectures which, with characteristic Heideggerian perversity, he makes preliminary to a reading of Plato’s Sophist. German readers need to remember h ere that, in English translations, Heidegger’s peculiar restricted use of the German word “Dasein”—to express the form of being each of us exhibits, the form of being that is mine—is not translated; it is just treated as an English word, “Dasein.” So in English translation, a fter many dark remarks on the nature of practical understanding or phronêsis, Heidegger brings things to a characteristic crescendo with the section heading “The Object of Phronêsis: Dasein Itself.”5 I think there is a characteristic mistake in this, a hidden Kantianism, but I like the form of the formula, or the question, what the object of phronêsis is. The opposing formula of naïve Aristotelianism is this: the object of phronêsis is the h uman itself. Or as we might equivalently put it: the object of phronêsis is specifically human life. What the phronimos understands or “knows”—what their intelligence has got right with—is life in the sense of h uman life, the life we find wherever we find an individual h uman being.
F o rms o f N ature
47
Here, again, the word “human” (man, anthropos, Homo, Mensch) is to be understood as a word expressing something that is somehow shared by me and a Trobriand islander but not by me and Kant’s Martians and Venusians; it catches hold of something that is in some sense on a level with Norway rat and sago palm. The “object of phronêsis” as I might possess it is not something that need enter into the thoughts of possible Martians. By contrast, again, the object of practical wisdom in Kant is reason itself or reason’s law—something, in any case, that pertains to would-be Martians as well as to myself, something with that kind of generality or abstractness. Rational being or person is a general category like organism or animal, or, equivalently, vegetative life and perceiving life; it is a category, if you like, not a concrete concept like Norway rat or sago palm. But notice that the same is true of the “object of phronêsis” on Heidegger’s understanding; this is his hidden Kantianism, and it accounts, I think, for the deadening abstractness and emptiness of his practical philosophy. Whatever Heidegger’s Dasein may be, exactly, it is perfectly clear that this form of being could be constituted on a distant planet inde pendently of our own. It is, as we might say, a formal or categorial concept, just as the concept of a rational being is. The concept is supposed to catch a form of being in the sense in which it belongs to metaphysics to apprehend the forms of being. In any case, the naïve reading says that if I assign an object to practical understanding, a controlling theme, that is the word I w ill have to use, namely “human,” or “human life,” or “specifically h uman life.” The bearers of practical understanding have a certain kind of knowledge, they’re right about something, and the thing they’ve got in line with is expressed by the word “human,” and nothing fancier, more abstract, or more extensive. It w ill of course follow that insofar as the phronimoi have possession of some kind of practical “principles” or any general forms of reason for doing something, what they grasp w ill have application only to human beings. Now, that either “the human” or “Dasein” could be put as the “object” of practical understanding or phronêsis might seem equally hopeless as an interpretation of Aristotle. Each seems to violate Aristotle’s own claim about what practical understanding is “about,” what it concerns, what it’s “peri,” as he says. The words “human” and “Dasein” express something somehow general or universal, so a ren’t both Heidegger and the naïve
48
N ature and “ S ec o nd N ature ”
MAJOR PARTS OF THE PART THAT HAS LOGOS epistēmonikon
logistikon
WHAT EACH MAJOR PART CONCERNS OR IS “PERI”
what is universal what is particular
what is not “variable” what is “variable”
RELATION OF EACH MAJOR PART TO WHAT IT’S ABOUT/“PERI”
not controlling
controlling
TYPES OF VIRTUES OR EXCELLENCES RESIDING IN EACH MAJOR PART sophia
epistēmē 1, epistēmē 2, epistēmē 3 . . .
phronēsis technē 1, technē 2, technē 3 . . .
(Note here that the lower forms of cognition are numerous: there are many epistēmai and technai I might attain, but the phronēsis and sophia I might seek are single.)
Figure 1
Aristotelian saying that the phronimos knows something about a certain universal? And it seems this must be wrong. It w ill be remembered that Aristotle operates with two distinctions in Book VI and four basic types of cognition that merit the title of intellectual excellence or virtue (Figure 1). All four cognitive types are habitual states—hexeis. The exercise of an epistêmê is in some kind of contemplation of the object of the epistêmê, a specific kind of concern with it. If I have epistêmê in connection with Norway rat, or equivalently, of Norway rats—a bit of scientific knowledge of them—then I have it in my sleep, it’s a habit or hexis. When I wake up and exercise it, it w ill be in thoughts about Norway rats in general. Exercises of sophia w ill be in general meditations on God and being and whatever comes into Aristotle’s metaphysics, I suppose.
F o rms o f N ature
49
By contrast, if I have shoemaking craft, I again have it in my sleep, and when I wake up and exercise it, the thoughts that fill my mind w ill concern these materials, the peculiarities of this foot, the one to be shod, and so forth. Similarly, exercises of practical wisdom, which are happening all over the world right now, God willing, w ill be in concern with the things the particular agent confronts, their possibilities of action in particular. The solution for the naïve or Heideggerian Aristotelian is plainly to distinguish claims about “what the state is concerned with,” what it’s peri— or what the state is forever on about in the exercise—from a claim about “what its object is,” a claim of the sort the Heideggerian Aristotelian is trying to make. What the shoemaker knows is the craft of shoemaking, or shoemaking, or how to make shoes, and this is something in some sense universal and general, something possessed by others in just the same way at any time, from one of whom they learned it; this is compatible with the fact that, as we said, the shoemaker’s exercises of this knowledge are not in meditations on shoemaking in general, they are not “peri” shoemaking- in-general, but in each case “peri” this leather and this foot and so on. So also, the Heideggerian Aristotelian and the naïve Aristotelian are at one in thinking that the practically wise or understanding person is onto something that is in some sense universal and exhibited by others, some long dead, who can be said to have or have had the same knowledge in the same way, from some of whom they acquired it, or some of what they have of it. This is consistent with the fact that it is not exercised in meditation on any universal; the exercises are not peri anything universal, but concern what is particular and admits of variation. uman or human life can The question, though, is whether the concept h be supposed to catch the universal in question, or catch its extension, and thus what Heidegger calls the “object” of the hexis—or whether instead a categorial abstraction like Heidegger’s Dasein or Kant’s rational being is all we can put in this place.
4. Aristotle’s View I w ill now give a short proof that for Aristotle the object of phronêsis as any of us might have it is the h uman itself, or h uman life itself, or something of the sort—something that belongs to Trobriand islanders as well as to me,
50
N ature and “ S ec o nd N ature ”
but not to a Martian. That is, I w ill prove that his Aristotelianism is naïve, and derive a few theorems that a contemporary Aristotelian, who rejects his cosmology, might affirm. In the hope of proving Aristotle’s naïveté, one might appeal to the celebrated ergon argument of Book I, about the so-called ergon of the human or “the function of man.” But that text is actually a bit ambiguous without further support; Kantians have been known to use it in their never-ending efforts to enlist Aristotle; they say that the ergon of man is just the ergon of a rational being in general. This is not so absurd given that Aristotle assigns an ergon to plants en bloc, and to brute animals en bloc. (This suggests the general abstract category view, for t hese are philosophical categories in Aristotle, as we see in De Anima.) The basic difficulty arises from the fact that for Aristotle, human beings, men, anthrōpoi are the only rational animals in the cosmos; his cosmos, a fter all, only runs a few hundred miles up. I am brought together with you u nder the formal concept of a rational animal and the concrete concept human being; which unity is Aristotle latching onto in assigning an ergon or function or whatever to anthrōpos? It w ill be better to stick to the materials of Book VI, the discussion of practical intellect. As I said, this is in any case the ground upon which sophisticated Aristotelians like Wiggins and McDowell (and Heidegger) base their readings. One of the main purposes of Aristotle’s ethical works is to sing the praises of sophia, or theoretical or philosophical wisdom, as Ross puts it in the standard Eng lish translation; sophia is supposed to be something higher and better and more sublime than phronêsis, or practical wisdom, as Ross translates it. Wisdom in theoretical m atters, sophia, is some kind of comprehensive grasp of the main points about the most exalted things; we learn later, of course, that it is shown especially in contemplation of the divine or in sharing in the divine self-thought. Its superiority over what ever it is that the merely practically wise have is a topic of Book VI, Chapter 7. One point in sophia’s favor is a formal feature of the state itself: Aristotle says that it is like the straight and the white, everywhere the same. His thought is presumably that t here’s only one God-and-the-world- order to apprehend, a fter all, only one divine self-thought to share in. Either a finite intellect has got it right or not. By contrast, Aristotle says,
F o rms o f N ature
51
phronêsis is like healthy and good—diff erent for man and fish. By the end of the passage, it is clear that Aristotle has inadvertently imagined some kind of rational fish—Barracuda sapiens, as we might call it; his thought must be that individual fishes of this extraordinary kind might be divisible into the practically wise and practically unwise too, but that it would be on a diff erent ground than would be used in our case. As we move, hypothetically, from Barracuda sapiens to Homo sapiens, we move from one shape of phronêsis to another. This is perhaps one of the first appearances of something with the role of the concept Martian in a written text. (Vaguely sensing the strangeness of this hypothesis, Aristotle then spoils things a bit, remarking with embarrassment that we do attribute something like phronêsis in a way to brute animals that exhibit a certain level of foresight.) It is clear that the difference between Barracuda sapiens and Homo sapiens, man and some kind of rational fish, which is a distinction of first nature, is supposed to be of decisive importance in the characterization of phronêsis. (Note h ere the difference between a view about the content of practical wisdom, insofar as it can be articulated, and a view about “what is known” or “understood,” what the “object” is, as Heidegger puts it: if we care to imagine a totally identical twin Earth, then for sure the content of practical wisdom there w ill be the same for the twin humans up there, in some sense; nevertheless, something diff erent w ill be known. Similarly, knowledge of a twin English that arose independently in the South Seas would have the same content, I guess, but what is known or grasped or understood, what the “object” is, is something diff erent, namely, another language.) Aristotle’s second argument in favor of the superiority of sophia directly concerns what is known through the state. The argument presupposes the background principle (which seems to have little g oing for it, frankly—but who knows?) that cognitions are better according as their objects are better. But, Aristotle says, there are a lot of better things in the cosmos than the human, for example, the constituents of the heavens. Ergo, phronêsis is not the best knowledge, sophia is. Sophia grasps those better t hings. But in all this reasoning Aristotle thinks he can form the contrast in terms of what Heidegger calls the “object” of the state, and it is clear that the word for this object is “human,” not Dasein in general, as Heidegger
52
N ature and “ S ec o nd N ature ”
thinks, and not rational being or rationality in general either. All of t hose would be exhibited in rational fish. Phronêsis : human :: sophia : God-and- the-w orld-order. There is an implicit proposition that if the human w ere the highest and most estimable object in the cosmos, phronêsis would be the highest knowledge. The thought that the only reason why phronêsis or practical wisdom isn’t the highest wisdom is that the human isn’t the highest thing, suggests an interesting result not stated by Aristotle. Suppose there is such a t hing as an epistêmê or systematic scientific knowledge in connection with the h uman, or h umans, or h uman t hings (ta anthrōpina), as I was supposing t here obviously can be an epistêmê in relation to Norway rat or Norway rats. This knowledge—suppose it to be possessed by a Martian— would not be as good a knowledge of that object as the phronêsis that the humans themselves might possess is. If human life were the best thing in the cosmos, knowledge of it would be the best knowledge, but it would not be an epistêmê pertaining to this life that would be the best knowledge of this best object, though such an epistêmê is there to be had: medicine has possession of some of it, for example. Rather the phronêsis that the individual bearer of this life might hope to have would be the best knowledge of that best object. Moreover, if there is no God, nothing divine, and the starry things in the heavens are burning rocks, but t here are such t hings as Homo sapiens and Barracuda sapiens and Martian, then these w ill be the best things— equally best, I suppose. And the best knowledge that the human can possess, namely, practical wisdom, w ill be diff erent from the best knowledge the barracuda can possess. That is, the best knowledges that are abroad among the men and barracudas would take different t hings as their objects; the second best knowledges w ill be epistēmai attaching to these same best objects, and they will be the same whoever bears them, w hether Martian or Venusian, like the straight and the white. Note h ere the contrast with Kant. In his system, once the moral law has been developed out of my own reason, I w ill insist on applying this law wherever in the cosmos (or outside it) I find reason in any form. Kantian practical rationality is like the straight and not like the healthy; it’s the same whatever y ou’re dealing with. For Kant, sophia and phronêsis would run along in lock step; each is available where the other is and each says the same in e very case. Any attempt to narrow the scope of the phronêsis that
F o rms o f N ature
53
I have, if I have any, any attempt to restrict the content of my knowledge to fellow humans, would saddle it with empirical dirt and destroy it.
5. Practical Knowledge in General and Anscombe’s Practical Knowledge It is perhaps a feeble defense of practical naturalism to show that Aristotle accepted it. I don’t think so myself. But our real difficulty is to develop an understanding of a certain kind of practical knowledge, which would have as its object something we might call the human, the specifically human. Or, equivalently, we must see how something with the character of a life form could be something in respect of which a practical knowledge is pos sible. We w ill see that Kant’s m istake, which everyone commits, is to misconstrue the concept of the h uman as an empirical concept. We can see that it is not an empirical concept, and that some knowledge into which it enters is not empirical, even in advance of seeing that it is also a practical concept, and that some of the knowledge into which it enters implicitly is practical knowledge. I w ill develop these thoughts in later sections. Here I want to make some preliminary remarks on the idea of practical knowledge. This is a very wide genus, I think. It includes the complex crafts or skills that Aristotle brings u nder the heading of technê; it also includes all the more elementary forms of knowledge of how to do something, for example, how to walk from one place to another, that were emphasized by Gilbert Ryle. It must also include the knowledge that comes with initiation into a practice more strictly construed. Moreover, I w ill suppose it includes in our case what Aristotle calls phronêsis, which is what we are attempting to characterize in a formal way. All of these forms of practical knowledge are in a sense general; the agent’s possession of them is what Aristotle calls a hexis, or Thomas a habitus; they are exhibited in action—that’s why they’re practical—but in indefinitely many individual actions. The cognition that is exhibited in action does not run out with any particular action it governs. By contrast, Elizabeth Anscombe, in her book Intention attempted to argue that we possess a special kind of knowledge, a knowledge “without
54
N ature and “ S ec o nd N ature ”
observation” of what we are d oing, when we are d oing it intentionally—and just as long as we are d oing it intentionally.6 It is exhibited in action, if you like, and is practical, but it is exhibited only in the one action upon which it bears. The idea that I have a special knowledge of “what I am doing” dropped out of the literature with a certain class of counterexamples. The decisive example, which destroyed the discussion, was of course due to Davidson. He argued that if I am trying to make seventeen copies of the document I am signing, by the use of interleaved carbon paper, it might be that I d on’t know w hether an impression is appearing on the seventeenth copy below. In that case I d on’t know whether I am making seventeen copies. But if in fact an impression is arising on the seventeenth copy, I am making seventeen copies, and we can say that I am making them “intentionally.” If I have a special nonobservational knowledge of anything, it’s not of what I’m doing intentionally, it’s of my intention or some other inner state. This example can be handled in a number of ways; some of what one says w ill depend on further details of the case. It seems to me that in the usual case, if one reflects on the a ctual practice of carbon copies that used to prevail, I do know all along that I’m making seventeen copies. That’s what the mortgage broker has told me to do. I start by making one g iant stack of interleaved documents and carbon paper. If I find out that I have not yet made an impression on the lower copies, I will start over again with those and write again, so that in the end I w ill have seventeen signed copies of the document in question. I had already begun making seventeen copies when I began writing on the w hole stack; I was still making my seventeen copies when I removed the unmarked lower pages, and continued with my making of seventeen copies when I finally did mark them. I knew all along that I was making seventeen copies, and Davidson has no counterexample. But there might be a case in which I w ill not have a chance to check and mark the lower copies. It’s not the usual mortgage broker business. Someone has me at gunpoint and says: all at once or you’re dead. I give it a try; and it works. I d idn’t know that I was making seventeen copies, but Davidson w ill say that I “did it intentionally.” This is fine. Davidson can say that. What we need to show has nothing to do with the adverb “intentionally” or any similar adverb that appears in Eng lish. We can invent another adverb, “anscombely,” so that we do things anscombely. In the first case I was making seventeen copies ans-
F o rms o f N ature
55
combely, and knew it all along; in the second case, at gunpoint, I wasn’t making seventeen copies anscombely, and I d idn’t know that I was making seventeen copies. Here it might seem that we are just introducing an adverb of our own. Anscombe’s thought, as I would understand it, is that there is a logically distinctive form of process with a subject, which is accompanied by a specific form of knowledge of it. Her thought, put another way, is that there is a special form of nonobservational knowledge, of self- knowledge, which is not of an inner state or merely psychical event, but of a genuine material process, as we might say. The material processes in question are indeed formally quite diff erent from others—we must make a division in logical space to represent them—but we do not leave the space of material processes and invent some entirely new category of “movements of the w ill” or something, if we want to supply this special self-k nowledge with a content. It might help orient the hearer if I say immediately that I w ill be attempting to make out a parallel between two relations. First, t here is the relation between my knowledge of my intentional action or my anscombely done d oings, on the one hand—my practical knowledge in Anscombe’s sense—and the action known, on the other hand, the one I am in the pro cess of performing intentionally or anscombely. Second, there is the relation between practical understanding or wisdom, and what it in some sense knows, or knows in a certain way, namely human life somehow taken generally, or, if you like, how to live it. In each case, there is a kind of knowledge of these things—what I’m d oing in the one case, and specifically human life on the other (the kind of life I’m living, so to speak)—that is empirical and external. But in each case there is another knowledge, which is, if you like “from the inside.” (This is of course an inept expression, which must be replaced in further reflection.) It is in e ither case in some sense productive of the thing known, and is thus practical. In other words, the naïve Aristotelian proposes that t here might be such a thing as knowledge of a kind of life, a kind of first nature, if you like, that is “from within,” an attainment that must of course be restricted to its particu lar bearers. It w ill inevitably belong to this kind of life to be governed by and prosecuted in the light of this kind of cognition, even if in particular cases the cognition is usually half-baked. This knowledge is quite diff erent in character from knowledge of a kind of life “from outside,” or empirically, such as I might have of Norway rat or sago palm or that
56
N ature and “ S ec o nd N ature ”
Martians might have of the human—and indeed that doctors, anatomists, and so on do have of the h uman. And this raises the worry voiced by Elizabeth Anscombe in connection with intentional action: If there are two knowledges, mustn’t there be two things known? The peculiarity of ethical knowledge, of phronêsis, inclines us to prise its object apart from anything of which anyone could have empirical knowledge. We want to make its object something like Dasein in general, or rationality in general, or the law of reason—and nothing that can be fitted into the same logical position as Norway rat or sago palm. This is like saying that the real content of the practical knowledge Anscombe was talking about is not a real material process.
6. Incredibly Brief, Schematic Characterization and Defense of Anscombe’s View of Knowledge of What One Is Doing Intentionally Let us briefly discuss Anscombe’s picture (as I understand it) and the place of practical knowledge in it. On this picture, intentional action description expresses a formally distinctive kind of thought. In it we predicate something of something, but in a particu lar way, which we can isolate schematically in the following tree (Figure 2), which may be taken as a division of forms of predication in thought and speech, or a division of forms of being in the sense of something’s being something. Notice the opposition X is Ving / X was Ving in the list of three possibilities given for the “grammar of events.” This certainly seems to be a s imple opposition of tense, like that of X is F and X was F. But in the pre sent case there is another possibility that obviously enters formally into the constitution of the opposed propositions with apparent tense grammar, viz. X V’d, which is quite diff erent. It would be a bit strange, but a philosopher might argue that the appearance of this third formal possibility renders the appearance of tense grammar in the opposition of X is Ving and X was Ving superficial and illusory. The correct position is surely that there is a common heading, opposition of tense, and then a division into, say, “process-a scriptions” and “state-a scriptions.” If you
57
F o rms o f N ature
predication in general x-F
temporal grammar
atemporal Fregean grammar
x was F x is F x wasn’t F x isn’t F
F(x) ~F(x)
event grammar (adding “perfective”)
mere state grammar
x was V-ing x is V-ing x V-ed
intention action grammar (adding concatenation by “reasons explanation” and the formal necessity of self-conscious or first-personal appearances)
mere event grammar
why? . . . I was doing A . . . why? why? . . . I am doing A . . . why? . . . I did A . . . why?
Figure 2
insist on the common heading and reject the division, you are surely utterly lost; a certain constriction w ill have wrongly befallen the concept of tense. We can imagine the same thought arising in connection with the grammar of the representation of self-k nowing intentional action as it is articulated in Anscombe’s Intention. Noticing its peculiarities, a philoso pher might insist that actions—what gets represented in the forms listed under that heading—a re not events. How could anything that is “happening” be something I was d oing? H ere again, a certain constriction has befallen the philosopher’s conception of an event; we ought rather to speak of a common heading, event or process descriptions, and a formal
58
N ature and “ S ec o nd N ature ”
division into descriptions of action and descriptions of “mere” events. Compare the elements in the sequence, beginning at the top: X-F X is F X is Ving I am d oing A As we move through this sequence, the network of opposing forms grows ever denser and more determinate. We have a clear example of being, in the sense of something’s being something, “sinking deeper into itself” as Hegel says. But granting that we have to do with new forms of predication or new forms of “being,” what is the connection of the “form of something’s being something” that we grasp through mastery of the forms found under the last leftward heading, to which Anscombe draws out attention, and knowledge? H ere we may note briefly the internal relation, employed by McDowell and o thers, between propositions of the form X knows that P and reasons explanations of the form X Ψ because P of X’s thoughts, actions, and so forth. It seems plain that you must grant that I know that it’s raining if you say that I’m doing something or think something “because it’s raining” in the “reasons” sense, or one of the “reasons” senses. If you thought I was wrong about it, you would explain the thought or action by reference to my belief that it is raining. That is, the fact that p, if it is known, is “such as” might be given in a true reasons account of my thoughts and operations. This rather formal thought has been supposed by some, for example, John Hyman, to make possible a reduction of knowledge that p to a certain sort of capacity, namely the capacity to think or act on account of the fact that p, in the reasons sense, and not simply the belief that p. As an elucidation, this seems hopeless apart from an account of “reasons explanation.” But it does seem that knowledge is always inter alia such a capacity. But, now, doing something “anscombely” is doing it (in my jargon) potentially-naïvely-rationalizingly.7 That is: whatever I am doing intentionally, at least in Anscombe’s sense, may be said to be “such as” might be given in a true reasons-account of other t hings I do, in particular the subordinate operations that are intuitively its phases. Thus, I w ill affirm the proposition that I am baking a cake in response to the successive
F o rms o f N ature
59
questions why I am measuring flour, why I am breaking eggs, and so forth. But—now applying the “rather formal thought”—if what I am doing when I am d oing something intentionally is something “such as” might be given in rationalization of the subordinate things I am doing—namely the very things in which the process resides at successive moments—then it is a condition of the existence of this process that I know I am engaged in it. Because the form of process we are characterizing under this last heading is precisely one that “rationalizes” its phases, binding them together in succession as phases of one process of a formally distinctive type, the type isolated by Anscombe, it is clear that the thing known only exists if the knowledge of it does. This is only a schematic argument for the conclusion that what is done intentionally is known by the agent, and evidently in quite another way than could be found in observational knowledge, for the simple reason that what is known depends on the knowledge for its actuality. Even if it w ere made less schematic, there would still be a question how far it elucidates the character of practical knowledge in Anscombe’s sense, rather than giving an indirect proof of one of her claims, via the “rather formal thought.”
7. Can Knowledge of a “Life Form” Be Practical, or: Can a Life Form Be Such as Might Be Prosecuted Characteristically in the Light of a Cognition of Itself? My effort w ill be to attempt to build up from below, as we might say, from the idea of life in general, to the idea of what we might call phronetic life, or potentially phronetic life. We are supposing that for certain things self- knowledge is possible in respect of some of the material processes of which the thing is subject, namely its intentional actions; such a “thing” is plainly formally very different from one that c an’t—for example, an individual sago. Its representation admits a formally distinctive possibility. We agreed also that the possibility of such a peculiar knowledge should not incline us to remove the t hing known from the category of a material process, but rather to develop a distinction within that category. So also a kind of life characterized by phronetic or ethical knowledge
60
N ature and “ S ec o nd N ature ”
ill no doubt be formally very different from one that is not—again, sago w palm, for example. (But also perhaps Martian; that Humeanism, or the no-w isdom view, is true of a sort of reasoning, thinking, “concept mongering” life, seems not to be a possibility that can be excluded formally.) But the peculiarities of ethical or phronetic knowledge should not incline us to remove its object from the category of a first nature, but rather to develop a distinction within the category of a first nature. If we can make sense of this idea, we w ill reach a form of naïve Aristotelianism that is not susceptible to McDowellian attack. 7.1. The Representation of Life in General An account of practical reason that wants to find practical significance in the concept human, considered as in one sense something to be put alongside the concepts sago palm and Norway rat and so forth, must find a genus to which t hese t hings all belong. It presupposes, in the first instance, a certain general conception of life and, if you like, of the grammar of the representat ion of life. The remarks that follow w ill be a bit dull for readers of some e arlier papers of mine, but as I develop this grammar, I w ill outline aspects of the material that w ill be employed by a naïve Aristotelian. We may begin, here as always, with the simple reflection that the represent ation of an individual living organism as living is everywhere mediated by an implicit represent ation of the species or life form under which the individual is thought to fall. Similarly, and more metaphysically, what states of affairs an individual organism may enter into depends on the life form it bears. Consider any concrete case in which you would say or think of an individual living being here and now that it is flying, or that it has leaves, or that it has eyes, or that it is blossoming, or that it is reproducing. It is clear that tokens of the same type of thing, eye or leaf or flight or whatever, might be constituted very differently in a form of life very different from the one you have i magined. Similarly, going upward, the same materials might constitute quite diff erent phenomena of life in sufficiently diff erent species. Consider, to take my favorite example, cell division or “mitosis,” a process described at length in elementary textbooks; its core is the pro cess of replication of genet ic material. Now, where we come upon this
F o rms o f N ature
61
process in bacteria and suchlike organisms—in which of course it first appeared, thus triggering the crush of terrestrial evolutionary history— in suchlike creatures, we w ill have come upon the process of reproduction, the coming-to-be of new bacteria from old, Aristotle’s genesis. But the same process is a part of growth and self-maintenance in a California condor or in a h uman being, Aristotle’s threpsis. Thus, what phenomena constitute reproduction, and what phenomena are constituted by mitosis, in turn depends on the life form in question. According to context, the same can constitute something diff erent, and something quite diff erent can constitute the same. Here I have been speaking metaphysically, so to speak, but it seems plain that something similar must be said at the level of thought: the repre sentation of given phenomena here and now as amounting to a process of reproduction, or as a phase of it, depends on a conception of how t hings stand with the individual’s so-called species or life form or “first nature.” This form-dependence or first-nature-dependence or species-dependence, or whatever you want to call it, extends, I think, to e very ordinary tensed description of an individual organism precisely as alive: for example, as eating or drinking or leafing out, or as having teeth or lungs or leaves, etc. The judger of particular vital phenomena relates them to the individual’s form. This relating of things to the form may of course be quite inarticulate. And I can have knowledge of vital facts about individuals though I have all sorts of wild ideas about the kind of life I’m dealing with, so long as they are irrelevant to the matter at hand. (Having introduced the idea of vital description, we may note something to be developed later: that everything that is given to a subject in what philosophers call self-k nowledge is to be formulated in a “vital description.” My height and weight are not m atters of self-k nowledge, but that I think, hunger, am walking across the street, and so forth, are.) Cognition of particular facts about an individual organism is in this way “mediated” in Hegel’s sense, and, as he points out, the phenomenon extends far beyond the representation of life. Its peculiar character in this case w ill arise from the peculiar character of the item to which we pass in making the mediate character of the cognition articulate. I have spoken of the life form or species or the “nature” in the sense of the “first” nature of a living t hing, having more or less picked t hese words up on the street. But in philosophy, I am thinking, we only r eally come to assign a clear content
62
N ature and “ S ec o nd N ature ”
to any such expression as “life form,” “species,” or “living first nature” by reflections like those I have just attempted. We can only become more articulate about what is contained in such expressions as “life form,” “species,” or “living first nature” if we consider the articulate representation of a particular such t hing. This sort of repre sentation is almost as quotidian as the representation of individual living beings as living. Where does the so-called life form itself become a topic? It is made articulate in a connected system of general judgments. True thoughts, which develop this system, we may call contributions to the “natural history” of the life form in question. The components of this system, the “natural historical judgments,” attribute to the life form, in a logically distinctive way, predicates, which can also intelligibly be attributed to individual organisms. “They have four legs,” we say of domestic cats or of domestic cat-form or of “the” domestic cat. “They bloom in spring,” we say of cherries or of cherry-form. “It has four legs,” we say of this cat hic et nunc; “It bloomed last spring” we say of the cherry tree in the garden. The properties expressed by these predicates may be said to “characterize” the life forms cat and cherry respectively. We may say, by contrast, that they “hold of” or “held of” the individual cat or cherry tree in question. The predicates may of course fail to hold of many individual bearers of the life forms they characterize. Just as there is an indefinite manifold of truths I may affirm about any definite cat h ere and now, t here is an indefinite manifold of natural historical remarks I might make about the kind of thing it is, or about its form, or about whatever it is that is expressed in the word “cat.” And just as there are certain relations of causal and other suchlike dependence among the things recorded in the propositions about this cat here, there will be a systematic relation among propositions characterizing cats period or “the cat.” The peculiarity of the characterizing propositions comes out especially in t hese relations and the way one goes about generating the system of them. We say how the four legs of the cat come to be—or how the four legs of cats come to be—and how they enter into the rest of cat life, and however much we elaborate on the m atter, we never leave this space of natural historical generality the theme of which might be written at the top of the page: CAT.
F o rms o f N ature
63
I do not want to labor them here, but some of the peculiarities of such a system of judgment are, I think, aptly brought out if we consider two points: first, though natural histories rise and fall in the course of evolutionary history, they are given to us in the first instance as propositions in the simple present: cats have four legs, cherries bloom in spring. But the instances of these “present tense” propositions can be present or past, and can extend as far as the case w ill allow into the past. I can say “Cats have four legs, for sure they do—for example, this cat, Charley, had four legs,” pointing to a photograph. H ere I am, of course, presupposing that I might have pointed to the same cat back then and said “Cats have four legs, for example, this one does,” where the opening subsentence “Cats have four legs” expressed exactly the thought I am expressing with it now. This is far from a definitive peculiarity of this form of judgment, but it should be noticed that it cannot be represented in the usual textbook forms of proposition. We get a little further if we see the system of such propositions as having some capacity to explain particular facts that line up with them. I constantly appeal to such generalities to explain what is g oing on with individuals. It is in the nature of the matter that I don’t need to refer to antecedent particular facts to explain the particular facts in question, at least in suitable cases. Knowledge of the universal aspect of the material suffices for the understanding of these suitable par ticular facts. The knowledge that is given in present tense generalities can also be employed in the comprehension of past facts about individuals. The temporal aspect of this sort of explanation exhibits some of the peculiarities we find in practical explanation, for example, that I can oing B; h explain why I did A by saying that I am d ere the explanandum is past. For example, I might say “I turned left there b ecause I am driving to Potsdam.” This of course presupposes that the explanatory proposition “I’m driving to Potsdam” could also have been asserted in the past, when I was making the left turn. But what that proposition expressed in the past is not diff erent from what it expresses now, I think, if I am still driving to Potsdam as I am saying I am. Of course, the explanatory character of natural historical generalities is not without parallels all over logical space. But it is of decisive importance in seeing how there can be such a thing as Aristotelian ethical
64
N ature and “ S ec o nd N ature ”
knowledge, though I w ill not be able to develop the point much h ere. In naïve Aristotelianism, as I would propound it, the explanatory relation that any life form bears to the individual organisms that come under it takes on a logically distinctive and more determinate form in certain sorts of “noninstrumental” explanation of action. B ecause the form of explanation is a form of “reasons explanation” the explanatory term must be represented as known. A more striking feature of this form of representation, from the point of view of a program of isolating it from all other forms, is the complete indifference of any particular characterizing proposition to any particular rate of instantiation or exemplification. Thus it belongs to mosquito-life, in any of its particular forms, that the egg develops by stages into an adult with wings. On the other hand, it belongs to mosquito-life that the female adult lays hundreds of eggs—a fter a suitable blood meal. We affirm t hese things though we know that the number of mosquitoes—the measure of its extension—has remained more or less the same for summers on end. The propositions about the developmental phases are affirmed without conditions qualifications like “unless a fish eats it” and so forth. It is clear that a natural history like this is formally consistent with any attribution of egg-number, no m atter how high. A natural history is thus compatible with any rate of instantiation in past and present cases. This brings out well the point that in such general predication we are making a contribution to a possible systematic totality of propositions: it is only b ecause many such propositions are realized in the individual case that others can fail to be realized. As Saint Thomas says, much a ctual goodness is a condition of any badness. And indeed we may say that where the characterizing predicates do fail to hold—where an individual cat has three legs or a cherry does not bloom—we have natural defect, a failure of elementary “natural goodness.” Thus judgments of goodness and defect make implicit reference to the species or life form, or whatever you might call it, that the individual bears. As Aristotle would say, t here i sn’t any one health, it i sn’t like straight or white, but diff erent healths according to what kind of t hing y ou’re talking about, t here is no medical art for everything, as he puts it in the passage discussed above. Such “evaluative” judgments are no more suspicious, metaphysically, than are the “positive” descriptions: the materials on which we base the evaluative judgments are implicitly present in positive
F o rms o f N ature
65
description. Though I w ill not be able to develop the point much here, it will be plain that the concept of the practical good on a practical naturalist account w ill be a specific more determinate form of this more abstract conception of good; it w ill be a form, as Foot says, of natural goodness. We may note that it is precisely in the cases where t here is no defect, deformity, lack, need, disaster, etc., that we are able to appeal to the system of general propositions in question to explain particular facts. Where there is defect, deformity, lack, need and disaster, I must explain the sorry phenomena here and now by reference to c auses there and then. Even though, as we say, it is statistically rare for a mosquito egg to grow into a flying creature, still, given a particular case in which it does, I explain each stage in the individual’s development by reference to the true doctrine of mosquito life. It is the particular failures that must be explained by such facts as that a fish ate the larva, that some bacterium overtook it, that there was a m istake in the replication of some crucial gene, and so forth. (Thus, on the naïve Aristotelian account, the concept of action on reasons, and that of action explicable in a certain way through the rational life form, and action founded on practical knowledge or understanding of that form w ill come together as more determinate forms of what we find elsewhere.) What is important here is the logical form of the universal in question, if we can speak of the “life form” as a universal. Certainly in e very case we will actually credit, the life form is associated with a universal. A universal, we might say, is a form of unity of the diverse, not as parts of a w hole. Parts are l imited by the w hole they go to constitute. But a universal can have in definitely many instances; it d oesn’t care how many t here are. Just as Frege says that the part-whole relation has numerous formally different subordinates—one form is at issue in Frege’s claim that thoughts are parts of other thoughts, and another in Caesar’s division of the parts of Gaul— so it seems to me that the unity of things as coming under a universal has diverse formally distinct subordinates, or equivalently, that the relation between an individual and some universal can take diverse forms, as Hegel, not Frege, is forever saying. The unity of several things as coming u nder a common life form is one quite special example. In a natural history, a certain sort of unity to which things are brought is made an object of in dependent reference and discussion. In it we operate with the idea of something that moves along through a tide of individuals and individual events and which are united under its heading. Of course, in the end it’s
66
N ature and “ S ec o nd N ature ”
g oing to crash and burn, and only ever have had a finite, l imited number of bearers, but as far as it’s concerned it might go on forever. But natural historical judgment isn’t the main or only way we have access to this sort of “unity.” I am thinking that the intellect operates with this specific form of unity and universality implicitly in all sorts of thought of individual t hings as alive and in a double way in certain sorts of judgment of defect and soundness. My thinking might thus relate to any such unity or universal in diverse ways. Naïve Aristotelianism, expressed in yet another way, is the view that t here is a certain such unity, a unity of this general logical shape, which is the one to which all our practical concepts and our practical philosophy are attached—namely the one expressed by the word “human.” Not the unity expressed by the words “rational being.” 7.2. Non-Practical Apprehension of One’s Own Form For the moment we need not bring human being and the human into the account. It is a formal feature of our i magined practical naturalism that it claims that the following idea is coherent: a natural life form—an item falling into the same logical category as domestic cat, cherry, California condor—might be characterized by reason, self-consciousness, and self- knowledge as a power or capacity—and by what might be called practical reason in particular (I w ill delay discussion of the practical). That is, self-conscious self-k nowing intellectuality might be just as much a “characteristic” of certain (possible) life forms or species as vision or the capacity to feel pleasure or pain is. In order for such a claim to hold of a given life form, it is not necessary (to repeat) that each single one of earers develop the power for intellectual representation. No more than its b it need be that every bearer of a sighted species develops the sight that is characteristic of what it is. Some individual cats, for example, are born blind. This aspect of Aristotelianism may of course seem to be a platitude. Why shouldn’t a self-conscious self-k nowing form of life be possible? It may, a fter all, also seem to be a platitude that the specifically human form in particular—the particular terrestrial life form that you and I together bear—is in this sense a practically reasoning one. But surprisingly many philosophers are explicitly or implicitly committed to accounts of intellect
F o rms o f N ature
67
and reason generally which entail that the self-same animal species might in one epoch, A, be devoid of concepts and reasoning, and then in a later epoch, B, develop these powers or “practices,” perhaps again losing them in a still later historical period C. Self-consciousness and intellectual representation, w hether theoretical or practical, are, on such views, to be compared with money and banking, or with a certain style of clothing. They do not characterize the life form as such; they do not attach to anything with the specific kind of generality I labored above; mention of these is formally prohibited from entering into a natural history. On such an account, it would be an open question whether the “intellectual interlude,” epoch B, in such an imaginary history should be viewed as a period of sickness and psychic deformation—as I suppose it is a sort of forcible deformation of chimpanzee-life, akin to the binding of Chinese women’s feet, to remove them from their characteristic environments, hold them in zoos, and “teach” them rudiments of American Sign Language. John Haugeland says in the introduction to his collected papers that the object of those essays, Dasein, is a few thousand years old; of course he knows that anthrōpos has been around for maybe a hundred and fifty thousand (here I think he is crossing his master Heidegger, who speaks of “primitive Dasein”); if this is right, shouldn’t the critical question be posed, whether one makes one’s d aughter sick in Daseining her?8 A brief study of the more alarming sections of Sein und Zeit would certainly suggest that one is! I do not have space to develop the point, and w ill assume that this is an analytical falsehood. The phenomenon of, say, reflectively bringing a perceptually given object under a concept in judgment is only possible in individual cases where the possibility characterizes the form of the animal, the perceiver, in question. To say that it has “broken out” in some individual or individuals is always properly speaking wrong, I think; we ought to say that a new life form has “broken out” and that these individuals simply bear it. It could no more break out in a lone individual or mob of individuals, as a peculiarity of theirs, than could eyes and visual perception in an individual earthworm. If I insist that the thing has eyes and sight, I must relate phenomena to a diff erent natural history. I must say that here we have a new form of earthworm, just now come into being. How I suppose this is possible is another matter. If I step on it, then I must also grant that this new form w ill only ever have had one bearer.
68
N ature and “ S ec o nd N ature ”
I assume that when McDowell invites us to imagine a rational wolf in “Two Sorts of Naturalism” he is not rejecting this point. (I w ill not yet speak to the practical aspect of this thought experiment.) He rather fastidiously begins with the supposition that “some wolves have become rational,” that is, that there are other suchlike wolves outside and alongside any one such wolf. This is presumably for what are vulgarly called Wittgensteinian reasons. 7.3. Excursus on the Form of Certain Claims of Wittgenstein Before returning to this rational wolf, I note in passing that Wittgenstein’s objections to a so-called private language run in a strict parallel with our present conception, which I might have called a private life argument if I were offering a detailed argument. Just as t here is no intrinsically private meaning, so t here are no intrinsically private leaves or processes of nutrition or states of hunger; if there can be one leaf, there can be another. Of course the forms of generality or potential generality or “mediation” are logically quite diff erent in the cases of language and life form. The crucial point is again that in reckoning that t hese phenomena are linguistic, I bring them to a formally distinctive sort of generic unity or potentially generic unity, the one Wittgenstein sometimes names with such words as “practice” and one use of “form of life,” or that McDowell points to, perhaps a bit indirectly, under the title of second nature. It too can be made a topic of some independent discussion of a general and quasi atemporal sort. There is a specific sort of doubling up in the representation of an individual as engaged in a linguistic operation, for example, as making a note of something; the other t hing I represent besides the individual, viz. their so-called language—the unity to which I bring them or their deeds— might be found in another. The representation cuts a space for that even if nothing fills it. T hese points, properly developed, are logical. That t here should actually be or have been others if such a unity is to find a place in nature as we know it—that is, if there is to be any unity to which I might bring individual phenomena—this is a separate point of a quite different character, though it is in a way more obvious than the deeper logical point on which it depends. Wittgenstein is not always clear, I think, on this distinction.
F o rms o f N ature
69
Our thoughts about life, applied to the case at hand, would entail that in the representation of someone as engaged in a linguistic operation there is not just a doubling but a tripling in the representation: I bring the note taker to the unity of a language (or practice more generally) and I bring them at the same time to the unity of a life form which might extend further than the former unity. That t here must actually be or have been others to bring to this second general or potentially general unity is again a separate and radically different sort of claim, which again is more obvious than the inevitably esoteric point about forms of unity or universality. ✣ ✣ ✣
But to return to McDowell’s rational reflective self-conscious self-knowing wolf. The position I am considering is that in imagining such a thing we are imagining a new form of wolf life, a new subject for natural historical predication, Lupus sapiens, so to say, to put alongside the Barracuda sapiens of Aristotle and the Martians of popular imagination. That t here should be a rational form of wolf-life is only slightly stranger than that t here should be a rational form of primate life. Now, one feature of the passage that puzzles me is McDowell’s distracting use of the word “wolf” in connection with the thought experiment. He speaks of “standing back from wolfish impulses” and of “what naturally matters to wolves, such as having plenty of meat to eat,” and of “what wolves need,” where the example mostly focused on is, for example, to hunt in packs.9 But what we are imagining is that wolf, lupus, has become a taxonomic genus that includes both rational and brute forms of life as subordinate. It has become like “primate” or “mammal” with the appearance of human beings. It can hardly be supposed to express the kind of unity that we have made our theme; “primate” is not the name of a first nature, no more than Indo-European is the name of any language we are speaking, though we are all speaking Indo-European languages and though there was a language, proto-Indo-European, that some people spoke—as once there was the ur-primate kind of life. We can say things like “primates have hair,” but it seems that this means that every kind of primate has hair, or is characterized by the presence of hair (even if some
70
N ature and “ S ec o nd N ature ”
individuals are hairless). It is not inferentially impotent in the way that a genuine natural historical judgment is; it is refuted by any case of a hairless kind of primate. Perhaps t here is another way of construing predications about primates, but I don’t think that any naïve Aristotelian has been so naïve as to think that any such predication should be mentioned in a system of practical philosophy. Thus bringing the impulses of Proto-lupus sapiens, that is, actually existing wolves, into the m atter is a complete red herring; whatever truth there may be about the reasons of a bearer of Lupus sapiens ought to be consistent with the extinction of the various forms of Lupus non-sapiens, or indeed with a creationist account. Now, we are considering the possibility of an individual self-conscious, self-k nowing animal, in the presence of the premise that this only is pos sible where it is characteristic of the form the individual b earers and the premise that vital predication about individuals relates them to a form. The question is what this tells us about the kind of life in question. Remember now, what was noted above, that all the predicable items anyone has ever imagined as attachable to oneself in self-knowledge are themselves vital predications: I’m in pain, I like this stuff, I think it’s going to rain, I plan on returning to the United States one day. Pleasure and pain, thought and intention are all phenomena of life, and animal life. One’s weight and height are not among the objects of possible self-k nowledge. On our premises, then, the self-k nower in apprehending such things is always and everywhere relating themself to their form. In representing any animal as thinking or as in pain, I bring it to a certain formally distinctive unity; in representing it as bearing self-k nowledge in respect of these things, I represent the animal as bringing itself to a unity of the same type. Self-consciousness is always implicitly form-consciousness. And on our present premises, this w ill have to be a feature of the life form itself: it belongs to the prosecution of such life to see t hings in the light of it, as we might say. Its representation is a part of it. And even this is to say too little: It might be characteristic of some form of rational life that its bearers come into relation to some other sort of life upon which their sort of life depends; if, for example, it belongs to our rational wolves that they can digest nothing but buffalo meat or something, and thus that they can only live by laying hold of buffalo and so on, then since such operations w ill involve the exercise of thought, we might say that it belongs to such a kind of animal to operate with the concept buffalo. Now the proposition I was
F o rms o f N ature
71
considering, which might be expressed in such terms as “It’s a characteristic feature of the form that the bearer comes into intellectual contact with this form itself,” might be validated if that sort of relation, the wolf- buffalo relation, w ere merely rendered reflexive. It would be like saying that every rational being must, in the course of its thoughts, sometimes get around to noticing itself and putting a name to itself. But if that were the w hole account of the origin of self-consciousness, the name presumably wouldn’t be the first person and its uses w ouldn’t be in self- knowledge—and maybe it wouldn’t have been a rational thinker we were thinking of to begin with. It is, I think, the same with the “first life form” concept, as we might call it, the concept implicit in self-conscious self-attribution of vital descriptions, the unity to which the thinker implicitly brings themself in such thought. It cannot be subtilized out of the materials of experience. It must precede the individual as characteristic of their life form. Of course, if the animal in question gets a l ittle articulate on this point, it can maintain all sorts of strange thoughts about this so-called form, the second ele ment in the doubling up—for example, that it never had any other bearer. It can suppose, for example, that the unity to which it brings its thoughts in self-consciousness does not extend to all the other intellectual animals it thinks it is perceiving—a ll of which are in fact “of the same species,” as it puts it—arguing that the unity is merely biological and classificatory. This is no different from the fact that, having achieved some level of articulateness about the peculiarities of the first person it deploys in its acts of self-k nowledge, the animal can easily work itself into the idea that maybe “this I,” the one it registers as thinking, d oesn’t have a body, or isn’t a body, but only associated with one causally, and so on. 7.4. The Truth of Certain Claims of the Young Marx In the Paris Manuscripts, Marx suggests that “species being,” or equivalently “species consciousness” (Gattungswesen or Gattungsbewusstsein)—t he registering of the (first natural) universal one comes under—is a condition of all universal representation; it is a condition of having concepts.10 “Human” is, for each of us, the original universal. Allen Wood, in his book on Marx, declares this proposition to be transparently false: it is obvious, Wood says, that we only have the capacity to represent the human in
72
N ature and “ S ec o nd N ature ”
general because we are first capable of universal concepts, like those of a star or a sago palm, for example.11 Given this power, we can form the concept of the human as well. I don’t think Wood would treat Kant’s claim that self-consciousness and apperception and the representation “I” are a condition of all conceptual representation with the same sneering disregard. Nor would he so treat the view that the categories are determinations of the I think, and all determinate concepts determinations of these. He wouldn’t say that Kant is an idiot, that it is obvious that it’s only because I can conceptually represent t hings in general, a star in the heavens for example, that I can arrive at the point of pointing a finger to myself in thought too. But Marx’s claim is an immediate consequence of Kant’s, in the presence of our Aristotelian premises. (These are premises, which Marx accepted in a different vocabulary from his graduate student days to his grave.) Wood can only see the concept human as an empirical concept gained by sifting the materials of external experience. Wood is thinking that the item in question is something one can only grasp from outside. It is an alien object. Free self-conscious thought, he thinks, must transcend any such t hing, it must position itself outside any such thing. And so on. Marx would, I think, be happy to paraphrase McDowell and say that this is a peculiarly modern intellectual disturbance of which Aristotle, his master, was quite innocent. There must be a way of possessing the concept human, of accessing this unity, that is not a sifting of the objects of external experience. There must be a way of grasping what I am calling one’s form that is not an interpretation of outer experience, but that shows itself in acts of self-predication. I believe that words like “human,” “anthrōpos,” “Homo,” “Mensch,” and so forth all express precisely this grasp, however idiotic the ones who come u nder the word may have been in associating it with an extension. Note that Marx’s use of the concept Gattungswesen has a peculiar abstractness and a certain ineptness. There are a lot of Wesen out there, a lot of essences or kinds of being; some of these are life forms or Gattungen. Any Gattung is a Wesen. A tarantula has a Gattung, which is also a Wesen, but not Gattungswesen. The only Gattungswesen around for the bearing is the one we call “human.” But of course there could be others, Aristotle’s Barracuda sapiens or McDowell’s Lupus sapiens, for example. Gattungswesen expresses a formal characteristic of certain possible Gattungen
F o rms o f N ature
73
or life forms or the sorts of “unity” that might get expressed in a bit of natural history. But though this formal category might be instanced in any number of concrete Gattungen—we now have Homo sapiens, Barracuda sapiens, Lupus sapiens, and the damned Martians—nevertheless the intellectual operation through which the individuals reach their respective so-called Gattungen is the same in all the self-conscious acts of any bearer of any of them—just as the first person, as an intellectual operation, is the same in all of them. The b earers of the diff erent kinds cotton onto different life forms through the first-life-form operation, as I might put it, slightly idiotically, just as the different individual rational animals latch onto different individual rational animals through the first- person concept. The intellectual operation is perfectly pure in either case. As I said above, Kant’s failure to grasp this fact is the source of his deafness to naturalism in ethical theory. He constantly complains against the tradition that relates ethical discourse to specifically human life, complaining that it brings an empirical concept into the basis of ethics. But this argument is simply invalid. I am suggesting that t here are certain logical peculiarities distinctive of the representation of any “rational form” or Gattungswesen: Does this mean we are no longer talking about a life form in the sense I have been using, no longer bringing t hings to that kind of unity? Shouldn’t we drop it and move to ideas of second nature or Geist or maybe Dasein and get on with things? No doubt we w ill have to get on to these things, or some of them. I suppose I am preparing spadework for that. But this does not mean dropping the idea of a rational life form or Gattungswesen. Marx saw quite clearly I think that the failure of Kant and even Hegel to make use of this idea was due to a “constriction the idea of a nature had undergone in their thinking,” and that it deprived them of access to the naturalism propounded in Aristotle’s practical philosophy.12 We noted that my peculiar “access” to myself in self-knowledge might lead me to worry how this “I” can be related to a body, how anybody could have this peculiar access to itself. Philosophers mock this sort of talk. They propose to be cool with their embodiment, and, very unkindly, to be unforgiving of any sign of Cartesian difficulties in a fellow philosopher. But I think the resistance to certain uses of the concept human—a concept to which we append various predicates as characteristic, for example that they have two feet, that t hese chemical processes take place in their cells—is
74
N ature and “ S ec o nd N ature ”
just another form of the Cartesian difficulty, and blocks the way to any genuine resolution of them. The idea that the subject of such multiply instantiable propositions might at the same time provide, let’s say, the unity of thought and the highest concept of ethical theory, and so forth, seems to them a barbaric false naturalism. It strains intelligence that what enters into such propositions—things of the logical shape of mountain goat, spirochete, and tobacco mosaic virus—could in any case at the same time appear to me a priori, so to say, in every thought. There must be a change of grammar or logical form when we speak of a unity tokened in diverse cases of self-conscious thought and in the light of which they are self- conscious thought. Something new has sprung out, and that must change everything. Maybe we have moved outside the categories of nature to something outside them, or something at their basis. But I see no reason not to say what we said in the case of the forms of predication of individual objects. Action is not something different from happening; it is a logically distinctive form of happening; there is a common genus and an operation. The opposition of the progressive judgments X is Ving and X was Ving is not something diff erent from an opposition of tense. It is just that oppositions of tense come in diverse types. Similarly the life forms that come to their bearers “from within,” through an a priori conception, are not something other than life forms; it is just that there is a formal distinction within that category. 7.5. Forms of Practical Life I have been laboring the self-conscious self-k nowing character of the animals we have to do with. We must now turn to the practical, though I confess it w ill be with absurd brevity. (I) Let us first ask: Could we have self-conscious self-k nowing conceptual representors without action? Here we have a contest of authorities, McDowell vs. Grundlegung I. Kant imagines certain “favored creatures” whose reason has not “broken out into a practical employment,” creatures who think and judge about things, but operate entirely on instinct. McDowell declares this to be absurd. I think I would side with McDowell. atter? I feel lost in t hese analytic But what do I r eally know about the m arguments, t hese thought experiments. The human form, my form, what
F o rms o f N ature
75
I am, the unity to which I bring my thoughts, etc., is evidently also exhibited in the specific unity, outlined by Anscombe, of concept-of-itself governed process, that is, of intentional action; coming into possession of this form is coming into possession of a w ill, if all goes well. I know that it is possible by sharing in it, that is, as a sort of factum of reason; I have no similar way of knowing that the opposed idea is impossible. Put another way: the progress of this life—whatever e lse might belong to it, for example, the beating of the heart—is in part through processes which are exercises of concepts of themselves, for example, driving to Potsdam. What shows that they are exercises of concepts is, for one thing, that the agent reasons how to realize them in the course of their realizing them, or might. The agent applies thoughts relating possibilities at hand to what they’re up to, which they must thus conceive—that is what their “being up to” those things consists in; that’s the kind of process or something’s-being-up that we have to do with here, and in Anscombe. It is thus that the agent can have self-knowledge in respect of some of what’s up, of what’s going on, of progressive truth. W hether any material to be caught in these categories is found in our i magined barracuda sapiens and lupus sapiens, or w hether instead they are Kant’s “favored creatures,” w ill be a m atter for enquiry. I admit I d on’t know how the enquiry would proceed if this possibility is granted. But why not just grant that it’s pos sible as far as anyone knows and say that you know it doesn’t stand so in your own case, and thus in the case of the kind of thing you are, viz. a human being. You don’t know this empirically, but rather as a Faktum of reason in self-conscious exercise of concepts, in realizing them—that is, in action. It is not a cognition you gain “from outside.” Even where I do gain knowledge of this sort from outside, observing rational fish or whatever, what I am imputing to the fish presupposes the concomitant imputation of a (possibly inarticulate pre-philosophical) knowledge that is not like my own, not from outside. So it is with the knowledge by observation of another’s particular intentional action as such: in it I commit myself to the actuality and possibility of a knowledge of the same matter of particular fact, the same “happening,” that is formally quite diff erent from my own knowledge of it. I empirically know the event precisely as not empirically known by its agent. (II) Enough of the “favored creature.” In what we may call a Humean being, a bearer of a Humean Gattungswesen, there will be thought and also
76
N ature and “ S ec o nd N ature ”
a genuine dependence of action on thought. But this dependence of action on thought always presupposes a more inclusive process or intention that might itself depend on thought in the same way; in the end, some actions w ill depend not on thought at all but on appetite. (That this possibility should be available r eally is, I think, an analytic claim about the idea of intentional action or concept-of-itself-governed process; this is I think shown by Anscombe; apart from the bringing of actions by thought to more inclusive actions or operations, we would have no reason to think of the processes as realizing concepts.) Again McDowell comes out against it implicitly in many places, including the discussion of the rational wolves. (Here he may seem to have Kant on his side. But McDowell appears to be giving analytic arguments that it must be so wherever reason breaks out into a practical employment. It is clear that Kant does not think there is any analytic absurdity in the idea, as the familiar footnote in the Religion and the w hole idea of a ‘fact of reason’ makes plain.)13 The reader is moved by the rhetorical description of the rational wolves as “standing back” from their wolfish impulses, insofar as they are rational. Such language is also found in lyrical passages of Korsgaard.14 But in a sense a merely Humean sort of agent might “stand back” from an objective, when they realize that it d oesn’t actually fit with the wider pro cess they’re in the course of realizing, or not as well as something else does. Even a so-called final end or objective one, which can only be explained by reference to appetite or pleasure, might be put to rest in the light of some new desire or prospective pleasure. I think this is not the sort of critical distance McDowell is imagining. That sort of critical distance presupposes the possibility of action depending on thought-about-it, action that is not governed by a wider objective or action that is u nder way. But, again, why bother lofting analytic arguments that it must be so wherever there is self- consciousness and reason and judgment? We are anxious to oppose Humeanism in ethical theory, but setting out to disprove it analytically of whatever applies thought to action seems a hopeless plan of attack. (III) And again, what is true is that I am capable of such a t hing. For example, I can do something b ecause I promised someone I’d do it, or think I did. Occasionally I even manage it. Here the action depends on thought—it is action upon a consideration or “reason”—but the thought does not link the action to a more inclusive action or objective or intention. It simply characterizes it as promised to someone, and thus as belonging
F o rms o f N ature
77
to justice. It is a condition of my life’s admitting this form of explanation that I am not a merely Humean being and the b earer of a merely Humean life form. Another possibility has been provided for. That this should be a possibility within some reasoning kind of life might be a possibility we can provide for in the abstract, so that maybe even Humean agents on Mars could grasp the possibility, if they are possible, and maybe apply it empirically to h uman beings. But it can only be realized at all if its bearers can grasp it other than empirically. And it can, I think, only be realized if the bearers can grasp it otherw ise than by analysis of the concept rational being. The Humean hypothesis about the character of the natural form I bear falls to a plain fact of reason. Again, whence the craving to show that it is characteristic of any Gattungswesen, or of any well-constituted individual “rational being,” that is capable of thought-dependent final objectives? The desire to make this analytic of rational agency turns on the incapacity to grasp the idea of a fact of reason—and the incapacity to recognize the possibility of a knowledge of determinations of the life form one bears, and in particular of its basic powers, from within, by means of such a “fact.” The opposing “logical Humean” position, as we might call it—that no operation of practical reason can ever do anything but fit action to objectives that are the contents of appetite—is just as absurdly extravagant as the opposed “logical Kantian” claim. One should make do with the thought that it might be so for some forms of practical reason, that is, in some kinds of rational life. Kant thinks that in moral knowledge I possess a fact, which shows something about the character of practical reason wherever it might arise. The idea of a Humean being is not inconsistent, for Kant; rather it is inconsistent with what I know about myself, my subjection to concepts of practical good and duty; and that this must be something I know pertains to practical reason period. Practical reason he then construes, inexplicably, as a single power potentially realized in individual b earers of diverse life forms. In my practical self-k nowledge I thus know something about rational fish, wolves, and Martians; and in the end I w ill clobber them all with the categorical imperative. In envisaging the possibility of action founded on thoughts that do not simply link the action to wider objectives, we have, I think, arrived at the concept of a phronetic kind of life. Such a kind of life is not merely mediated by a (perhaps inarticulate) representation of itself, as I have argued that
78
N ature and “ S ec o nd N ature ”
any self-conscious form of life must be. Rather it resides principally in the realization of itself as thus represented. Just as individual cases of d oing A intentionally reside in realizations of the concept doing A—a phenomenon we found even in Humean kinds of agents. Actions are not just cases of things falling u nder that concept. The agents in question can be viewed as judging these final objectives simply to be good; and the grounds upon which they estimate t hese actions and other objectives are “reasons” in the certain robust sense characteristic of contemporary philosophy. The reason they give for doing the thing is at the same time given as a reason for thinking it good. The good they register is evidently not the concept of the useful that is deployed by the Humean agent, but (I would argue) a more determinate form, a sinking deeper into itself, of “natural good”—as intentional action is a more determinate form of process. That they can operate with such a concept and register truth in it is again a fact of reason for which they can, in the nature of the case, find no external ground. (That we must not seek such an “external grounding” is precisely something that must be developed from the nature of the particu lar case; it must be shown to be like looking for an external grounding for the judgment that one is in violent pain, or the judgment that one is.) We might further suppose, I think following Kant structurally at least, that even the abstract possibility of such a thing, that a form of life could have this character at all, can only be “given” to them in this way. On the other hand, the grounds upon which they register things as good in this sense, and upon which they select, or might select, final objectives—they have no reason to expect t hese to operate in other forms of reasoning life, or even in other forms of phronetic life. The idea of good with which they operate is restricted to such as they are, to a kind of life, which is governed by reckoning with this kind of good and producing it. But it does extend across the bearers of that life: in knowing something about myself through phronêsis or my pitiful approximation to it—for example, that it would be good to look a fter my health a bit more—I must also know something that can be known and properly should be known by any who bears my form, though it might not be just that same thing, and might not be too easy to articulate. In bringing my actions to the standard of justice, for example, I know that it belongs to the human to do some such thing—unless, of course, I am in a state of illusion, due, for example, to
F o rms o f N ature
79
fraudulent upbringing, as Callicles argued I would have to be. But we have no reason to suppose this; it would have to be a kind of dogmatism, like the insistence that the Humean model fit every case of operation on the basis of thought. In forming ideas of phronêsis and practical good, we say nothing about other forms of rational life and nothing about reality as it is apart from such as we are, nor do we suppose our eyes are opened to features of such a thing.
✣
Pa rt I I
✣
R E A SON IN PERCEP TION A N D ACTION
Chapter Three ✣
The Rational Role of Perceptual Content ✣
Matthew Boyle Truth and semblance [Schein] are not in the object insofar as it is intuited, but in the judgment about it insofar as it is thought. Thus it is indeed correct to say that the senses do not err; yet not b ecause they always judge correctly, but because they do not judge at all. Hence truth, as much as error, and thus also semblance as what leads to the latter, are to be found only in judgments, i.e., only in the relation of the object to our understanding. —Kant, Critique of Pure Reason, A294 / B350
I
n Mind and World, John McDowell offered an influential defense of the thesis that the content of perceptual experience must be conceptual.1 He did not defend, but rather took for granted, that perceptual experience should be characterized as having some sort of content or other. McDowell was not alone in taking this for granted: the idea that perception presents us with specific contents—that it informs us about the world by representing aspects of the world as having certain definite features—was the orthodox position in late twentieth-century philosophy of perception. In the last two decades, however, a number of philosophers have questioned
83
84
R eas o n in P ercepti o n and A cti o n
this orthodoxy. Their criticisms vary, but a core idea for most of t hese critics is that perception itself does not take a stand on how things are: it does not represent perceived objects as being any particu lar way whatsoever. Rather, perception simply relates us to perceivable things in such a way that we, the subjects who judge, can justifiably represent them as having certain features. One prominent critic, Charles Travis, puts the point as follows: “Perception, as such, simply places our surroundings in here is no commitment to their view; it affords us awareness of them. T 2 being one way or another.” When philosophers describe perception as presenting us with contents, they are, according to such critics, mistakenly retrojecting a quasi-judgmental structure into the activity of perception itself.3 This challenge to the idea of perceptual content should be of interest to philosophers whether or not they are attracted by McDowell’s claim that perception has specifically conceptual content, but it holds a special interest for anyone who thinks McDowell was right about something. McDowell sought to establish a point about the conditions u nder which perception can give us reasons for judgment: it can do so, he argued, only if the conceptual capacities that enable us to think about reasons are already at work in our perceiving itself. But it is hard to see how this claim could be explained without employing the notion of perceptual content. Consider McDowell’s original, Kant-inspired way of explaining his thesis: When we enjoy experience conceptual capacities are drawn on in receptivity, not exercised on some supposedly prior deliverances of receptivity . . . In experience one finds oneself saddled with content . . . The content is not something one has put together oneself, as when one decides what to say about something . . . [Nevertheless, we] would not be able to suppose that the capacities ere manithat are in play in experience are conceptual if they w fested only in experience, only in operations of receptivity. They would not be recognizable as conceptual capacities at all unless they could also be exercised in active thinking . . . Minimally, it must be possible to decide whether or not to judge that things are as one’s experience represents them to be.4
T he R ati o nal R o le o f P erceptual C o ntent
85
The core idea h ere is what we can call McDowell’s Requirement: (MR) The same conceptual capacities drawn on in making judgments must also be drawn on in our perceiving the things concerning which we judge. But what does this mean? In clarifying (MR), it is natural to appeal, as McDowell does in the passage above, to the idea of perceptual content. Using that idea, we can express the significance of (MR) by saying that it is not sufficient that perception should merely present us with contents that we can affirm by exercising our conceptual capacities, say, by making judgments of the form (1) This F is G. (MR) demands in addition that our ability to be perceptually presented with such contents should itself require our having appropriately related capacities for conceptual thought. So the abstract requirement imposed by (MR) is specified as something like (MRC) A subject can have a perceptual experience with a certain content (for instance, This F being G) only in virtue of having appropriately related conceptual capacities (minimally, in this case, capacities to think of t hings as an F and as G).5 Suppose, however, that the skeptics about perceptual content are right: then we cannot clarify the significance of (MR) by stating some requirement on the relationship between the content of perception and the content of possible judgments. In that case, the demand expressed by (MR) becomes obscure: What are “conceptual capacities,” and what does it mean to say that t hese capacities must be “drawn on” in perception itself? Moreover, what sort of rationale might t here be for such a requirement? McDowell’s argument for (MR) is not easy to state rigorously, but the intuitive idea is clear enough: we might express it loosely by saying that, if perception is to provide reasons for judgment, it must speak in a language that thought can understand. But if it is wrong to conceive of perception as “speaking” at all, it becomes difficult even to frame a question about how perception can provide us with reasons for judgment.
86
R eas o n in P ercepti o n and A cti o n
It thus appears that, if we are to embrace (MR), we must also embrace the idea of perceptual content. In a recent restatement of his position, McDowell draws just this conclusion. Only by accepting (MR), he holds, can we avoid falling into what he (following Wilfrid Sellars) calls “the Myth of the Given.”6 Since he takes the Myth of the Given to be unacceptable, McDowell concludes that we must accept (MR). But in order to accept (MR), he insists, we must credit experiences with content: “It is often thought that when p eople urge that experiences have content, they are responding to a felt need to accommodate the fact that experiences can mislead us. But the proper ground for crediting experiences with content is that we must avoid the Myth of the Given. Making room for misleading experiences is a routine by-product.”7 McDowell’s claim is thus that only if we accept the idea of perceptual content can we explain the rational role of perceptual experience—its ability to supply us with a knowledge- grounding basis for specific judgments about our environment. This is the claim I want to examine in this essay.8 My interest in it is motivated, on the one hand, by sympathy for several of the thoughts that lead McDowell to endorse (MR), and on the other hand, by a conviction that the critics of the idea of perceptual content are right about something important. At a minimum, I believe they are right to question the clarity of this idea, and to suggest that, in its most familiar and straightforward forms, it blurs important distinctions between perception and judgment. Moreover, I am struck by the fact that Kant, who is widely—and, I think, rightly—taken to be the classic advocate of (MR), also appears to reject the idea that our senses present us with truth-evaluable contents. In what follows, I want to consider whether such a position is coherent: Can one both accept (MR) and yet reject the idea of perceptual content? I will argue that we can coherently accept such a combination of views, and w ill attempt, from a Kantian perspective, to make a case that we should.
1. McDowell’s Case for (MR) We have seen how natural it is to appeal to the idea of perceptual content in clarifying (MR). But what was McDowell’s argument for (MR), and exactly what conception of perceptual content does his argument motivate?
T he R ati o nal R o le o f P erceptual C o ntent
87
McDowell’s defense of (MR) was of course inspired by Kant’s famous claim that “intuitions without concepts are blind,” and by his elaboration of this idea in the thesis that “the same function that gives unity to the dif ferent presentations in a judgment also gives unity to the mere synthesis of dif ferent present at ions in an intuition.”9 As McDowell understands the latter claim, it amounts to the idea that when we have “empirical intuitions”—manifolds of sensibly-based presentations by which an object is given to us for cognition—this depends on our actualizing our conceptual capacities for understanding the world, the very capacities that are exercised, in a more reflective mode, when we make judgments about what is the case. Only in virtue of the fact that our sensible intuition actualizes these conceptual capacities does the sensibly-given material possess the kind of intelligibility that allows it to present an object for cognition. (Were this not so, our empirical intuitions would be “blind”: not capable of informing us about objects at all.) So it is not acceptable to conceive of our perceptual capacities as supplying us with representations of objects on which our conceptual capacities are brought to bear in a second, distinct step. Our perceptual capacities must themselves be informed by our capacities for rational understanding, in such a way that t here can be no cognitively significant actualization of the former that is not also an actualization of the latter.10 There are, I think, two distinguishable lines of thought in McDowell’s work that support this Kantian position. One begins from the idea that a satisfactory account of perception must explain how our perceiving can exercise an intelligible “constraint” on what we judge, so that our enterprise of forming a reasonable view about the world is not represented as an activity in which we proceed without input, a “frictionless spinning in a void.”11 McDowell takes this requirement of constraint to imply that perception must have a content that can be endorsed in judgment: perception itself must inform us that “things are thus-a nd-so.”12 But, McDowell goes on to argue, perception cannot exercise the right sort of constraint on judgment u nless its content is presented in a form on which we can reflect critically. Our perceiving something that is F must not merely cause us to judge that this t hing is F; it must supply us with something we can understand to be a reason for so judging, a ground that is intelligible from the subject’s own point of view as speaking in f avor of this judgment.
88
R eas o n in P ercepti o n and A cti o n
It is this demand for rational intelligibility from the subject’s own standpoint that is supposed to motivate (MR). McDowell’s thought is that, if the content of perception is to be available to the subject’s critical reflection, the subject’s conceptual capacities must be drawn on in the perceptual present ation of the relevant content. For if the subject’s conceptual capacities w ere not drawn on in their perceptual experience itself, then “the putatively rational relations between experiences . . . and judgments . . . [could not] themselves be within the scope of spontaneity— liable to revision, if that w ere to be what the self-scrutiny of active thinking recommends. And that means that we [could not] genuinely recognize the relations as potentially reason-constituting.”13 McDowell sums this up by saying that nonconceptual contents could at best supply “exculpations,” not “justifications,” for a subject’s judgments: they could at best make it explicable, and thus excusable, that the subject judged as they did; they could not constitute the subject’s reasons for so judging. So the same conceptual capacities drawn on in making judgments must also be drawn on in our perceiving the things concerning which we judge, on pain of perception’s not supplying us with reasons for judgment. Call this the Rationality Argument for (MR). There is much in this argument that deserves further scrutiny, but given that our interest is simply in the role played by the notion of perceptual content, we can pass over many contentious issues. The notion of perceptual content is not the focus of the Rationality Argument; it enters with little fanfare as a clarification of what it could mean for perception to “constrain,” and thus, in a suitably reflective subject, provide a “reason” for, judgment. The question we will need to consider is exactly what conception of perceptual content is motivated by this demand for rational constraint. Before considering that question, however, I want to mention a second McDowellian line of thought that supports (MR). This line focuses not on the conditions under which perception can present us with reasons for judgment, but rather on the conditions under which our senses can present us with mind-independent objects at all. McDowell argues, in effect, that it is a condition of the possibility of our senses presenting us with objective states of affairs that our capacities for conceptual understanding inform our capacities for receiving sensory impressions.14 He illustrates this point by considering our capacity to observe the colors of things:
T he R ati o nal R o le o f P erceptual C o ntent
89
No one could count as making even a directly observational judgment of color except against a background sufficient to ensure that she understands colors as potential properties of t hings. The ability to produce ‘correct’ color words in response to inputs to the visual system (an ability possessed, I believe, by some parrots) does not display possession of the relevant concepts if the subject has no comprehension of, for instance, the idea that these responses reflect a sensitivity to a kind of state of affairs in the world, something that can obtain anyway, independent of t hese perturbations in her stream of consciousness. The necessary background understanding includes, for instance, the concept of visible surfaces of objects and the concept of suitable conditions for telling what color something is by looking at it.15 McDowell’s topic in this passage is our capacity to make observational judgments about color, but the surrounding text makes clear that he means the point to extend to the visual experiences that provide the basis for such judgments.16 We can have visual experiences that present objective color properties, McDowell suggests, only insofar as our seeing is informed by an understanding of how vision bears on the properties of mind- independent objects. It is not merely that this understanding enables us to make reasonable judgments on the basis of visual experience; it informs our seeing itself, in such a way that a person with the relevant understanding can be visually presented with the colors of t hings, whereas a person lacking the relevant understanding would not be capable of seeing objective colors in this sense. McDowell does not put things this way, but we might think of this second line of thought as beginning from a traditional idea about how to distinguish perception from mere sensation. According to this idea, merely sensory states are simply modifications of the subject’s own experiential state, whereas perceptual states are sensory states that do not just modify the subject’s own experiential state in some way, but present the subject with some object. This conception of the distinction between perceptual and nonperceptual sensing is certainly one important source of the idea of perceptual content: characterizing perception as having “content” can be understood as simply an alternative way of expressing the idea that perception, in contrast with mere sensation, acquaints us with objects and their
90
R eas o n in P ercepti o n and A cti o n
properties (with the colors and shapes of spatially located t hings, with objective features of texture and hardness, with localized sounds, etc.). McDowell claims, in effect, that our understanding of the distinction between how things seem and how they really are—an understanding embodied not primarily in our ability to formulate this distinction abstractly, but in our sensitivity in practice to how perceptual appearances bear on the condition of objective things—is what enables our sensory capacities to function as perceptual capacities. But, McDowell goes on to suggest, this sensitivity must draw on our conceptual capacities. If this is right, then our conceptual capacities must be drawn on in any sensing that amounts to perceiving things in our environment. Call this the Objectivity Argument for (MR).17 There is obviously a g reat deal in the Objectivity Argument that needs further clarification and defense—above all, the crucial premise that we can perceive mind-independent t hings only by drawing on a conceptual understanding of the distinction between how t hings seem and how they really are. For present purposes, however, we can simply focus on the question what sort of “content” perception must have if this argument is sound. It seems to me that the conception of perceptual content required by the Objectivity Argument is considerably less specific than the one invoked in the Rationality Argument. The Objectivity Argument presupposes no more determinate conception of perception than this: that it presents us with mind-independent things and their properties. In what follows, I want to suggest that perception might be contentful in this sense—it might acquaint us with objects and their properties—w ithout being contentful in the more determinate sense invoked in the Rationality Argument—having the sort of content that implies a represent ation of something as possessing some determinate property. If so, we might re spect at least part of the motivation for (MR) without presupposing that perception has the latter sort of content.
2. Problems for Perceptual Content I now turn to some reasons for thinking that the notion of perceptual content is problematic. I w ill not attempt to summarize the arguments that other authors have offered against this notion; I w ill just try to formulate the line of criticism that I myself find most compelling.18
T he R ati o nal R o le o f P erceptual C o ntent
91
The notion of perceptual content about which I want to raise doubts is the idea that perception presents objects as being certain ways (or equivalently: as possessing certain properties or features). I take this to be the conception of perceptual content that is expressed, in a schematic way, when McDowell speaks of perception as presenting “things being thus- and-so”: the idea is that perceptual content involves a referential element (marked by “things”), on the one hand, and a classificatory element (marked by “thus-and-so”), on the other. Let us call any content of this sort represen tational content, and let us call a view that ascribes such content to perception representationalist. There are some contemporary invocations of the notion of perceptual content that are not obviously represent at ionalist in this sense: thus some philosophers hold that the content of perception might be given by a set of possible worlds, or by characterizing the color properties of different regions of a three-dimensional surface.19 On some ways of understanding these proposals, I believe they would amount to forms of representationalism, and would be subject to analogues of the objections I raise below. On others, they would r eally amount to ways of developing the kind of nonrepresentationalist position that I myself endorse. In the present essay, I w ill just focus on views of perception that are straightforwardly representationalist in the sense just specified. The idea that perception has representational content typically forms part of a certain picture of the cognitive role played by perception, one on which our senses, as it were, recommend determinate judgments about the things they present. We may indeed refrain from making the relevant judgments, for instance b ecause we have doubts about the reliability of our senses in such circumstances; but on the representationalist view, a given perceptual experience itself presents t hings as being a certain way: it is itself, so to speak, staked to a determinate view about what is the case. Hence a perception with representational content can be said to portray things veridically or nonveridically. This conception of perception in turn suggests simple and appealing models both of how perception can make a given judgment reasonable and of how it can mislead us. It can give us a reason to make a judgment (in the simplest case, anyway) by itself presenting the relevant judgable content as sound. It can mislead us by presenting as the case what is not in fact the case. In these ways, our senses function like informants who present us with purported information— hence the common metaphor of the “testimony” of the senses.
92
R eas o n in P ercepti o n and A cti o n
When he wrote Mind and World, McDowell held a particularly straightforward version of the representationalist view: that perceptual experiences have propositional contents expressible in declarative sentences. Thus, on the e arlier view, a perceptual experience might have the content: (1) This cube is pink. More recently, McDowell has suggested that we should think of perceptual experiences as having “intuitional content,” which is not supposed to have a propositional form, but rather to be (roughly) expressible in demonstrative noun phrases, as in (2) This pink cube. This latter sort of content cannot itself be evaluated as true or false. The demonstrative expression in (2) may refer or fail to refer, and it may present its referent accurately or inaccurately, but it does not express a proposition that can be true or false. Hence it is not a possible content of judgment. Nevertheless, it is obviously closely related to various judgable contents—for instance, (1). To have a perception whose content was (2) would be to have something in one’s environment presented to one as a pink cube, in a way that would intelligibly rationalize a judgment such as (1). McDowell’s reasons for changing his mind about the logical form of perceptual content are complex, and I am not sure I fully understand his revised view. An important factor in the change was his coming to think that perceptual content is not discursively “articulated” in the sense in which judgable contents are articulated.20 This led him to prefer the demonstrative noun phrase model of perceptual content—which McDowell takes from Wilfrid Sellars—because it registers the fact that the subject must take a further step to organize this perceived material into a specific judgment.21 But McDowell also remarks that Sellars’s noun phrase model is “useful only up to a point,” since it suggests that perceptual experience has a merely fragmentary discursive content, whereas in fact the content of perceptual experience is not discursive at all.22 We w ill need to consider what it means to say that perceptual content is not discursive at all, and how this coheres with the idea that perception has content. For the moment, however, I w ill set aside these complications and take McDowell at his word when he says that the Sellarsian model does capture his idea “up to a
T he R ati o nal R o le o f P erceptual C o ntent
93
point.” Like the propositional model of perceptual content, the Sellarsian noun phrase model conceives of perception as presenting an object as being a certain way. The difficulties I want to raise apply to any conception of perceptual content that makes this basic representationalist commitment.23 The most fundamental objection to any account that ascribes representational content to perception is what we might call the “What Content?” Challenge: Can we, for ordinary cases of perception, specify what plausible content the relevant perceptual experiences could have?24 Consider a case of visual perception. As I write this, I have a copy of Brentano’s Psychology from an Empirical Standpoint sitting on my desk adjacent to my computer. While I am focused on the computer screen, I discern the book only vaguely in the periphery of my visual field, but if I turn to look directly at it, I see the title standing out in clear white capitals on a dark green background. The green background forms a rectangle, which is surrounded by a white border extending to the edges of the book. There is a band of black hashing at the base of the cover, and the insignia of the publisher. But what is “the content” of my perceptual experience when I look directly at the book? Might it, for instance, be one of the following contents? (1) This book is green and white and black. (2) This green, white, and black book. It is tempting to say that my perceptual experience takes in a lot more than this! A fter all, I can see the quite particular patterns in the color that are present, the specificity of their shades, the mild bowing of the book cover, the placement and orientation of the book on the desktop, etc. McDowell sometimes speaks of the content of perceptual experience being “partly specifiable” by a proposition or noun phrase, presumably as a way of acknowledging that the content of a given experience w ill greatly exceed what is captured in any particular specification.25 But is it really plausible to suggest that my perceptual experience itself, either at a time or over some stretch of time, comprises any specific set of contents whatsoever? There is a tremendous amount that I could say about the book on the basis of visual perception, but on reflection it does not seem obvious that my visual experience of the book must (purport to) register some or all of these points independently of my considering the question of the shade of the book, its placement and orientation on the
94
R eas o n in P ercepti o n and A cti o n
desk, etc. I see the book, and in virtue of having it in plain view, I am in a position to answer any of these questions as they occur to me. No doubt the book was determinate in every respect before any particu lar question occurred to me: it lay in a certain spot on the desk, somewhat askew from the computer; on its cover was a certain pattern of green and white; e tc. But the idea that my perception of the book must have presented the book as determinate in some or all of these ways—that the book was represented as having certain specific properties, independently of my raising any par ticular question for judgment—builds an implausible and unnecessary determinacy into unreflective perception. I look down at the book: all the features I have named, and many others, are visible to me. Visib le: that is to say, in virtue of seeing the book, I can take notice of them. But it does not follow that, simply in perceiving the book, some part of me already has registered some set of features of the book, which registration I may judge to be veridical or not.26 Isn’t it enough for perception to present the object about which such determinations can be made, without needing, as it w ere, to propose answers to some specific set of questions about what it is like? As a way of pressing this question further, let us focus on one partic ular feature of the book: the color of the rectangle on its cover. I can see this color clearly: it is green, indeed dark green, a rather wintry shade of dark green, more toward blue than yellow. But what is the representational content of my color experience? Is the rectangle presented to me as green? Representing it as green affiliates it with a certain range of the color spectrum, but must I perceive the color of the rectangle as falling within this range to experience the rectangle as I do? (Think of p eople who have learned a set of color words that demarcate the spectrum differently: must they have some quite other experience of the actual color of this rectangle?) And in any case, isn’t my color experience much more specific than merely that the rectangle is green? It isn’t just any shade of green, a fter all! But which more determinate characterization of the color would capture how my perception presents the rectangle to be? Dark green again covers a wide range of colors, as does bluish dark green, etc. The same kinds of objections that spoke against putting green in the content of the experience will apply to these more specific characterizations as well. It may now appear that the only truly accurate way to express the content of my color experience is to say that I experience the rectangle as being that shade of color, where this demonstrative expression refers to the
T he R ati o nal R o le o f P erceptual C o ntent
95
utterly specific color that the rectangle on the surface of this book appears to exhibit. McDowell has in fact appealed to demonstrative expressions of this sort to capture the “fineness of grain” of perceptual experience, its power to present things as occupying a wholly specific place along some continuous dimension of variation.27 But it seems dubious whether this sort of appeal to demonstratives can save the idea that perception has representational content. Suppose a person tells me he perceives something to have “that shade of color,” and I want to know what color he means: Where do I look? Presumably, I look at the thing demonstrated and assay its color. This phrase designates a color, for me and also presumably for the subject who utters it, only in virtue of the fact that the object demonstrated does in fact have some specific color.28 The demonstrative phrase does not itself express a way of classifying the color in question; it just indicates where we would need to look to make a classification. The idea that my perceptual experience presents the object as having an utterly specific color, namely one designated by the phrase “that shade of color,” spuriously represents this potential for classification as if it were itself a maximally specific classification. So whereas some philosophers maintain that the content of color experience is too “fine grained” to be captured by a concept, and McDowell replies that the conceptual content of color experience might be captured by demonstrative expressions used on the occasion of perceiving specific hether it is right to think of perception itself as presenting colors, I query w objects as having any specific color whatsoever. The idea that the relevant color contents are too fine grained to be captured by a concept, and so must be given in some “nonconceptual” form, gives distorted expression to a point that comes into proper focus when we reject the idea of representa tional content altogether: perception simply presents us with the actual colors of things, which are not themselves classified in any way whatsoever. On this conception, seeing is not a species of representing something as thus-and-so; it is simply a way of being presented with something (in those aspects that are accessible to the vision). A defender of a representational conception of perception might reply that, even if it is not necessary for perception to propose classifications of the objects it presents, nevertheless it might in fact do so.29 That is, it might present us with objects for classification and also certain semblances pertaining to t hese objects: for example, this color might look like a shade
96
R eas o n in P ercepti o n and A cti o n
of red, this pair of tones might sound harmonious, this bit of birdsong might sound like the call of a robin, etc. Antirepresentationalists about perception should not deny, of course, that perceived objects often do seem to us to have particular, specifiable characteristics. Nor should they deny that t hese perceptually based semblances are distinguishable from actual judgments about the objects in question: a perceived color may look red, and yet I may know that the lighting conditions are misleading and so not judge it to be red; a bit of birdsong may sound to me like a robin’s call, and yet I may doubt that this is what it really is; etc. What antirepresenta tionalists should question is whether perception itself should be characterized in terms of such semblances. The cognitive function of perception, on everyone’s view, is to enable us to encounter and learn about features of our environment. The claim I am defending is that we can explain how perception performs this function—how it presents us with objects for judgment and enables us to make well-founded judgments about them—w ithout needing to attribute representational content to it. We can admit the existence of perceptually based semblances without compromising this point. To say that a perceived object looks, sounds, or feels a certain way is to say something about how its appearance disposes us to classify it. What semblances things present to us w ill undoubtedly depend on facts about our own constitution and background experience— for instance, a noise that sounds threatening to a person visiting an unfamiliar place may not sound threatening to a local. Moreover, we may find ourselves disposed to classify an object in a certain way and yet refrain from endorsing this classification in a judgment. Be that as it may, the characterization of what is perceived in terms of such semblances takes a step beyond what we actually perceive: perception itself presents the thing in my environment that I aim to learn about, not the classification it disposes me to make, and whose accept ance might constitute a bit of actual learning. We see, for instance, the shade of green that appears on the cover of this book. On the reading I am proposing, what follows the perceptual verb here is not in an intensional context: we see a certain actual shade of green, describable in indefinitely many ways, but not presented by sheer perception as being any way whatsoever, not even as green. To be presented with an actual shade of color is to be presented with something of unlimited determinacy in the following sense: for any coherent question about
T he R ati o nal R o le o f P erceptual C o ntent
97
this shade, t here is a determinate answer to this question. But this is not to say that our perception presents the shade as determinate in any such respect. Persons who possess appropriate color concepts may recognize the relevant shade as green, dark green, etc., but to see this actual color is simply to be acquainted with something classifiable, not to be presented with some proposal about how to classify it. An analogous line of thought will apply to any aspect of a perceptually- presented object. Let the aspect in question (shape, movement, texture, or whatever) be clearly perceptible via some sense modality. Now consider whether it is plausible to ascribe to perceptual experience itself some specific representational content or set of represent ational contents. Contents that make some determinate, non-demonstrative feature ascription w ill in general have two defects. First, any such content w ill be either too unspecific or else too committal and determinate. For, on the one hand, it is implausible to make the content too generic: what is perceived is surely not just a rough texture or an ovoid shape, but some determinate variety of such a thing. But on the other hand, it is equally implausible to attribute to the subject’s perceptual experience contents corresponding to all of the classifications they could make of the thing on the basis of perception, and arbitrary to restrict one’s attribution to any particular subset of such contents. Moreover—and this is the second defect—attributing to the subject’s perceptual experience a representation of the object as having some specific feature (as being rough, ovoid, etc.) will imply that their perception identifies the object as belonging to a certain class of possible objects of perception. But it is implausible to build this kind of affiliation with a determinate class of possible perceptibles into the perceptual presentation of this object. Given my mastery of the rough / smooth contrast, I can judge this perceptually-presented surface to belong to the class of rough surfaces, but why should experiencing the texture of this surface require affiliating it with any determinate class of other surfaces whatsoever? Including such a classification in the content of perception appears to build an implausible and unnecessary concern with comparison into the act of perception itself. And once again, we should not allow the representationalist to retreat to “demonstrative contents,” maintaining (for example) that perception presents a given surface as having “this texture.” “This texture” is not a classification of the texture in question: it
98
R eas o n in P ercepti o n and A cti o n
does not designate some way of being textured which this surface is presented as exhibiting. It indicates something classifiable, not a description under which the perceived thing is presented but something to be described. The representationalist thinks of perceptual experience as involving both a referential element and a classificatory element: an object is presented as having some specific property, in a way that is clearly marked in the Sellarsian “This such” model of perceptual content. Our reflections suggest that this model conflates two distinct moments in knowledge- acquisition, one of which is purely perceptual, the other properly cognitive. The basic form of perceptual awareness is not apprehending something-as-having-some-property. It is simply apprehending something for classification in various dimensions. Perception puts us in the sort of contact with an object that allows us to raise and answer questions about its characteristics as they occur to us. Specific perceptually-decidable concepts correspond to particular questions that can be raised concerning a perceived object (for instance, Is it red or not?). But perception can perform its cognitive function without itself proposing answers to any such question; it need only make the object available for judgment in certain characteristic respects (color, shape, movement, etc.). Judging whether an object has some property is an exercise of a specific classificatory skill in relation to the perceived object, but to include a purported classification in the characterization of perception itself is in effect to suggest that this particular classificatory skill is exercised simply in perceiving the relevant object. We have seen no reason to think this necessary, and some not inconsiderable reasons to think it implausible.
3. Kant’s Anti-Representationalism Where does this leave us? On the one hand, we have seen that McDowell has prima facie attractive reasons for insisting on (MR). On the other hand, we have found that, for reasons of both plausibility and principle, it is unattractive to attribute representational content to perception. But, as we observed in §1, it is not easy to explain what (MR) means without appealing to the representationalist conception of perception. So it might seem that we must either reject (MR) or embrace the idea that perception
T he R ati o nal R o le o f P erceptual C o ntent
99
has representational content. I want to suggest, however, that another resolution is possible, and that we can find precedent for such a resolution in Kant. I mentioned e arlier that McDowell’s advocacy of (MR) was inspired by Kant’s claim that “the same function that gives unity to the different presentations in a judgment also gives unity to the mere synthesis of dif ferent presentations in an intuition.”30 This remark certainly seems to imply that our understanding—our capacity to frame and apply concepts— is somehow at work in actualizations of our perceptual capacities themselves. Because Kant speaks of sensibility as supplying us with “presenta tions” (Vorstellungen) that are “unified” by “functions” grounded in our understanding, he is often regarded as a classic advocate of the view that perception has conceptual (and hence representational) content. It is therefore somewhat surprising to find him remarking, l ater in the same work, that “truth and semblance [Schein] are not in the object insofar as it is intuited, but in the judgment about it insofar as it is thought. Thus it is indeed correct to say that the senses do not err; yet not b ecause they always judge correctly, but b ecause they do not judge at all. Hence truth, as much as error, and thus also semblance as what leads to the latter, are to be found only in judgments, i.e., only in the relation of the object to our understanding.”31 This remark may at first appear compatible with the idea that perception has representational content, since it seems possible to distinguish the question w hether the senses judge from the question w hether they present representational contents for possible endorsement in judgment. But it is striking that Kant says not only that the senses do not judge, but that for this reason “truth and semblance” are not to be found in “the object insofar as it is intuited.” It is hard to see how someone who took perception to have representational content could say this. For suppose perception presents me with an intuitional content as of (2) This pink cube. If the relevant cube is in fact pink, then is this not a true representation of the cube? Or if we w ill not call it “true” b ecause it is not a proposition, is it not at least an accurate represent ation of the cube, in a way that would make it misleading at best for Kant to say that truth is not to be found in the object insofar as it is intuited? And by the same token, if the relevant
100
R eas o n in P ercepti o n and A cti o n
cube is not actually pink, then has not perception presented us with a misleading semblance of pinkness? If Kant thinks of perception as having representational content, how can he say that “semblance as what leads to error” is not present in what the senses supply?32 We can begin to see how Kant could have regarded his claim about “the same function” giving unity both to intuitions and to judgments as consistent with his claim that “truth and semblance are not in the object qua intuited” by recalling that Kant holds truth to consist in “the agreement of a cognition with its object.”33 A presentation is capable of truth, for Kant, only if it involves a certain duality: there must on the one hand be an object presented, and on the other hand some “cognition” capable of agreeing with this object. Judgments are capable of being true or false because a judgment is “the present ation of a presentation” (die Vorstellung einer Vorstellung) of an object: there is on the one hand some presen tation of an object (expressed, in a categorical judgment, by the subject term), and on the other hand some further presentation of that object as being a certain way (expressed, in a categorical judgment, by the predicate term).34 A judgment is true just if the latter presentation stands in the right kind of agreement with the object presented by the former. When Kant says that truth and error are not in the senses, he means, I believe, that sensible intuitions do not involve such a duality of object presented and presentation of that object as being a certain way. Intuitions, as Kant famously says, are singular presentations that relate immediately to an object.35 What this means, I suggest, is that empirical intuitions pre sent us with aspects of the actuality of some concrete individual (this is their singularity) without relying on any general representation to achieve this presentation (this is their immediacy). A “manifold of intuition” pre sents some a ctual perceptible scene in manifold aspects. I use the word “actual” in t hese formulations to mark the general applicability of a point we noted e arlier in the case of color: an a ctual color (that is, an a ctual case of something’s being a determinate color) is classifiable without limit, but to say that perception presents something’s actual color does not imply that it presents this color as classified in any way whatsoever.36 This potential for classification without limit is a reflection of a basic ontological difference between a t hing and a concept. Any a ctual thing is wholly determinate: it is such that, for any coherent question we may ask about it, t here is a determinately correct answer. Any concept, by contrast,
T he R ati o nal R o le o f P erceptual C o ntent
101
only partially determines the conception of an object to which it applies, while leaving further questions open: Yes, this is green, but is it a light or a dark green (etc.)? Kant himself draws just such a distinction between things and concepts in a passage from the Transcendental Dialectic: very concept, in regard to what is not contained in it, is indeE terminate, and stands u nder the principle of determinability: that of every two contradictorily opposed predicates, only one can apply to it, which rests on the principle of contradiction and hence is a merely logical principle, which abstracts from every content of cognition, and has in view nothing but the logical form of cognition. Every thing, however, as to its possibility, further stands under the principle of thoroughgoing determination; according to which, among all possible predicates of t hings, insofar as they are compared with their opposites, one must apply to it.37 When Kant says that an intuition is a singular presentation through which an object is given for cognition, I think we should understand him to mean that intuition presents t hings (Dinge) in this sense: actualities admitting of classification without limit, but not presented as classified in any way whatsoever. To pose and answer any particular question about the object given in intuition is to make a judgment, and although empirical intuition enables such judging, it does not in itself present any particu lar classifications of the perceived object. In raising a specific question about the object—“Is this pink?”, “Is this cubic?”—we introduce a concept that does not belong to intuition itself, but constitutes a rule we bring to bear on the object in making a particular judgment about it. The idea that concepts are rules is, of course, a Kantian one, and it is closely connected with Kant’s idea that concepts are representations that can never be simply given but must be made in an exercise of the “spontaneity of understanding.”38 This seems on its face to imply that if we consider only the given, nonspontaneous element in our cognition, we should find in it no concepts whatsoever. But if this is Kant’s view—a matter that is, admittedly, contentious—it is not easy to see how it can be reconciled with the idea that intuition, the given element in cognition, has representational content. To ascribe representational content to intuition is to suppose that general ways of being—ways any of a range of objects can be represented as being (for example, pink or cubic)—are presented in
102
R eas o n in P ercepti o n and A cti o n
intuition. A philosopher who ascribes such contents to intuition may stipulate that he w ill not call them “concepts” u ntil they have been “carved out” by our discursive capacities, but it is hard to see how this can be more than a verbal scruple if those very contents are supposed to be already there for the carving in what intuition presents.39 On the repre sentationalist picture, when we frame concepts on the basis of what perception presents, we are (in the simplest case, anyway) just taking notice of general ways of being that perception already presents as extant. We are not making a wholly new kind of representation, but simply making an aspect of our existing representation explicit. Contrast how t hings look if we deny that perception has representa tional content. On this view, what is presented is, for example, the a ctual color of a certain object: something classifiable without limit, but not per se classified in any way whatsoever. Suppose the actual color in question is (in fact) a shade of pink. Perceiving this color does not inform us that the property being pink—a general way of being, which other actual color pre sentations would also exemplify—exists or is instantiated here. To frame the concept pink is to constitute a representation of a general way for things to be by comparing this actual color with others, abstracting from various respects in which they differ, and reflectively arriving at a representation of a way of being they all share.40 This is not a “carving out” but a true making—a constituting of a general represent ation not given in the perceptual presentation of the matter at hand. A subject who has framed the concept pink w ill, of course, normally be able to recognize, on the basis of perception, that certain actual colors are instances of pinkness; but in so doing the subject w ill be bringing to bear a rule for classification that is not to be found in what is given, but must be produced through an act of spontaneity.
4. Perception and Rationality Yet, if perception does not itself propose classifications, how can perceiving something give me a reason to classify it in a certain way? This question brings us back to the Rationality Argument, which maintains that perception can rationally “constrain” judgment only if it has some determinate represent ational content that is accessible to critical reflection.
T he R ati o nal R o le o f P erceptual C o ntent
103
Underlying this argument is a certain picture of rationality, one that best fits cases of inference. An inference characteristically affirms some propositional content (the conclusion) on the ground that it is supported by some set of propositional contents (the premises) that are presumed to be true. The representationalist takes the simplest kind of perceptual judgment to relate to perception in a quasi-inferential way. The step is at best a limiting case of inference, since the content affirmed in judgment is identical to (or, on the Sellarsian noun phrase model of perceptual content, identical given a formal transformation to) the content of the perceptual premise, but even so, the type of rational relation that unites them is one of what we may call representational support: the claim- to-truth of one representation (a judgment) is supported by the presumed truth of another representation (a perception).41 But is this the only way that a judgment can be rationally well-founded? The idea that perception must supply us with “reasons” may artificially constrain our thinking here by encouraging us to look for some proposition-like ground that supports our judgment. But surely there is another variety of well-founded judgment, one that is grounded not on some other proposition or set of propositions, but rather on an exercise of a classificatory capacity in relation to a particular object.42 Having acquired the concept red, for instance, I can reliably judge w hether perceptually presented things are red. I do not make such judgments by first discerning any other independently specifiable feature of the relevant objects; the step I take in assigning this color classification is cognitively primitive. The step may appear less daunting if we suppose that perception itself presents certain things as red, but in fact this does not eliminate the need for a not-f urther-justifiable classificatory step, but merely shifts it from judgment into perception. There must be such a step at some point, for there can be exercises of propositional rationality only given some body of propositions represented as true, and the availability of such propositions itself rests (in the case of empirical propositions, anyway) on a transition from encounters with t hings to representations of how matters stand with those things. If this sort of transition from object to classification cannot be rational, then no further inference can be unconditionally rational either, for the rationality of affirming the conclusion of an inference is conditional on the rationality of accepting the premises of that inference.
104
R eas o n in P ercepti o n and A cti o n
The cognitive function of judgment, according to Kant, is precisely to subsume particular cases under general concepts, and thus to constitute propositions that can serve as the elements of inference.43 In the basic case, the cognitive step to judgment is not itself justifiable by appeal to some rule of application or inference. Kant himself makes this point: “The power of judgment is the faculty of subsuming under rules, i.e., of determining whether or not something stands u nder a given rule (casus datae legis). General logic contains no precepts at all for the power of judgment, and moreover cannot contain them . . . [I]f it wanted to show generally how one ought to subsume under these rules, i.e., to distinguish whether something stands under them or not, this could not happen except once again through a rule. But just b ecause this is a rule, it would demand another instruction for the power of judgment.” 44 The implication Kant draws from this argument is that the power of judgment is a distinct aspect of our rationality (or, as he puts it, a distinct faculty belonging to our “higher cognitive powers”), capable of a distinctive variety of soundness or unsoundness. Good judgment, he maintains, cannot be taught through rules but can only be developed through practice and strengthened through reflection on examples.45 It is this sui generis faculty for subsuming part icu lar cases u nder general concepts, I submit, that is exercised in making judgments on the basis of perception. A perceptually-based judgment does not affirm some proposition (or other form of represent at ional content) that perception has already presented as a candidate for affirmation; it determines whether a particular case falls under the general rule embodied in a certain concept, and thus formulates, for the first time, a representat ion of the case in question. The cognitive role of perception is simply to supply the case for judgment, not to propose some classification of it. Conceiving of perception as nonrepresent ational in this sense does not require us to suppose that our warrant for perceptual judgment is inaccessible to rational scrutiny. We can certainly subject our warrant to scrutiny, in at least two respects: we can examine the perceptually- presented case more carefully, and we can reflect on the reliability and the frailties of the skills we exercise in classifying it. Subjecting our warrant for perceptually-based judgments to scrutiny in these ways certainly draws on our rational capacities, but acknowledging this does not require crediting perception itself with representational content.
T he R ati o nal R o le o f P erceptual C o ntent
105
5. Perception and Objectivity On the view I am proposing, the perceptual states that Kant calls “empirical intuitions” do not have representational content, but they do pre sent objects in a way that enables us to make judgments about them. They do this not by representing these objects as having particu lar features, but simply by presenting the objects themselves, in all their manifold aspects, and thereby making it possible for us to exercise our conceptual capacities in classifying them. Does this picture leave any room for the idea that our capacity to perceive itself depends on our conceptual capacities? It might seem not: for if perception does not represent its objects as having particular properties, what work is there for concepts to do? Yet I want to suggest that, on Kant’s view, there is another kind of work for concepts to do in making empirical intuition possible, a work that precedes any determinate representation of the objects presented in intuition. This claim is both textually and philosophically contentious. Let me begin with a brief sketch of the textual case for attributing such a view to Kant. The case would begin by noting Kant’s famous claim that “the same function that gives unity to the diff erent presentations in a judgment also gives unity to the mere synthesis of diff erent presentations in an intuition, which, expressed generally, is called the pure concept of the understanding.” 46 Kant tells us that a “function” is “the unity of the act of bringing various present at ions u nder a common one,” and he associates such functions with our faculty of understanding, the “spontaneous” aspect of our cognitive power, which frames concepts and applies them in judgments.47 His claim in the passage quoted above appears to be that there is a single fundamental function of our understanding that makes possible not only the unity of diverse elements in a judgment, but also another kind of unity that constitutes, from a manifold of “presenta tions” (Vorstellungen), a single “intuition” that presents an object in its manifold aspects. He tells us that this fundamental function, in its application to intuition, is called “the pure concept of understanding,” and he goes on to argue that this most basic “concept of an object in general” can be divided into twelve distinguishable aspects or dimensions, for which he adopts the Aristotelian term “categories.” 48 Then, in his famous Transcendental Deduction of the Categories, Kant argues that t hese
106
R eas o n in P ercepti o n and A cti o n
categories make possible the unity in virtue of which empirical intuitions are given to our senses at all.49 In this sense, he appears to hold that these basic categories make it possible for us not just to make judgments about perceptually-presented objects, but to have the very perceptions (in his terminology, the “empirical intuitions”) that present objects as possible topics for judgment. It is this Kantian idea that inspired McDowell’s influential reflections on the role of conceptual capacities in perception. On the face of it, however, Kant’s point concerns the conditions, not of the representation of objects as having particu lar features, but of the mere presentation of objects in empirical intuition. I have suggested that, for Kant, this presentation does not involve any representation of objects as exemplifying general ways of being. If this is right, then Kant’s claim that the categories give unity to the presentations in an intuition amounts to the suggestion that the various sensory impressions we receive from objects are, so to speak, focused on particu lar objects in virtue of our relating them to t hese fundamental concepts. To see what this might mean, it helps to observe, first, that the special class of concepts that Kant calls “categories” have a radically different character from determinate empirical concepts. They are, we might say, not material concepts of particular ways for things to be, like red, round, or a potato, but formal concepts of general types of objective being, of which diverse material exemplifications are possible, concepts like substance, accident, reality, negation, unity, plurality, e tc.50 These categorial concepts stand on a diff erent level from the particular determinations expressed by empirical concepts: they express, not determinate ways for t hings to be, but general forms of being, considered in abstraction from any particu lar determination of that form. Thus a substance is any type of persistent being that supports changeable accidents; an accident is any changeable determination that inheres in a substance; a reality is any positive determination of the being of some object; a negation is any determination that characterizes an object merely through what it is not; e tc. We might understand Kant’s idea that the presentations in an empirical intuition are unified by the categories as amounting to the suggestion that our capacities for sense perception are oriented toward determinate forms of objectivity by the same basic functions of the understanding that allow us, at a diff erent stage in cognition, to frame determinate judg-
T he R ati o nal R o le o f P erceptual C o ntent
107
ments about the objects we perceive. It is, for instance, one thing perceptually to attend to a persisting substance, another to attend to one of its changeable accidents. On the reading of Kant’s claim that I am proposing, our understanding of the difference between a substance and its accidents, which is exercised in one way when we judge a certain substance to have a particu lar accident, is also exercised in another, more basic way when we simply train our perceptual capacities on a substance or one of its accidents: here we exercise our sensory powers in a manner informed by an understanding of the categorial form of their object. Kant’s doctrine, I am suggesting, is that this focusing of perceptual attention itself draws on our understanding of certain fundamental formal concepts of knowable being, the categories. In this way, our capacities for conceptual understanding are drawn on in our perception itself, rather than merely in our perceptually-based judgments. Nevertheless, this view does not attribute representational content to perception. It conceives of perception as merely focusing our perceptual attention on an object of a certain categorial type. What our senses then present to us is an actual object in the relevant category: an actual substance, an a ctual color, an a ctual shape, or whatever. These entities fall on the “classifiable” side of the classifiable / classification divide: they are representables rather than representations, whose specific mode of givenness we could express with a perceptual demonstrative, as in “this persisting thing” or “this changeable attribute” (or, more specifically, “this color,” “this shape,” e tc.). As we noted earlier, t hese demonstrative phrases do not themselves express any classification of the relevant entities; they simply focus our attention on the thing to be classified. When Kant says that a function of the understanding gives “unity” to “the mere synthesis of presentations in an intuition,” then, he need not be read as saying that the understanding shapes the deliverances of our senses into determinate representational contents.51 He can be read as saying that the understanding constitutes our manifold capacities for sensing as capacities for empirical intuition, which work together to make determinate objects available for judgment, without themselves posing, much less purporting to settle, any specific question about these objects. The “this” of empirical intuition is, we might say, the sensible enmattering of what Kant calls “the concept of an object in general”—the concept of something about which there can be cognition.52 In orienting our perceptual attention t oward such
108
R eas o n in P ercepti o n and A cti o n
objects, our understanding gives to intuition the unity in virtue of which our various senses are focused on actual objects that can constitute topics for judgment in all its various dimensions. Even if this w ere granted to be a defensible reading of Kant, there would remain the more basic question of whether it is a philosophically defensible view of perception. To answer this question, we would need to revisit the Objectivity Argument, which purports to show that a conceptual understanding of certain basic structures of objectivity is necessary, not merely as a condition of our making judgments about the things we perceive, but as a condition of our very perception of those things. McDowell’s own arguments on this point turn, as we have seen, on the idea that our capacity for objective perception depends on a grasp of the distinction between how things seem and how they r eally are. Our reconstruction of Kant’s argument turns rather on the idea that our capacity for objective perception depends on our grasp of certain basic categories that structure our perceptual orientation toward objects. Each of these claims is controversial.53 A defense of e ither would require detailed argumentation that I cannot offer here. For present purposes, it must suffice to observe that our discussion shows a nonrepresentationalist view of perception to be consistent with McDowell’s core idea: that our capacities for conceptual understanding are drawn on not merely when we make perceptually-based judgments, but in our perceiving itself. On the Kantian view we have developed, the operation of conceptual understanding is a necessary condition for the perceptual presentation of objects for judgment. This is certainly consistent with the spirit, though not with the original letter, of McDowell’s claims about the role of conceptual capacities in human perception.
Conclusion Having reached this point, I must make amends to McDowell, whom I have treated throughout this essay as a representationalist. In fact, since he wrote “Avoiding the Myth of the Given,” McDowell has held that perception has what he calls “intuitional content,” and his account of intuitional content makes any ascription of represent ationalism to him problematic.
T he R ati o nal R o le o f P erceptual C o ntent
109
McDowell holds that intuitional content is not “discursive,” and he explains what he means by making two points. First, “there are typically aspects of the content of an intuition that the subject has no means of making discursively explicit.”54 That is, a subject may have an intuition with content for which they have framed no corresponding concept. For instance, I might have an intuition as of a certain shade of color although I have as yet no concept of that shade. Having an intuition puts me in a position to frame a corresponding color concept, but my having it need not itself involve the actualization of any such concept. Secondly, intuitional content has a unity that is simply “given,” not a unity that involves the subject’s “putting significances together.”55 That is, intuition presents an object as having various features independently of the subject’s exercising their capacity to determine whether and how to attribute those features. McDowell does not himself illustrate the point, but I think he means something like this: I might connect a certain cube with pinkness in various ways, for instance by judging assertorically that it is pink, or problematically that it might be pink, and at any rate appears to be pink. How I judge the cube to be w ill be a m atter of my actively putting together certain significances (this cube and pink) in diff erent ways. But an intuition as of a pink cube itself leaves no scope for such activity: it has a kind of content of which the subject must make something, not a kind that is itself an expression of the subject’s unifying activity. Each of these points, and particularly the first, makes it problematic to call McDowell’s revised conception of perceptual content “representa tionalist.” McDowell speaks of intuitional content as not yet discursively “articulated,” but “present in a form in which it is already suitable to be the content associated with a discursive capacity.”56 He goes on to explain this form of presence in terms of the idea that intuitional content is unified under concepts akin to Kant’s categories: concepts of an object (or kind of object) whose features are potentially discursively articulable, but not actually discursively articulated, in judgments. This is obviously very close to the conception at which we have arrived, on which perception has a kind of unity that makes an object available for judgment, a unity that can be characterized by saying that in perception the concept of an object in general is enmattered by material drawn from our various senses. If there is a difference between McDowell’s position and the view I have defended, it is that I do not see how, having reached this point, it can
110
R eas o n in P ercepti o n and A cti o n
be illuminating to say that intuition presents “content” in a form such that “one could make it, that very content, figure in discursive activity.”57 This characterization makes it sound as if intuition already characterizes or classifies its objects, only not yet discursively, but in some implicit way that discursive activity can make explicit—for what e lse can the content be that is first present in intuition and then made to figure (“that very content”) in discursive thought? But I have urged that we should simply eradicate classification from our conception of intuitive (that is, perceptual) content. Perception may be said to be contentful in the broad sense that it presents objects in a way that makes them available for classification; but it does not have content in the narrower representationalist sense: it does not present t hings as classified in any way whatsoever. Is this position compatible with what is attractive in McDowell’s arguments for (MR)? I do not see why not.
Chapter Four ✣
Resolute Disjunctivism ✣
James Conant
J
ohn McDowell has suggested that his disjunctivist account of perceptual experience affords a useful way of reformulating the nub of the Kantian critique of the Cartesian conception of perceptual experience. The aim of this paper is to present an internal critique of McDowell’s version of disjunctivism and to elaborate a reformulation of it that more fully realizes its original Kantian ambitions.
1. Early McDowell on Cartesianism vs. Disjunctivism It is useful to start with some of McDowell’s earliest statements of a disjunctivist strategy in order to get clear about what its original philosophical ambitions were supposed to be. In his comparatively early article “Singular Thought and the Extent of Inner Space,” McDowell summarizes what he takes to be problematic in (what he calls) “a fully Cartesian picture” and what its subsequent effect on philosophy then came to be as follows: “In a fully Cartesian picture, the inner life takes place in an autonomous realm, transparent to the introspective awareness of its subject; 111
112
R eas o n in P ercepti o n and A cti o n
the access of subjectivity to the rest of the world becomes correspondingly problematic, in a way that has familiar manifestations in the mainstream of post-Cartesian epistemology.”1 This picture forms the primary target of this early article. Its various disastrous effects come in for illuminating discussion. The most disastrous of these is the effect of putting our “very possession of an objective environment in question.”2 It is in this essay—in his effort to spell out what is involved in making sense of the very idea of an alternative picture, one that is able to avoid this disastrous consequence by countenancing cases of the exercise of our perceptual capacity in which we enjoy direct access to external reality—that we find the following helpfully pithy statement of the original idea that animates McDowell’s disjunctivism: Short of that picture, the newly countenanced facts can be simply the facts about what it is like to enjoy our access, or apparent access, to external reality. Access or apparent access: infallible knowledge of how things seem to one falls short of infallible knowledge as to which disjunct is in question. One is as fallible about that as one is about the associated question how t hings are in the external world. So, supposing we picture subjectivity as a region of reality, we need not yet be thinking of the newly recognized infallibly knowable facts as constituting the w hole truth about that region. Of facts to the effect that things seem thus and so to one, we might say, some are cases of things being thus and so within the reach of one’s subjective access to the external world, whereas others are mere appearances.3 The disastrous assumption that underlies the Cartesian picture is this: our subjectivity (what goes on in our mind as a consequence of our exercising our cognitive capacities) constitutes a self-standingly intelligible “region of reality”—“an inner realm.” This conception of subjectivity, in turn, encourages the following idea: the only sort of fact that can be fully transparent to the mind—that can be fully, directly, transparently, and, as such, infallibly known—is a fact pertaining to the layout of this inner realm. The idea of an outer world then threatens to become the idea of a region of reality whose layout can be “known” only in a manner that contrasts sharply with the manner in which we have now, under the influence of Cartesianism, come to conceive what is involved in knowledge of
R es o lute D isjunctivism
113
our own minds. The layout of the outer world comes to be the realm of that which can only be putatively, indirectly, opaquely, and, as such, always only at best “fallibly” known. On this understanding of the meaning of the term “fallible,” it is predicated not just of the capacity to know (in order to express a characteristic of the capacity and a form of defect to which it is liable), but predicated distributively of its exercises—so that it applies imme atter how unimpeded diately to each and every exercise of the capacity, no m or undisturbed its operation. On this conception of a perceptual capacity, pending some supplemental rescue operation, the most that any act of the capacity itself can deliver off its own bat is always something that falls short of being a case of knowing how things are without the mind. A capacity for “knowledge” so conceived—one whose exercises as such must always fall short of yielding knowledge of anything that lies without the mind—is a “cognitive” capacity in name only. So here we come upon the first desideratum that must be satisfied by the sort of account of a cognitive capacity with which McDowell seeks to replace the one that figures in the Cartesian conception. On the alternative conception, “a capacity for knowledge” properly so-called must not be conceived as operating merely within—and hence only upon items available within—a self-standingly intelligible “interior” region of reality. It must allow for the idea that the exercise of such a capacity, when all goes well, affords an unencumbered glimpse of some aspect of the layout of outer reality. It must allow for the possibility of a form of exercise of our cognitive capacity that is fully, directly, and transparently of what is the case—and not only always merely of what seems to the be the case to the bearer of such a capacity. One burden of that early article of McDowell’s is to argue that many contemporary philosophical attempts to leave behind the idea of a Cartesian mental substance, a res cogitans, unwittingly retain what is most fateful in Descartes’s understanding of what it means to say of our cognitive capacity for knowledge of the world that it is fallible—namely, by leaving in place the assumption that we must “break the link” between the concepts of knowledge and of infallibility in the manner that the Cartesian picture does in its account of what must be involved in any attempt to direct our capacity for knowledge at anything without the Cartesian mind. This leads the modern philosopher to suppose that the following conclusion should be taken to be unobjectionable: each act of a perceptual
114
R eas o n in P ercepti o n and A cti o n
capacity—due to the infallibility inherent in the capacity—must as such yield something less than knowledge. To arrive at knowledge, the exercise of such a capacity cannot in and of itself suffice. It requires further supplementation—if not from Descartes’s God, then from an inference to the best explanation, or through a reliability inference, or through something e lse, with which its initial form of actualization must be supplemented—so that we are not only afforded an experience of the perceptual state in question “within the mind” but are entitled to regard it as accurately representing how things are without the mind. This leads to a second desideratum for a satisfactory alternative to the Cartesian account of a cognitive capacity (one that must be in place if we are to properly make good on the first desideratum mentioned above): the alternative account must allow that there can be exercises of such a capacity that are per se knowledgeable—that, through being the exercises of the sort of capacity they are, suffice to yield knowledge, without requiring further supplementation through acts on the part of an additional faculty seeking to remedy the insufficiency of the act of the original (putatively “cognitive”) capacity. McDowell sees that in order to make good on these two desiderata, the alternative form of account he seeks to recommend must relieve the post- Cartesian philosophical pressure to break the link between the concepts of knowledge and infallibility—a break that comes to be felt by modern philosophy to be necessary as soon as we turn our attention to cases in which our cognitive capacity purports to represent how things are without the mind. Fully to abandon the Cartesian picture, according to McDowell, means to allow that “the infallibly knowable fact—its seeming to one that things are thus and so—can be taken disjunctively, as constituted either by the fact that things are manifestly thus and so or by the fact that that merely seems to be the case.” 4 If we can allow this, then “the idea of things being thus and so figures straightforwardly in our understanding of the infallibly knowable appearance; there is no problem about how experience can be understood to have a representational directedness towards external reality.”5 This leads us to a third desideratum that our alternative account of a cognitive capacity must satisfy: it must allow us to retain or recover the link between the concepts of infallibility (or, as we shall call it below, indefeasibility) and knowledge—a link that comes to be severed in most post-Cartesian epistemology.
R es o lute D isjunctivism
115
The question of the remainder of this paper may now be stated as follows: How is the disjunctivist proposal broached in this comparatively early article of McDowell’s best spelled out if the ensuing conception is to fully vindicate these three interrelated philosophical ambitions?
2. Conjunctivism vs. Disjunctivism An important insight of McDowell’s critique of the Cartesian conception—one that it nonaccidentally shares with Kant’s critique of that conception—is the following: as soon as one endorses a crucial assumption underlying the Cartesian picture, it becomes impossible to head off the threat of skepticism. Classical forms of Cartesian skepticism about perception turn specifically on the challenge that it is impossible to maintain both of the following two thoughts at once: (1) In virtue of perception, we can come to know about the world around us. (2) A subject currently having a veridical perception is having an experience that is exactly the same as the experience she would be having w ere she instead having an indistinguishable nonveridical perception. The Cartesian skeptic about perception, in effect, attempts to argue as follows: since (2) is evidently correct, we must reject (1). Is it impossible to maintain both (1) and (2)? Is that right? Well, it depends in part upon how one understands (2)—in particular, upon what it means to claim that one perceptual experience is “the same” as another.6 There is an understanding of (2) according to which it expresses a truism. What McDowell seeks to show is that the Cartesian’s understanding of (2) turns it into something that is by no means platitudinous. A crucial logical feature of the picture McDowell wishes to target can be helpfully brought into focus by considering why it might be apt to speak of that picture as resting upon (what I w ill henceforth call) a conjunctivist analysis of experience—that is, a conception according to which the epistemically good and bad cases share a logically highest common f actor. According to conjunctivism, the good case consists of a conjunction of two factors: the factor it shares with the bad case (things seeming just that way
116
R eas o n in P ercepti o n and A cti o n
to the subject) plus an additional factor (things actually being just that way). Conjunctivism purports to furnish the first half of an overall story about how we acquire perceptual knowledge, while leaving it open exactly how we come to know that the second factor is indeed in place. The disjunctivist wants to expose how the very shape of such a story is hopeless: how the crucial move in the philosophical conjuring game has already been made if we allow for this philosophical two-step—one that both the Cartesian skeptic and the typical antiskeptic equally take to be a philosophically innocent move. The problem for conjunctivism may now be put as follows: Even if some marvelous additional story about considerations bearing on the second factor is supplied by the conjunctivist, given the epistemic impotence of the first f actor considered in isolation from the second factor, how can it ever hope to qualify as an account that recovers the truth of (1)? McDowell has taught us to reject any such conjunctive form of analysis in favor of one that partakes of a disjunctive form—one according to which there is a logical asymmetry between the successful and the failed exercise of a rational capacity. The Cartesian conception of experience contrives to make it seem as if, in order to count as a knower, the occupant of the good disjunct has to determine which of the two disjuncts she occupies and that she has to do so on some ground wholly independent of the mere exercise of her capacity for perceptual knowledge, no m atter how responsible. What McDowell seeks to show is that if we have a sufficiently firm grip on the concept of a nondefective exercise of a capacity for knowledge, then we can appreciate that the subject who successfully exercises the relevant capacity knows, without requiring any Cartesian detour to reach an entitlement to knowledge. We can now summarize a crucial aspect of the point of McDowell’s disjunctivism as follows: the Cartesian conception of the exercise of a perceptual capacity—a nd the conjunctive form of analysis upon which it rests—has the relative logical priority of the defective and the nondefective exercises of the capacity backward. What we require is a form of account that allows us to appreciate how a proper conception of the logical character of a defective exercise of the capacity presupposes prior conceptual clarity regarding the character of its nondefective exercise. We must come to appreciate how the relation between a defective and a successful exercise of our perceptual capacity is never that of a
R es o lute D isjunctivism
117
highest common f actor to a case in which something further has been added to that f actor, but rather always that of a privative exercise to one that is wholly free of any dimension of privation in its exercise.
3. McDowell on Perception as a Capacity for Knowledge In some of his most recent work on disjunctivism, McDowell invites us to consider what it means to claim (as he thinks we should) that perception is a capacity for knowledge. He encourages us to distinguish carefully between what goes on in these two fundamentally different sorts of pos sible exercise of such a capacity: the good case of its exercise which yields knowledge, and the bad case which does not. The aim of the passage we are about to consider is to show us how to conceive of these two cases as related to one another, if we wish to conceive that relation in properly disjunctivist terms. The passage occurs in the context of replying to Tyler Burge. It seeks to clarify what it means to say of a perceptual capacity that it is fallible—as well as how, if we wish so to conceive it, we ought to understand the character of the logical difference in the two respective forms of exercise of the capacity to be associated with each of the two disjuncts in the disjunctivist schema: A perceptual capacity, in the sense that m atters for the disjunctive conception I have sketched, is a capacity—of course fallible—to get into positions in which one has indefeasible warrant for certain beliefs. That is what the capacity is a capacity to do, and that is what one does in non-defective exercises of it, exercises in which its acknowledged fallibility does not kick in. For instance, a capacity to tell whether things in one’s field of vision are green is a capacity—of course fallible—to get into positions in which the greenness of things is visibly t here for one, so that one has indefeasible warrant for believing that they are green. That the capacity is fallible means that a possessor of it can be fooled; for instance, if the light is unsuitable for telling the colors of t hings, one can take something to be green when it is not. It is wrong to think it follows that even when one is not fooled in an exercise of this capacity, one’s position must fall short of having the greenness
118
R eas o n in P ercepti o n and A cti o n
of something visibly present to one, and thereby having an indefeasible warrant for believing the thing to be green. Fallibility is an imperfection in cognitive capacities. But the mistake I am pointing out is easier to recognize if we consider its analogue in application to other sorts of imperfection in capacities. Some people have a capacity to throw a basketball through the hoop from the free-throw line. Any instantiation of such a capacity is imperfect; even the best players do not make all their free throws. Burge thinks that t here cannot be a fallible capacity in whose non- defective exercises one gets to have indefeasible warrant for certain beliefs. One might as well think that t here cannot be a capacity—of course not guaranteed success on all occasions—in whose non-defective exercises one actually makes free throws.7 On McDowell’s understanding of the terms “perceiving” and “seeing,” an act of such a capacity presents me with a judgable content—one that may eventuate in knowledge through a further act in which I accept what I thus “perceive” or “see” as true. I do not wish to dispute that t here are scenarios in which my perceptual capacity is in act in such a manner—so that there indeed needs to be a logically distinct further act of acceptance in order for my cognitive activity to yield a form of perception-based knowledge. One question that w ill eventually occupy us in this paper w ill be the following: Should the unqualifiedly nonproblematic exercise of a capacity for “perceiving” or “seeing” be conceived as equally partaking of such a logically composite structure or should in such a case “perceiving” or “seeing” be taken to denominate a particular manner of judging—a partic ular way (as Kant would put it) of actualizing our general capacity for judgment? This, in turn, w ill lead us to a second question: Is “perceiving” understood along the lines McDowell here recommends (as the name of a capacity that puts me in—what he calls—“a position to know”) a form of capacity whose most basic exercise is an act of self-consciousness?8
4. Merely Seemingly Seeing vs. Successfully Seeing Before taking up t hese two questions, I would like to mark some distinctions and introduce some terminology for keeping track of them— terminology that w ill prove helpful later. While doing so, I would also like
R es o lute D isjunctivism
119
to underscore what I do not wish to object to in McDowell’s disjunctivism. For I agree with much of what McDowell intends the lesson of the above passage to be. We might put the most central point with which I wish to agree as follows: just as a proper appreciation of the fallibility of the practical capacity of shooting a f ree throw is no bar to allowing that when an appropriately accomplished player shoots the ball through the hoop, a proper comprehension of his practical capacity suffices to explain the possibility of its nondefective exercise, so too, a proper appreciation of the fallibility of our theoretical capacity for acquiring perceptual knowledge ought not to stand in the way of an acknowledgment that (for example, in a situation where I am entitled to claim “I see p”) my seeing is due to the nondefective exercise of my capacity to see what is so. The nondefective exercise of such a capacity suffices. Suffices for what? McDowell and I agree on the following: in the successful case, it suffices to ensure that t here is no gap to close between representation and world.9 We therefore agree on this point: the conjunctivist is wrong in thinking that t here is a highest common factor across the two disjuncts—across nondefective and defective exercises of our perceptual capacity—so that even in the good case the question remains no less open than in the bad, whether the repre sentation in question is appropriately world-involving. Only once one is forced to spell out one’s answer to the foregoing question—“Suffices for what?”—does the point at which McDowell and I diverge come fully into view. McDowell will say in answer to that question: a nondefective exercise of my capacity for perception (if it is to be conceived as a capacity for knowledge) suffices to put me in a position to know. I w ill say: an in-no-way defective exercise of this capacity suffices for me to know—this is what it means to say that it is an in-no-way defective exercise of a capacity for knowledge. We w ill explore what this disagreement r eally comes to below. Before we do, let us first get clearer about the wider background of agreement against which this intramural disagreement between two types of disjunctivist takes place. McDowell would rightly regard a certain sort of response on the part of a conjunctivist to his version of disjunctivism as utterly point-missing. I w ill raise the worry toward the end of this paper whether McDowell’s own way of seeking to resist my preferred conception of a perceptual capacity for knowledge replicates a feature of what is point-missing in the following possible response to disjunctivism on the part of a philosopher who remains in the thrall of the conjunctivist
120
R eas o n in P ercepti o n and A cti o n
conception. But first we need to be clear about exactly why this sort of response to McDowell is point-missing. To this end, imagine a conjunctivist who seeks to head off McDowell by saying the following: Consider a person who thinks her visual experience puts her in a position to say that she is seeing an oasis. But she later realizes she was wrong about that and says something on t hese lines: “I thought I was seeing an oasis but I now realize that what I was seeing was just a hallucination.” According to this quite intelligible remark, it was true at the relevant past time that she was seeing something. Indeed, she could have had that very same visual experience whether what she saw was a hallucination or a real oasis. If she subsequently learns that what she had been seeing was merely a hallucination then she w ill withhold her assent from the visual appearance that there was an oasis. But whether she subsequently withholds her assent from that visual appearance or not, what she then saw was the same. So, in such a case, there is a highest common factor between seeing a hallucination of an oasis and seeing an oasis—namely, a certain sort of visual appearance. Our imaginary conjunctivist interlocutor’s way of using the verb “see” here involves a conflation of two diff erent uses—which Gilbert Ryle teaches us respectively to recognize as the use of a verb of success and a mere- appearance use of that same verb. Let us call the latter sort of “seeing” merely seemingly seeing.10 Let us try calling the other sort successfully seeing. The conjunctivist is vulnerable to the charge that he fails to appreciate the grammatical relations that obtain between t hese two ways of speaking of what someone sees. The wrong sort of fixation on the surface grammar of these two ways of speaking, however, might well at first appear to confirm the conjunctivist’s picture: for, taken together, this pair of locutions invites the idea that they respectively refer to the members of a pair of ways of d oing one and the same thing (“seeing”)—merely seemingly doing it and successfully doing it—where each of the adverbs modifies the same common core of visual activity, albeit it in two very different ways. This goes with the following picture of how to understand the success use of the verb: it involves a linguistic device that allows its user through the employment
R es o lute D isjunctivism
121
of a single word (that is, without the need for an additional adverbial modifier) to single out a very particular way of d oing something (namely, doing it successfully) from a much larger manifold of possible ways of doing that same thing. To understand the point in this way would be precisely to misconstrue the relation of logical priority between success and nonsuccess uses of a verb. Ryle writes: “We shall see l ater that the epistemologist’s hankering for some incorrigible sort of observation derives partly from his failure to notice that in one of its senses ‘observe’ is a verb of success, so that in this sense, ‘mistaken observation’ is as self-contradictory an expression as ‘invalid proof’ or ‘unsuccessful cure.’ But just as ‘invalid argument’ and ‘unsuccessful treatment’ are logically permissible expressions, so ‘inefficient’ or ‘unavailing observation’ is a permissible expression, when ‘observe’ is used not as a ‘find’ verb but as a ‘hunt’ verb.”11 When one speaks of “mistaken observation,” one is not using the verb in its success sense. Conversely, if one is using the verb in its success sense, then the concept observation does not admit of the qualifier correct (for it in no way qualifies). One moral to be drawn from what Ryle says h ere is that in my foregoing use of the locution “successfully seeing,” the “successfully” is at best redundant and at worst misleading, so it should be scrapped. This is the right conclusion to draw, if the locution is understood to imply that “successfully” seeks to introduce a modification into what otherw ise would be a logically less complex description of an act. If one means to use the verb in its success sense, then the “successfully” is logically otiose. J. L. Austin, for one, would therefore have vehemently disapproved of my use of the adverb “successfully” and his worry would not have been entirely misplaced: “Only if we do the action named in some special way or circumstances, different from t hose in which such an act is naturally done (and of course both the normal and the abnormal differ according to what verb is in question) is a modifying expression called for or even in order.”12 Austin’s remark here is directed at those philosophers who he thinks manage to confuse themselves through the use of idle modifying expressions that in no way determine the verbs they modify. If we have no occasion to speak in such a manner in ordinary life, then, ceteris paribus, that qualifies for Austin as a sufficient ground for concluding that we ought not to speak that way as philosophers.
122
R eas o n in P ercepti o n and A cti o n
One can sympathize with Austin’s thought that only u nder the pressure of philosophy w ill someone be prone to go in for such nonmodifying uses of adverbs without thereby endorsing his conclusion that any instance of such a form of use ought to be placed on the index. Such merely apparently modifying expressions (that do not serve further to logically determine the concepts they apparently modify) play an essential role, for example, in the philosophical practices of both Kant and Wittgenstein. One w ill encounter such forms of use wherever these phi losophers seek to characterize (what Kant or early Wittgenstein call) the logical or (what later Wittgenstein calls) the grammatical form of a concept. In what we might call its formal use, the word “successfully” serves such a philosophical purpose. When so employed, the adverbial expression bears an altogether different logical valence than does an adverb (such as alertly) when used to materially characterize the particular manner in which one sees. So even if Austin is right about the logically otiose character of saying one is “successfully” φ-ing, where what is at issue is a material use of φ and φ denotes a verb of success, this still leaves room for a philosophically elucidatory use of such an adverb to assist in the task of bringing to logical self-consciousness an aspect of the formal character of the concept u nder investigation. In employing the expression “successfully seeing” to elucidate what is philosophically confused in the conjunctivist’s conception of “seeing,” the aim is precisely not to introduce a determination into a broader genus of seeing in order to isolate a species thereof. The role of the adverb is to highlight a formal aspect of the concept under investigation. Successfully seeing (so understood) stands to merely seemingly seeing for McDowell’s disjunctivist as the nondefective stands to the defective exercise of a capacity. That is to say, the expression “successfully” as used here merely indicates a form of exercise of the capacity that is free of any dimension of logical privation.13 The elucidatory role of the word “successfully” in this context is simply to make wholly explicit such an absence of logical privation—hence properly to isolate an act of the capacity that is in no way encumbered: that is fully spontaneous in the Kantian sense of the term. So employed, the formal use of “successfully” functions as a philosophical device for clarifying what a case of simply seeing comes to. One way to characterize the conjunctivist’s philosophical blind spot would be to say that he misapprehends the formal role of “successfully”
R es o lute D isjunctivism
123
in characterizing the logical grammar of perceptual verbs that are—in Ryle’s terminology—verbs of success. The Cartesian conception of sense experience requires one to construe the grammar of “successfully” as if it were on a formal par with the role of “merely seemingly”—that is, as if it indicated some form of modification of the logically most basic exercise of one’s capacity for seeing. One way to put the disjunctivist’s teaching would be to say that she seeks to show the conjunctivist how he comes to misconstrue the logical relation that obtains between these cases—and, in particular, to show him that merely seemingly seeing is logically parasitic on the success sense of (simply) seeing. When I (in the success sense) see p, I see what is so (not just what seems to be so); when I merely seemingly see p, there is no world-involving state of affairs p that I see. McDowell rightly assigns an absolute significance to this dimension of success in an account of the logical grammar of perceptual verbs of success. The way McDowell puts this is to say that for me to count as (really, simply, successfully) seeing, my seeing must reach all the way to the fact and must put me in a position to know; merely seemingly seeing does neither of these things—it neither reaches all the way to the world in the relevant sense, nor does it put me in a position to know. Hence, for the disjunctivist, any grasp of a merely seemingly seeing sense of the verb “see” would have to be logically parasitic on a prior grasp of the logically more demanding success sense of the verb.14 One way of putting what McDowell’s version of disjunctivism teaches would be as follows: in misconceiving how merely seemingly seeing stands to successfully seeing, what the conjunctivist takes for a highest common f actor is a philosophical illusion—one that can be dissipated through a grammatical investigation of the relations of logical priority that obtain between these two ways of using a cognitive verb and hence between the correlative (defective vs. nondefective) forms of exercise of a cognitive capacity they denote. As long as the conjunctivist’s picture remains in place, it can seem as if we have no option but to conclude that any nondefective exercise of our perceptual capacity must share in the nature of a defective exercise of that same capacity in the following respect: all any act of perception can ever yield is a form of representation restricted to the Cartesian conception of the perceptual sphere—hence one that does not reach all the way to the things themselves. McDowell’s disjunctivist analysis seeks to show that there is a fundamental asymmetry between a nondefective and a defective
124
R eas o n in P ercepti o n and A cti o n
exercise of a perceptual capacity, such that the former reaches all the way to the things themselves, whereas the latter does not. T hese two disjuncts stand to one another not as two species of a genus, but rather as stage thunder to thunder—as simulacrum to genuine article. McDowell’s point against the conjunctivist might therefore be put as follows: we should not read the logical structure of the defective exercise of the capacity into that of its nondefective counterpart, so that we end up conceiving of the latter as a composite of an appearance and a reality. The disjuncts do not share a common factor; they involve logically fundamentally distinct forms of representation—or, to put this point in the Kantian idiom that I have been employing: they result from logically distinct forms of exercise of and one the same cognitive capacity—the form of the first is that of a genuine appearance (in which we see what is so); the form of the other is that of a mere appearance (in which we at most only seem to see what is so). We can see McDowell here as attempting to remain faithful to the following philosophical maxim: Do not try to read the formal character of the logically primitive phenomenon off the model of its logically privative counterpart! The question that w ill occupy us in the remainder of this paper may now be put as follows: What does full faithfulness to this philosophical maxim require of us when elaborating a disjunctivist conception of how properly to discriminate the full range of logically distinct forms of exercise of a perceptual capacity for knowledge?
5. Putting Oneself in a Position to See vs. (Successfully) Seeing Once we become clear in this way about where the real nub of the disjunctivist critique of a Cartesian conception of experience lies, it opens up the following question: How should we conceive the formal character of the logically privileged disjunct? From what point(s) of view must it be possible for the perceiving subject to answer the question: Do I myself occupy that disjunct? Does it suffice if that question can be answered from a point of view outside the unity of apperception of she who exercises the self-conscious cognitive capacity? Or must the knowledge that I am an occupant of the disjunct in question be self- consciously available to me
R es o lute D isjunctivism
125
from within the scope of a successful act of the perceptual capacity itself? Before we take up this question, let us note that McDowell’s conception of (what is involved in my doing what he calls “non-defectively exercising” my capacity for) perceptual experience still leaves room for a highest common f actor across the following two cases: (a) the moderately happy case in which an exercise of my perceptual capacity successfully gets me into a position in which I have “an indefeasible warrant” for p but I still need to make a mediating reflective maneuver in order to exploit that warrant, and (b) the even happier case in which I simply exploit that same warrant and thereby immediately come to know p. One of the places in which McDowell takes particular pains to defend this structural feature of his overall conception is in a response to Barry Stroud. Stroud says this: “A person who sees that it is raining judges or believes or accepts or otherw ise puts it forward as true that it is raining.”15 McDowell comments in response to this: I think that is simply wrong about a perfectly intelligible notion of seeing that something is the case. And this other notion is the right one for my purposes. Certainly one w ill not say one sees that p unless one accepts that p. But one can see that p without being willing to say one does. Consider a person who thinks her visual experience does not put her in a position to say how things are in some respect. But she later realizes she was wrong about that, and says something on t hese lines: I thought I was looking at the tie u nder one of those lights that make it impossible to tell what color things are, so I thought it merely looked green to me, but I now realize that I was seeing it to be green. According to this quite intelligible remark, it was true at the relevant past time that she was seeing the tie to be green, but at that time she did not in any way put it forward as true that the tie was green. She does now, but that is irrelevant to Stroud’s claim that the seeing itself must have involved endorsement or accep tance. She withheld her assent from the appearance that the tie was green that her experience presented her with—an appearance
126
R eas o n in P ercepti o n and A cti o n
that was actually the fact that the tie was green making itself visually available to her. Stroud thinks withholding one’s assent from an appearance is incompatible with seeing that t hings are that way. But this person, as she now realizes, did see that the tie was green, though she withheld her assent from the appearance.16 McDowell’s example here involves a variation on Sellars’s famous parable of the tie shop. In the original Sellarsian tie-shop scenario, I can come to be in a position to know a tie in the tie shop is blue even though it appears green to me, thanks to an additional piece of knowledge that I can acquire: knowledge regarding how I am perceptually circumstanced—specifically knowledge to the effect that under the conditions of the yellow lighting in the tie shop a blue tie w ill appear green.17 McDowell’s example asks us to imagine someone who believes herself to be in Sellars’s tie shop, but does not realize that the proprietor has now arranged for the lighting to be fixed, so that blue ties now appear to be blue and green ones appear to be green. Absent this knowledge—still operating under her prior belief that she is in a tie shop in which blue ties appear green—a longtime customer of the establishment w ill not know right off when looking at a green tie that it is green. This is not something she w ill know—absent a further mediating inference—simply by looking at the tie and perceptually apprehending its (apparent) color. McDowell, in effect, suggests h ere that reflection on his modified version of the parable of the tie shop ought to enable us to distinguish two diff erent but equally valid paths by means of which one can come to occupy the good disjunct: there is the familiar direct route (in which one knows by just looking and seeing what is the case and accepting it to be so), but then there is also the route traveled by the protagonist of the parable (who can say “I now realize that I was seeing the tie to be green”). She, too, counts for McDowell as an occupant of the good disjunct. Admittedly, the situation of the parable’s protagonist is initially slightly less optimal than that of her more celebrated (and initially reflectively less unfavorably circumstanced) co-occupant, but McDowell in effect urges: Do not let that distract you from what matters! If the parable achieves its intended purpose, it w ill have shown you that our protagonist does not differ in any way that ought to matter for homing in on the rele-
R es o lute D isjunctivism
127
vant concept—namely, that of a successful exercise of a rational perceptual capacity. The crucial point is that these co-occupants of the good disjunct are equally able to pass the following test with flying colors: they can each truthfully claim to have “successfully” “seen” the tie to be green. We are invited to conclude that the sentential verbal twins that form the members of the pair of claims that they w ill thereby respectively utter bear a common logical form. That the respective perceptual accomplishments of our co-occupants are fully on a par (in the logically relevant re spect) w ill come clearly into focus for us, McDowell suggests, if we properly reflect upon the significance of the fact that they express one and the same claim when each of them says: “I saw the tie to be green.” That they are in equally good positions to make such an avowal shows them to have equally valid title to a perch within the good disjunct. For it shows them to have been equally able via a fully successful act of their perceptual capacity to take in what is the case. Is that right? This example brings out a critical feature of McDowell’s conception of what qualifies someone as an occupant of the good disjunct—of what constitutes a minimal case of someone’s having cleared the threshold for counting as having successfully seen what is so. Though, of course, he nowhere formulates the upshot of his discussion in these terms, McDowell’s midrash on the parable, in effect, depends upon the possibility of a revelation of a highest common factor across the two scenarios distinguished above. On this conception of perceptual experience, the sentence “I saw that p” expresses the very same perceptual thought p—equally available to the two co-occupants of McDowell’s good disjunct—a nd it does so regardless of whether the subject of the perceptual experience initially accepts p or not. This much seems right: it is equally linguistically appropriate for an erstwhile occupant of either of these scenarios in retrospect to conclude something that she may express by calling upon the words “I saw that p.” But that still leaves open the question whether these two fully apt employments of one and the same form of words express the same thought across these two circumstances of use. The question that we now need to explore in order to get clear about this is the following: Do these two uses of a perceptual verb across the two scenarios denote the same form of exercise of one’s perceptual cognitive capacity?
128
R eas o n in P ercepti o n and A cti o n
6. Self-Consciously Seeing vs. Self-Alienatedly Seeing Now I do think t here is a use of the verb “see” that approximates what is here desired by McDowell (when he speaks of “a perfectly intelligible notion of seeing that something is the case” which is—as he puts it—“the right one for my purposes”),18 but on the sense that the verb acquires through that form of use, it now admits only of certain forms of conjugation.19 We can begin to see why this is so by assembling a few observations about what is being said about the “perceiver” who “sees” on such an employment of the verb. Notice first: without the interpolation of a further perspective—either that of another perceiver or of a retrospective perspective of my own on my previous act of “seeing”—the only way I can know that the first of the following two claims is true is if I already know the second is true: (1) I am in a position to successfully see p. (2) I see p. Notice second: there is nothing that the following could amount to: my knowing (1) while not yet being able to claim the following: (3) I know p. Now notice third: if I am conjugating both of the two verbs at issue h ere (the verb for putting oneself in such a position and the verb for self- consciously seeing) in the first-person present indicative, there is no act of the mind that can express the joint affirmation of (1) and the negation of (3). On the use of the verb “to see” for which McDowell wants to allow, someone else can say of the perceiving subject who attempts to see the color of the tie under the reflectively unfavorable circumstances sketched in McDowell’s tie shop scenario: she is (according to “the right notion” for McDowell’s purposes) “seeing” it to be green (though she doesn’t realize that she is, b ecause she thinks the lighting in the tie shop c auses blue things to look green). Later, after she learns that the shop’s lighting does not cause blue things to appear green, she’ll be able to say in the first-person past indicative something that resembles what the observer of the original event of “seeing” can already say about her in the third-person present
R es o lute D isjunctivism
129
indicative. But do the two uses bear the same logical form? Let us suppose that they do. If so, then the logical grammar that the verb “see” bears on the supposed form of use common to t hese two speakers is one that does not admit of a first-person present indicative employment. A logical link that ordinarily obtains on the logically full-blooded success use of verbs for self-conscious forms of cognitive activity has here been severed—the link that, for example, ordinarily underwrites a third-person observer’s ability (say, in a normal tie shop) to move from claiming (about a customer sampling a tie) that “she sees p” to inferring that this same piece of knowledge is expressible by the customer herself in the form of a first- person indicative present tense avowal.20 I want to claim the following holds of the relation between the notion of seeing that Stroud seeks to elucidate and the notion of “seeing” that McDowell seeks to single out as putatively “the right one” for his purposes: the latter stands to the former as privative to nonprivative exercises of one and the same capacity. The very intelligibility of McDowell’s protagonist’s manner of employing a perceptual verb (where the aforementioned link is severed) is parasitic on a logically prior use of any such verb, where it expresses a fully successful exercise of a self-conscious capacity for knowledge—one that can be expressed in the form of a first-person present- tense avowal. It is the mark of the relevant class of verbs when employed in their logically full-blooded success sense not only that they admit of such a first-person present indicative use, but also that this use expresses the logically fundamental form of exercise of the rational capacity they self-ascribe. We might express this point as follows: on the logically full-blooded use of the verb, the “I see” must necessarily be able to accompany any p I can be said to see (just as for Kant, it is the mark of self-consciousness that the “I know” must necessarily be able to accompany any p I can be said to know). Let us call this the self-consciously seeing sense of “see.” On the logically derivative use of the verb we h ere seek to understand (the one that occurs in the protagonist’s remark “I now realize that I was seeing it to be green”), any p I can be said to see (by someone else or by my later self) is not only now no longer necessarily accompaniable by a first-person present indicative use of “I see”—it is not even possibly so accompaniable. The concept of “seeing” employed by the parable’s protagonist requires the following remarkable logical feature to be in place: that the subject lacks
130
R eas o n in P ercepti o n and A cti o n
(what we might call) perceptual unity of apperception. We require a logically peculiar concept h ere b ecause we want to speak of someone who suffers from a logically peculiar form of perceptual disunity of apperception. A variant on our logically full-blooded concept of “seeing” can come into play here, but only if that concept first comes to suffer privation along the requisite logical dimension.21 Let us call the logically secondary concept of seeing that comes to be generated in this manner that of self-alienatedly seeing. How does this concept differ from that of self-consciously seeing? It helps to register a difference in the two concepts of seeing that come into view here if we reflect on the forms of relation they respectively bear to Moore’s paradox.22 On the self-consciously seeing use of the verb, the following sentence is Moore-paradoxical: “I see p but I d on’t know that I am seeing p”—and so, too, is its first-person past-tense indicative counterpart. This formal dimension of the concept pervades the entire logical space of its use. On the self-a lienatedly seeing use of the verb, the following sentence is not Moore-paradoxical: “I saw p but at the time I d idn’t know I was seeing p”—a nd this entails that it has no first-person present-tense indicative counterpart. This moment of self-a lienation pervades the entire logical space of the use of this concept of what is to see.23 Someone might miss the force of this by trying to object along the following lines: The following is paradoxical for any r, provided the following sentence is well-formed: “I r but I don’t believe I am r-ing.” Meanwhile, the following sentence is never paradoxical: “I r-ed but at the time I didn’t know I was r-ing.” For example, “I walk but I don’t believe I am walking” strikes me as paradoxical, if well- formed, and “I walked but at the time I didn’t know I was walking” does not strike me as paradoxical. Yet, presumably, your argument above about “seeing” is not meant to extend to “walking.” So I am worried that your argument doesn’t work because it threatens to show that “walking” bears a logical peculiarity you want to attribute specifically to “seeing.”24 Contrary to what the objector here presumes, the point is meant to possess a degree of generality he supposes it cannot mean to have. The logical peculiarity in question is not peculiar merely to the perceptual cognitive
R es o lute D isjunctivism
131
achievements (as opposed to the nonperceptual ones) of a self-conscious being, but rather to all logically full-blooded exercises of self-consciousness (be they acts of our capacity to know the world or ones of our capacity to intellectually act within or upon the world). The point in question, if developed with sufficient care, may and should be extended to walking—or more precisely: to cases where the verb to walk expresses the full- blooded intentional exercise of the sort of capacity for practical knowledge that Anscombe elucidates in Intention.25 Our interlocutor, however, is right about the following: the point cannot be extended to anything anyone might call “walking” on any possible occasion of use of that verb. In particular, it w ill no longer apply when the verb ascribes to a rational subject something other than a full-blooded exercise of her capacity for intentional action. What the objection can serve to bring out—once the logical character of the use of the verb in question has been appropriately clarified—is a point that holds equally of seeing and walking. The negative way to formulate that point is as follows: when the capacity expressed by the verb is a rational capacity for knowledge (be it theoretical or practical), and if the form of the exercise is a self-a lienated one (as opposed to a fully self- conscious one), then there w ill in such a case no longer be anything paradoxical about an utterance of the following form: “I was r-i ng but at the time I didn’t know I was r-ing.” (Consider: “I was sleepwalking but at the time I d idn’t know I was sleepwalking.”) Moreover, since the principle of charity behooves us to try to make sense of the utterances of others, when someone utters something of this form, it is natural to construe the verb in such a context of use as bearing some sense other than its full- blooded, self-conscious, success sense. Once it is so construed, the interlocutor’s example o ught not even to appear to tell against the crucial point. To formulate the logical peculiarity that h ere equally characterizes a self-conscious being’s capacities for seeing and walking in positive terms: if the form of exercise of a capacity for knowledge that the verb r expresses is the one expressed by the logically full-blooded success sense of the verb, then an utterance of the form “I was r-ing but at the time I didn’t know I was r-ing” w ill necessarily be Moore-paradoxical. We can make the point more vivid by exploring the logical grammar of a concept that denotes a species of walking—as the concept of sleepwalking does—which exhibits a dimension of logical privation parallel to
132
R eas o n in P ercepti o n and A cti o n
that found in the concept of self-alienatedly seeing. The parallel may be put as follows: self-a lienatedly seeing stands to self-consciously seeing as sleepwalking does to (intentionally exercising one’s practical rational capacity for) walking. Just as it is not paradoxical for McDowell’s parable’s protagonist later to say of her former self “I was seeing it to be green, but at the time I d idn’t know I was seeing it to be green,” so too t here is nothing paradoxical about my later saying of my former self “I was sleepwalking, but at the time I didn’t know I was sleepwalking.” Correlatively, just as the verb for self-alienatedly seeing admits of no first-person present indicative sense, neither does the verb for sleepwalking. There simply is no form of knowledge which could be expressed by my saying—here and now, as I am doing it—“I am sleepwalking.” As with self-a lienatedly knowing, so too with the case of self-a lienatedly walking h ere under examination: there is no thought to think which could be expressed through the self-predication of the verb to sleepwalk in the first-person present indicative. Conversely, in order for someone to be brought under that concept, a logical gap must be opened up between the one whose capacity for sleepwalking is in energeia and the one who speaks of that actuality. This gap can be opened up by the one who speaks of the sleepwalker being someone other than the sleepwalker himself (in which case she can say of him in the third-person present indicative: “He is sleepwalking”). Alternatively, if I wish to self- predicate the verb, then the requisite logical gap can be opened up through a modulation of the tense or mood of the utterance (hence I can, for example, express retrospective knowledge of what I had been doing at a certain time by saying “At the time I was sleepwalking”). The former of these two sorts of employment of a verb for self-alienated activity is available to a clued-in observer of the perceptual activity of the protagonist in McDowell’s parable. On that employment of the verb, the words “she sees the tie to be green” (when uttered by such an observer) express the observer’s knowledge of the protagonist’s initially cognitively deficient relation to p. The initial applicability of the concept of self- alienatedly seeing h ere turns on the fact that it is not one and the same person who predicates the concept and of whom the concept is predicated. In order to understand the parable, we must enter into the perspective of the narrator: we, too, must become clued-in observers. Our prior grasp of the logical character of this third-person perspective (from which we initially comprehend what the protagonist may be said to see, even though
R es o lute D isjunctivism
133
she cannot yet say this of herself) is a condition of our being able to understand the sense that the protagonist’s own later ascription (to her earlier self) of “seeing” bears. The preceding meditation aims to bring out the importance of the following point: the logically primary use of our ordinary verbs for rational cognitive capacities is the first-person present indicative. This is the perspective from which we comprehend the capacity from the inside: from within and through its exercise. It is the perspective of Descartes’s thinker when he thinks the cogito, as well as that of Kant’s logical I who thinks that anything he can think is necessarily accompaniable by the “I think.” It is the perspective from which a proper philosophical elucidation of our cognitive capacities, as beings capable of rational thought, action, and speech, must begin. The logically primary use of the verb for the concept of seeing that figures in McDowell’s parable is the third-person present indicative. The perspective from which to begin to comprehend the peculiar nature of this sort of seeing—self-a lienated seeing—is from the outside: looking from sideways-on at a subject who exercises the capacity. Hence unlike Descartes’s thinker who is unable (while thinking his doubt) to doubt that he thinks, she who thus “sees” (while she is self-a lienatedly seeing) has no difficulty doubting that she sees. Unlike Kant’s logical I, who (in thinking p) thinks the necessary accompaniability of p by the “I think,” she who thus “sees” p (while she is in the act of self-a lienatedly seeing p) is unable to accompany p by the “I see.” Hence the perspective from which one must begin a proper philosophical elucidation of the logically peculiar concept of self-a lienatedly seeing is that of the third-person. From there one can work one’s way out to possibilities for employing the verb where it admits of conjugation in the first person. My ability to employ the verb for successfully seeing in the first-person present indicative form is a condition of its genuinely being a cognitive verb of success. Conversely, there being nothing that counts as having conjugated it in the first-person present indicative is a sign that the verb that McDowell’s parable’s protagonist employs is not—at least not in the sense of the term I seek to specify here—a verb of success. The self-a lienated sense that that latter verb bears flows from t here being literally nothing to think through the first-person present indicative use of the verb. There is no employment of it that can bear the logical form “I see p.” Indeed, it is her understanding of the nonavailability of such a form of thought that
134
R eas o n in P ercepti o n and A cti o n
enables McDowell’s parable’s protagonist to express, through the self- predicative form of use she does make of the verb, the logically peculiar thought she intends to express.
7. Resolute vs. Irresolute Disjunctivism Having arranged for such a use of the verb within the parable, when McDowell goes on to explicate “the” character of the thought his sample sentence expresses, the intended topic is the logical grammar of “seeing” as such. His aim is not to explicate just the character of the protagonist’s act of “seeing” within the frame of the parable. The aim is more general: to explicate the logical character of any act of seeing that can be expressed through the use of a verbal twin of the protagonist’s sentence—even when it is uttered by a reflectively favorably circumstanced speaker g oing about her ordinary life outside the frame of the parable. In outward sentential appearance, the deployments of the verbs for self-consciously and self-alienatedly seeing are identical, but not in inward logical form. If we fail carefully to attend to how such sentential verbal twins involve verbs for cognitive activity that are at best distantly related in logical form, then we will easily fall into the confusion of imagining that a single form of words expresses a common form of thought across these two contexts of use. Thinking this allows us to fall into a subtler version of the conjunctivist’s original confusion. We can rephrase a point made earlier about the conjunctivist as follows: the conjunctivist takes two uses of what outwardly is one and the same verb (“seeing”) to involve two different employments of one and the same concept, and hence to refer to a pair of ways of d oing one and the same t hing—merely seemingly d oing it and successfully doing it—where each of the adverbs modifies the same common core of visual activity, albeit it in two very diff erent ways. The conjunctivist then regards the logical features that the verb exhibits in one of these uses (in its logically derivative nonsuccess use) to involve a common core of visual activity that recurs as a highest common f actor in successful acts of seeing (where seeing p entails p). He concludes that all acts of seeing must involve such a logical moment of mere appearance.
R es o lute D isjunctivism
135
Now we can say this about McDowell’s version of disjunctivism: he takes two uses of what outwardly is one and the same verbal sign (“seeing”) to involve two different employments of one and the same concept (of seeing), and hence to refer to a pair of ways of doing one and the same thing. He then regards the logical features that the verb exhibits in one of t hese uses (in a logically alienated use: where I can see p without my accepting p) to exhibit a common core of visual activity that recurs in self-conscious acts of seeing (where my seeing p goes with my knowing that I see p and hence with my accepting p). He concludes from this that all acts of successfully seeing must involve a highest common logical moment of an act of seeing p without, in so seeing, thereby accepting p. What it is for me to know p in virtue of my seeing p thereby comes to be conceived as a form of cognitive achievement that is necessarily logically composite, comprising two acts: one of seeing and one of accepting p. I do not mean to deny that McDowell’s conception of the matter is far superior to that of the conjunctivist. But I do mean to raise a question regarding what it means strictly to think through the central insight that animates disjunctivism. McDowell’s conception of the matter is far superior because his conception of “a common core of visual activity” is nowhere near as logically anemic as that of the conjunctivist. McDowellian “seeing” improves on matters considerably because now seeing p entails p. But it still does not entail I know p. This is what allows t here to be a structural parallel between what McDowell wants to show the conjunctivist and what I want to show McDowell. Let us call the species of disjunctivism that McDowell arrives at through his critique of the Cartesian epistemologist irresolute disjunctivism in order to set the stage for this question: What shape would a fully resolute account of the insight underlying that disjunctivist critique have to assume? McDowell wants to show the conjunctivist that the two concepts of “seeing” with which the conjunctivist works stand in a logically asymmetrical relation to one another: one cannot work up to an account of perceptual success starting with the logically more anemic of t hese two concepts, where the first logical moment affords no more than mere Schein and the second supplements that common f actor with something that secures a relation to Sein. The form of seeing one self-ascribes in ascribing the logically anemic concept to oneself is not one to which something can
136
R eas o n in P ercepti o n and A cti o n
be added to turn it into a case of successfully seeing. Success is not something that can be merely added to merely seemingly seeing to yield a case of successfully seeing. The two logical moments of seeing a “mere appearance” of p and knowing p when conjoined do not jointly yield something that has the logical form of successfully seeing p. For, in such a case, their connection is not secured through the logical link constitutive of successfully seeing. If it is somehow secured at all, it w ill have to be through an assurance afforded by a mediating act of reflection, external to the act of “seeing” per se: one that renders it intelligible how there comes to be a coincidence of the p which figures in the so-called act of “seeing” with the p which figures in the subsequent act of knowledge—hence how it is that Schein happens to be in perfect alignment with Sein. (This is the role Descartes assigns to the idea of God, permitting the meditator reflectively to surmount the obstacle that would otherw ise forever block the very possibility of extracting knowledge from perception.) The ground of my knowing p, if it so obtained, is one that is not yet in place in (what is here called) my “seeing” p. It requires an additional source of ratification. It is precisely in this respect that it differs from the ground of the knowledge of p that one obtains if one successfully sees p. Now the resolute disjunctivist wants to show the irresolute disjunctivist something structurally parallel: self-a lienatedly seeing is logically less full-blooded than self-consciously seeing. It still severs the logical link between, for example, uses of the verb in the first-person past and in the first-person present indicative. Given the logically fundamental character of the severed inferential link, the overlap in surface grammar does not warrant an inference to an identity in the logical form of the concepts expressed by these diff erent uses of an outwardly identical verb. This asymmetry suffices to reveal that the concept of rational cognitive activity expressed by the logically more anemic verb involves a dimension of logical privation. The protagonist in McDowell’s parable both sees and knows, but when she “sees” p, she does not know either p or that she sees p. The self-ascription of such a form of seeing is only possible if it is ascribed either retrospectively (in the first-person past form) or imaginatively (in the subjunctive mood) by situating “oneself” within a very particular sort of frame. In e ither case, the form of seeing one thereby self-a scribes is not that of self-consciously seeing. Self-consciousness is not something that can be merely added to an act of non-self-consciously seeing to
R es o lute D isjunctivism
137
yield a case of self-consciously seeing. The two logical moments of (what McDowell calls) “seeing” p and “accepting” p, when merely conjoined, do not jointly yield something that has the logical form of self-consciously seeing p. For, in such a case, their connection is not secured through the logical link constitutive of self-consciously seeing. If it is somehow secured at all, it w ill have to be through an assurance afforded by a mediating act of reflection external to the act of “seeing” per se, that renders it intelligible how there comes to be a coincidence of the p which figures in the so-called act of “seeing” with the p which figures in the subsequent act of knowledge. (This is the role McDowell assigns to reflection in his narration of the parable, so his protagonist may reflectively surmount what would otherwise block the very possibility of her extracting perceptual knowledge from her circumstances.) The ground of my knowing p, if so obtained, is not yet in place in (what is here called) my “seeing” p. It requires an additional act of ratification. It is precisely in this respect that it differs from self-consciously seeing p. As with my earlier formal use of the locution “successfully,” my pre sent use of “self-consciously” w ill prove misleading if it is thought to introduce a modification into what otherw ise would be a logically less complex description of the act in question. H ere, too, we have a use of an apparently modifying expression that does not serve further to logically determine the concept it modifies, but rather to highlight a formal aspect of what is involved in the logically full-blooded case of seeing. The elucidatory role of the word “self-consciously” in this context—like that of “successfully” above—is to make explicit the absence of privation along a certain dimension in the logical form of the exercise of a rational cognitive activity. As before, so here too, the formal concept of self-consciously φ-ing functions as a philosophical device for bringing clearly into view yet a further dimension of what simply φ-ing comes to. Let us now bring together the following three points: (1) the role of a properly unqualified concept of cognitive success is to render philosophically perspicuous what it is for the capacity under investigation to operate fully spontaneously (free of any dimension of privation); (2) the concept of self-a lienatedly seeing p introduces a dimension of privation into the form of what it is to logically full-bloodedly see p; and (3) on the irresolute disjunctivist’s conception of perceptual success, self-alienatedly seeing p qualifies as a case of (what he calls) a “successful”—that is, a
138
R eas o n in P ercepti o n and A cti o n
putatively in no way defective—exercise of the capacity to see p. It follows from these points that the irresolute disjunctivist’s account of perceptual success fails to yield an adequate account of the concept we ought here to take as our starting point, if we wish properly to gather the logical form of the rational capacity u nder investigation. Just as the irresolute disjunctivist tries to show the conjunctivist that what figures as two logically separable moments in his analysis of success is really just one logical moment in an act of successfully seeing, so too the resolute disjunctivist will try to show the irresolute disjunctivist that what he analyzes as separate logical moments form one logical moment in the logically full-blooded case of success. Just as the irresolute disjunctivist tries to show the conjunctivist that the only case that ought to be accorded the honor of being sorted into the good disjunct is the one in which seeing p can serve as an indefeasible ground for knowing p, so here again: the resolute disjunctivist will try to show the irresolute disjunctivist that only if self-consciousness of the indefeasibility of the ground figures in perceptual consciousness of the ground do we have a case of an act of the capacity that o ught to be accorded the honor of being sorted into the most august disjunct. Just as the irresolute disjunctivist says to the conjunctivist: “What you take to be mere appearance does not stand to perceptual knowledge as the part stands to the w hole but rather as the bad does to good,” so too the one sort of disjunctivist says to the other: “What you take to be seeing does not stand to self-consciously knowing as part stands to whole, but rather as self-a lienation stands to self-consciousness.” As the one says to the other: “The proper understanding of what you take to be the logically complex success case (of appearance plus reality) needs to have the plus removed from it if we are to comprehend the logically basic act of the capacity; and what you take to be the logically s imple case (mere appearance) is the logically privative form of exercise of the capacity,” so too now the one sort of disjunctivist says to the other: “The proper understanding of what you take to be the logically complex success case (of seeing plus accepting) needs to have the plus removed from it if we are to comprehend the logically basic act of the capacity; and what you take to be the logically simpler case (something afforded through an exercise of one’s perceptual capacity to which a judgment stroke needs to be attached) is the logically privative form of exercise of the capacity.” On the resolute conception, t here are not two logically distinct acts that share a common content—one of seeing and one of knowing—rather,
R es o lute D isjunctivism
139
t hese are two dimensions of a single self-conscious act. Qua logical aspects of a single self-conscious exercise of the capacity, they are merely notionally, never r eally, separable. Just as the disjunctivist says to the conjunctivist: “In the nondefective exercise of the capacity, your seeing the tie to be green does not fall in any way short of the green tie that you see,” now the one sort of disjunctivist says to the other: “In the fully nonproblematic exercise of the capacity, your seeing the tie to be green does not fall in any way short of you knowing it to be green. In the logically basic exercise of the capacity, t here is neither room nor need for a logical separation of your seeing p from your accepting p.” As we noted earlier, when McDowell speaks of “seeing” in the passage quoted above, he wants his use of the word to comprehend the kind of “seeing” he illustrates with his tie-shop parable (in which seeing p and knowing p admit of the sort of separation of logical moments that cases of self-alienatedly seeing plainly exhibit), while also wanting his deployment of the verb not to refer just to such cases. He also wants what he means when he speaks t here of “seeing” to comprehend logically full-blooded self-conscious acts of seeing as well. He wants his use of the word “see” to cover both of these cases as if they comprised two species of a single formal genus. I have tried to show there is no such genus. If this is right, the relation between the two sorts of seeing that McDowell’s use of the verb “seeing” yokes together can be specified only disjunctively. Their logical relation can be properly elucidated only through a comprehension of how that which is logically constitutive of the primary case (self-consciousness) undergoes a form of privation (self-alienation) in the envisioned logically derivative employment of the verb. Much earlier, we saw the disjunctivist level the following critique of the conjunctivist: “What you construe as a highest common factor is a philosophical illusion engendered by your failing to distinguish properly between the cases of merely seemingly seeing and successfully seeing—an illusion that can be dissipated through a grammatical investigation of the relations of logical priority that obtain between t hese verbs and hence between the correlative (defective vs. nondefective) forms of exercise of a cognitive capacity.” This schema may be redeployed. The one sort of disjunctivist may say to the other: “What you construe as a highest common f actor (namely, the supposedly always logically separable moment of ‘perceptual experience’ to or from which I can give or withhold assent) is a philosophical illusion partially engendered by your failing to distinguish properly between your
140
R eas o n in P ercepti o n and A cti o n
protagonist’s self-alienatedly seeing and someone’s self-consciously seeing— an illusion that can be dissipated through a grammatical investigation of the relations of logical priority that obtain between these verbs and hence between the correlative (logically full-blooded vs. logically comparatively attenuated) forms of exercise of a cognitive capacity.” Just as there is no highest common factor between merely seemingly seeing and successfully seeing, so too t here is no highest common factor between self-alienatedly seeing and self-consciously seeing. Self-consciously seeing is not a form of cognitive capacity one comes to exercise by supplementing an act of seeing with a further act of self-consciousness. We have seen that McDowell’s version of disjunctivism admits of room for a highest common factor across the following two cases: the case of (what an analytic philosopher w ill call) “the veridical perceptual experience” that I am having now and to which I give my assent and the case of “the veridical perceptual experience” that I am having now and from which (due to reflectively unfavorable circumstances) I withhold my assent. On such a conception, what figures here as an exercise of a capacity to do something called “having a veridical perceptual experience” amounts to something that is less than an act of knowledge. The capacity for “perceptual experience” so conceived is not a concept of a self-conscious cognitive capacity. It is at best a concept of a form of exercise of the capacity that could be ingredient in a logically attenuated case of arriving at knowledge through an act of perception mediated by one of reflection. For perceptual experience, so conceived, only eventuates in knowledge if I do something beyond enjoying the experience: if I (paraphrasing Frege) advance from a grasp of the content of the thought delivered up to me through the perceptual experience to an acknowledgment of its truth- value. This, in turn, yields a conception according to which the exercise of my capacity to gain knowledge via perception stands to my capacity to have veridical perceptual experiences as something logically composite to something comparatively logically s imple. On McDowell’s version of disjunctivism, whenever I know via seeing, there are necessarily two logically (though not necessarily temporally) distinct moments: the one in which I see and the one in which I judge what I see to be so. A resolute disjunctivism w ill seek to expunge from its form of account this residual dimension of logical conjunctivism in the characterization of the logical form of the act of the capacity that is fully f ree of any modicum of defect or privation.
R es o lute D isjunctivism
141
8. Felicitous and Infelicitous Disjuncts vs. (Merely) Good and Bad Ones here are two philosophical lessons we can learn from Kant upon which T I have drawn in the foregoing discussion—one about the relations of our general cognitive powers to one another and another about the relation of act to power. The first of these lessons is that in the case of a self-conscious capacity, such as our capacity for perceptual knowledge, the operation of the capacity is not to be factorized into the joint operation of two self- standingly intelligible capacities—say, one of perceiving and one of judging. The second lesson is that the nondefective case of the exercise of such a capacity is not to be analyzed as (what I called above) an attenuated case of its exercise plus something. This is close to the lesson I extracted above from McDowell’s own critique of Cartesianism, but it cuts slightly deeper. Putting the various lessons together, we get this: to perceive (in the relevant success sense of the term) is to know (and to fail to know is for the capacity in question to encounter or suffer some kind of encumbrance, attenuation, or defect in its exercise). If we bring these lessons to bear on disjunctivism, then we arrive at the following reformulation: When I am an occupant of the good disjunct, the act of knowing does not differ from that of perceiving—there is one act; when I am an occupant of the bad disjunct, the exercise of my perceptual faculty is not per se an act of knowledge—if I am to acquire knowledge, further acts of mind are required. It is this feature of the latter disjunct that the Cartesian tries to read into the structure of even the best case of knowledge. We saw above that Cartesian skepticism results from the challenge that it is impossible to maintain both of the following two thoughts at once: (1) In virtue of perception, we can come to know about the world around us. (2) A subject currently having a veridical perception is having an experience that is exactly the same as the experience she would be having w ere she instead having an indistinguishable nonveridical perception. The way of understanding the relation of the good to the bad case now under consideration retains the original virtue of disjunctivist analyses:
142
R eas o n in P ercepti o n and A cti o n
it blocks the reading of (2) that makes it seem incompatible with (1). But it provides further clarification on how properly to understand (1), so as to yield a principle for distributing cases across disjuncts that differs from the one that McDowell’s version of disjunctivism invites us to adopt. In McDowell’s discussions of these matters, the good case is thought to be one in which I both perceive and know, and the bad case is thought to be one in which I perceive but do not know. The label of “goodness” may incline us to sort the cases in accordance with whether the story ends well hether the for the occupant of the disjunct or not—in accordance with w exercise of her faculty issues in knowledge or not. The resolute disjunctivist offers a different account of wherein the true virtue of (what McDowell calls) “the good disjunct” o ught really to be taken to consist. The proposed reconception of the disjunctivist schema turns on a more discriminating principle for sorting cases into some disjunct other than the optimal one. On McDowell’s version of disjunctivism, the philosophical joint at which the disjunctive distinction is carved is one that takes the crucial criterion to be a m atter of whether the occupant of the disjunct can be said by a third-person observer to be in a position to know. Does this suffice to vindicate the philosophical aspirations of disjunctivism? Does it suffice that from some (for example, later, or other) point of view on the perceptual appearances of the perceiver she can be said (by herself or someone e lse) to have been in “the right position” for knowing—or must, as the resolute disjunctivist insists, the (so-called) point of view from which this can be said be in the logically strongest possible sense hers? Let us henceforth refer to the two fundamentally different sorts of disjunct—one that is utterly f ree of any form of privation and one that is not—that the resolute disjunctivist’s alternative principle of classification yields as a felicitous disjunct and a nonfelicitous disjunct. According to our reformulated principle, I qua perceiver count as an occupant of an infelicitous disjunct, not only if what t here is to know and what I seem to perceive do not coincide, but also when they do not conflict (for example, in the manner allowed for by McDowell’s parable, where I both seem to “see” p and p is the case), yet my act of perceiving and my act of knowing are disjoined in such a way that in (doing something a philosopher may want to call) “perceiving” p, it still remains the case that I myself do not self-consciously perceive p. Any scenario, regardless of how “good” its final
R es o lute D isjunctivism
143
outcome might be, in which my act of perceiving and my act of knowing come apart (so that an exercise of additional capacities is required in order to arrive at knowledge) counts according to our new principle as a case to be sorted under the infelicitous disjunct. As we have seen, resolute disjunctivism holds that there is no room for a highest common factor across the following two cases: the case in which I know in seeing (where the act of assent is not dissociable from the act of seeing) and the cognitively attenuated case in which I am able to come to know only through the supplementation of an initially logically attenuated act of perception with a further effort of mind, so that “perception” eventuates in knowledge through the mediation of reflection. On this alternative conception, the difference between the two disjuncts must now be specified in terms of the character of their logical form. The logically primitive form of exercise of the capacity (in which perceiving is, as such, an act of knowledge) represents the first disjunct. Any other form of its exercise that in any way attenuates or disrupts (and hence lacks accord with the form of) the capacity is to be classified as belonging to some further disjunct. Only now, we can no longer hold that there are just two disjuncts: the good and the bad one. To paraphrase Tolstoy: felicitous disjuncts are all alike in being felicitous, whereas each sort of infelicitous disjunct is infelicitous in its own way. On this conception, perceptual error is only one of the grosser and more extreme forms of perceptual infelicity. For t here are many ways to deviate—to lack accord—w ith the form of felicity.26 It suffices to qualify as an occupant of the infelicitous disjunct if you are someone whose exercise of a perceptual capacity does not as such result in knowledge. This was, of course, true of the erstwhile occupant of the bad disjunct at issue in the original critique of conjunctivism: that “perceiver” suffers from a perceptual misapprehension without knowing that he so suffers. Hence, we may say of him (viewing him from the third- person point of view) that he is unable to distinguish the situation he is in from the one he takes himself to be in. But to say this, according to the new principle of classification, is not yet to furnish a full account of what places him into the infelicitous category. It is not simply his inability to discriminate his own situation from one of knowing (making it the sort of case upon which the Cartesian fixates in his preferred examples of defective exercises of a perceptual capacity). What places him in
144
R eas o n in P ercepti o n and A cti o n
the infelicitous category is a distinction that the Cartesian is unable to get into view. This classic occupant of the original bad disjunct is one of a whole possible range of possible cases of infelicity. Let us call the Cartesian’s favorite variant of infelicity the there-is-no-way-to-tell variant of the infelicitous disjunct. The standard disjunctivist diagnosis brings out how the conjunctivist illicitly contrives to make it seem as if the case of the good disjunct shared a highest common factor with the there-is-no-way- to-tell variant of the infelicitous disjunct. This is a philosophical step in the right direction, but for the resolute disjunctivist, the there-is-no- way-to-tell variant of the infelicitous disjunct does not exhaust the category of infelicity. What makes something a case of the felicitous disjunct is not that I both perceive and (also) know; it is that I am sufficiently favorably circumstanced to know simply in virtue of perceiving. It does not suffice to disqualify a case as one of infelicity, according to the resolute disjunctivist’s principle, simply to note that I am sufficiently well circumstanced that my perceptual engagement with the world is of a sort that it can serve as a step along the path from ignorance to knowledge. If the exercise of the capacity only “eventually”—rather than immediately— issues in knowledge, then it is not free of infelicity in the relevant sense. So a whole range of further cases of forms of exercise of my perceptual capacity w ill count as cases of infelicity according to the resolute disjunctivist’s principle. It w ill include, for example, any sort of case in which I have reason to stand back from a perceptual judgment that I might otherwise be inclined to make because of what I take myself to have reason to question about how I am circumstanced, reflectively or otherw ise. Indeed, it w ill include any sort of case in which a gap can open up between perceptual taking and judging, even if the path to knowledge is one that may be extremely easily disencumbered of its initial obstacle—as, for example, in certain situations in which things are just as they appear but in which further reflection is required to credit that t hings really are as they appear. I am thinking h ere of a sort of case for which philosophers love to arrange, one in which I later learn that the circumstances under which I perceive turn out not to be unfavorable in a way that I initially had reason to surmise they might be. Consider the following case: I point at features of something that I take to be the Müller-Lyer–style diagram I drew on the blackboard of my classroom at the beginning of the class and say to my students, “This line
R es o lute D isjunctivism
145
merely appears to be longer than that one,” strongly emphasizing the word “appears.” While doing so, I, on the one hand, simply assume that it is the very same diagram that I drew earlier (and thus that it is, in the usual way, occasioning an optical illusion in me), while, on the other hand, I find myself thinking that the one line r eally does now seem to be so much longer than the other that it is proving even harder than usual for me to credit that they are, indeed, of the same length. I then realize, aided by their mischievous laughter, that the students of my introductory epistemology class have, during a very brief absence on my part from the classroom, contrived to alter the diagram (lengthening the apparently longer line) and that they have been eagerly awaiting the moment when I would again say, while pointing at the line in question, “This line merely appears to be longer than that one.” In this sort of case, though I end up with knowledge (in this case that the one line is really—a nd not merely seemingly—longer than the other), it is not simply in virtue of “perceiving” what is the case: I do the (business we are here calling) “perceiving,” I hesitate, I marshal further considerations, and then I arrive at knowledge. In such a case, I both “perceive” and (in the end) I know, but the original act of “perceiving” is insufficient for the knowing. Let us call such a case the insufficiency- of- the-act-of-the-perceptual-capacity-for-telling variant of the infelicitous disjunct—or, for short, the insufficiency-of-the-perceptual-act variant. Something qualifies as belonging to this variant only if both of the following features obtain: First, unlike in cases in which blue merely looks green to the subject, in this case, the subject is (as McDowell likes to put it) “in a position” to conclude that things are just as they appear. Second, the subject in this variant comes to know things are just as they appear ere lies not (as in through an additional act of reflection.27 The problem h the classic Cartesian version of the bad disjunct) in the fact that the subject misperceives or fails to be in a position to know, but in how she comes know. What is h ere called “perceptual knowledge” is a form of coming-to-k now. It exhibits a logically processual character, comprising phases beyond that of the initial act of the capacity, such that the path from “perceiving” to “knowing” admits of possible interruption, for the overall achievement of knowledge requires a striving that exhibits the logical form of inquiry, comprising moments that bear imperfective and perfective aspect. The form of exercise of a cognitive capacity that is properly classified as
146
R eas o n in P ercepti o n and A cti o n
belonging to the felicitous disjunct for the resolute disjunctivist must not exhibit these logical features. Hence an employment of the verb “to see” that expresses the logically full-blooded act of self-consciously seeing is also one that does not admit of the imperfective form qua act of knowledge in its first-person present indicative use. The resolute disjunctivist’s ground for classifying the aforementioned insufficiency-of-the-perceptual-act case under the infelicitous (rather than felicitous) disjunct rests on the irresolute disjunctivist’s original reason for classifying cases under the bad (rather than the good) disjunct: namely, the insistence of the absence of a highest common factor across the felicitous disjunct and any infelicitous disjunct. To say that even this scenario ought to be sorted as a variant of an infelicitous disjunct is to claim that t here is no highest-common-ingredient act across the following two exercises of my perceptual capacity: (1) one in which I know that a is F in virtue of an exercise of my perceptual cognitive faculty, and (2) one in which it appears to me that a is F in virtue of an exercise of my perceptual cognitive faculty, but in which I only come to know that a is F—and thus come to know that I do not thereby perceptually misapprehend what is the case—through supplementing that act of perception with one of reflection. The temptation here, even for those who otherw ise take themselves to have digested the lesson of disjunctivism, is to say that t here a common act is ingredient in both of these exercises of my cognitive power—namely, “the” act that brings it about that things perceptually appear as a case of a’s being F when a really is F. The temptation is to think that if “the” appearance is “the same,” and there is no misapprehension “in” the appearance, and things are as they appear, and so I come to know that things are as they appear b ecause it is the t hings themselves that so appear to me, then all is fine: what more could you want? A ren’t all of the crucial elements for qualifying something as a case of epistemic goodness of a sort that we should care about in place? The point of the new principle of classification is to insist that the two cases (the nondefective and the somewhat attenuated exercise of a cognitive faculty) differ in Kantian transcendental-logical form.28 What we want to call, under the pressure of the aforementioned temptation, “the” act (allegedly equally ingredient in these two cases of exercising the same general capacity) is not a common act: in one case, the act in question is qua the very act that it is
R es o lute D isjunctivism
147
an act of knowledge; in the other, it is an act of the power of knowledge which, as such, is insufficient for knowledge. The underlying error in construing these two exercises of the perceptual capacity as having a highest common factor is au fond the same as the one that animates conjunctivist analyses: that of reading what comes to be a self-standingly intelligible moment in the structure of a less than fully felicitous exercise of the capacity into the structure of the logically more fundamental case of its fully felicitous exercise, thereby decomposing the unity of the fundamental form of act of the power. Now we can say, more precisely, what qualifies any case as one that ought to count as a variant of the infelicitous disjunct: the question is not whether I can be said to both perceive and (also) know, but w hether my acts of perception and knowledge form the requisite sort of unity or not. So, even if I end up knowing the truth and even if the exercise of my perceptual faculty figures as a crucial first step in my eventually getting t here, if the act of exercising my capacity for perceptual knowledge amounts to, qua the very act that it is, something less than an act of knowledge, then my knowledge is not grounded in a logically full-blooded act of Kantian perceptual judgment—but rather only in (what Kant would call) a problematic perceptual judgment, an act of the capacity that requires nonperceptual supplementation to issue in knowledge.29 The fundamental point here is just an application of the more general point about the logical priority of power to act in a Kantian account of the form of our cognitive power. What we need to understand is not simply how to sort cases into those in which I both exercise my perceptual capacity and know and t hose in which I exercise that capacity but fail to know, but rather how properly to differentiate between the formal character of the logically prior case of the exercise of the faculty (in which it suffers no attenuation or interruption) and its range of defective or attenuated forms of exercise (some of which may nonetheless eventuate in knowledge). The original disjunctivist diagnosis charges the Cartesian with illicitly equating a supposed feature of the case of the good disjunct with an a ctual feature of the there-is-no-way-to-tell variant of the infelicitous disjunct, thereby leaving the poor Cartesian with no way to tell knowing from nonknowing. This diagnosis, as indicated above, is fine as far as it
148
R eas o n in P ercepti o n and A cti o n
goes, but it is apt to leave the irresolute disjunctivist imagining that he has fully learned the lesson of disjunctivism while continuing to categorize the case of the insufficiency-of-perception-for-telling scenario as a fully happy case on the grounds that its occupant both “perceives” and knows. What the resolute disjunctivist’s more penetrating diagnosis suggests is rather this: the Cartesian too closely assimilates the difference between the defective and the nondefective exercise of a perceptual faculty to the difference between the following two variants of the infelicitous disjunct—the there-is-no-way-to-tell variant and the insufficiency- of-the-perceptual-act variant. The optimal case of perceptual knowledge that the Cartesian is able to envision (a fter we receive the required assistance from God) is one that consists in a form of exercise of a cognitive capacity that nevertheless retains the infelicitous structure of the insufficiency-of-perceptual-act case. What the Cartesian lacks and what a properly prosecuted disjunctivist diagnosis of his error should yield is a properly delimited concept of a fully felicitous exercise of our perceptual capacity. T here is nothing wrong with the idea that mundane cases of the insufficiency-of-the-perceptual-act-for-telling scenario exhibit just the order of epistemic priority that the Cartesian thinks every act of perceiving must stand in to an act of judging in order to result in knowledge. The problem with the Cartesian conception is that it writes this order of priority into the very idea of what it is to know something via perception. The Cartesian conception makes it look as if every case of perceptual knowing requires additional reflective entitlement in a sense in which the insufficiency-of-t he-perceptual-act-for-telling scenario actually does. So when we turn to the best case of perceptual knowledge, one in which no further reflective entitlement is required, it still looks as if, absent an act of judgment that accepts what the act of our perceptual capacity delivers up to us, we remain without something further we need in order for perception to yield knowledge. This is the feature of the Cartesian conception that the resolute disjunctivist seeks to bring within his target range: its conception of the composite character of even the best case of perceptual knowledge, so that even such a case must partake of a structure that enfolds a pair of logical moments—one in which something is given to the perceiving subject and one in which the subject accepts what is thus given.
R es o lute D isjunctivism
149
9. Fully Abandoning the Sideways-On Point of View In the foregoing discussion—in order to bring out the contrast between resolute and irresolute disjunctivism—the focus has been almost exclusively on two ways of departing from felicity: the fully defective exercise of the capacity and the self-a lienated one. The focus on the latter was in service of the following point: in a felicitous exercise of the capacity, t here is no room for a Fregean dissociation of the act of perceptual thought from that of perceptual judgment. This means that, as we have seen, cases that qualify for the irresolute disjunctivist as sortable u nder the good disjunct do not qualify for his resolute counterpart as sortable under the perfectly felicitous disjunct. For in the fully felicitous case, t here can be no logical separation of moments between what I see and what I judge to be so. McDowell’s own characteristic criticism of a g reat deal of modern philosophy is that it turns on a methodological fantasy of the priority of (what he suggestively calls) a sideways-on point of view on the relation of knowing subject to known world. Against this, he insists upon the fundamental methodological priority of those forms of knowledge that are available to us from within our cognitive perspective onto the world. The manner of reconceiving the fundamental insight underlying disjunctivism proposed here grows out of an attempt to execute as resolutely as possible this very methodological turn in philosophy that McDowell himself has so eloquently proposed and deftly defended. The reconception turns on an insistence that the logical priority upon which McDowell himself insists turns on a further dimension of priority that is logically nested within it: the availability of my first-person present indicative point of view onto my cognitive activity as a prior condition on the intelligibility of any other (third-person, and / or past, and / or subjunctive) point of view onto that same activity. McDowell’s original point against the conjunctivist is that we should not read the logical structure of the defective exercise of a cognitive capacity into that of its nondefective exercise (so that we end up conceiving of the successful case of knowledge as consisting of a composite of a factor of mere appearance and a further factor that somehow secures a relation
150
R eas o n in P ercepti o n and A cti o n
to reality). I suggested e arlier that McDowell, in so arguing, seeks to adhere to the following philosophical maxim: Do not try to read the character of the logically primitive phenomenon off the model of its logically alienated counterpart! It is on precisely these grounds that I have been concerned to argue that we should also not read the logical structure of what it is for me to have seen p under reflectively unfavorable circumstances into the nder fully felicitous circumstances, structure of what it is for me to see p u so that even in our account of the nonproblematic case we continue to take it that t here is an isolable (albeit veridical and genuinely world-involving) appearance awaiting combination with a logically distinct operation of judgment. This reconception of the relation between the logically primary disjunct and its comparatively logically privative siblings yields a differ ent—a nd, I contend, deeper—diagnosis from McDowell’s original one of wherein the seemingly innocent first step in the Cartesian philosophical conjuring trick actually lies. It lies not merely in the philosophically mistaken supposition that there is a logical symmetry across knowledge- yielding and non-k nowledge-y ielding ways of jointly exercising my cognitive capacities (though McDowell is absolutely right to think that there is a fundamental asymmetry and hence no highest common f actor to be found across these cases). It lies more deeply in the philosophically mistaken supposition that t here is a logical symmetry across fully self-conscious exercises of my cognitive capacities and less than fully self- conscious exercises of my cognitive capacities. The recommended conception accordingly holds that there is a fundamental logical asymmetry between those forms of knowledge that can be expressed through the use of a cognitive verb—for instance, know, judge, see, hear, etc.—used in its first-person present indicative grammatical form (and hence from what a philosopher may be inclined to call the first-person present indicative “point of view”) and those forms of knowledge that involve some dimension of self-alienation (and hence express a form of knowledge that is in the first instance available only from some other point of view). To appreciate why an exercise of our perceptual capacity for knowledge that issues in a judgable but unjudged “experiential content” is a case of its infelicitous exercise is to become fully clear about what is philosophically required in order not to fall back into even the subtlest of versions of a highest-common-factor conception of experience. This is the
R es o lute D isjunctivism
151
point that I take the following series of remarks from Barry Stroud to be making: The fact that judgments made in perceptual experience are revisable in the light of reflection or further experience can make it seem that continued acceptance of a perceptual judgment about the independent world cannot be based solely on the original experience that gave rise to it; further reflection or experience serves to support it as well. And this can make it tempting to suppose that the content of that original experience in itself must have implied nothing about the independent world. . . . . 30 But yielding to this temptation is another way of falling into the “highest common factor” view of perception and reasonable belief that McDowell has rightly rejected. . . . [T]he content of an experience alone cannot give a person reason, or be a person’s reason, to believe something. . . . It is not simply the content of a person’s experience that gives the reason to believe something; it is the person’s experiencing, or being aware of, or somehow “taking in” that content.31 This last remark could be paraphrased in our e arlier idiom as follows: the perceptual (Fregean) thought, considered as mere judgable content, cannot, apart from the perceiving subject’s recognizing it as true, amount to an act of knowledge; rather, it is only with the perceiving subject’s perceptual judgment that such and such is the case that the subject commits herself to occupying a determinate position within the space of reasons. Only if the capacity for judgment can be fully in act in perception (in the manner indicated by Stroud) can perception as such be a capacity for knowledge. The moment our description of what perception yields begins to shade into a description of anything less than this—into a description of the mere deliverance or “presentation” of something judgable which the judging subject may hold at a distance and regard as not yet constituting knowledge until a further active step is taken on her part—then what we have offered is, by the lights of the resolute description, a description of a form of exercise of our perceptual capacity for knowledge that ought to be sorted into a disjunct other than the one that we reserve for fully felicitous exercises of the capacity. Admittedly, it is not a “defective” exercise of the capacity, if what that means is that it is
152
R eas o n in P ercepti o n and A cti o n
one that results in error—say, in the mistaking of how things merely perceptually seem for how t hings are. However, it is also not a fully felicitous case of the exercise of the capacity because it at best merely puts us in a position to know—the act itself is not one of knowing. The form of the exercise of the capacity that should be accorded pride of place as fully felicitous, according to the resolute disjunctivist, is the one that reveals (what Kant calls) the form of the general capacity. The general capacity here at issue is one whose manifold of possible forms of actualization w ill be distributed across the manifold of possible disjuncts that a more complete and perspicuous form of disjunctivist account w ill need to furnish, once the disjunctivist begins to appreciate the importance of no longer simply equating the idea of freedom from perceptual error or illusion with the very idea of a best case of perceiving. For what a properly executed version of disjunctivism should want the sort of case that is sorted under its most privileged disjunct to reveal is the following: (what Kant calls) the logical form of the capacity. This is the form the capacity exhibits when it is able to operate under optimal conditions, hence fully unimpeded and unencumbered. It is only through such a case that its logically most basic form of exercise can be revealed. It is such a case—in which the form of the exercise is in perfect accord with the general form of the capacity— that a disjunctivist ought to call the truly good case.32
Chapter Five ✣
Control and Knowledge in Action Developing Some Themes from McDowell ✣
Markos Valaris
I
n the philosophy of perception, John McDowell has long championed an approach on which perceptual experience is to be understood as the exercise of conceptual capacities (understood as the capacities characteristic of self-conscious rationality) in sensory consciousness.1 More recently, he has championed what he himself sees as a parallel approach in the philosophy of action—an approach on which “physical activity can be rationality in action, as opposed to a mere result of exercises of rationality.”2 It is often assumed that h uman thinking and rationality are purely inner matters, that is, that they are constitutively indepen dent of bodily events (beyond the central nervous system, perhaps). On such a view, the bodily events that constitute our actions are, as such, merely contingent causal effects of suitable mental states and episodes.
153
154
R eas o n in P ercepti o n and A cti o n
McDowell’s suggestion, by contrast, is that we should think of our intentional actions themselves as “exercises of rationality”—in other words, as instances of thinking and reasoning of no lesser status than their inner counterparts. I find McDowell’s suggestion for an embodied and extended conception of the mind in action highly attractive. My aim in what follows w ill be to motivate and develop a way of thinking about some aspects of agency that follows this broad template. I begin in Section 1 by discussing an apparent puzzle about human agency that David Velleman has recently highlighted. 3 According to Velleman, what makes actions characteristically actions conflicts with what makes (normal, adult) human agents characteristically human. Overt actions are, according to Velleman, fundamentally bodily movements controlled by the agent. And what makes h uman agents distinctively human is a kind of self-consciousness—what Velleman calls “reflective consciousness.” 4 Unfortunately for us, however, these two conditions turn out to conflict with each other: reflection tends to get in the way of our meeting the demands of agential control, and so of agency as such. In contrast to Velleman, I do not think that there is inherently a conflict between reflection and agency. Self-consciousness does not have to get in the way of agential control. But this is not to say that Velleman’s concerns are unmotivated. Indeed, there is a seemingly natural way of understanding agential control—a picture that Velleman himself clearly accepts—on which the tension that Velleman identifies seems unavoidable. As I w ill be suggesting in Section 2, however, the lesson we should draw from this is not that human agency is inescapably alienated: it is, rather, that this seemingly natural conception of agential control ought to be rejected. But if this picture of agential control is rejected, what should we put in its place? The trouble with the picture of agential control that Velleman presupposes, and which I criticize, is precisely that it seems to assume the interiority of the agent’s mental contributions to their own actions. What the agent’s mind contributes to the control of action, in other words, falls short of the overt actions themselves. Following McDowell, I suggest that this is precisely the feature of this view that we need to reject. In Section 3, I sketch an account of agential control that does so.
C o ntr o l and K n o wledge in A cti o n
155
1. A Puzzle about Human Agency My aim in this section is to get clear on the puzzle about human agency that Velleman identifies. I w ill consider possible responses in the following section. The first step involves formulating general conditions for agency. As Velleman (following Frankfurt) points out, agency is not a uniquely h uman 5 phenomenon. It is not just that words such as “action” and “agency” (and their natural correlates in other languages) can be used very broadly, to characterize, for instance, the erosion of the coastline by the sea, or the effects of acid on some metal. More relevant to our purposes is the apparent fact that, in addition to this extremely broad sense of “action,” t here seems to be another, rather more robust one, which nonetheless applies to animals in general, not just h umans. As many philosophers seem to agree, there is an import ant sense in which both humans and, say, spiders are agents, while the ocean and a vat of acid are not.6 What, then, are the conditions for agency, in this sense? One answer, suggested by Frankfurt and adopted by Velleman, focuses on the idea of guidance or control.7 It is in virtue of being in control of (some of) the happenings in and around its body that a spider or a h uman being count as agents (in this robust sense). Conversely, it is in virtue of being guided or controlled by an agent that certain bodily happenings are actions: “Given a bodily movement which occurs under a person’s guidance, the person is performing an action . . . A nd he is not performing an action if the movements are not u nder his guidance as they proceed, even if he himself provided the antecedent c auses—in the form of beliefs, desires, intentions, decisions, volitions, or whatever—from which the movement has resulted.”8 Broadly speaking, the suggestion that we should think about agency and action in terms of control seems correct; in any case, I do not plan to question it here.9 But, once again, what is it for an agent to be in control of their bodily movements in the relevant sense? Velleman continues to follow Frankfurt on this: Behavior is purposive when its course is subject to adjustments which compensate for the effects of forces which would otherw ise
156
R eas o n in P ercepti o n and A cti o n
interfere with the course of the behav ior . . . The behav ior is in that case u nder the guidance of an independent causal mechanism . . . The activity of such a mechanism is normally not, of course, guided by us. Rather it is, when we are performing an action, our guidance of our behavior. Our sense of our own agency when we act is nothing more than the way it feels when we are somehow in touch with the operation of mechanisms of this kind, by which our movements are guided and their course guaranteed.10 In this passage, Frankfurt sketches a two-tiered account of agential control or guidance. On the first tier, t here is guidance by “independent causal mechanisms,” which suffices for purposive behavior but not, on its own, for action. To get to full-blown action (the second tier), Frankfurt suggests, it suffices to be “somehow in touch with the operation of mechanisms of this kind.” How should this “somehow” be spelled out? Velleman suggests that the answer is phenomenological: it has to do with how agency is “felt” or experienced.11 Since our discussion is meant to cover spiders as well as humans, however, this seems unwise: an account of agency should not commit us to any particular view about what it is like to be a spider, a fter all. For now, it seems better to work with the following schematic idea: Two-tiered conception of control: Agential control consists in being somehow in touch with the operation of independent causal mechanisms that guide our bodily movements. The question of how exactly to fill in the details w ill matter later on. Now, according to Velleman, this way of thinking about what it is to be an agent turns out to conflict with the distinguishing mark of specifically human agency, that is, with what he calls “reflective consciousness.”12 While this term is not explicitly defined, it is clear that what he is talking about is a capacity for self-consciousness, understood as a capacity for being aware of one’s own actions, along with a capacity for reflecting upon the reasons for them. The trouble, according to Velleman, is that self-consciousness inevitably alienates us (or, to use his term, “disidentifies” us) from our actions.
C o ntr o l and K n o wledge in A cti o n
157
Velleman gives a few reasons for thinking this. To begin with, the content of our reflective thoughts about our reasons for acting seems misaligned with the goals of the corresponding actions: if I am thinking about my thirst, I am ipso facto not thinking about getting a drink. Moreover, reflecting in this way is itself a form of activity, and so liable to distract us from our first-order actions.13 Neither of these reasons seems especially strong.14 But Velleman suggests that there is a deeper, structural reason why reflective consciousness is alienating: “Most importantly, though, consciousness just seems to open a gulf between subject and object, even when its object is the subject himself. Consciousness seems to have the structure of vision, requiring its object to stand across from the viewer—to occupy the position of Gegenstand.”15 On Velleman’s view, it seems, t here is just no way for reflective consciousness to feature in action that does not “open a gulf” between the agent and their action. And that is the tension he finds at the heart of our thinking about h uman agency: “What makes for agency in general, according to Frankfurt, is the . . . condition of being ‘in touch’ with the mechanisms guiding one’s behavior. And the feature that distinguishes a person from a lower animal now turns out to take the person out of touch with those mechanisms, since what distinguishes him as a person is reflective consciousness, which opens a gulf between him as the subject of reflection and his motives as the object.”16 By this measure, nonhuman animals do better at satisfying the general condition for agency, as their lack of reflective consciousness makes it impossible for them to fall out of touch with the mechanisms that guide their behavior. Is alienation, then, the inevitable fate of human agency? This seems unnecessarily pessimistic. Rather, I think, the trouble stems from Velleman’s acceptance of the two-tiered account of agential control. On that account, one’s control of bodily action seems to be outsourced to mechanisms that are, strictly, distinct from the agent themselves. The agent’s involvement in their own action, accordingly, is l imited to staying “in touch with” those mechanisms. But then, it seems, the agent falls short of actually doing anything in the public world. No wonder, then, that alienation looms. The problem Velleman identifies, therefore, really does seem to threaten human agency—understood, at least, in terms of the two-tiered
158
R eas o n in P ercepti o n and A cti o n
account of control. Moreover, this way of thinking about agential control is strikingly widespread. For example, in the course of defending his account of know-how and intelligent action, Jason Stanley, approvingly quotes the following passage by Jerry Fodor: here is a l ittle man who lives in one’s head. The l ittle man keeps T a library. When one acts upon the intention to tie one’s shoes, the little man fetches down a volume entitled Tying One’s Shoelaces . . . W hen the little man reads the instruction “take the left free end of the shoelace in the left hand,” he pushes a button on a control panel. . . . W hen depressed [the button] activates a series of wheels, cogs, levers and hydraulic mechanisms. As a causal effect of the functioning of these mechanisms, one’s left hand comes to seize the appropriate end of the shoelace.17 This is obviously tongue-in-cheek, but the point seems clear: we exercise agential control by setting in motion independent causal mechanisms—“a series of wheels, cogs, levers . . .”—which see to it that our intentions get executed. And, despite working with a rather different metaphysics of agency, Helen Steward writes: “I am in charge [of my bodily actions] in the way that a government minister is in charge of a department. No minister directly controls all of the work of a particular department or knows all the details of what is g oing on within it: tasks are delegated to particular civil servants who then delegate further work to more junior officials, and so on.”18 Once again, the point seems to be this: an agent exercises control over their bodily actions by setting high-level goals or para meters for action, and then trusting to lower-level mechanisms to carry them out in relative independence. And broadly similar suggestions can be found in, for example, Dretske, Enç, and Papineau.19 What accounts for the broad popularity of this picture of agential control? A full examination of this question is far beyond the scope of the present chapter.20 Broadly speaking, however, I think it is fair to say that it is felt to be mandated by empirical findings in the psychology and neuroscience of agency and motor control. Agents—including highly skilled expert agents—a re incapable of reporting on many of the fine details of their actions, even in cases where those details seem critical to the success of their actions.21 Indeed, it seems that motor actions can be
C o ntr o l and K n o wledge in A cti o n
159
guided successfully to completion even in cases where their agents are radically mistaken about these details.22 I will not have much to say about this line of argument in this chapter, although—as I w ill suggest below—I do not think it is conclusive.23 Before getting to my own positive alternative account of control, however, I want to consider our options, if we w ere to accept this more standard picture. More specifically, could the threat of alienation that Velleman identifies be avoided without questioning the underlying account of control? If such attempts fail—as I believe they do—then this should provide motivation for questioning the model of control on which they rely. Velleman identifies two potential solutions to the problem of alienation, one that he presents as a “half-measure,” and another that he pre sents as a kind of ideal of human agency. The “half-measure” consists in a reading of Frankfurt’s well-k nown hierarchical account of the w ill, and I am not going to discuss it further.24 It is Velleman’s ideal of h uman agency that is relevant to us.
2. Skill as “Higher Wantonness” Velleman’s ideal is of a h uman agent who transcends reflective consciousness—an agent who has achieved a “higher wantonness,” as he puts it. This higher wantonness is supposed to be characteristic of highly skilled action, and is illustrated with passages from the Daoist text the Zhuangzi: Butcher Ding was cutting up an ox for Lord Wenhui. Wherever his hand touched, wherever his shoulder leaned, wherever his foot stamped, wherever his knee pushed—w ith a zip! with a whoosh!— he handled his chopper with aplomb, and never skipped a beat. He moved in time to the Dance of the Mulberry Forest, and harmonized with the Head of the Line Symphony. Lord Wenhui said, “Ah, excellent, that technique can reach such heights!” The butcher sheathed his chopper and responded, “What your servant values is the Way [Dao], which goes beyond technique. When I first began cutting up oxen, I did not see anything but
160
R eas o n in P ercepti o n and A cti o n
oxen. Three years l ater, I couldn’t see the w hole ox. And now, I encounter them with spirit and don’t look with my eyes. Sensible knowledge stops and spiritual desires proceed. I rely on the heavenly patterns, strike in the big gaps, am guided by the large fissures, and follow what is inherently so. I never touch a ligament or tendon, much less do any heavy wrenching!”25 It is not to the point here to consider the interpretive puzzles such passages raise. What matters for us is that, according to Velleman, the ideal exemplified by such highly skilled agents consists, at least in part, in the abandonment of self-regulation and, with it, of reflective consciousness.26 Indeed, such agents are comparable to nonhuman agents, like the ones featured in this passage: The kui [a mythical one-legged beast] said to the millipede, “I go hippety-flopping on one foot, and there’s nothing like it! How do you manage those ten thousand feet of yours?” The millipede said, “It’s not like that . . . I just put my heavenly mechanism into motion. I don’t know how it works!” The millipede said to the snake, “I use this mob of legs to walk but still don’t match up to you with none at all. How do you do it?” The snake said, “The heavenly mechanism does it. What could be easier? What use would I have for legs?”27 hese creatures exemplify agency in virtue of being in touch with the T “heavenly mechanisms” that guide their movements. Similarly, highly skilled action, according to Velleman, involves an analogous reliance on mechanism: the skilled agent, like the millipede, just puts their “heavenly mechanism into motion,” trusting that it w ill be adequate to the demands of the situation. Of course, actual (as opposed to mythical) millipedes and snakes lack the capacity for reflective consciousness altogether. Skilled human agents do not. And indeed, according to Velleman, reflective consciousness still plays a role in skilled action, albeit a secondary and enabling one.28 For one thing, highly skilled agents normally have to work hard and deliberately in order to cultivate the relevant mechanisms. Moreover, abilities for reflection and self-regulation do not entirely go away in highly skilled agents. Rather, although they play no role in the normal course of skilled
C o ntr o l and K n o wledge in A cti o n
161
activity, they remain ready to intervene if something goes wrong: in such cases, explicit reflection and self-regulation may be required to work out a way to proceed.29 I think, however, that we should not be satisfied with this account of skilled agency. For one thing, the role reserved for reflection and self- regulation on this account is a curious one. Intuitively, we would expect a skilled agent’s skill to be especially manifest in difficult circumstances. There may be little to distinguish a novice bicycle rider from a skilled one on a smooth road, without traffic or obstacles. It is when conditions become hard that we expect skill to make a difference. The skilled rider should be able to take difficulties (up to a certain point) in stride, while the unskilled one is not. Things, however, look diff erent on Velleman’s account. On this account, skilled performance is associated with the absence of self- regulation and reflection. But then, it seems that skill deserts the skilled agent, precisely when the going gets tough—since it is at that point that the agent is supposed to fall back to self-regulation and reflection. Moreover, it seems open to question whether Velleman’s account captures the nature of skilled action, even when t hings are g oing well. When we praise an agent for a skilled performance, for example, we naturally take ourselves to be praising them for something that they are doing (or just did), namely, the performance itself. The problem, now, is that it is unclear that Velleman’s account can vindicate this natural thought. On that account, it seems, the agent’s role is simply to stand by and let the movements unfold under the guidance of the appropriate mechanisms. In that case, however, it is unclear that the agent really deserves praise for the performance itself (rather than, perhaps, for their work in cultivating the relevant mechanisms in the first place).30 Of course, Velleman could respond to this by pointing to the account of agential control sketched earlier, and suggesting that, so long as the agent is suitably “in touch” with the mechanisms responsible for the control of the relevant bodily movements, they count as the agent of t hose movements. To evaluate this claim, we need a story about what it is to be “in touch” with these mechanisms in the relevant sense. I have already explained why I do not think that Velleman’s suggestion that we should look to phenomenology for an answer is adequate. A more plausible suggestion is that we should look to these mechanisms’ responsiveness to the agent’s personal-level thought: unlike, say, the mechanisms controlling
162
R eas o n in P ercepti o n and A cti o n
our digestive processes, the mechanisms involved in our bodily actions work to implement our personal-level goals, often in accordance with plans formulated through personal-level reasoning. The thought, then, is that to the extent that the relevant subpersonal mechanisms are reliable enough in implementing our intentions and plans, we can be said to be in control of the ensuing actions.31 But while this claim might seem prima facie plausible, it does not withstand scrutiny. A reliable match between bodily movement and the content of our intentions is not enough for agential control, nor (a fortiori) for skill. We can make the problem vivid by considering the following caricature of skilled action. In the science fiction novel Nexus, nanotechnology has made it possible to run human-made programs on one’s own brain.32 Kade, one of the protagonists, ends up with an app on his brain that promises to give him the martial arts skills of Bruce Lee. Phenomenologically, this works like a video game. Kade picks a target using a pointer in his visual field, and selects an attack routine from a drop- down menu. At that point, the program takes over, and guides Kade’s limbs through the corresponding motions. Aside from the initial choice of an attack routine from the menu, Kade has no idea what moves his body is going to make. Nevertheless, we may suppose that the app r unning on Kade’s brain is very reliable in getting his body to execute the selected moves.33 Even so, we would be disinclined to say that Kade has become a skilled martial artist, or that he has the right sort of control over his martial arts moves. The challenge for an account like Velleman’s is to explain how Kade’s case differs from that of normal skilled performance: why is Kade less “in touch” with the mechanisms that execute his intentions than a true skilled agent?34 Although the example might be reminiscent of traditional deviant causal chain scenarios, the problem I am raising is not meant to be just another version of that problem. For one thing, I am not questioning whether Kade’s jumping back kick, say, counts as an intentional action: it may be a sort of “ballistic” action, which once initiated proceeds on its own with little or no further input from its agent. The question is whether Kade meets the conditions for agential control—and, if not, why not. Furthermore, the causal chain implicated in Kade’s case is not “deviant,” in the sense familiar from classic action-theoretic debates. Deviant causal
C o ntr o l and K n o wledge in A cti o n
163
chain scenarios generally involve fortuitous connections, which bring about the intended outcome by coincidence: a gunshot stampeding a herd of wild pigs, thereby killing the intended victim; or a mountaineer’s inadvertently loosening his grip on the rope, thereby sending his companion tumbling to his death. But fortuitousness is not the issue h ere. Using the Bruce Lee app is how Kade intends to execute his jumping back kick; indeed, it is the only way in which he can get to perform a jumping back kick.35 Kade is an example of an agent who has dispensed with “self- regulation,” and whose bodily movements are guided by highly reliable mechanisms that implement his high-level intentions—and yet, intuitively, his martial arts moves do not exhibit agential control, nor is he a skilled martial artist.36 What are our options, then? Someone committed to the two-tiered account of control might decide to bite this bullet and reject intuition here. Perhaps, as in so many other areas, common sense has been revealed to be just wrong: we r eally have no more control over our own actions than Kade does over his martial arts moves. It is, perhaps, only our long familiarity with our own bodies and their behavioral possibilities that inclines us to think otherwise. While not incoherent, I believe that the above response is a counsel of despair, to be endorsed only if all alternatives have been shown to fail. Fortunately, I believe there are alternative ways of thinking about agential control that can avoid this consequence.
3. An Alternative Picture: Knowledge in Action If what I have argued above is along the right lines, then we have reason to suspect that the two-tiered account of agential control should be rejected. On that account, as we saw, it suffices for agential control that the agent be somehow “in touch” with independent mechanisms which guide the relevant bodily movements. The trouble, as I have already suggested, is that this account seems to leave action-execution out of the agent’s purview: the agent’s involvement in their own actions seems to fall short of their actually d oing anything in the public world. How can we correct this mistake?
164
R eas o n in P ercepti o n and A cti o n
The account of agential control Velleman inherits from Frankfurt seems to rest on a dichotomy between, on the one hand, the agent’s mental states—such as beliefs, desires, and intentions—a nd, on the other, the mechanisms that actually guide the bodily movements that constitute their actions. The agents themselves, moreover, are clearly identified with the former—this is why staying “in touch with” these mechanisms becomes an issue in the first place. This dichotomy, moreover, seems to be an instance of a way of thinking with a long history in the theory of action. Consider the question Wittgenstein (himself a critic of this way of thinking about action) famously asks in the Philosophical Investigations: “What is left over if I subtract the fact that my arm goes up from the fact that I raise my arm?”37 That your arm goes up, Wittgenstein’s question implies, is just a brute physical fact about you. Your mind’s contribution to the action of raising your arm is what is “left over,” once we subtract this brute physical fact. Crucially for our purposes, this implies that the mind’s contribution to action falls short of the overt action itself: whatever exactly it is that is “left over” when we subtract the arm’s rising from the fact that you raise your arm, it cannot be something that entails that your arm rises. And this, of course, is not just a quirky feature of Wittgenstein’s discussion: most contemporary accounts of action still rely on such action-independent mental states—mental states, that is, whose presence does not entail that any overt action is taking place. The suggestion I want to explore involves rejecting this assumption. On the view I w ill suggest, we can see in agential control a kind of thinking that is not constitutively independent of overt physical action—rather it is, in a sense to be made more precise in what follows, embodied in the action itself. This suggestion, of course, makes contact with McDowell’s proposal, quoted at the beginning of this paper, that “physical activity can be rationality in action, as opposed to a mere result of exercises of rationality.”38 Moreover, there is a symmetry between this position and a position in the philosophy of perception that McDowell has long championed. 39 According to traditional views in the philosophy of perception, perceptual states are themselves purely inner states of organisms, in the following sense: being in any particular perceptual state entails nothing about your environment.40 The epistemic role of perceptual states is explained, then,
C o ntr o l and K n o wledge in A cti o n
165
in terms of their (contingent) relations to external objects and states of affairs: for example, one might argue that perceptual states may reliably covary with states of the environment, and so carry information about them.41 While still very influential, such views no longer seem to be as dominant as they once were. As McDowell has argued, such views make it hard to see how we could direct our thoughts on particular things in our environments at all.42 How could visual perception, for example, put you in a position to think about that tree over there, if a subject could be in just the same perceptual state, although no tree is visible to them at all? In response, McDowell and other critics of such views have argued that perceptual states are world-involving states that disclose or make manifest parts or aspects of reality to perceiving subjects. This implies, of course, that perceptual states are not purely inner: you cannot be in a state that discloses or makes manifest the world as being a certain way unless the world really is that way.43 My suggestion is that, just as, on the picture just sketched, perceptual states are world- involving, we ought to recognize action- involving exercises of our minds: a kind of thinking that is in the action. This is independently plausible. Consider an action most of us perform unreflectively, such as tying our shoelaces. Shoelace tying consists in a series of complicated finger movements, which collectively implement a solution to a relatively complex problem. Tying your shoelaces, then, plausibly manifests your grasp of a solution to this problem. It seems wrong, however, to suppose that your grasp of a solution to this problem must be a further, inner state, which is distinct from your actually tying your shoelaces. This is suggested, for instance, by the fact that you may have no informative way of verbally describing how you solve the problem of tying your shoelaces. (Of course, this is consistent with an action- independent representation of a solution to the problem being encoded in your nervous system. The existence of such a sub-personal representation, however, would not show that you possess an action-independent way of grasping the solution.) It seems better to allow that your finger movements themselves constitute your solving the problem of shoelace tying. But, of course, problem solving is a paradigm case of intellectual activity—a kind of thinking.44 Such action-involving thinking would help us understand what is going wrong in the case of Kade. As the case was described, Kade has no
166
R eas o n in P ercepti o n and A cti o n
idea what movements his body is g oing to perform given the input from the app’s menu. Moreover, even as his body is going through those movements, he experiences them as a passive bystander. In our terms, Kade’s movements do not embody his grasp of how to perform a jumping back kick, for example. This is so, even though the mechanisms responsible for implementing his choices may be as reliable as we like.45 So how should we understand action-involving thought? We can begin to approach this topic by drawing on Michael Thompson’s discussion of intentional action.46 Thompson’s topic is not the nature agential control, but rather the teleology of intentional action. Nevertheless, his discussion is useful h ere, because it is designed specifically to c ounter the assumption that the teleology of intentional action must be inherited from action- independent m ental states. Thompson begins with the observation that we often explain actions in terms of other actions: for example, my breaking an egg over a bowl in the kitchen may be explained in terms of the fact that I am making an omelet. My breaking of the egg is explained as a means, or instrumental part, of the “bigger” action of making an omelet. Other approaches in the philosophy of action would seek to explain this in terms of action- independent states (such as plans or intentions) that represent egg- breaking as a means to making an omelet. Thompson, by contrast, suggests that we take instrumental explanation in terms of actions (what he calls “naïve rationalization”) as metaphysically fundamental. Explanations of the form “S is G-ing because she is F-ing,” in other words, need no metaphysical backing by action-independent mental states, such as S’s intending to F by G-ing. This approach is especially helpful in cases, like the case of shoelace- tying, where it is implausible to attribute to the agent an action-independent representation of what they are doing. Despite this, the agent’s finger movements can be given a straightforward rationalizing explanation in terms of the bigger action of tying their shoelaces: they are moving their fingers the way they are because they are tying their shoelaces. Still, Thompson’s suggestion is not quite sufficient for our purposes. To see this, notice the following point: two agents may perform the same bodily movements, in the serv ice of the same larger actions, even though only one of them is acting with control. Consider two downhill
C o ntr o l and K n o wledge in A cti o n
167
skiers—one a novice, the other an experienced skier—both making the same sharp left turn on a slope. Just as before, Thompson’s framework allows us to give rationalizing explanations of both agents’ bodily movements as they make their respective turns: they both move their bodies in the relevant ways, because they are making a sharp left turn. The trouble, however, is that this similarity may mask an important difference: as we may suppose, only one of them—the experienced skier—accomplished the turn with control; for the novice, the success was very much down to luck. T here is nothing in Thompson account, as it stands, to explain this difference. I suggest that we can capture such distinctions by making explicit an epistemic dimension of Thompson’s proposal. Specifically: If you are G-ing b ecause you are F-ing, then you are (thereby) taking it, of your G-ing, that it is (part of) a way for you to F. A couple of clarifications are in order h ere. To begin with, and crucially, the “takings” in question are not meant to be self-standing mental states that are distinct from the actions themselves: as the shoelace tying case discussed above suggests, we can make sense of the idea that your grasp of a way of acting is manifested or embodied in the action, rather than directing it from the outside.47 Relatedly, t here is no need to suppose that you must be able to describe in words your G-ing, or how it fits into a way for F-ing. As I pointed out earlier, it is a familiar theme in the recent literature that even skilled agents are often unable accurately to describe the details of their actions. But this is perfectly consistent with the present proposal, as the thoughts in question can be de re, rather than descriptive. We can again appeal to an analogy with the philosophy of perception to make this point. Just as (according to many philosophers, at least) our thoughts can be about par ticu lar physical objects, not in virtue of our deploying any identifying description of those objects, but rather in virtue of our being in perceptual contact with t hose objects, our thoughts can be about particular ongoing actions of ours in virtue of our being in the process of performing those actions.48 Thus, while tying your shoelaces, you think (indeed, plausibly know) of your finger movements, that they are instantiating a way for you to tie your shoelaces.
168
R eas o n in P ercepti o n and A cti o n
Finally, a note about the concept of a “way” of acting, as used above. Any temporally extended action consists in a series of other actions, and ultimately bodily movements: for example, making an omelet partly consists in breaking and beating some eggs, and tying your shoelaces consists in a series of complicated finger movements.49 When you successfully F, the sequence of actions (action types) by means of which you F-ed constitutes the way in which you F-ed. At the same time, it is possible correctly to implement a way to F, and yet to fail to F—due, as we say, to circumstances beyond your control. Even an expert tennis player’s shot may be deflected out of bounds by a freak gust of wind, for example. Let us now return to the case of our two skiers. Both, as we have assumed, succeed in making a sharp left turn; and both, we may further assume, do so by implementing the same way of making a sharp left turn. Thus, on the view suggested above, while making their turn, they both take it that what they are d oing constitutes a way for them to make a sharp left turn. But, by hypothesis, the novice’s success in making the left turn was due mostly to luck. By the same token, the novice’s taking it that what they are doing constitutes a way for them to make a left turn is also due to luck. The novice’s actions, therefore, do not embody knowledge that they are implementing a way for them to make a sharp left turn. The expert skier’s success, by contrast, is not significantly due to luck; their actions, therefore, may plausibly be taken to embody knowledge. This, then, leads to a very natural approach to agential control: your level of control over your F-ing is determined by the epistemic standing of your taking it, of some G that you are currently doing, that it is (part of) a way for you to F. Being in control (tout court) of your F-ing consists in knowingly implementing a way for you to F: You are exercising control over your F-ing just in case you are G- ing because you are F-ing, and you thereby manifest your knowledge, of your G-ing, that it is (part of) a way for you to F. One might worry about this, on the grounds that agential control comes in degrees, while knowledge (arguably) does not. We can speak of an agent’s having more or less control over some relevant bodily movements, but can we also speak of them having more or less knowledge of those movements that they constitute a way for them to make a sharp left turn on their skis?
C o ntr o l and K n o wledge in A cti o n
169
There are a couple of things to say in response. The first point is that, even if we assume that “knows” does not function as a gradable verb in ordinary usage, this is not to say that we do not frequently draw epistemic distinctions between diff erent pieces of knowledge: one piece of knowledge may be more secure than another, for example. Similarly, it might be that two agents are both in control of their respective bodily movements, but one of them is more securely so. Furthermore, comparisons of epistemic status make sense even for subjects who fall short of the (likely vague) threshold required for knowledge. Comparing two novice skiers, for example, we can say that one exhibited more or less control over a certain move. This can be taken to correspond to their being closer to the threshold required for knowledge. Notice, also, that the link between control and knowledge helps connect the topic of control with the much-discussed topics of skill and know-how. One way to approach the connection is this: if control consists in actions that embody knowledge, then one might want to know how such knowledge is warranted—as knowledge intuitively needs to be. Broadly speaking, the answer is straightforward: the knowledge in question is warranted by standing knowledge-states of the agent’s, namely, their skills. The nature of such standing states is a m atter of ongoing controversy, of course, and I cannot discuss it further here. But my account seems compatible with many diff erent possibilities.50 Finally, one might worry that my approach—w ith its emphasis on knowledge and epistemic norms—conflicts with the fact that simple agents, such as nonhuman animals and human infants, are capable of acting with control. But I think there need be no special problem here. A fter all, we also ascribe to such agents mental states such as knowledge and belief, and describe them as acting for reasons—a nd do so in ways that are not obviously metaphorical, as ascriptions of malicious intent to a recalcitrant car engine clearly are. No doubt, t here are philosophical puzzles in this area; but I see no reason why the present proposal should face any special difficulties. If a dog can know where his bone is buried, he can also know, of his current furious digging, that it is a way for him to get to it. Let us take stock. Acting with control involves, in the general case, taking means to achieve your ends. This, on the view developed here, is an intellectual activity—it is a practical form of problem-solving. Moreover, it is so intrinsically, and not just in virtue of being accompanied by distinct,
170
R eas o n in P ercepti o n and A cti o n
“inner” episodes of deliberation. On this approach, agential control does not consist in simply being “in touch” with independent causal mechanisms that guide the respective movements of your body. As McDowell puts it in a passage I already cited earlier, it is, rather “rationality in action, as opposed to a mere result of exercises of rationality.” Accordingly, it is not the case that “reflective consciousness” as such tends to get in the way of agential control. Rather, as we just saw, agential control constitutively involves a kind of reflexive self-awareness—a recognition of the means-ends structure of one’s own actions.
Chapter Six ✣
Naturalism in the Philosophy of Action ✣
Jennifer Hornsby
J
ohn McDowell once offered me a possible way out of a puzzlement I felt. (This was in conversation, a long time ago.) I was puzzled about what philosophers suppose they mean by “actions,” and John suggested that it might help to speak in terms of happenings rather than events. Well, the suggestion didn’t immediately prove to cure my puzzlement. And I think that I may (at last) have come to understand why.1 Happenings play a prominent role in John’s lecture “Some Remarks on Intention in Action,” in which he draws attention to the importance of distinguishing things that happen from events.2 I find that I disagree with him t here. And I want to try to uncover the source of this disagreement in what follows. Thinking as I do that philosophers of action are bad at keeping track of what they’re talking about, my own remarks w ill not be aimed at John exclusively. Far from it: agreement with John w ill matter very much more than the par ticu lar disagreement—as is surely fitting for a volume in honor of a true philosopher whose work I could not hold in higher regard. McDowell dissociated himself from “much modern philosophy of mind” when he said that this “saw itself under an obligation to reintegrate 171
172
R eas o n in P ercepti o n and A cti o n
the thinking subject into a natural world from which it has come to seem alien.”3 So it is, I think, with much modern philosophy of action. And my hope is that even in the course of taking exception to an aspect of McDowell’s treatment of intention in his “Remarks,” I might strengthen an understanding of human agency of a kind McDowell himself has done much to promote. Insofar as I am concerned h ere with the questions in temporal ontology which have puzzled me for so long, that’s b ecause I think we need answers to them if a naturalism of a sort McDowell endorses is to prevail in the philosophy of mind-and-action.
1 1.1 In a paper originally published in 1987, McDowell asked, “What reason is there to suppose that the sense that we make of linguistic behavior is still available to us, if when we contemplate the behavior we are required to deny ourselves the terms in which we ordinarily make sense of it?” 4 No reason. And so it may seem to be with nonlinguistic behavior. When it comes to finding a fellow human agent’s doing what they do intelligible, one cannot take what McDowell came to call a “sideways-on” view—“ here the conceptual system, there the world.” In Mind and World, McDowell spoke of “a kind of understanding that involves seeing how the phenomena of our subjects’ lives can be organized in the order of justification, the space of reasons.”5 He said that “experiences are actualizations of our sentient nature in which conceptual capacities are inextricably implicated.” When he turned to action t here, he said, about “intentional bodily actions,” that t hese “are actualizations of our active nature in which conceptual capacities are inextricably implicated.” The implication of one’s conceptual capacities in one’s active nature ensures that the kind of understanding involved in taking a person to be a subject of experience is involved equally in treating them as an agent. Perhaps one can just about imagine a theorist who observes plenty of someone’s behavior but has no interest in seeing the someone as one who themself sees some point in doing what they do. Such a theorist would be indifferent to what the one whose behavior they observe might be up to: the theorist d oesn’t seek to know what reasons they have for behaving as
N aturalism in the P hil o s o phy o f A cti o n
173
they do, nor therefore can the theorist find them intelligible in the everyday way in which we may find one another intelligible. In order to find a person intelligible in that way, one evidently cannot adopt the stance of such a theorist, but must seek to share the agent’s point of view. And a philoso pher of action can no more adopt the theorist’s stance than a fellow human being can. Or at least the philosopher w ill not adopt such a stance if they acknowledge that any human agent is, like themself, a deliberator and actor who may tell another what they are doing, and who can find out more about what someone else is doing by asking them. The temporal perspective of someone who deliberates and acts is prospective—directed at any time toward the f uture from what is then the present. This perspective is introduced only when it is allowed that agents themselves know more than is open to the view of o thers about what they are doing, and thus what is happening—that they are d oing so- and-so. A simple example: you see that I have just walked across the street; you ask me where I’m going, and I say “I’m g oing to the library,” I then express an intention I have; and in making it known to you, I bring an anticipated future into view for you. (For all you had known, I might have been walking to the park.) But it’s not that you are now simply better able to predict what w ill happen next with me; nor, in finding out where I am headed, do you learn in what manner my body is going to move. In learning from someone what they are, or w ill be, d oing, one comes to apply to them the concepts they themself use in practical thought, concepts whose possession they manifest both in acting and in saying what they’re up to. When McDowell suggested to me (long ago, as I said) that one does well to speak in terms of happenings, he recognized that we couldn’t find one another intelligible if we supposed that the sole inhabitants of the temporal world were events. 1.2 Let me now say something about those who, so far as the temporal world is concerned, do want to confine themselves to talk of events (and “states” perhaps). It was things these philosophers said that surely led to the puzzlement I expressed to John as to what “actions” is supposed to mean. I’ll start at the beginning, with Donald Davidson’s papers from the 1960s.
174
R eas o n in P ercepti o n and A cti o n
I’ll then say a l ittle about the linguistic resources needed in any account of human agency. All too often in the past, I’ve drawn attention to the fact that “an action,” as used by Davidson, and as it continues often to be used by those who subscribe to a story of action of the sort he introduced, can strike one as ambiguous.6 Sometimes Davidson meant by “an action” a concrete unrepeatable individual thing, where such t hings are of a sort at least one of which exists if someone intentionally did something. But sometimes he meant something quite different: an action, when taken to be something someone did, appears to be a t hing of a sort that a person may have done. The ambiguity, if this is what it is, shows itself strikingly in Davidson’s earliest writing on the subject, where he wrote: “I flip the switch, turn on the light and illuminate the room . . . and also alert the prowler. . . . Here I do not do four things, but only one.”7 Davidson assured us that the “only one thing” he “does” is an action, actions being a species of events, events being variously describable concrete particulars. But what about the four things which Davidson listed in order to assert that he “does” not do three of them? If Davidson presented an actual example, then we have to think that he turned on the light (past tense)—that turn on the light was something he did. But we d on’t understand the example unless we take flip the switch, illuminate the room . . . also to be things he did.8 (What appears to be a simple pre sent tense of Davidson’s “I flip the switch,” “I do,” e tc. is often found in philosophy of action. Presumably it is meant to be understood generically somehow. But if it introduces a time, then it cannot be the present in which an agent has their presence, but has to be, as it were, an arbitrary past.) Why should Davidson have denied that he did three of the things that he had just said that he did? Well, he wanted to convey that he had told us about a single action in a sense suited to his causal theory—a particular, a “concrete individual.” T hose who f avor the concrete over the abstract as candidates for real existence might approve of such particulars but want to rule out any talk of things a person did. They might say that verb phrases like flip the switch and illuminate the room cannot really specify things. Of course it would be outlandish to suggest that there w ere as many really existing entities as there might be found verb phrases covering everything that someone at some time did. But given that “something,” “what,” and “that which” can be used to tell of something done, speaking of “things” here seems inevitable.
N aturalism in the P hil o s o phy o f A cti o n
175
Insofar as Davidson’s actions are redescribable events to which other items stand in the relation “cause,” they appear on the scene in his analysis of so-called action sentences, any of which says that someone did something. So even if events have a place in a story of what t here is when an action verb in the past tense is predicated of an agent, still no event is predicated of a person when what they did is specified. We can think, along with Davidson and many others, that if Ann ate an apple, there is an event, sc. Ann’s eating of an apple. But this could hardly lead us to think that, a fter all, eat the apple was not something Ann did. Nor could it lead us to deny that eating the apple was something Ann had been d oing. We are bound then to question whether Davidson really noticed any anomaly, or was entitled to his “presumably,” in the opening sentences of “The Logical Form of Action Sentences”: “Jones did it slowly, deliberately, in the bathroom, with a knife, at midnight. What he did was butter a piece of toast. We are too familiar with the language of action to notice at first an anomaly: the ‘it’ of ‘Jones did it slowly, deliberately, . . .’ seems to refer to some entity, presumably an action.”9 Having been told that butter the toast was what Jones did, should it r eally seem to us that the “it” of “Jones did it. . . .” refers to an entity which we should presume to be an action? There is a s imple explanation why Davidson for his part should have been ready to conflate things done by agents with actions (events).10 His focus was exclusively on a certain sort of teleological explanation, in which the end or goal that an agent had in acting shows up in the explanandum: “He φ-d. Why did he φ?” This focus ensured that actions in his intended sense could be circumscribed by using verbs definable, as he put it, “in terms of some terminal stage, outcome or consequence,” and by using these verbs in the simple past tense. Thus his action sentences: “I turned on the light,” “Jones buttered a piece of toast,” e tc. Call the verb phrases predicated here telic: one knows what it is for an agent’s telos to have been reached if they had intended to turn on the light, or to butter a piece of toast, or. . . . And it is surely right to think that in a case of successful agency, a past tense telic verb is predicable of the agent. In confining himself to past-tense occurrences of telic verb phrases in his account of what actions are, Davidson failed to make allowance for the fact that verbs, inflected in one or another way, are predicated of agents. He failed to allow that verb phrases are not all of them telic. And he failed
176
R eas o n in P ercepti o n and A cti o n
to allow that verb phrases, telic or other, can occur not only in saying what someone did or has done (past tense), but also in saying both what they will do (f uture tense), and again in saying what they will be doing, or are doing, or were doing. T hese last three make use of the progressive. Not only do they differ from one another in respect of tense (f uture, present, past), but all of them differ from “did” and “will do” in respect of aspect. They contain what linguists call “the progressive particle in ‘–ing.’ ” Notice that it is not only in predications that the perfective / imperfective distinction is marked. Thus infinitives are used when it is said what someone will do or did, but gerunds when it is said what someone w ill be, or is, or was doing. (“What w ill she do?” “What did she do?” Turn on the light. “What w ill she be doing?” “What is she d oing?” “What was she doing?” Turning on the light.) Again, the distinction is marked in nominals: “Ann’s eating of the apple” is a perfect nominal, whereas “Ann’s eating the apple” is an imperfect one.11 It is evident that the language Davidson drew on in order to give his own account of human agency falls very far short of the language of action as we know and use it. McDowell for his part of course appreciates that the progressive may be used in predications, both of telic and other verbs of action. Indeed, “the image of intentions as a kind of continuant” (presented in his “Remarks,” which I come to next) is designed to “accommodate prog ress in an action.” I’m g oing to suggest that McDowell conceals the full significance of the use of the progressive, despite the need for it in accommodating progress and in avoiding an exclusively “sideways-on” view.
2 2.1 In his “Remarks,” McDowell presented a picture of intention in action in order to provide an understanding of how a prior intention, which is an intention for the f uture, relates to an intention in action. Whereas a prior intention is taken to predate the action that t here w ill be if and when it is executed, an intention in action is one in the course of getting executed— “contemporaneous with d oing whatever it is an intention to do,” as McDowell initially characterizes it. McDowell was concerned to contrast his own picture of intention in action with the picture one gets from John
N aturalism in the P hil o s o phy o f A cti o n
177
Searle’s account of the “for the future” / “in action” relation. To that end, McDowell considered “an intention harbored by someone standing on a curb: an intention to cross the street when the light turns green.” He “worked with an image of intentions as a kind of continuant whose instances change their shape as time passes” so that “intentions for the f uture become intentions in action when the time for acting comes”; and “a persisting intention in action changes its shape in respect of how much of what is intended lies behind it and how much is still in prospect.” McDowell relied upon a distinction between t hings that have happened and things that are happening. This in turn allowed him to distinguish particu lar actions from, as he said, activity. Then he could claim that intention in action relates to these differently. His view of the difference emerges in two claims he made: (1) It must be wrong to suppose that someone’s intention’s being an intention in action consists in the presence of a de re relation to a particular action. (p. 9) (2) Intention in action can be seen as relating de re to relevant activity. (p. 10) I’m going to quarrel with the second claim, (2). I think that McDowell’s image of intentions as a kind of continuant relies on a reification of activity which stands in the way of a proper understanding of a h uman agent’s prospective perspective—the perspective of one who is in the course of doing something that, save for a change of mind or forgetfulness on their part or interference of one or another sort from outside, they w ill come to get done. So far as claim (1) is concerned, I want not merely to endorse it, but to suggest that it pinpoints the central problem with the idea that a story of human agency might belong where Davidson placed it—in what is called “the event-causal framework” (§3.2). We are told that this framework provides “by far the most widely accepted view in the con temporary philosophy of mind and action,” commitment to which is a “commitment to a very minimal and widely endorsed kind of naturalism.”12 This “very minimal naturalism” is of a sort which McDowell called “restrictive,” against which he set his own “liberal” naturalism in 2004.13 But I’m g oing to suggest that McDowell’s image of intentions depends upon a questionable diagnosis of the errors of those who have committed themselves to the “very minimal” naturalism. This will put me in a position not only to say both how right I think McDowell was in claim (1),
178
R eas o n in P ercepti o n and A cti o n
but also to suggest that once that claim is accepted, a conclusion emerges which is more far-reaching than the one he settled for in his “Remarks.” 2.2 McDowell’s understanding of happenings gave rise to the claim that I said I would quarrel with—to his saying “Intention in action can be seen as relating de re to relevant activity.” Here an activity seems to be treated as an object, a particular, a res. At least I think that McDowell relies upon such a treatment in sustaining his image of intentions as a kind of continuant. McDowell said: “I have framed the conception of the reality of partic ular actions . . . in terms of a mode of thought to which tenses matter. Actions are a species of events. And I am urging an asymmetry between things that have happened, which I am allowing as particulars, and t hings that are g oing to happen, or things that are happening, which I am not.”14 Here McDowell distinguished between that which he does, and that which he doesn’t, allow to be particulars—between events and things that are happening. And if the “relevant activity” of his claim (2) is taken to be something of a sort that may be happening, then it would seem now not to be a particular, not a res, and thus not something to which an intention in action (or anything e lse) might be related de re. Still, McDowell provides a reason for conceiving an activity as a res. He says: “Activities occupy time in a different way from actions. As soon as one has embarked on an activity, the activity is all t here, in a way that is indicated by the fact that one can attribute it with a use of the perfect tense: as soon as one is walking, one has walked. (Of course this does not imply that the current stretch of walking is over.) Since the activity is all there at any moment, we can, if we like, conceive it as an object, a res.”15 One could wonder exactly what McDowell intends by “an activity” h ere. Is, say, walking, or walking across the street, to be reckoned an activity? Or is it John’s walking, or John’s walking across the street, which is to be thought of as an activity? Whereas walking is something a person is d oing when they have embarked upon walking, so that “is walking” is predicable of that person, John’s walking is something happening so long as John is walking. Well, given McDowell’s speaking of an activity as occupying time, it seems that he must mean something that is or was ongoing—John’s walking as it might be, or John’s walking across the street as also it might be.
N aturalism in the P hil o s o phy o f A cti o n
179
What sense are we then to make of the activity as “all there at any moment”? Certainly, as McDowell says, the nontelic “is walking” and “has walked” can both be predicated of someone who is walking. This might be taken to be enough for their walking to count as “all there.” But John’s walking’s being “all t here” in this sense d oesn’t seem to be enough to ensure that some whole object is present. Indeed McDowell’s next sentence (which he put in parentheses) speaks to the contrary. If no stretch of walking need be over when someone has walked, then t here need be nothing fitting the idea of “a w hole” so long as someone is walking. If the w hole of some object, res, were r eally t here at any moment when someone is walking, then it seems we should have to say, absurdly, that there are ( / were) indefinitely many res whenever someone is ( / had been) walking. The idea of an activity’s being “all t here” is even harder to accept when an activity is described with a telic verb—as in John’s walking across the street. Certainly “is walking across the street” and “has walked across the street so far” are both predicable of John when he is walking across the street. But McDowell spoke of “the particularity that comes with completion.”16 And when someone is on their way across the street but not yet across it, there is as yet no completion of the sort required for there to be a particular: completion requires them to have got across. As McDowell said, “It may be true of one that one is crossing a street though one is not going to cross the street; perhaps halfway over one is going to be knocked down by a bus.”17 Indeed. Imagine someone who actually was knocked over by a bus. In such a case there is activity but no action. The imperfective “was crossing the street” is true of them, but not the perfective “crossed the street.” (The difference here, in respect of what is true of the person, is a difference of aspect rather than tense.) I quoted McDowell as urging an asymmetry between things that have happened and things that are happening. It seems now that in order to find such an asymmetry, one must have a use for the idea of completion of the sort t here is when an agent has reached an end of theirs.18 And the partic ular associated with an agent’s reaching of their end is only brought with the perfect tense of a telic verb. It surely must be doubted w hether someone’s walking across the street can be conceived as an object all there at any moment.
180
R eas o n in P ercepti o n and A cti o n
2.3. How does this affect the picture of intention in action that McDowell presented? here can certainly be something initially compelling about portraying T intentions as McDowell did—as a kind of continuant. If we think of a person’s prior intention to do something as something there is, then we won’t think of it as something which simply evaporates when it comes to be executed and they start to do what it was they’d intended to do. So it is easy to think of an intention to walk across the street as persisting until it comes to be an intention in action, when, in McDowell’s terms, it comes to be such as to change its shape. And if this is how we think, my objection to McDowell’s treating activities as he does could seem to be beside the point. Suppose that someone intended to φ. All that McDowell’s picture requires, it might now be thought, is that their intention to φ should come to relate them to activity consisting precisely in them φ-ing. W hether or not that which they may come to have done belongs in the category of object—is a res—might now appear to be neither h ere nor t here. If one thinks in this way, one assumes that a person’s intention to φ remains as it is until executed and is being executed just so long as they are φ-ing. It is then as if any intention an agent might have to do something would come to be fulfilled if they started to do and carried on d oing the thing. But it cannot be right to assume that any intention for the f uture (unless abandoned or forgotten) persists as it is from the moment it is first formed u ntil the moment it starts to be executed and then on u ntil it has been executed. This assumption could only seem right to one who forgot about the reasoning of human agents, the range of things they are able to do, and the means-end structure of intentional action. Before someone sets out on doing something they intend, they w ill often need to come to know what t hey’ll do in order to do it. Having accepted the invitation, you intend to go to the conference; and long before you take any actual steps which might count as your actually g oing there, you wonder which airport you want to land in on your way. You might come to intend to get to Pittsburgh by (inter alia) taking a plane to Toronto (where perhaps you can stop long enough to meet up with a friend). Or again: you intend to eat an apple, and before you begin to do so, you form the intention to eat this apple, the one you’ve set your eyes on in the bowl you’ve started to move toward. The filling in of means is not the
N aturalism in the P hil o s o phy o f A cti o n
181
preserve of prior intentions alone. Even when someone has started to act with a view to some end, there can be details to be settled. Someone who had decided to make tea might have put the k ettle on but not yet have opted for black tea or green tea. Or consider someone who is painting a picture. It could be that the character of the brush strokes they make at any time depends as much upon how they imagine their work might come to look in the f uture as it does upon how it then looks. As they paint, they have no recipe, and no intention to do this, then that, then the other determinate t hing. Someone who paints a picture is likely to do so thoughtfully. We do many things without much thought, without giving consideration to how to be d oing them. Still, that is not to say that we do them unknowingly. I made use of an example in which I was crossing a street and you asked me why I should be. U ntil you did, I might not have been paying much attention to where I’d got to or where I was going. When you asked, it came to my attention that I was going to the library. So long as I was crossing the street, I might not have settled w hether, having got across it, I would walk a bit north and then go west, or would immediately turn left and then, before long, head north. If I arrived at the library in due course, I relied on my knowledge of ways of getting t here, or perhaps on knowing that I could look up an exact way while en route. If it’s granted that it is seldom that anyone crosses the street just for the sake of it, it must be appropriate to consider my street crossing as part of a larger project. So we have an example in which d oing something is among the means of getting something else done. Such examples, in which one needs to do one t hing and then another and then perhaps a third and a fourth and a fifth t hing, and where it need have been no part of one’s original project to do all of these things, make a problem for McDowell’s picture of intentions as continuants. McDowell said that his “image of intentions as a kind of continuant, [although] helpful for some purposes . . . should not encourage us to conceive intentions as self- standing items.” Indeed, I suggest that the image could at very best help us think about isolated intentions—as when I get from one side of the street to the other simply by walking across it. Only if I intend nothing more than to walk across the street is that alone what I might need to give thought to. But could anyone leading a life ever be thought to intend just this?
182
R eas o n in P ercepti o n and A cti o n
3 3.1 Of course McDowell d oesn’t really think that an understanding of intention can be got from treating isolated intentions alone. T here must be more to be learnt from thinking about intentions as they play a role not in isolation but in the course of life. McDowell has his own example of a project with a more complex shape than that of simply walking across a street, which he turned to in order to show why we should pay heed to the passage of time. I’m going to suggest that his example shows why we must pay heed to an agent’s own temporal perspective. McDowell’s was an example of baking a cake: Suppose I am baking a cake. Perhaps a God’s-eye view, from outside the temporal unfolding of events, has in its scope a cake that is going to result from my cake-baking. . . . [I]t cannot be right to suppose that . . . “I am baking a cake”—something intelligible only at a time- bound perspective— has a relational form, amounting to “There is a cake that I am baking.” What one says when one says “I am baking a cake” is not proved false if no cake results from one’s activities. (Consider “I was baking a cake, but I was interrupted.”)19 Indeed: “I was baking a cake” does not entail “there is or was a cake x such that I was baking x.” And just as there is no cake until a cake has resulted from my activities, so “my baking of a cake” does not denote a particular until I’ve produced a cake. Well, actions no more spring into existence than cakes do. So the example suggests that the question what it is for a partic ular to come to exist concerns the events of the temporal world as much as the enmattered objects of the spatial world. “Bake a cake,” “build a h ouse,” and the like—so-called verbs of creation—have been thought to introduce some special difficulty for the theorist. But however one might attend to any such difficulty, the point here, about events, has to be a quite general one. It is only with activity through the passage of time that t here comes to be a particular which is someone’s baking of a cake.20 So, equally, it is only with the passage of time that t here comes to be a particular which is someone’s walking the w hole way to school, or which is someone’s folding of all the flour into the fat. In order to grasp that it is never the case that there is a
N aturalism in the P hil o s o phy o f A cti o n
183
genuine particular until time enough has passed for the agent to have finished, one has only to invoke the possibility of the agent’s being stopped. “I was doing such-and-such, but I was interrupted,” as McDowell says. When McDowell turned to his cake-baking, he remarked that “we need to introduce t hings to do with more complex relations to what one has already done.” “Having beaten the eggs,” he said, “one needs to fold in the flour and so on until the cake is baked.”21 Yes. But unless we take a God’s-eye view, we have to think of someone who could say “I am baking a cake,” and who needs themself to do further t hings in order that they should have baked one. They know what they are d oing, and can decide how to carry on. They may be relying on a r ecipe which o thers can use. But their role throughout is that of one who is in a position to know at any time what they themself intend to be d oing. (You might ask them what they intend to do next.) 3.2 Even if I disagree with McDowell about how we should think about what is involved in a project with much greater complexity than that of getting across the street, of course I take him to be right to say that “intentions in action are not directed de re at particular actions.” Indeed this now appears to be doubly true, as it were. First: so long as someone’s intention to do something is coming to be fulfilled, t here is as yet no relevant particular event for it (or for anything else for that matter) to be directed at. Second: anything that someone does, in the course of a multistep project as it might be, they no longer intend to be d oing once they have finished d oing it. This is already enough to put McDowell quite at odds with Davidson, and indeed to show why the story of agency told by those said to be the most minimal naturalists must be rejected. Davidson said that a belief which combines with a pro-attitude to yield an action is a belief that the action described in a certain way has a certain property.22 When he turned on the light, his belief was one which (paired with a want) he said caused his action—caused an event of his turning on of a light. But this belief had as its content that this event has the property of being just such an event. The problem with this is apparent. Until he has turned on the light, there is as yet no action for his belief to be about. If the problem does not show up clearly, that may be partly b ecause we are encouraged to consider t hings people do that it takes very little time to do. At any rate, we are invited to take what McDowell spoke of as a “God’s-eye view,” as if from outside the
184
R eas o n in P ercepti o n and A cti o n
temporal unfolding of events. That makes it easy to confuse someone’s belief relating to something they may come to have done with a belief about something that comes to exist only when they have done it. But in truth there could be no time at which an agent’s belief about how to do something presupposed the existence of a particu lar that did not yet exist. And likewise for intention: there could be no time at which someone’s having done something was presupposed to their intention to do it. To use McDowell’s words again: someone’s intention’s being an intention in action cannot consist in the presence of a de re relation to a particular action. ere brought into being ahead of Davidson treated actions as if they w their being caused by such m ental states as he took to cause them. He told a story of action which requires retroactive causation. The story is found in many variants, thanks to modifications made over the years by Davidson himself and by others. But we see that it must be rejected because it is unable to give any account of the predication of telic verbs with imperfective aspect. Evidently the story cannot provide an account of the agency of any animals that can make progress toward their ends. The story results from an attempt to subordinate the teleological character of agency to a certain conception of causal explanation. Those who tell the story often give the name of “means-end beliefs” to such beliefs as they take to play a crucial role in causing h uman actions. But an agent’s belief about how to do something—about means to their end—serves their purposes only insofar as it concerns how they themself may act at a time, situated as they know themself then to be. And it seems that if they are actually to succeed in reaching their end, except luckily perhaps, then they must know how they themself at the time of acting may do what they intend. Their perspective isn’t God’s. And they cannot rely just on facts about how something may be done which might be learned from a recipe. (First one beats the eggs. Then one folds in the flour . . .)
4 4.1 I quoted from Mind and World, where McDowell said that t here w ere “intentional bodily actions” in which “conceptual capacities” were “inextricably implicated.”23 And now again, in his “Remarks,” McDowell
N aturalism in the P hil o s o phy o f A cti o n
185
states his conclusion as one about bodily action. Of course a philosopher who works in a certain Cartesian tradition, who might find a problem about causation across a putative boundary between m ental and physical, w ill wish to pay particu lar attention to what’s bodily. But McDowell is very surely not such a philosopher, and indeed he presents his “Remarks” as “stand[ing] in contrast with familiar philosophical pictures of the relation between mind and body.” One might then expect him to think that someone who answers questions about why they are acting as they are, or how they w ill do what they need to do next, places what they do in an order of reason even if they make no allusion to their body or any of its parts. McDowell introduced motor intentions when he spoke of intentions as changing their shape: “Intentions to do t hings like crossing streets change from one determinate short-term motor shape to another as they express themselves in acting.”24 But how are motor intentions to be conceived? Is the “—” of “A is—ing intentionally” sometimes to be filled with words making mention of a part of A’s motor system? That which one is d oing intentionally, one intends to be d oing. But that which one intends to be doing, one ordinarily knows oneself to be doing. And few of us know anything about our motor systems, whose ins and outs have no place in saying why we should be acting as we are. McDowell also introduced “the ability to move one’s limbs as executing one’s intentions require.” There can be no quarrel with the idea that a human agent is able to move their body as required. But they also have a vast range of knowledge and skills whose specification makes no mention of limbs or of any other body parts—whether to find their way to the library or to bake a cake, or to speak English . . . No doubt one’s motor system is busily at work when one makes bodily movements of any sort. But as McDowell says himself, the determination of the workings of the motor system “is not a task for the practical thinking that intention belongs to.” No more, it seems, must the movements of bits of their body be caught up in the practical thinking of someone who is in a position to do something they are able to do or to exercise one or another skill. We have learnt how to do things which require our bodies to move, but we have no need to have learnt exactly how our bodies must be moving in order that we should be doing them, and if we have learnt, we might do better to have forgotten. I think that McDowell relies on his reification of activity to uphold the conception of intention found in his “Remarks.” When he invites us to think of “one’s intention concerning one’s current relevant activity,” it
186
R eas o n in P ercepti o n and A cti o n
seems as if that which one is d oing might be a t hing subject to change thanks to one’s having an intention which concerns it.25 But when someone is doing something intentionally, their d oing it is ongoing; and their intention can be simply to be doing it, along with such other things as conform with their current plan—as, for instance, someone who is crossing the street may have planned to go to the library in order to read the book they reserved the previous day. 4.2 At this point I should acknowledge a footnote in McDowell in which he questions whether the “exteriorization of intention stops at the bound aries of one’s body.”26 He mentions Anton Ford in the footnote, and Ford supports a conception of human agency found in G. E. M. Anscombe’s Intention, as McDowell himself surely does au fond.27 In light of this, it might seem that all I have done is to press upon McDowell a concession that he may be ready to make. What I have meant to argue, however, is that his image of an intention as a kind of continuant stands in the way of his making the concession. When McDowell articulates his image of intentions, he says: “One intends to do whatever is needed in order to finish d oing what one is d oing.” I presented examples in which an agent in the course of doing something may not yet know at all exactly what they w ill do in order to have finished. I note now that in order to make room for such examples, it w ouldn’t do any good to say that one intends to do whatever might prove to be needed in order to finish doing what one is doing. That might make allowance for the fact that even when one has started in action, one may still have work to do to determine how to carry on. But it could falsely suggest that whatever emergency might arise, and whatever inexcusable t hing one might find one would need to do in order to carry on, one should not be distracted from one’s project. So why does McDowell in his “Remarks” fail to allow that someone plays the role of rational agent throughout the execution of any of their intentions? I think that it is a consequence of his swallowing the idea, which he takes over from his adversary Searle, that intentions might be factored exhaustively into t hose that are prior to actions and t hose that are contemporaneous with the actions they are in. If my sole end w ere to be
N aturalism in the P hil o s o phy o f A cti o n
187
safely on the other side of the street, then the only intention I would need would be an intention to cross the street when safe for me to do so. But if I have a reason to get to the other side of the street—suppose that I won’t get to read the book I reserved at the library unless I continue across, then I now intend to be g oing to the library and to be crossing the street and to do whatever I intend to do at the library. When account is taken of an agent themself who needs to do one t hing and then another and another if an intention they have is to come to have been fulfilled, the distinction between f uture intention and intention in action, as McDowell takes it over from Searle, breaks down. (My intention to cross the street and my intention to get to the library were for a patch of time harbored simulta neously, but once I’ve got across the street, I no longer intend to cross it, although I still intend to be going to the library.) 4.3 Three further consequences of McDowell’s starting from the example of someone waiting on the side of the street for the light to turn green are worth noting. In the first place, it leads him to base his account of intention in action on a case in which an episode of someone’s activity comes to fruition (if it does) in what w ill be reckoned a single action. For Davidson, there are no other cases. For Davidson, nothing more is needed for a person’s action than a suitably caused movement of their body. Of course Davidson allows that, thanks to a bodily movement’s being in its turn the cause of various effects—thanks to the work of nature, as he would have it— an agent may do many things by making a single movement. But in a story like Davidson’s, there is no place for an example, as any lifelike example is bound to be, in which one’s reaching one’s objective is a m atter of one’s taking various means oneself, each in turn. By working with his image of intentions as continuants, McDowell may distract us from appreciating quite how badly astray Davidson has led philosophers. Also—a nd this now is the second unwanted consequence of the agendum McDowell takes over from Searle—the street-crossing example may lead one to buy into the assumption that the job of a philosopher of action is to say what it is for a bodily movement to be initiated. But given that one may remain stationary as one waits for the light to turn green and that one sometimes stops moving when thinking how to proceed, a
188
R eas o n in P ercepti o n and A cti o n
philosopher who starts with this assumption w ill at best provide an extraordinarily disjointed account of someone’s activity, taking it to cease whenever they are motionless and abruptly to break out when they start on d oing something. Knowing what it might amount to for the body of an intentional agent to start to move could hardly be the key to understanding the execution of intentions. Now for the third consequence. McDowell speaks of a prior intention as determining when the time for action arrives. Of course we would be unable to coordinate with one another, or to take the cake out of the oven when it’s cooked, u nless we could keep track of time. But starting to act on an intention one has is not always a matter, as McDowell says, of “the operation of an ability to keep track of time.” An intention need not be an intention to do something at a determinate time. You might intend to eat lunch at home t oday, but leave it open not only what to eat but when to start to prepare it. We can be spared from constantly monitoring time, and often enough we have no need to act on a signal. A realistic picture of a human agent must be of someone who may have many projects on the go at any time, and whose intentional activities cannot be accounted for only one by one. They lead a life, equipped with powers not only of self-movement, but also of perception and of thought and of imagination, any of which they may bring into play at any time they are awake. I think that McDowell’s treatment of his street-crossing example leads him not only to restrict the actualizations of our active nature to our bodily actions, but also to lose sight of the extended prospective perspective of any a ctual h uman agent.
5 When McDowell said that “knowledge and intentionality can be in view only in the framework of the space of reasons,” he sought to show how the skeptical questions of traditional epistemology can be set aside. The knowledge then in question was knowledge of how t hings are, and the intentional states of concern were “the propositional attitudes.”28 Let practical knowledge encompass the knowledge a person has of what they intend to be d oing or to get done, and the knowledge they have of how they can do this or that, and the knowledge they have of their own abilities and
N aturalism in the P hil o s o phy o f A cti o n
189
skills. If such knowledge has a place in the space of reasons wherein rational structures are found, there cannot be any obstacle to seeing a person’s conceptual capacities as implicated in all of their intentional activity, and their actions. Whereas the restricted naturalism imposed by the event causal framework cannot provide an account even of nonhuman animal agency, a liberal naturalist takes the natural world to be one from which “the mind” (that is, “the h uman mind”) could not possibly be segregated. I hope I made it clear at the outset that I didn’t mean to bring into dispute the mass of work by McDowell on human agency. In talks, in lectures, and in other publications, McDowell has engaged illuminatingly with many questions that have arisen in connection with Anscombe’s Intention.29 I’m more than sympathetic to his answers to t hese questions, but I find them to be in conflict with some of his “Remarks.” It may well be that McDowell no longer stands by everything he says there. But I hope to have shown that once one has abandoned philosophical prejudices such as McDowell counteracted in Mind and World, one can readily stand firm against any story of action of a kind to which he is clearly opposed. Insofar as a restrictive naturalism posits a world which cannot contain animal agents, it ought to be obvious that it cannot contain human beings, who, each one, like any animal, lives and acts in time, but also is apt to know what they themself are doing at any time.
✣
Pa rt I I I
✣
CONSEQU ENCE S FOR M ETA PH YSICS
Chapter Seven ✣
Perceiving the World ✣
Sebastian Rödl
M
stislav Rostropov ich, arguably the greatest cellist of the twentieth c entury, disclosed in an interview that it was his practice to play, every morning before breakfast, one of Johann Sebastian Bach’s “Suites für Violoncello allein.” There are six, one for each day of the week. Rostropov ich would rest on Sunday. If I were to give an interview, I would explain that she who desires to be a philosopher is to make it her practice to read, every morning, before breakfast, one of John Henry McDowell’s lectures on Mind and World. There are six, one for each day of the week. The philosopher may rest on Sunday. Just as no cellist w ill never be done playing these suites, so the philosopher w ill never be done reading these lectures. But I can give an interim report. In Mind and World, John McDowell proposes that we conceive perception in a certain way. According to that conception, what we perceive bears this form: we perceive that things are thus and so. As this is what we perceive, perception provides us with knowledge: perceiving that t hings are so, we can exclude that they are otherw ise and thus know that they are so. Through this conception of perception, and through it alone, we comprehend ourselves to know an objective world, a world that is as it is whether or not we know it to be so. 193
194
C o nsequences f o r M etaphysics
The above representation of McDowell’s proposal uses the first-person pronoun. And this is crucial. Mind and World does not approach knowledge and perception from sideways on. It seeks to comprehend knowledge from within: from within the knowledge so understood. The first-person pronoun indicates that what is expressed by means of it is thought, or known, from within. McDowell stresses that his account entails that a subject of knowledge and perception is self-conscious.1 This understates the significance of self- consciousness. It presents self-consciousness as characterizing that of which the account is an account: it treats of the kind of knowledge that belongs to self-conscious thinkers. Now the account certainly is of self- conscious knowledge. But it is that only b ecause it is nothing other than the self-consciousness of knowledge: it unfolds the concept of knowledge that is internal to knowledge, the concept thought in “I know”. The account is not other than that of which it is the account. This marks it as from within. Recently, anglophone philosophy has taken an interest in self- consciousness, an interest in it as an interesting phenomenon alongside others. (It has become an “area of competence”.) That self-consciousness is not an object of philosophy, but philosophy—as self-consciousness is its own comprehension and philosophy the bringing to language of this comprehension—this idea is foreign to the anglophone discourse of the past decades. Yet it underlies McDowell’s work. The physiognomy of his work, on language or on thought, on knowledge or on virtue, everywhere reflects his uncompromising adherence to this idea of philosophy: it seeks understanding from within. It seems to me that his reception has yet to take the measure of this orientation of McDowell’s thought. McDowell identifies an obstacle to accepting that what we perceive, at the deepest level, is that t hings are thus and so. It is a conception of nature as the realm of processes governed by natural laws. If we conceive of nature in this fashion, then we cannot place the subject of knowledge and perception in nature. Knowledge and perception are not comprehensible as processes governed by natural laws. McDowell goes on to propose that we enlarge our conception of nature; enlarging it in the right way, we can place in nature the subject who knows nature. This strikes me as an error. It is an error not because the subject is outside nature, in a diff erent place, as it were, supra nature.2 It is an error
P erceiving the W o rld
195
ecause the subject of knowledge and perception cannot be placed in b anything. This is a consequence of the impossibility of thinking knowledge from sideways on; it reflects the inner universality of knowledge and therefore of the human individual.3 Sections 1 to 3 develop the idea of knowledge comprehended from within. Sections 4 to 6 recapitulate the train of thought by which McDowell brings out that, comprehending ourselves to know how t hings are indepen dently of our knowing them to be so, we think of perception in this way: as perception of how t hings are. Sections 7 and 8 further unfold that idea of perception. It emerges that with it, we have left b ehind the frame that seems to require us to choose whether we place the human being in nature or above it.
1. Knowing and Knowing That One Knows It seems clear that it is one t hing for things to be a certain way, another for someone, Steffi, say, to know that they are that way. U nless things are that way, Steffi cannot know that they are. But it does not suffice, in order for Steffi to know that things are that way, that they be that way. Let us say that for Steffi to know that p is for it to be the case that Kp. While Kp contains p, it goes beyond it; it holds further matter, pertaining to Steffi. Kp is p and more. This raises a question. Must Steffi, in order to know p, know Kp? Must she, in order to know something, know that t hings are as they must be in order for her to know it? It makes no sense to answer “yes”. For it makes no sense to say that, in order to know something, one must know more than it. It equally makes no sense to answer “no”. For then someone may know something while failing to know that she knows it. And then t here is no such thing as saying “I know.” Let it be that I know p. I do not thereby know that I know that; I cannot say w hether I know it or not. However, if I cannot say that, I cannot assert p. This is not intelligible: “P, certainly. But I cannot say that I know that. For all I know, I have no idea whether p or not.” It may seem that I may come to know what I h ere declare myself not to know, namely, Kp. So let it be that I know Kp. However, as I do not know that
196
C o nsequences f o r M etaphysics
I know that, I cannot assert it. I cannot say “I know p.” What we just said about “p” holds for any sentence, among them “Kp”. As both answers to the question make no sense, the question makes no sense. Yet it seems it must make sense. It proceeds from the thought that it is one thing for things to be a certain way, and another for them to be known to be that way. And this is nothing more than: t hings are as they are anyway, whether or not they are known to be so. Still, the question cannot be all fine. There are these two t hings, we say: things’ being as they are and someone’s knowing them to be so. Our question represents both as objects of knowledge: it asks how knowing the one stands to knowing the other. Thus it treats someone’s knowing something as a reality to be known next to the reality of things’ being as she knows them to be. But if knowledge were a certain reality, why would one speak of knowledge and reality? If the mind were one of the things in the world, why would one speak of mind and world? We asked: Must one, in order to know something, know that one knows it? The question suggests that knowing that one knows something is knowing more than it. If this is right, then it makes no sense to answer “yes”. Yet the history of philosophy is replete with figures who seem to think that she who knows something knows that she does so. It follows that they reject the condition: that knowing oneself to know something is knowing more than it. If we see how to reject this, we see what is wrong with the question. It is one t hing for things to be thus and so; it is another for Steffi to know that they are, we said. This seemed clear. Yet when we say, it is one thing to know that t hings are so, it is another to know oneself to know that, this is not clear. For if in this sentence “oneself” is a first-person pronoun, then the sense of “I know”, which that very sentence deploys, dissolves. This we saw when we considered the negative answer to our question: knowledge that one knows something cannot be added to knowledge of it, for “I know p” is already implicated in the assertion, expressive of knowledge, of “p”. Consider this dialogue: “P.” “Really? Do you know that?” “I do not know. Perhaps not. Ask someone e lse. I was only speaking of p. I am not competent to say whether I know p. This is a different question.” This makes no sense. If I face the question w hether p—let us imagine the question is put to me—I cannot answer “Yes, p, but I have to leave it open whether I know that.” I cannot say this b ecause, answering “Yes”, I un-
P erceiving the W o rld
197
derstand myself, and am understood by her to whom I speak so to understand myself, to say something I know. If we see why this is so, we understand why knowing something is nothing less than knowing oneself to know it.
2. “I Know” Answering a question someone has put to me, I understand myself to answer her question. It is not that, on the one hand, I answer her question, and, on the other hand, understand myself to do that: answer her question.4 More generally, when I assert that t hings are so, w hether or not I do so in response to a question put to me, I understand myself to assert this. It is not that, on the one hand, I assert that things are so, and, on the other hand, understand myself to do that. If I did not understand myself to assert that t hings are so, then asserting this would not be what I do. The same holds for the act of mind I express in asserting that t hings are so: judging that things are so, I understand myself to do that, judge that things are so. Moreover, asserting that t hings are so, I take it to be right to assert what I, thereby, assert. The term “thereby” indicates that it is in asserting something that I take it to be correct to assert it. In the same way, I take it to be right to judge as I, thereby, judge. I not only understand myself to judge; I understand myself to judge correctly. So I do because I judge with, or through, the concept of judgment. Thus I comprehend my judgment to conform to its concept; I conceive it to be as it is to be, simply as judgment. The notion of correctness or rightness we used, saying that I think it right to assert what I, thereby, assert, and that I think it correct to judge as I, thereby, judge, signifies a measure of perfection to which a judgment or an assertion is subject being what it is: judgment or assertion. A judgment that conforms to this measure is as it is to be, simply as judgment. It conforms to what it is, or to its concept. We can put this by saying that judgment and its validity are self- conscious. “Self-consciousness”, here, does not signify consciousness of a special object, a self, should there be such a thing. It signifies the internality of the consciousness to that of which it is the consciousness. Judgment is self-conscious: I am conscious of judging as I, thereby, judge; being conscious of judging as I do and so judging are one act of the mind. The
198
C o nsequences f o r M etaphysics
validity of a judgment is self-conscious. For, I think it valid to judge as I, thereby, judge; thinking it valid to judge as I do and so judging are one act of the mind. As a valid judgment is understood to be valid in this very judgment, the thought of its validity is internal to its validity.5 The main perfection of judgment is its truth: it is correct to judge that things are so only if they are. A valid judgment is true. So in judging something, I take myself to judge truly. But this is not all. I take it to be no accident that it is correct to judge as I do. So I do b ecause I conceive my judgment to be correct in this very judgment. Hence I do not think that there is a condition of the correctness of judging as I do that lies outside that of which I am conscious in so judging. If there were such a condition, I would need to recognize that it is satisfied in order to think my judgment correct, and this recognition, ex hypothesi, would be a diff erent act of mind from that judgment. That of which I am conscious in judging is what I judge. And what I judge is not a circumstance in which my judgment may find itself, not an accident, or something that befalls my judgment. As I take it to be valid to judge as I do in so judging, I understand my judgment to be nonaccidentally true, or knowledge. In judging, I understand myself to judge, judge correctly, know. So I understand my judgment in this judgment. This understanding of my judgment is internal to my judgment; it is internal to what it understands. Therefore we may call it self-understanding, using the prefix “self-” as above: not to indicate what is understood, but to indicate the internality to what is understood of its being understood. The first-person pronoun expresses the internality to what is thought of its being thought. When someone thinks “I am F”, then she who thinks this is she of whom she thinks it. And not only does this identity obtain in fact. If what she thinks requires for its expression the first-person pronoun, then this identity is contained in what she thinks. Hence, what is thought in the first person contains its being thought by her whom the thought concerns. Conversely, if someone’s being or doing something contains its being thought by her, then this thought of hers is a first- person thought. Hence the expression in language of the understanding of judgment internal to judgment—its self-understanding—is “I judge”, “I judge truly”, “I know”. Now we see why, answering the question “p?” by “p”, I understand myself to express knowledge: answering a question, I understand myself to
P erceiving the W o rld
199
answer this question, and I take my answer to be correct. I think it right to assert what I, thereby, assert. And this means that I understand myself to know what I assert.
3. What Is Known in Knowing That One Knows Our question was: Must someone, in order to know something, know that she knows it? The answer “no” dissolves the thought expressed by “I know”. For “I know that t hings are so” expresses an understanding of one’s judgment that is internal to that judgment. Understanding myself to know that t hings are so is not a diff erent act of mind from judging that they are; I understand myself to know what I, thereby, judge. Hence, if I do know that things are so, then my knowing this and my knowing myself to know it are one act of mind. “I know”, if it gives voice to knowledge as opposed to error, gives voice to knowledge that is knowledge of itself. Thus we destroy the “I know” if we lay it down that knowing something may be without knowing that one knows it. Must someone, in order to know something, know that she knows it? We cannot answer “no”, if we speak of knowledge capable of being expressed by “I know”. For such knowledge is knowledge of itself. For the same reason, we cannot answer “yes”. The question presupposes that knowing that one knows something is knowing more than it. However, “I know p” expresses an understanding of myself as knowing p that is internal to my knowing p. (Provided I do know it. I drop this qualification in what follows, taking it to be understood.) This understanding is not a separate act of mind from the one that it understands. As knowing something and knowing that one knows it is one act of mind, no more is known in the latter than is known in the former. The right answer to our question rejects that question. It is wrong to say that someone must, in order to know something, know that she knows it. For this suggests that knowing that one knows something is knowing more than it. And knowing something is knowing oneself to know it.6 Knowing oneself to know something, being nothing other than knowing it, is not knowing more than it. The question, as it presupposes that knowing that one knows that things are so is knowing more than that things are so, attempts to think knowledge from sideways on. The
200
C o nsequences f o r M etaphysics
meaninglessness of the question reflects the fact that this attempt falls to the ground if it is directed at knowledge that is capable of being expressed by “I know”. The question lets us think of someone’s knowing that things are so as a further thing to be known, additional to what she knows, knowing that things are so. While this seemed sensible, it jarred with a sense that the concepts of knowledge and reality, mind and world, are on the same plane, a sense expressing itself in phrases such as “knowledge and reality” and “mind and world”. Now we see that plane. Knowing something, I know myself to know it. Thus I understand what I know to be such as to be known. I understand it as something that is, is real, reality. In knowing, understanding myself to know what I thus know, I think myself and what I know through the concepts of knowledge and reality. This is the origin of these concepts. They must be understood together and from this source: from within. We observed that many figures in the tradition take it that she who knows something knows that she knows it. They do not propound this as a thesis for which they marshal evidence or provide arguments. They treat it as a matter of course.7 So they do because the knowledge that is their topic is knowledge capable of being expressed by “I know”. This is explicit in the question governing their reflections: What can I know? It may be said that t here is knowledge that may be without knowledge of itself; it may be said that we have such knowledge. We need not decide whether there is such knowledge, nor who, if anyone, has it. We concern ourselves exclusively with knowledge that can be expressed by “I know”. So does McDowell. For he seeks to comprehend knowledge from within as opposed to from sideways on. “From within” means: from within the judgment that is to be knowledge. An inquiry into knowledge from within seeks to bring to the light of explicit language the understanding of knowledge that is internal to knowledge, the understanding expressed by “I know”. It is not that McDowell thinks there are two ways of comprehending knowledge (from sideways on and from within) and recommends the latter over the former. Rather, he holds that there is no understanding knowledge from sideways on. And this is indubitably right if the topic is knowledge that can be expressed by “I know”. Such knowledge is its own understanding and therefore is nothing but what it is understood to
P erceiving the W o rld
201
be in that understanding. McDowell follows t hose in the tradition who treat it as a matter of course that knowing something is knowing oneself to know it. Urging that we approach knowledge from within, he makes explicit what they did not think of saying, as it was part of the idea of philosophy that formed the medium of their reflections.8 Understanding knowledge is understanding it from within. Therefore, should we arrive at an account of knowledge, saying what it is in virtue of which someone knows when she does so, we w ill state, and state only, what is known by her who so knows. This is to say that our account w ill say only what anyone knows in knowing anything at all. Consequently, it cannot be understood to be a theory to which other theories might be opposed. There is no such thing as saying “no” to it: if the “no” were knowledge, it would, knowing itself to be knowledge, be knowledge of what the account says. In view of this, the account may be called quietist.9 For it leaves every thing as it is. This holds true of everything in Mind and World and of everything put forth in the remainder of this essay to the extent that it achieves its aim: to understand knowledge from within.
4. Knowledge of Things as They Are Anyway; The Concept of a Basis of Knowledge It is one t hing for things to be a certain way; it is another for someone, indeed, for anyone, to know that they are that way. We must not take this to invite us to approach knowledge from sideways on. Understanding knowledge is understanding it from within. Thus we must understand this from within: that it is one thing for things to be a certain way, another for them to be known to be that way. We must see how this is something we know in knowing what we know. We must see it to be contained in the self- understanding of knowledge, in what we understand knowledge to be, saying “I know”. It is one thing for things to be a certain way, it is another for them to be known to be so. If this is right, then it is known in any knowledge. In knowing that t hings are a certain way, I know myself to know a way which things are anyway, whether they are known to be that way or not. Let us consider how I know this.10
202
C o nsequences f o r M etaphysics
In judging that t hings are so, I understand myself to judge that they are so to the exclusion of judging that they are not so. Thus, in judging, I think the opposite judgment; I think it as the judgment I exclude. Hence, I understand what I judge to be such that there is saying “yes” to it and there is saying “no” to it. What I judge provides for opposition in judgment. Equivalently, I understand what I judge to be a way for things to be such that t hings may be that way or not. What I judge provides for opposition in being. In this sense my judgment may be valid or invalid, true or false: it may be invalid in the sense that t here is such a thing as g oing wrong with respect to what I judge. What I judge is such that the judgment of it is liable to error. In judging, I exclude the opposite judgment, thus understanding what I judge to be such that judging it may be right or wrong. This is to say that what I judge does not, on its own, provide insight into the correctness of judging it. If it did, there would be no such thing as erring with respect to it. For the opposite judgment would be of something that reveals the incorrectness of judging it; it would be judging something to judge which is to recognize it to be wrong so to judge. And t hese words do not add up to the description of an act of judgment. Furthermore, if the way I judge things to be provided for the recognition of the correctness of judging it, then there would be no leaving open the question w hether things are so. I could not take myself to be ignorant with respect to it. Hence, if what I judge provided for the recognition of the validity of judging it, I would, in judging it, understand t here to be no difference of things’ being as I judge them to be from my judging them to be so. Conversely, I understand there to be such a difference as I understand the way I judge things to be to be such that there is such a thing as judging wrongly, or being ignorant, with respect to it. This is how in judging (and thus in knowing, should my judgment be knowledge) I understand t hings’ being as I know them to be to be one thing, my knowing them to be so, another. This gives rise to the concept of a basis of knowledge. In judging, I understand my judgment to be valid. But what I thus pre sent as valid to judge is such that judging it may be valid or invalid. This has been expressed by the disconcerting metaphor that judging is “sticking one’s neck out”. What I judge is the m atter of a yes-or-no question. In judging, I answer this question. I say “yes”, thinking that this is the right answer. But how can I think that this is the right answer? What I judge is
P erceiving the W o rld
203
such that judging it may be valid or invalid. Given only what I judge, namely, that things are so, I have to say, things may be so, or not. But I say that they are so; I exclude that they are not so. What I judge does not provide insight into the validity of judging it. I do not, through what I judge, comprehend that things must be as I judge them to be, that they cannot be otherwise. Hence, in order to establish that things are as I judge them to be, I need to appeal to something other than what I judge. I need to appeal to something that shows that things are as I judge them to be and rules out that they are otherw ise. This is the origin of the concept of a ground: a ground on which I judge is to reveal t hings to be as I judge them to be. If indeed it shows that, it is a ground on the basis of which I know things to be so. I know p on the basis of q, that is: in the light of q, I recognize that this is how t hings must be: p. The fact that p does not shine that light upon itself; the fact that q shines that light upon the fact that p. This concept of ground is contained in the concept of knowledge that I deploy in “I know”. It is contained in the “I know” insofar as what I understand myself to know does not on its own provide for the recognition of the correctness of judging it, and that is, insofar as what I know is as it is, and is understood by me to be as it is, anyway, w hether or not I know it. A ground on the basis of which I know p is something in virtue of which it must be the case that p. Therefore a ground is something that is the case. Only something that is the case can render it necessary that something is the case. And not only this. A ground is something on the basis of which I know something, and I can know something only on the basis of something I know. Hence a ground not only is something that is the case; it is something that I know to be the case. Knowing q to be the case, I know p on the basis of q. It has been said that someone who knows something on a certain basis need know nothing of this basis. Someone who says this fixes it that she does not speak of knowledge that is capable of being expressed by “I know”. If that is the topic, then a ground on the basis of which someone knows something is something on the basis of which she understands herself to know it. Such a ground is something she knows. Judging on a ground is understanding that ground as a ground for judging what one, thereby, judges, and judges on that ground. This has been expressed by saying that a judgment is a response to reasons as such: a response that is a recognition
204
C o nsequences f o r M etaphysics
of the reason to which it responds as a reason. As judging something is the same as thinking it valid to judge it, my recognition that something shows it to be valid to judge as I, thereby, judge, is nothing other than that judgment.
5. The Oscillation It is one thing for things to be as they are; it is another for them to be known to be so. Things are as they are whether or not they are known to be so. We set out to see how this articulates an understanding of knowledge that is internal to the knowledge so understood. We found that it is internal to judgment as, in judging, I understand what I judge to be something judging which may be valid or invalid. What I judge does not provide insight into the validity of judging it. Yet, in judging, I understand it to be valid to judge what I, thereby, judge. As what I judge does not provide for this understanding, the question arises how I can so understand it. We answered: I understand it to be valid to judge as I do in the light of something that reveals that things must be as I, thereby, judge them to be. That is the basis on which I judge what I do, therein understanding myself to know it on this basis. What I judge does not provide insight into the validity of judging it. We said that this raises the question how I can think it valid to judge as I do. It is worthwhile to make explicit that this question is none other than the question how there can be judgment that understands itself to be of something that is as it is anyway. An answer to this question gives the concept of knowledge of something that is as it is independently of being known to be so; it gives the concept thought in “I know”. If we run into a difficulty with our answer to this question, it w ill be a difficulty of retaining our hold on the idea of knowledge of an independent reality. Our answer does run into a difficulty. McDowell develops the concept of perception by way of thinking through that difficulty. The difficulty arises in the attempt to comprehend knowledge from within. If one has abandoned that aim, there is no difficulty. However, if one has abandoned that aim, one no longer inquires into knowledge that is capable of being expressed by “I know”. One no longer speaks of our knowledge.
P erceiving the W o rld
205
The answer runs into a dilemma. The dilemma gives rise to what McDowell calls an oscillation. We oscillate between two ideas: recognizing one to be untenable, we turn to the other; recognizing the latter to be untenable, we return to the first. And so on. Here is the first term of the oscillation. I understand my judgment to rest on a ground, which reveals things to be as, on that ground, I know them to be. Now, in order to know something on a ground, I must know that ground. My knowing the ground w ill be my judging things to be a certain way. This may be put by saying that my judgment rests on another judgment. Of course, a ground on which I know something resides in t hings’ being so, not in my judging them to be so. The point is that, knowing p on the ground that q, I judge q. And I know p on that ground only if I know q. Now my judgment q bears the same character as the one that it is to ground: what it judges is such that judging it may be valid or invalid; what it judges does not provide for the recognition of the validity of judging it. Hence, when we say that knowing something is knowing it on a ground, we in effect say that knowing something is knowing it on the basis of something else, which one knows. And this makes no sense. Thus we move to the second term of the oscillation: we conclude that there must be judgments that do not require a ground, judgments that do not open up the space for the idea of a ground on the basis of which to judge what I so judge. That is, t here must be judgments whose validity I recognize through what I judge in these judgments themselves; what they judge places them outside the space of error. Furthermore, the judgments we seek must be capable of grounding judgments that do not share this character of theirs; they must be able to ground judgments whose validity I do not recognize through what I judge in them and which therefore do require a ground in the light of which I recognize that things are as I judge them to be. As grounds on which to know things that cannot be known through themselves, the privileged judgments anchor a chain of grounds. A judgment of the kind that we seek is to establish that things are a certain way, a way they are whether or not they are known to be so. Such a thing can be known only by affecting her who thereupon knows it. Hence the judgment must reflect a sensory affection of the subject. At the same time, what is judged in this judgment cannot be something judging which
206
C o nsequences f o r M etaphysics
may be valid or invalid. The validity of the judgment must be evident through what it judges alone. It must be, as we may put this, self-evident. When we conjoin t hese two marks of the judgment we seek, we form the idea of a judgment whose object does not go beyond a sensory affection, or sense impression.11 In a term he borrows from Sellars, McDowell calls the idea that there are such judgments the “Myth of the Given”. When Davidson discussed this idea, he found that, first, t here are no ere any, they would not justify any such judgments, and second, if there w judgment requiring a ground: a judgment about a way things are that is such that t hings may be that way or not, or equivalently, a judgment about a way t hings are that is such that things may be known to be that way or not. For, not going beyond a given sensory affection of a given subject, the judgment would be l imited to its subject and its time. Hence it could not constitute a basis for knowledge of something not so l imited. From the incoherence of the idea of a judgment that anchors a chain of grounds, we conclude that only another judgment (more precisely: what it judges) can provide a ground for a given judgment. As we realize that this makes no sense, we return to the idea of judgment which is a myth, the Myth of the Given. This is the oscillation brought on by the dilemma.
6. Dismounting the Seesaw Being caught in this oscillation, it is as though we were trapped on a seesaw. McDowell introduces a conception of perception that allows us to dismount the seesaw and break f ree of the oscillation. The relevant conception says: what is given in perception—what is perceived—is nothing other than what is judged on the basis of this perception. We can dismount the seesaw with this conception; we can steer clear of Scylla and Charybdis. First, what I perceive is that things are thus and so. And perceiving that t hings are so, I know that they are so. For, t hings cannot fail to be so given that I perceive that they are so. Knowing in this way that things are so, namely, by perceiving that they are, there is no need for me to appeal to something else that I know as a basis on which I so know. As I perceive what I thus know, what I know itself is the ground on which I know it. This avoids the first horn of the dilemma. Second, what I perceive is a way for t hings to be that is such that t hings may be that way or not.
P erceiving the W o rld
207
Therefore it is such as to be a basis on which something e lse may be known. This avoids the second horn. What I perceive is such that something else may be known on its basis. Therefore it is, conversely, such as to be known on the basis of something else. This brings out how McDowell dismounts the seesaw as opposed to attempting to force to the ground one side of it. Perception, as McDowell proposes we understand it, does not anchor a chain of grounds. It does not put a stop to a regress. The Myth of the Given attempts to do that, in vain. McDowell rejects a certain idea of what it is to know something that is as it is independently of being known to be so. According to this idea, to know something is to know it on the basis of something else, which one knows. Thinking of perception as McDowell proposes, we say to this: knowing that things are so may be perceiving that they are; in this case, what we perceive, and that is, what we know, itself provides the ground on which we know it. This does not stop a regress; it shows that there is no need to stop a regress. We can embrace the interminability of justification; it no longer poses a threat to our ability to comprehend ourselves to know something that is as it is anyway.12 In judging, I exclude the opposite judgment. Thus I understand what I judge to be such that a judgment of it may be valid or invalid. Therein I comprehend myself to judge how things are anyway, whether or not they are known to be so. This means, further, that what I judge does not provide insight into the validity of judging it. If it did, then to judge invalidly with respect to it would be to judge something judging which is recognizing the invalidity of judging it. Now I may understand my judgment to be valid as I judge on a ground in the light of which I recognize the validity of judging as I, thereby, judge. However, such a ground is something I know. Hence, resting my judgment on a ground cannot be the only way in which I can understand my judgment to be valid. As we take ourselves to judge how things are anyway, we show that we possess the concept of a further way in which we may think it valid to judge as we, thereby, judge. This is the concept of perception, at work in the thought of myself as knowing how things are by virtue of perceiving how they are. This concept is internal to any knowledge that is capable of being expressed by “I know” and comprehends itself to be knowledge of something that is as it is anyway. We perceive things to be a certain way. This thought is part of McDowell’s account of knowledge. And this account is an account from
208
C o nsequences f o r M etaphysics
within. Hence, what we said of such an account holds of what it says of perception: it is not a thesis; there is no such thing as saying “no” to it. It articulates the “I know” and thus says nothing other than what anyone knows in knowing anything at all (in knowledge capable of being expressed by “I know”). It leaves everything as it is. This is not to say that one cannot question McDowell’s account of knowledge and perception. It is to say that putting forth something that one intends to be a contrary thesis, contrary to a thesis one attributes to McDowell, perhaps a thesis to the effect that what is perceived, fundamentally, is a diff erent kind of thing from what may be judged: this is not questioning McDowell’s account. It is failing to engage it.
7. Knowledge of Perception This is how we dismount the seesaw. When I perceive that things are so, things cannot be otherw ise. Things cannot be otherw ise than I perceive them to be. Thus, perceiving that t hings are so, I can exclude that t hings are not so. I know that they are so. We must beware to understand this aright. We may be tempted to understand it as follows. If someone perceives that t hings are so, then things are so. Hence, and for instance, if I perceive that things are so, then they are so. So the fact that I perceive that things are so provides a ground on the basis of which I know that things are so. When we understand it in this way, we conceive knowing something on the basis of perceiving it as a case of knowing something on the basis of something e lse. We are back on the seesaw. And worse than that. We are proposing that I know p on the basis of the fact that I perceive p. (Given that I perceive p, it must be the case that p.) But a ground on which I know something is something I know; I can know p on the basis of the fact that I perceive p only if I know that I perceive p. But it does not seem that I can know that I perceive p u nless I know p. But I was supposed to know p on the basis of the fact that I perceive p. The proposed understanding of the way in which perception enables knowledge of how things are anyway approaches knowledge from sideways on. We said: if someone perceives that things are so, then they are so;
P erceiving the W o rld
209
therefore, and for instance, when I perceive . . . This cannot be the order of understanding if we understand knowledge from within. It cannot constitute a way of understanding knowledge that is capable of being expressed by “I know.” It is no surprise that it does not resolve a difficulty that afflicts our attempt to understand the knowledge that we understand ourselves to possess. If a way of understanding perception is to enable us to dismount the seesaw, it must be the way in which we understand perception from within. It must be the understanding of perception in play in “I perceive”. I am to know p on the basis of the fact that I perceive p. But on what basis do I know this: that I perceive p? It may seem that we can answer this question if we attend to the first-person character of perception. While I know that things are so on the basis of the fact that I perceive that they are so, I know that I perceive that not on the basis of any other fact; rather, I know that I perceive that t hings are so by perceiving that they are so. I need do nothing more than perceive that things are so in order to know that I perceive that. For my knowing myself to perceive that things are so is internal to my perceiving them to be so. This is marked by the expression in language of that knowledge, “I perceive”: this indicates the internality of the understanding so expressed to what it understands, the internality of my understanding myself to perceive what I do so to my perceiving it. This represents perception as something that puts an end to a regress of grounds. Grounds come to an end in something that through itself provides for the recognition of the validity of judging it. Then I need not appeal to anything other than what I judge in the light of which I recognize the validity of judging it. The thought above presents the judgment “I perceive that p” as fitting this description: what I judge in this judgment provides for my knowing it. In this respect it differs from what I judge in judging p: what I judge in that judgment does not provide for my knowing it. I can judge it only on the basis of something else, namely, on the basis of the fact that I perceive p. If what we just said above about perception w ere sound, we would not need to say it. We could say this. If someone knows that things are so, then things are so. Hence, for instance, if I know that things are so, then they are so. So the fact that I know that t hings are so provides a sufficient ground
210
C o nsequences f o r M etaphysics
on the basis of which I know that t hings are so. T here is no regress h ere because I know that I know that t hings are so by knowing that they are so. It is easy to see why this makes no sense: knowing oneself to know that things are so is nothing other than knowing that t hings are so. Therefore what is known in knowing the former cannot be a basis on which one knows the latter.13 And this shows what is wrong in the above notion of the way in which perception grounds knowledge. It distinguishes two things known: p, and I perceive p. It then exploits the special way in which I know that I perceive something in order to stop the regress. However, the way in which I know that I perceive p entails that there are not two things known. I know that I perceive p by perceiving p because knowing that p by perceiving it is the same act of mind as knowing myself to perceive p. As it is the same act of mind, it is the same knowledge, and what is known in knowing the latter cannot be a basis on which I know the former. Perception enables us to dismount the seesaw only if it is a ground in a different manner from the one that gives rise to a regress. It cannot be that “I perceive p” is an instance of “q” in “I know p on the basis of q”. My perceiving p cannot be something e lse that is the case, in the light of which I know p. Once we think this, we are lost. And then it does not help to add I know the fact that I perceive in a special way, spontaneously, say, or directly. Must I, in order to know that p by perceiving it, know that I perceive it? If we say “no”, then there is no intelligible use of “I perceive”. If we say “yes”, then it seems I must, in order to know p by perception, know more than p. The right response is to reject the question. We must reject its presupposition that what I know, by perceiving it, in knowing that things are so, is something other than what I know in knowing that I perceive that things are so. We saw that, if we accept this, then the concept of perception cannot free us from the oscillation. And we already know why it is wrong to accept it, provided we speak of perception capable of being expressed by “I perceive”. Understanding myself to perceive something and perceiving it are but one act of mind. As knowing that I perceive something is the same act of mind as knowing, by perceiving it, what I thus perceive, not more is known in the former than is known in the latter. I know that t hings are so in virtue of perceiving that they are. We may put this by saying that I judge that things are so on the ground that I
P erceiving the W o rld
211
perceive it. But when we put it in this way, we must not think that this introduces a further fact, over and above the fact which I perceive, as a ground on which I know the latter. T here is no such further fact. What I know, knowing that I perceive that things are so, is nothing other than what I know, knowing, by perceiving, that t hings are so. Thinking that I perceive that things are so, I do not go beyond what I perceive, and thus know, by perceiving it. If we say that I judge that things are so on the ground that I perceive that they are so, we must keep in mind that this is to say nothing other than that I judge that things are so on the ground that, manifestly or perceptibly, so they are. This is how McDowell presents it: as I perceive something, what I perceive, the perceived fact, rationally constrains my judgment. “By being taken in in experience, how things are anyway becomes available to exert the required rational control . . . on one’s exercises of spontaneity.”14 “Experience enables the layout of reality itself to exert a rational influence on what a subject thinks.”15 “The world itself must exert a rational constraint on our thinking.”16 A rational constraint is one that justifies what it constrains. The world itself, what I perceive, justifies my judging what I judge on the basis of perception. Of course, it is equally well to say that what justifies, or rationally constrains, is my perception. And McDowell says that, too. Thus he says, “We need to see intuitions as standing in rational relations to what we should think.”17 It does not m atter whether we represent the fact perceived or our perception of it as the basis on which we judge. It does not matter because, knowing, by perceiving it, that t hings are so, is nothing other than knowing oneself to perceive, and thus to know, that they are so. I said the ground on which I judge, judging on the basis of perception, is things’ being thus and so, perceptibly, or manifestly. One may object that this is a trick of words: “things are perceptibly so” means that they are so and that one can come to know that by means of perception; it conjoins these two separate things in one expression. But this objection makes no sense. If it did (we are compressing our reflections above), then knowing that it can be perceived that t hings are so would be knowing more than knowing, by perceiving it, that things are so.18 And then there would be no such t hing as knowledge of things as they are anyway, no such t hing as such knowledge that can be expressed by “I know”. There would be no asking what and how we know, or perceive.
212
C o nsequences f o r M etaphysics
8. Mind and World It is natural to conceive perception as something that it is possible to think and understand from sideways on. It seems a proper object of empirical inquiry, encountered in nature, as certain things, namely, animals, are affected by certain other things, perhaps also animals, perhaps those that the former hunt and eat. In Mind and World, McDowell presents the obstacle to embracing his conception of perception as residing in a constriction of our idea of nature to the realm of processes governed by natural law. He undertakes to remove this obstacle by enlarging our idea of nature, so as to enable it to encompass us: subjects of reason endowed with the power of perceiving how t hings are. Thus he appears to propose that, if only our idea of nature were liberal enough, there would be no difficulty placing knowledge of nature in nature. This seems wrong to me. There is no placing knowledge of nature in nature. Not because it must be placed elsewhere. But because it has no boundary.19 Perception involves sensory affection. In thinking “I perceive that that thing is so”, I relate myself to the object that I thus know to be so: I locate myself relative to it, and I understand myself to be affected by it. It may seem that this means that, knowing myself to perceive something to be so, I know myself to be an element of the objective order, standing in certain relations to other such elements. It may seem to secure a conception of the subject of knowledge as an element of the very objective world which that subject knows, and to secure it from within, from within that subject’s knowledge of this very world. In perceiving the tree, I place within the objective order of things that are in space and time both the tree and myself. I locate both the tree and myself in this order, and do so in one and the same act of mind. I know myself to be one of the things in space and time and to partake in the causal commerce of things in space and time. I know that not by perceiving myself, but by perceiving, for example, that tree. I say “for example”, because perceiving any other t hing would equally provide me with that knowledge of what I am and what I do and suffer. Our topic is perception expressible by “I perceive”, that is, perception that is such that knowing that one perceives some thing to be so is nothing other than knowing, and therefore is not knowing more than, that that thing is so, by perceiving it to be so. This is the concept that we now propose
P erceiving the W o rld
213
signifies a relation of two things. However, if knowing that one perceives things to be so is nothing other than knowing, by perceiving it, that they are so, then perception cannot be a relation. If the concept of perception signifies a relation, and perceiving that something is nothing other than knowing oneself to perceive it, then it holds true that, in perceiving anything at all, I know myself and my perception to be an element of the very world that I know through perception. T here are many things in the world. Among t hese is one that is such that, coming to know something about some thing, any t hing, is coming to know something about it, too; knowing the substance of some t hing, any thing, is knowing its substance, too. What a wonderful t hing that is! It must be inside any t hing, every thing, if inquiring into any t hing yields knowledge of it. More wonderful still: that thing, that soul and essence of everything, am I! And this is not the end of wonders: you are that t hing, too! We are raving. If perception were a relation, then knowing myself to perceive that that t hing h ere is so could not be the same as knowing, by perceiving it, that that t hing h ere is so. Knowing one t hing and how t hings stand with it is knowing less than knowing that and knowing that another thing is related to it in a certain way. Knowing “a is F” is knowing less than “a is F and a stands in R to b.” Therefore this cannot be the logical relation that “a is F” bears to “a is perceived by me to be F”. We are driven to raving when we conjoin the notion that perception is a relation of two things with the insight that knowing oneself to perceive that some t hing is so is nothing other than knowing (and thus is not knowing more than) that that thing is so by perceiving that. The raving is the misfiring of our attempt to think a sideways-on conception of perception from within. How shall we conceive perception, if not as a relation of two things, myself and the object I perceive? Knowing myself to perceive that things are so is not knowing more than knowing, by perceiving it, that t hings are so. Conversely, and equivalently, knowing, by perceiving it, that things are so is knowing nothing less than that I perceive, and thus know, that they are so. Perceiving that this pear tree is blooming, for example, I know myself to perceive, and thus know, that it is blooming. Perceiving that it is blooming, I know that t hings are as I judge them to be, judging the tree to be blooming. What I so judge, and thus what I perceive, is a way for things to be anyway. As I judge what I do through the concept of judgment, and
214
C o nsequences f o r M etaphysics
as I perceive what I do through the concept of perception, this describes what I perceive not from sideways on, but from within: this is how I understand what I judge in judging it, and this is how I understand what I perceive in perceiving it. In perceiving that the pear tree is blooming, I understand what I perceive to be an aspect of the objective world, a world that is as it is independently of being known to be so. Hence, what I perceive, perceiving anything, is the world. Of course, I only ever perceive part of the world. But its being a part of the objective world is internal to it. It is internal to it in the sense that knowing the part is knowing it to be a part. Such is the unity of the world. This explains a manner of speaking of Mind and World: McDowell describes perception as “impingements by the world on our sensibility”, he speaks of our “responses to the impacts of the world on sensibility” and suggests that we “take sensing to be a way of being acted on by the world.”20 To perceive is to be affected by the world. It might seem that “the world” serves as a means to speak in a general way about the things that affect our senses. That sensing is being acted on by the world then is a way of saying that sensing is being acted on by t hings, t hings in the world. But that makes it puzzling why one would introduce that term. One would not be tempted to say that certain particles collide with certain other particles in the world. Why, then, would one say that in perception we are affected by things in the world? Is it to make the point that we are affected not by a thing in the mind, but by a t hing in the world? But a t hing in the world is not a kind of thing. When there is a tree, and I dream of a tree, then t here are not two trees, one in the world, another in my dream. McDowell uses “the world” to characterize perception because the universality that this term expresses is internal to what is perceived. He makes this explicit in various passages: “The subject of the experience understands what the experience takes in . . . as part of a wider reality, a reality that is all embraceable in thought but not all available to this experience.”21 “The object of an experience, the state of affairs experienced as obtaining, is understood as part of a w hole thinkable world.”22 This understanding is internal to the experience that it concerns; it is the understanding expressed in “I experience”.23 The term “the world” is needed to describe perception because it articulates the understanding of her who perceives something, her understanding of herself as perceiving it, which is nothing other than her perceiving it. The universality of “the
P erceiving the W o rld
215
world” belongs to the comprehension of perception from within, to the comprehension of its object that perception itself is. Understanding myself to perceive that the pear tree is blooming, I relate myself to the pear tree. But this does not suffice to describe what I perceive. It leaves out that I understand the pear tree to be an element of the objective order, and so understand it in perceiving it. If perception were a relation of two things, myself and what I perceive, it would have to be a relation not only to the pear tree, but to the world. However, the world is not anything to which something can stand in a relation. Nothing stands in a relation to the world, not even the mind. The world is an illimitable totality. There is no placing it in a relation to anything. I perceive the world. I have the world in view. So do you. Thinking of you that you perceive that that pear tree is blooming, I think the world, the objective order. Thinking of anyone that she perceives that that pear tree is blooming, I think the objective world. So I do provided I think of perception that is capable of being expressed by “I perceive”. This does not mean that I place her, you, me in a relation to the world. There is no placing something in a relation to the world, not even the mind. It means that there is no thinking what she is, as she has perception, no thinking what she does and suffers, as she perceives how t hings are, that is less than thinking the world. In this sense, she, her perception, is the world. She, her perception, is universal. The same for you and me. In the same thought, in the thought of her, you, me as perceiving that that pear tree is blooming, she is particularized, you are particularized, I am particularized. Indeed, I am spatialized and temporalized, and so are you, and so is she. Perception is this unity of universality and particularity, the unity expressed in “I perceive that tree, that is, the world”. Being universal in being particular, and being particular in being universal, the subject of perception, the human being, you, I, Steffi, is an individual. The concept of perception, its concept from within, is a concept of that unity of universality and particularity; it is a concept of the h uman individual. To understand perception is to understand that individuality. McDowell suggests that what stands in the way of understanding perception as the manner in which the world exerts a rational constraint on judgment is a scientistic conception of nature. In consequence, he thinks we can place perception, and thus knowledge, insofar as it is from perception, in nature, if we abandon that conception. But perception,
216
C o nsequences f o r M etaphysics
insofar as it sustains knowledge that is an understanding of itself and expresses itself in “I know”, is not placeable in nature, b ecause it is not placeable, period. It is not any thing, but universality; it is not a relation of one thing to another thing, but an openness to the illimitable whole. Thinkers have attempted to register the uniqueness of knowledge—its universality—by placing it supra nature. It is not my purpose here to consider any of these thinkers. But they retain the high ground against any response to them that proposes to place knowledge in nature, no matter how broad nature is made out to be in that response. Therefore the obstacle to the explicit affirmation of McDowell’s understanding of perception does not lie in a scientistic idea of nature. It lies in the difficulty of seeing the concrete subject to be the world. The subject is the world not by being abstracted into a transcendental I, a universal self-consciousness, an objective self, a limit of the world, etc. On the contrary, it is the world by perceiving this, not that, being h ere, not t here, being me, not you. My purpose h ere was not to resolve, but to bring up this difficulty. It is the difficulty of understanding knowledge, understanding it from within. I hope to be following McDowell by refusing to put it away by stepping outside and considering the matter from sideways on.
Chapter Eight ✣
Seeing the World Moral Difficulty and Drama ✣
Evgenia Mylonaki
1 The topic of my paper is the right way to take John McDowell’s opposition to what he calls “the blueprint picture” of the practical reasoning of a virtuous deliberator. This is roughly the view that the virtuous deliberator must possess a kind of blueprint for virtuous action—a general, substantial conception of living well—whose content is independent of their choices in particular circumstances and which they can apply to these circumstances. My aim is to demonstrate that a proper appreciation of McDowell’s opposition to this picture paves the way for a deeper engagement with his ethics, which in turn promises a more radical reconfiguration of the landscape of moral philosophy than even McDowell seems to have anticipated. In the course of developing this line of thinking, I hope that even my partial departure from his thought can serve as a testament to the greatness of his philosophical spirit and a token of my gratitude for his guidance over the years. 217
218
C o nsequences f o r M etaphysics
In his later work, McDowell frames the view that he opposes as follows: The idea of the blueprint picture is that the content of a conception of the universal, doing well, is in principle available, and accessible for correctness, in abstraction from judgments or actions, in particular circumstances that we want to see as applications of it. . . . W hether some particu lar judgment or action was a correct application of the universal would be a question of what followed from the universal’s content together with the facts of the situation. So the question whether some conception of doing well was correctly applied in some particular case would be separable from the question w hether doing well so conceived was the right end to pursue. Correctness of application would be recognizable, in principle, from a stance that was neutral with re spect to the corresponding end. And for a deliberator to be getting things right, arriving at what really is an instance of doing well, both conception and application would need to be correct.1 He frames the alternative picture he is working t oward as follows: here is nothing wrong with saying that a possessor of excellence T [in deliberation] grasps the content of the universal, doing well. But we need not conceive that grasp as separable, even in principle, from the state, one aspect of which is a motivational propensity that results from having been properly brought up. . . . [W]e have the essentials of a contrasting picture, in which the content of the end cannot be pinned down in abstraction from the ability to put it into practice in recognizing specific occasions for action. In this contrasting picture there is nothing for the grasp of the universal to be except a capacity to read the details of situations in the light of a way of valuing actions into which proper upbringing has habituated one.2 McDowell’s rejection of the blueprint model has often been regarded as the motivation for embracing some variety of what is generally known as “moral particularism” (the view that a morally excellent individual is not so in virtue of their grasp of universals).3 However, McDowell insists that he does not object to the idea of there being general principles at work in
S eeing the W o rld
219
the virtuous agent’s taking certain features of their situation as reasons for action or judgment. He writes, “There is nothing wrong with saying that a possessor of excellence grasps the content of the universal, doing well.” 4 It has also been assumed that McDowell’s treatment of morality in terms of the excellence of h uman powers of deliberation and choice is equivalent to his espousal of a variety of moral generalism: namely, the “virtue ethics” view that a person of principle has access to general truths about human flourishing, which may be brought to light through the philosophical elucidation of h uman nature.5 However, his rejection of the blueprint model entails that if truths about human flourishing are to serve as principles of practical reasoning, then their content can be revealed only in the reasoning of those who embody these principles. In opposing the blueprint model, McDowell objects neither (as the moral particularist supposes) to the existence of a universal that one must grasp in practical reasoning nor (as the moral generalist supposes) to the idea that one’s grasp of the particular is of primary ethical significance. I believe that McDowell’s opposition to the blueprint view of practical reasoning has not yet received full appreciation. This no doubt relates to the haphazard classification of his work under the headings of moral particularism and virtue ethics. However, I believe that there is a further reason for this neglect. McDowell tends to frame the problematic idea at work in the blueprint model in terms of the assumed separation between the operation of the intellect and the actualization in particu lar circumstances of motivational propensities. Moreover, he draws, from the Aristotelian concepts of practical wisdom (phronêsis) as a kind of perception (aisthesis) and excellence of character (arête ethike) as the outcome of habituation (hexis), the elements of a contrasting story, on which the work of the intellect is inseparable from the actualization in practical excellence of motivational propensities molded in upbringing. T hese aspects of his ethics have been hastily incorporated into epistemological discussions concerning the intelligibility of moral discernment as a perceptual faculty and into moral-psychological discussions concerning the motivational aspect of the virtuous person’s grasp of reasons.6 While these aspects of his ethics are certainly crucial, I have come to believe that jumping to them without having reached a fuller appreciation of his opposition to the blueprint model risks obscuring the radical character of McDowell’s thought. In the following sections, I argue that, in rejecting
220
C o nsequences f o r M etaphysics
the blueprint model, McDowell opposes a certain view of the relation between the general and the particular in the practical reasoning of the virtuous. I also argue that, if we wish to be resolute in this opposition, then we must revise our conception of virtue in ways that may extend beyond McDowell’s account.
2 McDowell, like Aristotle, sees no problem in reframing the question concerning virtuous practical reasoning in terms of the practical syllogism.7 We may adopt Aristotelian language to provide a preliminary characterization of the practical syllogism as an argumentative structure consisting of the following components: the major premise, which is of the general (that is, the end); the minor premise, which is of the particular (that is, the circumstances); and the conclusion, which regards the application of the major premise to the particular stated in the minor premise (that is, the action or judgment of the virtuous person).8 This reframing is unproblematic because the question of moral practical reasoning does not concern the passage from thought-events to the virtuous action or judgment; rather, it involves the thought that the virtuous action or judgment embodies, what I henceforth refer to as the judgment of morality. If the question of moral practical reasoning did in fact concern the character of the passage from thought-events to the judgment of morality, then one would be at a loss as to the application of the apparatus of practical syllogism to cases in which there is obviously no such passage; as when, for instance, upon observing that you are hurt by my words, I simply fall into silence. However, as McDowell states, “the conceptual apparatus of universal end and application to the circumstances at hand still fits, even in the absence of any course of thinking that constitutes arriving at the application.”9 The schema of the practical syllogism still fits: this means that the virtuous action or judgment is not an undifferentiated particular, say a mere physical or mental occurrence. Rather, it is a calculation (logismos), one—according to McDowell (following Aristotle)— concerning how to live well. Thus reframing the issue of virtuous practical reasoning, though, may give the impression of an attempt to re-intellectualize ethics. In other
S eeing the W o rld
221
words, it may seem that talk of the calculation or the judgment of morality simply serves to redirect our attention to the aspect of moral thought as a mere exercise of the intellect, to the detriment of an understanding of moral experience as the living reality of historical individuals. I believe it is in this vein that philosophers who aim to do justice to the reality of moral experience often suggest that we should distance ourselves from explanations of such experience in terms of judgment. When discussing Henry James as a moralist, for instance, Cora Diamond approvingly asserts that judgment does not hold for James the central position that it does for philosophers. What she refers to as “judgment” h ere is a kind of evaluative thought in which “one has in mind something or other—act, person, character trait—a nd considers the application to it of some evaluative term.”10 Alice Crary draws a similar equation between phi losophers who hold moral judgment in high regard and t hose who view moral experience as applying concepts of some generality to particu lar persons, actions, features of a situation, and so on.11 However, this impression is itself rooted in the misunderstanding that discussing judgment in moral philosophy commits us to the undesired consequence that all that we might intelligibly say about moral thought or experience is that it is the application of a set of preexisting general concepts to the particulars of the world. McDowell’s work on practical reasoning addresses this misunderstanding by inviting us to reopen the question of the syllogistic character of moral thought, such that it is no longer within our right simply to assume that a discussion of calculation or judgment in moral philosophy constrains the realm of the moral beyond recognition. I said above that it is the character of the thought embodied in the action or judgment of the virtuous person that the practical syllogism articulates. If this is the case, then we must conceive of each element of the practical syllogism (major premise; minor premise; conclusion) as no more than syllogistically articulated aspects of this one thought— that is, of the judgment of morality. On this interpretation, the judgment of morality may be regarded either as the apprehension of an end, or as the apprehension of the circumstances, or as the apprehension of the end as applying in the circumstances. This is not to say much, however; for now, the question is precisely to explicate the articulation of these aspects of the judgment of morality.
222
C o nsequences f o r M etaphysics
3 In opposing the blueprint model of practical reasoning, McDowell rejects a particular view of the syllogistic character of the judgment of morality. As I mentioned in Section 1 above, McDowell frames the driving force of the account he rejects as the idea that “the content of a conception of the universal, doing well, is in principle available, and accessible for correctness, in abstraction from judgments or actions in particu lar circumstances that we want to see as applications of it.”12 In terms of the practical syllogism, the idea he opposes is that there is a logical gap between the major premise and the conclusion of the syllogism; specifically between the aspect in which the judgment can be represented as of the general (the end) and the aspect in which it can be represented as the application of the general to the particular (the circumstances). This would mean that the thought of the virtuous must be subject to two distinct standards of success: (1) getting t hings right with respect to the question what ends to pursue; and (2) getting things right with respect to the question how things stand in the world here and now. On the blueprint model, knowing the universal, d oing well, is such that one does not thereby know how things stand in the world here and now. For instance, one may count as knowing that lying is bad without being able to know that one is, in withholding information here and now, lying, and thus without being able to know how things actually stand in the world here and now. In a word, on the blueprint model, t here is no difficulty in granting one full grasp of the universal, while one falls short of knowing how things stand in the world h ere and now. Therefore, if moral knowledge is not of how things stand in the world here and now, then it must be of how things stand in the world in a qualified sense—say, the world of ta anthropina, that is, human affairs, or the world of morality. On this view, then, that about which one has knowledge in correctly grasping the universal end can be no more than a realm or dimension of reality (a part of the world in this sense), about which it is an open question how it relates to the realm or dimension of reality of the particulars. This open question is settled by completing one’s practical reasoning, that is, by calculating how one’s independently available grasp of the universal applies in the circumstances.
S eeing the W o rld
223
In this light, we can take McDowell’s objection to the blueprint model to be a rejection of the assumption that t here is a logical gap between the major premise and the conclusion of the practical syllogism in the judgment of morality. If there is no such gap, however, then this must be because the question what ends to pursue is a form of the question how things stand in the world h ere and now. In other words, if there is no logical gap between the major premise (concerning the end) and the conclusion (concerning what to do in the circumstances) in the syllogistic articulation of the judgment of morality, then in the grasp of the end, the virtuous person’s knowledge must not fall short of being knowledge of how t hings stand in the world here and now. Thus, in grasping the universal, on the McDowellian view, the person of excellence comes to know how t hings stand with respect to the world as such instead of with respect to a part of the world (a realm or dimension of reality) as the blueprint picture supposes. Another way of expressing the same thought is as follows: McDowell is denying that there is a logical gap between the aspect in which the judgment of morality is the virtuous person’s grasp of a universal (that is, stateable in the major premise) and the aspect in which this grasp of the universal affords us knowledge of the particulars of the situation (that is, stateable in the conclusion). In doing so, he is in effect denying that the general and the particu lar relate to each other as distinct realms or dimensions of reality and that the task of virtuous reasoning is to bring these distinct realms or dimensions together by applying one’s conceptions of the former to that of the latter. On my interpretation, the philosophical core of McDowell’s thought is that in the virtuous apprehension of one’s situation (what I refer to h ere as the judgment of morality), the world is one. Therefore, the question of moral knowledge ( phronêsis)—what ends to pursue or how to live well—is actually a species of the question of knowledge of the world as such (sophia)—how to bring the world into view.
4 If we understand the judgment of morality according to the above discussion (that is, if we deny the logical gap between this judgment as the grasping of an end and as a way of bringing the world into view), then the
224
C o nsequences f o r M etaphysics
mention of a distinctly moral falsehood may begin to sound queer.13 One attraction of the blueprint model was its seeming ability to make sense of moral falsehood. If we posit a logical gap between the aspect in which the judgment of morality is the grasp of an end and the aspect in which it as a way of bringing the world into view, then it would be possible to have done all that allows one to grasp a general moral thought (for example, that lying is morally bad) without actually seeing whether and how the thought yields knowledge of the circumstances. That is, it would be possible to grasp such a thought without having completed one’s practical reasoning or reached an answer to the question how things stand in the world here and now—say, whether withholding information right here right now is indeed lying or not. This creates the impression that we may have at our disposal a judgment (for example, that lying is morally bad) whose content we can fully determine to be moral (for example, to be morally bad) prior to figuring out w hether and how it applies in the circumstances. We may call this a “pure moral judgment,” to note that the term “moral” qualifies the subject m atter of the judgment—that which is independent of the truth or falsehood of the judgment in the circumstances. Thus, we may believe that we can explain moral falsehood as a judgment which is fully moral in its basis, but which happens to be false in the circumstances. Nonetheless, the ease with which the blueprint model seems to explain the intelligibility of moral falsehood is only apparent. As I suggested in Section 3 above, on the blueprint model, moral knowledge is essentially preoccupied with the nature of a dimension or realm of reality (the moral part of the world) that is separate from the dimension or realm occupied by particulars. It amounts to knowledge of particulars only in a secondary, derived sense: as applied to a conception of the particulars, which itself is derived from an exercise of a cognitive power that owes nothing to the agent’s ability to judge virtuously. However, if the knowledge embodied in one’s moral judgment when things go well is a hybrid of two knowledges—one moral and one nonmoral—then it is unclear why we should count the success or the failure (that is, the truth or falsehood) of the so- called moral judgment as an essentially moral success or failure. The reason is that, on the blueprint model, the difficulty of morality pertains to the success or failure of the judgment qua grasp of the moral part of the world (of the moral universal) rather than qua grasp of the particulars. Supposedly, on the blueprint picture, there is nothing dis-
S eeing the W o rld
225
tinctively moral about this latter grasp. The only difficulty with such a grasp is the conceptual one of applying a general concept to particulars, and this is not a difficulty we should have to recognize as moral. However, if the difficulty of morality concerns the success or failure of moral judgment not qua grasp of the particulars but only qua grasp of the general (the grasp of a universal as stated in the major premise of the syllogism), then it is not clear why we should count the truth or falsehood of any judgment that aspires to be a judgment of morality as an essentially moral truth or falsehood. In other words, if the difficulty of determining the moral truth in the circumstances is merely conceptual (that is, the difficulty of applying an independently available universal conception to a conception of the particulars), then we lose sight of the sense in which the truth or falsehood of such a judgment may be moral in essence (qua apprehension of the truth) and not merely in subject m atter (qua conception). Therefore, contrary to initial impressions, the blueprint model of virtuous practical reasoning fails even to bring into view—let alone to explain—the possibility of moral falsehood—that is, the possibility of an essentially moral failure of judgment. If one posits a logical gap between the success or failure of a judgment qua general conception (qua grasp of a universal end) and its success or failure qua application (qua grasp of the truth concerning the particulars or the circumstances), then no m atter how one spins the story afterward, the latter w ill not be moral in essence. One may thus still speak of the failure of a judgment to hit the moral mark in particular circumstances without thereby being entitled to speak of it as an essentially moral falsehood in the same way that one may speak of the failure of a judgment to hit the mark concerning the right way to clean clothes without thereby being entitled to speak of it as an essentially cleaning falsehood. With this line of thinking, in letting the difficulty of morality run f ree from that of bringing the world into view, we risk turning the very possibility of an essentially moral truth and falsehood into a chimera. Contrary to what one might expect, we cannot secure the intelligibility of moral falsehood by cutting off the realm of moral knowledge from the realm of knowledge of the particulars, as the blueprint view emphatically does; rather, we have to bring them closer together. For in separating moral knowledge from knowledge of the particulars, the blueprint model posits an insurmountable metaphysical schism between the realm of moral
226
C o nsequences f o r M etaphysics
generality and that of moral particulars. By insisting upon an alternative to the blueprint view, McDowell is in fact suggesting that in the judgment of morality, the issue of moral generality just is the issue of the particulars of the situation. Now the following difficulty arises: if we suppose that there is no logical gap between the aspect in which the judgment of morality is the apprehension of the general and that in which it is the apprehension of the particulars, then although we may explain the possibility of sound judgment which is moral as such, we shall still be hard-pressed to explain how a judgment that falls short of bringing the world into view may nevertheless count as moral. The blueprint model seemed to offer an easy answer to this question. As noted above, however, this answer cannot work. McDowell’s treatment of the difference between the phronimos (the person of practical wisdom) and the akratês (the incontinent person) on the one hand and the enkratês (the continent person) on the other is, I believe, best viewed as an attempt to respond to this difficulty. The general outline of his account of the difference between the phronimos, the akratês, and the enkratês is sufficiently well-k nown.14 McDowell believes that in the apprehension of the phronimos, the grasp of the universal is inseparable from what the blueprint view might consider its application to the particulars of the situation. This makes the judgment that the possessor of excellence is capable of at once moral knowledge and knowledge of the particulars. The akratês counts as failing precisely because the judgment of the universal that they fail to act on is now regarded as only partly available to them, and the same is true of the enkratês. The judgment of the universal that the akratês fails to and that the enkratês manages to act on is only partly available to both of them, precisely b ecause in being available to them, it fails to fully make available to them the particulars of the situation. It is precisely because the judgment of the universal in its full availability would make fully available the particulars of the situation that it counts as moral, and it is precisely because the judgment is not thus fully available to them, that both the akratês and the enkratês are considered to have failed in their judgment. This is roughly the shape of McDowell’s answer to the difficulty of explaining an essentially moral falsehood.
S eeing the W o rld
227
5 This story, however, still needs to be filled in. What might be distinctly moral about having the particulars of the situation made fully available to one? What might be distinctly moral about having the world come into view? We cannot appreciate this as long as we continue to assume that the matter of access to the particulars of a situation is a merely conceptual one of subsuming one’s independently available apprehension of the particulars under the correct conception of a universal. A fter all, McDowell insists, this is something that the enkratês can also do. However, the enkratês is set apart from the virtuous person precisely in this: the enkratês’s labor is a matter of conceiving of the particulars of their situation and bringing these particulars under the heading of an independently available conception of the universal. For instance, if the grasp of the universal w ere that of the value of honesty, the enkratês’s l abor would be a m atter of trying to figure out which of the independently conceived features of their circumstances would rightly be subsumed u nder the concept of lying. However, this leaves it entirely open that the enkratês might be able to see other features of their circumstances as falling under, instantiating, or embodying other concepts in the situation. The enkratês of the example might, for instance, conceive of withholding information here and now as falling under the concept of saving their life. We may say now that where the labor of the enkratês’s practical reasoning is at most the conceptual labor of subsuming independently available features of the particulars of a situation under “the correct” concept of the universal, the labor of the phronimos’s practical reasoning is not merely conceptual to begin with. We get the impression that it must be so b ecause we inadvertently take particulars to be internally undifferentiated, predictively determinable objects of the sort that would fall u nder concepts such as red or apple. It is not necessary, however, thus to conceive of what is registered in the minor premise of the practical syllogism. In what follows I w ill suggest an alternative picture, on which the particulars within the minor premise are determinate answers to questions that arise at the level of the contingent, historical, and individual reality of our lives with o thers, as Iris
228
C o nsequences f o r M etaphysics
Murdoch most emphatically suggests.15 Such questions include who I am to you, who you are to me, how our relationship affects my sense of myself and of things, what I do for a living and how this shapes the social world in which I move, what my fears are and how they shape my political views, what my traumas are and how they affect my most intimate relationships, and so on and so forth. As I w ill suggest in what follows, these are all particulars which can only be characterized in morally loaded terms and so whose recognition would itself constitute moral judgment. In this light, we may postulate that when McDowell claims that the phronimos’s practical reasoning silences other considerations (those speaking for the alternative concept in the enkratês’s grasp of the universal), he cannot simply mean that the virtuous person applies one concept where the enkratês applies two. There is, a fter all, no reason to believe that a virtuous person could not fully apply a number of concepts in the circumstances. Nor is it enough to state that the phronimos has been raised in such a way that her motivational and intellectual propensities simply do not pull her in opposing directions and that they speak instead with one voice, as this creates the impression that the labor of practical reasoning is already over before it begins. What McDowell must mean instead is that once the phronimos’s practical reasoning is completed, then the world is clearly in view and one hears what is being said without interfering noise, as it were. This leaves it open, though, that the practical reasoning of the virtuous involves labor; or e lse, that t here exists such a thing as the difficulty of morality for the virtuous. So, we need to ask ourselves: what precisely is this labor of practical reasoning—the moral-deliberative difficulty—that the phronimos f aces, if it is not the motivational difficulty that the enkratês f aces? That is, what is the difficulty of morality for the phronimos, if it is not that of seeing how to stick to one of the multiple moral concepts that may be applicable in the circumstances? My suggestion here is that the moral- deliberative difficulty for the phronimos is the difficulty of addressing the open issues of the contingent, historical, and individual reality of their life, which prevent them from seeing how t hings stand in the world here and now by not letting them appreciate what to fear, what to embrace, what to attend to, whom to treat justly, how to love, how to acknowledge the separateness of the other, and so on and so forth. On this suggestion, the core point of McDowell’s use of perception to describe
S eeing the W o rld
229
the deliberative power of the virtuous person lies in this: once the practical reasoning of the virtuous has been completed, that is, once life’s open questions have on each occasion been settled, all that one needs to do may be likened to opening one’s eyes and seeing. In this sense, and this sense only, answering the question of morality may be conceived of as a matter as simple as seeing. To appreciate the difficulty of morality for the phronimos and so to get fully into view the practical reasoning of the virtuous person, we must understand how this seeing is possible. My suggestion is that we cannot appreciate this u nless we acknowledge that the question of morality (that is, what ends to pursue) may not be settled in abstraction from the practical reasoning whose job is to settle the open questions that arise on the level of the particulars for us, as individual living beings. It is in order to address t hese questions so as to bring the world into view, and not in order to properly classify the internally undifferentiated, predictively determinable particulars in the world, that one must be courageous, generous, loving, kind, and so on and so forth. This simple thought often goes entirely overlooked in the moral philosophy of our times: bringing the world into view is often a m atter of working out one or some of life’s open questions, such as t hose briefly noted above. However, it is impossible fully to spell out what it would be to work out life’s open questions in a way that brings the world into view ahead of time. The only statements we can make are true for the most part and serve as a shorthand: for example, we may say that we cannot fully bring the world into view without courage, love, kindness, generosity, or justice. However, these virtues do not play the role of principles in the practical reasoning of the virtuous person qua conceptions waiting to be applied, specified, or instantiated. Rather, on the extension of McDowell’s standpoint developed here, these virtues are principles in the sense that they have emerged in our collective experience of life as the ways to bring the world into view in addressing life’s open questions. Therefore, for example, the activity of practical reasoning is not governed by the principle of courage in the sense that it calculates how to apply a concept of courageous action and judgment to the particulars of one’s circumstances. Rather, the activity is governed by the principle in that it calculates how to face one’s fears so as to bring the world into view. Furthermore, in this case, one’s practical reasoning is excellent (that is, courageous) to the
230
C o nsequences f o r M etaphysics
extent that one manages to find the best way to face one’s fears and bring the world into view. On this line of thought, the difficulty of morality is indeed that of sound practical reasoning and judgment. But this is precisely the difficulty of facing our fears, understanding our anxieties, tending to our wounds, remembering that the reality of the other with whom we are involved is separate from our own, overcoming our comforting fantasies concerning ourselves and others without being cruel to e ither, and so on; all these activities are unintelligible except as ways of bringing the world into view. In sum, to claim that the difficulty of morality is the difficulty of a kind of judgment is to say no less than that it is the difficulty of the ongoing pro cess of resolving life’s open questions. Therefore, it is certainly not a merely intellectual difficulty despite being a difficulty of judgment. It appears to be a merely intellectual difficulty b ecause the issue of practical reasoning (that is, the relation between moral generality and moral particularity) seemed to be the issue of seeing how to apply an independently available conception of moral universals to a conception of particulars. But now we need to ask ourselves: Who can manage such a feat of practical reasoning? That is: Who can come to s ettle life’s open questions virtuously? Who can come to see? In line with Aristotle’s aristocratic spirit, McDowell gives the impression that we may answer t hese queries by simply turning our attention to those who have been raised to have their motivational dispositions molded in such a way as to be able to see what needs to be done on each occasion. Such persons have been inculcated with the correct principles in their upbringing such that their intellectual and emotional powers speak with one voice, as it were; they have been habituated to a certain outlook and conception of doing well, and so on.16 McDowell’s belief is that in order to distinguish the virtuous person’s excellent exercise of practical reasoning (practical wisdom or phronêsis) from that of a mere intellectual power to set the blueprint for living well in stone and then apply it, we need to suppose that habituation to a certain outlook—a certain picture of the world—determines what counts as excellence in practical reasoning (or practical wisdom), at least partially. However, I believe that we need not and should not settle for this answer to the question: Who can settle life’s open questions virtuously? Who can come to see? In the following discussion, I draw from Henry James and his masterpiece, The Portrait of a Lady, in order to sketch the beginning of an
S eeing the W o rld
231
alternative modernist account: that someone who can come to s ettle life’s open questions virtuously is not someone who is habituated to the correct view of the world (independently of whether this view is an intellectual- cum-motivational achievement); rather, such a person is one who can let their world-view come into contact with life’s open questions in such a way that it might be placed in radical jeopardy.17 The moral of this story is that if we take in stride the most promising features of John McDowell’s account of the virtuous practical reasoning, we arrive at a view of moral experience as an openness not just to changing views of particu lar people and circumstances but also to a shifting perception of the world as a whole. For if having a clear view of the particulars is to the virtuous person nothing less than bringing the world into view, and the view of the virtuous person is to be contrasted with the view of the merely continent person in a similar fashion to that suggested above, then, as I s hall try to demonstrate, having a clear view of the particulars—that is, settling life’s open questions here and now—may on occasion come to no less than the total collapse of one’s w hole outlook on the world. We can and should radically oppose the blueprint model of practical reasoning, that is, the view on which the practical reasoning of the virtuous person involves—however indirectly—the possession and application of a specialized concept, whether this happens to be the concept of a moral universal, a blueprint for living well, or a moral view of the world itself. We can and should radically oppose this model which seeks to confine knowledge of morality to knowledge of a special domain. We can and should, instead, embrace the McDowellian standpoint, from which moral knowledge can be seen to be a form of knowledge of the world as a whole. But, I w ill suggest in what follows, to do so we must revise the McDowellian answer to the question: Who can settle life’s open questions virtuously? Who can come to see?
6 Isabel Archer, the heroine of The Portrait of a Lady, is an ideal fit for my purposes here for two reasons. First, this is a drama about how to see the world, as James explicitly states in his remarkable preface to the novel. Second, this heroine is—to use a phrase that James borrows from George
232
C o nsequences f o r M etaphysics
Eliot—a “frail vessel.”18 This is hence a drama in which the heroine is most emphatically not the kind of person whose motivational dispositions have been molded in upbringing to allow her (and the readers) to see what needs to be done from one occasion to the next; indeed, quite the opposite is true, as we shall see in what follows. Nonetheless, this is a drama in which the adventure lies in Isabel’s active sense of life—a sense of freedom manifesting as independence and seeing the world for herself—which finally, albeit tragically, comes to fruition in seeing. Thus, Isabel does a number of things in the conventional sense: she crosses the ocean following an invitation from Mrs. Touchett, her aunt who lives in England. There, she piques the interest of her cousin Ralph, who, unbeknownst to her, secures for her a g reat fortune; she also makes friends with Madame Merle, an old friend of her aunt’s. Despite her family’s protests, Isabel marries Gilbert Osmond, a friend of Madame Merle’s—a penniless widower with a young daughter. Much later, a fter her marriage has been revealed to be a disappointment, Isabel leaves Florence for England to see her d ying cousin Ralph. She contemplates leaving her husband, but eventually returns to Rome. However, these are not the real dramatic elements of the story. For James, like McDowell, there is another kind of action—that of coming to see—and it is this action that he places at the center of his drama. But what he shows in doing so is that the aristocratic answer to the question “Who can s ettle life’s questions virtuously?” is not needed because a modern answer to the question is available.19 But let me first say a bit more about the real drama. The real drama begins when Isabel, who has been in E ngland for some time, returns to the estate of her dying uncle, Mr. Touchett, after a trip to London. The solemn atmosphere of death is interrupted by m usic coming from the drawing room: a stranger is playing Schubert on the piano. “I am Madame Merle,” the stranger says by way of self-introduction in the first major recognition scene of the story. Here is a forty-year-old woman of no particular beauty but of considerable charm and experience, the “cleverest” and “most complete” w oman, as Ralph explains later on.20 The spell has been cast: Isabel’s sense of Madame Merle as an individual who is charming, cultured, and experienced as ideal is the first movement of the story’s true action of seeing the world. Some time a fter her marriage to Madame Merle’s friend, Isabel walks in on her husband and Madame Merle in a very strange scene that gives
S eeing the W o rld
233
her a fresh impression for no special reason: Madame Merle is standing while Mr. Osmond is seated, the two of them looking at each other with an intimacy which seems possible only through a hitherto unimagined bond that unites them. Soon afterward, Isabel experiences terrible visions that keep her awake all night. These include the vision of her relationship with her husband as one of deep mistrust over the very fabric of each other’s being rather than one of mere differences of opinion; that of his rare indifference to society (which she admired so much during courtship) as his covert desire that it should recognize his superiority; that of her life as an imprisonment in the damp, suffocating dueling that her husband’s many traditions have sucked her into, which she previously believed to be the culture, experience, charm, and fine taste of the old world. Later on, when Madame Merle interferes with her domestic arrangements, Isabel suddenly asks Madame Merle, “Who are you—what are you? . . . W hat have you to do with my husband? . . . W hat have you to do with me?” To this, Madame Merle replies, “Everything!”21 “Misery!” exclaims Isabel, and this exclamation reveals her recognition that Madame Merle has indeed arranged what she thought was her f ree decision to marry Mr. Osmond. Later that afternoon, Isabel wonders if “to this intimate friend of several years [Madame Merle] the g reat historical epithet of wicked were to be applied,” but soon lets out a “soft exclamation,” “Poor Madame Merle!,” in what seems to be yet another recognition that if the marriage she has arranged is also making Mr. Osmond unhappy, then Madame Merle, too, must have fallen from his favor in a way that must be making her very miserable indeed.22 Later, Isabel’s sister-in-law reveals to her that her stepdaughter, Pansy, is the product of an illicit love affair between her husband and Madame Merle. To her sister-in-law’s disappointment, Isabel reacts to this revelation with the exclamation, “Poor w oman—a nd Pansy who d oesn’t like her!”23 This announcement is Isabel’s recognition of Madam Merle’s ploy as the scheme of a woman who is forced to spend her entire life at such a distance from her child. Soon a fter, in a chance meeting between Isabel and Madame Merle, the former admits that she knows everything. In a “sudden rupture to her voice,” she sees Madame Merle’s recognition that now she knows everything, that “everything was at an end between them.”24 In her own recognition of Madame Merle’s recognition, Isabel can now see more intensely that, while she believed herself to be mostly f ree
234
C o nsequences f o r M etaphysics
and her own person, what she really was in fact was “a dull un-reverenced tool.”25 The image fills her mouth with bitterness, and a fter it dissipates, she finally sees Madame Merle for what she is: “the cleverest w oman in the world, standing there within a few feet of her and knowing as l ittle what to think as the meanest.”26 Nonetheless, Isabel is not done seeing yet. A fter her arrival in England for her cousin Ralph’s funeral, she meets an old American suitor who offers to “free” her from her unhappy marriage. In his words: Why shouldn’t we be happy—when it’s before us, when it’s so easy? . . . It would be an insult to you to assume that you care for the look of a thing, for what people w ill say, for the bottomless idiocy of the world. . . . We can do as we please; To whom under the sun do we owe anything? What is it that holds us, what is it that has the smallest right to interfere in such a question as this? Such a question is between ourselves—and to say that is to s ettle it! . . . The world’s all before us—and the world’s very big.27 Ironically, this is precisely Isabel’s own previous concept of freedom as in dependence and being in control of one’s destiny, outside of all societal norms—a concept that she now feels as a flash of a lightning toward the end of the adventure.28 Afterward, Isabel is plunged back into the darkness, where she is now free and knows where to go.29 While we are left in fearful wonderment for a page or two, we are informed that she has headed back to Rome. The narrator’s words here are clear enough: “She had not known where to turn; but she knew now. There was a very straight path.”30 The flash of lightning that Isabel feels in her old suitor’s violent possession is the final and thereby forceful return of a light—her previous idea of freedom as independence—which held her entire world-v iew intact. The reader, along with Isabel and her cousin Ralph, believes throughout the drama that this independence w ill ultimately acquaint her with life to a large extent. However, this concept of freedom as independence gradually collapses as the drama unfolds, and after the final recognition scene which signals the last stage of this collapse, Isabel can see for the first time. Notably, James does not reveal what she now sees, likely b ecause the moral labor involved in this seeing is now over. One might argue that what she now sees is simply what lies before her: the contingent, historical, individual reality of her life with its commitments, obligations, burdens,
S eeing the W o rld
235
complications, and open questions. There is nothing dramatic about this though. What made for drama was the moral labor of opening her eyes to see; of figuring out how to see; or of reasoning practically, in the philosophical terms of this paper.
7 On my reading, the dramatic setting of James’s novel is populated with a series of recognitions that are later seen to be misrecognitions or only partial ones that occasion further realizations. These further recognitions are themselves later seen to be misrecognitions or only partial recognitions, and so on and so forth, until the final dramatic scene, in which the narrator declares that Isabel now knows where to turn. At each juncture, nonetheless, the partial recognition or partial seeing is not explicable in terms of an outlook that finds expression in the circumstances. On the contrary, each recognition counts as such precisely because Isabel’s world- view—a habituation to a certain (American, perhaps) way of taking things—is jarred u ntil it finally collapses. Thus, it is Isabel’s arrest before the scene of the wordless communion between her husband and Madame Merle that is her first act of bringing the world into view, as we gradually come to see. As we come to see along with her, this is the world in which the cleverest and the most complete woman (Madame Merle) is not a person of freedom and independence; the world in which the most cultured and intelligent individual (Isabel’s husband) is not a man of knowledge and taste; the world in which the pursuit of independence is not what sets one free; and, finally, the world in which the w oman who returns to the deepest bondage (in the unhappiest of marriages)—that is, Isabel herself—may in fact own up to the contingent, historical, and individual reality of her life, and perhaps be thus set free. In my recounting of the story, the recognitions are stages in the gradual but certain collapse of Isabel’s pervasive concept of freedom as independence. Her former suitor’s words t oward the end of the novel, “We can do absolutely as we please,” appear to be a harsh and ironic indictment of this concept. On the Jamesian story, I might add, it is in Isabel’s openness to the collapse of her world-v iew, and not in the l abor of putting into practice her proper habituation of the ability to take certain t hings as reasons, that the
236
C o nsequences f o r M etaphysics
moral l abor of practical reasoning, of opening her eyes to see the world, lies. This is also why Isabel’s is a story of what she herself does—what I refer to as the moral labor of virtuous practical reasoning—and not merely a story of what befalls her. Let us return to two of the recognition scenes mentioned above: (1) When Isabel walks in on her husband and Madame Merle, the image gives her a fresh impression, in part because she turns her attention from her betrayed self to Madame Merle, whom she once viewed as an image of freedom as independence but who is now felt to be in a mysterious bond such as to shake Isabel’s concept of freedom as inde pendence. This jarring first affects her recognition with regard to her failed marriage during the night, when she is haunted by the horrible visions. In them, Isabel can see her failed marriage as a matter of mutual deception instead of a mere betrayal by her husband; even in the face of g reat horrors, she manages to consider matters from her husband’s perspective as well and not merely through the lens of her own embittered experience. (2) The same is true when Isabel discovers that Madame Merle has indeed arranged the unhappy marriage between Isabel and Mr. Osmond. For this recognition, too, rests on Isabel’s ability to raise the question of Madame Merle’s poor present condition in the face of all the pain caused by the new realization. It is, one might think, in virtue of Isabel’s ability to rise above her own suffering and bitterness to see the others for what they are in themselves, to each other and to herself, that she can, over the course of a series of agonizing recognitions and misrecognitions, have her view of the world collapse in a way that helps open her eyes to the world. While this may be the case, what matters for my purposes here is that, regardless of our interpretation, what Isabel does in each case is to work out the particulars of her situation (what I refer to in this paper as life’s open questions) such that this jeopardizes her entire view of the world (as held together by her concept of freedom as independence) and not merely a part thereof. For Isabel, and for each one of us, the difficulty of morality is immense: it is the difficulty of working out the particulars of our situation in a way that renders us vulnerable to the collapse of our world-v iew. In Isabel’s case, the collapsing world-view threads through the entire fabric of her life and being, such that a collapse of the former is inseparable from a collapse of the latter. What comes undone for her as the story progresses is none other than the thread that runs through the very fabric of her life: her friendships, her marriage, and her guardianship. In this case, the difficulty
S eeing the W o rld
237
involved in bringing the world into view is the same difficulty involved in being so radically vulnerable to the possibility of the collapse of one’s entire life. The possibility of this drama brings in an often overlooked danger endemic to our ability to bring the world into view. Here is how James articulates this danger in the story: on the night of Isabel’s visitation by “the terrors,” one of her visions involves her failure to view her husband as he really was during their courtship, in which the task she had set for herself was precisely that: to see him as he really was. The narrator’s description here is very telling: “She had a vision of him—she had not read him right. A certain combination of features had touched her, and in them she had seen the most striking of portraits.”31 The attentive reader w ill notice that in a novel entitled The Portrait of a Lady, a recognition described as “terror” is the heroine’s recognition that her most fundamental failure of seeing consisted in having seen a portrait instead of a man. As Iris Murdoch sees more clearly than anyone else in her Metaphysics as a Guide to Morals, we are creatures perpetually engaged in the task of bringing the world into view.32 In such a task of judgment, however, we are vulnerable to the danger of successfully bringing concerns, worries, and interests of our own to bear on this or that feature of our circumstances such that we end up forming a portrait—the mere picture or representation we see in a combination of features of the situation at hand. Such a picture or represen tation, in light of the account I sketch here, cannot bring the world into view. Bringing the world into view cannot be accomplished by applying or actualizing a concept (however general and rational) or a world-v iew such as the one McDowell supposes in his aristocratic answer to the question “Who can settle life’s questions virtuously?” No portrait of the world w ill do. Instead, bringing the world into view involves the endless, colossally difficult effort of addressing life’s open questions, in a way that may threaten to pull the thread with which we have woven our entire lives: the concept or concepts that form the underlying structure of our view of the world.
8 I want to end this paper with a few initial and at best suggestive remarks about the emerging view of the difficulty of morality, that is, the moral- deliberative labor of practical reasoning, and with one objection to my
238
C o nsequences f o r M etaphysics
reading of James’s story above. Answering this objection w ill allow me to point out the most controversial feature of the emerging view of practical reasoning. Let me start with the initial, suggestive remarks. On the picture that is emerging in the latter half of this paper, we may note the following: (1) The moral labor of practical reasoning does not consist in specifying what counts as living well in the circumstances, in the light of one’s view of living well expressed in one’s intellectual-cum-motivational capacities, a view that has been molded in one’s upbringing. On the contrary, the moral labor of practical reasoning is the labor of working out the particulars of one’s situation in a way that renders oneself open to the possibility of the collapse of one’s view of what makes life itself worth living. (2) Working out the particulars of one’s situation involves addressing life’s open questions, such as the one Isabel asks of Madame Merle in one of the recognition scenes above: Who are you to me? (3) Doing so may culminate in the utterance of a sentence, in a simple exclamation (as when Isabel says, “Poor Madame Merle!” upon recognizing Madame Merle to be her stepdaughter’s birth mother), an impression, or a sense. In any case, any one of these is the expression of a judgment which reveals one’s previous way of judging as faltering and falling apart. (4) The potential for this kind of judgment lies in each and every one of us, independently of whether we have been raised with a rich conception of how it befits a h uman being to live or whether we are “frail vessels.” In fact there are historical circumstances, such as the ones at the center of James’s story, in which a frail vessel is much more likely to reach this kind of judgment. But one may object that this view emerges from a very particular telling of Isabel’s story. Perhaps, one may think, Isabel’s story can be told as the account of a woman who was brought up into the ability to love the other in the sense of being able to treat the reality of the other as separate from
S eeing the W o rld
239
her own. Such an account could very well do justice to James’s heroine. Therefore, perhaps McDowell is right, a fter all, to raise the issue of the importance of upbringing and habituation in moral m atters. This is all fine. What is not fine is to fall back into a view on which the l abor and the difficulty of virtuous practical reasoning is l imited to that of activating an ability to be impressed by a part or aspect of the world, even if that ability is a second nature, and even if its actualization in the circumstances re uman powers. D oing so disregards quires the use of the entire arsenal of h what, as I have suggested in the first part of this paper, constitutes the philosophical core of McDowell’s ethics: the notion that the difficulty of morality is the difficulty of bringing the world into view by addressing life’s open questions. And as I have attempted to show in my retelling of the Jamesian story, this is as much a m atter of opening ourselves up to the changing perceptions of particu lar people and circumstances as it is a matter of opening ourselves up to the changing perception of the world as a whole. This helps me formulate the final and most crucial feature of the emerging view of the moral-deliberative l abor of practical reasoning: (5) The moral-deliberative issue of bringing the world into view may, at least on occasion, become the issue of upholding our world-v iew or letting it collapse on us. The difficulty of knowing which one we ought to be trying at every stage in our lives as well as the difficulty of trying either is immense. We see this in fiction more than we see it in philosophy, perhaps. But we need philosophy to explain how what we see is possible. And this is the topic of another essay.
✣
Pa rt I V
✣
HISTOR ICA L PR EC ED EN TS
Chapter Nine ✣
See the Right Thing “Paternal” Reason, Love, and phronêsis ✣
Jennifer Whiting
I
am pleased to honor a philosopher whose work has played so important a role in the development of my own that I am sometimes asked if I was his student. Far from it. My first encounter with McDowell was in Oxford in 1982, when he was teaching a seminar devoted to his criticisms of Plato’s Moral Theory by Terence Irwin, who was not only supervising my dissertation but sitting, then and there, by my side. I spent half the time thinking McDowell was dead wrong to read Aristotle as embracing Socrates’s identification of virtue with knowledge when Aristotle seemed so obviously concerned to establish the possibility of akrasia. I spent the other half wondering why McDowell thought it necessary to reject so much of what Irwin had to say about Aristotle’s method and why Irwin thought it necessary to reject so much of what McDowell had to say about Aristo tle’s positive conception of virtue. I was blessed with a second chance when, in 1986, I landed in Pittsburgh simultaneously with John and began to engage seriously with him.1 Even then I was slow to appreciate just how perspicuous his unorthodox readings of Aristotle tend to be. Light began to dawn as I read 243
244
H ist o rical P recedents
the work of Iakovos Vasiliou.2 But it was only when I started to work on what I had eventually to divide into “Locomotive Soul” and “Hylomorphic Virtue” that I was r eally moved: the more support I found for John’s views in texts that he himself rarely (if ever) cited, the more persuaded I was that he had his finger on Aristotle’s pulse.3 My conversion was not well received by those who pride themselves on their scholarship. “It’s all very fine what you say about De Anima, but why do you have to bring McDowell into it?” (Princeton Colloquium in Classical Philosophy, 1997). This explains the fact that “Locomotive Soul,” published in Oxford Studies in Ancient Philosophy, has only a McDowellian “Afterword,” while “Hylomorphic Virtue” appears in a nonhistorical volume devoted to “constitutivism” in ethics. The main argument of “Hylomorphic Virtue” is that Aristotle’s account of moral development is analogous to his account of how the various capacities of soul come to be in an embryo, where the blood of the mother provides the matter, or hulê, in which a creature with the form, or morphê, of the father comes to be. On this account, movements in the seminal fluid contain what we would call “genetic information” and they work on the matter in ways like those in which the movements of an artist’s tools work on matter suitable for realizing the form of whatever he aims to produce. When all goes well, the f ather’s form fully “masters” (kratei) the m atter and the result is a hylomorphic unity that resembles its f ather, even in sex. Things can of course go wrong and often do: fathers sometimes end up with d aughters, which is a good t hing, at least from the point of view of the species. Aristotle describes this as a case where nature “deviates” but says that this deviation is “necessary by nature,” in order that the kind should survive (Generation of Animals 4.3). It is not obvious that this is consistent with his natural teleology, but science is often perverted by prejudice. The crucial point is that Aristotle explains the genesis of female offspring in a way different from that in which he explains the genesis of male offspring. The mechanism is empirically respectable: what acts is generally affected by that on which it acts. But just as Aristotle treats a knife as for cutting in a way in which whatever it cuts is not for blunting, so too he treats movements in seminal fluid as for imposing their form (including their sex) in a way in which menstrual fluid is not for damping these
S ee the R ight T hing
245
movements. Sometimes, however, the menstrual fluid is too g reat in quantity and / or too cold, and the movements in the seminal fluid are “mastered” (krateitai) by the female principle and converted into their opposites (for example, something female or a son resembling his mother and her ancestors). This asymmetry is a feature, not a bug, of Aristotle’s teleology. When things go as they should teleologically speaking go, m atter is mastered by form in a way such that the result enjoys a kind of unity lacking in cases where m atter, acting to some extent in accordance with its own tendencies, prevents a form from being completely realized. When a form is only incompletely realized, t here remain two potentially competing elements in play, one acting in accordance with the form (to the extent that it is realized), the other acting in accordance with the matter (to the extent that it retains some degree of efficient-causal autonomy). I argue that Aristotle adopts a similar account of moral development. An infant starts with a bundle of autonomous desires oriented toward various forms of activity. Early on, it is mostly a bundle of bodily appetites, but spirited and other potentially reason-responsive desires soon emerge.4 Such desires are the “matter” for moral development, on which “form” must be imposed. And it is the practical intellect of the father, or some other parental figure, that initiates the process, in part by instilling in the child a sense of what is noble (kalon), and so to-be-pursued, and what is shameful (aischron), and so to-be-avoided. If all goes well, the child starts to internalize at least some of the relevant values and practical intellect begins to emerge in her. She is still subject to conflict between as yet untamed desires and the commands of her emerging intellect, so she still needs “parental” guidance: she is at best enkratic. But if she is on the right track, her intellect w ill develop in ways such that it can start playing a role analogous to that played by an embryo’s heart once that has come to be: it will govern later stages in her own development as she becomes increasingly able to see for herself what is required by her circumstances and thus wants to act accordingly. In the ideal case, she w ill achieve phronêsis, which is the paradigmatic virtue of practical intellect. McDowell describes this achievement as follows: [M1] It is not that practical wisdom issues o rders, whose content it determines by its own independent operations, to
246
H ist o rical P recedents
motivational propensities that have been separately moulded to obedience. A virtue of character, strictly so called, involves a harmony more intimate than that. Practical wisdom is the properly moulded state of the motivational propensities, in a reflectively adjusted form; the sense in which it is a state of the intellect does not interfere with its also being a state of the desiderative element.5 I concur. But I prefer to cast practical nous rather than phronêsis in the role of form. For I am concerned primarily with the distinctions Aristotle sees between the phronimos (whose desires are fully enformed by practical nous) and her akratic and enkratic counterparts (whose desires are only partially enformed by it).6 I am less concerned with the distinction between the phronimos and the akratês than with that between the phronimos and the enkratês. For it is McDowell’s account of the latter that is responsible for much of the controversy surrounding his view. The key features of this view, for which I aim to provide textual support, are: (1) the identification of virtue with phronêsis; (2) the requisite conception of phronêsis as a capacity that is si multaneously cognitive and desiderative; (3) the association of phronêsis with perceptual powers that distinguish the virtuous agent from o thers, powers acquired only through the sort of habituation involved in a proper upbringing; (4) the appeal to “silencing” to distinguish the psychic economy of the phronimos from that of her enkratic counterpart. McDowell’s appeal to “silencing” is perhaps the most controversial aspect of his view, so I begin and end with it.
1. Silencing and Evidence of Aristotle’s Commitment to It McDowell’s appeal to “silencing” is motivated largely by the need to distinguish the virtuous agent from her enkratic counterpart. Each (like the akratês) reaches what is in some sense the right prohairesis (or decision). And each (unlike the akratês) acts in accordance with that prohairesis. But
S ee the R ight T hing
247
the enkratês is conflicted in a way the virtuous agent is not: he struggles against the sort of desire by which the akratês is carried away. Yet the enkratês succeeds where the akratês fails: the enkratês acts in ways like those in which the virtuous agent acts. Viewed from the outside, his be havior may even be indistinguishable from hers. How, then, are they supposed to differ? McDowell’s argument is essentially that we cannot account for this distinction if we take the virtuous agent to reach the correct prohairesis by weighing the reasons for and against the virtuous course of action and then deciding in favor of the virtuous course, but can account for it if we take the virtuous agent to reach her prohaireseis in ways that involve “silencing.” We cannot account for this distinction by taking the virtuous person to weigh and then act in accordance with her judgment, because this is what the enkratês does. But we can account for the distinction if we take the occurrence of silencing to distinguish the virtuous from the merely enkratic agent: though risk to life and limb is generally a reason to retreat, the virtuous agent, unlike her enkratic counterpart, is “deaf” to this reason in situations where the stakes are such that courage requires her to stand firm.7 One way to model this is to say that the virtuous agent, unlike her counterparts, does not in the relevant circumstances see as good (and so to-be pursued) t hings that she ordinarily (and quite rightly) sees as good (and so to-be-pursued). Another way to model this is to say that the virtuous agent still sees the relevant things as good, but that (as McDowell puts it) “the inferential link” between ‘good’ and ‘to-be-pursued’ is, in her present circumstances, “broken.”8 McDowell prefers the second model because he reads Aristotle as “anxious to avert [the] misconception” according to which “the genuinely courageous agent simply does not care about his own survival.”9 But the first model is compatible with taking the virtuous agent to care about such things. For it does not follow from the fact that a virtuous agent fails to see such things as good in circumstances when having or pursuing them is incompatible with virtue that she does not care about such things. As long as she tends generally to see them as good and tends generally to pursue them, it is plausible to say that she cares about them—especially if she is allowed to wish, in circumstances that render them of no account, that the circumstances had been otherw ise.10
248
H ist o rical P recedents
I prefer the first model because it seems to me a better fit with the evidence there is for taking Aristotle to embrace some version of silencing. And I recommend this model to McDowell on the ground that it fits better with the central role played in his own account by a form of perception distinctive of the virtuous agent. My case for the first model begins with its pedigree. Plato’s Socrates argues against “weighing” in a famous passage where he responds to an imaginary interlocutor who asks w hether he is not ashamed to have led the sort of life that has put him at risk of death. Socrates’s reply begins: “You are wrong sir, if you think a man who is any good at all should take into account [hupologizesthai] the risk of life or death: he should look to this only in his actions, w hether what he does is right or wrong, w hether he is 11 acting like a good or a bad man.” His reply ends as follows: “I go around doing nothing but persuading both young and old among you not to care for your body or your wealth in preference to or as strongly as the best pos sible state of your soul, as I say to you: ‘Wealth does not bring virtue, but virtue makes wealth and everything e lse good for men, both individually and collectively.’ ”12 This is a version of the “conditionality thesis” that Socrates voices elsewhere in Plato: in Euthydemus 279–281 (where sophia and epistêmê stand in for aretê) and in Meno 87–88 (where nous and phronêsis stand in for it). The claim is that virtue (repeatedly identified with a form of knowledge but labeled in various ways) is the only thing that is unconditionally good for a h uman being: the goodness for an individual of all other things is conditional on that individual’s having the virtue (aka wisdom) that allows her to make virtuous (aka wise) use of them. Aristotle too seems to embrace some version of this thesis. [1] Just and fine t hings, which are the subject matter of politics, exhibit much variation and fluctuation, so they may seem to exist by convention only and not by nature. And good things too exhibit some such fluctuation on account of harms that come to many people from them; for some in fact have been destroyed on account of wealth, others on account of courage. [1094b14–19] The initial claim is familiar from Plato’s Socratic dialogues: it is difficult to define virtues in behavioral terms b ecause tokens of one and the same type of action (for example, standing firm in battle) are sometimes kalon
S ee the R ight T hing
249
and sometimes aischron. Aristotle adds that various so-called goods suffer a similar form of “bivalence”: wealth and even (natural) courage are sometimes beneficial and sometimes harmful to their possessors. And he seems elsewhere to connect this with some version of the conditionality thesis. [2] Since the unjust agent is grasping, he is concerned with goods, not all goods but whatever ones are matters of good and bad fortune, ones that are without qualification [haplôs] always good, but not for a particular agent [tini] always good. Men pray for these things and pursue them, but they ought not do so. They ought to pray that the things that are without qualification good are also good for them, but they should select the things that are good for them. [1129b1-6] here are questions about how exactly to understand ‘haplôs,’ which I T follow others in rendering “without qualification.” But “in the abstract”— meaning without regard to the “particulars” associated with any given ere. The instance of the type of thing in question—m ight be better h “particulars” are contingent features attached to individual instances of some type whose tokens are generally good and to-be-pursued. What the “particulars” are must, in each and every case, be determined by empirical investigation. This, as we s hall see, helps to explain Aristotle’s association of phronêsis with perception. The conditionality thesis provides the axiological basis for silencing: things that are generally good (and so to-be-pursued) are not in all circumstances good (and so to-be-pursued). Consider for example money and honor, which are generally good because of the activities they facilitate; or sexual pleasure, which is generally good insofar as it is tied to acts that perpetuate species. We can still say with (2) that some money (for example, blood money) and some honor (for example, honor from base sources, which may corrupt one’s character) is not in fact good; and that some pleasures (for example, adult pleasures taken in sex with young teens) are not in fact good. On the first model, Aristotle associates phronêsis with a context- sensitive power of perception that “sees” or “hears” the relevant things as good when and only when they are in fact good. In cases where the “particulars” switch the standard “valence” of some item, the virtuous agent
250
H ist o rical P recedents
fails to see that item as good, so she does not then and there desire it. She differs in this respect from her counterparts, to whom the relevant item appears good even in circumstances where reason tells them it is not. It is this appearance that draws her counterparts into the sort of “weighing” commonly associated with deliberation about what to do: as long as the item appears good, it w ill seem to her counterparts to be something whose goodness must be outweighed by an even greater good. We shall see how plausible this model is when we come to Aristotle’s distinction between deliberative and perceptual phantasia. What I want to stress here is the power of perceptual phantasia, not just in the life of the akratês, but also, and more importantly, in the everyday lives of nonrational animals, for whom following perceptual phantasiai is the norm. For the best way to appreciate McDowell’s reading of phronêsis is to situate it in the context of that norm. On his account, what moves the virtuous agent is a special case of what moves any animal: namely, perceptual appearances. But the virtuous agent is a very special case. For the appearances to which she is subject are products of what McDowell calls a “proper upbringing.”
2. Psychic Contingency in Plato’s Republic Let’s begin with Plato’s Republic, which provides crucial background for Aristotle’s view. My reading is unorthodox, but I defend it elsewhere.13 It rests on two key ideas, each of which figures in my McDowellian reading of Aristotle. The first posits an explanatory asymmetry between what happens in “good” and what happens in “bad” cases.14 There is a fundamental difference between what goes on in Republic II–I V, where Socrates is concerned primarily with well-ordered souls and states, and what goes on in Republic VIII–IX, where Socrates is concerned with corrupt souls and states. The models and metaphors that dominate II–I V are medical and musical; the key concepts are those of harmony and the proper blend of ele ments stressed in the Hippocratic tradition. The models and metaphors that dominate VIII–IX are political; the key concepts are those of distinct parties ruling and being ruled by one another.
S ee the R ight T hing
251
This difference is signaled in the verbs used in a passage toward the end of Republic IV where Socrates is concluding his discussion of justice and preparing to examine injustice. [3] To produce health is to establish the [components] of the body in a natural relation of mastering and being mastered [kratein te kai krateisthai], one by another, while to produce disease is to establish a relation of ruling and being ruled [archein te kai archesthai] contrary to nature . . . Then, i sn’t to produce justice to establish the [parts] of the soul in a natural relation of mastering and being mastered [kratein te kai krateisthai], one by another, while to produce injustice is to establish a relation of ruling and being ruled [archein te kai archesthai] contrary to nature. [Republic IV: 444d3–1] 15 In the “good” cases, one bodily and one psychic, Socrates employs the Hippocratic language of anonymous elements “mastering and being mastered by” one another in a proper blend. In the “bad” cases, Socrates adopts the political language of anonymous elements “ruling and being ruled by” one another. In a well-ordered soul, there are of course individual appetites, natural and moderate and each “doing their own work”: hunger securing appropriate nutriment, thirst securing proper hydration, sexual desire securing reproduction, etc. But these do not add up to an appetitive “part of soul” in the sense of “part” established by the argument in book IV. For the language of “parts” is a political metaphor drawn from the common use of ‘meros’ to refer to the sort of political factions that figure so prominently in books VIII–I X, where Socrates compares the souls of corrupt individuals to the constitutions of corrupt states. And the argument in IV turns on the presence of conflict within a soul analogous to various forms of conflict possible, but not necessarily present, within a state. On my reading, the existence of an appetitive “part”—whether or ganized around honor, wealth, or pleasure—is a sign that something has gone wrong. Such “parts” should no more exist in a well-ordered soul than their analogues should exist in a well-ordered state. This brings me to the second key: “radical psychic contingency.” Socrates’s view is not just that the so-called parts stand in diff erent relations to one another in different individuals, reason ruling in some,
252
H ist o rical P recedents
spirit in others, and appetite in many. And it is not just that what he calls ‘to epithumêtikon’ can differ in internal structure from one individual to another, being in some cases simply a collection of natural and moderate appetites, each doing their own work, while being in o thers something like an organized political party, pressing o thers for the satisfaction of its own demands. If an individual’s epithumêtikon counts as a genuine part only in the latter sort of case, then the soul of a just agent w ill have fewer parts than the soul of an unjust one. The Republic shows many signs of recognizing such contingency. Book X seems to allow for the possibility of additional parts beyond the canonical three: one (nourished by tragedy) that is prone to pity and one (nourished by comedy) that wishes to produce laughter.16 But the clincher is toward the end of IV, where Socrates says: [4] is not concerned with someone’s doing his own externally, but with what is inside him, what is truly his own. One who is just does not allow any [part] of himself to do the work of another [part] or allow the various classes [ta genê] within the soul to meddle with each other. He regulates well what is really his own and rules himself. He puts himself in order and is his own friend, and harmonizes the three [parts of himself] like three limiting notes of a musical scale—high, low, and middle. He binds together these [parts], and any others there may be in between, and from having been many things he becomes entirely one, moderate and harmonious. [Republic IV, 443c10–e2]17 ere Socrates not only acknowledges the possibility of additional parts H beyond the canonical three; he also recognizes precisely the form of psychic contingency I see in Aristotle. On my reading, Aristotle associates a proper upbringing with a stage in which practical intellect is emerging but has not yet fully “mastered” to orektikon. At this stage, to orektikon is still subject to some desires contrary to the dictates of practical intellect, which may however be “for the most part” correct. Whatever sort of practical knowledge the agent has achieved is still vulnerable to being “dragged about” or “overpowered” by desires, so she may sometimes behave akratically. And even when she resists temptation and succeeds in behaving as the virtuous person would
S ee the R ight T hing
253
behave, this is only because practical intellect has somehow secured the obedience of her orektikon. But not even the reliable obedience of her orektikon to practical intellect is sufficient for genuine virtue. In order to achieve that, her orektikon must “speak in unison” with her practical intellect. This means at least that in a fully virtuous agent, the orektikon agrees with practical intellect in the sense that it is moved by the very reasons by which practical intellect is moved. But the idea may be that they speak in unison because, “from having been two, they have become one.” In this case, phronêsis w ill have the sort of motivational efficacy required for its identification with virtue.18
3. Psychic Contingency in Aristotle’s Politics Before turning to Aristotle’s conception of phronêsis, I want to look briefly at some lines from his Politics that long puzzled me. For I now think that the proper reading of them supports McDowell’s view by undermining a common source of resist ance to it—namely, the idea that Aristotelian virtue presupposes two distinct parts of soul, one of which gives o rders that the other willingly obeys. The lines, which appear in Aristotle’s defense of natural slavery, run as follows: “the soul rules the body with masterly [despotikê] rule, while intellect rules desire with a political or a kingly rule” [1254b5–7]. It seems easy enough, at least prima facie, to understand the second bit. In the kingly case, nous presumably issues commands that orexis is supposed to obey. In the political case, the two must in some sense share in ruling and being ruled. Perhaps they divide the l abor, each giving some commands either to the other or to the agent as a w hole. But what could Aristotle mean in saying that the soul rules the body with a despotic rule? Could he think that the soul gives o rders to the body? Surely not in the case of nonrational animals. And even if his claim were restricted to rational animals, the idea that the soul rules the body by giving orders to it is hard to reconcile with his conception of the soul as the form and the essence of the organic body.
254
H ist o rical P recedents
Let’s look at the puzzling lines in context. [5] Politics 1.5 (a) Ruling is not only necessary but also advantageous . . . T here are many forms of rulers and ruled t hings . . . For whatever is composed out of many and becomes some one common thing [hen ti koinon] . . . in all such things the ruler and the ruled are manifest, and this from nature as a whole is present in ensouled beings. For even in things that do not share in life there is some ruling principle [tis archê], as for example in the case of harmony. [1254a21–33] It is clear from this example that talk of “rule” does not entail that one thing gives o rders to another. Even so, Aristotle might think order- giving is involved in the forms of rule found in animals. But before discussing these, Aristotle makes an import ant methodological point. He distinguishes “good” from “bad” cases and tells us to focus on the former. (b) But the animal is composed of soul and body, of which the former is by nature the ruler, the latter the ruled. One ought to investigate what is by nature [to phusei] in beings whose condition is in accordance with nature [kata phusin] and not in those who are corrupted. So one ought to study the human being who is in the best condition in both soul and body, in whom this [sc. what is by nature] is clear. For in those who are bad or badly disposed, the body often seems to rule the soul on account of its condition being base and contrary to nature [ para phusin]. [1254a22–b2] He continues: (c) It is then, as we’ve said, first of all in the animal that we can study despotic and political rule. For the soul rules the body with a despotic rule, while nous rules orexis with a po litical or a kingly rule. In these cases it is clear that it is in accordance with nature and advantageous for the body
S ee the R ight T hing
255
to be ruled by the soul and for the affective part to be ruled by nous and the part that has logos, while the that involves equal roles or a role reversal is harmful to all. [1254b2–10] The idea that the body might in “bad” cases rule the soul simply adds fuel to the fire: what, in the hylomorphic context, could be meant by this? The key to resolving this lies in combining points made in Politics 1.12 and Eudemian Ethics 8.3. But the resolution undermines my prima facie reading of Aristotle’s claim that nous rules desire with a “kingly or political rule.” Politics 1.12 distinguishes despotic rule from the other forms of rule that prevail in h ouseholds not by appeal to facts about who gives commands to whom, but (in a classic Platonic division along a single axis) by appeal to the proper beneficiaries of each. The proper form of fatherly rule is “kingly,” exercised with a view to the good of his children; the proper form of marital rule is “political,” exercised by husband and wife together for their common good; and the proper form of despotic rule is exercised for the good of the master. But the master should not himself give orders to his slaves, which is undignified. This is clear from EE 8.3, where Aristotle assimilates the relationship between theoretical and practical intellect to that between master and steward. He distinguishes two ways of ruling: either by giving orders or by being that for the sake of which orders are given. Theoretical intellect rules by being that for the sake of which practical intellect gives orders. Questions remain how to understand the original “kingly or political” disjunction. Is the idea that nous rules some desires with one sort of rule and others with another—perhaps appetitive desires with a kingly rule and partly reason-responsive desires with a political rule? Or is the idea that nous rules in different ways in differently constituted subjects—r uling together with desire for their common good in fully virtuous agents but ruling in a more kingly way in enkratic ones? I cannot explore these questions here. But whichever way Aristotle intends the disjunction, his talk of nous “ruling” orexis should not be taken to entail that nous gives commands to orexis. Practical nous may sometimes give commands, Odysseus-style, to orexis—as in “Endure my heart, you have suffered more shameful things than this.”19 But in such cases the
256
H ist o rical P recedents
agent is at best enkratic. And it does not follow that practical nous operates this way in the phronimos, whose orektikon has been fully mastered by practical nous. For Aristotle clearly allows for explanatory asymmetries between the “good” case and the various forms of departure from it. And in the “good” case, practical nous may give commands to the agent herself, not to some “part” of her.
4. Psychic Contingency in the Nicomachean Ethics I introduced the Politics passage because it, properly read, undermines a common objection to taking practical nous to stand to desire as form to matter. Let us turn then to Nicomachean Ethics 1.13, where (as the standard reading has it) Aristotle divides the human soul into four parts. Here again I want to call attention to the ways in which the hylomorphic reading allows us to resolve exegetical puzzles. The standard reading gives rise to a general puzzle: how can we reconcile the four-part division presented in 1.13 with the three-part division that dominates Aristotle’s De Anima and other scientific works (that is, the division into thinking, perceiving, and nutritive-reproductive “parts”)? I think we can best answer this by attending to puzzles internal to the argument actually given in 1.13. For t hese puzzles reveal how difficult it is to find evidence in 1.13 itself that Aristotle takes four parts to be the norm for a human soul. And this in turn supports taking the norm to involve only three parts, with practical nous serving as the proper form of the orectic part. The standard reading has Aristotle dividing the h uman soul in two stages. First comes the division into a part having reason (to logon echon) and a part lacking reason (to alogon). Each of these is then divided into two subparts. To alogon is divided into a vegetative part (which is described as having “no share in logos” and set aside as irrelevant for present purposes) and an “appetitive and generally desiring” part (which is said to “partake of reason in a way, insofar as it is attentive to it and disposed to obey”). Only then does Aristotle divide to logon echon into two subparts, one of which is said to have reason “strictly and in itself” and one of which is said to have reason “just as something which is such as to listen to its father does.”
S ee the R ight T hing
257
It is generally assumed that this division is the same as the one we find in EN 6.1 (= EE 5.1), where Aristotle divides to logon echon into the part that has scientific knowledge (to epistêmonikon) and the part that calculates (to logistikon). But 6.1’s argument for dividing to logon echon differs significantly from the one given in 1.13. So the assumption requires defense. The argument in 6.1 rests on two claims. First, cognition occurs “in accordance with some sort of similarity and appropriateness” of the cognizer to the cognized. Second, the objects with which to epistêmonikon is concerned differ in kind from those with which to logistikon is concerned. To epistêmonikon is that with which we contemplate t hings whose princi ples cannot be otherw ise than they are; to logistikon is that with which we contemplate things that can (or whose principles can) be otherw ise than they are. So to epistêmonikon and to logistikon must themselves differ in kind. They are, therefore, distinct parts of soul, one corresponding to theoretical, the other to practical, intellect. Note, however, how difficult it is to find any explicit distinction in 1.13 between theoretical and practical intellect. Aristotle begins by warning readers that the account he is about to present has only as much precision as required for the purposes of the politician [1102a23–26]. He continues: [6] Nicomachean Ethics 1.13 (a) Some t hings are said about [the soul] adequately in the nonspecialist accounts, and we should make use of t hese: for example, that one of it is nonrational while one has reason. W hether t hese are (a) distinguished in the way that parts of the body (or anything divisible into parts) are, or whether they are (b) two in account but by nature inseparable just as the convex and the concave in the periphery are, makes no difference for present purposes. [1102a27–33] (b) Of the nonrational , one seems to be common and vegetative, I mean the one responsible for nutrition and growth . . . But let us set to threptikon aside, since it is by nature without a share of human virtue. [1102a32–b12] (c) There seems to be some other nonrational nature in the soul, which partakes in a way of logos. For we praise the
258
H ist o rical P recedents
logos, and the part of soul having logos, of the enkratês and akratês. For it urges them correctly, towards the best things. It appears that in them [en hautois] there is also some other natural in addition to logos, which fights against and opposes logos. For as when someone decides to move to the right, paralyzed parts of the body simply go astray in the opposite direction, t owards the left, so too in the case of the soul: for the impulses of akratic tend in opposite directions. But in bodies, we see the that is g oing astray, while in the case of the soul we do not. Perhaps however we should suppose that in the case of the soul too, there is some in addition to reason, being opposed to this and resisting . [1102b12–25] (d) But how this is different makes no difference. And even this seems to partake of reason, as we have said. For in the enkratês, at any rate, it obeys reason. And it is presumably even more ready to listen in the case of the temperate and courageous person; for it agrees with reason in all things. [1102b25–28] So far, Aristotle’s focus is on that subpart of the nonrational part that (c) describes as “partaking in a way of logos.” (c) is primarily concerned with the akratês, whose reason, like that of the enkratês, is praised because it urges him t oward what is best. But t here is in him, as in the akratês, another element that opposes logos and pulls him in the opposite direction. I have italicized ‘in them’ to call attention to the possibility of reading (c) as making a claim only about what parts of soul must be present in the akratês and the enkratês, as distinct from any and every human agent. Politics 1.13 explicitly allows that not all h uman souls have the same parts: the (natural) slave lacks to bouleutikon. We should thus give serious consideration to the reading I propose: it is compatible with what Aristotle says in (c) that the fully virtuous and phronimos agent lacks a nonrational part that fights against and opposes logos. The fully virtuous agent w ill of course have desires, but it does not follow that t hese desires must be h oused in a part of soul distinct from that in which her practical nous resides.
S ee the R ight T hing
259
EN 1.13 continues by further elaborating the nature of the nonvegetative subpart of the nonrational part: (e) A nd it does seem that the nonrational is indeed two-fold. For the vegetative in no way shares in reason, but the appetitive and in general the desiring partakes in a way, insofar as it is attentive to it and disposed to obey. In this way too we say that we possess reason of our father or our loved ones but not of the mathematicians. [1102b28–33] (f) T hat the nonrational is persuaded in a way by reason, admonishment and all censure and encouragement bear witness. But if we must say that even this has reason, then to logon echon will also be twofold, one strictly and in itself and the other just as something such as to listen to his father does. [1102b33-1103a3] It is far from obvious that (f) divides to logon echon into theoretical and practical subparts. For the distinction between what has logos “strictly and in itself” and what has logos “in the way that something such as to listen to his f ather does” reads like one between something that has reason in the sense that it can work things out for itself and something that has reason in the sense that it can follow or obey the reasoning of another. It seems to leave open w hether what is being worked out or followed is theoretical or practical. Even so, I do not think the distinction is best read as topic-neutral one. Aristotle’s point seems in the end to turn on a distinction in kind between the contents of theoretical and practical thought. But to see this, we must attend to the argument actually given. Unlike the argument of 6.1, this argument does not appeal directly to a difference in kind between the objects of theoretical and practical intellect. It rests rather on (e)’s claim that to alogon has logos in the sense that it is attentive to logos and disposed to obey it. The main move occurs in (f), where taking desire to partake of reason is supposed to entail that to logon echon itself has two parts. And the easiest way to understand the alleged entailment is to assume that Aristotle takes a subpart’s having
260
H ist o rical P recedents
logos in the relevant way to render that subpart sufficiently rational to count as a citizen of to logon echon. This supports my proposal to take the theoretical-practical distinction to be implicit in (e)’s distinction between the reason of a mathematician and the reason of a father or some other loved one, and then to read (f) as repeating the same distinction in slightly different terms. The idea is that theoretical intellect has logos in the way the mathematician does, while practical intellect has logos in a different way, one associated with learning from those for whom we have affection because we have experienced their love for us (on which more below). The proposal rests largely on my reading EN 6.5, where Aristotle seeks to distinguish phronêsis both from epistêmê (the topic of 6.3) and from technê (the topic of 6.4). Each of the three is a hexis meta logou connected to some sort of truth. The objects of epistêmê are things whose principles cannot be otherw ise than they are, and epistêmê is a hexis apodeiktikê: it allows its subject to produce the sort of demonstrations discussed “in the Analytics” [1140a31–33]. The objects of phronêsis and technê are t hings that can (or whose principles can) be otherw ise than they are. Each is concerned with distinctively h uman goods, but technê is “productive” while phronêsis is “practical” [1140b20–21]. Aristotle seeks in what follows to clarify what he means by “practical.” In technê, the most choice-worthy practitioner has a kind of expertise that allows her to err at w ill; but in phronêsis the opposite (so to speak) is true. Aristotle adds, to further elucidate the difference, that phronesis is “not a hexis involving logos alone” [1140b28–30]. The idea, I think, is that desire is involved in phronêsis in a way it is not involved in technê: a skilled practitioner might or might not desire to bring about the end of her craft as such, but phronêsis involves desire in a way such that its subject cannot fail to act as phronêsis commands—unless, perhaps, something intervenes. My proposal, then, is to read (f)’s talk of a part of soul that has logos “strictly and in itself” as referring to theoretical intellect, whose activities do not involve any even potentially nonrational psychic states or attitudes as such. Theoretical intellect is meta logou monon in the sense that its subject’s ability to produce demonstrations does not depend in any essential way on the contents of that subject’s desires: the knave can do math even if a fool cannot.
S ee the R ight T hing
261
The contrast is with practical intellect: its proper hexis is oude meta logou monon in the sense that desires that tend to be only partly reason- responsive play an essential role in its subject’s grasp of the sort of truth with which it is concerned. Unlike physical pleasures and pains, whose ability to impede any and e very sort of thought is topic-neutral, desires that are only partly reason-responsive have the power to wreak topic- specific havoc on their subjects’ practical thoughts. They do not simply interfere with the actualization of a subject’s first actuality capacity for thought; they interfere with the capacity itself in a way that threatens to corrupt the contents of its actual thoughts. Aristotle clearly recognizes this phenomenon. [7] It is not every supposition [hupolêpsis] that pleasure and pain destroy or corrupt—not, for example, [the supposition] that the triangle does or does not have two rights—but rather those concerned with what is to be done. For the principles [hai archai] of what is to be done are that for the sake of which the things to be done . But to one who has been corrupted through pleasure or pain straightaway the principle fails to appear: it does not appear to him that for the sake of this or on account of this he o ught to choose and do all things. For vice is destructive of archê. So it is necessary that phronêsis is a true hexis, involving logos, that is concerned with h uman goods and practical. [1140b13–21] It is admittedly odd to read Aristotle as describing practical nous and desire as possessing logos in the same way; and perhaps even odder, given his characterization of phronêsis as ‘epitaktikê,’ to have him describing practical nous in terms of its capacity to attend to and obey logos in something like the way we might attend to and obey the logos of our f athers. The second oddity is easily mitigated. For phronêsis is not the only virtue of practical nous. There also sunesis, which is kritikê and involved in judging when someone else is speaking [1143a6–22]. Sunesis may be associated with judging the reasons given by another for his behavior; or with advising another about what, given how he describes his situation, he ought to do. Or it may be associated with comprehending and thus coming to accept reasons given by another about what one ought oneself to do. It may be all of the above. We cannot resolve this h ere. The crucial point is
262
H ist o rical P recedents
that t here is a virtue of practical nous that is exercised in listening to and perhaps also following the logos of another.20 We can mitigate the first oddity by reading Aristotle’s talk of practical nous and the desiring part of soul as possessing nous “in the same way” as referring to a single part which is referred to in diff erent ways in diff erent contexts. Aristotle might well talk this way if he takes the norm for a human soul to include a part that is, so to speak, “bi”: neither purely rational (as the epistêmonikon is) nor purely nonrational (as the threptikon is) but something with rational and nonrational aspects. In a fully virtuous agent, this “practical” part w ill be unified in what Aristotle calls the strictest sense: its rational aspect w ill stand to its nonrational aspect as actuality to potentiality [De Anima, 412b6–9]. This may explain the “no difference” claim in [6](a). Most commentators seem to assume that Aristotle takes the relations among the various so-called parts to be uniformly distinguished either as parts of the body are or as the convex and concave are but suspends judgment on a question whose answer is not immediately relevant. But I read him as leaving room for the kind of psychic contingency I see in Plato’s Republic, where what are at early stages of development “many” can “become one” but may nevertheless remain to some extent distinct. In the latter case, Aristotle might say that they are related in something like the way parts of the body are: they are capable of pulling in different directions even when they in fact pull the same way. If Aristotle allows such contingency, then we can easily reconcile the apparently four-part division in EN 1.13 with the three-part division that prevails in his scientific works. The scientific division yields: (1) nutritive, (2) perceptive, and (3) thinking soul. For reasons explained in “Locomotive Soul,” I identify (3) with theoretical intellect and place practical intellect (along with imagination) in (2). This allows Aristotle to identify a nonrational animal’s act of perceiving (or imagining) something as pleasant with an active desire to pursue that thing, and its act of perceiving (or imagining) something as painful with an active desire to avoid that thing. Such unity is the norm among nonrational animals. On the hylomorphic reading, such unity is also the norm for rational animals. But the role played by thought in producing the images on which they act allows for split cognition and so for split motivation. Aristotle discusses such splits in De Anima 3.11, where he distinguishes deliberative
S ee the R ight T hing
263
phantasia (to which only rational animals are subject) from perceptive phantasia (to which most if not all animals, including rational ones, are subject).21 In rational animals, perceptive phantasia may present a certain course of action as pleasant (and so to-be-pursued), while deliberative phantasia presents its as bad (and so to-be-avoided). Or perceptive phantasia may present a certain course of action as painful (and so to-be- avoided), while deliberative phantasia presents it as good (and so to-be- pursued). Such complications render rational animals vulnerable to the forms of psychic conflict associated with akrasia and enkrateia. But these are anomalies in the animal world. The unity of practical cognition with desire is the norm. This norm supports McDowell’s identification of phronêsis with a perceptual capacity whose exercise moves its subject immediately to act. But the identification is controversial, so let’s turn to it.
5. The “Deductivist Prejudice”: From “Uncodifiability” to the Perceptual Conception of Phronêsis McDowell’s argument for this identification turns primarily on the “uncodifiability” of the principles by which the phronimos is guided. McDowell associates this with the idea that these principles hold, as Aristotle says, only “for the most part.” But t here is more to the story than that: perception of particulars plays a special role in phronêsis, one dif ferent in kind from the role it plays in sciences whose principles hold only “for the most part.” We can best appreciate the general structure of McDowell’s argument by starting with his early, mostly in propria persona “Virtue and Reason.” This is surely the primary source of the scholarly animus against McDowell’s Aristotle, less I suspect for its appeal to Wittgenstein than for its extended quotation of Stanley Cavell. But we cannot appreciate McDowell’s reading of Aristotle if we do not understand what is g oing on h ere, where he aims to “exorcise” a “prejudice” about the nature of rationality that fuels resistance to his purportedly Aristotelian account of phronêsis. I say “purportedly” not to cast doubt on the attribution to Aristotle, but to register the fact that McDowell is explicitly rejecting standard
264
H ist o rical P recedents
accounts of Aristotle’s practical syllogism. According to such accounts, the actions of the phronimos are to be explained in terms of an interaction, which is assimilated to deductive inference, between the agent’s grasp of some universal knowledge of how she o ught to act (expressed in the major premise) and the agent’s grasp of some perceptible fact about the situation in which she finds herself (expressed in the minor premise). The prejudice, which McDowell calls “deductivist,” is the idea that “acting in the light of a specific conception of rationality must be explicable in terms of being guided by a formulable universal principle.”22 I italicize “guided by” so as to call attention to the way in which the issue is not simply (as often assumed) w hether or not t here is a formulable universal principle: the issue is rather what is involved in being guided by a principle, whether formulable or not in universal (or even “for the most part”) terms. McDowell begins with Wittgenstein’s example of a case where there is a formulable rule: “Add 2.” He then follows Wittgenstein in questioning the idea that grasping the rule and continuing to apply it involves two factors: (1) a psychological mechanism that reliably produces correct applications of the rule; and (2) something “objectively t here,” like rails, to keep this mechanism “on track.” His strategy is to undermine resis tance to the idea that one can be guided by a principle that is not formulable in universal (or even “for the most part”) terms by dispelling the idea that these factors are present in cases where there is such a principle. McDowell thus moves from the case of addition, where there is a formulable rule, to a diff erent sort of case, the competent use of words. It is h ere that he quotes Cavell: We learn and teach words in certain contexts, and then we are expected, and expect others, to be able to project them into further contexts. Nothing insures that this projection w ill take place (in particular, not the grasping of universals nor the grasping of books of rules), just as nothing insures that we w ill make, and understand, the same projections. That on the w hole we do is a matter of our sharing routes of interest and feeling, modes of response, senses of humor and of significance and of fulfilment, of what is outrageous, of what is similar to what e lse, what a rebuke, what forgiveness, of when an utterance is an assertion, when an appeal, when an explanation—all the whirl of organism Wittgenstein
S ee the R ight T hing
265
calls “forms of life”. Human speech and activity, sanity and community, rest upon nothing more, but nothing less, than this.23 McDowell draws on this in answering an objection to his conception of a proper Aristotelian upbringing as “initiation into a conceptual space [one organized by the concepts of the kalon and the aischron] by being taught to admire and delight in actions in the right way.” The objection targets the idea that what is being instilled in such habituation is a conceptual framework in which there is, relative to this or that particular set of circumstances, an objectively correct move to be made. As McDowell puts the objection, the product of such habituation “looks rather, like a congruence of subjectivities, with the congruence not grounded as it would need to be to amount to an objectivity.”24 McDowell draws on Cavell to stress similarities with the ways in which we are initiated into other forms of conceptual space, including mathematical ones. But his talk of admiring and delighting in actions in the right way signals the crucial difference to which I think Aristotle himself alludes when he distinguishes phronêsis from mathematical and other forms of epistêmê by saying that phronêsis is not meta logou monon. Aristotle’s point is not just that the principles by which the phronimos is guided hold, at best, only “for the most part.” His point is also that these principles, to the extent that they can be formulated, tend to involve concepts that give rise to “hard cases,” cases where (as McDowell puts it) disagreements about whether or not a concept applies “resist resolution by argument.” The difficulties here are due in part to the forms of bivalence introduced in [1]. Even if there are some dangerous situations where calling the police (such as they are) is the t hing to do, t here are many dangerous situations where that is exactly the wrong thing to do: m ental health workers would better “serve and protect.” But there are further difficulties due to the ways in which pathê are involved in coming to grasp and in applying the relevant concepts. It is especially in “hard cases” that pathê, including desires, are likely to interfere with an agent’s capacity to “read” the particular circumstances in which she finds herself. The operative pathê can be negative, as when fear leads police to see threats more readily in the behaviors of unarmed Black men than in similar behaviors on the part of armed white men. But
266
H ist o rical P recedents
t hese pathê can also be generally positive, as when love misleads a father whose legally minor daughter has suffered statutory rape to press charges when doing so w ill only intensify her trauma in ways other options would not. His love may be the very thing that makes it hard for him to hear what she is saying when she resists his decision. And Aristotle might describe his failure as one of sunesis, a virtue of the practically oriented logistikon exercised in grasping what others say. Or, to consider another case, a father’s love may distort his grasp in a more solicitous way, misleading him into giving consent for his d aughter, still underage, to marry a man three times her age who has impregnated her. The roles played by pathê in allowing agents to read correctly the par ticular situations in which they find themselves are what make it plausible to take perception of particulars to play a role in phronêsis that differs in kind from the one it plays in sciences whose principles hold only “for the most part.” We s hall return to this point. My hope is that t hese examples, even if they do not persuade, lend some respectability to McDowell’s argument: [M2] It is not to be supposed that the appreciation of the partic ular instance that is explicitly appealed to in the second kind of case [where t here is no formulable principle] is a straightforward or easy attainment on the part of t hose who have it: that e ither, on casual contemplation of an instance, one sees it in the right light, or e lse one does not, and is then unreachable by argument. First, “Don’t you see?” can often be supplemented with words aimed at persuasion. A skillfully presented characterization of an instance w ill sometimes bring someone to see it as one wants; or one can adduce general considerations, for instance about the point of the concept a particular application of which is in dispute. Given that the case is of the second kind, any such arguments w ill fall short of rationally necessitating acceptance of their conclusion in the way a proof does. But it is only the prejudice I am attacking that makes this seem to cast doubt on their status as arguments: that is, appeals to reason. Let us turn then to McDowell’s reading of Aristotle.
S ee the R ight T hing
267
6. McDowell’s Key Texts: EN 1.3–4 McDowell rests the case for his reading on two sorts of passages: methodological ones that distinguish forms of knowledge by appeal to differences in their respective contents; and ones that make substantive points, especially about cognition and moral development. Prominent among the former are passages where Aristotle says that the subject m atter of practical thought admits of less precision than the subject matter of scientific thought, and where he contrasts the contingent, variable, and particu lar objects of practical thought with the necessary, unchanging, and universal objects of scientific (especially mathematical) thought. Prominent among the latter are passages where Aristotle suggests that logoi in the ethical-political domain, unlike those in scientific domains, can be grasped only by someone who has had a proper upbringing. It is not always clear how t hese points are related, but Aristotle surely takes empeiria to serve as a kind of m iddle term. His idea seems to be that only someone who has had the relevant kind of experience is in a position to grasp the contingent, variable, and particular truths that are the proper objects of phronêsis. Let’s begin with EN 1.3–4, where Aristotle himself seems to connect the dots. In the lines following [1]’s assertion of bivalence, he moves from (a) points about the need to seek only so much precision in any domain as the nature of the subject admits; to (b) claims about why young p eople, who lack experience in the actions of life, are not appropriate students of politics; and concludes (c) that it is necessary for one who is to benefit from instruction about “fine and just t hings” to have had what McDowell calls “a proper upbringing.” The points in (a), together with points in EN 6, support McDowell’s claim that the principles by which the phronimos is guided are uncodifiable. But t here is more to the uncodifiability than the idea that “for the most part” premises yield only “for the most part” conclusions. For that idea fails to distinguish phronêsis from the natural sciences, which (like mathematical sciences) are meta logou monon. And though it might seem from 6.8 that Aristotle locates mathematics on the one hand and politics on the other at opposite ends of what is simply a continuum that runs from the most abstract forms of knowledge (which are more easily taught and
268
H ist o rical P recedents
acquired) to the least abstract, with the natural sciences (involving experience of particulars) falling in between, this fails to explain the connections envisaged in 1.3–4. If what differentiates phronêsis from the various forms of epistêmê is primarily its position on a continuum from the more to the less abstract, why does EN 6 not say simply that it takes even longer to become phronimos than to become phusikos? Why does (c) require that students of politics have had a proper upbringing? Of what relevance are their habits? Aristotle cannot appeal simply to the importance of self-control and good study habits, broadly construed to include the sort of practice involved in mastering a sport or musical instrument: such habits are impor t ant in all domains. The idea seems to be that the experience required of the would-be phronimos differs in kind from that required of the would-be phusikos. For Aristotle says in (b) that what prevents young folk from achieving phronêsis is the fact that they tend to live in accordance with pathos, pursuing each thing in accordance with it. The problem is not just that they need to see more action: they need more of the sort of experience involved in acting, including experience of the consequences of their own actions. And they especially need experience acting in accordance with reason. In other words, young folk need the sort of experience afforded by repeatedly hearing and following the advice of a trusted mentor who has herself reached some approximation of phronêsis and can provide the sort of skillful characterization, or reminders of the point of some concept, to which [M2] refers. But the uncodifiability of the relevant principles means that the mentor cannot teach these principles in the way she might teach the universal principles of mathematics or even the “for the most part” principles of physics: the relevant principles can be acquired only through habituation, whose credentials as a mode of acquiring principles Aristotle recognizes (along with those of perception and induction) in EN 1.7. Opponents of this reading might latch onto 1.3’s comparison of young folk to akratic agents, who fail to benefit from whatever knowledge (gnôsis) they have achieved. For this comparison might be taken to show that 1.3 is concerned only with the disposition to act on the relevant knowledge and not with the forms of comprehension involved therein. But that is just what McDowell denies when he reads Aristotle as identifying the knowledge in question with virtue: neither the akratês nor the enkratês comprehends quite
S ee the R ight T hing
269
what the virtuous agent comprehends. And t here is evidence, both h ere and in EN 10.9, that Aristotle’s point is at least partly about comprehension. [8] Argument and teaching would not prevail in all , but it is necessary to prepare the soul of the student by means of habits for enjoying nobly and hating , just as someone nourishing the seed earth. For one who lives kata pathos would not listen nor generally understand an argument turning him away [ou gar an akouseie . . . oud’ au suneiê]. [1179b23–28] Aristotle has already hinted, back in EN 1.4, at possible reasons for such failures of comprehension. A fter citing with approval Plato’s distinction between paths to and paths from the principles of a subject matter, Aristotle says that we must begin from things that are known / knowable (gnôrima). He then distinguishes t hings known / knowable to us (hêmin) from things known / knowable without qualification (haplôs) and goes on to say: [9] Whence it is necessary for one who is to learn sufficiently about fine and just things, and political m atters in general, to have been properly [kalôs] habituated. For the archê is “the that” [to hoti], and if this should appear adequately, there w ill be no need in addition for “the why” [tou dihoti]. For such a person has or might easily acquire the archai. [1095a4-8] As Vasiliou notes, Aristotle elsewhere asserts the importance of starting from t hings known to us but does not in scientific contexts follow up with the demand for anything comparable to a proper upbringing. But Vasiliou quotes a passage that indicates the natural scientific analogue: [10] People who are puzzled about whether one ought to honor the gods and love one’s parents need punishment, while t hose who are puzzled as to whether snow is white or not need perception.25 In the scientific case, failure to know of some subject of investigation that it has some property for which the “because” is sought is remedied by perception of suitable number of untainted samples. The methodological point of [5](b) goes without saying: we do not study yellow snow. In the
270
H ist o rical P recedents
practical case, t hose who are puzzled about the “thats”—those, for example, who do not see that loving one’s parents is (at least “for the most part”) kalon—need affective training (and perhaps even, in cases where their grasp of the relevant principles has been corrupted, punishment, though it is not clear that this r eally works). The argument in [8] leads directly to a discussion of the role played by fathers in a proper upbringing, on which more below. We must return h ere to the reasons for taking 1.3 to be concerned with comprehension and not just with acting on what one comprehends. Initiation into the conceptual space of to kalon and to aischron is complicated not just by the bivalence mentioned in [1]. It is also complicated by the unity of “strict” or “authoritative” (kuria) virtue, as distinct from the sort of “natural” virtue Aristotle seems to have in mind when he says in [1] that people are sometimes destroyed by courage.26 This unity plays an important role in McDowell’s view. [M3] If a genuine virtue is to produce nothing but right conduct, a simple propensity to be gentle cannot be identified with the virtue of kindness. Possession of the virtue must involve not only sensitivity to facts about others’ feelings as reasons for acting in certain ways, but also sensitivity to facts about rights as reasons for acting in certain ways; and when circumstances of both sorts obtain, and a circumstance of the second sort is the one that should be acted on, a possessor of the virtue of kindness must be able to tell that this is so. So we cannot disentangle genuine possession of kindness from the sensitivity that constitutes fairness. And since t here are obviously no limits on the possibilities for compresence, in the same situation, of circumstances of the sorts proper sensitivities to which constitute all the virtues, the argument can be generalized: no one virtue can be fully possessed except by a possessor of all of them, that is a possessor of virtue in general. Thus the partic ular virtues are not a batch of independent sensibilities. Rather, we use the concepts of the particular virtues to mark similarities and dissimilarities among the manifestations of a single sensitivity, which is what virtue, in general,
S ee the R ight T hing
271
is: an ability to recognize requirements that situations impose on one’s behavior. It is a single complex sensitivity that we are aiming to instill when we aim to inculcate a moral outlook.27 McDowell speaks here in his own voice, but it is clear from EN 6.12–13 that Aristotle is committed to such unity. For he treats the “natural” virtues as separable from one another in a way that genuine virtues (sic) are not. The “natural” virtues are modular tendencies to engage in the sorts of behavior stereot ypically associated with some particular virtue. The fact that such tendencies can destroy their subjects is part of the argument for refusing to count them as genuine virtues, but what lies b ehind this fact is (as 6.13 says) their failure to be enformed by practical nous. The naturally courageous agent may rush into battle when nothing of value is r eally at stake. And the naturally generous agent may give so much to those whose needs are modest, or to those unlikely to make effective use of his gifts, that he has nothing left to give when he encounters people in great need who would surely make effective use of his gifts. These are the sorts of mistakes a young person might make, and they may impede comprehension of arguments about fine and just things. It may be hard for a young person to grasp that certain forms of giving are not in fact good, and so not in fact generous. Never mind the why. For there are countless possible “whys” h ere: a gift may undermine the self-respect of the intended beneficiary or facilitate forms of dysfunction that harm him and / or others, and so on.28 Such mistakes are surely among the proper objects of gnômê, which is a form of understanding in virtue of which Aristotle takes mature agents to be “forgiving” (sungnômôn). For such mistakes are all too intelligible given the bivalence endorsed in [1]. Comprehension is further complicated by the fact that is not behav iors themselves that bespeak virtue, but rather the prohaireseis that enform the behaviors: the virtuous agent chooses to engage in the relevant sort of behavior because doing so is kalon, while the enkratês (or an agent on his way to virtue) may engage in it b ecause failing to do so would be aischron. So determining when an action of a certain type is genuinely kalon can be tricky. If someone gives to a good cause because he thinks doing so w ill win him an office he can exploit for his own advantage, then
272
H ist o rical P recedents
what he does is not the same as what the genuinely virtuous person does when she gives the same amount, or the same percentage of her income, to the same cause. And even when an agent’s motives are admirable, he may still go wrong. If, for example, he fails to see that the gift w ill do more harm than good to its recipient, then his act of giving w ill fail to count as kalon.29 For t hese and other reasons, acquiring the “thats” is generally more difficult in practical domains than in scientific ones. But, as [9] says, someone who has the “thats” “either has or can easily acquire” the princi ples by which the phronimos is guided. This is largely because what is being done is a function of the prohairesis that enforms it. So it is only if the student grasps the relevant prohairesis that she fully grasps what is being done. But the prohaireseis that enform the virtuous person’s actions w ill be opaque to someone who lacks basic familiarity with the concept of to kalon and its pedagogic partner to aischron. Saying this does not require us to read Aristotle as taking actual virtue to be necessary for proper comprehension of logoi about ethical / political matters. Aristotle’s point is simply that without some experience of one’s own and others’ actions as kalon (and so to-be-pursued and / or praised) or as aischron (and so to-be-avoided and / or blamed), the student w ill not be in a position to comprehend the relevant logoi. Here, however, it is crucial to honor the range of phenomena to which ‘logos’ might refer: not just the sort of deductive argument a mathematician might give but also the sort of skillful characterization a trusted advisor might give so as to persuade her advisee to “see” a course of action as she herself “sees” it. Of course the latter works only if the advisee has an adequate grasp of the concepts with which the speaker works: but we should not confuse an “adequate” grasp with a “complete” one. The former puts one on the way to achieving a complete grasp in the only way in which one can achieve that. Just as the heart, once it comes to be, controls the subsequent development of the embryo, so too what Vasiliou calls “proto-phronêsis” must control the development of phronêsis strictly so called. For this is required if the reasons for which the phronimos acts are to count as her own. McDowell seeks in “Some Issues in Aristotle’s Moral Psychology” to explain what, in lieu of formulable principles, the virtuous agent is supposed to possess: namely, a correct but largely uncodifiable conception of what it is “to live well” or to achieve to eudaimonia, a conception ac-
S ee the R ight T hing
273
cording to which “living well” consists largely in “doing well.” McDowell treats the contents of this conception as the “major premise” (or premise of “the good”) on which the agent acts. Equipped with this conception, the phronimos is able to “see,” on a case-by-case basis, what really matters in each of the situations in which she finds herself and so what counts as “doing well” then and there. On this account, the role played by the perceptual capacity with which Aristotle associates phronêsis is not to reveal the truth of the minor premise (or premise of “the possible”). For that—for example “here is an opportunity to comfort my friend”—is something whose truth anyone, whether virtuous or not, might grasp. The perceptual capacity is rather what allows the virtuous agent to determine which of the various partic ular truths she grasps are what “really matter” in the present situation. Perhaps what it would take to comfort her friend here and now would serve simply to reinforce the very prejudices that make him so uncomfortable in the present racially charged situation, so that her commitment to promoting racial justice requires her to speak up in ways that increase rather than mitigate her friend’s discomfort, which is, h ere and now, of no account. The sort of perception involved in seeing that speaking up is kalon is what McDowell associates with phronêsis. It is this sort of perception that leads the agent to select, from among the various particular truths she grasps, which is to serve as the minor premise on which she w ill act. McDowell supports this reading by appeal to two passages in EN 6. Let’s turn then to t hese.
7. Aristotle on Phronetic Perception The first passage appears at the end of EN 6.8 where Aristotle is discussing the capacity to deliberate well that he takes to be characteristic of phronêsis. He makes some remarks here about why young people can be mathematicians and geometers and wise in such matters but seem not to become phronimos: “the reason is that phronêsis is also of particulars, which come to be known / knowable from experience, and a young person is not experienced” [1142a11–15]. Then, a fter noting that natural science presupposes more experience than mathem atics does, he contrasts phronêsis with epistêmê:
274
H ist o rical P recedents
[11] That phronêsis is not epistêmê is clear. For it is of the last t hing . . . for what is to be done is such. [Phronêsis] corresponds to [antikeitai] nous. For nous [viz. theoretical nous] is of the terms, of which there is no logos, but it [viz. phronêsis] is of the last thing, of which there is not epistêmê but rather aisthêsis, not aisthêsis of proper but the sort of aisthêsis by which we perceive that the last thing in mathematical is a triangle. For it [aisthêsis] will come to a stand there as well. But this [sc. the sort by which we perceive the last thing in mathematical domains] is more aisthêsis than phronêsis is, but a diff erent form of that [sc. of aisthêsis]. [1142a23–30] I cannot defend this admittedly interpretative translation in detail. The case for it, especially for construing the reference to nous as a reference to theoretical nous as such, rests largely on my reading of the second passage. [12] Nicomachean Ethics 6.11 (a) Nous is concerned with last things in both . For nous, and not logos, is concerned with the first terms and the last ones. The kind that operates in accordance with demonstrations [viz. theoretical nous] is of unchanging terms and first t hings, while the kind that operates in practical is concerned with what is last and capable of being otherwise and with the other [sc. the minor] premise. For these are starting points [archai] of that for the sake which . For from the particulars the universals . Of these then [viz. the particulars] it is necessary to have perception, and this is nous [viz. practical nous]. [1143a35–b5] (b) W hence t hese seem also to be natural, and though no one is by nature sophos, to have gnômê and sunesis and nous. A sign of this is that we believe that these follow particul ar periods of life, and that a certain period has nous and gnômê, as if nature w ere the cause. So we o ught to attend to the undemonstrated sayings and opinions of t hose
S ee the R ight T hing
275
who are experienced and older and phronimos, no less than to demonstrations. For on account of having the eye of experience, they see correctly. [1143b6–14] Aristotle is concluding his discussion of the intellectual virtues; (b) summarizes his discussion of the hexeis that are especially concerned with particulars. He has said at the outset that gnômê is the hexis in virtue of which people are said to be “forgiving,” which connects the relevant form of comprehension to the pathê of its subject. But the crucial point here is the way (a) supports my reading of [12] as focused on the relationship of phronêsis to theoretical nous as such. Theoretical nous grasps the definitions from which its proofs (which are demonstrative) begin, which definitions cannot themselves be proved; practical nous grasps the contingent things (principally actions) by means of which the end is achieved. The idea in [12](a) is that the two forms of nous, although they are distinct, are similar insofar as each involves a form of cognition that is not itself the product of deductive argument. This is the basis of my decision to render [12]’s use of ‘antikeitai’ with “corresponds to” rather than the more common “is opposed to.” H ere, as in [12](a), Aristotle calls our attention to an important respect in which the operations of phronêsis are similar to those of theoretical nous, at least in some of its domains. It is not entirely clear what Aristotle means when he speaks of the way in which “we perceive that the last t hing is a triangle” or what exactly is supposed to “come to a stand there as well.” But given the context, with its focus on the sort of deliberation which it is the task of the phronimos to do well, it is natural to connect this with EN 3.3, where Aristotle assimilates deliberation to the form of investigation in which a mathematician working with a diagram is engaged. In both cases, some end is set and the investigator searches for a way to reach it. And in both cases Aristotle seems to think that there is a kind of analysis that proceeds backward from the specification of the end to “the last t hing in the analysis, which is the first thing in the coming to be” of the end to be reached [1112b20–24]. In the mathematical case, it may be that the first step in constructing the requisite proof is to construct a triangle. In the practical case, “doing well” is the end and the archai mentioned in [12](a) are contingent actions the agent might (or might not) perform in order to realize that end. The
276
H ist o rical P recedents
virtuous agent perceives the particular circumstances in which she finds herself and does so in a way that enables her to discover what she can best do, in those circumstances, to act on the conception of living well that sets the terms, so to speak, of her deliberation. Let us return then to [12], which is one of the strongest bits of evidence for McDowell’s view. Its value lies in the way it both assimilates phronetic to geometrical perception (so as to establish the cognitive credentials of phronêsis) and distinguishes phronetic from geometrical perception (so as to recognize the role played by desire in phronetic perception). The assimilation comes in the comparison of phronêsis to two forms of perception, each of which is generally recognized as instantiating a form of knowledge: (1) perception of proper perceptibles, such as colors, sounds, flavors, etc.; and (2) whatever sort of perception is involved in the mathematical case where one perceives (for example) that the last t hing is a triangle. Aristotle says in the De Anima that the perception of proper perceptibles either does not admit of error (2.6) or (in what is perhaps a correction of this) is among the various forms of perception least susceptible to error [428b18]. And we have just seen a plausible way to understand the form of perception involved in the geometrical case. But however exactly we understand that, the point of the last sentence is that phronêsis involves a kind of perceiving that resembles the other two in that neither its contents nor its subject’s grasp of these contents can be secured by means of an argument. But this does not undermine the status of these forms of perception as exercises of knowledge. So we should not take the perceptual character of phronêsis to undermine its status as an exercise of knowledge. But what about the alleged difference between phronetic and mathematical perception at the end of [12], where the latter is said to be “more aisthêsis” than the former “but a diff erent form of it”? I read this as referring back to Aristotle’s claim, at the end EN 6.5, that phronêsis differs from epistêmê (strictly so called) in that phronêsis is oude meta logou monon. Insofar as exercises of phronêsis involve desire, phronetic perception differs in eidos from mathematical perception, which is “more aisthêsis” in the sense that it is exclusively cognitive and without the conative aspect characteristic of phronetic perception. Someone might object that there is a better explanation, readier to hand, of the alleged difference in eidos: namely, the special role played by
S ee the R ight T hing
277
experience of particulars in the acquisition of phronêsis, a role to which Aristotle has just called attention. I agree that this is part of the explanation. But taking this to exhaust the explanation seems to me problematic because I do not see how this suffices to explain the claim that the mathematical form is “more perception” than the phronetic form is. So I read [12] as saying that mathematical perception is more exclusively perceptual than phronêsis is in the sense that it, unlike phronetic perception, does not involve desire or other pathê. We can best appreciate this by returning to some points made in my discussion of EN 1.13.
8. Acquiring Practical Nous: Affection and Epistemic Trust My reading of EN 1.13 turns on aligning the distinction between theoretical and practical intellect with the distinction between a part of soul that has reason in the way in which someone who comes to have the reason of a mathematician has it and a part of soul that has reason in the way in which someone who comes to have the reason of a parental figure has it. I take the point about the perfected state of practical intellect, as distinct from the process that leads to it, to be as follows. Mathematicians can give proofs, so a part of soul has reason in the way a mathematician does when it can generate proofs—that is, when its hexis is apodeiktikê. But our fathers are not expected to give proofs about the sorts of things we learn from them qua fathers—that is, practical matters. For such t hings tend not to admit of proof. So a part of soul can have reason in the way a father does without being able to generate proofs. To have reason in the way a (decent) f ather does, a part of soul must be such as to generate (decent) prohaireseis. But prohairesis, unlike apodeixis, involves desire. Aristotle in fact characterizes prohairesis as a form of desire: it is nous orektikos or (in other words) orexis dianoêtikê [1139b4–5]. So exercises of practical intellect involve desire in a way that exercises of theoretical intellect do not. But let us approach the perfected state of practical intellect by way of some things Aristotle says about how it comes about.
278
H ist o rical P recedents
The mathematician, unlike the phronimos, is expected to demonstrate the truth of her conclusions by showing that they follow logically from principles proper to mathematics (together, of course, with a few axioms common to all the sciences). In order to grasp her proofs in the way that someone acquiring mathematical epistêmê must do, the student must grasp the relevant principles and axioms, but he need not grasp anything outside of these. That is part of the point in [6](f), where Aristotle describes the relevant subpart of to logon echon as having reason “strictly and in itself.” But Aristotle contrasts this way of having logos with the way in which we have the logos of our fathers and other loved ones (henceforth “parental figures”). His point is not simply—as in [12](b)—that in order to learn the sorts of things that do not admit of demonstration we must attend to the undemonstrated (anapodeiktois) sayings and opinions of those who are experienced or practically wise in the relevant m atters. His point is also that with respect to at least some of these things, affection for those from whom we learn them plays an ineliminable role. This is clear from EN 10.9. In the lines immediately following [8], Aristotle dismisses the power of arguments, on their own, to lead people to virtue. Arguments have force only with “free” youth whose character is noble and who are truly lovers of what is kalon. The many, by contrast, obey fear rather than shame, and refrain from base deeds not b ecause these deeds are shameful but on account of concomitant punishments. So Aristotle turns to the role of law as a compelling force, not just in keeping the many on track but also in cultivating love of what is kalon among free youth. But he recognizes that good laws are not sufficient, and that individual citizens must cultivate such love in their own children. Here Aristotle compares the strength that laws and customs have in cities to the even greater strength that paternal arguments and characters have in households on account of kinship and the father’s good deeds. Because the children already have affection for their father, they are disposed by nature to obey him [1180b3–7]. The idea seems to be that children, because they have affection for their f athers, w ill obey their f athers’ commands even when they do not fully grasp the reasons b ehind them.30 They will trust that the f ather has a point even though he is in no position to prove it. The truth of these claims depends of course on paternal relationships being at least “for the most part” in accordance with nature, with f athers
S ee the R ight T hing
279
ruling their children in a kingly (rather than tyrannical) way. But [5](b) gives us the methodological license to focus on the normative case. Children can of course resist their parents’ commands. Some w ill disobey; o thers will obey, e ither grudgingly or, when they love their parents, gladly. In the ideal case, where good parents prescribe what is decent, children w ill come eventually to understand and accept the (nondemonstrable) reasons for their parents’ commands, with the result they end up being able to work t hings out for themselves. They are then prepared to deal with the sort of “hard” cases they are likely to confront when they no longer have the benefit of sound parental advice. But when they themselves reach that point, they w ill be no more able than their parents were to give the sort of proofs that are rightly expected of mathematicians but wrongly demanded of parents. They w ill “have” the reason of their parents not in the sense that they have mastered the contents of their parents’ practical thoughts but in the sense that they w ill have the capacity to arrive at similar thoughts on their own. There is of course room in a proper upbringing for imparting judicious “for the most part” principles, and that is something McDowell’s Aristotle can surely allow. But given that such principles admit of exceptions whose precise contours cannot be specified in advance, they are no substitute for the sort of judgment expected of grown-ups. It also takes time—because it takes experience—to comprehend the “undemonstrated” sayings of parental figures. The crucial point here is that the experience in question is diff erent in kind from that which explains why it takes longer to become phusikos to become mathematikos. It is not just that what we learn from parents is not fully codifiable: even where it admits of approximate codification, a special kind of experience may be required before we can begin to appreciate its contents. Our pathê are involved not just in the sense that we must have affection for t hose from whom we learn, so that we trust long enough to “try on” advice whose merits we do not yet appreciate. Our pathê are themselves part of the curriculum. We cannot learn from parental figures without having experienced at least some import ant pathê and without having grasped the roles played by such pathê in the lives of our fellow creatures. And the less experience we ourselves have, the more we need to rely on the experience of others, especially those whose powers of discernment we trust.
280
H ist o rical P recedents
When searching for a real-life example to illustrate this point, I settled on an occasion when I, in my forties, was visiting my husband in the hospital. I was by then sufficiently experienced in our “form of life” to appreciate the kind of value patients place on privacy when they are trapped in vulnerable positions, and I felt strongly that I should not be looking at the elderly and speechless man whose face was peering through the gap in the curtain that separated him from us. Surely there is very strong “for the most part” truth to the effect that I o ught not look. But he kept staring at me, and I could not help glancing back now and then, trying to read his gaze. I was not creeped out in the way I tend to be when strangers stare at me. I was unnerved but did not feel his gaze inappropriate. I felt, increasingly, that he wanted something and I thought about calling the nurse. But that didn’t seem quite right: if he wanted something, he could surely push the call button. I realized eventually that the nurse had simply thrown his food down and run off, and that he might want help eating. Experience surely helped here: I had been hospitalized in my thirties, with an elderly and speechless roommate, and was shocked by the way the nurses treated her by comparison with me. Even so, I was reluctant to approach this man with an offer of help. And my husband—relying no doubt on solid “for the most part” truths—felt strongly that I should not. But I do not always obey the reason even of my loved ones, so I eventually went over and asked the man if he wanted help eating. He continued to stare without speaking, but gave no sign of resistance, so I cautiously picked up the spoon and started feeding him. Tears then came to his eyes, but since he was eating with gusto I did not shrink away. I felt like I had in the end succeeded in perceiving what was required of me then and t here. Things could have played out otherwise and I might well have ended up feeling chastened. The point is not that my action stemmed from genuine phronêsis: I have screwed up too many times to think that.31 The point is simply that this is what an exercise of phronêsis is like. There would be lots of cognitive content, some in the form of “for the most part truths,” some perhaps in the forms of doubts, e ither about the applicability of t hose truths in the present situation or about what would constitute their application then and t here. It may be that some of these truths are such that the relevant “for the most part” generalizations could be taught by a parental figure who is able to articulate them in ways accessible to someone who is not yet phronimos. But it is not clear that the relevant truths are best acquired in this way.
S ee the R ight T hing
281
A child learns much from observing the behavior of others, especially if, as she matures, she listens to the sorts of (nondemonstrable) reasons her models give for what they are doing. In cases where the model is virtuous and the child admires the model, the child may come to behave in ways similar to t hose she observes. But if what the child imitates is mere behavior—if the child does not have a sense of the prohaireseis that enform the model’s actions and never asks about t hose aspects of the situation or behaviors she does not understand—then she is unlikely to end up with a stable disposition to act in ways relevantly “like” t hose in which the virtuous model acts. Her praxeis, in the special sense in which praxis is associated with prohairesis, w ill differ in kind from those performed by the model. And this w ill be true, no m atter how superficially similar the external behaviors are, as long as her prohaireseis differ in form from those of her model. What the child needs to learn is to see things for herself in something like the way the decent model sees t hings. And affection for the model does not suffice. The child must experience pathê akin to those the model experiences, thus learning to deal with her own pathê. She must also achieve some appreciation of the various pathê experienced by others, so that she can deal appropriately with them. There are many real-life phenomena I could cite as evidence for the inseparability of the cognitive from the orectic. For example, it seems clear that our concerns play a role in explaining what we notice about the par ticu lar situations in which we find ourselves and that our concerns can thus facilitate or impede proper appreciation of these situations. This is especially import ant when it comes to our ability to understand one another. People previously indifferent to or even biased against members of a particu lar demographic often have their eyes opened by friendship with an individual who belongs to that demographic, especially if they refuse to engage in the sort of “exceptionalism” that allows them to maintain their prejudice or indifference. And trusting what another says about experiences of a kind one has not oneself had—perhaps even could not oneself have had—is arguably an exercise of sunesis. The sort of trust of which I speak is akin to the epistemic deference that children should in the normative case have for parental figures, but it allows for deference to flow in other directions as well. Many patriarchically raised fathers, for example, testify that they have come to appreciate the talents—and the struggles—of women through coming to see
282
H ist o rical P recedents
the world, to some extent at least, through the eyes of their daughters. More such deference to the experience of others, regardless of age, formal education, and other such “hierarchigenic” factors would clearly make the world a better place. But I mention such deference and trust primarily to render more familiar the difficulty of separating the cognitive from the motivational aspects of such phenomena. In sum, it is sufficiently difficult to separate the cognitive from the orectic that we cannot take it for granted that the burden of proof—should proof be in the ballpark—lies with those who reject their separability. And any attempt to impose this burden by citing examples of akratic (or even enkratic) agents simply begs the question in the hylomorphic context. For psychic elements and processes that are distinct in such agents need not be distinct in the phronimos.
9. Conclusion Silencing is largely a function of the role played by desire—a nd certain absences thereof—in the psychic economy of the phronimos. Absences, especially of desires tied to the sorts of pleasure and pain that threaten an agent’s grasp of the archai, play a crucial (but often neglected) epistemic role. But the roles played by desire and the relevant absences thereof are not supplementary to the roles played by cognition. The phronimos correctly sees what is required by the particular circumstances in which she finds herself and is thereby motivated to act accordingly. Not only is no auxiliary desire required to move her. She suffers no opposed desires against which she might need to struggle, most importantly no desires of the sort that might distort cognition in the ways to which [7] alludes. Should she find herself in circumstances that require her to forego things she ordinarily desires and legitimately pursues—things like honor and physical safety—the considerations that normally count in favor of pursuing such t hings are, then and t here, “silenced”: they count, then and there, for nothing. So she is not even tempted. We can see why Aristotle might embrace this view if we return to his distinction between deliberative and perceptive phantasia. Here, however, we must keep in mind a point made in his initial account of phantasia: namely, that it is possible for an object about which a subject has true
S ee the R ight T hing
283
beliefs to appear to that subject to be otherwise than she believes it to be. For example, the sun may appear to a subject to be a foot across even though that subject is persuaded (that is, believes) that it is larger than the inhabited world [428b2–4]. Similarly, one good (for example, a pleasure) may appear to a subject to be greater than another even though that subject believes—perhaps as result of reasoning, perhaps as a result of deferring to some authority—that it is in fact smaller. In other words, the valence of an object’s presentation in perceptual phantasia may differ from the valence of its presentation in deliberative phantasia. There is a thus a distinction in the experience of a rational animal between seeing a situation as requiring a certain sort of action and having to calculate the reasons for and against acting in the relevant way. McDowell’s Aristotle can allow that a considerable degree of reasoning, not necessarily deductive, may be involved in coming to see a situation as the virtuous person sees it. But once the virtuous agent sees her situation in the relevant way, her perception tends to be firm in a way that of the akratês is not: it is not “clouded, or unfocused, by a desire to do otherw ise.”32 So the virtuous agent w ill—g iven normative unity of perception (or imagination) and desire—act accordingly. I am not sure what to say about the distinction between the akratês and the enkratês, who are anomalous in the animal world. Aristotle’s talk, at EN 1150b21ff, of the way in which the enkratês is able to prepare himself to deal with temptation suggests that the enkratês may resemble someone who has had to reason his way to the conclusion that the sun is larger than the inhabited world, and keeps the arguments in mind in ways such that he is able to employ them whenever the appearances tempt him to think the sun smaller. This ability may be what distinguishes him from the akratês, whose grip on the reasoning may be less secure. Or the akratês may simply “say the words” b ecause she has accepted the conclusion on someone else’s say-so. Maybe she undergoes constant changes in gestalt: sometimes she achieves the perspective from which it appears (adjusting for distance) to be larger, sometimes the adjustment for distance is just too hard, so her position is unstable, not unlike that of someone who tries to hold the duck image in place while the rabbit image constantly crops up. The virtuous agent, however, is like someone who simply sees the sun, when she sees it, as larger than the inhabited world. It may “look” like t hings that are actually smaller, but she has no temptation to believe it is.
284
H ist o rical P recedents
I’d like to conclude by calling attention to the special challenge posed by appetitive desires, which tend not to be reason-responsive. My hypothesis is that the perceptual model appeals to Aristotle precisely because appetitive desires are so recalcitrant to argument. He may think it is especially with respect to their objects that veridical perception is required, with the phronimos perceiving their characteristic objects as good when and only when they are in fact good. This hypothesis is meant not to replace but rather to complement McDowell’s defense of the perceptual model by appeal to uncodifiability. It simply points to another aspect of the general reading for which I have sought to provide scholarly support.
Chapter Ten ✣
Self-Consciousness and the Idea of Bildung Hegel’s Radicalization of Kant ✣
Andrea Kern
1 In Mind and World, John McDowell characterizes Kant’s conception of the relationship between sensibility and understanding in a way that accords with Hegel’s reading of Kant, especially as developed in the early text “Faith and Knowledge.”1 McDowell agrees with Hegel on two points: (1) Kant achieves a genuine insight in viewing the faculties of sensibility and understanding as mutually dependent on one another, and (2) Kant is not in a position fully to comprehend his own insight into the relationship between these two faculties. It is, I think, due to McDowell’s work that contemporary philosophy has come to appreciate that our conception of how sensibility and understanding are related to
285
286
H ist o rical P recedents
one another is not just one of many topics in philosophy that one might come to reflect on, but of ultimate significance for any conception of the nature of all human activity, in that it pervades everything we come to think about this activity. This is not to say that he was the first to make this point. On the contrary, this point, as I w ill argue, is Hegel’s deepest concern. But it is only through McDowell’s philosophical work that the philosophical pervasiveness of this Hegelian point has been made available to us. In what follows, I w ill argue that McDowell does, however, depart from Hegel’s understanding of Kant in a decisive respect and that this departure has significant consequences for his view of the role that upbringing and education—Bildung—plays in our understanding of the unity of sensibility and understanding. McDowell believes that we can only understand the idea of a “real connection” between understanding and sensibility if we equip ourselves with a robust concept of second nature.2 Kant’s shortcoming, on McDowell’s diagnosis, is that he lacks such a concept of second nature. Hegel does not share this diagnosis. This is b ecause he has a different understanding of the problem that confronts the Kantian exposition of the unity of sensibility and understanding. Hegel’s solution to this problem, as I w ill show, leads him to a concept of Bildung as an “immanent moment of the absolute.”3 As an “immanent moment of the absolute,” Bildung’s role cannot be to render the “real connection” between understanding and sensibility comprehensible. Rather, the task Hegel ascribes to the idea of Bildung can be characterized only once we have entitled ourselves to a point of view from which the connection between sensibility and understanding is conceived in a manner that has no room for the question of how this connection comes to be “real.” My argument w ill proceed in three parts. First, I w ill sketch a reading of Kant on which his aim is to grasp the connection between sensibility and understanding in a way that undercuts the very question that McDowell believes we need a robust concept of second nature in order to answer. Second, I w ill explain why the Kantian account of the unity of sensibility and understanding is inadequate, in Hegel’s eyes, and what consequences Hegel draws from this. Third, I w ill demonstrate what it means to understand Bildung as an “immanent moment of the absolute.”
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 287
2 Hegel and McDowell are united in thinking that Kant’s task in the Transcendental Deduction is to entitle himself to the claim that a subject who possesses represent ations that refer to objects must possess two capacities, each dependent on the other. She must possess a faculty of sensibility, through which she can receive represent ations of objects by being affected by t hose objects. And she must possess a faculty of understanding, through which she can form conceptual representations of objects that may be combined to form judgments. Kant’s central insight, according to both Hegel and McDowell, is that the contributions these faculties make to the formation of object-related representations are only comprehensible if we understand them as mutually dependent on one another. Hegel characterizes this mutual dependence as one in which “the Kantian forms of intuition and the forms of thought do not lie separated from one another as particular, isolated faculties, as is commonly i magined. One and the same synthetic unity . . . is the principle of intuiting and of the understanding; the understanding is only the higher potency [Potenz], in which the identity, which is totally immersed in manifoldness in intuiting, simultaneously sets itself against [the manifoldness] and constitutes itself in itself as universality, which is what makes it the higher potency.” 4 Thus, Hegel understands Kant’s thesis of mutual dependence between the faculties to claim that the acts of both faculties are determined by the same principle—the principle of a “synthetic unity” of representations. According to Hegel’s interpretation of the dependence thesis, we must understand judging and sensible intuiting as capacities that, though dif ferent, are not based on two separate principles that determine their acts but are rather two diff erent manners of realizing one and the same principle. Hegel calls the understanding the “higher potency” of sensibility. If we view both faculties with an eye to their underlying principle, we cannot distinguish one from the other. What distinguishes the faculties from one another only comes into view when we consider the manner in which each respectively realizes this (single) principle. While sensibility realizes the principle of “synthetic unity” in a way that is bound to the given object in which it “immerses” itself, the understanding, by contrast, realizes the principle independent of a given object.5 Understanding is not dependent
288
H ist o rical P recedents
on anything but the capacity to realize this principle. That explains why Kant identifies the understanding with the capacity to combine represen tations according to this principle.6 The understanding is the “higher potency” in that its realization of this principle is not dependent on anything other than itself. McDowell takes a position strikingly similar to Hegel’s when he writes that the Kantian idea according to which empirical knowledge results from the cooperation of sensibility and understanding only makes sense “if we can achieve a firm grip on this thought: receptivity does not make an even notionally separable contribution to the co-operation. . . . We should understand what Kant calls ‘intuition’—experiential intake—not as a bare getting of an extra-conceptual Given, but as a kind of occurrence or state that already has conceptual content.”7 Kant’s insight, according to McDowell, is that we must reject the idea of an even notional separability of understanding and sensibility. To reject this idea is to see that any activity that results from the sensibility of a creature that is in possession of the power of understanding, must, as such, actualize the very conceptual unity in terms of which the understanding, qua capacity for judgment, is defined. However, McDowell thinks that Kant does not have the means necessary to comprehend his own insight that is at the center of the Transcendental Deduction: namely, that sensible acts as such, that is, qua manifestations of sensibility, realize exactly the same principle that defines the understanding. McDowell’s diagnosis of Kant’s failure has two steps. First, he credits Kant with the insight that the idea of a synthetic unity of a manifold of representations that purport to refer to objects requires us to assume not only a “notional connection” between intuitions and concepts but also a “real connection.”8 One example of a position that grants a “notional” connection but not a “real” one is Gareth Evans’s view. Evans claims that certain sensible acts are dependent on the understanding—namely, those acts whose role it is to provide experiences that are relevant for the content of judgments. He denies, however, that sensible acts as such are dependent upon the understanding.9 Evans wants to say that a sensible act can count as an “experience” of a sort that is relevant to the content of judgments only if one is able to make a judgment about this sensible act, that is, only if one is able to make a judgment whose content is the presence of this sensible act. The content of the sensible act
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 289
itself, however, should be of a kind that does not presuppose the presence of a faculty of judgment.10 According to McDowell, Kant saw clearly that a merely notional connection between sensibility and understanding is insufficient to make intelligible the idea of a synthetic unity of representations that purport to refer to an object. Instead, we must understand the faculty of judgment as a faculty whose principle of synthetic unity of representations is involved in determining the content of sensible acts themselves. McDowell’s second step is to claim that Kant ultimately cannot make sense of the idea of a “real connection” between understanding and sensibility because he has no robust concept of second nature at his disposal, which would allow him to understand this connection as part of the nature of a particu lar kind of being. This avenue is closed to Kant because he identifies nature with that realm whose laws are the object of empirical knowledge. In order to make good on his insight, he must seek another way out. Kant seeks this expedient, according to McDowell, in the idea of a supersensible realm that affects sensibility through the understanding in a transcendental manner. Yet this idea does not meet Kant’s own standards of intelligibility, according to McDowell: Kant is peculiarly brilliant here. Even though he has no intelligible way to deal with it, he manages to hold on to the insight that a merely notional connection of concepts with intuitions w ill not do. That forces him into a way out that is unintelligible by his own lights. The real connection has to be that spontaneity is involved in the transcendental affection of receptivity by the supersensible. And now the good thought that our sensibility opens us to a reality that is not external to the conceptual can show up ere only in a distorted form, as if the ordinary empirical world w constituted by appearances of a reality beyond.11 ecause Kant lacks a robust concept of second nature, according to McB Dowell, he is compelled to conceive of the real connection between understanding and sensibility in a way that scuttles the very notion of sensibility needed for his own insight into sensibility’s real connection with the understanding. Instead of conceiving how sensibility can be a faculty of empirical knowledge, he demotes the empirical world to a world of mere appearances b ehind which an inaccessible “true reality” is located.
290
H ist o rical P recedents
3 While Hegel does not share this diagnosis, he, too, thinks that Kant characterizes the unity of understanding and sensibility in a way that undermines his own enterprise of (among other t hings) understanding how representations are possible that refer to objects, which are as they are independently of how they are represented. Hegel’s diagnosis of Kant’s failure runs in another direction. According to Hegel, Kant’s undoing results from his assumption that a subject’s faculty of judgment can only be a unity consisting of a priori elements if it is a “pure unity”—pure in the sense that it is a unity that is “not originally synthetic.”12 Kant supposes, so Hegel’s diagnosis runs, that a subject’s faculty of judgment can consist of a priori elements only if these form a unity that, as such, excludes all elements of sensibility. This conception of a subject’s faculty of judgment means, according to Hegel, that Kant must characterize the understanding as a merely subjective human capacity for which “the manifold of sensibility, empirical consciousness as intuition and sensation, is in itself something uncombined, the world something disintegrating, which only gets objective coherence and stability, substantiality, multiplicity, even actuality and possibility, through the good offices of h uman self-consciousness and intellect—an objective determinateness that is man’s own perspective and projection.”13 According to Hegel, rendering the faculty of judgment subjective is the unavoidable flip side of Kant’s assumption that the idea of a faculty of judgment consisting of a priori elements rules out that this capacity, qua capacity, might contain sensible elements. It follows that Kant cannot avoid downgrading the faculty of a judgment to a mere subjective human characteristic that confronts the empirical world as something extrinsic to it.
4 In what follows, I w ill distinguish two interpretations of the thesis that Kant lacks the means to comprehend the real connection between sensible intuiting and judging. According to the first interpretation, the thesis implies that Kant is not in a position to grasp how the faculty of judgment
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 291
can be a unity that depends upon a faculty of sensibility. That is McDowell’s reading. Thus, McDowell thinks it sufficient to arrive at a position that understands the faculties to be unified in this sense. The relevant idea of dependance, according to this position, is faculty-dependance. On the second interpretation, by contrast, the thesis implies that Kant is not in a position to grasp how the faculty of judgment can be a unity that depends upon an act of sensibility. That is Hegel’s reading. According to Hegel, the relevant idea of dependance is act-dependance. According to the former position, the unity of a faculty of judgment is something that is conceived to be dependent upon a faculty of sensibility for being an actual unity. This implies that it is possible to think the unity of a faculty of judging without thinking an a ctual unity. According to the second position, this is incoherent. Hegel w ill argue that the thinking of the kind of unity that a faculty of judging is, is nothing other than the thinking of an actual unity: the a ctual unity that the thinker is. Thus the distinction between faculty- dependence and act-dependence, I w ill argue, turns out to be crucial for how we conceive of the kind of unity that the faculty of judgment is. Since McDowell does not distinguish these two readings from one another, as I w ill suggest, he cannot fully see the point of Hegel’s critique of Kant. So I w ill first reconstruct the Kantian position insofar as it is the target and basis of Hegel’s critique. Then I w ill highlight the consequences of this critique for Hegel’s conception of self-consciousness and the role of Bildung. Kant opens the division of the first Critique called “Transcendental Logic” by claiming that he is about to undertake a task that has hardly ever been attempted—namely, an “analysis” (or, more literally, “dissection”: “Zergliederung”), not of the concepts of understanding, but of “the faculty of understanding itself.”14 In light of this characterization of his enterprise, we can describe the result of the Transcendental Deduction of the concepts of the understanding as follows: the idea of the understanding is the idea of a faculty for judgment that a sensible being can possess only if she is equipped with certain “a priori concepts,” the so-called categories. As we saw above, Hegel accuses Kant of having a false conception of the idea of the a priori. Indeed, much of our understanding of Hegel’s critique of the Kantian position depends on exactly what idea of the a priori he attributes to Kant. According to Hegel, the task of the Transcendental Deduction consists, among other things, in more closely specifying the
292
H ist o rical P recedents
relevant idea of a priori concepts that are the object of the deduction. Kant first attains his conception of the a priori, in Hegel’s view, with the conclusion of the deduction—that is, once we have understood how such a priori concepts are possible. For it is the “possibility” of such concepts that we were unable to understand prior to the deduction.15 We enter the deduction with a preliminary understanding of a list of so-called a priori concepts whose usage is conceived to be universal and necessary on account of the particular role they play in our thought. Their role is, namely, to serve as “rules of thinking” that the understanding must follow in order to combine a manifold of representations so as to produce (1) the unity of a judgment or (2) the unity of an intuition whose object it (the understanding) is able to think.16 It follows from the role of t hese concepts that they cannot be acquired from experience with objects. For they are understood to be concepts that are constitutive of the ability to enjoy experiences whose objects can be thought. It is with this preliminary understanding of a priori concepts that we plunge into the Transcendental Deduction, which is now tasked with allaying a worry: namely, the worry that these “subjective conditions of thinking” might be merely subjective, that is, that they might not be valid of objects as they are given to us in experience.17 In order to alleviate this worry, Kant shows that these concepts, through which we think objects, are not merely subjective conditions of thinking but rather “conditions of the possibility of experience in general [überhaupt] and are therefore also valid [gelten] a priori of all objects of experience.”18 In light of this outcome, Kant can then conclude that a subject that possesses a faculty of judgment is eo ipso in possession of a priori knowledge of the objects of experience. For possessing a faculty for judgment means possessing concepts whose use in judging itself involves an awareness that objects of experience are only possible, as such, through these concepts. It means possessing concepts whose deployment in judging involves a consciousness that objects of experience are in necessary agreement with t hese concepts. The deployment of these concepts in judging therefore provides knowledge of objects of experience—a nd, indeed, knowledge that is a priori in the following sense: namely, as knowledge that is contained in e very knowing of objects of experience as that which makes the latter possible. The idea of a capacity for judgment is therefore nothing other than the idea of a ca-
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 293
pacity, the very possession of which entails that one has, and is conscious of having, the capacity to know objects of experience. This reading departs in a decisive respect from a widely held understanding of Kant. According to the latter, the work of the Transcendental Deduction consists essentially in providing an a priori justification for the use of concepts whose a priori status as subjective conditions of thought we recognized before this deduction. The idea is that this a priori justification elevates these subjective elements to the status of a priori knowledge. Hegel cannot share this reading b ecause of the par ticular way he thinks Kant understands the unity of sensibility and understanding. According to Hegel, the task of the deduction cannot consist in providing an a priori justification for the use of concepts whose status as subjective conditions for thinking we have recognized independently of and prior to this justification. For this would imply that the understanding, qua faculty, is logically independent of sensibility. In order to think that it is possible to recognize certain concepts as necessary for the activity of the understanding while holding open the question whether these concepts are valid of objects of experience, one must think that it is possible to specify the activity of the understanding independently of the idea of objects of experience. According to Hegel’s reading of Kant, this is what Kant denies. Therefore the task of the Transcendental Deduction cannot be to provide a justification for the employment of concepts whose necessity for thought has been established prior to this deduction. Hegel understands the task Kant sets for himself in the deduction otherw ise, namely, as the task of making it comprehensible how t here can be a priori concepts in the sense of subjective conditions of thinking at all. In Hegel’s view, the deduction is about understanding how a priori concepts—in the sense of concepts that we are conscious of as subjective conditions of thinking—are possible in the first place. As Hegel sees it, Kant aspires to show that the viewpoint we must take up in order to appreciate the possibility of a priori concepts is a viewpoint from which the question that troubled us at the outset of the Transcendental Deduction can no longer be formulated without having already been answered. On Hegel’s reading, it is one and the same t hing to understand what it means that there are certain a priori concepts that are
294
H ist o rical P recedents
“subjective conditions of thinking” and to be justified in using t hese concepts in judgments that lay claim to objectivity. According to this reading, it is part of the point of the Kantian argument that we are precluded from thinking that certain concepts that we are conscious of as subjective conditions of thinking are merely subjective. Rather, the idea that these concepts could be merely subjective is, according to this reading, irreconcilable with the idea that they are “rules of thinking.” Kant says this quite explicitly at the beginning of the deduction. To represent these concepts as merely subjective would mean to represent them as concepts that are “entirely empty, nugatory, and without meaning.”19 Kant’s thought h ere is not that when we represent t hese concepts as merely subjective, we represent to ourselves a particular kind of concepts, that is, t hose that are “entirely empty, nugatory, and without meaning.” Instead, he is saying that, thus regarded, these concepts would be nothing more than “mere phantom[s] of the brain.”20 That is, they would be something that only appears to be an a priori concept, a “subjective condition of thinking,” though in truth it is no such t hing.21 On this reading, it is a condition of the intelligibility of an a priori concept of judgment that it is represented as a concept that is in necessary agreement with objects of experience and eo ipso has the status of knowledge. The worry at issue in the deduction therefore cannot be that a being might possess a priori concepts and be conscious of them as rules for thinking yet without being justified in employing them in judgments that make knowledge claims about objects of experience. The worry of the deduction is rather that it seems possible that those concepts which we believe we recognize as “subjective conditions of thinking” are not such conditions as all, but rather “mere phantom[s] of the brain.”
5 Kant’s analysis of the faculty of judgment culminates in the thought that judgment, qua faculty, comprises a priori knowledge of objects of experience and that the subject is conscious of such knowledge a priori. Having a capacity for judgment means having a consciousness of represen tations of objects of experience that is contained in all one’s empirical cognitions (knowings) as that which makes empirical knowledge possible.
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 295
The fundamental characterization of the understanding, then, is that it is a faculty for knowing objects of experience that is a priori conscious of itself. It follows from this that the understanding can never be merely an object of consciousness b ecause there is no faculty of understanding inde pendently of this consciousness (of itself). It is on the basis of this structure that Kant characterizes the consciousness in which the faculty of understanding consists as self- consciousness, and identifies the idea of the understanding with the idea of self-consciousness.22 Kant calls this act of self-consciousness, in which the understanding constitutes itself as the capacity for knowing objects of experience, an act of “original apperception,” in order to make clear that this act does not consist in the exercise of a faculty of self-consciousness but rather constitutes the capacity for self-conscious acts, regardless of their content. This self-consciousness is therefore inscribed in all acts of the understanding, no matter what content they have. It is, as Kant says, “in all consciousness one and the same.”23 With this “original” self-consciousness, Kant purports in §§14–15 to have identified the fundamental unity on which all acts of judging by means of the categories as well as the categories themselves depend. It is the consciousness that a faculty for knowing objects of experience has of itself—one that is thus present “in all consciousness” in which this faculty is exercised as “one and the same” and which therefore “accompanies,” as he says, all exercises of this faculty.24 According to Hegel’s and McDowell’s reading, this is Kant’s deepest insight into the nature of the understanding whose self-consciousness he proposes to articulate. It is the insight that the understanding, and hence the capacity whose self-consciousness he seeks to articulate, and the capacity for being sensibly affected by objects on which its knowledge of objects depends are not two self-standing faculties, each of which makes its contribution to cognition independently of the other, but two interdependent elements of an “original synthetic unity” that is, as such, conscious of itself. However, whereas McDowell thinks that Kant’s failure is that he cannot explain how this “original synthetic unity” could ever become something actual because he is not in a position to exploit the notion of Bildung for this purpose, Hegel, by contrast, thinks that Kant’s failure goes deeper: that Kant fails because he has a wrong conception of this “original synthetic unity.” Or so I w ill argue.
296
H ist o rical P recedents
6 McDowell holds that our concern must be to entitle ourselves to a position he calls “relaxed naturalism.”25 According to such a “relaxed naturalism,” the role and meaning of the concepts of upbringing and Bildung consist in describing a transformative process in which a sensible being, who is at first only potentially linked to the understanding, is transformed into a being whose sensibility acquires the character of a self-conscious cognitive capacity.26 It is the role of upbringing to realize a unity between two faculties that is initially only potential and not yet a ctual. This thought runs through McDowell’s works as a kind of leitmotif. It is foregrounded at various places, as, for example, in the following: Our nature is largely second nature, and our second nature is the way it is not just b ecause of the potentialities we w ere born with, but also b ecause of upbringing, our Bildung.27 Human individuality is not just biological, not exhausted by the singleness of a particu lar human animal. A fully fledged human individual is a free agent. . . . Freedom is responsiveness to reasons. It is not a natural endowment, not something we are born with.28 The idea of participation in a communal form of life is needed for a satisfactory understanding of responsiveness to reasons. . . . Responsiveness to reasons . . . marks out a fully-fledged human individual as no longer a merely biological particular, but a being of a metaphysically new kind.29 The child, so runs the idea, is only potentially a rational animal and not yet so in reality. The child’s doings and being first attain this sort of rational realization as the child is brought up, or educated. Education, Bildung, is, according to McDowell, a process in which an individual who is, at first, a merely sensible being undergoes a metaphysical transformation. This means that the human individual, according to this view, is born a merely animal creature whose governing principle is sensibility, but becomes a specifically rational creature whose governing principle is understanding in virtue of Bildung. In this sense a human individual acquires a new,
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 297
essentially diff erent capacity for sensibility in acquiring Bildung, namely a capacity for sensibility that is then dependent on understanding. Bildung, according to this view, is a process through which an individual of a metaphysically new kind comes into being. On this conception, a child who has yet to be ‘brought up’ has a different sort of actuality or reality from that manifested in the activity and existence of an individual who has gone through a process of Bildung. As we w ill see in a moment, Hegel disagrees with this idea. According to Hegel, the role of Bildung does not consist in transforming the potentiality of a metaphysical kind that characterizes the child into a reality. An adult human being does not differ from a child by belonging to a “metaphysically new kind.”30 Hegel rather understands Bildung as an “immanent moment of the Absolute.”31 This understanding rules out the possibility that the reality that results from Bildung might fundamentally differ from the one from which it emerges. Hegel arrives at this understanding of Bildung by recognizing that the standpoint we achieve when we possess Kantian self-consciousness cannot be that of “absolute knowledge.”32 The fundamental act of Kantian self- consciousness, as we saw above, consists in the consciousness that a capacity for knowing objects of experience has of itself. As such, it contains a consciousness of the unity of its capacity for judgment with its capacity for sensible intuitions, both of which it understands as aspects of that capacity of which it is conscious. McDowell thinks this is enough to entitle oneself to a position that can affirm the unity of both capacities in this sense. In contrast, Hegel’s critique of Kant rests on his insight into the inadequacy of an understanding of self-consciousness as a mere faculty- consciousness of the sort described. In Hegel’s view, Kantian self-consciousness gets entangled in a “contradiction.”33 This contradiction consists—from the perspective of the Phenomenology of Spirit—in the inability of such self-consciousness to comprehend its own existence. Kantian self-consciousness is unable to be “certain of itself in its existence.”34 A self-consciousness, however, whose mere constitution does not yield certainty of its own existence fails to grasp its own concept. For it has a represent ation of self-consciousness that is not identical with the reality of what it represents. Hegel calls a self- consciousness that does, by contrast, understand its own concept “the self-k nowing spirit.”35 The “self-k nowing spirit” is a self-consciousness
298
H ist o rical P recedents
that has a conception of itself through which it is, qua being what it thinks it is, certain of its own existence. Kantian self-consciousness is not in a position to achieve this certainty because it conceives self-consciousness as the consciousness of a unity whose elements—the faculties of understanding and sensibility—provide no answer to the question of how t here can be an act of such self- consciousness, that is, an act that manifests this unity. This is b ecause the unity, consciousness of which constitutes Kantian self-consciousness, is a unity of mere faculties. As mere faculty-consciousness, self-consciousness cannot explain its own reality through the elements of the unity whose consciousness it is supposed to be. For as mere faculty-consciousness it does not represent a unity that contains a reality. That is to say, it represents something whose reality self-consciousness qua self-consciousness cannot grasp. In “Faith and Knowledge,” Hegel expresses this objection by saying that self-consciousness h ere has been reduced to an “absolute intellectual point,” a “fixed intellectual One.”36 The inner contradiction in which this conception of self-consciousness finds itself entangled, according to Hegel, is as follows: Self-consciousness that contains no consciousness of its reality cannot be identical to that reality of which it is conscious. Yet these have to be identical in order for it to be self-consciousness. A self-consciousness that does not contain consciousness of its reality is constrained only ever to understand itself as the condition of certain acts—namely, acts that are exercises of the very faculty, in the consciousness of which self-consciousness consists—but not as something that is identical to the reality of particu lar acts. Hegel’s thinking through of the Kantian position consists in resolving this contradiction. Self-consciousness, Hegel concludes, must grasp its “realization” “through itself.”37 Hegel’s question therefore runs as follows: How must self-consciousness be constituted such that it can grasp its own reality?
7 We can understand the progression of the argument that Hegel unfolds in the Phenomenology of Spirit as a step-by-step attempt to free the concept of self-consciousness from those misunderstandings that prevent us from
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 299
fully comprehending it—including the Kantian misunderstanding, among others. According to Hegel, the failure of Kantian self-consciousness to explain, through itself, the consciousness in which it consists shows two things. First, it shows that the faculty-consciousness Kant describes cannot itself be fundamental, but must be derivative of fundamental self- consciousness. Second, it shows that the assumption on which the associated conception of self-consciousness rests must be erroneous. This erroneous assumption, which Hegel identifies as a presupposition of this conception of self-consciousness, is familiar from our previous discussion. It consists in the idea that the understanding can be a unity that contains a priori elements only if, qua unity, it does not contain acts of sensibility. This is the error Hegel attributes to Kant in “Faith and Knowledge.” And the difficulties confronting Kant’s position make it clear that and why this must be an error. To f ree oneself from this Kantian misunderstanding, according to Hegel, requires one to develop an understanding of self-consciousness that links consciousness of a priori elements to consciousness of a sensible act. Hegel’s aim in the Phenomenology of the Spirit, beginning in chapter IV, is to develop such a conception of self-consciousness. For our purposes, I can limit myself to two essential steps of his argument, which suffice for an understanding of how he grasps the relation between the idea of self- consciousness and the idea of Bildung. Hegel’s first step is to characterize self-consciousness as the consciousness of a unity of elements that has the character of a “movement.”38 Hegel describes this “movement” as follows: self-consciousness consists in an act of consciousness that is directed at an object in the sensible world, an act that is, as such, conscious of itself as an act of a faculty whose exercise is enabled through the sensible object that is the content of this faculty and which Hegel describes as a faculty that contains perception and understanding, that is, a faculty for perceptual knowledge. In performing this “movement,” self-consciousness contains, as such, the consciousness of a faculty as well as the consciousness of a sensible act. It is the unity of these two forms of consciousness, each of which depends on the other. Hegel writes: “The being of what is meant, the singularity and the opposed universality of perception, as well as the empty inner of understanding, t hese are no longer essences
300
H ist o rical P recedents
[Wesen] but are moments of self-consciousness, that is, as abstractions or distinctions that are at the same time nugatory for consciousness.”39 If we understand self-consciousness as a unity with this structure, we begin to grasp self-consciousness as something that “realizes [itself] through itself.” For it then consists in a unity of elements that, as such, contain a consciousness of the reality of that unity in which self- consciousness consists. Self-consciousness contains, as one of its “moments,” an act of consciousness through which it directs itself to the sensible world insofar as it is conscious of this act as a realization of its faculty for perceptual knowledge. Hegel infers from this that, for a self-consciousness that comprehends itself in this way, the object to which it refers qua consciousness has “become life.” 40 What Hegel wants to say with this is that we thereby think of consciousness as an activity that relates to its object in a way that exemplifies the structure of life as such. The idea of life, as Hegel employs it here, is the idea of a distinctive kind of unity whose parts are what they are in being “for” each other.41 It is the idea of a unity whose parts are understood through concepts that one cannot understand independently of one another but only through the unity in which they figure. Hegel explicitly notes that the concept of life as he employs it here remains completely abstract and minimal. At this juncture, it means nothing more than that the object to which self-consciousness refers qua consciousness, is, in its character, “reflected being.” 42 In other words, the object at which this consciousness is directed is not something external to it but forms a “unity” with it.43 In this sense, Hegel says, “the object has become life through this reflection in itself.” 44 Yet this does not yet yield a full understanding of the concept of self- consciousness. And the reason it does not, according to Hegel, is because a self-consciousness that comprehends itself in this way conceives of “life” as an object of perception and hence has an object that, qua object of perception, is distinct from the consciousness of this object. Hegel concludes that self-consciousness cannot yet comprehend its own reality, and thus cannot comprehend itself, if it comprehends its reality as an object of perception. For this would mean that “life,” which is its object, is conceived as something from which self-consciousness “divides” itself precisely through the consciousness it has of it.45 A self-consciousness that understands itself as consciousness of life, conceived as an object of
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 301
perception, can only “distinguish [life] from itself as [something] existing” and, hence, can only understand life as something not identical with itself.46 Hegel concludes from this—and this is the second step of his argument—that self-consciousness must consist in a unity that understands life—to which self-consciousness refers qua consciousness—not as an object of perception that it differentiates from itself as existing, but rather as something that, of its own accord, “points to consciousness.” 47 Self-consciousness must comprehend life, of which it is conscious, as an object that is, on its own account, determined as the object of a consciousness. It must comprehend the life of which it is conscious as one that realizes itself in the consciousness that it (self-consciousness) has of it (life). A life that had this structure would be actualized in activities— whatever they might be, let them be dancing and laughing, eating and swimming, walking and writing—whose actuality would have to be explained through the consciousness that their b earers have of the faculties of these activities and the role they play in their life. A self-consciousness that is conscious of a life of this sort—Hegel calls it “this other life, for which the genus [Gattung] as such exists and which is genus for itself”—understands its own reality as a realization of a concept of life whose reality qua reality is self-conscious.48 This means that a life that has this structure cannot be described from a standpoint other than the one that is occupied by those beings whose life it is. For the concepts that one would have to employ to articulate this life—that is, to describe the capacities possessed by its b earers and their characteristic objects— could only refer to what they refer to if that to which they refer were recognized as part of the very concept of life through which one explains oneself, that is, one’s own existence and activities. With this understanding of self-consciousness, Hegel argues, self- consciousness reveals itself as essentially “living self-consciousness.” 49 In its most fundamental use, according to Hegel, the concept of self- consciousness characterizes a form of life not in terms of a particular faculty that individuals belonging to this life-form typically possess. Rather, self-consciousness characterizes a formal aspect of this form of life in that it determines the distinctive manner in which it realizes itself.50 In this sense, the concept of self-consciousness denotes, as Michael Thompson expresses this, the “form” of a life-form.51
302
H ist o rical P recedents
When Hegel says that the fundamental object of self-consciousness is a form of life in that it specifies the distinctive sense in which this life has a form, and that is, a principle of explanation through which the life activities of its bearers are explained, he means to give an “abstract” account of the object of self-consciousness.52 Abstractly conceived, the object of self-consciousness is a form of life whose form it specifies. When Hegel thinks of this as the “abstract object” of self-consciousness, he wants to say that this characterization can be “enriched” by a further “articulation” [Entfaltung] of it. This more articulated understanding of self- consciousness would differ from its abstract understanding in that it would no longer abstract from those elements in the concept of a self- conscious form of life that are contained in it qua the concept of a reality.53 We can disregard for our purposes the individual steps in Hegel’s “articulation” of this initially abstract object of consciousness. It is enough to bring into view the structure of consciousness in which self-consciousness consists, according to Hegel. The central insight Hegel gleans from the Kantian misunderstanding is that the concept of self-consciousness, in its fundamental employment, describes neither an “intellectual point” nor a particular ability of a sensible being, but instead the form of a form of life, whose intrinsic manner of realization it lays out. According to Hegel, there is no self-consciousness that is not consciousness of a form of life, and hence that is not consciousness of something that contains, as such, an explanation of its own reality. It follows from this that we cannot employ the concept of self- consciousness without describing the reality of a consciousness—a consciousness that, qua consciousness, is conscious of its reality. Hegel will say later, a fter he has further unfolded the concept of self-consciousness, understood as the concept of the form of a form of life, that this concept reveals itself to be the concept of an “in itself universal self-consciousness that is a ctual in another consciousness in such a way that the latter has complete independence, or is a t hing for it, and that it is precisely therein conscious of the unity with it, and is, in this unity with this objective being [gegenständlichen Wesen], self-consciousness.”54 Hegel emphasizes two points here. A form of life whose form is self-consciousness is a form of life that contains a “universal self-consciousness.” In other words, it is a form of life that has a consciousness whose subject is not a particular individual (who differentiates herself from o thers through this consciousness), but is
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 303
instead a general subject. If we conceive this “universal self-consciousness” that constitutes such a form of life merely as something general, however, we have not yet understood it as a real consciousness. And in that case we have not yet understood it as self-consciousness either, but instead merely as the “abstraction” thereof.55 A life-form’s universal consciousness is comprehended as self-consciousness only when it contains its own reality qua consciousness. Hegel concludes from this that self-consciousness must consist in the unity of the consciousness of a form of life with the consciousness of an “objective being” that actualizes this unity through its consciousness.56 Self-consciousness is the consciousness of a form of life in virtue of which an individual of this form of life grasps her “doing and being” as an actualization of that form of life which she thereby, through that self- consciousness, actualizes.57 Hegel’s unfolding of the “concept of self-consciousness” thus leads to the thought that the fully unfolded concept of a self-conscious form of life consists in the “life of a people [Volk].”58 According to Hegel, the “life of a people” presents the shape of a self-conscious form of life in which the concept of such a form of life attains its “complete reality”: “It is in fact in the life of a p eople that the concept of self-conscious reason’s realization— beholding [anzuschauen] in the independence of the Other complete unity with him, or having as my object this free thinghood of an Other, which I come across and which is the negative of myself, as my own being-for- myself—has its complete reality.”59 What is decisive h ere for our purposes is how Hegel hits upon the idea of the “life of a people”—namely, by developing the initially abstract concept of a self-conscious form of life to its full expression. The “life of a people” is supposed to be the complete realization of this concept. On the one hand, this means that the concept of the “life of a p eople” is not identical to the concept of a self-conscious form of life. But on the other hand, it means that the concept does not describe something else. Hegel’s claim is that the concept of the “life of a people” characterizes the shape a form of life assumes, when the concept of that life-form realizes itself completely. This means that it represents the shape of a self-conscious life by means of which we understand the concept of such a form of life completely, since we understand the concept in this shape as one that no longer presupposes the reality it describes in any sense, but rather comprehends this reality through itself. The concept of a self-conscious form of life is abstract compared to the
304
H ist o rical P recedents
concept of the “life of a p eople” insofar as it abstracts from all t hose determinations of this concept in virtue of which it is the concept of a reality. What is left then is the concept of a reality that describes the form of that reality. For our purposes, it suffices to consider the abstract concept of a self- conscious form of life in order to understand Hegel’s revision of the Kantian concept of self-consciousness and the consequences this revision has for the concept of Bildung. For if we understand self-consciousness as a concept that specifies the form of a form of life, then we no longer follow Kant in understanding self-consciousness as a consciousness of a unity of faculties that cannot, qua what it is conscious of, account for its own reality. We instead understand it as a consciousness that is identical with the d oing 60 and being of a particular kind of individual. A mere faculty-consciousness cannot account for the reality of this consciousness because that which it is conscious of is a mere potentiality, and hence something other and less than a reality. That is to say, it represents something whose reality self- consciousness qua self-consciousness cannot grasp but can only presuppose. By contrast, a self-consciousness that consists in a distinctive form of life is conscious of an object—namely, the vital activities of the form of life whose form it specifies—that is not only something real—for example, a man walking along the street, or a w oman riding a bike or making a calculation—but something whose reality as a doing of this or that partic ular kind is explained through the very consciousness that its b earers have of their capacity for this or that kind of activity, and in this sense, is explained through nothing other than itself. As we saw above, Hegel reproaches Kant for holding a false conception of the a priori on account of his misguided assumption that the understanding can be a unity containing a priori elements only if it does not contain any sensible acts. The concept of a self-conscious form of life is, in Hegel’s view, the concept of a kind of unity that Kant holds to be impossible: the unity of a sensible consciousness with an a priori consciousness. For a self-conscious life-form is one that realizes itself in the life of individuals who are conscious of their d oing and being as the manifestation of that life-form which they realize. An individual who is conscious of her doing and being as the manifestation of a form of life, precisely insofar as she is thus conscious of her d oing and being, has an a priori consciousness of that life-form whose manifestation she under-
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 305
stands her d oing and being to be. Yet since this a priori consciousness of her life-form just is the manner in which she actualizes her life-form, it follows that we cannot describe the content of this consciousness without invoking acts of sensibility—namely, those sensible acts in the performance of which the life of the relevant individual consists. Self-consciousness is consequently the consciousness of a unity of an a priori consciousness and the consciousness of a sensible act. This life-form-consciousness is a priori in a thoroughly Kantian sense. It is a consciousness that an individual cannot justify through experience, for it is the condition of the reality of the doing and being of an individual who realizes a self-conscious form of life. In this sense, it is both necessary and universal. Yet since this consciousness consists in nothing other than the self-consciousness that an individual has of itself, it is likewise irreducibly bound to the consciousness of that individual.61 Here is the “true a priori,” as Hegel calls it, the conception of the a priori once it has been cleansed of the Kantian misunderstanding.62 In this way, self-consciousness is understood as containing a consciousness of its own reality, since the latter is none other than the reality of the doing and being of an individual who manifests a self-conscious form of life. As a concept that describes the form of a form of life, self-consciousness no longer shrinks from knowledge of its own reality, but is, as Hegel says, a “self-consciousness for a self-consciousness.”63
8 For Hegel, the concepts of upbringing and education—Bildung—have their place in elucidating the realization of the concept of a self-conscious form of life. Hegel describes the role of Bildung in his Philosophy of Right as follows: Education [Bildung], in its absolute determination, is therefore liberation and work towards a higher liberation; it is the absolute point of transition to the infinitely subjective substantiality of ethical life, which is no longer immediate and natural, but spiritual and at the same time raised to the shape of universality. . . . But it is through this work of education that the subjective w ill
306
H ist o rical P recedents
attains objectivity even within itself, that objectivity in which alone it is for its part worthy and capable of being the actuality of the Idea.—Likewise, this form of universality to which particularity has worked its way upwards and cultivated [heraufgebildet] itself, i.e. the form of the understanding, ensures at the same time that particularity becomes the veritable being-for-itself of individuality; and, since it is from particularity that universality receives both the content which fills it and its infinite self- determination, particularity is itself present in ethical life as f ree subjectivity which has infinite being-for-itself. This is the standpoint from which it becomes plain that education is an immanent moment of the absolute, and that it has infinite value.64 In designating Bildung as an “immanent moment of the absolute,” Hegel means to say that the “work of Bildung” is an essential characteristic of a self-conscious form of life. Without Bildung, there is no self-conscious life- form, and vice versa. A self-conscious form of life is one whose individuals become “worthy and able” to realize this life-form in the way that corresponds to its concept—namely, in a self-conscious manner. That is the meaning of the concept of “Bildung.” According to Hegel, Bildung means performing activities whose point it is to develop or shape (bilden) the singular consciousness of an individual into one that has the “form of universality” and, in so d oing, to develop or shape the universal consciousness itself for which the individual consciousness provides the “fulfilling content.” Through the work of Bildung, indeed, the individual transforms herself. Yet according to Hegel, the transformation the individual undergoes in completing the work of Bildung does not consist in transforming what is initially a merely sensible being into an individual of a metaphysically new kind. Rather, Bildung is the transformation of an individual whose doing and being initially realizes a self-conscious form of life in a “immediate, natural” way into one who realizes this form of life in a “spiritual” way. Hegel does want to say that the consciousness a small child has of the unity of its d oing and being with the life-form it realizes differs from the consciousness of an adult who has gone through the work of Bildung. The child’s consciousness of this unity with the form of life it realizes has not yet achieved the “shape of universality.” The child is
S E L F - C O N S C I O U S N E S S A N D T H E I D E A O F B I L D U N G 307
conscious of the unity of its existence with its form of life, but not in virtue of any general represent ation of its form of life and concomitant consciousness of the difference between its individual life and the universality of its form of life. The child’s self-consciousness consists in a unity of elements, of whose differences it has only an “immediate” conscious. The work of Bildung consists in transforming the child’s immediate oing and being consists, into a consciousness of this unity, in which its d “spiritual” consciousness of that unity. An individual possesses this sort of spiritual consciousness when her consciousness of the unity of her doing and being with the life-form that she realizes is mediated by a universal representation of this form of life. If she has such a universal representation of her form of life, then her consciousness of this unity takes on another, spiritual character. The transformed character of the individual’s consciousness of her unity with the life-form she bears consists in the fact that she now grasps her own d oing and being as the realization of that life-form of which she has a universal representation. With this, self-consciousness is no longer anything other than consciousness of the reality of its own concept. Thus, when Hegel claims that the realization of a self-conscious form of life is essentially a m atter of Bildung, he does not simply want to point out that Bildung is a necessary condition for the existence of self-conscious individuals. He rather means to claim that Bildung is a moment in the realization of a form of life whose reality—on all levels of its realization— consists in being a self-conscious reality. When we understand Bildung in this way, according to Hegel, we understand it as an “immanent moment of the absolute.”
9 In the exposition of a self-conscious form of life, there is no place for the idea of a process that transforms a sensible being who is initially only potentially capable of self-conscious knowledge into a self-consciously knowledge-capable individual. For the concept of self-consciousness, as the characterization of the form of a form of life, does not describe an ability that can be said to represent a mere potentiality (such as a child might have) for some sort of doing and being. Nor does it describe an ability
308
H ist o rical P recedents
that one might describe as innate to an individual. It simply does not belong to t hose things that can be said to characterize an individual e ither qua first or qua second nature. Self-consciousness describes the intrinsic mode of realization of a life-form whose description as an “absolute point of transition [Durchgangspunkt]” requires the concept of Bildung. On this conception, Bildung is not a condition on which the reality of a particu lar individual’s self-consciousness depends and whose reality (that is, Bildung’s reality) it cannot think otherwise than as something that it must simply presuppose in the form of another self-conscious individual that enables this particu lar individual that is not yet self-conscious to undergo a process of Bildung through which it comes to acquire the capacities needed for self-consciousness. Rather, the concept of Bildung describes the intrinsic manner in which a self-conscious form of life realizes itself. This concept of Bildung, which reveals it to be an “immanent moment of the absolute,” rests on an understanding of self-consciousness according to which one cannot employ the concept of self-consciousness to describe something that is to be realized without thereby describing the reality of that being for which self-consciousness is something to be realized: namely, as a reality to be perfected. I have presented Hegel’s account of self-consciousness and its correlated account of the role of Bildung in a manner that emphasizes the differences between his and McDowell’s account. Yet, as I have also tried to bring out, the difference between them rests upon a deeply shared common insight. This is the insight that one cannot think about h uman life unless one acknowledges a standpoint that, right from the start, takes into account the idea of self-consciousness that enables one to think anything at all. I have argued that we should reject a specific feature of McDowell’s account of the relation between self-consciousness and Bildung on the ground that it threatens his own project. This suggests that we should not think of Hegel’s account of self-consciousness and Bildung as constituting a criticism of McDowell, but rather as an invitation for McDowell to endorse it on the ground that it deepens his own, most fundamental insight—as an invitation to be resolute about self-consciousness and the role that it plays in human life. (Translation from German by Daniel Smyth)
Chapter Eleven ✣
The Idealism in German Idealism ✣
Robert Pippin
1. Logical Idealism My suggestion in the following is that we can find the “idealism” in German Idealism, at least Hegel’s version, by following a path laid out by John McDowell in Mind and World, the path whose name is the “unboundedness of the conceptual.”1 At least I have found it extremely helpful and have been much influenced by it. The formulation is inspired by Wittgenstein’s remark that “when we say, and mean, that such-a nd-such is the case, we—a nd our meaning—do not stop anywhere short of the fact; but we mean: this–is–so.”2 McDowell’s diff erent formulation of the same point is that “there is no ontological gap between the sort of thing one can mean, or generally the sort of thing one can think, and the sort of thing that can be the case.”3 Or, even more to the point of this discussion, “The constraint [on thinking] comes from outside thinking, but not from outside what is thinkable.” 4
309
310
H ist o rical P recedents
To follow McDowell further: there is a form of this sort of claim of identity in Mind and World that can also be helpful h ere. This is so b ecause in his view we can certainly distinguish thinking from what is thought (the world is not a thought-thing; thinking is a discursive activity; the world is not a discursive activity) and still insist that the world “is made up of the sort of thing one can think.”5 (That discursive activity is, in its unity, the unity of anything that can be known, would be expressed on the “object side” by a determinate object being articulable as a single unity.) Or, for example, the profound-sounding claim that t here is no ontological gap between thought and world just comes to the fact that “one can think, for instance, that spring has begun, and that very same thing, that spring has begun, can be the case.”6 What I think when I know (think truly) that something is the case is simply what is the case. It is thus a truism of sorts that “the forms of thought are the forms of things.” This will have a Kantian spin if we note that “the form of things” for Kant includes the forms of intuition, space and time, but that route requires a qualification, that the forms are “only subjective.” This seems to take back with one hand (the subjective forms) what had been given with the other (the form of the extraconceptual, the form of t hings). This is, quite rightly by my lights, how McDowell wants to treat the preferability of Hegel’s over against Kant’s (official or received) form of idealism. The distinction between “conditions on the possibility of knowledge of things” and “conditions on the possibility of things themselves,” which Henry Allison uses to characterize Kant’s idealism, should be rejected “on the ground that the relevant conditions are inseparably both conditions on thought and conditions on objects, not primarily either the one or the other.”7 All of this w ill require qualification as we go on. McDowell, as he has explained subsequently, meant the phrase to apply to cases of our access to the world in experience, to cases of empirical knowledge, not to cases of “philosophical knowledge.” But McDowell admits that the model can be used for spontaneity, thinking being the “self-moving source of its own concepts.”8 This is what I propose to do in the following in a discussion derived from the recent claims made in Hegel’s Realm of Shadows.9 The determinate bearing of this on Hegel and “philosophical knowledge” becomes clear when we turn to the book he regarded as the heart of his systematic philosophy, a book he published in his early forties, more than a decade a fter his revolutionary Phenomenology of Spirit. This
T he I dealism in G erman I dealism
311
was a two-volume, three-part book that he wrote while teaching classical gymnasium students in Nuremberg between 1812 and 1816, at times teaching versions of the book itself to the no doubt bewildered high schoolers.10 It was called The Science of Logic, which he called a “science of pure thinking.” The first volume is called an “objective logic” and it contains a “logic of being,” and a “logic of essence.” The third part, the second volume, is called the “Subjective Logic” and it consists of a “logic atter in the extreme: this book still of the concept.” To understate the m awaits its full contemporary reception. Aside from occasional dust-ups about its beginning argument or its “movement,” and occasional attempts at an overview, or summary formulations of its purpose, it has not inspired the range of detailed engagement found in work on Kant’s Critiques or Hegel’s own Phenomenology of Spirit or Philosophy of Right. This is not to say that there have not been excellent philosophical reconstructions of various claims made in the Logic, especially in postwar German work, in the Heidelberg school in particular.11 It is just that the book does not have a living tradition of animated, mutually informed scholarship. It is very rarely taught in graduate seminars, and certainly not in undergraduate courses, and the quantity of scholarship pales in comparison with other historical figures like Aristotle, Descartes, or Kant. This is so even though there are now a number of other examples of philosophical reflection on logic that form something like a context within which Hegel’s project ought to be comprehensible as an alternative. I mean not just Kant’s Transcendental Logic, and Fichte’s Wissenschaftslehre, but Frege’s Begriffsschrift, Wittgenstein’s Tractatus, and Husserl’s Logical Investigations.
2. The Basic Problem This is what I propose in the following five sections. A fter this sketch (§2) of what I consider the basic challenge in understanding and defending this idealism, I w ill discuss (§3) its central claim of an identity between the forms of thought and the forms of being (this is where the bearing of the “unboundedness” claim will be most evident); (§4) its similarities with and differences from Kant’s idealism; (§5) in what sense it is or is not correct to characterize Hegel’s position as an “objective idealism”; and (§6) the developmental structure of such a “logical idealism.”
312
H ist o rical P recedents
This idealism, at least the thread that travels from Kant, through Fichte, to Hegel (Schelling’s “idealism” is another issue) has three components. The first is the claim that a priori knowledge of the world, the ordinary spatiotemporal world, as well as “objects” and practices in it like art, religion, and the state, is possible; this is knowledge about that world, but achieved independently of empirical experience. Idealism in this sense is primarily a critique of empiricism (not of empirical knowledge, although it is sometimes confused with such a critique; empiricism is itself an a priori position, intended to explicate what any possible knowing amounts to). The second component is where all the interpretive controversies begin. It is the claim that this a priori knowledge, while in some sense to be specified, ultimately about the world that exists independent of thought, consists in thinking’s or reason’s knowledge of itself; thinking’s determination of thinking or, as Hegel designates it, a “science of pure thinking.” (Kant’s allegiance to this principle is manifest in The Critique of Pure Reason: “reason has insight only into what it itself produces according to its own design.”12 ) It is understandable, but also quite false, to think that these two components can only be jointly claimed if either (a) objects of knowledge depend for their existence on being thought, or (b) access to objects requires some sort of mind-imposed unification of sensory elements, resulting in a “subject-mediated” product, an appearance, not the thing as it is in itself. There are many versions of this existential dependence, or subject-mediated interpretation of German Idealism in the extant literat ure. This view no doubt stems from the understandable but false inference that if such a conceptual structure is not derived from experience, it must be contributed by, or “imposed by,” us. This impositionism must be so, whether the claim is that objects depend for their availability as possible objects of experience on such “mind- imposed” unity, or, in a diff erent tack, in what is known as “objective idealism,” if the claim is that what there is is, in some sense or other, “really” a concept. (On this view, the idealism in Hegelian idealism refers to the ideal, nonsensible or noetic true nature of reality itself.)13 A variation on this notion of restricted knowledge is the claim that philosophy can determine only the finitely knowable aspects of the in-principle knowable, requiring us to admit that we do not know objects as they are fully in themselves. (This looks to Hegel, and should look to us, like a claim that we possess knowledge, but not really.) I do not believe any of the major later
T he I dealism in G erman I dealism
313
idealists believed anything like this, and I have come to believe that Kant, in his better moments, did not believe it e ither. But there is clearly a question to be answered, and it amounts to the third dimension of idealism: how the first two components could possibly be true (that objective a priori knowledge is possible, and that what pure reason knows in such knowledge is “itself,” thinking itself), if the standard versions of the third component are not true too. Since interpretations that opt for the first alternative are rightly dismissed as “too Kantian” a reading of Kant, the standard interpretation of Hegel on this point was indicated above, that these two claims can be jointly assertable only if what there “really” is, the “really real world,” what is accessible only to pure reason alone, consists in abstract entities, thoughts, “thinking moments.” Something like the Absolute’s or God’s thinking itself, an inherent, evolving noetic structure. This is sometimes what scholars are insisting on when they insist, against what they perceive as “nonmetaphysical” readings, that Hegel was certainly (and for many quite obviously) a “metaphysician” in just this sense. But these options do not exhaust the relevant alternatives. It is true: Hegel most certainly was a “metaphysician,” but he lived nowhere near this vaguely neo-Platonic neighborhood, nor in the mind-imposed unity camp. The most import ant watchword for Hegel’s Logic, once we realize that no form of “object dependence on subject” is at stake in that project (an extremely widespread general view of what idealism must be to count as idealism), is that we are not talking about any dependence but about an “identity” (a “speculative identity” to be sure) between the forms of pure thinking and the forms of being, an identity compatible with maintaining a difference between thinking and being.14 This is just the identity at stake in McDowell’s unboundedness claim. There has been, to my mind anyway, so much confusion around the idea of a “nonmetaphysical” reading of Hegel that this point about identity should be stressed. Hegel is a metaphysician because he believes that the forms of thought are the forms of being, and he is an idealist b ecause he believes in the autonomy of, the nonderived status of, such forms of thought, and in its spontaneous self-unfolding in the Logic. Using Kant in my 1989 book was, I thought, a way of suggesting that the self-standing self-a rticulation by pure thinking of itself was where first philosophy begins and ends for Hegel (the Critique of Pure Reason is a critique by reason of itself, a self-authorization and self-constraint), but b ecause I did
314
H ist o rical P recedents
not emphasize sufficiently the former “metaphysical” claim, it was easy enough for readers to assume that I was ascribing to Hegel a “subjective idealism.”15 That would be what I called here “impositionism” and it is very clearly not Hegel’s position. Anyone can see that from almost any page anywhere in Hegel, but nevertheless it is what a “Kantian” or “epistemological” reading (mine, supposedly) was supposed to attribute to Hegel. I note also the widespread view that this Hegelian project is doomed from the start, that t here is not and cannot be such a topic, “pure thinking.” Since Hegel’s contemporary, Schelling, began this line of attack, it has kept reappearing in the European tradition down to the present, with the popularity of “new realisms” and speculative materialisms and the influence of the cognitive and neurosciences. The criticism is that thinking must always be understood as grounded on, or dependent on, or an epiphenomenon of, the materiality or contingency or unconscious source or instinct of the thinker. Thinking always has as its condition “the nonthought,” or is always embodied and so conditioned by body or nature. There are several versions of this. But this criticism is wrongheaded and question-begging from the start. The topic of pure thinking, the logic of thinking, has nothing to do with, in no way rests on some view about, the thinker, the subject, consciousness, the mind. The topic so understood rather raises again as a problem the possibility of the intelligibility of what ever is being touted as source or hidden origin, the conditions assumed in any such determinate identification. Any such criticism, insofar as it is a thinking, a judging, and a claim to know, is always already a manifestation of a dependence on pure thinking and its conditions, and such “moments” of pure thinking are to delimit the normative domain of intelligibility (what can rightly be distinguished from what, as one example) and not to describe any process or series of events that goes on in supposed indepen dence of the empirical world. Any thinking that is on to objects (anything at all thinkable) in any way assumes “what it is for thinking to be on to objects,” and this question cannot be an empirical question. Any such inquiry could only determine “what goes on” when thinking is on to objects, not what it is for it to be. It can only be a pure or a priori question, as is its complement: what must thinking be, determinately, for it to be on to objects, for it to be able to issue in judgments that are truth b earers? Pure thinking is neither dependent on nor independent of the empirical
T he I dealism in G erman I dealism
315
or materiality or the brain or whatever new absolute comes into fashion. That question already manifests a misunderstanding of the question of pure thinking itself. This is not to deny that any reference to thinking presumes a thinker, indeed a living, purposive, rational thinker. It is to argue for the autonomy of the question of “any thinking at all.” That is, it is to insist on the priority and autonomy of “logic.”
3. Idealist Identity The difficult issues concern not what I know when I know what is the case, but how I can be in a position to know what is the case, e ither empirically (because, in McDowell’s example, say, I have seen the signs of spring) or, especially, when the claim is specific, about a particular “condition” known a priori—for example, that every event has a cause. (We could also say that while this must be what it is to think what is the case, we have not thereby explained how it is possible to do so. For Hegel, we can so explain it only by explaining the identity and difference of thinking and being, a point he w ill address in the “determinations of reflection” in the Essence Logic.) Here is Hegel’s formulation: The older metaphysics had in this respect a higher concept of thinking than now passes as the accepted opinion. For it presupposed as its principle that only what is known of things and in things by thought is r eally true [wahrhaft Wahre] in them, that is, what is known in them not in their immediacy but as first elevated to the form of thinking, as things of thought. This metaphysics thus held that thinking and the determination of thinking are not something alien to the subject matters, but are rather their essence, or that the things and the thinking of them agree in and for themselves (also our language expresses a kinship between them); that thinking in its immanent determinations, and the true nature of things, are one and the same content.16 It will be important (for Hegel, at any rate) that this account of an “identity” (“one and the same content”) is true of philosophical or speculative thinking, thinking that has as its subject m atter “true being” or “actuality.” Hence the identity within difference of being and thinking, the
316
H ist o rical P recedents
core principle of Hegel’s version of idealism. In other words, there is an “identity” between “pure thinking’s moments” (suitably well defined) and “any possible object of pure thinking,” or pure thinking’s “truth.” It is an identity within difference because the speculative claim does not mean that the world, what seems other than thinkings, must nevertheless be “thinkings.” It is not, any more than true thoughts (judgments) are true by virtue of the world being thoughts. Once we understand the necessary dependence of any true thinking about anything on pure thinking, and once we understand what constitutes pure thinking (especially its “spontaneity”), and once we understand the “moments” necessary for pure thinking to be pure thinking, we have established thereby the truth about what there is, what there is in its intelligibility.
4. Kant and Hegel As noted at the outset h ere, the context of much of McDowell’s own discussion is questions about perception and perceptual knowledge claims, so, while a valuable guide, it is of limited use for the details of The Science of Logic, which is some sort of claim to a priori knowledge on its own, in dependent of the role of pure concepts in perception (which certainly is an issue for Hegel in the Phenomenology of Spirit and the Philosophy of Subjective Spirit). Moreover, the relation of thinking and the thinkable in the Logic faces a problem not germane to McDowell’s concerns in his book. The problem is: How to account systematically for the determinate moments of “any thinking of the thinkable,” such that they count as the determinate moments of the thinkable itself, being?17 The core of the idealism issue (the status of pure thought and its determination of itself), for both Kant and Hegel, had especially to do with the implications of a relationship that they both insisted on between thought’s determination of what it is to be thinking (in the sense of knowing) and the conceptual content required for thinking to be successful in judging (in the sense of knowing)—that such content must be understood as what it is only by being known a priori to be such content. Distinctly philosophical knowledge, for that is what we are discussing, is and must determine for itself its own possible objectivity. More prosaically put, if there is to be such a priori knowledge, it c an’t be made true by
T he I dealism in G erman I dealism
317
something “outside” of judgment, to which it can be compared. (Here is where the relevance of McDowell’s perceptual discussion begins to be of less use.) Any such appeal to objectivity is “inside” judgment’s self- determination, a point made the most of, among the idealists, by Fichte. (To say that the Ich posits itself and in so doing posits the nicht-Ich is his way of making this point. It has nothing to do with creating an external world by thinking it. He meant that thought’s measure of its own possibility, as knowledge of the nicht-Ich, could only be an internal measure of itself.)18 Thought was in this sense “self-determining”; concepts could be said, and w ere frequently said by Hegel, to “produce their own content.” This involves revolutionary claims made first by Kant about the exclusively productive or spontaneous nature of thinking and the inherently self-conscious nature of all thinking. Both claims were enthusiastically embraced by Hegel, although the implications for philosophical truth were quite diff erent for him from those Kant drew.19 Kant’s case for such synthetic a priori knowledge was not traditional rationalism, because Kant denied any receptive relation between a domain of objects and pure thinking. Hegel agreed. Pure thinking only has itself as its proper object. It is an endlessly interesting aspect of this claim (and almost impossible to hold in mind properly) that thinking so conceived, is not conceived as an entity or event, w hether psychological or immaterial. Thinking is its own “object” only in the sense of what pure thinking is about—the activity of thinking necessary for it to be thinking, and necessary for it to be a thinking of (in the sense of knowing) objects. But, when the argument is formulated this way, Kant also seemed to say that philosophy could determine something about the objects of thought only insofar as they were subjected to “subjective” or “our” conditions, to conditions for the applicability of thought, a condition which itself required something other than forms of thought; that is, it required forms of sensible intuition unique to finite, human, rational knowers. There is no question that Hegel rejects what he understood to be such “subjective idealism,” and so rejects the Kantian claim that we know only phenomena, not things in themselves. In fact, Hegel never seems to tire of rejecting it at e very chance he gets, noting that Kant seems to give with one hand (knowledge) what he takes away with the other (“merely” of appearances), and that this is partly why Kant drew the inference from his demonstration that the exercise of our form of thought produces
318
H ist o rical P recedents
contradictory results, generates antinomies. Kant should have been led to question the adequacy of his understanding of our “forms of thought.” And Hegel does not reject Kant’s idealism by arguing that we do know the “things in themselves” that Kant said we c ouldn’t. The clearest statement of his position is in §44 of the Encyclopedia Logic, where he rejects the concept altogether. He basically notes that if we define the notion by asserting that, if we abstract away any means we have for knowing, we should not be surprised that we end up with something unknowable, a caput mortuum. “Equally simple, however, is the reflection that this caput mortuum is itself merely the product of thought, more specifically, [the product] of thought that has progressed to pure abstraction, [the product] of the empty I that makes this empty identity of itself into an object for itself.”20 And yet Hegel also never tires of saying that his own theoretical philosophy is like Kant’s, in that at its heart is a logic, an enterprise in which our thinking has itself as its proper object. He is also clear that such a theory of pure thinking is a successor to, not a further episode in, the history of modern rationalist “metaphysics.” He also never tires of mentioning his debt to Kant, even given the strong disagreements. This sets the stage for the interpretive task. The following remark is typical— so typical that fidelity with it o ught to function as something like a necessary condition for the possibility of work on the Kant-Hegel relation: “The primary concern [das nächste] of the Kantian philosophy is thus that thinking is supposed to investigate itself, the extent to which it is capable of knowing. Nowadays, the Kantian philosophy has been left behind, and everybody wants to be at a point further on [weiter]. To be further along [Weitersein], however, has a double meaning: both to be further ahead and to be further b ehind. Looked at in clear light, many of our philosophical endeavors are nothing but the method of the old metaphysics, an uncritical thinking along [unkritisches Dahindenken] in a way everyone is capable of.”21 This task is also connected to various senses in which the appeal to the “ideal” in German Idealism is understood. As already noted, if it is possible to identify some common core to the tradition classified as “German Idealist,” it would at the very least be as a sustained critique of empiricism— the denial that the basic intelligibility of the world can be explained by the sheer deliverances of sensibility alone, together with the subsequent
T he I dealism in G erman I dealism
319
putative organi zation of such deliverances by abstraction, association, generalization and so forth. This is connected as well with a rejection of empiricist accounts of action, as if the basic engine of action is conative, with reason only of strategic use, only a means to satisfy desires. As noted e arlier, the best way to understand what makes German Idealism idealism is this anti-empiricism, now, post-Sellars, understood as the rejection of the possibility of, and so any foundational reliance on, givenness. And this aspect is quite relevant to McDowell’s project. “The given” is understood as noninferentially warranted cognitive states, such that any such putative state counts as knowledge of a kind, or a basis of knowledge, just by being had, or experienced. By contrast—and here is the heart of the idealist claim—a ny sensory interchange with the world (or even any intellectual receptivity) that is to be able to play any role in cognition can play such a role only if always already informed by a nonempirically derived categorical structure. So of course the central questions have to be: what is the nature and status of this categorical structure (or pure thinking’s knowledge of its own structure and possibility)? And what does “already informed (or en-formed) by” mean? Moreover, while it might seem natural to think that, if this categorical structure cannot be said to be derived from, contributed by, sense-experience, it must be contributed “by us,” must be subjective, as if “imposed” by us on such sensible deliverances, this implication is clearly rejected by Hegel. Hegel thinks that this is Kant’s position and, again, t here is no question that he rejects that sort of subjective idealism (whether he is right that this is Kant’s position or not).22 What possibility is left?
5. Objective Idealism? It is perfectly clear to Hegel how much he is asking of us, how natural one (empiricist) or the other (subjective idealist) alternative seems, but he nonetheless insists we must leave this assumed alternative behind. “We must then reject the opposition between an independent immediacy in the contents or facts of consciousness and an equally independent mediation, supposed incompatible with the former. The incompatibility is a mere assumption, an arbitrary assertion. All other assumptions and postulates
320
H ist o rical P recedents
must in like manner be left behind at the entrance to philosophy, whether they are derived from the intellect or the imagination.”23 This might suggest—and has, to many—another alternative, already mentioned, another sense of idealism. The “ideal” in German Idealism could also refer to some roughly Platonic or neo-Platonic sense of an ideal or noetic (nonsensible) structure of reality, something supposedly more real than, underlying, manifest imperfectly in, sensible deliverances in experience and itself accessible only to pure reason. Hegel’s own characterizations of idealism sometimes do not invoke the identity formulations noted above, but assert that idealism is constituted by the “unreality” or “untruth” of the finite. By contrast, what is “truly real” is the infinite, the Concept, the Absolute. We would need to understand those passages in their proper context, but in merely introducing his basic claims, it is already clear how difficult it w ill be to pin down Hegel’s core position, as Hegel also rejects an “objectivist” interpretation of such a categorical structure. This is of course not surprising. Hegel is supposed to be a dialectical thinker, and we would expect each one-sided alternative to be “sublatable” into a higher unity. His dissatisfaction with such one-sided alternatives is something frequently voiced by Hegel. For example, in the Addition to EL §24, Hegel says (at least in this context, channeling his inner Schelling), “the logical is to be sought in a system of thought- determinations in which the antithesis between subjective and objective (in its usual meaning) disappears. This meaning of thinking and of its determinations is more precisely expressed by the ancients when they say that nous governs the world, or by our own saying that there is reason in the world, by which we mean that reason is the soul of the world, inhabits it, and is immanent in it as its own innermost nature, its universal.”24 This passage already suggests that the subjective idealism / objective idealism alternatives are not the only alternatives, that this very formulation of the alternatives has missed something crucial to Hegel. What any cognitive thinking does is to render something intelligible, a task that has many diff erent dimensions. The truth of any such claim is often, of course, a matter of empirical experience, but the question of what any such account is, what it is to render successfully intelligible in any of t hese and many more senses, and what relation such renderings have to each other, is not, cannot be, an empirical m atter. And “what it is to render successfully intelligible” need not at all be limited to “what we count as
T he I dealism in G erman I dealism
321
having done so.” Understanding what it is to render something intelligible is just thereby to understand the intelligibles, what t here is to be thought, being. So a science of logic is a science of pure thinking. Pure thinking’s object is, and only is, itself. But this “object” is not a nature, an object. As noted above, Hegel’s Logic has nothing to do with “the mind” as a substance or thing. As in so many cases, Hegel is following both Aristotle h ere and Kant, for whom the claim that the “I think” must be able to accompany all my represent ations is a logical point—that is, expresses the form of thought—and is not a claim about how the mind actually operates.25 If that were the case, and Hegel were making a claim about the mind’s nature, knowledge would be l imited by its “instrument,” something Hegel had been vigorously denying since the Introduction to the Phenomenology. In knowing itself, what pure thought knows is the possible intelligibility, the knowability, of anything that is. But the intelligibility of anything is just what it is to be that thing, the answer to the “what is it” (tode ti) question definitive of many sciences since Aristotle. So in knowing itself, thought knows of all t hings, what it is to be anything. Again, as for Aristotle, the task of metaphysics is not to say of any particular t hing what it is. That is the task of the individual sciences. It is rather to determine what must be true of anything at all, such that what it is in particular can be determined by the special sciences (what in scholasticism w ere called the transcendentalia). Or: it is to know what is necessarily presupposed in any such specification.26 Put another way, the task of metaphysics is to understand what it is to say of anything what it is.
6. Logical Movement Further, in Hegel’s treatment, as in Kant’s, rendering intelligible in any of these senses is not just to grasp a content; it is not an intuitional reception. (Pure thinking’s object is itself, but not as an object to be grasped or “seen.” Pure thinking interrogates itself by thinking. What is being denied h ere is again the heart of rationalist epistemology, its reliance on noetic intuition, the “light of reason.”) Thinking is discursive; to think what is the case is to assert that it is (the basic unit of intelligibility in Hegel’s account is the judgment; assertion is its linguistic manifestation), and that is
322
H ist o rical P recedents
something always open to challenge and interrogation. This means that a science of thinking is also a science of “reasons,” of ways of giving reasons in rendering anything genuinely or properly intelligible. It is in this sense— thinking’s determining for itself what it is to think, in the sense of judging what is the case for appropriate reasons—that there can be a “science of logic.”27 So there is at least to be a common claim in these alternatives, something like: there is some sort of rationally articulable, categorical structure without which the world would not be even empirically intelligible at all. We could even, at this stage, call this categorical structure (in a way very different from the contemporary understanding of formal logic) the world’s “logic,” and this in the oldest sense of the logos inherent in all things, as in the quotation about nous above. This categorical structure is neither subjective nor objective, or somehow both subjective and objective. (Logic is not “formal” in Kant’s sense, that is, empty; rather, it is said by Hegel to be formal because it is a science of determinate form. However, as we s hall see, this does not mean that it is a science of special objects, thing-like entities. Such forms are said to be instead moments, Denkbestimmungen, thought-determinations, even “pulsations,” a claim that requires more unpacking than is possible here.28) Further, a Hegelian logic is also not an Aristotelian or scholastic logic, not a descendant of the Port Royal logic, not a version of a Leibnizean / Wolffian conception of logic. It is closer to, but must still be distinguished from, both what Kant calls “general logic” and that Kantian invention: “transcendental logic.” And it clearly cannot be interpreted in the terms fashioned by Frege for a predicate logic, or any version of a mathematical or symbolic logic. But we d on’t get very far by understanding what the Logic is not. The structure of the Logic involves ever more adequate self-consciousness about the determination of a kind of conceptual content, that kind specifiable by thinking alone. In the Logic, that kind is the content appropriate to the conceptual capacities required for the thought of anything at all. (Given the judgmental character of intelligibility for Hegel, by “the thought of anything at all,” I mean the thought of—in the sense of the knowledge of—anything’s being the case, anything’s being such and such, anything happening.) T here can certainly be other determinants of content, empirical experience, for example. But The Science of Logic is not about concepts like horse or Ferris wheel, but about concepts,
T he I dealism in G erman I dealism
323
or conceptual capacities, necessary to render intelligible any objective content. This w ill not mean that such content “depends” for its intelligibility wholly on “subjective capacities,” as if the content amounts to “rules” of thought understood as such a subjective capacity. It is Hegel’s position that the structure of reality is conceptual, is its intelligibility. And “in us” there is an actualization of a conceptual power, an actualization that is not in the nonspeculative sense identical with, the same as, the being-at-work in the t hing (the t hing is not trying to make itself intelligible), but that, in the right understanding of that being-at-work, its actualization in thought, is the same “it” that is actualized in being, the principle of its intelligibility. This can be understood in the same way that, according to Aristotle, in sense perception there can be a single “actualization” of two distinct potentialities, in the perceptible and in the perceiver.29 Or, with the proper qualifications, it can be expressed as Kant would express categories; that, for example, causality refers as much to a way of understanding as it refers to a feature of any possibly intelligible world. So it is perfectly appropriate to say such t hings as that for Hegel reality “has a conceptual structure” or “only concepts are truly real,” as long as we realize we are not talking about entities, much less separable, immaterial, abstract entities, but about the “actualities” of beings, their modes or ways of being what determinately and intelligibly they are. To say that “any object is the concept of itself” is to say that what it is in being- at-work-being-what-it-is can be determined, has a logos, and this is the other side of the coin as “every judging is the consciousness of itself as judging.” We can even say that reality comes to self-consciousness of itself in us; or that the light that illuminates beings in their distinct being-at- work is the same light that illuminates their knowability in us, as long as we do not mean a light emanating from individual minds.30 Such an inquiry w ill culminate in something like “the concept of the concept itself,” and so concerns a kind of logical (not psychological) full self-consciousness.31 (According to Hegel, the concept is the “ground and the source of all finite determinateness and manifoldness.”32) Again, pure thinking has only itself as its object. That is, the dynamic of the Logic is internally self-critical. Implicit but ungrounded assumptions about such conceptual determination are exposed and shown to be inconsistent with some initial conceptual articulation, thus requiring revisions and r eformulations. The final “concept logic” makes conceptuality itself its
324
H ist o rical P recedents
theme, and, whatever else is going on in the Science of Logic, it is certainly the case that assumptions about conceptuality (especially about conceptual content or determinacy) had been at work in the book’s prior accounts, but not attended to as such until the final volume. In EL §17, Hegel says that the “unique purpose, deed and goal” of his Logic is “to arrive at the Concept of the concept and so to arrive at its return [into itself] and contentment [Befriedigung].”33 And in the Addition to §83 he says simply, “Only the Concept is what is true, and, more precisely, it is the truth of Being and of Essence. So each of these, if they are clung to in their isolation, or by themselves, must be considered at the same time as untrue.” This impression—of a self-critical, self-correcting, internal pro cess of reflection—is only strengthened toward the end of the EL, when Hegel speaks, in summation of the point he has reached, of “the concept that has carried itself out [sich selbst ausgeführt] in its objectivity, or the object that is inner purposiveness, essential subjectivity [wesentliche Subjektivität].”34 This result could easily be misinterpreted. The absolute idea, expressed in Hegel’s terms as the identity of logic and metaphysics, could be understood as some sort of direct inference from the logical structure of thought. The basic form of rendering intelligible, one might reason, is the one place categorical judgment, S is P. This simply requires, if it is to be intelligible, that the world be structured as substances and properties. QED. But that would be dogmatism, and would be rejected by Hegel. The characteristic and necessary features of judgment must be derived with a claim to necessity from the simplest, most immediate manifestation of any contentful thought, “Being!,” which is the first moment of the Being Logic and of the book itself. This internal derivation of more complex conceptual moments in order for thought to be rightly onto objects, and the kind of necessity claimed, is what answers in Hegel to Kant’s insistence on a “transcendental deduction” of the objectivity of the categories. While it is always possible to suspect that, in any such derivation, we are specifying only “what we must think” or even “must believe” in order to judge rightly that something is the case, such a suspicion is arbitrary if there is no reason to suspect such parochialism, as if thinking w ere a kind of species- characteristic capacity. The radicality of Hegel’s presuppositionless beginning and the necessity of the derivation is supposed to eliminate such a suspicion from the outset, and the self-negating and self-correcting
T he I dealism in G erman I dealism
325
derivation is supposed to preserve such purity. He realizes that the avoidance of any such parochialism, the establishment of pure thinking just as such as the “truth” of being, w ill disappoint anyone used to a more substantive version of metaphysics. In the Encyclopedia Logic, in a rather informal anticipation of what he senses might be his reader’s frustration with the abstractness and nonsubstantive character of such an absolute, Hegel notes, “When one ere finally the substantive speaks of the absolute idea, one can think that h must come to the fore, that h ere everything must become clear.”35 He is anticipating t hose interpreters who want Hegel to defend a substantive absolute, God, a necessary being, a substance that is also subject and so forth, and he is taking steps here to close off explicitly such an interpretation. The absolute idea, he notes, “has shown itself to amount to this, namely that determinateness does not have the shape of a content, but that it is simply as form, and that accordingly the idea is the absolutely universal idea. What is left to be considered h ere therefore, is thus not a content as such, but the universal character of its form—that is, method.”36 Appropriately, the truth is the truth being demonstrated throughout the book, recollected here as such, the identity of the forms of thought and the forms of being, now thought as such, and not any determination of content (any determination of which quality, what essence, what cause, e tc.).37 That truth, though, the absolute idea, just is self-conscious conceptuality, or the right understanding of the implications of the logical structure of apperception, or purely logical knowledge, and in this purity the manifestation of absolute freedom. We are very far here from a Hegel committed to deriving the content of the world from pure thought alone, or who believes he has determined once and for all “the purpose of the world.” But, however “shadowy,” Hegel’s Logic (he had characterized it as “the realm of shadows”) is hardly a thin formalism. It, the idea of thought’s pure self-determination, the right understanding of thinking generally, stands against many other “absolute ideas” (whether characterized that way or not) that are prominent in modern philosophy: empiricism, dogmatic rationalism, reductionism, scientism, consequentialism, moralism. And the implications of the account of pure thinking, of philosophy itself, once t hose implications are drawn out in the rest of the Encyclopedia, are of g reat importance in philosophy, for how we should understand the conceptual
326
H ist o rical P recedents
structure of any investigation of nature, and especially for how human freedom should be understood, what counts as its realization—as in, for example, this formulation from McDowell: “What is exciting about Hegel is that he finds a way to bring freedom, as something that makes human beings metaphysically special, within the scope of knowable reality, by overcoming the Kantian idea that knowable reality coincides with the realm of Nature.” 38
N o t es Ack n owle d g ment s Co n t r i but o rs I n d ex
Notes
Introduction 1. Edited volumes on McDowell’s work include Reading McDowell: On Mind and World, ed. Nicholas A. Smith (London: Routledge, 2002); McDowell and His Critics, ed. Cynthia MacDonald and Graham MacDonald (Malden, MA: Blackwell, 2006); John McDowell: Experience, Norm, and Nature, ed. Jakob Lindgaard (Malden, MA: Blackwell, 2008); and Mind, Reason, and Being-in-the-World: The McDowell-Dreyfus Debate, ed. Joseph K. Schear (London: Routledge, 2013). A list of article-length discussions would be too extensive to include here. 2. John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994). 3. Stephen Pyke, Philosophers (Manchester: Cornerhouse Press, 1993). 4. Compare this well-known passage from Christine Korsgaard’s The Sources of Normativity: A lower animal’s attention is fixed on the world. Its perceptions are its beliefs and its desires are its w ill. It is engaged in conscious activities, but it is not conscious of them . . . But we human animals turn our attention on to our perceptions and desires themselves, on to our own m ental activities, and we are conscious of them . . . A nd this sets us a problem no other animal has . . . For our capacity to turn our attention on to our own mental activities is also a capacity to distance ourselves from them, to call them into question. I perceive, and I find myself with a powerf ul impulse to believe. But I back up and bring that impulse into view 329
330
N O T E S T O PAG E S 4 – 9
and then I have a certain distance. Now the impulse d oesn’t dominate me and now I have a problem. S hall I believe? Is this perception really a reason to believe? I desire and I find myself with a powerf ul impulse to act. But I back up and bring that impulse into view and then I have a certain distance. Now the impulse doesn’t dominate me and now I have a problem. S hall I act? Is this desire r eally a reason to act? (Christine Korsgaard, The Sources of Normativity [New York: Cambridge University Press, 1996], 92–93) 5. We call this picture “Platonic” because it is often attributed to Plato and is naturally suggested by some of his remarks about the relationship between the parts of the soul in the Republic and other dialogues. We do not claim that this is really Plato’s view: his true view is, we think, more subtle. 6. Gareth Evans, The Varieties of Reference (Oxford: Clarendon Press, 1982), 158–159. 7. McDowell, Mind and World, 64. 8. For further discussion of the contrast between “additive” and “transformative” accounts of rationality, see Matthew Boyle, “Additive Theories of Rationality: A Critique,” European Journal of Philosophy 24, no. 3 (2016): 527–555. 9. See esp. McDowell, Mind and World and “Naturalism in the Philosophy of Mind,” in his The Engaged Intellect: Philosophical Essays (Cambridge, MA: Harvard University Press, 2009), 257–275. 10. Ludwig Wittgenstein, Philosophical Investigations, trans. G. E. M. Anscombe, ed. P. M. S. Hacker and Joachim Schulte (Malden, MA: Blackwell, 2009). 11. Esp. John McDowell, “Wittgenstein on Following a Rule,” Synthese 58 (1984): 325–364. 12. Esp. John McDowell, “Virtue and Reason,” Monist 62 (1979): 331– 350, and John McDowell, “Deliberation and Moral Development in Aristotle’s Ethics,” in The Engaged Intellect, 41–58. 13. See also John McDowell, “Two Sorts of Naturalism,” in Virtues and Reasons: Philippa Foot and Moral Theory, ed. Rosalind Hursthouse (New York: Clarendon Press, 1995). 14. See esp. McDowell, Mind and World, lecture IV. 15. This is also a theme in Andrea Kern’s “Self-Consciousness and the Idea of Bildung,” which appears in Part IV. 16. See also McDowell, Mind and World, and John McDowell, “Conceptual Capacities in Perception,” in Kreativität, ed. G. Abel (Hamburg: Felix Meiner Verlag, 2006). 17. See also John McDowell, “The Logical Form of an Intuition,” Journal of Philosophy 95, no. 9 (1998): 451–470, and John McDowell, “Avoiding
N O T E S T O PAG E S 1 0 –1 6
331
the Myth of the Given,” in Having the World in View: Essays on Kant, Hegel, and Sellars (Cambridge, MA: Harvard University Press, 2009), 256–272. 18. Notably, John McDowell, “Criteria, Defeasibility, and Knowledge,” Proceedings of the British Academy 68 (1982): 455–479; John McDowell, “Knowledge and the Internal,” Philosophy and Phenomenological Research 55, no. 4 (1995): 877–893; John McDowell, Perception as a Capacity for Knowledge (Milwaukee, WI: Marquette University Press, 2011); and John McDowell, “The Disjunctive Conception of Experience as Material for a Transcendental Argument,” in Disjunctivism: Perception, Action, Knowledge, ed. Fiona Macpherson and Adrian Haddock (Oxford: Oxford University Press, 2008), 376–389. 19. McDowell, Mind and World, lecture V; John McDowell, “What Is the Content of an Intention in Action,” Ratio 23, no.4 (2010): 415–432; John McDowell, “Some Remarks on Intention in Action,” Studies in Social Justice (2011): 1–18; John McDowell, “Anscombe on Bodily Self-Knowledge,” in Essays on Anscombe’s Intention, ed. Anton Ford, Jennifer Hornsby, and Frederick Stoutland (Cambridge, MA: Harvard University Press, 2011), 128–146. 20. John McDowell, “Some Remarks on Intention in Action,” The Amherst Lecture in Philosophy 6 (2011): 1–18; available at: http://w ww .a mherstlecture.org/mcdowell2011/. 21. See esp. McDowell, Mind and World, lectures II and VI. This idea is also a theme in Robert Pippin’s “The Idealism of German Idealism,” which appears in Part IV. 22. See also McDowell, “Deliberation and Moral Development in Aristotle’s Ethics.” This McDowellian theme is also discussed in Jennifer Whiting’s “See the Right Thing: ‘Paternal’ Reason, Love, and Phronêsis,” which appears in Part IV. 23. See esp. McDowell, “Hegel’s Idealism as a Radicalization of Kant,” in Having the World in View, 69–89. 24. See also McDowell, Mind and World, lecture IV, esp. §4. 25. McDowell, Mind and World, 111. 26. Ibid., 27. 27. Robert Pippin, Hegel’s Realm of Shadows: Logic as Metaphysics in “The Science of Logic” (Chicago: Chicago University Press, 2018). 28. McDowell, Mind and World, ix.
1. Skepticism and Quietism about Meaning and Normativity This paper descends from an e arlier paper, “The Significance of Signposts: A Challenge to Quietism about Meaning,” written for a workshop on Wittgenstein’s rule-following considerations at the University of Leipzig in 2013. Subsequent versions of that paper w ere presented at the University of Riverside,
332
N O T E S T O PAG E S 1 9– 2 1
the University of Copenhagen, the Ohio State University, and Johns Hopkins University. Some of the material in that paper was revised, under the title of the present paper, for the conference “Skeptical Solutions” at the University of Bonn, and presented again in various forms at the University of Salzburg, the University of Dresden, the University of Vienna, the University of California, San Diego, and the University of Tuebingen. I am grateful to audiences on all those occasions for questions and discussion. I would also like to thank Evgenia Mylonaki for her extensive and insightful comments on the penultimate version.
1. See John McDowell, “Wittgenstein on Following a Rule” and “Meaning and Intentionality in Wittgenstein’s Later Philosophy,” both reprinted in John McDowell, Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998); “How Not to Read Wittgenstein’s Philosophical Investigations,” reprinted in John McDowell, The Engaged Intellect (Cambridge, MA: Harvard University Press, 2009); and “Wittgensteinian “Quietism,’ ” Common Knowledge 15 no. 3 (2009): 365–372. 2. See Saul Kripke, Wittgenstein on Rules and Private Language (Cambridge, MA: Harvard University Press, 1982). All page references to Kripke are to this book. McDowell’s understanding of Kripke’s interpretation of Wittgenstein, and his challenge to Kripke, are most clearly articulated in §§3–4 of “Following a Rule” and §§1–6 of “Meaning and Intentionality.” 3. For simplicity, I refer simply to “McDowell’s” quietism, although McDowell puts it forward not as his own, but as Wittgenstein’s. In a related simplification, I refer to the skeptical problem which Kripke ascribes to Wittgenstein as “Kripke’s” skeptical problem. 4. I first encountered John McDowell when, as a second-year undergraduate at Oxford in the spring of 1978, I attended a graduate seminar that he and Michael Dummett conducted on the topic of realism and anti- realism in the philosophy of language. This seminar (in which, incidentally, McDowell consistently seemed to me to have the upper hand in his debate with Dummett) was one of the formative experiences of my time as an undergraduate. My next serious encounter with McDowell’s ideas was soon a fter the 1994 publication of Mind and World, which I read in the context of working on Kant. This book was a tremendous inspiration to me, both for the specific questions that it raised about the normative relation between thought and the world, and for the seamless way in which it incorporated reflection on historical figures, especially Kant, into discussion of contemporary issues. Since then, I have continued to find McDowell’s work consistently inspiring, exciting, and stimulating. I have learned a g reat deal both from critical engagement with his writings and from lively and memorable discussions with him at conferences and workshops. I am very grateful to him for the contributions he has made to my philosophical thinking over the years. 5. Kripke, Wittgenstein on Rules and Private Language, 9.
N O T E S T O PAG E S 2 1– 24
333
6. Ibid., 13 (meaning plus rather than quus); 10, 15 (“unjustified leap in the dark”). 7. Ibid., 21. 8. Ibid., 13, 21 (past, present, and f uture meaning); 19 (meaning of linguistic expressions). 9. Ibid., 15–18 (general instructions for the use of “plus”); 16, 12 (definition of “plus”); 22–37 (disposition to give the sum in answer to “plus” questions); and 41–51 (introspectable quale or mental images). 10. Ibid., 51–53. 11. Ibid., 107. 12. See McDowell, “Following a Rule,” 235, 248; McDowell, “Meaning and Intentionality,” 264. 13. See McDowell, “Meaning and Intentionality,” 265; McDowell, “How Not to Read Wittgenstein,” 100. 14. See McDowell, “Following a Rule,” 226–227; McDowell, “Meaning and Intentionality,” 264–265; McDowell, “How Not to Read Wittgenstein,” 100. 15. See McDowell, “Following a Rule,” 228–220; McDowell, “Meaning and Intentionality,” 267–268; McDowell, “How Not to Read Wittgenstein,” 101. 16. This assimilation is suggested by McDowell’s characterization of Kripke’s “sui generis” proposal as a “saving grace” of his reading of Wittgenstein (“Intentionality and Interiority in Wittgenstein,” in Mind, Value, and Reality, 298n4). 17. See McDowell, “Wittgensteinian Quietism,” 367–369. 18. Ludwig Wittgenstein, Philosophical Investigations, trans. G. E. M. Anscombe, rev. P. M. S. Hacker and Joachim Schulte (Malden, MA: Blackwell Publishing, 2009), §85, §198. McDowell discusses the signpost example in §7 of “Following a Rule,” §8 of “Meaning and Intentionality,” §§3–7 of “How Not to Read Wittgenstein,” and §2 of “Wittgensteinian Quietism.” 19. Wittgenstein, Philosophical Investigations, §85. 20. See McDowell, “Meaning and Intentionality,” 276; see also McDowell, “How Not to Read Wittgenstein,” 104. 21. Of course, as McDowell indicates at “Wittgensteinian Quietism,” 368, some signposts are ambiguous. If a signpost is poorly placed, I may need to interpret it to determine which of two possible paths I am supposed to follow. No interpretation is needed, however, to determine whether I should follow it in the direction from post to tapered end or from tapered end to post, and, as the reference to Investigations §506 at p. 276 of “Meaning and Intentionality” shows, that is the relevant aspect of the example. 22. Note that I am understanding the “use” of an expression in a way which does not presuppose the idea of the expression’s being used with a par ticular meaning, that is, in such a way that we can adequately describe the use of the sign “+” on some occasion by saying that someone responded to “what is
334
N O T E S T O PAG E S 2 5 – 2 6
68 + 57?” with “125,” as opposed to saying that she used “+” to assert that 68 plus 57 is 125. This clarification is necessary, given that McDowell invokes a semantically loaded notion of “use” on which the use of an expression is properly described by saying that it is used to express a specific thought or make a specific assertion. See, e.g., John McDowell, “In Defense of Modesty,” in Meaning, Knowledge, and Reality (Cambridge, MA: Harvard University Press, 1998), 87–107, at 99–100; “Another Plea for Modesty,” (in ibid., 108–131) at 123–124; and “Anti-Realism and the Epistemology of Understanding” (in ibid., 314–343), at 317–321. A related conception of use is invoked by Barry Stroud, for whom we can describe the use of the expression “Add 2 each time” by saying “The words ‘Add 2 each time’ are used by us to mean that two is to be added each time” (Barry Stroud, “Meaning and Understanding,” in Wittgenstein and the Philosophy of Mind, ed. Jonathan Ellis and Daniel Guevara [Oxford: Oxford University Press, 2012], 19–36, at 27). One can acknowledge the possibility of this semantically loaded notion of use compatibly with recognizing a more traditional conception of the use of an expression as characterizable without appeal to its meaning; it is the more traditional conception which I am assuming here. For more on the distinction between these two conceptions of the use of an expression, see Hannah Ginsborg, “Inside and Outside Language: Stroud’s Nonreductionism about Meaning,” in The Possibility of Philosophical Understanding: Reflections on the Thought of Barry Stroud, ed. Jason Bridges, Niko Kolodny, and Wai-hung Wong (Oxford: Oxford University Press, 2011), 147–181, at 150–154, and William Child, “Meaning, Use and Supervenience,” in Wittgenstein on Philosophy, Objectivity, and Meaning, ed. James Conant and Sebastian Sunday (Cambridge: Cambridge University Press, 2019), 211–230, at 211. 23. The reading of Kripke sketched in this and the next two paragraphs is explained and defended more fully in Hannah Ginsborg, “Leaps in the Dark: Epistemological Skepticism in Kripke’s Wittgenstein,” in Skepticism: Historical and Contemporary Inquiries, ed. G. Anthony Bruno and Abby Rutherford (New York: Routledge, 2018), 149–166. 24. Kripke repeatedly restates the challenge in this form, for example at Wittgenstein on Rules and Private Language, 11, 13, 15, 21, 37, 40 and 42. 25. A charge levelled by many commentators on Kripke, beginning with Colin McGinn, Wittgenstein on Meaning (Oxford: Blackwell, 1984), 150ff. For more on the charge, see Ginsborg, “Leaps in the Dark.” 26. Kripke, Wittgenstein on Rules and Private Language, 22. 27. Is this move tacitly verificationist? I do not think so. We might suspect verificationism if, as on some readings of Kripke, the argument moved from my not knowing that I meant, or mean, addition, to t here not being a fact that I meant, or mean, addition. (For examples of such readings, see Crispin Wright, “Kripke’s Account of the Argument against Private Language, Journal of Philosophy 71, no. 12 [1984]: 759–78, at 761–762, and Alexander Miller,
N O T E S T O PAG E S 2 9– 3 2
335
Philosophy of Language [New York: Routledge, 2018], at 200.) However the move I am making here has the nonfactuality of meaning be a consequence not of my failure to know that I meant addition, but of my failure to know, on any given occasion of use, what I ought to say in light of past uses of “+.” It assumes that part of what it is for me to mean something by an expression is for me to be in a position to know that I am going on as I ought in my use of the expression. This does imply a dependence of a certain metaphysical question (is there a fact of my meaning addition by “+”?) on an epistemological question (do I know that I am now using “+” appropriately?), but not in the verificationist sense in which the obtaining of a fact depends on my knowing that same fact to obtain. I am grateful to an anonymous reviewer for raising this question. 28. I say more about the idea of “going on” in Wittgenstein, and its connection with language-learning, in Hannah Ginsborg, “Wittgenstein on G oing On,” Canadian Journal of Philosophy 50, no. 1 (2020): 1–17. 29. Here I go substantially beyond Kripke, since he makes only passing references to language-learning in the course of developing the argument, and I am suggesting that it is central to the line of thought motivating the skeptical paradox. (The references that he makes include a characterization of Wittgenstein as having apparently shown that “all language, all concept formation” is impossible, where the reference to concept “formation” suggests that it is learning which is at issue, and passages at 59n45, 59n46, and 117 which hint that the problem is connected with language-learning.) If this suggestion is thought to be insufficiently grounded, then my issue with McDowell’s quietism can be recast in terms of its failure to engage the skeptical argument I have described, whether or not it is Kripke’s. 30. McDowell, “Following a Rule,” 253. 31. What if, as in fact happened in some parts of England during the Second World War, signposts are turned around to confuse potential invaders? In that case, we may “follow” a right-tapering signpost by g oing to the left, and even take this response to be (in one sense) normatively called for by the signpost, but we w ill still continue to see it as pointing to the right, and indeed will think of ourselves (in a more fundamental sense) as following the signpost in the wrong direction. 32. Apparently Polish truck drivers accompanying less experienced colleagues on a first trip to Germany like to tease them by pretending that “Ausfahrt” (the German term for a highway exit) is the name of a very large town (Magdalena Bartłomiejczyk, “Text and Image in Traffic Signs,” Linguistica Silesiana 34 [2013]: 111). It is hard to imagine, by contrast, that a driver could convince even the most gullible colleague that right-pointing arrow signs in Germany mean “Go to the left.” 33. “The temptation to start on [the regress], and its disastrous consequences, are the same w hether we are considering non-d iscursive
336
N O T E S T O PAG E 3 2
expressions of rules, such as sign-posts, or discursive expressions, such as—for the same rule—someone saying ‘To follow the trail at this point you must go to the right’ ” (McDowell, “How Not to Read Wittgenstein,” 100). Since this claim is aimed against Brandom’s view that signposts exhibit a kind of normativity which is more fundamental than that associated with the explicit use of linguistic expressions, it is worth noting h ere that while I agree with Brandom in taking the normativity associated with understanding signposts to be at a lower level than that involved in language use, I do not agree that it is to be made out in terms of social practices of giving and asking for reasons. For more on the independence of this kind of normativity from reasons, in the context of McDowell’s views, see Hannah Ginsborg, “Empiricism and Normative Constraint,” in In the Light of Experience: New Essays on Perception and Reasons, ed. Johan Gersel, Rasmus Thybo Jensen, Morten S. Thaning, and Søren Overgaard (Oxford: Oxford University Press, 2018), 101–138. 34. This was an advertising slogan for a children’s drink made by the Dutch firm Nutricia. Another example is the illuminated “I fart” (“In motion”) signs that used to be found in Danish elevators; in a possibly apocryphal story, these signs had to be covered up in the elevators Queen Elizabeth II used on a visit to Denmark. McDowell might object that t hese are not cases where, say, a linguistic expression in Dutch can be understood as bearing another meaning in English, but rather cases where a linguistic expression in Dutch can be mistaken for a diff erent linguistic expression in English which, taken in isolation, looks just like it. However, even if this is granted, the fact that we cannot make this kind of m istake for signposts is enough to undermine the analogy on which McDowell is relying in his use of the signpost example. 35. H ere I disagree with Robert Brandom, according to whom “a practice of pointing requires a great deal of social stage-setting—the untrained may be unable to transfer their attention beyond the tip of the pointing finger, or may perversely trace the line of indication in the wrong direction, from fingertip to base, and so take it that something behind the one pointing has been singled out” (Robert Brandom, Making It Explicit [Cambridge, MA: Harvard University Press, 1994], 461). There is a hint that McDowell might also take this view, since he treats pointing gestures and utterances of “Go to the right” as alike in their potential to “lapse . . . into normative inertness” (McDowell, “How Not to Read Wittgenstein,” 100), thereby suggesting that they are also alike in requiring training to be understood. But Brandom’s view strikes me as plainly false: I know of no evidence that small c hildren ever make istake that Brandom describes, and if they did, it is hard to the kind of m imagine how we could ever train them to understand pointing gestures correctly (would we point in the direction we expected them to follow the pointing gesture?). Brandom ascribes this position to Wittgenstein, but that again seems false: according to Wittgenstein “it is part of human nature to understand pointing with the finger in the way we do” (Ludwig Wittgenstein,
N O T E S T O PAG E S 3 2 – 3 8
337
Philosophical Grammar, ed. Rush Rhees, trans. Anthony Kenny [Oxford: Blackwell, 1974], §52). 36. McDowell, “Following a Rule,” 239. 37. McDowell, “How Not to Read Wittgenstein,” 101. 38. The idea that the use of an expression can accord with past uses of that expression, in a sense that does not amount to accordance with what those expressions meant, is implicit in Cavell’s discussion of learning and projecting words in ch. 7 of Stanley Cavell, The Claim of Reason (Cambridge, MA: Harvard University Press, 1979). I am grateful to an anonymous reviewer for pointing out this connection. 39. In fact he sometimes appears simply to identify conformity to past usage with conformity to past meaning, as when he moves from saying “the skeptic questions w hether my present usage accords with my past usage” to identifying the problem as that of “how . . . I know that ‘68 plus 57’, as I meant ‘plus’ in the past, should denote 125” (Kripke, Wittgenstein on Rules and Private Language, 12; Kripke’s emphasis removed). The same identification is suggested when he says “If I meant [addition], then to accord with my previous usage I should say ‘125’ when asked to give the result of calculating ‘68 plus 57’ . . . i f I meant [quaddition], I should say ‘5’ ” (ibid.) It appears h ere that accordance with my previous usage just is accordance with what I meant. Given the apparent identification of past usage with past meaning, it might be wondered whether I am right to put so much emphasis on conformity to past usage as distinct from past meaning: perhaps the skeptical problem is simply about how “125” can conform to what I meant, so that McDowell is right to construe the fundamental puzzlement as being about how a use of an expression can accord with our understanding of it, past or present, not about how it can accord with past uses. But in that case it is hard to see why I should be bothered by the skeptical challenge, and specifically why I should see it as threatening the “correctness” of my use of “125.” Why should it m atter to me whether or not I am according with my past meaning as long as I am according with what the expression means now? It is only if we assume that accordance with past meaning is necessary for me to be going on correctly from previous uses, where this in turn is required for my having mastered the meaning of the expression, that it makes sense to regard it as a form of correctness and so to suppose, as Kripke mistakenly does, that I need to defend it from skeptical attack. 40. See, for example, §3 of “Noncognitivism and Rule-Following,” in Mind, Value and Reality. 41. As helpfully pointed out by an anonymous reviewer, there has been much empirical research into small c hildren’s participation in t hese kinds of norm-governed activities. I discuss some of the relevant research, and its implications for ascribing primitively normative attitudes to prelinguistic children, in Hannah Ginsborg, “Conceptualism and the Notion of a Concept,”
338
N O T E S T O PAG E S 4 1–76
in Concepts in Thought, Action, and Emotion: New Essays, ed. Christoph Demmerling and Dirk Schröder (New York: Routledge, 2021), 42–59, and in Hannah Ginsborg, “Spontaneity without Rationality: A Kantian Approach to Self-Consciousness and Perceptual Content,” in Perceptual Knowledge and Self-Awareness, ed. Andrea Giananti, Johannes Roessler, and Gianfranco Soldati (forthcoming).
2. Forms of Nature 1. Philippa Foot, Natural Goodness (New York: Oxford University Press, 2001). 2. John McDowell, “Two Sorts of Naturalism,” in Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 167–198, 167. 3. Arthur Schopenhauer, The Two Fundamental Problems of Ethics (Oxford: Oxford University Press, 2010), 145. 4. David Wiggins, “Truth, Invention, and the Meaning of Life,” Proceedings of the British Academy 62 (1976): 331–378, 375. 5. Martin Heidegger, Plato’s Sophist (Bloomington: Indiana University Press, 1997). 6. G. E. M. Anscombe, Intention (Cambridge, MA: Harvard University Press, 1957). 7. See Michael Thompson, “Naïve Action Theory,” in Life and Action (Cambridge, MA: Harvard University Press, 2008). 8. John Haugeland, Dasein Disclosed: John Haugeland’s Heidegger, ed. Joseph Rouse (Cambridge, MA: Harvard University Press, 2013). 9. McDowell, “Two Sorts of Naturalism.” 10. Karl Marx, Economic and Philosophic Manuscripts, ed. Dirk J. Struik, trans. Martin Milligan (New York: International Publishers, 1964). 11. Cf. Allen Wood, Karl Marx (New York: Routledge, 2012), 19. 12. McDowell, “Two Sorts of Naturalism,” 167. 13. Immanuel Kant, Religion within the Boundaries of Mere Reason, trans. Allen Wood and George di Giovanni (New York: Cambridge University Press, 1996), 50–51 (6:26n in the Academy pagination). 14. Christine Korsgaard, The Sources of Normativity (New York: Cambridge University Press, 1996).
3. The Rational Role of Perceptual Content I am greatly indebted to John McDowell and Charles Travis for feedback on an earlier version of this paper. I am also indebted to comments from participants in a conference on “Intuitional Content and the Myth of the Given” held in
N O T E S T O PAG E S 8 3 – 8 6
339
Patras, Greece, and to the comments of two anonymous reviewers for Harvard University Press.
1. John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994). 2. Charles Travis, “The Silence of the Senses,” Mind 113, no. 449 (2004): 57–94, 65. 3. Thus, another prominent critic of perceptual content, Bill Brewer, suggests that the very idea of perceptual content involves “an unwarranted intrusion of conceptual thought about the world presented in perception into the . . . account of the most basic nature of perception itself.” Bill Brewer, “Perception and Content,” European Journal of Philosophy 14, no. 2 (2006): 165–181, 174. 4. McDowell, Mind and World, 10–11. 5. This, as I understand it, was the view McDowell held in 1994. In more recent work, he has retained his commitment to (MR) while revising his account of what it implies about the nature of perceptual content and its relation to specific conceptual capacities. I w ill come to these revisions in due course. 6. Thus McDowell writes that “Givenness in the sense of the Myth would be availability for cognition to subjects whose getting what is supposedly Given to them does not draw on capacities required for the sort of cognition in question” (John McDowell, “Avoiding the Myth of the Given,” in Having the World in View [Cambridge, MA: Harvard University Press, 2009], 256–274, 256). Specifically, where “the sort of cognition in question” is the cognition of a rational subject—one capable of reflecting on their own reasons for judgment—this means that what perception presents w ill count as “given” in the mythical sense if its availability for cognition does not draw on the rational capacities paradigmatically expressed in reasons-based judgment. So, in effect, we must accept (MR) to avoid the Myth of the Given. Having characterized Givenness in the way just quoted, McDowell remarks: “If that is what Givenness would be, it is straightforward that it must be mythical. Having something Given to one would be being given something for knowledge without needing to have capacities that would be necessary for one to be able to get to know it” (ibid.). I do not understand this argument. I can see that, if it is admitted that certain capacities are required for any cognition of a given sort, it would be incoherent to hold that t here can be cases of that sort of cognition that do not draw on those capacities. But it is not obviously incoherent to hold that something might be given for cognition—not actually cognized, but made available as a ground on which cognition can be based—w ithout the involvement of those capacities. W hether this is possible is simply the point in question, and I cannot see how an abstract argument of the kind McDowell sketches here could settle the matter. I do, however, believe
340
N O T E S T O PAG E S 8 6 – 8 8
that McDowell has other, more compelling arguments for (MR). I w ill come to these shortly. 7. McDowell, “Avoiding the Myth of the Given,” 268. 8. I w ill set aside other motivations for attributing content to perception, such as the aim of explaining how perceptual experience can mislead us. I believe that critics of the notion of perceptual content have shown that it is pos sible to account for perceptual illusion and hallucination without appealing to this notion (see, e.g., Bill Brewer, Perception and Its Objects [Oxford: Oxford University Press, 2011], ch. 5, and Charles Travis, “Unlocking the Outer World,” in Perception: Essays after Frege [Oxford: Oxford University Press, 2013], 223– 258), but I will not defend this position here. Taking my cue from McDowell, I will focus exclusively on the question whether we must appeal to perceptual content to account for how perception can put a subject in a position to make true and reasonable judgments about their environment. 9. Immanuel Kant, Critique of Pure Reason, trans. Paul Guyer and Allen W. Wood (Cambridge: Cambridge University Press, 1997), A51 / B75; A79 / B105. The German word that I translate as “presentation” is Vorstellung, standardly rendered as “represent ation.” “Present ation” is, however, etymologically defensible and not implausible as a rendering of the term (compare Werner Pluhar’s translation of the first Critique (Immanuel Kant, Critique of Pure Reason, trans. Werner S. Pluhar [Indianapolis: Hackett Publishing Company, 1996], 22n73). I prefer “present ation” here because it does not prejudge the question whether having a Vorstellung of an object involves having some representation of that object as being a certain way. Kant classifies intuitions (Anschauungen) as Vorstellungen (see A320 / B376–377), but I s hall argue below that they are not representations in the latter sense. 10. It is import ant to emphasize that t hese are claims about our capacities for sense-perception. McDowell does not deny that nonrational animals can learn about objects through their senses without drawing on rational capacities for understanding the world, but he holds that, given that we are rational animals, our sensing must take a special form. 11. McDowell, Mind and World, 11. 12. Ibid., 9. 13. Ibid., 52. 14. Ibid., 108–126. Note that this is again a claim about the conditions under which we rational animals can have perceptual experiences of objects and their properties. It would be very implausible to suggest that nonrational animals merely undergo a flux of subjective states without objective significance. But this is not McDowell’s claim: his idea is that the object-directedness of our sensory awareness is mediated by a rational understanding of the nature of objectivity. He holds the analogous but diff erent object-directedness of nonrational sensory awareness to be mediated by a nonrational analogue of such understanding.
N O T E S T O PAG E S 8 9– 9 3
341
15. Ibid., 12. 16. Thus, in the paragraph that introduces the discussion of color judgments, McDowell writes that “the capacities that are drawn on in experience are recognizable as conceptual only against the background of the fact that someone who has them is responsive to rational relations, which link the contents of judgments of experience with other judgable contents. These linkages give the concepts their place in a possible view of the world” (ibid., 11–12). H ere the topic is conceptual capacities drawn on not just in observational judgments, but in experience itself. 17. McDowell does not himself treat this line of thought as a distinct argument for (MR). In Mind and World, he presents it simply as a way of emphasizing the variety and complexity of the conceptual capacities drawn on in perceptual experience, having already defended (MR) via the Rationality Argument. But the Objectivity Argument seems to me to be relevant, not just to the question of the extensiveness of the rational capacities drawn on in perception, but to the motivation for and interpretation of (MR) itself. 18. The debts of the following discussion to the discussions of Travis and Brewer should be clear to anyone who knows their work. A full treatment of the question w hether perception has representational content would need to engage with important recent responses to the nonrepresentationalist view, notably Susanna Siegel, The Contents of Perceptual Experience (Oxford: Oxford University Press, 2010), and Susanna Schellenberg, The Unity of Perception: Content, Consciousness, Evidence (Oxford: Oxford University Press, 2018). Since the present discussion is oriented specifically t oward McDowell’s work, I focus my discussion exclusively on his way of developing and defending representationalism. 19. For an example of the former suggestion, see Robert Stalnaker, “What Might Conceptual Content Be?,” in Essays on Non-Conceptual Content, ed. York H. Gunther (Cambridge, MA: MIT Press, 2003), 95–106. For an example of the latter, see Christopher Peacocke, “Scenarios, Concepts, and Perception,” in Essays on Non-Conceptual Content, 107–132. 20. McDowell, “Avoiding the Myth of the Given,” 262ff. 21. For this way of representing the content of an intuition, see Wilfrid Sellars, “Some Remarks on Kant’s Theory of Experience,” Journal of Philosophy 64, no. 20 (1967): 633–647, and Wilfrid Sellars, Science and Metaphysics: Variations on Kantian Themes (London: Routledge, 1968). 22. McDowell, “Avoiding the Myth of the Given,” 270. 23. In presenting his revised view of perceptual content, McDowell writes that “in a visual intuition, an object is visually present to a subject with those of its features that are visible to the subject from her vantage point. It is through the presence of those features that the object is present” (ibid., 265). He also writes that “experience directly reveals t hings to be as they are
342
N O T E S T O PAG E S 9 3 –1 0 0
believed to be in perceptual beliefs” (ibid., 270). The talk of the presence of “features” in the former remark, and the language of “revealing t hings to be as . . .” in the latter, suggest that McDowell continues to think of a perceptual present ation of an object as involving a present ation of it as having certain features. To this extent, his revised view remains representationalist. 24. My development of this challenge owes a debt to Brewer’s discussion of the Müller-Lyer Illusion in his 2006, though I think the points that Brewer makes about this case stand out more clearly when they are made about more ordinary cases of perception. 25. For examples, see McDowell, “Avoiding the Myth of the Given,” 260, and also McDowell, “Conceptual Capacities in Perception,” in his Having the World in View, 127–144. 26. One thing that might make the representationalist view tempting is the observation that I can recall the features of something I formerly perceived, as I do, for instance, when I assist a forensic sketch artist by recalling the facial features of a person who assaulted me. Doesn’t this imply that in perceiving my assailant’s face, I registered some of its features? In fact this sort of example lends no additional support to the representationalist view. In recalling what I formerly perceived, I recall my assailant’s face—v ividly or hazily—a nd answer questions on the basis of this recollection. This need not imply that the face was originally presented to me as having certain specific features. It suffices that the face itself was presented to me, and did in fact have such features. Memory can simply put me back into a relation with the face itself on the basis of which I can judge of its features here and now. 27. For the appeal, see McDowell, Mind and World, 56ff, and McDowell, “Avoiding the Myth of the Given,” 263, 270. McDowell was not himself appealing to demonstratives to save the idea of perceptual content: both he and his opponents took it for granted that perceptual experience has some sort of content; their dispute was about whether this content must be conceptual or nonconceptual. But the sorts of doubts about conceptual content to which McDowell was responding closely parallel the doubts we have canvassed about representational content. 28. An object can of course appear to have a color it does not have. In this sort of case, indicating what apparent color is meant by “that shade of color” will be more complicated: it might, for instance, require indicating what swatch of color on a palette would match the object’s appearance. Still, the phrase “that shade of color” designates the color of some actual t hing; it does not itself express a classification of the color in question. 29. This paragraph responds to a question raised by an anonymous reviewer. 30. Kant, Critique of Pure Reason, trans. Guyer and Wood, A79 / B105. 31. Ibid., A294 / B350. 32. In a recorded lecture called “Are the Senses Silent?,” McDowell suggests that, in the passage I am discussing, what Kant meant was that the
N O T E S T O PAG E S 1 0 0 –1 0 3
343
senses by themselves do not present us with content (see http://w ww .youtube.com/watch?v =f BQHEGg5JSo, at the 3:02 mark). Kant’s real point, McDowell proposes, is that the senses present us with content only insofar as their operations are informed by functions grounded in the understanding—so while the senses by themselves do not present us with truth and semblance, they do present us truth and semblance when they are thus informed. I have two objections to this reading. First, Kant does not include any qualification like McDowell’s “by themselves” in his statement of the claim: he simply says that truth and semblance are not found in what the senses pre sent. Secondly, and more tellingly, he is willing to restate the point in his own technical vocabulary: “truth and semblance are not in the object insofar as it is intuited, but in the judgment about it insofar as it is thought.” This remark does not appear to leave room for the idea that empirical intuitions present truth and semblance (in virtue of being not mere sensible presentations, but sensible presentations informed by functions of the understanding). Kant says explicitly that truth and semblance are not in the object qua intuited (I take that to mean they are not present in empirical intuition), but come on the scene only with thought about an object. 33. Kant, Critique of Pure Reason, trans. Guyer and Wood, A58 / B82. 34. Ibid., A68 / B93. 35. Ibid., A320 / B377. 36. This conception of Kantian “empirical intuitions” is closely related to the one developed in Lucy Allais, Manifest Reality: Kant’s Idealism and His Realism (Oxford: Oxford University Press, 2015). I have been greatly helped by Allais’s work, though I disagree with her about how to conceive of the relation of categorial understanding to empirical intuition. I w ill not attempt to detail t hese differences h ere: I hope to discuss Allais’s views more fully in f uture work. 37. Kant, Critique of Pure Reason, trans. Guyer and Wood, A571–572 / B599–600. 38. See also Ibid., A106, A722 / B750n; A68 / B93, cf. JL 9:93. 39. See also McDowell, “Avoiding the Myth of the Given,” 263. 40. See also Immanuel Kant, “Jaesche Logic,” in Lectures on Logic, trans. J. Michael Young (Cambridge: Camridge University Press, 1992), 9:94–95. 41. McDowell denies that he conceives of the relation between perception and judgment as inferential, since on his view, the content of perception is—in the simplest case, anyway—identical to the content judged true (McDowell, “Conceptual Capacities in Perception,” 131). This is a terminological issue: even if we reserve the term “inference” for nontrivial steps from one content to another content, the substantive comparison between the rationality of inference and the rationality of perceptual judgment (as McDowell conceives it) stands.
344
N O T E S T O PAG E S 1 0 3 –1 1 8
42. See also Charles Travis, “Reason’s Reach,” European Journal of Philosophy 15, no. 2 (2007): 225–248, §3, and Brewer, Perception and Its Objects, 157. 43. See also Kant, Critique of Pure Reason, trans. Guyer and Wood, A132 / B171, A247 / B304. 44. Ibid., A132–133 / B171–172. 45. See also ibid., A133–4 / B172–3. 46. Ibid., A79 / B104–105. 47. Ibid., A68 / B93. 48. Ibid., A79–80 / B105. 49. See also ibid., B144–145. 50. For the idea that categories are distinctively formal concepts, see for instance ibid., A129–30, A245, B288, A253 / B309, A310 / B367. 51. Ibid., A79 / B104–105. 52. See also ibid., B128, B158. 53. For diff erent kinds of opposition to McDowell’s Objectivity Argument, see for instance Tyler Burge, “Perceptual Objectivity,” Philosophical Review 118, no. 3 (2009): 285–324, and Charles Travis, “Unlocking the Outer World,” in Perception: Essays a fter Frege (Oxford: Oxford University Press, 2013), 223–258. 54. McDowell, “Avoiding the Myth of the Given,” 263. 55. Ibid. 56. Ibid., 261. 57. Ibid., 264.
4. Resolute Disjunctivism 1. John McDowell, “Singular Thought and the Extent of Inner Space,” in Subject, Thought, and Context, ed. Philip Petit (New York: Clarendon Press: 1986): 236–237. 2. Ibid. 3. Ibid, 240–241. 4. Ibid, 242. 5. Ibid. 6. For a helpful extended discussion of this point, see Thomas Lockhart, “Motivating Disjunctivism,” in Rethinking Epistemology, vol. 2, ed. Günter Abel and James Conant (Berlin: De Gruyter, 2012): 309–347. 7. John McDowell, “Tyler Burge on Disjunctivism II,” Philosophical Explorations: An International Journal for the Philosophy of Action 16, no. 3 (September 2013): 245–246. 8. The central argument presented in the remaining portions of this paper occurs somewhat differently arranged, and developed in far greater
N O T E S T O PAG E S 1 1 9–1 2 8
345
detail, in part II, section IX of Sofia Miguens, ed., The Logical Alien: Conant and His Critics (Cambridge, MA: Harvard University Press: 2020). The pre sentation of the argument there also draws more explicitly on philosophical distinctions and forms of terminology which, though also employed here, are laid out more painstakingly in e arlier stretches of part II of that book, especially in section VIII and in the first half of section IX. 9. A less compressed discussion of the analogy between the form of the capacity to shoot a free throw and that of seeing what is so may be found in Miguens, The Logical Alien, 634–636. 10. When one “sees a hallucination” of an oasis sufficiently vividly as to be misled by it momentarily, then one does not see a merely m ental (rather than a worldly) “something”; rather, one merely seemingly sees an oasis. 11. Gilbert Ryle, The Concept of Mind (New York: Routledge, 2009), 114. 12. J. L. Austin, “A Plea for Excuses: The Presidential Address,” Proceedings of the Aristotelian Society 57, no. 1 (June 1957): 1–30; 16. For a more detailed discussion of how (and how not) to understand Austin’s point h ere, see James Conant, “Three Ways of Inheriting Austin,” in La philosophie du langage ordinaire: Histoire et actualité de la philosophie d’Oxford / Ordinary Language Philosophy: The History and Contemporary Relevance of Oxford Philosophy, ed. Christoph Al-Saleh and Sandra Laugier (Hildesheim: Olms Verlag, 2011), 395–415. 13. The same could be said, for example, about a w hole range of coordinate expressions (such as “in no way attenuated” or “nonproblematic” or “logically full-blooded”) as they occur in t hese pages. For a more careful discussion of these expressions, and the senses in which I seek to employ them, as well as of what I mean h ere by “logical privation,” see Part II, section VIII of Miguens, The Logical Alien. 14. Just as the nonsuccess use of the verb push is logically parasitic on the success sense of the verb (where pushing x does require that x moves), and as the nonsuccess use of the verb tell is logically parasitic on the success sense of the verb (where telling S that p requires that S comes to be informed that p), e tc. 15. Barry Stroud, “Sense-Experience and the Grounding of Thought,” in Philosophers Past and Present (Oxford: Oxford University Press, 2011), 273– 289. This article of Stroud’s is specifically concerned to advance a criticism of the conception of perceptual experience advocated by McDowell in John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994). 16. John McDowell, “Reply to Barry Stroud,” in Reading McDowell, ed. Nicholas H. Smith (New York: Routledge, 2002), 277–278. 17. For the original example, see Wilfrid Sellars, Empiricism and the Philosophy of Mind (Cambridge, MA: Harvard University Press, 1956), 37–39. 18. McDowell, “Reply to Barry Stroud,” 277–278.
346
N O T E S T O PAG E S 1 2 8 –1 3 0
19. I explore briefly below in this paper what is required to conjugate this verb respectively in the third-person present and in the first-person past. For a far more detailed exploration of these forms of use of the verb, including an exploration of what is involved in a first-person subjunctive form of use of this verb, see Miguens, The Logical Alien, 708–730. 20. This is a characteristic mark of a rupture or defect in self- consciousness—a mark that characterizes a whole spectrum of cases of self- alienation, e.g., cases in which another person might rightly ascribe anger, fear, or desire to me that I am unable to express in a self-ascription. According to David Finkelstein, this kind of inability to express one’s own state of mind— an inability to express it by self-ascribing it—distinguishes fully self-conscious states of mind from a range of other cases. He initially develops this point in connection with elucidating the difference between (what he calls) conscious and unconscious states of mind (see, e.g., David Finkelstein, Expression and the Inner [Cambridge, MA: Harvard University Press, 2003], and David Finkelstein, “Making the Unconscious Conscious,” in The Oxford Handbook of Philosophy and Psychoanalysis, ed. R. Gipps and M. Lacewing [Oxford: Oxford University Press, 2019], 331–346). We should, he argues, understand the distinction between conscious and unconscious states of mind in terms of a claim that may be put (roughly) as follows: someone’s state of mind is conscious just in case she is able to express it by ascribing it to herself. In forthcoming work, Finkelstein focuses on how the connection between consciousness and expression should bear on our thinking about perceptual knowledge and, more specifically, on disjunctivist thinking about perceptual knowledge. In this endeavor, he defends a view according to which: (1) a person’s perceptual knowledge that p is nondefective only when it is fully conscious, and (2) its being fully conscious is a m atter of the person’s being able to express it by self-ascribing it, e.g., by saying, “I see that p,” or, “I saw that p.” 21. This is something we do all the time as natural language speakers. We form logically secondary concepts by subtly removing or modifying one or more of the logico-g rammatical conditions on the use of its logically primary counterpart. This leads to a great many verbal twins who are not full- blooded logical twins. One of the g reat sources of philosophical confusion is that we then to try to understand the logical grammar of the primary case on the assumption that one of the secondary uses of the concept must be logically more basic. We start our investigation with a concept such as (mere) “bodily movement” or (dead) “sign,” etc., and then trying to add a further factor to work our way up to an understanding of the concept that occasions philosophical perplexity in us: intentional bodily movement, the meaningful use of a sign, e tc. 22. The paradox came to public attention through Wittgenstein’s discussions of it. The closest thing we have to something by Moore himself dis-
N O T E S T O PAG E S 1 3 0 –14 3
347
cussing it (albeit at this point in response to Wittgenstein’s comments about it) is the piece, tentatively dated 1944, titled “Moore’s Paradox” (in G. E. Moore: Selected Writings, ed. T. Baldwin [Abingdon: Routledge, 1993], 207–212), which comes from an untitled and incomplete manuscript from the Moore papers in the University Library, Cambridge, and which Moore never prepared for publication. 23. That the verbal twins “I saw p” (used in accordance with the idiom of the parable’s protagonist) and “I saw p” (used in accordance with the language game of the full-blooded verb of success) resemble one another as much as they do is certainly no accident. For they are both members of the same family of cognitive concepts—whose other siblings (“I see my reflection in the mirror”), whose distant cousins (“The cat sees p”), not to mention the descendants of their siblings (“I see what you mean”), and those of their cousins (“The cat sees this means trouble”) bear a range of crisscrossing resemblances to one another. But that does not alter the point that matters for our present purpose: our verbal twins are not logical twins. 24. I owe this objection to an anonymous referee. I have reproduced the objection as it was stated. But I have responded to it on the assumption that r is supposed to stand h ere for a form of activity that can be done self- consciously. It is certainly right that if the capacity whose exercise r ascribes to the subject is not such a capacity, then the following w ill never be paradoxical: “I r-ed but at the time I d idn’t know I was r-ing.” (Consider: “I bled but at the time I d idn’t know I was bleeding.”) Moore-paradoxicality can so much as arise in connection with an act of r-ing, only if r-ing is the kind of t hing that can be done self-consciously—a nd hence can also involve privation along this dimension through the manner in which one so does it. If r in the objector’s schema is supposed to be neutral not only with respect to whether the form of the exercise of the capacity is a self-conscious one, but also with re hether r is the act of a rational or a nonrational capacity, then that spect to w introduces the possibility of a whole further range of confusions I do not seek to disentangle here. I focus below solely on the sort of confusion that arises if the objector wishes his use of r to be neutral with regard to the degree of self-a lienation the act of r-ing in question might involve. 25. G. E. M. Anscombe, Intention (Oxford: Blackwell,1957). 26. The infelicity can be due to a defective exercise of the capacity (i.e., the case that clearly falls into McDowell’s conception of what belongs in the bad disjunct). The defect, in turn, can be traceable to infelicity in e ither the material or formal conditions. It can be due to a suspension of (rather than an error in) judgment. Such a problematic act of judgment might arise from a concern about the adequacy of either the material or the reflective conditions u nder which one seeks to vindicate the candidate for judgment. But it may simply be due to insufficient evidence. Then the manner in which the capacity of judgment is in act w ill have to do with the manner in which one seeks to
348
N O T E S T O PAG E S 14 5 –1 5 1
move from belief to knowledge. It would be no less a fool’s errand to attempt to convert the foregoing cursory list into a complete inventory of possible disjuncts than it would be to attempt a complete specification of all of the pos sible ways in which a human being can fall short of the condition of perfect health. 27. For a more detailed discussion of what the resolute disjunctivist ought to say about the case of a perceiving subject who is initially reflectively unfavorably circumstanced and the implications of the proper treatment of such a case for an understanding of the concept of indefeasible warrant, see Miguens, The Logical Alien, 741–749. 28. For further discussion of what I mean here by “Kantian trans cendental-logical form,” how such a conception of logical form differs from what we might call a broadly Fregean conception of logical form, why it presupposes a non-Fregean conception of logical generality, and how such a conception rests a properly Kantian account of the nexus of capacity and act, see Miguens, The Logical Alien, 602–611. 29. McDowell writes, “In experience one takes in, for instance sees, that things are thus and so. That is the sort of thing one can also, for instance, judge” (McDowell, Mind and World, 9). My point may be put as follows: If what is h ere called the “seeing” requires a further act of judgment—so that, as McDowell puts it, one not only sees but also judges—in order to amount to knowledge, then the case u nder consideration is one in which “seeing” and knowing do not form the requisite unity to qualify as a fully felicitous exercise of my perceptual capacity. There are cases—such as the insufficiency-of-the- perceptual-act case—in which, in order to arrive at knowledge, I must first look and then also judge. But this division of labor across two distinct acts (one of looking and one of judging) should not be read back into the structure of the fully spontaneously virtuous disjunct. 30. I cannot imagine that Stroud thinks that this sentence can serve as a summary of that conception—that he thinks that McDowell himself supposes that the content of a perceptual experience, even when it belongs to the happy disjunct, implies nothing about the independent world. For McDowell to think this would require that he completely have lost track of what was right in his original disjunctivism. So I take it that the structure of Stroud’s argument against McDowell is supposed to have the form of a dilemma—only one horn of which is concerned with pressing the following worry: If we conceive of the perceptual content presented to us in experience as fully divorced from the exercise of our capacity for judgment, then we fall back into the Cartesian conception. This then allows for the second horn of the dilemma to be an argument along the lines I sketched above—regarding whether even a tiny gap, as it were, may be allowed to open up between the judgable content and the judgment in our characterization of a case without
N O T E S T O PAG E S 1 5 1–1 5 2
349
its thereby falling into the infelicitous disjunct. The point of the second horn of the dilemma is to illuminate the character of the logically most basic case of (what Stroud calls) seeing what is so. 31. Stroud, “Sense-Experience and the Grounding of Thought,” 88–89. 32. To fully think through the requirements of such a conception, however, requires not only rethinking the account of the relation between perception and judgment with which McDowell works, but also that of the relation between thought and judgment. A fundamental obstacle to McDowell’s embracing any version of resolute disjunctivism in the philosophy of perception is his commitment to a Fregean version of conjunctivism in philosophical logic. On Frege’s account, there is a logical act (and a correlatively shared content) that is a highest common f actor between thinking p (i.e., the act of merely grasping the content of p) and judging p (the transition to a logically—though not necessarily temporally—distinct act of acknowledging the truth or falsity of p). A clear-sighted commitment to this form of conjunctivism (regarding the relation of thought to judgment) in philosophical logic places an indefeasible limit on how deeply into the structure of the mind one’s disjunctivism in the philosophy of perception may be permitted to penetrate. Conversely, a commitment to the version of disjunctivism in the philosophy of perception recommended here requires that one develop a conception of judgment that does not presuppose the force / content distinction in the form that it has been bequeathed to us in the philosophical logic of Frege and Geach. It requires a philosophical logic that admits of a parallel form of disjunctivism in the philosophy of judgment: one in which the case of the logically fundamental exercise of our capacity for judgment (in which, in and through thinking, I judge) stands to the case of the comparatively attenuated exercise of that same general capacity (in which I withhold a verdict on the truth of what I think) as fully felicitous to less than fully felicitous disjunct. For a fuller discussion of these topics I refer the reader to sections VIII and following in part II of Miguens, The Logical Alien. In those sections I further explore the differences between McDowell and Stroud in connection with these topics. McDowell clearly sees the impingement of these issues in the philosophy of perception and in the philosophy of judgment on one another. He takes this to be a reason not to allow one’s version of disjunctivism in the philosophy of perception to run into conflict with the truth of Fregeanism (regarding the relation of thought to judgment) in philosophical logic. Stroud does not appear to appreciate the impingement of these issues on one another. This makes it easier for him to endorse a version of resolute disjunctivism in the philosophy of perception, even though—as I attempt to demonstrate in The Logical Alien— he remains sympathetic to the counterpart form of conjunctivism in his philosophy of logic.
350
N O T E S T O PAG E S 1 5 3 –1 5 5
5. Control and Knowledge in Action I thank Evgenia and Matt, the editors of this volume, for feedback and support. I also thank two anonymous referees for their comments.
1. E.g., John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994); John McDowell, “Avoiding the Myth of the Given,” in Having the World in View: Essays on Kant, Hegel and Sellars (Cambridge, MA: Harvard University Press, 2009), 256–274; and John McDowell, Perception as a Capacity for Knowledge (Milwaukee, WI: Marquette University Press, 2011). 2. John McDowell, “Some Remarks on Intention in Action,” The Amherst Lecture in Philosophy 6 (2011): 1–18, 17. 3. David Velleman, “The Way of the Wanton,” in Practical Identity and Narrative Agency, ed. Catriona Mackenzie and Kim Atkins (New York: Routledge, 2008), 169–193. 4. In what sense is reflective consciousness the distinguishing mark of humanity? While most human adults are self-conscious in the relevant sense, not all human agents are: for example, very young children and adults with advanced dementia are not. So, the “distinguishing mark” of humanity is not a necessary condition of humanity (and not a sufficient one e ither, since we cannot rule out a priori the possibility of nonhuman reflective agents). Nonetheless, Velleman is surely right that paradigmatic instances of human agency do exhibit self-consciousness, of a degree and possibly a kind we have no evidence of elsewhere. 5. Harry Frankfurt, “The Problem of Action,” American Philosophical Quarterly 15, no. 2 (1978): 157–162, 161–162. 6. Xabier E. Barandiaran, Ezequiel Di Paolo, and Marieke Rohde, “Defining Agency: Individuality, Normativity, Asymmetry, and Spatio- Temporality in Action,” Adaptive Behavior 17, no. 5 (2009): 367–386. For some prominent recent examples of philosophers (other than Frankfurt and Velleman) making this case, see Christine Korsgaard, Self- Constitution: Agency, Identity, and Integrity (New York: Oxford University Press, 2009); Helen Steward, A Metaphysics for Freedom (New York: Oxford University Press, 2012); and John Hyman, Action, Knowledge and Will (Oxford: Oxford University Press, 2015). Frankfurt speaks of actions in this sense as “intentional movements,” but he does not take this to imply “self-conscious assent” (Frankfurt, “The Problem of Action,” 159). Aristotle in the Nichomachean Ethics famously thought that nonhuman animals and infants could act “voluntarily,” but not from “choice” (NE 3.1). Barandiaran, Di Paolo, and Rohde attempt to give an analysis of a notion of agency that could apply to complex self-sustaining systems in general, rather than just biological organisms. 7. Velleman, “The Way of the Wanton,” 176–178.
N O T E S T O PAG E S 1 5 5 –1 5 7
351
8. Steward, A Metaphysics for Freedom, 32–33; Frankfurt, “The Problem of Action,” 159. There may be actions that do not involve bodily movements, such as a doctor’s killing a patient by purposely withholding life- saving treatment. These cases would require separate discussion. I also set aside mental actions such as performing arithmetic in one’s head, although— given that agents are material beings—they too involve bodily happenings. 9. Rowland Stout, “Ballistic Action,” in Process Action and Experience, ed. Rowland Stout (Oxford: Oxford University Press, 2018), 210–228. A potential complication concerns “ballistic” actions, that is, actions that a fter an initial segment seem to proceed with little or no input from their agent. Consider an agent swinging a golf club. Intuitively, the entire movement would count as an action of the agent’s. And yet, once the downward portion of the swing has begun, the agent seems to have little ability to stop or significantly alter the rest of the movement. So, Frankfurt’s criterion is not quite right as it stands. Nonetheless, such cases do not undermine the broader point, which is that agency is constitutively linked with control: this can be true even if not every part of every action manifests control. 10. Frankfurt, “The Problem of Action,” 160. 11. Ibid., 177. 12. Ibid., 178. 13. Ibid., 179. 14. There is empirical work that suggests that consciously attending to one’s actions interferes with performance, at least in certain settings. See, for example, Sian Beilock, Thomas H. Carr, Clare MacMahon, and Janet L. Starkes, “When Paying Attention Becomes Counterproductive: Impact of Divided versus Skill-Focused Attention on Novice and Experienced Performance of Sensorimotor Skills,” Journal of Experimental Psychology 8, no. 1 (2002): 6-16; Sian Beilock and Rob Gray, “From Attentional Control to Attentional Spillover: A Skill-Level Investigation of Attention, Movement, and Performance Outcomes,” Human Movement Science 31, no. 6 (2012): 1473–1499. This research, however, tends to focus on exceedingly simple tasks (e.g., single golf putts) in the lab. For this reason, it is consistent with maintaining that actively directed attention has a big role to play in the performance of more complex tasks where questions of strategy and the sequencing of subtasks are relevant, especially outside the lab. See Barbara Montero, “Does Bodily Awareness Interfere with Highly Skilled Movement?,” Inquiry 53, no. 2 (2010): 105–122; Barbara Montero, Thought in Action: Expertise and the Conscious Mind (Oxford: Oxford University Press, 2016); Wayne Christensen, John Sutton, and Doris McIlwain, “Putting Pressure on Theories of Choking: Towards an Expanded Perspective on Breakdown in Skilled Performance,” Phenomenology and the Cognitive Sciences 14, no. 2 (2015): 253–293; Wayne Christensen, John Sutton, and Doris McIlwain, “Cognition in Skilled Action: Meshed Control and the Varieties of Skill Experience,” Mind and Language
352
N O T E S T O PA G E S 1 5 7–1 6 0
31, no. 1 (2016): 37–66; David Papineau, “Choking and the Yips,” Phenomenology and the Cognitive Sciences 14, no. 2 (2015): 295–308. 15. Velleman, “The Way of the Wanton,” 176–178. 16. Ibid., 180. 17. Jerry Fodor, “The Appeal to Tacit Knowledge in Psychological Explanation,” Journal of Philosophy 65 (1968): 627–640, 627; Jason Stanley and Timothy Williamson, “Skill,” Nous 51, no. 4 (2017): 713–726, 718. 18. Steward, A Metaphysics for Freedom, 51. 19. Fred Dretske, Knowledge and the Flow of Information (Stanford, CA: Center for the Study of Language and Information, 1983); Berent Enç, How We Act: Causes, Reasons, and Intentions (New York: Oxford University Press, 2006); Papineau, “Choking and the Yips.” 20. Part of its popularity, no doubt, has to do with the prominence of the Causal Theory of Action as a metaphysics of agency. But, as examples like Steward’s show, this cannot be the whole story. See Steward, A Metaphysics for Freedom. 21. Michael Brownstein and Eliot Michaelson, “Doing without Believing: Intellectualism, Knowledge-How, and Belief-Attribution,” Synthese 193, no. 9 (2016): 2815–2836. 22. E.g., Pierre Fourneret and Marc Jeannerod, “Limited Conscious Monitoring of Motor Performance in Normal Subjects,” Neuropsychologia 36, no. 11 (1998): 1133–1140; and Anthony Marcel, “The Sense of Agency: Awareness and Ownership of Action,” in Agency and Self-Awareness: Issues in Philosophy and Psychology, ed. Johannes Roessler and Naomi Eilan (New York: Oxford University Press, 2003), 48–93. 23. I respond to this line of argument in Markos Valaris, “Knowing What You Are Doing: Action-Demonstratives in Unreflective Action,” Ratio 33, no. 2 (2020): 97–105. 24. Harry Frankfurt, “Freedom of the W ill and the Concept of a Person,” in The Importance of What We Care About: Philosophical Essays (Cambridge: Cambridge University Press, 1988), 11–28. 25. Edward Slingerland, Effortless Action: Wu-Wei as Conceptual and Spiritual Ideal in Early China (New York: Oxford University Press, 2003); Philip Ivanhoe and Bryan van Norden, eds., Readings in Classical Chinese Philosophy (New York: Seven Bridges Press, 2001), 219–220. Velleman draws on the much-discussed Daoist notion of wu-wei. Wu-wei is literally translatable as “nonaction,” but Velleman follows the common practice of interpreting it as a kind of highly skilled action. 26. As Velleman points out, the Zhuangzi’s ideal has ethical aspects that are not directly relevant to the present discussion. For more on self- regulation, see Velleman, “The Way of the Wanton,” 188. For reflective self- consciousness, see ibid., 186.
N O T E S T O PAG E S 1 6 0 –1 6 2
353
27. Ivanhoe and van Norden, Readings in Classical Chinese Philosophy, 238–239. 28. Velleman, “The Way of the Wanton,” 187–188. 29. It is worth noting here that this picture of skilled action, as normally excluding reflective consciousness or even thought but calling upon it in training and when the going gets tough, is far from peculiar to Velleman. See, for instance, Hubert Dreyfus, “Intelligence without Representation: The Relevance of Phenomenology to Scientific Explanation,” Phenomenology and the Cognitive Sciences 1, no. 4 (2002): 367–383; Hubert Dreyfus, “The Return of the Myth of the Mental,” Inquiry 50, no. 4 (2007): 352–365; Hubert Dreyfus, “The Myth of the Pervasiveness of the Mental,” in Mind, Reason, and Being-in-the-World: The McDowell-Dreyfus Debate, ed. Joseph K. Schear (Abingdon: Routledge, 2013), 15–40; Papineau, “Choking and the Yips.” 30. Cf. Ellen Fridland, “Knowing-How: Problems and Considerations,” European Journal of Philosophy 23, no. 3 (2015): 703–727. Fridland makes a very similar point, in the course of arguing against so-called intellectualist views of know-how and skill: “But clearly, when we call the performance of a musician or an athlete ‘brilliant’, ‘stunning’ or ‘banal’, we are talking about the musician or the athlete and not the passive mechanisms that occur inside the agent—that is, that which determines the suitability of normative intellectual ascriptions to agents cannot be a brute, passive process b ecause otherw ise we could only ascribe intelligence to the process, but not to the agent.” I doubt, however, that it is intellectualism that causes the problems here: what causes the problems, I think, is the underlying picture of agential control, which can be shared by intellectualists and anti-intellectualists alike. 31. For an especially stark statement of this sort of view, consider the following, by Berent Enç: “The goal directed system commands a package be havior, the way one o rders a packed lunch from a h otel for a day’s hike, being confident that what goes into the package w ill be selected by competent personnel” (Enç, How We Act, 65). See also the passage from Fodor (“The Appeal to Tacit Knowledge in Psychological Explanation,” 627) quoted e arlier. Explicit appeals to control as reliability can be found in the following sources: Jesús Aguilar, “Agential Systems, Causal Deviance, and Reliability,” in Causing Human Actions: New Perspectives on the Causal Theory of Action, ed. Jesús Aguilar and Andrei Buckareff (Cambridge, MA: MIT Press, 2010); Jesús Aguilar, “Basic Causal Deviance, Action Repertoires, and Reliabilities,” Philosophical Issues 22, no. 1 (2012): 1–19; Joshua Shepherd, “The Contours of Control,” Philosophical Studies 170, no. 3 (2013): 395–411; Joshua Shepherd, The Shape of Agency: Control, Action, Skill, Knowledge (Oxford: Oxford University Press, 2021); Imogen Dickie, “Skill before Knowing,” Philosophy and Phenomenological Research 85, no. 3 (2012): 737–745; Imogen Dickie, Fixing
354
N O T E S T O PAG E S 1 6 2 –1 6 4
Reference (Oxford: Oxford University Press, 2015); Ernest Sosa, Judgment and Agency (Oxford: Oxford University Press, 2015). 32. Ramez Naam, Nexus: Nexus Arc Book 1 (New York: Angry Robot, 2012). 33. In the book, Kade loses all his fights against real martial arts experts. This, however, is not necessarily due to the lack of the reliability (in the relevant sense) of the app; it might just be down to physical strength, for example. In any case, we can stipulate a highly reliable version of the Bruce Lee app for our purposes. 34. Dickie, Fixing Reference, 96–99, suggests that, starting with an account of guidance that places reliability at its core, we can get full-blown agential control simply by requiring the guiding states to be personal-level conscious states. Kade’s example, however, shows that this does not suffice. Kade’s selections from the menu are personal-level conscious events; and yet Kade is not in control of the ensuing movements of his body. 35. Shepherd, “The Contours of Control,” 407–408, proposes to deal with deviant causal chains by requiring that, in exercising control, the agent’s intention must lead to the intended outcome through “normal” (for the agent) causal channels. But the Bruce Lee app would indeed seem to be Kade’s normal way to execute a martial arts move. Of course, the Bruce Lee app is not “normal,” if by that we mean something like normal for a true martial arts expert. A true martial arts expert has agential control over their martial arts moves, which Kade intuitively lacks. But understanding what agential control consists in is our challenge, so we cannot just appeal to an understanding of normality that takes this for granted. 36. As an anonymous referee pointed out to me, some care is needed here. Of course, an agent who plays a video game is exercising agential control in doing so (and perhaps considerable skill). Similarly, Kade may be exercising control over some aspects of his martial arts–related behavior: for example, he still needs to select a target and an attack routine from the app’s drop-down menu. The point is just that he is not exercising control over his martial arts moves themselves. 37. Ludwig Wittgenstein, Philosophical Investigations, trans. G. E. M. Anscombe (Englewood Cliffs, NJ: Prentice Hall, 1958), §621. 38. McDowell, “Some Remarks on Intention in Action,” 17. 39. McDowell, “Singular Thought and the Extent of ‘Inner Space,’ ” in Meaning, Knowledge and Reality (Cambridge, MA: Harvard University Press, 1998), 228–259; McDowell, Mind and World; John McDowell, “The Disjunctive Conception of Experience as Material for a Transcendental Argument,” in Disjunctivism: Perception, Action, Knowledge, ed. Fiona Macpherson and Adrian Haddock (Oxford: Oxford University Press, 2008), 376–389; McDowell, “Avoiding the Myth of the Given”; McDowell, Perception as a Capacity for
N O T E S T O PAG E S 1 6 4 –1 6 6
355
Knowledge; John McDowell, “Perceptual Experience: Both Relational and Contentful,” European Journal of Philosophy 21, no. 1 (2013): 144–157. 40. This claim is often defended as a consequence of the variability of perceptions and the existence of illusions and hallucinations, or alternatively by inference from the practice of perceptual psychology. Bertrand Russell, The Problems of Philosophy (1912; repr. Oxford: Oxford University Press, 2001); Tyler Burge, “Disjunctivism and Perceptual Psychology,” Philosophical Topics 33, no. 1 (2005): 1–78. 41. Dretske, Knowledge and the Flow of Information. 42. McDowell, “Singular Thought and the Extent of “Inner Space’ ”; McDowell, “Perceptual Experience.” 43. For such considerations, see, in addition to McDowell, John Campbell, Reference and Consciousness (New York: Oxford University Press, 2002); Mark Johnston, “Better Than Mere Knowledge? The Function of Sensory Awareness,” in Perceptual Experience, ed. Tamar Szabó Gendler and John Hawthorne (Oxford: Oxford University Press, 2006), 260–290; Mark Johnston, “On a Neglected Epistemic Virtue,” Philosophical Issues 21 (2011): 165–218; and Bill Brewer, Perception and Its Objects (Oxford: Oxford University Press, 2011). The views these authors end up defending still diverge in impor tant details, of course: in particu lar, Brewer and Campbell deny that perceptions have representational content, while McDowell does not. But we need not enter into this debate here. 44. Dickie (Fixing Reference) similarly distinguishes between an action’s manifesting a piece of propositional knowledge and its being guided by a piece of knowledge. In the latter case, the knowledge must presumably be distinct from the action itself; in the former, the knowledge may be simply embodied in the action in the sense I am recommending h ere. For an account of knowledge-how that is framed in related terms, see Joshua Habgood-Coote, “Knowledge-How, Abilities, and Questions,” Australasian Journal of Philosophy 97, no. 1 (2019): 86–104. 45. Of course, through repeated observation or by other means, Kade may come to learn what bodily movements go into, say, a successful flying back kick. Even so, such knowledge would accompany his bodily movements, rather than be embodied in them. 46. Michael Thompson, Life and Action (Cambridge, MA: Harvard University Press, 2008), ch. 2.; G. E. M. Anscombe, Intention (Cambridge, MA: Harvard University Press, 1957). Thompson’s approach is inspired by a reading of Anscombe’s Intention, and, of course, the theme of knowledge in action I w ill be emphasizing below is also a major Anscombean theme. T here is a difference: Anscombe’s focus is on knowledge of what one does, whereas my focus is on knowledge of ways or means of d oing it. The two topics are not unrelated, but I w ill not try to sort out the relations here.
356
N O T E S T O PA G E S 1 6 7–1 6 8
47. Thus, this proposal should be distinguished from the superficially similar proposal defended by “intellectualists” regarding know-how and skill, including Stanley and Williamson, “Skill”; Carlotta Pavese, “Know-How, Action, and Luck,” Synthese 198 (2021), 1595–1617; Carlotta Pavese, “Probabilistic Knowledge in Action,” Analysis 80, no. 2 (2020): 342–356; Carlotta Pavese, “Knowledge, Action, Defeasibility,” in Reasons, Justifications, and Defeaters, ed. Jessica Brown and Mona Simion (Oxford: Oxford University Press, 2021), 177–200. Dickie (“Skill before Knowing”) also emphasizes the distinction between knowledge being manifested in action, rather than guiding it from the outside. I return briefly to her proposal below. 48. Cf. McDowell, “Some Remarks on Intention in Action,” 7; Eric Marcus, Rational Causation (Cambridge, MA: Harvard University Press, 2012); Thompson, Life and Action. McDowell also considers and rejects a similar claim, on the grounds that, while the action is ongoing, t here is no par ticular action for us to think about. But I confess that I do not see the difficulty here. Of course, while the action is ongoing, there is no complete action to direct our thoughts at. But why should we balk at the idea that an incomplete, or ongoing, action is a particu lar we can think of de re? Intuitively, at least, we would seem to have no difficulty using ordinary demonstrative reference to direct our thoughts to one rather than another ongoing pro cess. For example, supposing that two football games are on the same TV screen at the same time, you can certainly select to direct your thoughts to this game rather than that one. Moreover, what you attend to in each case would seem to be the ongoing game as such, not simply the stretch of game that has gone on before. I suggest, therefore, that we should not balk at the prospect of de re thoughts about ongoing actions. (For broader considerations in favor of allowing ongoing occurrences as particulars, see Zoltán Gendler Szabó, “Things in Prog ress,” Philosophical Perspectives 22 [2008]: 499–525; Helen Steward, “Processes, Continuants, and Individuals,” Mind 122, no. 487 [2013]: 781–812.) 49. Cf. Will Small, “Practical Knowledge and the Structure of Action,” in Rethinking Epistemology, vol. 2, ed. Günter Abel and James Conant (Berlin: De Gruyter, 2012), 133–227; Markos Valaris, “The Instrumental Structure of Action,” Philosophical Quarterly 65, no. 258 (2015): 64–83. Tying your shoelaces may be a “basic action,” in one sense of this term: it is something most of us can do with no deliberation or conscious planning. In this sense, the claim that tying your shoelaces is a basic action is entirely consistent with the claim in the text, which implies that the finger movements in which your shoelace-tying consists are actions of yours (they are not mere muscle spasms, for example). Thus, even for basic actions, t here w ill be such a t hing as a way in which you perform them, and so the principle in the text can apply. Sometimes, however, this use of the term “basic action” is conflated with the
N O T E S T O PAG E S 1 6 9–1 7 2
357
rather diff erent idea of an action that is not performed by doing other things, and so not in any “way” (Small, “Practical Knowledge and the Structure of Action,” and Valaris, “The Instrumental Structure of Action,” call these “atomic” actions). Tying your shoelaces is not a basic action, in this stronger sense—indeed it is doubtful w hether anything is (Douglas Lavin, “Must T here Be Basic Action?,” Noûs 47, no. 2 [2013]: 273–301; Thompson, Life and Action). If there are actions such that we can perform them without there being any way we perform them in, then my account would fail to apply to them. But, I submit, it is likely that the concept of agential control would also fail to apply to them. 50. It seems compatible with intellectualist accounts like those offered by Jason Stanley and Timothy Williamson in “Knowing How,” Journal of Philosophy 98, no. 8 (2001): 411–444; Stanley and Williamson, “Skill”; and Jason Stanley, Know How (Oxford: Oxford University Press, 2011), but also some nonintellectualist ones, such as Kieran Setiya, “Practical Knowledge Revisited,” Ethics (2009): 128–137. Dickie (“Skill before Knowing”) develops a similar account of knowledge embodied in action, and argues that it shows that “skill comes before knowledge.” T here is a reading of this claim with which I can agree: skill is a standing state that can warrant knowledge embodied in action. Dickie, however, embeds her claim within a broadly reliabilist account of skill and control, which (for the reasons discussed in Section 3) I reject. A related view is defended by Habgood-Coote (“Knowledge-How, Abilities, and Questions”), who suggests that know-how consists in the ability to have particular bits of knowledge of the sort discussed here.
6. Naturalism in the Philosophy of Action 1. I say “at last” because I have things in print in which, as it now seems to me, I said stuff that should have left me puzzled. 2. John McDowell, “Some Remarks on Intention in Action,” The Amherst Lecture in Philosophy 6 (2011): 1–18; available at: http:// w ww .a mherstlecture.org/mcdowell2011/. Henceforth I refer to this lecture as “Remarks.” Where no page references are given for quotations, the quotations are to be found in the lecture’s printed Abstract. 3. See McDowell’s “Naturalism in the Philosophy of Mind,” in Naturalism in Question, ed. Mario de Caro and David Macarthur (Cambridge, MA: Harvard University Press, 2004), 91–105. 4. John McDowell, at 113 in “Another Plea for Modesty,” in Meaning, Knowledge, and Reality (Cambridge, MA: Harvard University Press, 1998), 108–131.
358
N O T E S T O PAG E S 1 7 2 –1 76
5. John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994), 157. The two quotations which follow are from p. 35 and p. 90. 6. I first drew attention to this at the Fourth International Wittgenstein Symposium in 1979: see Jennifer Hornsby, “Actions and Abilities,” in Language, Logic and Philosophy, ed. R. Haller (Vienna: Hölder-Pichler- Tempsky, 1980), 387–391. 7. Donald Davidson, “Actions, Reasons, and C auses,” Journal of Philosophy 60 (1963): 685–700, 686. 8. I’ve left out alert the prowler, which Davidson distinguished from the other three things on his list of four, it being something he did unintentionally. Davidson’s own assumption was that philosophy of action’s primary task is to find conditions for the correct application of “intentionally”—an assumption that has to be abandoned when it is acknowledged that an enquiry into human agency must introduce the agent’s perspective at its start. And I take it that that which one does unintentionally (as opposed to simply lacking any intention to do) is done in doing something else one means to do. 9. Donald Davidson, “The Logical Form of Action Sentences,” in The Logic of Decision and Action, ed. N. Rescher (Pittsburgh, PA: University of Pittsburgh Press, 1967), 81–88 (italics mine). In “Actions, Reasons, and Causes,” Davidson had said “ ‘I turned on the light’ clearly refers to a partic hole sentences refer, ular event,” but in this l ater piece, he came to deny that w saying instead that such a sentence as “Jones buttered the toast” has a logical form in which is found an existential quantifier over events and a predicate having one more place than the surface version would suggest. Given Davidson’s own views about semantic compositionality, one must be puzzled by his apparently thinking that his three-place predicates—“ turned on,” “buttered,” etc.—stand in no relationship to “turn on” or “butter” as they recur in other sentences, whether in an infinitive or differently inflected. 10. One still might wonder how Davidson got away with denying that he did as many things as he said he did. Well, he said: “Only the question ‘Why did you (he) do A?’ has the true generality required” (Davidson, “The Logical Form of Action Sentences,” 686n2). And h ere something done (specified with instances of “A” [no need for “do A”]) is conflated with the event (the agent’s d oing of A), which Davidson wanted us to think the “why” question seeks an explanation for. 11. Perfect nominals take adjectives and can be pluralized, bearing witness to the particularhood and the countability of their denotations. Imperfect nominals, by contrast, are adverbially modifiable and negatable. They have within them a verb which, as Zeno Vendler put it, is “alive and kicking” (see Zeno Vendler, “Facts and Events,” in Linguistics in Philosophy [Ithaca, NY: Cornell University Press, 1967], 122–146). I note that sometimes English doesn’t have the words for a perfect nominal distinguishable from the
N O T E S T O PA G E S 1 7 7–1 8 4
359
imperfect one, so that, e.g., “Her walking across the street” can play both roles. The idea that there are, as McDowell puts it, two styles of predication of telic verbs is captured by Michael Thompson when he speaks of event-or process- descriptions. See chapter 8 of Michael Thompson, Life and Action (Cambridge, MA: Harvard University Press, 2008). 12. So we are told in Markus Schlosser, “Agency,” in The Stanford Encyclopedia of Philosophy (Winter 2019), ed. Edward N. Zalta, https://plato .stanford.edu/a rchives/w in2019/entries/agency/. 13. McDowell, “Naturalism in the Philosophy of Mind.” 14. McDowell, “Remarks,” 8. 15. Ibid., 10. 16. Ibid., 7. 17. Ibid. 18. It could be thought that use of the perfect tense even with a nontelic verb introduced completion of a sort: if someone has walked to some specifiable place or throughout some specifiable interval (to here, or until t2, as it might be), then there is a delimited stretch of walking (from there to here, or from t1 to t2), and such a stretch might be called a complete one. But what ever such stretches might be called, they are not particulars (although t here is no doubt that, delimited as they are, they are particu lar, in the sense of specific, stretches). When someone stops walking and not b ecause they have finished a walk they’d been taking or reached somewhere they’d intended to walk to, there is an end to their walking in a purely temporal sense. And t here might then be said to be an event—of their walking u ntil they stopped. But such an event is not something on which Davidson officially bestowed the title action. Nor is it something McDowell allowed to be a particu lar when he framed his conception of the reality of what he called particular actions. 19. Ibid., 9. 20. It can be a question when a cake first comes to exist. A reviewer of a draft of this paper invited me to consider the statement “The cake now in the oven w ill need more time there before it is baked.” Here the cake in question exists, but it is not yet baked. “Make a cake” (which speakers of British En glish are more likely to use than “bake a cake”) certainly provides for unfinished cakes. “I’ll put the cake in the oven in a bit” might be said at a time when the cake (as then it is thought of) consists of dough distributed between two baking tins. (Someone whose intention it was to make a birthday cake would be unlikely to think of themself as having finished even when a cake they had made was cooked: in this case they would probably want not merely to have produced a baked cake, but to have produced a decorated one, appropriately iced maybe.) 21. Ibid. 22. Davidson, “Actions, Reasons, and Causes.” 23. McDowell, “Remarks,” 35, 90.
360
N O T E S T O PAG E S 1 8 5 – 2 01
24. Ibid., 11. 25. Ibid., 10. 26. Ibid., 13n15. 27. I say more relating to McDowell’s views on Anscombe in my conclusion (§5). The first edition of Anscombe’s Intention was published in 1957; the second edition (1963) was republished by Harvard University Press in 2000. The paper by Ford which bears directly on the matter at hand is Anton Ford, “The Province of Human Agency,” Noûs 52, no. 3 (2018): 697–720. 28. McDowell, “Naturalism in the Philosophy of Mind,” 95. 29. Ibid.
7. Perceiving the World 1. John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1996), 47n1. 2. McDowell recommends his “relaxed naturalism” (Mind and World, 89) as enabling us to avert “the threat of supernaturalism” (79). 3. In the Introduction to the Phenomenology of Spirit, Hegel remarks that we must not assume that knowledge is something other than we, nor that knowledge, and that is, we, is something other than the absolute. 4. It may happen that I answer someone’s question unwittingly: “Hey, you just answered my question!” “Oh, did I? I had no idea.” But this cannot be so in general. 5. If a judgment is invalid, then the thought of its validity cannot be internal to its validity, for it has no validity. 6. In this sentence, “oneself” is a form of the first-person pronoun; it is that pronoun in oratio obliqua. The English language has no special form for this. I ask the reader to supply it whenever it is needed. 7. It may be thought that recent discussions of the KK-principle has fi nally subjected to scrutiny what was an uncritical assumption. This is quite wrong. The KK-principle is understood to represent a relation of entailment of two distinct propositions. Understood in this way, the principle denies what in the tradition was taken for granted. The discussions of the KK-principle make no contact with the traditional idea that knowing something is knowing oneself to know it. In a seminal paper, Alston diagnosed “level-confusions in epistemology”: the confusion was to reside in a failure to distinguish knowledge from knowledge of knowledge (see William P. Alston, “Level Confusions in Epistemology,” Midwest Studies in Philosophy 5, no. 1 [1980]: 135–150). The idea that there are such levels is the single most disastrous confusion that ever afflicted philosophical thought about knowledge. 8. McDowell describes his topic as knowledge that is an act of spontaneity, or freedom. This is the same as saying that the topic is knowledge that is
N O T E S T O PAG E S 2 01– 2 0 6
361
capable of being expressed by “I know.” The realm of freedom, or spontaneity, is the realm of what is the thought, the knowledge, the understanding of itself. 9. I am not sure that the label is a happy one. 10. McDowell writes: “That is precisely not threatening to common sense, to the conviction that the world is independent of our thinking” (Mind and World, 83). It would be good if we replaced “conviction” by “knowledge”. An assertion that threatens our convictions differs from a claim that disagrees with what we know in that the latter, just as such, is no threat to us. The knowledge in question surely is common. But it is infelicitous to call it “common sense”, for two reasons. First, many t hings are common sense, many of which lack the dignity and the depth of the idea that knowledge is of something that is as it is anyway. The term “common sense” does not bring out the character of the knowledge that it is the task of philosophy to bring to explicit self-comprehension. Second, the term suggests that philosophy ought to be responsive to the ideas harbored by the man in the street, his sense. However, confronted with a philosophical question, the man in the street w ill likely say things that make no sense. This is not to slight the intelligence of the man in the street. It is to observe that philosophy is a discipline. It requires exercise. 11. More recently, McDowell modified his original idea that concepts are at work in sense perception in the following way (see John McDowell, “Avoiding the Myth of the Given,” in Having the World in View [Cambridge, MA: Harvard University Press, 2009], 256–274). First there is a capacity of sense perception; what is given in acts of it is thought through formal, or logical, concepts such as—McDowell’s example—the concept of an animal. Then, what is so given is thought according to material concepts such as— again McDowell’s example—the concept of a cardinal, in acts of a separate capacity, which McDowell calls a recognitional capacity. So the modification resides in the separation of the power of sense perception from recognitional capacities, the limitation of the former to logical concepts and the relegation of material concepts to the latter. McDowell explains that he was driven to these modifications by objections to his original idea made by Charles Travis. I cannot find in these objections anything that would motivate modifying the original idea. Moreover, I find the modification ill-conceived. The distinction of the power of sense perception from recognitional capacities is unsound because the application of a logical concept in sense experience, that is, the thinking of something given to the senses through a formal concept, is the use of a material concept: “this animal” signifies a material concept. As Kant puts it, the material concept is the logical concept in concreto: in what is given in sensory experience (see Immanuel Kant, Critique of Pure Reason, A567 / B595). Therefore McDowell’s conception of the power of sense perception entails that it is—a nd thus is not diff erent from—a totality of recognitional capacities.
362
N O T E S T O PA G E S 2 0 7– 2 14
12. It may seem that McDowell presents perceptions as halting a regress of justification. For he speaks of perception as providing “ultimate grounds of empirical judgments” and credits himself with having shown how to “reinstate friction without undermining the very idea of ultimate grounds, as the Myth of the Given does” (Mind and World, 67). But we must not overlook that the idea of an ultimate ground has undergone a transformation as we have rejected the Myth of the Given. McDowell explains that an empirical concept, as such, is an understanding of the way in which it is integrated with o thers in an idea of the world as a w hole. That idea is articulated in principles according to which one thing depends on another: “rational linkages” which “give the concepts their place as elements in possible views of the world” (ibid., 12). It follows that my knowledge of what I perceive here and now is no firmer than my knowledge of the principles through which I unite it with everything that I know in a thought of the world as a whole. And McDowell asserts that our obligation self-critically to reflect on t hese principles is perpetual. 13. Neither can I know the former on the basis of the latter. There is no inference of the form: “P. Therefore I know p.” 14. McDowell, Mind and World, 26. 15. Ibid. 16. Ibid., 42. 17. Ibid., 68. 18. McDowell says that “perceptible facts are essentially capable of impressing themselves on perceivers” (Mind and World, 28). This is not something he found out about perception, from sideways on. Rather, perceiving a fact is knowing the described essence of it; it is knowing it to be perceptible. For perceiving something and thereby knowing it is the same act of mind as knowing oneself to perceive it. So it is if the perception in question is capable of being expressed by “I perceive”. 19. McDowell recommends his relaxed naturalism as enabling us to avoid the “threat of supernaturalism” (Mind and World, 78). 20. Ibid., 10 and 12, respectively. 21. Ibid., 32. 22. Ibid., 36. 23. It may seem that McDowell means to describe an understanding that must be in the background of the experience it concerns, rather than inside it. He writes: “No one could count as making even a directly observational judgment of colour except against a background sufficient to ensure that she understands colours as potential properties of things. . . . The relevant background understanding includes, for instance, the concept of visible surfaces of objects and the concept of suitable conditions for telling what colour something is by looking at it” (Mind and World, 12). The two sentences differ in that the first one does not place the understanding of which it speaks in the background of the observational judgment; rather, what is said to be in
N O T E S T O PAG E S 2 1 8 – 2 1 9
363
the background is what ensures that the understanding is there. Thus the sentence places no obstacle in the way of conceiving the observational judgment, and by extension, the observation on which it is based, as putting this understanding to work. And surely this is how we must understand it. For McDowell’s point is that conceptual capacities are drawn into operation in experience. The understanding in question is constitutive of the relevant concepts; hence, if these concepts are operative in experience, so is that understanding. The understanding of what one perceives, perceiving something to be a certain color, as an object bearing a certain property, is at work in perception; it is not a background to it. The second sentence, by contrast, speaking of a “background understanding”, represents the understanding as a background. This strikes me as a slip. McDowell corrects the slip in the third lecture, where, discussing Evans, he makes it plain that placing experiences in the context of a background understanding does nothing to show how they enable knowledge of an independent reality. See also Mind and World, 54–55.
8. Seeing the World For their comments on an earlier draft of this paper, I want to thank Matt Boyle, Tal Brewer, Irad Kimhi, Anselm Mueller, Kieran Setiya, and John Hacker- Wright. I want to thank Robert Pippin for a number of lectures and conversations on art and philosophy in my seminar in Athens in 2018. I also want to thank the audience of the Iris Murdoch Centenary Conference at the University of Oxford in 2019, and in particular Megan Laverty for a number of discussions on the thought of Iris Murdoch. Finally, I want to thank two anonymous reviewers for Harvard University Press for their comments. Irad Kimhi’s book Thinking and Being develops a non-Fregean account of thought’s being of the world, which I believe is congenial with the line of thinking I pursue in this paper. The shape of this paper would undoubtedly have been diff erent and much improved if I had encountered Kimhi’s book e arlier during my writing.
1. John McDowell, The Engaged Intellect: Philosophical Essays (Cambridge, MA: Harvard University Press, 2009), 44. 2. Ibid., 46. 3. See, for instance, Michael Ridge and Sean McKeever, “Moral Particularism and Moral Generalism,” in The Stanford Encyclopedia of Philosophy (Winter 2016), ed. Edward N. Zalta, https://plato.stanford.edu/a rchives /w in2016/entries/moral-p articularism-g eneralism/. 4. McDowell, The Engaged Intellect, 46. 5. See, for instance, Rosalind Hursthouse and Glen Pettigrove, “Virtue Ethics,” in The Stanford Encyclopedia of Philosophy (Winter 2018), ed. Edward N. Zalta, https://plato.s tanford.edu/archives/w in2018/e ntries/e thics -v irtue/.
364
N O T E S T O PAG E S 2 1 9– 2 3 3
6. For one such case, see Jeffrey Seidman, “Two Sides of ‘Silencing,’ ” Philosophical Quarterly 55, no. 218 (2005): 68–77. 7. Aristotle introduces the apparatus of the practical syllogism in his discussion of akrasia in the Nicomachean Ethics (1147a24–1147b19), but the topic is also discussed in several other works (e.g., De Anima 434a15–20, and De Motu Animalium 701a8–33). 8. See, for instance, p. 57 of John McDowell, “Virtue and Reason,” Monist 62 (July 1979): 331–350, reprinted in John McDowell, Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998). 9. McDowell, The Engaged Intellect, 49. 10. Cora Diamond, “Henry James, Moral Philosop hers, Moralism,” Henry James Review 18, no. 3 (1997): 243–257, 268. 11. Alice Crary, Inside Ethics: On the Demands of Moral Thought (Cambridge, MA: Harvard University Press, 2016). 12. McDowell, The Engaged Intellect, 44. 13. My use of the term “moral falsehood” as opposed to “moral error” is deliberate and meant to evoke an analogy with Plato’s account of falsehood in the Sophist. 14. The locus classicus of this distinction is McDowell, “Virtue and Reason.” 15. For my account of Murdoch’s work in this regard, see Evgenia Mylonaki, “The Individual in Pursuit of the Individual; A Murdochian Account of Moral Perception,” Journal of Value Inquiry 53 (2019): 579–603. 16. See his essays on Aristotle’s ethics in McDowell, Mind, Value, and Reality. 17. All the references to the text in what follows are from the Library of America edition: Henry James, Novels 1881–1886: Washington Square, The Portrait of a Lady, The Bostonians, ed. William T. Stafford (New York: Library of America, 1985). 18. In the preface to the novel, James calls his heroine a “frail vessel” (borrowing George Eliot’s term). See Henry James, The Art of the Novel: Critical Prefaces (Chicago: University of Chicago Press, 2011), 49–50. Additionally, in a very uncanny move early in the novel, the narrator paints a ruthless portrait of Isabel as a person with “a general idea that p eople w ere right when they treated her as superior,” whose thoughts are “a tangle of vague outlines never connected,” with an “unquenchable desire to think well of herself,” and who is inconsistent and hypocritical, r unning “the danger of keeping up the flag a fter the place has surrendered” (James, The Portrait of a Lady, 240, 241, 242). 19. For a reading along these lines, see Robert B. Pippin, Henry James and Modern Moral Life (Cambridge: Cambridge University Press, 1999). 20. James, The Portrait of a Lady, 372. 21. Ibid., 723.
N O T E S T O PAG E S 2 3 3 – 24 4
365
22. Ibid., 725, 726. 23. Ibid., 751. 24. Ibid., 759. 25. Ibid. 26. Ibid. 27. Ibid., 798. 28. Ibid., 799. 29. Ibid. 30. Ibid. 31. Ibid., 631. 32. Iris Murdoch, Metaphysics as a Guide to Morals (New York: Viking, 1992).
9. See the Right Thing EN: Aristotle, Nicomachean Ethics EE: Aristotle, Eudemian Ethics Except where indicated, translations of passages from Greek texts are my own, with angular brackets indicating things not explicitly in the original but relatively clearly intended.
1. For the fruits of my extended conversations with Irwin and McDowell, see Jennifer Whiting, “Strong Dialectic, Neurathian Reflection, and the Ascent of Desire: Irwin and McDowell on Aristotle’s Methods of Ethics,” Proceedings of the Boston Area Colloquium of Ancient Philosophy 17, no. 1 (2002): 61–122. The main ideas t here are (1) that one can reasonably read Aristotle as employing in the service of the moral psychology McDowell ascribes to him something like the method of “strong dialectic” that Irwin ascribes to him; and (2) that this sort of reading is close to the one that McDowell himself gives when he allows that Aristotle’s virtuous agent may seek “reflective reassurance” that her ethical outlook is correct by engaging in Neurathian reflection. The crucial points, according to McDowell, are that the reflection be in some sense internal to the ethical outlook in question and that such reflection play no direct role in the motivational economy the genuinely virtuous (and so phronimos) agent, who must be moved directly by the valuations internal to that outlook. Reflective reassurance seems useful primarily to agents who, though on their way to virtue, are at least sometimes tempted to rationalize departures from what they in some sense “know” they should do. 2. See Iakovos Vasiliou, “Perception, Knowledge, and the Sceptic in Aristotle,” Oxford Studies in Ancient Philosophy 14 (1996): 83–131, and Iakovos Vasiliou, “The Role of Good Upbringing in Aristotle’s Ethics,” Philosophy and Phenomenological Research 56, no. 4 (1996): 771–797.
366
N O T E S T O PAG E S 24 4 – 24 8
3. See Jennifer Whiting, “Locomotive Soul: The Parts of Soul in Aristotle’s Scientific Works,” Oxford Studies in Ancient Philosophy (2002): 141–200; and Jennifer Whiting, “Hylomorphic Virtue: Cosmology, Embryology, and Moral Development in Aristotle,” Philosophical Explorations 22, no. 2 (2019): 222–242. 4. Though thumos is often rendered “spirit” and associated with various forms of self-assertion, there are places where Aristotle seems to use it to cover emotions more generally, including pity and fear. What justifies grouping desires that stem from such emotions together and distinguishing them from appetites is the fact that they are potentially reason-responsive in a way appetites are not. For more on this distinction in Plato, from whom Aristotle takes his point of departure, see Jennifer Whiting, “Psychic Contingency in Plato’s Republic,” in Plato on the Divided Self, ed. Rachel Barney, Tad Brennan, and Charles Brittain (Cambridge: Cambridge University Press, 2012), 174–208. 5. John McDowell, “Some Issues in Aristotle’s Moral Psychology,” in Companions to Ancient Thought 4: Ethics, ed. Stephen Everson (Cambridge: Cambridge University Press, 1998), 107–128, section 11. Reprinted in John McDowell, Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998). Aristotle’s term for the desiring element (in general) is to orektikon (on which more below). 6. I use the term “counterpart” to capture Aristotle’s view that each in some sense reaches the correct prohairesis, while leaving open the exact sense in which each does so. References to prohairesis are ambiguous, sometimes referring to a propositional-attitude-like state of soul (i.e., nous orektikos or orexis dianoêtikê 1139b4–5), sometimes to the content of that state, which may however be the content of some other state, such as belief or recommendation. So an akratic agent might reason his way to the belief that φ-ing is the t hing to do without however coming to have the sort of desire to φ that is involved in prohairesis qua state of soul. If φ-ing is in fact the t hing to do, he reaches the correct prohairesis in the sense that he comes to accept the content of the correct prohairesis: but his attitude is one of belief, not prohairesis. 7. John McDowell, “Virtue and Reason,” Monist 62 (1979): 331–350, §3. 8. Personal communication. 9. McDowell, “Virtue and Reason,” §3. 10. McDowell has, in conversation, allowed this. 11. Plato, Apology, in Plato, Five Dialogues: Euthyphro, Apology, Crito, Meno, Phaedo, 2nd ed., trans. G. M. A. Grube, rev. John M. Cooper (Indianapolis, IN: Hackett Publishing Company, 2002), 21–34, 28b–c. 12. Plato, Apology 30b. I follow Cooper’s revision of the Grube translation, with one small change. For consistency with the rest of my text, I use “virtue” rather than “excellence” for “aretê.” There has been extended controversy about how to take the last sentence here. But Burnyeat makes a
N O T E S T O PAG E S 2 5 0 – 2 7 8
367
compelling case for this version, according to which virtue makes wealth and everything e lse good for their possessors. M. Burnyeat, “Apology 30b2–4: Socrates, Money, and the Grammar of γίγνεσθαι,” Journal of Hellenic Studies 123 (2003): 1–25. 13. Whiting, “Psychic Contingency.” 14. There is an obvious affinity here to McDowell’s disjunctivism, discussed in James Conant’s contribution to this volume. 15. Translation by G. M. A. Grube, rev. C. D. S. Reeve (Indianapolis, IN: Hackett Publishing Company, 1992). 16. Plato, Republic X, 606b9 (tragedy) and 606c6 (comedy). 17. Grube-Reeve translation, slightly modified. All occurrences of ‘part’ have been bracketed; neither ‘meros’ nor any equivalent of it appears in the Greek. 18. See note 6. The agent w ill act “from” prohairesis understood as orexis dianoêtikê, and not simply (as the enkratês acts) “in accordance with” prohairesis understood as the propositional content of the proper orexis. 19. Homer, Odyssey XX.17–18. 20. One’s relationship to the speaker m atters here. We are more likely to listen in ways that facilitate comprehension if the speaker is someone we love and respect, and comprehension is more likely to come with the sort of sympathetic (or at least open-minded) listening that goes with such attitudes. See below, the section “Acquiring practical nous.” 21. For more on this, see Jennifer Whiting, “The Mover(s) of Rational Animals: De Anima 3.11 in Context,” in Aristote et l’âme humaine, Lectures de De Anima III offertes à Michel Crubellier, ed. G. Guyomarc’h, C. Louguet, and C. Murgier (Louvain-la-Neuve: Peeters, 2020). 22. McDowell, “Virtue and Reason,” section 4 (from which all quotes in this section are taken). 23. Stanley Cavell, Must We Mean What We Say? (Cambridge: Cambridge University Press, 1976), 52. 24. McDowell, “Virtue and Reason,” 331. 25. Aristotle, Topics 1.11, 105a4–8. See Vasiliou, “The Role of Good Upbringing.” 26. He may also be thinking of one of the hexeis, discussed in EN 3.8, that are sometimes confused with genuine courage, but that would not affect my point. 27. McDowell, “Virtue and Reason,” 333. 28. I am thinking of an occasion when I, around ten years of age, asked my father why he offered to buy a meal for a man who was begging but would not give him money with which to buy the meal. 29. For more on this conception of virtue, see Whiting, “Eudaimonia, External Results, and Choosing Virtuous Actions ‘for Themselves.’ ” 30. See Plato, Republic 401e–402a.
368
N O T E S T O PAG E S 2 8 0 – 2 9 0
31. I am thinking here, with shame, of an occasion on the subway in NYC when I attempted, self-protectively, to ignore a crying w oman who turned out to be asking only for a tissue. 32. McDowell, “Virtue and Reason,” 54.
10. Self-Consciousness and the Idea of Bildung CPR: Immanuel Kant, Critique of Pure Reason, translated by the author PS: G. W. F. Hegel, Phenomenology of Spirit, trans. A. V. Miller (New York: Oxford University Press 1988) I am grateful to Matthew Boyle and Thomas Oehl for their helpful comments on the paper. The paper draws on material from Andrea Kern, “Selbstbewusstsein und die Idee der Bildung als, immanentes Moment des Absoluten’. Über einige Unterschiede zwischen Kant, Hegel und McDowell,” in Thomas Oehl and Anton Kok, eds., Objecktiver und absoluter Geist nach Hegel. Kunst, Religion, und Philosophie innerhalb and außerhalb von Gesellschaft und Geschichte (Leiden: Brill, 2018), 872–894.
1. G. W. F. Hegel, Faith and Knowledge, trans. Walter Cerf and H. S. Harris (Albany: State University of New York Press, 1977). 2. See John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994), 98: “Since he [Kant] does not contemplate a naturalism of second nature, and since bald naturalism has no appeal for him, he cannot find a place in nature for this required real connection between concepts and intuitions.” 3. G. W. F. Hegel, Elements of the Philosophy of Right, trans. H. B. Nisbet (Cambridge: Cambridge University Press, 1991), §187, 253(D) / 226(E). Page numbers refer to the standard German (D) and English (E) editions of Hegel’s works. Translations have been emended without comment. 4. Hegel, Faith and Knowledge, 305(D) / 70(E). 5. Ibid., 305(D) / 70(E). 6. CPR B134–135; B153. 7. McDowell, Mind and World, 9. 8. See also ibid., 98. 9. Gareth Evans, The Varieties of Reference (Oxford: Clarendon Press, 1982), 227ff. 10. Some interpreters ascribe this sort of account to Kant. I discuss and criticize this interpretation in Andrea Kern, “Spontaneity and Receptivity in Kant’s Theory of Knowledge,” Philosophical Topics: Analytical Kantianism 34, no. 1–2 (2006): 145–162. 11. McDowell, Mind and World, 98. 12. Hegel, Faith and Knowledge, 309(D) / 73(E). See also 329(D) / 92(E). 13. Hegel, Faith and Knowledge, 309(D) / 74(E).
N O T E S T O PAG E S 2 9 1– 2 9 8
369
14. CPR B90. 15. CPR B90. On the idea that demonstrating the possibility of a priori concepts is the central theme of the Transcendental Deduction, see Stephen Engstrom, “Understanding and Sensibility,” Inquiry 49 (2006): 2–25. 16. CPR B76. 17. CPR B122. 18. CPR B161. 19. CPR B123. 20. Ibid. 21. Ibid. This reading of a priori concepts should not be confused with the denial of concepts that have no possible exemplification in intuition. Such concepts of pure reason (or “ideas”) are among the main topics of the Transcendental Dialectic (Soul, World, God, etc.). What Kant denies, according to this reading, is not that there can be such concepts that go beyond the conditions of possible experience and, although they do not express cognitions, may have quite definite cognitive roles to play. Rather, he denies that one can recognize a concept as a subjective condition for thinking without recognizing its objective validity. 22. On the identification of the understanding with self-consciousness, see the extraordinarily helpful and important essay by Stephen Engstrom, “The Unity of Apperception,” Studi Kantiani 26 (2013): 37–54, to which the above reflections on Kantian self-consciousness are indebted. 23. CPR B132. 24. Ibid. 25. See McDowell, Mind and World, 89. 26. On this point, see my critique in Andrea Kern, “Kant über selbstbewusste Sinnlichkeit und die Idee menschlicher Entwicklung,” in Andrea Kern and Christian Kietamann, Selbstbewusstes Leben. Texte zu einer transformativen Theorie der menschlichen Subjektivität (Berlin: Suhrkamp, 2017), 270–301. 27. McDowell, Mind and World, 87. 28. John McDowell, “Towards a Reading of Hegel on Action in the ‘Reason’ Chapter of the Phenomenology,” in Having the World in View: Essays on Kant, Hegel and Sellars (Cambridge, MA: Harvard University Press, 2009), 166–184: 166. 29. Ibid., 172. 30. Ibid. 31. Hegel, Philosophy of Right, §187, 253(D) / 226(E). 32. See PS §808, 530–531(D) / 492–493(E). 33. Hegel, “Faith and Knowledge,” 318ff.(D) / 81ff.(E). 34. PS §793, 519(D) / 482(E). 35. PS §807, 529(D) / 492(E). 36. Hegel, “Faith and Knowledge,” 319(D) / 83(E).
370
N O T E S T O PAG E S 2 9 8 – 3 0 6
37. PS §347 supra, 233(D) / 211(E). 38. PS §167, 121(D) / 105(E). 39. Ibid. 40. PS §168, 122(D) / 106(E). 41. PS §168, 123(D) / 107(E). 42. PS §168, 122(D) / 106(E). 43. Ibid. 44. Ibid. 45. Ibid. 46. Ibid. 47. PS §172, 125(D) / 109(E). 48. Ibid. 49. PS §176, 127(D) / 110(E). 50. Ibid. 51. Ibid. See also Michael Thompson, Life and Action (Cambridge, MA: Harvard University Press, 2008), ch. 1. 52. PS §173, 125(D) / 109(E). 53. Ibid. 54. PS §349, 234(D) / 212(E). 55. Ibid. 56. Ibid. 57. Ibid. 58. PS §350, 234(D) / 212(E). 59. Ibid. 60. PS §350, 234–235(D) / 212(E). 61. Ibid. 62. Ibid. See also Hegel, “Faith and Knowledge,” 309(D) / 73–74(E), 310(D) / 75(E), 326(D) / 89–90(E). 63. PS §177, 126(D) / 110(E). 64. Hegel, Philosophy of Right, §187, 252–253(D) / 225–226(E).
11. The Idealism in German Idealism EL: G. W. F. Hegel, The Encyclopedia Logic. Citations are to the German edition in Werke, ed. Eva Moldenhauer and Karl Markus Michel (Frankfurt: Suhrkamp, 1970–1971), vol. 8. The translation used is The Encyclopedia Logic, trans. T. F. Geraets, W.A. Suchting, and H. S. Harris (Indianapolis, IN: Hackett Publishing, 1991). SL: G. W. F. Hegel, The Science of Logic. Citations are by volume and page number of the edition in Gesammelte Werke, ed. Rheinisch-Westfälische Akademie der Wissenschaften (Hamburg: Meiner, 1968). The Cambridge
N O T E S T O PAG E S 3 0 9– 3 1 6
371
translation of the Logic, by George di Giovanni (Cambridge: Cambridge University Press, 2010) has helpfully listed the volume and page number of the Gesammelte Werke in the margins, and this translation is the one used.
1. John McDowell, Mind and World, 2nd ed. (Cambridge, MA: Harvard University Press, 1996). 2. Wittgenstein 1973, §95. I discuss this so-called identity theory of truth at greater length in Robert Pippin, Hegel’s Realm of Shadows: Logic as Metaphysics in “The Science of Logic” (Chicago: University of Chicago Press, 2018). 3. McDowell, Mind and World, 27. 4. Ibid., 28. 5. Ibid., 27–28. 6. Ibid., 27. 7. John McDowell, Having the World in View: Essays on Kant, Hegel, and Sellars (Cambridge, MA: Harvard University Press, 2009), 153. 8. John McDowell, “What Is the Phenomenology About?,” in McDowell and Hegel: Perceptual Experience, Thought and Action, ed. Federico Sanguinetti and André J. Abath (Cham: Springer, 2018), 29–43; 242–243. 9. Pippin, Hegel’s Realm of Shadows. 10. The first paragraph of the Philosophy of Right is typical of the claim made for the centrality of the Logic. 11. I mean, for example, the work of Dieter Henrich, Hans Friedrich Fulda, Michael Theunissen, Rüdiger Bubner, Konrad Cramer, Rolf-Peter Horstmann, and o thers. 12. Immanuel Kant, Critique of Pure Reason, trans. Paul Guyer and Alan Wood (Cambridge: Cambridge University Press, 1999), Bxiii. 13. As we s hall see, such a view is not entirely wrong, but it crudely misstates the true position. 14. See, for example, Charles Larmore, Das Selbst in seinem Verhältnis zu sich und zu anderen (Frankfurt: Klostermann, 2017). 15. Robert Pippin, Hegel’s Idealism: The Satisfactions of Self- Consciousness (Cambridge: Cambridge University Press, 1989). 16. SL 21.29. 17. I do not detect in McDowell a robust interest in the possibility of a priori knowledge, but am unsure of his position. From a Hegelian point of view as I understand it, everything claimed in Mind and World about the relation between mind and world is an a priori claim. They are certainly not empirical claims. I think it is fair to say that the book assumes such claims could only be a priori claims, and that the objections to McDowell from critics like Tyler Burge and Christopher Peacocke, in appealing to “what we have learned” about this relation “from science,” misunderstand the issue at stake. 18. See Robert Pippin, “Fichte’s Alleged One-Sided, Subjective, Psychological Idealism,” in The Reception of Kant’s Critical Philosophy: Fichte,
372
N O T E S T O PA G E S 3 1 7– 3 2 0
Schelling and Hegel, ed. Sally Sedgwick (Cambridge: Cambridge University Press, 2000), 147–170. 19. This issue, and most of the other issues raised in this paper, are discussed at much greater length in Pippin, Hegel’s Realm of Shadows. 20. EL §44. 21. EL §41A. 22. It is not only Hegel’s view, but a very common one, inspired by what Kant himself says about the beginning of the critical period in his philosophy, as announced in his famous letter to Markus Herz of February 2, 1772: Therefore the pure concepts of the understanding must not be abstracted from sense perceptions, nor must they express the receptivity of representations through the senses; but though they must have their origin in the nature of the mind, they are neither caused by the object nor bring the object itself into being. In my Dissertation, I was content to explain the nature of intellectual representation in a merely negative way, namely to state that they were not modifications of the mind brought about by the object. However, I silently passed over the further question of how a representation that refers to an object without being in any way affected by it can be possible. (Immanuel Kant, Correspondence, trans. Arnulf Zweig [Cambridge: Cambridge University Press, 1999], Ak. 10:130–131; my emphasis). 23. EL 78. 24. See also SL 21.34: Anaxagoras is celebrated as the man who first gave voice to the thought that Nous, thought, is the principle of the world; that the essence of the world is to be defined as thought. In this, he laid down the foundation for an intellectual view of the universe, the pure shape of which must be logic. Logic has nothing to do with a thought about something which stands outside by itself as the base of thought; nor does it have to do with forms meant to provide mere markings of the truth; rather, the necessary forms of thinking, and its specific determinations, are the content and the ultimate truth itself. 25. J. Kreines, Reason in the World: Hegel’s Metaphysics and Its Philosophical Appeal (Oxford: Oxford University Press, 2015). “That part of the soul, then, which we call mind (by mind I mean that part by which the soul thinks and forms judgment) has no a ctual existence u ntil it thinks” (Aristotle, On the Soul, trans. W. S. Hett [Cambridge, MA: Harvard University Press, 1995], 429a20–25). Kreines has developed a reading of what he calls Hegel’s “meta-
N O T E S T O PAG E S 3 2 1– 3 2 3
373
physics of reason,” in a way that illuminates passages like t hese. But I am not sure how he understands “by virtue of” in “things are what they are by virtue of their concept” (Kreines, Reason in the World, 35), and just how we are to understand the status of The Concept and its various moments, or what such Denkbestimmungen are. 26. There is a fine summary of Hegel’s Aristotle lectures in the history of philosophy lectures by Alfredo Ferrarin in Hegel and Aristotle (Cambridge: Cambridge University Press, 2007), 105–195. 27. Even t hese s imple formulations already introduce a deep complication. In ordinary contexts, we want to say that one has somehow to “see” what one can, by virtue of such “seeing,” assert. Hegel wants to characterize this situation in terms of what he calls the logic of reflection: a “positing” self-conscious spontaneity, or a positing reflection; an “external” reflection that provides the ground of this positing; resulting in a “determinate” reflection that expresses the unity of these moments. Hegel himself admitted that this is the most difficult aspect of his Logic, and the issue w ill reappear in what follows. 28. See the useful account, and several clear formulations in Willem deVries, “Hegel’s Logic and Philosophy of Mind,” in The Age of German Idealism, ed. Robert C. Solomon and Kathleen M. Higgins (New York: Routledge, 1993), 216–253. 29. See also G. R. G. Mure, An Introduction to Hegel (Oxford: Clarendon Press, 1940), 23. 30. See Hans-Peter Falk, Das Wissen in Hegels Wissenschaft der Logik (Munich: Karl Alber, 1983), 170–171, for a good statement of this point. See also Rolf-Peter Horstmann, “Hegel on Objects as Subjects,” in The Oxford Handbook of Hegel, ed. Dean Moyar (Oxford: Oxford University Press, 2017), 121– 138. When Horstmann says that objects must be understood in Hegel as “constituted by their own conceptual activity,” that concepts are “active principles within subjects,” and so that objects are, for Hegel, subjects (132), I don’t take him to mean that Hegel thinks objects are thinking or conscious, but to be referring to what he later calls a “modernized version of Aristotle’s conception of objects as entelechies” (133). I understand that in the manner defended in this paper, but I am not sure Horstmann does. His formulations sometimes make it hard to understand how Hegel deals with the availability of such objects to self-conscious rational subjects, as if the latter is not a Hegelian question. See his characterization of Hegel “exporting (as it were) the object-forming conceptual principles out of the h uman subject and importing them into the Concept as the internal ground of the being of the object (135).” I think Hegel’s position is more like Aristotle’s: substantial form is actualized in the object and in the subject. 31. This difference between “pure” thinking, with only itself as its object, and thinking considered psychologically, is an import ant one throughout
374
N O T E S T O PAG E S 3 2 3 – 3 2 6
ere, and is insisted on just as strongly by Hegel in the Phenomenology of h Spirit §465 and §467. 32. SL 12.23. 33. EL 41; E 63. “Befriedigung” or satisfaction is an important term. It signals that the question of what providing a sufficient ground or reason consists in, what counts as a successful explanation, is a matter of such satisfaction, of when the demand for explanation has been satisfied. This is not something logically obvious, or the same under the diff erent assumptions that govern Hegel’s three main “logics.” See the discussion in Pippin, Hegel’s Realm of Shadows, ch. 7. 34. EL §214, A. See also the end of the Remark to §215 on “overgrasping” [übergreifende] subjectivity, meant to distinguish it from the “one- sided” form with which it is often confused. Michael Theunissen, Sein und Schein: Die kritische Funktion der Hegelschen Logik (Frankfurt: Suhrkamp, 1978), has a number of compelling things to say about Hegel’s general use of the notion of “übergreifen.” 35. EL §237A. 36. SL 12.237. 37. Angelica Nuzzo, “The End of Hegel’s Logic: Absolute Idea as Absolute Method,” in Hegel’s Theory of the Subject, ed. David Carlson (New York: Palgrave Macmillan, 2005), 187–205. In this sense, Nuzzo is quite right that “Hegel replaces the metaphysical Absolute with a theory of absolute cognition, whereby knowledge of the absolute turns into absolute knowing . . . The term absolute for Hegel is no longer substantive but only adjective . . . absolute knowing, absolute idea, absolute spirit (188).” Likewise with her claim that by the absolute idea, Hegel simply means the logic itself (191). I have a stronger notion of Hegel’s language of logical striving (Streben) or impulse (Trieb) than Nuzzo. See again ch. 7 of Pippin, Hegel’s Realm of Shadows. 38. John McDowell, “Responses,” in Sanguinetti and Abath, McDowell and Hegel, 231–258.
Acknowledgments
The prog ress of this volume from conception to completion has taken a number of years, and it might never have been completed had it not been for the help and advice of several people who assisted at key moments. We, the editors of the volume, would like to express our gratitude to Lindsay W aters of Harvard University Press, who guided and supported the volume through its early stages, and also to Ian Malcolm, who became our editor on W aters’s retirement and played a critical role in bringing the volume across the finish line. We are deeply grateful also to our two graduate student editorial assistants, Claudia Hogg-Blake and Kristen De Man, both of whom did crucial work in preparing the volume for publication. And we are indebted to Jim Conant, Jennifer Hornsby, and Jennifer Whiting for advice and assistance at diff erent points along the way. Some of the papers collected in the volume were originally presented at conferences on John McDowell’s work organized by Costas Pagondiotis at the University of Patras, Greece, and we would like to express our gratitude to Professor Pagondiotis for organizing this excellent series of conferences, which provided a vital stimulus for reflection on McDowell’s work and its ongoing significance. Evgenia Mylonaki wants to add her special thanks to Aristides Baltas of the University of Athens, who introduced McDowell’s thought in Greece and who forged an enduring alliance between the academic communities in Pittsburgh and in Greece. Finally, we wish to express our gratitude to John McDowell himself, whose work and whose teaching have had a profound impact on each of us. A volume of criticism is perhaps a poor sort of repayment for that debt, but it is the kind of tribute that philosophers know how to offer, and we hope that McDowell w ill recognize, behind the various quibbles expressed by our 375
376 A ckn o wledgments
contributors, our shared appreciation of his work and the indispensable role it has played in holding open a space for thoughts and questions that might otherw ise have had no place in philosophy—the space in which our responses have been able to take shape. ✣ ✣ ✣
Michael Thompson’s essay was originally published as: Michael Thompson, “Forms of Nature,” in Freiheit: Stuttgarter Hegel-Kongress, ed. Gunnar Hindrichs and Axel Honneth (Frankfurt: Vittorio Klostermann, 2013), 701–735. We would like to thank the publisher, Vittorio Klostermann, for permission to reprint this essay. Andrea Kern’s essay was originally published in German, in slightly dif ferent form, as: Andrea Kern, “Selbstbewusstsein und die Idee der Bildung als ‘immanenetes Moment des Absoluten,’ ” in Objectiver und absoluter Geist nach Hegel, ed. Thomas Oehl and Arthur Kok (Leiden: Brill, 2018), 872–894.
Contributors
Matthew Boyle
Professor of Philosophy, University of Chicago
James Conant
Chester D. Tripp Professor of Philosophy, University of Chicago
Hannah Ginsborg
Willis S. and Marion Slusser Professor of Philosophy, University of California, Berkeley
Jennifer Hornsby
Professor of Philosophy, Birkbeck, University of London
Andrea Kern
University Professor in the History of Philosophy, University of Leipzig
Evgenia Mylonaki
Assistant Professor in Practical Philosophy, University of Patras
Robert Pippin
Evelyn Stefansson Nef Distinguished Service Professor of Social Thought and Philosophy, University of Chicago
Sebastian Rödl
University Professor in Practical Philosophy, University of Leipzig
Michael Thompson Professor of Philosophy, University of Pittsburgh Markos Valaris
Senior Lecturer in Philosophy, University of New South Wales
Jennifer Whiting
Distinguished Professor of Philosophy, University of Pittsburgh
377
Index
a priori concepts, 291–294 a priori knowledge, 292–294, 312–317 absolute (Hegelian term), 286, 297, 305–308 accord (use of an expression in accordance with previous uses), 20–37 action, 11–14, 53–59, 65, 74–79, 217–224, 232 agential control, 11, 154–170 akratês, 226, 246–247, 258, 283 Allison, Henry, 310 Anscombe, G. E. M., 53–59, 75–76, 131, 186, 189, 355n46 Aquinas, Thomas, 53, 64 Aristotle, 8, 14–15, 44–55, 61, 63–64, 69, 72–73, 219–220, 230, 243–279, 282–284, 321–323 Austin, J. L., 121–122 Bildung, 15, 42, 286, 291–299, 304–308 Brewer, Bill, 9, 339, 342n24, 355n43 Burge, Tyler, 117, 118, 372 capacities: cognitive, 6, 8, 112–114, 118, 123–150, 296; conceptual, 9, 84–90, 105–108, 172, 322–323, 363; for knowledge, 10, 117, 150–152; perceptual, 6, 9–10,
14–15, 87, 105–107, 119, 147. See also content: perceptual Cartesianism, 73–74, 111–116, 123–124, 135, 141–150, 185, 348 causal mechanism, 156–166, 170 Cavell, Stanley, 263–265, 337n38 conjunctivism, 116, 123–124, 134–149 content: conceptual, 9, 84, 95, 316; perceptual, 9–10, 83–98, 343n41; representa tional, 91–107, 355n43 Crary, Alice, 221 Dasein, 46–47, 51, 56, 67, 73 Davidson, Donald, 54, 173–177, 183–184, 187, 206, 359n18 deliberation, 8, 11, 13, 46, 170, 173, 273–276; “blueprint model” of practical deliberation, 13, 217–226, 230–231; deliberative phantasia, 250, 262–263; moral-deliberative difficulty, 228–239 Diamond, Cora, 221 discursive activity, 310, 321 disjunctivism, 10, 111–124; resolute vs. irresolute, 134–152 Dretske, Fred, 158 empiricism, 41–43, 318–319 enkratês, 226–228, 245–247, 255–271, 283
379
380 I N D E X Evans, Gareth, 4–5, 288, 363 events: grammar of, 56–58; vs. happenings, 171–178 Fichte, Johann Gottlieb, 311–312, 317 first person (grammatical), 128–136, 146, 149–150, 194–198 first person perspective, 3, 71–73, 209 Fodor, Jerry, 158, 353n31 Foot, Phillipa, 41–42, 65 Ford, Anton, 186 Frankfurt, Harry, 155–159, 164 Frege, Gottlob, 1, 45, 57, 65, 140–151, 311, 322, 344n53, 349n32 Gattungswesen, 71–77 grammar, 56–60, 74, 120–123, 129–136 grounds, regress of, 209 Haugeland, John, 67 Hegel, G. F. W., 1, 14–16, 58, 61, 65, 73, 285–326 Heidegger, Martin, 46–51, 67 hexis, 48–53, 219, 260–261, 275–277 human, 2–15, 31, 38–39, 40–56, 71–79, 153–159, 254–265, 326; h uman agency, 153–159, 185–189, 258; h uman individual, 296–297; h uman soul, 256–258, 262 Humeanism, 60, 75–79 Hylomorphism, 15, 244, 255–256, 282 Hyman, John, 58 idealism, 15–16, 309–320; subjective vs. objective, 310–321 impression, sensory, 88, 106, 206 indefeasibility, 114–118, 125, 138, 348n27, 349n32 infallibility, 112–114 inference, 103–104 intention, 11–12, 53–59, 75–78, 131–132, 154–166, 172–189, 346n21 interpretation, 21–24, 30–34, 37, 72 intuition, 44, 87, 99–110, 287–297 James, Henry, 13, 221, 230–232, 234–235, 237–239
judgment, 9, 83–110, 197–209, 237–238, 349n32; faculty of, 290–294, 317; natural- historical, 62–70; of morality, 220–226; perceptual, 147–152 kalon, 245–248, 265–278 Kant, Immanuel, 1, 9, 14–15, 41–53, 72–78, 83–109, 111–152, 285–326, 332n4 Knowledge: being in a position to know, 10, 30, 118–126, 142–145, 152, 183, 315; capacity for knowledge, 10, 113–131, 150–151, 293; knowledge “from within,” 12–13, 55, 74, 77, 125, 149, 194–216 Korsgaard, Christine, 76 Kripke, Saul, 19–38 life: form of, 53, 60–78, 280–296, 301–308; representation of, 60–61 linguistic meaning, 20, 28–29, 36–37 logic, 92, 100–104, 213–226, 361n11; in Hegel, 16, 311, 321–326; logical form, 63–74; logical gap, 222–226; logical space, 63 Marx, Karl, 41, 71–73 Moore’s Paradox, 130–131 moral experience, 13, 221, 231 moral falsehood, 224–226 moral knowledge, 77, 222–226, 231 Myth of the Given, 86, 206–207 naïve Aristotelianism, 40ff, 41, 43–46, 50, 60, 64, 66 naturalism: liberal, 6, 11–12, 177, 189, 212; minimal, 177, 183; restrictive, 12, 177, 189 nature: first vs. second, 239, 286–296, 308 normativity: inertness vs., 22–23, 30–32; of meaning, 21–22 oscillation, 204–210 Papineau, David, 158 perception: conceptualist view of, 1–2, 83, 85, 86–90, 153, 172; conjunctivist view of, 115–124, 134–140; disjunctivist view of (see disjunctivism); highest common factor
INDEX 381 conception of, 115–151; states of, 89, 105, 114, 164–165; veridical vs. illusory, 91–94, 115, 140–141, 284 perceptual capacity, 6, 9–10, 14–15, 87, 99, 138; Cartesian conception of, 113, 116, 143; defective vs. non-defective use of, 117–127; felicitous vs. non-felicitous use of, 141–152 phronêsis, 14, 219–230. See also practical understanding phronimos, 46–48, 226–229, 246, 272–275 Plato, 1, 4, 46, 243–269, 313, 320, 364n13 practical intellect (practical nous), 43–49, 245–246, 256–263 practical knowledge, 44–45, 53–56, 59, 65, 131, 188, 252 practical naturalism, 40–43, 53, 66 practical reason, 40–41, 60, 66, 77; practical reasoning, 217, 219–222, 224–225, 227–231, 236–239 practical syllogism, 220–227, 264 practical understanding, 45–47, 55. See also phronêsis pure thinking, 311–325 Pyke, Stephen, 2 quietism, 7, 19–20, 33, 37–39, 201 rationality: additive vs. transformative conceptions of, 3–11, 69–71, 251–252, 253–256, 256–263, 287–290, 307–308 reflective consciousness, 154–160, 170 represent at ionalism about perception, 91–110 res, de re, 167, 177–178, 183–184 Ross, W.D., 50 rule, 7, 19–23, 101–104, 264, 335–336n33; of thinking, 292–294, 323 Ryle, Gilbert, 53, 120–123 Schelling, Friedrich Wilhelm Joseph, 312, 314, 320 Schopenhauer, Arthur, 44 scientism, 2–3, 215–216 Searle, John, 176–177, 186–187
second nature, 8, 15, 40–43, 68–69, 286, 296. See also nature: first vs. second self-a lienation, 138–140, 149–159 self-consciousness, 12–13, 66–67, 70–79, 154, 156, 194, 197–198; exercises of a capacity, 128–140; universal, 302–307 Sellars, Wilfrid, 86, 92–93, 98, 103, 126, 206, 319 semantics, 24, 333–334n22 sensation, 89 sensibility, 99, 285–299 sideways on (view from), 133, 149–152, 172–173, 194–195, 199–201, 208–209, 212–216 skepticism, 33–39, 115, 141 skill / skilled agency, 159–169, 185–188, 260. See also technê sophia, 48–52, 223, 248 Stanley, Jason, 158, 356n47, 357n50 Steward, Helen, 158 Stroud, Barry, 119, 125–126, 129, 151 subjectivity, 13, 112, 290–294, 305–306, 324–326. See also idealism: subjective vs. objective technê, 48, 53, 260–272. See also skill teleology, 166, 175–176, 184, 244–245 Thompson, Michael, 7–8, 166–167, 301, 358–359n11 Transcendental Deduction, 105, 287–294, 324 Travis, Charles, 9, 84, 341n18, 361n11 two-tier accounts, 156–157, 163 understanding, 214–216; and sensibility, 285–290, 293, 298; faculty of, 291–300, 342–343n32 unity: of apperception, 124, 130; synthetic, 287–289 universal: concepts, 48–49, 65–66, 71–72; in ethics, 218–231, 264–274; in perception, 214–216; universality of self-consciousness. See also self-consciousness Vasiliou, Iakovos, 244, 269, 272 Velleman, David, 154–164
382 I N D E X virtue, 8, 13–14, 45–48, 229, 243–250, 253, 270–273, 348n27. See also phronêsis warrant, 104, 117–118, 125 Wiggins, David, 45–46, 50
Wittgenstein, Ludwig, 1, 7, 19–20, 23–32, 68–69, 122, 164, 263–265, 309, 311, 346–347n22 Wood, Allen, 71–72 world, 12–13, 212–216, 222–223, 237–239; external, 112–113