Conversations on Human Action and Practical Rationality [1 ed.] 9781443850032, 9781443847889

Citation preview

Conversations on Human Action and Practical Rationality

Conversations on Human Action and Practical Rationality

Edited by

Carlos Mauro, Sofia Miguens and Susana Cadilha

Conversations on Human Action and Practical Rationality, Edited by Carlos Mauro, Sofia Miguens and Susana Cadilha This book first published 2013 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 © 2013 by Carlos Mauro, Sofia Miguens and Susana Cadilha 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): 1-4438-4788-7, ISBN (13): 978-1-4438-4788-9

TABLE OF CONTENTS  Introduction ................................................................................................. 1 Sofia Miguens and Susana Cadilha Interviews Alfred R. Mele ........................................................................................... 29 Hugh J. McCann ........................................................................................ 51 Michael Bratman ....................................................................................... 85 George Ainslie ........................................................................................... 95 Daniel Hausman ...................................................................................... 119 Joshua Knobe .......................................................................................... 133 Appendix ................................................................................................. 143 References of Project “Conversations on Human Action and Practical Rationality”

INTRODUCTION SOFIA MIGUENS AND SUSANA CADILHA

The interviews collected in the present volume originated in a research project entitled Conversations on Human Action and Practical Rationality, which was conducted at the University of Porto, Portugal, between 2007 and 2011 by MLAG (the Mind, Language and Action Group, a research group of the Institute of Philosophy1). The project was designed as a practical project about practical rationality: our aim was to have authors who work on practical rationality and human action answering a set of questions, speaking about their own work and discussing the theoretical differences separating them. The idea was to create an opportunity for some prominent authors in the area to speak for themselves as to what they do assume. The present volume has the philosophy of action as its domain of reference. Philosophy of action is a field in which issues such as the nature and explanation of actions, the nature of intention, deliberation and decision, the relation between reasons and causation or the possibility of akrasia are discussed. Yet in this book, as was already the case with other activities and publications within the project Conversations, what one finds is not so much work on specific issues in the philosophy of action but rather a result of our attempts to explore the direct links between philosophy of action and fields such as moral philosophy, cognitive psychology or the philosophy of economics. Agency and rationality are common denominators to these and several other domains, and that was the true focus of our project2. That Project Conversations was conceived as a practical project meant above all that it was designed around interviews, ‘conversations’ on practical rationality and human action, which we intended to pursue with

 1

The project was funded by FLAD (Fundação Luso-Americana,http://www.flad. pt/). 2 Action, Agency and Rationality is one of the areas of research of MLAG; it has issues of rationality at its core and reaches from philosophy of action to moral philosophy, political philosophy and philosophy of economics.

2

Conversations on Human Action and Practical Rationality

several people. Thus the script of the interview was particularly important for the whole process. Our script had six introductory questions: 1) In your view, what are the most central (or important) problems in the philosophy of action? 2) For some or all of the following problems - action, agency and agent - what do they contrast with most significantly? 3) Which of these are liable to be rational/irrational? 4) In what sense is the thing to do to be decided by what is rational? Are there limits to rationality? 5) What explains action, and how? What is the role of deliberation in rationality? 6) How is akrasia possible (if you think it is)? Specific questions about each author’s own work were then to be asked, so that the interviews would follow their natural course. These started with: 7) How do you think your own work has contributed to the field? What do you consider are your most important contributions? What are your plans for future research? The idea was that the script would prompt the exploration of each author’s thought. Some of the authors whose work we had in mind when we first started considering the script were those interviewed for the present volume: Alfred Mele, Michael Bratman, Joshua Knobe, Daniel Hausman, Hugh McCann and George Ainslie. Among them, some are philosophers, some philosophers with a leaning towards cognitive science, one a psychiatrist and behavioral economist. A longer list of authors helped us delineate the domain of the project: our interest in, and our discussions of, very diverse writings concerning the philosophy of action in contemporary philosophy was our entrance way into the project.3 We must reiterate the fact that in the project we intended to address issues as philosophers, but with an eye to conceptual and practical connections with other domains, ranging from cognitive science to

 3

In a volume in Portuguese (Sofia Miguens & Susana Cadilha eds., 2012, Acção e Ética, Lisboa, Colibri), which was the first publication of the Project, a compilation of more extended references of the project may be found, as well as interviews with several Portuguese philosophers (António Zilhão, Ricardo Santos, João Alberto Pinto, Vasco Correia).

Introduction

3

economics, in which questions regarding action, agency and rationality come up. We were interested in locating connections between pure philosophy of action and empirically-minded and experimental work. This was, in fact, the main rationale for our ‘one script’ strategy. Looking at the final results, we believe it was fruitful to ask the exact same questions to philosophers of action (Alfred Mele and Hugh McCann), to a philosopher of economics (Daniel Hausman), to philosophers somehow closer to cognitive science (Michael Bratman, who works in the Stanford Symbolic Systems Program; Joshua Knobe, appointed in the Program in Cognitive Science at Yale and whose work applies experimental methods to philosophy) and to a psychiatrist who (among many other things) has conducted research on preference reversal in animal behavior (George Ainslie). An introduction to some of the themes of the interviews, at some points referring to passages from the interviews themselves, follows – we built it around a brief profile of each interviewee.

Delineating the Issues Alfred Mele Alfred Mele has been, since 2000, William H. Lucyle and T. Werkmeister Professor of Philosophy at Florida State University. He works in philosophy of mind, philosophy of action, metaphysics and Greek philosophy; human behavior is the common denominator to his very diverse philosophical interests. He started his philosophical career as an Aristotle scholar, and some of the reasons for his interest in Aristotle, such as his approach to akrasia, remained a fixed point throughout his career. For Mele, philosophy of action is a sub-domain of the philosophy of mind with numerous connections to moral philosophy – such connections are visible in topics such as free will, moral responsibility, akrasia or motivation for action. In his first book, Irrationality – An Essay on Akrasia, Self-Deception and Self-Control (1987), Mele addressed several problems regarding two forms of irrational behavior: self-deception and weakness of will or akrasia. In a later book, Self-Deception Unmasked (2001), he would again take on the topic of self-deception, discussing empirical work which he believes supports the positions previously defended in Irrationality. In his second book, Springs of Action – Understanding Intentional Behavior (1992) he set out to understand what explains actions, focusing on the role of beliefs, desires and intentions in the production of human action. In the

4

Conversations on Human Action and Practical Rationality

following book, Autonomous Agents – From Self-Control to Autonomy (1995), he addressed the very possibility of free or autonomous action, and later, in Free Will and Luck (2006), he faced a major theoretical challenge to the idea that sometimes we act freely: the threat posed by luck and manipulation. In his fifth book, Motivation and Agency (2003), he developed a theory of the role of motivation in the life of intelligent agents. In his latest book, Backsliding – Understanding Weakness of Will (2012), Mele once again deals with akrasia. Along with akrasia, free will has been a recurrent topic in Mele’s work: in another recent book, Effective Intentions: The Power of Conscious Will (2009), he examines alleged scientific evidence in favor of the idea according to which free will is an illusion. Such view is supposedly supported by the fact that our brain 'decides' what we do before we are even aware that a decision has been made; there is a small 'time window for free will', of about 100 ms, yet all that can be done in this time window is to veto decisions or intentions. If things are indeed so, our decisions and intentions do not have any role in the production of corresponding actions (e.g., my intention to raise my arm does not have any role in the fact that now my arm is raising) and the idea that we are free agents, and as such worthy of praise or blame, is simply an illusion. The idea that free will is an illusion is obviously not new in the history of philosophy: suffice it to think of Spinoza or Nietzsche, who, in their works, have done much to deconstruct the phenomenology of the will. Both thought that our sense of making things happen, i.e. our sense of agency, is misleading in that it leads us to think we are masters in our house – ‘our house’ being our mind and our will. Still, even if that idea is not new, discussing it is, according to people like Mele, made more urgent by the explosion of research on the brain. But how, exactly, could an attack on free will pose a challenge to the philosophy of action? That could happen in several ways. One may, for instance, doubt that our sense of agency has any actual relation to causation. Notice, as an aid to imagination at this point, that an ‘evil genius’ scenario does not arise only for the representation of a world outside one’s mind – it arises also for the capacity of willing: it might be the case that an 'evil neuroscientist' makes me believe that I am doing the things I am doing, when in fact I do nothing for myself since everything I ‘do’ is controlled by him. Even in the first-person case, we could think it is a mere illusion that there is such a thing as the special role agency is meant to perform in the genesis of what we do. Seeing something is subject to illusion: for any given case of this, it is conceivable that things should seem just as they do where no such thing was seen, or even,

Introduction

5

perhaps, where no seeing was going on at all. Are we not also subject to illusion when it comes to (our role in) bringing something about? And why should it not be in the nature of this case that such illusion be systematic? This worry aside, to what extent must we know what we are doing in order to be doing it? To what extent must one not be in the grip of an illusion in order to be acting? And how might such immunity be in the cards? These are some of the worries about free will arising for a philosopher of action – and Mele, as coordinator of reference books such as Mental Causation (with J. Heil, 1993) and Free Will and Consciousness: How Might They Work? (with R. Baumeister & K. Vohs, 2010), and currently the leader of a large project on free will, called Big Questions In Free Will4 (2010-2013), has paid much attention to all of them. In the interview Mele closely examines what is meant by 'action'. One important thing to keep in mind when considering the expression 'human agents' is that humans do not always act, i.e. humans are not permanently and in every circumstance agents. Here are some of the examples given by Mele in the interview: if an evil neuroscientist uses Sam’s brain as a calculator to do multiplication, it is not Sam himself who is calculating when the results of operations occur to him – somehow such results cannot be regarded as his own thoughts. Something occurs in Sam which could be an action of Sam, but in fact is not. If we want to understand what an action is, Mele thinks, it is important to understand why we can and should say that someone like Sam is in such case merely a scenario of events, not a real agent. Or think of Uma: she may calculate something (for instance, a 15% gratuity at the restaurant) using her usual formula of multiplication, or she may do the same calculations compulsively, for no reason – compulsion, in the second case, significantly changes what we want to say about Uma as an agent. What is at stake here is the fact that something which may seem to be an action – purposeful movements of one’s own body, a seemingly intentional sequence of events – is not really an action if the agent ‘fails’ to be there, i.e. if the so-called ‘action’ merely happens in him, to him, or through him. For something to be an action, it must be caused in the right way. For some people, this is a historical matter, regarding what actually causes what: actions must be caused by beliefs, desires and intentions of the agent or they simply will not be actions. That is what the cases above illustrate and such is the outline of a general causalist thesis, defended by many authors in the philosophy of action, Mele included. Intentional

 4

This is a project sponsored by the John Templeton Foundation.

6

Conversations on Human Action and Practical Rationality

actions can only be adequately accounted for in terms of beliefs, desires and intentions; and explanations of actions evoking such beliefs, desires and intentions are causal explanations. Once agents (considered as entities which perform certain body movements on the basis of beliefs, desires and intentions) are in place, the question regarding their responsibility for actions arises. According to Mele, only human agents are responsible agents, and, even so, only for some of their actions. This is so because in order that an agent be responsible, it is not sufficient that her action is intentional; self-control is yet another necessary condition (one of Mele’s examples here is that of a man who compulsively washes his hands several times a day). Also, it should be considered that the agent might lack the motivation to use selfcontrol. One extra step needed is autonomy; plus, it should be the case that the agent ‘could have done otherwise’. Rationality and irrationality of agents in acting may then be assessed in relation with their reasons for acting (e.g., if I throw a heavy vase from the balcony when a man I dislike is passing on the street because I have the desire to kill him and I believe that a heavy vase thrown at his head does the job, I am definitely being rational – yet there is something wrong with me if in the same circumstances I throw a feather or a quail’s egg at him). But how is it that an agent ends up having reasons for acting in a certain way? Paths can be diverse, yet a particularly important one is the process of deliberation. Deliberating is something like carefully considering what to do. We usually think of deliberations as leading to the formation of intentions, in contrast to a situation where an agent predicts what will be the case with him (if I jump from a building, I will fall – the agent thinks). Mele sees things here the following way: by deliberating the agent puts forward a decision regarding the practical problem of what to do; this may be seen as an inference process leading to an evaluative conclusion. An agent may deliberate rationally or irrationally; in fact, it is not a necessary condition for the agent to act rationally that the action in question be a direct product of a deliberation. Anyway, if an agent deliberates and ends up thinking that A is the thing to do, then she acquires (by default, says Mele) the intention to do A. Practical decisions, i.e., decisions about what to do, are then, according to Mele, momentary mental actions of intention formation; intentions, in turn, are executive attitudes toward plans. Not all intentions, Mele thinks, are actively formed by me in an act of deciding (or in his terminology, not all intentions are ‘actionally acquired’): if I open the door of my office every day, when I approach this same locked door today I can 'non-actionally' form the intention to open it. Such is the way things go by default with agents in what concerns

Introduction

7

deliberation and intentional action. Yet there may be conflicts or even clashes between processes of formation of intentions and the agent’s motivation – this is what leads to akrasia. According to Mele, akrasia concerns the relation between an agent’s best judgment and her motivation to act, and is basically poor self-control. Typically, decisive better judgments regarding what is to be done are formed on the basis of our evaluation of the objects of our desire. Yet the motivational force of our desires is not always in line with our evaluation of the objects of our desires. It should not be surprising, then, that there are situations where although we think it is better to do A than to do B, we are strongly motivated to do B rather than A. This is what accounts for akrasia. Once one has considered belief-desire mechanisms, deliberation, decision and the motivation of agents, one might think there is still one more question, a global question, as it were, regarding our nature as agents: are we free? Is there such a thing as free-will? As we said above, this is another topic Mele has worked much on, although he doesn’t discuss it in the interview. Yet, in a way, that question drives all his work, since he believes that by considering issues regarding desires, motivation, reasons or deliberation the question of free-will is in fact being addressed. Thus, he says: “I myself have no special use for the noun «will»”. In this, he contrasts with other authors we interviewed, such as George Ainslie or Michael Bratman, who do not refrain from saying that in dealing with practical rationality what they are ultimately interested in is a theory of the will. Mele thinks there is no need for such a theory; all that we need in order to render our talk of free-will legitimate is to look closely at all the mechanisms involved.

Hugh McCann Hugh McCann is a Professor at the Department of Philosophy of Texas A&M University. His main interest is action – it is action that brings together the topics of philosophy of mind, philosophy of language, metaphysics and philosophy of religion he works on. These topics range from the ontology of events and change to the nature of causality and time, divine action, creation and eternity, and to the problem of evil. He is the author of The Works of Agency: On Human Action, Will and Freedom (1998) and Creation and the Sovereignty of God (2012). He has also authored numerous articles on quite diverse topics related to the nature of action, e.g. “Divine Sovereignty and the Freedom of the Will”, “Volition and Basic Action”, “Creation and Conservation”, “The Author of Sin?”, “Resisting Naturalism: The Case of Free Will”, “The Will and the Good”

8

Conversations on Human Action and Practical Rationality

and “Pointless Suffering: How to Make the Problem of Evil Sufficiently Serious”. He is also the author of the Stanford Encyclopedia of Philosophy article on "Divine Providence". It takes only a quick glance to realize that McCann positions himself against the currently dominant naturalism. In fact, that was the origin of our interest in his work: we wanted to understand how such kind of anti-naturalism would influence a view of action. McCann stands in a long tradition of connecting problems of action with issues in theology, such as those concerning sin and responsibility, or the relation of human freedom to divine foreknowledge and omniscience. One way to approach his work is precisely, as per his own suggestion in the interview, to try to understand how his way of addressing topics which are central to the philosophy of action (such as, for instance, the nature of basic actions5) could bear on the answer to traditional philosophicaltheological problems such as the problem of evil. One very important contrast between McCann and Alfred Mele is McCann’s opposition to the causalist approach. McCann’s volitional theory is best understood if seen under the light he himself proposes in the interview: tracing the history of the philosophical concept of action from logical behaviorism on, through the causal theory of action, now widespread and widely accepted, he presents his own position, centered on the concept of volition, as a criticism of the latter. Volition is McCann’s signature-concept as a philosopher of action; it is volition which allows him to explain how action differs from other events and to talk about free will and responsibility. Volition is, according to McCann, the inner activity by which agency is exercised when we perform overt actions. It is not a momentary act, but an activity: the activity of willing. It is volitions that are, according to McCann, basic actions. To fully understand the concept, one might consider an example from the interview (this is in fact a kind of example pervasive in McCann’s writings, since it is crucial for making his case regarding volitions): a patient, suffering from a particular neurological condition, wants to raise his arm, is willing to raise his arm, but to no avail, since his arm does not raise. He is paralyzed, yet he ‘wills’. McCann thinks the theory of action should account for such possibility – there must be some conception of what the case is in such situation. McCann sees nothing wrong with the idea that an action undertaken for a reason is intentional, nor with the idea

 5

I.e. those actions which do not take place by means of the agent doing something else (in contrast with non-basic actions such as firing the gun by pulling the trigger, or pulling the trigger by moving one’s finger).

Introduction

9

that beliefs and desires of the agent are reasons for action; all he wants is that such ideas do not close our eyes to the fact that acting is mainly willing to act, wanting to make something happen. In other words, the essential thing about action is not what happens subsequently but the fact that an agent produces it by willing (to do something). Beliefs and desires may be necessary conditions for an action to take place, but only the volitional act is a sufficient condition. The phenomenological aspect of this is very important for McCann – he thinks that to engage in volition is to feel spontaneously active, not something that merely happens to us. It is important to understand the way McCann sees the relation between volition and intention, as well as the relation between beliefs, desires and intentions. First, for McCann, intentions cannot simply be combinations of beliefs and desires. Here is another example from the interview: I may want to see Rome, and I believe I can do it by going on a holiday to Italy this summer, yet it hardly follows from this that I intend to go to Italy this summer. What is missing? What would be necessary for me to intend to go? What is missing is intention – intending to do something requires that I have decided, that I am committed to doing something and that I actually form a plan for it. But if we have intentions only, what we have are states, not events, and actions are supposed to be events. On the other hand, if we think, for instance, about a movement of one of our limbs (for instance, our arm’s upward motion), what we have is an event but not, by itself, an action (it may either be an action or not). This is the space filled by volition, the means by which we execute intentions. An intention, being a state, says McCann, cannot execute itself; that is why volitions are needed. Another important difference here, then, is the difference between desire and volition. A desire is always something which befalls us, something we find ourselves having, while a volition is, in McCann’s term, ‘actional’, it is itself active, not the sort of thing that ever befalls a person. Even if moral and theological implications of the discussion of action are, obviously, of great interest to McCann – he is interested in issues such as the duties and obligations of God and men, in why God created a world in which there is suffering and sin, as well as in questions of guilt – he is careful to emphasize in the interview that his views leave open the question whether the will is caused, either by reasons or by intentions. Are we free agents? If the will is subject to causation can we possibly be free? For McCann the problem of free-will is one of the problems of philosophy of action, a problem he sees as concerning the forming of intentions (not executing them). He is an incompatibilist, a libertarian who believes having free-will is forming intentions. The libertarian must argue that

10

Conversations on Human Action and Practical Rationality

reasons have explanatory power without causally determining the formation of intentions; he also has to explain why a decision is not an accident for which the agent is not responsible. A compatibilist, in turn, must find a way to distinguish between free and non-free actions in a deterministic world, and to offer a theory of responsibility and guilt that is compatible with determinism. Neither has a simple task. This seeming impasse between libertarians and compatibilists leads McCann, who is himself a libertarian but who considers the notion of agent causation a vacuous notion, to evoke a practical perspective on action, that of the phenomenology of acting. As he puts it in the interview, “From the practical perspective of the agent (…) it is conceptually impossible for decision or willing to count as ‘irruptions’, because it is conceptually impossible for these phenomena to occur accidentally or inadvertently. To see this, simply imagine the example I mentioned before: how you would react if a student came up to you and said, «I’m sorry to have missed your lecture yesterday Professor (fill in your name). I accidentally decided not to come»”. Finally, in what respects akrasia, McCann thinks that it happens frequently and that it is not a particularly problematic issue for the philosophy of action – akrasia simply happens when agents lack the motivation to carry out the intentions they formed. Contrary to Davidson’s famous proposal according to which there is nothing specifically moral in the problem of akrasia6, McCann does not flinch at the idea that the concept does have moral connotations – in his opinion akrasia is often a failure of the righteousness (fortitude) of a person and as such says something about her character.

Michael Bratman Michael Bratman is Durfee Professor in the Department of Philosophy at the School of Humanities and Sciences at Stanford University. Stanford University is an important place for the history of cognitive science in the second half of the twentieth century, and Bratman himself is close to its Symbolic Systems Program, a program of cognitive science which includes several disciplines, from philosophy to artificial intelligence, whose goal is the study of computers and minds or, in other words, the study of natural and artificial systems which use symbols to represent information. His philosophical work led him to develop a model of human practical reasoning, the BDI (Belief-Desire-Intention) model, which is used in many

 6

Cf. D. Davidson, “How is Weakness of the Will Possible?” In Essays on Actions and Events, Oxford, Oxford University Press, 1980.

Introduction

11

areas, including artificial intelligence. In 2008 he received the IFAAMAS (International Foundation for Autonomous Agents and Multi-Agents Systems) Award for Influential Papers, an award marking key contributions to research on agents and multi-agent systems. How does a philosopher end up inspiring work in artificial intelligence in such a direct way? Bratman believes the Western philosophical tradition has focused mainly on mind and knowledge rather than on agency. Yet agency, in the sense of ‘behavior potentially under the control of an oriented mechanism, which pursues a particular purpose’, is a concept common to a large number of disciplines, for all of which philosophical studies of agency are of interest. Such disciplines range from artificial intelligence, and cognitive science in general, to the social and political sciences. Bratman’s first book, Intention, Plans, and Practical Reason (1987), with its focus on intentions and plans, goes a long way in answering the question of how a philosopher ends up inspiring work done in artificial intelligence and having a software model developed to program intelligent agents inspired by his investigations. The book had a strong impact, calling attention to the complex and constitutive role of intentions and planning in rational agency. Intending, as it interests Bratman, is not just intention in action, the kind of phenomenon people such as Elizabeth Anscombe and Donald Davidson paid most attention to – it is also planning, i.e. developing stable and future-directed intentions. Moreover, according to Bratman, the phenomenon of intending to act is not just the result of practical reasoning: it is also a fundamental element of the information input of agents. Understanding intentions and plans is essential to understand the nature of agency and a key to characterize the crucial difference between practical rationality, concerning what to do, and theoretical rationality, concerning what to believe. The BDI model is in fact a proposal about the nature of practical rationality, according to which at its centre lies the phenomenon of intending to act; it separates two distinct capacities, the capacity of selecting a plan and the capacity of implementing current plans; according to it, agents are able to balance the time they spend deliberating on plans and implementing plans. Bratman continued his investigation in Faces of Intention – Selected Essays on Intention and Agency (1999), and Structures of Agency: Essays (2007), exploring how characteristically human agency is related to issues of self-determination, self-government and autonomy. In fact, according to Bratman, the issues at the core of philosophy of action concern the basic structures of human agency, and his work has extended from questions regarding the nature of plans and intention to specific phenomena such as shared intention and shared agency. These are, he believes, particularly important for understanding

12

Conversations on Human Action and Practical Rationality

human coordination and sociality. His work on shared agency, shared intention and shared valuing has been influential in the philosophy of law and political theory, and naturally so, since it aims at characterizing basic human capabilities, such as those of sociality and self-government, without reference to which there could be no such fields. These capabilities are indeed decisive in making us human, and according to Bratman the most crucial of them is shared intention. Bratman regards his planning theory as a 'modest theory of the will' in fact, he sees it as an alternative to other theories of the will, for example George Ainslie’s. According to Bratman, and contrary to what incompatibilists take it to be the case, the will is not something extraneous to the causal order: creatures with a 'will' are simply creatures endowed with certain psychological structures, namely capacities for imposing a structure on their thought and action. The question is to understand such structures, and Bratman thinks understanding the nature of planning and associated norms is crucial there. Given the fact that human lives are extended in time and socially coordinated, and given the fact that this happens in a context of cognitive limitations and limited access to information, planning is fundamental (it’s worth keeping in mind here, in order to better place Bratman in the history of cognitive science, that he was deeply influenced by Herbert Simon’s theory of bounded rationality7). Acts of choosing and deciding are the standard ways we use to delineate plans in our minds. Human agents are planning agents and Bratman insists that planning agents are a very special type of purposive agents (this last category encompasses many other animals, e.g. rats or bats or cows). Such purposiveness, which characterizes all agents as agents, is not to be identified with the phenomenon of planning that Bratman is interested in. In planning theory, intentions have a special relevance, in contrast with classical philosophical discussions of human motivation to act, where noticeably more attention is paid to desires. Even in 20th century philosophy of action comparatively little attention was paid to intention in Bratman’s sense. In the interview, Bratman says he sees his work, and its focus on planning, as resulting from a discussion about agency and intention between Anscombe and Davidson which has never actually taken place. He directs our attention to the many points of

 7

Economist and cognitive scientist Herbert Simon (1916-2001) put forward the idea of bounded rationality in contrast with conceptions of rationality as optimization (one example being the rational choice theory’s idea of maximization of expected utility). The idea of bounded rationality intends to capture the doings of a rational agent in situations of limited access to information which are quite common in the real world.

Introduction

13

agreement between them: both admit 'multiple descriptions' of what happens and they agree on the topic of individuation of actions. The classic points of conflict between the two concern causal explanations of action (Davidson defends them, Anscombe does not), and the non-observational knowledge which agents such as ourselves have of themselves while acting (something upon which Anscombe insists much and which is absent in Davidson). Yet, according to Bratman, there is a kind of tacit agreement between Anscombe and Davidson about intending to act: none of them thought that going that way would be very fruitful. But this is precisely the way to go, according to Bratman. Even if both intentions and desires are, in Davidson’s terminology, 'pro-attitudes' (mental attitudes related to action), intentions are special in that they involve commitment. It is such commitment which makes persistence of plans possible, and which makes new plans possible based on those with which the agent is already committed. Also, intentions are, in contrast with desires, subject to requirements of rationality; together with beliefs, intentions have the function of providing a consistent model of the future of the agent, for the agent. As Bratman puts it in "Intention, Belief, Practical, Theoretical": “Central to the planning theory is the idea that intentions – in contrast with ordinary desires – are both embedded in characteristic regularities and are subject to distinctive rational pressures for consistency and coherence. There is, in particular, a rational demand that one’s intentions, taken together with one’s beliefs, fit together into a consistent model of one’s future. There is, further, a rational demand that one’s intentions be means-end coherent in the sense, roughly, that it not be true that one intends E, believes that E requires that one intend means M, and yet not intend M. And these norms of consistency and coherence are operative in a planning agent’s practical reasoning.” Understanding the nature of those rational requirements is a central concern of Bratman. And to characterize intentions this way is to see them as elements of futuredirected plans, which are central to practical rationality. Yet another reason for the great interest Bratman’s work has for the studies of rationality and human action is his exploration of the epistemology of practical rationality, which, as he shows, contrasts with theoretical rationality in many interesting ways. Bratman focuses for instance on the fact that plans are typically partial, i.e. they have to be filled in by reality and context, and also on the fact that in contexts of practical decision-making a cognitive attitude of 'acceptance in a context' takes over (which contrasts much with belief8).



8 Cf. M. Bratman 1992 ‘Practical Reasoning and Acceptance in a Context’, Mind 101, 1-14.

14

Conversations on Human Action and Practical Rationality

George Ainslie George Ainslie is an american psychiatrist, psychologist and experimental economist and the author of two books which were especially important in Project Conversations: Picoeconomics – The Strategic Interaction of Successive Motivational States Within the Person (1992) and Breakdowns of the Will (2001). The idea of 'picoeconomics' is the following: the same way classical economics aims at describing negotiation for limited resources between institutions and microeconomics aims at describing negotiation for limited resources between individuals, picoeconomics aims at describing interactions for the control of the finite behavioral capacity of an individual within that individual. As in the other two cases, the interactions for the control of the finite behavioral capacity of an individual within that individual resemble a negotiation between parties. In Breakdowns of the Will, Ainslie applies the picoeconomics approach to a model of the self (or person) and of the will. This is what allows him to explain the so-called ‘collapses’ of the will. Ainslie’s medical background clearly marks his approach: he always has in mind not only cases which are the classic object of interest of philosophers, such as more or less abstractly conceived cases of akrasia, but clinical conditions, namely addiction (to drugs or alcohol, for example, but many other types of addiction as well). His proposals regarding the nature of action, self and will, are in fact intimately related to his studies of addictive behavior, with which he dealt very closely in his clinical work9. He actually believes that the ambiguity which characterizes addictive behavior (the willing and not willing to do what one sets about to do, as in the case of the addict who wants and does not want the heroine he is about to take) is quite pervasive even outside clinical contexts. What is particularly relevant for understanding the self and the will in the phenomenon of ambiguity is the fact that agents are very often seduced by urgent pleasures with destructive objects, thus interfering with their own long-term stable choices. It is as if there was a self always alert to the possibility of immediate, strongly motivating, compulsive, pleasure – such as in ‘I want these drugs now’ – and such self often won and controlled action, even if that meant failing to give satisfaction to the long term interests of another self that is also oneself, and failing to respond to what that other self is committed to. If Ainslie is right in thinking that this is a particularly important structure for the study of action and agents, that means that understanding the preferences of agents such as ourselves



9 He was Chief Psychiatrist in the Veteran Administration Medical Center, Coatesville, Pennsylvania.

Introduction

15

involves much more than the computations of relative values considered by rational choice theory. A basic idea of Ainslie is that different parts or aspects of personality are in conflict in individuals (this is something he explicitly relates to the Freudian theory of the id, the ego and the superego) and that has to be considered always, in choice behavior. This is particularly clear in addictive behavior, thus making addictive behavior particularly significant in the study of action. Ainslie’s ideas spread to behavioral economics, especially through Richard Thaler’s10 theory of multiple selves. Also, together with Drazen Prelec’s11 investigations, Ainslie's work was one way through which studies on operant conditioning joined other approaches to decision making, creating a challenge to mainstream economic thinking, centered on rational-choice theory. In the interview, again evidencing his medical and clinical background, Ainslie puts forward a definition of action quite different from those of the other authors’ interviewed: for him actions are teleological processes, which are repeatable and reward-responsive. Among the examples he gives are bulimia and other kinds of addictive behavior. It is the 'rewardresponsiveness' of actions he focuses on and it is within such framework that he answers our questions regarding choice, deliberation, akrasia, etc. Since he does not recognize a distinction in kind between actions and passions, or between actions and thoughts, but only degrees of motivatedness and degrees of deliberateness, Ainslie cannot simply classify actions into rational and irrational. He believes that the test for rationality which comes closest to the ordinary meaning concerns whether a choice serves the long range interest of the self. In approaching the issue, he prefers to focus on motivation, which, he says, implies a universal currency, best called ‘reward’, which is necessary to settle the competition between members of the class of processes that can be substituted for one another. He refers to the hypothetical space at which these processes compete for expression as the motivational marketplace. Speaking about ‘reward’ in this way is a mark of Ainslie’s experimental investigation of operant conditioning in animals. Ainslie has worked with psychologist Howard Rachlin12, and investigated, in particular, the phenomenon of inter-temporal choice in pigeons. He was first to demonstrate experimentally the phenomenon of preference reversal in agents who have immediate benefits in view, a phenomenon he explained

 10

American economist, b. 1945, University of Chicago Booth School of Business. Professor of economics in the MIT Sloan School of Management. 12 American experimental psychologist, Emeritus Professor at SUNY. 11

16

Conversations on Human Action and Practical Rationality

in terms of hyperbolic discounting of prospective rewards. This means that the agents’ valuing of prospective reward stands in inverse proportion to its delay (this is an idea stemming from Rachlin’s work). Ainslie then integrated such views into experimental and theoretical work on intertemporal choice. In fact, the term ‘picoeconomics’ is often used to describe the implications of one specific experimental discovery: the tendency people have of showing strong preference for immediate payoffs in alternative to long-term benefits, a tendency which is stronger the closer both situations are to the present moment. In other words, given two (comparatively) similar rewards, humans show a preference for the one which arrives earlier. For example, many people, when offered a choice between getting $50 now and $100 a year from now choose $50 now. However, given a choice between $50 in 5 years, and $100 in 6 years, they all choose $100 in 6 years, even if this is the same choice only seen at a greater distance. It is thus said that humans discount the value of the later reward by a factor which increases with the increasing delay of the benefits. A large number of experiments confirmed that spontaneous preferences of human and non-human subjects followed a hyperbolic curve rather than the conventional exponential curve which would reflect consistent choice over time. Ainslie’s book Picoeconomics covers such topics, but one can also find there many insights derived from the philosophical tradition of reflection on the passions, a tradition which connects the Stoics, David Hume and (very centrally) Freud. Anyway, a central point of Ainslie’s is that a formal way of thinking about choice, such as the utilitarian way of thinking about choices in terms of maximization of expected utility, taking as a touchstone simple monetary choices, simply does not capture the complexity, nor the ambivalence, of most human choices. It is in his book The Breakdowns of Will that Ainslie deals more directly with akrasia. The idea of hyperbolic discounting of prospective rewards suggests a model of the self as a population of processes in search of reward – the short-term and long-term interests of agents are seen as reward-seeking processes competing with each other, according to a rule of maximization of expected rewards discounted at every moment of choice. This is combined with an inter-temporal bargaining model of the will. In the context of such competition, the ‘will’ is not (contrary to what rational-choice theory holds) a superfluous concept: it has a crucially important function: maintaining the agent’s preference for LL [larger and later], as opposed to SS [smaller and sooner]. As for human action par excellence (to use D. Velleman’s term), how is it that, according to Ainslie, consciousness, deliberation or freedom

Introduction

17

come to be? Ainslie believes the question should be posed in terms of the relation between motivation and deliberation: “Most of the processes that seem to have been discussed by philosophers are not only motivated but also deliberate”, and he thinks that deliberate actions have been tested in the marketplace of reward; in fact one chooses them while conscious both of them and of their immediate alternatives. In this sense a deliberate action is one chosen “all things considered”. Still how could such collection of processes be a person, and ‘one and the same person over time’, as Locke would put it? According to Ainslie a (single) person comes into existence out of populations of competing interests: “These properties of action permit a theory of how a person, that is, a population of reward-seeking processes, can form higher mental processes – ‘ego-processes’ and an ‘ego-identity’.” A person is thus a single entity that extends over time and who knows that she cannot always count on herself: temptations come by, and it is very possible that she succumbs to them in the future. What she has to do then is 'negotiate with herself' such inter-temporal relations and preferences. In the interview Ainslie evokes Jon Elster’s emblematic case of Ulysses and the Sirens13: Ulysses does not want to be tempted by the Sirens, yet since he believes he will in fact be tempted by the Sirens, he decides to bind himself to the mast, planning his action in order to keep himself from succumbing to temptation when the time comes. With his theory of the self and the will, inspired by behaviorist psychology on the one hand and by Freudian themes on the other, Ainslie makes clear not only the difference the biological nature of (at least some) agents might make, but also that a model of the self is crucial when addressing issues of practical rationality.

Daniel Hausman Daniel Hausman is Herbert A. Simon Professor in the Department of Philosophy of the University of Wisconsin – Madison, and a prominent name in the philosophy of economics. For people with a specific interest in the philosophy of science, economics is a particularly exciting subject: it shares many features with the natural sciences, including the use of mathematical methods, while its objects are social phenomena. Also, economic theories such as theories of welfare and social choice often involve substantive (moral, political) philosophical commitments. Work in



13 Jon Elster, Ulysses and the Syrens, Cambridge, Cambridge University Press, 1979.

18

Conversations on Human Action and Practical Rationality

the philosophy of economics is thus bound to require quite diverse theoretical skills and interests, and Daniel Hausman’s education path is a good example of that. He attended Harvard College majoring first in biochemistry and then in History and English Literature. After teaching intermediate school in the Bronx and obtaining a Master of Arts in Teaching at NYU (New York University), he spent two years studying moral science at Gonville and Caius College at Cambridge University, in the United Kingdom, completing a PhD in philosophy at Columbia University, New York, in 1978. Hausman’s research has focused on methodological issues, as well as on metaphysical and ethical issues arising at the borders of economics and philosophy. In collaboration with Michael McPherson he founded the journal Economics and Philosophy. He was editor of the journal for the first ten years of its existence. He has also coordinated an important anthology, The Philosophy of Economics (2007), and published several books, namely Capital, Profits, and Prices: An Essay in the Philosophy of Economics (1981), The Inexact and Separate Science of Economics (1992), Economic Analysis and Moral Philosophy (co-authored with Michael McPherson, 1996), Causal Asymmetries (1998), Economic Analysis, Moral Philosophy and Public Policy (co-authored with Michael McPherson, 2006) and, more recently, Preference, Value, Choice and Welfare (2011). In his Stanford Encylopedia of Philosophy article on "Philosophy of Economics", Hausman maps the problems of philosophy of economics, characterizing the philosophy of economics in an illuminating way as consisting of three main types of investigations: (a) investigations of rational choice, (b) investigations on the appraisal of economic outcomes, institutions and processes, and (c) investigations on the ontology of economic phenomena and the ability to acquire knowledge of them. According to Hausman, although these inquiries intersect in many ways, it is useful to keep such branching in mind, especially since the branches can be seen respectively as (1) a branch of action theory, (2) a branch of ethics (or normative, social and political, philosophy) and (3) a branch of philosophy of science. In the interview, Hausman begins by bringing to our attention that the very identification of the main problems of philosophy of action depends on the interest with which one approaches the field. In his case, having economics as a background and rational choice theory as a reference, the major problems are the nature of preferences, the difference between preferences and value judgments and the connection between preferences and wants, beliefs, desires, choices and action (or, in other words, the relation between rational choice theory concepts, as used by economists,

Introduction

19

and common descriptions of human action, closer to common sense psychology). It is hardly disputed that rational choice theory involves a specific perspective on practical rationality. For example, when asked about the role of deliberation in rationality, Hausman pointed out that this is itself controversial: in economics it is assumed that agents have completed their deliberations, that they already have a definite and clearly fixed ranking of preferences. They then choose according to this ranking – that is what acting rationally is. In such conditions, to deliberate is simply to calculate – nothing more needs to be said about deliberation. Eventually, at a later stage, problems concerning the formation or the change of preferences may arise (or even questions about the very relevance of the notion of ‘preferences’ – Oxford philosopher John Broome, for example, would prefer to simply speak of ‘goods’, so as not to skew or bias questions in advance14). But such reflections mostly do not come up in economists’ work. A bit more should be said about agents and the ranking of preferences, given the fact that, from the perspective of economics, that is what rational agents are: entities characterized by a ranking of preferences, acting rationally when choosing according to it. In the Stanford Encyclopedia article, Hausman, while analyzing the history of economic thought from Aristotle, through the Physiocrats, David Hume and Adam Smith, to the present, calls attention to the fact that 20th century economists have generally abandoned earlier hedonistic formulations of choice and preference, which were closely linked to utilitarian philosophy and which focused on the happiness of agents. It was in the place of such formulations that talk of a ranking of preferences became common. Rational agents are characterized by their preferences; preferences are rankings of objects of choice. To deliberate is to create a ranking of the alternatives we face. It then becomes a crucial feature of rational agents that they are able to make a consistent ranking. This means that rankings are complete – for two alternatives x and y, the agent either prefers x to y, or prefers y to x, or is indifferent. Also, the rational agent's preferences are transitive. Thus we have the economist’s picture of the rational agent: he or she has complete and transitive preferences and armed with such preferences he or she chooses among alternatives. This angle of approach is in fact essential to distinguish economics from other social sciences: what is distinctive of economic investigations of social reality is supposed to be precisely the (rational) nature of the preferences of economic agents.

 14

Cf. John Broome, Weighing Goods, Oxford, Blackwell, 1995.

20

Conversations on Human Action and Practical Rationality

Admittedly, something like desire for wealth and consumption should be present behind such rational preferences – this, however, is simply assumed. Basically, wealth and consumption are what makes humans happy; preferences such as the ones of the ascetic man are, as Hausman comments in the interview, quite 'strange'. What we have here is the outline of a theory of rationality, and thinking about the nature of such theory in its multiple branches and domains is part of the occupation of philosophers of economy. Still one may argue, for example, that this is too weak a theory, as it says nothing about beliefs, or about what rationality implies when agents do not know everything relevant to the choices they make. One might also consider it too strong: one might for instance argue that there is nothing irrational in having incomplete preferences in situations which involve uncertainty – situations which definitely abound in the lives of agents such as us. And it may be reasonable to suspend judgment or not to choose among alternatives whose nature is not completely understood – why should such a stance be considered irrational? Moreover, as for the transitivity of preferences, while it does seem a plausible requirement of rationality, it also seems to face abundant experimental evidence that people's preferences are not in fact transitive. Arguably this doesn’t imply that agents themselves are irrational, since it is only a part of their behavior that could be characterized in that manner. Anyway, the interpretation of data regarding change of preferences and ‘irrational’ choices is a core issue in studies of rationality, calling for the combination of abstract models with empirical research. Extending the theory of rationality to circumstances involving risk and uncertainty calls for yet more principles and technical instruments, such as Bayesian ones, whose nature is also up to the philosopher of economics to analyze. In any case, it is important to keep in mind that rational choice theory involves the formalization of the conditions of rationality, allowing for formal manipulations. If we have agents with complete and transitive preferences who also satisfy an additional condition of continuity, then they can be represented by the ordinal utility function. One can define a function that represents the preferences of the agent such that U (X) > U (Y) if and only if the agent prefers X to Y and U (X) = U (Y) if and only if the agent is indifferent between X and Y. The function represents the ranking of preferences only – it contains no additional information. Any transformation of "U" that preserves the order also represents the preferences of the agent. When, in addition to that, the agent’s preferences satisfy the condition of independence, and other technical conditions, they can be represented by the expected utility function. This is a function

Introduction

21

which has important properties much discussed by economists and philosophers of economics. The above is just an elementary sketch of what goes on in the philosophy of economics. Connections between philosophical issues about action, agency and rationality and the field of economics were at the very origin of project Conversations15, and Daniel Hausman’s work helped us making such connections clear.

Joshua Knobe Joshua Knobe is currently a Professor in the Cognitive Science Program at Yale University and a major figure in experimental philosophy. His case for experimental philosophy has had plenty of coverage in mainstream and online media, from The New York Times, to Slate or bloggingheads.tv. Experimental philosophy advocates setting aside philosophers’ appeal to (their own) intuitions trying instead to understand how ordinary people think by means of empirical studies. Knobe’s name is associated with the "Knobe Effect" or "Side-Effect Effect". The “Knobe Effect” emerged in a much-discussed study of intentional action16. In the study Knobe confronted people on the street (people spending time in a Manhattan public park) with the following scenario (each subject read a vignette): “The vice-president of a company went to the chairman of the board and said, ‘We are thinking of starting a new program. It will help us increase profits, but will also harm the environment.’ The chairman of the board answered, ‘I don’t care at all about harming the environment. I just want to make as much profit as I can. Let’s start the new program.’ They started the new program. Sure enough, the environment was harmed.” The subjects were then asked how much blame the chairman deserved for what he did and whether he had intentionally harmed the environment. A vast majority of people (82%) said the chairman harmed the environment intentionally. This is the 'harm condition'; in the 'help condition', the vignette is exactly the same, except the word ‘harm’ is replaced by ‘help’. Thus, subjects read: “The vicepresident of a company went to the chairman of the board and said, ‘We are thinking of starting a new program. It will help us increase profits, but will also help the environment’. The chairman of the board answered, ‘I

 15

Carlos Mauro, one of the editors of this book, and an economist by training, who did a PhD in Philosophy on the topic of rationality in action (Porto, 2009), was instrumental in conceiving of Project Conversations on Practical Rationality. 16 Cf. Joshua Knobe, 2003, “Intentional Action and Side Effects in Ordinary Language”, Analysis 63, 190-193.

22

Conversations on Human Action and Practical Rationality

don’t care at all about helping the environment. I just want to make as much profit as I can. Let’s start the new program.’ They started the new program. Sure enough, the environment was helped.” Asked whether the chairman had intentionally helped the environment, a majority of subjects (77%) said ‘no’. The asymmetry between the 'harm' and the 'help' scenarios is known as the "Knobe Effect". Based on it, Knobe has argued that so-called folkpsychological attributions of intentionality are not morally neutral, and that, if such is the case, one should not regard them as neutral tools for predicting and explaining behavior as is often done in philosophy and cognitive science. Instead, the common concept of ‘intentional action’ should be regarded as something like a 'multipurpose tool', with different cognitive uses. The example above is not in itself a particularly controversial piece of experimental philosophy; more needs to be said about the reasons why experimental philosophy became so controversial within academic philosophy17. Even if much of what is actually done in experimental philosophy can be regarded as cognitive science research on topics ranging from intention and consciousness to free-will and the emotions, the fact is that experimental philosophy began with an extra purpose: to militate against traditional armchair philosophy. Experimental philosophers claim that in order to understand the ways humans think and act, philosophers should 'go out and run empirical tests', rather than stick to their own intuitions. This is the reason why experimental philosophy triggered a heated debate about what one does when one does philosophy. Now, ‘traditional’ philosophers who are less than enthusiastic about experimental philosophy mostly think there is nothing wrong with the empirical research of the topics referred to above – what they dispute are the reasons why anyone should consider that Knobe, or any other experimental philosopher, is doing philosophy, as opposed to social or cognitive psychology. The controversy around experimental philosophy can easily be reconstituted by reading the Manifesto Knobe co-wrote with Shaun Nichols18, and critical responses to it. Some such responses are Antti Kauppinen’s (“The Rise and Fall of Experimental Philosophy”19) or

 17

The following paragraphs express the views of the authors of the present Introduction, and not those of Carlos Mauro, who is himself a practitioner of experimental philosophy. 18 Cf. Joshua Knobe & Shaun Nichols, Experimental Philosophy, New York, Oxford University Press, 2008. 19 Antti Kauppinen, 2007, “The Rise and Fall of Experimental Philosophy”, Philosophical Explorations 10 (2): 95-118.

Introduction

23

Timothy Williamson’s, in his book The Philosophy of Philosophy20. Many other comments can be found on the Web. Antti Kauppinen ends his article by saying that "At best, survey results provide food for thought – but we are better nourished if instead of designing artificial setups we pay close attention to what is said in real-life situations of language use, as conscientious philosophers have done at least since Socrates". Among other things, this is a methodological observation about the actual role of language in philosophical investigations. In fact, something that might exasperate armchair philosophers (i.e. those analytic philosophers whom experimental philosophers see as representatives of the ‘technical mainstream’) is the fact that experimental philosophers (at least in the Knobe-Nichols Manifesto) characterize the 'method of conceptual analysis' as consisting in attempts to ‘identify precisely the meaning of a concept by breaking the concept into its essential components’21. This is a surprisingly simplistic description, which basically overlooks the whole history of analytic philosophy since its late nineteenth century beginning with Frege. Not only it skips the role logic has had in investigating the nature of thought, as a response to scientism and psychologism rising in late nineteenth century, but also disregards the fact that philosophical analysis of thought simply is not, for the founders of the analytic tradition, analysis of concepts in people’s minds. Yet the fact that at the origins of analytic philosophy logic and language did come together in a conception of philosophical method (the philosophical significance of this is obviously open to discussion22) seems to go unnoticed by experimental philosophers: for them 'method' seems to denote merely experimental method, the only method which can be ‘scientific’. Still, an experimental philosopher could argue that the dispute is not only, or mostly about method but about issues: in the interview, Joshua Knobe stressed the importance not only of the methods of experimental philosophy for the philosophy of action, but also the fact that experimental philosophy is reviving classic topics of philosophy, such as those addressed by a philosopher like Nietzsche, topics concerning how human beings really are, in contrast with the abstract technical issues of armchair

 20

Timothy Williamson, The Philosophy of Philosophy, Oxford, Blackwell, 2007. Cf. Knobe & Nichols, 2008, p. 4. 22 In fact it is historically more accurate to see the history of analytic philosophy since its fregean beginnings as an ongoing dispute around several conceptions of method, all having logic and language as their reference. Such dispute has to do, namely, with diverging conceptions of the relation between formal languages and natural languages, which obviously bears on the conception of what an ‘analysis’ by means of logic does. 21

24

Conversations on Human Action and Practical Rationality

philosophers. One example Knobe gives of such Nietzschean questions is that of why people believe in free-will (the answer might lie, according to Nietzsche, in our desire to justify acts of punishment: Nietzsche speaks of the 'metaphysics of the hangman'). As stated in the Manifesto: “The ultimate hope is that we can use this information to help determine whether the psychological sources of the beliefs undercut the warrant for the beliefs. The basic approach here should be familiar from the history of philosophy. Just take a look at nineteenth-century philosophy of religion. At the time, there was a raging debate about whether people’s religious beliefs were warranted, and a number of philosophers (Marx, Nietzsche, Feuerbach, etc.) contributed to this debate by offering specific hypotheses about the psychological sources of religious faith. These hypotheses led to an explosion of further discussion that proved enormously valuable for a broad variety of philosophical issues. But then something strange happened. Although arguments of this basic type had traditionally been regarded as extremely important, they came to occupy a far less significant role in the distinctive form of philosophy that rose to prominence in the twentieth century. The rise of analytic philosophy led to a diminished interest in questions about, for example, the fundamental sources of religious faith and a heightened interest in more technical questions that could be addressed from the armchair. The shift here is a somewhat peculiar one. It is not that anyone actually offered arguments against the idea that it was worthwhile to understand the underlying sources of our beliefs; rather, this traditional form of inquiry seems simply to have fallen out of fashion. We regard this as a highly regrettable development. It seems to us that questions about the sources of our religious, moral, and metaphysical beliefs are deeply important questions and that there was never any good reason to stop pursuing them. Our aim now is to return to these questions, this time armed with the methods of contemporary cognitive science.”23 One thing that could be said about this is that experimental philosophers seem to be aware of Anglophone philosophy only. If they were aware of a wider philosophical tradition it would be obvious to them that such topics, as well as Nietzsche himself as a major philosopher, were in fact never abandoned. Yet what continental philosophers inspired by Nietzsche (such as Foucault or Deleuze) aim at is certainly not in the least similar to the experimental philosophers’ proposal of bringing the methods of cognitive science to bear on the underlying sources of our beliefs. Looking beyond the controversial issue of experimental philosophy, it

 23

Knobe & Nichols 2008, p. 7.

Introduction

25

is important to notice that someone sharing Knobe’s views is also arguing that the normative models of rationality often subscribed to by philosophers are not the right models to describe how creatures like us are rational. Evoking the classic topic of the naturalistic fallacy, Knobe alerts to the danger of the 'non-naturalistic fallacy': the mistake of thinking that experimental results are completely irrelevant, and that philosophers can simply ignore everything we know about human nature (Knobe is thinking of empirical investigations here) in answering philosophical questions (e.g., questions about action, agency or morality). This idea – that real rational agents are not ideal rational agents – was at the very origin of our projects about rationality, and inasmuch as cognitive-science minded research on practical rationality and human action pursues such goals, its results are of great interest per se, regardless of methodological controversies. Alfred Mele, Hugh McCann, Michael Bratman, George Ainslie, Daniel Hausman and Joshua Knobe are some of the authors with whom we completed interviews in Project Conversations on Practical Rationality and Human Action. Those interviews make up the central part of the present volume, which we hope conveys discussions that took place within the Project. Over the years we dealt with extremely varied topics, ranging from free-will and akrasia, to methodological discussions around psychoanalysis or experimental philosophy, to the theological implications of the theory of action. We are convinced that issues of practical rationality and human action are a key to vast and fertile areas of contemporary philosophy. We hope that readers will think the same. Sofia Miguens Susana Cadilha References regarding the work of Alfred Mele, Hugh McCann, Michael Bratman, George Ainslie, Daniel Hausman and Joshua Knobe can be found at the end of this volume, in Appendix: References of the Project “Conversations on Practical Rationality and Human Action”.

INTERVIEWS

ALFRED R. MELE FLORIDA STATE UNIVERSITY

1. In your view, what are the most central (or important) problems in the philosophy of action? The most basic question in the philosophy of action is about the nature of action: What is an action? Another central question is about the explanation of actions: How are actions to be explained? The first question directly raises two others: How do actions differ from nonactions? And how do actions differ from one another? A philosopher who asks the question about explanation may be looking for a theory about how to explain why agents perform the actions they perform, a theory about how actions are produced, or both. Of course, there are lots of other interesting questions in the philosophy of action. Here is a small sample. What is it for an action to be intentional? What is it to do something freely? Do we ever act freely? Are we morally responsible for any of our actions? What is it to act for a particular reason?1 2. For some or all of the following problems—action, agency and agent—what do they contrast with most significantly? In my view, an agent is just a being that acts and agency is simply the property of being an agent2. This makes action the most basic of these three notions. That is, our answer to the question “What is an action?” will tell us what an agent is and what agency is. I will say a bit more about this, drawing on the chapter just mentioned, before taking up the issue of contrast. What is it to be a human agent? There are fancier and less fancy answers in the literature. My preferred answer is far from fancy: to be a human agent is to be a human being who acts. Human agents are not always acting. When they are not, they are not functioning as agents. 1

I have written several introductory articles on the philosophy of action that take up the questions identified above and describe various competing views about how they are to be answered. They include Mele 2005, 2006b, 2009b, 2009c, 2009d. 2 See my Motivation and Agency, 2003, ch. 10.

30

Conversations on Human Action and Practical Rationality

Consider the following cases. CASE 1. By means of direct electrical stimulation, a neuroscientist uses Sam’s brain as a calculator to do multiplication problems. Sam has no desire to do mental arithmetic and he does not try to do multiplication problems in his head. Sam is conscious of various results of the manipulation. He has conscious thoughts like the following. “Problem: 17 x 11. Answer: 187.” Sam has no idea that he is being manipulated. These thoughts occur with great frequency. Sam thinks he is going crazy. CASE 2. The calculations occur spontaneously in Sam’s brain, but everything else is the same, including Sam’s lacking a desire to do mental multiplication. Again, he thinks he is going crazy. CASE 3. Uma has practiced and learned a useful method for solving multiplication problems in her head. Just now, in utterly ordinary circumstances, she used the method to solve 113 x 15 in order to calculate the gratuity on a dinner bill. CASE 4. This case differs from the preceding one in that Uma has compulsive desires to solve multiplication problems in her head and she was motivated by such a desire to solve 113 x 15. Uma does not like having those desires. She regards them as alien forces. In cases 1 and 2, Sam, in Thomas Nagel’s words, is “merely . . . the scene” of the pertinent events (1986, p. 113). He is not solving multiplication problems; he is not playing an agent’s role in their being solved. Uma, in case 3, does solve a multiplication problem; she plays an agent’s role. Uma solves the same problem in case 4. In that case, as in case 3, she performs the mental action of solving that problem, and insofar as she does this, she plays an agent’s role. The difference between the two cases is not that she acts in one but not in the other. Rather, the difference is that her action is not compulsive in one and is compulsive in the other. This may be disputed. It may be claimed that it is not Uma who acts in case 4, but something else. I disagree. Uma has the psychological problem, and that problem of hers manifests itself in the compulsive desire that motivates her solving the multiplication problem. If it is something other than Uma that acts in case 4, what is it? Certainly neither her desire to solve the problem nor her compulsion. Desires and compulsions simply are not capable of performing the pertinent action, because desires and compulsions have no idea how to solve multiplication problems, or how to

Alfred R. Mele

31

do anything at all for that matter. (It may be replied that electronic calculators solve multiplication problems even though they have no idea how to do so. However, electronic calculators do not act, in any sense of “act” that is of special importance to the philosophy of action.) David Velleman reports that “The cases of defective action that occupy [Harry] Frankfurt’s attention are cases in which the agent fails to participate because he is ‘alienated’ from the motives that actuate him . . .” (1992, p. 470). Here, a distinction is in order between failing to participate in an action and participating defectively in an action. The task of accommodating human action par excellence – a task Velleman sets for himself – arises whether one conceives of human agents as human beings who act or as something fancier. On the less fancy view, one can safely say that Uma, a human agent, acts defectively in case 4, and one may also say that the defect is such that her action is not a human action par excellence. On a fancier view, one can say that Uma qua agent and Uma qua human being are different things and that the agent does not act, in which case neither the agent nor the human being engages in human action par excellence. I myself have never felt the need to use the expression “human agent” to mean something fancier than “human being who acts.” The question posed to me, again, was this: For some or all of the following problems, what do they contrast with most significantly: action, agency, agent? Because I take action to be the notion in terms of which “agent” and “agency” are to be defined, I take the deepest question raised by this one to be the following: What do actions “contrast with most significantly”? The answer is nonactions. But how do actions differ from nonactions? This question, as I said in my reply to question 1, is directly raised by a fundamental question in the philosophy of action: What is an action? According to the answer I favor to the question how actions differ from nonactions, actions are like sunburns and money in an important respect. The burn on Al’s back is a sunburn partly in virtue of its having been caused by exposure to the sun’s rays; a burn that looks and feels just the same is not a sunburn if it was caused by a heat lamp. The piece of paper with which Al just purchased a drink is a genuine U.S. dollar bill partly in virtue of its having been produced (in the right way) by the U.S. Treasury Department. A duplicate bill produced with plates and paper stolen from the Treasury Department is a counterfeit dollar bill, not a genuine one. Similarly, a certain event is Al’s purchasing a drink – an action – partly in virtue of its having been appropriately caused by mental items (or their neural correlates). An event someone else covertly produces by remote control – one including visually indistinguishable bodily motions not

32

Conversations on Human Action and Practical Rationality

appropriately produced by mental states of Al’s – is not a purchasing of a drink by Al, even if it feels to him as though he is in charge. According to this view, one important difference between actions and non actions is historical – a matter of how they are caused. I lack the space to defend my own view of the matter here. Readers who are curious about the details of my causal view of what actions are and my defense of it may wish to consult Motivation and Agency. 3. You say that when we are faced with a case of compulsion we still, nevertheless, have an agent performing an action. Is it the case that in any situation in which we can talk about agency, we can also talk about (moral) responsibility? Does that extend even to cases of compulsion? Or there is no necessary connection between the two notions of agency and responsibility? Agents, as I use the term, are simply beings that act. In my opinion, many agents are not morally responsible for anything – for example, bluejays, beavers, toads, and trout. In my view, human beings are morally responsible for some – but not all – of their intentional actions. As I see it, when a compulsive hand washer washes his hands for the fiftieth time in a single day, he is intentionally washing them, but he is not, at the time, morally responsible for washing them. Here’s a necessary connection between agency and moral responsibility: Necessarily, only agents perform actions for which they are morally responsible. But this is true because the following is true: Necessarily, only agents perform actions. 4. Which of the following are liable to be rational/irrational: action, agency, agent? Many different conceptions of rationality are evident in the literature on the topic, and the same goes for irrationality. If agency is the property of being an agent (as I say in response to question 2), then no legitimate conception of rationality will yield the result that agency itself is rational – or irrational. Even the property of rationality is not itself rational; and the property of irrationality is not irrational. However, if Joe is a rational agent, he has the property of being a rational agent – that is, rational agency is among his properties. The notion of rational agency derives from that of rational agent.

Alfred R. Mele

33

The term “rational” is opposed not only to “irrational” but also to “nonrational.” Plants, pots, and paramecia are nonrational, but not irrational. Irrational beings are imperfectly rational. A being wholly devoid of rationality is nonrational and therefore exempt from the charge of irrationality. Even if acting for a reason requires or entails some rationality, it may be irrational of us to do some of the things we do for reasons. To say that an agent is rational rather than nonrational is perhaps to say that he, she, or it can assess reasons for action and act on the basis of the assessments. What does it mean to say that an agent is rational rather than irrational? This is a hard question. Donald Davidson contends that any interpretable human agent is largely rational, in the sense that his beliefs, intentions, and the like generally cohere with one another and with his behavior (1985). If this is right, interpretable irrational human agents fall somewhere within the class of largely rational human agents; the difference between irrational and rational agents is a matter of degree. Davidson’s “paradox of irrationality” is interesting in this connection: “The underlying paradox of irrationality, from which no theory can entirely escape, is this: if we explain it too well, we turn it into a concealed form of rationality; while if we assign incoherence too glibly, we merely compromise our ability to diagnose irrationality by withdrawing the background of rationality needed to justify any diagnosis at all” (1982, p. 303). I have tackled this paradox in two domains – action that displays weakness of will (akratic action; see my reply to question 11) and motivationally biased belief. In Irrationality (1987) and elsewhere, I defend the view that when agents act akratically they act for reasons and that, in central cases of such action, they act contrary to rational judgments: “the background of rationality” required for that is in place. But insofar as their uncompelled actions are at odds with their rational judgments, they act irrationally. Similarly3, motivationally biased believers test hypotheses and believe on the basis of evidence. Again there is a background of rationality. But, owing to the influence of motivation, they violate general standards of epistemic rationality. To say that an action is irrational is to say that it was irrational of the agent to perform it. The kind of irrational action that has interested me the most is akratic action; it is discussed in my response to question 11.

3

See Irrationality and Self-Deception Unmasked, 2001.

34

Conversations on Human Action and Practical Rationality

5. You consider that the most basic notion to be defined in philosophy of action is precisely the notion of action. But when you address the issue of rationality/irrationality, you prefer to speak about agents and not about actions. Is this because you consider that the irrationality of an action always depends on the desires and beliefs of an agent? Is it the case that ‘rationality’ or ‘irrationality’ can only have an instrumental meaning? How would you define a position according to which there can be irrational beliefs or desires? An irrational belief of mine, as I understand the notion, is a belief that it is irrational of me to have. The same goes for an irrational desire. The irrationality attaches to the person – me, in the present example – for the belief, desire, intention, action, or whatever. I don’t see that there is anything special about action in this particular connection. 6. In what sense is the thing to do to be decided by what is rational? Are there limits to rationality? Depending on how rationality is to be understood, judgments to the effect that something is a rational course of action may rarely settle what “the thing to do” is. When I get to my office in the morning, there ordinarily are lots of different things that it would be rational for me to do, in an ordinary sense of “rational,” and way too many such things for me to do all of them. For example, in an ordinary sense of “rational,” it might be rational of me to spend several hours on this interview today even though spending those hours instead on any one of five other projects would also have been rational. I’m inclined to understand “rational course of action” as “rationally permissible course of action” (an analogue of “morally permissible course of action”). Because of my own limitations, I’m ducking the question about limits of rationality. 7. Are there reasons to act which are unconscious? What is the difference between acting consciously and acting for conscious reasons? Philosophical interest in reasons seems to have hit an all-time high. The topic is so lively now that the philosophical study of reasons may even deserve a name of its own. One may look to Latin or Classical Greek for a name; but because English is the language in which the overwhelming majority of this work is published, a more English-sounding name would be appropriate. In Mele 2007, I suggested Reasonology.

Alfred R. Mele

35

Philosophical work on what its authors call “reasons for action” tends to be guided by concerns with two distinct but related topics: the explanation of intentional actions; and the evaluation of intentional actions or their agents. In work dominated by the explanatory concern, reasons for action tend to be understood as states of mind (for example, as certain kinds of combinations of beliefs and desires à la Donald Davidson). In some work dominated by the evaluative concern, typical reasons for action are understood as states of, or facts about, the agent-external world. On both conceptions of reasons, it would seem that some reasons for action may be “unconscious”, in the sense that the agent who has a reason to do something is not conscious of it. For example, if reasons are desirebelief pairs, Joe may have an unconscious desire to offend a certain person who insulted him yesterday, and that desire may be part of a reason for which Joe offends him. And if reasons are states of, or facts about, the external world, it would seem that we may be unconscious of some reasons we have to do things. The fact that Joe’s plane is leaving two hours later than scheduled may be a reason for Joe not to hurry to the airport, even though he is not conscious of this fact because he forgot to check on the flight before he left home. I am not sure how to interpret the expression “acting by conscious reasons”. Doing something consciously – for example, consciously signing my name on a petition – is just a matter of doing it and being conscious of doing it. I may consciously sign my name on a petition without being conscious of some reason for which I am doing it. 8. What explains action and how? What is the role of deliberation in rationality? The first question here is, in a way, the topic of two of my books – Motivation and Agency, which I already mentioned, and Springs of Action (1992). I suppose that one of my limitations is that I don’t know how to answer the question in anything shorter than a book. Ann asks: “Why did Bob fight with Carl? What explains his doing that?”. Donna replies: “Carl dared Bob to fight with him.” This might satisfy Ann; perhaps all she wanted to know was what prompted Bob to fight with Carl. But she might have wanted to know more. In fact, she responded to Donna’s reply as follows: “Well, often people who are dared to fight just walk away. So I feel that you haven’t actually answered my question.” Donna understood Ann’s complaint. She offered more information and highlighted a belief: “Bob believes that backing down in situations like this is a sign of weakness and that signaling such weakness

36

Conversations on Human Action and Practical Rationality

makes one’s life harder.” Donna assumed that Ann would understand that Bob wanted to avoid making his life harder. In my view, intentional actions can be adequately explained in terms of such things as agents’ beliefs, desires, and intentions, and adequate explanations in this sphere are causal explanations. If Donna managed to explain why Bob fought with Carl, then such facts as that Bob believed that backing down in situations of the sort in which he found himself would make his life harder and that he wanted to avoid making his life harder are causally relevant to his fighting with Carl.4 I turn to the second part of the question. It is a truism that practical deliberation is deliberation about what to do. One might say that it is reasoning conducted with a view to answering the question “What shall I do?” (That question must be distinguished from “What will I do?” The former calls for a decision, or at least the acquisition of an intention. The latter calls for a prediction.) Practical deliberation, as I understand it, is an inferential process, involving evaluative premises, driven at least partly by motivation to settle on what to do, and it may take the form of reasoning about what it would be best to do or what it would be “good enough” to do5. The settling motivation that I mentioned disposes agents to intend in accordance with the reasoning’s evaluative conclusion, given that agents become settled on a course of action in forming or acquiring an intention. Their being so disposed supports the primary point of practical deliberation, which is to lead to a satisfactory resolution of one’s practical problem. An agent who judges it best to A – or “good enough” to A – but is still unsettled about whether to A has not resolved his practical problem. An agent may deliberate rationally or irrationally, and it is not a necessary condition for an agent’s acting rationally at a time that the action at issue was the relatively direct product of deliberation. Ordinarily, a lifeguard who reacts to a glimpse of a child in danger of drowning would not need to deliberate about what to do in order for it to be true that his rescuing the child was a rational action.

4

For details and argumentation, see Springs of Action and Motivation and Agency. See Springs of Action, chapter 12, Autonomous Agents, 1995, chapter 2, and Motivation and Agency, chapter 4.

5

Alfred R. Mele

37

9. Is the result of deliberation an intention or the action itself? If deliberation ends in an intention, then what happens between the producing of the intention and the action itself? On the other hand, if deliberation ends in an action, how is akrasia possible? As I mentioned in my response to the previous question, practical deliberation, as I understand it, is an inferential process, involving evaluative premises, driven at least partly by motivation to settle on what to do, and it may take the form of reasoning about what it would be best to do or what it would be “good enough” to do. As I see it, deliberative conclusions are beliefs – for example, the belief that it would be best to quit smoking, beginning today6. How are these beliefs related to intentions? In my view, a common route from a belief that it is best to A produced by practical evaluative reasoning to an intention to A is a default route7. Consider default procedures in computing – for example, a standard procedure in common word-processing programs for the spacing of text. When authors create new files, any text they type will be displayed single-spaced, unless they preempt this default condition of creating a file by entering a command for an alternative form of spacing. When authors do not issue a preemptive command and their programs and hardware are working properly, entering a new file systematically has the identified result. Similarly, in the absence of preemptive conditions (for example, strong opposing desires) in normally functioning human beings, their acquiring a belief that it is best to A might systematically issue in an intention to A. Acquiring such a belief might figure importantly in the production of an intention to A in particular cases, even if the transition from such beliefs to such intentions sometimes is blocked – even if the disposition to make that transition is defeasible. The basic idea is that “normal human agents are so constituted that, in the absence of preemption, judging it best . . . to A issues directly in the acquisition of an intention to A” (Mele 1992, p. 231). In simple cases involving little or no motivational opposition, the transition from judgment to intention is smooth and easy. In such cases, agents who conclude that it is best to A have no need to think about whether to intend to A; nor, given their motivational condition, do they need to exercise self-control in order to bring it about that they intend to A. No special intervening effort of any sort is required. The existence of a default procedure of the sort at issue in 6

See Springs of Action, chapter 12, Autonomous Agents, chapter 2, and Motivation and Agency, chapter 4. 7 See Mele 1992, chapter 12.

38

Conversations on Human Action and Practical Rationality

normal human agents would help to explain the smoothness and ease of the transition. Indeed, we should expect an efficient action-directed system in beings who are capable both of making deliberative judgments and of performing akratic actions to encompass such a procedure. Special energy should be exerted in this connection only when one’s better judgments encounter significant opposition. If and when there is an akratic failure to intend in accordance with one’s better judgments, opposition is encountered: something blocks a default transition; something preempts the default value of the judgment. In Mele 1992 (pp. 230-34), I distinguish among three kinds of case in which an agent’s conclusion that it is best to A (or his concluding Afavoring better judgment) is opposed by competing motivation: (1) a default process unproblematically generates a belief-matching intention even in the face of the opposition; (2) a belief-matching intention is formed even though the default route to intention is blocked by the opposition; (3) the motivational opposition blocks the default route to intention and figures in the production of an akratic intention. What is needed is a principled way of carving up the territory. In Mele 1992, I suggest that a belief-matching intention is produced (in the normal way) by default, as opposed to being produced via a distinct causal route, when and only when (barring causal overdetermination, the assistance of other agents, science fiction, and the like) no intervening exercise of self-control contributes to the production of the intention (p. 233). (Sometimes opposing motivation is sufficiently weak that no attempt at self-control is called for.) If the move from belief to intention does not involve a special intervening effort on the agent’s part, the intention’s presence typically may safely be attributed to the operation of a default procedure. In my view, self-control figures importantly in explaining why, when a default route from better judgment to intention is blocked, we sometimes do, and sometimes do not, intend on the basis of our better judgments. Barring the operation of higher-order default processes, overdetermination, interference by intention-producing demons, and so on, whether an agent intends in accordance with his better judgment in such cases depends on his own efforts at self-control. In simple cases of self-indulgence, the agent makes no effort at all to perform the action judged best, or to form the appropriate intention. In other cases in which an agent believes it best to A, he might attempt in any number of ways to get himself to A or to intend to A. He might try focusing his attention on the desirable results of his A-ing or on the unattractive aspects of his not A-ing. He might generate vivid images of both, or utter self-commands. If all else fails, he might seek help from a behavioral therapist. Whether his strategies work will

Alfred R. Mele

39

depend on the details of the case; but strategies such as these can have a salutary effect, as empirical research on delay of gratification and behavior control amply indicates8. Why do we reason about what it would be best to do? Sometimes, at least, because we are concerned to do what it would be best to do and have not yet identified what that would be. (Often, we may settle – even rationally settle – on the first alternative that strikes us as good enough: for example, when we take little to be at stake and suppose that the cost required to identify the best alternative would probably outweigh the benefits.) In such cases, if things go smoothly, better judgments issue in corresponding intentions. And it is no accident that they do, given what motivates the reasoning that issues in the judgments. Of course, if common sense can be trusted, things do not always go smoothly: we can identify the better and – owing partly to the influence of recalcitrant desires – intend the worse. If this happens, the fact that it does would show, not that better judgments have no role to play in the etiology of intentions and intentional behavior, but rather that, in human beings as they actually are, an agent’s judging it best to A does not ensure that he forms or acquires a corresponding intention. In Mele 1987, I attempt to explain how this can be true, how better judgments may be rendered ineffective by competing motivation. It is tempting to speculate about how a default procedure of the sort that I have sketched might have emerged in us. Any speculation about how agents like us come about – agents who sometimes reason about what it would be best to do with a view to settling on what to do and then intend and act on the basis of their better judgments – should attend to the emergence in such agents of what mediates between judgment and action. Agents like us would be well served by a default procedure of the kind sketched: a procedure of this kind conserves mental energy, obviating a need for a special effort or act, in each case, to bring it about that, having judged it best to A, one also intends to A. Special efforts would be required only under special circumstances.

8

For discussion, see Mele 1987 and 1995.

40

Conversations on Human Action and Practical Rationality

10. A) But how are intentions and decisions related? In which moment is the “better judgment” formed? Can we say that a decision is the “better judgment”? B) If akrasia happens because there is a gap between the assessment of the agent and the strength of her desires, can one still say that akrasia is an inconsistency between reason and desire? When we are faced with a case of akrasia, is it that what happens is that there is a stronger desire, which is in itself motivating, and which erases the evaluative belief according to which a certain action is better? In Mele 2003, I motivate the view that decisions to do things – practical decisions – are momentary mental actions of intention formation. In my view, intentions are executive attitudes toward plans; plans – which range from simple representations of prospective “basic” actions to complex strategies for achieving remote goals – constitute the representational content of intentions (Mele 1992). The momentary action of intention formation in which deciding to A consists is, more fully, an action of executive assent to a pertinent first-person plan of action (Mele 2003, ch. 9). I discuss executive assent shortly. Deciding to A, as I understand it, is not to be confused with any process that issues in deciding to A, including, for example, deliberation about what to do, in the case of deliberation-based deciding. And deciding to A, as I conceive of it, does not precede the onset of the intention to A formed in the act of deciding. Instead, what it is to decide to A is to form – actively – an intention to A. The intention arises in that momentary intention-forming action, not after it. Are all intentions formed in acts of deciding? Consider the following: “When I intentionally unlocked my office door this morning, I intended to unlock it. But since I am in the habit of unlocking my door in the morning and conditions . . . were normal, nothing called for a decision to unlock it” (Mele 1992, p. 231). If I had heard a fight in my office, I might have paused to consider whether to unlock the door or walk away, and I might have decided to unlock it. But given the routine nature of my conduct, there is no need to posit an action of intention formation in this case. My intention to unlock the door may have been acquired without having been actively formed. In short, it might have been nonactionally acquired. If, as I believe, all decisions about what to do are prompted partly by uncertainty about what to do9, then when there is no such uncertainty, no decisions will be made. This is not to say that, in such situations, no intentions will be acquired. 9

See Mele 2003, chapter 9.

Alfred R. Mele

41

I mentioned that, in my view, deciding to do something is an action of executive assent to a first-person plan of action. My notion of executive assent is straightforward. If you tell me that Ronaldo is an excellent football player and I express complete agreement, I thereby assent to your claim. This is overt cognitive assent. If you propose that we watch Ronaldo play tonight at the stadium and I express complete acceptance of your proposal, I thereby assent to your proposal. This is overt executive assent: I have agreed to join you in executing your proposal for joint action. Possibly, my overt act of assenting to your proposal was a matter of my giving voice to a nonactionally acquired intention to join you in watching Ronaldo play. For example, upon hearing your proposal, I might not have been at all uncertain about what to do; straightaway, I nonactionally acquired an intention to join you, and I voiced that intention in an overt act of assent. Or I might have weighed the pros and cons, judged that it would be best to join you, and, on the basis of that judgment, nonactionally acquired an intention to join you. However, there is also a distinctly different possibility. Perhaps, because I already had plans and because your offer was attractive, I was uncertain about what to do. Perhaps, upon reflection, I judged that I could revise my plans without much inconvenience but was still uncertain about what to do, because my prior plans were attractive as well. And perhaps I performed a mental action of assenting to your proposal and then expressed that inner assent to you. In performing that mental action, if that is what happened, I decided to join you: my mentally assenting to your proposal was an act of intention formation, an act of settling on joining you to watch Ronaldo play tonight. You will have noticed that I have used “decisions [to A]” as a synonym for “decidings [to A].” That is one use of the term. “Decision” may also be used to refer to the intention formed in an act of deciding and to what someone decides, as in “His decision was to go to the arena.” I make these points to forestall confusion. I reject the identification of practical decisions with better judgments. One can decide to do something without judging it best to do it, and one can judge it best to do something without deciding to do it.10 Also, in paradigmatic cases of akratic action, as I understand them, the belief that it is best to A is not “erased.” Instead, the belief persists while the agent performs an uncompelled, intentional action that it contrary to it11. To answer the question about “consistency,” I’d need to know what the questioner means by the term in this connection. 10 11

For argumentation, see Mele 1987, pp. 19-20, 43-44. See Mele 1987 on strict akratic action.

42

Conversations on Human Action and Practical Rationality

11. So, how do you think akrasia is possible? The classical Greek term akrasia is formed from the alpha privative (basically, a negation sign) and kratos, meaning strength or power. The power at issue is the power to control oneself in the face of actual or anticipated temptation. So akrasia is deficient self-control. Self-control, in this sense, may be understood as constituted primarily by a robust capacity to see to it that one does what one believes to be best or better on the whole when tempted to do otherwise. The self-controlled person, Aristotle writes, “is in such a state as . . . to master even those [temptations of a certain kind] by which most people are defeated,” and the akratic person “is in such a state as to be defeated even by those . . . which most people master” (Nicomachean Ethics, 1150a11-13). In Plato’s Protagoras, Socrates says that the common view about akratic action is that “many people who know what it is best to do are not willing to do it, though it is in their power, but do something else” (352d). Here he raises (among other issues) the central question in subsequent philosophical discussion of akrasia: Is strict akratic action possible? That is, is it possible for people to perform free intentional actions that they consciously believe at the time of action to be inferior to another live option? The belief at issue is based on the agent’s own values and opinions. A feature of paradigmatic strict akratic actions that typically is taken for granted and rarely made explicit is that the beliefs with which they conflict – for example, that it would be better to study tonight than to attend a party – are rationally acquired. In virtue of clashing with the agent’s rationally acquired beliefs of this kind, akratic actions are subjectively irrational (to some degree, if not without qualification). There is a failure of coherence in the agent of a kind directly relevant to assessments of the agent’s rationality. To some theorists, the threat that strict akratic action poses to our ability to make sense of human action seems so severe that they deem strict akratic action conceptually or psychologically impossible. Many others try to accommodate strict akratic action in a general theory of human action. Much of my first book, Irrationality, was devoted to this project. I’ll offer a brief summary of my view about how strict akratic action is possible. My view rests partly on two theses, both of which I defended. 1. The motivational force of our desires is not always in line with our evaluation or assessment of the “objects” of our desires (i.e., the desired items).

Alfred R. Mele

43

2. Typically, decisive better judgments are formed, in significant part, on the basis of our evaluation or assessment of the objects of our desires.

If both theses are true, we should not be surprised that sometimes, although we judge it better to A than to B, we are more strongly motivated to B than to A. Given how our motivation stacks up, it should also be unsurprising that we B rather than A. Thesis 1, as I explained in Irrationality (ch. 6), enjoys considerable empirical support. It is confirmed as well, I argued, by common experience and by various thought experiments (chs. 2 and 3). Desire-strength is influenced not only by our assessment or evaluative rating of the objects of desires, but also by such things as the perceived proximity of prospects for desire-satisfaction, the salience of desired objects in perception or in imagination, and the manner in which we attend to desired objects. Thesis 2 is a major plank in a standard conception of practical reasoning. In general, when we reason about what to do, we inquire, not about what we are most motivated to do, but rather about what it would be best, or better, or “good enough” to do. When we ask such questions while possessed of conflicting desires, our answers typically rest significantly on our assessments of the objects of our desires, assessments that need not be in line with the motivational force of those desires. A few hours ago, an agent judged it better on the whole to A than to B, but he now has a stronger desire to B than to A. Two versions of the case merit attention. In one, along with the change in desire strength, there is a change of judgment. For example, last night, after much soul-searching, Al formed a decisive judgment favoring not eating after-dinner snacks for the rest of the month and desired more strongly to forego them than to indulge himself; but now, a few hours after dinner, Al’s desire for a snack is stronger than his desire for the rewards associated with not snacking, and he decisively judges it better to have a snack than to refrain. In another version of the case, the change in relative desire strength is not accompanied by a change of judgment. Al retains the decisive judgment favoring not eating after dinner, but he eats anyway. Assuming that Al eats intentionally and is not compelled to eat, this is a strict akratic action. Empirical studies of the role of representations of desired objects in impulsive behavior and delay of gratification (reviewed in Irrationality, pp. 88-93) provide ample evidence that our representations of desired objects have two important dimensions, a motivational and an informational one. Our decisive judgments may be more sensitive to the informational dimension of our representations than to the motivational dimension, with the result that such judgments sometimes recommend courses of action that are out of line with what we are most strongly motivated to do at the

44

Conversations on Human Action and Practical Rationality

time. If so, strict akratic action is a real possibility – provided that at least some intentional actions that conflict with agents’ decisive judgments at the time of action are not compelled. A discussion of compulsion would lead quickly to the issue of free will, which I will not discuss here12. It is worth noting, however, that unless a desire is irresistible, it is up to the agent, in some sense, whether she acts on it; and it is typically thought that relatively few desires are irresistible. Of course, a proper appreciation of the latter idea would require an analysis of irresistible desire13. It may suffice for present purposes to suggest that, often, when we act against our better judgments, we could have used our resources for self-control in effectively resisting temptation. Normal agents can influence the strength of their desires in a wide variety of ways. For example, they can refuse to focus their attention on the attractive aspects of a tempting course of action and concentrate instead on what is to be accomplished by acting as they judge best. They can attempt to augment their motivation for performing the action judged best by promising themselves rewards for doing so. They can picture a desired item as something unattractive – for example, a chocolate pie as a plate of chocolate-coated chewing tobacco – or as something that simply is not arousing. Desires typically do not have immutable strengths, and the plasticity of motivational strength is presupposed by standard conceptions of self-control. Occasionally we do not act as we judge best, but it is implausible that, in all such cases, we cannot act in accordance with these judgments.14 12. How is a belief-desire-intention vocabulary related to the notion of will? In Mele 2006, in explaining why I define free will as the ability to act freely and treat free action as the more basic notion, I complain that “I often cannot tell what authors mean by ‘will’ in ‘free will’” and I report my opinion that “I am blameless for this ignorance” (p. 17). I myself have no special use for the noun “will”. If I were to be presented with a specific account of the will, I might be able to say how the will, so understood, is related to such things as beliefs, desires, intentions, decisions, exercises of self-control, and trying, as I understand them.

12

But see my Free Will and Luck, 2006a. For an attempted analysis, see Springs of Action, chapter 5. 14 This suggestion is defended in Irrationality, chapter 2 and Autonomous Agents, chapter 3. 13

Alfred R. Mele

45

13. What do you think is the relation between folk psychology and the philosophy of action? How do you think it works? One project in the philosophy of action is constructing a theory about how intentional actions are caused in terms of such items as beliefs, desires, and intentions (or their neural realizers). The terms “belief”, “desire”, “intention”, and “action” are all folk psychological terms associated with folk conceptions of these things. Extant work on this project involves various attempts to clarify the meaning of the terms just mentioned, and the starting point is ordinary usage. Some philosophers of action say that they are concerned to understand or analyze various “folk concepts.” Others are willing to refine (apparent) folk concepts for certain explanatory purposes. I’ll take intentional action as an example. Joshua Knobe, Thomas Nadelhoffer, and others have done some interesting work on what lay people do and don’t count as intentional actions15. Some of this work provides evidence that most lay folk conceive of intentional action in such a way that some foreseen “side-effect” actions count as intentional actions – for example, harming the environment in a case in which one knows that one will harm it by executing a profit-making plan one intends to execute but neither wants nor intends to harm it. Philosophers with the project of analyzing “the folk concept” of intentional action should definitely take note of this work. What about philosophers whose primary business is building an adequate theory about how agents’ intentional actions are caused? Most such philosophers are concerned with a species of goaldirected action, and “intentional action” may be the ordinary English expression that most nearly names their target. If it is discovered that most lay folk are happy to count some non-goal-directed actions as intentional and if it is decided that “the folk concept” of intentional action embraces such actions, philosophers with the project of developing a theory of action-production can retain their target and accommodate ordinary usage by modifying their name for it; they can call it “intentional goal-directed action.” If that sounds redundant, bear in mind that agents of goal-directed actions may be as simple as mosquitos; and mosquitos presumably do not act intentionally.

15

See Knobe and Nichols 2008 for references.

46

Conversations on Human Action and Practical Rationality

14. What is your opinion regarding experimental philosophy? Specifically in the philosophy of action, what would be its use? I personally find Experimental Philosophy – specifically, the kind that features surveys of lay people – most useful in the philosophy of action in connection with the claims some philosophers make that they are analyzing the folk (or ordinary) concept of intentional action, intention, or the like. For example, Hugh McCann, in a paper defending the Simple View – that is, the thesis that, necessarily, S intentionally A-ed only if S intended to A – says that “the Simple View . . . pertains to the everyday concept of intending, not a stipulated one” (1991, p. 33). Presumably, he would have said the same, if asked, about the concept of intentional action to which the view pertains. And some survey studies give us good reason to doubt that McCann’s conception of intentional action matches “the folk concept” of such action (see my response to question 1 above). As I mentioned in my response to question 13, I do not believe that philosophers who work on intentional action are required to try to capture “the folk concept” of intentional action. Indeed, I believe – partly on the basis of some survey experiments that Fiery Cushman and I conducted16 – that there is no such thing as the folk concept of intentional action and that, instead, there are different interconnected conceptions of intentional action in folk thought. 15. In your opinion, what are the most relevant links between, respectively, philosophy of action and philosophy of mind and philosophy of action and moral philosophy? I see much of the philosophy of action as part of the philosophy of mind. One who regards the subject matter of the philosophy of mind as having at its core some aspect of what lies between environmental input to beings with minds and behavioral output may be inclined to see the philosophy of action as concerned only with the output end of things. That would be a mistake. Many intentional actions depend for their development on the processing of input – both from the environment and from the body. Try to imagine yourself cooking a meal or driving to work without processing such input. Conceptualizing and explaining intentional actions of intelligent agents – or some central subset of such actions – is the primary business of the philosophy of action, and properly conducting that business requires a firm understanding of most of the ins and outs of the philosophy of mind. 16

See Mele and Cushman 2007, Cushman and Mele 2008.

Alfred R. Mele

47

There are many connections between the philosophy of action and moral philosophy. Indeed, philosophers of action and moral philosophers write about some of the same topics: for example, moral responsibility, free will, weakness of will, the nature of reasons for action, the nature of motivation (including moral motivation), and the explanation of actions. Cross fertilization here can be very productive. The action theoretic book of mine that extends furthest into moral philosophy is Motivation and Agency (2003). There I take up such issues as the role of first-person moral ought judgments in action production and whether such judgments essentially encompass motivation to act. The work in moral philosophy that has had the biggest impact on me is Aristotle’s Nicomachean Ethics; and in my action theoretic work on such topics as weakness of will, motivation, and reasons for action, I pay close attention to the work of such moral philosophers as Jonathan Dancy, R. M. Hare, Thomas Nagel, T. M. Scanlon, and Michael Smith. 16. How do you think your own work has contributed to the field? What do you think are your most important contributions? What are your plans for future research? I suppose I consider my most important contribution to the field to be something approaching a general philosophy of action. I started with pretty specific questions: for example, questions about how to account for akratic actions, what is involved in exercising self-control, whether intentions play a role in producing actions that goes beyond any role that can be played by combinations of beliefs and desires, what it is for a desire to be irresistible, how decisions are related to intentions, what is it for a state essentially to constitute motivation to act, what can be added to an ideally self-controlled agent (self-control being understood as the contrary of akrasia) to yield an autonomous or free agent, and how compatibilists and libertarians about free will should reply to apparent problems that different kinds of luck pose for their theories. And I now have, among other things, something approaching a general theory about how actions are to be explained that includes accounts of such explanatory items as desires (including action-desires), intentions, and decisions. My philosophy of action also yields judgments about such things as what makes intentional actions intentional and what makes free actions free. I recently completed a book entitled Effective Intentions: The Power of Conscious Will (2009). In it, I take up such claims as the following, each of which has been defended in the scientific literature on free will and consciousness: your brain routinely decides what you will do before you

48

Conversations on Human Action and Practical Rationality

become conscious of its decision; there is only a 100 millisecond window of opportunity for free will, and all it can do is veto conscious decisions, intentions, or urges; intentions never play a role in producing corresponding actions; and free will is an illusion. I show that the evidence offered to support these claims is sorely deficient. I also show that there is strong empirical support for the thesis that some conscious decisions and intentions have a genuine place in causal explanations of corresponding actions. In short, there is weighty evidence of the existence of effective conscious intentions or the power of conscious will. My final chapter wraps things up with a discussion of imaginary scientific findings that would warrant bold claims about free will and consciousness of the sort examined in the book. I plan to continue working on issues at the intersection of the philosophy of action and various sciences. I will work as well on some purely conceptual issues in the philosophy of action, including the nature of action that exhibits weakness of will, epistemic requirements on moral responsibility, and the nature and conceptual possibility of free will. 17. If you had to choose just five research topics, connected to your work, for the next 30 years, what would you choose? Thirty more years of work by me is optimistic! But here are the topics I would choose, if I had to choose right now: Free will; the bearing of findings in neuroscience, social psychology, etc., on whether we ever act freely and whether we are morally responsible for any of our actions; roles played by conscious and unconscious intentions in producing actions; roles of consciousness more generally in action production; action explanation. I discuss many of these issues in my latest book, Effective Intentions: The Power of Conscious Will.

References Aristotle. 1915. Nicomachean Ethics. Vol. 9 of W. Ross, ed. The Works of Aristotle. Oxford University Press. Cushman, F. and A. Mele. “Intentional Action: Two-and-a-half Folk Concepts.” In Knobe and Nichols 2008. Davidson, D. 1982. “Paradoxes of Irrationality.” In R. Wollheim and J. Hopkins, eds. Philosophical Essays on Freud. Cambridge University Press. —. 1985. “Rational Animals.” In E. LePore and B. McLaughlin, eds. Actions and Events, Basil Blackwell, pp. 473-480.

Alfred R. Mele

49

Knobe, J. and S. Nichols, eds. 2008. Experimental Philosophy. New York: Oxford University Press. McCann, H. 1991. “Settled Objectives and Rational Constraints.” American Philosophical Quarterly 28: 25-36. Mele, A. 1987. Irrationality: An Essay on Akrasia, Self-Deception, and Self-Control. New York: Oxford University Press. —. 1992. Springs of Action: Understanding Intentional Behavior. New York: Oxford University Press. —. 1995. Autonomous Agents: From Self-Control to Autonomy. New York: Oxford University Press. —. 2001. Self-Deception Unmasked. Princeton University Press. —. 2003. Motivation and Agency. New York: Oxford University Press. —. 2005. “Action.” In F. Jackson and M. Smith, eds. Oxford Handbook of Contemporary Philosophy. Oxford University Press. —. 2006. Free Will and Luck. New York: Oxford University Press. —. 2006b. “Action.” In D. Borchert, ed. Encyclopedia of Philosophy: Second Edition. Macmillan. —. and F. Cushman. 2007. “Intentional Action, Folk Judgments, and Stories: Sorting Things Out.” Midwest Studies in Philosophy 31: 184201. —. 2007. “Reasonology and False Beliefs,” Philosophical Papers 36: 91118. —. 2009. Effective Intentions: The Power of Conscious Will. New York: Oxford University Press. —. 2009b. “Action and Mind.” In P. Calvo and J. Symons, eds. Routledge Companion to the Philosophy of Psychology. Routledge. —. 2009c. “Free Will.” In W. Banks, ed. Encyclopedia of Consciousness. Elsevier. —. 2009d. “Intention and Intentional Action,” in B. McLaughlin, A. Beckermann, and S. Walter, eds. Oxford Handbook of Philosophy of Mind. Oxford University Press. —. n.d. “Causation, Action, and Free Will,” in H. Beebee, C. Hitchcock, and P. Menzies, eds., Oxford Handbook of Causation. Oxford University Press, forthcoming. Nagel, Thomas. 1986. The View from Nowhere. Oxford University Press. Plato. 1953. Protagoras. In B. Jowett, trans. The Dialogues of Plato. Clarendon Press. Velleman, J. David. 1992. “What Happens When Someone Acts?” Mind 101: 461-81.

HUGH J. MCCANN TEXAS A&M UNIVERSITY

1. In your view, what are the most central (or important) problems in the philosophy of action? The most important problem in the philosophy of action is to determine the nature of action—that is, to determine how it is that actions differ from other events. The problem was famously formulated by Wittgenstein in his question, “What is left over if I subtract the fact that my arm goes up from the fact that I raise my arm?”1 This problem is directly related to a second important problem, that of understanding the nature of agency. An account that sheds light on one of these issues is bound to illuminate the other, and in turn to assist us with other problems about action, most especially that of free will, and other issues having to do with responsibility and practical rationality. As to what differentiates an action such as raising one’s arm from the mere event of the arm rising, there has tended to be a reasonably broad consensus on two points. First, when an agent performs an action, intention or purpose is somewhere in the picture, a point often made by saying that when an agent acts “what he does” is, under some description, intentional. Second, action theorists have tended to say that part of the difference between actions like arm raising and mere events like the arm going up is that in the former case the agent “brings about” the motion of the arm, whereas in the latter the arm’s motion is owing to other causes, or at least to a causal process which is such as to compromise agency. From here on, however, there is serious disagreement, the contours of which will begin to emerge if we focus on the phrase “what he does” in the first of the points just mentioned. There are actually two issues to be considered here. The first concerns the fact that it is frequently possible for an agent to accomplish many things in a single exercise of agency. Thus, to cite an example of Davidson’s, an agent might by moving his finger flip a switch, turn on a light, illuminate a room, and alert a prowler. Or, someone might by raising 1

Ludwig Wittgenstein, Philosophical Investigations (New York: Macmillan, 1953), sec. 621.

52

Conversations on Human Action and Practical Rationality

an arm signal a turn while driving. This leads to the problem of the individuation of action: whether the many things “done” in a single exercise of agency are to be counted as just one action, as multiple discrete actions, or as overlapping segments of causal chains. But the problem of individuation, although once the focus of great attention, is largely independent of the central problem of the nature of action, the problem of determining what exactly is “done” when an agent performs an overt movement like raising an arm—of how, if at all, we are to differentiate between the upward motion of the arm and the act of raising it. Early efforts to answer this question were powerfully influenced by logical behaviorism, which in one form or another held the allegiance of a great many so-called “ordinary language” philosophers. These thinkers largely disdained the idea that overt actions should be viewed as arising out of acts of physical exertion or muscle flexing, or from mental acts of volition. Rather, they viewed actions like raising an arm or moving a finger as basic actions: that is, roughly, actions not done by doing anything else. To adopt this stance is to hold in effect that ontologically speaking the action of raising the arm is identical with the arm’s rising, which counts as an action provided the circumstances are right. Ordinary language philosophers tended to say that the arm’s motion was an action when viewed from the perspective of the “language game” of action, which was implicitly antideterministic, and was held to have legitimate standing even if from a mechanistic point of view the motion of the arm might be held to be biologically caused.2 From the actional perspective mechanistic principles were held to have no standing; considered as action, the motion of the arm was instead to be explained teleologically, in terms of the reasons for which the agent had acted. There are, however, major problems with this account, of which we may consider two. First, the sequestering of the language of action from that of mechanistic explanation appears artificial, and to the extent that the former is antideterministic simply fails to address the obvious clash between the two. Second, since it eschewed talk of anything that might be going on within the agent, this view could offer no account of what it is for an agent to act for one reason rather than another. This is important, because someone could in principle have multiple reasons that would justify an action of a certain type, yet perform the action only for one of them. For example, someone sitting in a seminar might wish to ask a question, and he might also wish to relieve some stiffness in his shoulder. Both can be 2

For this kind of view see A. I. Melden, Free Action (London: Routledge & Kegan Paul, 1961).

Hugh J. McCann

53

accomplished by raising his arm, but the agent might raise the arm for only one of these reasons. What would be the difference between his raising an arm in order to signal a question, and his raising it in order to relieve stiffness? If we are unable to appeal to anything mechanistic a causal answer is excluded, and if we refuse to allow talk of any attendant mental activity we can offer no account that would treat the agent himself as settling upon the reason for which he acts. But then we are left with no understanding of how it is that agents are able to act for some reasons but not others. Owing to these and other failures, ordinary language treatments of action were eventually abandoned, usually in favor of either of two other accounts. The first, which may be called the causal theory of action, is based on the idea that only by invoking some sort of causal relationship would it be possible to resolve the problem of what it is to act for one reason rather than another. And the simplest solution was to postulate a direct causal relationship between the appropriate reason and the bodily motion held to be performed “for” it. Suppose we think of the agent’s reason as consisting of a desire on his part—say, to signal at a seminar he is attending that he wishes to ask a question—and a belief that he can accomplish this end by raising his arm. A causal theory of action might hold that if this desire and belief cause the agent’s arm to rise, then its movement counts as an action of the agent raising his arm, and that the action is performed by the agent for the reason that he wished to ask a question and believed he could signal the question by this means. Or, we could simply say that the agent raised his arm in order to signal a question. Again, however, the account is wanting in two respects. One is that as it stands, it assumes what amounts to a reductive stance on the nature of intention, implying that intention can be understood simply as a combination of desire and belief. This is not entirely wrongheaded. It is fair to say that an action performed for a reason is intentional; and there is nothing wrong with treating an agent’s desires and beliefs, separately or together, as reasons for action. But the account as it stands implies that there is nothing to intention beyond desire and belief, that to have an intention to act a certain way is simply to have a reason so to act. This however is false. I may desire to see Rome, and believe I can do so if I vacation in Italy this summer. It hardly follows, however, that I intend to go there. Intention requires that I have made up my mind: that I be committed to going, which in turn implies a commitment to forming an effective plan for going and to carrying out that plan in a timely way.3 So even if the 3 Michael Bratman, Intention, Plans and Practical Reason (Cambridge, MA: 1987), ch. 2.

54

Conversations on Human Action and Practical Rationality

causal theory is on the right track, it must be elaborated to permit independent states of intending: to allow that agents who have reasons to act have yet to settle the question of whether they will act, and if so why. They need to decide or otherwise make up their mind what to do, thereby forming the intention that will govern their action. But this alone will not suffice, because it is possible for mental states, whether of desire and belief or of intention, to cause bodily changes of the indicated sort, but in such a way that no action at all occurs, much less one that is intentional. Thus, in an example of Davidson’s, a mountaineer desires to secure his own safety and knows he may do so by releasing his hold on a rope from which a fellow climber is dangling. The desire and belief so unnerve him that his grip on the rope is loosened, allowing his companion to fall.4 The mountaineer performs no action at all in this example, and neither would there be any action if instead of a desire and belief, it was an actual intention to let the companion fall that unnerved him. Yet behavior appropriate to an action of allowing the companion to fall occurs, and is caused by just the sort of states the causal theory specifies. There are numerous cases of this kind—cases of wayward causal chains—and they show that mere causation by mental states that encapsulate an agent’s reasons, or even by states of intending, cannot make the difference between action and mere bodily motion. A fortiori, such causation cannot explain what agency consists in. Whether the pathway from having reasons to acting is causal or not, something is being left out here, something that must be understood if the nature of agency is to be elucidated. What is being left out, in my view, is volition, the inner activity by which agency is normally exercised when we perform overt actions. It is important to note in this connection that, independent of the problems with the causal theory, there is plenty of reason for denying that overt movements like raising an arm count as basic actions. Everyone knows that it is possible to try to raise an arm and fail to move it at all, and the same sort of thing can be said for any type of voluntary movement. Indeed, the clinical literature on testing for neurological deficiency makes it quite clear that even paralytics can try to perform movements without any success at all.5 What is it, then, to try to perform a bodily movement? Trying is not a species of action. Rather, “trying” is a general name for going about the performance of an action, a term usually employed in 4

Donald Davidson, “Freedom to Act”, in Essays on Freedom of Action, ed. Ted Honderich (London: Routledge & Kegan Paul, 1973), 139-56, pp. 153-54. 5 See for example William W. Campbell, DeJong’s The Neurologic Examination 6th ed. (Baltimore: Lippincott Williams and Wilkins, 2005), ch. 27.

Hugh J. McCann

55

cases of failure, or where success is doubtful. And anything we might do can count as an attempt to act depending on the circumstances. When John Hinckley tried to assassinate Ronald Reagan, his attempt consisted in shooting Reagan. When persons of normal physical ability try to perform bodily movements their attempts consist in acts of muscular exertion aimed at getting the movement accomplished. When the attempt succeeds the movement is performed by engaging in the exertion, so that the exertion counts as a more basic activity than the movement itself. But a completely paralyzed person is not capable even of muscular exertion. Thus the attempts of paralytics consist in a still more basic activity, the activity of volition, out of which all voluntary exertion and movement ultimately arises.6 Volition as I understand it is not a momentary, staccato act but an activity: that of willing the physical exertion appropriate to the bodily changes brought about in performing overt actions. It continues as long as the exertion in question needs to be sustained, and is the means by which we are able to alter our efforts in response to sensory feedback as to how the action is proceeding, so as to insure success. It is well suited to its role as basic action, in that it has the qualities of action inherently. To engage in volition is to feel spontaneously active; phenomenologically, at least, it is not the sort of thing that ever befalls a person. And volition is intrinsically intentional: we cannot will any exertion without meaning to will it, and to will it exactly when and as we do. The answer to Wittgenstein’s question, then, is this: what is left over when I subtract the fact that my arm goes up from the fact that I raise it is the voluntary exertion that causes the arm to rise. The exertional act is founded in turn upon volition, which is the causally basic activity that normally underlies all overt action. At bottom, then, it is by willing the muscular exertion suitable to raising his arm that the agent “brings about” the upward motion of arm when he raises it, and similarly for other overt movements. And finally, Wittgenstein’s question cannot be renewed for the activity of volition itself, for unlike actions of bodily movement volition does not consist in bringing about some nonactional event like the motion of an arm. Volition possesses intrinsically the characteristics of action, but it is only a means to bringing other events about. In itself, to will is not to bring about anything. It is to be noted that while this account sheds light on the nature of overt action and the phenomenon of agency, it is not a reductive account: it does not seek to define action or agency in other terms, and it is fully 6

A more elaborate argument for these conclusions can be found in my The Works of Agency (Ithaca, NY: Cornell University Press, 1998), chs.3-5.

56

Conversations on Human Action and Practical Rationality

consistent with the idea that in full-blown cases of action, intention formation mediates between having reasons and acting. Finally, and perhaps most important, this account leaves open the question whether, in acting, the will is caused to do what it does, either by reasons or by intentions. That is as it should be. The question whether the will is subject to causation is, after all, a debated one, and we presently have no definitive answer to it. But none of us has difficulty discerning when the motions of our bodies and their parts are voluntarily brought about and when they are not, nor do we have any difficulty discerning when we have acted for a reason and, except in pathological cases, what the reason is. It is plausible to think, therefore, that whether the will is subject to causation or not, our knowledge of what reasons guide our actions is not gotten by tracing causal relationships. This leads us to a third great problem in the philosophy of action: the free will problem. The basic positions in this dispute are well known. As was noted above, spontaneity seems essential to operations of the will. When we act we feel that we are beginning something new to the general order of things, that what we do and the consequences of it are in some important way up to us. And many are disposed to believe that if this were not so we would not be responsible for our actions, either morally or legally. Seemingly opposed to this is the idea that our actions are causally determined—remotely by heredity and environment but proximately by prior psychological states, over which we have at best only secondary control.7 The plausibility of this idea is not what it would have been, say, two hundred years ago; if present day science is on the right track at all, the world is not, as far as microphenomena are concerned, a determined place. But although no one has shown that mental acts of will are determined, neither has anyone shown that indeterminism at the subatomic level extends to phenomena in the brain that might correspond to recognizable mental states and events.8 Thus, in the absence of decisive empirical evidence, philosophers are left to argue the issue on phenomenological and conceptual grounds. Some are incompatibilists: 7

By secondary control I mean that the psychological states that may be held to determine our acts of will are subject to our influence only via earlier acts of will—which, if determinism is true, means that we are able to influence our own state of mind only in a way that is ultimately subordinated to complete determination by states we have not controlled at all. 8 But for interesting discussion along these lines see Robert Kane, The Significance of Free Will (New York: Oxford University Press, 1996), ch. 8; and David Hodgson, “Quantum Physics, Consciousness, and Free Will,” in The Oxford Handbook of Free Will, ed. Kane (New York: Oxford University Press, 2002), ch. 4.

Hugh J. McCann

57

they hold that legitimate free will and the thesis that our acts of will are causally determined cannot be reconciled. If they are prepared to trust the phenomenal character of experiences like deciding and willing they will be libertarians. That is, they will hold that the will is free in a contracausal sense, and they will likely emphasize the point that if things were otherwise then our acts of deciding and willing could in principle be traced to causes that existed long before we were born, in which case we must surely not be responsible for anything we decide or do.9 The other stance available to incompatibilists is hard determinism. Hard determinists agree that free will cannot be reconciled with determinism; but they are disposed to believe that acts of decision and volition are after all determined—and are therefore disposed also to conclude that, difficult though it may be to accept, we are not responsible for anything we ever do. Perhaps the most popular argument against libertarianism holds that, far from what the libertarian alleges about determinism ruling out responsibility, it is in fact the absence of determinism that would exclude our being responsible. For, it is claimed, it makes sense to hold an agent responsible only if his actions arise from his character. Yet if decision and volition are not caused by prior mental states then they do not arise from our character. Indeed they seem to come from nowhere at all—that is, they appear as random events, events that befall the agent rather than counting as anything he may be said to bring about or control. And surely agents are not to be held responsible for decisions or for volitional endeavors that befall them at random. This argument—we may call it the randomness argument—may be made either by hard determinists or by determinists of other varieties. I shall have more to say about it below, in my response to question 16. The most common alternative to hard determinism is compatibilism or so-called soft determinism, according to which even if our decisions and actions are caused we can still be free in performing them, or if not free then at least responsible. In its classical form compatibilism sought to reconcile determinism with freedom, holding that freedom was only a conditional or hypothetical matter. For example, it might be held that even though our seminar participant’s act of raising his arm was caused, it still counts as free provided only that had he willed differently some other act would have resulted. Similarly, it might be held that a decision—say, Hinckley’s decision to assassinate Reagan— is free provided just that it arose from Hinckley’s strongest motive, so that 9

For detailed development of this sort of argument see Peter van Inwagen, An Essay on Free Will (New York: Oxford University Press, 1983), ch. III; and van Inwagen, “Free Will Remains a Mystery”, The Oxford Handbook of Free Will, ch. 7.

58

Conversations on Human Action and Practical Rationality

had his strongest motive been to do something other than assassinate Reagan, the opposite decision would have ensued. Classical compatibilism is also a position that faces major difficulties. One is that it seems to misrepresent the distinction between free and unfree decisions almost entirely. Cases in which we think an agent’s freedom to decide differently than he did was impaired are nearly always cases in which we see one motive as too influential, as so strong that the value of other actions available to the agent became, as it were, submerged. The person who under mortal threat hands his wallet to a robber, the alcoholic who seems unable to resist a drink, the abuser of drugs who seems hopelessly addicted, are all persons who, we may presume, would have made opposite decisions and acted in opposite ways had their predominant motives been otherwise. Yet it is precisely because their behavior is so strongly motivated in the direction it does take that we view their freedom as at least limited. It does not appear that a classical conditional analysis can easily rectify this problem. Accordingly, recent compatibilists have tended to offer alternative solutions: to hold that whether an agent is responsible for his behavior depends on whether the motive by which he was guided truly represents his character, on whether he himself would have preferred to be guided by it, or perhaps simply on whether, in the circumstances in which the agent found himself, it would be reasonable for us to expect of him that his decision and action be guided differently.10 Committed libertarians and hard determinists are unlikely, however, to be persuaded by any of these moves. For, they will argue, no matter what condition the compatibilist might propose as crucial to whether a decision or action was free, the agent will not in fact be free unless he was able to control that condition. Let it be true that had an agent willed differently he would have acted differently, or that had his strongest motive been otherwise he would have decided otherwise. That does not matter, the incompatibilist will argue, if the agent was not able to control his willing and his deciding, which if determinism is true he ultimately was not. Similarly, whether the agent approved of his own motives does not matter if neither the motive, nor the approval, nor their influence on his behavior was ultimately within the agent’s power to alter—which, again, if determinism is true would seem not to have been the case. And as for whether it would be reasonable for us to have expected opposite behavior given the agent’s circumstances, the argument would simply be that if determinism is true, no such expectation could be reasonable in any circumstance. 10

A nice summary of alternative forms of compatibilism can be found in Ishtiyaque Haji, “Compatibilist Views of Freedom and Responsibility”, The Oxford Handbook of Free Will, ch. 9.

Hugh J. McCann

59

The upshot of the problems afflicting the various stances on the free will problem is a kind of impasse. The libertarian, if he is to make good on his claims about freedom and responsibility, faces a twofold task. He needs to show that reasons or motives can have explanatory force without causally determining the formation and execution of intention; and he has to offer plausible reasons for thinking decision and volition do not count as accidents for which the agent cannot be responsible. Determinists have more than one alternative available: they may attempt to rescue compatibilism, by finding a persuasive way of making the distinction between free and unfree decision and action in a determinist setting; or they may move to a hard determinist position and try to develop an account of responsibility that preserves as much as possible of our usual practices of praise and blame, reward and punishment. Needless to say, none of these agendas is likely to prove easy to execute. 2. For some or all of the following problems—action, agency and agent—what do they contrast with most significantly? The response I would give to this question is implicit in my discussion of question 1. Action is to be contrasted with events that are not action. In the overt realm, the relevant contrast is between the action which is the bringing about of some change and the change that is brought about: thus, between my raising my arm and the arm’s upward motion. As I see things, overt actions are made actional by the mental activity that is basic to them—namely, in the great majority of cases, by volition, which is intrinsically actional. There is, however, another contrast that pertains to the mental realm, between those occurrences in which we are active and those in which we are not. Decision and volition are actional: so is the business of directing our attention to or concentrating upon some thought content that is already before the mind. By contrast, things like the onset of desire or the occurrence of a memory are not actional. The difference is that the latter lack the spontaneity of doing—what Carl Ginet has called the “actish phenomenal quality”—and the intentionality or purposiveness that are intrinsic to mental doing.11 These, as I see it, are the positive features of agency, so that the contrast between events that display agency and those that do not lies in the presence or absence of these features. Finally, an agent is simply an individual capable of engaging in activities in which the essential features of agency are displayed. To the extent that an entity is incapable of such activities, it is not an agent. 11

Carl Ginet, On Action (New York: Cambridge University Press, 1990), p. 13.

60

Conversations on Human Action and Practical Rationality

3. Is it the fact that you see the “spontaneity of doing” as a central feature of agency that makes you think there must be something besides beliefs and desires (and even besides the intention created from these) for an action to take place? Do you think there is something missing from the definition of action if we refer to intention only? But if intention is already commitment to a plan, doesn´t it already imply such spontaneity? What does volition really add to it all? If we refer to intention only, we refer to a state, not an event, whereas actions are events. By contrast, the simple motion of a finger or a limb, though certainly an event, is not an action in itself. The phenomenon of volition fills a gap here, since it serves as the standard means by which we execute prior intentions. An intention, being a state, cannot execute itself. Furthermore, evidence from cases of paralysis indicates that volition is a genuine phenomenon, something that really occurs. For this reason alone, an empirically sound philosophy of action needs to include it. Some authors speak of what they call “occurrent intention” or “intention in action.” These are terms of art that normally simply substitute for “volition”, since Ryle made many people afraid of that word. 4. But is it really possible that a volitive act should take place without being accompanied by beliefs and desires or without following them? It is not possible to commit oneself to an incipient plan unless the plan is already before the mind. In addition, the goals embodied in the plan need to be apprehended as valuable for a decision in its favor to make sense. The usual way in which these conditions are satisfied is for the plan to be presented to the mind via beliefs and desires, or via beliefs and felt obligations. I suppose it is in principle possible for them to be satisfied in some other way, but I have no suggestions to offer at present. 5. It is admittedly possible for me to have the appropriate belief and desire to go through with an action, and still the action does not take place (as in Davidson’s example of the mountain climber). But don’t you think it is also possible that a volitive act exists and still we end up going against it? This is certainly possible, as is illustrated by cases of paralysis, where volition occurs but often has no bodily consequences beyond the brain. The difference is that volition or willing still counts as action. And we are responsible for it—responsible, for example, for obeying the neurologist’s

Hugh J. McCann

61

instruction to try to move. As stated, I hold volition to be actional. That is, it is an activity with all the crucial features of intentional action. When it has the intended physical consequences volition is the initial part of the bodily action in question. (This is unless we take actions to be tropes. If we do that, I suppose volition stands alone, in a way, since the movement performed by means of it would have to count as a different trope. But the motion of the arm, were it to occur, would still not count as an action—most especially according to trope theory.) 6. So desire is in a certain way something passive in us whereas the volitive act is not? Exactly. 7. But couldn’t we be deceived or betrayed by a volitive act too? For example, let us imagine that by means of a device set up in the brain, someone is led to carry out a given action – it is as if it they were controlling that movement, as if it was a pure volitive act. Do you think that in that case there would be a volitive act (since the criterion is somewhat “phenomenological”) and yet not an action? This is essentially an empirical question—first and foremost a question whether cases of the type described are empirically possible. There have, I think, been cases of motions of digits and limbs being brought about by simulation of the motor cortex, but I understand that the subjects of such experiments deny that they bring about the motion in question. What else may be possible I do not know. We shall have to wait and see. 8. Is the volitive act a pure mental act? I am uncertain what is meant by this question. The volitive act is certainly mental, and it is surely related to goings on in the brain. We do not know how. I do not take the volitive act to include such things as the tensing of muscles or the motions of parts of the body. 9. Which of the following are liable to be rational/irrational: action, agency, agent? The concept of rationality has far broader application than just to the realm of action. Most saliently, it pertains also to the cognitive realm: we may be rational or not in our beliefs, or reasoning, our investigative

62

Conversations on Human Action and Practical Rationality

procedures, and so forth. And even if we confine our intention to the realm of action—what we might call the realm of practical rationality— there are numerous issues that might be considered. There is, for example, a dispute over whether it is rational to have and to act upon intentions that one believes one cannot fulfill.12 But the most important dimensions of practical rationality have to do with exercises of agency—that is, primarily, with our decisions and willings. Whether an agent is rational can of course depend on other things as well: on his habits of deliberation, for example, or his tenacity in pursuing a course of action. In the end, however, practical rationality is a matter of what and how we decide and will. And here I think there are three main things deserving of consideration. The first is whether deciding and willing are themselves ratiocinative —that is, whether they occur not as a more or less proximate, and perhaps causal, consequences of deliberation, but as an actual part of it, as the outcome of a process of argument. Aristotle seems to have thought they did13, but it is not obvious that many contemporary philosophers would say the same. Second, if decision and volition are themselves ratiocinative in the way they arise is it possible, evaluatively speaking, to make rational sense of the process? Most especially, is it possible to think of reasoning that issues in a decision or act of will as in any way logically valid, or is there no such thing as thinking that is at once intrinsically practical and recognizably logical? This turns out to be a profoundly important matter, as we will see in considering question 16 below. Finally, there is the question of the premises. One supposes that truly respectable willing ought to be based on sound foundations—that is, on values that will somehow pass muster when it comes to reflective appraisal. And someone might believe that this too is a question of practical rationality. Which brings us to... 10. In what sense is the thing to do to be decided by what is rational? Are there limits to rationality? Some philosophers appear to have felt that practical rationality consists entirely in finding appropriate means to ends that are fixed by something other than reason. Aristotle, for example, believed that all of us naturally 12 For an argument that such intentions are irrational see Bratman, Intention, Plans, and Practical Reason, chs. 3 and 8; also Alfred R. Mele, “Intention, Belief, and Intentional Action”, American Philosophical Quarterly 26 (1989): 19-30. I have defended an opposing view; see for example The Works of Agency, ch. 10. 13 See especially De Motu Animalium, 701a7-36.

Hugh J. McCann

63

seek happiness14, and he treated the study of ethics as a matter of discerning how a person of practical wisdom might contrive to achieve that end. Similarly, Hume held that our ends in acting were set by our passions, and that although reason might at times disclose facts that would arouse passion within us, its proper role in the practical sphere was to discover causal connections that would enable us to act so as to achieve the ends passion sets.15 Suppose then that I have a fixed objective: I wish to drive to the nearest large airport, 100 miles away in Houston, Texas. In order to achieve my objective I must decide upon the route I shall take to get to the airport. Now if practical rationality is simply a question of finding and choosing means suited to fixed ends, then any route that will get me to the airport would appear to fit the requirements of rationality. It turns out, however, there are at least three separate routes I could take from my home to the airport. So far, therefore, rationality does not dictate any single decision for me. We can, if we wish, try to rectify this situation. It may be, for example that one route is quicker than the others, or safer, or more scenic, or more fuel efficient. Notice, however, that none of these things will matter unless I change my objective to include these additional ends. Now of course I may alter my objective in this way, but does practical rationality require that I do so? Someone might think it does; it could even be supposed that I will be deciding rationally only if I choose the best route to the airport, the route that will maximize the total benefit to be achieved by the trip. This seems to me, however, to be entirely wrong. For one thing, it is not obvious that all the goods just mentioned are even commensurable with each other. How does one measure the good of greater safety against the good of better scenery? Still less does there appear to be any one standard by which all these goods could be compared, and that is not even to mention whether the comparison could be done efficiently, within a reasonable period of time. The idea that my decision should maximize the good has, in short, no meaningful application to the case. Still less does there seem to be any a priori reason for thinking considerations, say, of safety should enter into my deliberation, whereas those fuel efficiency or speed should not. So there seems to be no way of selecting among the other objectives that might enter into my decision, except arbitrarily. Finally, and above all, there is an empirical point to be made: the simple fact is that we do not, in the great majority of cases where we deliberate over what to do, seek to maximize anything. I have driven to Houston 14

Aristotle, Nichomachean Ethics, 1097a28-1097b5. David Hume, A Treatise of Human Nature, ed. L.A. Selby-Bigge (Oxford: Oxford University Press, 1888), second ed. P. H. Nidditch (1978), pp. 459-60. 15

64

Conversations on Human Action and Practical Rationality

International Airport many times, sometimes choosing one route and sometimes another. I could not begin to say which is the best route to take, and far from feeling that this constitutes irrationality on my part, I am inclined to think the truly irrational thing would be to try to reach an answer to that question. And I believe the same applies to most agents most of the time. The truth is that in decision making we are not maximizers but sufficers: we seek means that are adequate to our ends, and then move forward. Only rarely do we concern ourselves with what is best, and even when we do it is a relative thing: what we have in mind is not some ideal standard of utility but an arbitrarily chosen set of our own ends. That is, we seek a means to achieve one of our objectives while achieving others as well, as when I decide that the quickest route to the airport is “best” today because I want to save time for other projects. It appears then that ultimately, our objectives are things we bring to the task of practical reasoning; they are not decided by it, except insofar as they serve as means to still further ends. I can think of two difficulties that might be raised against this stance. First, sometimes people have objectives that are so unrealistic that to intend them at all appears to represent a failure of the person’s mental faculties. Suppose, for example that someone intends to be the first person to swim the Atlantic. Here is a goal that cannot be achieved, so that unless some other aim of the agent’s could be achieved merely by pursuing it (he might wish, for example to get his name into the news) we might be tempted to pronounce him irrational even to have adopted this intention. I do not think, however, that this sort of irrationality represents a failure of practical reason. It concerns a broader conception of irrationality, a fundamental failure to appreciate what constitutes a worthwhile objective. A similar thing occurs in the cognitive realm. The mental patient who thinks he is Napoleon need not have reasoned wrongly; he may simply be deluded as to the facts of the case. We can be similarly deluded as to suitable ends; indeed, perhaps our would-be Atlantic swimmer belongs in the bed right next to our Napoleon. So I think there is a use of the term “irrational” that applies to cases like this. But I do not think the Atlantic swimmer’s “irrationality” represents a failure of practical reasoning properly so called. The second difficulty has to do with the provenance of moral principles. Certain philosophers, most notably Kant, have thought that reason is capable in itself of generating injunctions that would be morally binding on all rational agents. If this is correct then reason alone would be able, if not to dictate a single end for every case of deliberation, at least to circumscribe our ends, in the sense that there would be some goals it would be improper to pursue. And to the extent this was so, an agent

Hugh J. McCann

65

whose acts of will violated these restrictions could be held to be behaving irrationally. This is far too big an issue to go into here, but a few points deserve to be made. First, I am suspicious of this project. I am inclined to suspect that rationality alone can dictate my duty only if I have a prior duty to behave in accordance with rationality, and I do not see how reason can provide for such a duty. Second, I think it highly unlikely that reason alone, practical or otherwise, can produce a single solution to every deliberational problem. I see no plausibility whatever in the idea that, if only I understood the workings of reason well enough, I would see that I am after all duty-bound to choose just one of the available routes to Houston International Airport. But third, and on the other hand, it is not at all unreasonable to think that duty sometimes provides ends to which decision making is supposed to conform. Indeed, this would seem to be the whole point of the idea of duty. Suppose, for example, that I have promised to take my family on a trip to the Azores this summer. If promises are to be kept, then surely it is my duty to form and to carry out the intention of making the trip. So if the duty of keeping one’s promises can be shown to arise out of reason alone, then I would be acting contrary to the dictates of reason, and hence in this sense irrational, not to take my family on the promised vacation. As stated, however, the question whether reason alone possesses such resources must be left for another time. 11. You think that by reason alone we are not able to establish goals or the “best thing to do.” In that case, do you think it is impossible to regard as irrational something that I have reasons (belief +desire) to do? Can we as much as talk about normative reason in any domain of philosophy? We are certainly able to establish goals through the use of reason. For example, only a moment ago I established for myself the goal of answering this question before getting dinner. The problem of establishing the objectively best thing to do, once values are in place, is simply a matter of lack of information. Nevertheless I see no reason to suppose that the decisions we do make are, by and large, anything but rational, in the sense that they represent a reasonable application of the values and information available to us. But I sense that this question is really concerned with an entirely different matter—namely, whether reason can serve as the ultimate arbiter of value, in a sense that, once the task was complete, the vexing questions of moral epistemology would be settled. I suspect not, but not for any reason having to do with difficulties about practical decision making.

66

Conversations on Human Action and Practical Rationality

12. Does it make sense to talk about a “normative sense of practical rationality” or about a “descriptive model of practical rationality” (or equivalents concepts)? Does the distinction between normative and descriptive play an important role when we approach the issue of practical rationality? Like the last question, this one seems to be concerned more with questions of ethics that with the way practical reason actually operates. My concerns as an action theorist are with the latter, and in this regard I take my enterprise to be in essence a descriptive one. But “rational” is certainly an evaluative term, so no doubt I am dependent on a certain common sense ability, in both myself and my readers, to recognize a rational progression of reasoning when we see it. Whether the distinction between normative and descriptive should assume a more important role I do not know. If we wish for practical reason to provide a foundation for our values, no doubt it should. But that seems to me to have little direct bearing on problems of action theory. 13. George Ainslie told us, in his interview, that 'we can say that rational action is that which maximizes longest-term reward'. What is your opinion about such a perspective? Do you believe there is something such as “long term interest” which can be established through the analysis of our behavior, or this makes no sense in philosophy? We can of course say anything we like, and it may well be that to adopt the perspective of longest term maximization will make possible the development of a theory of what is rational that will be of normative interest and also predictive value. My concerns as an action theorist are, however, descriptive, not normative, and I am convinced that in our actual practical decision making, maximizing longest-term reward is not a conscious objective of ours. Now of course it may be that by approaching decision making as sufficers, we actually do maximize long term gain, regardless of whether it is in our interest to do so. That would be an interesting finding if it could be demonstrated. There is another dimension to the issue here: Whether it is in fact the case that to maximize longest-term reward is what an agent ought to be doing, in some sort of all-in evaluative sense, if he is to be deemed “rational.” Perhaps so, but in order to weigh in on this matter I would need to hear the argument.

Hugh J. McCann

67

14. What is your opinion about models such as the one described above, according to which the notion of will should be considered in a purely functional manner? Do you think we should rather talk about “preferences” in order to deal with such issues? Or should we consider will as something like a mechanism of stabilization of choices along the years? Do you believe that the dialogue between these two approaches could be fruitful? Or this is rendered impossible by the fact that in approaches such as these the will is determined (and this is something you reject)? “Hyperbolic discounting suggests a model of the self as a population of reward-seeking processes, which includes explicit hypotheses about these attributes. Shorter and longer range interests can be seen as rewardseeking processes that compete according to a rule of strict maximization of discounted expected reward at each moment of choice, and are motivated by game-theoretic considerations to form the very deliberative processes that we have been wondering about. (…). In a more precise term of art, will is what stabilizes plans in the face of temptations.” —George Ainslie, Conversations on Human Action and Practical Rationality.

The model of hyperbolic discounting is not familiar to me, and I have no idea what successes it may have had, so I am not able to address these questions very well. I certainly do not object to efforts to treat the will in functional or modular terms, as long as this is not taken to exclude treating decision and volition in other terms as well. How successful the model may be in shedding light on fundamental issues of decision and action is of course an empirical question, to which only time can provide the answer. As for whether there can be fruitful dialog between these approaches, of course there can, as long as both sides are willing to learn from the other. What is likely to be fruitless, rather, is dialog between and among parties who agree on every particular. 15. But don´t you think that in assessing rationality it is a good thing that factors such as cognitive conditions of agents, the environment in which they are, etc., are weighed? Yes.

68

Conversations on Human Action and Practical Rationality

16. What explains action and how? What is the role of deliberation in rationality? I think the best way to pursue this question is to return to the subject of practical reasoning. It is not unusual to see practical reasoning portrayed simply as a branch of theoretical reasoning—that is, as a matter of deducing propositional conclusions from propositional premises. Suppose, for example, that I desire to hear some music of Beethoven. Realizing this, and knowing there is to be an all-Beethoven concert tonight, I might reason as follows: I desire to hear some Beethoven. I will hear some Beethoven if I go to the all-Beethoven concert. Therefore, I should go to the all-Beethoven concert.16

It has to be admitted, I think, that inferences like this often occur as part of deliberation. They are, however, problematic. One difficulty that leaps to the eye is that the conclusion does not follow from the premises. We could of course begin to fix this if the only means available for me to hear some Beethoven were to go to the concert. As was noted in the answer to question 10, however, we are often presented with more than one means to achieve an objective, and this is a case in point. I can easily hear some Beethoven simply by playing a CD. Moreover, even if attending the concert were my only means of hearing some Beethoven the conclusion still would not follow. The two facts that I have the desire in question and that it can only be filled by going to the concert yield no conclusion whatever about what I “should” do. And there is a still deeper problem: Why should this argument be considered an instance of practical reasoning at all? Certainly its intrinsic features do little to justify such an appellation. Admittedly, the subject matter is practical. But the point of practical reasoning is not just to bring me to judgments about what I should do, but to settle, as far as reason can, what I am going to do. That is, it is to settle my objectives, which are a matter not of belief but of intention. If the closest I can get to a settled intention is an advisory conclusion drawn from premises that do not justify it, then practical reasoning is an impotent activity indeed, one whose study will offer us next to no insight into how intentions are formed. Indeed, the above syllogism need not even occur as a part of deliberation. It could as easily occur in contexts where my concern is not intention formation, but rather 16 Compare what Robert Audi calls the simplest basic schema for practical reasoning: Audi, Practical Reasoning (New York: Routledge, 1989), p. 99.

Hugh J. McCann

69

something theoretical: an exercise, say, in moral psychology or selfanalysis, undertaken for its own sake. If so, my reasoning would still be practical in its subject matter, but not in its point. Contrast this with a situation where, in cognizance of my desire to hear some Beethoven and knowing about tonight's concert, I actually make up my mind to go. Here, I reach an intrinsically practical conclusion: I decide in light of my reasons. This still presents itself as a ratiocinative procedure, but one which carries me much further in the direction of action than what is described above. A syllogistic representation of it might be as follows: Would that I hear some Beethoven. I will hear some Beethoven if I go to the all-Beethoven concert. Therefore, I shall go to the all-Beethoven concert.

The practical point of this reasoning is inescapable, because it is embedded essentially in the elements of which the syllogism is composed. The major premise here is not a proposition but a kind of optation: it encapsulates the content of a state of desire—what I actually think when the desire is experienced or rehearsed. And the conclusion represents another sort of optation: the content of my act of deciding to go to the concert, which is identical with the intention thereby formed. Though propositional in its representation, my intention is not just a proposition, because in one dimension it is not rendered “false” if, as things develop, I am unable to carry it out. It signifies a state of resolution or commitment on my part, which unless I rescind it is in force whether I will make it to the concert or not. Unlike the first of our two syllogisms, this second one is intrinsically practical: it portrays intention formation as a ratiocinative procedure. The execution of intention can be represented along the same lines. Recognizing that I intend to raise my arm, and knowing that I will do so if I will the appropriate sort of muscular exertion, I proceed to engage in the appropriate volition. Mutatis mutandis, this reasoning has exactly the structure of the decision-making argument given above, so that the entire progression from having a reason, to forming the intention it justifies, and then to acting upon the intention can be seen as one of reasoning. Moreover, the progression has a logic of its own, one that turns out to be surprisingly familiar. For consider the conclusion of the practical syllogism given above—that is, my intention to attend the Beethoven concert. It turns out that if I fulfill this intention, and given that the belief recorded in the minor premise of the syllogism is correct, then by simple modus

70

Conversations on Human Action and Practical Rationality

ponens, my desire to hear some Beethoven will be satisfied.17 Someone might object that this is an improper approach to the validity of an argument. And of course we do not, in judging the validity of theoretical reasoning, move from the conclusion to the premises. But with intrinsically practical reasoning it is precisely the way to go. This is because whereas theoretical reasoning aims at leading us to truth, which is a matter of bringing our mind into conformity with the world, practical reasoning has the role of assisting us in bringing the world into conformity with our mind—that is, with the way we want the world to be. Thus practical reasoning has what we might call practical validity when, as in the present case, its conclusions are appropriate to achieving the goals framed in its conative premises. A further point to be noticed is that this approach to the logic of our practical syllogism is completely undamaged by the fact that there are other ways available for me to hear some Beethoven. That is precisely as things should be if, as was argued in my response to question 10, practical rationality is concerned not with what is best but with what will suffice. But there is much more to be added. Consider the first of the two tasks that, at the end of the response to question 1 above, were said to face the libertarian: the task of accounting for the explanatory force of an agent’s reasons without invoking deterministic causation. In light of the practical syllogism given above, we can see how to carry out this task, for it depicts a ratiocinative procedure through which the agent himself settles the question of what reasons account for his intention, and eventually his action. From the practical point of view of the agent his decisions are made not because of reasons but in light of them: when I decide to attend the Beethoven concert my intention is formed in cognizance of my reasons. In it, I progress from merely having reasons to having an intention —that is, to having a commitment to the goal those reasons embody. The effect is that the content of my reasons—the content, in this case, of my states of desire and belief—is reformulated by me into the content of an intention. I come to have the intention to go to the Beethoven concert. I conceive this undertaking, moreover, as a means to hearing some Beethoven, and so come to intend to do this as well.18 Later, when the volitional stage of my progression toward action is reached something 17

For this approach to evaluating practical arguments see Anthony Kenny, Will, Freedom and Power (Oxford: Basil Blackwell, 1975), ch. 5, especially pp. 80-82. 18 In other situations I will have already decided to hear some Beethoven. When that is so, the intention to hear some Beethoven serves as the major premise of the practical syllogism, and is implicitly reaffirmed by my deciding to go to the concert.

Hugh J. McCann

71

similar occurs. The sorts of action that will get be to the concert are already pretty routinized for me, so I need not reach independent decisions as to what means I shall take to get to the concert. Instead, cognizant of my intention and of the means to carry it out, I can set my plan of action into motion by entering upon the volitional sequence it embodies. Here again the idea is that the content of the reasons is reformulated as the content of the practical conclusion, which in this case is willing. So if the first stage of my going to the concert is to walk to my car, we can think of my practical reasoning as going roughly like this: I shall attend the Beethoven concert. I will progress toward attending the Beethoven concert if I will the exertion appropriate to walking to my car Therefore, let the appropriate exertion occur.19

There is no mention whatever of nomic causality in this account, for several good reasons. The first is a practical one: the fact is that we simply do not know whether desire and belief, in the sense of mental states, are nomically related to decision and action. Each of us is, however, quite expert at being able to discern and report the reasons that explain our own decisions and actions. So whatever relation serves as a basis for this knowledge cannot be the alleged causal relationship. Second, what serve as reasons in practical syllogisms are not mental states but the content of those states—that is, abstract entities, which could not serve as nomic causes anyway.20 Third, and most important, the nature of decision and volition are such that no causal tie to reason states is necessary. It is not an extrinsic or accidental feature of deciding that it counts as a step toward doing what the agent decides upon, a step to be explained by the reasons that will eventually explain his action. Rather, it is intrinsic to my deciding to attend the Beethoven concert that it counts as commitment to so doing, and to achieving the ends encapsulated in the reasons in cognizance of which I decide. And because these things are intrinsic to my decision, no relationship of nomic causality is necessary for the reasons in question to explain my decision. All that is necessary is for me to make the decision and it will be made for those reasons, precisely because it is a commitment made in cognizance of them, and a 19 It is worth mentioning that since volition counts as basic action, this conclusion reflects the Aristotelian idea mentioned earlier, that the conclusion of the practical syllogism is action. 20 This is a point that has long been known. See Thomas Reid, Essays on the Active Powers of the Human Mind (Cambridge MA: MIT Press, 1969), p. 283.

72

Conversations on Human Action and Practical Rationality

commitment to pursue the ends they embody. The explanation those reasons provide will accordingly be teleological not causal, and we will be able to tell what reasons explain my decision simply by knowing the intention formed in it. Essentially the same thing applies to the volitional activity by which I will get to the concert. Volition too is intrinsically a commitment to pursue the ends that guide us in controlling what is willed. Complex though the exertional sequence that will take me to the concert may be, therefore, once I enter into willing it I can only be acting with the intention of getting to the concert. Thus my basic action will be explainable, again teleologically, in terms of my intention and the reasons that ground it. Notice, by the way, something that it is not being claimed here: it is not part of the view I am espousing that it is part of the content of my decision or of my willing that it is to be done for the reasons that explain it. Such claims are sometimes made about mental acts21, but they are unnecessary. The intrinsic nature of both decision and willing is enough to secure the result that when they occur, they will be undertaken for the reasons that explain them, the reasons that form the premises of the practical reasoning in which the agent engages when he decides or wills. This alone is sufficient to settle the matter as to what reasons explain what the agent does. Thus, to return to the example given earlier, it would be foolish to try to discern what reasons explain our seminar participant’s raising his arm by searching for a nomic relation between his desire states and his decision to raise it, or between them and his action of raising it. We don’t even have to ask what his desires and beliefs were. All we have to do is ask what intention or intentions he undertook to fulfill in deciding and willing as he did. If we know what he intended to do, we will know what reason or reasons explain his behavior. I mentioned a second problem that libertarians face: that of providing grounds for thinking decision and volition, even if uncaused, do not count as accidents. Does an account of the nature of decision and volition offer help with this issue? Well, imagine this situation: a professor enters his classroom, whereupon a student who was absent from the previous class walks up to his desk. Looking the professor straight in the eye, the student says: “I’m sorry I missed your class the other day Professor (fill in your name). I accidentally decided not to come.” Such an excuse is unlikely to gain the Professor’s indulgence, for one simple reason: it is selfcontradictory. Decision, as was stated earlier, is intrinsically active and 21

See for example Ginet, On Action, p. 143; also Timothy O’Connor, “Agent Causation”, in O’Connor, ed., Agents, Causes, & Events (New York: Oxford University Press, 1995), pp. 173-200, p. 192.

Hugh J. McCann

73

intentional, and the same goes for volition. From the subjective perspective of the agent, the phenomenology of active willing is such that when it is engaged in, we feel spontaneous, and we must mean to be doing precisely what we are doing, and precisely at the time we are doing it. From the practical perspective, then, it is conceptually incoherent to think of either volition or willing as accidental, or to think of the agent as in any way helpless in their occurrence. There must however be more to say here, for philosophers are not in the habit of uttering what can only be taken as conceptual nonsense. What, then, could opponents of libertarianism have mind in saying a decision that is not nomically caused has to be considered an accident? I think it is this: that an event that is not nomically caused is in a certain way discontinuous with all that precedes it. All of us have long range objectives and behavioral habits, so that someone who knows us well may often be able to foretell with high probability how we will decide and act. But to the extent that decision and action are uncaused, no such prediction is certain and no behavior is controllable. Thus a mere onlooker or scientific observer may look upon the decision or action as a kind of theoretical accident regardless of its practical status, and regardless of its teleological justification, and he may feel quite helpless to control the phenomenon, even if our experience as agents is one of total control. And there is this point to be made as well: many philosophers are disposed to think of natural causation as a matter of existence conferral, a process by which the world is able, as it were, to bootstrap itself from the present into the future—so that if we understand the nomic cause of a thing, we are able to account for its existence. Take that away, and the existence of things becomes a mystery, one that philosophy should try to avoid if at all possible. Thus, it might be argued, even if from the agent’s practical perspective the idea that his actions lack any nomic cause presents no problem, they still require a cause from the theoretical point of view, from the point of view of our understanding where they come from. Not only determinists but libertarians too have, I think, been impressed by considerations like these, and some have sought to fill the theoretical gap by postulating another sort of causation, namely agent causation, to account for the existence of our behavior.22 I think however that this is a mistake, for at least two reasons. First, the notion of agent causation seems to be vacuous, both theoretically and empirically. To my knowledge no 22

A full treatment of theories of agent causation is impossible here, but for extended defenses see O’Connor, Persons and Causes: The Metaphysics of Free Will (New York: Oxford University Press, 2000); and Randolph Clarke, Libertarian Accounts of Free Will (New York: Oxford University Press, 2003).

74

Conversations on Human Action and Practical Rationality

one has ever described any activity by which we are able to confer existence on our own doings, nor has anyone ever reported detecting such an activity. All we know is that our decisions and actions take place, that we engage in them. Second, suppose there were such an activity— something we do in order to confer existence on our doings. If so, it would have to occur either as a part of the doing—say, a decision—that it is supposed to cause, or as an independent event. If it were an independent event, then it would in turn become the focus of our concerns over freedom. We would have to worry about how to account for its existence, so that a vicious regress would threaten. Hence it will not do to treat agent causing as a kind of independent doing. But neither will it do to consider it a part of the doing it is supposed to explain. For consider: if my activity of agent-causing counts as part of my decision to attend the Beethoven concert, then until the decision is on hand the activity can cause nothing. Yet once the decision is present there is nothing for it to do, for there is nothing left to be caused. It can only be concluded, I think, the idea of agent causation is a false step. We are responsible for our decisions and actions because it is we who are active in their performance, and because in performing them we mean to be doing exactly what we do. But we do not bring our decisions and actions into existence. If this is correct, then the usual means by which libertarians might try to account for the existence of our decisions and actions fail. This, however, is not as damning a verdict as it may appear, for as far as I am able to tell determinists are no more able to account for the sheer existence of things than are indeterminists. This is because event causation, considered as an operation of existence conferral, is every bit as hopeless a notion as agent causation. It might be thought that this has to be wrong, that scientific laws, at least if they are deterministic, guarantee that the past will lend existence to the future in an orderly way. This, however, is completely misguided. Scientific laws, classically at least, are not even diachronic. Newton’s first law does not tell us that an object which is not acted upon by a net force at t will be at rest or in uniform motion a moment later; it tells us the object is at rest or in uniform motion at t. The same applies to the other two laws. Ideally considered, action is simultaneous with reaction, and acceleration occurs simultaneously with the application of force. To the extent, then, that it is erected upon these principles, familiar science does not and cannot speak of a world that is self-propagating.23 Furthermore, just as is the case with so-called agent 23 Whether contemporary physics is different in this respect I am not qualified to say, but I know of no persuasive argument or evidence to this effect.

Hugh J. McCann

75

causation, we know of no process by which the world is able to sustain itself, no operation by which one physical event is able somehow to confer existence upon another. And there is no reason to think the case will prove any different for mental events. In short, prospects that a deterministic account of the world, whether as pertains to natural events or to mental phenomena, would account for the mere existence of human decision and willing are decidedly bleak—as bleak, in my estimation, as the prospects for agent causation. Deterministic science does, however, enjoy one great advantage: where successful, it captures the dynamic properties of the world so well that, if we only presume the world will continue to exist, we can extrapolate with complete security from a description of the present to a description of the future. The effect is present each event as completely at home in the world; we need only understand what surrounds it spatiotemporally, and we will understand perfectly the event itself. In this more limited way, deterministic explanations do tell us where events come from, by portraying them as arising naturally and with full predictability from what went before. Teleological explanations of decision and willing cannot hope to match that, because they can never pin things down altogether as to what occurs. My reasons for attending the Beethoven concert are such as to make my decision to attend it sensible and justified. But no matter how much sense and justification they provide, they do not guarantee that I will so decide, or for that matter that I will reach any decision at all. And if my decision is not causally determined there are going to be other possible worlds that are precisely the same as our own in all that occurs up to the point when I decide, but in which another outcome ensues. Only a deterministic account of my decision can change that. The problem is, however, that a deterministic account would make my decision depend entirely on states of mind in whose provenance I am ultimately completely passive, and in whose immediate consequence—that is, the decision itself—I am in fact passive as well. My decision, that is to say, would issue from my reasons instead of being made in consideration of them. This is what gives rise to the free will problem, because it flies in the face of the phenomenal experience of deciding. It makes deciding an event in which I am being acted upon, whereas the experience of deciding is one of acting spontaneously. The question whether the phenomena of active willing are finally determined or not is largely an empirical one, although perhaps not entirely so. Kant, for example, would deny it. Before turning to the next question, however, it well to consider whether, as is sometimes said, it is in the end only a matter of luck that agents whose acts of will are

76

Conversations on Human Action and Practical Rationality

undetermined behave as they do. As mentioned above, this is not even possible from the agent’s practical perspective. The phenomenology of deciding is such that it is impossible for me to feel lucky in deciding to attend the Beethoven concert. Someone might argue, though, that I am lucky from the theoretical perspective, since I live in a world in which I do decide to go to the concert, rather than in one of the presumably infinite number of possible worlds that are the same up to the point of my decision but in which I decide differently. There is, however, a point to be made on the other side, one that is seldom noticed. True, if our world is a determined place then there are no possible worlds that are the same up to the point of my decision but differ as to the decision itself, or for that matter in any other way. There are, however, an infinite number of worlds in which I decide differently than I do in this one—namely, all of those in which, even if they extend infinitely into the past, the conditions that determine my decision differ throughout the history of time. And of course I do not control which of these worlds exists. But then it must be that from the theoretical perspective, I am in the end as lucky to decide as I do in a determined world as I am in an undetermined one. The only real difference is that if the present world is deterministic I should feel lucky from the practical perspective as well, because then the phenomenology of willing is a deception. My decisions and volition, although they are my doing, are not active in the way the experience of engaging in them indicates, but are instead something that befalls me. In the long run, then, determinism puts me far more at the mercy of luck than does libertarianism. 17. So, you claim that explanation of action does not have to be causal explanation. There is no cause-effect relation, since you assume that if there were, then my decision would be determined by mental states over which I do not have any control, i.e. the decision would not really be mine. Yet, on the one hand, causation does not have to be deterministic and, on the other, there may be an explanation for the beliefs and desires that I have, i.e. they do not come up out of nothing, they are related to my life, depend on my character, etc. This makes them mine and also makes me have some control over them (I can reflect on my beliefs and desires and eventually even change them). Or do you think differently? I don’t think I have ever claimed that action does not have a causal explanation. That seems to me to be an empirical question that we are not presently in a position to answer. I have tried to articulate a non-causal view of the etiology of decision and action, because I do think that if these

Hugh J. McCann

77

things were caused there would be an important sense in which they would not be ours—namely, that our sense of being active in deciding and willing rather than passive as we are when we experience desire would be deceptive, and our decisions and volition would not count as the initiation of new causal streams in the history of the universe. This does not of course prevent decision and volition from being mine in other ways. After all, any mental state counts as mine strictly in virtue of being predicated of me. And it is also true that many mental states and events can be indirectly “controlled,” if this means only that they may occur as consequences of other things that I do. But that does not mean they can be controlled in the sense that libertarians take decision and volition to be controlled. As for the idea that causation does not have to be determinist, that seems to me to be a subject unto itself, too complicated to go into here. I do think, however, that in examining this sort of claim we need to ask what we mean by “causation,” and whether what we mean has any empirical content. 18. Also, if there were no causation or determination, wouldn’t decision seem like an irruption, something that happened by chance, by luck? There is a distinction to be made here between the point of view of an observer who wishes to understand and control an agent’s behavior, and that of the agent himself as agent. The first is a theoretical perspective, the second a practical one. From the theoretical perspective, any undetermined event counts as irruption, in the sense that prior to the event, there are in any given case no steps that can be taken by any outsider which will guarantee that one only one outcome will occur. In the case of decision and volition, however, it does not follow that the event cannot be understood. It may be understood teleologically, in terms of the purposes for which the agent acted. Furthermore, it is impossible for us to make our way through life without having long term goals, and long term goals tend to engender regular patterns of behavior aimed at their achievement. There is no need to treat this as a matter of causation or determination. It is simply an issue of practical rationality. It does, however, mean that statistical generalizations will be found to apply to the decisions and actions of rational agents, caused or not. And in this sense, those events need be no more incomprehensible than phenomena that are said to be “statistically caused”. There is, of course, an irreducible residue of unpredictability

78

Conversations on Human Action and Practical Rationality

about these doings. And that should trouble us theoretically, but no more than unpredictability in any other sphere. From the practical perspective of the agent, on the other hand, it is conceptually impossible for decision or willing to count as “irruptions,” because it is conceptually impossible for these phenomena to occur accidentally or inadvertently. To see this, simply imagine the example I mentioned before: how you would react if a student came up to you and said, “I’m sorry to have missed your lecture yesterday Professor (fill in your name). I accidentally decided not to come”. 19. You say that we decide by considering our reasons (beliefs and desires), but not because of them. What does it mean to say that someone acted due to a certain reason and not another, other than that reason has a causal role in determining her action? I may decide after considering my reasons, but I never decide by considering my reasons, if by this is meant a step that brings about my deciding. I decide simply by deciding. As for what it means to decide for one reason rather than another, one thing it does not mean is that the one reason but not the other caused my decision. If it did, then I would have to find out the reason why I decided by examining causal relations in my mind. And how would I do that? By searching for relationships of necessitation? On this matter Hume was right: there are none to be observed. By looking for generalizations that are known to cover the case? But no useful universal generalizations are known, and statistical ones are useless for this purpose. By examining what would occur in nearby possible worlds? Ridiculous. There simply is no concept of causation that is of any practical value for deciding this kind of question. How then do I determine the reason for which I decided? Very simple: I consult the intention formed by my decision. For example, if I decided to go shopping for the reason that I wanted some cheese and not for the reason that I wanted some wine then I will, in deciding, have formed the intention not just to go shopping but also to shop for cheese. And there is no reason to think the intention arises through determination from my reason. Rather, to decide is simply to consciously recast my reasons in a new modality of thought, that of intending. In effect, my reasons present to me an incipient plan of action. To decide is simply to process in a certain way the information my reasons convey to me, to apply to the content of my reason states a new modality of thought. Notice that not a

Hugh J. McCann

79

single word about causation is needed to describe any of this—and it had better not be, first because the notion of causation is of no practical value for settling problems like this; and second because what counts as a “reason” in deliberation is not a state or states of desire and belief but rather the abstract content of those states, which as Reid long ago observed is not even capable of functioning as a cause. 20. Do you believe that non-conscious or non-cognitive factors can be important in explaining an action? It is certainly possible for decisions and actions to be motivated by conative states that we wish to conceal from ourselves, and so would deny having. But this does not mean that, in influencing our decisions, such motives do not enter into consciousness in any way. Here we need to remember that it is one thing to be thinking something, and quite another to be self-aware in the matter—that is, to think that we are thinking the thing in question. I am inclined to believe that what are often called “unconscious” desires and beliefs are mental contents that we think, but do not think that we think. There are in addition such things as habits, routines, and long standing intentions upon which we act. When we do, our actions are indirectly influenced by motives of which we were conscious in the past, but which need not manifest themselves in thought at the time of action. 21. Are there other reasons to act which are unconscious? What is the difference between acting consciously and acting by conscious reasons? Whether there are reasons other than conscious ones—that is, conscious in the broad sense given in my answer to the last question—I am inclined to doubt, but I would be open to suggestion. To act consciously is to engage in decision, volition, or selective direction of attention to one content or another. To act by conscious reasons is to do these things for reasons that occur in consciousness. 22. According to you, is intentionality a characteristic of conscious systems and their actions only, or can it be attributed to any system in whose behavior we find a purpose or goal? I take it that the notion of a purpose or goal would apply to any system whose behavior could be explained teleologically, but I don’t think I

80

Conversations on Human Action and Practical Rationality

would apply the notion of intentionality that broadly. To me, the sense of “intentionality” that applies to human decision and action implies consciously undertaken commitment, so I would want to confine that idea to conscious systems. 23. Do you think that someone who shelves his books while sleepwalking acts intentionally? Why would that be so? Sleep does not exclude any and every form of consciousness, as the phenomenon of dreaming demonstrates, so I don’t see any reason why sleepwalkers might not be in some way conscious, and so capable of intentional action. But my real answer to this question has to be that, not being a sleepwalker, I do not know. 24. So, given the way you assess the process of intention formation that leads to action, how do you think akrasia is possible (if you think it is)? I think it is, and that in light of the foregoing discussion we can see exactly how akrasia occurs. The problem is that we can reach a judgment as to what is best, or more likely about what is obligatory (which need not be the same thing) yet lack the motivational resources to form, or alternatively to execute, the appropriate intention. The latter sort of case occurs when we bravely make up our mind to do what is demanded, thus forming the relevant intention, but end up so preoccupied with other motives that the value of the original course of action is lost to our attention, and only the initially rejected alternative retains the support of truly practical reason. That, I suppose, is what happens in cases of alcoholism and addiction, among others. Every alcoholic knows that it is the easiest thing in the world to resist temptation . . . once. The problem is to resist it again, and again, and again. For with time the perceived value of drinking so fills the mind that the motives to abstain either fade to insignificance, or are lost completely in the distraction of the temptation, so that only motives that support drinking retain effective force, making it, from the practical point of view, the truly reasonable alternative. Thus the original attention is abandoned, and the hand reaches for the glass. The case is similar for intention formation. Suppose I reach the moral judgment that it is my duty to help the homeless, and that I believe I will do so if I contribute to a local homeless shelter. Based on the discussion under question 16 above, it will be seen that premises like these can lead to a rationally formed decision only if we can make logical sense out of this argument:

Hugh J. McCann

81

I have a duty to support the homeless. I will support the homeless if I contribute to the homeless center. Therefore, I shall contribute to the homeless center.

We have already seen, however, that there is no logical sense to be made of this sort of argument. In order for me to decide intelligently to contribute to the homeless center my major premise—that is, the first premise in the practical syllogism—has to be conative in its content. In this connection, there are two points to be considered. First, some philosophers have espoused a view sometimes called motivational internalism, according to which a judgment that it is one’s duty to perform a certain action counts in itself as motive to perform the act, or guarantees conceptually that such motivation will be present. Both versions of this position seem to me to be wishful thinking. On its very face my judgment that I ought to support the hungry is not conative and hence has no motive force; and I challenge anyone to infer something that does have motive force from it. What does seem to me to be true is that it would be highly unusual for a normal person to reach the conclusion that he was dutybound to perform some action and yet to be moved not at all in the direction of performing it. But that is a point about human psychology, not the logic of decision making, and to claim that a complete absence of such motivational force is psychologically impossible is, at least for me, to beg too many questions having to do with the pathology of wrongdoing.24 I think it will not do, then, to take the major premise of the above syllogism as either constituting or entailing anything conative. What is needed is a genuine motive, something that counts intrinsically as a denizen of the practical side of the mind. Here, however, there is a second point to be made. It is, I think, wrong to suppose that a suitable motive for doing one’s duty is that one desire to do it, that one finds the idea attractive. It is offensive to think that the force of duty should depend on what amounts to no more than mere subjective inclination. Indeed, the whole point of duty is to override mere desire—to make sure, as I am fond of telling my students, that you do not get what you want. What is needed, then, is an intrinsically deontic conation, a subjective sense of what Kant would call the practical necessity of an action. In the present case, we might express it as I must support the homeless—where this is understood as another sort of optation, a conative content that captures the felt 24

For critiques of motivational internalism see Robert Audi, “Moral Judgment and Reasons for Action”, in his Moral Knowledge and Ethical Character (New York: Oxford University Press, 1997) ch. 10; and Sigrún Svavarsdóttir, “Moral Cognition and Motivation”, The Philosophical Review 108 (1999): 161-219.

82

Conversations on Human Action and Practical Rationality

urgency of duty. Making this adjustment, our practical syllogism would be: I must support the homeless. I will support the homeless if I contribute to the homeless shelter. Therefore, I shall contribute to the homeless shelter.

Here we have a piece of genuinely practical reasoning, whose logic is the same as that of our earlier practical syllogisms pertaining to decision and volition. Suppose, however, that the crucial premise is either unavailable, or so weak in its influence that it would be all but completely unreasonable for me to make it a basis for intention formation. There are all sorts of ways for me to believe it is my duty to help the homeless, yet experience little or no felt obligation to do so. Perhaps I was taught as a child that the plight of the poor is their own fault, and that to help them would only be only to encourage them in wrongful ways; perhaps I think that if you help the homeless you will eventually wind up with an undesirable element in your neighborhood, which I would abhor; or perhaps I was once homeless myself, and want nothing whatever to do with the entire subject. But whatever the reason may be, it is easily possible for me to recognize my duty in this case, and yet be either unmotivated by it, or so weakly motivated that a decision to favor duty is, though possible, very unlikely. This, I think, is the classic case of akratic decision making, and I think it is eminently understandable on the principles I have defended. It is worth noting, finally, that nothing in this treatment of akrasia calls for a deterministic account of decision or willing. Most especially, I do not adhere to the view that in akratic action, or in other cases where we are likely either to excuse an agent or to mitigate blame and punishment, determinism somehow takes over, so that an action that might otherwise have been free is now “caused”. Such a view of the will seems to me to be highly unreasonable—rather like saying, in the scientific context, that the phenomenon of beta decay is not deterministic, and then looking for ways to bring it under control. If beta decay is indeterministic, it is intrinsically so; anything that might bring it under control would simply destroy its nature, so that it would cease to be beta decay. As I see it, the same applies to active willing: if decision and volition are free as pertains to nomic causation, they are intrinsically free. Accounts of willing that we are inclined to call unfree should be in terms not of causal determinism,

Hugh J. McCann

83

but of the ways in which intrinsically practical reasoning can misfire.25 25. You seem to see the problem of akrasia as mainly a moral problem – is it not possible that akrasia could be a type of irrationality without any moral connotations? I doubt that there can be cases of akrasia with no moral connotations of all, because weakness of will is in itself a failure of fortitude, and so counts against a person’s moral character. I do however think there are some cases of akrasia that do not involve failure to observe a moral duty. A mountaineer who resolves to make a dangerous climb but then backs out at the last minute due to fear exhibits akrasia, but he violates no moral duty. My earlier answer was hasty in not allowing for this sort of case. 26. According to you, the problem of akrasia appears both at the level of intention formation and in the carrying out of the action. But if the intention is not formed, it does not seem that we are faced with a typical case of akrasia, in the sense that the agent does not go against her own judgment. Or do you think otherwise? Can someone be akratic before there is an action? If one thinks it is wrong commit adultery but, out of moral weakness, makes up one’s mind to do so, then one has already set oneself upon a course that one judges to be wrong. This does count as going against one’s better judgment, and thus in my view deserves to be called akrasia. But this may be partly a verbal matter. Perhaps we should speak of two types of akrasia, akratic decision and akratic action. 27. If I think that I should kill someone, and I form the corresponding intention and then I do not act according to it, am I being akratic? If it is truly my best judgment that I should kill someone, I still may or may not form the intention to do so. But once I form the intention, then, unless my judgment changes, not acting to carry out the intention will be akratic provided my not acting is owing to something like, say, a failure of nerve. It is said that this sort of thing happens with some frequency on the 25

For further discussion of akrasia see my The Works of Agency, ch. 11; for an excellent summary discussion of akrasia by an author who has done extensive work on the subject see Alfred R. Mele, “Autonomy, Self-Control, and Weakness of Will”, The Oxford Handbook of Free Will, ch. 24.

84

Conversations on Human Action and Practical Rationality

battlefield; we tend to recoil instinctively from killing another human being, even a mortal enemy. There may of course be other explanations for my not acting, such as missed opportunity or the like, and then the failure will not be akratic. 28. How do you think your own work has contributed to the field? What do you think are your most important contributions? What are your plans for future research? I don’t imagine that I am the best person to judge my contributions to the field or their value. Such as they are, my contributions are largely summarized in the answers to these questions; many are developed more fully in the papers comprised in my The Works of Agency, to which I have already referred probably too often. I have in the past been especially active in discussions of volition and basic action; in treatments of the formation and execution of intention as an activity of practical reasoning; in debates about the relation between belief and rational intention; and to some degree in discussions of free will and the explanation of action. I have always been interested in the philosophy of religion, and have lately been active in discussions about the relation between human freedom and divine sovereignty, and of how the notion of free will might be useful to attempts at resolving the problem of evil. I am at work on a book manuscript that deals with these subjects among others, and I have another book project, still in its early stages, devoted solely to the subject of free will. If I am to tout my own work I should acknowledge its indebtedness to the advice and assistance of others. I have benefitted greatly from discussions with friends and students over the years. Their names are far too numerous to list here, but I should say that my work has gained a lot from the comments and criticisms—even if I often refused to be swayed by them—of Robert Audi, Myles Brand, Hector-Neri Castañeda and Alfred Mele, among a great many others. I am deeply grateful to them all.

MICHAEL BRATMAN STANFORD UNIVERSITY

1. In your view, what are the most central (or important) problems in the philosophy of action? We are both knowers and doers. In each case our minds are appropriately connected to the world. Or so it seems. These commonsense remarks are the background of three central areas of philosophy: epistemology, philosophy of mind, and philosophy of action. In all three cases the issues are conceptual, metaphysical and normative. We seek concepts adequate to an understanding of the phenomena; we seek to understand in what these phenomena consist; and we seek to understand relevant norms – both theoretical and practical – and their relation to matters conceptual and metaphysical. The problems throughout are deep and of great importance in our understanding of who we are; and these problems across these different areas of philosophy interact in complex ways. But I think it is clear that the western philosophical tradition has – with some notable exceptions – tended to focus more intensely on knowledge and mind than on action and agency. One powerful source of the attraction and excitement of the philosophy of action to me is that the target of its inquiries – our nature as agents – is gripping within selfreflection and of fundamental importance within a wide range of human concerns (some of which I mention below). A second source of the attraction to me of the philosophy of action is the sense that, as compared with our philosophical understanding of knowledge and mind, this is less well understood – though equally fundamental – territory, and so territory that is more susceptible to our seeing things in new and newly fecund ways. What is it to be an agent, and not just an object in the natural, causal order? What is it to act? to act intentionally? to act for a reason? What is it for the agent to be the source of the activity, for the activity not just to be the outcome of an event-causal process? What kind of knowledge do we have of our own agency? How is the idea of a normative reason – a reason that is relevant to the justification of action – connected to the concepts and metaphysics of agency and action? to the phenomenon of motivation of action? to the explanation of action? How is the bearing of rationality

86

Conversations on Human Action and Practical Rationality

on these practical domains – as we say, practical rationality – related to the bearing of rationality on the domains of knowledge and belief – as we say, theoretical rationality? We think that some agents govern their own lives; but what is that? How do these matters look when we turn from the actions and agency of individual human agents to the actions and agency of groups – to our singing the duet, or – to move from small to large – to “we the people”? These, anyway, are some (I do not say all) of the deep, difficult, and complexly inter-related questions that are central to the philosophical study of human agency. 2. For some or all of the following problems – action, agency and agent – what do they contrast with most significantly? And how was it exactly that thinking about agents and the structure of agency led you to your planning theory? An initial point is that there are many different kinds of agents, including spiders, cats, infant humans, chimpanzees, and adult humans. If we try to say what is common and peculiar to all cases of agency we will probably arrive at an idea that Harry Frankfurt, drawing on work Ernest Nagel, has sketched: agency involves behavior that is potentially under the control of a guidance mechanism – a guidance mechanism that tracks a given end – where that mechanism is one whose operation constitutes the activity of the agent and not merely of a sub-system.1 And the point for present purposes is that these ideas about agency are broadly generic. We should not think that all that we can say about human agency is provided by such a generic model of agency. In particular, reflection specifically on human agency leads us to the idea of choice, or decision, or intention – of, as we can say, will. This idea seems distinctively relevant to our understanding of human agency, and in a way that distinguishes acting from knowing. Choosing or deciding to believe in most cases seems problematic in ways that we begin to get at in our talk of wishful thinking. But choosing or deciding to act and so acting accordingly seems a central case of agency. But what is this idea of will, and what in the world (if anything) does it point to? One aspect of this idea that is a part of a certain amount of common talk is the thought that a choice or decision in some sense breaks the causal, deterministic chain from the past to the future. I myself find that this “incompatibilist” thought – at least, as a thought about what is 1

Harry Frankfurt, “The Problem of Action”, American Philosophical Quarterly 15 (1978): 157-62.

Michael Bratman

87

essential to choice or decision – looses its grip when we reflect on how breaking that chain could matter to the metaphysics of the forms of agency we care about. On reflection, what seems to me most important about the will – about choice, decision, and the like – is not whether it is outside the causal order (how would that help us live our lives?), but that these willlike phenomena settle practical questions in ways that impose distinctive structure on our downstream practical thinking and acting. Creatures with what we might as well call “a will” are creatures with the psychological structures and capacities that are involved in imposing this structure on thought and action. What structure, and why bother? Here I think both the concept and the phenomenon of planning, and associated norms, are fundamental. Intentions are elements of typically partial and almost-always at least partly future-directed plans. Plans and planning play fundamental coordinating roles in our temporally extended and social lives, especially given our cognitive limitations. That is why we bother with plans. Choice and decision are standard ways we have for arriving at such plan-like states of mind. To understand what the will is we need to understand what it does; and to understand what it does we need to understand how such planning functions in our temporally extended and social agency. Our understanding of that does not depend on seeing the will as outside the natural, causal order. But it does depend on understanding the distinctive roles of planning structures in our practical thinking and action. And that is a matter both of understanding how those structures function, what the relevant norms (norms, for example, of consistency of intentions and means-end coherence of plans) are that apply to and sometimes guide this functioning, and why it matters whether we (try to) conform to those norms. Or so I have over the years argued, in developing what I have called the planning theory of intention and our agency – a theory that can be thought of as a modest theory of the will.2

2

See Michael E. Bratman, Intention, Plans and Practical Reason (Cambridge: Harvard University Press, 1987; reissued by CSLI Publications, 1999); Michael E. Bratman, Faces of Intention: Selected Essays on Intention and Agency (Cambridge: Cambridge University Press, 1999); and Michael E. Bratman, Structures of Agency: Essays (New York: Oxford University Press, 2007). For a recent overview discussion see my “Agency, Time, and Sociality”, Proceedings and Addresses of the American Philosophical Association 84:2 (2010): 7-26.

88

Conversations on Human Action and Practical Rationality

3. What explains action and how? What is the role of deliberation in rationality? And where does your planning theory stand in the philosophy of action debate, considered from a historical point of view? In a way, the planning theory is a response to a debate that didn’t quite happen. Philosophy of action as we know it was formed on the crucible of the interactions – primarily in the 1960s – between Elizabeth Anscombe and Donald Davidson about intention and agency.3 They agreed about the individuation of action – about the idea that normally when I B by A-ing I perform one action, multiply describable. They disagreed about whether the explanation of action by appeal to the agent’s reasons for so acting was a kind of causal explanation.4 And they also disagreed about Anscombe’s claim that in acting intentionally one necessarily has a distinctive kind of non-observational and non-inferential knowledge of what one is doing. The disagreement about reasons and causal explanations shaped much of what we know today as the philosophy of action. Davidson’s enormously influential desire-belief causal model of intentional agency – a model that sees purposiveness, rather than a distinctive mode of self-knowledge, as at the heart of agency – is one of the main fruits of this debate. In contrast, the disagreement about self-knowledge in intentional agency stayed for awhile on the back burners, though – due to work of, among others, David Velleman5 – it has lately returned to center stage (together with implications for our understanding of the relation between practical and theoretical rationality). But there was also an underlying agreement – or anyway, a truce – about the very idea of intending to act. For different reasons, neither saw this idea as getting at something fundamental, though in later work Davidson did seek a conservative correction in this skepticism.6 But 3 Elizabeth Anscombe, Intention, second edition (Oxford: Basil Blackwell, 1963); Donald Davidson, Essays on Actions and Events (Oxford: Oxford University Press, 1980). 4 See also Elizabeth Anscombe, “The Causation of Action”, in C. Ginet and S. Shoemaker, eds. Knowledge and Mind (Oxford: Oxford University Press, 1983): 174-90. 5 For a recent discussion see his “What Good is a Will?” in A. Leist, ed., Action in Context (Berlin: Walter de Gruyter, 2007): 193-215. 6 For a slightly different take on this see Kieran Setiya, “Intention”, Stanford Encyclopedia of Philosophy http://plato.stanford.edu/entries/intention/. Setiya calls this “Davidson’s conversion” and has a somewhat different understanding of its significance within Davidson’s overall philosophy of action. For my understanding of this feature of Davidson’s work, see my “Davidson’s Theory of Intention” as reprinted in Faces of Intention.

Michael Bratman

89

in a different, concurrent tradition of research on agency – associated, for example, with work of Hector Neri-Castañeda7 – the concept and phenomenon of intending to act were seen as fundamental. In particular, intending was seen not just as an output of practical reasoning (which is how Davidson eventually came to think about intending) but as a fundamental element in the inputs to practical reasoning. (A central example is reasoning from intending an end to intending means, though this is not an example highlighted by Castañeda.) The question of how to draw on the insights both of the Davidsonian desire-belief model and of the focus on intending as a basic phenomenon tended to be ignored, though it was brought to the fore in Gilbert Harman’s groundbreaking 1976 essay, “Practical Reasoning”.8 The planning theory was my response to this incompletely articulated debate. 4. How would you characterize the distinctive way your theory addresses the fundamental questions in the philosophy of action? The planning theory has implications for many of the issues noted above, issues that are at the heart of the philosophy of action. One example is that it supports a view of intending as importantly different from believing. Intending p will play different roles in downstream planning, means-end reasoning, and action than will believing p (though here I disagree, in different ways, with both Harman and Velleman). As I see it, this also helps us avoid collapsing the norms on intending that are central to our planning agency to theoretical norms, and thereby helps support an important distinction between theoretical and practical rationality.9 This is tied to recent debates about the nature and ground of what John Broome, in an important series of papers, calls rational requirements – for example, rational requirements of consistency of belief, and of intention.10 And one outstanding question here is whether the planning theory can put us in a

7 Hector-Neri Castañeda, Thinking and Doing (Dordrecht, Holland: Reidel, 1975). And see my “Castañeda’s Theory of Thought and Action,” as reprinted in my Faces of Intention. 8 Gilbert Harman, “Practical Reasoning”, Review of Metaphysics 29 (1976): 43163. 9 See my “Intention, Belief, Practical, Theoretical,” in Simon Robertson, ed., Spheres of Reason: New Essays on the Philosophy of Normativity (Oxford University Press, 2009): 29-61. 10 See e.g., John Broome, “Reasons,” in R. Jay Wallace et. al., eds., Reason and Value (Oxford: Oxford University Press, 2004): 28-55.

90

Conversations on Human Action and Practical Rationality

better position to assess and respond to recent challenges from Joseph Raz and Niko Kolodny to the normative force of such rational requirements.11 I also think that the planning theory can make significant contributions to our understanding of basic human capacities for self-governance and sociality – capacities that are central to our humanity. I think, first, that when we highlight these planning structures as deeply embedded in the kind of agent we are, we can make progress in understanding what it is for an agent to govern her own life. This is in part because the possibility of appealing to planning structures helps us answer a deep question that is in the background of our thinking about self-governance: what psychological structures are such that when they guide thought and action the agent directs and governs? (I call this the question of the agential authority of those psychological structures.12) And this is in part because of the central roles of planning structures in constituting and supporting major forms of cross-temporal organization of our practical thought and action. Appeal to these planning structures gives us a way of understanding the role of the will in self-governance without seeing the will as outside the natural, causal order. The planning theory also helps us better understand the kind of shared agency illustrated by cases like that of our singing the duet, or Margaret Gilbert’s example of our walking together – cases of what I call modest sociality.13 A central problem here is how to understand the distinctive way you and I are related when we walk together, and how this differs from the way in which you and a stranger are related to each other when you are both merely walking down the same street at the same time without bumping into each other. Broadly speaking, the game-theoretic tradition aims to understand our special relation to each other when we walk together as a matter of a kind of equilibrium in a context of common knowledge. But this seems ill-suited to make the contrast between you 11 Joseph Raz, “The Myth of Instrumental Rationality”, Journal of Ethics and Social Philosophy 1 (2005); Niko Kolodny, “The Myth of Practical Consistency”, European Journal of Philosophy 16 (2008): 366-402. For my response to these concerns see my “Intention, Practical Rationality, and Self-Governance”, Ethics 119 (2009): 411-443. 12 See Structures of Agency. 13 See the quartet of essays on shared agency in my Faces of Intention, my “Modest Sociality and the Distinctiveness of Intention”, Philosophical Studies 144 (2009): 149-65; Margaret Gilbert, “Shared Intention and Personal Intentions”, Philosophical Studies 144 (2009): 167-187, and John Searle, “Collective Intentions and Actions” in Cohen, Morgan, and Pollack, eds., Intentions in Communication (Cambridge: MIT Press, 1990), pp. 401-415.

Michael Bratman

91

and me when we walk together, on the one hand, and, on the other hand, you and the stranger merely walking down the same street at the same time without bumping into each other. An alternative view, associated with Gilbert’s work, is to see what is special as including certain mutual, reciprocal obligations between us and entitlements of each to hold the other accountable for failing to play her role. In contrast, the planning theory gives us the conceptual, metaphysical, and normative resources to chart a path in the territory between these two approaches: our modest sociality is not just a matter of equilibrium given common knowledge; but it does not essentially involve mutual obligations and entitlements – though such obligations and entitlements are enormously common. Indeed, there can be such characteristic obligations and entitlements in the absence of shared agency – just think about cases of mutual but insincere promises to join in a shared activity. Very roughly, what is essential to our shared agency are certain structures of interrelated intentions of the participants; and we understand such intentions within the planning theory. 5. Given what you just said, do you think your work can be a useful link between philosophy of action and other areas, like social sciences, collective action theory, ethics, etc.? The conjecture is that basic planning capacities support the temporal organization of our agency and are at the bottom of two other fundamental human practical capacities: our capacities for self-governance and for important forms of sociality. Whether these planning theoretic approaches to self-governance and to modest sociality will be successful is a matter for further research. But it does seem to me a potential advantage of the planning theory that it promises to help us characterize important features of human agency that may well turn out to be central to these further basic capacities for self-governance and modest sociality. An underlying idea is that a central target of our reflections in the philosophy of action is basic structures – basic capacities and modes of thought – characteristic of (what we can call) the will and thereby of our human agency. We get at these structures by studying how they function in particular cases of practical thinking and intentional action, and by studying the norms that apply to such functioning. But it is these general structures of agency – structures that help constitute and support our temporally extended and social lives – that are our primary target. And the conjecture is that these planning structures help us understand other basic human practical capacities – including, in particular, capacities for selfgovernance and sociality.

92

Conversations on Human Action and Practical Rationality

Questions about how to understand our agency – about the needed concepts, the underlying metaphysics, and the associated norms – are at the foundation of our understanding of who we are both as individuals and as participants in sociality. Our answers to these questions will impact our approach and answers to a wide range of issues. The concepts and phenomena of agency and intention – both individual and shared – are at the heart of an enormous range of targets of human reflection. There are countless examples from moral philosophy (there is, of course, the very idea of moral responsibility and accountability; and think also about the distinction between intending and expecting that is built into the principle of double effect), from literature and the arts (think of the ideas of authorial intent, and of artistic collaboration), from social scientific studies (which tend to presuppose certain models of individual agency and practical rationality), from law (think of appeals to intent in the criminal law, and to the very idea of a conspiracy), from artificial intelligence (what exactly are we looking for an artificial version of?), and from efforts to understand the interaction between neuroscience and our self-understanding as agents, both in our commonsense and in domains of accountability, such as the law. For example, if the planning theory can help us answer the question about agential authority noted earlier – what psychological structures are such that when they guide the agent governs? – then perhaps we can make better sense of how a neuroscience of mind and behavior, one that traffics solely in an event causal order, can be compatible with the reality of agency, of agents as the source of actions. And we can ask whether this can be part of the effort to understand the compatibility of such a science with fundamental norms of accountability. A related question is whether this helps us better understand current debates, both in philosophy (deriving from Anscombe’s work) and in psychology (for example, some of the skeptical views of Daniel Wegner14), of the nature, scope and limits of our knowledge of our own agency. At another extreme, we can ask whether, and to what extent, our understanding of what I have called modest sociality can help us to understand larger forms of social agency. Prominent examples here include the social activities central to a legal system and/or to democratic politics. Is it fruitful to think of law, and/or democracy, as a distinctive kind of shared activity? (Scott Shapiro has recently been exploring the

14

Daniel M. Wegner, The Illusion of Conscious Will (Cambridge: Books, MIT Press, 2002).

Bradford

Michael Bratman

93

first question, about law, and Anna Stilz has been exploring the second query, about democracy.15) 6. How do you think your own work has contributed to the field of philosophy of action? And how does that relate to the way you think philosophy should be done? I am a “let many flowers bloom” sort of philosopher: I doubt we can know in advance which specific research projects will lead to important results. Indeed, in philosophy it is hard to know how significant the results are even once one has arrived at them – these judgments of significance can require years of further reflection on the part of the philosophical community. But it does seem clear to me that our understanding of our agency, both individual and shared, is as fundamental as things get in philosophy. And I am hopeful that a modest theory of the will such as that sketched in the planning theory can help us better understand how to think about our agency and about related matters of practical rationality, self-governance, sociality, and the location of our agency in the natural order.16

15

Scott Shapiro, Legality (Harvard University Press, 2011); Anna Stilz, Liberal Loyalty: Freedom, Obligation, and the State (Princeton University Press, 2009). 16 Thanks to Manuel Vargas for helpful comments on an earlier draft.

GEORGE AINSLIE DEPARTMENT OF PSYCHIATRY COATESVILLE VA MEDICAL CENTER    1. In your view, what are the most central (or important) problems in the philosophy of action? First a defining question: What should be the boundaries of our category, “action”? To my mind the most serviceable test for inclusion is whether the process is selected for repetition by what follows it. Roughly, is it a teleological process? This is somewhat like Dennett’s “intentionality”; but his is an experiential test. Many processes are shaped by what follows them without being consciously goal-directed, or even conscious, so my definition is much broader than the term might imply, probably broader than most people’s definitions. I summarize the factor(s) that selects processes for repetition as reward. This concept, too, is broader than conventional usage implies—broader than “pleasure”, and, perhaps, “utility”, although the latter term does not necessarily imply conscious valuation. The definition of an action as a reward-responsive process does not always sort experiences into binary categories. There are physiological processes that are driven mostly by hardwired stimuli, but can be changed by learning, as in biofeedback and some kinds of yoga. Bulimics have to induce vomiting by physical stimulation at first, but may eventually learn to vomit merely by intention. I discuss partial reward-responsiveness further in question 10. Now, the most central problems are: a) Is there a kind of deliberateness beyond simple reward dependency? How would this deliberateness relate to reward dependency? b) What explains inconsistency of preference (cf. question 9)? c) Is there a kind of self-control beyond the deliberateness of (a)? How would such self-control be related to conventional notions of will? d) How do non-deliberate actions differ from deliberate ones? e) If action is determined by motivation, is choice strictly determined by prior causes (vs. “free will”)? If so, are people morally responsible for their choices? (cf. questions 12 and 13) f) Can self-control be irrational? (cf. question 14)

96

Conversations on Human Action and Practical Rationality

2. You say that reward, pleasure and utility are similar. Couldn’t it be that pleasure and satisfaction on the one hand and reward and utility on the other are closer to each other? Why do you think that is so? (Both in case of a positive and a negative answer.) Utility is a term of art in economics, which considers only relatively long term rewards—the “utility” of biting your fingernails or emitting a tic would sound odd. Most people would say that of “pleasure” as well, although Freud, who used pleasure as a term of art, would not. My own suggestion is to define two terms of art: x Reward is whatever makes a behavior that it follows more likely to recur. The behavior in question may be just a mental process, and may even be the gratification of an urge that is subjectively aversive, for instance to panic. x Satisfaction is that subset of reward which operates over such a long term that it creates no longer-term incentive to forestall it. This would be closest to philosophy’s “rational” reward or true preference. Note that its term may still be fairly short (that is, it might pay off soon), as long as the long term prospect of it does not motivate avoidance. Utility and pleasure connote subsets of reward that operate over the middle-to-long term, and I would not try to make them technical. 3. Do you think we could say that reward is some kind of second order desire (as it would be called in mentalistic language), or does the concept of desire already include the pleasure and satisfaction linked to the fulfillment of desire? Which of the two scenarios makes the most sense: i) I have the desire to eat an ice-cream (1st order) and I have the wish to feel the typical pleasure of eating an Ice-cream (2nd order); or ii) I have the desire to eat an ice-cream, which already involves the property of giving one a characteristic sort of pleasure? There may be a distinction to be made here—between an action tendency (or urge to act) independently of reward and the selection of actions by reward. Some neurophysiologists (Kent Berridge, for instance) seem to have dissected the two apart, so that a subject may evince desire for a reward but not initiate behavior to take it. In the kind of example mentioned, though, I don’t see the usefulness of the distinction. But do you have an experience or thought experiment in which the two are felt separately?

George Ainslie

97

Certainly a person can fix her mind on a behavior and stop weighing incentives for or against it for a period of time—usually because she believes that further deliberation would just waste effort, but sometimes as the impulse-avoiding device I classify as attention control. In these cases the person’s intention to act might be said to be temporarily divorced from the differential reward for acting; but this divorce doesn’t have much functional importance. Even in the latter case the impulse-avoiding device is unstable, vulnerable to increasing urges to re-evaluate prospective reward as time passes or as current reward level decreases. 4. When we talk about utility and reward it seems clear that we have to use the concept of belief in order to say that the agent specifically believes the reward will be X, or the utility will be Y. Or don’t we? What is the role played by the mind in the production of such (for example, the mental states and not just the intentional states)? Yes, choice is governed by prospective reward. Any prospect can be called a belief at some level—but including Gendler’s concept of “alief.” Behavior, posture for instance, can be shaped by differential comfort without conscious awareness, even while asleep. I’m not clear how you’re bounding “mind”—or “state.” We try out scenarios in imagination—or scenarios arise and compete in imagination— until one passes some kind of threshold. David Redish has neurophysiological movies of this happening in a mouse’s hippocampus as it pauses at a fork in a maze. The difference in people is probably that our vastly increased computing power leads not only to more imagination and imagination of more remote events, but also modeling of other people’s imagination (“theory of mind”) and the recursive inclusion of our own choices in what we imagine (giving rise to the will—as I imagine it). At the phenomenological level I know that beliefs and intentions are conventionally examined separately. But if our intentions also function as cues—for instance as we evaluate our current choice as a test case of selfcontrol—then our intentions form part of our beliefs. 5. It seems clear that preferences in rational-choice theory are quite similar to belief-desire pairs (or the belief-desire binomial) in the theory of action in philosophy. Do you agree with that? How would you include the notion of preference as belief-desire in your explanation of action and rationality? Grossly, desires (or desired actions) compete on the basis of a set of beliefs, and the winner is “preferred.” So yes, the same thing. However,

98

Conversations on Human Action and Practical Rationality

beliefs also compete on the basis of how they fulfill our desires, depending variably on our restraints against wishful thinking. (Hyperbolic discounting supplies some constraints on doxastic voluntarism, but that is a whole other topic.1) And, as I just said, our observation of how our desires are doing in this competition is among our significant beliefs. A straightforward belief-desire-action model will be limited in scope. I see preference as the same thing as expecting relatively more discounted reward. 6. For some or all of the following problems - action, agency and agent - what do they contrast with most significantly? The main alternatives to the intertemporal bargaining model of will, as described below, are that will is a superfluous concept, that will exists but consists only in the recognition of the superiority of choosing in categories, that it is a muscle-like organ directed by a superior ego, and that it consists of a direction of attention or perception that limits the person’s re-appraisal of choices (“resolute choice”). The last of these closely resembles the intertemporal bargaining model. The main alternative to a comprehensive motivational marketplace, as described below, is visceral reward theory or one of its variants. The difference here hinges on technicalities of psychological experiment and probably should not be discussed here. 7. Which of these are liable to be rational/irrational? These alternatives do not differ in their rationality. They generally depict the use of willpower as the most rational choice, a conclusion to which intertemporal bargaining theory adds serious qualifications. 8. In what sense is the thing to do to be decided by what is rational? Are there limits to rationality? It is hard to divide the products of intertemporal bargaining strictly into rational and irrational. The test for rationality that most closely conforms to ordinary speech is whether a choice serves your long range interest. Prospective choices that seem best from a vantage point well in advance, which will tend also to be the choices that seem best in retrospect, should 1

See my “Recursive self-prediction in self-control and its failure” In Till GrueneYanoff and Sven Ove Hansson, eds, Preference Change: Approaches from Philosophy, Economics, and Psychology. Springer, 2009, pp. 139-158.

George Ainslie

99

be called rational. However, pursuit of the best long range outcome may lead you close to temptations that will derail your plan, making choice of the second-best long range outcome more realistic. For instance, the best long range diet may be to choose food according to what you will be gladdest of in retrospect, but this criterion may be so vague in actual practice that you are motivated to eat too much from moment to moment. The plan that actually achieves weight control may be to adopt explicit personal rules about food choice, which causes some opportunities to be lost because of the innate rigidity of rules, but which recruits enough motivation to overcome temptations. In ordinary speech, “rationality” is agnostic as to whether long or short range goals are better. Accordingly, rationality might be applied to a short range decision to evade a long range plan, taking the short range goal as a given and testing how it can “rationally” be obtained. “If you want to get drunk, it is rational to find a bar with a lenient bartender.” This usage distracts from the tougher question of the rationality of motives, which is what I addressed above. 9. How do you think your own work has contributed to the field? What do you think are your most important contributions? My work has been to search for the shape of the function that describes how future prospects are devalued with delay, and to explore the implications of the resulting nonexponential shape. The shape was roughly shown to be more bowed than an exponential curve by thought experiments (“would you rather have $50 now or $100 in two years? What about $50 in six years or $100 in eight years?”), and was demonstrated to be specifically hyperbolic by behavioral experiments in both humans and nonhuman animals. Some economists and psychologists still say that these findings are artifactual, but they are now widely accepted in both fields. Hyperbolic discount curves predict reversals of preference from larger, later (LL) rewards to smaller, sooner (SS) alternatives as the delay to them gets shorter. Therefore willpower is not a superfluous concept (as rational choice theory, RCT, holds), but has a necessary function—the maintenance of preference for LL alternatives over time. Various theories of willpower have ascribed it somewhat vaguely to a manipulation of attention or a muscle-like faculty, but hyperbolic discounting supplies a mechanism – intertemporal bargaining among successive motivational states, or selves – that specifically accounts for several of its familiar properties without needing an overarching controller or homuncular ego:

100

Conversations on Human Action and Practical Rationality

x The perception of your current choice as a test case for how you will choose in similar future cases defines a repeated prisoner’s dilemma-like situation that creates an incentive for intertemporal cooperation. This boost in patience from choosing in categories, first described by Aristotle, has been demonstrated by human and nonhuman behavioral experiments. The bundling of choices into categories by perceiving each as a test case cannot be looked for experimentally, but is arguably demonstrated by thought experiments such as Kavka’s, Newcomb’s, and Monterosso’s problems. x The recursive self-prediction created by intertemporal prisoner’s dilemmas makes your choice unpredictable even with complete knowledge of the incentives you face, and even by yourself. I have argued that this introspective opacity accounts for the experience of freedom of will, without appealing to any kind of indeterminacy. x The major writers on willpower have described properties that fit intertemporal bargaining: – that it is a new source of strength, – that it takes the previously weaker side, – that it unites actions under a common rule, – that it involves no diversion of attention, and – that it is strengthened by repetition and is exquisitely vulnerable to single lapses. x Intertemporal bargaining tends to make your incentive structure both more rigid and more brittle, leading to several negative consequences that are familiar but that were previously not recognized as side-effects of willpower: – a tendency to evaluate choices more as signals to yourself than as experiences in their own right; – circumscribed areas of dyscontrol where intertemporal defections (lapses) have ruined prospects for cooperation, including the patterns called phobias and addictions; – an incentive not to detect lapses, which fosters repression and denial; and – the competitive advantage of concrete criteria over more subtle or humanistic criteria for what to interpret as cooperation, a factor that motivates compulsiveness. These phenomena suggest a new framework for understanding the person. Models based on populations or legislatures date back to Plato’s Republic, and they are widespread in folk psychology. Experiential words are naturally borrowed from, and lent to, populations and legislatures. I

George Ainslie

101

readily think of myself as arguing with myself, making deals with myself, blaming myself, being of two minds, etc., just as I speak of nations having a character, pride, or will. As Walt Whitman said, “I contain multitudes.” However, scientific psychology has regarded these usages as romantic metaphors. More mechanized analogies have included ant colonies and chaotic, weather-like systems, but by excluding a central decision-maker these have seemed not to correspond to human experience. As Jerry Fodor cautioned about modular models, “If there is a community of computers living in my head, there had also better be somebody who is in charge; and, by God, it had better be me.” Fodor deals with cognitive modules; hyperbolic discounting creates motivational modules, raising different problems for the concept of selfhood but also suggesting solutions to them as well as to his worry about who is in charge. The trouble with population models has been in specifying how the motives of the different units conflict, or, if they are seen as autonomous centers, how they come to agree. That is, if the partagents that make up a person are in only limited conflict with each other, what are the rules for this conflict—how is it provoked, and how is it limited? Aside from picoeconomics’ specific predictions about how motivations change over time, it provides a rationale for how part-selves can split off and reunite. And since the main mechanism for uniting is self-prediction, the process would not be expected to feel external or mechanical, “like an epileptic seizure” as Robert Kane once complained, but like part of getting along with yourself. In the absence of the intertemporal bargaining process, ad hoc theories of splitting off part agents have appealed to unmotivated factors, such as the happenstance of conditioning. However, under controlled conditions conditioning has been found to amount to no more than the learning of associations, and adapts to new associations quickly. The need to explain why people choose in ways that go against their own identified best interests has burdened the concept of conditioning with the creation of additional motives, often in haphazard fashion, imposing a tyranny of coincidence. The conditioned urge is a deus ex machina that discourages further inquiry. Motivational contingencies may be unknown, but at least they imply the possibility of systematic search for them. For instance, the development of sudden cravings after a chance exposure to a stimulus, or with no external occasion at all, can be modeled. In addition to will in the sense of willpower and population-like properties of the self, there are two other problems in the theory of action that were not previously well defined, but that led to tacit assumptions framing most inquiry into motivation. I have called them the problem of

102

Conversations on Human Action and Practical Rationality

pain and the problem of pleasure. Hyperbolic discounting has suggested solutions to both. I count these solutions as accomplishments, even though they have not been widely accepted, or even discussed, by theorists of action. They are clearly also areas of future research. The problem of pain is how aversive events and their prospect can lure attention but deter motor behavior. The common intuition is that you choose to participate in pleasurable experiences but that you are forced to participate in aversive ones. Thus conventional theory says that attention to aversive events is hardwired, like a reflex, and attention to their prospect is a conditioned reflex. But alternatives to fear and pain can often compete with them on the basis of reward, for instance during sports or combat, or even routine daily activities, the absence of which leaves you vulnerable when you are trying to sleep; and not entering into aversive emotions is a learnable skill, especially well developed in behavioral treatments for panic disorder and in techniques of natural childbirth. Furthermore, all stimuli that can induce conditioning have a motivational valence as well, an association that is unlikely to be mere coincidence. However, theoretical attempts to make all mental selection depend on motivation have been stymied by the problem of how to attract attention with a negative valence. An unmotivated process such as conditioning has seemed necessary. Hyperbolic discount curves offer an alternative mechanism: Just as addictive binges are driven by a cycle of temporarily preferred, transient reward and a consequent hangover of nonreward, a more rapid cycle of urges that reduce ongoing reward can drive itches, obsessive-compulsive (OCD) symptoms, tics, and other “wanted-but-not-liked” behaviors. The motivational pattern of these itch-like behaviors suggests that incentives which attract attention but deter behaviors with longer latencies may consist of similar cycles that are more rapid than our flicker-fusion threshold. Such very rapid cycles may let aversive urges be irresistible, for instance to panic or to give in to the emotional (protopathic) component of pain, despite an almost instant plunge in net reward— “given-in-to-but-not-wanted.” To the extent that aversive experiences force you to participate in them they do so by an overwhelming urge to do so, “an offer you can’t refuse,” resistance to which is a difficult but valuable skill. In this way the theory of action can do without undimensioned causes that operate reflexively, and can handle all behavioral selection with a competitive marketplace model. The common selective factor, best summarized as reward, operates over a longer continuum of durations than “pleasure,” but is not of different valence.

George Ainslie

103

Incentives at its longer end are experienced as pleasures, those at its shorter end as unpleasant urges. The problem of pleasure is that positive emotions are clearly accessible without physical stimulation, and can even be trained to occur at your will, for instance in actors; but people still depend largely on external events to occasion these emotions. As with aversive emotions, conventional theory invokes an automatic, unmotivated response of emotions to certain prepared stimuli, and a transfer of these emotions by conditioning to other stimuli. It has seemed only common sense that pleasure has to be released by external events—otherwise people could sit and reward themselves ad lib, short circuiting the adaptive reward contingencies by which the environment motivates behavior. But people do have a great capacity to experience imaginary scenarios emotionally, some “fantasy prone personalities” to a maladaptive degree. The constraint on this process seems not to lie in the limited evocativeness of imagination, but in the decreased emotional impact of a scenario with repetition (a decrease that is apparently less in the fantasy-prone). This decrease in impact occurs also with physically based pleasures, the more so the more they have a component of anticipation. It occurs despite long intervals of engaging in some other activity, which rules out neuronal fatigue as a cause. This decrease is what demands explanation. Again hyperbolic discounting suggests a motivational model: To evoke strong, positive emotion, scenarios must include periods of deprivation, in which suspense or longing builds up. A hyperbolic discounter will temporarily prefer lesser, earlier payoffs to the more intense ones that entail initial deprivation, and so will be lured to harvest emotional reward as soon as it becomes even slightly available. She will thus stay at a high level of satiety, just as someone who feeds herself by continual grazing will not get much pleasure from food. In response to this tendency we learn to cue our emotions to relatively infrequent occasions that are outside our control. Even so, when these occasions recur predictably our attention jumps ahead to their high points, and any suspense component is lost. Since attention cannot be restrained by willpower (monitoring a rule not to think about X entails thinking about X), the most effective occasions will have to be surprises. Thus the best model for how we optimize emotional reward is gambling. To the extent that we gamble on challenging jobs, unpredictable relationships, and even our own creativity, the outcomes reward us more intensely than even the greatest assured attainments. The motivational constraint on pacing emotional reward arises from our intrinsically poor ability to generate surprise.

104

Conversations on Human Action and Practical Rationality

These properties of action permit a theory of how a person that is a population of reward-seeking processes can form higher mental processes—“ego-processes” and an “ego-identity.” Familiarity with your own short-term pleasure-seeking will lead to increasing foresight and an incentive to select among these processes from longer and longer perspectives. Your action patterns form puzzles for your own solution. At some stages the solutions are apt to be explicit intertemporal prisoner’s dilemmas, that is, rule-based disciplines; but these lead in turn to ruleworship and invite searches for utility at a still higher level. The action patterns that that reward in the longest term often take the form of character, religious beliefs, or subordination of all tasks to one great love. Since these patterns govern a pyramid, not of solid stone but of dynamically balanced self-predictive contingent rewards, they are apt to discourage probing questions that might undermine the whole edifice. Reward maximizing for such patterns often takes the form of the paradoxical statement, “It doesn’t matter what you believe, as long as you believe it exclusively.” The need for surprise makes an operational formula for rationality impossible. We can say that rational action is that which maximizes longest-term reward, but such maximization requires the operation of two processes that are necessarily contradictory: Maximization of physical pleasures and the occasions for pleasant emotions is achieved by instrumental efficiency, which in turn often requires systematic selfcontrol. But these measures reduce surprise and thus the value of the pleasures obtained. Refreshment of appetite requires the risk of defeat, loss, pain, and other unpleasurable processes, which, to stay credible, must sometimes be realized. To court such losses is difficult and is often procrastinated. As Konrad Lorenz said of a life of leisure: “The whole glorious amplitude of the waves of human emotions, all that makes life worth living, is dampened down to a scarcely perceptible oscillation between scarcely perceptible tiny displeasures and pleasures. The result is an immeasurable boredom.” However, if periods of asceticism were an adequate solution it could be learned. Farmers learn to let fields lie fallow, after all, and heroin addicts learn to detoxify periodically just to cheapen their habits. But to see your highest goal as maximizing pleasure would be to undermine the discipline of the higher values that make success in your maximizing projects unique and thus uniquely satisfying. There are lots of alternative ways to get even long range pleasures. If you gamble on an outcome that is uniquely worthwhile it is irrational to invite losses, and if you catch yourself at this necessary risk-taking you threaten your belief in the original task. Gamblers must not try to lose, and yet it is the general .

George Ainslie

105

observation of people who depend on gambling for their satisfaction that they avoid declaring the game won and leaving it. Without the persistent risk there is not the intermittent satisfaction. But the gambler’s paradox is everyone’s paradox, because we only maximize satisfaction by choosing tasks that often defeat us, while trying as hard as we can not to be defeated. 10. What explains action and how? What is the role of deliberation in rationality? An action can be seen as any process that originates within the individual, the operation of which eliminates a possible alternative. When you kick with your leg you are unable not to be kicking. While you are succumbing to fear you are unable to be calm. While you think of a white bear you are unable not to think of a white bear. I do not recognize a distinction in kind between actions and passions, or between actions and thoughts. I do recognize degrees of motivatedness, and degrees of deliberateness. By motivated I mean the appropriate object of Dennett’s intentional stance. In most cases motivation implies goal-directedness, but the word “goal” implies both consciousness and some separation in time between the act and its outcome. The sleeper shifts her posture to avoid discomfort; this behavior is selected by its outcome and thus, in my usage, motivated, but do we want to call it goal-directed? Harder, the subaltern pays attention to the pain of snuffing a match with his fingers, not having learned Lawrence of Arabia’s trick (in the film) of not minding it; do we say that his attention to the pain is goal-directed? I don’t care, as long as we can say it is motivated. An action may be explainable as having been selected by its expected consequences, or having been driven by mental/neurological states that are insensitive to expected consequences, or by a mixture of both kinds of mechanism. If you kick with your leg because you are told to do so it is intentional; if you kick because the doctor has struck your knee with a hammer, it is unintentional; if you try to kick when the doctor strikes your knee with a hammer, it is apt to have both components. Emotions also may have both components. There are known to be instinctive fears, such as of heights; but these turn out, like other emotions, to be strong urges that are gratified in some sense by obeying them, as shown by the sometime experience of developing the skill to resist them, or just by finding yourself distracted from them by a more motivating event. Thus unless they are driven by an electrode in the brain I count emotions as at least

106

Conversations on Human Action and Practical Rationality

partially motivated in the broad sense I gave above. This is a break from the conventional theory that emotions are motivating but not motivated. I do not mean that all internal processes that are selected by outcomes are motivated. For instance, the increase in your metabolism in response to a fall in body temperature is not motivated, but the response of stamping your feet is. The difference is that metabolism responds only to specific, presumably hardwired feedback circuits, while the increase in comfort that rewards stamping your feet could select for any other process that might be substituted for stamping. “Motivation” implies a universal currency, best called “reward,” that is necessary to resolve the competition between all members of the class of processes that can be substituted for one another. I refer to the hypothetical site at which these processes compete for expression as the motivational marketplace. Membership in the class of substitutable processes is best seen as a matter of degree. Shivering is a borderline case, since you can resist the urge to shiver if you pay attention, but otherwise you will shiver involuntarily. Similarly, you breathe or hold your breath with voluntary muscles, but if you don’t interfere you breathe autonomously. The role of motivation is even harder to be sure of in attention itself. In many cases there seems to be no chance to avoid a thought—for instance, the thought of a white bear when you see a picture of a white bear, or even when you deliberately try not to think about white bears—so the possibility that the thought is ultimately selected by reward cannot be tested. Under some circumstances motivation can keep you from thinking about white bears, as when you are afraid you are about to fall off a cliff, or when a hypnotist has instructed you (and you are a good subject), but this does not refute the possibility that from some places in your train of thought, the thought of white bears will follow regardless of consequence. Still, the longer the train of thoughts that leads to a given thought, the more the occurrence or not of the final act can be influenced by reward. This consideration is important in the motivation of belief. Most of the processes that seem to have been discussed in philosophers are not only motivated but also deliberate. I shrink from a hardheaded definition, but the general idea is that deliberate actions have been tested in the marketplace of reward to the extent that you choose them while conscious both of them and of their immediate alternatives. Thus in a sense an action is deliberate if it is chosen “with all things considered;” but that phrase has usually included the assumption that the things considered do not change their relative values as a function of elapsing time. All things considered at this moment, you may want to tell your boss what you think of her, but that does not mean you would have wanted to do so when

George Ainslie

107

considering the same things at an earlier moment. Your wholehearted decision to confront her is deliberate, which is to say that contrary motives had more chance to stop it than they had of your just feeling anger or remembering occasions for anger (which are also actions in the broad sense I use). I am not much interested in defining the boundary between actions that are deliberate and those that are merely motivated, nor do I think the boundary is a sharp one. What I find important is the question of how the process of deliberation may be motivated, that is, how it can itself be one of the competitors in the marketplace: A hyperbolic pattern of discounting expected reward describes conflicting interests within the person. Conventionally we could say that a person has an interest in a present pleasure that conflicts with her later welfare, and has to decide between them. However, this way of putting it includes no hypothesis about who or what this person consists of, how she weighs the alternatives, whether or not she is bound to choose according to differential reward, or indeed whether her choice is “free,” is predictable but unpredicted, is unpredictable in principle, or meets some other test upon which society might base questions of personal responsibility. Hyperbolic discounting suggests a model of the self as a population of reward-seeking processes, which includes explicit hypotheses about these attributes. Shorter and longer range interests can be seen as rewardseeking processes that compete according to a rule of strict maximization of discounted expected reward at each moment of choice, and are motivated by game-theoretic considerations to form the very deliberative processes that we have been wondering about. To put this another way: The long tails of hyperbolic discount curves approach proportionality to the values that the various alternative rewards would have when immediate, and so motivate the learning of farsighted “ego functions” that forestall or cultivate the patterns of action that they foresee. These functions may be learned from the bottom up and become active or not in particular cases on the basis of the expected differential reward for them, without the supervision of any central organ. The most accessible example of a governing faculty that can be derived from individual motives has been the will. In ordinary speech multiple meanings of “will” are poorly differentiated. The will may be the faculty that connects thought to action, or the property of being owned by the actor, or the means by which temptations are resisted. An action is often spoken of as willed if it is merely intentional, but the connotation that will makes action consistent over time, as in “strong” or “weak” wills, is only tangentially captured in this usage. In a more precise term of art, will is what stabilizes plans in the face of temptations, without using physical

108

Conversations on Human Action and Practical Rationality

commitments or manipulating attention or emotion — “with the eyes open” as William James said. The possibility of such a function, or even the need for one, has been controversial in philosophy (and psychology and economics), but I appeal to empirical findings. The tendency of all vertebrates that have been studied to value prospective reward in inverse proportion to its delay (hyperbolic discounting) predicts both a need for stabilization of choice over time and a mechanism by which self-aware individuals can sometimes achieve this. Awareness that your current choice predicts the same choice in similar future circumstances should tend to bundle perceived choices into categories, the aggregate prospective reward for which will be increased or decreased by your current choice. To some extent this bundling effect should happen spontaneously, but it will be greatest when a conscious resolution defines what kinds of choices are relevant test cases and what options are consistent with the resolution. I have argued elsewhere that a similar process underlies the sudden eruption of appetites and emotions, but evidence for these non-deliberate examples is more tenuous. I nowhere mean to argue that preferences arise from a blank slate through even-handed trial and error. Examples abound of inborn preparedness for one motor response, emotion, or perception over another. But most arguably operate through motivation, and can be modified or overridden by contrary motivation. Likewise, although a child may learn many, or most, of her ego functions from the wisdom of her culture, they will still thrive or not on the basis of the differential reward that follows their activity. 11. So, let us call a “motivated” action any action performed by our body. Is it that there is among such action a special class of “deliberated”, conscious actions? Sure. Or at least there is a continuum from unthinking (or unstoppable) responses to responses we withhold until we have explored every objection. I don’t have a specific theory about how a threshold of action gets set, but among other things the extent to which we look at our current action as a test case must be one determinant.

George Ainslie

109

12. If action is determined by motivation, is choice strictly determined by prior causes (vs. 'free will')? Complete lack of causation is unimaginable, and random causation would not lead to a person’s sense of owning an action. The experiences that convince people that their wills are free are 1) the unpredictability of a choice given only their prior incentives for it, and 2) their active instigation of the choice2. The recursive self-prediction that I’ve discussed in previous answers such as 9 predicts both phenomena. Interpreting current choices as test cases for how you can expect to decide similar choices in the future (intertemporal bargaining) makes recursive self-prediction a major factor in decision-making. Choosing with an awareness that you are signaling yourself leads not only to strength of will but also what is experienced as freedom of will. This is best illustrated when the incentives for alternatives are closely balanced; then small changes in the prospects for future cooperation swing the decision between cooperation and defection. Belief about the direction of the present choice will be a major factor in estimating future outcomes; but this estimate in turn affects the probability that the present choice will be in that direction. Thus the decision process is recursive— not logically circular, but continuously fed back in real time like the output of a transistor to its own input. Here, I argue, is the hiding place of the various demons and soul-particles and Geiger counters that libertarians have hypothesized to be the first causes of choice. If the person's predictions about her propensity to make the choice in question are at all in doubt, this feedback process may play a bigger role in her decision than any pre-existing incentive, external or internal. For instance, a smoker trying to quit is offered a cigarette, guesses that she will be able to resist it, experiences this guess as an increase in the likelihood that she will reap the benefits of quitting, and thus has more prospective reward to stake against the temptation. Then she sees a new excuse to have a smoke— for example, she is facing a once-in-her-career exam that is especially stressful. The person’s worry that she will try the excuse and not get away with it—that is, that she will subsequently judge her choice to have been a lapse— will reduce her expectation of future abstentions and thereby her stake against lapses. This fall may be so great as to make the expected values of smoking this time vs. abstaining about equal, until some other consideration tips her self-prediction one way or the other. When the choice is closely balanced it might be argued to be modeled by Newcomb’s problem, with smoking the cigarette in the small box and 2

Cf. Richard Holton, Inquiry 52: 412-428, 2009 and many others.

110

Conversations on Human Action and Practical Rationality

long term cessation in the large. The being who can predict her choice is the person herself. I have argued elsewhere that a one-box strategy is not magical thinking as has been suggested, but an intuitive application of recursive self-prediction to this seeming paradox3. Intertemporal bargaining does not explain only choices in which the incentives are closely balanced. For example, psychologist David Premack described the example of father who put off picking up his children in the rain to get a pack of cigarettes, and, when he noticed what this meant about his character, gave up smoking on the spot.4 Such sensitive dependence on interpretations of small observations or thoughts lends itself to theories of an overarching ego, but it can be fully derived from intertemporal bargaining. Any choices that the person perceives as sensitive to recursive self-prediction should be experienced as free, even where there is no conflict or indecision. Furthermore, the experience of freedom of will is not an illusion. When we engage in intertemporal bargaining we are doing a real thing that keeps us from being a throughput for pre-existing incentives, and which separates us from—indeed defines—“lower” animals that lack the necessary self-awareness. The reason why people are only tangentially aware of intertemporal bargaining itself is another topic5, but I would argue that the feeling of freedom is one way in which we sense it. 13. So, are people morally responsible for their choices? Yes, but this answer will require yet another hypothesis. The intertemporal bargaining model of will—how it can be functionally free while being strictly determined by prior factors—does not answer a moralist’s charge that strictly determined choices can’t be blameworthy. However, since the bargaining model does depict a process of selfcongratulation and self-blame, it does lay the groundwork for answering this charge. First we need to notice its implications for self-blame. The threat that faces a person in bargaining with her future selves is not their retaliation but a prediction of poorer reward in the future as those selves respond to the implications of a lapse. When she lapses she does not make a choice to blame herself, but suffers from her awareness of a loss that has 3

Cf. my “Can thought experiments prove anything about the will?”, In D. Spurrett, D. Ross, H. Kincaid and L. Stephens, eds., Distributed Cognition and the Will: Individual Volition and Social Context, MIT, 2007, pp. 169-196. 4 Cf. “Mechanisms of self-control”. In W. A. Hunt, ed., Learning Mechanisms in Smoking, Aldine, 1970, p. 115. 5 Cf. e.g. my Breakdown of Will, Cambridge, 2001, chapter 7.

George Ainslie

111

already happened—a loss of her trust in herself and the consequently poorer prospects for self-control, an awareness that we label self-blame. It might then make sense to ask whether our sense of social blameworthiness is modeled on our internal experience. Might social blame be vicarious self-blame, a reversal of the conventional view that self-blame is internalized social blame? At first this explanation seems to run counter to the truism that parental authority is the original source of impulse control. However, obedience serves two purposes: to make your choices fit your parents’ wishes, and to protect you from temporary preferences that you yourself expect to regret. You have two prospects at stake in obeying rules—your expectation of avoiding external blame, and your expectation of avoiding self-blame by containing impulses that would be harmful in their own right. As you become able to escape the scrutiny of parents and others, the second kind of stake—avoiding self-blame— becomes a separate incentive that has to stand on its own if you are to be seen as having “internalized” the rules. Such intertemporal bargaining is a practical tool for self-control that would not be undermined by an awareness of determinism. As you make your choice, an omniscient being might know whether or not your self-control is about to suffer a setback, but you don’t know it. Getting an estimate about this is one of the expected outcomes of your choice, and thus one of the incentives for this choice. Your choice is based upon your imperfect self-prediction; and this is true whether or not you believe that the outcome is predetermined. The new hypothesis is that people perceive social blame as an empathic extension of their personal processes of self-blame: “I blame her because in her shoes I would blame myself.” Granted, blame is widely used as a social deterrent, but the view of social blame simply as manipulation is too cynical to fit the common intuition of it, as authors such as Railton, Smilansky, and Strawson have pointed out. Despite academic theories about when a person is blameworthy, people tend to base judgments on a sense of deservingness. For instance, theoretical tests for legal insanity vary widely from state to state in America, from merely being impelled by a mental disease to not even knowing right from wrong, but the rate at which juries accept this defense does not vary much from state to state, or within a state when the charge to the jury is changed. In practice the assignment of social blame appears to be quick and sure, a process that Greene & Haidt describe as more of “affective intuition than

112

Conversations on Human Action and Practical Rationality

deliberative reasoning”6. Furthermore, punishing someone just so as to make an example of her is regarded as unfair. Of course, we are not in other people’s shoes, so there are many ways that vicarious self-blame can be unfair. We try to correct them by means of juries of peers and objective tests of how much control a subject had over her behavior; but these means are imperfect, as, indeed, are our tests for our own self-blame. The important point is that strict determinism does not undermine either kind of responsibility. 14. Can self-control be irrational? I define self-control as impulse control, the avoidance or attempted avoidance of temporary preferences for small, sooner over larger, later goods (or for larger, later over smaller, sooner bads). I define rational as “in the person’s longest range interest,” what she looks forward to at a distance with welcome and backward upon with approval— this as opposed to a formal test involving logical consistency, transitivity, and so forth. And I would not call what is merely mistaken “irrational.” By this definition any self-control that served your longest range interests would never be irrational. However, this definition excludes two common usages of “self-control”: 1) Efforts at self-control are often overdone, badly done, or misapplied because of innocent error or lack of skill. For example— x Copying what you take to have been your parents’ personal rules may not suit your own needs. x Avoiding some temptations may not be worth the effort it takes, or the cost in terms of lost opportunities. x An unavoidable feature of using intertemporal bargaining (personal rules) is that explicit rules are more enforceable than subtle ones and thus may come to replace them despite a consequent rigidity and loss of opportunities. Often the benefits are not worth the costs. To some extent everyone makes mistakes in choosing controls, but sometimes such mistakes result in major pathology. An example is dieting, which I described in question 8 as a departure from what might be an optimal eating pattern in the long run, but as the choice that best serves your long range interest if the greater enforceability of explicit criteria for 6 Cf. “How [and where] does moral judgment work?” Trends in Cognitive Science 6, 517-523, 2002, p. 517.

George Ainslie

113

choice is necessary to control your overeating. But dieting is carried too far in anorexia nervosa, which usually arises in people who have had impulses to eat too much, and who adopt draconian rules to control them. Anorectics’ severe rules typically wind up costing more than the original problem by interfering with social and emotional satisfaction—or sometimes by killing them. Still, a patient who believed that her anorexia was the best deal she could get—that her extreme self-control measures were necessary to prevent even greater harm to her longest range interests— should be called mistaken but not irrational. 2) Ostensible efforts at self-control can also come to serve short range or mid-range interests. For instance, anorectics may discover that they have a weapon against parents, a sense of power, even a source of metabolic intoxication. These short range rewards are protected by the rationale of a self-control project, and they make this project hard to give up. In effect an addiction has come to masquerade as self-control. Mid-range interests may be even harder to identify—where a person has “sold her soul to the Devil” to get a narrow kind of success despite some sense that it will disappoint her. Workaholics, misers, and other compulsives are prominent examples. This pattern is logically distinct from the mistaken use of overly demanding rules in category (1), but empirically these patterns are hard to discriminate. Overestimation of the need for self-control, lack of skill in applying it, and impulses using it as a disguise are the commonest targets of psychotherapy. In common speech all might be called irrational, but error is not itself irrationality, and disguised impulsiveness is not self-control. However, to the extent that a mid range interest succeeds in controlling a short range interest, but thereby defeats a long range interest, it might be called irrational. I have found “rationality” to be an awkward term of art for discussing intertemporal bargaining. I would gladly return it to the logicians. 15. How is akrasia possible (if you think it is)? In my thinking akrasia is the failure of a person, acting in her longer range interest, to protect that interest from an opposing shorter range interest. If Ulysses sails onto the rocks because he has failed to have himself lashed to the mast, he is akratic. A stricter, alternative usage might say that someone was akratic if she could not protect her longer

114

Conversations on Human Action and Practical Rationality

range interest using will alone. Then Ulysses would be said to be akratic with respect to Sirens in any case, but be rationally able to compensate for his akrasia with an external committing device. I am indifferent between these usages. 16. Your conception of akrasia is quite different from the notion most commonly used in philosophy. A somewhat standard notion of akrasia, in philosophy, would be the following: On doing X, a person acts incontinently (akratically) if and only if: (a) does X intentionally; (b) believes that there is an alternative action Y that can be done; and (c) believes that, all things considered, it would be best to do Y rather than X. In this case, Y could be buying a new car using all her savings and X could be buying a cheaper car and saving for the future/retirement. Is it possible to match that notion of akrasia with your approach to the subject? According to you, should the attribution of akrasia to an action/agent be based on external criteria (such as “saving in the long run is better than spending all your money right now”, in case that is the case) or internal criteria, related to beliefs-desires consistency in the agent himself? If we interpret “all things considered” to mean “taking a long term perspective,” than my concept of akrasia is the standard one. “All things” would include the likely effect of a person’s remembering her current choice on similar future choices, that is, its meaning as a test case of her self-control. When she considers this she is choosing, in effect, according to a personal rule, which approximately defines her belief “that it would be best.” (I say approximately because she may be aware that there are sideeffects of will, especially oversensitivity to such test cases—compulsiveness — that may keep it from maximizing long term reward.) However, if “belief it would be best” means “expectation of maximal reward” it needs to be qualified in one of two ways to account for the akrates’ seeming failure to maximize this expectation. In one case her belief “that it would be best” would be equated to her expectation of long term reward, the discounted value of which has simply fallen below that of her expectation of imminent reward (possibly because she has not taken into account the test case effect—a Davidsonian solution). In the other case she identifies an exceptional condition, such as a special occasion, that she falsely believes will prevent her in the future from interpreting her choice as a

George Ainslie

115

lapse of self-control and thus reducing her expectation of keeping resolutions. The difficulty of predicting such future interpretations represents a source of introspective opacity that changes beliefs into bets (and which, I argue elsewhere, is a mechanism for free will7). My criteria are entirely internal. If the person would prefer long in advance to buy the more expensive car (or, as a rough test of this, does not regret the choice afterwards), then buying it would not be akratic. The distinction is most important with the addictions. There are objective tests of addiction such as behaving despite self-harm, or with very strong motives, or in the presence of physiological tolerance, but I call a choice addictive or impulsive only if the person is motivated to avoid it in the more distant future, and/or expects to regret it. The person herself must be ambivalent; as a matter of observation, the ambivalence is always between obtaining an early payoff and avoiding deferred costs. 17. In a common scenario, from which moment may we consider that the agent’s “interest in the long run” is defined? Is it expressed by the agent himself or is it something that we can perceive by analyzing the behavior of the agent? Reward is always evaluated from the viewpoint of the present moment. An interest in the long run can exert influence on choice only to the extent of the current discounted values of the rewards that comprise it. This influence might be reportable, but inference from behavior is more reliable to the extent that we can observe the agent’s behavior (and is necessary in the case of nonhuman animals)—recall that I include mental processes in behavior. 18. Is it because you think that there is no centre – no central unit that can be called “the agent” - that you don’t think it is useful, for your model, to employ notions such as “decision”? A whole person can be an agent just like a legislature can be an agent, that is, can make decisions. Furthermore, deciding can be more than just bidding in a marketplace. Like members of a legislature, the interests in a person can agree upon and support executive functions—analyzing, planning, even directing attention, emotion, and self-perception in such a way as to favor some interests over others. My claim is that there is no 7

Cf. my “’Free will’ as recursive self-prediction: Does a deterministic mechanism reduce responsibility?” In George Graham and Jeffrey Poland (eds.) Addiction and Responsibility. MIT, 2011.

116

Conversations on Human Action and Practical Rationality

need to reify these functions as inborn faculties, and, conversely, that these functions lack the capacity to function without regard to reward. Just as parliamentary governments need votes, internal executive functions need prospective reward. I claim further that the conditions that elicit executive functions – the emergence of conflicting internal interests and the limited warfare relationship among them – are predicted by the hyperbolic discounting of prospective reward. 19. Philosophy and economics both use folk psychology as a source for their theories of action. Do you agree with this formulation (this way of putting things)? What is the role played by folk psychology in your research? To a great extent, I agree. Now we have experimental philosophy and neuroeconomics, but I don’t think either one has changed their theories of action much. I admit to being soft on folk psychology. I think common experience is part of what motivational science has to explain, even if that experience is colored by wrong-headed theorizing. As with thought experiments, we should distinguish between findings and interpretations, that is, between what people notice and what they make of it8. 20. What is your general opinion about experimental philosophy? More precisely, we’re talking about X-PHI, an area of philosophy whose important representatives are people like Stephen Stich, Shaun Nichols, Joshua Knobe, among others. Do you think that the use of experimental methods by philosophers could be beneficial to philosophy? X-PHI is definitely a healthy step. In my youth philosophers were not allowed to experiment and behaviorists were not allowed to philosophize (except, oddly, within the framework of Wittgenstein), to the detriment of both. What are people’s notions of free will? How many teeth does a horse have? There is an obvious way to answer such questions. The tough job arises when the answers are seemingly inconsistent with each other. Whatever the source of the findings, this will remain the core job of philosophy. I don’t know how X-PHI will affect philosophy as a profession. It erodes the boundaries between philosophy and social psychology, in a way that I have enjoyed so far, but I am apprehensive. Social psychologists 8

See my 2007 chapter in Spurrett et.al., Distributed Cognition and the Will, MIT.

George Ainslie

117

look for how their techniques can elicit findings, and are awkward in referring these findings to larger questions. Experimental philosophers define larger questions that need to be answered, and look for experiments that might help to answer them. These experiments are often cruder than those of the psychologists, but I would not be unhappy at being called an experimental philosopher. Certainly it is good to test assumptions where possible; and philosophy, which I regard as a behavioral science, needs to accept the outcomes of well-conducted tests. However, I have noticed how techniques tend to capture their fields—econometrics in economics, concurrent reward schedules in behavioral psychology, magnetic resonance imaging in neuropsychology, and so on. There is a kind of Gresham’s Law in research whereby the concrete tends to drive out the subtle. With the academic boundaries down I would hate for experts in experimental design and inferential statistics to bully philosophers into diverting energy from the analysis of meanings to increasingly fine-tuned tests of assumptions. Not a matter for panic, I think, but vigilance.

DANIEL HAUSMAN UNIVERSITY OF WISCONSIN – MADISON

1. In your view, what are the most central (or important) problems in the philosophy of action? I find it hard to say which are the most central or important problems in the philosophy of action, because I think the answer depends so heavily on the interests of particular investigators. Those interested in the philosophy of psychology may look to action theory to help clarify conative notions such as desire or intention in order better to understand the mind. Those interested in the nature of virtue may look to the philosophy of action for insight into rationality and the nature of the will. Those interested in the nature and problems of social cooperation may be mainly interested in what action theory has to say about planning, intention, and the interpretation of the circumstances in which one must act. My own interests in the philosophy of action derive from my interests in economics, which is constructed around a specific and highly idealized theory of rational action. Looking at the philosophy of action from this peculiar perspective, I see the main problems as the following: 1. 2. 3. 4. 5. 6.

What are preferences? How do preferences differ from evaluative judgments? How are preferences related to tastes or likings? How are preferences related to belief, desire, and choice (or action)? How should one explain and appraise changes in preferences? How does the economist's model of (rational) choice map on to more standard accounts of human action?

2. What, then, are preferences on your view? Preferences, like desires, play a large role in human life. People have preferences and desires about almost everything, and they constantly express their preferences and desires in their speech and actions. Before they can talk, children reach for one object rather than another. Animals

120

Conversations on Human Action and Practical Rationality

also express preferences and desires, as my dog does when I mention going for a ride in the car. Preferences, unlike desires, are comparative. To prefer something is always to prefer it to something else. If there are only two alternatives, one can desire both, but one cannot prefer both. Because they are comparative, preferences, unlike desires, require that one weigh alternatives. They are thus generally more cognitive, more like judgments, than are desires. When English speakers talk about preferences, they do not need to use the words, "prefer" or "preference." The waitress may ask me, "Which would you like?" rather than "Which would you prefer?" We often ask about preferences by asking people which alternative they would choose, which they like better, or which they think would be better. In ordinary English, the words “prefer” and “preference” can express four different concepts: 1. Enjoyment comparisons. When I say that I prefer Coke to Pepsi, I am talking about what gives me more enjoyment. An enjoyment comparison such as this one may compare either overall enjoyment or enjoyment in some particular regard. 2. Comparative evaluations. When a politician such as Churchill expresses a preference for resistance over surrender to the Nazis, he is not reporting which policy he enjoys more. He is talking about which policy he judges to be superior. Comparative evaluations such as this can be partial – implying a ranking with respect to some specific criterion – or total – implying a ranking with respect to every relevant consideration. From the perspective of a wartime leader, Churchill might prefer to resist, while as a lover of medieval architecture and monuments, he might prefer to surrender. When, as in this case, partial evaluations conflict, the agent must adjudicate among the different considerations if he or she is to construct a total ranking. Making a total comparative evaluation is more cognitively demanding than making an enjoyment comparison, and there is more room for mistakes. A total comparative evaluation takes into account every consideration the agent judges to be relevant. Economists almost always mean by the term, “preference”, a total comparative evaluation. 3. Favoring. When people say that affirmative action calls for racial preferences, they mean that it favors racial minorities – that is, it gives them a better chance to be hired by firms or admitted by universities. Favoring does not imply comparative evaluation nor does comparative evaluation imply favoring. Favoring also has nothing to do with any sort of enjoyment comparison.

Daniel Hausman

121

4. Choice ranking. Finally, when a waiter asks me whether I would prefer soup or salad with my meal, he wants to know my choice, not my evaluation or my views on my enjoyment. “Preferences” can thus also be choice rankings. The four senses of "preference" are distinct from one another. Preferences in these different senses are not even in the same category. Enjoyment comparisons and comparative evaluations are mental attitudes, while favoring and choosing are actions. These senses often come apart, and there is no reason why the same word should be used for all these senses. Indeed in most other languages the word that best translates the English word, “preference” cannot be used to express all of these concepts. Regardless of the words they use, people are constantly talking about their preferences, considering what their preferences ought to be, and inquiring about each other's preferences. Sometimes preferences require little thought. It doesn’t take much contemplation to determine which of two ice-cream cones tastes best. On the other hand, a decision concerning which profession one prefers to pursue may require weeks of reflection and discussion. It may be hard to know what one prefers. The most important and common meaning of “preference” is comparative evaluation. Individuals can evaluate alternatives in specific regards (i.e. an Obama presidency is worse from the perspective of affluent taxpayers), or they can make overall or total evaluations. In everyday usage, preferences are typically “overall” comparative evaluations. In an overall evaluation, agents compare alternatives with respect to most of what matters to them rather than in some specific regard or with respect to everything that matters to them. In an overall comparative evaluation, in contrast to a total comparative evaluation, people regard some of the factors that affect their evaluation of alternatives as competing with preferences rather than as influencing preferences. For example, I may turn down an invitation to a movie that I am eager to see, because of a promise I made to have lunch with my sick aunt. If asked why I turned down the invitation, I might say either that, owing to my promise I preferred to have lunch with my aunt, or I might say that I preferred to go to the movie, but I could not do so owing to my promise. In ordinary language we can take facts such as promising as influencing action via influencing preferences or we can regard them as leaving one’s preference ranking untouched and, like a constraint, ruling out alternatives. In their deliberations, people often treat moral considerations as constraints rather than as factors to be balanced against

122

Conversations on Human Action and Practical Rationality

others, and it is probably more natural to treat promises as competing with preferences in determining choices rather than as determining choices via their influence on preferences. But ordinary usage permits both usages. In economics, in contrast to ordinary language, everything that influences choices apart from beliefs and constraints does so via its influence on preferences. 3. And how do preferences differ from evaluative judgements? Preferences in economics motivate action and, for the most part, this is also the case in everyday language. An agent’s beliefs and preferences jointly determine which among the feasible alternatives the agent will choose. Unlike judgments (on one common view of judgment), preferences have a conative element: they are action-guiding. If one understands “evaluative judgments” as in this way motivating action, then preferences are one kind of evaluative judgment. There are clearly other kinds of evaluative judgments, which, unlike preference rankings, are not comparative. If, on the other hand, one understands evaluative judgments as not themselves intrinsically motivating, then, although incorporating evaluative judgments, preference rankings would not themselves be evaluative judgments. 4. How do you think preferences are related to beliefs, desires and choice (or action)? The short answer is that beliefs and desires are among the factors that determine preferences and that beliefs and preferences are among the feasible objects of choice that determine choices. Let me elaborate, beginning with the latter claim. Suppose that an individual, call her “Ann,” faces a choice among four alternative feasible actions C1, C2, C3, and C4, and suppose that Ann prefers C3 to C4, that she is indifferent between C1 and C2, but she prefers both C1 and C2 to C3 and C4. There are other actions that Ann would prefer to any of these four alternatives, but she knows that these other actions are not feasible. (In this way, constraints influence action both directly by making some actions impossible and indirectly by causing agents to believe that some alternatives are impossible.) If Ann believes that C1, C2, C3, and C4 are all feasible, then her preferences combined with her beliefs will cause her to choose C1 or C2. If, on the other hand, Ann falsely believed that only C3 and C4 are feasible, then Ann’s preferences combined with her beliefs will

Daniel Hausman

123

cause her to choose C3. Constraints, beliefs, and preferences among the alternatives jointly cause actions. If preferences among the alternatives available for choices were treated simply as givens, the paragraph above would say all that needs to be said. But the account of human choice would consist of the uninteresting platitude that among the alternatives agents believe to be available, they choose what they most prefer. Fortunately, preferences among the alternatives available for choices are not treated simply as givens. They depend on beliefs about the consequences and the properties of alternatives, preferences among the consequences, and a panoply of other psychological factors. So, for example, the expected utility of an action (which is an indicator of where that action is located in the agent’s preference ranking) is the sum of the expected utilities of the outcomes of the action the agent believes to be possible weighted by the agent’s subjective probabilities (degrees of belief) concerning the possible outcomes. 5. How should one explain and appraise changes in preferences? Changes in preferences have several sources, and depending on the sources, one may offer different appraisals of the changes: x Changes in preferences owing to changes in states of affairs. Suppose that I generally prefer going to movies to attending concerts. Owing to an accident, I lose my eyesight, and I now prefer attending concerts to going to the movies. It is easy to rationalize this change in preferences in terms of more general unchanged preferences for pleasant and rewarding experiences. It would be more difficult to explain and defend a change in preferences among flavors of ice cream consequent to a loss of vision – though (of course) preferences among flavors of ice cream are not in need of much defense. x Changes in preferences owing to changes beliefs. These are also often easy to explain and to justify. Someone who initially preferred Obama over Romney for president might come to believe that Obama is a Kenyan-born Muslim terrorist and thus come to prefer Romney instead. Although one might have reasonable doubts about the justification and truth of this agent’s beliefs, given his changed beliefs, there is nothing puzzling about the change in his preferences. x Changes in preferences owing to new experiences. Having seen and touched raw oysters, Gerald may prefer eating almost anything to consuming raw oysters. Challenged by his friend Geraldine, he

124

Conversations on Human Action and Practical Rationality

tries one and finds it delightful; and his preferences immediately change. Whether or not one assimilates this source of preference change to the previous one in terms of belief change, there is nothing problematic about this sort of belief change. x Changes in preferences owing to deliberation and miscellaneous psychological processes. Without having new experiences or changing one’s beliefs, people can come to feel differently about alternatives via discussion and reflection. This process may be rational, as in a case in which the arguments a friend makes help to make certain facts salient, or it may be irrational, as in a case in which an advertisement with an attractive couple enjoying a bowl of jello leads one to prefer jello to crème brulé for desert. Psychologists have uncovered a panoply of factors that influence preferences, many of which do so in ways that are hard to defend. For example, individuals who express a preference for policy A over B when the policies are described in terms of the number of lives they save may express the opposite preferences when the very same policies are described in terms of the numbers of individual whose deaths they permit.1 6. Some philosophers believe the following: decision leads us to a specific intention and this, in turn, leads us to act. Do you think that could be the way things unfold, without having to imagine that between decision and action everything in the world remains constant? It seems difficult to defend the view according to which today’s decision to buy a dog tomorrow depends, in order to be effective (i.e. in order to actually bring me to action), on all things remaining constant. What distinguishes a decision which brings me to action from another which doesn’t? Is it reasonable to think that we continue deliberating until the very instant which precedes action? Let me paraphrase the question as follows. Suppose that today I form the intention to do something tomorrow, such as buy a dog. Sometimes such an intention leads to my buying a dog tomorrow. Sometimes it doesn’t. So it appears that deliberation does not come to an end with the formation of an intention. In that case, why mention intention at all? Regardless of my intention-forming today, my action tomorrow depends on deliberation then. 1

Amos Tversky and Daniel Kahneman, "The Framing of Decisions and the Psychology of Choice," Science 211(1981): 453-8.

Daniel Hausman

125

I think that this question is mistaken. My reasons are basically those of Michael Bratman’s,2 and I have nothing to add to his work in this regard. An intention is a plan that permits reconsideration if one acquires new information or experiences or has a sufficiently strong reaction when one is about to carry out the intention. So the fact that people may reconsider is not inconsistent with forming a settled intention. On the other hand, unless there is cause for reconsideration, a settled intention determines action without further deliberation. So it is far from otiose to invoke intentions. 7. So, how is akrasia possible (if you think it is)? My views on akrasia are rather heterodox. Akrasia purported occurs when (a) an agent – call her Barbara – has a settled preference for one course of action or a secure intention to carry out an action, yet (b) when the time to act arrives without any new information or constraints, Barbara voluntarily acts in a different way, and (c) immediately afterwards she regards herself as failing to do what she had intended and regrets the failure. Doubts about whether akrasia is possible are doubts about whether these three conditions can consistently hold. Those who doubt the existence of akrasia will maintain that either Barbara did not have a settled preference or a secure intention, or she received new information or faced new constraints, or the purportedly akratic action was involuntary, or she did not really regard herself as failing. In my view, in contrast, once the (purported) phenomenon has been described in terms like those above, then the only problems remaining concern either the advantages of alternative theoretical redescriptions or the psychological mechanisms responsible for the phenomenon. I don't think that there is in fact much of philosophical interest in the problem. There are different costs and benefits in describing akrasia as reflecting a failure of voluntary choice, as due to inconsistent preferences, as due to unadjudicated conflicting desires, as a rapid-fired preference change and then change back, and so forth; and the theories that result will have different virtues and drawbacks. But having admitted the apparent phenomenon, as we must, it does not seem to me of great importance whether we maintain that akrasia is real or merely apparent.

2 Michael Bratman, Intentions, Plans, and Practical Reasoning. Stanford: Center for the Study of Language and Information, 1999.

126

Conversations on Human Action and Practical Rationality

8. In your view, what explains action and how? What is the role of deliberation in rationality? This is obviously a huge, difficult, and controversial question. Let's take folk-psychology as our starting point. According to folk psychology beliefs and desires and intentions explain actions. Agents have some set of desires and aversions for the outcomes of actions and perhaps also some desires or aversions to particular actions themselves; and they have beliefs about what consequences actions are likely to have. They may also have plans and intentions, which are not reducible to beliefs and desires. Deliberation consists in considering whether to form, modify, or retain an intention and in considering which action will best satisfy the agent's desires. This last claim tells us what deliberation accomplishes, but gives few hints on how deliberation works. The economist's model of rational action is much in the spirit of folk psychology, but it is much simpler. The economist begins by supposing that agents have already completed almost all of their deliberations and have managed to adjudicate in the light of their beliefs among their intentions, plans, and desires to form a ranking of alternatives in terms of everything relevant to evaluation and choice. If this ranking extends to the immediate objects of choice, then there is nothing left for the economist to do except to say that agents choose whatever feasible alternative lies at the top of their rankings. Apart from describing the formal properties of such rankings, there is nothing to be said about deliberation or rationality. What gives economists something to do and prevents their model of rational action from becoming trivial is that economists tackle one part of the problem of preference formation. They suppose that preferences over the outcomes of choices are already given, but not preferences over the immediate objects of choice themselves. Those preferences are derived from preferences over outcomes and beliefs (or subjective probabilities) concerning the consequences of the immediate objects of choice. So there is some room for deliberation in the economist's "refinement" of folk psychology, but deliberation has been reduced to calculation. 9. Rational choice theory is a normative theory, right? Can you explain what this means? What role does such theory play in the explanation and prediction of individual human actions? Rational choice theory is a normative theory of action, preference and belief. What this means is that it does not consist in generalizations concerning how people actually behave, or what properties their preferences and beliefs actually have. It instead prescribes certain structural properties

Daniel Hausman

127

that their beliefs, preferences, and actions ought to possess. It thereby functions as a standard according to which one can assess the actions, beliefs, and preferences of individuals. An individual’s beliefs, preferences, and choices are not justified if they do not conform to the requirements of rational choice theory. The prescriptions laid down by rational choice theory are not moral requirements. Its standards are not moral standards, and the justification of belief, preference, and action that comes with conforming to those standards is not moral justification. What sort of normativity is it then? The unhelpful answer is that it is rational normativity. Someone who fails to adhere to the standards of a fully adequate rational choice theory would show a rational failing. They would be foolish or imprudent. To the extent that the actual formulations of principles of rational choice may be faulty, then it is (of course) possible that violations are not irrational. If one asserts that individuals are in fact rational, then the principles governing rational belief, preference and choice will constitute generalizations concerning features of people’s actual beliefs, preferences and choices. There are more and less stringent versions of rational choice theory. The least stringent view takes people’s preferences to be complete and transitive. It idealizes and supposes that people have complete knowledge, which means that beliefs match facts and need not be discussed further. Finally, it takes choices among feasible alternatives to be determined by preferences. More stringent versions require consistency in preference across different sets of alternatives and impose new conditions such as independence. When the assumption of perfect knowledge is relaxed, degrees of belief are required to conform to the axioms of the probability calculus, and preferences among lotteries should be positively responsive to larger probabilities of preferred outcomes. Actions should depend on beliefs and preferences in the way discussed above in the answer to query 4. One might argue that rational choice theory itself plays no role whatsoever in the explanation or prediction of individual human actions. For example, to explain and predict behavior, one relies not on the claim that the preferences of an agent ought to be transitive, but on the empirical generalization that the preferences of agents are in fact transitive. But I think that this is a mistake. One of the grounds for believing that the empirical generalization is a reasonable approximation to the truth is the conviction that people are largely rational and that transitivity is a condition of rationality. Furthermore, in explaining and predicting behavior, we are often concerned with the reasons why people action, not merely with the causes of their actions. Talk of reasons raises questions

128

Conversations on Human Action and Practical Rationality

about whether behavior is justified and requires that agents be depicted as, to some extent, rational. 10. Which of these – action, agency and agent – are liable to be rational/irrational? Actions, agency, and agent can all be characterized as rational or irrational, although the contrast between what is rational and what is irrational differs in each case. Despite the fact that economists more or less identify action (choice) with rational action, they are not committed to the counterintuitive view that there is no such thing as irrational action. The most common sources of irrational action are irrational beliefs, irrational preferences, or a failure of the connection that should exist between an agent's actions and her beliefs and desires. Although some of these irrationalities have formal characterizations, irrationality is often a substantive matter. Those who believe that Michelle Obama is a Martian, like those suffering from Parfit's "future Tuesday indifference", need not be guilty of any inconsistency in order to count as possessing irrational beliefs or desires. Although not unrelated, the contrasts between rational agency and irrational agency and between a rational agent and an irrational agent are not quite the same. In addition to rational action, rational agency requires that the outcome in question be rationally intended. (A rational action that has the outcome in question as an accidental and unintended result is not an instance of rational agency.) A rational agent, on the other hand, possesses a certain sort of psychology. Rational agents are likely to act rationally, but there is nothing about being a rational agent that precludes having irrational beliefs and acting irrationally. And irrational agents need not act irrationally. 11. In what sense is the thing to do to be decided by what is rational? Are there limits to rationality? If "the thing to do" is the thing there is most reason to do and what is most rational is what there is most reason to do, then the thing to do is whatever is most rational to do. Even if one sets aside qualms about the two "ifs" in the previous sentence, that sentence doesn't fully answer the question, because something can be rational without being most rational. Although there are some models of rationality, such as those in economics, that deny that it can be rational to do A if it would be more rational to do B, our everyday standards of rationality are much weaker than this. Rationality is a matter of being "properly" as opposed to "perfectly" responsive to reasons.

Daniel Hausman

129

There are several further reasons to doubt that "the thing to do is to be decided by what is rational." On a purely formal view of rationality, the bearing of rationality on decision is in a sense trivial. What determines what is the thing to do lies in the substance of the various reasons, not in rationality itself. Secondly, "the thing to do" need not be identified with "what there is most reason to do". It could instead, depending on the context, mean "the most prudent thing to do", "the right thing to do", "the thing most people do", and so forth; and rationality does not coincide with all of these, if indeed it coincides with any of these. Finally, the claim that what is most rational is the same as what there is most reason to do is questionable on several grounds. First, what there is most reason to do is, plausibly, a matter of what the facts are, while what is most rational to do depends on what the agent believes rather than what is in fact the case. Second, consider a case where the agent's beliefs are correct and comprehensive and there is a choice between adhering to a plan and doing what there is most reason to do. In such a case, it is at least arguable that it may be more rational to adhere to the plan. I am not sure what is meant by asking in this context, "Are there limits of rationality?" Perhaps what is meant is, "Does rational deliberation always tell us what to do?" If that is the question, then the answer is "No." There are many instances where we simply cannot tell what is the best thing to do. In choosing between two opaque boxes, one empty and one containing $10,000, the thing to do is to choose the box containing $10,000, but rationality gives me no guidance at all about which to choose. 12. For some or all of the following problems – action, agency and agent – what do they contrast with most significantly? As to some extent an outsider to the literature on the philosophy of action, I am not sure exactly what these three "problems" are. I take the problem of action to be the problem of articulating what exactly an action is and how an action differs from "mere behavior." I interpret the problem of agency to be linked to concerns about responsibility, both moral and causal. Finally I take the problem concerning agents to be linked to psychological questions concerning what must be true of something in order for it to be capable of action and of agency. 13. How do you think your own work has contributed to the field? What do you think are your most important contributions? What are your plans for future research? Although implicit in earlier work, it is only during the past decade that I've begun to contribute significantly to action theory, broadly conceived.

130

Conversations on Human Action and Practical Rationality

Because the literature on philosophy of economics overlaps very little with the literature in action theory, I suspect that my work has thus far had little influence on philosophers who would identify themselves as mainly concerned with action theory. My most significant contributions are the following: 1. An account of the way in which the theory of choice around which mainstream economics is constructed relates to folk-psychological views of belief, desire, preference and action. 2. Works probing the concept of preference economists employ and criticizing the main views of preference that economists defend. These include criticisms of so-called "revealed-preference" theories, and theories that identify preferences with expected benefits. 3. An account of the strategy economists employ to explain and to predict actions that makes clear the role that preferences play. 4. A clarification of the relations between preference, self-interest and welfare. This includes an examination of Amartya Sen's notions of sympathy and commitment and his defense of the possibility of counterpreferential choice. I am currently writing a book on preferences in which these themes are developed further. This book will be addressed both to economists and to those interested in the philosophy of action. 14. Given recent developments in experimental philosophy, do you see any chance of economists, experimental philosophers and psychologists joining efforts to study human action? If you think there are any, what do you think would be the main issues addressed? I am not enthusiastic about experimental philosophy. Although those who espouse it are right to challenge the cavalier use of “intuition” – that is to say, the philosopher’s inclinations concerning what to say – I doubt that there is much of philosophical interest to be learned about the variety of ways that people use words such as “knowledge.” For example, the fact that people often speak of someone knowing P without any commitment to P’s truth does not undermine the view that there is a crucial distinction between belief and knowledge, in part precisely because knowing P implies P. I do believe that there are many fruitful possibilities for collaboration between economists, philosophers and psychologists. Consider, for example, some marvelous experiments reported by Ernst Fehr and

Daniel Hausman

131

Simon Gächter3. In these experiments subjects in groups of 4 with changing membership repeatedly played public goods games (much like prisoner’s dilemmas, where cooperation is best for the group but defection is a dominant strategy for each individual). Cooperation is initially high, but after 10 rounds, virtually everybody defects. Results like these led economists previously to suggest that once everyone is self-interested and that initial cooperation is merely the result of confusion. What Fehr and Gächter then did is to allow the groups to play another ten rounds, but this time, at a cost to their own payoff, subjects were allowed to punish defectors. With this option, which has no selfinterested benefit, cooperation now became prevalent, indeed in some groups universal. So what one sees in such experiments is not “human nature” but the outcome of a complicated interaction between psychological propensities and institutional facts. Such work has implications for economic research and policy and for political philosophy. I do not know what are the limits to the possibilities of such collaboration, but I conjecture that questions in ethics and political philosophy are likely to be particularly fruitful areas. Consider, for example, the contemporary work by economists, psychologists, and philosophers on subjective well-being,4 on altruism,5 or on so-called 3

“Cooperation and Punishment in Public Goods Experiments,” American Economic Review 90 (2000): 980-94. 4 See D. Hausman, “Hedonism and Welfare Economics,” Economics and Philosophy, 26 (2010) 321–344; Dan Haybron, The Pursuit of Unhappiness: The Elusive Psychology of Well-Being. Oxford: Oxford University Press, 2010; Daniel Kahneman, “Objective Happiness,” In Well-being: Foundations of Hedonic Psychology, ed. D. Kahneman, E. Diener and N. Schwarz, 3–27. New York: Russell Sage Foundation Press, 1999; Daniel Kahneman, “Experienced Utility and Objective Happiness: A Moment-based Approach,” in Choices, Values and Frames, ed. D. Kahneman and A. Tversky, 673–692. NewYork: Cambridge University Press and the Russell Sage Foundation, 2000; Daniel Kahneman and Robert Sugden, “Experienced Utility as a Standard of Policy Evaluation,” Environmental and Resource Economics 32 (2005): 161–181; Daniel Kahneman and Alan Krueger, “Developments in the Measurement of Subjective Wellbeing,” Journal of Economic Perspectives 20 (2006): 3–24; and Daniel Kahneman and Richard Thaler, “Utility Maximization and Experienced Utility,” Journal of Economic Perspectives 20 (2006): 221–234. 5 Elliott Sober and David Sloan Wilson, Unto Others. Cambridge, MA: Harvard University Press, 1999; C. D. Batson, Altruism in Humans. New York: Oxford University Press, 201l; James Andreoni, “Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving,” Economic Journal 100 (1990): 464-77.

132

Conversations on Human Action and Practical Rationality

“nudges.”6 7

6

Richard Thaler and Cass Sunstein, Nudge: Improving Decisions About Health, Wealth, and Happiness. New York: Penguin, 2009; Luc Bovens, “The Ethics of Nudge,” in T. Grüne-Yanoff and S.O. Hanssen, eds. Preference Change: Approaches from Philosophy, Economics and Psychology. New York: Springer, 2009, pp. 207-19; Dan Hausman and Brynn Welch, “To Nudge or not to Nudge,” Journal of Political Philosophy 18(2010): 123-36. 7 The works in which these contributions are developed are the following: Economic Analysis, Moral Philosophy, and Public Policy (jointly with Michael McPherson, Cambridge University Press, 2006; first edition 1996), chapters 4, 5, 6 and 8. "Preference, Belief and Welfare," American Economic Review, Papers and Proceedings, 84 (May, 1994) (with Michael McPherson), pp. 396-400. "Rational Choice and Social Theory--A Comment," Journal of Philosophy 92(1995): 96-102. "The Impossibility of Interpersonal Utility Comparisons." Mind 104 (1995): 47390. "Revealed Preference, Belief, and Game Theory." Economics and Philosophy 16(2000): 99-115. "Sympathy, Commitment, and Preference," Economics and Philosophy 21(2005): 33-50; forthcoming in Fabienne Peter and Hans Bernhard Schmid, eds. Rationality and Commitment. Oxford: Oxford University Press, 2008, pp. 4969. "Consequentialism, and Preference Formation in Economics and Game Theory," Philosophy 59 Supplement (2006): 111-29. "The Philosophical Foundations of Mainstream Normative Economics", The Philosophy of Economics: An Anthology (3rd edition) (co-authored with Michael McPherson). Cambridge: Cambridge University Press, 2007, pp. 22650. And especially, my Preference, Value, Choice, and Welfare. Cambridge University Press, 2011.

JOSHUA KNOBE YALE UNIVERSITY

1. In your view, what are the most central (or important) problems in the philosophy of action? The philosophy of action is now undergoing a radical shift that involves a return to the issues that traditionally animated the field. If one looks back at the major figures in the history of philosophy – all the way from Aristotle to Nietzsche – one finds a deep concern with questions about what human beings are really like. Then, for a brief period in the twentieth century, there was a retreat from these traditional questions and a preoccupation with more abstract technical issues. (For some reason, it was thought that the traditional questions were ‘not truly philosophical.’) It seems to me that this period is now coming to an end and that we are witnessing a resurgence of the traditional emphasis on deeper explorations of human nature and human thought and feeling. But of course, a great many things have changed since philosophers first took up these questions. We now have access to experimental methods that enable us to make a new kind of progress on these traditional issues. Instead of simply reasoning from the armchair about how the human mind appears to work, we can now make use of all the resources of contemporary cognitive science. So I think what we are seeing at this point is an attempt to go after the old questions using a set of new methods. 2. Which are the “old questions” that you think are coming back in the domain of philosophy of action? Would you say questions such as “What is an action?” or “How can we explain actions?” should no longer be answered by philosophy alone? Do you think that there is need for a research programme which includes, for instance, psychology? I certainly hadn’t meant to suggest that philosophers ought to stop working on these more technical questions, only that it has been exciting to see the way philosophers in recent years have shown a willingness to

134

Conversations on Human Action and Practical Rationality

return to the broader, more interdisciplinary kinds of issues that traditionally lay at the heart of philosophy. Perhaps the best way of conveying the tone of this new work is just to give an example. One of the most prominent of the traditional questions that is now returning is the question as to why people believe in free will. Nietzsche suggests that the answer might lie in our desire to justify certain acts of punishment. On his view, we start out by inflicting harm on wrongdoers, and then when we find ourselves trying to justify these acts of punishment, we end up saying that the wrong-doers must have had some special capacity for free will. (For this reason, Nietzsche describes the belief in free will as ‘the metaphysics of the hangman.’) Now, it is certainly quite difficult to figure out whether Nietzsche is right on this point, but it does seem like a deeply important and provocative claim, and philosophers are now conducting experimental studies that can help us figure out whether or not it is actually true. 3. What do you think is the relation between folk psychology and the philosophy of action? How do you think it works? In the late twentieth century, there were two completely distinct groups of researchers working on questions about the concepts of intention, freedom, action, and so forth. First, there were researchers in psychology who were examining these concepts using empirical methods (in the field known as ‘folk psychology’ or ‘theory-of-mind’). Second, there were researchers in philosophy who approached seemingly similar questions using more a priori approaches (in the field known as ‘philosophy of action’). In more recent years, researchers from these two disciplines have been working more closely with each other, and questions naturally arise about where exactly the boundaries of these disciplines might lie and how they might relate to each other. In my view, it would be better if everyone thought a lot less about questions like these. It simply doesn’t seem helpful to focus on these disciplinary distinctions and to worry which topics fall in which discipline. Instead, it might be better if we could try to stop thinking about the boundaries of our disciplines and just try to work together to address questions of mutual interest.

Joshua Knobe

135

4. Which of the following are liable to be rational/irrational: action, agency, agent? Contemporary research points to a pervasive irrationality that appears to infect just about all aspects of human thought and behavior. Our most fundamental decisions seem to be determined in large part by nonconscious stereotypes, quick and dirty heuristics, mental shortcuts. Over the course of the past few years, there has been an explosion of exciting philosophical work exploring the implications of this irrational aspect of human cognition. 5. In what sense is the thing to do to be decided by what is rational? Are there limits to rationality? I do believe that there are limits on the human capacity for rationality. The sorts of normative models proposed by philosophers just don’t represent real possibilities for creatures like us. These models seem simply to be changing the subject. They cannot be taken seriously as attempts to grapple with the question as to how human beings actually ought to think about their lives. Still, there is definitely room for serious work on these topics. Although it is no use proposing models that completely ignore the limitations of our human nature, there has been some very interesting research examining the ways in which human beings actually can come to think more rationally. Here is one example. Systematic experiments show that studying formal logic does not actually help people to think more logically. Since the techniques employed in formal logic are so different from the way human beings actually think about real problems, this sort of study only serves to make people better at solving puzzles written out in logical notation. But there is still room for hope: the experimental results also show that studying statistical reasoning actually does help people to think more rationally. Hence, if we can arrange for students to receive rigorous training in statistics, we truly can enable them to think more rationally about the problems they encounter in their lives.

136

Conversations on Human Action and Practical Rationality

6. If normative models of rationality are in fact distant from the agents’ psychological reality, i.e. from the way people actually deliberate and act, what do you think the consequence of this should be? Should they be kept as an ideal? Should they be changed? Take the case of logic – as you yourself pointed out, many techniques employed in formal logic are quite different from the ways human beings actually go about thinking about problems. Yet, that doesn’t mean we should revise them, or give them up. Or does it? Do you think that such a thing as a descriptive theory of rationality is possible, or necessary, or even to be desired? It certainly isn’t enough just to describe the way people ordinarily think; one also wants to engage with questions about how people ought to think and, in some cases, to criticize people’s ordinary modes of thought. One can see a real flowering of this sort of reflection within early modern philosophy, where many thinkers argued that there might be something wrong with our ordinary tendency to defer to authority and something far more right about trying to actually work out certain intellectual problems for oneself. One also finds a similar concern in contemporary philosophy, where there has been a tremendous amount of interesting work about the ways in which we should and should not be guided by our immediate intuitions. But, of course, one can only really be addressing the question of how human beings ought to think and reason if one in some way takes into account our limited cognitive capacities. If someone tells me that it would be rational to have a set of beliefs that was completely logically consistent, or that it would be rational to believe all the logical consequences of one’s own beliefs, then it seems like this person isn’t even engaged in the project of trying to figure out how actual human beings ought to go about thinking and reasoning. So even though the questions we face here are normative questions -- questions about how people ought to think or reason -- our answers to these questions are inevitably constrained by empirical truths about human nature.

Joshua Knobe

137

7. Given the fact that research on rationality (beginning with Kahneman and Tversky) could be taken to show that we are much more irrational than we think we are, in what sense is the thing to do still to be decided by what is rational? Why do we care so much about rationality after all? What is the relevance of models of rationality for philosophy of action, moral psychology, and so on? The fact that human beings so often act irrationally certainly doesn’t mean that the whole topic of rationality is not worthy of serious consideration. Far from it: rationality is an ideal we can strive for in certain domains, and we now know far more than we used to about how to actually become more rational. Here is a simple example. If it is a dreary and rainy day outside, people will tend (irrationally) to conclude that their lives as a whole are not going quite so well. It is as though they are thinking: ‘I feel kind of sad right now... That must be because my life as a whole is not what it should be.’ However, there is an easy way to avoid this problem. If you just think for a moment about the weather -- noticing how dreary and rainy it is -- studies show that you will not be nearly so influenced by it when thinking about the overall value of your life. Perhaps this is just the sort of example you were looking for in your previous question, a case in which there are things we can do -- things that truly are possible for ordinary human beings -- that make us able to think more rationally about the world around us. 8. What explains action and how? What is the role of deliberation in rationality? In some corners of philosophy, there has been an almost absurd overemphasis on the role of reflection and deliberation. The focus has been on our capacity to stand back, reflect on the value of our own desires, apply certain rational principles… Now, it is certainly true that people can think about certain problems in this way (just as people can think about certain problems by solving partial differential equations). But if one wants to engage in the study of the ways in which people might actually live their lives, it seems that these sort of processes must play only a very peripheral role. A proper study of human action needs to explore a broader array of factors: non-conscious and automatic processes, sudden flashes of emotion, all the subtle yet powerful influences that shape our lives.

138

Conversations on Human Action and Practical Rationality

9. How do you see the status of normative reasons from the point of view of the justification of action? Do you think a normative reason can be a reason for me only or does it, in order to justify my action, have to be somehow universal? You seem not to attach too much importance to our capacity to “stand back, reflect on the value of our own desires, apply certain rational principles”, etc. But how could we keep doing ethics, or economics, and not appeal to normative reasons and rational principles and constraints? I completely agree with your point here. Researchers should not be content just to reach a better understanding of the way we ordinarily think and act but should always aspire to help us improve our modes of thought and action. Still, it would be a mistake just to assume that this ameliorative project has to involve anything like the application of conscious principles. Thus, to take one prominent example, some people behave more morally than others, but the empirical evidence does not suggest that this difference arises because some people are better than others at applying conscious moral principles. It looks like some people have a better way of thinking about moral questions -- and perhaps we could learn how to think this way ourselves -- but it does not appear that we will make much progress here just by trying to formulate moral principles and tell people how to apply them. 10. Do you think that someone who shelves his books while sleepwalking acts intentionally? Why would that be so? Are there other reasons to act which are unconscious? What is the difference between acting consciously and acting by conscious reasons? These are all very interesting questions, and it would be wonderful to try conducting some experimental studies to see how people might answer them. One possible method here would be to look at the difference between people’s intuitions about the use of expressions like ‘because’ (which can be used to indicate merely causal explanation) and their use of expressions like ‘in order to’ or ‘on the grounds that’ (which can only be used to indicate reason explanation). Now, suppose that a person uses a certain tone of voice because he has a non-conscious desire to hurt his wife’s feelings. People might then say ‘He used that tone because he wanted to hurt his wife’ (a causal explanation), but would they also say ‘He used that tone in order to hurt his wife’ (a reason explanation)? It would be exciting to try running some studies to find out.

Joshua Knobe

139

11. How is akrasia possible (if you think it is)? I think that the most well-supported account is the sort of picture one finds in Plato. On this view, the mind is composed of a number of distinct parts. One of these parts – what Plato called thumos – serves as a kind of power that enforces the decisions of our reason and pushes down the inclinations of our appetites. Recent empirical work appears to be lending strong support to this basic picture. We now have overwhelming evidence that the mind truly is composed of distinct parts, suggesting that Plato’s general approach was on the right track. Moreover, there is now a considerable body of research pointing toward the existence of a part of the mind that plays precisely the role that Plato assigned to thumos. (Here, I am thinking especially of the innovative new experimental work from Roy Baumeister and colleagues.) 12. Do you think that akrasia, and other forms of traditional thinking about irrationality, is a real psychological phenomenon? From your answer it seems that you see akrasia as a fight between reason and appetites – is that really so? How do you think akrasia is possible from a psychological point of view? Actually there are a whole series of new and fascinating studies on this topic from Alfred Mele, Joshua May, Richard Holton, Carlos Mauro, Paulo Sousa and others. What these studies seem to indicate is that people do have an ordinary notion of something like ‘weakness of the will’ but that this notion doesn’t correspond to anything like the sort of thing we might include in a scientific theory of the mind. Instead, it is beginning to appear that people’s ordinary notion of weak will actually includes a role for moral considerations. So, for example, suppose I believe that it would be right to punch someone in the face, but then when I go to do it, I feel a burst of compassion and am unable to go through with my original plan. Many people might be reluctant to say that my inability to do what I planned to do was the result of any kind of weakness on my part. Yet, this intuition they have seems not merely to reflect a purely psychological conclusion. It seems that people’s intuitions in such cases can actually be influenced by judgments about what it would be morally right to do.

140

Conversations on Human Action and Practical Rationality

13. How does an experimental philosopher argue against the naturalistic fallacy? Or does he avoid it? (Do you think there is such thing as the naturalistic fallacy?) The ‘naturalistic fallacy’ arises when people suppose that one can somehow infer directly from purely scientific facts to claims about moral issues. This is definitely something to avoid. Clearly, it would be a mistake to say that we have gotten such-and-such experimental results and then to infer straightaway that a certain kind of moral view must be right or wrong. But there is also another sort of fallacy in the opposite direction, which we might call the ‘non-naturalistic fallacy.’ This is the fallacy of thinking that experimental results are somehow completely irrelevant to moral questions, so that philosophers can just ignore everything we know about human nature and answer moral questions completely from the armchair. This latter fallacy is surely just as pernicious as the former. We cannot use experimental results in some completely direct way to answer moral questions, but if we learn more about the psychological mechanisms underlying our moral intuitions, then this information -- in combination with a great deal of complex philosophical reflection -- can help us get a better handle on the fundamental questions of morality. 14. Do you think experimental philosophy is helpful in dealing with the problems of philosophy of action? Can experimental philosophy be useful in other domains, such as deciding on public policies, etc.? In my view, experimental philosophy has already proved enormously fruitful in thinking through questions in the philosophy of action. There are numerous experiments exploring people’s intuitions about free will, about intention, about weakness of the will, about group agency, and philosophers have developed complex and very promising accounts of the ways in which such data can bear on the relevant philosophical questions. I’m glad you also asked about the relevance of such work for more concrete questions involving public policy, etc. Unfortunately, I myself have no background in public policy, and I wouldn’t be qualified to give you a good answer. But that is what collaboration is for! The key to addressing these issues with the seriousness they deserve is presumably not so much for experimental philosophers like me just to speculate about possible policy implications but rather for people in philosophy to team up with researchers from other disciplines who are well positioned to go after these questions in a more rigorous fashion.

Joshua Knobe

141

15. How do you think your own work has contributed to the field? What do you think are your most important contributions? What are your plans for future research? It might initially appear that our ordinary capacity for understanding action is something like a scientific theory – a system of concepts and principles that allow us to predict and explain the behaviors we observe. The key idea behind my research is that this initial assumption leaves out something very important. Recent experimental research indicates that our ordinary way of understanding action is actually suffused with moral considerations. In other words, even when we are simply trying to understand what happened in a given case, we tend to conceptualize the behaviors we observe in terms of their moral status. Perhaps the best way to explain the basic idea here is just to describe one of my earliest experiments. In that study, subjects were randomly assigned to one of two conditions. Subjects in the ‘harm condition’ received the following story: The vice-president of a company went to the chairman of the board and said, ‘We are thinking of starting a new program. It will help us increase profits, but it will also harm the environment.’ The chairman of the board answered, ‘I don’t care at all about harming the environment. I just want to make as much profit as I can. Let’s start the new program.’ They started the new program. Sure enough, the environment was harmed.

After reading this vignette, they were asked the seemingly straightforward question: ‘Did the chairman intentionally harm the environment?’ To form the story for the ‘help condition,’ we can leave almost everything the same but simply replace the word ‘harm’ with ‘help.’ The story then becomes: The vice-president of a company went to the chairman of the board and said, ‘We are thinking of starting a new program. It will help us increase profits, and it will also help the environment.’ The chairman of the board answered, ‘I don’t care at all about helping the environment. I just want to make as much profit as I can. Let’s start the new program.’ They started the new program. Sure enough, the environment was helped.

142

Conversations on Human Action and Practical Rationality

After reading this second story, subjects were asked: ‘Did the chairman intentionally help the environment?’ The results showed a surprising asymmetry. Subjects in the harm condition said that the chairman harmed the environment intentionally, while subjects in the help condition said that he helped the environment unintentionally. Yet it seems that the only major difference between these two cases lies in the moral status of the behaviors that the agent performed. So it appears that people’s moral judgments can somehow influence their intuitions about the seemingly non-moral question as to whether or not a behavior was performed intentionally. At first, it was thought that this phenomenon might simply be due to some odd little quirk of the concept of intentional action in particular, but subsequent research has shown that the very same pattern emerges in people’s intuitions about many other concepts: causation, reason explanation, doing and allowing, valuing, and so forth. It is definitely beginning to look as though people’s moral judgments truly do affect their whole way of thinking about action.

APPENDIX REFERENCES OF THE PROJECT “CONVERSATIONS ON HUMAN ACTION AND PRACTICAL RATIONALITY”

Alfred Mele – Selected Bibliography Books (Author): Effective Intentions: The Power of Conscious Will. Oxford University Press, 2009. Free Will and Luck. Oxford University Press, 2006. Motivation and Agency. Oxford University Press, 2003. Self-Deception Unmasked. Princeton University Press, 2001. Autonomous Agents: From Self-Control to Autonomy. Oxford University Press, 1995. Springs of Action: Understanding Intentional Behavior. Oxford University Press, 1992. Irrationality: An Essay on Akrasia, Self-Deception, and Self-Control. Oxford University Press, 1987. Books (Editor): Free Will and Consciousness: How Might They Work? (co-edited with R. Baumeister & K. Vohs). Oxford University Press, 2010. Rationality and the Good (co-edited with M. Timmons and J. Greco). Oxford University Press, 2007. The Oxford Handbook of Rationality (co-edited with J. P. Rawling). Oxford University Press, 2004. The Philosophy of Action (Oxford Readings in Philosophy series), Oxford University Press, 1997. Mental Causation (co-edited with J. Heil). Oxford: Clarendon Press, 1993.

144

Conversations on Human Action and Practical Rationality

Papers: “Vetoing and Consciousness” in T. Vierkant, J. Kiverstein & A. Clark (eds.), Decomposing the Will, Oxford University Press (forthcoming). T. Stillman, R. Baumeister & A. Mele, “Free Will in Everyday Life: Autobiographical Accounts of Free and Unfree Actions,” Philosophical Psychology (forthcoming). “Libet on Free Will: Readiness Potentials, Decisions, and Awareness,” in L. Nadel & W. Sinnott-Armstrong (eds.), Libet, Free Will, and Responsibility, Oxford University Press (forthcoming). “Teleological Explanations of Actions: Anticausalism vs. Causalism,” in J. Aguilar & A. Buckareff (eds.), Causing Human Actions: New Perspectives on the Causal Theory of Action. MIT Press, 2010, pp. 183-198. “Scientific Skepticism about Free Will,” in T. Nadelhoffer, E. Nahmias, & S. Nichols (eds.), Moral Psychology: Classical and Contemporary Readings, Blackwell, 2010, pp. 295-305. “Conscious Intentions” in J. Campbell, M. O'Rourke, & H. Silverstein (eds.), Action, Ethics, and Responsibility. MIT Press, 2010, pp. 85-107. “Weakness of Will and Akrasia” Philosophical Studies 150 (2010): 391404. “Moral Responsibility for Actions: Epistemic and Freedom Conditions “Philosophical Explorations 13 (2010): 101-111. “Testing Free Will,” Neuroethics 3 (2010): 161-172. “Conscious Deciding and the Science of Free Will,” in R. Baumeister, A. Mele, & K. Vohs (eds.), Free Will and Consciousness: How Might They Work? Oxford University Press, 2010, pp. 43-65. “Moral Responsibility and History Revisited”. Ethical Theory and Moral Practice (2009) 12: 463-475. “Mental Action: A Case Study,” in L. O'Brien & M. Soteriou (eds.), Mental Actions and Agency Oxford: Clarendon Press, 2009, pp. 17-37. “Delusional Confabulations and Self-Deception,” in W. Hirstein (ed.), Confabulation: Views from Neuroscience, Psychiatry, Psychology, and Philosophy. Oxford University Press, 2009, pp. 139-157. “Have I Unmasked Self-Deception or Am I Self-Deceived?,” in C. Martin (ed.), The Philosophy of Deception. Oxford University Press, 2009, pp. 260-276. “Manipulation, Compatibilism, and Moral Responsibility”. Journal of Ethics (2008) 12: 263-286. F. Cushman and A. Mele, “Intentional Action: Two-and-a-half Folk Concepts,” in J. Knobe & S. Nichols (eds.), Experimental Philosophy. Oxford University Press, 2008, pp. 171-188.

Appendix

145

“Proximal Intentions, Intention-Reports, and Vetoing,” Philosophical Psychology 21 (2008): 1-14. “A Libertarian View of Akratic Action,” Studies in Philosophy and the History of Philosophy 49 (2008): 252-275. “Persisting Intentions,” Noûs 41 (2007): 735-757. “Self-Deception and Three Psychiatric Delusions: On Robert Audi's Transition from Self-Deception to Delusion,” in M.Timmons, J. Greco, & A. Mele (eds.), Rationality and the Good, Oxford University Press, 2007, pp. 163-175. “Reasonology and False Beliefs,” Philosophical Papers 36 (2007): 91118. “Self-Deception and Hypothesis Testing,” in M. Marraffa, M. De Caro, & F. Ferreti (eds.), Cartographies of the Mind, Kluwer, 2007, pp. 159167. “Decisions, Intentions, Urges, and Free Will: Why Libet Has Not Shown What He Says He Has,” in J. Campbell, M. O’Rourke, & D. Shier (eds.), Explanation and Causation: Topics in Contemporary Philosophy, MIT Press, 2007, pp. 241-263. “Free Will: Action Theory Meets Neuroscience,” in C. Lumer (ed.), Intentionality, Deliberation, and Autonomy: The Action-theoretic Basis of Practical Philosophy, Ashgate, 2007, pp. 257-272. “Practical Mistakes and Intentional Actions,” American Philosophical Quarterly 43 (2006): 249-260. “Free Will: Theories, Analysis, and Data,” in S.Pockett, W. Banks, & S. Gallagher (eds.), Does Consciousness Cause Behavior? An Investigation of the Nature of Volition, MIT Press, 2006, pp. 187-205. “Libertarianism, Luck, and Control,” Pacific Philosophical Quarterly 86 (2005): 395-421. “Agnostic Autonomism Revisited,” in J. Taylor (ed.), Personal Autonomy, Cambridge University Press, 2005, pp. 109-123. “Acção humana par excellence (“Human Agency Par Excellence”) in F. Mão de Ferro (ed.), A explicação da interpretação humana, Lisboa: Ediçoes Colibri, 2005. J. Gert & A. Mele, “Lenman on Externalism and Amoralism: An Interplanetary Exploration.” Philosophia, 32 (2005): 275-283. “The Illusion of Conscious Will and the Causation of Intentional Actions,” Philosophical Topics 32 (2004): 193-213. “Can Libertarians Make Promises?” in J. Hyman & H. Steward (eds.), Agency and Action, Cambridge University Press, 2004, pp. 217-241. “Action: Volitional Disorder and Addiction,” in J. Radden (ed.), The Philosophy of Psychiatry, Oxford University Press, 2004, pp. 78-88.

146

Conversations on Human Action and Practical Rationality

“Outcomes of Internal Conflicts in the Sphere of Akrasia and SelfControl,” in P. Baumann & M. Betzler (eds.), Practical Conflicts, Cambridge University Press, 2004, pp. 262-78. “Agents’ Abilities,” Noûs 37 (2003): 447-470. “Emotion and Desire in Self-Deception,” in A. Hatzimoysis (ed.), Philosophy and the Emotions, Cambridge University Press, 2003, pp. 163-179. “Soft Libertarianism and Flickers of Freedom,” in D. Widerker & M. McKenna (eds.), Moral Responsibility and Alternative Possibilities, Ashgate, 2003, pp. 251-264. “Philosophy of Action,” in K. Ludwig (ed.), Donald Davidson, Cambridge: Cambridge University Press, 2003, pp. 64-84. H. Beebee & A. Mele, “Humean Compatibilism.” Mind, 111 (2002): 201223. “Acting Intentionally: Probing Folk Notions,” in B. Malle, L. Moses, & D. Baldwin (eds.), Intentions and Intentionality: Foundations of Social Cognition, Cambridge: MIT Press, 2001, pp. 27-43. “Goal-Directed Action: Teleological Explanations, Causal Theories, and Deviance,” Philosophical Perspectives, 14 (2000): 279-300. “Deciding to Act,” Philosophical Studies 100 (2000): 81-108. “Reactive Attitudes, Reactivity, and Omissions,” Philosophy and Phenomenological Research 61 (2000): 447-452. “Responsibility and Freedom: The Challenge of Frankfurt-Style Cases,” in M. Betzler & B. Guckes (eds.), Autonomes Handeln, Berlin: Akademie Verlag, 2000, pp. 25-38. “Twisted Self-Deception,” Philosophical Psychology 12 (1999): 117-137. “Ultimate Responsibility and Dumb Luck,” Social Philosophy & Policy 16 (1999): 274-293. Reimpresso em E. Paul, F. Miller, & J. Paul (eds.), Responsibility, Cambridge: Cambridge University Press, 1999. “Motivation, Self-Control, and the Agglomeration of Desires,” Facta Philosophica 1 (1999): 77-86. “Is There a Place for Intention in an Analysis of Intentional Action?” Philosophia 27 (1999): 419-432. “Autoinganno e Controllo Delle Ipotesi” (Self-Deception and Hypothesis Testing; translated by Massimo Marraffa), Sistemi Intelligenti 11 (1999): 503-519. “Motivational Strength,” Noûs 32 (1998): 23-36. “Noninstrumental Rationalizing,” Pacific Philosophical Quarterly 79 (1998): 236-250. “Motivated Belief and Agency,” Philosophical Psychology 11 (1998): 353-369.

Appendix

147

“Flickers of Freedom,” Journal of Social Philosophy 29 (1998): 144-156. “Two Paradoxes of Self-Deception,” in J. Dupuy (ed.), Self-Deception and Paradoxes of Rationality, Stanford: CSLI Publications, 1998, pp. 3758. “Practical Irrationality: Two Kinds of Akratic Action,” Cadernos de Filosofia 4 (1998): 9-37. “Agency and Mental Action,” Philosophical Perspectives 11 (1997): 231249. “Passive Action,” in G. Holmström-Hintikka & R. Tuomela (eds.), Contemporary Action Theory, vol. 1, Dordrecht: Kluwer, 1997, pp. 135-143. P. Livingston & A. Mele, “Evaluating Emotional Responses to Fiction,” in M. Hjort & S. Laver (eds.), Emotion and the Arts, New York: Oxford University Press, 1997, pp. 157-176. “Internalist Moral Cognitivism and Listlessness,” Ethics 106 (1996): 727753. “Socratic Akratic Action,” Philosophical Papers 25 (1996): 149-159. “Rational Intentions and the Toxin Puzzle,” Proto Sociology 8/9 (1996): 39-52. Also in G. Preyer & G. Peter (eds.), The Contextualization of Rationality, Mentis: Paderborn, 2000. A. Mele & S. Sverdlik, “Intention, Intentional Action, and Moral Responsibility.” Philosophical Studies (1996) 82: 265-287. “Conceptualizing Self-Control,” Behavioral and Brain Sciences 18 (1995): 136-137. “Justifying Intentions,” Mind 102 (1993): 335-337. “History and Personal Autonomy,” Canadian Journal of Philosophy 23 (1993): 271-280. “Intending for Reasons,” Mind 101 (1992): 327-333. F. Adams & A. Mele, “The Intention/Volition Debate.” Canadian Journal of Philosophy 22 (1992): 323-338. J. Heil & A. Mele, “Mental Causes.” American Philosophical Quarterly 28 (1991): 49-59. “Irresistible Desires,” Noûs 24 (1990): 455-472. “Exciting Intentions,” Philosophical Studies 59 (1990): 289-312. “He Wants to Try,” Analysis 50 (1990): 251-253. “Errant Self-Control and the Self-Controlled Person,” Pacific Philosophical Quarterly 71 (1990): 47-59. “Intention, Belief, and Intentional Action,” American Philosophical Quarterly 26 (1989): 19-30. “She Intends to Try,” Philosophical Studies 55 (1989): 101-106.

148

Conversations on Human Action and Practical Rationality

“Intentions by Default,” Pacific Philosophical Quarterly 70 (1989): 155166. “Akratic Feelings,” Philosophy and Phenomenological Research 50 (1989): 277-288. “Motivational Internalism: The Powers and Limits of Practical Reasoning,” Philosophia 19 (1989): 417-436. Also in R. J. Wallace (ed.), Reason, Emotion, and Will, Ashgate, 1999. “Against a Belief/Desire Analysis of Intention,” Philosophia 18 (1988): 239-242. A. Mele & M. Smith, “The New Paradox of the Stone.” Faith and Philosophy 5 (1988): 283-290. “Are Intentions Self-Referential?” Philosophical Studies 52 (1987): 309329. “Intentional Action and Wayward Causal Chains: The Problem of Tertiary Waywardness,” Philosophical Studies 51 (1987): 55-60. “Is Akratic Action Unfree?” Philosophy and PhenomenologicalResearch 46 (1986): 673-679. “Aristotle on Akrasia, Eudaimonia, and the Psychology of Action,” History of Philosophy Quarterly 2 (1985): 375-393. Also in N. Sherman (ed.), Aristotle’s Ethics: Critical Essays, Rowman and Littlefield, 1999. “How to Represent Aristotelian Deliberation Syllogistically,” The New Scholasticism 59 (1985): 484-492. “Aristotle on the Roles of Reason in Motivation and Justification,” Archiv für Geschichte der Philosophie 66 (1984): 124-147. “Aristotle’s Wish,” Journal of the History of Philosophy 22 (1984): 139156. “Aristotle on the Proximate Efficient Cause of Action,” Canadian Journal of Philosophy, Supplementary Vol. X (1984): 133-155. “Choice and Virtue in the Nicomachean Ethics,” Journal of the History of Philosophy 19 (1981): 405-423. “The Practical Syllogism and Deliberation in Aristotle’s Causal Theory of Action,” The New Scholasticism 55 (1981): 281-316. “Aristotle on Akrasia and Knowledge,” The Modern Schoolman 58 (1981): 137-157. “On ‘Happiness and the Good Life’,” Southwestern Journal of Philosophy 10 (1979): 181-187.

Appendix

149

Some Encyclopedia Entries: “Free Will”, International Encyclopedia of Ethics, Blackwell, forthcoming. “Action Theory”, Encyclopedia Americana, forthcoming. “Self-Deception”, in E. Craig (ed.), Routledge Encyclopedia of Philosophy, Routledge, 2010. http://www.rep.routledge.com/article/V033 “Freedom of Will”, in M. Binder, N. Hirokawa & U. Windhorst (eds.), Encyclopedia of Neuroscience, Springer, 2009, pp. 1631-34. “Free Will”, in W. Banks (ed.), Encyclopedia of Consciousness, Elsevier, 2009, pp. 265-277. “Action”, in D. Borchert (ed.), Encyclopedia of Philosophy: Second Edition, New York: Macmillan, 2006, vol. 1, pp. 14-22. “Weakness of the Will”, in D. Borchert (ed.), Encyclopedia of Philosophy: Second Edition, New York: Macmillan, 2006, vol. 9, pp. 728-732. “Agent Causation” (p. 18); “Belief and Desire” (pp. 85-86); “Counterexample, Philosophy By” (p.181); “Moral Motivation” (p. 622); “Practical Syllogism,” (p. 747), in T. Honderich (ed.), Oxford Companion to Philosophy, 2nd ed., Clarendon Press, 2005. “Action, Philosophical Issues About”, in L. Nadel (ed.), Encyclopedia of Cognitive Science, Nature Publishing Group, 2003, vol. 1, pp. 20-23. “Self-Control” (pp. 1548-51); “Temperance” (pp. 1693-96), in L. Becker & C. Becker (eds.), Encyclopedia of Ethics, 2nd. ed., New York: Routledge, 2001. “Accidie” (p. 6), “Control” (p. 184), “Motivation” (pp. 591-92), “Socratic Paradoxes” (p. 861), “Toxin Puzzle” (pp. 924-25), in R. Audi (ed.), Cambridge Dictionary of Philosophy, 2nd ed., Cambridge: Cambridge University Press, 1999. “Intentionality”, in P. Werhane & R. Freeman (eds.), Encyclopedic Dictionary of Business Ethics, Oxford: Blackwell, 1997, pp. 339-340. “Basic Action”, in J. Kim & E. Sosa (eds.), A Companion to Metaphysics, Oxford: Blackwell, 1995, pp. 44-46.

Hugh McCann – Selected Bibliography Books (Author): The Works of Agency: On Human Action, Will and Freedom. Ithaca, New York: Cornell University Press, 1998.

150

Conversations on Human Action and Practical Rationality

Creative and Critical Thinking, 2nd ed. (textbook). With W. Edgar Moore and Janet McCann. Boston: Houghton Mifflin, 1985.

Papers: “Is Raising One’s Arm a Basic Action?” The Journal of Philosophy 69 (1972): 235-249. “Volition and Basic Action.” The Philosophical Review 83 (1974): 451473. “Trying, Paralysis, and Volition.” The Review of Metaphysics 28 (1974-75): 423-442. “Nominals, Facts, and Two Conceptions of Events.” Philosophical Studies 34 (1979): 129-149. “On Mental Activity and Passivity: A Reply to Thalberg.” Mind 88 (1979): 592-596. “The Trouble with Level-Generation.” Mind 91 (1982): 481-500. “Individuating Actions: The Fine-Grained Approach.” Canadian Journal of Philosophy 13 (1983): 493-512. “Rationality and the Range of Intention.” Midwest Studies in Philosophy 10 (1985): 191-211. “Intrinsic Intentionality.” Theory and Decision 20 (1986): 247-273. “Divine Conservation and the Persistence of the World.” (with Jonathan Kvanvig). In T. V. Morris, ed., Divine and Human Action: Essays in the Metaphysics of Theism. Ithaca, New York: Cornell University Press, 1988, pp. 13-49. “Intending and Planning: A Reply to Mele.” Philosophical Studies 55 (1989): 107-110. “Practical Rationality: Some Kantian Reflections.” Journal of Philosophical Research 15 (1990): 57-77. “Settled Objectives and Rational Constraints.” American Philosophical Quarterly, 28 (1991): 25-36. Also in A. R. Mele. ed., Readings in the Philosophy of Action. New York: Oxford University Press, 1997, pp. 204-222. “The Occasionalist Proselytizer: A Modified Catechism.” (with Jonathan Kvanvig). In J. E. Tomberlin, ed., Philosophical Perspectives 5, Philosophy of Religion. Atascadero, California: Ridgeview Publishing Co., 1991, pp. 587-615. “The God Beyond Time.” In L. P. Pojman, ed., Philosophy of Religion, second edition. Belmont, California: Wadsworth, 1993, pp. 231-245. “Dretske on the Metaphysics of Freedom.” Canadian Journal of Philosophy 23 (1993): 617-628.

Appendix

151

“Intenzione e Forza Motivazionale.” Discipline Filosofiche 2 (1993): 251272. “Paralysis and the Springs of Action.” Philosophia 23 (1993): 193-205. “Divine Sovereignty and the Freedom of the Will.” Faith and Philosophy 12 (1995): 582-598. “On When the Will Is Free.” In G. Holmström-Hintikka and R. Tuomela, eds., Contemporary Action Theory, vol. I. Dordrecht, Netherlands: Kluwer Academic Publishers, 1997, pp. 219-232. “Sovereignty and Freedom: A Reply to Rowe.” Faith and Philosophy 18 (2001): 110-116. “Edwards on Free Will.” In P. Helm and O. D. Crisp, eds., Jonathan Edwards: Philosophical Theologian. Aldershott, England: Ashgate Press, 2003, pp. 27-43. “The Author of Sin?” Faith and Philosophy 22 (2005): 144-159. “Intentional Action and Intending: Recent Empirical Studies.” Philosophical Psychology 18 (2005): 737-748. “Resisting Naturalism: The Case of Free Will.” In A. Corradini, S. Galvan and E. J. Lowe, eds., Analytic Philosophy Without Naturalism. Boston: Routledge, 2006, pp. 225-240. “Metaethical Reflections on Robert Audi’s Moral Intuitionism.” In J. Greco, A. Mele, and M. Timmons, eds., Rationality and the Good. New York: Oxford University Press, 2007. “The Will and the Good.” In C. Lumer, ed., Intentionality, Deliberation, and Autonomy: The Action-theoretic Basis of Practical Philosophy, Ashgate, 2007. “Pointless Suffering: How to Make the Problem of Evil Sufficiently Serious.” In J. L. Kvanvig, ed., Oxford Studies in the Philosophy of Religion, vol. 1. New York: Oxford University Press, forthcoming.

Some Encyclopedia Entries: “Action, Philosophy of,” “Practical Reason,” “Practical Reasoning,” and “Reasons for Action.” In R. Audi, ed., The Cambridge Dictionary of Philosophy. New York: Cambridge University Press, 1995. “Creation and Conservation.” In C. Taliaferro and P. L. Quinn, eds., A Companion to the Philosophy of Religion. Oxford: Basil Blackwell, 1997, pp. 306-312. “Creation.” In A. Hastings, ed., Oxford Companion to Religious Thought. New York: Oxford University Press, 2000, pp. 143-144. “Divine Power and Action.” In W. Mann, ed., The Blackwell Guide to

152

Conversations on Human Action and Practical Rationality

Philosophy of Religion. Oxford: Blackwell, 2005, pp. 26-47. “Creation and Conservation.” In D. Borchert, ed., Encyclopedia of Philosophy, 2nd ed. Detroit: Macmillan Reference USA, 2006.

Michael Bratman – Selected Bibliography Books (Author): Structures of Agency: Essays. Oxford University Press, 2007. Faces of Intention: Selected Essays on Intention and Agency. Cambridge University Press, 1999. Intention, Plans, and Practical Reason. Harvard University Press, 1987.

Books (Editor): Introduction to Philosophy: Classical and Contemporary Readings. John Perry and Michael Bratman eds., Oxford University Press, 1986.

Papers: “Individuation and Action,” Philosophical Studies 33(1978):367-375. “Practical Reasoning and Weakness of the Will,” NOUS 13(1979):153171. “Simple Intention,” Philosophical Studies 36(1979): 245-259. “Intention and Means-End Reasoning,” The Philosophical Review 90 (1981): 252-265. “Castaneda's Theory of Thought and Action,” in James Tomberlin (ed.), Agent, Language and the Structure of the World: Essays Presented to Hector-Neri Castaneda with His Replies, Indianapolis: Hackett, 1983, pp.149-169. “Taking Plans Seriously,” Social Theory and Practice 9 (1983):271-287. Also in R. Jay Wallace (ed.), Reason, Emotion, and Will, Ashgate, 1999 e em Elijah Millram, ed., Varieties of Practical Reasoning, Cambridge, Mass.: MIT Press, 2001. “Two Faces of Intention,” The Philosophical Review 93 (1984):375-405. Also in A. Mele (ed.), Philosophy of Action (the Oxford Readings in Philosophy Series), Oxford University Press, 1997. “Davidson's Theory of Intention,” in Bruce Vermazen and Merrill Hintikka (eds.), Essays on Davidson: Actions and Events, Oxford: Oxford University Press, 1985, pp. 13-26.

Appendix

153

“What Is Intention?” in Philip R. Cohen, Jerry Morgan and Martha E. Pollack (eds.), Intentions in Communication, Cambridge: MIT Press, 1990, pp. 15-31. (With D. Israel e M. Pollack) “Toward an Architecture for ResourceBounded Agents,” CSLI Research Report #104 (1987), and SRI International Technical Note 425. “Intention and Personal Policies,” Philosophical Perspectives III (1989), 443-469. (With D. Israel e M. Pollack) “Plans and Resource-Bounded Practical Reasoning,” Computational Intelligence 4 (1988): 349-355. Also in R. Cummins and J. Pollock (eds.), Philosophy and AI: Essays at the Interface, Cambridge: MIT Press, 1991. “Cognitivism about Practical Reason”, Ethics 102 (1991): 117-128. “Shared Cooperative Activity,” The Philosophical Review (1992): 327341. German translation: Hans Bernhard Schmid & David Schweikard, eds., Collective Intentionality, Suhrkamp, 2009, pp. 176-193. Spanish translation: Jorge Fabra, ed., Derecho y Convencionalidad, Editorial Universidad Libre, 2010. “Practical Reasoning and Acceptance in a Context,” Mind 101 (1992): 114. “Shared Intention,” Ethics 104 (1993): 97-113. “Intention partagee et obligation mutuelle,” (translated by Joelle Proust) in Les limites de la rationalite: I. Rationalite, ethique et cognition, Paris: Decouverte, 1997, pp. 246-266. “Pur une theorie modeste de l'action planifiee: response a Gauthier et Dupuy,” (translated by Jean-Pierre Dupuy) in Jean-Pierre Dupuy and Pierre Livet, eds., Les limites de la rationalite: I. Rationalite, ethique et cognition, Paris: Decouverte, 1997, pp. 75-87. “Moore on Intention and Volition,” University of Pennsylvania Law Review 142 (1994): 1705-1718. “Planning and Temptation,” in Larry May, Marilyn Friedman, & Andy Clark (eds.) Mind and Morals: Essays on Ethics and Cognitive Science (Cambridge: Bradford/MIT, 1996): 293-310. “Following Through with One's Plans: Reply to David Gauthier,” in Peter Danielson, ed., Modeling Rationality, Morality, and Evolution (Oxford: Oxford University Press, 1998): 55-66. “Toxin, Temptation, and the Stability of Intention,” in Jules L. Coleman and Christopher W. Morris, eds., Rational Commitment and Social Justice: Essays for Gregory Kavka (Cambridge University Press, 1998): 59-83. “I Intend that We J,” in Raimo Tuomela and Ghita Holmstrom-Hintikka,

154

Conversations on Human Action and Practical Rationality

eds., Contemporary Action Theory vol. II (Dordrecht: Kluwer -Synthese Library Series, 1997): 49-63. German translation: Hans Bernhard Schmid & David Schweikard, eds., Collective Intentionality (Suhrkamp, 2009): 333-355. “Identification, Decision, and Treating as a Reason,” Philosophical Topics 24 (1996): 1-18. Also in Laura Waddell Ekstrom, ed., Agency and Responsibility: Essays on the Metaphysics of Freedom (Boulder, Colorado: Westview Press, 2001). “Valuing and the Will” Philosophical Perspectives: Action and Freedom 14 (2000): 249-265. “Hierarchy, Circularity, and Double Reduction,” in S. Buss and L. Overton, eds., Contours of Agency: Essays on Themes from Harry Frankfurt (Cambridge, Mass.: MIT Press, 2002): 65-85. “Nozick on Free Will,” in David Schmidtz, ed., Robert Nozick (New York: Cambridge University Press, 2002): 155-174. “Two Problems About Human Agency,” Proceedings of the Aristotelian Society (editor: A. W. Price) 101 (2001): 309-326. “Shapiro on Legal Positivism and Jointly Intentional Activity,” Legal Theory 8 (2002): 511-517. “Autonomy and Hierarchy,” Social Philosophy & Policy 20 (2003): 156176. Also in E. Paul, F. Miller, & J. Paul, eds., Autonomy (Cambridge University Press, 2003): 156-176. “Shared Valuing and Frameworks for Practical Reasoning,” in R. Jay Wallace, Philip Pettit, Samuel Scheffler, & Michael Smith eds., Reason and Value: Themes from the Moral Philosophy of Joseph Raz (Oxford University Press, 2004): 1-27. “A Desire of One's Own,” Journal of Philosophy 100 (2003): 221-242. “Planning Agency, Autonomous Agency,” in James Stacey Taylor, ed., Personal Autonomy: New Essays on Personal Autonomy and Its Role in Contemporary Moral Philosophy (Cambridge University Press, 2005): 33-57. “Personal Rules and Rational Willpower,” San Diego Law Review 42: 1 (2005): 61-68. “What is the Accordion Effect?” The Journal of Ethics 10: 1-2 (2006): 519. “Three Theories of Self-Governance” in John Fischer, ed., Philosophical Topics 32: 1 and 2 (2004): 21-46. “Dynamics of Sociality,” Midwest Studies in Philosophy: Shared Intentions and Collective Responsibility XXX (2006): 1-15. “Intention, Belief, Practical, Theoretical” in Simon Robertson, ed., Spheres of Reason: New Essays on the Philosophy of Normativity (Oxford

Appendix

155

University Press, 2009): 29-61. “Anchors for Deliberation,” in C. Lumer e S. Nannini, eds., Intentionality, Deliberation and Autonomy (Aldershot, etc.: Ashgate, 2007): 187-205. “Intention, Belief and Instrumental Rationality,” in David Sobel e Steven Wall, eds., Reasons for Action (Cambridge: Cambridge University Press, 2009): 13-36. “Shared Agency,” in Chris Mantzavinos, ed. Philosophy of the Social Sciences: Philosophical Theory and Scientific Practice (Cambridge: Cambridge University Press, 2009): 41-59. “Reflections on the Philosophy of Action” in Jesus Aguilar & Andrei Buckareff, eds., Philosophy of Action: 5 Questions (Automatic Press/VIP, 2009): 17-23. “Setiya on Intention, Rationality and Reasons,” Analysis 69 (2009): 510521. “Intention Rationality,” Philosophical Explorations 12 (2009): 227-41. “Agency, Time, and Sociality,” Proceedings and Addresses of the American Philosophical Association (forthcoming)

Some Encyclopedia Entries: “Intention” in Robert Audi, ed., The Cambridge Dictionary of Philosophy (Cambridge: Cambridge University Press, 1995): 380-381. “Intention” in Jaegwon Kim & Ernest Sosa, eds., A Companion to Metaphysics (Oxford: Basil Blackwell, 1995): 243-244. “Intention” in Samuel Guttenplan, ed., A Companion to Philosophy of Mind (Oxford: Basil Blackwell, 1994): 375-379.

George Ainslie – Selected Bibliography Books (Author): Picoeconomics. Cambridge: Cambridge University Press, 1992. Breakdown of will. Cambridge: Cambridge University Press, 2001.

Papers:1 “Impulse control in pigeons”. Journal of the Experimental Analysis of Behavior 21: 485–489, 1974.

1

CF. http://www.picoeconomics.org

156

Conversations on Human Action and Practical Rationality

“Specious reward: A behavioral theory of impulsiveness and impulse control”. Psychological Bulletin, 82: 463–496, 1975 (with Herrnstein, R. J.) “Preference reversal and delayed reinforcement”. Animal Learning and Behavior 9: 476–482, 1981. “A behavioral economic approach to the defense mechanisms: Freud's energy theory revisited”, Social Science Information 21: 735-779, 1982. “Beyond microeconomics: conflict among interest in a multiple self as a determinant of value”. In J. Elster (ed.), The Multiple Self, Cambridge University Press, pp. 133-175, 1986. “Freud and Picoeconomics”, Behaviorism 17: 11-19, 1989. “Derivation of “rational” economic behavior from hyperbolic discount curves”. American Economic Review 81: 134–140, 1991. “Intertemporal choice. Derivation of “rational” economic behavior from hyperbolic discount curves”, American Economic Review 81: 334-340, 1991. “A utility-maximizing mechanism for vicarious reward: Comments on Julian Simons “Interpersonal allocation continuous with intertemporal allocation”, Rationality and Society 7: 393-403, 1995 “Studying self-regulation the hard way”, Psychological Inquiry 7, 1621, 1996 J. Monterosso, G. Ainslie, P. Toppi Mullen, B. Gault, “The fragility of cooperation: A false feedback study of a sequential iterated prisoner's dilemma”, Journal of Economic Psychology 23, 437-448, 2002 (com J.R. Monterosso), Building blocks of self-control: Increased tolerance for delay with bundled rewards, Journal of the Experimental Analysis of Behavior 79, 27-48, 2003 “Uncertainty as wealth”. Behavioural Processes, 64: 369-385, 2003. (with J. Monterosso), Hyperbolic discounting as a factor in addiction: A critical analysis. In Rudy Vuchinich & Nick Heather, eds., Choice, Behavioural Economics and Addiction, Pergamon, 2003. “The dangers of willpower: A picoeconomic understanding of addiction and dissociation”. In J. Elster & O-J. Skog, (eds.), Getting Hooked: Rationality and Addiction, Cambridge University Press, pp. 65-92, 1999 J. Monterosso, G. Ainslie, The behavioral economics of will in recovery from addiction, Drug Alcohol Depend. 2007 September; 90 Suppl. 1, 2007

Appendix

157

Papers in Progress: (2009) A. Hofmeyr, G. Ainslie, R. Charlton, D. Ross, “The relationship between addiction and reward bundling: An experiment comparing smokers and non-smokers” (2005) G. Ainslie, “Emotion as a motivated behavior”

Daniel Hausman – Selected Bibliography Books: Capital, Profits and Prices: An Essay in the Philosophy of Economics. New York: Columbia University Press, 1981. The Philosophy of Economics: An Anthology. (ed.) New York: Cambridge University Press, 1984. The Inexact and Separate Science of Economics. Cambridge: Cambridge University Press, 1992. Essays on Philosophy and Economic Methodology. Cambridge: Cambridge University Press, 1992. Economic Analysis and Moral Philosophy. Cambridge University Press, 1996 (with Michael S. McPherson). Economic Methodology: Crossing Disciplinary Boundaries. London: Macmillan, 1998 (with Roger Backhouse, Uskali Mäki, and Andrea Salanti). Causal Asymmetries. New York: Cambridge University Press, 1998. Economic Analysis, Moral Philosophy, and Public Policy. Cambridge University Press, 2006 (with Michael S. McPherson).

Papers: “How to do Philosophy of Economics,” in P. Asquith and R. Giere, eds., PSA 1980. East Lansing: Philosophy of Science Association, 1980, pp. 352-62. “Are General Equilibrium Theories Explanatory?” in J. Pitt, ed., Philosophy in Economics. Dordrecht: Reidel, 1981, pp.17-32. “John Stuart Mill's Philosophy of Economics,” Philosophy of Science, 48 (l981): 363-85. “Constructive Empiricism Contested,” Pacific Philosophical Quarterly, 63 (l982): 2l-28.

158

Conversations on Human Action and Practical Rationality

“Causal and Explanatory Asymmetry,” in P. Asquith and T. Nickles, eds., PSA 1982, vol. 1. East Lansing: Philosophy of Science Association, 1982, pp. 43-54. “The Limits of Economic Science,” in N. Rescher, ed., The Limits of Lawfulness. Pittsburgh: Center for Philosophy of Science, University of Pittsburgh, 1983, pp. 93-100. “Defending Microeconomic Theory,” Philosophical Forum, 15 (1984): 392-404. “Classical Wage Theory and the Causal Complications of Explaining Distribution,” in J. Pitt, ed., Change and Progress in Modern Science, Dordrecht: Reidel, 1985, pp. 171-97. “Is Falsificationism Unpracticed or Unpractisable?” Philosophy of the Social Sciences 15 (1985): 313-19. “Liability, Responsibility and Harm.” Ethics 97 (1986): 262-69. “Health Care: Efficiency and Equity,” in S. Spicker, S. Ingman and I. (eds.) Ethical Dimensions of Geriatric Care: Value Conflicts for the 21st Century. Dordrecht: Reidel, 1987, pp. 67-78. “Economic Methodology and Philosophy of Science,” in R. Teichgraeber e G. Winston, eds., The Boundaries of Economics. Cambridge: Cambridge University Press, 1987, pp. 88-116. “An Appraisal of Popperian Economic Methodology,” in N. de Marchi, ed., The Popperian Legacy in Economics. Cambridge: Cambridge University Press, 1988, pp. 65-86. “Economic Methodology in a Nutshell,” Journal of Economic Perspectives 3(1989): 115-27; rpt. in B. Caldwell, ed. The Philosophy and Methodology of Economics. Cheltonham: Edward Elgar, 1993, vol. 1, pp. 275-87 “The Insufficiency of Nomological Explanation,” Philosophical Quarterly 39 (1989): 22-35. “Explanatory Progress in Economics,” Social Research 56 (1989): 36181; rpt. in B. Caldwell, ed. The Philosophy and Methodology of Economics. Cheltonham: Edward Elgar, 1993, vol. 3, pp. 460-80. “Ceteris Paribus Clauses and Causality in Economics,” in A. Fine e J. Leplin, eds. PSA 1988, vol. 2. East Lansing: Philosophy of Science Association, 1989, pp. 308-17. “Are Markets Morally Free Zones?” Philosophy and Public Affairs 18 (1989): 317-33. “Decision Theory and the Deductive Method,” Richerche Economiche 43 (1989): 199-217. “What Are General Equilibrium Theories?” in W. Sieg, ed., Acting and Reflecting. Dordrecht: Kluwer, 1990, pp. 107-114.

Appendix

159

“Making Interpersonal Comparisons Coherently,” Economics and Philosophy (com Martin Barrett) 6 (1990): 293-300. “On Dogmatism in Economics: The Case of Preference Reversals,” Journal of Socio-Economics 20 (1991): 205-25. “Thresholds, Transitivity, Overdetermination, and Events,” Analysis 52(1992): 159-63. “On the Conceptual Structure of Neoclassical Economics - A Philosopher's View,” in Essays on Philosophy and Economic Methodology (1992), pp. 25-32. “Why Don't Effects Explain Their Causes?” Synthese 94 (1993), pp. 22744. “Liberalism, Welfare Economics, and Freedom,” Social Philosophy and Policy 10(1993): 172-97; also in E. Paul, F. Miller, and J. Paul, eds. Liberalism and the Economic Order. Cambridge: Cambridge University Press, 1992, pp. 172-97. “Taking Ethics Seriously: Economics and Contemporary Moral Philosophy” (with Michael McPherson) in The Journal of Economic Literature. 31 (June, 1993), pp. 670-731; also in A. Hamlin, ed. Ethics and Economics. Cheltenham: Edward Elgar, 1996. “Kuhn, Lakatos and the Structure of Economics,” in R. Backhouse, ed. Contemporary Issues in Economic Methodology. London: Routledge, 1994, pp. 195-215. “The Impossibility of Interpersonal Utility Comparisons.” Mind 104 (1995): 473-90. “The Composition of Economic Causes,” The Monist 78 (1995): 295-307. “Why Does Evidence Matter So Little to Economic Theory?” in Structures and Norms in Science. ed. M.L. Dalla Chiara, K. Doets, D. Mundici, and J. van Benthem. Dordrecht: Kluwer, 1997, pp. 395-407. “Problems with Supply-Side Egalitarianism,” Politics and Society 24 (1996): 343-51; Rpt. in Erik Wright, ed. Recasting Egalitarianism. London: Verso, 1998, pp. 75-85. “Economics as Separate and Inexact,” Economics and Philosophy 12 (1996): 207-20. “Rationalité, Bien-être, et Economie Normative,” tr. J.S. Lenfant, in Hubert Brochier, Roger Frydman, Bernard Gazier, & Jérôme Lallement, eds. L'Economie Normative. Paris: Economica, 1997, pp. 219-29. “Rationality and Knavery,” in Werner Leinfellner and Eckehart Köhler, eds. Game Theory, Experience, Rationality; Foundations of Social Sciences; Economics and Ethics: In Honor of John C. Harsanyi. Dordrecht: Kluwer, 1998, pp. 67-79.

160

Conversations on Human Action and Practical Rationality

“Liberty and its Value,” in J. F. Laslier, N. Gravel, A Trannoy & M. Fleurbaey, eds. Liberty. Routledge, 1997, pp. 45-48. “Economists' Responses to Anomalies: Full Cost Pricing versus Preference Reversals,” in John Davis, ed. New Economics and its History, NC: Duke University Press, 1988, pp. 255-72 (with Philippe Mongin). “Problems with Realism in Economics,” Economics and Philosophy 14 (1998): 185-213. “Ontology and Methodology in Economics,” Economics and Philosophy 15 (1999): 283-88. “Realist Philosophy and Methodology of Economics: What Is It?” Journal of Economic Methodology 7 (2000): 127-33. “Revealed Preference, Belief, and Game Theory.” Economics and Philosophy 16(2000): 99-115. “The Basic Problems of Economic Methodology,” Économies et Sociétés 7 (2000): 147-55. “In Defense of Two Policy Platitudes,” In Macroeconomics and the Real World, vol. 2, Keynesian Economics, Unemployment, and Policy. Ed. Roger Backhouse e Andrea Salanti. Oxford: Oxford University Press, 2000, pp. 269-76. “Does Economics Need Laws?”, spanish translation: “¿Necesita Leyes la Economía?” Argumentos de Razón Técnica 3 (2000): 115-37. “Critical Realism and Theories of Open Systems”, Spanish translation: “El Realism Crítico y las Teorías de Sistemas Abiertos.” Argumentos de Razón Técnica 3 (2000): 61-92. “A New Era for Economic Methodology,” Journal of Economic Methodology 8 (2001): 65-68. “Tendencies, Laws and the Composition of Economic Causes,” in Uskali Mäki, ed. The Economic World View. Studies in the Ontology of Economics. Cambridge: Cambridge University Press, 2001, pp. 293307. “Explanation and Diagnosis in Economics,” Revue Internationale De Philosophie 55 (2001): 311-26. “The Limits to Empirical Ethics,” in Christopher Murray, Joshua Salomon, Colin Mathers, and Alan Lopez, eds. Summary Measures of Population Health: Concepts, Ethics, Measurements and Applications. Geneva: World Health Organization, 2002, pp. 663-68. “Rational Belief and Social Interaction.” Behavior and Brain Sciences 26 (2003): 163-64. “Philosophical Foundations of Normative Economics,” in Melvin Ayogu and Don Ross, eds. Development Dilemmas: The Methods and

Appendix

161

Political Ethics of Growth Policy. New York: Routledge, 2005, pp. 4061. “Filosofia dell’economia,” in Filosofia delle scienze, ed. Nicla Vassallo. Turin: Giulio Einaudi, 2003, pp. 71-97. “Testing Game Theory.” Journal of Economic Methodology 12(2005): 211-23. “Valuing Health,” Philosophy and Public Affairs 34 (2006): 246-74. “The Philosophical Foundations of Mainstream Normative Economics,” The Philosophy of Economics: An Anthology (3rd ed.) (with Michael McPherson). Cambridge: Cambridge University Press, 2007 “Protecting Groups from Genetic Research.” Bioethics 22 (2008): 157-65. “Philosophy of the Social Sciences” in Diego Rios and Christoph Schmidt-Petri, eds. Philosophy of the Social Sciences: Five Questions.Roskilde, Denmark: Automatic Press, 2008, pp. 57-67. “Market Failure, Government Failure, and the Hard Problems of Cooperation.” Éthique et économique/Ethics and Economics, 6 (1), 2008. http://ethique-economique.net/ “Experimenting on Markets and the World,” Journal of Economic Methodology 15 (2009): 209-16. “Preference Satisfaction and Welfare Economics,” Economics and Philosophy 25 (2009): 1-25 (with Michael McPherson). “Equality of Autonomy,” Ethics 119 (2009): 742-56. “To Nudge or Not to Nudge,” Journal of Political Philosophy 18 (2010): 123-36 (with Brynn Welch) “The Philosophy of Economics,” in Fritz Allhoff, ed. Philosophies of the Sciences. West Sussex: John Wiley & Sons, 2010, pp. 324-55.

Some Encyclopedia Entries: “Economics and Ethics,” Routledge Encyclopedia of Philosophy. ed. Edward Craig. London: Routledge, 1998, vol. 3, pp. 205-11 (with Michael S. McPherson). “Economics, Philosophy of,” Routledge Encyclopedia of Philosophy. ed. Edward Craig. London: Routledge, 1998, vol. 3, pp. 211-22. “Philosophy of Economics,” Stanford Encyclopedia of Philosophy; on line http://plato.stanford.edu/ “Falsification,” New Palgrave Dictionary of Economics. “Theory Appraisal,” New Palgrave Dictionary of Economics (with Ellery Eells).

162

Conversations on Human Action and Practical Rationality

“Mindless or Mindful Economics: A Methodological Evaluation,” in Andrew Caplin and Andrew Schotter, eds. Handbook of Economic Methodology. Oxford University Press, 2008, pp. 125-51. “Laws, Causation, and Economic Methodology,” in Harold Kincaid e Don Ross, eds., The Oxford Handbook of Philosophy of Economics. Oxford: Oxford University Press, 2009, pp. 35-54. “Philosophy of Economics,”International Encyclopedia of the Social and Behavioral Sciences. Elsevier, 2001, pp. 4159-4165.

In Progress: “Why Not Just Ask? Preferences, “Empirical Ethics” and the Role of Ethical Reflection,” in From Fairness to Goodness, ed. Dan Wikler, World Health Organization. “Probabilistic Causation and Practical Causal Generalizations,” In Ellery Eells and James Fetzer, eds. The Place of Probability in Science. “Ethics and Economics.” Encyclopedia of Philosophy, 2nd ed. Routledge. “Social Scientific Naturalism and Experimentation in Economics,” In Uskali Mäki, ed. The Handbook of the Philosophy of Economics. North-Holland, Elsevier, 2006. “Morgenbesser's Schemata,” in Sandford Goldberg, ed. Sidney Morgenbesser's Contributions. “Egalitarianism Reconsidered,” Journal of Moral Philosophy (co-authored with Matt Waldren).

Joshua Knobe – Selected Bibliography2 Books (editor):

Ǥ ‘„‡ Ƭ Ǥ ‹…Š‘Ž•ǡ Experimental Philosophyǡ ‡™ ‘”ǣ šˆ‘”† ‹˜‡”•‹–›”‡••ǡʹͲͲͺǤ

Papers: B. Malle & J. Knobe, “The folk concept of intentionality”, Journal of Experimental Social Psychology, 33 (1997), pp. 101-121. B. Malle & J. Knobe, “The distinction between desire and intention: a folk-conceptual analysis”, in B. Malle, L. Moses & D. Baldwin (eds.),

2

Cf: http://pantheon.yale.edu/~jk762/publications.html

Appendix

163

Intentions and Intentionality: Foundations of Social Cognition, Cambridge, MA: MIT Press, 2001. “Intentional action and side effects in ordinary language”, Analysis, 63 (2003), pp. 190-193. “Intentional action in folk psychology: an experimental investigation”, Philosophical Psychology, 16 (2003), pp. 309-324. “Intention, intentional action and moral considerations”, Analysis, 64 (2004), pp. 181-187. “Folk Psychology and folk morality: response to critics”, Journal of Theoretical and Philosophical Psychology, 24 (2004). J. Knobe & G. Mendlow, “The good, the bad and the blameworthy: understanding the role of evaluative reasoning in folk psychology”, Journal of Theoretical and Philosophical Psychology, 24 (2004), pp. 252-258. “Theory of mind and moral cognition: exploring the connections”, Trends in Cognitive Sciences, 9 (2005), pp. 357-359. “The concept of intentional action: a case study in the uses of folk psychology”, Philosophical Studies, 130 (2006), pp. 203-231. A.Leslie, J. Knobe & A. Cohen, “Acting intentionally and the side-effect effect: ‘Theory of mind’ and moral judgment”, Psychological Science, 17 (2006), pp. 421-427. J. Knobe & A. Burra, “Intention and intentional action: a cross-cultural study”, Journal of Culture and Cognition, 6 (2006), pp. 113-132. J. Knobe & A. Burra, “Experimental philosophy and folk concepts: methodological considerations”, Journal of Culture and Cognition, 6 (2006), pp. 331-342. D. Pettit & J. Knobe, “The pervasive impact of moral judgment”, Mind & Language, 24:5 (2009), pp. 586-604. Y. Inbar, D. Pizarro, J. Knobe & P. Bloom, “Disgust sensitivity predicts intuitive disapproval of gays”, Emotion, 9:3 (2009), pp. 435-439. “Cognitive processes shaped by the impulse to blame”, Brooklyn Law Review, 71 (2005), pp. 929-937. J. Knobe & B. Fraser, “Causal Judgment and moral judgment: two experiments”, in W. Sinnott-Armstrong, Moral Psychology, Cambridge, MA: MIT Press, 2008, pp. 441-448. F. Cushman, J. Knobe & W. Sinnott-Armstrong, “Moral appraisals affect doing/allowing judgments”, Cognition, 108:1 (2008), pp. 281-289. C. Hitchcock & J. Knobe, “Cause and Norm”, Journal of Philosophy, 11 (2009), pp. 587-612. “Folk judgments of causation”, Studies in the History and Philosophy of Science, 40:2 (2009), pp. 238-242.

164

Conversations on Human Action and Practical Rationality

S. Nichols & J. Knobe, “Moral responsibility and determinism: the cognitive science of folk intuitions”, Nous, 41 (2007), pp. 663-685. J. Knobe & S. Nichols, “Free will and the bounds of the self”, in R. Kane (ed.), The Oxford Handbook of Free Will, 2ª ed., New York: Oxford University Press (forthcoming). J. Knobe & J. Doris, “Strawsonian variations: folk morality and the search for a unified theory”, in J. Doris et al., The Handbook of Moral Psychology, Oxford: Oxford University Press (forthcoming). “Reason explanation in folk psychology”, Midwest Studies in Philosophy, 31 (2007), pp. 90-107. “Folk psychology: science and morals”, in D. Hutto & M. Ratcliffe (eds.), Folk Psychology Reassessed, Kluwer/Springer Press, 2007, pp. 157174. J. Knobe & J. Prinz, “Intuitions about consciousness: experimental studies”, Phenomenology and Cognitive Science, 7:1 (2008), pp. 6783. J. Knobe & E. Roedder, “The ordinary concept of valuing”, Philosophical Issues, 19:1 (2009), pp. 131-147. J. Phillips, L. Misenheimer & J. Knobe, “The ordinary concept of happiness (and others like it), Emotion Review (no prelo). “Experimental philosophy and philosophical significance”, Philosophical Explorations, 10 (2007), pp. 119-122. “Experimental Philosophy”, Philosophy Compass, 2:1 (2007), pp. 81-92. B. Malle & J. Knobe, “Which behaviors do people explain? A basic actorobserver asymmetry”, Journal of Personality and Social Psychology, 72 (1997), pp. 288-304. B. Malle, J. Knobe, M. O’Laughlin, G. Pearce & S. Nelson, “Conceptual structure and social functions of behavior explanations”, Journal of Personality and Social Psychology, 79 (2000), pp. 309-326. J. Knobe & B. Malle, “Self and other in the explanation of behavior”, Psychologica Belgica, 42 (2002), pp. 113-130. B. Malle, J. Knobe & S. Nelson, “Actor-observer asymmetries in explanations of behavior: new answers to an old question”, Journal of Personality and Social Psychology, 93 (2007), pp. 491-514. J. Knobe & B. Leiter, “The case for Nietzschean moral psychology”, in B. Leiter & N. Sinhababu (eds.), Nietzsche and Morality, Oxford: Oxford University Press, 2007, pp. 83-109. “Answers to Five Questions”, in J. Aguilar & A. Buckareff (eds.) Philosophy of Action: 5 Questions, London: Automatic Press, 2009.

Appendix

165

J. Knobe & S. D. Kelly, “Can one act for a reason without acting intentionally?”, in C. Sandis (ed.), New Essays on the Explanation of Action, Basingstoke: Palgrave Macmillan, 2009, pp. 169-183.