Naturalism and Constructivism in Metaethics [1 ed.] 9781443859370, 9781443856737

In this collection of essays, several authors, belonging to different generations and philosophical traditions, discuss

163 81 1MB

English Pages 212 Year 2014

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Naturalism and Constructivism in Metaethics [1 ed.]
 9781443859370, 9781443856737

  • 0 0 1
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Naturalism and Constructivism in Metaethics

Naturalism and Constructivism in Metaethics

Edited by

Sofia Bonicalzi, Leonardo Caffo and Mattia Sorgon

Naturalism and Constructivism in Metaethics, Edited by Sofia Bonicalzi, Leonardo Caffo and Mattia Sorgon This book first published 2014 Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2014 by Sofia Bonicalzi, Leonardo Caffo, Mattia Sorgon and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-5673-8, ISBN (13): 978-1-4438-5673-7

TABLE OF CONTENTS

Preface ....................................................................................................... vii Part One Empathy and Nature .................................................................................... 2 Leonardo Caffo Naturalism, Free Will, and Different Approaches to the Is/Ought Problem ...... 7 Andrea Lavazza Some Elements for a Critical Review of the Concept of “Naturalistic Fallacy” ..................................................................................................... 24 Michele Borri Empathy and Sympathy: From Description to Prescription ...................... 51 Sarah Songhorian Part Two Defining Practical Reasoning: Constructivism and Instrumental Reason ....................................................................................................... 74 Sofia Bonicalzi Practical Reasoning Between Abstraction and Idealisation: Onora O’Neill and John Rawls .................................................................. 85 Francesca Vitale Towards a Gendered Rational Choice Theory ......................................... 118 Martina Belmonte Part Three Naturalism, Deontic Logic and Cognitive Science .................................. 146 Mattia Sorgon

vi

Table of Contents

The Debate on Naturalism in Contemporary Ethics ................................ 152 Luciana Ceri The OIC/PAP Dispute: Two Ways of Interpreting ‘Ought’ Implies ‘Can’ ........................................................................................... 172 Guglielmo Feis Cognitive Neuroscience and Animal Consciousness............................... 182 Matteo Grasso Contributors ............................................................................................. 204

PREFACE AND NOW, ETHICS SOFIA BONICALZI, LEONARDO CAFFO AND MATTIA SORGON

The best things always happen by chance: they just happen, they fall outside the meticulous projects in which we pretend to locate every manifestation of our life. In line with this tradition, the book you are holding in your hands stems from a conference whose papers, at least initially, were not intended for publication. In June 2012 we organised, a conference called “Workshop on Ethics: a Junior-Senior Debate” at the Department of Philosophy of the University of Milan, which we think it is worth saying a few words about. First of all we have to thank the institutions that made this possible: apart from the already mentioned Department of Philosophy of the University of Milan we must express our gratitude to the editorial staff of Rivista Italiana di Filosofia Analitica Junior (Junior Italian Journal of Analytical Philosophy) among the official organs of the Italian Society of Analytical Philosophy, for helping us in the refereeing process of the contributions that make up this book. Although its activities have ceased for contingent reasons, we also wish to thank the Research Centre for Philosophy “Doiè”: the enthusiasm and economic contribution of the people involved made the development of the conference possible. One final institutional heartfelt thanks goes to the ESAP, European Society for Analytic Philosophy, for welcoming our workshop among its prestigious international events. The subject matter of the conference was obviously ethics in its various declinations, but you might wonder why there was so much excitement over one particular conference on ethics of the many held all over the world. The answer is simple and it lies in the unique and innovative formula of the event. Divided into three sections, it had six young researchers, Master’s and PhD students, engaging in dialogue with three renowned professors: Carla Bagnoli, Luciana Ceri and Mario De

viii

Preface

Caro. To them, we would like to extend our sincerest gratitude for the enthusiasm and keenness they showed in this complex project which we can now consider not only successful, but also worthy of a new edition in the near future. This book is not just a mere re-proposition of the papers discussed on that marvellous day, it is instead the result of a collaboration that was shaped and inspired by that day: some authors joined in later, others were unable to participate, but the fundamental idea has always been that of comparing two philosophical generations in the context of the most decisive issue for the “life to come”: ethics. A complex dialogue, but a crucial one, not only for philosophy but also for everyday life: we do think we succeeded in the task we set ourselves. We shall let the readers judge this for themselves.

A Note This book is dedicated to Luca Magni, too fragile for this world, and to Ettore Brocca, who has given up philosophy but, aware of it or not, started all this with his intelligence and the passion of a deep friendship. We are deeply indebted to Tommaso Bertolotti who first revised and refined this manuscript, providing insightful and constructive comments, and to Merope Ippiotis for her ultimate professional proofreading. We would like to express our deepest gratitude to Carlo Sandroni, for his invaluable support and encouragement. Further thanks go to Sarah De Sanctis, who translated from Italian and revised the greater part of this book. Her intelligence, competence and humility are the proof that Quine was right: there is something.

PART ONE

EMPATHY AND NATURE LEONARDO CAFFO (LABONT/UNIVERSITY OF TORINO)

1 The purpose of this section is to reflect upon two entities—”empathy” and “nature”—in relation to contemporary ethics. The main thesis of this field of study is inscribed in an analysis of the nature of mind, applied to the foundation of morals. The unconscious part of the human mind is the key to decipher in what way morals and ethics, the private and the social, are influenced by our biological makeup. That is, they are influenced by the frames built into that machine we call Homo Sapiens. These innate aspects of the mind connect and join us to one another at a deep level, defining the general meaning that allows the very existence of our species’ ethical life. They establish an ontological foundation of human morality, which, since the discovery of mirror neurons, has been fostering the (now unexciting) idea that if on the one hand ethics is essentially a social construction, on the other hand morality—on which human coexistence is built—is more innate and absolute than even Plato could have imagined. (However, the realism or semi-realism defended by Lavazza in this volume, about the is/ought question, is the best theory to describe how this whole issue is still an open question). The rigorous arguments presented in the essays of this section are developed within this framework, which is significantly supported by contemporary literature, and trigger the reflection about what makes ethical enunciates absolute or relative. Let us consider the following notion: if there are valid arguments in favour of an “altruistic” thesis in a technical (cognitive) sense, for which—against Hobbes and those philosophical traditions that have elaborated on his views—humans live for other humans and can develop their nature only within a reciprocal cooperation, then a moral theory which aims at dividing rather than uniting human beings (favouring for instance egoism and interest rather than benevolence and altruism) is wrong not only from certain viewpoints, but in an absolute sense. The power deriving from this use of cognitive science (and analytic philosophy) should be clear: by

Leonardo Caffo

3

entering the vast domain of moral theory, it can guarantee an empirical validation of widespread intuitions. Here is an example.

2 When Umberto Bossi—unfortunately a famous Italian politician—calls for secession, he is not only acting “inelegantly” towards Italian democracy, he is also violating human nature, which reaches its goal and is thus fulfilled in the exclusive coexistence of different individuals. In fact, various human beings join together in a single epiphenomenon called “empathy”, which regulates our revulsion at the abandonment at sea of illegal immigrants, unjustified killings, etc. However, political practices violate the nature of our species in many other ways. In modern times everything is regulated through the social object we call “money”. As a social object, money is construed on the basis of a “weak textualism” [(I am herein uncritically accepting the theory proposed by Derrida/Ferraris (Ferraris 2012)] and, unlike stones or seas, it would not exist if human beings disappeared from the Earth. Can ethics, i.e. acting for the other, follow the trend of money, markets, and other phenomena which were invented to serve us in our daily lives, but which now—in a very disheartening turn of events—have become masters of their own creators? To answer this question all we have to do is “look outside:” the following analyses will illustrate that we can build social and ethical relations on psychologically founded frames, and that “morality” means “wellness” inasmuch as it stands for the wellness of community. As discussed, for instance, in an essay by Sarah Songhorian, here lies the ultimate confutation of the egoism that regulates the wretched power relations of our age. Paraphrasing Lakoff’s scientific thinking into a political setting, to appeal to the masses in an attempt to display some sort of factual evidence is useless. What really matters is to foster coexistence and mutual aid by encouraging the individual’s intrinsic moral capability, without encountering the typical naturalistic fallacy, as the essay by Michele Borri (in this book) aptly demonstrates. Until we will learn to associate our good with the common good, we will always have a “stranger” at our door: at first it will be “the nigger”, then “the southerner”, and finally “the woman”. Cognitive science tells us that we live through metaphors, and that human beings need to go beyond the brutality of the world in order to saturate the need for desire that is characteristic of our species. Man is not a “wolf among men”, but a being that exists, acts and speaks as part of a complex structure populated by other living beings, which complete each other especially on the basis of their difference. As, after Chomsky,

4

Empathy and Nature

language cannot exist for a Robinson-Crusoe-like figure born and grown up in isolation, similarly, life only makes sense if there is cooperation among humans who act and live because others have acted and lived, “dwarfs standing on the shoulders of giants.” In contemporary research, the analysis of ethics through scientific data is almost entirely in fieri, hence this volume is meant to fill a relevant gap in the existing and still developing literature on these matters.

3 The theses exposed in this book are, in fact, of great importance to the recent and expanding field of study that stems from the interaction between moral philosophy and cognitive science. We are not referring to Neuroethics, as some might think, as the latter should investigate the connection between neuroscience and the individual’s moral cognition. The essence of our discussion is more philosophical and aims to answer— and better express—questions such as: in there anything like a just life, beyond all interpretations? Can morality exist regardless of social constructions? To what extent is ethics natural and to what extent is it cultural? There is one key underlying idea that runs through these studies: the moral point of view stands far above our personal interests, as it is objective rather than contingent. This claim is clearly deployed against the relativistic perspective in philosophy, for which truth as a theoretical entity does not exist (I am thinking, for instance, of Harman’s arguments). The intuition we are referring to becomes even more crucial if we consider Sarah Songhorian’s analysis, which moves the axis of empathy from a descriptive to a prescriptive dimension. Obviously, we are not arguing in favour of an ingenuous objective criterion, as we are well aware of the possibility of correlation between objective moral assertions and subjective decisions. (Kutschera 1991, 65 ff.; Kant 2007, “Analytics,” par. 7). The ethics underlying the advocated stance could probably be better defined as a metaethics, arguing that Right is set close to the biological dimension of humans—i.e. to act socially or for the other—and Wrong gets bigger the further one gets away from this biological being (recalling Hobbes and the argument against capitalism). In a 1958 study (Baier 1958) which philosophers today are very familiar with, Kurt Baier argued that moral perspective “can show some ‘good reasons’ to distinguish what is right from what is wrong” (Da Re 2010, 14). This thesis is today widely accepted, though more “reasons” are necessary in order to claim that social acts are right when they are close to human nature and wrong when they move in the opposite direction.

Leonardo Caffo

5

4 A world where human actions are moved by a logic that is egotistic or self-interested is an unhappier world for everyone. Nonetheless, we all know that our reality is shaped on this model: even though self-interest is not always blind to judicious moral reflections and is surely directed at social construction, in such a social sphere the human being is totally subdued by social facts—while the growing volume of studies we are analysing would rather have it the opposite way. Whereas Baier argues that the moral point of view is not really a “point of view”, the former model (caring for others) is necessarily better than the latter (egotism), as the former is closer to objectivity—that is, to “morality in itself” (Baier 1958, 181). The strength of such an assertion is self-evident: existing norms are neither intangible nor unchangeable, rather, they necessitate a constant replacement with “other possible, ideal norms” (Ib., 174). Reading this, one might wonder: what kind of ethics should we have, then? How can we do justice to a social and moral project that is founded on human biology? And what are the boundaries, if we are to act for the other? In The Republic Plato presents for the first time what can be defined as “group ethics”. Polemarchus follows Cephalus in the stressful dialogue with Socrates (Plato, The Republic, 332 a-b) and states that justice is founded on giving and returning to everyone what they deserve, in a reciprocal gesture of help. However, for Cephalus this should be a preferential kind of treatment reserved for friends, while enemies should be fought until death. Of course Socrates, insatiable as always, pushes Cephalus to yield to the idea that damaging another is always despicable and that we can never tell for certain our friends from our enemies. Rather than an objection, Socrates’ statement is the extension of an ethical theory that, if properly discussed, can be helpful to us. Paraphrasing what Cephalus said on the basis of the course we have charted in this chapter, we can obtain an ethics for which, on the one hand, one should give and return to others what they are entitled to (for example, a fair salary or the possibility to actively make use of social facts). On the other hand, such an ethics would make acting for others necessary, because if the other is inscribed in us (as argued by neuroscience and empathy studies, supported by the recent discovery of mirror neurons), and if reciprocal action rightly develops our nature, then the conflict between enemy and friend will be overcome. Group ethics becomes ethics of the other. The very other has to be slowly rejected as other than oneself (as Jacques Derrida claimed in his The Animal that Therefore I Am) in order for the discourse to take shape. If we consider ethics in this light, Hegel’s distinction between “ethics”

6

Empathy and Nature

(Sittlichkeit) and “morality” (Moralität) falls apart—as specialized literature has already almost entirely accepted. In fact, if morality is the personal moral dimension, while ethics is the realization of morality within the sphere of customs and institutions like family, society and state, then both entities cannot be considered from an internal/external perspective anymore, because they collapse into one single dimension. In his Oneself as Another, Paul Ricœur finely argues that otherness is intimately implied in ipseity and viceversa (Ricœur 1993, 75-79). The desire to live happily with ourselves coincides with the desire to live happily with and for others, so that —as Ricœur reminds us—we may live within just institutions in which we know the following things to be necessary: the dimension of freedom, seen as the saturation of the void of desire (freedom as development), and the subjugation of social facts to what is human—and not the opposite. But what are the boundaries, if there are any, of this “other than oneself” which we have discussed? For whom are we to act? And with whom are we to realize our biological course? When can we know we have committed an “immoral” action, i.e. contrary to the mores we are outlining on the basis of cognitive studies? Specialized studies and some of the arguments included in this chapter and throughout the book aim to answer these questions. This volume aimss to contribute to a growing research field, which aims at showing (in analytic philosophy, but not only) what can be defined as a proper “ethical turn”.

Bibliography Baier, Kurt. 1958. The Moral Point of View. A Rational Basis of Ethics. Ithaca (NY)-London: Cornell University Press. Da Re, Antonio. 2010. Le parole dell’etica. Milano: Bruno Mondadori. Kant, Immanuel. 2007. Critique of Practical Reason. Cambridge: Cambridge University Press. Kutschera, Franz von. 1991. Fondamenti dell’etica, Milano: Franco Angeli. Ferraris, Maurizio. 2013. Documentality: Why It Is Necessary to Leave Traces, New York: Fordham University Press. Ricœur, Paul. 1993. Sé come un altro. Milano: Jaca Book.

NATURALISM, FREE WILL, AND DIFFERENT APPROACHES TO THE IS/OUGHT PROBLEM ANDREA LAVAZZA (CENTRO UNIVERSITARIO INTERNAZIONALE, AREZZO)

Abstract The chapter considers recent attempts to derive a prescription from a description or, more generally, from a matter of fact on the basis of neuroscience findings about free will and brain functioning. But it is argued that—within the framework of the attempted naturalization of morals— we must face improper inferences from scientific descriptions and explanations to normative concepts. The inferences taken into consideration are improper first of all because relevant knowledge is insufficient or insufficiently corroborated. Secondly, inferences are improper because, as we attempt to argue, it is not legitimate to move from descriptions of average phenomena to general prescriptions. This is not motivated only by the impossibility of deriving ought from is contained in the literal interpretation of Hume’s Law, but also because of the very nature of normative judgments expressed by human beings.

1. Introduction The metaphysical idea of free will has always been at the centre of the debate on the meta-ethical criteria that can legitimately be adopted in a moral system. Nevertheless, incompatibilist views—which negate free will as at odds with determinism as a metaphysical thesis and empirical reality—have never promised to subvert the moral reflection as credibly as they are starting to do now. The well-known experiments of Benjamin Libet (Libet et al. 1983) and John-Dylan Haynes and his research group (Soon et al. 2008; 2013) attempt to demonstrate that, rather than being agents able to set their own course of action and take responsibility for it, we —together with our conscious awareness—are passive spectators who watch the unfolding of actions that our brain has “launched” before we even realized it (Prior to these studies, we had been fully convinced that

8

Naturalism, Free Will, and Different Approaches

such an action could only be the fruit of our determination.). According to many students of empirical psychology, the phenomenon known as confabulation, by which we try to motivate ex-post choices made on the basis of environmental clues that orient us “automatically”, is the flip side of our being “autonomous”, by which we respond to impulses from the physical and human world according to models we learned over the course of our evolutionary history in order to adapt to our environment and reproduce. Obviously, this is a simplified description of an extreme case, but it is not far from the description of our “moral” life provided by some naturalistic approaches. They are descriptions in which classical categories are challenged by a new, naturalized understanding of how human behaviour arises, and to what extent many of the meta-ethics and moral prescriptions adopted over the course of human history are “unrealistic” and thus “irrational”, with the exception of those that can be directly traced back to our species’ adaptive mechanisms. It is interesting to note that the so-called Hume’s Law appears to be an obstacle for this type of naturalism, and yet it is perceived to be an ally of non-cognitivism, and thus questioned by moral realists. The impossibility of deriving a prescription from a description or, more generally, from a matter of fact, is one of the most controversial meta-ethical principles, and has been interpreted in many ways and questioned by many authors on the basis of various strategies. Hume’s Law establishes that one cannot derive a prescriptive or normative statement from a descriptive statement, or from a matter of fact, without making a categorial mistake. In every system of morality, which I have hitherto met with, I have always remarked, that the author proceeds for some time in the ordinary ways of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when all of a sudden I am surprised to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, 'tis necessary that it should be observed and explained; and at the same time that a reason should be given; for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it (Hume 1739, 3,1,1)

From the point of view of formal deductive logic, “deriving” means building a valid argument that moves from true premises to true conclusions. An example is the classic syllogism “All men are mortal,

Andrea Lavazza

9

Socrates is a man, so Socrates is mortal”. It is thus not possible to derive (in terms of a formally valid logical argument) a proposition on how we should act from a series of descriptions or events. The so-called is/ought problem was also recently challenged and — whether implicitly or explicitly—declared obsolete, or deemed nonbinding by supporters of certain forms of the naturalization of morals, since “you cannot derive an ought from an is” is considered a powerful argument against the possibility of deriving moral prescriptions from that which is truthfully described by physical sciences (These are the most recent meta-ethical trends, which tend to set themselves apart from the naturalism debate launched by Moore). In this chapter I will thus consider certain attempts to circumvent Hume’s Law, including a few that I think should be subject to criticism and one that seems to be more plausible. These attempts are framed in various different philosophical and naturalistic approaches.

2. Neurophilosophy In her recent book Braintrust, Patricia S. Churchland proposed a way to move beyond Hume’s Law. She claims that: In a much broader sense of “infer” than derive you can infer (figure out) what you ought to do, drawing on knowledge, perceptions, emotions, and understandings, and balancing considerations against each other. We do it constantly, in both the physical and social worlds. In matters of health, animal husbandry, horticulture, carpentry, education of the young, and a host of other practical domains, we regularly figure out what we ought to do based on the facts of the case, and our background understanding. I have a horrendous toothache? I ought to see a dentist. There is a fire on the stove? I ought to throw baking soda on it. The bear is on my path? I ought to walk quietly, humming to myself, in the orthogonal direction. What gets us around the world is mainly not logical deduction (derivation). (…) The important point for my project, therefore, is straightforward: that you cannot derive an ought from an is has very little bearing so far as in-theworld problem solving is concerned. (Churchland 2011, 6-7).

Formally, this type of inference resembles the attempts of ethical naturalists (who affirm that moral truth exists) to respond to Hume’s Law, in particular those of A. MacIntyre. They maintain that for person P to achieve objective O, P should undertake action A. I should point out that I am not interested in staking out a position on cognitivism or noncognitivism in ethics, or on moral realism. What I aim to highlight is that it is not possible to overcome the is-ought problem in this way in order to

10

Naturalism, Free Will, and Different Approaches

naturalize, in a neuroscientific sense, certain matters that are part of the moral sphere. Indeed, the argument put forward by Churchland is only valid to the extent that it has an external criterion of “truthfulness”, linked to the efficacy of the action on the world. For example, when one has a toothache, going to see an ophthalmologist neither soothes the pain nor helps in treating the abscess. A dentist’s care, instead, is effective, because dentists possess the required knowledge about teeth, based on the current scientific consensus. If it were a common occurrence to find fake dentists who attempt to sell themselves as professionals without possessing the right skills, the answer to the question “what should I do when I have a toothache?” would be: visit a dentist who can show you their degree certificate. Similarly, when we run across a hungry bear, our goal is to get away unharmed. The means adopted to achieve these goals are those available in that specific moment. Not all means, of course, but just those that are effective at keeping us safe. The efficacy of such means will be clear to all when they see us coming back safe and sound from the bear’s den (or from the dentist’s parlour!). If at time t1 we are facing the hungry bear and adopt a certain behaviour, and if at time t2 we are safe and sound, assuming all other variables remain constant, no one will be able to doubt that the behaviour we adopted is effective and recommended if we want to survive a bear attack. This is what we mean by the external criterion of “truthfulness”, and it is also how science works. Science attempts to describe how the world works, and on the basis of this description we can draw inferences on the best way to achieve results, defined as states of affairs in the world. However, these types of inferences become problematic precisely when they are applied in “moral” contexts. For the purposes of this paper, a minimal, ad hoc description of moral contexts suffices. By moral contexts we mean those in which results are at stake, and defining “results” as states of affairs in the world that can be prohibited, allowed, or mandatory, with social or legal enforcement. Let us return to the type of inference introduced by Patricia Churchland. Suppose that empirical psychology tells us (as indeed it seems to do) that the best and easiest way to win an election is to denigrate one’s adversary and appeal to the simplest, unconscious emotions of citizens and their stereotypes (including those on the alleged superiority of one race over another). This entails that the answer to the question “What should a candidate do during an election?” must be to denigrate one’s adversary and appeal to the simplest, unconscious emotions of citizens and their stereotypes (including those to do with the alleged superiority of one race over another).

Andrea Lavazza

11

While all may agree about the efficacy of such a campaign, many would argue that it constitutes “bad politics”, and that it leads to a violent debate in which the true problems of society are overlooked. The outcome of the electoral process would probably be disastrous, also from a strictly “non-moral” point of view, since it could lead to such a tense and unruly social climate—with clashes between different groups of citizens (along ethnic fault lines, for example)—that the very politicians who won the election could hardly manage the country at all, as they would be unable to meet the approval of their very voters. The final outcome would be that, much like aspiring witch-doctors, they would fail to be re-elected. It is thus not unreasonable to suppose that even a certain number of aspiring politicians would reject this advice on how to conduct an election. They would thus reject the inference that moves from an is to an ought, even though it is, in some aspects, an effective inference, which delivers the results it promises, this being the “ought to be case” one aims for, based on the “is the case” of those preferences and predispositions displayed by voters when they choose which candidate to support. The difference lies in the assessment of how the result is achieved and also in the result itself, namely the election of candidates who show a disinterest in public issues and adopt an aggressive, rather than thoughtful, political attitude. This is the moral context in which an inference that is less rigid compared to Hume’s Law turns out to be controversial and, in the final analysis, unconvincing. Of course, one could object that Churchland’s arguments can also be taken to the extreme, and thus that the counterexample of neuro-politics cannot be generalized. For example, one could escape unharmed from the bear’s den by killing the bear. This is a conduct that many would disapprove of—at least in the absence of clear and present danger that could constitute grounds for self-defence. The commonest behaviour is instead to escape, which can take a number of different forms. Nevertheless, those inferences, as postulated by Churchland, aim to overcome the obstacle of Hume’s Law while moving towards ethics deriving from neuroscience. They are only valid within the framework of practical knowledge embodying the “truthfulness” criterion. The latter is to be understood as a reflection of the actual state of the world, and thus universally accepted and prevailing over criteria adopted in “moral” frameworks, about which the assessments differ, whatever the origin of these differences. Indeed, it could be argued that different people have different moral intuitions, or that they explicitly adhere to different moral codes, and that any convergence on shared moral rules takes place on the

12

Naturalism, Free Will, and Different Approaches

basis of reason and conviction, or on the basis of emotional states. But it does not seem that rules of behaviour, as they include specific purposes and specify what is preferable, can be adopted on the basis of explanations of the state of the world (This does not mean that prescriptions should not take into account material contingencies, such as a scarcity of medical resources when one is forced to choose who to treat and who not to treat, or the psychological discoveries concerning the “force of will” in the education of children). One may also argue that Churchland aims to propose these inferences in a prescriptive way precisely in order to overcome the differences that emerge in controversial situations. If we move from is to ought, consensus can only be reached thanks to science and “truthfulness” criteria that are objectively and intersubjectively verifiable. But this is a conventional proposal, which in many cases clashes both with emotivism (moral intuitions are stronger than scientific descriptions) and with rationalism (which is contrary by definition to such a proposal). What would remain is a petition of principle, an attempt at rhetorical persuasion. The acceptance of Churchland’s inferences thus becomes an empirical question that can be measured over time in different social contexts. Perhaps the equation between is and ought will be made by a small group of neuroscientists, while failing to convince all others, since it appears to lack the coherence and force of a persuasive argument. It is interesting to note that recent arguments have been proposed against the supposed obligation to change one’s belief about ethics because of discoveries in neuroscience (Kaposy 2010). In that case, criticism was aimed at the “normative claim that ethical thought ought to reflect the conclusions of neuroscience that contest concepts such as free will, selfhood, and personhood.” Kaposy’s main argument is that “from the perspective of instrumental rationality, it is rational to preserve our belief in free will, selfhood, and personhood”, even though scientific evidence seems to question these very concepts, which play an essential role in the moral sphere. My line of argument differs from that of Kaposy—whose basic approach I share—in two aspects. The first concerns the need for a coherent framework for one’s own beliefs, which cannot disregard what science tells us about reality, both with regards to the physical world and to human beings and their functioning. This does not mean that we must adhere to a complete naturalization of knowledge, which places scientific research (generally intended as comprising natural sciences only) as the sole source of genuine knowledge or understanding of the world. Therefore, if at the basis of certain concepts—including some that are

Andrea Lavazza

13

“highly valued components of our ethical worldview” (Kaposy 2010)— there are elements questioned by neuroscience, those concepts should be modified. The second aspect concerns the nature of the concepts under consideration. One, free will, undoubtedly has an empirical component that cannot be ignored. The others, selfhood and personhood, are in my opinion essentially normative concepts, for which criticism based on neuroscience is either off the mark (personhood) or only partially applicable (selfhood). One could say that the recent success of scientific naturalism is underpinned by the so-called “success of science” argument (cf. De Caro and Macarthur 2010). Indeed, there is no doubt that science and its applications can be portrayed as a series of epistemic successes that allowed us to establish our domain over fields that appeared mysterious and threatening just a few centuries ago. According to proponents of scientific naturalism, a sort of super-induction is thus possible, which allows us to conclude that even phenomena that currently appear not to be “naturalizable”—meaning that they can’t be explained in terms of natural science—will become so one day (or at least may become so in principle). However, concepts that are intrinsically normative in character, such as personal identity, moral responsibility, intentionality, and—at least in part—free will itself, continue to elude the possibility of a scientific explanation. Along these lines, Sam Harris maintains that morality “relates to the intentions and behaviours that affect the well-being of conscious creatures” (2010), and that science is thus equipped with the best tools to identify all that contributes to this well-being, since the latter is linked to cerebral states. The point here is that well-being as a criterion has been on offer on the metaethics market for a long time. It represents the point of view of classical and contemporary utilitarianisms (in their various shades) and it is engaged in an unsettled competition with other meta-ethical points of view, such as deontological virtues or morals, according to which what is obtained (well-being) is not as important as how it is obtained (that is, whether the right procedures have been followed). Science therefore does not seem to be any more able to settle the meta-ethical question today than it was in the past, in spite of increased empirical knowledge.

3. From Neuroscience to the “Abolition” of Personhood There is widespread consensus about the fact that the concept of personhood has a significant prescriptive value and is not, by itself, a natural category, unlike human beings as a biological species. A testament

14

Naturalism, Free Will, and Different Approaches

of this is the fact that many scholars question the equation between human being and person, as they do not deem it proper to confer the ‘title’ of person to each and every member of the species Homo sapiens sapiens at all stages of their lives. Nevertheless, thanks to the rapid progress of cognitive neuroscience, several researches began to use neurobiological criteria in order to reassess or discredit the concept of personhood by defining it as illusory. In particular, Farah and Heberlein (2007) maintain that there is an innate cerebral network, comprising four specific areas of the brain, which automatically produces the perception of a particular category of objects that are then defined as persons. This hypothesis is based on an increasing body of experimental data, which are individually well supported by the extant evidence. Due to the difficulties associated with defining personhood in the most controversial bioethical cases, they suggest that the concept itself should be abandoned as the outcome of an evolutionary and adaptive mechanism that has become inadequate in the light of the dilemmas created by contemporary medicine. In other words, Farah and Heberlein erase the is/ought distinction— albeit not explicitly—in an inverse direction compared to what Churchland proposed. While the latter moved from neurobiological evidence to a moral prescription, the former uses neurobiological evidence to argue against a widely used prescriptive concept. According to Farah and Heberlein, one could argue that we come into the world genetically programmed to represent people as something distinct from the other entities of the external world. This hard-to-silence system is autonomous, inasmuch as it fires, when activated by a wide range of stimuli, some of which–—such as drawings, geometrical shapes, including irrelevant or counterintuitive ones—we know not to be persons, thanks to the correcting activity performed by other parts of our brain. Farah and Heberlein thus maintain that the distinctive criteria of the ‘person’ concept are not easily set apart if compared, for example, to the ‘plant’ category, which possesses a degree of objective reality that a person would not have. The brain is innately equipped to deal with certain kinds of stimuli in a special manner, and the perception of certain activating characteristics, such as human faces, bodies, or movements, sets which mode of processing should be used. The result is that we perceive and ponder such characteristics using a dedicated cerebral system, and we do so in an innate, automatic, and insuppressible manner. Our impression that the world holds two radically different categories of things—persons and non-persons—might be the result of the periodical activation of the neural network that helps us recognize persons by certain stimuli, and does

Andrea Lavazza

15

not reflect a fundamental distinction between stimuli that activate it and stimuli that do not. Mental representations—Farah and Heberlein write— can exist and be activated by stimuli even without systematically grasping fundamental categories of the natural world. Obviously, they note, certain “things” have minds, and others do not (although they do not specify what they mean by mind, given their naturalistic presuppositions). Farah and Heberlein stress two important aspects of the “person recognition network”. The first is its separation from systems used for the recognition of other entities. This causes the illusion that persons and nonpersons are two different categories of things, in spite of our inability to establish, in principle, a distinction between the two. In other words, we have the intuition that—although human beings, animals, and computers may have varying degrees of intelligence, ability to communicate, and self-awareness, and although we are unable to identify a break in the continuum between a healthy individual, an individual in a coma, a stupid machine, and an intelligent machine—we have the impression that certain entities are people and others are not. The second significant aspect has to do with the autonomy of the “person recognition network”, with its tendency to be activated by certain stimuli (faces, behaviours). For this reason—they claim—it is difficult to abandon the idea of person even in front of a patient in a vegetative state or a foetus. If we had a separate cerebral system for plants—this is the paradox they use—we would have the impulse to smell the flowers on our friends’ Hawaiian shirts, or to water green rugs. We evolved this way— Farah and Heberlein argue—, because it was fundamentally important for us to interact with our fellow humans, and to recognize them always and under all circumstances. Identifying an object as a human being is more functional to our survival and a more effective adaptation that identifying a human being as an object. Additionally, such an adaptation must have arisen in a physical world where uncertain or ambiguous cases were reduced to a minimum: there were no foetal ultrasounds, people did not live long enough to develop Alzheimer’s disease, and traumas that today cause vegetative states were invariably lethal. To sum things up, the concept of person is the result of a sort of illusion, much like a visual illusion: the result of cerebral mechanisms that depict the world in an untruthful manner under certain circumstances. The authors themselves acknowledge that their analysis is nihilistic, since it undermines many ethical systems. If persons are not part of the basic landscape of the world, then there are no tangible elements that define the status of a given being as a person, and thus there is no chance of obtaining objective criteria for the ‘person’ concept itself (cf. Strawson

16

Naturalism, Free Will, and Different Approaches

1959). This underlies their suggestion to fall back on a utilitarian approach (even if it does not resolve the concepts that were meant to be tackled), which considers the ability of living beings to express certain psychological characters (intelligence, self-awareness...) and attempts to protect the interests of these beings. The fact remains that, as individuals immersed in our social lives, it matters little to us that the ‘person’ concept is an illusion. Indeed, we cannot biologically re-programme ourselves (at least not for now...). Furthermore, we cannot deny that the ‘person concept,’ as inadequate as it is in a world that is far different from the one we evolved in, is still of use to us in many practical fields (from child care to showing respect for others…). The fact that, in this case, it does not seem possible to make Hume’s Law obsolete may be demonstrated by presenting a defence of the ‘person’ concept that does not deny the neurobiological data presented by the two authors. Indeed, the ‘person’ concept seems to contain a cultural, prescriptive component that does not arise solely from the functioning of specific cerebral areas. For example, without straying from our cultural traditions, it was recently summarized—on the back of a vast trove of anthropological, historical, and critical literature—that in the Iliad and the Odyssey depictions of human beings are not based on the distinctions between organic and inorganic, external and interior, physical and spiritual, but rather on different areas and surfaces of the body (thymos, phrenes, noos…) which give life, together with all the other elements they interact with, to a kaleidoscopic game1. In Homer’s work, the human body is never depicted by a single term, but as a plurality of moving limbs. In the Iliad, human and divine events are told through the use of similes comparing them to animal characters and behaviours, atmospheric phenomena, or simple states of affairs. The human figures, whose deeds are recounted by the epic poems, are unique entities moved by a series of impulses provoked by the deities who dominate the scene. On the one hand, the human body hybridizes with and is often contaminated by the forms assumed by deities; on the other, human prerogatives such as menos (energy, strength, yearning) are also attributed to natural things and elements (lances, fire, and rivers are also endowed with menos). All of this evidently clashes with the innate tendency of cerebral functions to automatically categorize living things, capable of telic actions according to beliefs and desires, as separate from things, which are completely inert and inanimate2. In fact, these mechanisms have played a fundamental role in human evolution: in the savannah, recognizing a predator at first glance can be a life-saving skill, and those among our progenitors who did so

Andrea Lavazza

17

faster and more automatically that the others, and whose genes were better adapted to do so, had a better chance of transmitting such genes to a great many descendants. It follows that the ‘person’ concept is historically “new”, and appears gradually at different times and in different places. It also gradually emerges—this is obvious and undeniable if we are not dealing with an entirely a priori concept—on the material basis of our perceptive tools and the cerebral architecture that allows us to reason. But the idea that every human being has a prima facie right to equal respect and attention under rules of equity does not seem to arise automatically from only the biological factors wired into our brains. There are people who kill and torture specific “categories of persons” or presumed “non-persons”, and argue that their actions are legitimate on the basis of an explicitly-affirmed difference from other “categories of persons”, an argument that instinctively repulses most 21st century humans3. Remaining in the field of the critique of “universal concepts” arising from our neurological make-up as shaped by Darwinian evolution, we can claim that accounting for the genesis of a given phenomenon is not the same as explaining its meaning or its importance. The cerebral wiring behind the ‘person’ concept is comparable to the wiring behind morality, as argued by Marc Hauser (2006) (cf. also de Waal 2006). Whether there are moral universes is questionable. For example, the precept “help the children and the weak” does not hold true in those cultures that discriminate between superior and inferior ethnicities. Additionally, if there is such a thing as a ‘person concept’ network, it is not only a tool for passive recognition, but also a spur to action, otherwise it would have never evolved and bioethics would not be concerned with it. Indeed, to classify people as persons means to treat them as persons. We must thus ask why the moral obligations relating to the person are violated with extreme frequency, as can be empirically demonstrated. If anything, Farah and Heberlein describe the enabling conditions for moral behaviours and statements. Such conditions are very varied and, most importantly, none is able to account for the specifically moral content of human practices. It is also evidently insufficient to consider evolutionary advantages only. This takes us back to E.G. Moore’s naturalistic fallacy: we cannot bridge the gap between being and having to be. Nevertheless, as stated above, we are attempting a categorial critique of naturalism. Therefore, our comment on the empirical falsification of the automatic and universal character of the ‘concept’—since it is produced by our brain—shall suffice. Actually, the ‘person’ concept being tried is not a basic concept; it is already a construct that blends “automatic” intuitions together with a “data

18

Naturalism, Free Will, and Different Approaches

point” that comes from cultural stratification and intellectual reflection. In order words, if we come into the world and grow up in an anthropized environment, we continuously receive messages containing concepts that are already structured and interact with our innate endowments, to which we may add the invention/discovery of new ideas. This does not imply negating that the human experience may have been meaningfully organized prior to (and independently from) language and acquired concepts, nor does it deny that primary conceptual structures are endowed with meaning, as they are based on experience with actual objects and situations brought about by general capabilities, such as gestalt perception, motor movements, and the formation of mental images. In this perspective, basic physical experience provides the preconceptual foundation of language and other cognitive functions.

4. Putnam Versus the Fact/Value Dichotomy Hilary Putnam is an authoritative critic of the fact/value dichotomy, namely the position that does not allow for deriving a normative statement from a descriptive statement, or from a matter of fact. In his opinion, facts and values are connected in a non-obvious manner. Facts and values are entangled in at least two senses. First, factual judgments, even in physics, depend on and presuppose epistemic values. One would think that this ought to be uncontroversial, but in fact all the leading positivists – joined here by Popper, in spite of his frequently touted disagreements with Carnap and Reichenbach – made what I regard as pathetic attempts to evade this fact. What the logical! positivists were shutting their eyes to, as so many today who refer to values as purely “subjective” and science as purely “objective” continue to do, is obvious: the fact that judgments of coherence, simplicity (which is itself a whole bundle of different values, not just one “parameter”), “beauty”, “naturalness”, and so on are presupposed by physical science. But coherence, simplicity, and the like are values. All of the standard arguments for noncognitivism in ethics could be repeated without any change whatsoever for noncognitivism in epistemology; for example, Hume’s argument that ethical values are not “matters of fact” (because we do not have a “sense impression” of goodness) could be modified to read “epistemic values are not matters of fact because we do not have a sense impression of simplicity or a sense impression of coherence.” Disagreements about the beauty or “inner perfection” (Einstein’s term) of a theory could certainly be described as “differences in attitude”. And when it comes to fields less subject to experimental control than physics, fields like history or economics, for example, it is utterly simplistic to suppose that such disagreements can always be settled by “induction and deduction”. In fact,

Andrea Lavazza

19

after the publication of Nelson Goodman’s “The New Riddle of Induction”, the idea that there is such a thing as the method of “induction” has been seen by philosophers of science to be extremely problematic (Putnam 2012, 291-292).

Another one of Putnam’s lines of argument concerns purely logical aspects, and seems to capture an important component of ethical statements with a normative component (that is, the “positive” or “negative” sense they are endowed with). Indeed, they seem to require a double understanding that is impossible if the descriptive plane is separated from the prescriptive one. A second way in which values and facts are entangled might be described as “logical” or “grammatical”. What is characteristic of “negative” descriptions like “cruel”, as well as of “positive” descriptions like “brave”, “temperate”, or “just” (note that these are the terms that Socrates keeps forcing his interlocutors to discuss), is that to use them with any discrimination, one has to be able to understand an evaluative point of view. That is why someone who thinks that “brave” simply means “not afraid to risk life and limb” would not be able to understand the allimportant distinction that Socrates kept drawing between mere rashness or foolhardiness and genuine bravery. It is also the reason that, as Iris Murdoch stressed, it is always possible to improve one’s understanding of a concept like “bravery” or “justice”. If one did not at any point feel the appeal of the relevant ethical point of view, one would not be able to acquire a thick ethical concept, and sophisticated use of it requires a continuing ability to identify (at least in imagination) with that point of view (Putnam 2012, 292).

The point is that, for Putnam, this implies a conceptual pluralism at odds with classical ethical relativism. As aptly underlined by Mario De Caro and David Macarthur, editors of his most recent collection of essays, One of Putnam’s most important insights regarding the question of fact and value is to see that one has a subjectivist attitude toward moral values, according to which they are incapable of genuine truth and justification, then consistency dictates that one must adopt the same subjectivist attitude toward the cognitive values of consistency, reasonableness, simplicity, and the like. These values are presupposed by reason in the areas of science, epistemology, and logic that the metaphysician takes for granted. So if all values were subjective then so, too, would be all the ”facts” (De Caro and Macarthur 2012, 15).

It follows that scientistic opponents of Putnam’s position must grapple with an acute dilemma: either concede his point or treat these paradigmatically

20

Naturalism, Free Will, and Different Approaches

cognitivist domains as non-cognitive, “effectively sawing off the branch upon which they are sitting since surely no-one will argue that all discourses are non-cognitive, incapable of genuine truth and justification” (Ibid.). As Putnam himself recounts, “a few years ago, speaking to an audience that contained at least fifty Nobel Prize winners, [he] said the following:” I have argued that even when the judgments of reasonableness are left tacit, such judgments are presupposed by scientific inquiry. (Indeed, judgments of coherence are essential even at the observational level: we have to decide which observations to trust, which scientists to trustsometimes even which of our memories to trust). I have argued that judgments of reasonableness can be objective. And I have argued that they have all of the typical properties of “value-judgments” In short, I have argued that my pragmatist teachers were right: “knowledge of facts presupposes knowledge of values”. But the history of the philosophy of science in the last half century has largely been a history of attempts-some of which would be amusing, if the suspicion of the very idea of justifying a value judgment which underlies them were nor so serious in its implications—to evade this issue. Apparently any fantasy—the fantasy of doing science using only deductive logic (Popper), the fantasy of vindicating induction deductively (Reichenbach), the fantasy of reducing science to a simple sampling algorithm (Carnap), the fantasy of selecting theories given a mysteriously available set of “true observation conditionals”, or, alternatively, “settling far psychology” (both are Quine’s)—is regarded as preferable to rethinking the whole dogma – the last dogma of empiricism?—that facts are objective and values are subjective and never the twain shall meet (Putnam 2012, 47-48).

On that occasion none of the scientists had any objections. This does not mean that Putnam’s position was widely shared, on the contrary. Nevertheless, it brings us closer to an idea that does not reduce ethical prescriptions and judgments to mere subjectivity unrelated to facts, and it does not elevate “facts” to a level of objectivity that forces us to maintain the “present state of affairs” as the normative code for our behaviour. An example can help us understand this attempt. Consider the sex selection of foetuses. Some feel that (leaving aside for a moment the thorny issue of the destruction of embryos) selecting the sex of new-born infants is legitimate, and that it is up to the parents to decide whether they prefer a boy or a girl, regardless of the fact that in the specific case of those parents (or of that single parent) nature itself had “randomly drawn” a boy or a girl. There is thus a moral decision in agreement with Hume’s Law: a prescription cannot be derived from a state of affairs. Indeed, a

Andrea Lavazza

21

couple might prefer a girl for a series of justifiable reasons, regardless of the concept in which they find themselves. What might happen, however—and this should be kept in mind when assessing the morality of foetal sex selection—is that the single independent decisions of parents, taken as a whole, might result in a certain geographical area (or even an entire country) finding itself with a generation characterized by a highly skewed gender ratio. In such a scenario, the overwhelming preponderance of men (or women) would have serious repercussions both on individuals and on society as a whole: these problems are so evident they do not need to be listed here. In light of this, not allowing the selection of foetal sex seems more reasonable and morally justifiable, unless one were to impose “quotas” of boys and girls on the parents, which would constitute a serious violation of their autonomy in a way that almost anyone would deem unacceptable. It would thus seem that the best solution would be to let nature take its course, since this would ensure an almost even sex ratio, most likely as a result of selective and adaptive evolution. In this case, it would appear that a moral decision made to protect future generations from serious personal and societal problems arises from the understanding, appreciation and acceptance of a natural mechanism, in apparent contradiction of the is/ought, fact/value distinctions. Yet, a careful examination shows this not to be the case, because the crux of the matter still concerns assessing the preferability of states of affairs that depend on the considered judgment of the involved subjects, which subjects must take into account the physical context in which they find themselves, but are not bound by their decisions to that state of affairs. It is only thus that one can go “beyond” a restrictive understanding of Hume’s law, in the spirit of Putnam’s proposals. But this does not push us into the realm of the strong naturalization proposed by, amongst others Churchland, Farah or Heberlein.

5. Conclusion My position is thus to try to show that—within the framework of the attempted naturalization of morals—there are today improper inferences from scientific descriptions and explanations to normative concepts. In my opinion, the inferences we take into consideration are undue first of all because relevant knowledge is insufficient or insufficiently corroborated. Secondly, inferences are undue because, as we attempt to argue, it is not legitimate to move from descriptions of average phenomena to general prescriptions. This is not motivated only by the impossibility of deriving

22

Naturalism, Free Will, and Different Approaches

ought from is contained in the literal interpretation of Hume’s Law, but by the very nature of normative judgments expressed by human beings. These judgments are defined by the idea that, in many situations, it is preferable to achieve a certain state of affairs—how to treat a person, how to distribute a good—independently from how this generally takes place, and independently from the origins of both the mechanisms underlying natural events and the genesis of moral judgments. In other words, there may be moral dissent over what we ought to do, but I do not think it is sustainable to derive a new consensus from a total naturalization of normativity according to a perspective underpinned by the simple description of the world—including the human world—on the part of science.

Bibliography Churchland, Patricia Smith. 2011. Braintrust. What Neuroscience Tells Us about Morality, Princeton (NJ): Princeton University Press. De Caro, Mario and David Macarthur (eds.). 2010, Naturalism and Normativity, New York: Columbia University Press. —. 2012. “Hilary Putnam: Artisanal Polymath of Philosophy.” In. H. Putnam, Philosophy in an Age of Science. Physics, Mathematics, and Skepticism, Cambridge (MA)-London: Harvard University Press. de Waal, Frans. 2006. Primates and Philosophers. How Morality Evolved. Princeton (NJ): Princeton University Press. Farah, Martha J. and Andrea S. Heberlein. 2007. “Personhood and neuroscience: naturalizing or nihilating?” American Journal of BioethicsNeuroscience, 1: 37-48 Harris, Sam. 2010. The Moral Landscape: How Science Can Determine Human Values, New York: Free Press. Hauser Marc D. 2006. Moral Minds. How Nature Designed Our Universal Sense of Right and Wrong, New York: Ecco Books. Hume, David. 1739/1888. A Treatise of Human Nature, L.A. Selby-Bigge (ed.), Oxford: Oxford University Press. Kaposy, Chris. 2010. “The Supposed Obligation to Change One’s Beliefs about Ethics Because of Discoveries in Neuroscience.” The American Journal of Bioethics-Neuroscience, 1: 23-30. Libet, Benjamin, Curtis A. Gleason, Elwood W. Wright and Dennis K. Pearl. 1983. “Time of conscious intention to act in relation to onset of cerebral activity (Readiness-Potential): The unconscious initiation of a freely voluntary act.” Brain, 106: 623-642. Onians, Richard B. 1988. The Origins of European Thought: About the Body, the Mind, the Soul, the World, Time and Fate, Cambridge:

Andrea Lavazza

23

Cambridge University Press. Plunkett Kim, Jon-Fan Hu and Lawrence B. Cohen. 2008. “Labels can override perceptual categories in early infancy.” Cognition, 106: 66581. Putnam, Hilary. 2012. “The Fact/Value Dichotomy and Its Critics.” In Id., Philosophy in an Age of Science. Physics, Mathematics, and Skepticism, Cambridge (MA)-London: Harvard University Press. Saghoff, Mark. 2007. “A transcendental argument for the concept of personhood in neuroscience.” American Journal of BioethicsNeuroscience, 7: 55-57. Snell Bruno. 1953. The Discovery of the Mind: The Greek Origins of European Thought, trans. T.G. Rosenmeyer, Oxford: Oxford University Press. Soon, Chun Siong, Marcel Brass, Hans-Jochen Heinze and John-Dylan Haynes. 2008. “Unconscious Determinants of Free Decisions in the Human Brain.” Nature Neuroscience, 11: 543-545. Soon, Chun Siong, He AH, Stefan Bode and John-Dylan Haynes. 2013. “Predicting free choices for abstract intentions.” Proceedings of the National Academy of Sciences USA, 110: 6217-6222. Strawson, Peter F. 1959. Individuals: An Essay in Descriptive Metaphysics, London: Methuen. Tenenbaum, Joshua B. 2000. “Rules and Similarities in Concept Learning.” In S.A. Solla, T.K. Leen and K.-R. Müller (eds.), Advances in Neural Information Processing Systems 12. Cambridge (MA): The Mit Press. Vernant, Jean-Pierre. 2006. Myth and Thought among the Greeks, trans. J. Lloyd and J. Fort, New York: Zone Books.

Notes 1

See for example Onians 1988; Snell 1953; Vernant 2006. Not all categorizations are automatic and innate. See for example the role of “labels”, which can play a causal role in the formation of categories during childhood. Pertinent labels improve categorizations, while non-pertinent ones significantly damage it. Before children are able to use language, the labels they listen to can influence the way they categorize objects; cf. Plunkett et al. 2008. See also the research on how concepts are generalized and the debate on the alternative abstraction of rules, and assessments of similarity, with an exemplary element; cf. Tenenbaum 2000. 3 This, in part, is the objection made by Saghoff (2007), whose argument then shifts to a different field. 2

SOME ELEMENTS FOR A CRITICAL REVIEW OF THE CONCEPT OF “NATURALISTIC FALLACY” MICHELE BORRI (UNIVERSITY OF MILANO-BICOCCA)

Abstract One of the main objections to ethical naturalism is supposed to be the naturalistic fallacy argument. In this writing I develop a critical discussion about the argument, starting from its original formulation in the context of Moore’s Principia Ethica. In the course of time, the concept of naturalistic fallacy has gone through many different interpretations giving rise to much controversy but, nowadays, one of its most influential versions is probably the one affirming the existence of a significant connection between the naturalistic fallacy and the so called Hume’s Law. On the contrary, as I am going to argue in the present work, these two conceptions are substantially different as far as both their cultural origins and their ultimate goals are concerned. Furthermore, if we consider them separately and on the basis of their own merits, we shall see that they can hardly be considered as convincing criticisms against ethical naturalism.

1. The Definition of Naturalistic Fallacy in Moore’s Principia Ethica 1.1 In 1903 George Edward Moore published his Principia Ethica, a seminal text in the history of contemporary Moral Philosophy: establishing a basilar connection between “the science of Ethics” and the linguistic analysis of moral terms and judgments, he paved the way for a new discipline called meta-ethics. In order to gain a better understanding of Moore’s contribution to ethics, anyway, we need to place it within the broader context of his whole philosophical conception, as the two things are closely related.

Michele Borri

25

From this general standpoint, then, it is easy to see that he distinguished his position very clearly from both positivistic empiricism and metaphysical idealism—namely, the two main academic philosophical traditions at that time—to give rise to a new form of conceptual realism which, at first, he developed together with Bertrand Russell. Following this theoretical perspective, a fundamental realistic ontology was set forth to affirm the existence of such nonempirical and objective concepts as truth, good and numbers: in this way, a clear-cut distinction was made between the object of knowledge, on the one hand, and the cognitive processes, on the other. Based on these major premises, Moore was trying to lay the foundations for nonrelativistic ethics and gnoseology, saving the good and the truth from the domain of what is mutable, uncertain and subjective. Starting from this very brief outline, may now help us to understand more clearly some rather resolute and rigorous assertions which can be found in Moore’s Principia: If I am asked ‘What is good?’ my answer is that good is good, and that is the end of the matter. Or if I am asked ‘How is good to be defined?’ my answer is that it cannot be defined, and that is all I have to say about it (Moore 1903, § 6).

Even if in ordinary life the term good can be used in many different ways, Moore contends that it designates a single property existing per se. So, when we say that something is good, we are just recognizing that such property is present in some given object: My business is solely with that object or idea, which I hold, rightly or wrongly, that the word is generally used to stand for. What I want to discover is the nature of that object or idea (Moore 1903, § 6).

“Good denotes a simple and indefinable quality” (Moore 1903, § 10) and, as such, it cannot be identified with anything but itself. This radical ontologization of good constitutes one of the basic assumptions of Moore’s Principia, starting from which he criticizes most of the previous ethical theories saying that they all commit the same error, designated as naturalistic fallacy. Coherently with this ultimate conception of good, Moore develops a science of ethics that is absolutely independent from any form of empirical knowledge and is based on the objectivity of ethical values; as a consequence, ethical judgments do not need any kind of empirical verification.

26

Critical Review of the Concept of “Naturalistic Fallacy”

In the first place, what is good is an object of direct intuition1 and the only thing we can do, in the case of any basic ethical judgment, is just to check—by means of intuition—whether certain objects or actions do really possess the ethical quality that we are asserting. From this point of view, therefore, we can say that the general conception of ethics supported by Moore is mainly objectivistic, intuitionistic and cognitivistic; furthermore, we can add that our ethical judgments are also descriptive judgments, although we cannot verify or falsify them by means of any kind of empirical procedure. This means that our knowledge of good cannot be of the same sort as our sensory knowledge, because good does not really exist in time and, for this reason, it is not a natural object: this primary distinction—between those objects of knowledge that are natural and those that are not—was of the greatest importance for Moore in defining that kind of error he pointed out as the naturalistic fallacy. Moore himself says: No truth does, in fact, exist; but this is peculiarly obvious with regard to truths like ‘Two and two are four’, in which the objects, about which they are truths, do not exist either. It is with the recognition of such truths as these -truths which have been called ‘universal’- and of their essential unlikeness to what we can touch and see and feel, that metaphysics proper begins. Such ‘universal’ truths have always played a large part in the reasonings of metaphysicians from Plato’s time till now; and that they have directed attention to the difference between these truths and what I have called ‘natural objects’ is the chief contribution to knowledge which distinguishes them from that other class of philosophers—‘empirical’ philosophers—to which most Englishmen have belonged (Moore 1903, § 66).

It is not my intention to discuss the conceptual contents of this short excerpt in any greater depth here, but I would like to highlight at least two different points which seem to be particularly useful in delineating an adequate reference frame for some of Moore’s most significant philosophical views. The first point is Moore’s explicit remark concerning the predominant “empirical” orientation of most English philosophers: such an affirmation appears to be especially interesting if we consider that—as we soon will see—these philosophers represent the main target for his accusation of naturalistic fallacy.

Michele Borri

27

In the second place, we notice very clearly in this case the strong influence exerted—in a more or less direct way—on Moore’s philosophical conception of the Principia by the manifold Platonistic attitudes, circulating at the time in many clubs and societies of the University of Cambridge. Similarly, the same influence was also quite evident in Bertrand Russell’s first works and, about twenty years later, in Wittgenstein’s Tractatus logico-philosophicus. It is just by keeping in mind this kind of very subtle and pervading Platonizing ontology that we can now go on to analyse the concept of naturalistic fallacy as it appears in Moore’s Principia Ethica.

1.2 On the whole, the text of Principia Ethica presents itself as quite complex and multifaceted, containing several cross-references, and many terms are frequently defined with different shades of meaning, depending on the context in which they appear: these general features of the text are clearly reflected also in the variable definitions of naturalistic fallacy which we can find in different sections and paragraphs of the work. However, the reconstruction and detailed analysis of the various meanings of naturalistic fallacy we can found in Moore’s Principia is not the real purpose of this short essay and, for this reason, I shall take into consideration here only its basic and best known version.2 As I mentioned before, according to Moore, good is a simple and indefinable object or idea: “‘Good,’ then, if we mean by it that quality which we assert to belong to a thing, when we say that the thing is good, is incapable of any definition, in the most important sense of that word;” and, after a few lines, he goes on saying: “It is one of those innumerable objects of thought which are themselves incapable of definition, because they are the ultimate terms by reference to which whatever is capable of definition must be defined.” (Moore, § 10). When, starting from these theoretical definitions, he moves on to develop his critical analysis of the most commonly known ethical theories, Moore makes use of such paramount principles as a powerful means to challenge those theories. Accordingly, the first and most obvious error that they all commit is trying to define something (i.e. the good) that on the contrary—according to Moore himself—is indefinable on principle. On this ground, the critique is shaped around the subdivision of ethical theories into two main groups: naturalistic ethics and metaphysical ethics.

28

Critical Review of the Concept of “Naturalistic Fallacy”

A) Naturalistic ethics These theories try to define the good by identifying it with some natural property: in doing this, they totally disregard the autonomy of ethics and, even more, they attempt to reduce ethics to something different from itself. Depending on the different properties they consider, such theories can take different names and shapes: i) when good is identified with “general happiness”, we have that kind of Utilitarianism upheld by J. Bentham; ii) if the property at issue is “pleasure”, then we have J. S. Mill’s Psychological Hedonism; iii) if by good we refer to something that is “more evolved”, we are facing an Evolutionistic Ethics like the one supported by H. Spencer. What we can maintain for all of these theories—at least, conforming to Moore’s perspective—is that good comes to be identified with a natural property that can be perceived in virtue of some empirical experience and, for this reason, the above mentioned theories commit the naturalistic fallacy, which consists precisely in the pretension to consider the good as something identical to a natural quality.3 Moore sees it as absolutely unacceptable that, following this kind of naturalistic ethics, we end up likening ethics to the positive sciences and, moreover, accepting that ethics can continue to employ just the same criteria and methods of research employed by empirical sciences. B) Metaphysical ethics In this group we basically find those theories identifying good with some property inherent in those objects whose existence is postulated in a supersensible reality, e.g. the eternal life, a pure will or a kingdom of ends. The examples that Moore sees as most representative of this type of theories are those of the Stoics, Spinoza and Kant, even if in such cases the properties assimilated to the good are not natural in the same sense as those belonging to the first group. Nevertheless—according to Moore— the metaphysical theories still commit the naturalistic fallacy, because the metaphysical properties in question imply some peculiar and fundamental assumptions pertaining to the very nature and ultimate structure of reality itself.4

Michele Borri

29

1.3 After the publication of Moore’s Principia, some of the chief ideas and definitions set forth in the book stimulated an intense debate both within and outside the academic community. At first, the philosophical theories advanced by Moore became a subject of research and discussion for such authors as H. Prichard and W. D. Ross, who shared Moore’s basic intuitionistic assumptions. In the following years, however, the debate also involved different philosophers, like A. J. Ayer and C. L. Stevenson, who came to be known as the most representative interpreters of the emotivist approach to ethics.5 If in the first half of the XX century the field of meta-ethical studies appears to be dominated by the discussions between intuitionists and emotivists, a new orientation took shape during the course of the 50’s, based on the work of Richard Mervyn Hare: prescriptivism. Hare partakes of the same general view, which was equally supported by intuitionists and emotivists, of ethics as a discipline oriented in the first place towards the analysis of moral language; but whereas for Moore and the intuitionists ethics is still concerned with “knowledge”, Hare sides with the emotivists on this question affirming that moral judgments are not cognitive ones, and thus upholding a non-cognitivistic conception of ethics as well. The emotivists, on their part, had pointed to emotions and feelings, apt to influence and modify somebody’s attitudes and conduct, as the distinctive feature of ethical language. But, from Hare’s point of view, the main feature of moral judgments consists in their prescriptivity, and their function is to control and to guide conduct.6 As Hare himself asserts: “The language of morals is one sort of prescriptive language” (Hare 1952, § 1.1.1). This prescriptive aspect becomes the fundamental criterion used by Hare to advance his critique of naturalistic theories: their basic error— according to Hare—is that they do not recognize the prescriptive element which is typical of moral judgments, while trying to account for it by means of some factual or descriptive statement; consequently, Hare concludes that his theory is not a naturalistic one, since it takes into serious consideration the prescriptive factor. But there is still a very important trait of Hare’s ethical theory—that has not been mentioned yet—and this has some interesting implications for the naturalistic fallacy.

30

Critical Review of the Concept of “Naturalistic Fallacy”

2. The Controversy Surrounding Hume’s Is-Ought Paragraph 2.1 As we have seen, the first important function of the prescriptive element—in the context of Hare’s line of reasoning—is to differentiate his theory from the naturalistic theories of ethics. Following this, the second crucial function of the prescriptive element is to reaffirm the existence of a radical separation between “descriptive statements” (proper for the empirical sciences) and “prescriptive statements” (that are typical of moral language): such a rigid distinction appears to be very strictly connected to the basic theoretical assumptions of logical positivism, reformulated from the viewpoint of analytic philosophy.7 Furthermore, adhering to the interpretation given by Hare, the grounds for that separation are to be found in Hume’s A Treatise of Human Nature and, more specifically, in a short paragraph where Hume claims the impossibility of logically deriving prescriptive propositions from descriptive ones. This interpretation was exposed, still with some caution, in The Language of Morals (first published in 1952), where we can find a few hints at a “rule” or certain “observations” made by Hume to this effect; but in a following essay dedicated to the subject of “universalisability” Hare speaks without hesitation about what he explicitly calls Hume’s Law (Hare 1955). Moreover, from a conceptual point of view, Hare aims to establish some sort of direct continuity between such rule/law and Moore, saying that: “In more recent times this rule was the point behind Professor G. E. Moore’s celebrated ‘refutation of naturalism,’ as we shall see later” (Hare 1952, § 1.2.5). So, just to sum up very briefly the main features of Hare’s interpretation: in his Treatise, Hume was bringing into effect a logical “rule” (that Hare would eventually call a “law”) stating that it is impossible to logically deduce “ought” propositions from “is” propositions, and the same rule would also be the true basis for the refutation of naturalism (that is, for the formulation of the error called naturalistic fallacy) accomplished by Moore about one hundred and fifty years later.8 Within this philosophical framework, then, Hare’s prescriptive theory appears to be like a sort of logical outcome, because it completes the

Michele Borri

31

fundamental separation between the descriptive and prescriptive spheres envisaged by both Hume and Moore’s antinaturalistic positions. In this way, Moore’s antinaturalism is supposed to have its roots in the same logical separation endorsed by Hume. I have deliberately lingered on this aspect of Hare’s theory because it is just starting from this rendition of the subject that, in the following years, the so called “standard interpretation” will move towards some kind of permanent association (or, sometimes, identification) between Moore’s naturalistic fallacy and Hume’s presumed Law. Yet, if we want to disentangle (at least partly) this intricate question, then it might now be appropriate to take a step backward and briefly consider a very well-known paragraph taken from Hume’s Treatise.

2.2 The paragraph in question is taken from the Third Volume of Hume’s Treatise of Human Nature, “Of Morals”: I cannot forbear adding to these reasonings an observation, which may, perhaps, be found of some importance. In every system of morality, which I have hitherto met with, I have always remark’d, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz’d to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, ‘tis necessary that it shou’d be observ’d and explain’d; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it. But as authors do not commonly use this precaution, I shall presume to recommend it to the readers; and am persuaded, that this small attention wou’d subvert all the vulgar systems of morality, and let us see, that the distinction of vice and virtue is not founded merely on the relations of objects, nor is perceiv’d by reason (Hume 1739-1740, § 3.1.1.27).

The interpretation of this puzzling paragraph has been the subject of great controversy from the second half of the XX century onwards: very briefly, the critical point is represented by the opposition between the “standard interpretation” and those who criticize it.9 The most important thesis proposed by the “standard interpretation” is that Hume, in this short paragraph, affirms: that starting from non-moral

32

Critical Review of the Concept of “Naturalistic Fallacy”

premises we cannot reach any kind of moral conclusions; or rather, that we cannot logically deduce any kind of prescriptive statements (oughtpropositions) from descriptive statements (is-propositions). If the conclusion includes some prescriptive elements, there must be some prescriptive elements in the premises as well. So, in the final analysis, moral discourse is endowed with its own peculiar characteristics, which cannot be reduced to the typical modalities of descriptive language and, vice versa, descriptive language cannot account for ethics’ unique features. As a consequence, there is an irreparable divergence between empirical sciences, which are based on descriptive statements, and ethics, which is grounded in prescriptive statements. It should now appear clear why, from the viewpoint of the “standard interpretation”, the two theoretical models that we have considered so far—namely, the naturalistic fallacy and the so called Hume’s Law— virtually converge: both of them, in fact, assert the existence of a deep schism between the two different realms. For this reason, over the years, a new and arguable philosophical tradition has come into being, looking upon Hume as a forerunner of logical positivism and analytic philosophy. Although, from a conceptual perspective, such an interpretation of Hume’s philosophy is definitely very sharp and innovative, it nevertheless has the limitation of not taking adequately into account the social and cultural context in which Hume was writing and, therefore, it presents us with an image of Hume and his philosophy that seems to be the result of an effort to interpret his works in the light of a series of patterns of thought and principles that are quite different from his own coordinates. This problem was probably most clearly identified by A. Flew when he stated that: The Treatise and the two Inquiries seem often to be read as if they had been written for submission to the journal Analysis. But to understand Hume it is necessary always to remember that he thought of himself as contributing in these works to a would-be Newtonian science of man (Flew 1969).

2.3 Reckoning with the preceding observations, I will now suggest an alternative interpretation of the Is-Ought Paragraph, trying to restore the excerpt to its original context.

Michele Borri

33

As we have seen before, the paragraph in question is taken from the Third Volume of the Treatise of Human Nature, which deals explicitly with morals. The first part of the book (whose title is “Of virtue and vice in general”) is divided into two sections: the first one being entitled “Moral distinctions not deriv’d from reason”, and the second “Moral distinctions derive’d from a moral sense.” Within this general textual structure, the Is-Ought Paragraph is exactly the last paragraph of the first section, before the beginning of the second section. As regards the textual content, in the paragraph immediately before this, Hume had just stated that “morality consists not in any relations, that are the object of science” and, a few lines after, that “morality is not an object of reason.” (Hume 1739-1740, § 3.1.1.26). In the very next paragraph (the first of the second section), he proceeds with his reasoning by saying: “Thus the course of the argument leads us to conclude, that since vice and virtue are not discoverable merely by reason, or the comparison of ideas, it must be by means of some impression or sentiment they occasion, that we are able to mark the difference betwixt them;” and, furthermore: “Morality, therefore, is more properly felt than judg’d of” (Hume 1739-1740, § 3.1.2.1). In this frame of reference, we may wonder why Hume only places the paragraph concerning Is and Ought just at the end of the first section, and before the beginning of the second. Unless we assume that such a collocation is altogether fortuitous (which seems to be somewhat improbable), the most plausible answer is that it was his purpose to provide, in this paragraph, a few observations aimed at facilitating the transition from the first to the second section: that is, from the discussion about the fact that moral distinctions are not derived from reason to the assertion that they are derived from a moral sense instead. If we turn now to the content of the paragraph itself, we notice that, after observing that “the usual copulations of propositions, is, and is not” are usually replaced by propositions that “are connected with an ought, or an ought not”, Hume continues by saying that “this ought, or ought not, expresses some new relation or affirmation” and that such new relation cannot be “a deduction from others, which are entirely different from it.” But to what kind of other relations is Hume referring to, in this case? If we read it in the light of what he has just said in the previous paragraph, the “entirely different” relations—from which we cannot deduce the “new relation” expressed by the ought and ought not—are

34

Critical Review of the Concept of “Naturalistic Fallacy”

those “that are the objects of science” and can be investigated through the use of reason.10 At the same time, Hume is also saying that the “new relation” expressed by the ought and ought not “shou’d be observ’d and explain’d”, if we want to understand “what seems altogether inconceivable:” namely, the fact that such “new relation” seems to be deduced from a completely different type of relations. So, in this case, Hume is not really declaring that there is not (and cannot be) any possible kind of connection between is and ought: he is just saying that we cannot really figure out the real connection existing between is and ought if we take into consideration only the ordinary functions being performed by the analytical reason; if we want to explain how it happens that we make a shift from is to ought, then we need to observe and understand the authentic nature of this new relation. If the readers will use this precaution—Hume sums up—they will see that moral judgments (that is “the distinction of vice and virtue”) are not “founded merely on the relations of objects” and are not “perceiv’d by reason”, and this comprehension, in the end, will “subvert all the vulgar systems of morality.”11 Having ascertained this, in the subsequent section—which discusses the “moral sense”—Hume goes on to show that what cannot be explained by the “vulgar systems of morality” (i.e. the rationalistic systems) can be very well explained by his theory: in fact, the key element of moral perception is the moral sentiment, and passing from is to ought is not impossible but—more simply—it cannot take place just on the basis of rational procedures, because it essentially depends on the moral sense. In conclusion, it is exactly this moral sentiment that can positively account for the “new relation” expressed by the ought and ought not.12

2.4 Now we may be in a better position to assess whether the Is-Ought Paragraph really does contain the necessary presuppositions to back up what we nowadays call Hume’s Law. a) If we look at the textual and formal structure of the paragraph in point, it seems hard to concede that Hume is positing any law of logic here. What we can reasonably acknowledge is that he is criticizing the assertions made by the ethical rationalists, and putting his readers on guard against their way of reasoning.

Michele Borri

35

The real function of the paragraph could rather be that of showing the inconsistencies of ethical rationalism and introducing what Hume, on his part, holds to be the true foundational elements of morality. b) Hume’s perspective is scarcely related to the basic principles of XX century’s metaethics: its focus is not really on the logical analysis of moral terms and judgments given that, from Hume’s point of view, morality is a primary constituent of human nature and is mainly linked to passions, feelings and motives for action. As a consequence, Hume’s approach seems to be much closer in this respect to such contemporary disciplines as psychology, sociology and anthropology. c) On the whole, Hume is not interested in the construction of logically valid syllogisms to lay the foundations for a rational morality, because the moral quality of an action cannot be established on the basis of logical or rational requirements: morality and values are primarily a matter of feelings and passions. Even if it were possible to construct a solid foundation out of logically perfect syllogisms in order to develop a completely rational moral system, from Hume’s viewpoint this would be out of the question because we would still be on the level of demonstrative reasoning (which characterizes analytical reason), and from that level it would be in any case impossible to pass to the moral level. If Hume was around to express his opinion regarding the philosophical project of developing a prescriptive ethics grounded on logical foundations, he would probably criticize that project for the same reasons that brought him to dissociate himself from the ethical rationalism of his times: reason cannot give origin to morality, because morality derives from moral sense. Therefore, in the end, the transition from is to ought can only take place in virtue of moral sense. d) As a last point, we can observe that if (following what Hare says) prescriptivity is the trait allowing us to distinguish non naturalistic from naturalistic ethics, then Hume’s vision of morality is absolutely naturalistic because it is not founded on prescriptivity but on feelings, emotions and passions. In conclusion, we might say that there are not really many elements supporting the hypothesis that in the Is-Ought Paragraph we can find the premises for what, in the XX century, has been called Hume’s Law. For this reason, when we think of the problem concerning the impossibility of deriving prescriptive statements from descriptive ones, all things considered, we should not make reference to Hume’s Law but, rather, to “Hare’s Law”.

36

Critical Review of the Concept of “Naturalistic Fallacy”

3. Comparing and Contrasting the Naturalistic Fallacy with Hume’s Law 3.1 So far, we have considered the two notions of naturalistic fallacy and Hume’s Law separately, but now—remembering what Hare affirmed about the fact that the same logical rule should be considered as the true basis for both Hume’s Law and Moore’s “refutation of naturalism”—we need to make a comparison between the two different structures, in order to appraise whether Hare’s standpoint has solid foundations or not. In so doing, I will not take into account the discrepancies between the Is-Ought Paragraph and its interpretation in terms of Hume’s Law that I have pointed out before, and I shall very briefly compare and contrast the two models here outlined. a) As regards the naturalistic fallacy, we can summarize it as follows: good is a simple and indefinable quality that can be acknowledged only by means of direct intuition and, as a consequence, it cannot be defined by way of some natural properties investigable through empirical procedures and means. b) In its most essential version, Hume’s Law states that: starting from observations and descriptions concerning states of things, it is impossible to logically deduce any evaluation and prescription with respect to objects and actions.

3.2 About the so called Hume’s Law, first of all, we can observe that the main problems associated with it are to do with propositional logic, the different typologies of statements and the necessary conditions to draw some logically valid conclusions from certain premises: therefore, we are chiefly in the domain of the logical analysis of language. When we look at the naturalistic fallacy, on the other hand, we realize that in this case we are facing a different kind of problem as, in his Principia, Moore does not in any way undertake a systematic discussion concerning what are the correct (and incorrect) logical inferences among different types of statements. Instead, Moore’s fundamental problem seems to be to uphold the existence of two ontologically distinct realms: on the one side, good is defined as an object that does not exist in time and is not part of nature; on the other side, we find all the natural and empirical qualities pertaining to the physical world. Since the ontological sphere defined by

Michele Borri

37

the good is constitutively and qualitatively dissimilar from the sphere of natural qualities. It follows that the former cannot be investigated and explained making use of the same methods employed by the empirical sciences. This radical diversity in perspective appears to be further confirmed by the fact that while, from the perspective of Hume’s Law, the implications occurring among statements are logical ones (either the statements are logically deducible from one another or they are not), in the case of the naturalistic fallacy the mechanisms through which the naturalist philosophers are supposed to erroneously reduce the ethical judgments to descriptive statements are not logical deductions but, rather, some kind of arbitrary processes of identification and substitution. More specifically: Moore is not claiming that, when they pass from descriptive statements to prescriptive ones, the naturalist philosophers are making inferences that are logically invalid, but that they are confusing good with something else; so they are not making a logical error but a definitional one. The fallacy consists in identifying good with something else which is not good, but this fallacy is not due to any kind of logically incorrect deduction.13 As a matter of fact, it would not make sense to affirm that from some non natural properties (like the good) we cannot logically deduce any type of natural properties (or, inversely, that starting from some natural properties we cannot logically deduce any kind of non natural properties), because in this case the separation occurring between natural and non natural is not a logical but an ontological one; we may possibly confuse, mistake or identify what is natural with what is not (as it happens, according to Moore, in the case of the naturalist philosophers), but we could obviously not logically deduce any natural object from non natural entities, and vice versa. Finally, we can observe that while the so called Hume’s Law entails a prescriptive/normative definition of ethics, the same is not true in the case of the naturalistic fallacy, as according to Moore the real basis for ethics consists in the intuitive knowledge of what is good; hence we can say that the first one is a non cognitivist approach to ethics, while the second is a cognitivist one.

3.3 Having considered all the above, we should now to turn back to the opening question of this section, in order to evaluate in greater detail if what Hare asserts about the existence of some sort of continuity between

38

Critical Review of the Concept of “Naturalistic Fallacy”

what has been called Hume’s Law and Moore’s naturalistic fallacy really is a reliable hypothesis or not. i) As we have seen earlier, in the context of Hare’s theory, what he calls Hume’s Law is strictly connected to the founding of a prescriptive understanding of ethic. On the contrary, in the case of the naturalistic fallacy, the prescriptive element is almost absent; indeed, according to Moore, the main feature of ethics is the investigation into the nature of good, which is not identified (as it happens with Hare) with the modalities of prescriptive language. The point here is not that Moore did not recognize the essentially prescriptive character of moral language—as it was maintained by Hare— but, more likely, that we are facing two entirely different conceptions and definitions of ethics: one is prescriptivistic and the other one is not. So, as regards this aspect, we can see that the two models are not really in a relation of continuity with one another. ii) The second distinctive trait of Hare’s theory, which is rigorously dependent on the former, is its non naturalism in ethics and, in this case, there seems to be a greater similarity with Moore, who claims to be a non naturalist as well. However, the critical point about this matter seems to concern the different ways in which the two philosophers discern between what is natural and what is not: as regards Hare, non naturalism is characterized by the fact of expressly taking into account the prescriptive element; as for Moore, non naturalism in ethics rests on the intuition of an ontologically independent property, which is substantially different from empirical properties and cannot be investigated with the same methods and means of experimental sciences. In this respect, the two perspectives manifest—at least on the surface— a greater affinity but, at the same time, they also appear to be oriented towards very different goals. iii) From my personal point of view, I deem that the major difference between the two authors lies in the fact that Moore’s ethics presupposes an ontological substratum, which is completely absent from Hare’s theory. More in general, we might observe that Moore’s philosophy is inspired by a subtle ontological dualism, whose ideal origins might be traced back to Descartes and Kant: in fact, the distinctions between res cogitans and res extensa and between noumenon and phenomenon bear an interesting resemblance to Moore’s differentiation between good and natural properties.14 So, in my opinion, the notion of naturalistic fallacy might be better understood if we put it in a relation of continuity with this philosophical

Michele Borri

39

tradition, instead of considering it just like a prosecution of a dubious interpretation of Hume who, incidentally, did not theorize about any sort of radical dualism at all.

4. Three Distinct and Autonomous Models One of the main results of the discussion so far, is that the three conceptual structures in question—namely, the Is-Ought Paragraph, the naturalistic fallacy and the presumed Hume’s Law—seem now to be a long way off from converging towards the same far horizon but, conversely, they appear as three separate and independent patterns, stemming from very different cultural/historical backgrounds and aiming at divergent objectives, as I am going to suggest in the next few paragraphs. A) Hume’s ethics and the Is-Ought Paragraph It would be pointless to search through Hume’s Treatise for any proof of the fact that he sought to establish that we cannot logically deduce any ethical conclusion from purely factual premises and, therefore, if we want to reach some kind of ethical conclusion, we need to add to the factual premises at least one ethical premise: Hume’s purpose, in all likelihood, was not to reduce the sphere of morality to something like a set of rules to resolve a problem of formal logic but, rather, to show that the ethical questions cannot be explained by means of rational reasoning because they depend on a moral sentiment. In his celebrated paragraph concerning the is-ought propositions, Hume is just saying that we should give a reason making it possible to account for the surprising shift occurring from is to ought and such a reason—as it appears clearly in the next section—is precisely the “moral sense”. According to what Hume says, morality is deeply rooted in those passions and desires which are integral parts of human nature and, consequently, it is just through the study of such natural desires and passions that we can reach a better understanding of what morality is. As was explicitly declared by Hume himself in the subtitle of his Treatise, his attempt was “to introduce the experimental method of reasoning into moral subjects”, so his research programme was to develop a scientific investigation concerning the natural origins (i.e. sentiments, desires and passions) of those moral values constituting the true basis of morality.

40

Critical Review of the Concept of “Naturalistic Fallacy”

As it was very well expressed by John Dewey, almost two centuries after the publication of Hume’s Treatise: “A morals based on the study of human nature instead of upon disregard for it would find the facts of man continuous with those of the rest of nature and would thereby ally ethics with physics and biology” (Dewey 1922, “Introduction”). Dewey was also convinced that Hume’s aim was to study the moral phenomena with methods of inquiry similar to those of the natural sciences (Welchman 2010). So if we look at Hume’s conception of morality from this point of view, the interpretation provided by Hare affirming that we can establish a significant relationship of continuity between Hume’s philosophy and Moore’s “refutation of naturalism” doesn’t seem to correspond with the general orientation of Hume’s thought. And on the whole, we might say, it seems fairly hard to recognize in Hume the precursor of both Moore’s and Hare’s antinaturalism. B) Moore’s antinaturalism and the naturalistic fallacy Turning to Moore’s Principia, it is clear to see how throughout the whole text he often repeats that good is a simple property, indefinable, non natural and ontologically autonomous. In practice, however, he has not way of confirming his assertions and, in the end, the overall impression is that he is taking for granted what, on the contrary, should be demonstrated. It is one thing to affirm that some properties are unique, that they should not be confused with others, and to keep them clearly distinct from one another (we would not have any great problem agreeing with Moore about this matter); but it is an entirely different question to attest that certain properties are endowed with an autonomous existence in a non natural (or supernatural) realm, and that they are completely separate and independent from natural facts and, furthermore, that good is one of such properties. Do we have any valid reason to believe that good is really a property of that sort? As was stated by A. Walter, if we follow exactly the same method of reasoning adopted by Moore, we may end up defending the opposite position (Walter 2006). For example, if I proceed by starting from the reverse assumption, namely that good is unquestionably a natural property, I can in the same way reach the incontrovertible conclusion that Moore is in error because he is confusing what is in truth a natural property with some kind of imaginary non natural (or supernatural) entity and, as a consequence, he is

Michele Borri

41

committing the anti-naturalistic fallacy (i.e. he is confusing or identifying or substituting something that is natural with something else which is not). If we do not accept, from the beginning, Moore’s indemonstrable presuppositions, then his objections towards naturalism in ethics appear to be altogether insubstantial. Even so, some opponents of naturalism might very well reply that they are not really using the naturalistic fallacy argument in the original sense that was established by Moore in his Principia, and this might be true, because nowadays many alternative versions of the naturalistic fallacy do indeed circulate. In his study on this subject, O. Curry has distinguished at least eight different meanings attributed to the naturalistic fallacy within the contemporary debate, showing that they do not actually constitute convincing argumentations against ethical naturalism. Therefore, when someone is making use of the notion of naturalistic fallacy, we should be very cautious and try to understand what she is really talking about. As Curry himself asserts: The naturalistic fallacy…seems to have become something of a superstition. It is dimly understood and widely feared, and its ritual incantation is an obligatory part of the apprenticeship of moral philosophers and biologists alike. But if the arguments presented above are correct, then it is surely time to dispense with this superstition. To that end I make the following recommendation: Whenever someone uses the term ‘naturalistic fallacy,’ ask them ‘Which one?’, and insist that they explain the arguments behind their accusation. It is only by bringing the ‘fallacy’ out into the open that we can break the mysterious spell that it continues to cast over ethics (Curry 2006).

As regards Moore, unlike Hume he was undoubtedly critical of the possibility of developing a naturalistic investigation about ethics, but nowadays the postulations he made in order to support his position, appear to be largely disputable and absolutely not provable. C) Hare’s prescriptivism and Hume’s Law As I have said before, in the case of what has been called Hume’s Law we cannot really attest that Hume, in his Treatise, has posited any such “law” but, more reasonably, we have to acknowledge that that law is mainly the result of a much later interpretation, based on a very short excerpt taken from Hume’s Treatise.

42

Critical Review of the Concept of “Naturalistic Fallacy”

This interpretation consists essentially of a series of logical rules that are substantially independent of both Hume’s Is-Ought Paragraph and Moore’s naturalistic fallacy, as I have argued in the former sections of this work. Consequently, we should say—more properly—that “Hare’s principle” establishes what kind of premises we need to have within an ethical argumentation if we want to reach some prescriptive conclusions resting upon logically valid inferences and, in this way, it tries to lay the foundations for an ethics that—inasmuch as it is based on logically valid inferences—claims to be universalistic and prescriptive. Nevertheless, as O. Harman states: Yes, ought may not be derivable from is in the narrow logical sense, but that bears little on how we actually operate. Rather than machines for churning out syllogisms—humans are mortal, Socrates is human, Socrates is mortal—we are constraint satisfiers who infer courses of action from what we perceive. It’s not that deductive logic plays no role in our lives, nor that heuristics don’t figure, it’s just that a lot of problem solving in our physical world is of a different kind (Harman 2012).

Very often, in our daily life, we draw conclusions about what we ought to do, starting from some simple observations of empirical facts. Let us make a few practical examples: “Today there’s a very thick fog I ought to drive very carefully”, “I’ve hurt my hand ĺ I ought to disinfect it and put on a sticking plaster”, “My azalea plant is withering ĺ I ought to water it as soon as possible”. When we reason like this, we are starting from a descriptive premise and we turn directly to a prescriptive conclusion. Normally, in cases like these, no one of us in everyday life would start wondering whether she is making a logically valid deduction or not, or if in the reasoning embeds any hidden or implicit prescriptive premises, and so on: most of the times, just on the basis of our perceptions relating to different facts and situations, we make choices and decide what to do under the circumstances. The point is that, in daily life, we shift from the descriptive to the prescriptive mode easily enough, insofar as the structure of human action does not necessarily reflect the laws of logic. The multiform connections occurring among perceived facts, duties, prescriptions and actions are not obligatorily of the logical kind and, quite often, the actions that we accomplish may very well be moral or ethical ones even if they do not comply with the laws of logic.

Michele Borri

43

Actually, we do not hold an action to be moral simply on the basis of it being the result of a logically correct reasoning as to its premises and conclusions (the opposite might also happen very easily: a reasoning can be formally correct, but the action based on it is not a moral one): we can accomplish some highly moral actions even if we do not follow the rules of logical reasoning and rational calculation. There are value systems in every society regulating the conduct of its members, and the observance of such values and rules—most of the time—does not necessarily depend on rational or logical factors but, instead, on such motives as habits, feelings, instincts, imitation, interests, emotions, and so forth. As several different studies on the issue highlight, “moral perception” and moral behaviour are not mainly logical and rational competences but they are, to a great extent, the composite effect of many concomitant factors like empathic attitudes, education, selective attention, inner dynamics and many others (Vetlesen 1994; Crary 2007; Cullison 2010). As a consequence, the overall impression is that Hare’s principle— concerning the logical non-deducibility of prescriptive statements from descriptive statements—is, after all, only scarcely able to account for the reasons giving rise to moral conduct or, at the least, too drastically reductive compared to the real complexity of the problem.

5. Some Brief Concluding Remarks But what are the most important results of these reflections, when we turn to the problem of evaluating the real grounding of such kinds of critiques using the naturalistic fallacy argument against ethical naturalism? First of all I would say that, when one is being accused of committing the naturalistic fallacy, we need to proceed very carefully and ascertain exactly what it is we are talking about. Is the person who is making the criticism pointing at Hume’s Is-Ought Paragraph, at Moore’s naturalistic fallacy or at Hare’s principle? As I showed earlier, in fact, these are three clearly distinct issues, relating to three considerably different philosophical views. a) In the first case—namely, the Is-Ought Paragraph—we are not facing any law of formal logic but, rather, a series of observations aimed at introducing the readers to Hume’s theory, affirming that morality is not derived from reason but from a moral sense.

44

Critical Review of the Concept of “Naturalistic Fallacy”

This general conception of morality, in turn, is intended to give real substance to an extensive comprehension of human nature, in accordance with the methods of experimental sciences. So, from this point of view, morality should be investigated from a naturalistic perspective as well. For this reason, taking Hume’s paragraph out of context in order to support some sort of antinaturalistic positions seems to eventually be at odds with the general orientation of his works. b) As regards Moore’s naturalistic fallacy, this is definitely a frontal attack on the possibility of applying the methods of experimental sciences to the study of ethics and, in this respect, it looks like being utterly divergent from Hume’s perspective. But the critical point of Moore’s theory is represented by the fact that, in order to justify his fundamental dualism between “natural” and “nonnatural”, he is bound to appeal to some kinds of entities, allegedly external to time and not available to sensorial perception, and to profess, in addition, that good is no doubt one of such mysterious entities, which we can seize only by means of direct intuition. The problem is that the main explanation provided by Moore to corroborate his assertions consists just in repeating time and again that good is good, and this is all he has to say about it, but an argumentation of this sort is clearly implausible from a scientific point of view. Therefore, the objections to ethical naturalism evoking the naturalistic fallacy argument seem to involve also a more or less convinced adherence to the same obscure and unexplainable assumptions constituting the foundation of the argument itself. c) In the case of Hare’s principle, at last, we are facing a model identifying with prescriptivity the central feature of moral language (differing significantly from both Hume’s and Moore’s ethics, which do not claim to have a primarily prescriptive character), and affirming the logical non-deducibility of prescriptive statements from descriptive ones. So it is important to clearly understand if the person who is taking Hare’s principle as the basis for her criticism towards ethical naturalism is chiefly referring to the prescriptive or to the logical-deductive aspect of the model. i) If the criticism sets out to affirm that ethical naturalism is not adequate to account for the prescriptive character of morality, in the first place we can reply that morality is indeed a very wide and internally articulated field of study, and that prescriptivity is only one among several different constitutive features. But even supposing that prescriptivity was the central character, still we cannot affirm on a purely speculative basis

Michele Borri

45

that ethical naturalism lacks the capacity to explain the prescriptive element, because this is not a matter that can be decided on the abstract level of general theoretical principles. If we want to make such a working hypothesis, then it must be empirically tested, we need to allow the empirical sciences to investigate prescriptive phenomena and analyse their origins and functions; we have to take into consideration their results, and only in the end can we evaluate whether such results are satisfactory or not. ii) If the criticism, on the other hand, is that naturalistic ethics would result in the consequence of conceding to derive prescriptive statements from descriptive ones—contravening the principle which does not allow one to logically deduce the former from the latter—we might respond observing that very often, at the practical level, morality is not really founded on logical rules and that, most likely, the vast majority of the moral decisions and actions that we accomplish in our daily life are not dependant at all on any kind of logical deductions. The naturalistic approach to ethics, in fact, should be based mainly on the study of actual human behaviour in different circumstances, and not on a few conjectural assumptions concerning the correct logical outcome of a rigidly prescriptive and syllogistic way of reasoning. For these reasons, I will conclude by suggesting that the recurrent criticisms, appealing to the naturalistic fallacy as a reliable and deciding argument against ethical naturalism, quite often turn out to be inappropriate and insubstantial instead.

Bibliography Antiseri, Dario. 1966. Dal neopositivismo alla filosofia analitica. Roma: Abete. Árdall, Páll S. 1966. Passion and Value in Hume’s Treatise. Edinburgh: Edinburgh University Press. Broiles, David R. 1964. The Moral Philosophy of David Hume. The Hague: Martinus Nijhoff. Carcaterra, Gaetano. 1969. Il problema della fallacia naturalistica. Milano: Giuffré. Castignone, Silvana. 1969. “Il problema del rapporto tra ‘Is’ e ‘Ought’. Saggio di bibliografia critica con riferimento al pensiero di David Hume (1959-1967).” Annali della Facoltà di Giurisprudenza della Università di Genova, 3: 239-261. Celano, Bruno. 1994. Dialettica della giustificazione pratica. Saggio sulla legge di Hume, Torino: Giappichelli.

46

Critical Review of the Concept of “Naturalistic Fallacy”

Crary, Alice. 2007. Beyond Moral Judgment, Cambridge (Mass.): Harvard University Press. Cullison, Andrew. 2010. “Moral Perception.” European Journal of Philosophy, 18(2): 159-175. Curry, Oliver. 2006. “Who’s Afraid of the Naturalistic Fallacy?.” Evolutionary Psychology, 4: 234-247. Dewey, John. 1922. Human Nature and Conduct. An Introduction to Social Psychology. New York: Henry Holt and Co. Flew, Antony. 1969. “On the interpretation of Hume.” In William D. Hudson (ed.), The Is-Ought Question, London: MacMillan. Frankena, William F. 1939. “The Naturalistic Fallacy.” Mind, 48(192): 464-477. Gill, Michael B. 2006. The British Moralists on Human Nature and the Birth of Secular Ethics. Cambridge: Cambridge University Press. Hare, Richard M. 1952. The Language of Morals. Oxford: Clarendon Press. — 1955. “Universalisability.” Proceedings of the Aristotelian Society, 55: 295-312. Harman, Oren. 2012. “Is the Naturalistic Fallacy Dead (and If So, Ought It Be?).” Journal of the History of Biology, 45: 557-572. Hudson, William D. (ed.). 1969. The Is-Ought Question. London: MacMillan. Hume, David. 1739-1740. A Treatise of Human Nature. London: Thomas Loncman. Hunter, Geoffrey. 1969. “Hume on is and ought.” In William D.Hudson (ed.), The Is-Ought Question, London: MacMillan. Kemp, Jonathan. 1970. Ethical Naturalism: Hobbes and Hume. London: MacMillan. Lecaldano, Eugenio. 1970. “La fallacia naturalistica e l’etica inglese del Novecento.” Rivista di Filosofia, 61: 191-212. Moore, George E. 1903. Principia Ethica. Cambridge: Cambridge University Press. —. 1903-4. “Kant’s Idealism.” Proceedings of the Aristotelian Society, 4: 127-140. Nakhnikian, George. 1963. “On the Naturalistic Fallacy.” In Hector-Neri Castaneda and George Nakhnikian (eds.), Morality and the Language of Conduct. Detroit: Wayne State University Press. Nowell-Smith, Patrick H. 1954. Ethics. London: Penguin Books. Paltrinieri, Gianluigi. 1997. La natura come giustificazione. Kant e la fallacia naturalistica di George Edward Moore. Venezia: Il Cardo.

Michele Borri

47

Pigden, Charles R. (ed.). 2010. Hume on Is and Ought. Basingstoke: Palgrave Macmillan. Rohatyn, Dennis A. 1987. The reluctant naturalist: A study of G.E. Moore’s Principia Ethica. Lanham: University Press of America. Vetlesen, Arne J. 1994. Perception, Empathy and Judgment: An Inquiry into the Preconditions of Moral Performance. University Park: Pennsylvania State University Press. Walter, Alex. 2006. “The Anti-naturalistic Fallacy: Evolutionary Moral Psychology and the Insistence of Brute Facts.” Evolutionary Psychology, 4: 33-48. Welchman, Jennifer. 2010. “Dewey’s moral philosophy.” In Molly Cochran (ed.), The Cambridge Companion to Dewey, Cambridge: Cambridge University Press.

Notes 1

In the Preface to his Principia, Moore says that those propositions which he calls intuitions “are incapable of proof”, and that he implies “nothing whatever as to the manner or origin of our cognition of them” and not even about “the exercise of any particular faculty;” all that he says about his version of Intuitionism is that: “I am not an ‘Intuitionist,’ in the ordinary sense of the term” ((Ibid.), Preface). Following these minimal and elusive clues, it would be quite difficult to specify in more detail the real meaning attributed by Moore to the term intuition. 2 As to the question of the different meanings of naturalistic fallacy within the text of Principia Ethica, for example, G. Nakhnikian distinguishes six different meanings (Nakhnikian 1963); in a more recent study dedicated to the subject, D. A. Rohatyn identifies nothing less than twelve distinct definitions of the same term, although some of them seem to be similar to one another (Rohatyn 1987). 3 Once again, to understand the distinction between natural and non-natural properties affirmed by Moore more clearly, we can turn back to his underlying ontology: behind the whole material reality there is a non-material structure of concepts that cannot be reduced to what we can perceive through our senses and good is one of those concepts. Therefore, when Moore asserts that good is a nonnatural property, he is saying that good does not exist in time because it is beyond time and its presence cannot be ascertained—as it happens in the case of natural properties—by means of sensory perception or through the instruments of empirical analysis, but only by way of intuition. 4 This association of both naturalistic and metaphysical ethics in the same error of naturalistic fallacy has aroused many doubts among the readers and interpreters of Moore’s Principia. However, if we look at this issue from a broader perspective, we can possibly figure it out more clearly by remembering that much of metaphysical ethics is typically underpinned by an overall conception of the natural world alluding to a divine will, or to some natural laws that are supposed to be the manifestation of a divine order or of an all-encompassing cosmic harmony.

48

Critical Review of the Concept of “Naturalistic Fallacy”

From this viewpoint perhaps—although the connections are undeniably much less evident and direct than in the case of naturalistic ethics—we can elucidate more explicitly what Moore is really thinking of, if we understand his affirmations as a critique of those theories inclined to identify the good with some properties referring to a transcendent (or otherworldly) comprehension of nature. 5 It is not my purpose, in this work, to elaborate a reconstruction of the different meanings ascribed to the notion of naturalistic fallacy during these years; G. Carcaterra and E. Lecaldano have written some excellent studies on this matter (Carcaterra 1969; Lecaldano 1970). 6 With regard to this subject, it may be worth pointing out that both emotivism and prescriptivism adopt a definitely reductionist perspective of the complexity and plurality of possible usages and functions of ethical language, as they try to characterize it on the base of just one quality among its several different ones; as a matter of fact, moral language can also arouse emotions, it can also prescribe something, but it can do many other things as well and it cannot be reduced to just one of these two functions. This was properly remarked, for instance, by P. H. Nowell-Smith in his Ethics (§ 7.1): “The words with which moral philosophers have especially to do, which are usually called ‘value-words’, play many different parts. They are used to express tastes and preferences, to express decisions and choices, to criticize, grade, and evaluate, to advise, admonish, warn, persuade and dissuade, to praise, encourage and reprove, to promulgate and draw attention to rules; and doubtless for other purposes also. These activities form the complex web of moral discourse and our problem is to trace the connexions between them and to come to understand how it is that the same word can be used in all these different ways. What a man is doing with a particular value-word at a particular time can only be discovered by examining what he says in its context, but it would be just as absurd to suppose that there is no connexion between these activities as to suppose that the same expression can only be used to do one job.” (Nowell-Smith 1954). 7 A thorough and detailed study of the multifarious connections between logical positivism and analytic philosophy was the one published by D. Antiseri almost fifty years ago (Antiseri 1966). 8 Regarding the convictions expressed by Hare, it is important to stress that there is no possible textual evidence at all that we can find in order to substantiate his interpretation: as a matter of fact, there is not one single reference in the whole text of Moore’s Principia either to Hume or to his works not to mention any references to the problem of the impossibility of logically deducing prescriptive propositions from descriptive propositions; the overall impression, instead, is that such questions were simply not present in the realm of Moore’s theoretical interests and problems. Furthermore, if we consider some other essays that Moore dedicated to Hume’s thought, we find that in those writings Moore argues against some consequences of Hume’s empiricist scepticism, but he never takes into consideration any aspect of Hume’s moral philosophy. As a consequence, unless we assume that what Hare says about the existence of some kind of more or less direct continuity connecting Moore’s ethics with Hume’s philosophy is the result of some attestation or personal conversation with Moore himself (but, if this is the

Michele Borri

49

case, Hare does not reveal anything about it), we have to conclude that the reconstruction proposed by Hare is purely speculative. 9 In this short essay I cannot even attempt a concise reconstruction of the many and complex problems related to an in-depth analysis of the Is-Ought Question. One of the best collections of selected writings on the matter was published by W. D. Hudson at the end of the Sixties (Hudson 1969), and in the same year S. Castignone issued a very interesting critical bibliography discussing the subject (Castignone 1969). Coming to more recent times, one of the most exhaustive and systematic introductions to the question is the one by B. Celano (Celano 1994); for an updated prospect of the problem, a new volume by C. R. Pigden collects the proceedings of a recent conference on the subject at issue (Pigden 2010). 10 Here Hume is referring to that kind of analytical reason, which was the main subject of the First Volume of his Treatise. Hence, in this passage he is saying that the “new relation” expressed by the ought and ought not is substantially different from the other relations that are the “object of reason” and, because of this diversity, the new relation cannot be a deduction from rational relations. 11 It must be remembered that, from an historical point of view, Hume’s ethical views were part of the great intellectual debate taking place in those times, between “moral sense theorists”—like, for instance, Shaftesbury and Hutcheson— and “moral rationalists” like Cudworth, Wollaston and Clarke. And that was also, by the way, one of the most important and significant periods in the whole history of English moral philosophy—as it was conveniently demonstrated by M. B. Gill’s fundamental study about that epoch (Gill 2006). In that context, one of the main purposes of the Third Volume of Hume’s Treatise was the confutation of the theories which were championed at the time by the moral rationalists; so when in the paragraph Hume alludes to the “vulgar systems of morality” he is speaking—in actual terms—of the rationalistic theories. 12 Dealing with one of the classical texts of our philosophical tradition—such as Hume’s Treatise—any interpretation is clearly objectionable and open to criticism. In this case, I’ve tried to provide an interpretation that was—as much as possible— close to the structure of the text and mindful of the cultural and historical context in which Hume was writing his works. When I was developing this working hypothesis, I availed myself of the influential contributions of such authors as P. Árdal, D. R. Broiles, J. Kemp and G. Hunter (Broiles 1964; Árdal 1966; Hunter 1969; Kemp 1970). 13 In his remarkable essay, dedicated to a critical analysis of the naturalistic fallacy, W. Frankena expounded some general considerations fairly similar to those I am discussing here. According to Frankena, Moore’s naturalistic fallacy should be regarded as a specific case of a more general definist fallacy consisting in the process of “confusing or identifying two properties, of defining one property by another, or of substituting one property for another. Furthermore, the fallacy is always simply that two properties are being treated as one” (Frankena 1939). This kind of fallacy, in the words of Frankena, “has no essential connexion with the bifurcation of the ‘ought’ and the ‘is’”, and it does not necessarily imply violating the principle that we cannot deduce ethical conclusions from non ethical premises: admitting that it is an error to confuse or identify two different properties, then it is

50

Critical Review of the Concept of “Naturalistic Fallacy”

not an error depending upon some incorrect logical deduction. Therefore, in Frankena’s view too, the notion of naturalistic fallacy is not intrinsically connected with the is-ought question and they remain two distinct and independent problems. 14 As was aptly shown by G. Paltrinieri in his extensive study on this subject, in Moore’s philosophy we can find a significant Kantian legacy and, at any rate, Kant is an author whose name recurs frequently in many of Moore’s works (Paltrinieri 1997). For instance, Moore’s explicit acknowledgement of “synthetic a priori judgments” as “an exceedingly important discovery of Kant’s” (Moore 1903-4), might be seen as the theoretical basis for his declaration that “the fundamental principles of Ethics must be synthetic propositions, declaring what things, and to what degree, possess a simple and non-analysable property that may be called ‘intrinsic value’ or ‘goodness’” (Moore 1903, § 35). And, furthermore, the definition of good, as a foundational and indefinable quality, might be understood more clearly if—coherently with the Kantian perspective—we take it as a transcendental condition allowing the formulation of ethical judgments.

EMPATHY AND SYMPATHY: FROM DESCRIPTION TO PRESCRIPTION SARAH SONGHORIAN* (VITA-SALUTE SAN RAFFAELE UNIVERSITY)

Abstract This work has two distinct aims. The first is that of defining the terms “empathy” and “sympathy” in a univocal way in order to underline the difference between them. This has to be achieved by means of a dialogue between the results of neuroscience, developmental psychology, and other types of empirical data on the one hand, and philosophy on the other. Much of the misconception regarding these terms has to do with the fact that they have been used by these different traditions without due interaction and dialogue between them. The second aim regards the possibility of interpreting the concepts, as they are defined here, as providing some elements for the centenarian “is/ought question”. Do the two concepts show the passage between a description of how we are (“is”) and a prescription of how we should be (“ought”)? The answer to this last issue is twofold, since I will need to introduce between these two levels a metaethical one.

1. Relevance of the Topic The interest in the concepts of “empathy” and “sympathy” grows from two different perspectives. On the one hand, there is the philosophical tradition and its usage of these concepts for different purposes and with different meanings. Studying the works of David Hume, Adam Smith, Edith Stein and Max Scheler, I could find several definitions of the terms but no agreement either on a single one, or on the proper distinction between the two concepts involved, or on the role they should play in aesthetics and in ethics. On the other hand, neuroscience has been expressing a growing interest in these issues. Since the discovery of the so-called Mirror Neuron System

52

Empathy and Sympathy: From Description to Prescription

(see Rizzolatti et al. 1988; Rizzolatti et al. 1996; Gallese et al. 1996; Iacoboni et al. 1999; Rizzolatti et al. 2001; Rizzolatti et al. 2004) several experiments were carried out in order to understand our capacity to comprehend and mirror the emotional reactions of others (see De Vignemont, Singer 2006; Wicker et al. 2003; Keysers et al. 2004; Singer et al. 2004; Jackson et al. 2005; Morrison et al. 2004; Avenanti et al. 2005; Botvinick et al. 2005). I believe the attempt to combine these two different perspectives can prove useful for both, since a deep philosophical analysis can provide more precise definitions and a deeper understanding of conceptual differences, while, on the other hand, empirical data (such as those stemming from neuroscience or from developmental psychology, just to give a few examples) can provide new objects for philosophical discussion and a more objective picture of how we actually are— with our faculties and their natural limits—and of the world around us. What makes this attempt interesting is the fact that there is no agreement on any of the issues that can be raised concerning the two concepts involved, and that the two disciplines have used them almost ignoring the fact that at least another tradition was working on them. The first aim of this paper is, thus, that of delivering working definitions that can prove useful both because they aim to provide a bridge between these two research fields and because, since “empathy” and “sympathy” are particularly widespread concepts, we can easily lose track of what they really mean, and of the etymological difference that we should keep in mind. The second aim of this work, which explains the subtitle, is the attempt to outline whether and how the distinction between these two concepts can be relevant also for the centenarian “is/ought question”. Will the question be the same which Greene asks himself in the famous article “From neural “is” to moral “ought”: that is, what are the moral implications of neuroscientific moral psychology?” (Greene 2003)? If the definition of “empathy” as a certain neural circuit proves consistent and convincing, will I be drawing a normative conclusion from a simple description of a natural device? Will “sympathy” be drawn from “empathy” in this precise way? The definition of the two concepts will help us understand this passage and will show that the question asked by Greene is not exactly the one addressed here.

Sarah Songhorian

53

2. From “Is” to “Ought” The attempt to link the centenarian debate on the “is/ought question” to the work that can be done on the concepts of “empathy” and “sympathy” is twofold. First of all, if the definitions of the concepts provided here prove useful and consistent, there will not be an actual “is/ought passage”, since “sympathy” will be defined as a metaethical mechanism and not as a normative concept itself. The fact that this metaethical tool can be used to build our ideas of what we ought or ought not to do, and so with normative consequences, is not directly the purpose of this work. Obviously, it will be interesting to analyze in greater detail the passage between this alleged metaethical tool and its normative consequences, but for now I will focus on the relation between a neural mechanism (i.e. “empathy”) and the metaethical one (i.e. “sympathy”). In Joyce’s words: Even if there were an a priori prohibition on deriving evaluative conclusions from factual premises, this need not stand in the way of metaethical implications being drawn from factual premises, for a metaethical claim is not an ethical “ought” claim; it is more likely to be a claim about how we use the word “ought” in ethical discourse—which is a perfectly empirical matter (Joyce 2008, 371).

I deny the fact that there is such an a priori prohibition of the derivation of evaluative conclusions from factual premises, but it is not a matter that can be fully solved by the concepts that I am going to analyze here. The shift from the description of a neural mechanism to metaethical implications is interesting in itself, and can be complicated by the fact that metaethics does not only concern our usage of the ethical language, but sometimes provides us with a criterion for assessing the correctness of our usage of such language. In this latter sense, it has something to do with prescribing. In so far as some metaethicists offer prescriptions about how the word “ought” ought to be used, metaethics sometimes steps beyond the descriptive. Even in such cases, however, metaethicists are still not pushing ethical “ought” claims (Joyce 2008, 371).

So, even though it will not deplete the entire issue of the “is/ought question”, I believe the relation between the two concepts analyzed here can provide some further insight, and can represent a mediate way to reach a solution. It is not a matter of deriving a normative conclusion from a

54

Empathy and Sympathy: From Description to Prescription

pure description of how our brain “is”, but to derive it by means of a metaethical criterion that takes into account both how we are actually wired and the way in which we use the evaluative language. Hume’s concern was the absence of any explanation or justification for the passage from sentences containing “is” and “is not” to propositions connected with “ought” or “ought not”. I cannot forbear adding to these reasonings an observation which may, perhaps, be found of some importance. In every system of morality, which I have hitherto met with, I have always remark'd, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz'd to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought or an ought not. This change is imperceptible; but is, however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, 'tis necessary that it shou'd be observ'd and explain'd; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it. But as authors do not commonly use this precaution, I shall presume to recommend it to the readers; and am persuaded, that this small attention wou'd subvert all the vulgar systems of morality, and let us see, that the distinction of vice and virtue is not founded merely on the relations of objects, nor is perceiv'd by reason (Hume 1969, 469-470).

The issue deals precisely with this lack of explanation of the passage: if it is possible to move from a descriptive sentence to a prescriptive one, then you need to justify that change of planes. The second and related aspect of the issue concerns the misrepresentation of the “is/ought question” by some neuroscientists like Joshua Greene. Notwithstanding the evident relevance of his experiments concerning human beings’ actual capacity to produce moral judgments, he also drew rash conclusions concerning the relevance of such an analysis for the “is/ought question” debate: If these empirical claims are true, they may have normative implications, casting doubt on deontology as a school of normative moral thought (Greene 2003, 36).

The misconception derives from the fact of disregarding the metaethical level. The denial of such a level of analysis counts as an unjustified passage from description to normativity. For this reason, I believe the concepts of “empathy” and “sympathy” could provide a more

Sarah Songhorian

55

detailed picture of the relations between “is” and “ought”, since “sympathy” will not count as a normative concept, but as a metaethical one. The next step leading from this very concept to a normative moral theory will need some further work, which there is not enough space for here.

3. Definitions The first step that needs to be taken regards the definition of the concepts involved in this analysis since, as we said, there is no agreement on it. In De Vignemont and Singer’s words: There are probably nearly as many definitions of empathy as people working on the topic (De Vignemont and Singer 2006, 435).

The same could be said for the concept of “sympathy”, and for the distinction between the two of them and other related ideas. Some efforts in the definition of these ideas has been made, in particular by Stephanie Preston and Frans de Waal (Preston and De Waal 2002)1 and by Frédérique De Vignemont and Tania Singer (De Vignemont and Singer 2006), but there is some room for further examination, in particular for a philosophically well-informed analysis.

3.1 Etymology and a Brief History Let us start from the etymological difference between the two ideas. Ɣ “em-pathy” (ΤȞʍȑȚıțį): to feel within oneself someone else’s feelings. Ɣ “sym-pathy” (IJȤȞʍȑȚıțį): to feel with someone else. To put oneself into someone else’s situation and to judge it. But how exactly can we use this etymological difference? What does it tell us about our ability to share and understand other people’s emotions? Let us take a brief look at the history of the terms, in particular how they entered the English lexicon. “Empathy” is the more recent term in English, entering the vocabulary only in the last century. The credit for translation into English goes to Edward Bradford Titchener […], who coined the word empathy to express the difference between einfühlung (in-feeling) and mitgefühlung (with-

56

Empathy and Sympathy: From Description to Prescription feeling) […], which was already in the English lexicon as sympathy (Escalas, Stern 2003, 567)2.

3.2 Empathy Interpreting this etymological difference, we can say, as a first approximation, that: There is empathy if: (i) one is in an affective state; (ii) this state is isomorphic to another person’s affective state; (iii) this state is elicited by the observation or imagination of another person’s affective state (De Vignemont and Singer 2006, 435).

A further condition could be added: (iv) the subject is not necessarily conscious of the difference between subjects, between himself and others. The fourth condition (iv) I just expressed is radically different from the one present in De Vignemont and Singer’s work, that is: (iv) one knows that the other person is the source of one’s own affective state (De Vignemont and Singer 2006, 435).

I believe there are at least two related reasons for privileging my condition (iv) to the one presented by De Vignemont and Singer. The first is that allowing empathy to be unconscious, at least on some occasions— though my idea is that it is unconscious in most cases—helps distinguish different phenomena, such as, for example, emotional contagion, identification, theory of mind, perspective-taking, cognitive empathy and so on. To examine such distinctions in depth and reach a univocal definition of the terms just quoted is a task that cannot be accomplished here. However, more precise, univocal, and narrow definitions of the terms in use within a certain literature are crucial in order to gain a better understanding of the phenomena at issue. The second reason for differentiating my definition of “empathy” from De Vignemont and Singer’s one has to do more precisely with the matter of this work, i.e. the relation between the concepts of “empathy” and “sympathy”. The definition of “empathy” I propose is a very narrow definition of affective empathy, without any contribution from cognitive appraisal or perspective-taking (see Lamm, Batson and Decety 2007 and Batson et al. 1997). These aspects will show a substantial difference between “empathy” and “sympathy”, as we can see both in the quote below by Escalas and Stern and in Table 1, which has also been drawn from the same article and that

Sarah Songhorian

57

provides, in my opinion, a good example of how we can distinguish the two concepts. Empathy refers to a person’s capacity to feel within or in another person’s feelings (Langfeld 1967, 138), and most researchers now consider it “an emotional response that stems from another’s emotional state or condition and that is congruent with the other’s emotional state or situation” (Eisenberg and Strayer 1987, 5). […] whereas sympathy stems from the perspective of an observer who is conscious of another’s feelings, empathy stems from that of a participant who vicariously merges with another’s feelings (Escalas and Stern 2003, 567).

Primary component Control Self-other differentiation Self-other orientation Relation to other Opposite Attitudinal effects

TABLE 1 DEFINITIONAL ISSUES Empathy Sympathy Affective Cognitive Feeling Thinking Involuntary Voluntary Absent Present Self Other Loses self in other Distance from other Merging with other Understanding of other Nonempathy Antipathy Indifference Hostility Direct Indirect

I am not in complete agreement with all the aspects the authors take into account, or at least I would probably not have given exactly the same answer to every category. For example, as we already noted, I would say that it is possible for an empathic response to be conscious—it is not usual, but still possible. On the contrary, Escala and Stern say that the selfother differentiation is always absent.

3.3 Sympathy But what exactly do I mean by the term “sympathy”? First of all, we have to remember the etymological meaning of the term, that is: withfeeling, or the ability to feel with someone else. To put oneself into someone else’s situation and to judge it. I will try to suggest a step forward, consisting in distancing from a purely emotive ability. This is consistent with the idea expressed in Table 1 that the primary component

58

Empathy and Sympathy: From Description to Prescription

of sympathy is cognitive. It does not mean, obviously, that the affective component is not present, it is simply not the primary one as it is for empathy. Adam Smith rightly pointed out the difference between “sympathy” and “compassion”, which are often considered synonyms, notwithstanding the evident difference that the etymological definition suggests. Pity and compassion are words appropriated to signify our fellow-feeling with the sorrow of others, sympathy […] may […] be made use of to denote our fellow-feeling with any passion whatever (Smith 1759, 49, italics mine).

Sympathy, like empathy, is completely indifferent to the content of the feeling. What these two concepts take into consideration has nothing to do with the type or intensity of someone else’s feelings, but with the ability to experience it. The ways in which they enable us to experience another’s feeling, however, are completely different. Empathy deals with an automatic and mostly unconscious ability to experience someone else’s emotional reactions; sympathy, on the other hand, is a more complex metaethical phenomenon. Sympathy is the basis of the moral sentiments for Smith because “to sympathize with” means “to align with the estimation of right or wrong based on fellow feeling.” The nuances that arise are many and varied; but Smith is more consistent than he is generally credited in standardly using sympathy as the source of intuitions about the merit (or demerit) of other individuals (Agosta 2011, 4, italics mine).

I am sympathetic with this perspective, with the only difference that I would prefer to talk about a “mechanism that balances different beliefs and emotions” rather than “intuitions”. I do not believe Smith is an intuitionist in this regard. The relevance of Agosta’s point on Smith is, from my perspective, the fact that he introduces an evaluative element within the concept of “sympathy”. It is not merely a matter of basic ability to enter into a relation with co-specifics, but a way to judge behaviours and characters and decide which considerations we should take into account in order to act morally. Smith’s subtitle makes this role of “sympathy” particularly clear: An Essay Towards an Analysis of the Principles by which Men Naturally Judge Concerning the Conduct and Character, first of their Neighbours, and Afterwards of Themselves.

Sarah Songhorian

59

“Sympathy” represents such a principle of moral sentiments. It can be seen as a metaethical criterion to judge and decide, that is, the point of view we have to adopt in order to pronounce any moral judgment. But how exactly does this “principle” work? Though sympathy is very properly said to arise from an imaginary change of situations with the person principally concerned, yet this imaginary change is not supposed to happen to me in my own person and character, but in that of the person with whom I sympathize. When I condole with you for the loss of your only son, in order to enter into your grief, I do not consider what I, a person of such a character and profession, should suffer, if I had a son, and if that son was unfortunately to die; but I consider what I should suffer if I was really you; and I not only change circumstances, but I change persons and characters. My grief, therefore, is entirely upon your account, and not in the least upon my own. It is not, therefore, in the least selfish (Smith 1759, 501-502, italics mine). The ideal spectator runs a cognitive simulation in which one may indeed begin with one’s own characteristics as input, but quarantines one’s own peculiarities in favor of those of the other (Agosta 2011, 5, italics mine).

There are different things to consider regarding these two quotations. First of all, sympathy, as opposed to empathy, arises when the subject puts himself into a particular position in relation to the state of the other he is considering. Sympathy does not imply the mirroring of someone else’s emotions, it is a much more complicated phenomenon implying both an emotive reaction, probably empathic and most of the times unconscious, and the fact of taking a certain position in order to be able to express a judgment. The subject imaginatively changes his position with that of the other person: he does not simply imagine himself in that position, but he considers all the peculiarities of the person he is trying to understand as if they were his own. The second element I would like to consider here is the difference between a) the fact of immediately empathizing with someone’s feelings as if they were our own, that is the fact of taking a first-person perspective, a sort of identification of oneself with the other; and b) the more complex and pondered ability to assume a third-person standpoint, i.e. the one of the so-called ideal spectator, who is capable of judging fairly both before and after the performance of the action, thus allowing moral judgment on the one hand and moral motivation on the other. In the phenomenon of sympathy I try to integrate both the emotional reaction I derive from my “empathic brain” (i.e. my automatic ability to resemble other people’s emotions) and some information about the subject that is undergoing such

60

Empathy and Sympathy: From Description to Prescription

a situation, since the aim of sympathizing is that of understanding how that particular person, with that particular character, should have behaved. Not only is the point of view different, but so are the sources of the information that the two mechanisms elaborate: empathy concerns only emotive reactions, while sympathy needs to integrate different kinds of information, i.e. emotive responses, knowledge of the person and of her character, comprehension of the situation, its features and—as far as possible—its consequences (see also Batson et al. 2003). I believe that sympathy, together with the concept of the ideal spectator, constitutes the principle Adam Smith had in mind for establishing a criterion to judge and to direct moral behaviour.3 In this sense, they are not normative concepts in the strict sense of normativity, but they surely are metaethical ones. A further point regards the last sentence of Smith’s quotation, that is, the fact that sympathy is not “in the least selfish”. Sympathy is somehow directed towards acting in favour of others, while empathy can be selfish in a certain way since, despite the fact that the emotion has its source in another person, empathy has nothing else to do with others. It does not promote any sort of attitude towards them. To feel someone’s emotion per se does not imply certain active behaviours in his favour. The relation between sympathy and acting in favour of others is simply closer than the one linking beneficent behaviour to empathy. This last point has to do with the idea, proposed by Agosta, that “the ideal spectator runs a cognitive simulation.” Surely, there is a cognitive contribution in order to consider the other person’s character and situation properly; it is less clear, although Agosta’s thesis is surely plausible, that this contribution is a simple simulation. It could be a simulation, a projection or a combination of the two strategies moving from a more natural and basic experience, such as empathy, in which, supposedly, the mirror mechanism is just a projection of ourselves (since we do feel as if those emotions were our own), to a more complicated balance between simulating a character and a situation and projecting ourselves there, although quarantining our peculiarities, as if we were someone else. Obviously, the issue about whether it is a projection, a simulation, or a combination of the two needs further study.

4. A Single Step? Interpretation and Two Issues I have, therefore, interpreted the term “empathy” as referring to the automatic and (mostly) unconscious capability to mirror someone else’s emotional state, regardless of the type or intensity. If we see someone in

Sarah Songhorian

61

sorrow we automatically feel that sorrow as if it were our own. We mirror that emotion (see Gallese 2001; Gallese 2003; Gallese et al. 2004; Rizzolatti and Sinigaglia 2006; Pfeifer et al. 2008). Our brain is activated exactly as it is when we are experiencing that kind of emotion first-hand. The intensity of that neural activation will obviously be modulated depending on whether its source is internal, that is if the emotion is really ours, or external, being provided to us by means of an empathic experience. Intensity can also vary depending on the proximity of the person whose emotion we are empathizing with, both physically and affectively. If the person in need is physically close to us we are usually more prone to help her compared to the situation in which she is far from us. An everyday life example is the fact that we actually are keener to help a beggar than to send the same amount of money to children in need in Third World countries. The same can be said, obviously, for affective proximity: we are more prone to help a relative than a stranger (see Meyer et al. 2012). We can for sure recognize this bias and work against it, but it is clear that we do suffer from such a bias in the first place. This proximity bias can be caused either by the fact that those that are physically closer are usually easier to help, it costs us less time and effort to give some money to a beggar than to send it to a charity; or it can be explained by means of a proximity bias within empathy: when we see somebody in pain this provides us with a stronger impression than when someone tells us about such pain (see Singer et al. 2004 and Lamm et al. 2007). The mechanism is the same, but the intensity is rather different. This latter hypothesis on the cause of the proximity bias looks more convincing. At this level of analysis the cost/benefit explanation is still plausible. In sum, empathy is some kind of neural activation that occurs automatically and often unconsciously, enabling us to mirror other people’s emotions. It is, though, a very basic ability. In order to understand people’s ability to enter into a relation with cospecifics, and to engage pro-social behaviours, we need to take some further steps, adding something to this initial and very simple capacity. We do not just mirror the emotions of others, we understand them and we act according to them. You can catch an emotion, just as you can catch a cold, without knowing whom you caught it from. Facial empathy will be of further help once we have the capacity to refer our secondhand response to the particular other individual whose visage is inducing the response, or in other words to mark the response as part of our representation of that individual. Once we have the capacity to index our response in this fashion and to integrate it with other information about the same individual, facial empathy can assist

62

Empathy and Sympathy: From Description to Prescription us in interpreting, predicting, and explaining behaviour. We can also, at least to a degree, segregate these “copy-cat” responses from responses that arise in the normal way from our own perceptions and memories. Specifically, we can allow the copy-cat responses to run “off-line”, as it were, so that the emotion’s typical effects on one’s behaviour are diminished if not entirely eliminated (Gordon 1995, 730).

Despite the difference in the term used by Gordon to refer to empathy, the above quotation is a good example of the kind of step that is necessary in order to make this automatic ability something operatively relevant in our daily lives. Even infants have the capacity to mirror their mother’s expressions or the cry of other new-borns (Goldman 1993, 352; see also Simner 1971; Meltzoff, Moore 1977; Meltzoff, Moore 1983), but that cannot count as an occurrence of the kind of superior ability we are interested in. From this very primeval and embedded ability, we need to reach a complex and stratified capacity to understand the emotions and beliefs of others in order to imagine ourselves in their position, assuming their character and peculiarities and being thus able to judge their behaviour or, if the behaviour we try to judge is one in which we ourselves are planning to engage, to decide whether it is the right one or not. Sympathy can, thus, be defined as the criterion we can use to judge a posteriori both the behaviour of others and our own and to decide a priori—by means of the imaginary projection of an ideal spectator—whether this conduct is appropriate or not. In order to reach this last metaethical criterion, some steps must be taken. There will not be enough room for them here, since they are not the focus of this work. Yet, it is clear that the transition from basic and emotive empathy to sympathy as a criterion of moral judgment and moral decision-making, concerning not only emotional responses, but also beliefs and principles, in a sort of reflective equilibrium (see Rawls 1971; Rawls 1999), is not a direct one. Concepts such as theory of mind, perspective-taking and cognitive empathy need to be explained since they can play important roles in the transition from empathy to sympathy. The same can be said about such concepts as emotional contagion, identification, simulation, and projection, or as respect, cure, and prosocial behaviours, which are respectively relevant for a better understanding of empathy and its functioning, of the related phenomena and of the kind of conduct enabled by sympathy. The analysis just conducted, though hardly complete, brings out two different issues: a descriptive research field and a metaethical one, leading to the consideration of normative issues.

Sarah Songhorian

63

4.1 The Descriptive Issue If the definition of “empathy” I have provided above works, then, on a purely descriptive level of analysis, what we will find is that, when we see someone in an emotional state, there is an activation of exactly the same areas of our brain that are active when that emotional state is our own. The intensity may vary both from felt emotions to observed ones, and according to what I have called the “proximity bias”. Neuroscience can provide a great amount of experimental data and, concerning the matter at issue here, there is a lot of work still to be done. In particular, minimal experimental tasks must be conceived, so that the different elements of a broad empathic response can be properly identified and distinguished—much of the data we have now were acquired by means of complex experimental tasks involving both emotional responses and a conspicuous amount of mentalization and conceptualization. So, one of the things that can be done in order to achieve a better understanding of the phenomenon of empathy is to design minimal tasks and see if this basic and emotive ability correlates only with the activation of subcortical areas corresponding to the so-called “emotional brain” (Dalgleish 2004). It will also be useful to analyze empathic activations when the subjects see others in joyful or happy situations, since most of the experiments have been done analyzing the observation of painful or distressful situations. Interesting data could be provided also by the analysis of the so-called “empathy disorders” both in children and adults. A general “empathy disorder” has been suggested to be a characteristic component of many other disorders including autism, sociopathy, prefrontal damage, fronto-temporal dementia, and even anorexia nervosa (Preston and de Waal 2002, 14)4.

If a correspondence between these disorders and some neural deficits or reduced activations is found, this data could count as proof of the role of such areas in empathic responses. Some studies have been conducted, in particular concerning autism, but there is more to be investigated not only with regard to autism but also, and especially, to all the other disorders mentioned. The last disorder I think it would be interesting to consider is sadism. Notwithstanding the exclusion of sadism from the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM) since the fourth edition, it is at least an interesting conceptual counterexample that needs to be considered. I define “sadism” as the

64

Empathy and Sympathy: From Description to Prescription

pattern of behaviour of a person that “is amused by, or takes pleasure in, the psychological or physical suffering of others” (Criterion 4 in DSM IIIR). The hypothesis, which has to be verified, is that a sadist does not have a neural impairment in, or a reduced activation of, what can be called the “empathic circuit”, but that the disorder must be found at a different level. With regard to a pure description of the neural activity, my guess is that a sadist will not be substantially different from everyone else with respect to the “empathic circuit”. The level of the disorder is quite different from that of autism or sociopathy: a sadist understands other people’s feelings (while an autistic child has severe difficulties), but he has a response that is the opposite of the typical one—when he sees someone in pain, he rejoices. This kind of research will provide a more precise picture of the specific neural areas that are involved in our typical capacity to empathize with others and will try to give some sort of explanation as to when and why deviations from the standard occur.

4.2 The Metaethical Issue and the Normative One The second issue has more to do with the concept of “sympathy” rather than with that of “empathy”, since we have defined the first as that metaethical mechanism enabling us to judge a posteriori both other people’s behaviour and our own, and to decide a priori—by means of the imaginary projection (and/or simulation) of an ideal spectator—whether this conduct is appropriate or not. Is this an overall normative theory? Of course it is not, but the combination of the ability to put ourselves in someone else’s situation and in that of the ideal spectator—that directs towards ourselves the same kind of considerations we apply to others when judging them from a sympathetic point of view—can constitute the criterion to balance beliefs, emotions, principles and so on. It does not tell us how we should behave in a strict sense, but it shows how we judge behaviours a posteriori and how we can educate ourselves to foresee that kind of judgment and make it relevant to our own decision-making. This issue has to do with a different level of description, compared to the one involved in the description of neural circuits, but it is still description unless we attribute an evaluative element to it. If we do so, if we show that sympathy is somehow a good, or the best, way to judge and decide in morally-laden situations, then the step from “is” to “ought” is less radical. The idea is this: the fact that we ordinarily have (save in cases of impairment or of reduced activation) an “empathic circuit” in our brain does not imply that we will necessarily endorse a sympathetic criterion of

Sarah Songhorian

65

judgment and decision-making. To endorse such a criterion will be the best “choice”—it is obviously not a choice that a subject can make in a completely conscious way, since we are partially the result of the interaction between deterministic factors, but it is possible to educate ourselves to it. But, if it is a “choice”, there has to be an alternative. My idea is that different degrees of sadism are the alternative. At least conceptually, sadism is the opposite of a sympathetic attitude: this makes you feel a certain emotion as if you were in the other subject’s situation, you feel it with him, exactly as he feels it; sadism, on the other hand, leads you to rejoice in other people’s pain. In this sense, if sympathy ought to be the criterion to judge and decide morally (as opposed to a sadistic attitude), then the transition from “is” to “ought” is twofold: the first move is a sort of shift within the descriptive domain, from the neural description to the metaethical one; the second is the move from sympathy as the criterion we actually use, to a normative theory prescribing the pro-social attitudes—attention, respect, care—as normative. Nothing has been said as far as the second aspect of the “is/ought question” is concerned; my focus has only been on the first one. Given the natural ability we are endowed with, our education, culture, and the environment in general can make us direct this capacity towards one of two possible attitudes: either sympathy or sadism. Obviously it is not an all-or-nothing alternative and it is not a choice than can be made once and for all: we are always merged in a continuum and we will face every situation with a certain degree of sympathy and sadism—interpreted in the broadest sense possible: everyone at least once in his life has experienced rejoicing in the failure of an opponent. The metaethical issue is, therefore, related to the concept of sympathy and shows the shift from a descriptive level to a different one that has an evaluative element, i.e. the fact that sympathy is desirable, it ought to be the direction that we give to our ability to empathize. This sense of “ought” is quite broad, and it is not the sense we want for a normative ethical theory. A normative theory requires a connection with the metaethical concept of sympathy and needs to derive from it. But this connection will not be enough, since the meaning of “ought” we desire for a moral theory is more binding than the one that can be ensured through “sympathy” alone. My idea is that such an aim could be obtained by means of the concepts of attention, respect and care, which are concepts linking subjects to each other, and that—I believe—can be normative in a strict sense.

66

Empathy and Sympathy: From Description to Prescription

5. Empathy: A Necessary, but not Sufficient Condition So, on this account, empathy can provide the natural and embedded basis for our ability to understand those who are similar to us, but it is not moral until we somehow decide that we are going to take the emotions of others as relevant for our moral judgment and decision-making, and to give a precise direction to this consideration (that is to say, whether we enact sympathy or sadism). As long as it is just an instinctive reaction, it is far from being a moral theory that tells us how we should behave towards others. The innate mechanism is, thus, necessary for the sympathetic consideration of other people, but it is not a sufficient condition to make that relationship moral: having such a mechanism is totally compatible with making immoral decisions and judgments, which leaves human beings with some freedom—it is, for example, totally compatible with some degree of sadism. If we lack such an automatic mechanism for sharing other people’s emotions, it is much more complicated to understand them, though it is not impossible by means of a more complex and mentalizing mechanism that cannot provide an immediate and emotional sharing, but an intellectual comprehension through the attribution of mental states to others. So, besides the existence of some neural circuits, a person always has the possibility of a moral option: neurosciences do not imply determinism or the abolition of free will. The neural circuits give us a choice: we can either act morally or not. And it is the concept of “possibility” that explains the first passage—that I have defined as a shift—between description and normativity, in the broad sense in which a metaethical tool can be considered normative. Is this a naturalization of morality? We can speak of naturalizing morality, if by naturalization we mean the fact of tracking down the conditions of possibility, as I have tried to do here. But we cannot, of course, if we mean a reduction of normativity to the mere description of how we actually are, of our neural endowment.

Bibliography Agosta, Lou. 2011. “Empathy and Sympathy in Ethics.” In Internet Encyclopedia of Philosophy. Accessed May 2012. http://www.iep.utm.edu/emp-symp/. Avenanti, Alessio, Domenica Bueti, Gaspare Galati, and Salvatore M. Aglioti. 2005. “Transcranial Magnetic Stimulation Highlights the

Sarah Songhorian

67

Sensorimotor Side of Empathy for Pain.” Nature Neuroscience 8: 955960. doi:10.1038/nn1481. Baron-Cohen, Simon. 2011. The Science of Evil: On Empathy and the Origins of Cruelty. New York: Basic Books. Batson, C. Daniel, Shannon Early, and Giovanni Salvarini. 1997. “Perspective Taking: Imagining how Another Feels versus Imagining how You Would Feel.” Personality and Social Personality Bulletin 23: 751758. doi: 10.1177/0146167297237008. Batson, C. Daniel, David A. Lishner, Amy Carpenter, Luis Dulin, Sanna Harjusola-Wevv, E. L. Stocks, Shawna Gale, Omar Hassan, and Brenda Sampat. 2003. “‘…As you Would Have Them Do Unto You’: Does Imagining Yourself in the Other’s Place Stimulate Moral Action?” Personality and Social Psychology Bulletin 29: 11901201. doi: 10.1177/0146167203254600. Botvinick, Matthew, Amishi P. Jha, Lauren M. Bylsma, Sara A. Fabian, Patricia E. Solomon and Kenneth M. Prkachin. 2005. “Viewing Facial Expressions of Pain Engages Cortical Areas Involved in the Direct Experience of Pain.” Neuroimage 25: 312319. doi:10.1016/j.neuroimage.2004.11.043. Carr, Laurie, Marco Iacoboni, Marie-Charlotte Dubeau, John C. Mazziotta, and Gian Luigi Lenzi. 2003. “Neural Mechanisms of Empathy in Humans: A Relay from Neural Systems for Imitation to Limbic Areas.” Proceedings of the National Academy of Sciences of the United States of America 100/9: 54975502. doi:10.1073/pnas.0935845100. Celano, Bruno. 1994. Dialettica della giustificazione pratica. Saggio sulla legge di Hume. Torino: Giappichelli. Cremaschi, Sergio. 2005. “Adam Smith antiutilitarista.” La società degli individui 24/VIII/ 3: 116. Dalgleish, Tim. 2004. “The Emotional Brain.” Nature Reviews Neuroscience 5: 583589. doi:10.1038/nrn1432. Dapretto, Mirella, Mari S. Davies, Jennifer H. Pfeifer, Ashley A. Scott, Marian Sigman, Susan T. Bookheimer, and Marco Iacoboni. 2006. “Understanding Emotions in Others: Mirror Neuron Dysfunction in Children with Autism Spectrum Disorders.” Nature Neuroscience 9: 2830. doi:10.1038/nn1611. De Vignemont, Frederique, Tania Singer. 2006. “The Empathic Brain: How, When and Why?.” TRENDS in Cognitive Sciences 10/10: 435441. doi:10.1016/j.tics.2006.08.008.

68

Empathy and Sympathy: From Description to Prescription

Eisenberg, Nancy. 2000. “Empathy and Sympathy.” In Handbook of Emotions. Edited by Michael Lewis, Haviland-Jones Jeannette M., Lisa Feldman Barrett, 677691. New York/London: Guilford Press. — and Janet Strayer. 1987. “Critical Issues in the Study of Empathy.” In Empathy and Its Development. Edited by Nancy Eisenberg, and Janet Strayer, 3-16. Cambridge: Cambridge University Press. —. 1987. Empathy and Its Development. Cambridge: Cambridge University Press. Escalas Jennifer E., and Barbara Stern. 2003. “Sympathy and Empathy: Emotional Responses to Advertising Dramas.” Journal of Consumer Research 29/4: 566578. doi: 10.1086/346251. Gallese, Vittorio. 2001. “The «Shared Manifold» Hypothesis: From Mirror Neurons to Empathy.” Journal of Consciousness Studies 8/5-7: 3350. —. 2003. “The Roots of Empathy: The Shared Manifold Hypothesis and the Neural Basis of Intersubjectivity.” Psychopathology 36: 171180. doi: 10.1159/000072786. —, Luciano Fadiga, Leonardo Fogassi, and Giacomo Rizzolatti. 1996. “Action Recognition in the Premotor Cortex.” Brain 119: 593609. doi: 10.1093/brain/119.2.593. —, Christian Keysers, and Giacomo Rizzolatti. 2004. “A Unifying View of the Basis of Social Cognition.” Trends in Cognitive Sciences 8: 396403. doi:10.1016/j.tics.2004.07.002. Goldman, Alvin I. 1993. “Ethics and Cognitive Science.” Ethics 103/2: 337360. Gordon, Robert M. 1995. “Sympathy, Simulation, and the Impartial Spectator.” Ethics 105/4: 727742. Greene, Joshua D. 2003. “From Neural “Is” to Moral “Ought”: What are the Moral Implications of Neuroscientific Moral Psychology?.” Nature Reviews Neuroscience 4: 847850. doi: 10.1038/nrn1224. —. 2009. “The Cognitive Neuroscience of Moral Judgment.” In The Cognitive Neurosciences IV. Edited by Michael S. Gazzaniga, 9871000. Cambridge: MIT Press. Greene, Joshua D., and Jonathan Haidt. 2002. “How (and Where) Does Moral Judgment Work?.” TRENDS in Cognitive Sciences 6/12: 517523. doi:10.1016/S1364-6613(02)02011-9. Greene, Joshua D., R. Brian Sommerville, Leigh E. Nystrom, John M. Darley, and Jonathan D. Cohen, 2001. “An fMRI Investigation of Emotional Engagement in Moral Judgment.” Science 293: 21052108. doi: 10.1126/science.1062872.

Sarah Songhorian

69

Hume, David. 1739-1740. Treatise of Human Nature, London: John Noon-Thomas Longman. Re-edited 1969. London: Penguin. Iacoboni, Marco. 2009. “Imitation, Empathy, and Mirror Neurons.” Annual Review of Psychology 60: 653670. doi: 10.1146/annurev.psych.60.110707.163604. Iacoboni, Marco, Roger P. Woods, Marcel Brass, Harold Bekkering, John C. Mazziotta, and Giacomo Rizzolatti. 1999. “Cortical Mechanisms of Human Imitation.” Science 286: 25262528. doi: 10.1126/science.286.5449.2526. Jackson, Philip L., Andrew N. Meltzoff, Jean and Decety. 2005. “How Do We Perceive the Pain of Others? A Window into the Neural Processes Involved in Empathy.” Neuroimage 24: 771779. doi:10.1016/j.neuroimage.2004.09.006. Joyce, Richard. 2008. “What Neuroscience Can (and Cannot) Contribute to Metaethics.” In Moral Psychology Volume 3: The Neuroscience of Morality: Emotion, Brain Disorders, and Development. Edited by Walter Sinnott-Armstrong, 371394. Cambridge: MIT Press. Keysers, Christian, Bruno Wicker, Valeria Gazzola, Jean Luc Anton, Leonardo Fogassi, and Vittorio Gallese. 2004. “A Touching Sight: SII/PV Activation during the Observation and Experience of Touch.” Neuron 42: 335346. doi: 10.1016/S0896-6273.04.00156-4. Lamm, Claus, C. Daniel Batson, and Jean Decety, Jean. 2007. “The Neural Substrate of Human Empathy: Effects of Perspective-taking and Cognitive Appraisal.” Journal of Cognitive Neuroscience 19/1: 4258. doi:10.1162/jocn.2007.19.1.42. Langfeld, Herbert Sidney. 1967. The Aesthetic Attitude. Port Washington, New York: Kennikat. Meltzoff, Andrew N., and A. Keith Moore. 1977. “Imitation of Facial and Manual Gestures by Human Neonates.” Science 198/4312: 7578. doi: 10.1126/science.198.4312.75. —. 1983. “Newborn Infants Imitate Adult Facial Gestures.” Child Development 54: 702709. Meyer, Meghan L., Carrie L. Masten, Yina Ma, Chenbo Wang, Zhenhao Shi, Naomi I. Eisenberger, and Shihui Han. 2012. “Empathy for the Social Suffering of Friends and Strangers Recruits Distinct Patterns of Brain Activation.” Social Cognitive and Affective Neuroscience 19. doi: 10.1093/scan/nss019. Morrison, India, Donna Lloyd, Giuseppe Di Pellegrino, and Neil Roberts. 2004. “Vicarious Responses to Pain in Anterior Cingulate Cortex: Is Empathy a Multisensory Issue?.” Cognitive, Affective & Behavioural Neuroscience 4: 270278. doi: 10.3758/CABN.4.2.270.

70

Empathy and Sympathy: From Description to Prescription

Pfeifer, Jennifer H., Marco Iacoboni, John C. Mazziotta, and Mirella Dapretto. 2008. “Mirroring Others’ Emotions Relates to Empathy and Interpersonal Competence in Children.” NeuroImage 39: 20762085. doi:10.1016/j.neuroimage.2007.10.032. Preston, Stephanie D., and Frans B. M. De Waal. 2002. “Empathy: Its Ultimate and Proximate Bases.” Behavioural and Brain Sciences 25: 172. Rawls John. 1971. A Theory of Justice. Cambridge: Harvard University Press. —. 1999. Collected Papers. Cambridge: Harvard University Press. Rizzolatti, Giacomo, Rosolino Camarda, Leonardo Fogassi, Maurizio Gentilucci, Giuseppe Luppino, and Massimo Matelli. 1988. “Functional Organization of Inferior Area 6 in the Macaque Monkey. II. Area F5 and the Control of Distal Movements.” Experimental Brain Research 71: 491507. doi: 10.1007/BF00248742. Rizzolatti, Giacomo, and Laila Craighero. 2004. “The Mirror-Neuron System.” Annual Review Neuroscience 27:169192. doi: 10.1146/annurev.neuro.27.070203.144230. Rizzolatti, Giacomo, Luciano Fadiga, Vittorio Gallese, and Leonardo Fogassi. 1996. “Premotor Cortex and the Recognition of Motor Actions.” Cognitive Brain Research 3: 131141. doi: 10.1016/09266410(95)00038-0. Rizzolatti, Giacomo, Leonardo Fogassi, and Vittorio Gallese. 2001. “Neurophysiological Mechanisms Underlying the Understanding and Imitation of Action.” Nature Reviews Neuroscience 2: 661670. doi: 10.1038/35090060. Rizzolatti, Giacomo, and Corrado Sinigaglia. 2006. So quello che fai. Il cervello che agisce e i neuroni specchio, Milano: Raffaello Cortina. Scheler Max Ferdinand. 1923. “Wesen und Formen der Sympathie.” In Gesammelte Werke. Vol. VII. Edited by M. Frings, Bern und München: Francke Verlag. Translated by P. Heath. 2008. The Nature of Sympathy. New Brunswick, New Jersey: Transaction Publishers. Shamay-Tsoory, Simone G., Judith Aharon-Peretz, and Daniella Perry. 2009. “Two Systems for Empathy: A Double Dissociation between Emotional and Cognitive Empathy in Inferior Frontal Gyrus versus Ventromedial Prefrontal Lesions.” Brain 132: 617627. doi: 10.1093/brain/awn279. Simner, Marvin L. 1971. “Newborn’s Response to the Cry of Another Infant.” Developmental Psychology 5: 136150. doi: 10.1037/h0031066.

Sarah Songhorian

71

Singer, Tania, Ben Seymour, John O’Doherty, Holger Kaube, Raymond J. Dolan, and Chris D. Frith. 2004. “Empathy for Pain Involves the Affective but not Sensory Components of Pain.” Science 303: 11571162. doi: 10.1126/science.1093535. Singer, Tania, Ben Seymour, John O’Doherty, Klaas E. Stephan, Raymond J. Dolan, and Chris D. Frith. 2006. “Empathic Neural Responses are Modulated by the Perceived Fairness of Others.” Nature 439: 466469. doi:10.1038/nature04271. Smith, Adam. 1853. The Theory of Moral Sentiments or An Essay Towards an Analysis of the Principles by which Men Naturally Judge Concerning the Conduct and Character, first of their Neighbours, and Afterwards of Themselves, London: Henry G. Bohn. Stein, Edith. 1917. Zum Problem der Einfühlung, Halle. Translated by Stein W. J. 1989. On the Problem of Empathy. The Collected Works of Edith Stein Sister Teresa Bendicta of the Cross Discalced Carmelite Volume Three. New York: Springer. Wicker, Bruno, Christian Keysers, Jane Plailly, Jean-Pierre Royet, Vittorio Gallese, and Giacomo Rizzolatti. 2003. “Both of us Disgusted in My Insula: The Common Neural Basis of Seeing and Feeling Disgust.” Neuron 40: 655664. Zahavi, Dan. 2001. “Beyond Empathy. Phenomenological Approaches to Intersubjectivity.” Journal of Consciousness Studies 8/5-7: 151167. doi: 10.3846/coactivity.2010.08. —. 2007. “Expression and Empathy.” In Folk Psychology Reassessed. Edited by Daniel D. Hutto, Matthew Ratcliffe, 2540. Dordrecht: Springer. —. 2008. “Simulation, Projection and Empathy.” Consciousness and Cognition 17/2: 514-522. doi: 10.1016/j.concog.2008.03.010. —. 2010. “Empathy, Embodiment and Interpersonal Understanding: From Lipps to Schutz.” Inquiry 53/3: 285330. doi: 10.1080/00201741003784663.

Notes * Ph.D. Student in Cognitive Neuroscience and Philosophy of Mind, Vita-Salute San Raffaele University, Milan and IUSS Ne.T.S., Pavia; CeSEP Center for Public Ethics Studies Vita-Salute San Raffaele University, Milan; PERSONA Research Centre in Phenomenology and Sciences of the Person, Vita-Salute San Raffaele University, Milan. 1 The most important difference between the definitions I would like to propose here and those exposed by Preston and De Waal is that I have a restrictive concept of “empathy”, while they have a very broad one (Preston and De Waal 2002, 4,

72

Empathy and Sympathy: From Description to Prescription

figure 1). In their view the concept of “empathy” includes a variety of different phenomena. My idea is that a more restrictive definition of all the related items will prove more useful since it will enable us to use different words for different (but related) phenomena, providing a more detailed conception of them.  2 The fact that the word “empathy” was introduced later than “sympathy” in the English lexicon can help in the interpretation of the works of some authors like David Hume or Adam Smith: the fact that they do not distinguish between the two concepts could be seen as a mere consequence of the fact that they did not have two different words (Agosta 2011).  3 A very interesting project would be to investigate more deeply the relationship between Adam Smith’s concept of “sympathy” and that of “the ideal spectator”. I believe that the ideal spectator can be interpreted as an interiorization of the experience done with real spectators judging the subject’s behaviour, so that the ideal spectator represents the subject’s own conscience in progressive formation. I already noted that Adam Smith could not distinguish between “empathy” and “sympathy” since the first term was not in use in the English lexicon at the time he wrote his Theory of Moral Sentiments; so he used the term “sympathy” both for what I have considered empathy as a natural and embedded capacity and for the metaethical criterion (different examples of this double use can be found in the text).  4 Simon Baron-Cohen will add to this list: borderline personality disorder, antisocial personality disorder, and narcissism (Baron-Cohen 2011).

PART TWO

DEFINING PRACTICAL REASONING: CONSTRUCTIVISM AND INSTRUMENTAL REASON SOFIA BONICALZI* (UNIVERSITY OF PAVIA)

1 On what basis can we say that a society is not rational if it unfairly limits the personal freedom of its members? Why is it usually considered more rational to work hard in order to get good grades at school, rather than to develop criminal skills? In what sense is it rational to take the shortest way home when we have little time available? All these issues seem to belong to the sphere of practical reasoning: this kind of reasoning asks what we should do in certain situations of choice, and aims at reaching normative solutions suitable to direct our life and inspire our behaviour. The attempt to adapt our behaviour to a certain standard could have originated from different sources: we might have wanted our life to embody principles and values considered as important (independently from the ontological status we confer to them), or we might have realised that conforming to certain criteria helped us to reach some goals that we wished to achieve. Probably, when we analyse the features of a political institution, the rationality we seek is associated with the conformity between the choices that are made and the concept of justice, or the system of laws, in which we believe. If our objective is to find the quickest way to get home, rational choice seems instead the result of a calculation of costs and benefits associated with the options that are actually available. In the first case, rationality is thought to be strongly connected to questions of normative ethics that affect how we assess the world, investing issues related to the role of ethical principles and to the power of practical reasoning to produce motivations for action. In the latter, rationality is seen as independent from moral concerns and rather related to

Sofia Bonicalzi

75

instrumental requirements—a solution that sometimes demands a form of specialized knowledge. In both cases the choices and assessment we make are thought to have consequences on our life, representing a precedent for the development of our future behaviour. Sometimes acting in accordance with reason seems to reflect the need to behave considering what is thought of as healthy or normal, so to be respectful, in a way, of the standards of our common humanity. Otherwise, practical rationality can be understood as the capacity to make the best use of the means at our disposal. Nevertheless, the reduction of these questions to a common ground, founded on the mere distinction between practical (action oriented) and theoretical (predictive and explanatory) reasoning, seems to be a challenging endeavour (about the consistency of the agential perspective as autonomous from the scientific one, see De Caro 2010, 101). Even though theoretical reasoning about truth also requires us to adhere to it, when we recognise the outcomes of practical reasoning as consistent we are usually pushed to conform our existence to their requests, in a way that looks intuitively different from the natural tendency to believe in propositions theoretically accepted as valid. Are these approaches mutually exclusive? What is—if there is one— the bond between practical reasoning and normativity? Which is the best explanation for our everyday choices? Is the reference to the constant need to increase our personal utility sufficient to explain our behaviour? Is practical reason also able to provide the motivational set that inspires the action, as the Platonic-Kantian tradition suggests, or is it something “inactive”, which “can never be the source of so active a principle as conscience, or a sense of morals” (Hume 1978, 3.1.1.11)? Does morality itself have an intrinsic value or an instrumental one? In seeking to avoid an improper juxtaposition between the concepts of “reason” (as a human faculty) and of “rationality” (as the ability to develop a specific strategy), this section attempts to reflect on some of the main declinations in which the discourse on practical reason can be articulated and understood.

2 Francesca Vitale’s contribution explores the versions of moral constructivism developed by John Rawls (1971; 1999) and Onora O’Neill (1989; 1996), proposing a valuable comparison of their ethical-political accounts based on the contrast between the concepts of “idealisation” and “abstraction”. These key elements are understood as the theoretical tools

76

Defining Practical Reasoning

adopted by the two philosophers in order to articulate their own conception of practical reasoning. Constructivism probably represents the main heritage of Kantian ethical thought in the contemporary debate (Rawls 1980), extending its influence both in meta-ethical reflection and in moral theory, and thus contributing to the debate about the ontological status of moral values and shaping our normative accounts of morality. This is clearly not the place to reconstruct the multiple ramifications of Kantian constructivism, nor to stress analogies and differences between Kantian practical thought and the subsequent expression of Kantianism in ethics, focusing on the autonomy and objectivity of morality, and of its inner connection with human freedom. As a meta-ethical standpoint, opposing both Humean sentimentalism and rational intuitionism, constructivism takes the role of practical reasoning as central in defining the principles of morality, even though not everyone believes it to be capable of expressing a clear and consistent perspective (Darwall-GibbardRailton 1992). Suffice it to say that the same attribution of a form of constructivism to Kantian thought is questioned (Krasnoff 1999; Audi 2004; Kain 2006b; Irwin 2009), and the core element of Kantian ethical reflection (the moral law expressed by the categorical imperative) can be considered not as an autonomous product of human practical reason, but as the description of the norm that rules God’s will (Fonnesu 2006, 23; Landucci 1993, 32). Usually, at least in its more influent versions (Korsgaard 1996a, 36; O’Neill 1989, 206), constructivism is presented as a third way between moral realism and anti-realism (Bagnoli 2007, 257), emphasizing the intrinsic value of morality in connection with practical reasoning (Darwall 2006), which provides an objective anchor to a morality whose principles are not grasped, but built by reason itself (Mordacci 2008, 167). At the same time, the Kantian theory of moral obligation represents one of the unavoidable terms of comparison in contemporary political and ethical debate (Copp 2006, 17), embodying one of the main alternatives in the discussion about normativity, alongside consequentialist ethics, virtue ethics and other forms of deontologism. Vitale firmly includes the constructivist proposals by Rawls and O’Neill in the range of deontological ethics, based on the priority of duty over the other moral concepts. As is well known, Rawls offers a normative (contractualist) interpretation of Kantianism, paying less attention to the metaethical disputes over the status of moral properties and values: the reflection on practical reasoning stems directly from the need to justify the principles that structure a well-ordered society, formed by free and autonomous Kantian agents. Vitale carefully outlines the terms of the Rawlsian thought

Sofia Bonicalzi

77

experiment relating to the particular choice situation (original position, veil of ignorance, reflective equilibrium) from which the first principles of justice derive. How to ensure that the agreement reached by the participants in social cooperation is based on fair principles? The plausibility of such an agreement is based on the nature of the reasonable constraints imposed on the original choice situation: the development of a context of choice that is immune from irrelevant information represents the source of a public conception of justice (for rational and equal human beings living in a wellordered society) founded on an objective and sharable basis. The first principles of justice come necessarily from the rightness of the procedure in the “purity of a consequence independent system” (Sen 2002, 279). But is it true that a coherent moral or political theory can be structured eliminating the allegedly unnecessary information? In other words, could the universalist pretension to produce a synthetic moral account be satisfied, or is it necessary to take into account the contextual dimension from which ethical demand arises, as particularist thinkers (Taylor 1985; MacIntyre 1984; Sandel 1982; Williams 1985; Dancy 2004; Hooker & Little 2000) claim? Vitale discusses the outcomes of Rawlsian thought, deconstructing the process and analysing their validity in light of the criticism (proposed primarily by communitarians like Sandel, 1982) of an excessive dose of abstraction, thus being inadequate in explaining the mechanisms that structure complex societies, while also offering an account solely modelled on Western democracies. Taking the objection seriously, the author addresses the issue employing a double strategy, proposing firstly the reconstruction of the structural points of Rawls’ theory according to the methodological model (Idealizational Theory of Science) developed by the School of Poznan, and in particular by Leszek Nowak (1980; 2000; 2004), one of the main protagonists of that intellectual experience known as the Marxist Polish School, which tried to apply scientific methods to human science. Then Vitale proceeds to compare Rawlsian constructivism to Onora O’Neill’s model which, in the attempt to challenge both the theorizations of ethical universalists and the traditional claims of particularists, avoids the idealisation procedure (accused of distorting the truth about human identity with arbitrary assumptions, falling into indeterminateness) in favour of abstraction, intended as a fundamental aspect of every form of practical reasoning. Focusing on O’Neill’s emphasis on abstraction, Vitale displays the importance of practical reasoning (act-oriented) for the construction of universally valid basic principles coherently sharable and followable by all (in the name of Kantian cosmopolitan reason), on the

78

Defining Practical Reasoning

basis of certain minimal assumptions that preserve the truth about individual identity.

3 The Rawlsian reflection offers us a bridge towards the other contribution presented in this section of the volume. The connection between practical reasoning and the principles of justice is evident in Rawls’ fundamental Kantian Constructivism in Moral Theory (1980), where the philosopher distinguishes two elements contained in the idea of social cooperation (528). The reasonable is what people that are part of the mechanism of social cooperation are expected to accept, independently from their personal ends and life plans. It is something related to the fair terms of cooperation that free and equal members of a well-ordered society should share according to an idea of reciprocity and mutuality. This expectation is understandable only with reference to the intersubjective sphere, while the rational (subordinated and, in a way, organized by the reasonable) is connected to the individuality of each participant who calculates the value of social cooperation according to the principles of rational choice, considering the real opportunities of reaching his or her personal ends within a given community. There could certainly be difficulties on the side of the rational—the Weberian zweckrational— that may be related to the problematic identification of a clear definition of rationality, but the real challenge for Kantian constructivism, in Rawls’ interpretation, derives from the realm of the reasonable: the model of rational choice can be intended as the method by which the contractors choose the first principles in the original position, but they are immediately associated with the Kantian imperative, whose validity is independent from our specific ends; “Kantian constructivism begins from a conception of the person and of practical reason, the ideal of free and equal moral persons who are both reasonable and rational” (Freeman 2003, 27). Martina Belmonte’s contribution leaves the sphere of the reasonable and enters the pragmatic perspective adopted by decision theory and theory of rational choice, extensively employed in contemporary economic, political and social disciplines. As we will see, here too one of the main challenges is to do with a problem of inclusion and exclusion: which elements should the theory consider in order to provide a convincing explanation of a certain practical phenomenon (“Any principle of choice uses certain types of information and ignore others.” Sen 2002, 348)? How to employ precious tools such as idealisation and abstraction

Sofia Bonicalzi

79

(Cherniak 1995)? What are the acceptable implicit or explicit information constraints (Sen 1970a; b; 1979)? According to the approaches explored by Belmonte, individual rationality is intended as a tool to solve technical and practical problems including, if necessary, the contributions of specific scientific disciplines. In this sense, individual rationality is not intended as a superior authority that has the role of defining personal and collective goals and objectives (or principles), maybe referring to (internal or external) values. It is rather the calculator that fulfils the function of deciding which is the best way to achieve a specific goal (in the least expensive manner), thus maximizing personal utility. The survey on the nature of rationality allows predictions about the ways in which individuals interact with the environment (Nozick 1993). The rational choice model of action has been widely applied in social, political and economical sciences, because it has been seen as appropriate to explain the agents’ behaviour in different contexts (M. Friedman 1953), even though its predictive power in many fields has been strongly criticized (Barry 1970; Green and Shapiro 1994). According to the normative interpretation of the theory, the agent’s choices may be subject to criticism not because they deviate from principles recognized as valuable by our reason (unless maximizing personal utility itself is treated as a normative value), but if the particular choice made will not turn out to be the best possible option in view of a specific purpose (it is worth noticing that Kant seems instead to exclude this sphere from the same realm of practical reasoning, coincident with morality itself, considering instrumental considerations as implication of theoretical, and not practical, thought. See Fonnesu 2007, 17). In other words, in this sense, the eventual inability to choose the alternative that maximizes personal utility is considered irrational, as this maximization is in itself the objective the agents have to achieve: it is deeply incoherent to want the thing A, considering B necessary to reach A, and deciding not to do B (Darwall 2006, 294). Even though the definition of rationality offered by the theory of rational choice seems plain, there are some points to explore. Firstly it is not uncontroversial whether, according to its traditional structure, the image of individual rationality provided by the above-mentioned theory is valid. Moreover, we could also ask whether the assumptions about the way in which the agents are supposed to behave in a context of choice are correct. Belmonte suggests the possibility of integrating the results of the theory of rational choice with the outcomes offered by gender studies with respect to certain contexts of choice. The theory of rational choice inevitably discloses a set of useful tools to delineate the phenomenology of

80

Defining Practical Reasoning

the actors’ behaviour. However, even though we agree to exclude from the competences of practical reasoning the definitions of people’s duties, it is doubtful if this theory could also work as a general explanation of human rationality. It seems particularly unable to explain attitudes such as trust, altruism or the natural tendency of human beings to form groups independently, at least in appearance, from their expected utility. Moreover, in some cases, following the model of rational choice seems to lead to unpleasant consequences (e.g. collectively suboptimal equilibria), also contrasting with the more general conception of what is reasonable (Rescher 1988, vii, 6; McClennen 1990, 4-5; Gauthier 1985b, 85-86). From an empirical-epistemological perspective, further limitations appear to be related to the assumption that human beings have a clear understanding of the available alternatives, being able to order them in a hierarchy and choose the best option in view of the maximization of personal utility (according to requisites of completeness and transitivity). The basic assumptions of the theory outline a specific vision of the agents’ identity, conceived as individuals who base their behaviour on rational calculations (and thus entering into intersubjective relations only if cooperation favours an increase in the utility rate), act rationally and make choices that are thought to maximize their well-being (the maximum utility). The theory of rational choice implies a concept of “rationality” certainly narrower compared to the one included in ordinary language and perception, excluding the reference to duties and moral values. “On such a view all action must inevitably aim at preference satisfaction, and further reason would be needed to show why the satisfaction of any particular preferences was ethically desirable” (O’Neill 1998, 34). Moreover, according to the well-known axiom of revealed preference, firstly applied to analyse consumer behaviour, individual preferences can be directly assimilated to choices, since the evident behaviour of individuals in the context of choice immediately reveals their preferences (the preferred options), regardless of any reference to internal motivations or states of mind. As previously considered, the main point highlighted by Belmonte’s analysis does not concern the idea that rationality is not reducible to the utilitarian calculation of costs and benefits of the available options (since, in a broader sense, it may contribute itself to the determination of goals and projects, also providing the set of motivations for action), but that the narrow concept of personal identity—at the basis of the theory—is insufficient in explaining people’s behaviour in certain contexts. What we may question is whether an instrumental rationality theory, which explains our behaviour taking into account only revealed preferences, could really

Sofia Bonicalzi

81

be a useful key for understanding our conduct, and simultaneously aspiring to cover a normative function (guiding the action distinguishing between right and wrong on the basis of utility rate). How could we refine this conception? Considering the traditional account as unsatisfactory, Belmonte proposes some integrations to the theory, revising Amartya Sen’s objections (1973; 1977; 1993; 1995) to the axiom of revealed preferences in the light of the acquisitions of gender studies. Sen is sceptical towards the basic presumption of explaining the mechanisms of choice by focusing only on the internal coherence of the choice function, without any reference to elements external to (i.e., irrelevant to) the mechanism of choice, such as personal and social motivations, or other psychological and contextual factors. Belmonte accepts the reply of those who propose to save the traditional account of the theory, reformulating Sen’s case studies in a way that reduces their intuitive force (Binmore 2009), but presents some choice situations that seemingly require a more substantial explanation, not reducible to the outcome of the mere balance of alternative options. So, as we will see, in her widely documented account, the gender of the subject involved in the choice process—usually not considered by the theory of rational choice (as irrelevant to the characterization of the individual as a rational self-interested entity)—is interpreted as a key element in order to understand the outcome (gendered epistemic value of the menu) and to explain the mechanisms that structure the society we live in, thus leading to a general revision of the theory itself.

Bibliography Audi, Robert. 2004. The Good in the Right: A Theory of Intuition and Intrinsic Value. Princeton: Princeton University Press. Bagnoli, Carla. 2007. Il Costruttivismo kantiano. In Oggettività e Morale, edited by Giorgio Bongiovanni, Milano: Bruno Mondadori. Barry, Brian. 1970. Sociologists, Economists and Democracy. London: Collier-MacMillan. Binmore, Ken. 2009. Rational Decisions. Princeton: Princeton University Press. Cherniak, Christopher. 1995. “Rationality.” In A Companion to the Philosophy of the Mind, edited by Samuel Guttenplan, 526-531. Cambridge: Blackwell Publishers. Dancy, Jonathan. 2004. Ethics without principles. Oxford, New York: Oxford University Press.

82

Defining Practical Reasoning

Darwall, Stephen. 2007. Morality and Practical Reason: A Kantian Approach. In The Oxford Handbook of Ethical Theory, edited by David Copp, 282-320. U. S.: Oxford University Press. Darwall, Stephen, Allan Gibbard, and Peter Railton. 1992. “Toward Fin de siècle Ethics: Some Trends.” Philosophical Review, 10: 115-189. De Caro, Mario. 2010. “Naturalismo e normatività: prospettiva scientifica e prospettiva agentiva.” In Etiche antiche, etiche moderne. Temi di discussione, edited by Stefano Bacin, 101-118. Bologna: Il Mulino. Gauthier, David. 1985. “Maximization Constrained: The Rationality of Cooperation.” In Paradoxes of Rationality and Cooperation: Prisoner’s Dilemma and Newcomb’s Problem, edited by Richmond Campbell, Lanning Sowden, 75-93. Vancouver: University of British Columbia Press. Green, Donald P. and Ian Shapiro. 1994. Pathologies of Rational Choice Theory: A Critique of Applications in Political Science, New Haven: Yale University Press. Hill, Thomas E. 2007. “Kantian Normative Ethics.” In The Oxford Handbook of Ethical Theory, edited by David Copp, 480-514, U. S.: Oxford University Press. Hooker, Brad and Margaret Little. 2000. Moral Particularism, Oxford: Clarendon Press. Hume, David. 1978. A Treatise of Human Nature. Edited by Selby-Bigge, L. A. Oxford: Clarendon Press. Friedman, Milton. 1953, “The Methodology of Positive Economics” Essays in Positive Economics, Chicago: University of Chicago. Irwin, Terence. 2009. “Kant: Meta-ethical Questions.” The Development of Ethics: Volume III: From Kant to Rawls, ed. by Terence Irwin, 147173, Oxford: Oxford University Press. —. (2006), “Realism and Anti-Realism in Kant’s Second Critique.” Philosophy Compass, 1: 449-465. Korsgaard, Christine. M. 1996, The Sources of Normativity. Cambridge (MA): Cambridge University Press. Krasnoff, Larry. 1999. “How Kantian is Constructivism.” Kant Studien, 90: 385-409. Fonnesu, Luca. 2006. Storia dell’etica contemporanea: da Kant alla filosofia analitica. Roma: Carocci. Landucci, Sergio. 1993. La Critica della ragion pratica di Kant: introduzione alla lettura. Roma: NIS. MacIntyre, Alasdair. 1984. After virtue: a study in moral theory. Notre Dame, Indiana: University of Notre Dame Press.

Sofia Bonicalzi

83

McClennen, Edward F. 1990. Rationality and Dynamic Choice: Foundational Explorations. Cambridge: Cambridge University Press. Mordacci, Roberto. 2008. Ragioni Personali. Saggio sulla normatività morale. Roma: Carocci. Nozick, Robert. 1993. The Nature of Rationality. Princeton: Princeton University Press. Nowak, Leszek. 1980. The Structure of Idealization. Towards to Systematic Interpretation of the Marxian Idea of Science. Dordrecht: Reidel. —. 2004. L’approccio idealizzazionale alla scienza: una rassegna. In La realtà modellata. L’approccio idealizzazionale e le sue applicazioni nelle scienze umane, edited by Francesco Coniglione and Roberto Poli, Milano: Franco Angeli. Nowak, Leszek, Nowakowa, Izabella 2000. Idealization X: The Richness of Idealization. Amsterdam-Atlanta: Rodopi. O’Neill, Onora. 1989. Constructions of Reason. Explorations of Kant’s Practical Philosophy. Cambridge: Cambridge University Press. —. 1996. Towards Justice and Virtue. A Constructive Account of Practical Reasoning. Cambridge: Cambridge University Press. —. 2003. “Constructivism in Rawls and Kant.” In The Cambridge companion to Rawls, ed. by Samuel Freeman, 347-367. Cambridge: Cambridge University Press. Rawls, John. 1999, A Theory of Justice: Revised Edition, Oxford: Oxford University Press. —. 1971 A Theory of Justice. Cambridge: Harvard University Press. —. 1980. “Kantian Constructivism in Moral Theory.” The Journal of Philosophy, Vol. 77, No. 9 (Sep.): 515-572. Rescher, Nicholas. 1988. Rationality: A Philosophical Inquiry in the Nature and the Rationale of Reason. Oxford: Clarendon Press. Sandel, Michael. 1982. Liberalism and the Limits of Justice, Cambridge: Cambridge University Press. Sen, Amartya. 2002. Rationality and Freedom. London: Belknap Press. —. 1970a. Collective Choice and Social Welfare. San Francisco: HoldenDay, —. 1970b. “The Impossibility of a Paretian Liberal.” Journal of Political Economy. 78: 152-157. —. 1977. “Rational Fools: A Critique of the Behavioural Foundations of Economic Theory.” Philosophy and Public Affairs, 6, 4: 317-344. —. 1973. “Behaviour and the Concept of Preference.” Economica, 40, 159: 241-259. —. 1993. “Internal Consistency of Choice.” Econometrica, 61, 3: 495-521.

84

Defining Practical Reasoning

—. 1995. “Is the idea of purely internal consistency of choice bizarre?” In World, Mind, and Ethics, edited by J. E. J. Altham and Ross Harrison, 19-31. Cambridge: Cambridge University Press. Taylor, Charles. 1985. Philosophy and the Human Sciences. Cambridge: Cambridge University Press. Williams, Bernard. 1985. Ethics and the limits of philosophy. London: Fontana Press/Collins.

Notes * I am grateful to Luca Fonnesu who gave this introduction a close reading providing many determining suggestions. Further thanks go to Federico Faroldi who contributed to the linguistic revision of this section of the book.

PRACTICAL REASONING BETWEEN ABSTRACTION AND IDEALISATION: ONORA O’NEILL AND JOHN RAWLS FRANCESCA VITALE (UNIVERSITY OF SALENTO)

Abstract The paper deals with some arguments proposed by Onora O’Neill to support a constructivist thesis of practical reasoning (ethical principles are the result of a rational construction process, not of a “discovery”): there are no normative truths leading the action, regardless of the subject’s reason. I will set up a comparison between O’Neill’s thesis and the argument developed by John Rawls in A Theory of Justice. He argues that the principles of justice are the outcome of a rational construction “procedure” carried out by ideal agents in ideal conditions. I will claim that O’Neill’s approach is preferable to Rawls’. The comparison between O’Neill and Rawls on specific themes will reveal the differences that are “internal” to ethical constructivism. The internal analysis approach captures the numerous declinations of Constructivism. O’Neill’s constructivism maintains some Rawlsian traits—given the common Kantian matrix—but it diverges from Rawls’ in terms of how ethical principles should be determined. They move away from each other towards a different conception of practical reasoning: Unlike O’Neill, Rawls rejects any form of “idealisation” of the initial situation in which the agent chooses the guiding principles of his action, in favour of a process of “abstraction” (bracketing) to be applied to all aspects of agents potentially involved in the action. O’Neill supports a conception of practical reasoning that aims to establish principles that can be “followable” by all the agents on a “minimal basis” of shareability, an initial and necessary condition for the chosen principles to serve as a concrete guide for the action of each subject. This “minimal” level, in contrast to Rawls’ argument, does not presuppose any idealised conception of the agent

86

Practical Reasoning Between Abstraction and Idealisation

(original position, veil of ignorance, etc.), but it makes use of a procedure of abstraction that differs from idealisation and from any metaphysical reasoning underlying it. This paper explores some differences in the articulation of practical reasoning in O’Neill and Rawls, in order to give a picture, albeit perhaps partial, of Constructivism in ethics and the variety of its construction procedures.

Introduction The term method, as its very etymology suggests (ȝİIJȐ-ȠįȩȢ), indicates the way to be taken in the research, or rather the set of procedures to follow in order to formulate and solve a problem. In the contemporary philosophical context, ethical constructivist views—while remaining within a Kantian theoretical constellation—have taken the shape of a variegated family of analytical perspectives, in which every possible and different theoretical articulation is characterised by the acquisition of a distinctive method of construction of moral principles, its own way to be taken in normative research. Moral theories such as naturalism and intuitionism (employing a theoretical cognitive methodology for the first principles of ethics) and theories of normativity such as expressivism and sentimentalism (advocating the meta-ethical thesis that the motive of moral action is mainly based on the agent’s emotional dispositions) represent the first and necessary opportunity for comparison from which constructivism cannot shirk. To outline the peculiar features of the constructivist reflection through the comparison to the ethical theories just mentioned is an unavoidable step in to getting an overview of the main, I would say “external”, differences that mark the constructivist attempt to leave the field of cognitivism and non-cognitivism. In fact these external differences are the outcome of a comparison that adopts the forms of cognitivism and non-cognitivism as a reference system and the constructivist position as an inertial and external point of view on the system itself. The method of comparison, therefore, will be based on a centrifugal benchmarking process. In the present article I will draw a comparison that uses constructivism as a reference system and the multiplicity of its theoretical declinations as an inertial and inside point of view on the system itself. The differences that the comparison will bring to light will be defined as “internal”, as emerging from a movement of centripetal comparative analysis.

Francesca Vitale

87

I will consider Onora O’Neill and John Rawls as the poles of a comparison relationship inside the constructivist system, and I will particularly focus on the methodological tools adopted in the elaboration of their normative theories. The distinctive aspects of constructivist systems generally reflect the Kantian conception of practical reason as the source of moral normativity. Yet, within this complex system, various perspectives stir and ripen their own autonomous theoretical particularities in the interpretation of Kant’s dictation as well as their own choice of methodological criteria through which to initiate the “procedure of construction” of moral principles. In this work I will use the distinction made by Sharon Street between two principal characterisations of constructivism (Street 2010, 363-84): 1. The proceduralist characterisation of constructivism. 2. The practical standpoint characterisation of constructivism. In the first typology of constructivism, I think we can locate the procedure of construction used by Rawls, in sight of the two principles of justice, characterised by the method of idealisation. I sustain that the methodology of construction of ethical principles adopted by O’Neill belongs to the second form of constructivism, characterised by the operation of abstraction. The purpose is to highlight the way in which both methodologies converge on assigning to practical reason the fundamental role of moral normativity, while retaining their own distinct articulation. The paper will be articulated as such: The first part will offer a general overview of normative ethics and identify the elements characterising the formal structure of a constructivist theory. The second part will analyse, following the hermeneutic suggestions of the School of PoznaĔ, the idealisational model applied by Rawls to his procedural constructivism. The third part will focus on the aspects of practical reasoning and the method of abstraction that is prominent in the elaboration of O’Neill’s modal constructivism.

1. Elements of Normative Ethics In the contemporary debate various terminological distinctions have been introduced, in order to try to adequately define all the possible determinations of ethical knowledge. In philosophy, the term “ethics” refers to a set of principles, criteria and issues according to which people act. Ethical knowledge as such is constituted by two fundamental reflective moments:

88

Practical Reasoning Between Abstraction and Idealisation

1. The meta-ethical reflection: this assumes as its object of investigation the meaning of morality, its ontological statute, the nature of principles and moral values, the linguistic sense of moral language and the possibility of justifying morality itself. 2. The normative reflection: it is concerned with the justification of the different formal formulations of judgments and moral norms. Between the two forms of reflection, often subject to overlapping, it is useful to maintain a line of demarcation in order to avoid conceptual misunderstandings. More specifically, we must traditionally distinguish three types of normative theories: 1. Deontologism: this defends the thesis according to which human action is endowed, in and for itself, with moral value, because it is founded upon “duty”, a concept that has absolute priority over other moral concepts (good, aim, virtue, value). 2. Teleologism or consequentialism: this affirms the thesis according to which an act is “right” on the grounds of its aim, since the concept of “duty” is subordinate to that of the “good.” 3. Virtue ethics: this qualifies as ethics of the person, since it is not the moral act or its possible consequences that have axiological importance, but the agent itself. The constructivist proposals by Rawls and O’Neill rightly enter into the first theoretical normative typology. Ethical constructivism, in fact, puts forward the thesis that the sources of moral authority are referable to the individuals’ ability to build ethical principles through practical reasoning, rational deliberative procedures and reflective processes. To complete the tripartition presented above, I will offer a schematisation of the internal structure of a normative theory. The structure of a normative theory, like that of any theory, is modulated according to internal rules, which determine its form. Specifically, the structure of a normative theory is constituted by the following elements: x Assumptions/presuppositions/hypothesis/general facts (inputs). x Final principles/norms (outputs). x Procedures of regulation of the derivation of final principles from assumptions (derivation). In other terms, a normative theory should justify a set of principles from which some rules are inferred that are able to determine and to drive

Francesca Vitale

89

the action. As for the derivation, the sources of principles refer to the normative scheme to which the basic assumptions, contingent hypotheses and the general facts belong. According to these considerations even constructivism, as a normative theory with its own formal configuration, is translatable into the following elements1:1 x a, b, c: assumptions/conceptions/minimal hypothesis not object of discussion and bargaining. x p: procedure of construction of moral principles; a, b, c converge in p as materials or constraints for the process of construction. x n1, n2, n3: moral principles obtained through the process of construction. The formal characterisation I just described necessarily entails further details. The internal structure of a constructivist theory is shaped from a plurality of fundamental assumptions (or rather first assumptions, which are not further referable to other ones). Therefore, a normative theory having these structural presuppositions would hardly reduce the construction of ethical principles to a single primary assumption/conception translatable into the following formula: a ĺ p ĺ n1, n2, n3... Moreover, the plurality of moral principles n1, n2, n3 must not be considered as the direct conclusion following from assumptions a, b, c, because it is not included in the initial assumptions, but depends on the procedure of construction p that determines its origin: it is not by mediating a process of analytical deduction from the basic assumptions a, b, c that the articulation of normative principles n1, n2, n3 is enacted, but through the procedure of construction p. We will see, in fact, that O’Neill’s constructivist position belongs to a theoretical antifoundationalism frame, because it is opposed to any attempt of a foundation of ethics on metaphysical presuppositions or particularistic conceptions.

2. Rawls’ Idealisation in Modelling the Original Position The observations on moral theory carried out by Rawls in A Theory of Justice (1971) regard the structure of a theory of justice, which, in its initial stage, is submitted to methodological rules like any other theory. Particularly, a theory of justice is a “theory of the moral sentiments setting

90

Practical Reasoning Between Abstraction and Idealisation

out the principles governing our moral powers or, more specifically, our sense of justice” (Rawls 1999, 44). According to Veca’s interpretation, Rawls’ theory of justice is an attempt at a rational reconstruction of people’s moral sentiments through the method of “reflective equilibrium”, which allows individuals to find a point of provisional equilibrium between facts and theory, between their intuitive judgments and the principles derived from theory itself. Veca asserts: What justifies a conception of justice is nothing more than its congruence with our deepest and coherent understanding of our judgments, because our tradition is built into public life. It would be a pure intellectual eccentricity to formulate a very sophisticated moral theory that was radically divergent from the most general and properly distinctive traits of our intuitive moral considerations. It would seem, in fact, like the language of another world, of little profit and interest in comparison to our flexible and articulated natural languages (Veca 2010, 69, Translation mine).

This observation was already made in Political Philosophy (1998): You can think of […] a process of mutual adjustment between principles and judgments until it is reached at an equilibrium point, inevitably provisional (Veca 1998, 54).

This interpretation implicitly claims not to dwell only on the hypothetical dimension of the original position, but also on its intrinsic potentiality to be simulated every time that the individuals of a society are prepared to appraise or to justify the institutional order to which they belong. Rawls offers the following methodological indications: Moral philosophy must be free to use contingent assumptions and general facts as it pleases. There is no other way to give an account of our considered judgments in reflective equilibrium. This is the conception of the subject adopted by most classical British writers through Sidgwick. I see no reason to depart from it (Rawls 1999, 44-45).

The focus of Rawls’ moral theory resides, therefore, in the relationship between two different justifying strategies: 1. The original position: ideal situation of rational choice of the principles of justice, interpretable according to theoretical-deductive methodological parameters.

Francesca Vitale

91

2. The reflective equilibrium: inductive method of research of an “Archimedean point” between our considered judgments regarding what is right and the regulatory principles which inform them. The method of reflective equilibrium in relation to the original position shows that Rawls’ theory, like any theory, establishes a solid relation with “general facts”. The use of a hypothetical conceptualisation of the initial situation doesn’t compromise the connection between the “theoretical construction” and the “sense of morality” (Maffettone 2010, 78-79). As Rawls states: A correct account of moral capacities will certainly involve principles and theoretical constructions that go much beyond the norms and standards cited in everyday life; it may eventually require fairly sophisticated mathematics as well. […] Thus the idea of the original position and of an agreement on principles there does not seem too complicated or unnecessary. Indeed, these notions are rather simple and can serve only as a beginning (Veca 1998, 42).

I will now analyse the theoretical construction of the original position, characterised by proceduralism and idealisation. For this purpose I shall avail myself of the methodological contribution of the School of Poznàn, which offers an innovative interpretation of the procedure of idealisation.1 Among the greatest merits of one of the main theorist of the School of Poznàn, L. Nowak, is to his proposed application of an idealisational methodology to the field of human sciences (history, sociology, political theory, psychology. See Brzechczyn 2009; Coniglione and Poli 2004). The interpretation of Rawls’ procedure of construction in the light of the methodological contributions of the Polish School makes it possible to overcome the principal objections to Rawls in relation to the hypotheticalideal conceptualisation of the original position. Here, as we have said, the parties, as free and equal citizens, rationally establish an agreement to choose unanimously the principles of justice: a hypothetical agreement, though, is not an agreement, because it is ineffective as a constraining factor in a real situation. (Dworkin 1975, 17-21). An agreement is able to establish constraints and restrictions only if actually instituted in reality. What is an idealisational methodological model? Idealisation differs from abstraction, even if the two procedures are often confused. This terminological confusion is bound, according to Nowak, to the use that philosophy of science makes of abstraction with an empirical perspective. As Nowak distinguishes:

92

Practical Reasoning Between Abstraction and Idealisation Abstraction consists in a passage from properties AB to A, idealization consists in a passage from AB to A-B. For instance, the move from the notion of an open capitalist economy (CEO) to the notion of a capitalist economy (CE) is an act of abstraction, whereas an act of idealization would consist, for instance, in a passage from CEO to the notion of a closed capitalist economy (CE-O). […] The researcher, introducing idealizing conditions of the form p(x) = 0 eliminates factors thought to be secondary. What remains is the factor considered principal for the determined magnitude (Nowak and Nowakowa 2000, 116-117).

The core of the distinction revolves around the different modality with which the two procedures, abstraction and idealisation, are compared to the object. Abstraction puts aside some aspects of the object considered, while idealisation eliminates some characteristics of the object in matter in order to emphasise more important ones. Following an abstractive operation, the object doesn’t suffer any alteration; instead, following an idealisational procedure, the object undergoes a modelling that sets itself on a different level than the starting one. The theoretical procedure of idealisation is based on three different, but complementary, methodological moments: 1. The use of idealisational assumptions. 2. The process of concretisation of idealisational assumptions. 3. The empirical check of the theory. According to Nowak, the phases of idealisational methodology are “mutually connected” (Nowak and Nowakowa 2010, 204) because the theoretical content of every phase can be questioned by the following one. According to the classical interpretation of the theoretical project built in A Theory of Justice (1971), Rawls’ theory offers: 1. An internal structure submitted to methodological rules like any scientific theory. 2. A contractualist perspective of justice, based on procedures and decisional rules of rational character, which lead individuals to choose the principles of justice on the basis of a unanimous agreement. 3. A merely hypothetical contractualist nature, because the agreement between the parties stems from the expository device of the original position and the veil of ignorance; it does not reflect a historical fact that really happened. The subscription of such an agreement, in fact, is argued through propositions and conditional theoretical constructions.

Francesca Vitale

93

4. A sequence of hypothetical agreements based on a decisional process directed at the rational choice of fundamental principles of justice. Therefore, Rawls’ theory of justice sets itself up as a hypotheticalcontractual-decisionalist theory. The problem that every standard interpretation of Rawls’ theory has to face is tied to the third aspect described above: namely the purely hypothetical and ahistorical character of the agreement stipulated in the original position. The individuals who enter the agreement are idealised agents, submitted in turn to idealised conditions, which assure impartiality in the choice of principles of justice. In fact, the use of hypothetical conceptualisation of the original position, as a heuristic and expositive expedient of normative assumptions, raised numerous criticisms in the contemporary philosophical debate. The veil of ignorance, for example, would seemingly deprive the individuals of their particular characteristics, reducing them to disincarnated agents, not actually belonging to the real world. Therefore, the principles of justice chosen in a situation of merely ideal agreement cannot have effectiveness and applicability in an actual existing society. Sandel undoubtedly raises one of the most authoritative criticisms of Rawls’ theory of justice in his essay Liberalism and the Limits of Justice of 1982. Sandel analyses the essential points of Rawlsian theoretical construction, pointing out the anthropological difficulties that the ex hypothesi theorisation of the original situation raises: the Rawlsian subject, called to rationally decide the principles of justice, does not embody any value or specific end, because he is deprived of his identity requirements by the informative tie of the veil of ignorance (For Sandel’s objections to Rawls’ theory of justice see Dalle Fratte 1995). Sandel’s critique can be enclosed in the following question: if the premises of the original position do not derive from an empirical reasoning, but from some purely hypothetical postulations, what can they be worth for real human beings, who are responsible for approving their plausibility? The counter-factual theorisation of the original position is, therefore, one of the most problematized thematic cruxes for Rawls’ interpreters and critics. According to the interpretation of Rawls’ procedure of construction offered by the School of PoznaĔ, the elements related to the counterfactual theorisation and the respective idealisational methodology play a key role: Rawls’ theory of justice shapes itself as a hypotheticalcontractual-decisionist-idealisational theory (For this series of adjectives I

94

Practical Reasoning Between Abstraction and Idealisation

refer to Przybyz 2004, 276). How does Rawls apply idealisation as the method of construction of the original position? Following the hermeneutic view of the Idealizational Theory of Science (hereafter ITS), Rawls’ theory is articulated as a scientific theory: it is based on a sequence of idealisational statements (models), which are followed in decreasing order, according to their degree of abstraction and generality, up to the attainment of a final level of concretisation. The final level, where the sequential movement comes, is the one that draws closest to reality.2 Tracing the phases described by the ITS model, Rawls’ theory develops in the following way. Going into the detail: the original position moves from a counter-factual theorisation based on a series of assumptions in which the individuals and the circumstances of the choice of the principles of justice possess an idealising character. Rawls defines as “subjective circumstances of justice” the aspects characterising the individuals, while the elements external to them are “objective circumstances of justice”: The circumstances of justice may be described as the normal conditions under which human cooperation is both possible and necessary [...] there are the objective circumstances which make human cooperation both possible and necessary. Thus, many individuals coexist together at the same time on a definite geographical territory […] natural and other resources are not so abundant that schemes of cooperation become superfluous, nor are conditions so harsh that fruitful ventures must inevitably break down […] The subjective circumstances are the relevant aspects of the subjects of cooperation, that is, of the persons working together. Thus while the parties have roughly similar needs and interests, or needs and interest in various ways complementary, so that mutually advantageous cooperation among them is possible, they nevertheless have their own plans of life. These plans, or conceptions of the good, lead them to have different ends and purposes, and to make conflicting claims on the natural and social resources available (Rawls 1999, 109-110).

In the circumstances of justice the terms cooperation and conflict are central, because they qualify a scheme of society subtended to Rawls’ theory of justice: for the problem of social justice to emerge, a specific conception of society must be presupposed, containing the circumstances determining the necessity to build principles of justice. The set of these objective circumstances and those characterising the plurality of individuals in the original position are part of the basic assumptions of Rawls’ theoretical system. Rawls describes the aspects relating to the initial situation right through these assumptions in the third chapter of A

Francesca Vitale

95

Theory of Justice (1971), where all the conditions of the original position are grouped under different determinations: the circumstances of justice, the veil of ignorance, the rationality of the contracting parties and the formal constraints to the concept of justice. Following the ITS paradigm, the distinction3 between subjective circumstances and objective circumstances can be respectively translated in the form of: x Idealising assumptions: they contain and delineate characters of modelled objects, or rather not really existing. x Realistic assumptions: they contain and delineate some elements belonging to the field of reality. Rawls resorts to a series of idealising assumptions to describe the ownerships of the individuals involved in the deliberative procedure, putting in place a modelling of the decisional subject that is not interested in historical/real determinations of the subject himself. Although Rawls seems to neglect the importance of realistic assumptions in the theorisation of the original position, their presence, instead, turns out to be functional to the construction of his theoretical model. The first step towards the reconstruction of the original position can, therefore, coincide with the identification of the realistic assumptions (hereafter RA) used by Rawls (1999, 109-110): RA.1: the individual X lives at the same time with other individuals in a defined territory. RA.2: the individual X is united with the other individuals by the same physical and mental ability, so that he is not able to enact any attempt at domination over them. RA.3: the individual X is able to prove hostile towards another individual, possibly asking other individuals for support. RA.4: the individual X has a specific plan of life. RA.5: the individual X seeks fundamental social goods. RA.6: the individual X coexists with other individuals in an environment where the demand of the primary social goods overcomes the offer.

RA.1 is realistic because it expresses a judgment of existence: it is assumed that a plurality of individuals coexists in a determined geographical area. RA.2 is realistic, because it possesses the form of an existential assumption based on the attestation of a relation of similarity among the individuals, excluding the realization of expressions of mutual dominance. This result, therefore, is hardly disputable from a realistic point of view: an

96

Practical Reasoning Between Abstraction and Idealisation

individual endowed with his own physical and mental strengths could not realistically dominate all the other individuals present in society. RA.3 is realistic because it expresses a purely human dimension: fear connected with the perception of hostility from a group of individuals towards a single individual, and the need to seek help in others. RA.4 is strongly realistic because it refers to the evident difference of perspectives, purposes and needs characterizing the individuals inside a society. RA.5 mirrors undoubtedly the highest form of realism. It appears evident, in fact, that individuals of every society, regardless of their personal life plan, seek the acquisition of primary social goods such as rights, freedom, opportunity, wealth and income, with an equitable distribution. RA.6 shows a realistic nature that emerges from an observation of the ‘moderate scarcity’ of the natural resources that should satisfy the needs of individuals in a society. Continuing to abide by Przybysz’s interpretation on the basis of the ITS model, the second step towards reconstruction of the original position consists in the recognition of idealising assumptions (hereafter IA) employed by Rawls (1999, 11-125): IA.1: the individual X rigorously observes the principles he has chosen, led by a sense of justice that he has in common with the other individuals in society, and that is of public knowledge. IA.2: the individual X trusts in the observance of principles by the other individuals in society. IA.3: the individual X is rational. IA.4: the individual X does not show a form of personal interest towards the other individuals. IA.5: the individual X is concerned about his descendants for future generations. IA.6: individuals do not know their social status, the political and economic situation of society, nor the stadium of civilisation and culture reached and the natural resources available.

IA.1 is of idealising nature, because Rawls postulates that individuals are characterised by the propensity to show loyalty to the chosen principles. Nevertheless the counter-factual nature of this assumption is unequivocal: social order is constantly mined by the non-observance of principles and rules by its members. IA.2 idealises the tendency among individuals to reciprocally trust and to observe the chosen principles in accordance with the previously reached agreement. Nevertheless reality denies the content of this assumption,

Francesca Vitale

97

because a society often fails to hold a sense of trust among individuals. Consequently, all apparently stable agreements may face a breakdown. IA.3 has an idealising nucleus that is more complex to analyse. Rawls maintains that the rationality of individuals in the original position refers to the instrumental and economic reason, or rather to the ability to recover the fittest means for the attainment of purposes and satisfaction of principal needs. Nevertheless, the principle of rationality described by Rawls would idealise the individual involved in rational deliberation. His conception of rational person implies that the arrangements of preferences for all the individuals implicated in the choice are the same, since they originate from, and are determined by, a same problem of choice. Indeed, a series of empirical results, obtained while studying the so-called cognitive economy, shows that individuals do not always follow a ‘coherent set of preferences’ if they are involved in a deliberative process (Motterlini and Piattelli 2005, 11-24). IA.4 possesses a high degree of idealisation. Rawls, in fact, postulates that the individuals involved in the choice of the principles of justice are not affected by feelings like envy, shame and humiliation. Envy would be, above all, ‘collectively disadvantageous’. The parts in the original position would behave in a rationally selfless manner, trying not to cause any damage and to mutually assign themselves some benefits. This procedure is not realistic, because in interpersonal relationships, in order to make a decision, the individual must also take into account other people’s positions, which often produce conflicts of interest and a range of unpleasant feelings. Mutual (well-meaning) indifference in a society is not a realistic condition. IA.5 refers to an ideal form of original interest that the individual would display towards his future descendants during the choice of principles. Besides, this assumption is antithetical to assumption IA.4, in which postulates a mutual indifference of individuals in the original position. IA.6 is based on the idealisation of the veil of ignorance, a postulation of minimisation of knowledge, which, by imposing some cognitive restrictions onto the individuals implicated in the original position, guarantees a unanimous choice of the principles of justice. The counterfactual aspect of this assumption is easy to identify: individuals do not live in absence of information related to the institutional, political, economic and cultural order of their society. Besides, as flesh and blood individuals, they direct their actions towards purposes that cannot be separated from their personal needs and from their particular characteristics as real living people.

98

Practical Reasoning Between Abstraction and Idealisation

The set of realistic and idealising assumptions RA.1-RA.6 and IA.1IA.6 emphasises Rawls’ use of idealising propositional constructions, finalised to the description of subjective and objective circumstances of justice. In light of these idealising assumptions, the individual involved in the original position displays the following characteristics: the veil of ignorance deprives the individual of any knowledge related to his social status, to the cultural, economic and political level of his society and about the characteristics of other individuals (IA.6); the assumption of existence of a set of primary social goods (RA.5 and RA.6) refers to the tendency of the individual to seek the satisfaction of his own needs through instrumental reason (IA.3) and to design his own plan of life, made of purposes and aspirations (RA.4); besides, regarding assumption (IA.4), the individual is not affected by hostile feelings towards the other members of society; so he chooses the principles of justice in a condition of absolute indifference towards other people’s preferences, while considering the well being of future generations (IA.5); after the principles are chosen, the choosing individual and the other members of society observe them coherently and rigorously (IA.1-IA.2). The choice of the two principles of justice represents the heart of Rawls’ theory. The method of idealisation allows Rawls to build an original position in which the procedure of principles choice happens on the basis of idealising assumptions that impose some restrictions on the individuals involved in the deliberation. The Theoretical Nucleus (hereafter TN) can be thus formulated: (TN) The individual X, implicated in the situation postulated by the assumptions RA.1-RA.6 and IA.1-IA.6, chooses one of the following principles of justice: a. ‘each person is to have an equal right to the most extensive scheme of equal basic liberties compatible with a similar scheme of liberties for others’ (Rawls, 1999, 53). b. ‘social and economic inequalities are to be arranged so that they are both (a) reasonably expected to be to everyone’s advantage, and (b) attached to positions and offices open to all’ (Rawls, 1999, 53). After defining the ideal context, therefore, the choice of principles follows coherently. Liberty, as the first and fundamental principle of justice, has priority over the second principle of justice that pertains to the equitable distribution of wealth and the equal opportunity of access to public positions. It is important to underline that both principles of justice are guaranteed by idealising assumptions: assumption IA.4, postulating that the individual is in a state of indifference towards the others, deprived

Francesca Vitale

99

of any form of envy, determines that there cannot be a situation in which someone may want more profit for himself than for the others, although according to assumption IA.3 he will not even allow his profit to suffer detractions either. Moreover, assumption IA.6 fortifies the maximin rule, because the individuals in the original position are unaware of their own future social status (they do not know if they will be among the more or the less advantaged) and they rationally choose the most equitable solution from the moral point of view. Following the maximin rule, in fact, the first principle of justice, sustaining that all people have equal right to the fundamental liberty, prevails over other possible principles that found the unequal subdivision of liberty, whatever the real role of the individual in the society is (assumption IA.6 postulates that the individual has no knowledge about it), and it guarantees a minimum satisfaction. After rebuilding the original position’s model through the idealisational constructs used by Rawls as the premises to the choice of principles of justice, it is now possible to proceed to analyse the phase that concerns the concretisation of the chosen principles, keeping in mind the logical passages of the ITS. It is also opportune to specify that the attempt to reconstruct the process of concretisation of the principles of justice, following the phases of progressive reduction of the idealising character of all the assumptions hitherto exposed, represents an interpretative proposal that meets some evident analytical difficulties (as clarified in Przybyz’s article. See Przybyz 2004, 288-289). In fact, in Rawls’ theory the problem of a gradual removal of the veil of ignorance from the individual’s characteristics seems to differ from the related demand for concretisation of the two principles of justice.4 Surely, the possibility of applying principles of justice to a society characterised by less favourable circumstances than those idealised in the original position is bound to the procedure of reflective equilibrium, as it has been discussed. In fact, the more realistic form of concretisation that Rawls seems to delineate, corresponds to the effort of making the rationally and reasonably chosen principles coincide with the individuals’ convictions, and vice versa. We can now turn to the reconstruction of the procedure of concretisation. The progressive approach from the ideal individual of the original position to the real individual in a society must happen together with a gradual concretisation of the principles of justice in constitutional and subsequently legislative acts. To ensure such a conception of justice (delineated in the picture of the original position) acquires more and more realistic features, it is necessary to start from the verification that the order of a really existing society sets its members in front of a more extensive

100

Practical Reasoning Between Abstraction and Idealisation

knowledge of the circumstances than idealised by Rawls, who described the concretisation of the principles of justice as follows: Thus I suppose that after the parties have adopted the principles of justice in the original position, they move to a constitutional convention. Here they are to decide upon the justice of political forms and choose a constitution: [...] the persons in the convention have, of course, no information about particular individuals; [...] But, in addition to an understanding of the principles of social theory, they now know the relevant general facts about their society, that is, its natural circumstances and resources, its level of economic advance and political culture, and so on (Rawls 1999, 172-173).

The procedure of the concretisation of the principles of justice is based on the elimination of the veil of ignorance postulated in the idealising assumption IA.6, although Rawls underlines the demand to preserve a good part of the impartiality requisite. In fact, he still maintains the idealising assumptions IA.1, IA.5, and part of the assumption IA.6, so that the individual does not completely lose those characteristics of faithfulness to the principles he chose during the original agreement: rationality, lack of any form of envy, responsibility and lack of knowledge about his own and other people’s peculiar affairs and preferences. Rawls’ purpose is to elaborate a scheme that contains the phases of application of the principles of justice, without reducing the idealising assumptions to realistic ones excessively simplified by the political order, or to the individualistic ones typical of the economic theory. The best scheme that a society can achieve is that of an “imperfect procedural justice” (Rawls 1999, 173). The constitutional level is the field of concretisation of the first principle of justice, denominated as the level of equal liberty. The procedure of concretisation will lead to the second formulation about the Theoretical Nucleus (hereafter TN’) (Rawls 1999, 220): (TN’) The individual X implicated in the situation postulated by the assumptions RA.1-RA.6 and IA.1-IA.6 chooses two principles of justice: a. ‘Each person is to have an equal right to the most extensive total system of equal basic liberties compatible with a similar system of liberty for all.’ Priority rule: Liberty can be restricted in two cases: 1. ‘a less extensive liberty must strengthen the total system of liberty shared by all.’ 2. ‘a less than equal liberty must be acceptable to those citizens with the lesser liberty.’

Francesca Vitale

101

b. ‘social and economic inequalities are to be arranged so that they are both (a) reasonably expected to be to everyone’s advantage, and (b) attached to positions and offices open to all.’

The partial removal of the veil of ignorance has determined the concretisation of the principle of “basic liberty” chosen in the original position within a new principle that guarantees a “system of basic liberties”. The passage from the ideal plan of liberty to the real one is marked by a gradual specification of the various typologies of liberty of which the members of a society can benefit under specific limitations. The legislative level, instead, is the area of application of the second principle of justice, which affirms the equitable distribution of the income and the equal opportunities of access to public positions. The procedure of concretisation will lead to the following formulation: (TN’’) The individual X implicated in the situation postulated by the assumptions RA.1-RA.6 and IA.1-IA.6 chooses two principles of justice: a. ‘Each person is to have an equal right to the most extensive total system of equal basic liberties compatible with a similar system of liberty for all’ (Rawls 1999, 220). Priority rule: Liberty can be restricted in two cases: 1. ‘a less extensive liberty must strengthen the total system of liberty shared by all.’ 2. ‘a less than equal liberty must be acceptable to those citizens with the lesser liberty’ (Rawls 1999, 220). b. Social and economic inequalities must serve: 1. ‘for the greatest benefit of the least advantaged, consistent with the just savings principle, and 2. attached to offices and positions open to all under conditions of fair equality of opportunity’ (Rawls 1999, 266-267). Second priority rule: Two restrictions exist: 1. ‘an inequality of opportunity must enhance the opportunities of those with the lesser opportunity; 2. an excessive rate of saving must on balance mitigate the burden of those bearing this hardship’ (Rawls 1999, 266-267).

From this formula we can infer that individuals, at this stage of social reality, require a complete knowledge of the economic structure of society in order to establish some laws. Similarly to the concretisation of the first principle, part of idealising assumption IA.6 is removed with the purpose of allowing citizens to have a general overview of the economic subdivisions of the social structure they belong to. The problem of

102

Practical Reasoning Between Abstraction and Idealisation

economic inequality, ideally postulated in the original position, finds a solution in the concretisation of the activity of distribution of inequality in order to benefit the less privileged. The idealising assumption of open access to public positions and functions is concretised in a conferment of equal opportunities in the access to public offices. Finally, the total elimination of the veil of ignorance will allow individuals to pass from the legislative stage to the civil one, as they come to know the contingent factors that would have made any agreement about the principles of justice impossible in the original position. The reconstruction of Rawls’ theory of justice proposed here, in the light of the theoretical gains obtained by the School of PoznaĔ, manages to underline the central role played by idealisational methodology inside the Rawlsian theoretical model. Rawls uses idealising constructions in the theorisation of the original position with the purpose of offering a vision of social relationships in their initial structure, made up of less complex conditions and characteristics than those of a real society. In general, the use of idealising procedures of construction is indeed widespread in contractualist theories, such as utilitarianism and liberalism. Summing up, through the interpretative lenses of the ITS Rawls’ proceduralism seems to be able to withstand, at least partly, the accusations of being an “excessively idealised” theory, “inapplicable” to the real world. We have found, in fact, that the progressive reduction of idealising aspects of theoretical assumptions formulated by Rawls leads, through the procedure of concretisation, to a more realistic configuration of the subjective and objective circumstances of justice. Nevertheless, the possibility and modality of application of the principles of justice remain, in A Theory of Justice (1971), a rather obscure field of investigation, given the intricate elaboration offered by Rawls himself.

3. Onora O’Neill’s Modal Constructivism: Practical Reasoning and Abstraction Towards Justice and Virtue (1996), the most complex among O’Neill’s works in theoretical terms, introduces the following subtitling: A Constructive Account of Practical Reasoning. In 1989, the author published an ample anthology of essays on Kantian ethics under the title Constructions of Reason. Explorations of Kant’s Practical Philosophy (1989). It is evident that the nucleus of O’Neill’s reflection intends to elaborate a conception of practical reasoning within a constructivist context.

Francesca Vitale

103

In the light of the above-mentioned reconstruction of the idealisational methodology used by Rawls in A Theory of Justice (1971), the comparison with O’Neill’s theory is almost necessary. The words of the author follow: Those universalists who try to orient the practical reasoning by assuming idealized conceptions of persons, reason and action, or transcendent moral ideals or real moral properties, do not test their arguments sufficiently against the limitations of the human condition and the constraints of human knowledge (O’Neill 1996).

From this consideration it is clear that O’Neill’s position is in stark contrast with the procedural model introduced by Rawls with regards to the choice of the principles of justice. We have seen, in fact, that the theorisation of the original position resorts to idealised assumptions and constructions in order to describe the individuals involved in the deliberative procedure, and the circumstances external to them. Although initially the two compared authors agree not to formulate their thesis on the basis of metaphysical or realistic arguments, they differ in their articulation of practical reasoning. O’Neill claims that Rawls’ constructivism is ‘unsatisfactory’. The author, in fact, introduces the lines of a “more guarded constructivism” (O’Neill 1996, 48), which is far from Rawlsian constructivism in two decisive points: 1. The description of the agent and the conditions of the action is based on the method of abstraction, not of idealisation. 2. The articulation and the justification of her conception of practical reasoning do not appeal to ideals or unjustified particular factors. According to O’Neill, the notions of society and person underlying the conceptualisation of the initial situation in terms of original position are the consequences of the idealising method used by Rawls. In fact, the ideal of a free and equal moral person that follows a scheme of cooperation in a democratic society, based on a constitutional structure, faces the risk of not including all those individuals belonging to a society whose institutional order is different. The purpose of the author is to develop a conception of practical reasoning that has as its paradigm the Kantian idea of cosmopolitan reason, able to direct the construction of moral principles towards universalistic issues, not confined within the limits dictated by a narrow social order. The whole set of these reflections leads O’Neill to consider Rawls closer to Rousseau’s contractualist tradition rather than Kantian cosmopolitanism (O’Neill 2003, 324).

104

Practical Reasoning Between Abstraction and Idealisation

After specifying theoretical points in which O’Neill’s constructivism differs from Rawls’, I will focus on one in particular: O’Neill’s use of the method of abstraction in opposition to the idealisational method applied by Rawls in A Theory of Justice (1971). First, however, I will offer a general overview of what O’Neill means by “practical reasoning:” “practical reasoning is reasoning which is used to guide action, and is contrasted with theoretical reasoning, which is used to guide thinking” (O’Neill 1998, 613). She distinguishes between four typologies of practical reasoning: 1. Reasoning directed at ends in which reason has an instrumental value (end-oriented reasoning). 2. Reasoning directed at ends in which reason individualises ends considered objective (end-oriented reasoning). 3. Reasoning directed at actions in which reason appeals to social norms (act-oriented reasoning). 4. Reasoning directed at actions in which reason appeals to universally valid principles (act-oriented reasoning). The first typology of practical reasoning is typical of the ethical theories that consider reason incapable of determining the ends of morality, since its “only task is to show how the pursuit of the passions, of subjective ends is to be organized effectively and efficiently” (O’Neill 1996, 50). The central aspect of instrumental reason is its specificity in finding the most adequate means for the attainment of particular ends, personal conceptions of good, subjective preferences and individual desires. The instrumental role that Hume assigns to reason, in opposition to the motivational strength of passions, is paradigmatic (See Hume 1751). The conception of practical reasoning described in the second grouping belongs to the so-called ‘Platonist’ perspectives, a vast set of theories (such as perfectionism and moral realism) individualising ends that are held objective on the basis of “real moral properties or metaphysically grounded moral ideals” (O’Neill 1996, 50). The third typology of practical reasoning is typical of “Particularist” positions, according to which actions possess a rational orientation if they conform to “actual norms and commitments” in a circumscribed and interwoven context of traditions. On completion of the typological subdivision of practical reasoning O’Neill introduces the Kantian theories that believe action to be rational if it is “informed by principles that all in the relevant domain can follow” (O’Neill 1996, 50).

Francesca Vitale

105

O’Neill considers only the last form of practical reasoning valid for construction of universal moral principles, because the other three typologies suffer from arbitrariness, incomprehensibility and conditioned rationality (O’Neill 1996, 51). For a practical reasoning to be considered adequate to the construction of universal moral principles, it must satisfy what O’Neill defines as the “requisite of followability”, or rather a reasoning oriented to action that is sharable by all firstly in thought. O’Neill specifies: For a principle to be followable by others is a matter of its being followable in thought. Here the demand is for intelligibility […] Practical reasoning cannot aim only at intelligibility, which is the proper object of theoretical understanding. It must also aim to recommend or prescribe action, to warn against or proscribe action: it must be action guiding (O’Neill 1996, 57).

I believe that the central idea of this definition lies in the articulation of two fundamental steps within practical reasoning: shareability both from the point of view of action and from an intellectual point of view. Every “stretch” of reasoning is sharable in action if people can act on the basis of reasoning itself. Likewise, every trait of reasoning is sharable in thought if, as O’Neill often specifies, it turns out to be intelligible and acceptable, accessible and convincing, or if it can coherently be adopted by others (O’Neill 1996, 57). It is important to highlight that, O’Neill’s assertion that every form of reasoning must satisfy the requisite of followability in thought, while only the first-order practical reasoning requires both followability in thought and followability in action. Nevertheless, the requisite of followability in thought is, according to the author, crucial within constructivist theory, because it imposes some constraints on reasoning itself. O’Neill formulates the expression “doubly modal structure of practical reasoning” to underline this strong demand for complementarity between the sphere of intelligibility and that of action (O’Neill 1996, 58). In brief, O’Neill’s conception of practical reasoning consists of two modal aspects that are the specification of the notion of followability: 1. Intelligibility: practical reasoning must be coherently accessible to all and coherently acceptable by all. The relevant criteria of accessibility and acceptability are set by reason itself on the basis of the constraints imposed by the Categorical Imperative, whose formality forbids choosing maxims of action that others cannot accept, that is, follow.

106

Practical Reasoning Between Abstraction and Idealisation

2. Applicability: practical reasoning must guide action. Its practicability concerns the possibility that universally accepted principles translate into particular actions, because the universality of principles does not affect the extension of the action that is the realisation of the uniformity content of the principles, although this realisation is not uniform in turn. Principles do not determine action; they rather underdetermine them by the virtue of their universality. To complete the description of aspects characterising practical reasoning, it is useful to underline that the principles of coherence and adoptability, emerged within the requisite of followability, refer to the plan of inter-subjective justification. It is easy to notice in the author’s considerations a frequent reference to the first formulation of the Kantian Categorical Imperative, given in the Foundation of Metaphysics of Morals (1785): Practical reasoning is formulated from the reasoner’s point of view. In offering reasons for action agents endorse principles that are thought by the agent to be possible principles for all others in some domain. […] practical reasoners need therefore judge only that the principle itself is followable by all in that sphere, in the sense that its universal adoption in the relevant domain would not be incoherent (O’Neill 1996, 58-59).

If a moral theory intends to build, through practical reasoning, principles that are “followable” by all, then it must necessarily adhere to forms of inter-subjective justification, in order to avoid relativist or particularist drifts. O’Neill, in fact, often connects the requisite of followability to that of publicity (See O’Neill 1996, 58-59; 2003, 358; 1997). These are, therefore, the characteristics that, according to O’Neill, practical reasoning must possess in order to determine principles guiding moral actions. The heart of O’Neill’s criticism towards Rawls concerns the use of the method of idealisation to introduce the well-known “expository device” of the original position. According to O’Neill, Rawls’ process of construction related to the choice of the principles of justice has always suffered the accusation of being excessively abstract, unable to reflect the real conditions in which the members of a society live. Nevertheless, in the author’s opinion, abstractness should not be the true target of criticism towards the hypothetical theorisation of the original position, but idealisation. Indeed, O’Neill thinks that, in the elaboration of a moral theory, the method of abstraction is in itself “unavoidable and innocuous” (O’Neill 1989, 209). O’Neill’s clarifications follow:

Francesca Vitale

107

Abstraction, taken strictly, is simply a matter of detaching certain claims from others. Abstract reasoning hinges nothing on the satisfaction or nonsatisfaction of predicates from which it abstracts. [...] All uses of language must abstract more or less: The most detailed describing cannot abolish the indeterminacy of language. There is no general reason to object to an account of justice that argues from abstract premises to abstract principles (O’Neill 1989, 208).

Moreover, according to O’Neill, if theoretical moral elaborations did not resort to the method of abstraction, they would risk falling into arbitrariness of empirical-contextualist perspectives. Reconstructing the author’s judgement, we can claim that the idealisation employed by Rawls does not correspond to a pure one, regardless of the individual characters of the people, or of some aspects of reality: it is, instead, an unjustified procedure of selection of certain elements of reality, subsequently typified and conceptually exasperated. Idealisation, in other terms, ends up conferring to the subject characteristics that he does not empirically possess and that are destined to take form only contingently and in postulatory way. Abstraction, on the contrary, consists of the simple exercise of “bracketing”, without denying, some predicates of the object under discussion (O’Neill 1996, 40). According to the author, in order for reasoning to be defined practical, it is necessary that its premises are able to satisfy the requirement of applicability. A reasoning whose assumptions ignore this requirement (and consequently the reality in its totality) is to be defined utopian rather than practical. To clarify as much as possible what O’Neill means by abstraction in reasoning, and what by idealisation, let us consider an example. An abstract sentence could be the following: “a book is a series of consecutive printed or handwritten sheets, related to each other and provided with cover.” This proposition is abstract because the definition contained in it provides “general” descriptive elements of the object in hand. No reference is made to the genre of the book, to its plot, to smoothness or slowness, or the emotional content it could have for the reader. The description that emerges from this proposition may hold true for every book in every time and every place. This is the task of abstractive reasoning. Compare the sentence just described with the following one: “a book is a series of consecutive printed or handwritten sheets, related to each other and provided with a cover, which tells gripping stories and exciting characters that can pleasantly engage the reader.” This type of sentence cannot be considered abstract, because the descriptive information it contains may not apply generally to every book. The proposition is rather idealised, because it contains what O’Neill calls “enhanced predicates”. In fact, it cannot be said that all books leave the

108

Practical Reasoning Between Abstraction and Idealisation

reader with a feeling of pleasure, or that their characters and stories are all necessarily gripping and exciting. O’Neill’s opinion is now clear: Idealization […] can easily lead to falsehood. An assumption and derivatively a theory, idealizes when it ascribes predicates—often seen as enhanced, ‘ideal’ predicates—that are false of the case in hand, and so denies predicates that are true of that case (O’Neill 1996, 41).

O’Neill underlines that it is dangerous if, in the field of natural sciences, the method of idealisation conceptually increases theoretical reasoning within practical reasoning. She, in fact, stresses that: Practical reasoning seeks to guide action whose aim is to fit the world (in some measure) to its recommendations, rather than its conclusions to the world. If the world is to be adapted to fit the conclusions of practical reasoning, and these assume certain idealizations, the world rather than the reasoning may be judged at fault. More concretely, agents and institutions who fail to measure up to supposed ideals may be blamed for the misfit (O’Neill 1996, 42).

O’Neill’s consideration sheds further light on her critique of the method adopted by Rawls to build practical reasoning: the idealising assumptions and constructions that make up Rawls’ theoretical model is composed produce idealising conclusions that pretend to provide just as idealising principles, whose principal purpose is, yet, their application to the real world. There is an unquestionable theoretical and practical difficulty in this auspicious passage from the ideal to the real. We have broadly analysed the points of Rawls’ theory that undergo the most the effects of his idealisational methodology. It is sufficient to point out that, for O’Neill too, the principal aporias of Rawls’ idealising assumptions concern the invention of the veil of ignorance, the ideal of mutual independence of individuals in the original position, the conception of a free and equal person as the moral ideal of constitution-based democratic societies. According to O’Neill, ethical reasoning must not found its premises neither on idealised constructions, nor metaphysical arguments, nor on substantive conceptions of the good. Instead, must: Fall back on modal questions about the possible choices of abstract agents, and construct an answer to the question ‘what principles must a plurality of abstractly characterized agents reject?’ Finally, rejection of nonuniversalizable principles can guide action by requiring that we ensure that the agents actually affected, with their particular identities and

Francesca Vitale

109

vulnerabilities, can genuinely choose or refuse those principles (O’Neill 1989, 218).

Therefore, building practical reasoning means: avoiding every form of argument based on the idealisation of the agent’s aspects and the conditions of choice to which it should and could adapt; using the method of abstraction in order to prescind only from inclinations, sensitive desires and particular abilities of the agents; directing theoretical-argumentative structure towards action; describing the elements of action in an intelligible way for the people who are prepared to reflect on them; leading to universal principles, which, abstracting from numerous predicates, do not need to prescribe a uniform action, but are able to prescribe a diversified action through the underdetermination of action itself; setting in evidence that underdetermination of action does not coincide with a notdetermination of action, taken generally, because ethical principles “prescribe constraints rather than regiment action, they recommend types of action, policy and attitude rather than providing detailed instructions for living” (O’Neill 2001, 18; 1996, 89). O’Neill points out that practical reasoning, if built in conformity to the set of requirements listed so far, can vindicate the objective validity of the method of abstraction and the authoritativeness of the universal practical principles built through the reasoning itself: Those who deny that general principles are important for moral deliberation have mostly misunderstood how principles work. Principles do not give us algorithms for living. They identify broad requirements we must live up to, but they do not actually tell us what to do. We are left instead to craft responses that honor our general commitments using the materials of the case at hand (O’Neill 1997, 15).

It is necessary, now, to analyse how the abstractive process is articulated and how the construction of practical reasoning occurs in O’Neill’s thought. O’Neill’s constructivism is usually defined modal, because the process of construction of practical principles is bound to some essential conditions imposed by reason itself, so as to establish the modality by which practical reasoning must be articulated formally, without falling in indeterminateness. Such binding conditions are dictated from the first formulation of Kant’s Categorical Imperative, which the author adopts as a [w]ay to tasting maxims for their conformity to the basic requirements of a possible community of beings who are and remain capable of action,

110

Practical Reasoning Between Abstraction and Idealisation despite the vulnerability to one another’s action of their capacities to act (O’Neill 1996, 142).

The central idea, according to O’Neill, is that normative authority of universal practical principles must not be inferred from supposed grounds external to reasoning, but from the principle of universal adoptability drawn from the Categorical Imperative. O’Neill inherits the ‘metaphor of construction’ from the Kantian pages of the Transcendental Doctrine of Method: If we look upon the sum of all knowledge of pure speculative reason as a building for which we have at least the idea within ourselves, it can be said that in the ‘Transcendental Doctrine of Elements’ we have made an estimate of the materials, and have determined for what sort, height and strength of building they will suffice. Indeed, it turned out that although we had in mind a tower that would reach the heavens, yet the stock of materials was only enough for a dwelling-house—just roomy enough for our tasks of the plain of experience and just high enough for us to look across the plain. The bold undertaking had come to nothing for lack of materials, quite apart from the babel of tongues that unavoidably set workers against one another about the plan and scattered them across the earth, each to build separately following his own design. Our problem is not just to do with materials, but even more to do with the plan. Since we have been warned not to risk everything on a favorite but senseless project, which could perhaps exceed our whole means, yet cannot well refrain from building a secure home, we have to plan our building with the supplies we have been given and also to suit our needs.5

Among the main objections to constructivism, one of the sharpest is directed at the metaphor I just exposed: procedures of construction demand certain starting materials on which to build the respective ethicaltheoretical structures; consequently, there is an inevitable vicious circle, caused by the choice not to assume any realist conception at the basis of practical reasoning. O’Neill’s philosophical project, in fact, does not refrain from using some “basic materials” for her own construction. Nevertheless, accusations of circularity can be rejected if some opportune clarifications are made on both the nature of “materials” at stake and the methodology adopted for their composition. The materials she uses are minimal presuppositions at the basis of the subject agent’s experience and the methodology for their formulation is abstraction. The author replaces Rawls’ idealising method with the abstractive one in order to individuate the starting conditions of the action in its general aspects, without idealising them. These basic conditions necessarily concern all the agents

Francesca Vitale

111

that are implicated in any action. First of all, for practical reasoning to be built, it must satisfy the requisite of accessibility for all the agents involved in reflection, and that it follows the negatively modal principle, according to which the principles that others could not follow are not to be considered reasonable (O’Neill 2000, 25). Secondly, the method of abstraction allows the delineation of the basic assumptions on which the development of practical reasoning depends. These minimal assumptions represent the starting materials that serve the purpose of constructing practical reasoning. The author, in fact, tries to individuate the set of conscious and unconscious assumptions upon which every activity is based. These abstract presuppositions have the function of unavoidable conditions of human action in its general aspects, because agents necessarily assume them when undertaking any activity. The basic assumptions formulated by O’Neill claim that (O’Neill 1996, 101): 1. there are others (considered as separate from the agent); 2. those others are nevertheless connected to the agent (either or both can act on the other); 3. those others have limited but determinate powers. Those assumptions refer respectively to plurality, connection and finitude. It must be remembered that: the first one has a higher priority than the other two because, without the recognition of the presence of the other in reality, the problem of inter-subjective justification of moral principles would not occur, even before the ethical problem itself; the assumption of connection derives directly from the first one and it expresses existing relationships among different subjects within the dynamic of action; the third and final assumption refers to the idea of plurality as a diversified totality of agents, each having a limited and circumscribed power, so as not to overpower other people. Every activity presupposes these three assumptions. Firstly, none of us could do any action without recognising the existence of others as subjects that are also capable of acting and reacting. This acknowledgement can take place in a conscious or unconscious way. Secondly, admitting a plurality of others in carrying out an action means to recognize that there is a form of interaction between us and the plurality of potential subjects agents. Their sphere of action comes in connection with ours whenever their activity involves and implies us, and vice versa. The assumption of connection requires that we consider the agency of others as strictly related to our own. Thirdly, our activity presupposes finitude in terms of vulnerability: the others, as sources of initiative and action are subjects having an

112

Practical Reasoning Between Abstraction and Idealisation

undefined set of capabilities. At the same time, though, as potential direct or indirect recipients of our activity, they are vulnerably exposed to the effects of our actions. The last assumption, therefore, offers a reason for morally considering others, setting certain general prudential constraints on our activity. Finally, those three same core assumptions represent the conditions on which any activity takes place. What is clear from these considerations is that, according to O’Neill, the consequences of human action are not totally predictable, but the starting assumptions of plurality, connection and finitude are helpful for “fixing the appropriate scope of ethical consideration.” (O’Neill 1996, 100-1). She specifies that the basic presuppositions do not contain “states of consciousness” and that every meaning in them refers exclusively to abstract aspects of an action, whether the agents are conscious of the given action or not. So, O’Neill firstly pursues a project that aims at constructing a set of “regulatory principles”, which contains conceptions of the world and of the people that live in it: these principles must proceed from basic assumptions about a plurality of others, the societies they belong to, the general capacities, capabilities and vulnerabilities that they hold. Following the author, therefore, these abstractive assumptions are not objects of construction but starting materials upon which the procedure of construction of moral principles can develop. It is also important to note that the method of abstraction used in Towards Justice and Virtue (1996) in order to formulate the minimal assumptions of activity is the same that O’Neill applies in Constructions of Reason (1989) to outline the circumstances of justice. O’Neill, like Rawls, offers a view of the circumstances of justice entailing a conception of society and people but, unlike Rawls, she avoids idealising premises: Constructivism might begin simply by abstracting from the circumstances of justice, meagerly construed. The problem of justice arises only for a plurality of at least potentially interacting agents. It does not arise where there is no plurality, or no genuine plurality of agents, and hence no potential for conflict. […] Some assumptions about the agents of such a plurality are needed if any construction is to be possible. If idealization is to be avoided, these assumptions must only be abstract. […] A more Kantian constructivism must then start from the least determinate conceptions both of the rationality and of the mutual independence of agents (O’Neill 1989, 212).

O’Neill proposes a practical reasoning directed by a set of “bare” concepts of society and people in order to construct universal principles. She puts into effect an explicit binding between the fundamental

Francesca Vitale

113

assumptions characterising the circumstances of justice and the aforementioned ones relating to plurality, connection and finitude. The basic and abstract concept of society is that of a plurality of “potentially interacting agents:” for the problem of justice to arise, a set of initial conditions of interaction is necessary, because without the possibility of interaction between people the notion of justice itself would be void of meaning. The basic and abstract concept of the person is characterised by the following aspects: .

1. “A meager and indeterminate view of rationality” (O’Neill 1989, 212): by abstracting from the particular characteristics of the agents, it is possible to achieve a conception of the person that is based on a form of rationality focusing on the most efficient means to pursue ends as well as an unspecific capacity to “understand and follow some form of social life” (O’Neill 1989, 212). 2. “A meager and indeterminate view of the identity and of the mutual independence of agents” (O’Neill 1989, 213): human beings are structurally characterised by different degrees of dependence and independence in their daily activities: in fact, O’Neill makes use of the concept of “interdependence” to explain how agents are able to establish various interdependent relationships with each other. Interdependence is to be considered a form of social relation in which each member is mutually dependent on the others, as autonomous participants of the relationship itself. The concept of interdependence is essential to any understanding of the nature of the interaction between several agents in a system of social and moral life. Therefore, we find that according to O’Neill it is impossible to elaborate a conception of people and of society that does not admit the minimal assumptions I have just outlined: firstly, it would be rather difficult, in the development of the concept of the person, if the process of abstraction could lead to assumptions that do not presuppose a generic capacity of individuals to participate in social life, to achieve their ends and to enjoy some degree of dependence and interdependence. Secondly, it would be equally difficult to develop a fundamental conception of society that fails to consider the interaction between people as a prerequisite for justice to take shape. Every conception of society must presuppose an interacting plurality in order to contribute to the formulation of the circumstances of justice, which could not be constructed if we posited a “pre-established harmony” between agents, or the impossibility of an interactive and pluralistic system as the basis of every social apparatus.

114

Practical Reasoning Between Abstraction and Idealisation

Conclusion I have shown the difference between Rawls and O’Neill in the elaboration of the basic assumptions underlying the constructive process of practical reasoning, through the description of the methodological characteristics of idealisation in opposition to those of abstraction. Such opposition has pushed O’Neill to underline the theoretical demand to establish a clear-cut distinction between the two methodological procedures at stake, because their identification brings some dangerous conceptual misunderstanding within ethical reasoning itself. I have highlighted how O’Neill attempts to create a constructivist account of ethics from a set of minimal presuppositions entailing only abstraction and not idealisation. Rawls, instead, employs an idealised procedure of construction of the two principles of justice, ruled by several assumptions about the subjective and objective conditions of the original position that are not reproducible in a real society. However, the perspective offered by the Idealizational Theory of Science helped me provide some attempts at a concretisation of Rawls’ idealised assumptions. The previous reconstruction of the Rawlsian theory of justice, on the basis of the Przybysz’s interpretation, has shown how central the contents of idealising assumptions are in order to reflect Rawls’ conception of the person and the structure of social interactions between human agents. We could summarise the different approaches of the authors considered in this paper by resorting to the wish to achieve a “more Kantian constructivism”, expressed by O’Neill in Constructions of Reason (1989). She sketches the outlines of what a procedure of construction should be: It surely cannot guide an answer to the hypothetical question ‘What principles would a plurality of agents, with minimal rationality and indeterminate capacities for independence, choose to live by?’ Materials for answering hypothetical questions are lacking. However, the blueprint may permit a beginning of an answer to a modal question such as ‘What principles can a plurality of agents of minimal rationality and indeterminate capacities for mutual independence live by?’ (O’Neill 1989, 213).

As the discussion has pointed out so far, O’Neill constructivism aims at tackling the last question: the procedure of construction must be modal in order to lead to principles that could be coherently adopted by all, since they have passed the Kantian universalizability test. Rawls’ constructivism, instead, aims at addressing the first question: the procedure of construction is a hypothetical device for determining which principles of justice all the individuals could accept under several idealised circumstances of choice.

Francesca Vitale

115

Throughout this article I have tried to focus on a first level of construction that is structured around the methodological development of basic assumptions (starting materials), from which the configuration of a second level of construction depends: the latter is about the identification of a set of practical principles capable of guiding human action. It is now clear that Rawls and O’Neill structure the first level of construction from different methodological standpoints: Rawls’ first level of construction starts with “idealisation”, while O’Neill’s first level of construction starts with “abstraction”. However, both their basic accounts of constructivism will flow into the second level of construction, where the thin universal principles formerly built should be thick enough to embrace at least in part the moral life of human beings. The passage from the first to the second level of construction is a sensitive topic that undoubtedly deserves a broad and complex analysis that I cannot take up here.

Bibliography Bagnoli, Carla. 2007. “Il costruttivismo kantiano”. In Oggettività e Morale, edited by Giorgio Bongiovanni, Milano: Bruno Mondadori. Brzechczyn, Krzysztof. 2009. Idealization XIII: Modeling in History. PoznaĔ Studies in the Philosophy of the Sciences and the Humanities, vol. 97. Amsterdam-New York: Rodopi. Coniglione, Francesco and Roberto Poli. 2004. La realtà modellata. L'approccio idealizzazionale e le sue applicazioni nelle scienze umane. Milano: Franco Angeli. Coniglione, Francesco. 2004. “Between Abstraction and Idealization: Scientific Practice and Philosophical Awareness.” In Idealization XI: Historical Studies on Abstraction, edited by F. Coniglione, Roberto Poli, and Robert Rollinger, 59-110, Atlanta-Amsterdam: Rodopi. Da Re, Antonio. 2008. Filosofia Morale. Milano: Bruno Mondadori. Dalle Fratte, Gino. 1995. Concezioni del bene e teoria della giustizia: il dibattito tra liberali e comunitari in prospettiva pedagogica, Roma: Armando Editore. Dworkin, Ronald. 1977. “The Original Position.” In Reading Rawls, edited by Norman Daniels, 16-53. New York: Basic Books. Hill, Thomas E. Jr. 1989. “Kantian Constructivism in Ethics.” Ethics, Vol. 99, 4: 752-770. Hume, David. 1751. An Enquiry concerning the Principles of Morals. London: printed for Andrew Millar. Kant, Immanuel. Kritik der reinen Vernunft. In Kants Gesammelte Schriften, (A = 1781) vol. III: 1-552, (B 1787) vol. IV: 1-252.

116

Practical Reasoning Between Abstraction and Idealisation

Maffettone, Sebastiano. 2010. Introduzione a Rawls. Laterza, Roma-Bari: Laterza. Mordacci, Roberto. 2008. Ragioni Personali. Saggio sulla normatività morale. Roma: Carocci. Motterlini, Matteo and Massimo Piattelli Palmarini. 2005. “Introduzione.” In Critica della ragione economica. Milano: Il Saggiatore. Nowak, Leszek. 1980. The Structure of Idealization. Towards to Systematic Interpretation of the Marxian Idea of Science. Dordrecht: Reidel. Nowak, Leszek and Izabella Nowakowa. 2000. Idealization X: The Richness of Idealization. Amsterdam-Atlanta: Rodopi. Nowak, Leszek. 2004. “L'approccio idealizzazionale alla scienza: una rassegna.” In La realtà modellata. L'approccio idealizzazionale e le sue applicazioni nelle scienze umane, edited by Francesco Coniglione and Roberto Poli, Milano: Franco Angeli. O’Neill, Onora. 1989. Constructions of Reason. Explorations of Kant’s Practical Philosophy. Cambridge: Cambridge University Press. —. 1996. Towards Justice and Virtue. A Constructive Account of Practical Reasoning. Cambridge: Cambridge University Press. —. 1997. “Political Liberalism and Public Reason: A Critical Notice of John Rawls.” Political Liberalism. The Philosophical Review, 106: 411-428. —. 1998. “Practical Reasoning and Ethics.” In Routledge Encyclopedia of Philosophy, vol. 7: 613-620. London-New York: Routledge. —. 2000. Bounds of Justice. Cambridge: Cambridge University Press. —. 2001. “Practical Principles and Practical Judgment.” Hastings Center Report 31, 4: 15-23. —. 2003. “Constructivism in Rawls and Kant.” In The Cambridge Companion to Rawls, edited by Samuel Freeman, 347-367, Cambridge: Cambridge University Press. Pirni, Alberto. 2005. Filosofia Pratica e Sfera Pubblica. Percorsi a confronto. Reggio Emilia: Diabasis. Przybyz, Piotr. 2004. “Idealizacja w Teorii sprawiedliwosci John Rawlsa.” In La realtà modellata. L'approccio idealizzazionale e le sue applicazioni nelle scienze umane, edited by Coniglione Francesco, and Roberto Poli, Milano: Franco Angeli. Rawls, John. 1971. A Theory of Justice. Cambridge: Harvard University Press. —. 1980. “Kantian Constructivism in Moral Theory.” Journal of Philosophy, 77: 515-572.

Francesca Vitale

117

—. 1999. A Theory of Justice: Revised Edition. Oxford: Oxford University Press. Sandel, Michael. 1982. Liberalism and the Limits of Justice. Cambridge: Cambridge University Press. Stemplowska, Zofia. 2008. “What’s Ideal About Id eal Theory?.” Social Theory and Practice, 34: 319-340. Street, Sharon. 2010. “What is Constructivism in Ethics and Metaethics?.” Philosophy Compass, 5: 363-384. Veca, Salvatore. 1998. La filosofia politica. Roma-Bari: Laterza. —. 2010. La società giusta. Milano: Il Saggiatore. Verrucci, Gianluca. 2010. Ragion pratica e normatività. Il costruttivismo kantiano di Rawls, Korsgaard e O’Neill. Milano: Mimesis. Vigna, Carmelo. 2003. Libertà, giustizia e bene in una società plurale. Milano: Vita e Pensiero.

Notes 1

Nowak starts from a basic assumption according to which there is an essential difference between abstraction and idealisation. The distinction among these two theoretical procedures lies in the fact that while abstraction is an operation through which the intellect draws the universal from the specific, idealisation proceeds from the simplifications of the aspects of phenomenal reality, which are considered secondary, with the purpose of gathering what is essential about the phenomenon itself: science works with ‘ideal’ models and objects in order to overcome the discrepancy between reality and appearance (for an introduction to this perspective see Nowak and Nowakowa 2000; Nowak 1980; Coniglione 2004). 2 The succession can be so formulated: Mk, Mk-1, …, AMi, where k, k-1, …, i point out the idealisational quantity present in statements. The model Mk-1 represents the concretisation of the precedent model Mk, while Ami indicates the model as closer to reality, including empirical-descriptive aspects (see Nowak 1980). 3 This distinction is underlined by Przybyz Piotr, who inspired the present contribution. I recognise my intellectual debt towards his theses, although I tried, within the limits imposed by logical-scientific formulas, to implement them and make them my own (see Przybyz 2004, 273-298). 4 About the necessity of the process of concretisation see Rawls, 1999, 215-21. 5 This is the translation of Kant’s famous passage offered by O’Neill herself, in Constructions of Reason, 11; See Kant. Kritik der reinen Vernunft, in Kants Gesammelte Schriften. (A = 1781) vol. III: 1-552, (B = 1787) vol. IV: 1-252.

TOWARDS A GENDERED RATIONAL CHOICE THEORY MARTINA BELMONTE

(UNIVERSITY OF EAST ANGLIA1)

Abstract Studies on gender and those on rationality have few exchanges with each other. In this paper, my aim is twofold: on the one side, to show that gender studies offer much empirical evidence to support and actualize a debate about the definition of rationality based on revealed preference; on the other side, to show that choices made out of gender identification can be tackled with a rational choice approach. To do this, I will introduce a problem that is particularly interesting both for feminists and for rational choice theorists, namely the gender epistemic value of the menu, and I will argue that it can be accounted for with a combined approach. The gendered epistemic value of the menu case pushes for extending the revealed preference account and hence for refining the definition of normative rationality for a specific class of problems in which gender membership matters for the agents. I argue that a combined approach is a promising way both to make sense of the critique coming from Sen (1993) to the internal consistency of choice as unique criterion to define rationality, and to give a more rigorous tool to the feminist debate about individual free choices that turn out to be disadvantageous when considered collectively. The paper is structured as such: in the first section I will refer to the feminist debate about rational choice and I will note that, in the individual context, rational choice theory is generally dismissed either because egotistical or because formal. In the second section I will present the tools of revealed preference and the definition of rationality as consistency. I will then present Sen’s criticism to internal consistency with various examples of violation of the independence axiom. At a first analysis, Binmore’s claim, which consists in redescribing the choice problem, is

Martina Belmonte

119

successful. In the third section, I will present gender as an interpersonal category, together with the problem of the advancement of women, both as a collective and as an individual issue. The problem is that choices that seem to flow from women’s preferences lead to undesirable outcomes. In order to describe this phenomenon, the psychological research has devised concepts like “stereotype threat” and “gender schema”, while the philosophical research refers to the notion of adaptive preference, which still lacks a coherent account. In the fourth section I will merge the previous ones, and I will present a gender choice that has the same structure as Sen’s examples. I will call this gendered epistemic value of the menu and I will argue that its meaning is remarkably different from other cases of violation of the independence axiom. The redescriptive strategy is not appropriate for the new case because, if the theorist has gender concerns, too much information is lost. Finally, I will argue that, for some kind of choices, gendered rational choice theory promises fruitful developments.

Introduction Studies concerned with gender and rationality generally run along two different paths. Some exceptions on the gender studies side are England (1993), Friedman and Diem (1993), Folbre (1993; 1994), Anderson (2001), Peter (2003), Thalos (2005), Driscoll and Krook (2008). Nevertheless, none of them shared the perspective on choice I am considering here: a philosophical approach to individual rationality. Rational choice theory is often dismissed by feminist literature because of egotistical and utilitarian (for example England 1993), and the theory itself is confused with its applications in economics. But, as Anderson (2001) suggests, it is important to draw a distinction between the formal and the rhetorical aspect of rational choice theory: The formal theory of rational choice says that people tend to maximize their utility; that is, that they adopt as their end the maximal satisfaction of their overall scheme of preferences. The theory is formal in that it concerns only the relative rankings of a person’s preferences, disregarding the contents of these preferences or the individual’s reasons for having them. It says that an individual’s preferences fit into a single, complete, transitive ordering, and that individuals choose to satisfy the top ranked preference in that ordering (Anderson 2001, 373-374).

The rhetorical theory, on the contrary, builds an agent that is “selftransparent, opportunistic, cool and calculating, autonomous and self-

120

Towards a Gendered Rational Choice Theory

confident.” Anderson argues that, whereas the formal theory could be accepted, the rhetorical theory could not, because of those egoistic assumptions. Nevertheless, she also argues that the main task of the theory, which is identified in explaining human actions, is carried out by the rhetorical theory. According to her interpretation, the concept of preference understood in a purely formal sense has little normative power. So, although her distinction between rhetorical and formal rational choice is worthy in order to reject some accusations to rational choice, nevertheless it is not enough to support its use in tackling gender issues. In the following essay, I will argue that the normative power of rational choice is given by axioms and by their ability to rule out irrational behaviour. Axioms, however, can be criticized through some cases that present intuitively reasonable choices that are yet considered as irrational by the theory. I will present a case that, while showing that the axiom of independence is inappropriate, pushes for an extension of revealed preference.

1. Revealed Preference Rational choice theorists analyse agents’ choices and seek the conditions for a choice to be rational. Let us think of a choice situation: Example 0: Ann has just finished a PhD in philosophy, and she faces two options: either to accept an offer for a post-doc position and pursue an academic career, or to accept an offer from a local firm and work in HR as a recruiter. This is an example of individual choice problem, described by an agent, Ann, and two alternatives. It is defined under certainty because the alternatives are certain: if Ann accepts an offer, she will certainly get the job. I will focus on this kind of problem because more complex problems are modelled as an extension of individual problems under certainty.

1.1 Formal Tools In the context of revealed preference, it is possible to define what is rational for Ann to do with two basic tools: choice function and preferences. A preference is a relation over alternatives, and it is defined by some axioms, depending on the choice situation. Given a finite and nonempty set of alternatives A, a choice function, C : A ĺ P (A), is a nonempty subset of alternatives: ‫ = ׎‬C (A) ‫ ك‬A. The idea behind the

Martina Belmonte

121

revealed preference approach is to directly derive the agent’s choices from the agent’s preferences (1.1.1), and, inversely, to directly derive the agent’s preferences from the agent’s choices (1.1.2). Choosing rationally is thought of as ranking all of the alternatives and picking the top one(s). 1.1.1 From Preferences to Choice I will consider strong preference, which is defined by the following set of axioms: asymmetry, completeness and transitivity. This means that the agent is idealised in the sense that she has clear preferences towards all the alternatives and she is never indifferent to them. The symbol for preference is P, and e, f, g belong to the set of alternatives A, so ePf means the alternative e is preferred to the alternative f . Preferences are asymmetric when, if ePf then fŃPe; complete when, if eŃPf then fPe; and transitive when, if ePf and fPg then ePf. Having a nonempty set of alternatives, the preferred alternative is the optimal element: e*Pe, for all e ‫ א‬A. This can be written also in the form of a function C(A)=optA={e* ‫ א‬A | e*Pe, for all e ‫ א‬A} that can be easily proved to be a non-empty set, and therefore a choice function. So, the goal of defining a choice function from a preference relation has been achieved. 1.1.2 From Choice to Preference Revealed preference is defined as the relation that is expressed by and only by the choice: xPy ļ x=C({x,y}). In order to achieve the purpose of defining a preference relation from a choice function, it has to be proven that P in the definition of revealed preference is an ordering. From the fact that x=y and that C =‫ ׎‬it can be respectively derived that P is asymmetric and complete. However, in order to prove that P is also transitive, I have to introduce the following axiom: let x, y ‫ א‬X and X ‫ ك‬Y. If x=C(X) then y=C(Y). This is called axiom of independence of irrelevant alternatives,2 and if the choice function satisfies it, then the relation P is transitive. A choice function that satisfies the axiom of independence is called consistent, and this is the only characteristic for a choice function that is necessary in order to derive an ordering. In other words, every consistent choice function can refer to asymmetric, complete and transitive preferences over alternatives.

122

Towards a Gendered Rational Choice Theory

1.1.3 Revealed Preference: “Behaviour without Anything Other than Behaviour” The slogan “behaviour without anything other than behaviour”3 is in a certain sense true as revealed preference is concerned. A distinctive feature of revealed preference is to avoid any reference to the agent’s motivation: the agent simply chooses what she prefers and prefers what she chooses. As Binmore writes, modern decision theory succeeds in accommodating the infinite variety of the human race within a single theory simply by denying itself the luxury of speculating about what is going on inside someone’s head. Instead it pays attention only to what people do (Binmore 2009, 8).

This theoretical position makes revealed preference a technical notion that should be distinguished from the notion of preference used in ordinary language. The word preference has many nuances that are deliberately not taken into account by the notion of revealed preference (Hausman 2007). For example, a preferred alternative can be thought of as an alternative chosen because the agent likes it, rather than because she has the duty to choose it, or because she holds some principles that support it. In this sense, preference is opposed to duty or commitment (Sen 1977). Or, an agent could have a preference because she expects an advantage from the alternative she prefers. So, in this sense, a preferred alternative is the most advantageous one. Preference may also be used in relation to the agent’s self interest as opposed to the agent’s altruism. Or, preference may refer to the alternative the agent would have hypothetically chosen. Revealed preference deliberately neglects these differences and considers as preferred alternative the chosen one, regardless of the agent’s motives. 1.1.4 Rational Choice With the representation theorem it is possible to represent consistent choice functions with numerical functions, called utility functions. This is possible through the properties of ordering. A utility function is a function from the set of alternatives to real numbers, u: X ĺ R, and it represents the preference relation if and only if it pairs a higher number with the preferred one among two alternatives: xPy ļ u(x) > u(y). The representation theorem states that for every ordering relation, a utility function does exist.4 Therefore, consistency is the necessary and sufficient condition for a choice function in order to derive a utility function. As it is defined here, the concept of utility is a ‘thin’ one: the utility function may

Martina Belmonte

123

have whatever argument, because the important feature is the structure of maximization. As Binmore writes, “in the theory of revealed preference, utility functions are no more than mathematical devices introduced to help in solving choice problems” (Binmore 2009, 14). The representation theorem provides an argument for the utility function only once either preferences or choices are given. Now, it is possible to define the rational choice under certainty as the choice that maximizes the agent’s utility function. This model of rationality does not entail that the agent who chooses rationally deliberately maximizes her utility function; rather it entails that the rational agent behaves as if she were maximizing the utility function that it is possible to build from her consistent behaviour. So, in the example above, on the one side, Ann’s rational choice depends on her preferences: if she prefers to work in academia, then the rational choice will be to accept the post-doc position, which will have a higher utility. On the other hand, if Ann rationally chooses the offer as a recruiter, it is possible to build a utility function that represents her preference toward the local firm which has higher utility.

1.2 Revealed Preference: Psychological Assumptions Sensibly Chosen If it is true that revealed preference does not commit itself with motivations and reasons, nevertheless it is untrue that the psychological sphere is completely banned. Since the ordering expressed by P is called revealed preference, and since the theory has normative and explanatory power, revealed preference must be interpreted in relation to the psychological sphere. In other words, if the choice of an alternative did not express a psychological attitude—positive towards the chosen alternative and negative towards the rejected alternative—the consistency of choice as condition of rationality would not have any normative or explanatory justification. Sen argues that: The whole framework of revealed preference analysis of behaviour is steeped with implicit ideas about preference and psychology. I would, therefore, argue that the claim of explaining «behaviour without reference to anything other than behaviour» is pure rhetoric, and if the theory of revealed preference makes sense it does so not because no psychological assumptions are used but because the psychological assumptions used are sensibly chosen (Sen 1973, 243-244).

124

Towards a Gendered Rational Choice Theory

The assumption is that it is possible to encompass all of the agent’s attitudes towards an alternative and to coherently express them in the form of preference. Contrasting reasons may exist, but it is possible to sensibly merge them in revealed preferences. Hausman (2007) calls this sense of preference “all-things-considered-rankings”, which are an [o]verall evaluation of the objects over which preferences are defined. This evaluation implies a ranking of these objects with respect to everything that matters to the agent: desirability, social norms, moral principles, habits—everything relevant to evaluation (Hausman 2007, 37).

All the properties of alternatives are merged together and form a single point called alternative, towards which the agent has a preference. Then the choice problem is maximally abstract, since alternatives are presented without dimensions. In the normative model of rationality that I am considering, the task of modelling the problem and the task of choosing are held as separate (Hosni 2013). Abstract problems and idealised agents differ from real problems and agents, but between them a relation of approximation exists. However, approximation is a notion that is far from being unproblematic. As Kreps puts it: No one that I know of would seriously maintain that individuals do conform exactly to the sorts of axiomatic systems that will be studied here. [...] At best then, individual behaviour approximates the axiomatic based behaviour that we shall study. Why then does it make sense to study the behaviour of systems (economies, organizations) where the actors are presumed to satisfy exactly the axiom systems? The answer must be: because if their behaviour is approximately what is modeled, then the model will tell us something about how their behaviours interact or intertwine in the system. This, the reader will surely note, takes a somewhat large leap of faith (Kreps 1988, 6).

The relation of approximation does not have a solid theoretical foundation. What I want to challenge is the idea that a single relation of approximation—or, in other words, the same degree of abstraction—fits all kinds of decision problems. I will do this, presenting a case in which, by maximally abstracting a decision problem, we lose a relevant bit of information that we do not want to lose.

Martina Belmonte

125

1.3 A Debate on Consistency 1.3.1 Sen’s Criticism As seen above, to be consistent is the only requirement a choice function must satisfy in order to represent rational behaviour. Sen refers to this definition of rationality as the thesis of internal consistency of choice. Consistency properties, such as the independence of irrelevant alternatives, are “internal to the choice function in the sense that they require correspondence between different parts of a choice function, without invoking anything outside choice (such as motivation, objectives, and substantive principles)” (Sen 1993, 495). According to Sen (1973; 1977; 1993; 1995), there is something deeply misleading in this conception of rationality, even something false, because there is no completely internal way of determining whether behaviour is consistent. It follows that the axiom of independence of irrelevant alternatives is not apt to ensure the rationality of choice. Sen’s strategy to prove his argument is to present some real choice situations in which it seems intuitively rational to violate the axiom of independence. I can identify three groups of examples that have been presented by Sen and others. Example 1 (Epistemic value of the menu) A man meets an acquaintance who proposes to have tea together. The agent then refrains from going directly home and kindly accepts. But, when the acquaintance widens the options and offers him cocaine, the agent regrets his previous choice and decides to go directly home (Sen 1993). Example 2 (External norm) A person sitting at a dinner table has to choose a piece of fruit from a basket. Unfortunately the basket is very poor and contains only one apple. The agent decides then to behave decently, and takes nothing. But, as soon as another apple is added to the basket, he immediately picks up the old apple and eats it with relish (Sen 1993). Example 3 (Reason based) A European student has to decide where to go for his Master’s. He has been offered a place at Princeton University and at the London School of Economics. Since he wants to go to America, he chooses Princeton. But, as he realizes that a place at Chicago University has become available too, after pondering for a while, he eventually opts for remaining in Europe, and going to London (Rubinstein 2006).

Sen argues that whereas two statements can be inconsistent to each other, two acts cannot. Only by having elements external to the choice that allow us to understand what agents are trying to do, it is possible to refer acts to statements, and to consider them as inconsistent. The agent’s

126

Towards a Gendered Rational Choice Theory

choices are intelligible only inside a context-the menu-that influences the alternatives’ meaning. Sen therefore extends his criticism to the revealed preference approach that [e]ssentially underestimates the fact that man is a social animal and his choices are not rigidly bound to his own preferences only. [...] An act of choice for this social animal is, in a fundamental sense, always a social act (Sen 1973, 253).

The formal theory in itself, and particularly the formal notion of revealed preference, is empty and its explanatory power derives only from the substantive interpretation of the choice situation. 1.3.2 Binmore’s Reply According to Binmore, Sen’s criticism relies on a misunderstanding of the normative model. For instance, considering the example of epistemic value of the menu, in Sen’s account the alternatives are ‘tea with the acquaintance’ (t), ‘to go home’ (h) and ‘snorting cocaine’ (c). The alleged inconsistency derives from employing the following choice function: C({t,h}) = t and C({t,h,c})=h. This violates the axiom of independence of irrelevant alternatives and, according to the definition of rationality given above, the choice is irrational. But, as Binmore claims, the reason that she violates the Independence of Irrelevant Alternatives without our finding her behaviour unreasonable is that snorting cocaine isn’t an irrelevant alternative for her. The fact that cocaine is on the menu changes her beliefs about the kind of person she is likely to meet if she accepts the invitation. If we want to apply the theory of revealed preference to her behaviour, we must therefore find a way to formulate her decision problem in which no such hidden relationships link the actions available in her feasible set A with either her beliefs concerning the states in the set B or the consequences in the set C (Binmore 2009, 11).

An appropriate way to model the decision problem would, for example, distinguish between the alternatives of having tea with a cocaine addict and having tea with a non-addict person, whatever the epistemological status of the decision maker is. Therefore, Sen’s criticism can be resolved in seeking adequate descriptions for the alternatives of the choice problem. Once these have been established, the alleged violation of the independence axiom disappears. Binmore’s argument seems convincing, and if Sen’s claim refers only

Martina Belmonte

127

to the interpretation of the choice problem it fails to undermine the revealed preference approach and rationality as consistency.

2. Gender Issues Whereas in the previous section I referred to abstract problems, now I want to introduce an empirical category, i.e. gender, and gendered phenomena.

2.1 Gender, Among Other Categories In the context of the so-called third way feminism, gender is considered an analytical category besides others, like race, class, sexuality, etc. These are social categories linked, though in a non-deterministic way, to a material basis, which, in the case of gender, is the traditional sexual division between male and female. The reason for using the category of gender to analyse reality is that reality itself is structured along the gender line: children are socialized as females or males, women and men have different rights and duties (think of retirement or maternity leave), commodities are advertised targeting men or women, and so on. However, gender is not a homogeneous category, because other factors like race, class, nationality, religion, disability, etc. may play a role. So, for instance, the distance between a middle class man and a middle class woman may be shorter than the distance between a middle class woman and a working class woman (Spelman 1988; Damaske 2011). Therefore, employing the category of gender to draw general conclusions seems fallacious and, as a result, the partiality of the gender-based knowledge feeds gender scepticism (Bordo 1993). Indeed, once the only category of gender turns out to be insufficient as the unique parameter to analyse women’s (and men’s) situations, the problem is to set the right number of parameters to be used. The advantage of analysing the situations of a single person in gendered terms is to highlight trans-individual dynamics that involve men and women as such, rather than individuals, and to draw general considerations for men and women, rather than for individuals. It is clear that for complex phenomena, the more parameters are added, the more the analysis is accurate; nevertheless the transindividual space identified by the gender line, the race line, the religion line, the sexuality line, and so on, is increasingly narrowed. Therefore, the description of a phenomenon becomes so specific that general considerations are unlikely to be drawn and the advantage of using transindividual categories is lost.

128

Towards a Gendered Rational Choice Theory

The only way out of this paradoxical situation seems to discard transindividual categories and to return to studying either general and neutral cases or individual cases. This conclusion, however, is undesirable for at least three reasons: all gendered processes that have been studied by feminist research would thus be disregarded; it neglects the fact that sometimes people’s actions are indeed determined by their gender; and finally it is politically weak because: in a culture that is in fact constructed by gender duality [...] one cannot be simply ‘human’. This is no more possible than we can ‘just be people’ in a racist culture. One cannot be gender-neutral in this culture (Bordo 1993, 241-2).

The challenge from gender scepticism to gender analysis is twofold: on the ontological side, the relationship between individuals and gender requires some clarification; on the epistemological side, the kind of gender-based knowledge has to be specified. I consider the relationship between individuals and gender as a relationship of membership to an open group. Young (1990) defines gender as a social group, namely as a group that exists only in relation to at least another group and that expresses social relations. It constitutes the background where individual identities take place (Young 1990, 1994). Open groups are social groups whose open boundaries are emphasized. Indeed, individuals belong simultaneously to more than one group, whose relevance depends on contextual factors. Members are tied together by a common knowledge of norms (gender norms in case of gender groups) that generates expectations about gender roles and social sanctions if expectations are not met. Groups can affect individual behaviour because people face a set of choices constrained by norms, and people can act in accordance with or against these norms. Thus open groups differ from mere aggregates of people tied together by external properties, like having blue eyes. Moreover, unlike collective agencies, open groups do not exist aside from the actual people belonging to the group: if the Italian Parliament exists as an institution above its members, gender groups depend on actual women and men. Given these features of open groups, the kind of knowledge that is drawn from their employment in social sciences is contextual: it depends on empirical elements, in the sense that the more empirical evidence supports the presence of an open group, the stronger the analysis is. Furthermore, since people do not belong exclusively to a group, then a complete analysis based on one parameter is unlikely. However, a partial analysis is useful when the researcher’s interest lies in highlighting one specific aspect—the gender one.

Martina Belmonte

129

2.2 The Advancement of Women The problem of the advancement of women is a gendered phenomenon and refers to the slower career progress of women compared to men. Research shows that women face invisible difficulties in rising to the top in an apparently egalitarian environment and these constitute the so-called glass ceiling (Valian 1998). The concentration of women in low positions in an occupational sector is called vertical occupational sex segregation (Browne 2006). A recent report from the British Philosophical Association and the Society for Women in Philosophy UK, “Women in Philosophy in the UK” (2011)5 shows the problem of the advancement of women in the philosophical field in British universities.6 Statistics show that at the undergraduate level the percentage of women and men is similar, whereas going up at further levels the percentage of women decreases at every stage, and at the professorship level women are at only 19%.

2.3 Psychological Phenomenon Since women are not forced to give up their advancement in careers, these numbers are the result of individual choices. Psychological research tries to explain such choices and here I will consider stereotype threat and gender schema. Stereotype threat is defined as “the experience of being in a situation where one faces judgments based on societal stereotype about one’s group” (Spencer, Steele and Quinn 1999, 5). Experiments focus on the effect of, and on the reaction to, societal devaluation expressed by stereotypes in ability performances, like math tests. In a study at the University of Michigan, for example, Spencer, Steele and Quinn (1999) analyse two groups of students with a mathematical test. The first group was told that generally no gender difference has come up, and the result was that women and men performed equally. The second group was told that generally a gender difference was recorded, and the result was that men outperformed women. They concluded that the existence of a negative stereotype towards a group (for instance, that women are less capable than men in hard mathematics) led to underperformances once the stereotype was triggered. If the stereotype is widely known and concerns gender groups, then even merely stressing the group membership can affect behaviour. In everyday life, there are many ways to stress group membership: the colours used, the dress code, etc. Also ‘tokenism’ might be a feature to highlight group memberships, and experiments have shown that the proportion of women and men in a group can negatively affect

130

Towards a Gendered Rational Choice Theory

women’s performance when they are underrepresented (Inzlicht and BenZeev 2000). While stereotype threat concerns negative beliefs about group members, also neutral beliefs about gender matter. Valian calls ‘gender schema’ [a] set of implicit, or noncounscious, hypotheses about sex differences [which] plays a central role in shaping man’s and women’s professional lives. These hypotheses [...] affect our expectations of men and women, our evaluations of their work, and their performance as professionals. Both men and women hold the same gender schemas and begin acquiring them in early childhood (Valian 1988, 2).

Her experiments point out how schemas affect beliefs, for example in assessing men and women’s height: when they were shown photos of men and women of the same height, participants in the experiment judged men to be taller than women. According to Valian (1998), gender schemas are psychological devices—similar to role schemas—helpful in forming and making sense of exceptions and to explain successes and failures. However, unlike role schemas (father schema, teacher schema, etc.) gender schemas are oversimplifications because they polarize traditionally supposed masculine and feminine traits, making them contradictory, rather than compatible. Valian employs the accumulated disadvantage induced by gender schema as an explanation for glass ceiling phenomena. Similarly, Babcock and Laschever (2003) explain the problem of the advancement of women (or lack thereof) by the phenomenon of avoiding negotiation, which they consider a gendered phenomenon: women are less likely to start a process of negotiation, or less likely to recognize it when it is already in progress (cf. also, Croson and Gneezy 2009). They call ‘not asking’ a phenomenon which includes low expectations, low selfevaluation, low goals, sense of gratitude and inadequacy, fear of asking, the influence of stereotypes, etc., whose causes are identified in a process of self-censorship due to education and gender norms. The cost of avoiding negotiation is high for women, both because of the accumulation of disadvantage and because it also changes the attitude of the people that women fail to ask.

2.4 Adaptive Preference In philosophy, the term of adaptive preference has been employed to analyse the phenomena just reported. ‘Adaptive preferences’ is the name given to those preferences expressed by an agent in response to her set of

Martina Belmonte

131

feasible options. The exemplary case are the sour grapes found in Aesop’s famous fable, The Fox and the Grapes (Elster 1985): the fox, realizing that the alternative of eating grapes was unavailable, changed his preference towards the grapes, deciding that they were still sour and he did not want them, after all. Forming adaptive preferences is a phenomenon that often concerns groups of people whose set of alternatives is narrowed, and it may be detected through a comparison either with preferences expressed before the narrowing of alternatives, or with preferences expressed by the members of other groups. Nussbaum (2000) and Sen (1990) have emphasized that many women in developing countries adapt their preferences to hardly bearable conditions. However, in developed countries too, the phenomenon of the adaptive preference is largely spread, and this occurs on a gender basis. There are, for example, experiments that show how women’s preferences, especially social preferences, tend to be more context-sensitive than men’s preferences (Croson and Gneezy 2009); there are also theories that rely on this diversity (Hakim 2007). However, even if the phenomenon of adaptation is intuitively familiar, its analytical accounts differ. Three issues are at stake: what exactly adaptive preferences are, whether they represent a problem for rationality, and whether they represent a problem for justice. About the definition, adaptive preference can refer both to a change in preference after an agent has realized that the chosen alternative was unavailable, and to the formation of the preference itself. If it can be easy to identify a change of preference for adaptive reasons, an adaptively formed preference is more difficult to detect: indeed, if we assume that adaptive preferences are all of those preferences expressed after having considered the set of available alternatives (Levey 2005), then we are compelled to consider all preferences as adaptive. But this is counterintuitive, and a further specification is required to distinguish adaptive preferences from realistic ones. Bruckner (2009) considers adaptive those preferences that explicitly downgrade unfeasible and upgrade feasible alternatives. He reports the example of a girl who, educated in traditional female roles and therefore unaccustomed to traditional male activities, ends up preferring female activities. However, establishing when such downgrade happens and when it is merely ignorance about alternatives remains problematic. Another issue is the rationality of adaptive preferences. Many authors point out that it is often rational to hold adaptive preferences, but they differ in the standard of rationality. Elster (1985), for example, considers as irrational the preferences due to a causal process occurring non-

132

Towards a Gendered Rational Choice Theory

consciously—like the one concerning sour grapes—and as rational the intentional adaptation, typical for instance of character planning. Bruckner (2009) considers as rational the adaptive preferences endorsed upon reflection, aside from the causal mechanism that has brought them about. A different issue is the political meaning of adaptive preferences. Sen (1999) and Nussbaum (2000) use the phenomenon of adaptation as the main objection to the idea that well-being may consist in the satisfaction of subjective preferences. The problem is whether it is possible to consider something as unjust, if it is adaptively preferred by the agent.7

2.5 A debate on Choice Adaptive preferences and psychological self-censorship phenomena undermine the idea of autonomous agents choosing on the basis of preferences. The debate about choice feminism (Gill 2007; Snyder-Hall 2010; Ferguson 2010; Hirschmann 2010), which consists in the thesis that “feminism should simply give women choices and not pass judgement on what they choose” (Snyder-Hall 2010, 255), emphasizes the tension between individual preferences and collective outcomes. The problem is the following: interpreting choices as expressions of single women’s preferences on the one hand safeguards women’s autonomous decision making, but on the other leaves the problem of collective unfair outcomes unresolved. On the contrary, interpreting choices as the result of a structure that systematically disadvantages women on the one hand explains unfair collective outcomes, but on the other hand impairs single women’s decision making (Browne 2006). The debate about individual choice reflects the structure of gendered phenomena: there is both an individual and a collective aspect. Collective outcomes cannot exist without individual choices, but the sum of features of individual choices is not enough to explain all the traits of collective outcomes. In particular, the slower advancement of women is a collective outcome of women’s choices that is unequal compared to men’s advancement. There are strong intuitions for holding the unequal pace as an unjust difference, but the unjust feature cannot be derived from individual choices, since women’s choices are not coerced. Moreover, a woman that has a preference towards a low work position to save time to spend with her family cannot be accused of having unjust or ill-formed preferences. So, there is a gap between the analysis at the micro-level and at the macro-level (Folbre 1993), and in order to account for inequality, the structure of the constraints in which women and men make choices is

Martina Belmonte

133

generally considered. There is much feminist economic analysis that addresses the problem of choices in constrained environments (Folbre 1993; Fraser 1996; Williams 2001; Crompton 2007), but I am interested in the philosophical aspect of such choices, in relation to the definition of rationality based on revealed preference.

3. Gendered Choices In the second section, I presented Sen’s criticism to internal consistency as well as his thesis that, in order to define rationality, an external dimension must be taken into account. I argued that if the external dimension is led back to interpreting and modelling the choice problem, then the critique is a blank shot, because the definition of consistency applies only once the problem has been modelled. In the third section I presented the problem of the advancement of women as a gendered issue and the debate about choice feminism. I argued that preferences and choices are insufficient tools to analyse women’s choices that lead to unjust collective outcomes. In this section I will merge the previous ones and I will argue that gendered choices constitute an original reason to extend the revealed preference structure.

3.1 A new Example of Violation of Independence Let us look again at the first example in the second section, and let us add some context: Example 5 Three University students—two women, Laura and Ann, and a man, Luke—are wondering about their future plans. They are close to completing their PhDs and so far all of them have had a brilliant career and the same possibilities to keep working in academia. Discussing their preferences, Ann says that she would like to pursue a university career, to work hard and travel a lot, having a job that is both demanding and challenging. Luke expresses the same ambition. Both of them indeed have already received a job offer from a local firm, but they choose to turn it down. Laura claims she wants to have a family and be a full time mum. She would be glad to stop working, then, and to become a housewife. In response, Ann and Luke say they also want a family, but they want to work and to cultivate their interests as well, outside the family. Luke is confident enough that he can well combine his family life and his career in academia, while Ann

134

Towards a Gendered Rational Choice Theory

starts thinking that the job at the local firm would not be so bad for her after all and that it would be better than staying at home or working in academia. A first attempt to model Ann’s choice is CA ({academia, firm}) = academia and CA ({academia, firm, home}) = firm. Her choice is clearly inconsistent and since it cannot be represented by a utility function, it turns out to be irrational. Like in previous examples in the second section, the problem is that the choice does not seem unreasonable. We understand that Ann, recognizing that she belongs to a gender group, downgrades the alternative that most differs from gender roles and opts for an alternative that better allows her to combine family and work. The new alternative on the menu of choice, namely being a housewife, triggers the common knowledge about gender norms that affects the degree of desirability of some alternatives for the agent who belongs to the gender group. The norm in question may be that women should be primarily care-givers and men primarily bread-winners and the reasons to adhere to the norms may be manifold: Ann, for example, may agree with the norm, or she may think that the cost of not conforming is greater than the cost of conforming.8

3.2 Other Examples of Violation of Independence I claim that Anne’s case is different from other cases of violation of the independence axiom. All the previous cases (1.3.1) manifest apparently inconsistent choices that can be adjusted with a different model containing different descriptions for the alternatives. The decision maker chooses among alternatives, but, in abstract situations, she does not decide which alternatives are available: this is the modeller’s task instead. The modeller’s and the decision maker’s decisions may require different conditions of rationality (Hosni 2013), and an inconsistency in modelling is not necessarily inconsistency in choosing. Below I consider in more depth the previous examples. • Epistemic value of the menu: in the narrower menu the agent chooses the alternative ‘having tea with a non-addict person’, whereas in the wider menu the agent rejects ‘having a tea with a cocaine addict.’ Since the alternatives are distinct, the agent’s choice function is consistent. What kind of person the acquaintance is constitutes menu-dependent information (Bhattacharyya, Pattanaik, and Xu 2011), i.e. a piece of information acquired only with the new alternative.

Martina Belmonte

135

• External norm: the agent prefers an apple in a basket with other pieces of fruit to an apple alone. The example shows a menu-dependent criterion (Bhattacharyya, Pattanaik, and Xu 2011), which explains why the agent’s preferences towards the same apple change in different menus. • Reason based: the agent prefers easy decisions to difficult ones. At the simple level of difficulty, like the one in the first menu, he prefers attending an American university. However, when the level of difficulty increases, as it does when there are two American universities, he prefers to avoid a difficult choice and chooses the simplest alternative, i.e. London. His criterion for choosing depends on the composition of the menu.

3.3 Gendered Epistemic Value of the Menu My claim is that the new example above constitutes a new case, which I call gendered epistemic value of the menu (GEVM) and that it is remarkably different from the other cases. I argue that for GEVM Binmore’s redescriptive solution for avoiding inconsistent choices is inappropriate. • Like the epistemic value of the menu, GEVM involves a new piece of information brought along by the new alternative. However, if the epistemic value of the menu example involves a specific piece of information (e.g. being a cocaine addict) previously ignored completely, on the contrary GEVM refers to generic information (knowledge about gender norms), already owned by the agent: the new alternative indeed triggers the common knowledge which thus becomes relevant. It happens that an alternative generally linked to women’s role sets up a group membership function. So information is more indirect and complex in GEVM than in the simpler epistemic value case, because it does not concern the alternatives directly, but the agent’s position towards the alternatives, in relation to degrees of suitability: some alternatives suit better than others. • Both the external norm case and the GEVM refer to a norm: a decency norm the former and a gender norm the latter. But for the decency norm, the agent’s adherence as an individual is enough: she individually decides to behave decently. The gender norm on the contrary is considered only insofar as the agent recognizes her membership in a gender group. Thus the norm is not merely individually endorsed, but endorsed by the agent as a woman. The agent would prefer an academic career, but she recognizes herself as a woman, and because of gender norms, she

136

Towards a Gendered Rational Choice Theory

eventually chooses the local firm. Unlike the case of the poor basket, the notion of all-considered-preference seems particularly inappropriate to describe Ann’s attitude toward the alternative. Moreover, the decency norm states what one should not do, and allows all choices except one, whereas the gender norm states what to do, and dismisses all alternatives except one. However, there are several ways to be influenced by the gender norm, without completely endorsing it. • Just as the student chooses avoiding difficult decisions as his reason, so Anne chooses on the basis of gender norm reasons—yet, not only. She is primarily trying to fulfil her aspirations, and her membership to the gender group is not the only reason for her choice. GEVM has a compromise dimension that is absent in the reason-based choice. As a result of the above comparison, GEVM has something in common with all the other examples, but it cannot be reduced to any of them. The solution of redescribing the alternatives more appropriately is then not applicable to it as it is to them. Precisely, Binmore’s strategy is applicable in principle and indeed the alternatives can be described with two menus as follows: M1 = {academy, firm} and M2 = {academyg, firmg, house}, where academyg and firmg are different from academy and firm because linked to gender norms. Nevertheless, the redescription is indeed so vague that it leads to questions such as where the gender norm comes from and whether it was present even before, what exactly the gender norm is, and what is meant by the agent choosing (and preferring) the firm when it is linked to gender norms. Redescribing the alternatives turns out to be an inadequate strategy for tackling these kinds of inconsistent choices, because too much information gets lost. Reducing Ann’s preferences to a unique ordering, which is the equivalent of the notion of all-consideredpreferences, misses the fact that her preferences are adaptively formed. I call external preference the preference brought in by the new alternative and I call internal the preference that one holds before the change in the menu of choice. Initially the internal preference is the academic career over the local firm. With the new alternative, the gender norm involved is that women are primarily caregivers, so the external preference is to stay at home over working in academia. For transitivity from internal and external preferences, Ann should then prefer staying at home rather than accepting the job in the local firm, but the agent’s internal preference is the local firm instead. So, to avoid inconsistency, Anne modifies her first internal preference and ends up preferring the local firm to the job in academia. The final adaptive preference is then a mixture of internal and external preferences, which would get lost with a unique ordering.

Martina Belmonte

137

3.4 Revealed Preference Theory and Gendered Choices Since we consider Ann’s choice intuitively reasonable, while it is irrational according to the formalization by the revealed preference, GEVM presents normative conditions for rationality that differ from the consistency alone. To make sense of this I propose another strand of Sen’s criticism, which does not address the independence axiom, but the axiom of revealed preference itself. He argues that there are some kinds of actions, like commitment-motivated actions, which cannot be modelled with the revealed preference approach. [The] theory has too little structure. A person is given one preference ordering [...] that is supposed to reflect his interests, represent his welfare, summarise his idea of what should be done, and describe his actual choice and behaviour. Can one preference ordering do all these things? A person thus described may be rational in the limited sense of revealing no inconsistencies in his choice behaviour, but if he has no use for these distinctions between quite different concepts, he must be a bit foolish (Sen 1977, 335-6).

In this passage Sen emphasizes the limits of avoiding motivations at all. The point however is when and how to consider reasons, still thinking in normative rather than in descriptive terms. To tackle gender choices, revealed preference turns out to be thin rather than false, and GEVM shows that for certain situations a thin model misses relevant information. Therefore there is no need to reject them, but they have to be enriched with new elements. In other words, revealed preference considers maximally abstract problems. So in order to tackle gendered choices one needs to relax the degree of abstraction of the problem and enrich the model with additional structures. Revealed preference accounts for choice problems in their most abstract form, and so it is too general for some kinds of problems. As the case of GEVM suggests, problems can be grouped into classes (e.g. self-censorship out of group membership), and modelled with additional structures. The criterion for building classes of problems depends on the aim of the theorist and on the aspects she wants to highlight. If the person who models has concerns about gender justice, it will be useful to point out choices made out of self-censorship. Gender concerns matter in individual choice because individual decisions that involve membership to social groups do not derive only from individual preferences. Thus, merging for example Ann’s aspiration with gender norms in the notion of all-considered-preference is particularly inappropriate. There is an interpersonal dimension of the

138

Towards a Gendered Rational Choice Theory

choice that reflects the agent’s self-position among others and among alternatives. This dimension has been highlighted at a social level, for example by Peter (2003) and Sen (1989). Peter (2003) considers collective choices and argues that the kind of relationship between people whose preferences are considered (siblings, strangers, how they were brought up in their attitudes towards others’ preferences, etc.), should be taken into account. Sen (1989) considers cooperative conflicts, like those happening in families, and claims that interpersonal factors should be considered as well. These are the perceptions of an agent’s contribution to collective achievement and of personal interests, and the breakdown, which is the state agents would find themselves in, if cooperation fails. I claim that such an interpersonal dimension is still present at the individual level. Agents occupy different positions in relation to one another, which are often established by membership of social groups. Because social groups affect individual choices, and in turn groups as social actors are possible only through individual choices, when an interpersonal dimension is present, there are not only issues of individual rationality, but also issues of justice. If theorists are interested in gender justice, then having models that point out the gendered feature of choice is helpful.

3.5 An Account of External Norms The aim of my paper is to present GEVM and to show that, for some kinds of choices, the normative conditions of rationality can be different from those presented in the first section. My claim is that some problems would be better modelled with an extended account of revealed preference. Here I do not want to propose any account, yet I will consider a recent proposal to account for a case of violation of internal consistency, precisely the case of external norm. Bossert and Suzumura (2011) define a set of norms N, interpreted as the set of all pairs (S, w) composed of a set of feasible alternatives S and an element w of S. So, if (S, w) ‫ א‬N, then the choice of w from S will be prevented by the norm. Having the set N, they define a norm-conditional choice function as a choice function C , such that the choice function over the set of alternatives S is a subset of the set S, except that element that, paired with S, is forbidden by the norm: C(S) ‫{ \ ك‬z ‫ א‬S | (S,z) ‫ א‬N}. Bossert and Suzumura directly refers to Sen (1993)’s example: ({apple, nothing}, apple) belongs to the set N, and so apple cannot be chosen when the basket contains only the apple, but ({apple, banana, nothing}, apple) does not. This approach highlights that the relevance of a norm depends on the menu of alternatives, however it is restricted only to

Martina Belmonte

139

those norms that explicitly prohibit an alternative (or a set of alternatives). On the contrary, GEVM involves a norm that shows a favoured alternative. The agent downgrades the alternative that is the furthest from the favoured one, but the favoured one is not automatically chosen because the adherence to the norm is not her only motive. Moreover this account, which may quantitatively catch all pairs of menus and forbidden alternatives, nevertheless make it impossible to distinguish between different norms: every norm is defined by a specific pair of a set of alternatives and the prohibited alternative, without any other information that could help to detect when the same norm affects choices in different menus. For example, as the pair ({apple, nothing}, apple) belongs to N, so the pair ({banana, nothing}, banana), does. If an agent followed the norm of not buying any newspaper in case it held a monopoly on information, even if the only newspaper available were the one that he would have chosen if others were available, also the pair ({The Mars Tribune, nothing}, The Mars Tribune) would belong to N for the agent. But this latter choice is based on a very different norm from the decency norm about fruit. Since to be able to qualitatively distinguish or link two kinds of norms is a desirable feature of an account for membership-based norms, I do not consider Bossert and Suzumura’s external norm account a suitable solution to GEVM.

4. Conclusion The main goal of this paper was to present the gendered epistemic value of the menu. This contributes to rational choice theory by strengthening the criticism of the internal consistency of choice. Supported with random examples, like those presented by the literature so far, this criticism turned out to be a blank shot. On the contrary, by highlighting the gender parameter in women’s choices, the condition of consistency results in an ineffective criterion for rationality. GEVM is also interesting for feminists because it represents a case of self-censorship and genderaffected choice that leads (among other things) to the problem of the slow advancement of women. GEVM identifies a class of problems and the main question that rises is what the normative conditions of rationality for such class of problems are. In conclusion, here I have defined GEVM, I have analysed it thoroughly, and I have paved the way for further developments

140

Towards a Gendered Rational Choice Theory

Bibliography Anderson, Elizabeth. 2001. “Should Feminists Reject Rational Choice Theory?” In A Mind of One’s Own (2d. ed.), edited by Louise Anthony and Charlotte Witt, Boulder: Westview Press. Babcock, Linda and Sara Laschever. 2003. Women don't ask. Negotiation and the Gender Divide, Princeton: Princeton University Press. Bhattacharyya, Aditi, Prasanta K. Pattanaik, and Yongsheng Xu. 2011. “Choice, internal consistency, and rationality.” Economics and Philosophy 27: 123-49. Binmore, Ken. 2009. Rational Decisions, Princeton: Princeton University Press. Bordo, Susan. 1993. “Feminism, Postmodernism, and Gender Skepticism.” In Unbearable Weight, edited by Susan Bordo, 215-244 Berkeley: University of California Press. Bossert, Walter, and Kotaro Suzumura. 2011. “External norms and rationality of choice.” Social Choice and Welfare 37 4:729-41. —. 2010. Consistency, Choice and Rationality. Cambridge: Harvard University Press. Browne, Jude, and Marc Stears. 2005. “Capabilities, resources, and systematic injustice: a case of gender inequality.” Politics, Philosophy and Economics, 4: 355-373. Browne, Jude. 2006. Sex Segregation and Inequality in the Modern Labour Market. Bristol: The Policy Press. Bruckner, Donald W. 2009. “In defence of adaptive preferences.” Philosophical Studies 142: 307-324. Cudd, Anne. 1994. “Oppression by Choice.” Journal of Social Philosophy, 25: 22-44. Crompton, Rosemary. 2007. “Gender inequality and the gendered division of labour.” In The Future of Gender, edited by Jude Browne. Cambridge: Cambridge University Press. Croson, Rachel, and Ury Gneezy. 2009. “Gender Difference in Preferences.” Journal of Economic Literature, 47: 448-474. Damaske, Sarah. 2011. “A ‘Major Career Woman’? How women develop early expectations on work.” Gender and Society, 25: 409-430. Driscoll, Amanda, and Mona Lena Krook. 2008. “Feminist and Rational Choice Theory.” European Political Science Review, 1: 1-22. Elster, Jon. 1985. Sour Grapes, Studies in the Subversion of Rationality. Cambridge: Cambridge University Press. England, Paula. 1993. “The Separative Self: Androcentric Bias in Neoclassical Assumptions.” In Beyond Economic Man: Feminist Theory

Martina Belmonte

141

and Economics, edited by Marianne A. Ferber, Julie A. Nelson, 37-53. Chicago: Chicago University Press. Ferguson, Michaele L. 2010. “Choice Feminism and the Fear of Politics.” Perspectives on Politics 8: 247-253. Fraser, Nancy. 1996. “After the Family Wage: a Post-Industrial Thought experiment.” In Justice Interruptus. Critical reflections on the postsocialist, edited by Nancy Fraser, 41-67. New York: Routledge. Folbre, Nancy. 1993. “Micro, Macro, Choice and Structure.” In Theory on Gender/Feminism On Theory, ed. by Paula England, 323-42. New York: de Gruyter. —. 1994. How pay for kids? Gender and the structures of constraint. London: Routledge. Friedman, Debra, and Diem, Carol. 1993. “Feminism and the Pro(Rational) Choice Movement: Rational-Choice Theory, Feminist Critiques, and Gender Inequality.” In Theory on Gender/Feminism On Theory edited by Paula England, 91-114. New York: de Gruyter. Gill, Rosalind C. 2007. “Critical Respect: The Difficulties and Dilemmas of Agency and 'Choice' for Feminism.” European Journal of Women’s Studies 14: 69-80. Hakim, Catherine. 2007. “The politics of female diversity in the twentyfirst century.” In The Future of Gender, edited by Jude Browne, 191227. Cambridge: Cambridge University Press. Hausman, Daniel M. 2007. “Sympathy, Commitment, and Preference.” In Rationality and Commitment, edited by Fabienne Peter and Hans Bernhard Schmid. Oxford: Oxford University Press. Hirschmann, Nancy J. 2010. “Choosing Betrayal.” Perspectives on Politics,8: 271-278. Hosni, Hykel. 2013. “Toward a Bayesian Theory of Second-Order Uncertainty: Lessons from Non-Standard Logics.” In David Makinson on Classical Methods for Non-Classical Problems, edited by Sven Ove Hansson. New York: Springer. Inzlicht, Michael, and Talia Ben-Zeev. 2000. “A threatening intellectual environment: Why females are susceptible to experiencing problemsolving deficits in the presence of males.” Psychological Science, 11: 365-371. Kreps, David M. 1988. Notes on the Theory of choice. Boulder: Westview Press. Levey, Ann. 2005. “Liberalism, Adaptive Preference and Gender Equality.” Hypatia 20: 127-143. Nussbaum, Martha. C. 2000. Women and Human Development. The capabilities approach. Cambridge: Cambridge University Press.

142

Towards a Gendered Rational Choice Theory

Peter, Fabienne. 2003. “Gender and the Foundations of Social Choice: The Role of Situated Agency.” Feminist Economics 9: 13-32. Rubinstein, Ariel. 2006. Lecture Notes in Microeconomic Theory. Princeton: Princeton University Press. Sen, Amartya. 1977. “Rational Fools: A Critique of the Behavioural Foundations of Economic Theory.” Philosophy and Public Affairs, 6: 317-344. —. 1970. Collective Choice and Social Welfare, San Francisco: HoldenDay. —. 1973. “Behaviour and the Concept of Preference.” Economica, 40: 241-259 —. 1990. “Gender and Cooperative Conflicts.” In Persistent Inequalities, edited by Irene Tinker. Oxford: Oxford University Press. —. 1993. “Internal Consistency of Choice.” Econometrica, 61: 495-521. —. 1995. “Is the idea of purely internal consistency of choice bizarre?.” In World, Mind, and Ethics edited by Altham J., Harrison, 19-31. Cambridge: Cambridge University Press. Spelman, Elizabeth V. 1988. Inessential women. Problems of Exclusion in Feminist Thought, Boston: Beacon Press. Spencer, Steven. J., Claude. M. Steele, and Diane M. Quinn, 1999. “Stereotype threat and women’s math performance.” Journal of Experimental Social Psychology, 35: 4-28. Thalos, Miriam. 2005. “What is a Feminist to do with Rational Choice?.” In A Companion to Rationalism, edited by Alan J. Nelson, 450-467. Hoboken: Wiley-Blackwell. Valian, Virginia. 1999. Why So Slow? The Advancement of Women. Cambridge: The MIT Press. Williams, Joan. 2001. Unbending Gender. Why family and work conflict and what to do about it. Oxford: Oxford University Press, Young, Iris Marion. 1990. Justice and the Politics of Difference. Princeton: Princeton University Press. Young, Iris Marion. 1994. “Gender as seriality. Thinking about Women as a Social Collective.” Signs 19: 713-738.

Martina Belmonte

143

Notes 1

I would like to thank my supervisor, Hykel Hosni, for his trust, patience and shared ideas. Obviously, inaccuracies and mistakes are purely my own responsibility. 2 There are many ways of writing the axiom of independence, according to the specific form of preference relation considered. See Bhattacharyya, Pattanaik and Xu (2010). 3 Originally it was said by Little in relation to Samuelson’s revealed preference theory, quoted in Bossert and Suzumura (2010). 4 For the proof of the representation theorem, see for example Kreps (1988). 5 The Report can be found online: http://goo.gl/cckbVI 6 For similar analyses see for example Valian (1998); Browne (2006). 7 The answer to the problem of the political meaning of adaptive preference depends on what is considered as the problematic point in adaptive preferences: their being adaptive, their being preferences towards unjust things or their being ill-formed. In the case of adaptive preferences concerning gender groups as a consequence of gender norms, a guide to the answer might be their systematic and unequal distribution among gender groups. So, if adaptive preferences are a consequence of a social norm, the conformity to the social norm is not a problem in itself, but the gendered way in which the conformity is expressed, as well as the concrete issue involved (for example economic power), makes it a problem. As Levey puts it: What is wrong with gendered preferences seems to lie in their distribution. Wanting to be a nurse rather than a doctor, or preferring to forgo a good income in order to spend more time with one’s kids are not intrinsically problematic preferences. However, the distribution pattern of these preferences does seem problematic, as does the difference in men and women’s socioeconomic prospects that arise from choices based on these preferences (Levey 2005, 137). 8 Cudd (1994), for example, points out that for a family that anticipates future scenarios in deciding the division of labour among partners, it is advantageous to conform to social gender norms in a society that on average pays more for male labour. Also Browne and Stears (2005) present cases in which agents who have preferences differing from those prescribed by gender norms are more likely to make an effort to satisfy them in society.

PART THREE

NATURALISM, DEONTIC LOGIC AND COGNITIVE SCIENCE MATTIA SORGON (ROMA TRE UNIVERSITY)

1 Ethics is usually defined as a field of philosophy, also called moral philosophy, that “involves systematizing, defending, and recommending concepts of right and wrong behaviour” (Fieser 2003). Following this statement, we can identify three distinct general areas in which the debate about ethics has been developed: metaethics, normative ethics and applied ethics. The first one concerns the origin and meaning of ethical principles, the second aims to define moral standards and criteria that regulate our conduct, while the last considers certain specific and controversial issues of our political, social and scientific experience. The three contributions constituting this section are widely connected with these subject areas. Concerning fundamental topics of the contemporary ethical debate, each paper aims to reflect on a specific issue referred to the domain of metaethics, deontic logic and cognitive science: The debate on naturalism in contemporary ethics by Luciana Ceri analyses the role and the definition of the term “naturalism” in ethical theories; The OIC/PAP Dispute: Two Ways of Interpreting ‘Ought’ implies ‘Can’ by Guglielmo Feis explores the logical relationships between the terms “ought” and “can;” Cognitive Neuroscience and Animal Consciousness by Matteo Grasso provides a thorough exposition of investigations about animal consciousness in contemporary neuroscience. Therefore, this third section aims to take a closer look at the different connections in which the philosophical debate on ethics can be involved.

2 The first contribution provides a detailed analysis of the concept of “naturalism”. Starting from a first triadic definition of the term, Luciana Ceri recognizes a cosmological naturalism, characterized by its opposition

Mattia Sorgon

147

to supernaturalism and idealism, a substantive naturalism, referred to as a metaethical theory, and a methodological naturalism, advocating the application to ethics of those methods pertaining to empirical science. Hence, the author shows how the contemporary debate, Moore’s position for example, is focused on the discussion of substantive naturalism. Within this metaethical theory, two main different developments are identified: the first one, called analytic, regards a theory of moral language that states a definition of moral terms based on non-moral terms (semantic naturalism) and the derivation of moral judgements from descriptive statements (logic naturalism); the second one, called ontological or synthetic, concerns instead a theory of the properties denoted by moral terms based on the rejection of a dichotomy between facts and values. Against these views—each of which denies a specific kind of autonomy of ethics—Moore, together with noncognitivist and Kantian philosophers, propose to maintain a clear distinction between facts and values, and between moral and non-moral statements by appealing to the so-called Hume’s law, the logical view according to which no “ought” can logically be derived from an “is”. Highlighting and further exploring the differences between the various upholders of the autonomy of ethics, Luciana Ceri shows how the contemporary discussion is mainly based on moral ontology, in particular on issues concerning the existence and nature of moral properties. Through the exposition of the realist and antirealist views, the relationship between semantic and ontological autonomy and the reductionist and nonreductionist distinction, she outlines a thorough account of the contemporary debate. Then the discussion moves to the best known objection to naturalism, the open question argument of the naturalistic fallacy (Moore 1903). Having exposed Moore’s original version of the argument, which claims that the definition of “good” in terms of “pleasurable” is mistaken, the author provides a comparison between subsequent proposals and critics of the objection reviewing, among others, the positions of Frankena (1967) and Hare (1952). Focusing on the practicality of moral judgements and the neutrality of metaethics, Luciana Ceri concludes her analysis of naturalism. In the first of the last two points the author examines the criticism based on the question of whether naturalism can account for the motivational force of our moral judgements. Through a reflection on the moral judgement internalism and the Humean theory of motivation, the theme of moral practicality is considered through a comparison between the concepts of motivational and normative force. Finally, the latter point leads to the

148

Naturalism, Deontic Logic and Cognitive Science

conclusion reporting the best known discussions on the issue of the neutrality of metaethics, closely linked with Hume’s law, based on the works of Murdoch (1998) and Scarpelli (1962, 1982).

3 The second contribution offers a review of the debate concerning the derivation of the principle of alternate possibilities (PAP) from the “ought” implies “can” thesis (OIC). The OIC, as Guglielmo Feis clearly shows, states that we cannot be asked to do the impossible: whenever what we ought to do is an impossible act, we are free not to do that , “not can” indeed implies “not ought”. PAP on the other hand concerns the principle according to which we are responsible for what we have done only if we could have acted otherwise. Starting from a general overview of the OIC/PAP debate, based on the definition provided by Frankfurt (1988), the author explores the various truth-values combinations that these two principles can involve, approaching the discussion concerning the possible solutions to the debate. In this regard, Guglielmo Feis proposes to distinguish two new different versions of the OIC principle: the alternate OIC, in which “ought” implies that there is the possibility of avoiding the required action, and the commissive OIC, in which “ought” implies both the possibility of doing and not doing the required action. Hence the author, after highlighting some reasons backing his proposal in the works of Frankfurt (1988), Widerker (1991), Yaffe (2005) and Copp (2008), provides three different study cases based on the possibilities that one version of the principle could hold while the other not. Developing in this way the diving case (where the commissive OIC holds while the alternate OIC fails), the will-making case (where the alternate OIC holds while the commissive OIC fails), and the interplay case (where both principles play a role), Guglielmo Feis concludes his review of the OIC/PAP debate by advocating the introduction of a new distinction in the OIC principle.

4 The third contribution proposes a thorough exposition of the latest investigations around animal consciousness in neuroscience. Matteo Grasso starts from two main questions: “Are there conscious animals beside human beings?” and “Can we know what the experiences of those animals are like?”

Mattia Sorgon

149

The former question, called the distribution question (DQ), concerns the necessity to establish a valid criterion in order to distinguish between conscious and unconscious organisms; the latter, called the phenomenological question (PQ), focuses instead on the qualitative aspects of conscious experience, such as being conscious of the shape of an object, or experiencing desires and fears. These themes, regarding both cognitive neuroscience and philosophy of mind, are closely linked to the hard problem proposed by Chalmers (1995) and based on the concept of phenomenal consciousness introduced by Block (1995), which refers to the subjective and qualitative nature of each experiential state of consciousness, whether it belongs to humans or not. Two different types of theory are proposed as answers to (DQ): discontinuity theories, which state that consciousness emerged at a certain stage of the phylogeny and a sharp line divides conscious living organisms and unconscious ones, and continuity theories, which claim that consciousness is a property that has always characterized living beings and that it is distributed in different degrees throughout the natural world. The author discusses the most important examples of behavioural arguments: these are proposals characterized by the establishment of a criterion based on similarities and dissimilarities among living organisms obtained through the observation of their behaviour. Forsaking the behavioural arguments due to the inconclusiveness of their approach (Baars 2005), Matteo Grasso focuses on recent developments in cognitive neuroscience, providing non-behavioural methods for assessing consciousness (Laureys and Tononi 2009). Showing a detailed review of the most important results of neuroscience regarding the problem of consciousness (mainly based on the analyses of neuroanatonomy and neurophysiology of human consciousness), the author compares this data through sets of homologies and analogies, referring to, respectively, structural and functional similarities between different species. The first set regards phenotypic characters that are similar in different species because of a common ancestor, while the second set concerns phenotypic characters whose similarity across different species is due to a convergent evolution. Matteo Grasso hence mentions the “Cambridge Declaration of Consciousness” (Low et al. 2012), a document of central importance for the animal consciousness debate. This declaration, signed by a group of internationally renowned scientists, summarizes many experimental observations and aims at setting some undisputable boundaries on the issue of animal consciousness.

150

Naturalism, Deontic Logic and Cognitive Science

Finally, considering the theme of PQ from Nagel’s perspective (Nagel 1974), the author concludes that, although the answer to PQ might be impossible in principle, cognitive neuroscience may provide an answer to DQ that could represent a base for the debate on animal consciousness.

Bibliography Baars, Bernard J. 2005. “Subjective experience is probably not limited to humans: The evidence from neurobiology and behaviour.” Consciousness and Cognition 14: 7-21. Block, Ned. 1995. “On a confusion about a function of consciousness.” Behavioural and Brain Sciences 18(2): 227-287. Chalmers, David J. 1995. “Facing up to the problem of consciousness.” Journal of Consciousness Studies 2(3): 200-219. Copp, David. 2008. “‘Ought’ Implies ‘Can’ and the Derivation of the Principle of Alternate Possibilities.” Analysis, 68: 67-75. Fieser, James. 2003. “Ethics.” In Internet Encyclopedia of Philosophy. Accessed January 2013. http://www.iep.utm.edu/ethics Frankena, William K. 1967. „The Naturalistic Fallacy.” In Theories of Ethics, edited by Philippa Foot, 50-63. Oxford: Oxford University Press. Frankfurt, Harry G. 1988. “What Are We Morally Responsible For.” In The Importance of What We Care About, Harry G. Frankfurt, 95-103. New York: Cambridge University Press. Hare, Richard M. 1952. The Language of Morals. Oxford: Clarendon Press. Laureys, Steven, and Giulio Tononi. 2009. “The Neurology of Consciousness: Cognitive Neuroscience and Neuropathology.” Oxford: Academic Press. Low, Philip, Panksepp, Jaak, Reiss, Diana, Edelman, David B., Van Swinderen, Bruno, and Christof Koch. 2012. “The Cambridge Declaration on Consciousness.” Accessed October 30, 2012. http://fcmconference.org Moore, George E. 1903. Principia Ethica, Cambridge: Cambridge University Press. Murdoch, Iris. 1998. Existentialists and Mystics. Writings on Philosophy and Literature, edited by Peter Conradi. New York: Penguin Books. Nagel, Thomas. 1974. “What Is It Like to Be a Bat?.” The Philosophical Review 83(4): 435-450. Scarpelli, Uberto. 1962. Filosofia analitica, norme e valori. Milano: Edizioni di Comunità.

Mattia Sorgon

151

—. 1982. “La meta-etica analitica e la sua rilevanza etica.” In Uberto Scarpelli, L’etica senza verità, 73-112. Bologna: il Mulino. Widerker, David. 1991. “Frankfurt on ‘Ought Implies Can’ and Alternate Possibilities.” Analysis, 51: 222-24. Yaffe, G. 2005. “More on “Ought” Implies “Can” and the Principle of Alternate Possibilities.” Midwest Studies in Philosophy, 29: 307-12.

THE DEBATE ON NATURALISM IN CONTEMPORARY ETHICS LUCIANA CERI (UNIVERSITY OF PARMA)

Abstract “Naturalism” is an ambiguous term that may be used to indicate a standpoint opposed to both supernaturalism and idealism (cosmological naturalism), or a metaethical theory (substantive naturalism), or the view that the research methods of empirical sciences can also be applied to ethics (methodological naturalism). As regards the naturalistic theories, the contemporary philosophical debate is concerned with substantive naturalism and moral ontology rather than with semantic analysis. One of the objections raised against naturalism is that it implies the naturalistic fallacy, which George Edward Moore illustrated with the open question argument. Although this argument is still controversial, it does not lack supporters at all. Another criticism of naturalism is related to the practicality of moral judgements. Practicality can be considered either as a motivating force or as a normative force. I shall examine an argument intended to show that naturalism cannot account for the motivating force of moral judgements, and I shall consider the objections that the naturalist might raise against it. Then I shall turn to the normative force of moral judgements, which naturalism allegedly fails to account for, and I shall discuss the readings of the debate on naturalism offered by Iris Murdoch and Uberto Scarpelli. According to Murdoch and Scarpelli, the source of the rejection of naturalism, and of the defence of Hume’s law, is the claim that human beings are autonomous free agents. Agreeing with such a reading implies the abandonment of the claim that metaethics is completely value-neutral. The neutrality thesis can be understood in two different ways, and only one of them is to be agreed with. The other one involves a violation of a principle which is, so to say,

Luciana Ceri

153

specular to Hume’s law: the principle that one cannot derive an “is” from an “ought”, i.e., no descriptive statement can be derived from any purely evaluative or normative statement.

1. The Meanings of “Naturalism” The term “naturalism” was long used with reference to those theories that explained natural phenomena and human actions by only referring to natural entities. In the early twentieth century, George Edward Moore introduced in his Principia Ethica a narrower sense of that term: according to Moore, naturalistic theories were the theories that defined moral terms by other terms denoting natural properties, and identified goodness with any one of those properties. The former kind of naturalism, which is called cosmological naturalism (Couture and Nielsen 1995, 4, 23), is a philosophical view opposed to both supernaturalism and idealism. The latter kind of naturalism, which is called substantive naturalism, is a metaethical theory. Moreover, there is a third kind of naturalism, methodological naturalism, which involves a commitment to apply the methods of inquiry of empirical sciences to all intellectual subjects (including ethics).1 Methodological naturalism is usually combined with cosmological naturalism, and it is compatible with the rejection of substantive naturalism: one can consistently claim (as Allan Gibbard does) that moral terms are not definable by terms denoting natural properties, and that goodness is not a natural property, while holding that a naturalistic account of morality can indeed be provided (Gibbard 1993, 58). Moore’s discussion and the contemporary debate are focused on substantive naturalism. Two different varieties of this theory can be distinguished: on the one hand, analytic naturalism, that is, a theory of moral language, according to which moral terms can be defined by merely using non-moral terms (semantic naturalism), and moral judgements can be derived from purely descriptive statements (logical naturalism); on the other hand, ontological or synthetic naturalism, which is a theory about the nature of the properties that are denoted by moral terms.2 According to this theory, there is no dichotomy between facts and values, and the entities to which moral terms refer are natural entities, i.e., they are just those entities that natural and social sciences deal with. So, naturalism rejects the thesis of the semantic, logical, and ontological autonomy of ethics, which was prevailing in the first half of the twentieth century and still has many upholders:3 both Moore and noncognitivists, together with those who hold a Kantian view, claim that

154

The Debate on Naturalism in Contemporary Ethics

ethics is not reducible to natural sciences or psychology. So, they make a sharp distinction between facts and values, and they hold as true the logical view related to such a distinction, namely “Hume’s law”. According to that law, no “ought” can be logically derived from an “is;” in other words, there is a gap between descriptive and evaluative statements that cannot be filled. However, the autonomy of ethics is upheld in different ways. While Moore adopted a realist view about moral ontology, and a cognitivist view of moral epistemology, noncognitivists put the thesis of the autonomy of ethics in terms of the function of moral judgements. For noncognitivists, moral judgements do not state anything about moral properties or facts, but they express and tend to arouse certain feelings, or attitudes, or guide conduct.4 Moral language, and evaluative language in general, is different from descriptive language: evaluating and describing are essentially different activities. Moreover, noncognitivists hold an antirealist position, according to which there are no moral properties at all. So, from the noncognitivist point of view, both the naturalists’ claim that moral terms denote natural properties and Moore’s claim that such terms denote certain (non-natural) properties are mistaken. In the 1950s, however, criticisms against naturalism were questioned, and many doubts were raised about both the distinction between descriptive and evaluative language, and the fact/value dichotomy. The writings of Elizabeth Anscombe, Peter Geach, Philippa Foot, and Iris Murdoch gave rise to a revival of ethical naturalism (Anscombe 1969; Geach 1972; Foot 2002; Murdoch 1998). In the last two decades of the twentieth century, then, new forms of naturalism have been put forward, as a result of the changes that have taken place in metaethical theorizing. Metaethics is no more reducible to the analysis of moral language, but it includes in its domain issues about the epistemology and metaphysics of morals (Darwall, Gibbard and Railton 1992, 125-126).

2. The New Naturalism At first, the defence of naturalism focused on the nature of moral concepts and language. Anscombe, Geach, and Foot worked out an Aristotelian version of naturalism referring to the teleology of living beings. On the contrary, the more recent naturalistic theories do not appeal to such a teleology: therefore, they may be called scientific forms of naturalism (Magni 2004, 11). Moreover, contemporary naturalistic theories do not only deal with semantic issues: naturalists and their critics do not merely discuss the

Luciana Ceri

155

nature of moral language, but also, and especially, the existence and nature of moral properties. So, ontology is now at the centre of a debate: whereas in the first half of the twentieth century the relevant opposition in metaethics was between naturalism and antinaturalism, now the field is divided between realism, which is held by naturalists (among others), and antirealism, which is advocated by noncognitivists (among others). According to naturalists, moral properties are natural properties: there are no moral properties or facts other than, or not constituted by, the properties and facts that natural and social sciences deal with. There is a continuity between ethics and science, and there is no sharp distinction between facts and values. Yet, most naturalists distinguish the ontological thesis that moral properties are natural properties from the semantic thesis that moral terms have the same meanings as nonmoral terms, and they claim that the ontological thesis does not imply the semantic thesis. That is, they claim that one can consistently maintain the semantic autonomy of ethics, while at the same time rejecting its ontological autonomy (Pigden 1989, 149); in other words, one can consistently claim that moral terms are not definable by merely using nonmoral terms, and that these two kinds of terms denote the same properties (Sayre-McCord 1988, 22). Indeed, the identity of properties does not imply the synonymity of the terms that refer to those properties (Kripke 1980, 106-155; Putnam 1981, 84-85, 206-208). The term “water” and the formula “H2O”, for instance, denote the same element, but they have different meanings; so, understanding the former does not imply understanding the latter. Moreover, one may consistently agree with the view about the logical autonomy of ethics, while holding that moral properties are natural properties, since the thesis of logical autonomy of ethics is entailed by the thesis of semantic autonomy of ethics. The impossibility of defining moral terms by merely using terms denoting natural properties entails the impossibility of drawing conclusions that contain moral terms merely from premises containing only nonmoral terms. However, as Charles Pigden has pointed out, the logical autonomy is not a distinctive feature of moral language. Every kind of discourse has this autonomy, because logic is conservative. So, the conclusion of an inference must not include anything that is not present, at least implicitly, in its premises (Pigden 1989, 148). Two sorts of naturalism can be identified: reductionist and nonreductionist naturalism, depending on what the verb “to be” is taken to mean in the statement that moral properties are natural properties.

156

The Debate on Naturalism in Contemporary Ethics

Those who argue for reductionism (Harman 1977; Brandt 1979; Railton 1986a; Lewis 1989) do not simply hold that moral properties are related to, or dependent on, natural properties, but rather that the former are identical or reducible to the latter ones. To say that moral properties are identical or reducible to natural properties implies that moral properties are constituted by different combinations of natural properties. Yet the constitution thesis does not imply the identity or reducibility thesis. That is the view of nonreductionist naturalism (Sturgeon 1988; Boyd 1988; Brink 1989, 156-160). Both reductionism and nonreductionism accept methodological naturalism, according to which the very same scientific methods of inquiry can be employed in ethics. Nonetheless, there is an important difference between those two varieties of naturalism. Reductionist naturalism eliminates moral properties: no explanation of moral judgements and actions needs appeal to moral properties or facts (Harman 1977, 21); an examination of nonmoral facts (including psychological facts, such as moral beliefs and moral sensibility) is enough to explain such judgements and actions. Nonreductionist naturalism, on the contrary, does admit the existence of moral properties and facts, on the ground of their explanatory role (for example, an appeal to generosity may help to explain why someone tried to help someone else in need).

3. The Naturalistic Fallacy Several objections may be raised against naturalistic theories. The best-known, and still controversial, is Moore’s objection that naturalism makes a logical mistake, which Moore himself called naturalistic fallacy. Moore showed such fallacy by the “open question argument”, which works this way. Suppose that “good” is defined in terms of “pleasurable”. One can always ask whether something that is pleasurable is also good. However, if that definition were correct, then the answer to such a question would be obvious. Moreover, saying that a certain thing is pleasurable, but it is not good, would be a contradiction, for that would be the same as saying that the thing in question was pleasurable, but not pleasurable too. Yet there is no contradiction in saying that something is pleasurable, but it is not good (Moore 1903, § 13). So, the definition of “good” in terms of “pleasurable” is mistaken. And Moore thought that one could show in the same way that any other definition of “good” in terms of nonmoral terms was mistaken. The open question argument has been employed by other intuitionists as well, and (in a different way) by Alfred Jules Ayer, Charles Leslie

Luciana Ceri

157

Stevenson, and Richard Mervyn Hare (Ayer 1951, 104-105; Stevenson 1963, 15, 30; Hare 1952, 81-93). Hare argues that if someone says “this is a good strawberry”, and we hold a naturalistic view, then we will say, for example, that the judgement “this is a good strawberry” means “this is a sweet strawberry”. Here “good” is defined in terms of sweetness. However, were someone to ask why that strawberry was good, we could not reply that it was good because it was sweet; otherwise our statement would be a tautology, for it would be equivalent to the statement “that is a sweet strawberry because it is sweet”. But moral statements are not tautologies: we make moral statements in order to commend or condemn something, or to guide other people’s or our own actions; and tautologies cannot carry out those functions, since they have no content at all. Moral terms are not definable by merely using nonmoral terms, since these two kinds of terms do have different functions. According to Hare, naturalistic theories overlook the evaluative or prescriptive meaning of moral terms, and therefore they fail to account for the conceptual connection between moral judgements and actions. Were we to agree to any naturalistic definition of moral terms, we could not use moral language in order to guide conduct any longer. Moore’s refutation of naturalism has been questioned on different grounds. First, in his essay The Naturalistic Fallacy (1939) William Frankena made an attempt to clear up the real issue between intuitionism and naturalism. Frankena remarked that defining a property, such as the property for which the term “good” stood, in terms of another property was not a logical mistake at all. Furthermore, in Frankena’s view such a way of defining “good” would have been better labelled the “definist fallacy” (Frankena 1967, 56), inasmuch as Moore claimed that any definition of “good” in terms of metaphysical properties would be as mistaken as any definition in terms of natural properties. Moreover, Frankena argued that, if one were to claim that any definition of the term “good” was fallacious (in the aforementioned sense) because the term was not definable, then one would have to show that the property to which “good” referred was a simple property and, therefore, it was not possible to define it. Otherwise, the alleged refutation of naturalism would turn out to be a question-begging argument (Frankena 1967, 59). Recently, Nicholas Sturgeon has noted that we know right from the start that one term at least, that is “good”, refers to the same thing as “good” does. So, if one is to claim that “good” does not refer to the same thing to which terms denoting natural properties refer, then one must assume that “good” does not denote any of those properties (Sturgeon

158

The Debate on Naturalism in Contemporary Ethics

2003, 536). Yet that is the very point at issue between Moore and naturalists. It has also been pointed out that the open question argument is meant to establish that there is no synonymity between moral and nonmoral terms. This claim, however, may be questioned for a number of reasons. Firstly, there may be a hidden synonymity between two terms, so that one does not immediately see it. That is so when the definition of a term is particularly complex; then, through a semantic analysis one may find something new about the meaning of that term (Brandt 1959, 163-166). So, there may be some correct definitions that leave open questions (Smith 1987, 293-294; Smith 1994, 38-39; Jackson 1998, 151). Moore’s argument, on the contrary, is based on the assumption that meanings are transparent, and that the synonymity between two terms, if present, is always plain. Secondly, someone rejects conceptual analysis and holds that the open question argument is aimed at refuting semantic naturalism, but not metaethical naturalism more generally (Harman 1977, 19). Ontological naturalism just claims that moral properties are identical to (or constituted by) natural properties; and it does not commit one to accept any reductionist definition of moral terms. Moreover, some naturalists, such as Gilbert Harman, agree that moral language might not be analytically reducible to any simpler language (Harman 1977, 20).5 It would be a mistake, however, to suppose that the open question argument has disappeared from philosophical debate once and for all. Indeed, someone does accept it in its original form (McGinn 1997, 10-20; Regan 2003), and others (for example, Terry Horgan and Mark Timmons) argue that it might be adapted to the refutation of ontological naturalism (Horgan and Timmons 1992). Moreover, some naturalists (such as Peter Railton) are willing to acknowledge that the open question argument plays a critical role (Railton 1989, 158), for it helps to see the reason why many people find it hard to accept naturalistic definitions of moral terms (Darwall, Gibbard and Railton 1992, 116).

4. The Practicality of Moral Judgements Another question that may be raised is whether naturalism can account for the practicality of moral judgements, that is, for the relation between an agent’s moral views and her actions. According to the naturalist, moral judgements are descriptive statements about moral properties and facts, which are natural properties and facts. So, when someone says that a certain act is right, she is

Luciana Ceri

159

ascribing a certain feature to that act; and, if her judgement is sincere, then she is expressing the belief that the act in question has that feature. However, noncognitivists hold (following Hume) that facts and cognitive mental states lack any motivational force. In other words, knowing some fact or having a belief can move someone to perform a certain action only if she has a desire to perform it, or a pro-attitude toward performing that action. So, if moral judgements were descriptive statements, and they expressed beliefs, then they could not motivate anyone to act, as indeed they do. This criticism is based on 1) moral judgement internalism, and 2) the Humean theory of motivation. 1) Moral judgement internalism holds that moral judgements have an intrinsic motivational force. According to this view, there is an internal, conceptual, and, therefore, necessary connection between one’s moral judgements and one’s motivation to act in accordance with them (Darwall 1992, 155; Darwall 1995, 9-11). This connection derives from the very meaning of moral judgements. So it is not conceptually possible that one sincerely makes (or assents to) a moral judgement without being motivated to act accordingly. That is a logical, conceptual impossibility, since it depends on the meaning of that moral judgement. Therefore, moral judgements have an intrinsic motivating force. 2) The Humean theory of motivation6 is so called because it presupposes Hume’s view that beliefs (or reason) lack in themselves the capacity to motivate, that is, they cannot move anyone to action. According to the Humean theory of motivation, desire is not the mental state of being motivated to act (contrary to Jonathan Dancy’s views7), but rather it is what moves someone to perform an action: desire is a motivating mental state. According to this theory, motivation is internal to desires, and it essentially depends on them. The desire needed if an agent is to be moved to do something is what drives the agent, and it is the source of her motivation to act, while beliefs about facts merely give a direction to the pull exerted by desire, that is, to the agent’s motivation (and, if those beliefs are true, they show the agent the way for the satisfaction of her own desire). So it is not enough for one to have certain beliefs (or other cognitive mental states) about an action if one is to be moved to perform it, but one must also have a desire to perform the action in question, or some other non-cognitive mental state toward it.8 The Humean theory does not hold that motivation to act depends wholly on the agent’s desires, and Hume himself merely claimed that reason alone cannot be a motive to any action. Both desire and belief are necessary to yield motivation. Nonetheless, they play two quite different,

160

The Debate on Naturalism in Contemporary Ethics

asymmetrical roles in it: desire is an active mental state, which moves the agent to action, and it is the very source of motivation, while belief is passive—it merely gives a direction to the pull exerted by desire and indicates to the agent what to do. Thus, desire plays a ruling role in motivation. The Humean theory of motivation is based on the view that desire and belief are very different from one another. And such a difference explains why desire plays a dominant role in motivation. That difference relates to the direction of fit of desires and beliefs.9 Desire is a mental state with which the world must fit; that is, desire has a world-to-mind direction of fit. Belief, by contrast, has a mind-to-world direction of fit: it must fit actual states of affairs, for the peculiar function of any belief is representing something that does exist. To put it in another way, desires are mental states that aim at satisfaction, and they cannot be said to be true or false, whereas beliefs aim at tracking the truth and representing existing states of affairs. A desire is satisfied if the state of affairs it presents becomes actual; a belief is true if the state of affairs it represents corresponds to the actual one. Let us now go back to the objection to naturalism that if moral judgements were descriptive statements, and they expressed beliefs, then they would not have the motivational force that they have. This objection is sound if the two theses it is based on are valid. Yet their validity is controversial. Upholders of naturalism may meet such an objection in two different ways. On the one hand, they may accept moral judgement internalism, while rejecting the Humean theory of motivation. So naturalists may hold that one cannot make a moral judgement without being motivated by that judgement itself to act accordingly, and that cognitive mental states have an intrinsic motivational force. On the other hand, naturalists may agree with the Humean theory of motivation and, at the same time, reject moral judgement internalism: they may hold that beliefs cannot by themselves move anyone to act, and that the motivational force of moral judgements is not conceptually necessary, but it is a contingent psychological fact that is dependent upon the content of one’s own moral views, desires, and attitudes (Brink 1989, 49). In the latter case, naturalists agree that cognitive mental states (and facts) cannot motivate to action, but they uphold moral judgement externalism, which is the thesis that one’s motivation to act in accordance with one’s own moral judgements is not internal to the judgements themselves. According to moral judgement externalism, the motives which get someone to do what she judges she ought to do lack a necessary

Luciana Ceri

161

connection with her moral judgement. Therefore, someone may think that she ought to perform a certain action without having any motive to perform it, and without feeling motivated to perform it at all. According to externalism, moral judgements cannot by themselves move an agent to do anything. This view enables naturalists to account for weakness of will, that is, for those cases in which one’s motivation to act in accordance with one’s own moral opinions does not have a sufficient force to yield action. Such cases might seem to be inexplicable by those who hold an internalist view about the relation between moral judgements and motivation. That is not so, however, because two forms of internalism can be distinguished: weak internalism and strong internalism.10 According to weak internalism, a moral judgement provides one with a motive to act in a certain way, but that motive is not necessarily stronger than other motives one does have to act in a different way. So, it may be that although one thinks that a certain action is morally right, or that it ought to be performed, her motive to perform it does not override the motives she has to do something else. According to strong internalism, by contrast, the motivation coming from moral judgements is stronger than the other motivations one has, and it is sufficient to produce action. The need to account for weakness of will only arises for those who hold a strong form of internalism. The view about the internal connection between moral judgements and motivation, however, does not commit one to holding strong internalism. It merely entails that it is not conceptually possible that one is completely indifferent to the moral judgements that she sincerely makes. After all, the supposition that there is an internal connection between moral judgements and the motivation to act in accordance with them, a connection that does not come from anything external to the moral judgements themselves, seems to be supported by the fact that when we behave in a different way from the way we regard as right, we often feel the need to give other people, or ourselves, some justifying reasons for our own conduct; and sometimes we experience feelings such as guilt or remorse. The cases in which someone acts contrarily to her own moral views might be described as conflicts between the motivation deriving from her own moral views and the motivation coming from desires, or inclinations, which lead her to do something opposite to what she judges to be morally right. The naturalistic account of the connection between moral judgements and motivation does not seem to be reconcilable with the capability people have to act in accordance with their own moral opinions, even when such opinions clash with their desires. Moreover, naturalism fails to account for

162

The Debate on Naturalism in Contemporary Ethics

some negative feelings that people often have when they act against what they think it is right or what they believe they ought to do. Such feelings show that moral judgements do have some motivational force, though it is not an overriding force. The “practicality” of moral judgements, however, is not to be understood only as a motivational force. The thesis that moral judgements are practical can also be taken to mean that moral judgements can guide the agent’s choices and actions. So, the practicality of moral judgements can be considered as a normative force. Such a force lies in giving the agent reasons that justify a certain action. Justifying reasons are considerations in favour of an action, and they must be distinguished from motives. Motivational and normative force of moral judgements, or the sphere of motivation and the sphere of justification, are—at least conceptually— separated: every action is motivated (that is, it is done for certain motives), but not all actions are justified (that is, supported by justifying reasons). It seems that naturalism fails to account for the normativity of moral language, while acknowledging it. In Natural Goodness, for example, Philippa Foot writes: “[Noncognitivist] theories were devised to take account of something that really is a feature of moral judgement: the ‘action-guiding’ character of morality […]” (Foot 2001, 9). If moral facts did exist and were natural facts, and if moral judgements were statements describing those facts, then in order to account for the normative force of moral judgements, one should hold that those facts guide one’s own choices and actions and are a source of reasons for action. That is, one would have to admit the existence of entities endowed with what John Mackie called “authoritative prescriptivity” (Mackie 1977, 39). But how is it possible for any fact to guide action? How could the mere knowledge of any state of affairs provide someone with a reason for performing a certain action, if she did not see in that state of affairs any reason to perform it? Such questions present an obstacle to naturalism. This obstacle is brought out by the open question argument (Rosati 2003, 501, 503; Railton 1989, 158). In Hare’s view, this argument shows that any naturalistic definition of moral terms fails to grasp their evaluative meaning. Moreover, according to some opponents of noncognitivism (such as Christine Korsgaard), the open question argument shows that neither analytic naturalism nor synthetic naturalism can account for the human beings’ capacity to wonder about what they ought to do. The human mind is self-conscious in the sense that it is essentially reflective. So, if human beings are to do anything, then they need some reasons: human beings need to understand why certain moral facts justify certain behaviours and

Luciana Ceri

163

not others (Korsgaard 1996, 43, 46-47).11 As Connie Rosati has noted, the force of the open question argument comes from the fact that human beings are able to make evaluative judgements and act autonomously (Rosati 2003, 506-507).

5. Hume’s Law and the Neutrality of Metaethics It follows from what has just been argued that the source of the rejection of naturalism, and of the view that it is not logically possible to derive any moral conclusion from purely descriptive premises, lies in an attempt to uphold the autonomy and freedom of human beings—which would be undermined if values were identical or reducible to facts, as naturalism claims. It follows from this reading of the debate on naturalism that the point at issue is not only theoretical, but also substantive. For this debate is not merely about the validity of the logical thesis that moral judgements cannot be derived from purely descriptive statements, but also about the defence of particular values.12 This way of seeing the issue about the relation between facts and values, or between descriptive statements and moral judgements, has been suggested for different purposes by both Iris Murdoch, in her essays Vision and Choice in Morality (1956) and Metaphysics and Ethics (1957), and Uberto Scarpelli, in his essay La meta-etica analitica e la sua rilevanza etica [“Analytic Metaethics and its Ethical Relevance”] (1980) (Murdoch 1998, 65-68, 70-74, 92-98; Scarpelli 1982, 108-110).13 Murdoch rejects the fact/value dichotomy; Scarpelli, on the other hand, argues in favour of it, and he sees in Hume’s law the “more specifically and aggressively liberal aspect of metaethics” (Scarpelli 1982, 109). Ever since Filosofia analitica, norme e valori [“Analytic Philosophy, Norms, and Values”] (1962) Scarpelli has expressed his own dissatisfaction with the alleged reasons for defending Hume’s law (Scarpelli 1962, 74).14 Scarpelli has pointed out that this law has often been taken as an obvious principle, as if it were a law of reason that could not be rejected at all. Yet, those who argued for it also claimed that logic was conventional, that is, that logical rules were valid because we did accept them. Scarpelli argues that in such a case we could nevertheless bring in some logical rules that would allow us to derive evaluative or normative statements from purely descriptive statements. So, what are the reasons not to bring in those rules? The reason, as Scarpelli has noted, is that by bringing them in

164

The Debate on Naturalism in Contemporary Ethics we would block our own choices once and for all, or at least the possibility that we might express our own choices in front of the situations and the events which were described by those statements [that is, by descriptive statements] (Scarpelli 1962, 75).

It is better, therefore, to have logical rules that allow the discourse to have a sufficient elasticity so that we can express various kinds of reactions and make various choices in any situation. The way Scarpelli has formulated the issue about the validity of Hume’s law is acceptable, as Gaetano Carcaterra remarked (Carcaterra 1969, 578-579), in so far as logical conventionalism is well-grounded. Apart from its validity, however, what is essential from Scarpelli’s point of view is that the defence of Hume’s law has its own source in an attempt to argue for human freedom: we have to accept Hume’s law not because human beings are free (and Hume’s law does respect their freedom), but rather because we want to make them free, we do have the value of freedom in our own ethics (Scarpelli 1982, 110).

The theories rejecting Hume’s law, by contrast, “bear the denial of freedom as a poisonous seed” (Scarpelli 1982, 112). It is worth noting that the way Scarpelli sees both the defence of Hume’s law and the rejection of metaethical naturalism is a reverse of the way Hare presented that defence in Freedom and Reason. Hare held that his aim was “the [...] one of showing how the fact of moral freedom is what gives moral language one of its characteristic logical properties” (Hare 1963, 61, my italics), that is, prescriptivity. Murdoch’s views about the nature of the debate on facts and values are similar to Scarpelli’s views (and they have been questioned by Hare15). Murdoch argued that the thesis of the distinction between facts and values, and between descriptive and evaluative meaning of moral terms, was not a merely conceptual or logical thesis, but was based on a moral assumption. According to Murdoch, claiming that the evaluative meaning of moral terms is distinct from their descriptive meaning, and that the former is independent from the latter (that is, from the criteria according to which people apply moral terms), is just another way of claiming that everyone should be free to apply moral terms to whatever they want. In other words, in Murdoch’s view, the thesis that the evaluative meaning of moral terms is independent from their descriptive meaning is just a way of claiming that everybody is completely free in the choice of the criteria of moral evaluation. Murdoch wrote: “I think that much of the impetus of the

Luciana Ceri

165

argument against naturalism comes from its connection with, and its tendency to safeguard, a Liberal evaluation [...]” (Murdoch 1998, 95).16 Thus, according to Murdoch, any argument in favour of the fact/value dichotomy, and the view that no evaluative judgement can be derived from any description of facts, reveals a definite moral attitude. Indeed, this kind of argument is a way of defending the view that human beings are not bound by facts in their own evaluative judgements, and they are autonomous agents. Murdoch acknowledged that such a view of human nature had its own merits, for it inspired political liberalism. Yet she pointed out that good political philosophy was not necessarily good moral philosophy (Murdoch 1998, 366). So, Murdoch proposed a different conception of individual freedom, and she thought that by so doing she was holding a normative view that was different from the one held by noncognitivism. According to Murdoch, freedom does not merely lie in one’s choice of the facts to which a certain moral concept will be applied, but rather in one’s ability to reorganize or change one’s own moral concepts. Freedom is thus conceived by Murdoch as a mode of reflection, that is, as a function of the attempt to reach the clearest possible vision of reality. Therefore, freedom cannot be separated from knowledge of the facts. Moral choices are not a product of our own will, but of a careful examination of our surroundings: The true naturalist [...] is one who believes that as moral beings we are immersed in a reality which transcends us and that moral progress consists in awareness of this reality and submission to its purposes (Murdoch 1998, 96).17

It follows from Murdoch’s views that philosophical images of morality are deeply influenced by moral attitudes (Murdoch 1998, 98).18 To agree to the way Scarpelli and Murdoch put the question about the relation between facts and values, or between descriptive statements and moral judgements, involves dropping the thesis that metaethics is neutral from the evaluative (and normative) point of view, that is, the thesis that the analysis of moral language, and the relation between facts and values, does not require or presuppose any moral principle at all. The neutrality thesis about metaethics, as Michael Gorr has remarked by pointing out its ambiguity (Gorr 1985, 136, footnote 2), may also be taken as the view that no moral principle can be derived from metaethical inquiries, for example, from an examination of the relation between facts and values. Different normative theories can, therefore, share the same view of such a relation.

166

The Debate on Naturalism in Contemporary Ethics

Let us now briefly consider the former way of understanding the neutrality thesis that was just mentioned. Its denial can be seen in two very different ways: 1) it can be seen as the view that one proposes certain solutions for metaethical issues because one defends certain values, as the defence of any metaethical theory is motivated by the adoption of certain values, and it can be explained by referring to such values; 2) it can be seen as the view that some normative or evaluative premises are needed for one to justify or provide a foundation for the metaethical theory that one holds. One may agree with the thesis that metaethics is not value-neutral,19 but only as long as the lack of neutrality is taken in the former way abovementioned: the defence both of the fact/value dichotomy and of the irreducibility of values to mere facts, for example, can be explained by referring to the adoption of certain values. The sphere of explanation, however, must be distinguished from the sphere of normative justification (or of foundation). Indeed, claiming that certain values have to be presupposed in order to justify a particular metaethical thesis would mean deriving theoretical and, in a wide sense, descriptive conclusions (about the nature of the relation between facts and values, for example) from evaluative premises. So, it would mean violating a principle that is symmetrical, so to say, to Hume’s law: the principle that one cannot logically derive any factual or descriptive statement from an evaluative or normative judgement—a principle that can hardly be questioned.

Bibliography Anscombe, Gertrude E.M. 1969. “Modern Moral Philosophy” (1958). In The Is-Ought Question, edited by William D. Hudson, 175-195. London: Macmillan. Ayer, Alfred Jules. 1951. Language, Truth and Logic (1936). London: Victor Gollancz Ltd. Bloomfield, Paul. 2009. “Archimedeanism and Why Metaethics Matters.” In Oxford Studies in Metaethics, edited by Russ Shafer-Landau, Volume Four, 283-302. Oxford: Oxford University Press. Boyd, Richard N. 1988. “How to Be a Moral Realist.” In Essays on Moral Realism, edited by Geoffrey Sayre-McCord, 181-228. Ithaca— London: Cornell University Press. Brandt, Richard B. 1959. Ethical Theory. Englewood Cliffs, N.J.: Prentice-Hall.

Luciana Ceri

167

Brandt, Richard B. 1979. A Theory of the Good and the Right. Oxford: Clarendon Press. Brink, David O. 1989. Moral Realism and the Foundations of Ethics. Cambridge: Cambridge University Press. Carcaterra, Gaetano. 1969. Il problema della fallacia naturalistica. La derivazione del dover essere dall’essere. Milano: Giuffrè. Couture, Jocelyn, and Kai Nielsen. 1995. “Introduction: The Ages of Metaethics.” In On the Relevance of Metaethics. New Essays on Metaethics, edited by Jocelyne Couture and Kai Nielsen, 1-30. Calgary: University of Calgary Press. Dancy, Jonathan. 1993. Moral Reasons. Oxford: Blackwell. —. 2000. Practical Reality. Oxford: Oxford University Press. Darwall, Stephen L. 1992. “Internalism and Agency.” Philosophical Perspectives, 6: 155-174. —. 1995. The British Moralists and the Internal «Ought» 1640-1740. Cambridge: Cambridge University Press. Darwall, Stephen L., Allan Gibbard and Peter Railton. 1992. “Toward Fin de siècle Ethics: Some Trends.” The Philosophical Review, 101: 115189. Foot, Philippa. 2001. Natural Goodness. Oxford: Clarendon Press. —. 2002. Virtues and Vices and Other Essays in Moral Philosophy (1978). Oxford: Clarendon Press. Frankena, William K. 1967. “The Naturalistic Fallacy” (1939). In Theories of Ethics, edited by Philippa Foot, 50-63. Oxford: Oxford University Press. Geach, Peter T. 1972. “Ascriptivism” (1960). In Peter Geach, Logic Matters, 250-254. Berkeley—Los Angeles: University of California Press. —. 1972. “Assertion” (1965). In Peter Geach, Logic Matters, 254-269. Berkeley—Los Angeles: University of California Press. Gibbard, Allan. 1993. “Reply to Railton.” In Naturalism and Normativity, edited by Enrique Villanueva, 67-73. Atascadero, CA: Ridgeview. Gorr, Michael. 1985. “Reason, Impartiality and Utilitarianism.” In Morality and Universality. Essays on Ethical Universalizability, edited by Nelson T. Potter and Mark Timmons, 115-138. Dordrecht: D. Reidel Publishing Company. Hampton, Jean. 1995. “Naturalism and Moral Reasons.” In On the Relevance of Metaethics. New Essays on Metaethics, edited by Jocelyne Couture and Kai Nielsen, 107-133. Calgary: University of Calgary Press.

168

The Debate on Naturalism in Contemporary Ethics

Hare, Richard M. 1952. The Language of Morals. Oxford: Clarendon Press. —. 1963. Freedom and Reason. Oxford: Oxford University Press. —. 1985. “Ontology in Ethics.” In Morality and Objectivity. A Tribute to J.L. Mackie, edited by Ted Honderich, 39-53. London: Routledge and Kegan Paul. —. 1999. “Objective Prescriptions.” (1993). In Richard M. Hare, Objective Prescriptions and Other Essays, 1-18. Oxford: Clarendon Press. —. 1999. “A New Kind of Ethical Naturalism?” (1996). In Richard M. Hare, Objective Prescriptions and Other Essays, 67-86. Oxford: Clarendon Press. Harman, Gilbert. 1977. The Nature of Morality. Introduction to Ethics. New York: Oxford University Press. Horgan, Terry, and Mark Timmons. 1992. “Troubles for New Wave Moral Semantics: The ‘Open Question Argument’ Revived.” Philosophical Papers, 21: 153-175. Humberstone, Lloyd. 1992. “Direction of Fit.” Mind, 101: 59-83. Jackson, Frank. 1998. From Metaphysics to Ethics. A Defence of Conceptual Analysis. Oxford: Clarendon Press. Korsgaard, Christine M. 1996. The Sources of Normativity. Cambridge: Cambridge University Press. Kramer, Matthew H. 2009. Moral Realism as a Moral Doctrine. Wiley: Blackwell Publishers. Kripke, Saul. 1980. Naming and Necessity. Oxford: Blackwell. Lewis, David. 1989. “Dispositional Theories of Value.” Proceedings of the Aristotelian Society, Suppl. Vol. 63: 113-137. Mackie, John L. 1977. Ethics: Inventing Right and Wrong. Harmondsworth: Middlesex, Penguin. Magni, Sergio Filippo. 2004. “Il dibattito contemporaneo sulla natura dell’etica.” Oltrecorrente, 8: 7-24. Marks, Joel. 1986. “The Difference between Motivation and Desire.” In The Ways of Desire. New Essays in Philosophical Psychology on the Concept of Wanting, edited by Joel Marks, 133-147. Chicago: Preceding Publishing. McGinn, Colin. 1997. Ethics, Evil, and Fiction. Oxford: Clarendon Press. Moore, George E. 1903. Principia Ethica. Cambridge: Cambridge University Press. Murdoch, Iris. 1998. Existentialists and Mystics. Writings on Philosophy and Literature, edited by Peter Conradi. New York: Penguin Books.

Luciana Ceri

169

—. 1998. “Vision and Choice in Morality” (1956). In Iris Murdoch, Existentialists and Mystics. Writings on Philosophy and Literature, edited by Peter Conradi, 76-98. New York: Penguin Books. —. 1998. “Metaphysics and Ethics” (1957). In Iris Murdoch, Existentialists and Mystics. Writings on Philosophy and Literature, edited by Peter Conradi, 59-75. New York: Penguin Books. —. 1998. “The Sovereignty of Good Over Other Concepts” (1970). In Iris Murdoch, Existentialists and Mystics. Writings on Philosophy and Literature, edited by Peter Conradi, 363-385. New York: Penguin Books. Pigden, Charles R. 1989. “Logic and the Autonomy of Ethics.” Australasian Journal of Philosophy, 67: 127-151. —. 1991. “Naturalism.” In A Companion to Ethics, edited by Peter Singer, 421-431. Oxford: Basil Blackwell. Putnam, Hilary. 1981. Reason, Truth and History. Cambridge: Cambridge University Press. Railton, Peter. 1986a. “Moral Realism.” Philosophical Review, 95: 163207. —. 1986b. “Facts and Values.” Philosophical Topics, 14: 5-31. —. 1989. “Naturalism and Prescriptivity.” Social Philosophy and Policy, 7: 151-174. Regan, Donald H. 2003. “How to Be a Moorean.” Ethics, 113: 651-677. Rosati, Connie. 2003. “Agency and the Open Question Argument.” Ethics, 113: 490-527. Sayre-McCord, Geoffrey. 1988. “Introduction: The Many Moral Realisms.” In Essays on Moral Realism, edited by Geoffrey SayreMcCord, 1-23. Ithaca-London: Cornell University Press. Scarpelli, Uberto. 1962. Filosofia analitica, norme e valori. Milano: Edizioni di Comunità. —. 1971. “La ‘grande divisione’ e la filosofia della politica.” Introduction to Moral Principles in Political Philosophy (1968), italian translation Etica e filosofia politica, by Felix E. Oppenheim, vii-xxxiv. Bologna: il Mulino. —. 1982. “La meta-etica analitica e la sua rilevanza etica” (1980). In Uberto Scarpelli, L’etica senza verità, 73-112. Bologna: il Mulino. Smith, Michael. 1987. “Should We Believe in Emotivism?.” In Fact, Science and Morality. Essays on A.J. Ayer’s Language, Truth and Logic, edited by Graham Macdonald and Crispin Wright, 289-310. Oxford: Blackwell. —. 1987. “The Humean Theory of Motivation.” Mind, 96: 36-61. —. 1994. The Moral Problem. Oxford: Blackwell.

170

The Debate on Naturalism in Contemporary Ethics

Staude, Mitchell. 1986. “Wanting, Desiring, and Valuing: The Case against Conativism.” In The Ways of Desire. New Essays in Philosophical Psychology on the Concept of Wanting, edited by Joel Marks, 175-195. Chicago: Preceding Publishing. Stevenson, Charles L. 1963. “The Emotive Meaning of Ethical Terms” (1937). In Charles L. Stevenson, Facts and Values. Studies in Ethical Analysis, 10-31. New Haven-London: Yale University Press. Sturgeon, Nicholas L. 1985. “Gibbard on Moral Judgment and Norms.” Ethics, 96: 22-33. —. 1988. “Moral Explanations” (1985). In Essays on Moral Realism, edited by Geoffrey Sayre-McCord, 229-255. Ithaca-London: Cornell University Press. —. 2003. “Moore on Ethical Naturalism.” Ethics, 113: 528-556.

Notes 1

For more on the distinction between substantive and methodological naturalism, see Railton 1989, 155-156, and Hampton 1995, 108. 2 For more on the distinction between analytic and synthetic naturalism, see Hare 1999, 67-68. 3 As regards the distinction between the three kinds of autonomy just mentioned, see Pigden 1991, 425. 4 It is worth noting, however, that Richard Hare, one of the main upholders of this theory of moral language, rejects the term “noncognitivism”. Cognitivism maintains that moral statements can be true or false, and that one can know them to be true or false, and Hare does accept such theory. Yet, he points out that moral statements are not purely descriptive statements; so, the theory according to which moral judgements do not state anything about moral properties or facts should be called “nondescriptivism”. See Hare 1999, 4; see also Hare 1985, 50-52. 5 See also Sturgeon 1985, 25-26; Railton 1986b; Brink 1989, 9, 166. 6 This theory is sometimes called “conativism”, or “noncognitivism”, in contrast to pure cognitivism, according to which the motivation to act does not come from one’s desires, but from one’s cognitive mental states; see Marks 1986, 133, and Staude 1986, 175. 7 Cf. Dancy 1993, 12, and Dancy 2000, 85. 8 For a defense of this theory, see Smith 1987. 9 About the notion of “direction of fit”, see Humberstone 1992; see also Dancy 2000, 78; Smith 1987, 50-56, and Smith 1994, 111-116. 10 For more on the distinction between weak and strong internalism, see Brink 1989, 41-42. 11 See also Hare 1963, 6: “[...] it is because we are free agents that we need to ask prescriptive questions [...].” 12 See Kramer 2009, 2: “Most of the reasons for insisting on the objectivity of ethics are ethical reasons.”

Luciana Ceri 13

171

See also Scarpelli 1971, xi-xiii, xxxii. See also Carcaterra 1969, 578. 15 See Hare 1963, 192: “[...] there is this much of truth in these allegations, that I am a liberal [...]. But it is simply not true that the things which I have said about the logic of moral language are peculiarly tied to any particular moral standpoint. To say that moral and other value-judgements are prescriptive [...] is not, by that alone, to commit oneself to any particular moral opinion.” 16 As regards the link between the argument against naturalism and the Liberal stance see also Murdoch 1998, 66, 68, 70-71. 17 See also Murdoch 1998, 70: according to the naturalist, “freedom is not an open freedom of choice in a clear situation; it lies rather in an increasing knowledge of his own real being, and in the conduct which naturally springs from such knowledge.” 18 See also Murdoch 1998, 67, 70. 19 For a detailed discussion of the relation between metaethics and normative theory, and a defence of the neutrality thesis about metaethics, see Bloomfield 2009. 14

THE OIC/PAP DISPUTE: TWO WAYS OF INTERPRETING ‘OUGHT’ IMPLIES ‘CAN’ GUGLIELMO FEIS (UNIVERSITY OF MILAN)

Abstract I will review the debate concerning the derivation of the principle of alternate possibilities (PAP) starting from the “Ought” implies “Can” thesis (OIC). I propose to explain the varieties of the positions in the debate by singling out two different versions of OIC: the alternate OIC and the commissive OIC. The alternate OIC implies the possibility of choosing something different from the commission or the omission of an action, whereas the commissive OIC implies both the possibility of the commission of the action and the possibility of not doing the action. This allows me to consider a broader spectrum of cases as compared to the standard debate, i.e. cases in which both OICs fail, cases in which both OICs work and intermediate cases in which only one OIC holds. In the last part of the paper I will consider the interplay between these two OICs and reply to some objections.

Introduction In what follows I will approach the topic of the derivation of the principle of alternate possibilities (henceforth PAP) from the thesis according to which “Ought” implies “Can” (henceforth OIC). The former—PAP—is the principle according to which we can hold someone responsible/blameworthy for something she did only if she could have acted otherwise; the latter states the thesis according to which we cannot be asked to do the impossible. Instead of providing a new OIC/PAP derivation or criticize someone else’s derivation, I will try to review the

Guglielmo Feis

173

whole debate by way of adding a new distinction inside OIC itself. I claim that this distinction can help us add new situations to the debate. Addressing more situations is important because this will reveal the complexities of OIC itself, showing that it requires more analysis before we use it as an unquestionable premise of a derivation. The paper is organized as follows: I first (§1) introduce the OIC/PAP debate, taking (§2) a closer look at the two principles (OIC and PAP). In §3 I display my proposal, distinguishing the alternate OIC and the commissive OIC, and in §4 I take a closer look at the new kinds of situation we are enabled to address by way of using the two principles. In the last section (§5), I draw some conclusions.

1. An Overview of the Debate The standard controversy concerning the OIC/PAP derivation started with Frankfurt (1988) who held that there was no derivation between the two theses. Frankfurt did not provide a formal argument but addressed a categorical difference between OIC and PAP: With respect to any action, Kant’s doctrine [OIC] has to do with the agent’s ability to perform that action. PAP, on the other hand, concerns his ability to do something else (Frankfurt 1988, 95-96).

The debate on whether such a thesis was correct or not was held mainly on the journal Analysis—chronologically: Widerker (1991), Yaffe (1999), Blum (2000), Schnall (2001), Fischer (2003), Copp (2008). This debate spread over to other journals and articles—chronologically: Widerker (1995), Copp (1997), Speak (2005), Widerker (2005)—and, at least according to Speak (2005), this kind of debate became part of what is called “PAPistry”.

2. OIC and PAP: a Closer Look at the Two Principles Before going on, I will present: 1) how OIC and PAP are stated in the present debate; 2) what combinations of truth-values for the two principles set the stage for the debate; 3) what the possible positions in the debate are.

174

The OIC/PAP Dispute

2.1 Ought Implies Can (OIC) The aim of the “Ought” implies “Can” (OIC) thesis is to show that there are some situations in which we lack a possibility and in which the act we ought to perform is an impossible act, and such situations can discharge us from an obligation. The best way to get this meaning from the formula is to read it through the law of contraposition, that is to say as “not can” implies “not ought”. In the OIC/PAP debate, there is not much inquiry into the problems of OIC:1 OIC is taken as the starting premise from which you try (not) to derive some sort of PAP.

2.2 Principle of Alternate Possibilities (PAP) The principle of alternate possibilities (PAP) was formulated by Frankfurt (1969: 829) as follows: “a person is responsible for what he has done only if he could have done otherwise.” In the debate there are different formulations of the principle and there is an attempt to construct a taxonomy of the different PAPs and their different strength. The main differences regard the predicate you want to attribute (being responsible or—morally—blameworthy or praiseworthy) and the way in which you evaluate the possibility (you can evaluate possibilities for doing something, for performing some act or even possibilities for having acted or having done otherwise). A whole list of these different PAPs, and the way in which these PAPs enter the derivation will have to be addressed in another paper.

2.3 The Truth-Value Assignments The truth-values for the two different theses are the following: OIC is assumed to be true and PAP—as the result of Frankfurt’s counterexamples that are considered successful—must be considered false. Given these initial assumptions, we have to consider what the consequences of a possible derivation of PAP starting from OIC could be. If we assume that there is a derivation from OIC to PAP we will manage to derive a false statement (PAP) starting from a true one (OIC). This means that we either have to review the truth-values of the two theses or that we have to give up on the derivation between the two theses.

Guglielmo Feis

175

2.3.1 Possible Solutions If we try to show, as Frankfurt did, that there is no derivation from OIC to PAP, we are able to justify the different truth-values assigned to the theses. This is the position held by Frankfurt (1988) and Yaffe (1999). Nonetheless, if we say that such an OIC/PAP derivation is possible, we have to choose between the following alternatives: 1) We can go against OIC by endorsing Frankfurt’s PAP counterexamples: given the derivation from OIC to PAP, we use the counterexamples against PAP as counterexamples against OIC as well. This is the position of Widerker (1991) and Fischer (2003), also Graham (2011: 343) recognized the possibility of such a move; 2) We can support OIC by going against Frankfurt’s PAP counterexamples: given the derivation from OIC to PAP, we extend the truth of OIC also on its consequence (PAP), claiming that this is enough to dismiss Frankfurt’s counterexamples. This is the position of Blum (2000) and Copp (2008). The literature, as I said, concentrates mainly on the steps of the derivation. My proposal is to distinguish two different kinds of OIC theses that are somehow mixed in the debate. My reason for doing so is that this distinction will enable us to distinguish new situations and intermediate cases between Frankfurt’s cases—in which the agent has no control at all on her oughts—and the ordinary OIC situations in which such a control is available.

3. My Proposal: Distinguishing Two New Forms of OIC Here I investigate a less considered topic: the nature and the functions of OIC. I think that, given the variety of argumentations for and against the OIC/PAP derivation, we need to better clarify our starting point. You may say that the different spectrum of conclusions is due to the fact that, as acknowledged before, there are different kinds of PAPs derived from the same OIC, but I think that there is more beyond OIC than previously noticed. Here I propose to distinguish two different kinds of OIC, the alternate OIC and the commissive OIC. Alternate OIC:2 this is the OIC in which ought (A) implies the possibility of choosing something different from the commission or the omission of A. Another way to put it is that ought (A) implies that there is

176

The OIC/PAP Dispute

the possibility to avoid the action that is required by ought (A), i.e. you can avoid the “prescriptive space” of the ought. Commissive OIC: this is the OIC in which ought (A) implies both the possibility of the commission of A and the possibility of not doing A, i.e. of the commission of (not A), for example, by way of omitting A. I think that this distinction is somehow implicit in the OIC/PAP debate. Here I provide some textual reasons for the distinction: 1) Since Frankfurt’s take on OIC and PAP we know that: With respect to any action, Kant’s doctrine [OIC] has to do with the agent’s ability to perform that action. PAP, on the other hand, concerns his ability to do something else (Frankfurt 1988, 95-96);

2) Widerker introduces commissive aspects in what he calls “PAP2”: An agent S is morally blameworthy for performing a given act A only if it is within S’s power not to perform A. (Widerker 1991, 223);

3) Yaffe acknowledges this distinction: We often quite adequately discharge our obligations not to act simply by passively not acting [commissive OIC], and not by actively performing some alternative [alternate OIC]. (Yaffe 2005, 308).3

4) Copp comments on his old formulation that showed commissive aspects: ‘A person is morally required to do A or not to do A in a situation only if she could have done A or she could have avoided A in that situation’ (Copp 2003: 275). More perspicuously, (Maxim-C) (a) A person is all-in morally required to do A in a situation only if she can do A in that situation and (b) a person is all-in morally required not to do A in a situation only if she can do something other than A in that situation. (Copp 2008, 70-71).

Nonetheless, my interest does not lie in the derivation but in the spectrum of examples that the alternate OIC vs. commissive OIC distinction will enable us to consider.

Guglielmo Feis

177

4. Intermediate Cases After distinguishing these two OICs we have to respond to the possible objection according to which the distinction is somehow superfluous. This requires us to answer the following two questions: 1) Are there cases where alternate OIC holds but commissive OIC does not hold? 2) Are there cases where commissive OIC holds but alternate OIC does not hold? Before answering these questions, I think it is useful to apply the alternate/commissive distinction to the cases we already know, namely Frankfurt’s cases and the normal cases. In a normal case such as “do not smoke” both OICs are satisfied: it is possible both to perform the action of smoking (A) and the action of not smoking (not A), as well as it is possible for you to do something else like walking into another room where there is no rule concerning (not) smoking. On the other hand, in Frankfurt’s cases both OICs fail. Commissive OIC fails because you have no possibility of refraining from the performance of A (it is impossible for you not to do A); alternate OIC fails because you do not have any chance of escaping your destiny in performing A or acting otherwise. In the next two paragraphs I will offer situations in which one kind of OIC is available but the other is not.

4.1 Commissive OIC without Alternate OIC: the Diving Case Here I will provide a case in which commissive OIC holds but alternate OIC fails. In this case we need to have access to the performance of both the commission of an action A and of its negation not A, without any chance to escape this request to somehow respond to ought (A). In order to obtain such a case, we just have to add to the normal situation some sort of restriction that will prevent us from running away from the prescriptive space determined by ought (A). You may conceive this case as a weak case of Frankfurt’s: we want to remove the possibility of acting otherwise but it is still possible to both do A and do not A. Nonetheless, there is no need to introduce some sort of Count(erfactual) (inter)Vener, we just need some sort of practical impossibility as defined, for example, by Oppenheim (1991).

178

The OIC/PAP Dispute

Example: I am standing on the diving platform of a swimming pool. Also imagine that there is only one stair to bring people up (you go down diving) and that, given that the platform is fairly high, you ought to dive. I stand on the platform, facing the following obligation: ought (dive). It is possible for me to dive and also not to dive, so commissive OIC holds but I am also facing a practical impossibility to do something else rather than diving or not diving. Given that I have no way out to do something else rather than diving (A) or not diving (not A), alternate OIC fails.

4.2 Alternate OIC without Commissive OIC: the Will-Making Case Here I will provide a case in which alternate OIC holds but commissive OIC fails. The case we need should be one in which it is possible to have a way out concerning the action but, once we choose to do that action, we will be forced to act exactly in that way and not in another. Example: consider the Italian Civil code and the act of will-making. Will-making is ruled by article 602 of the Italian Civil code that states: Art. 602: Il testamento olografo deve essere scritto per intero, datato e sottoscritto di mano del testatore. La sottoscrizione deve essere posta alla fine delle disposizioni. (The holographic will must be written entirely, dated and hand-signed by the testator. The signature must be written at the end of the will dispositions).

In Italy, if you want to finalize your final will, you have no other way of doing it than according to art. 602, i.e. by signing it. When you face the obligation to sign the will in order to make a holographic will, there is no way for you not to sign the will. You do not have access to a way of not A when A is “you ought to handwrite and sign the document in order to make a holographic will.” Without the signature there will be no holographic will, so you will have no access to action not A, i.e. not to sign your will, because there is no such thing as a handwritten but unsigned will. Given this situation, commissive OIC fails. Nonetheless, you can still perform any other sort of action when facing the obligation expressed by ought (sign your will): you have plenty of possible ways out of it, you can keep on living without entering the prescription space of will-making and, even if you face it (as you did when you first read article 602), you can decide not to make any will.

Guglielmo Feis

179

This shows that there are also cases in which commissive OIC fails but alternate OIC is available. I think this shows that the alternate/commissive distinction is worth making, and it enables us to consider a wider variety of examples than the one discussed in the debate. Moreover, we can investigate possible interplays of these different OICs.

4.3 OIC Interplay: on Will-Making Again In the previous paragraphs I showed that we can have situations in which both OICs are satisfied, both OICs fail and in which one OIC is satisfied and the other one fails. Here I want to present a case in which there is an interplay between the two OICs, namely a case in which commissive OIC is used as a test to evaluate some skills that will admit you to a certain situation, an intermediate case where you have only alternate OIC. This may sound paradoxical because it is thanks to the commissive OIC that you can benefit of a possibility in which there is no commissive OIC. Consider again the will-making case of the art. 602 of the Italian Civil code, presented above, and imagine the following situation. You need to have some kind of abilities in order to make a will, i.e. before facing the ought of art. 602 you need to “pass” a certain test in which you show that you possess some abilities. These abilities amount to proving yourself capable of discernment and possessing other mental faculties. In order to evaluate abilities we use commissive OIC: you ought to do A in order to show that you have the relevant ability. With regard to ought (A) you can both do A and not A, so commissive OIC holds. In the case at hand, your being capable of discernment is assessed through commissive OIC. This leads you to access the possibility of making a will, i.e. which is a case in which commissive OIC fails.

5. Conclusion I have presented the OIC/PAP debate and, after framing the positions, I proposed to review the debate introducing a new distinction in the OIC. I then showed that the distinction is not merely verbal but real, because there are cases in which one of the principles holds and the other does not. I also drew textual evidence from part of the debate to motivate the distinction, and I introduced new situations into the debate. A possible objection that may arise is that it is possible to recognize the alternate/commissive distinction but, instead of using it to distinguish two OICs, you may say that the alternate OIC and commissive OIC are PAPs

180

The OIC/PAP Dispute

instead.4 I reply saying that PAP is used to evaluate the subject with respect to the action she did (not) do, while OIC is used directly on actions and obligations. My two OICs issue no evaluation on subjects but allow us to distinguish two different types of contexts of action. There is still some work to be done to see how my OIC distinction can affect the OIC/PAP derivation, and whether the two kinds of intermediate cases allow for new possible derivations. This is the topic of another paper.5

Bibliography Blum, Alex. 2000. “The Kantian versus Frankfurt.” Analysis, 60: 287-88. Copp, David. 1997. “Defending the Principle of Alternate Possibilities: Blameworthiness and Moral Responsibility.” Nous, 31: 441-56. —. 2008. “‘Ought’ Implies ‘Can’ and the Derivation of the Principle of Alternate Possibilities.” Analysis, 68: 67-75. Fischer, John M. 2003. “Ought-Implies-Can, Causal Determinism and Moral Responsibility.” Analysis, 63: 244-50. Frankfurt, Harry G. 1969. “Alternate Possibilities and Moral Responsibility.” Journal of Philosophy, 66: 829-39. —. 1988. “What Are We Morally Responsible For.” In The Importance of What We Care About, Harry G. Frankfurt, 95-103. New York: Cambridge University Press. Martin, Wayne. 2009. “Ought but Cannot.” Proceedings of the Aristotelian Society, 59: 103-128. Oppenheim, Felix. 1991. The Place of Morality in Foreign Politics, Lexington Books, Lexington (Mass.). Ryan, Sharon. 2003. “Doxastic Compatibilism and the Ethics of Belief.” Philosophical Studies, 11: 47-79. Schnall, Ira M. 2001. “The Principle of Alternate Possibilities and ‘Ought’ Implies ‘Can’.” Analysis, 61: 335-40. Sinnott-Armstrong, Walter. 1984. “‘Ought’ Conversationally Implies ‘Can’.” Philosophical Review, 43: 249-61. Speak, Daniel. 2005. “PAPistry: Another Defense.” Midwest Studies in Philosophy, 29: 262-68. Stocker, Michael. 1971. “‘Ought’ and ‘Can’.” Australasian Journal of Philosophy, 49: 303-16. Vranas, Peter B. M. 2007. “I Ought, Therefore I Can.” Philosophical Studies, 136: 167-216. Widerker, David. 1991. “Frankfurt on ‘Ought Implies Can’ and Alternate Possibilities.” Analysis, 51: 222-24.

Guglielmo Feis

181

—. 1995. “Libertarianism and Frankfurt’s Attack on the Principle of Alternate Possibilities.” The Philosophical Review, 104: 247-61. —. 2005. “Blameworthiness, Non-robust Alternates, and the Principle of Alternate Expectations.” Midwest Studies in Philosophy, 29: 292-306. Yaffe, Gideon. 1999. “‘Ought’ Implies ‘Can’ and the Principle of Alternate Possibilities.” Analysis, 59: 218-22. —. 2005. “More on “Ought” Implies “Can” and the Principle of Alternate Possibilities.” Midwest Studies in Philosophy, 29: 307-12.

Notes 1

Nonetheless, see Stoker (1971) Sinnott-Armstrong (1984), Ryan (2003) and Martin (2009) for criticism and Vranas (2007) for some solutions. 2 I will call it ‘alternate OIC’ rather than ‘alternative OIC’ to stress the connection with Frankfurt’s PAP definition. 3 I added the [commissive OIC] and [alternate OIC] to show where I think my distinction is somehow present in Yaffe’s paper. 4 I received this remark from Gideon Yaffe in private correspondence. 5 I would like to thank Wojciech ĩeáaniec and Federico Faroldi for comments and observations on a previous version of this paper as well as all the participants and organizers of the “Workshop on Ethics” held at the Università Statale di Milano.

COGNITIVE NEUROSCIENCE AND ANIMAL CONSCIOUSNESS MATTEO GRASSO (ROMA TRE UNIVERSITY)

Abstract The problem of animal consciousness has profound implications on our concept of nature and of our place in the natural world. In philosophy of mind and cognitive neuroscience the problem of animal consciousness raises two main questions (Velmans, 2007): the distribution question (“are there conscious animals beside humans?”) and the phenomenological question (“what is it like to be a non-human animal?”). In order to answer these questions, many approaches take into account similarities and dissimilarities in animal and human behaviour, e.g. the use of language or tools and self-recognition in a mirror (Allen and Bekoff, 2007); however, behavioural arguments do not seem to be conclusive (Baars, 2005). Cognitive neuroscience is providing comparative data on structural and functional similarities, respectively called “homologies” and “analogies”. Many experimental results suggest that the thalamocortical system is essential for consciousness (Edelman and Tononi, 2000; Tononi, 2008). The argument from homology states that the general structure of the thalamocortical system remained the same in the last 100-200 million years, because it is neuroanatomically similar in all the present and past mammals, and it did not change much during phylogeny (Allen and Bekoff, 2007). The argument from analogy states that the key functional processes correlated with consciousness in humans are also present in all other mammals and many other animals (Baars, 2005). These processes are information integration through effective cortical connectivity (Massimini et al., 2005; Rosanova et al., 2012), and elaboration of information at a global level (Dehaene and Changeux, 2011). On this basis, the “Cambridge Declaration on Consciousness” stated that all mammals, birds, and many other animals (such as octopuses) possess the neurological substrates of consciousness (Low et al., 2012).

Matteo Grasso

183

Conscious experience is private (Chalmers, 1995; Nagel, 1974), therefore the answer to the phenomenological question may be impossible. Nevertheless, cognitive neuroscience may provide an answer to the distribution question, showing that conscious experience is not limited to humans since it is a major biological adaptation going back millions of years.

1. Two Questions on Animal Consciousness Consciousness is the most intimate and patent aspect of the experiences we undergo daily. From when we wake up to when, lying in bed, our senses are blurred and we sink into sleep, consciousness is a matter of fact, and so it seems perfectly obvious to us. In our lives we are spontaneously able to establish whether other people are conscious or not: we recognize when someone faints from fear, we are aware of pathological conditions characterized by the loss of consciousness such as coma, and we rely on the anaesthetist before an important operation so that the drugs administered erase, together with consciousness, all the pain and the unpleasant memories of the surgery. Despite this daily competence, it is hard to explain the criteria we use to determine whether an individual is actually conscious or not. Most of the times we just have to rely on observed behaviour: individuals who interact with the exterior environment and with other people are evidently conscious, especially if they are able to react consistently to stimulations and respond to our questions in a meaningful way. However, an objective judgment given from an external point of view, from the “third person”, is not enough: would we say that sleepwalkers are conscious just because they interact with the external environment? In an intuitive and simplistic way, we could say that individuals are conscious only if it is something to be them. Given that during deep dreamless sleep, or under general anaesthesia, we do not feel anything, we could say that in such cases consciousness disappears completely. However, the ability to evaluate consciousness might go beyond the limits of our own species: are animals such as dogs and cats conscious, too? They show complex behaviour, we empathize with them, we believe they understand our emotions and our feelings and many times we even talk to them. Are animals phylogenetically more distant from us (such as birds and fishes) conscious as well? Where does the boundary lie between animals that are conscious and animals that are not? With respect to these questions, our intuitions are useless and unfit, and they find us totally unprepared and unable to provide even the slightest response. Precisely

184

Cognitive Neuroscience and Animal Consciousness

because of the elusiveness of consciousness, even scientists and philosophers are in deep water when considering these issues. The problem of animal consciousness has profound implications for the concept of nature and the place of man within it. For a long time, anthropocentrism has been characterizing the view that humankind has of itself as a privileged species at the top of the so-called Scala naturae, as was considered to be the only being endowed with intellect, soul or simply mind and consciousness. The problem of animal consciousness comes up again, now more than ever, in the philosophical and scientific debate, bringing along several implications about the centrality of humankind in the natural world as well as ethical implications. In fact, many animal species are used in research and industry, and many of them are subject to intensive farming. The assumption that our ethical obligations towards animals depend on their mental life is quite common and well-grounded in intuition (Farah 2008), and the ethical issue concerning animals is of primary importance. In fact, the rejection of the hypothesis of animal consciousness seems less and less grounded, given the increasing amount of scientific models of affective and emotional experiences both in humans and other animals (Panksepp 2005). In recent years, two fundamental questions about animal consciousness have been at the centre of a big debate, involving both philosophy of mind and cognitive neuroscience (Allen and Beckoff 2007). The first is the socalled distribution question (DQ): (DQ): Are there conscious animals beside humans?

Many animal species share the ability to respond to environmental stimuli with intelligent behaviour, and the organisms phylogenetically closer to Homo sapiens share with us several anatomical structures. However, there is no objective criterion available for determining with certainty whether there are other conscious animals besides humans. Consciousness might in fact be a common feature of many species, including the class of mammals, or the entire group of vertebrates; nevertheless, without a reliable criterion we cannot establish any useful test to distinguish between conscious and unconscious organisms. Being conscious is not just being awake and reactive, or being able to respond to environmental stimuli and show intelligent behaviour. Being conscious also means having experience of the world, of the actual shape of a flower, the colours in the sky, the smells and flavours of food; being conscious means having desires, feeling pain from the wounds, feeling fear of predators, and so on. Thus, there is a second important question on

Matteo Grasso

185

animal consciousness that focuses on the qualitative aspects of conscious experience and is expressed in the form of the so-called phenomenological question (PQ): (PQ): Can we know what, if anything, the experiences of those animals are like?

Many philosophers think that consciousness is the last surviving mystery, and that science cannot face up to the problem of consciousness because of its subjective and phenomenal aspects. According to Chalmers (1995), conscious experience constitutes the very and only hard problem that science has to tackle. Both the hard problem and the PQ are based on the concept of phenomenal consciousness, introduced by Ned Block (1995) and widely debated in the philosophy of mind. This concept refers to the subjective and qualitative nature of each experiential state of consciousness, whether it belongs to humans or not. What does a wasp feel when glancing at the colours of a flower? What does a cow experience while looking into the eyes of the farmer who milks her daily? Can we make a comparison between the conscious and subjective experience of being a man with the experience of being another animal? These questions are loaded with both metaphysical and ethical implications, and these issues constitute the core of deep philosophical reflections. The DQ relates to the issue of the centrality and superiority of mankind in the natural world at its core and the question admits two contrasting answers: “Yes, there are conscious animals besides humans” or “No, there are not”. Two types of theories about the distribution of consciousness in nature have been proposed in order to answer the DQ (Velmans 2007), namely discontinuity and continuity theories. Discontinuity theories state that consciousness emerged at a certain stage of the phylogeny. The models that equate this stage with the evolution of the genus Homo usually reply negatively, claiming that only humans are endowed with consciousness and that the distribution of consciousness is discontinuous, since it belongs only to them. The models which instead place such a gap elsewhere are inclined to reply in a positive way, claiming that in nature there are many living organisms endowed with consciousness. According to this view, this phenomenon has emerged at a specific stage of phylogeny and it is a shared feature of either the entire taxon of vertebrates, or the sole group of mammals, or the primates alone, depending on the particular view. In general, discontinuity theories share a vision of nature that presents jumps and in which the living organisms endowed with consciousness are sharply divided from all the others. As we will see in the following paragraphs,

186

Cognitive Neuroscience and Animal Consciousness

many discontinuity models have recently been proposed. On the contrary, continuity theories always reply “Yes, there are conscious animals besides humans”; indeed, they think there are plenty of them. Continuity theories argue that consciousness is a property that has always characterized living beings, and that it is distributed in different degrees throughout the natural world. These theories often embrace a form of panpsychism, arguing that consciousness is a fundamental property that characterizes, to a certain extent, even matter itself, albeit at low and irrelevant levels, and that it co-evolves in complexity with matter and its aggregates (Chalmers 1995; Velmans 2007; Whitehead 1929). Among the many proposals aimed at addressing the distribution question, some are based on the identification of similarities and dissimilarities in the behaviour of humans and other animals, but behavioural observations do not seem to be a reliable and satisfactory criterion (Baars 2005). However, the two questions about animal consciousness have caused even more discussion since additional nonbehavioural arguments appeared in the scientific and philosophical debate. In recent years, cognitive neuroscience has been shedding light on the neural structures essential for consciousness in humans, raising several questions about the distribution of consciousness in nature and strongly influencing the debate on animal consciousness. A new neuroscientific line of argument is based on the similarities found at the neural level in species different from Homo sapiens, and it is deeply connected with the identification of those structures and processes that constitute the neural substrate of consciousness in human beings. In this article, I will firstly discuss the most important examples of behavioural arguments and I will argue that such arguments are unsatisfactory. Then, I will show how cognitive neuroscience has contributed so far to the study of animal consciousness, and I will show the conclusions that can be drawn in the light of these findings. Finally, I will argue that cognitive neuroscience is the only reliable base to answer philosophical questions such as the distribution question. Nevertheless, neuroscientific progress cannot lead to the solution of philosophical issues related to subjective conscious experience, since it cannot contribute to the formulation of an answer to the phenomenological question.

2. From Behaviour to Consciousness Whenever we have to give a reason why we consider a human conscious as opposed to, say, a computer, we most likely begin with noticing the following: a human being is capable of complex behaviour

Matteo Grasso

187

that goes well beyond the mechanical and automatic processes a computer is able to carry out with flawless precision. The observed behaviour is one of the most important criteria, so much that artificial intelligence research, since its origins, was based on the idea that building a computer able to implement typically human behaviours such as language or reasoning would correspond to recreating mental processes similar to those that occur in the human brain. Indeed, one of the fathers of artificial intelligence, Alan Turing, considered the behavioural criterion as the most reliable basis for developing a test of machine intelligence, now known as the Turing test (Turing 1950). In the same way, the debate has always been focused primarily on behavioural observations also with respect to the issue of animal consciousness, and many arguments based on the similarities between the behaviour of humans and other animals have been proposed (Allen and Beckoff 2007). One of the most evident abilities that characterize humans is the use of language. Our species, through the use of historical-natural languages, has been able to communicate and build the characteristic complex network of social ties. Some philosophers, such as Descartes, argued that the use of articulate languages is a clear sign of the ability to reason, and therefore constitutes the basis for inferring the presence of consciousness in other living beings (Descartes 1984-91). Humans, however, are not the only animals that use language. There have been many instances of non-human animals able to use a language in order to communicate. Professor Irene Pepperberg spent decades studying the linguistic skills of an especially talented African grey parrot (Psittacus erithacus). The grey parrot Alex showed many similarities with the use of language typical of human beings, in stark contrast to the Cartesian idea that animals are not equipped with language, or are only capable of a very limited use of it. Alex had a vocabulary of a hundred and fifty words (similar to that of a two-year old child), he could identify fifty different objects and recognize quantities up to six, he could distinguish seven colours and five shapes, he understood the concepts of zero and said “I want to go back to my cage” when he was tired (Pepperberg 2002). Linguistic skills have also been found in many other animal species besides African grey parrots, in the form of the capacity to use a modified sign language or symbols of the Yerkes Laboratory keyboard system (chimpanzees, bonobos), or alarm calls (lemurs, vervet monkeys, social mongooses and prairie dogs) (Griffin and Speck 2004). The use of language, although largely correlated with the complex cognitive processes Homo sapiens is capable of, might not be certain evidence of the presence of consciousness. In fact, discussing the example

188

Cognitive Neuroscience and Animal Consciousness

of parrots, Descartes (1984-91) argued that any process put in place by these and other animals is in principle conceivable as purely mechanical. Parrots, in this case, could talk without being aware of it, and this would also explain the limited use of language that distinguishes all other species compared to humans. Besides the use of language, the use of tools is often considered a second indication of intelligence and complex mental faculties. Once again, the animal kingdom presents many examples of species that show this capability. Among the various species, certain crows represent perhaps the most interesting example. New Caledonian crows (Corvus moneduloides) show the ability of “crafting” tools, such as wooden hooks, that they subsequently use to probe for invertebrates in crevices. In several experiments it was observed that these crows, when placed in proximity of a 30-cm-long cylindrical container with a single opening and containing a favourite food, are able to immediately retrieve one of the sticks available in the cage and, holding it with their beak, slide it into the cylindrical tube to extract the food (Hunt and Gray 2004). In nature, the crows insert sticks into cavities and drag out the preys that would be difficult or impossible to flush out otherwise. If faced with a vertical cylinder containing a bucket of seeds, such crows are able to grasp with their beak a metal stick, bend it leveraging on one end so as to fabricate a hook, and use it to extract the bucket of seeds from the cylinder (Weir, Chappell and Kacelnik 2002). Despite the ability to communicate and use languages or tools, further types of behaviour might seem to be solid evidence of the presence of consciousness in non-human animals. Less than fifty years ago, in order to ascertain the presence of self-awareness and meta-consciousness, Professor Gordon Gallup devised the so-called “mirror test”. It has long been known that species phylogenetically close to Homo sapiens, such as chimpanzees, are very attracted by the use of mirrors, and contemplate themselves in their own reflected image. Starting from this evidence, Gallup has developed the mirror test in order to determine whether these animals are actually able to recognize themselves in the reflected image, or they are interested in it only because they recognize the shape of one of their own kind. For this purpose, Gallup put several chimpanzees in front of a mirror, leaving them contemplating their own reflected image for a few days. After an initial period in which the animals responded by attacking the image or getting frightened, they eventually got used to it and began to simply contemplate the reflections in the mirror. At that point, Gallup marked the bodies of the animals with odourless paint , in a place visible only in the mirror. The animals showed evidence of understanding that the stain was on their own body, and would in fact turn

Matteo Grasso

189

the stained part of the body towards the mirror for a better look and inspected the same body part with a limb while looking in the mirror (Gallup 1970). According to Gallup, mirror self-recognition is an indicator of self-awareness. Furthermore, he claims that the ability to infer the existence of other individuals’ mental states (namely, having a “theory of mind”) is a by-product of self-awareness. He describes the connection between self-awareness and theory of mind by saying that, being selfaware, an individual is in a position to use his own mental experience to infer and model the existence of comparable mental and intentional processes in others (Gallup et al. 2002). It is clear that much additional work will be required before a complete report. However, a list of the animals that have passed the mirror test so far includes humans (from the age of 18 months) as well as all the great apes such as chimpanzees, gorillas, bonobos and orangutans (Miller 2009), some marine mammals such as bottlenose dolphins (Marten and Psarakos 1995) and orca whales (Delfour and Marten 2001), but also elephants (Plotnik, de Waal and Reiss 2006) and European magpies (Prior, Schwarz and Güntürkün 2008). Thus, many approaches take into account similarities and dissimilarities in animal and human behaviour. Nonetheless, behavioural arguments do not seem to be conclusive (Baars 2005). Even though behaviour might appear as a reliable criterion for assessing the state of consciousness in humans as well, recent developments in cognitive neuroscience have shown that there is no substantial proof. For instance, neurology and coma science have provided many examples in which the assessment of the level of consciousness is a puzzling clinical issue. This is the case of noncommunicative patients who, despite being perfectly aware, are unable to prove it by means of any observable behaviour. Indeed, neuroscience has recently started providing methods and tools for assessing consciousness that are not behaviour-based (Rosanova et al. 2012). These tools are proving their usefulness in the diagnosis and treatment of disorders of consciousness (and related conditions) such as the locked-in syndrome, and for the reasons outlined so far they might be very useful in the study of animal consciousness as well.

3. Neuroanatomy and Neurophysiology of Human Consciousness The study of animal consciousness is the centre of a renewed interest because of the appearance of a new series of arguments. These arguments are based not on the observed behaviour, but on the animals’ brain

190

Cognitive Neuroscience and Animal Consciousness

structures and processes, which are similar to human ones, that can be investigated using the tools of neuroscience. Cognitive neuroscience is gradually leading to a better understanding of the neural correlates of consciousness in humans and, in doing so, it is offering at the same time the opportunity to start investigating consciousness in other animals on a completely new basis. In recent years, several theoretical and experimental models attempted to identify the structures and processes responsible for consciousness in humans. On the one hand, it is generally believed that the basic structure for the emergence of consciousness is mainly the thalamocortical system as a whole, with a specifically important role played by the prefrontal, cingulate and parietal cortices (Edelman and Tononi 2000; Laureys and Tononi 2009). On the other hand, besides anatomical structures, the fundamental processes of consciousness are cortical connectivity through information integration (Tononi 2008; Edelman and Seth 2009) and elaboration of information shared and processed at a global level (Dehaene and Changeux 2011; Tononi and Koch 2008). The 100 billion neurons that constitute the central nervous system are organized into different structures that are important for the emergence of consciousness. Considering only the cerebral mass and the number of neurons, for example, one could think that consciousness emerges from the interaction of an exorbitant number of cells that are able to exchange electrical signals by means of an even greater number of intricate synaptic connections. However, extensive neurological evidence has shown that consciousness does not depend exclusively on the number of neurons. The 100 billion neurons contained in the central nervous system are divided into two distinguishable structures: the thalamocortical system and the cerebellum. The cerebellum is a single body with a complex structure and consists of about 70 billion neurons, while the thalamocortical system is composed of about 21 billion nerve cells (Pakkenberg et al. 2003; Llinàs, Walton and Lang 2004). Despite their similarity, these two structures contribute in a very different way to consciousness because, contrary to the cerebellum, the thalamocortical system is crucial in supporting consciousness (Edelman and Tononi 2000). The thalamocortical system is composed of the thalamus, a deep structure located in the diencephalon, and the cerebral cortex, the thin layer of neurons that constitutes the outer (and most recent) part of the telencephalon in vertebrates. This structure is considered responsible for complex cognitive processes such as memory, perception, thought, and language. As many cases studied by clinical neuropsychology demonstrate, injuries to this structure of the central nervous system can cause the loss of cognitive functions such as

Matteo Grasso

191

perceptive abilities, memory, thinking and reasoning, as well as the complete loss of consciousness (Laureys and Tononi 2009). Even a slight damage to some areas of the thalamocortical system (such as the brainstem or the ascending reticular activating system) can cause a total loss of consciousness, leading the subject into a state of coma, vegetative state or even brain death (Vincent 2000). The cerebellum too is responsible for important functions, such as motion control and motor learning, but this structure does not seem to play any significant role with respect to conscious experience. In fact, as a result of “cerebellectomy”, a surgery which consists in the total removal of the cerebellum, patients show serious deficits in posture and gait control, linked to the impaired control of limbs and eye movements (cerebellar ataxia); however, none of the cognitive functions seem to be damaged in these patients (Tononi 2008). In fact, the cerebral cortex and the cerebellar cortex have a very different structural organization: the modules of the cerebral cortex, as highly specialized, are abundantly connected to each other, while individual modules in the cerebellar cortex tend to be activated independently of one another, with little long-range interaction possible between distant modules (Tononi 2004, 2008). This shows that the integration of information through a complex network of long-range connections is fundamental for the emergence of consciousness. Recently, Stanislas Dehaene proposed the “Global Neuronal Workspace” model in order to explain which fundamental property is responsible for the emergence of consciousness from the thalamocortical system. Dehaene and his colleagues argue that the thalamocortical system is responsible for the emergence of consciousness as composed of a particular structure, the “global neuronal workspace”, consisting in a group of cortical pyramidal neurons with excitatory function and with long-range cross-cortical axons. A high density of this cell type is located in the prefrontal, cingulate, and parietal cortices, and, according to this model, together they would form a neuronal “workspace” connecting a large number of peripheral processing systems and specialized modules, which otherwise always work isolated and on a subconscious level (Dehaene and Changeux 2011). In this model, the prefrontal cortex is a root node of a global network formed by long-distance synaptic connections. This assumption is confirmed by the fact that the prefrontal cortex is connected with a number of very different sensory areas through cortico-cortical connections, and this massive two-ways connectivity justifies the important role played by this particular structure in the hypothesis of a global workspace connecting many areas of the thalamocortical system and supporting consciousness.

192

Cognitive Neuroscience and Animal Consciousness

There is very strong evidence that the thalamocortical system supports consciousness. However, although the presence of such brain structures is a necessary condition, it is not yet sufficient in attributing consciousness to a subject. We are witnesses to the fact that every night consciousness disappears when we fall asleep in a dreamless sleep, but of course the anatomy of our nervous system does not change from day to night. The changes responsible for this daily loss of consciousness are instead due to the availability of neurotransmitters and the particular activity pattern of the billions of neurons that form the thalamocortical system. As shown by electroencephalographic (EEG) measurements, during stage 4 of NREM sleep the firing rate of many cortical and sub-cortical neurons is characterized by oscillatory synchrony and bistability: neuromodulatory changes (e.g. low acetylcholine) trigger a modification in the intrinsic and synaptic conductance, therefore cortical neurons enter every few hundred milliseconds in a hyperpolarized down-state and cannot sustain firing for a short period of time. Shortly afterward, they inevitably return to a depolarized up-state and start their action potential activity again (Steriade, Timofeev and Grenier 2001). According to Giulio Tononi’s “Integrated Information Theory”, consciousness fades in deep sleep because of a decrease in the capacity to integrate information that characterizes the whole thalamocortical system in this state. This capacity is greatly reduced in deep NREM sleep because the thalamocortical system breaks down into causally independent modules and shrinks its repertoire of possible responses, therefore losing effective cortical connectivity and the capability to elaborate high amounts of information at a global level (Tononi and Massimini 2008; Massimini et al. 2005). Moreover, the importance of the thalamocortical system is also shown by the fact that several consciousness disorders are due to neurological damage to this structure. The most common disorders that produce at least transient coma are structural brain lesions (cortical or white matter damage, and brainstem lesions), metabolic and nutritional disorders, exogenous toxins, central nervous system infections such as septic illness, seizures, temperature-related disorders (hypothermia or hyperthermia), and trauma (Laureys and Tononi 2009). All these injuries or conditions interfere with the ascending reticular activating system (ARAS). The ARAS plays a central role in disorders of consciousness because it is the system responsible for the arousal function. As it is represented by structures in the brainstem, the diencephalon and projections to the cerebral cortex (Vincent 2000), the state of unarousable unconsciousness named “coma” is the effect of an impairment in this system (Laureys and Tononi 2009). In addition, recent studies show impaired functional

Matteo Grasso

193

connections between distant cortical areas and between the thalami and the cortex in vegetative patients (Laureys et al. 2002). More importantly, these studies suggest that a restoration of this cortico-thalamo-cortical interaction causes a recovery of cortical effective connectivity (Rosanova et al. 2012) and therefore the recovery of consciousness (Laureys et al. 2000). Given the base these findings constitute, is it possible to deal with the problem of determining the distribution of consciousness in nature? How confident can we be in determining, on the chronological and phylogenetic tree of life, the position of the line that separates conscious organisms from unconscious ones?

4. Homologies and Analogies Cognitive neuroscientists have provided comparative data on structural and functional similarities between different species, respectively called “homologies” and “analogies”. Homologies are phenotypical characters, namely bodily structures, similar in different species because they were inherited from a common ancestor possessing those characters. Analogies are phenotypical characters similar in different species because of convergent evolution instead of common ancestry. These characters evolved independently but are thought to be analogous for the similar function they play and were selected for. The arguments based on neural homologies state that the general structure of the thalamocortical system remained the same in the last few hundred million years, for it is anatomically similar in all currently existing and past mammals and did not change much during phylogeny. Moreover, these arguments state that the key functional processes correlated with consciousness in humans are also present in all other mammals (Edelman, Baars and Seth 2005). According to this hypothesis, consciousness is a biological adaptation dating back many millions of years (Baars 2005). Even on the basis of neuroscientific arguments, however, it is still very hard to formulate a precise answer to the question about the distribution of consciousness in the natural world. To this day, many hypotheses have been proposed about consciousness in non-human animals, covering different taxa such as mammals, vertebrates and invertebrates as well. According to Baars (2005), the comparison between the thalamocortical structures of man and those of other mammals suggests the possibility of attributing consciousness to this whole category of living organisms: all mammals, in fact, have a highly developed thalamocortical system. The study of cranial fossil remains and gene conservation across species

194

Cognitive Neuroscience and Animal Consciousness

suggests, moreover, that the fundamental structures of the thalamocortical system have not undergone major changes over the last 100-200 million years. Furthermore, it is possible to detect in all mammals the same EEG activation patterns observed in man, so that the results of EEG studies on mammals are often directly applied to humans. During the waking state, the EEG of all mammals shows low-voltage, fast and asynchronous neural activity throughout the whole thalamocortical system. On the contrary, during deep sleep the EEG reveals slow, synchronous and high-voltage activity. Moreover, the wake-sleep cycle has many similarities among all mammals: with a few exceptions such as whales, most mammals have both NREM sleep and REM sleep—a stage of sleep characterized by the rapid and random movement of the eyes and the higher presence of dreams. The variation of electrical brain activity in the transition between waking and sleep states is one of the features that humans and other mammals have in common: in humans synchronous oscillatory activity corresponds to the lack of consciousness, as shown during deep sleep, general anaesthesia and in other pathological conditions such as coma and epileptic seizures, and this kind of activity contrasts sharply with the presence of consciousness during the waking state. The distinction between waking and sleep therefore constitutes one of the most reliable criteria to confirm the presence or absence of consciousness in humans and other animals (Baars 2005). Homologies are necessarily remoter in non-mammals, which do not share the mammalian complex thalamocortical system. Avian species exhibit a broad range of complex behaviours, like the use of tools by New Caledonian crows, but what do neuroanatomical studies tell us about it? The overall organization of the central nervous system can be traced back to some 520 million years ago and it seems a common feature of some lower vertebrates such as reptiles, as well as birds and mammals (Smith 1999). Birds have anatomical structures that represent homologies and analogies compared to the thalamus and to the cerebral cortex of mammals: in the avian dorsal pallium, for example, the somatomotor circuitry seems a clear homology of the mammalian basal ganglia-corticothalamic loop (Medina and Reiner 2000). Moreover, they have waking EEG patterns similar to those detectable in mammals during the waking state (Edelman et al. 2005). Nevertheless, EEG patterns of avians during sleep are different from those of mammals, although their sleep is characterized by both REM and NREM sleep (Ayala-Guerrero 1989). Since reptiles show less homologies and analogies with brain structures and processes that are fundamental for consciousness in humans and other

Matteo Grasso

195

mammals, it is unclear whether reptiles are capable of conscious awareness. However, mammalian and avian species show homologous structures or similar analogous arrangements, suggesting the hypothesis that consciousness emerged approximately 300 million years ago, probably twice and in an independent way, after the divergence of the two reptilian lines that led alternatively to birds and mammals (Kardong 1995; Edelman et al. 2005). Further studies about the possibility of attributing at least primary consciousness to other organisms belonging to invertebrate species are currently being discussed, and many of them focus on cephalopods like the octopus (Merker 2005). Octopuses (Octopus vulgaris) possess refined cognitive capabilities. They show the ability of classifying different shaped objects in the same manner as vertebrates such as rats do. They show also memory, learning and complex decision-making skills. Cephalopods like the octopus have complex nervous systems similar to those of some vertebrates, at least with respect to the number of constituent neurons alone: the relative brain size of many cephalopods exceeds that of many lower vertebrates (Hanlon and Messenger 2002). Nevertheless, contrary to birds, cephalopods show a different organization of the nervous system that poses many problems in identifying the necessary structures underlying consciousness (Edelman et al. 2005). Moreover, a significant fact is that many cephalopods show EEG patterns akin to those present in awake and conscious vertebrates (Bullock and Budelmann 1991), and they are the only invertebrates in which this was demonstrated, besides fruit flies. So far, experimental evidence has been insufficient to draw any definitive answer to the DQ. However, as we increase our knowledge we may be able to make more profound and reliable predictions that apply not only to some birds and reptiles, but also to large-brained invertebrates and possibly to many other species (Seth and Baars 2005).

5. The Cambridge Declaration on Consciousness Recently, the hypothesis of animal consciousness has given rise to a huge debate. Philosophers and scientists of every background and orientation are facing up to the related theoretical issues, and they are discussing the experimental evidence that has gradually accumulated. However, the problem of animal consciousness might not remain unresolved. In fact, as I showed so far in this paper, consciousness studies have increased dramatically over the past few decades. A huge amount of brand new data is available today and might shape reflection on this issue.

196

Cognitive Neuroscience and Animal Consciousness

2012 was undeniably an important year for the studies on animal consciousness because it witnessed an important event. A group of eminent scientists from different countries met and signed a document of central importance for the animal consciousness debate: the “Cambridge Declaration on Consciousness” (Low et al. 2012). The team of scientists included neuroscientists, neurophysiologists, neuroanatomists, neuropharmacologists and computational neuroscientists. The Declaration was proclaimed in Cambridge, UK, on July 7, 2012, at the Francis Crick Memorial Conference on Consciousness in Human and Non-Human Animals, at the Churchill College, University of Cambridge, by Philip Low, David Edelman and Christof Koch. In this two-page document, these prominent neuroscientists summarized many experimental observations and made several assertions that can be now unequivocally claimed about the matter of animal consciousness. Firstly, the Declaration states that new techniques are available for the study of the neural correlates of consciousness in humans and non-human animals, and these studies have shown that it is possible to identify homologous brain circuits whose activity correlates with conscious experience, and which can be selectively facilitated or disrupted to determine whether they are necessary for conscious experience or not. According to the authors, a lot of evidence indicates that non-human animals possess the neuroanatomical, neurochemical and neurophysiological substrates that sustain consciousness and the ability to carry on complex intentional behaviours. The neural circuits that support attention, the sleep-wake cycle and decision-making are found in many animals, such as insects and cephalopods, and date back to the invertebrate radiation. Birds, in particular, represent a remarkable case of parallel evolution of consciousness, because they show evidence of human-like levels of consciousness in their behaviour, supported by strong similarities in the neurophysiological processes and neuroanatomical structures of their nervous system. In conclusion, the subscribers of the “Cambridge Declaration on Consciousness” clearly state the following: Convergent evidence indicates that non-human animals have the neuroanatomical, neurochemical, and neurophysiological substrates of conscious states along with the capacity to exhibit intentional behaviors. Consequently, the weight of evidence indicates that humans are not unique in possessing the neurological substrates that generate consciousness. Nonhuman animals, including all mammals and birds, and many other creatures, including octopuses, also possess these neurological substrates. (Low et al. 2012, 2).

Matteo Grasso

197

6. What Is It Like To Be a Bat? The phenomenological question focuses on the concept of “phenomenal consciousness”, namely the subjective and qualitative aspects of conscious experience, on the “first-person” point of view that we would not ascribe, for example, to a robot only because it seems capable of complex calculations and responses. What is it like to be a non-human animal? It is hard to say, first of all because animals do not talk and cannot tell us, or at least because we do not know how to communicate with them. In 1974 Thomas Nagel put the emphasis on this aspect in his famous article What is it like to be a bat?, in which he argued that conscious experience is different from every other property of the natural world because it is private. According to the author, even granting the fact that other animals might feel something, we will never know what it is like to be one of them. Indeed, even if one could imagine it, he would know what it is like to be a bat for a man, and not what it is like to be a bat for a bat (Nagel 1974). The privateness of others’ mental experience outlines a limit so clear and insurmountable that it could cast doubts on the very fact that it is really something to be another animal. However, this privateness does not only characterize the mental states of non-human animals, but also the conscious experience of our own fellow conspecifics. The philosophical problem of “other minds” leads to the sceptical conclusion that, by observing the behaviour of other human beings, we could never conclude that they, too, have a conscious mind (Nagel 1986). Only the subject himself can be sure of possessing a mind and of being conscious, as he alone has privileged access and experiences first-hand the fact of being conscious. For this reason, the privateness of conscious experience and the problem of other minds lead to two alternative ways to answer the question: the first is the sceptical and solipsistic conclusion that each subject believes to be the only one having for sure a conscious mind, while the second consists instead on relying on an inference to the best explanation admitting that, despite the unbridgeable gap between the first and the third person point of view, there are good reasons to believe that there are other conscious minds as well besides that of the subject himself. If you accept the second view, and this is the most interesting point, the conscious experience of other humans would not seem more accessible than that of non-human animals. Therefore, the reasons we have to infer that our fellow humans are conscious, since they share the same anatomical structures and implement the same types of behaviour, constitute the basis for inferring that many non-human animals are equally

198

Cognitive Neuroscience and Animal Consciousness

conscious. Thus, the privateness of conscious experience does not limit the possibility of answering the question about the distribution of consciousness in the natural world. It is clear that much additional work will be required in order to identify which (and to what extent) non-human animals are conscious, but it seems that the distribution question will have in any case a positive answer. However, the privateness of conscious experience will probably remain an insurmountable limit to our understanding of what it is like to be such other animals, leaving the phenomenological question completely unanswered.

7. Conclusions The problem of animal consciousness is perhaps one of the most relevant issues that philosophy of mind and cognitive neuroscience are facing today. The new research paradigm based on neuroscience has achieved important results in just a few decades, and this seems to be just the beginning of an era full of new and astonishing results. Cognitive neuroscience is broadening our understanding of the neural basis of consciousness in man, through the study of consciousness in normal conditions as waking and sleeping, and it is shedding light on which mechanisms regulate its disappearance in pathological conditions such as coma, providing a basis to improve the diagnostic process and the clinical treatment of such diseases and conditions. In the past, the debate on animal consciousness was mainly based on behavioural observations, and many arguments for or against animal consciousness were grounded in behavioural similarities and dissimilarities of other animals with respect to human beings. As we have seen, human consciousness studies today, constitute the basis for a new type of arguments, based no longer on behavioural observation but on the direct comparison between different kinds of anatomical structures and brain processes that, in humans, are known to give rise to consciousness. The neuroscientific research of interspecific homologies and analogies is suggesting that man is not the only animal endowed with consciousness. The basic structure for its emergence, i.e. the thalamocortical system, was already present for about 150-200 million years before Homo sapiens appeared, and characterizes the entire class of mammals. Moreover, in many animals we can find the same brain processes that are essential for consciousness in man (such as the presence of electrical brain activity in distinct stages of waking, NREM and REM sleep), and this evidence suggests that animals distant from man from a phylogenetic point of view,

Matteo Grasso

199

like birds, or even invertebrates such as cephalopods, possess at least a primary form of consciousness. Given that conscious experience is private, we cannot explain what it is like to be another animal. Therefore, answering to the phenomenological question may be impossible in principle. Nevertheless, cognitive neuroscience may provide an answer to the distribution question, showing that conscious experience is not limited to humans but is a major biological adaptation going back many millions of years. The further progress of neuroscience will clarify how wide we should consider the class of living organisms endowed with consciousness. However, at the moment the evidence is strong enough to consider all mammals and some vertebrates such as birds as fully belonging to that class. Only future research and the synergy between philosophy and cognitive neuroscience will eventually explain the evolutionary history of consciousness, showing a discontinuity in its distribution, if it emerged at a precise stage of phylogeny, or showing its continuity across the natural world. Therefore, from now on, cognitive neuroscience will probably represent the base for the debate on animal consciousness. As the “Cambridge Declaration of Consciousness” argues, many conclusions about animal consciousness can already be stated unequivocally, and further data and experimental evidence will foster even more—and on a scientific basis—the ethical and philosophical debate on the relationship between humans and other animals.

Bibliography Allen, Colin, and Mark Bekoff. 2007. “Animal consciousness.” In The Blackwell Companion to Consciousness, edited by Max Velmans, and Susan Schneider, 58-71. Malden: Blackwell Publishing. Ayala-Guerrero, Fructuoso. 1989. “Sleep patterns in the parakeet Melopsittacus undulates.” Physiology and Behavior 46(5): 787-791. Baars, Bernard J. 2005. “Subjective experience is probably not limited to humans: The evidence from neurobiology and behaviour.” Consciousness and Cognition 14: 7-21. Block, Ned. 1995. “On a confusion about a function of consciousness.” Behavioral and Brain Sciences 18(2): 227-287. Bullock, Theodore H., and Bernd U. Budelmann. 1991. “Sensory evoked potentials in unanesthetized unrestrained cuttlefish: A new preparation for brain physiology in cephalopods.” Journal of Comparative Physiology 168(1): 141-150.

200

Cognitive Neuroscience and Animal Consciousness

Chalmers, David J. 1995. “Facing up to the problem of consciousness.” Journal of Consciousness Studies 2(3): 200-219. Dehaene, Stanislas, and Jean-Pierre Changeux. 2011. “Experimental and theoretical approaches to conscious processing.” Neuron 70(2): 200227. Delfour, Fabienne, and Ken Marten. 2001. “Mirror image processing in three marine mammal species: killer whales (Orcinus orca), false killer whales (Pseudorca crassidens) and California sea lions (Zalophus californianus).” Behavioural processes 53(3): 181-190. Descartes, René. 1984-91. The Philosophical Writings of Descartes. Translated by Cottingham, John, Stoothoff, Robert, and Dugald Murdoch. New York: Cambridge University Press. Edelman, David B., Baars, Bernard J., and Anil K. Seth. 2005. “Identifying hallmarks of consciousness in non-mammalian species.” Consciousness and Cognition 14: 169-187. Edelman, David B., and Anil K. Seth. 2009. “Animal consciousness: a synthetic approach.” Trends in neurosciences 32(9): 476-484. Edelman, Gerald M., and Giulio Tononi. 2000. A Universe of Consciousness. New York: Basic Books. Farah, Martha J. 2008. “Neuroethics and the problem of other minds: Implications of neuroscience for the moral status of brain-damaged patients and nonhuman animals.” Neuroethics 1: 9-18. Gallup, Gordon G. Jr. 1970. “Chimpanzees: self-recognition.” Science 167: 86-7. Gallup, Gordon G. Jr., Anderson, James R., and Daniel J. Shillito. 2002. “The mirror test.” In The cognitive animal: empirical and theoretical perspectives on animal cognition, edited by Mark Bekoff, Colin Allen, and Gordon M. Burghardt, 325-333. Cambridge: MIT Press. Griffin, Donald R., and Gayle B. Speck. 2004. “New evidence of animal consciousness.” Animal Cognition 7: 5-18. Hanlon, Roger T., and John B. Messenger. 2002. Cephalopod Behavior. Cambridge: Cambridge University Press. Hunt, Gavin R., and Russell D. Gray. 2004. “The crafting of hook tools by wild New Caledonian crows.” Proceedings of the Royal Society of London B (Suppl.) 271: S88-S90. Kardong, Kenneth V. 1995. Vertebrates: Comparative anatomy, function, and evolution. Dubuque: W. C. Brown. Laureys, Steven, et al. 2002). “Brain function in the vegetative state.” Acta Neurologica Belgica 102(4): 177-85. Laureys, Steven, Faymonville, Marie-Elisabeth, Luxen, André, Lamy, Maurice, Franck, George, and Pierre Maquet. 2000. “Restoration of

Matteo Grasso

201

thalamocortical connectivity after recovery from persistent vegetative state.” Lancet neurology 355: 1790-91. Laureys, Steven, and Giulio Tononi. 2009. The Neurology of Consciousness: Cognitive Neuroscience and Neuropathology. Oxford: Academic Press. Llinàs, Rodolfo, Walton, Kerry D., and Eric J. Lang. 2004. “Ch. 7 Cerebellum.” In The Synaptic Organization of the Brain, edited by Gordon M. Shepherd, 271. New York: Oxford University Press. Low, Philip, Panksepp, Jaak, Reiss, Diana, Edelman, David B., Van Swinderen, Bruno, and Christof Koch. 2012. “The Cambridge Declaration on Consciousness.” Accessed October 30, 2012. http://fcmconference.org Marten, Ken, and Suchi Psarakos. 1995. “Evidence of self-awareness in the bottlenose dolphin (Tursiops truncatus).” In Self-awareness in Animals and Humans: Developmental Perspectives, edited by Uue T. Parker, Robert W. Mitchell, and Maria L. Boccia, 361-379. Cambridge: Cambridge University Press. Massimini, Marcello, Ferrarelli, Fabio., Huber, Reto, Esser, Steve K., Singh, Harpreet, and Giulio Tononi. 2005. “Breakdown of cortical effective connectivity during sleep.” Science 309: 2228-2232. Medina, Loreta, and Anton Reiner. 2000. “Do birds possess homologues of mammalian primary visual, somatosensory and motor cortices?” Trends in Neuroscience 23(1): 1-12. Merker, Bjorn B. 2005. “The liabilities of mobility: A selection pressure for the transition to consciousness in animal evolution.” Consciousness and cognition 14(1): 89-114. Miller, Jason. 2009. “Minding the Animals: Ethology and the Obsolescence of Left Humanism.” Accessed October 30, 2012. http://www.americanchronicle.com/articles/view/102661 Nagel, Thomas. 1974. “What Is It Like to Be a Bat?” The Philosophical Review 83(4):435-450. Nagel, Thomas. 1986. The View from Nowhere. Oxford: Oxford University Press. Pakkenberg, Bente, Pelvig, Dorte, Marner, Lisbeth, Bundgaard, Mads J., Gundersen, Hans Jorgen G., Nyengaard, Jens R., and Lisbeth Regeur. 2003. “Aging and the human neocortex.” Experimental gerontology 38(1-2): 95-99. Panksepp, Jaak. 2005. “Toward a science of ultimate concern.” Consciousness and cognition 14: 22-29. Pepperberg, Irene M. 2002. “Cognitive and communicative abilities of grey parrots (Psittacus erithacus).” In The cognitive animal: empirical

202

Cognitive Neuroscience and Animal Consciousness

and theoretical perspectives on animal cognition, edited by Mark Bekoff, Colin Allen, and Gordon M. Burghardt, 247-253. Cambridge: MIT Press. Plotnik, Joshua M., de Waal, Frans B. M., and Diana Reiss. 2006. “Selfrecognition in an Asian elephant.” Proceedings of the National Academy of Sciences 103(45): 17053-17057. Prior, Helmut, Schwarz, Ariane, and Onur Güntürkün. 2008. “MirrorInduced Behavior in the Magpie (Pica pica): Evidence of SelfRecognition.” Public Library of Science Biology 6(8): e202. Rosanova, Mario, Gosseries, Olivia, Casarotto, Silvia, Boly Mélanie, Casali, Adenauer G., Bruno, Marie-Aurélie, Mariotti, Maurizio, Boveroux, Pierre, Tononi, Giulio, Laureys, Stephen, and Marcello Massimini. 2012. “Recovery of cortical effective connectivity and recovery of consciousness in vegetative patients.” Brain 135(4):13081320. Seth, Anil K., and Bernard J. Baars. 2005. “Criteria for consciousness in humans and other mammals.” Consciousness and Cognition 14: 119139. Smith, Andrew B. 1999. “Dating the origin of metazoan body plans.” Evolution and Development 1(3): 138-142. Steriade, Mircea, Timofeev, Igor, and François Grenier. 2001. “Natural waking and sleep states: A view from inside neocortical neurons.” Journal of neurophysiology 85: 1969-1985. Tononi, Giulio. 2004. “An information integration theory of consciousness.” BMC Neuroscience 5: 42-72. —. 2008. “Consciousness as integrated information: a provisional manifesto.” The Biological Bulletin 215(3): 216-242. Tononi, Giulio, and Christof Koch. 2008. “The neural correlates of consciousness: an update.” Annals of the New York Academy of Sciences 1124: 239-261. Tononi, Giulio, and Marcello Massimini. 2008. “Why does consciousness fade in early sleep?” Annals of the New York Academy of Sciences 1129: 330-334. Turing, Alan M. 1950. “Computing machinery and intelligence.” Mind 59: 433-460. Velmans, Max. 2007. “The co-evolution of matter and consciousness.” Synthesis Philosophica 22(44-2): 273-282. Vincent, Steven R. 2000. “The ascending reticular activating system: From aminergic neurons to nitric oxide.” Journal of Chemical Neuroanatomy 18(1-2): 23-30.

Matteo Grasso

203

Weir, Alex A. S., Chappell, Jackie, and Alex Kacelnik. 2002. “Shaping of hooks in New Caledonian crows.” Science 297: 981. Whitehead, Alfred N. 1929. Process and Reality. New York: Macmillan.

CONTRIBUTORS

Martina Belmonte, University of East Anglia Sofia Bonicalzi, University of Pavia Michele Borri, University of Milano-Bicocca Leonardo Caffo, University of Torino Luciana Ceri, University of Parma Guglielmo Feis, University of Milano Matteo Grasso, Roma Tre University Andrea Lavazza, Centro Universitario Internazionale, Arezzo Sarah Songhorian, Vita-Salute San Raffaele University Mattia Sorgon, Roma Tre University Francesca Vitale, University of Salento