UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Rational agency Campbell, Peter G. 1988

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1988_A8 C35.pdf [ 4.55MB ]
Metadata
JSON: 831-1.0097898.json
JSON-LD: 831-1.0097898-ld.json
RDF/XML (Pretty): 831-1.0097898-rdf.xml
RDF/JSON: 831-1.0097898-rdf.json
Turtle: 831-1.0097898-turtle.txt
N-Triples: 831-1.0097898-rdf-ntriples.txt
Original Record: 831-1.0097898-source.json
Full Text
831-1.0097898-fulltext.txt
Citation
831-1.0097898.ris

Full Text

RATIONAL AGENCY By PETER G. CAMPBELL B.A., The University of B r i t i s h Columbia, 1983 A THESIS SUBMITED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i n THE FACULTY OF GRADUATE STUDIES (Department of Philosophy) We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA A p r i l , 1988 © Peter G. Campbell, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of P h i l o s o p h y The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date A p r i l 15, 1988 DE-6(3/81) ABSTRACT I t i s claimed that action discourse provides us with a c r i t e r i o n of adequacy for a theory of action; that with action discourse we have a family of concepts which a theory of action must accommodate. A f t e r an exegesis of Davidson's essay "Agency", i t i s argued t h a t h i s semantics of a c t i o n i s incompatible with our concepts of motivation and r e s p o n s i b i l i t y for action and of a t t r i b u t i o n s of action and agency, and must, t h e r e f o r e , be r e j e c t e d . A theory of r a t i o n a l agency i s presented within which are to be found accounts of intention, coming to i n t e n d , i n t e n t i o n a l a c t i o n , and an a l t e r n a t i v e semantics of action which connects the action e s s e n t i a l l y to agency. The theory of r a t i o n a l agency i s then used to i l l u m i n a t e the concepts of t r y i n g , compulsion, autonomy and involuntariness, mistake, accident, and the so-called active-passive d i s t i n c t i o n . i i TABLE OF CONTENTS ABSTRACT i i TABLE OF CONTENTS i i i LIST OF FIGURES . . . v ACKNOWLEDGEMENT v i INTRODUCTION . . . 1 CHAPTER I THE SEMANTICS OF ACTION 4 A. DAVIDSON'S "AGENCY" 5 1. The C r i t e r i o n of Action and the Expression of Agency 5 2. Agency. The Role of Causation i n Action, Part I . 10 3. The Role of Causation i n Action, Part II 17 4. Summary 26 B. DAVIDSON'S SEMANTICS REJECTED 26 C. SOME IMPLICATIONS FOR A SEMANTICS OF ACTION 39 CHAPTER II RATIONAL AGENCY 43 A. PROPERTIES OF INTENTION 44 1. Davidson on Intention 44 2. Bratman on Intention 49 B. A THEORY OF RATIONAL AGENCY 52 C. INTENTIONAL ACTIONS 71 i i i CHAPTER III APPLICATIONS 77 A. AGAINST THE SIMPLE VIEW 77 B. TRYING 84 C. NON-STANDARD ACTIONS AND EXCUSES 90 1. Mistake 90 2. Accident 93 3. Compulsion 95 4. Involuntariness 96 D. THE ACTIVE - PASSIVE DISTINCTION 98 1. Intentional Letting Die 100 2 Negligent Letting Die 102 BIBLIOGRAPHY 106 i i v LIST OF FIGURES Fig. 1 52 Fig. 2 53 F i g . 3 58 F i g . 4 58 F i g . 5 60 F i g . 6 60 F i g . 7. 61 F i g . 8 62 F i g . 9 64 F i g . 10 64 F i g . 11 65 F i g . 12. A MODEL OF RATIONAL AGENCY: RAAG 66 Fig. 13. RAAG PLAYING THE VIDEO GAME 82 Fig. 14. RAAG MAKES A MISTAKE 91 Fig . 15. RAAG ACCIDENTALLY SHOOTS HIS NEIGHBOUR'S DONKEY 95 V ACKNOWLEDGEMENT I would l i k e to thank Professor S.C. Coval for two years of work i n action theory. This thesis i s the product of our close col l a b o r a t i o n which began with Professor Coval's action theory seminar i n 1986, and which continues with further research on the problems of action, and a p p l i c a t i o n s of our theory to a r t i f i c i a l i n t e l l i g e n c e . We are working with Professor J.C. Smith and Doug Arnold of the U.B.C. Law School to construct a computer model of a r a t i o n a l agent. Chapter II of t h i s thesis w i l l form part of a paper co-authored by Professor Coval and I i n which we discuss our work on the computer model. I would l i k e to express my appreciation both to the U.B.C. Philosophy Department f o r f i n a n c i a l support i n the form of teaching a s s i s t a n t s h i p s , and to the U n i v e r s i t y of B r i t i s h Columbia for a Graduate Summer Fellowship, 1987. v i INTRODUCTION This essay i s concerned with some of the main problems i n the philosophy of action. These include the semantic issues of the nature and i d e n t i t y of actions and action descriptions, a c t i o n e x p l a n a t i o n s , both c a u s a l and t e l e o l o g i c a l , and assessment of agents. I take i t as a given that there i s mental causation. With mental causation we accept the notion of mental representation with cognitive content and causal e f f i c a c y . I argue that there i s a set of properties which the agency state must have i n order for there to be action. However, even though the nature of mental states and events i s a concern of the philosophy of mind, I claim that nothing I have to say about r a t i o n a l agency decides, i n any way, any of the t r a d i t i o n a l problems i n the philosophy of mind. That i s , the theory of action which I present i n Chapter II does not beg any questions for or against d u a l i s t , m a t e r i a l i s t or f u n c t i o n a l i s t accounts of mind. I t i s rather that with action, we accept a set of concepts which any viable theory of mind must accommodate. Our p r e - t h e o r e t i c a l action discourse presents us with a c r i t e r i o n of adequacy f o r a theory of action. With action discourse, we have a set of concepts and d i s t i n c t i o n s which we employ i n complex ways. We are at home with causal and 1 2 t e l e o l o g i c a l explanations of actions. We assess agents i n order to determine degrees of r e s p o n s i b i l i t y , and we wield excuses and defenses. I t i s a deep-seated fact of our conceptual scheme that purposeful behaviour i s amenable to t h i s kind of scrutiny. Therefore, a theory of action ought to accommodate and c l a r i f y these concepts which are the given of ac t i o n theory. And without good reason to do otherwise, we ought to minimize the extent to which the shape of a theory d i s r u p t s our pre-t h e o r e t i c a l concepts. What counts as an action must pay o f f to our concern for c l a s s i f y i n g events and agents i n r e l a t i o n to our concern with r e s p o n s i b i l i t y and explanation of action. This i s not to say that we ought to l e t pragmatic considerations such as our concern with h o l d i n g persons r e s p o n s i b l e f o r events determine our scheme of c l a s s i f y i n g events as i n t e n t i o n a l actions or otherwise. That i s , we ought not allow the inference from the claim that you are responsible for some event A to the claim that you intended A, nor to the claim that you desired A, and therefore not to the claim that you did A i n t e n t i o n a l l y . Nevertheless, any theory of action whose semantics or other t h e o r e t i c a l aspects i s inhospitable to action discourse v i o l a t e s the c r i t e r i o n of adequacy. Unless there i s some good reason to the contrary, such a theory ought, therefore, to be rejected. In Chapter I, I argue that Donald Davidson's semantics of a c t i o n i s a c t u a l l y incompatible with our family of action concepts. Afte r an exegesis of his theory, I argue that i t cannot accommodate the notions of r e s p o n s i b i l i t y and motivation, 3 nor the c l a s s i f i c a t i o n s mentioned above, and therefore cannot accommodate agency. In Chapter II, I present a theory of r a t i o n a l agency within which are to be found accounts of i n t e n d i n g and a c t i n g i n t e n t i o n a l l y , and an a l t e r n a t i v e semantics of action which connects the action e s s e n t i a l l y to agency. The theory requires t h a t we a c c e p t such " m e t a p h y s i c a l l y s u s p e c t " items as intensional objects, but t h i s , I argue, we already accept with action. In Chapter I I I , I apply the theory of r a t i o n a l agency, f i r s t to an objection of Michael Bratman's, and then to some of the long-standing problems i n a c t i o n theory, namely trying, compulsion, autonomy and involuntariness, mistake, accident, and to the so-called active-passive d i s t i n c t i o n . CHAPTER I THE SEMANTICS OF ACTION 1 begin with an exegesis of Davidson's "Agency" 1, i n which he develops h i s semantics of action and analyses the notion of agency. A close look at "Agency" w i l l reveal some of the main questions i n a c t i o n theory and the problems which Davidson reveals i n the p o s i t i o n s of others. Furthermore, no other action t h e o r i s t i s as powerful on, for example, the issue of the nature and i d e n t i t y of actions. However, Davidson i s the p r i n c i p a l proponent of the view of action which I oppose. For Davidson, a c t i o n s are events. Events are metaphysically respectable items which, l i k e objects, may be described i n various ways.2 Therefore, actions are extensionally conceived events. I f so, I argue, the semantics of action i s cut o f f from in t e n t i o n with the consequence that actions cannot support the normal i n f e r e n c e s between what i s done, and f o r example, motivation and r e s p o n s i b i l i t y . Following the exegesis, I begin to work my way out of the grip of the Davidsonian semantics with several arguments against his extensionality. In Chapter II, I 1 Donald Davidson, "Agency," i n Essays on Actions and Events, (New York, N.Y.: Oxford University Press, 1980), 43. 2 Ibid., 105-203. 4 5 present an al t e r n a t i v e semantics of action within a theory of r a t i o n a l agency. A. DAVIDSON'S "AGENCY" "Agency" i s concerned with the semantics of a c t i o n . Davidson's stated aim i s to f i n d an analysis of agency. This he takes as the task of giving an analysis of the r e l a t i o n that holds between an agent and h i s actions, and an analysis of the relata which stand i n t h i s r e l a t i o n . There are, roughly, three sections to Davidson's essay: the f i r s t addresses the nature and i d e n t i t y of the second term of the r e l a t i o n - the action. The second and t h i r d sections deal mainly with the r e l a t i o n i t s e l f . By h i s own admission 3 Davidson does not give us an analysis of agency, but does show what i t cannot be by revealing some mistakes of other action t h e o r i s t s . 1. The C r i t e r i o n of Action and the Expression of Agency The f i r s t , and for our purposes most important section of "Agency"4 addresses the question "What i s an action?" Davidson assumes that "there i s a f a i r l y d e f i n i t e subclass of events which are actions." 5 This gives us the genus of actions: they are events. The problem i s to di s t i n g u i s h the action subclass. Davidson's f i r s t candidate f o r the d i s t i n g u i s h i n g mark of actions i s that of being i n t e n t i o n a l . Since there are actions 3 Ibid x i i i . 4 Ibid r 43-47. 5 Ibid 44. 6 which cannot be anything but i n t e n t i o n a l , e.g., asserting, c h e a t i n g and l y i n g , 6 i n t e n t i o n i m p l i e s agency. But, the converse does not hold: that what you did was an action does not imply that you did i t i n t e n t i o n a l l y . I f . . . I i n t e n t i o n a l l y s p i l l the contents of my cup, mistakenly thinking i t i s tea when i t i s coffee, then s p i l l i n g the coffee i s something I do, i t i s an action of mine, though I do not do i t i n t e n t i o n a l l y . 7 Therefore, since agency does not imply intention, an event's being i n t e n t i o n a l i s not s u f f i c i e n t as a mark of action. Davidson's treatment of mistakes i s t e l l i n g . Mistakes cannot be done i n t e n t i o n a l l y . Making a mistake, he says, i s f a i l i n g to do what one intends, i s "doing something with the i n t e n t i o n of achieving a r e s u l t that i s not forthcoming." 8 Mistakes are, therefore, unintentional. Yet for Davidson they are a c t i o n s . 9 For example, a misreading i s a reading, a misinterpreting i s an i n t e r p r e t i n g . So, making a mistake i s doing something else intentionally:1 0 an u n i n t e n t i o n a l m i s r e a d i n g i s an i n t e n t i o n a l r e a d i n g ; an u n i n t e n t i o n a l misinterpreting is an i n t e n t i o n a l i n t e r p r e t i n g . And, mistakenly 6 And not acting I 7 Davidson, 45. 8 I b i d . , 46. 9 This w i l l be a point of dispute between us. 1 0 As w i l l be apparent, Davidson does not r e a l l y mean "something else". 7 s p i l l i n g the coffee i s i n t e n t i o n a l l y s p i l l i n g the contents of the cup. What i s common to a l l of the above cases of agency i s that the agent i s doing something i n t e n t i o n a l l y . What cancels agency, f o r Davidson, i s when the event i n question was caused externally: "I am the agent i f I s p i l l the coffee meaning to s p i l l the tea, but not i f you j i g g l e my hand." 1 1 In the former case I am doing something i n t e n t i o n a l l y , i n the l a t t e r I am not. (Of course, I am doing several things i n t e n t i o n a l l y i n the l a t t e r case, including holding the cup so as not to s p i l l i t s contents. What Davidson means i s that the bodily event which causes the event of the contents s p i l l i n g cannot be described so as to make the s p i l l i n g i n t e n t i o n a l for me. ) So, my s p i l l i n g the contents of the cup was i n t e n t i o n a l . This very same act can be redescrUbed as my s p i l l i n g the coffee. On the b a s i s of t h i s a n a l y s i s , Davidson o f f e r s the following causal c r i t e r i o n of agency, and so of action: C r i t e r i o n I A man i s the agent of an act i f what he does can be described under an aspect that makes i n i n t e n t i o n a l . 1 2 That i s , doing something which can be described under an aspect that makes i t int e n t i o n a l i s s u f f i c i e n t f or agency or action. 1 1 Davidson, 46. 1 2 Ibid.. 8 What makes t h i s possible i s that a t t r i b u t i o n s of intention are inte n s i o n a l : co-designating singular terms which r e f e r to what one does intentionally cannot be substituted, salva veritate. I f t h i s were not so, then from 1) Hamlet i n t e n t i o n a l l y k i l l s the man behind the arras, and 2) The k i l l i n g of the man i s i d e n t i c a l to the k i l l i n g of Polonius. we could i n f e r 3) Hamlet i n t e n t i o n a l l y k i l l s Polonius. which i s f a l s e . Whether or not a doing i s an in t e n t i o n a l doing depends on how i t i s described. I s p i l l the contents of the cup. This action can be redescribed as my s p i l l i n g the coffee. Under the f i r s t description, my action was i n t e n t i o n a l . Under the second description, the very same act was unintentional. This doesn't mean, of course, that the event which i s the action both had and la c k e d a c e r t a i n p r o p e r t y - th a t of being i n t e n t i o n a l . I t means that the event, the agent, and a certa i n description have a r e l a t i o n that does not obtain between the same event, the agent, and a different d e s c r i p t i o n . 1 3 So, while there i s a class of events which are actions, there i s neither a c l a s s of i n t e n t i o n a l actions, nor a c l a s s of unintentional actions: what I did w i l l be in t e n t i o n a l under one description, and not under another. The c r i t e r i o n of agency and a c t i o n , put i n terms of sentences and descriptions of actions i s : 1 3 I b i d . , 195. 9 C r i t e r i o n I I . A person i s the agent of an event i f and only i f there i s a description of what he did that makes true a sentence that says he did i t i n t e n t i o n a l l y . 1 4 That i s , i t i s s u f f i c i e n t and necessary for agency and action that there be a single description of what a person did that makes true a sentence that says he did i t i n t e n t i o n a l l y . W h i l e a t t r i b u t i o n s of i n t e n t i o n are i n t e n s i o n a l , a t t r i b u t i o n s of agency or action are purely extensional: "the expression of agency i s i t s e l f purely e x t e n s i o n a l . " 1 5 Singular terms which r e f e r to what an agent did are substitutable, salva veritate. As we have seen from Davidson's examples, being an action i s a t r a i t which p a r t i c u l a r events have independently of how they are described: my s p i l l i n g the contents of the cup i s i d e n t i c a l to my s p i l l i n g the coffee; Hamlet's k i l l i n g the man behind the arras i s i d e n t i c a l to h i s k i l l i n g Polonius; an o f f i c e r ' s sinking of that ship, which he mistakenly thinks i s the T i r p i t z , i s i d e n t i c a l to h i s sinking the Bismark. "The r e l a t i o n that holds between a person and an event, when the event i s an action performed by the person, h o l d s r e g a r d l e s s of how the terms are described. 1 , 1 6 1 4 Ibid., 1 5 Ibid., 1 6 Ibid.. 46. 47. 10 Therefore, we can say that a person does as agent whatever he does i n t e n t i o n a l l y under some description. This i s Davidson's semantics of a c t i o n . Actions are extensional events which s a t i s f y his c r i t e r i o n of agency and action. 2. Agency. The Role of Causation i n Action, Part I The second main section of "Agency" 1 7 i s a search for an analysis of agency that does not appeal to that mysterious and, as yet, unanalyzed notion of intention. What Davidson wants i s an analysis of agency, of the r e l a t i o n that holds between an agent and his actions. One way to arr i v e at an analysis of a r e l a t i o n i s through an analysis of the r e l a t a - of the things which can stand i n the r e l a t i o n . We know now that the second term, for Davidson, i s an extensional event. Which events are ju s t those which s a t i s f y the c r i t e r i o n . The c r i t e r i o n of agency and action appeals to the notion of intention. So, i f we had an analysis of intention (a good place to begin an analysis of agency) we could ask for an analysis of the r e l a t i o n that holds between i n t e n t i o n s and t h e i r corresponding a c t i o n s . But Davidson had no analysis of intention, discounting the "old" one from "Actions, Reasons and Causes" 1 8, where the Intention with which an action is done i s analyzed i n terms of the desires and b e l i e f s which r a t i o n a l i z e the action. Therefore, the question 1 7 I b i d . , 47-55. 1 8 I b i d . , 3-19. 11 Davidson asks i s "what can we say, without an ana l y s i s of intentions, about the r e l a t i o n between an agent and his action t h a t w i l l e l u c i d a t e doing, acting, and so agency?" His suggestion i s that we pick up the notion of the expression of agency. With the extensionality of the expression of agency we have the idea that i f an event i s an action, then the r e l a t i o n that holds between the agent and the event holds regardless of how the terms are described. 1 9 Since intentions are intensional and agency not, i t looks as though the concept of agency i s simpler than that of intention. And, the r e l a t i o n i t s e l f looks, to Davidson, a l o t l i k e ordinary event c a u s a l i t y . This being the case, an examination of the role of causation i n action ought to c l a r i f y the concept of agency. a. There are Primitive Actions The r e l a t i o n of event c a u s a l i t y i s the r e l a t i o n that holds between actions and t h e i r upshots, and according to Davidson, i t i s t h i s which allows the a t t r i b u t i o n of the upshots to the agent: i t i s Jbecause an action causes an upshot that the upshot i s a t t r i b u t a b l e to the agent. So some events which we attribute to an agent are events re l a t e d causally to some other event of which he i s the agent. However, not every action attributed to an agent i s so attributed because i t i s caused by another act of the agent. Some actions must be primitive actions - that i s , 1 9 Ibid., 47. 12 not caused by some causally p r i o r act of the agent. Otherwise, agency could never get started. Therefore, event cau s a l i t y cannot be the r e l a t i o n which p r i m i t i v e actions bear to the agent, and so event c a u s a l i t y cannot support the f i r s t a t t r i b u t i o n of agency. b. What are Primitive Actions? Davidson i d e n t i f i e s primitive actions as bodily movements. ( o r , perhaps b e t t e r , bodily movings, as opposed to mere movements of the body). He considers two main objections, the f i r s t of which argues t h a t b o d i l y movements are n e i t h e r p r i m i t i v e nor actions, the second of which argues that primitive actions involve more than body movement. It may be argued that i n order to move my finger I must f i r s t do something that causes my f i n g e r to move, namely, contract the appropriate muscles, and, doing t h i s requires that I f i r s t make c e r t a i n brain events occur. Since neither of these causally p r i o r events are bodily movements, bodily movements are not p r i m i t i v e actions. A c c e p t i n g t h a t movements of the body are caused by contractions of ce r t a i n muscles, which i n turn are caused by c e r t a i n brain events, Davidson's response i s : "Doing something that causes my finger to move...is moving my f i n g e r " . 2 0 The t r i c k i s turned, here, by d i s t i n g u i s h i n g the event of the 2 0 Ibid., 49-50. 13 movement of the finger from the action of moving the finger: moving the finger - that primitive action/bodily movement is doing whatever i t takes to cause the event of the fi n g e r movement. Chisholm 2 1 has two objections, one addressing Davidson's p a r t i c u l a r point here, the other addressing extensionality, and both based on the claim that knowing that one i s doing A i s a necessary condition for one's doing A. I f so, making cerebral events and muscle contractions occur, when these are unknown, are not actions. And, since consequences of actions are often unknown, there are e f f e c t s of actions which are not actions (contra Davidson). 2 2 Davidson has a counter example 2 3 to the necessity claim: Suppose that a man intends to make 10 carbon copies as he writes but does not know or believe with any confidence that he i s s u c c e e d i n g . I f he does, Davidson c l a i m s , he does so i n t e n t i o n a l l y , and without knowing that he i s . Therefore, knowing that one i s A-ing i s not a necessary condition of one's A - i n g . 2 4 2 1 Roderick Chisholm, "Freedom and Action, " i n Freedom and Determinism, ed. K. Lehrer (New York: Random House, 1966), 28-44. 2 2 My arguments against Davidson's e x t e n s i o n a l i t y w i l l support Chisholm's claim. 2 3 Davidson, 50, 92. 2 4 For a de t a i l e d discussion of t h i s matter, see Chapter III, Trying. 14 Knowing that one i s A-ing i s not a necessary condition for A-ing, for Davidson. Nevertheless, that h i s c r i t e r i o n of action must be s a t i s f i e d e n t a i l s that one must know whatever one i s doing under some description or other. Suppose 0 moves his finger i n t e n t i o n a l l y . I f so, 0 intends to do with h i s body whatever i s needed i n order to produce the event of the finger movement, where the "whatever needed" i s the cerebral events and muscle contractions. Therefore, i n intending to do whatever i s needed, 0 intends and knows about the cerebral and mental events under the description "whatever i t i s I do when I move my finger". The second main o b j e c t i o n to Davidson's c l a i m t h a t p r i m i t i v e actions are bodily movements i s the claim that some p r i m i t i v e actions include more than b o d i l y movements. For example, tying one's shoes includes the finger movement and the movement of the laces. The question i s , can these be separated into two events, the f i r s t of which i s the pri m i t i v e action? Since events are distinguished by t h e i r space-time locations, the separation here i s not the problem. The problem i s whether there i s a description of what i s done that r e f e r s to the event of the body moving and not to the shoelaces. Davidson's response i s the same as he made to the l a s t objection. The descr i p t i o n "I move my body i n just the way required to t i e my shoelaces" s a t i s f i e s the c r i t e r i o n , i s known to the agent, and ref e r s to the bodily event. It i s a description of an event with the r i g h t e f f e c t s , but no less a description of a bodily event which i s a primitive action. 15 c. Agent Causality Returning to the question of the r e l a t i o n between an agent and h i s action, we can now take, as the second term, primitive actions. As we have seen, Davidson argues that while ordinary causation can explain how agency spreads from primitive actions to actions described i n further ways, i t cannot explicate the r e l a t i o n between the agent and h i s p r i m i t i v e actions. The question now i s whether there i s a sui generis agent causality25 which r e l a t e s agents and t h e i r a c t i o n s , and which w i l l , therefore, e x p l i c a t e the r e l a t i o n between an agent and h i s pri m i t i v e actions. Another p o s s i b i l i t y which Davidson doesn't consider i s that there i s a term which, being unique as a causal term (desire), helps explicate that r e l a t i o n . Davidson's treatment of agent c a u s a l i t y i s important for my thesis since i t reveals the ground upon which our b a t t l e w i l l be fought. Agent c a u s a l i t y i s supposed to be that which relates agents to events and by v i r t u e of which these events are actions. Davidson poses the following dilemma for friends of agent c a u s a l i t y . Either the causing of a primitive action i s an event discrete from the primitive action, or i t i s not. On the f i r s t horn, agent c a u s a l i t y introduces an event separate from, and p r i o r t o , the p r i m i t i v e a c t i o n . There are two p o s s i b i l i t i e s . F i r s t , t h i s p r i o r event i s an action. If so, then contrary to the assumption, the primitive action i s not 2 5 From Thalberg. 16 p r i m i t i v e . And, i f agent c a u s a l i t y i s to serve i t s function, then t h i s p r i o r event which i s an action must be rel a t e d to another event, separate from and p r i o r to i t . And we have a regress. Therefore, agent c a u s a l i t y could never do i t s work, and agents could never act. The second p o s s i b i l i t y for the f i r s t horn of the dilemma i s that the separate and p r i o r event i s not an action. But i f so, i t i s an event which, because of agent causality, i s an agent-causing which i s not a doing. Therefore, agent c a u s a l i t y i s not doing the job i t was introduced for, since agent-causings are not a c t i o n s . Davidson concludes that agent c a u s a l i t y i s therefore superfluous. It i s important to notice that Davidson's own p o s i t i o n i s one i n which primitive actions are events which are causally r e l a t e d to separate and p r i o r b r a i n events which are not actions. Since the causally p r i o r events are not actions, and given the nature of brain events, for Davidson, he can s t i c k with good o l d ordinary event causality, as the r e l a t i o n which they bear to pr i m i t i v e actions. On the second horn of the dilemma, agent c a u s a l i t y does not introduce an event separate from and p r i o r to the primitive action. If so, the agent i s not related by an event of agent c a u s a l i t y to his primitive actions. Thus the notion of agent c a u s a l i t y would play no role i n agency, and saying that one i s rel a t e d by agent c a u s a l i t y to his actions would be saying no 17 more, and be nor more explanatory, than saying he acted. Again, agent c a u s a l i t y i s superfluous. The problem Davidson must face i s that i f the agent i s not r e l a t e d by some notion of agent c a u s a l i t y to his actions, then we are without grounds to support the claim that they are a c t i o n s . I f the agent i s merely c a u s a l l y r e l a t e d to h i s actions, and since we as agents are related causally to any number of events which are not actions, then there would seem to be nothing to support the d i s t i n c t i o n between actions and events which are not actions. 3. The Role of Causation i n Action, Part II a. The Accordion E f f e c t So far, Davidson has argued that causation i s central to the concept of agency, t h a t i t i s g a r d e n - v a r i e t y event causality, and i t concerns the effect of actions and not t h e i r causes. The feature of action discourse c a l l e d the accordion e f f e c t 2 6 demonstrates t h i s role of causation i n action. Suppose the following: 1. 0 moves his finger i n t e n t i o n a l l y . 2. 0 causes the switch to be f l i p p e d . 3. 0 causes the l i g h t to come on. 4. 0 causes the room to be illuminated. 5. 0 causes the prowler to be alerted. From Feinberg. 18 Since O's moving of his finger i s i n t e n t i o n a l , i t q u a l i f i e s as an a c t i o n by Davidson's c r i t e r i o n . But, whether or not 0 intends any of the consequences 2-5, Davidson says that 1-5 e n t a i l : 6. 0 f l i p p e d the switch. 7. 0 turned on the l i g h t . 8. 0 illuminated the room. 9. 0 alerted the prowler. A l l of 6 - 9, as well as 1, are action descriptions, and are at t r i b u t a b l e to 0 because of the accordion e f f e c t , which reveals that causation transfers agency from actions to the e f f e c t s of actions, whether or not the agent intends those e f f e c t s : ...once he has done one thing (move a f i n g e r ) , each consequence presents us with a deed; an agent causes what h i s actions cause. 2 7 The a c c o r d i o n e f f e c t demonstrates that causation i s what t r a n s p o r t s agency, though not i n t e n t i o n , and i n t e n t i o n i s i r r e l e v a n t as to how the accordion i s played. 2 7 Davidson, 53. Except for the d i f f i c u l t i e s which Davidson footnotes, he would l i k e to be able to say that an agent does what h i s a c t i o n s cause. I t i s t h i s s o r t of feature of Davidson's theory that I object to when I object to h i s extensionality. 19 The accordion e f f e c t works because of causation between e v e n t s . But i t i s a f e a t u r e of a c t i o n discourse o n l y . Therefore, the r e l a t i o n i t r e v e a l s i s not or d i n a r y event c a u s a l i t y . Causation i s a t r a n s i t i v e r e l a t i o n between events: i f event A causes event B, and B causes event C, then A causes C. But i t i s not t h i s property of causation that powers the accordion: ( i ) Suppose the bat struck the b a l l which broke the window. I f so, we can say that the movement of the bat broke the window, but not that the bat broke the window. ( i i ) Suppose the cue struck the cue b a l l which moved the 8-ball. We can say that the movement of the cue caused the movement of the 8-ball, but not that the cue moved the 8-ball. ( i i i ) Suppose that 0 f l i p p e d the switch which caused the l i g h t to go on. We can say Jboth that the movement of hi s finger caused the l i g h t to go on and, that 0 turned on the l i g h t . The r e l a t i o n between events i n ( i ) , ( i i ) and ( i i i ) i s event c a u s a l i t y . The r e l a t i o n between 0 and h i s action of turning on the l i g h t i s not. Just what the r e l a t i o n i s between an agent and h i s non-primitive actions i s addressed i n the next section on "Agency". 20 b. Doing With and Bo±ng By The t h i r d main section of "Agency" 2 8 i s concerned with the r e l a t i o n between an agent and h i s non-primitive actions, which i s the r e l a t i o n of "doing with" and "doing by" i n the following: A. Brutus k i l l e d Caesar by stabbing him. (2 non-primitive actions) B. Jones k i l l e d Smith by s t a r t l i n g Smith by opening the door by turning the key. ( A l l non-primitive actions) C. With one movement of h i s hand, Jones did a l l of the things i n B. (A primitive action and several non-prim i t i v e actions) Davidson has argued t h a t p r i m i t i v e a c t i o n s cause t h e i r consequences, and that t h i s i s ordinary event c a u s a l i t y . If some event A i s a primitive action (that i s , an event caused i n the r i g h t way) and i f events are rela t e d causally to t h e i r e f f e c t s , then primitive actions are related causally, to t h e i r e f f e c t s . S i m i l a r l y , we can see that non-primitive actions cause t h e i r consequences where again t h i s i s event ca u s a l i t y : the stabbing of Caesar by Brutus caused the death of Caesar. Davidson, 55-61. However, we as yet have no analysis of the r e l a t i o n that holds between and among primitive actions and non-primitive actions. F i r s t , we want to know how, e.g., the primitive action of moving one's hands i s related to the non-primitive action of tying one's shoes. Secondly, we want to know how non-primitive actions are r e l a t e d to c e r t a i n c o r r e l a t i v e events: e.g., how i s the non-primitive action of tying one's shoes related to the event of one's laces being tied? Or, how i s the non-primitive action of Brutus k i l l i n g Caesar related to the death of Caesar? F i n a l l y , we need to know what the r e l a t i o n i s between non-p r i m i t i v e actions: e.g., what i s the r e l a t i o n between the two n o n - p r i m i t i v e actions of Brutus stabbing Caesar and Brutus k i l l i n g Caesar? The obvious candidate i s causation. But Davidson shows why t h i s cannot be. c. Feinberg, Austin and Danto. On F e i n b e r g ' s v e r s i o n of the a c c o r d i o n e f f e c t , the squeezing down and puffing out of an accordion i s done over events, so that an action can be stretched out to include an e f f e c t . But, as Davidson notes, i f the squeezing and the stretching changes the time span of the event, then i t cannot be the same event, and so cannot be the same action. Therefore, t h i s i s not a c o r r e c t a n a l y s i s of the accordion e f f e c t . Feinberg also has a d i s t i n c t i o n between simple and causally complex acts. A simple act i s one which requires us to do nothing else. A causally complex act i s one which requires us to do something f i r s t as a means. For example, i n order to open the door I must first do something else which w i l l cause the 22 door to open. Therefore, on Feinberg's view, there are causally connected sequences of acts. J.L. Austin and Danto have s i m i l a r views. For Austin ... a single term descriptive of what he did may be made to cover either a smaller or a larger s t r e t c h of events, those excluded by the narrower description being then c a l l e d "consequences" or " r e s u l t s " or "e f f e c t s " or the l i k e of h i s a c t . 2 9 Therefore, actions can cause actions, for Austin. Danto makes the d i s t i n c t i o n between basic and non-jbasic actions, where the d i s t i n c t i o n i s marked causally: A basic action, such as moving a hand, causes the non-basic action of moving the stone. 3 0 Davidson argues that these views are mistaken; that actions are not c a u s a l l y r e l a t e d . While i t i s true that primitive actions cause t h e i r consequences, so that when I, for example, close the door, some action of mine causes the event of the door closing , i t i s f a l s e that some action of mine causes my action of c l o s i n g the door. I f not f a l s e , then i n order to close the door - that action - I would have to do two things. But i t seems c l e a r l y f a l s e that when I close the door by moving my hand I do two things. And, i f there are not two things, then they cannot be related, causally or otherwise. 2 9 J.L. Austin, "A Plea for Excuses," Philosophical Papers, ed. J.O. Urmson and G.J. Warnock (Oxford: Oxford Univ. Press, 1961), 145. 3 0 A . Danto, "Basic Actions," American Phil. Quarterly, 2 (1965): 141-8. 23 The heart of Davidson's p o s i t i o n here i s that a primitive action i s not numerically d i s t i n c t from i t s correponding non-pr i m i t i v e action. If so, then there are not two events to be causally re l a t e d . Otherwise, i f there were two actions causally related, we would have to say that, for example, the queen's moving her hand caused the queen's k i l l i n g the king, and that the queen caused the queen to k i l l the king. Suppose t h i s were the case. We s t i l l want an analysis of the action of the k i l l i n g . Here, the k i l l i n g i s the queen's action of moving her hand i n j u s t that way i n those circumstances. But, t h i s was sufficient i n those circumstances, to k i l l the king. If so, there was no room for any further action of the queen: Is i t not absurd to suppose that, a f t e r the queen has moved her hand i n such a way as to cause the king' s death, any deed remains for her to do or to complete? She has done her work; i t only remains for the poison to do i t s . 3 1 The action of the queen's k i l l i n g the king cannot be analyzed as the action of the queen's moving her hand causing the action of the queen's k i l l i n g the king. Actions are not related causally. Davidson has argued that i t i s a mistake to suppose that an action begins with a primitive action but ends l a t e r . So, the k i l l i n g of the king does not begin with the queen's primitive action of moving her hand and end with one of i t s consequences, 3 1 Davidson, 57-58. 24 the king's death. It i s true that the hand moving caused the death, but "the queen moving her hand i n that way, i n those circumstances" and "the queen doing something that caused the death of the king" are two descriptions of the same event. And doing something that causes a death i s causing a death. Causing a death i s , for Davidson, a k i l l i n g . Therefore the k i l l i n g of the king i s the queen's moving her hand i n that way, i n those circumstances, with that e f f e c t . d. Davidson's Accordion Davidson's diagnosis of the d i f f i c u l t y which others have had with a c t i o n i s the tendency to mistake features of a description of an event with features of the event described. An event can be described i n terms of i t s causes or of i t s e f f e c t s . Here are two d e f i n i t e descriptions of some event E 2 : (1) the event caused by event E1 (2) the event which caused the event E 3 I t i s a mistake to suppose that the consequence, E 3, i s included i n the d e f i n i t e description i n (2). What changes between (1) and (2) i s the description, not the event described: (1) and (2) are c o - r e f e r e n t i a l d e f i n i t e descriptions of one event, E 2 . As applied to action discourse, we can describe a single action i n terms of i t s causes or of i t s e f f e c t s . What gets described, i f we do, i s an event which i s the action. So, for Davidson, the accordion which endures the squeezing and the stretching of the accordion e f f e c t i s a single event which i s an action. What 25 changes i s the description of the action. Therefore, a l l of the f o l l o w i n g correspond to a s i n g l e descriptum, which i s the action: ( i ) "The queen moved her hand." ( i i ) "...thus causing the v i a l of poison to empty into the king's ear." ( i i i ) "...thus causing the poison to enter the body of the king." (i v ) "...thus causing the king to die." (v) "The queen moved her hand thus causing the king to die." (vi) "The queen k i l l e d the king." ( v i i ) "The queen emptied the v i a l into the king's ear." ( v i i i ) "The queen k i l l e d the king by pouring poison i n his ear." (ix) "The queen poured poison i n the king's ear thus causing his death." 3 2 The single descriptum i s the bodily event which i s the primitive action - the queen's moving her hand. Therefore, Davidson says, somewhat paradoxically, that there are only p r i m i t i v e actions: We never do more than move our bodies; the rest i s up to nature. 3 3 3 2 On the account of action which I develop (Chap. I I ) , n e i t h e r ( i ) , ( i ) + ( i i ) , ( i ) + ( i i i ) , nor ( v i i ) count as descriptions of the queen's action. 3 3 Davidson, 59. 26 The concept of being primitive i s , l i k e the concept of being i n t e n t i o n a l , intensional: under some description, namely ( i ) above, what the queen did was primitive; under another, say ( v i ) , i t was not. Likewise, the concept of being non-primitive i s i n t e n s i o n a l : under some description, e.g., ( v i i ) , what the queen did was non-primitive; under another, e.g., ( i ) , i t was not. 4. Summary The p o s i t i v e parts of Davidson's theory of action which he develops i n "Agency" are, then, as follows: 1. The C r i t e r i o n : Actions are events which are inte n t i o n a l under at least one description. 2. The ro l e which causation plays i n action i s the ro l e i t plays everywhere; namely, i t i s the r e l a t i o n between causes and e f f e c t s , and i t i s the r e l a t i o n of causation which permits re-description of actions i n terms of t h e i r causes and e f f e c t s . 3. Primitive actions are a l l the actions there are: "There are no further actions, only further d e s c r i p t i o n s . 1 , 3 4 And, non-primitive actions are a l l the actions there are. 4. Agency: The r e l a t i o n which holds between a person and an event which i s his action i s independent of how the terms of the r e l a t i o n are described. B. DAVIDSON'S SEMANTICS REJECTED In t h i s section I argue that Davidson's c r i t e r i o n of action and agency cannot work together with his semantics within a theory of action. I then suggest an alte r n a t i v e , leaving i t s 3 4 Davidson, 61. 27 f u l l development to the next chapter. F i r s t , i t w i l l pay to restate his notions of criterion and expression. Davidson's c r i t e r i o n of action, which t e l l s us whether something was done, and so whether agency i s on hand, i s that an event i s an action i f and only i f i t i s effected through agency. This i s a causal c r i t e r i o n , the causal term being an intentional e n t i t y or an agent. Since we want to d i s t i n g u i s h arm rai s i n g s from arm-risings, i t i s necessary that events which are actions be causally r e l a t e d i n the right way to an agent, where "the ri g h t way" i s cashed i n terms of intention: the agent must have intended the event under one true description. In other words, an event i s an action i f and only i f i t i s i n the r i g h t r e l a t i o n to the r i g h t states of an agent. Davidson's e x p r e s s i o n of agency, on the other hand, addresses another aspect of the r e l a t i o n between action and the agent: i t t e l l s us what was done, and so i d e n t i f i e s what was caused by agency. The expression of agency addresses the semantics of ac t i o n . Davidson believes that extensionality reigns among actions, that what you do when you act i s your action under any of i t s true descriptions. I f you s p i l l the contents of your cup mistaking i t for tea when i t i s coffee, then you s p i l l the coffee. I f the o f f i c e r sinks the ship mistaking i t for the T i r p i t z when i t i s the Bismark, then he sinks the Bismark. Because the expression of agency i s purely extensional, d e s c r i p t i o n s o f what was done f u n c t i o n e x t e n s i o n a l l y . A l l n o n - e q u i v a l e n t but c o - d e s i g n a t i v e 28 descriptions of an event which s a t i s f i e s the c r i t e r i o n pick out the same action. Therefore, as long as an event i s an action-so long, that i s , as i t meets the c r i t e r i o n - i t can be ascribed to the agent as his action under any description true of i t . The concept of an action i s an event extensionally conceived, for Davidson. We can begin to see that the c r i t e r i o n and the expression cannot work together i f we consider the following two points. F i r s t , we need a conception of action, and so of agency, which w i l l allow us to d i s t i n g u i s h cases i n which full agency has been at work, from those cases i n which partial or diminished agency has functioned. This d i s t i n c t i o n i s employed i n a r i c h way i n a c t i o n discourse. I t i s that which grounds the d i s t i n c t i o n between f u l l and p a r t i a l r e s p o n s i b i l i t y , which i n t u r n accommodates the workings of excuses and defenses. The f u l l / p a r t i a l agency d i s t i n c t i o n , and so excuses, requires that we r e c o g n i z e ( i s the r e c o g n i t i o n ? ) that acting standardly d i f f e r s from acting non-standardly. 3 5 A c r i t e r i o n of action and agency ought, therefore, to mark t h i s d i s t i n c t i o n . Davidson's c r i t e r i o n cannot. Indeed, any c r i t e r i o n based, as his i s , on ordinary event ca u s a l i t y where the terms of the r e l a t i o n are treated extensional l y i s , for this reason, incapable of marking the above d i s t i n c t i o n . For Davidson, the acting which occurs i n standard actions does not d i f f e r from the acting i n non-standard 3 5S.C. Coval and J.C. Smith, Law and Its Presuppositions (London: Routledge and Kegan Paul, 1986), Chap. I. 29 actions, nor would his c r i t e r i o n allow the d i s t i n c t i o n . The fa c t of i t being an action i s based only on the causal r e l a t i o n . Secondly, the d i s t i n c t i o n between f u l l and p a r t i a l agency i m p l i e s a corresponding d i s t i n c t i o n among a t t r i b u t i o n s of actions: there are standard actions and non-standard actions; those of the l a t t e r type being deviants of the former. This d i s t i n c t i o n i s what i s marked by " i n t e n t i o n a l " and "unintentional" i n action a t t r i b u t i o n s . There are inten t i o n a l a c t i o n s and u n i n t e n t i o n a l a c t i o n s , l i k e m i s t a k e s , and a c c i d e n t s . 3 6 Conceptually we cannot make sense of the non-standard without a theory of the standard. 3 7 The extensionality of Davidson's e x p r e s s i o n of agency w i l l not permit the d i s t i n c t i o n since what gets attributed to an agent as h i s action i s , i n standard and non-standard cases, the e x t e n s i o n a l l y conceived event. I have three arguments to support these claims against extensionality. 1 i The Argument from Motivation An adequate theory of action should leave undisturbed the normal i n f e r e n c e s between a c t i o n - what was done - and 3 6 The term "unintentional action" i s highly misleading. If ac t i o n s are n e c e s s a r i l y i n t e n t i o n a l , then with f a i l u r e of i n t e n t i o n a l i t y we have something other than an action: we have an unintentional event. 3 7 See Sam Coval and D.D. Todd, "Adjusters and Sense-Data, "American Phil. Quarterly 9 (Jan. 1972): 107. I provide an account of standard action and acting i n the sequel. 30 motivation. To describe an event as an action of an agent i s to impute a motive to that agent, the content of which i s embedded i n the desc r i p t i o n of the action attributed to him. That an action was a standard action implies, among other things, that i t was s u c c e s s f u l . To be successful i s f o r the action to satisfy the motivation which prompted i t - where t h i s implies that the motivation was the cause of that which s a t i s f i e d i t . What s a t i s f i e s the desires which constitute the motivation w i l l t e l l us what the d e s i r e s were f o r and w i l l therefore, i n standard cases, be a truth-condition on what was intended. S a t i s f a c t i o n i s an intensional notion, and so w i l l occur o n l y i f some event i s cognized, and cognized under the descr i p t i o n under which i t was desired. For example, the desire that the government be embarrassed w i l l be s a t i s f i e d by an event cognized under that d e s c r i p t i o n , and not by the same event cognized under the d e s c r i p t i o n "the government v o l u n t a r i l y changing i t s p o l i c y " or "the government recessing for holidays". Therefore, what you desire when you desire the government's embarrassment i s not an event under any of i t s t r u e descriptions. What you intend, when you intend to embarrass the government, i s not to produce an event under any of i t s true d e s c r i p t i o n s , nor all of i t s true d e s c r i p t i o n s . What you intend, when you intend to produce that which w i l l s a t i s f y the motivating desire, i s an aspect of an event. If so, and i f we are to s a t i s f y the desideratum that what you intend to do and what you do i n t e n t i o n a l l y are t i g h t l y linked ( i . e . , i f we are to have a unitary theory of intending and doing i n t e n t i o n a l l y ) , then what you do as the r e s u l t of a successful intention to produce an aspect of an event i s to produce an event-under-a-description, where the description i s the one embedded i n the p r o p o s i t i o n a l content of the d e s i r e ( s ) which motivated the doing. 3 8 Davidson's semantics of a c t i o n , of what i s done, i s inadequate i n not providing f o r the e s s e n t i a l motivational i n g r e d i e n t of a c t i o n . Since motivation i s an i n t e n s i o n a l notion, the normal inferences between ac t i o n and motivation cannot be supported by an extensional semantics of action. If the expression of agency i s purely extensional, then what an agent does i s what he intends, whether or not he intends i t under the description contained i n the a s c r i p t i o n . Therefore, the connection between the content of the desires and intention i n the causal ancestry of the action and the action, i s severed. Thus even Davidson's c r i t e r i o n of action i s so severed from the semantics that i t cannot play a motivational r o l e . Under the l i m i t s of a merely causal c r i t e r i o n of action, which i s what Davidson's becomes, since i t i s separated from the expression of agency, there are no means to explain why or to tes t whether the resultant event was intended. These questions make sense only when they e s s e n t i a l l y implicate some description of the event. "What did you intend?" and "Why did you do that?" cannot be answered ex t e n s i o n a l l y . I f the r i g h t i n f e r e n t i a l r e l a t i o n s 3 8Bratman has an objection to " t i g h t - f i t " which I address i n the sequel, where the r e l a t i o n of desire and b e l i e f to intent i o n i s developed. 32 im p l i e d by motivation are to be ret a i n e d , neither can the question "What was done?" be answered extensionally. 2. The Argument from Responsibility To describe an event as an action of an agent i s to ascribe r e s p o n s i b i l i t y f or that event to the agent. This gives us one r i g h t answer to the question about what we are responsible for. We can a l s o be h e l d r e s p o n s i b l e f o r events which are not actions. The charge of negligent commission ascribes to the agent r e s p o n s i b i l i t y f or a consequence of an action, t y p i c a l l y when the consequence i s negative, and only when one should not have caused the consequence. Negligence implies a standard of care with respect to action according to which one should have foreseen the l i k e l i h o o d of the negative outcome of an action, and should have, f o r this reason, r e f r a i n e d from so acting. Where one f a i l s to meet the standard of care and thus f a i l s to foresee the negative outcome, or acts despite the b e l i e f that there would be negative consequences, you are said to have acted n e g l i g e n t l y and are h e l d r e s p o n s i b l e f o r the n e g a t i v e consequence. The charge of negligent omission ascribes to one r e s p o n s i b i l i t y f o r an event when one should have acted, counterfactually, to prevent some, t y p i c a l l y negative, outcome. Where the charge of negligent omission i s warranted, one i s held responsible for a consequence, not of one's action, but of some event which you f a i l e d to prevent, and should have prevented. 3 9 3 9 For a d e t a i l e d d i s c u s s i o n of the r e l a t i o n between negligence and r e s p o n s i b i l i t y , see Chap. I l l , The Active-Passive D i s t i n c t i o n . 33 Therefore, we are responsible for our actions, and can be h e l d r e s p o n s i b l e f o r the consequences of our ac t i o n s and inactions. This i s to say that r e s p o n s i b i l i t y can carry further down the causal chain of events than does agency. The degree to which one i s held responsible for an action varies with the conditions of acting. One can be held f u l l y or only p a r t i a l l y responsible for some event. The d i s t i n c t i o n between f u l l and p a r t i a l r e s p o n s i b i l i t y goes hand i n hand with the d i s t i n c t i o n between f u l l and p a r t i a l agency. Responsibility for an event diminishes as agency of the event diminishes. It i s the job of excuses and legal defenses to reduce the degree to which one i s held responsible for some event, and the reduction, where j u s t i f i e d , i s j u s t i f i e d because agency i s at work i n the causal genesis of the event i n some diminished way. An adequate theory of action should leave undisturbed the normal inferences not only between motivation and action but al s o between a c t i o n and what one i s held responsible f o r . C l e a r l y , one i s h e l d r e s p o n s i b l e f o r one's a c t i o n s . On Davidson's extensional account of action I am held responsible for what I, as agent, cause however described so long as there i s a t r u e d e s c r i p t i o n of the event under which i t was in t e n t i o n a l for me. But t h i s prevents me from employing as a defence against the a s c r i p t i o n of r e s p o n s i b i l i t y , the fact that I did not intend that event under the description contained i n 34 the a s c r i p t i o n . 4 0 I am so prevented because i t i s that extensional event, however described, which i s my action. I f , therefore, Davidson i s ri g h t , and actions are the sheer events we cause as agents, and i f we are responsible for our actions, then the description under which I intended the event should be i r r e l e v a n t to how we gauge r e s p o n s i b i l i t y . But, i t i s not i r r e l e v a n t t o r e s p o n s i b i l i t y . The s i g n i f i c a n c e f o r r e s p o n s i b i l i t y of the des c r i p t i o n of events which figure i n one's inte n t i o n i s that i t i s just t h i s which allows excuses to work i n determining degrees of r e s p o n s i b i l i t y . Whenever an event which i s caused through my agency i s ascribed to me under some desc r i p t i o n other than that which figured or should have figured i n my intention, then an excuse applies and I am not responsible for that event under that description p r e c i s e l y for t h i s reason. Just such a move i s blocked i f actions are conceived of extensionally. Consider the following argument. We are not responsible for our mistaken actions. If I knowingly and i n t e n t i o n a l l y , shake the hand of the man I t r u l y b e l i e v e to be the bank manager, then shaking his hand i s an action of mine, and I am therefore, responsible for i t . Suppose that he i s , unbeknownst to me, the bank robber. On Davidson's view, shaking the bank manager' s hand i s the very same action as shaking the bank robber's hand. If we are responsible for our actions, then I ought to be responsible for that very same action, shaking the 4 0 The difference between intending and not intending i s l i k e the difference between desiring and not de s i r i n g . 35 hand of the bank robber. But I am not. Therefore, I am not responsible for an event under a description under which i t was not intended. That we employ excuses and defenses as we do suggests that one i s responsible, not f o r sheer events, but only f o r events under the d e s c r i p t i o n under which i t o c c u r r e d i n one's s u c c e s s f u l i n t e n t i o n . And, i f we are responsible f o r our a c t i o n s , then, by s u b s t i t u t i o n , an a c t i o n i s an event essentially under the description which fi g u r e d i n i t s cause. To c a l l an event an action is to so describe i t . It would appear that Davidson i s without a theory of excuses and so of how r e s p o n s i b i l i t y i s gauged, and i s without an account of how desires and b e l i e f s function i n agency so long as he holds on to his purely extensional account of agency; so long as he holds onto h i s purely extensional account of what was done. 3. The Argument from A t t r i b u t i o n To describe an event as an action i s to make an a t t r i b u t i o n to the agent. But ju s t what the a t t r i b u t i o n consists i n has separated action t h e o r i s t s . And not su r p r i s i n g l y , since the nature of a c t i o n a t t r i b u t i o n w i l l depend on the natures of action and agency: they define what we at t r i b u t e and to what i t i s a t t r i b u t e d . According to Davidson's expression of agency, what we at t r i b u t e to an agent when we att r i b u t e an action to him i s an event under a l l true descriptions of i t . Actions, for Davidson, 36 are e x t e n s i o n a l l y conceived events. Which events we can at t r i b u t e to an agent as his actions are those which s a t i s f y Davidson's criterion. Only those events which stand i n the r i g h t r e l a t i o n to the agent q u a l i f y as his actions: that i s , only those events which are caused by an agent's intention count as actions of h i s . Therefore, while an ac t i o n a t t r i b u t i o n a t t r i b u t e s an extensional event to an agent as his action, the truth of the claim e n t a i l s that the c r i t e r i o n was s a t i s f i e d , which i n turn e n t a i l s that the event was caused by an agent's inte n t i o n . Davidson's c r i t e r i o n of ac t i o n not only requires that a c t i o n s be events with the r i g h t s o r t of cause, but also specifies something about the nature of the causing and so of the cause. This something i s ju s t what allows us to say that the event was in t e n t i o n a l under at least one description. An event i s i n t e n t i o n a l under a description i f i t was caused by an inten t i o n of an agent. This makes the causing an intentional causing - a causing with a purpose - that purpose being to s a t i s f y the cause, the desire which motivated the action. That the cause i s satisfiable implies that i t has a representation of i t s object - that which would s a t i s f y i t . Intentions have both cognitive content and causal e f f i c a c y , and what they cause i s what would s a t i s f y them. In specifying t h i s about the nature of the cause of an action, an action a t t r i b u t i o n c l e a r l y s p e c i f i e s a limit to the nature of the cause. We at t r i b u t e c e r t a i n desires and b e l i e f s -37 those which cause and explain the action - with t h e i r cognitive content. Just as c l e a r l y , we do not a t t r i b u t e a l l possible desires and b e l i e f s causally relevant to an extensional event. That i s , we do not a t t r i b u t e an i n t e n t i o n f o r every true desc r i p t i o n of the event. In a t t r i b u t i n g an action to an agent, we a t t r i b u t e a s p e c i f i c mental event, and so a s p e c i f i c cognitive content, and causal e f f i c a c y which causes that which w i l l s a t i s f y i t . Therefore, to take, as Davidson does, an extensional event as the thing attributed i n action a t t r i b u t i o n without taking i n t o consideration the nature and the l i m i t s of the agency, i s , I w i l l argue, both an over and an under attribution. The source of t h i s d i f f i c u l t y f o r Davidson i s the s e p a r a t i o n of the c r i t e r i o n of action from the expression of agency. 4 1 I t i s just t h i s separation which allows him the e x t e n s i o n a l i t y of the event, and which prevents him from accommodating theories of motivation and r e s p o n s i b i l i t y within h i s theory of action. That the separation of the c r i t e r i o n from the expression i s a r t i f i c i a l w i l l be apparent i f we r e c a l l the function of the c r i t e r i o n . Its job was to get intention, and so purposefulness, and so agency into the picture. The c r i t e r i o n i s what allows us to d i s t i n g u i s h a r m - r a i s i n g s from a r m - r i s i n g s , p u r p o s e f u l behaviour from mere bodily movements. Purposeful behaviour i s behaviour with a purposeful cause, and a causing which, i n 4 1Davidson, 46-47, 120-121, 147, 195. 38 standard cases, produces that which s a t i s f i e s the cause. But what w i l l s a t i s f y a desire i s not an event under a l l true descriptions. What w i l l s a t i s f y a desire i s what figures i n i t s representational content. That w i l l be something l i k e aspects of events - the e x t r a - l i n g u i s t i c correlates of the descriptions which occur i n the mental events which constitute intentions. To claim that actions are events under a l l true description i s to a t t r i b u t e too much to an agent. Davidson's extensionality i s one kind of over-attribution. I t i s an over-attribution of descriptions of what one does as agent. Each new consequence of an action furnishes us with a new desc r i p t i o n 'of what was done. This i s Davidson's accordion e f f e c t . Another way for an action a s c r i p t i o n to over-attribute i s f or i t to at t r i b u t e too many actions. I t i s t h i s charge which Davidson r i g h t l y l e v e l s at Feinberg. Feinberg's accordion stretches over events, and each new event causally r e l a t e d to an action presents us with a new deed. His s i n i s to multiply actions; h i s at t r i b u t i o n s , then, at t r i b u t e too many actions to the agent. Davidson's e x t e n s i o n a l i t y of a c t i o n f o r c e s an under a t t r i b u t i o n to the agent by leaving the i n t e n t i o n a l i t y of the cause out of the act i o n a t t r i b u t i o n . Remember, i t i s the expression and not the c r i t e r i o n which i s attributed to an agent when we ascribe an action to him. Once the event which was caused by agency has been severed from the nature of i t s cause, and treated extensionally, there i s no way to re-connect such a 39 thing with agency, and so with intention. To claim that we at t r i b u t e extensional events to agents as t h e i r actions i s an under a t t r i b u t i o n since i t i s not such a thing that figures, or could figure, i n intentions. Davidson's extensionality forces an under s p e c i f i c a t i o n of the intention entailed by an action a t t r i b u t i o n . Just what the c r i t e r i o n of action was introduced for i s frust r a t e d by i t s separation from the expression of agency. The separation i s what allows Davidson to get intention, and so intension, back out of the picture. But, with the separation, we face a dilemma: either we connect the c r i t e r i o n to the action so that i t i s part of what i s attributed with action, i n which case what i s a t t r i b u t e d to the agent i s both the i n t e n t i o n a l i t y and the c a u s a l i t y , and t h e r e f o r e not mere extensional causation; or, the c r i t e r i o n i s not part of what we at t r i b u t e with action, i n which case action i s supernumerary to a l l or most of the conceptual connections of action discourse. Davidson would f i n d himself on the l a t t e r horn, but, as I have argued, t h e o r e t i c a l considerations favour the former. C. SOME IMPLICATIONS FOR A SEMANTICS OF ACTION My arguments against Davidson's extensionality, i f sound, have p o s i t i v e implications for a theory of action. I here sketch the alte r n a t i v e semantics of action suggested above, and which, I argue, meet the requirements on a theory of action. These b r i e f remarks w i l l be developed more f u l l y as I develop a 40 theory of intending, intention and acting i n t e n t i o n a l l y i n the sequel. Davidson, i t would seem, has i d e n t i f i e d the wrong sort of thing as the action. An event i s , according to the l i t e r a t u r e , supposed t o be the s o r t of t h i n g under no d e s c r i p t i v e r e s t r i c t i o n s : e x t e n s i o n a l i t y supports non-equivalence of d e s c r i p t i o n s . 4 2 I have argued that t h i s semantics won't do for action. Unless we assume that the i n t e n s i o n a l i t y of intention implies a representational or intensional r e s t r i c t i o n on the action, i t w i l l be devoid of the i n t e n t i o n a l content without which theories of motivation, r e s p o n s i b i l i t y and a t t r i b u t i o n are unworkable. We need, therefore, some additional semantic room over and above a world of extensional events and ordinary event c a u s a l i t y for the i n t e n t i o n a l content which we have found to be ineliminable from action discourse. While Davidson i s r i g h t i n that to a t t r i b u t e an action to an agent e n t a i l s a t t r i b u t i o n of causality, he i s wrong, I have argued, to think that i t i s mere causal e f f i c a c y . That he does i s evident i n h i s separation of the c r i t e r i o n of action, and so of intention, from the semantics of action, or of what we do. With intention we who engage i n action discourse accept some s t a t e of the agent which i s both c o g n i t i v e and c a u s a l . Therefore, we need a notion of cognitive e f f i c a c y or e f f i c a c i o u s 4 2 Davidson, 105-203. 41 c o g n i t i v i t y . 4 3 This, I claim, i s j u s t the r o l e desire plays i n our ordinary b e l i e f - d e s i r e psychology. Desires are j u s t that explanatory concept, both cognitive and causal, which we need to explain action. And action i s j u s t that notion with which we a t t r i b u t e t h i s cognitive e f f i c a c y to an agent, and therefore, a l o n g w i t h i t , the e f f e c t o f t h a t a t t r i b u t i o n , s i n c e a t t r i b u t i o n s of action are a t t r i b u t i o n s of agency. Next, we need a notion which i s the objective correlate of cognitive e f f i c a c y - that i s , what such a cause causes. That w i l l be whatever i s the e x t r a - l i n g u i s t i c c o r r e l a t e of the descriptions which occur i n the intentions which cause action, and which provide the e s s e n t i a l i n t e n t i o n a l content of actions. Descriptions correspond to properties or a t t r i b u t e s . Therefore, the concept we need f o r a c t i o n i s not an event under a d e s c r i p t i o n , but an event-under-a-description, an e s s e n t i a l l y i n t e n s i o n a l item. Which d e s c r i p t i o n the action e s s e n t i a l l y embeds w i l l be j u s t that one which occurs i n the c a u s a l l y successful intention. F i n a l l y , the r e l a t i o n which holds between these e s s e n t i a l l y i n t e n s i o n a l items, must be i n t e n t i o n a l causation: cognitive e f f i c a c y or e f f i c a c i o u s c o g n i t i v i t y functioning causally. This we accept when we accept that reasons are causes of actions. 4 3 This may be Dretske ' s notion of information. 42 We saw, i n the exegesis of "Agency", that Davidson was t r y i n g to f i n d an analysis of agency and action without having an account of intention. But, i t seems cle a r that once we take intentions as the cause of actions, then the r e l a t i o n between an agent and h i s actions w i l l be the r e l a t i o n between the agent's in t e n t i o n and h i s actions. The nature of the r e l a t i o n w i l l be explicated by one of i t s terms - desire. Agency i s that faculty of an agent by which his epistemic, t e l e o l o g i c a l and in t e n t i o n a l states produce behaviour. I turn therefore to the problems of intention, intending, and acting i n t e n t i o n a l l y . CHAPTER II RATIONAL AGENCY In t h i s chapter I present a theory of r a t i o n a l agency i n which w i l l be found an account of i n t e n t i o n , intending and coming to intend. Taking r a t i o n a l agency as basic, I diverge both from the t r a d i t i o n i n philosophy of action which begins with Anscombe taking what an agent does as the basic case by w h i c h t o u n d e r s t a n d i n t e n t i o n , 1 and from Bratman's "methodological p r i o r i t y of future-directed i n t e n t i o n s " . 2 I take a middle road, or rather both roads, with agency - the forming of an intention which i s e f f i c a c i o u s of an action - as the centre of attention. I claim that we are less susceptible to d i s t o r t i o n s both of action and of intention i f we maintain, what I argue i s , t h e i r e s s e n t i a l connections, and which are necessary i f we are to maintain the normal inferences which attend both notions. I begin with the common ground for requirements on a theory of intention. 1G.E.M. Anscombe, Intention (Oxford: B a s i l Blackwell, 1957), esp. 9. 2 M i c h a e l Bratman, "Two Faces of Intention," The Philosophical Review 93 (1984): 379. 43 44 A. PROPERTIES OF INTENTION There i s a family of concepts which i s connected with i n t e n t i o n and which a theory of intention must accommodate. Among them we include motivation, r a t i o n a l i t y , commitment to a c t i o n and the v a r i o u s modes of p r a c t i c a l reasoning. In a d d i t i o n , there i s the r e l a t i o n which i n t e n t i o n s bear to actions. In t h i s regard, there i s some common ground. A l l who accept Davidson's t h e s i s 3 that reasons are causes and who accept common-sense b e l i e f - d e s i r e psychological explanation accept that desires and b e l i e f s figure i n the causal ancestry of actions. But how they f i g u r e , and t h e i r r e l a t i o n to i n t e n t i o n s and i n t e n t i o n a l actions i s a matter of dispute. With intention, we accept the concept of a state of agency which i s causal of actions. What more we accept and need I w i l l discuss by f i r s t c o n s i d e r i n g what Donald Davidson and Michael Bratman have written on the subject. 4 1. Davidson on Intention In h i s paper "Intending" Davidson's concern i s with the p r o p e r t i e s of i n t e n t i o n s . Among these i s the conceptual connection between intending to A and being committed to A-ing. 3DonaldDavidson, "Actions, Reasons andCauses,"in Essays on Actions and Events (New York, N.Y. : Oxford University Press, 1980), 3-19. 4 Davidson, "Intending," Essays on Actions and Events, 83-102; Michael Bratman, "Intention and Means-end Reasoning," The Philosophical Review 90 ( 1 9 8 1 ) : 252-265; "Taking Plans Seriously," Social Theory and Practice 9 (1983): 271-287; "Two Faces of Intention," The Phil. Review 93 (1984): 375-405. 45 Commitment i s a necessary ingredient of intentions: 0 intends to A only i f 0 i s committed to A-ing. It i s t h i s aspect, at least, of intentions which distinguishes them from mere desires, which lack t h i s property. Since intentions contain commitment and desires do not, Davidson concludes that intentions cannot be reduced to desires. This marks a departure for Davidson from h i s e a r l i e r account of i n t e n t i o n i n "Actions, Reasons, and Causes", 5 where inte n t i o n a l actions were seen as those -which stand i n an appropriate r e l a t i o n to the agent's desires and b e l i e f s . 6 But for the above reason, Davidson, i n "Intending", has a non-reductive view of intentions. Intentions are l i k e d e s i r e s , i n t h a t both are of the genus pro-attitude, but intentions are a l l out or unconditional value judgments, whereas desires and wants are only prima facie value judgments. 7 The other property of intentions with which Davidson deals i n "Intending" i s what we may c a l l purity. Purity i s that property of intentions by v i r t u e of which we may separate them conceptually from the other components of agency and of action. F i r s t , we can intend to A without A-ing. And, for Davidson, we can abstract intentions from any antecedent or consequent event 5 Davidson, 3-19. 6 Others who hold a reductive view of intention include Robert Audi, "Intending," Journal of Philosophy 70 (1973): 387-403; Monroe Beardsley, "Intending," i n Values and Morals, ed. A. Goldman and J. Kim, (Dordrecht: Reidel, 1978); Paul Churchland, "The Logical Character of Action-Explanations," The Phil. Review 79 (1970): 214-236. 7 See below, p. 69 f., where I argue that commitment i s not an a l l or nothing concept. 46 such as deciding, deliberating, choosing, or any other form of coming to intend. This separation of intentions from the other components of agency, together with the claim that intentions are i r r e d u c i b l e , i s a view of intentions which we may c a l l the sui generis view. One way t o a p p r e c i a t e t h e f o r c e o f D a v i d s o n ' s i r r e d u c i b i l i t y claim i s to appreciate the following i r r e g u l a r i t y i n the attempt to explain actions by some account of mental causation. According to the reductive view, an i n t e n t i o n i s analyzable as a d e s i r e - b e l i e f complex, and i n t e n t i o n a l actions are those which stand i n the appropriate r e l a t i o n to these desires and b e l i e f s . The d i f f i c u l t y with t h i s view i s that i t does not, apparently, explain why i t i s that only some of an agent's desires are e f f i c a c i o u s of actions. That i s , i t doesn't explain why we have mental causation when we do. Davidson i s r i g h t to f i n d t h i s view of intention incomplete since i t leaves out the n o t i o n of commitment which i s c o n s t i t u t i v e of i n t e n t i o n s . With commitment, i n t e n t i o n s can do what mere desires can't: cause actions. It i s not c l e a r that with the notion of commitment we have an explanation of agency that we didn't have with desires. In p a r t i c u l a r , we have as y e t no e x p l a n a t i o n of why some commitments don't cause actions. Davidson's response could be that such cases may be explained by the intervention of some other component of agency, such as b e l i e f . But t h i s move i s 47 open to the reductive d e s i r e - b e l i e f theory of intention, and so Davidson's point does not t e l l against such a view. Commitment and p u r i t y do pose problems for a reductive account of intending. I w i l l show however that they are not insurmountable. But f i r s t , there i s room for disagreement on Davidson's characterization of the property of purity. It i s undeniable that one can intend to A without A-ing. This i s only to accept the commonplace that one's mind can change with respect to some action, where t h i s may be that one's degree of commitment can change: commitment i s the sort of thing which can wax and wane. I t would be a strength of a theory of intending that made e x p l i c i t how t h i s might work. However, in t e n t i o n s are not without e s s e n t i a l , i . e . , impure r e l a t i o n s both to antecedents and consequents of the system whose intentions they are. As noted above, there i s a conceptual entailment between intentions and commitment to action. That i s , intentions e n t a i l r e s t r i c t i o n s on the e f f i c a c y of competing desires. To intend to A i s to be i n a state of agency such that competing desires, should there be any, are held i n check, and such that one w i l l , ceteris paribus, A. Therefore, intentions r e s t r i c t which other intentions a r a t i o n a l system can have. As such, intentions are intimately, and i n e x t r i c a b l y connected to the rest of agency. One of the strongest connections i s with r a t i o n a l i t y . What I suggest i s that when we pick up a l l the r e l a t i o n s between a l l the other concepts e s s e n t i a l to agency, we w i l l have intention. 48 I f p u r i t y , f o r Davidson, means that intentions are not e s s e n t i a l l y r e l a t e d to the other components of agency, then, presumably, one could intend to A without having come to intend via some r a t i o n a l process between and among one's desires, b e l i e f s and other intentions- I f so, one could "parachute into" a state of intending to A; that i s , j u s t find oneself intending to A, with a l l that e n t a i l s . But such a separation of intention from the r e s t of agency i s incompatible with an important e s s e n t i a l ingredient of intending. It must be rational for an agent to intend to A, where to be r a t i o n a l i s for the intention to be compatible with the agent's desires, b e l i e f s and other i n t e n t i o n s . T herefore, what d e s i r e s , b e l i e f s and other intentions an agent has r e s t r i c t which further intentions he can form. I f so, intentions are, at the very least, not without i n e x t r i c a b l e connections to the other components of agency. Therefore, one could not ju s t f i n d oneself intending to A, where t h i s means that agency has been circumvented. Intentions are not pure i n t h i s sense. So f a r , then, we have found that a theory of intention must accommodate the f o l l o w i n g p r o p e r t i e s of i n t e n t i o n : the unconditional value-judgement, r a t i o n a l i t y , and the possible separation of the intention to A, from A-ing. 49 2. Bratman on Intention Michael Bratman deals, i n his recent work on i n t e n t i o n , 8 with several of i t s aspects. Among them i s the property of r a t i o n a l i t y by v i r t u e of which intentions must be consistent with one another, a property which does not extend to desires. This demand on intentions allows what he has c a l l e d the co-ordination of intentions i n planning. Bratman also deals with the r e l a t i o n of what one intends to what one does i n t e n t i o n a l l y . Though I cannot here do f u l l j u s t i c e to Bratman's work on these issues, I w i l l sketch h i s p o s i t i o n and r a i s e some concerns I have about c e r t a i n of his views which have relevance to t h i s project. In the next chapter, when we w i l l be better equipped to do so, I argue d i r e c t l y against one of h i s main points. One of Bratman's strategies i s to take planning as central to agency, and so to the concepts of action and intention. We are, he says, planning creatures. The c o - o r d i n a t e a b i l i t y of intentions into plans i s a r e s u l t of the r a t i o n a l i t y constraint on intentions. Since intentions must be consistent, and desires not, intentions are, he says, i r r e d u c i b l e to mere d e s i r e - b e l i e f complexes. Planning i s a function of agency which cannot, Bratman argues, i n consort with Davidson, be analyzed i n terms of b e l i e f s and desires. The other main issue with which Bratman deals i s the content of intentions. Bratman i s concerned to preserve the 8Bratman, op. c i t . . Zi 50 conceptual connections between a c t i o n s , r e s p o n s i b i l i t y and i n t e n t i o n s . In p a r t i c u l a r , he i s concerned w i t h the implications of ascriptions of r e s p o n s i b i l i t y for the content of intentions and the r e l a t i o n between what one intends and what one does i n t e n t i o n a l l y . According to what he c a l l s the "tight f i t " theory, there i s an intimate connection between the content of one's intention and what one does i n t e n t i o n a l l y , so that one does i n t e n t i o n a l l y only what one intends to do. But Bratman fi n d s that t h i s threatens the connections we employ between action and r e s p o n s i b i l i t y , since one can be held responsible for what one does not do i n t e n t i o n a l l y . Through a serie s of counter examples, 9 the connections between what one intends and what one does i n t e n t i o n a l l y are so weakened that one can do i n t e n t i o n a l l y what one does not intend to do. Thus weakened, ascriptions of r e s p o n s i b i l i t y are preserved for cases where we want to hold someone r e s p o n s i b l e f o r what they d i d not i n t e n d , but nevertheless caused. To f i l l the gap l e f t i n the r e l a t i o n between intention and inte n t i o n a l action, Bratman introduces the notion of the motivational potential of an intention: A i s i n the motivational pot e n t i a l of my intention to B, given my desires and b e l i e f s , just i n case i t i s possible for me i n t e n t i o n a l l y to A i n the course of executing my intention to B. 1 0 Motivational pot e n t i a l i s what connects agents to t h e i r actions. 9 See below Chapter III, sec. A. 1 0Bratman, "Two Faces", 395. We n o t i c e that Bratman's weakening of the connection between what one intends and what one does i n t e n t i o n a l l y i s l i k e Davidson's s e p a r a t i o n of i n t e n t i o n from the expression of agency. I r e s i s t t h i s weakening for the same reasons I re s i s t e d Davidson's separation, and w i l l argue against i t af t e r I have set out the theory of r a t i o n a l agency. For now, l e t us notice that once the separation i s made, we f i n d ourselves capable of a t t r i b u t i n g to agents f u l l r e s p o n s i b i l i t y f or events for which agency i s responsible only i n a diminished way. And, i t i s pos s i b l e to recognize the diminution, and important for the theory of intention and action to mark that difference. The issue i s not whether we should c a l l such cases actions. Rather, i t i s that we f i n d ourselves equipped to make important and subtle d i s t i n c t i o n s based on the re l a t i o n s h i p between the mind of the agent and those events which he e f f e c t s i n the world, and we should do so. After a l l , t h i s i s j u s t what we must be concerned with as wielders of action discourse, and therefore j u s t what an action theory should f i n d . 1 1 The other c r i t i c i s m I have of Bratman, i s a c r i t i c i s m of strategy. Planning cannot be central to action and intention when we can both act and intend to act without planning. This i s not to deny the importance of the uses of reason, consistency and planning for mature agents. But, the strategy d i s t o r t s the outcome. The view of intentions which I o f f e r is a reductive See below, sec. C. 52 one, but i s nevertheless, I argue, one which s a t i s f i e s a l l the acceptable requirements on a theory of intention. B. A THEORY OF RATIONAL AGENCY I s h a l l develop a theory of what i t i s for an agent to intend - to have an intention - and to act i n t e n t i o n a l l y . I begin with a simple case of action, and ask what p a r t i c u l a r s , properties, and r e l a t i o n s are necessary for the agent to have so acted. To t h i s end, i t i s h e l p f u l to construe the task as that of d e s c r i b i n g and constructing a system, both r a t i o n a l and t e l e o l o g i c a l , which i s capable of action. Such a system we may c a l l an agent. In order to f a c i l i t a t e the task I introduce an action schema, which w i l l be developed as the d e s c r i p t i o n of the system develops. Consider the man, 0, who releases the secret papers i n order to embarrass the government. I t i s true of 0, that he desires, d x , to embarrass the government and believes, bx , that the release i s a means to that end. That he so acted means, among other things, that dx and b x caused, >, the desire, d 2, to release the secret papers. This i s the causal compliment of A r i s t o t l e ' s p r a c t i c a l syllogism. In terms of an action schema we have: Fig. 1. d x & b x > d 2 This represents the mental side of things for 0. Add to t h i s a representation, by 0, of the world of events which he effected 53 t h r o u g h h i s agency: the e v e n t of t h e government's embarrassment, Ex , caused by the event of the release of the secret papers, E 2, which was caused by E B, the event of O's hand moving. We now have: Fig. 2. A1 & bx > d 2 . . . E B > E 2 > Ex . Thus we have, on the l e f t hand s i d e o f the schema, a representation of O's mentality, and on the r i g h t hand side, a representation of the world. The e l l i p s i s i n F i g . 2 w i l l be f i l l e d i n as we proceed. It i s important to note that as yet we do not have an action represented since the E's, as they occur here, represent extensional events i n the causal nexus. Before i t can be c a l l e d a schema of O's action, the r i g h t side w i l l have to be a r t i c u l a t e d to bring i n the e s s e n t i a l i n t e n t i o n a l content of actions. What we have so far represented of O's mental l i f e are the propositional attitudes of b e l i e f and desire. Therefore, the d's and b's of the schema represent states of the agent with p r o p o s i t i o n a l content, d i f f e r e n t s u b s c r i p t s to r e p r e s e n t d i f f e r e n t content. With desire we have a r i c h notion of a state of agency which i s cognitive and motivational, and meant to be s u f f i c i e n t to i n i t i a t e the processes of agency. We can ask what properties such a state must have i n order to function as i t does. F i r s t , i t must have causal power to move the system 54 through the processes of agency, 1 2 and to e f f e c t basic bodily movements by which agents af f e c t the world i n order to cause the states of a f f a i r s described i n the propositional content of t h e i r desires. That desires are causal with respect to t h e i r content implies that the causal property of a desire i s voided by the cognition that i t s object has occurred. Desires are self-extinguishing i n t h i s way. With the notion of an item which i s causal of a state of a f f a i r s which i s i t s object and which w i l l void i t s causal power, we have both the notions of satisfaction of a desire by the recognition of the occurrence of i t s object, and a state of pre-satisfaction i n t e g r a l to the "pro-ness" of the pro-attitude of the desire. With pro-attitude we have the notion of being attracted toward, or cognizant of the attractiveness of, the object of the desire. This i s the basis of value. And f i n a l l y , with desire, we have a state of agency which the agent may report upon. So we have, with desire, these properties: 1. cognitive e f f i c a c y 2. v o i d a b i l i t y of causal power 3. s a t i s f i a b i l i t y 4. p r e - s a t i s f a c t i o n 5. "pro-ness", or value, or attractiveness 6. r e p o r t a b i l i t y With b e l i e f we have the concept of a cognitive state with content, and the capacity to a f f e c t desires, and through t h i s , to e f f e c t other desires and states of a f f a i r s i n the world. B e l i e f s can channel the c o g n i t i v e e f f i c a c y of a d e s i r e by 1 2 This w i l l be elaborated upon shortly. 55 providing the means to i t s s a t i s f a c t i o n . 0 only desired to release the secret papers, and only did so, Jbecause he believed that would cause the embarrassment of the government. Although the function of b e l i e f s i s wide-ranging wi t h i n agency, the influence of b e l i e f s on states of a f f a i r s beyond the body i s always transmitted through desire. That i s to say that desire, being both cognitive and causal, i s our representation of the idea of the points of interface between the cognitive and the physical at hoth ends of agency. This i s our idea of agency. With b e l i e f , then, we so far have the following: 1. capacity to a f f e c t desires 2. capacity, i n conjunction with desires, to e f f e c t other desires, other states of agency, and through them, states of a f f a i r s . 1 3 We can now ask whether we have on hand a set of properties and r e l a t i o n s of O's state of agency which i s s u f f i c i e n t to cause his action. For i t to be s u f f i c i e n t , d x bt , and d 2 would have to be s u f f i c i e n t to constitute motivation. But i t seems that as yet they are not. It i s widely held that desires, or desire and means-end b e l i e f pairs constitute motivation. To i d e n t i f y a desire as a state with cognitive e f f i c a c y i s to be committed to the view that desires are the source of motivation - the motive force. And, for a r a t i o n a l system with agency, a means-end b e l i e f i s necessary to serve the desire i n r a t i o n a l action. But, from the occurrence of a prime motivant, a d x, or of a d 1 together with a means-end b e l i e f , a ^ , nothing follows 1 3 This l i s t w i l l be added to shortly. 56 i n the way of action without a b e l i e f to the e f f e c t that the desired state of a f f a i r s i s not or may not yet be the case. Without t h i s b e l i e f , one would normally, i . e . , r a t i o n a l l y , not yet be i n a state of motivation. Furthermore, one w i l l be i n a state of cognitive e f f i c a c y toward an object state of a f f a i r s only if one believes that the state of a f f a i r s i s a possible one. This i s not to say that i t i s necessary to know how to produce the goal, or to believe you have the means, i n order to be motivated, but only that the agent believe that the goal i s not impossible. This requirement constitutes part of the reason why one cannot intend what one b e l i e v e s to be impossible. Therefore, cognition of the way things are and of how they could be i s presupposed by motivation: agency, and so purposefulness, presupposes cognition. As an a s i d e , c o n s i d e r b r i e f l y the s t a t e which i s characterized as having some of the properties of desire, but because of the lack of the b e l i e f that the goal i s a possible one, i s not yet a state of desirous e f f i c a c y . This would be a pro-attitude, and so have the properties of p r e - s a t i s f a c t i o n and "pro-ness" or value, but would lack causal power. We may c a l l such a state a wish or a hope, a state of the genus pro-attitude, but lacking the cognitive e f f i c a c y of a desire. A wish or a hope could become a desire, but i n so doing would cease to be a mere wish or hope. To corrupt a phrase, a wish may be father to a desire, but only i f e f f i c a c y i s the mother. And, a desire may degenerate into a wish or hope, indeed will 57 for a r a t i o n a l system, when the system acquires the b e l i e f that i t s object i s not possible. This gives us another function of b e l i e f . B e l i e f s have the power to void the e f f i c a c y of a desire. The b e l i e f that the d e s i r e d s t a t e of a f f a i r s i s impossible, or i s current, i s s u f f i c i e n t to cancel the causal power of the desire. With t h i s analysis of the p r i n c i p l e agency state, we can see what more we need to add to the action schema. With the pre-supposed non-actuality and p o s s i b i l i t y b e l i e f s , we must add a set of background b e l i e f s . In addition, the fac t that 0 acted on dLj^  e n t a i l s , for a r a t i o n a l system, the b e l i e f that so acting would not f r u s t r a t e more important or more powerful desires of 0. That i s , before the e f f i c a c y of a desire may go through, the question of consequences f o r the agent's other desires of s a t i s f y i n g the desire i n question must be asked, and i f not, i s a departure from r a t i o n a l i t y , and one form of i r r a t i o n a l behaviour. This e n t a i l s that the mature agent (a system with more than one desire) has a set of desires which are ranked i n two ways: according to causal power, and attractiveness. It also e n t a i l s a set of b e l i e f s about the set of desires and the rankings. We can add to the schema the set of desires <d>, and a set of b e l i e f s which includes the pre-supposed b e l i e f s and the b e l i e f s about desires, <b>. We are now i n a p o s i t i o n to describe the process of coming to intend. The f i r s t step of the process w i l l be that point at 58 which d1 becomes "active" for the agent. We begin, therefore with the set of desires, <d>, and the set of b e l i e f s <b>. Desire d.,^  w i l l emerge from the set of desires either because of some cognitive cause, such as the b e l i e f that an opportunity has arisen to s a t i s f y i t - i . e . , that the world has " l i n e d up" i n such a way that s a t i s f a c t i o n i s possible - or because of d x ' s circumstantial or r e l a t i v e causal strength: Fig . 3. <d> This represents step 1, where d± "drops down" to engage i n the subsequent stages of the action process. Step 2 of the process i s for the occurrence of the a c t i v a t i o n of d1 to prompt the system to evaluate the consequences for O's other desires of acting so as to s a t i s f y dx . To repeat, t h i s i s a r a t i o n a l i t y requirement on the system: Fig. 4. <d> <b> An answer to t h i s question e n t a i l s a ranking of O's desires. If acting so as to s a t i s f y d± has more value for 0 than s a t i s f y i n g any other of h i s desires, or more value than those desires which may be fr u s t r a t e d by so acting, then the answer to the question I 59 about compatibility with the members of <d> w i l l be p o s i t i v e , " + ", and d x w i l l , under the impetus of i t s causal e f f i c a c y and because of t h i s f i r s t v a l i d a t i o n , proceed, step 3, to the next stage. I f the answer were negative, then the process of coming to intend w i l l cease, i f r a t i o n a l . This i s to say that O's b e l i e f s about the consequences of ^ proceeding i n the process constitute an on/off switch for d x . The causal power of d x may not be voided, but i f not, i t w i l l be held i n check by O's b e l i e f s . This i s to say, I w i l l argue, that a negative answer at t h i s point w i l l prevent the system from intending dj . The negative answer, i n the form of a b e l i e f i s a gate preventing dx from entering the further pathways of the system. Since an action schema i s also a r a t i o n a l i t y schema, where dx proceeds despite an "of f " , we have a form of i r r a t i o n a l i t y c a l l e d compulsion. In these terms, compulsion i s for a desire to be e f f i c a c i o u s because i t i s too strong for the system. S i m i l a r l y , we could have a system whose b e l i e f s are too weak to prevent desires from being e f f i c a c i o u s of actions. Or, an immature system may not have s u f f i c i e n t desires and b e l i e f s to constitute competition for some desire, which therefore passes t h i s stage unimpeded. Returning to the process, we have dj which has survived t h i s f i r s t stage i n the vetting process: 60 The occurrence of d x ' s a r r i v a l here, on what I s h a l l c a l l the " I - l i n e " , prompts the r a t i o n a l system to ask, step 4, f o r implementation of ^  . That i s , O's b e l i e f s concerning the means available to him, <b >, are queried: F i g . 6, <b> From among t h i s set, a set of b e l i e f s about the means to s a t i s f y i n g d x i s i d e n t i f i e d because they are believed by 0 to be the best available, because most e f f e c t i v e . The members of t h i s set are themselves ranked by effectiveness. This e n t a i l s a set of second-order b e l i e f s for 0. From among the set of most e f f e c t i v e means, 0 selects, step 5, the means which he believes, b x , to be most e f f e c t i v e . Given that he desires the end, 0, at step 6, desires, d 2, the means. Next, he re f e r s t h i s desire for means to h i s b e l i e f s about his desires, step 7, and asks again, as i n step 2, how the c o r r e l a t i v e event would a f f e c t h i s other desires: This process requires that for each member of the set of means available, O's b e l i e f s about h i s desires are queried as to the a c c e p t a b i l i t y of the occurrence of the state of a f f a i r s which i t represents, together with the occurrence of the state of a f f a i r s represented i n d^^ . This w i l l involve a ranking of effectiveness and compatibility or a c c e p t a b i l i t y of the events envisaged, by summing the values which each component has. If the answer to the query at step 7 i s p o s i t i v e for some means, then 0, again i f r a t i o n a l , w i l l be caused, step 8, to desire, on the whole, that means. For our agent 0, the means a v a i l a b l e to him i s the r e l e a s e of the s e c r e t papers. Therefore, 0 i s caused to desire the release, d 2, because i t i s the means to s a t i s f y i n g d x : If the answer at step 7 i s negative, either because there are no means a v a i l a b l e , or because none of those a v a i l a b l e are acceptable to 0, then the causal e f f i c a c y of i s voided, i n the f i r s t case, or held i n check, i n the second, by O's vetting b e l i e f s . In coming to desire, d 2, the release of the secret papers, 0 need not desire that event under a l l of i t s true descriptions. What i s required of 0, i f he i s to be r a t i o n a l , i s that he desire the means to s a t i s f y i n g d x . Therefore, d 2 may occur for 0 under t h a t minimal description - d e s i r i n g the means to embarrassing the government, or desiring the means-ness of the release of the secret papers. Of course t h i s event may be d e s i r e d f o r other of i t s p r o p e r t i e s , under other of i t s descriptions. But, 0 may desire the means-ness of the event while d e s i r i n g that i t would not have other of i t s properties. 0 may believe that the release of the papers w i l l be the same event as the cause of the loss of his job, or that which w i l l s t i g m a t i z e h i s fami l y . But, the d e s i r e to embarrass the government may be so strong, for 0, that he desires the means 63 despite these undesired consequences. What he cannot do, i f r a t i o n a l , i s d e s i r e the means when that event under that d e s c r i p t i o n i s more undesirable than the goal i s desirable. What he must do, i f r a t i o n a l , and so long as the v e t t i n g process has yielded p o s i t i v e r e s u l t s , i s desire the means. So d 2 must have at least t h i s minimal content. Just as the occurrence of d x on the I - l i n e prompted the system to ask "How?", so the occurrence of any desire for means may occasion the question of implementation. Indeed, i t w i l l occasion t h i s question for any desire which the system as yet does not know how to implement, and where the system i s r a t i o n a l . But for any system which i s capable of a f f e c t i n g the world, there must come a point at which the question of implementation does not a r i s e . That point, for our man 0, w i l l be the one where he "knows", either by learned patterns of behaviour or background a b i l i t i e s , how, for creatures such as we, to engage the world with h i s body. I t i s the point i n the process of coming to act where further means do not enter. 0 at step 9 asks, of h i s b e l i e f s about means, how to release the secret papers, and repeats steps 5 to 7 at steps 10 to 12. I t turns out that he can, b 2, move his hand, which holds the secret papers, toward the newspaper reporter, and so i s caused to desire that, d 3, at step 13: 64 Fig. 9. <d> <b> <b > <b> In order to s a t i s f y d 3 , 0 need not ask f o r means. It i s s u f f i c i e n t f or d 3 to be e f f i c a c i o u s of some bodily event that i t be vetted as above i n the process of agency. That t h i s i s so, allows us to c a l l d 3 a basic desire, or d B . Basic desires are those which e f f e c t the world d i r e c t l y through our bodies via the background of a b i l i t i e s we have to move our bodies. This must be true for some desires or agency could not emerge. Just as there are pri m i t i v e actions antecedent to which we perform no other, there are mental events a f t e r which occur no relevant others before we act. This i s represented i n the schema as: Fi g . 10. where E B i s the basic event i n the world of, i n our example, O's hand moving. Where an agent was successful through action, the relevant d e s i r e s , screened and channelled by b e l i e f s i n the v e t t i n g process, w i l l have caused those states of a f f a i r s i n the world which s a t i s f y the e f f i c a c i o u s desires. That i s , the e f f i c a c i o u s mental events, with t h e i r representations of possible events, 65 w i l l have caused a corresponding set of actual events i n the world. Taking the bottom I - l i n e of the l e f t side of the schema, we may represent t h i s as: Fi g . 11. d x . . . d 2 . . . d B > E B > E 2 > E1 where, i n our example, E 2 i s the event of the release of the secret papers, and E1 i s the event of the government being embarrassed. We do not, as yet ( f i g . ' s 9-11), have represented an action schema. What we have yet to add i s the notion of s a t i s f a c t i o n . S a t i s f a c t i o n , I have argued, i s a cognitive and intensional matter. What s a t i s f i e s a desire i s not some extensional event but some s t a t e of a f f a i r s or set of p r o p e r t i e s which i s represented i n the cognitive content of the desire. I t i s an event-under-a-description, an e s s e n t i a l l y intensional item which s a t i s f i e s a desire. Therefore, before we can represent the matching of the c a u s a l l y e f f i c a c i o u s d e s i r e s with t h e i r corresponding objects i n the world, we must construe the E's of our schema as the i n t e n s i o n a l items d e s c r i b e d above. Furthermore, what " s a t i s f i e s " the instrumental beliefs of the schema w i l l be the r e l a t i o n which holds, i n successful actions, between these E ' s . We represent the s a t i s f a c t i o n r e l a t i o n with '<—S—'. A standard action w i l l have the following form: 66 F i g . 12. A MODEL OF RATIONAL AGENCY: RAAG The Mental jThe World of External Events <b> <d>: dx: <b>: I: <bm >: m <b > m -*2 <-S—: the the the the the the the the the the the the the the U set of RAAG's desires and intentions desire which becomes causally active set of RAAG's b e l i e f s about <d> l i n e of intending set of RAAG's b e l i e f s about means b e l i e f that (such and so) i s the means to d x desire for the means to d x b e l i e f that (such and so) i s the means to d 2 basic desire basic event under the description i n d B event under the description i n d 2 event under the description i n d x r e l a t i o n of in t e n t i o n a l causation r e l a t i o n of s a t i s f a c t i o n 1 2: 3: 4: 5: 6: 7: 8: 9 10: 11 12: 13: 14: d x becomes causally active d x i s vetted by RAAG's b e l i e f s about his other desires d x becomes an intention RAAG's b e l i e f s are searched for means to s a t i s f y dj a b e l i e f , bx , about means emerges RAAG i s caused to desire the means the means i s evaluated by RAAG's b e l i e f s about h i s desires d 2 becomes an intention RAAG's b e l i e f s are searched for means to s a t i s f y d 2 a b e l i e f , b 2, about means emerges RAAG i s caused to desire the means the means i s evaluated by RAAG's b e l i e f s about h i s desires d B becomes an intention RAAG's inten t i o n i s e f f i c a c i o u s 67 Thus we have a model of r a t i o n a l agency. And with the aid of the schema we can see how to separate these concepts. Rationality i s what comes afte r d x ' s f i r s t appearance i n the schema. The r a t i o n a l i t y of agency i s that which serves the s a t i s f a c t i o n or i t s motivating desires. Agency i s d x , plus r a t i o n a l i t y , plus E B . I t i s e f f i c a c y of an event under a d e s c r i p t i o n according to the process described above. And Autonomy i s non-interference with, and non-aberration of, the process and i t s constituents. Of course the form of RAAG i n 12 i s not meant to be r i g i d . I t can be as simple as an occurrence of a d B causing an E B, or complicated by i t e r a t i o n s of the module that i s steps 4 - 8 , the "means-module". This would be the case for complex r a t i o n a l agency. In the full-blown schema 12, the bottom l i n e of desires, d x , d 2, d B, i s the l i n e which represents that the causal power, as well as the other relevant properties of desire, are in t a c t , by v i r t u e of the vetting process of the agent's b e l i e f s and desires. Therefore, intending begins only a f t e r the desire, d 1, has had i t s f i r s t v e t ting at step 2. That i s , intending begins at step 3, where ^ reaches the I - l i n e . This marks the fact for the agent that s a t i s f a c t i o n of dx i s , according to h i s b e l i e f s , on the whole d e s i r a b l e . Before the process of r a t i o n a l i t y begins, the agent merely desires but does not intend. Intending i s the r e s u l t of t h i s f i r s t stage i n the r a t i o n a l i z a t i o n of desires. The t e s t for whether a state of agency i s an intention 68 i s whether the agent can be i n that state and not intend. I claim that at step 3 above, he cannot. The agent also intends at those points at which d 2 and d B reach the I - l i n e - that i s , wherever the v e t t i n g process y i e l d s a p o s i t i v e r e s u l t . Thus we may say that intending - having an intention - i s having a desire functioning causally and p o s i t i v e l y via the b e l i e f s i n the v e t t i n g process. To report on the fact that one intends to A i s to report that one' s desire to A i s i n the vetting process and has had only p o s i t i v e vetting r e s u l t s so f a r . At any point i n the process, of course, a query as to the a c c e p t a b i l i t y of the events under consideration may y i e l d a negative answer, i n which case the causal e f f i c a c y of the desire, and so of the intention, i s e i t h e r voided or held i n check, i n a r a t i o n a l system. Therefore, one may indeed intend to A without A-ing. The schema thus elucidates how intending to A may not be s u f f i c i e n t for A-ing. I t f u r t h e r e l u c i d a t e s the intimate r e l a t i o n s between intentions and the rest of agency, the consistency requirements on i n t e n t i o n s , the commitment which i s c o n s t i t u t i v e of intentions, and the r a t i o n a l i t y of intentions. I t also makes p l a i n how plans and planning may function, since plans w i l l be sets of i n t e n t i o n s f o r future actions and, since they are intentions of a r a t i o n a l system, must be consistent. Future i n t e n t i o n s may be "stacked" i n <d> pending b e l i e f s about o p p o r t u n i t y and o t h e r a s p e c t s o f p l a n n i n g f o r t h e i r s a t i s f a c t i o n . 69 The I - l i n e of schema 12 represents the i n t e n t i o n with which, i n our example, 0 acted. At ^ , 0 intended to embarrass the government, and so, by the nature of i n t e n t i o n s , was committed to so acting. At d 2 , 0 intended, not simply, or merely, to release the secret papers, but intended to release the s e c r e t papers so as to embarrass the government. The content of the I - l i n e represents the content of the agent's inten t i o n . And so we see that the content of O's intention accretes as he learns how to s a t i s f y h i s desire and then intends to so s a t i s f y i t . At d B, O's intention i s not merely to move his hand. I t i s to move his hand i n ju s t the way which he believes w i l l be to release the secret papers so as to embarrass the government. This i s the intention with which 0 a c t s . 1 4 Committment i s a concept which admits of degree. Just as the content of O's intention accretes as we move from l e f t to r i g h t a l o n g the I - l i n e , so h i s committment may ev o l v e accordingly. I f , as 0 comes to intend to act, he discovers that the event he w i l l cause w i l l have further desireable properties, or fewer undesirable properties than he may have suspected, his committment may increase. Conversely, h i s commitment may decrease as he discovers undesirable consequences and side-e f f e c t s of the envisaged action. However, commitment i s not ju s t a function of desire. I t i s rather a function of the processional v e t t i n g of a desire. The further along one i s , in the process, the fewer impediments stand i n the way of the 1 4 This accretion of descriptions i s the mental side of the accordion e f f e c t . 70 vetted desire, and e f f i c a c y i s s t i l l "on" i n l i g h t of the v e t t i n g b e l i e f s . Therefore, commitment can increase, even though desire decreases. 1 5 With the f u l l - b l o w n schema, we have represented, i n schematic form, a l l the various modes of p r a c t i c a l reasoning, which w i l l o c c u r at the v a r i o u s nodes of the schema. Deliberating about how to act so as to s a t i s f y a desire i s the means-module, steps 4 - 8 . Deliberating about whether or not to act, where t h i s amounts to an evaluation of competing desires, i s step 2, 7, etc. . Choosing to act may be where no clear ranking of desires emerges, and some p r i o r motivant, such as the desire to act i n some way, or t h i s , together with a b e l i e f that opportunity w i l l be l o s t i f one does not act soon, causes a d to "drop down" to engage i n the process. Choosing a means suggests that no cle a r ranking emerges, and, as above, the motivants already i n the process cause the system to desire one of the e q u a l l y acceptable means a v a i l a b l e . W i l l i n g to act i s an in t e r e s t i n g case. I i d e n t i f y w i l l power with the causal power of desires. Therefore, one always acts out of w i l l power. Yet the claim that one acted out of will power suggests that the motivants f o r the a c t i o n i t s e l f were i n s u f f i c i e n t to move through the system, and some antecedent desire was involved i n coming to act. Doing one's duty may be such a case, where the 5 See p. 44 f. above, for Davidson's account of commitment. 71 desire to do one's duty i s the motivant, and not the desire for the action i t s e l f . 1 6 This, then, i s the theory of r a t i o n a l agency, a r a t i o n a l process of v e t t i n g of a desire with cognitive e f f i c a c y . To assert t r u l y that one intends to A i s to assert that the mental process sketched above, which involves desires, b e l i e f s and t h e i r r e l a t i o n s , has taken, or i s taking place. The phrase "the agent i s intending to A" i s true i f and only i f the r e s u l t of the v e t t i n g process for the desire to A has been "go", and not "stop" . C. INTENTIONAL ACTIONS We can see that the concepts of intending and coming to intend are concepts which q u a l i f y mental events. I turn now to an analysis of i n t e n t i o n a l actions - where intention i s used to q u a l i f y actions - and the r e l a t i o n which holds between intending and i n t e n t i o n a l action. F i r s t , there i s the d i s t i n c t i o n between standard and non-standard actions. I claim that a standard action i s one that i s standardly caused by an agent, where t h i s e n t a i l s that the agent's int e n t i o n was causally e f f i c a c i o u s , i n the way described above, and successful. That i s , where an agent intends that such and so be the case, and where t h i s intention causes that such and so i s the case, then the agent has acted standardly. 1 6 1 believe we have the means here to sort out the problems of weakness of w i l l . I leave that task to a future paper. 72 Standard actions are actions which are caused, in the right way, by an agent's i n t e n t i o n . Intentional actions are standard actions, and acting i n t e n t i o n a l l y i s for one's intention to be causal and s u c c e s s f u l . This i s to say that there i s no d i f f e r e n c e between a c t i n g , a c t i n g i n t e n t i o n a l l y and acting standardly. We may put t h i s i n terms of the action schema. For some event to be an action of an agent i s for that event to be caused by an agent's intention where a l l the matchings are i n place: the content of the e f f i c a c i o u s desires are matched by the events-under-a-description, and the instrumental b e l i e f s are matched by the causal r e l a t i o n s between these. This entails that the instrumental beliefs were true. Therefore, an action i s an event-under-a-description which i s caused i n the right way by an agent's desires and true b e l i e f s . This i s the analysis of action. What i s ascribed to an agent i n an action a t t r i b u t i o n i s , f i r s t , that which s a t i s f i e s h i s e f f i c a c i o u s desires and b e l i e f s - the i n t e n s i o n a l item represented i n the agent's e f f i c a c i o u s cognition - and which the agent caused. Secondly, we a t t r i b u t e to the agent the mental causation of the event-under-a-description, which e n t a i l s that the schema i s i n order for the agent. A l l non-standard actions w i l l be those for which the process I have schematized has f a i l e d i n i d e n t i f i a b l e ways. We w i l l f i n d that we have, or could have, concepts for every type of aberration of the process of action. 73 In order to give an account of i n t e n t i o n a l action, we have found i t necessary to appeal, at every point, to the mens of the agent. This i s , I claim, because a c t i o n i s a diagnostic concept: action i s a concept which e n t a i l s a diagnosis of an agent - namely, hi s mental causation. I have argued that actions are ineluctably i n t e n t i o n a l , and therefore i n t e n s i o n a l . This i s to say that there can be no r e f e r e n c e to an a c t i o n t h a t i s not a r e f e r e n c e to the ineliminable i n t e n s i o n a l i t y of the content of the agent's mental causation. Therefore, a t t r i b u t i o n s of i n t e n t i o n a l i t y are not intensional, contra Davidson. Since action i s t i e d e s s e n t i a l l y to the description(s) under which i t was intended, any reference to the action i s a reference to those properties which the agent desired. Thus a t t r i b u t i o n s of intention and action are not i n t e n s i o n a l . Describe the action any way you l i k e , i f what you are describing i s the action, then i t i s what the agent desired to bring about, and believed..., etc. Action e s s e n t i a l l y i s , t h e r e f o r e , what the agent intended. Because a c t i o n s are e s s e n t i a l l y i n t e n s i o n a l , s u b s t i t u t i o n of n o n - e q u i v a l e n t descriptions of the action, i n action a t t r i b u t i o n s , is truth-preserving. This may explain how i t was that Davidson came to i d e n t i f y the wrong sort of thing as an action. An analysis of action d i s c o u r s e r e v e a l s t h a t t h e r e i s e x t e n s i o n a l i t y t h e r e . Davidson's conclusion i s that actions are extensional. His 74 strategy i s the same i n "Causal Relations", 3- 7 where the presence of extensionality i n causal discourse i s enough to j u s t i f y the c l a i m t h a t c a u s a t i o n i s an e x t e n s i o n a l r e l a t i o n between extensional events. But, that there i s extensionality i n action and causal discourse does not s e t t l e the matter as to the nature of the items involved. I return at l a s t to our agent 0 and the i d e n t i t y of his action. As I argued above, the intention with which 0 acted was the conjunction of the content of O's e f f i c a c i o u s desires. That i s , 0 intended to embarrass the government by releasing the secret papers by moving his hand i n just the way needed i n his circumstances. Or, 0 intended to move h i s hand i n ju s t the way needed so as to release the secret papers so as to embarrass the government. Now we can ask, with Davidson, whether the basic b o d i l y event was the acti o n . Well, under the d e s c r i p t i o n "moving the hand" i t i s not. 0 does not intend to move h i s hand s i m p l i c i t e r , i . e . , not under that l i m i t e d description. What he intends i s a bodily event with the r i g h t consequences. Nor does he intend the release of the secret papers, where t h i s i s treated extensionally. He intends to move his hand so as to release the secret papers so as to.... This brings out ju s t how misleading Davidson's claim i s that there are only primitive actions. Clearly, for creatures such as we, who can engage the world only with our bodies, actions w i l l , of necessity, require a basic bodily event. But t h i s i s not the action. Where the 1 7 Davidson, 149-162. 75 agent merely intends to move h i s body, the action w i l l be the bodily movement under the description i n his intention. For 0, Jbecause the event of his hand moving caused the event of the paper's release which caused the event of the government's embarrassment, we can describe his action as embarrassing the government. We can do t h i s because the corresponding event occurred i n the right way, which i s to say, among other things, that the embarrassing was i n the content of O's successful intention. Now s u p p o s e t h a t t he e v e n t o f t h e government's embarrassment caused the event of the government's changing p o l i c y . What makes i t false to at t r i b u t e the change i n p o l i c y to 0 as his action i s that nowhere i n his intention, l e t us suppose, does a d e s i r e occur f o r 1 that event. We cannot a t t r i b u t e to O's agency the change i n the government's p o l i c y although we can att r i b u t e i t to him causally. While intentions give us, as agents, a "way into" the world of events, j u s t how f a r i n t o that world our agency takes us, j u s t how far the concept of action goes down the causal chain, i s determined by the content of our successful intentions. Agency tr a v e l s down the causal chain of events j u s t as far as inten t i o n does. Of course, r e s p o n s i b i l i t y , sometimes being a matter of negligence, may t r a v e l further, but t h i s does not e n t a i l that agency does. Therefore, contra Davidson, each consequence of an action does not present us with a deed. 1 8 See above, p. 18. 76 With t h i s account of r a t i o n a l agency and therefore of action we are equipped to deal with some of the t r a d i t i o n a l problems of action theory. I turn, then, to applications of the theory to non-standard actions, the active-passive d i s t i n c t i o n , t r y i n g , and to a challenge to t h i s account from the l i t e r a t u r e . i CHAPTER III APPLICATIONS A. AGAINST THE SIMPLE VIEW As was noted above, Bratman has a view of i n t e n t i o n a l action which separates the content of the intention with which an action i s done from what i s done i n t e n t i o n a l l y so that one can A i n t e n t i o n a l l y without having intended to A so long as one intended to B for some appropriate B. This i s a weakening of the connection between intention and int e n t i o n a l action and so runs contrary to RAAG. Bratman defends t h i s separatist view of i n t e n t i o n a l action by an argument against what he c a l l s the "Simple-View", which states that: f o r me i n t e n t i o n a l l y to A I must intend to A; my mental states at the time of action must be such that A i s among those things I intend. 1 I have been arguing against t h i s separation of inten t i o n and in t e n t i o n a l action and so have been defending the Simple View. I w i l l give Bratman's argument against the Simple View and show why I f i n d i t unconvincing. 1Bratman, "Two-Faces," 377. 77 78 The argument consists of three examples the t h i r d of which i s meant to do the t r i c k . In case 1 we are asked to imagine someone, l e t ' s c a l l him M, playing a d i f f i c u l t video game which requires f i r i n g a "missile" at a target. M i s quite s k i l l e d at i t but nevertheless i s doubtful of success. He aims and f i r e s . As i t happens, he "succeeds", says Bratman, i n ju s t the way he was t r y i n g , which means: 1. h i t t i n g the target was what he wanted to do 2. the h i t t i n g depended on his s k i l l s and so was not a matter of luck 3. h i s perception of the h i t t i n g terminated his attempt (and so i t was not inadvertence) I t i s c l e a r to Bratman that even though M i s doubtful of success, and i s t r y i n g , that i f he "succeeds", he does so i n t e n t i o n a l l y . I f so, then on the Simple View he must have intended to h i t the target which, for the Simple View, i s a thoroughly acceptable r e s u l t . Now suppose that, i n case 2, a second game of t h i s type i s added, and our ambidextrous M plays them simultaneously. Again because they are equally d i f f i c u l t , he i s doubtful of success at eit h e r game, and equally doubtful since he i s equally s k i l l e d with eit h e r hand. Suppose he misses target T2 but "succeeds" i n h i t t i n g target TI, where conditions 1 - 3 above obtain. If so, says Bratman, he h i t s TI i n t e n t i o n a l l y . On the Simple View, M must have intended to h i t TI. I f so, then because the case i s e n t i r e l y symmetrical with respect to T2, he must have intended to h i t T2 as w e l l , even though t h a t i n t e n t i o n was not successful. So, on the Simple View, M intended to h i t each target. 79 F i n a l l y , i n case 3, we are asked to imagine that these two games have been linked so that i t i s impossible to h i t both targets; that i f both are about to be h i t , the game shuts down and the player looses, and a l l of t h i s i s known to M. S t i l l , there i s a reward for h i t t i n g either target and since i t i s d i f f i c u l t t o h i t e i t h e r , M d e c i d e s to p l a y both games simultaneously: he reasons that the r i s k of shutting down the machines i s outweighed by the increased chance of h i t t i n g a target. So he t r i e s to h i t target 1, and t r i e s to h i t target 2. Suppose that M " h i t s " target 1, where conditions 1 - 3 above apply. Therefore, just as i n the f i r s t two cases, he h i t s the target i n t e n t i o n a l l y , says Bratman. On the Simple View, he must have intended to h i t target 1. But given the symmetry of the case, M must have also intended to h i t target 2. And here i s the problem for the Simple View. Since M knew that he could not h i t both targets, i f he intended to h i t both, which he must have a c c o r d i n g t o the Simple View, he would be c r i t i c i z a b l y i r r a t i o n a l , s i n c e h i s i n t e n t i o n s would f a i l one of the r a t i o n a l i t y constraints on intentions. I t i s a requirement of r a t i o n a l i t y that intentions be strongly consistent: My i n t e n t i o n s are strongly consistent relative to my beliefs i f a l l my intentions could be put together i n t o an o v e r a l l plan that i s consistent with those b e l i e f s . 2 Since M f a i l s t h i s strong consistency requirement, i f the Simple View i s r i g h t , he i s g u i l t y of i r r a t i o n a l i t y . But Bratman sees 2 Ibid., 380. 80 no reason to so charge the player. The strategy he employed maximized h i s chances of winning given the d i f f i c u l t y of h i t t i n g e i t h e r target. Therefore, the Simple View i s f a l s e : M did not have both the intention to h i t TI and the intention to h i t T2. But i f so, then since the case i s symmetrical with respect to both targets, M had neither intention, and so the Simple View i s f a l s e : The Simple View imposes too strong a l i n k between i n t e n t i o n and i n t e n t i o n a l action, a l i n k that i s i n s e n s i t i v e to differences i n the demands of p r a c t i c a l reason. 3 I think that some of Bratman's premises are true but that h i s conclusions are f a l s e . In arguing against Bratman I w i l l accept the strong consistency requirement, since intentions are r a t i o n a l items, and I w i l l agree that the game player i n t h i s t h i r d case was not g u i l t y of a form of i r r a t i o n a l i t y . A l l the same, I defend the Simple View, which i n consequence I rename the Unified View. The force of Bratman's argument against the Un i f i e d View turns on h i s claim that M h i t s TI i n t e n t i o n a l l y i n case 3. If M doesn't do i t i n t e n t i o n a l l y , then of course there i s no challenge to the Unified View. I s h a l l argue that we have some good reason to deny i n t e n t i o n a l i t y to the h i t t i n g of TI i n case 3, and that the consequences of such a denial are not as serious (or as counterintuitive) as at f i r s t they may seem. According 3 Ibid., 383. 81 to RAAG, M d i d not h i t T l i n t e n t i o n a l l y i n case 3. And, applying the theory and the schema w i l l show why and how we may withhold the a s c r i p t i o n of i n t e n t i o n a l i t y to cases 1 and 2. To begin to cast case 3 i n terms of the action schema i t w i l l be h e l p f u l , and I think unobjectionable, to say that M i s no longer playing two games. He i s playing and t r y i n g to win one game, the rules of which Bratman has provided. We assume that M desires, d Q, to win and believes that to win i s to h i t T l or T2 but not both, and so desires, d x, to h i t T l or T2 but not both. M comes to intend to do t h i s at dx , step 3. Since M i s a r a t i o n a l system, the a r r i v a l of dx at step 3 - the intention-prompts M to search his b e l i e f s about means for how to achieve such a h i t , step 4. Now, M believes that i n order to win he must h i t either T l or T2 but not both. But, since h i t t i n g e i t h e r i s d i f f i c u l t for M, he i s doubtful of h i t t i n g either. Therefore, i t i s r a t i o n a l for him to believe that the best means available to him to s a t i s f y d x i s to try to h i t both T l and T2. Furthermore, he believes, that to t r y to h i t T l and T2 i s to aim and f i r e the guns at the targets and so believes, bx , that to do so i s the means. At step 8, he comes to intend to so aim and f i r e . So far, there i s no v i o l a t i o n of the requirement of strong consistency on intentions, since he does not intend two incompatible actions, and does not have inconsistent intentions: he can intend to try to h i t both targets without inconsistency since intending to t r y i s not to intend to do. Next, M asks for means again, and believes, b 2 , that to move his body i n some ce r t a i n way, i n h i s circumstances, w i l l cause the guns to be 82 aimed and f i r e d at the targets, and comes, at step 13, to intend, d B, to do so. The basic desire, d B , causes a chain of events under the following descriptions: M's body moves, E B, i n ju s t the way described i n d B, which causes the guns to be aimed and f i r e d , E 2, which causes a missi l e to h i t TI, and another to miss T2. The schema of t h i s i s as follows: F i g . 13. RAAG PLAYING THE VIDEO GAME The Mental The World of External Events •S d x : the desire to h i t TI or T2 but not both bt : the b e l i e f that to aim and f i r e the guns at the targets i s a means to s a t i s f y i n g d x d 2 : the desire to t r y to h i t TI and T2, i . e . , to aim and f i r e at both b 2 : the b e l i e f that to move h i s body i n ju s t the r i g h t way w i l l aim and f i r e the guns at the targets d B : the desire to do so (as i n b 2) move h i s body E B : M's body moving i n the way he desired d B E 2 : the guns being f i r e d at both targets E x : the event of TI being h i t and T2 not 83 Now, according to the Unif i e d View M h i t s TI i n t e n t i o n a l l y only i f he intends to h i t TI, which causes the h i t t i n g . But M does not so intend. He intends, d x , to h i t TI or T2 but not both which i s not to intend to h i t TI. And, M intends to try to h i t both TI and T2, which again i s not to intend to h i t TI. Therefore, on the Un i f i e d View, M does not h i t TI i n t e n t i o n a l l y . Nor does M h i t T l i n t e n t i o n a l l y according to RAAG. What M does i n t e n t i o n a l l y i s what i s caused i n the r i g h t way by h i s e f f i c a c i o u s intention. At d B , step 13, M intends to h i t T l or T2 and not both by t r y i n g to h i t both T l and T2 by moving his body i n ju s t the way he believes w i l l do the t r i c k . A l l of d x, d 2 , and d B are matched by E's which satisfy them, and s i m i l a r l y for bx and b 2 . Therefore, we can say of M that he intentionally hit Tl or T2 but not both. To deny that M h i t T l i n t e n t i o n a l l y i s not to deny that the h i t t i n g was the r e s u l t of M employing his considerable s k i l l s . Nor does i t e n t a i l that M does not deserve c r e d i t f o r h i t t i n g T l . The claim of unin t e n t i o n a l i t y i s the claim that the action has deviated from the standard. The standard case of h i t t i n g T l would be where the agent intends to h i t i t and believes that such and so i s the means to h i t t i n g i t , where these comprise an int e n t i o n which i s successfully e f f i c a c i o u s of an event of the mi s s i l e h i t t i n g T l . Case 3 i s a deviation of the standard i n two respects: f i r s t M has no such i n t e n t i o n ; second, M's instrumental b e l i e f s are not functioning standardly, since he i s doubtful of success. I turn to t h i s issue i n the next section. 84 I t w i l l be apparent that the method employed here i s i n sharp contrast to Bratman's strategy. The strategy of taking agency as basic, as the place to begin sorting out hard cases, i s c l a r i f y i n g and so helps sort out c o n f l i c t i n g i n t u i t i o n s . B. TRYING Given the above analysis of case 3 what should we say of M while he was en acte? Since he i n t e n t i o n a l l y h i t T l or T2 but not both, we can say that he was hitting T l or T2 but not both. We can also say of him that he was trying to h i t T l and trying to h i t T2. And M would agree that at the time he was trying to h i t T l and not hitting T l , since he was doubtful of success. To report that one i s t r y i n g i s to report that e f f i c a c y i s "going through" despite one's lack of confidence i n one's means. Thus we have an a n a l y s i s of a t r y i n g which admits s e v e r a l d i s t i n c t i o n s . F i r s t , there i s a t r y i n g which may or may not be a doing. Where an instrumental b e l i e f i s tentative and yet e f f i c a c y proceeds despite the lack of confidence, the agent comes, at step 6 or 11, to a "doubtful desire" for the means. This i s a case of tentative cognition. Furthermore, where an instrumental b e l i e f i s tentative, i t s influence on the process of coming to intend may y i e l d tentative causation, where the agent engages the world h a l t i n g l y . Secondly, there i s a t r y i n g that i s not a doing, where to t r y i s to fail by having one's d e s i r e s be e f f i c a c i o u s under the impairment of a false instrumental b e l i e f . When someone intends to A but mistakenly 85 B's because of a f a l s e instrumental or i d e n t i t y 4 b e l i e f , we can say of him neither that he was A-ing, nor that he was B-ing; what we can say of him i s that he was t r y i n g to A but mistakenly B-ed. S i m i l a r l y , when someone intends to A but a c c i d e n t a l l y B's because of a f a l s e b e l i e f , or lack of relevant true b e l i e f , we can say of him that he was t r y i n g to A. These cases of t r y i n g share a d i m i n u t i o n e i t h e r of c o g n i t i o n or of e f f i c a c y : cognition may be "on" but tentative; e f f i c a c y may be "on" but tentative; e f f i c a c y may be "on" but impaired. Therefore, we ought to be reluctant to say of someone who i s i n t e n t i o n a l l y A-ing that he i s trying to A. He i s not t r y i n g ; he i s doing. Since M was doubtful of success i n both cases 1 and 2, we can say of him that he was t r y i n g to h i t T l . Assuming that M's t r y i n g i s a case of tentative cognition, i t i s a deviation from the standard. Therefore, we have reason to deny that M h i t T l i n t e n t i o n a l l y i n cases 1 and 2. I t i s c l e a r to Bratman's i n t u i t i o n s , but not so c l e a r to mine (and I don't think t h i s should count for much), that M does h i t T l i n t e n t i o n a l l y i n cases 1 and 2. But i t i s not enough to guarantee i n t e n t i o n a l i t y that the target was h i t , since t h i s fact does not guarantee that the h i t t i n g was due to agency. We cannot j u s t assume success. Where i t turns out that the agent's b e l i e f s were true and that therefore e f f i c a c y was successful i n j u s t the way the agent envisaged, i . e . , where Bratman's condition 2 obtains, we would a t t r i b u t e i n t e n t i o n a l i t y . Such a case would be a t r y i n g which 4 See below, sec. C ( l ) . 86 was a doing. But where a l l we know i s that the agent was tr y i n g to achieve a r e s u l t which was forthcoming, the question of i n t e n t i o n a l i t y i s inconclusive. With the i n c o n c l u s i v i t y of t r y i n g we can a t t r i b u t e not u n i n t e n t i o n a l i t y , but non-intentional ity . On t h i s a n a l y s i s , to report that you are t r y i n g i s to report on the i n c o n c l u s i v i t y of your instrumental b e l i e f s . To a t t r i b u t e a t r y i n g t o a n o t h e r i s t o a t t r i b u t e e i t h e r i n c o n c l u s i v i t y due to a "doubtful desire", or non-efficacy due to a f a l s e instrumental b e l i e f . On the theory of r a t i o n a l agency which I am presenting, i n order for an action to be i n t e n t i o n a l , i t i s s u f f i c i e n t and necessary that agency functioned i n an appropriate way i n the causal genesis of some correspondingly appropriate event, where the appropriate way i s for the agent's desires and true b e l i e f s to function causally i n the process of vetting and adjudication which I have outlined. In standard cases of action, the process w i l l i n c l u d e , among o t h e r t h i n g s , i n s t r u m e n t a l b e l i e f s functioning i n t h e i r p a r t i c u l a r way. What Bratman has described i n cases 1 and 2 i s something less than t h i s . In cases 1 and 2 the player does not believe that he w i l l be successful since he doubts h i s means. But, since h i s desire to win i s strong enough, he chooses the best means available which i s to employ hi s considerable s k i l l s : he intends to t r y . In other words, the instrumental b e l i e f s which would otherwise appear i n an action schema do not enjoy M's f u l l confidence: therefore, the 87 diagnosis we must make of M's cognitive e f f i c a c y i s that i t has not functioned standardly. So, while M intends to h i t T l , and while h i s instrumental b e l i e f s turn out to be true, so that his e f f i c a c i o u s d e s i r e s are matched by states of a f f a i r s which s a t i s f y them, and h i s b e l i e f s are matched by the re l a t i o n s between these states of a f f a i r s , we may deny i n t e n t i o n a l i t y , so long as doubt remains, as a recognition of the i n c o n c l u s i v i t y which accompanies t r y i n g s . The importance of marking t h i s a l t e r a t i o n of agency w i l l be reinforced when we come to discuss excuses and r e s p o n s i b i l i t y , where the diagnosis of the mind of the agent i s of central importance. 5 Support for t h i s view of t r y i n g as a cognitive concept comes from the fact that i n Bratman's case 3, M can intend to t r y to h i t T l and T2 without v i o l a t i n g the c o n s i s t e n c y requirement on intentions. The intention to t r y to h i t T l i s not incompatible with the intention to t r y to h i t T2, given M's b e l i e f s about the i m p o s s i b i l i t y of h i t t i n g both T l and T2, since to intend to t r y to do both i s not to intend to do both, nor i s i t to desire to do both, nor to believe that you w i l l do both. Trying i s the f i r s t of the ways I w i l l examine i n which the left-hand side of the schema - the agent's mens - can be out of order. The above a n a l y s i s of Bratman's case a p p l i e s , mutatis mutandis, to Davidson's claim that one can do i n t e n t i o n a l l y what 5 This analysis of t r y i n g can be applied with some success to Analysis "Problem" No. 16. See Analysis 38, 113. 88 he doubts he i s doing. 6 In h i s example, a man, i n writing heavily on h i s page, i s intending to make 10 carbon copies. He does not know or believe with any confidence t h a t he i s succeeding. But, since t h i s i s what he wants to do, then, i f he does "produce" 10 copies, he does so, say Davidson and Bratman, i n t e n t i o n a l l y . On my analysis, he i s tr y i n g to make 10 copies, and again, the question of i n t e n t i o n a l i t y i s inconclusive. The r e s u l t we may c a l l good f o r t u n e , and h i s a c t i o n a non-int e n t i o n a l making of 10 copies. Part of the purpose to which Davidson puts the above example i s to show that one can do i n t e n t i o n a l l y what one does not believe one i s doing. Therefore, there i s no such b e l i e f requirement on a c t i n g i n t e n t i o n a l l y . But i f one can act i n t e n t i o n a l l y without bel i e v i n g that one i s , then, say Davidson and Bratman, one can intend to A without b e l i e v i n g that one w i l l A.7 This i s to argue against what Bratman c a l l s the strong belief requirement on intending to A. According to the weaker version of t h i s thesis, i f one intends to A, then one must believe that one w i l l A. On the stronger version, to intend to A is to believe that one w i l l . The thesis i s meant to explain the oddness of remarks l i k e the following: "I intend to go to the concert, but I may not go." 8 Davidson and Bratman take 6 Davidson, 50, 60, 91 f f . 7 Ibid., 90 f f . ; Bratman, 383 - 385. 8H.P. Grice, "Intention and Uncertainty, " Proceedings of the British Academy 57 (1971): 263 - 279. 89 examples l i k e the above to defeat both versions of the strong b e l i e f requirement on intending and doing. The view I have been arguing for may seem to include the strong b e l i e f requirement, but i t does not. I have argued that i n order to act i n t e n t i o n a l l y , one's instrumental b e l i e f s must be true and function for the agent i n a standard way. This e n t a i l s that the agent believe that he has the means to his goal. In a standard case of h i t t i n g the video game target, the shooter believes that i f he moves his body i n j u s t the way he has i n mind, the m i s s i l e will h i t the target. But t h i s i s not to believe that he w i l l h i t the target, although t h i s l a t t e r b e l i e f i s e n t a i l e d by the former. Therefore, my requirement on i n t e n t i o n a l i t y i s weaker than e i t h e r v e r s i o n of the strong b e l i e f requirement. The entailment r e l a t i o n i s important for another puzzle of agency. If someone intending standardly to A were to r e f l e c t on his state of mind, he might come to believe that he w i l l be s u c c e s s f u l . Therefore, i f he were to p r e d i c t h i s future behaviour, he would say that he w i l l A. This explains the oddness of one saying "I intend to A but I may not", an oddness which led some the o r i s t s to the strong b e l i e f requirement on intending to A. Since intending implies confidence i n success, and commitment to action, intending to A e n t a i l s the b e l i e f that one w i l l A. To say that you intend to A when you also believe that you may not be successful, either because you believe you cannot, or are not able, i s to mislead the hearer. 90 C. NON-STANDARD ACTIONS AND EXCUSES I have j u s t given an account of one form of aberration i n r a t i o n a l agency which we c a l l a t r y i n g which i s not a doing. It w i l l be a further test of the theory of r a t i o n a l agency that i t f a c i l i t a t e s the leg a l and moral d i s t i n c t i o n s between standard and non-standard actions. Non-standard actions are those cases i n which excuses, and so defenses, apply. 9 Here I deal with only some of these d i s t i n c t i o n s . 1. Mistake As I noted above, a mistaken action i s a case of cognitive e f f i c a c y "going through" under the impairment of a f a l s e instrumental b e l i e f . Since i t i s an impairment, mistakes are unintentional. I t w i l l be help f u l to show how t h i s comes out i n terms of the action schema. So we take the following example: 1 0 Suppose that I desire dx , to meet Groucho Marx and come, at step 3, to intend to do so. I believe that Groucho i s next door and so believe, bx , that going there i s a means to meeting him, and therefore desire, d 2 , to go next door, and come, at step 8, to intend to do so. I believe, b 2, that i f I perform the fa m i l i a r movements I w i l l go next door and come, at step 13, to intend, d B , to so move. At d B , I intend to meet Groucho by going next door by moving i n the old f a m i l i a r ways. This "intention" i s e f f i c a c i o u s of the following events under these descriptions: 9S.C. Coval and J.C. Smith, Law and Its Presuppositions (London: Routledge and Kegan Paul, 1986), Chaps. 1 and 2. 1 0 This i s an example of Professor Coval's. E B , the event of my moving i n the way I have i n mind i n d B which causes the event, E 2, of my a r r i v i n g next door which causes the event, Ex , of my meeting my neighbour. F i g . 14. RAAG MAKES A MISTAKE The Mental <b > • m <b> V 1 4 The World of External Events E B —>E2—yE1 the desire to meet Groucho the b e l i e f that going next door i s a means to meeting Groucho The desire to go next door the b e l i e f that walking next door i s a means to going next door the desire to move i n that way necessary to walk next door the event of my walking that way the event of my a r r i v i n g next door the event of my meeting my neighbour In t h i s case bx i s f a l s e . The intention with which I act i s the inten t i o n to move i n ju s t that way i n order to go next door i n order to meet Groucho. But because bx i s false, my goal i s not forthcoming. That i s , not only do I not meet Groucho, but I do not move i n just that way i n order to go next door i n 92 order to meet Groucho. Therefore, neither d x , d 2 nor d B i s s a t i s f i e d , nor are b x and b 2 s a t i s f i e d . Consequently, what I do i n t h i s case i s mistakenly meet my neighbour. Since i t was the false b e l i e f , b± , that, together with d x , caused me to desire, d 2 , to go next door, I do not t r u l y desire to go next door. It i s a " f a l s e d e s i r e " . I f so, then d 2 at step 8 i s not an intent i o n . I t i s a mistntention. On Davidson's analysis of mistakes, making a mistake i s doing something else i n t e n t i o n a l l y . 1 1 That i s , there w i l l be a d e s c r i p t i o n o f what the agent does under which i t i s in t e n t i o n a l . For RAAG t h i s i s f a l s e . Now i t may seem that E B the event of my moving i n some way, i s i n t e n t i o n a l . I t would be i f I desired to move i n that way s i m p l i c i t e r . But I do not. The content of my intention at d B , step 13 i s not j u s t to move my body i n some way. It i s to move i n some way so as to arrive next door so as to meet Groucho. But to move the way I do i s not to move so as to come to meet Groucho. Therefore i t i s u n i n t e n t i o n a l : i n mistakes the b a s i c b o d i l y movement so described i s u n i n t e n t i o n a l . 1 2 The reason we recognize mistake as an excuse i s obvious, since making a mistake i s not for agency to have functioned standardly. If not, one ought not be held responsible for one's 1 1 Davidson, 45. 1 2 This analysis of mistake applies to Davidson's examples, which include s p i l l i n g the coffee, the o f f i c e r s i n k i n g the Bismark, and Hamlet k i l l i n g Polonius. Davidson, 44 - 45. 93 mistakes to the degree to which we are i n standard cases, since how agency f u n c t i o n s i n the non-standard case i s e n t i r e l y d i f f e r e n t than i n the standard case. Where some standard of care applies which you do not meet and should have met i n that you should have been more careful about your epistemology, about which b e l i e f s you have, and, therefore, about which b e l i e f s you allowed to function i n consort with desires i n action, then you are c r i t i c i z a b l e for the breach of the standard, and may be held responsible for the consequences. This i s t y p i c a l l y the case where the consequences of acting out of f a l s e b e l i e f s are negative. To f a i l to meet a standard of care i n the above way i s one way of being negligent. It i s to be negligent with respect to your b e l i e f s . 2. accident With m i s t a k e , what c a n c e l s agency i s the f a l s e instrumental b e l i e f . In accident, what does the c a n c e l l i n g i s the intervention of another cause. As Davidson says " . . . i f I s p i l l the coffee because you j i g g l e d my hand, I cannot be c a l l e d the agent." 1 3 The diminution of agency involved i n accidents i s eit h e r a lack of true instrumental b e l i e f s , or the presence of f a l s e instrumental b e l i e f s . In the above example, I may either not foresee that you w i l l j i g g l e my hand, or I may f a l s e l y believe that, e.g., i f I hold my cup away from you, the coffee won't get s p i l l e d . In either case I cannot be c a l l e d the agent of the s p i l l i n g , since I did not intend i t : there i s no desire 3 Davidson, p. 45. 94 to s p i l l the coffee, and no b e l i e f that i t w i l l s p i l l . To i l l u s t r a t e , I w i l l schematize the following example. 1 4 Suppose I intend to shoot my donkey and believe that aiming and f i r i n g the gun at him w i l l shoot him. But, at the moment of the shooting, my neighbour's donkey steps into the l i n e of f i r e and t r a g i c a l l y i s shot. In t h i s case I a c c i d e n t a l l y shoot my neighbour's donkey, since I lack both the b e l i e f that he w i l l intercept the f i r e and the desire to shoot him. This d i f f e r s from mistakenly shooting him, which would be to f a l s e l y believe, for example, that that donkey i s mine, and not my neighbour's. But as with mistake, i n accident none of the e f f i c a c i o u s desires and b e l i e f s are s a t i s f i e d . This i s an example of J.L. Austin's. F i g . 15. RAAG ACCIDENTALLY SHOOTS HIS NEIGHBOUR'S DONKEY The Mental <b> <b> 1 4 d R—> The World of External Events 71 Ei E B—>E' the desire to k i l l my donkey the b e l i e f that to aim and f i r e the gun at him w i l l k i l l him the desire to aim and f i r e the gun at him the b e l i e f that to do such and so with my body w i l l aim and f i r e the gun the desire to move my body the event of my body moving the event of the gun f i r i n g the event of my neighbour's donkey being shot 3. Compulsion A compulsion i s a desire whose e f f i c a c y i s so powerful that i t can override the r a t i o n a l processes of agency. That i s , at the point i n the r a t i o n a l process of coming to intend where desires are vetted by b e l i e f s , a compulsive desire can proceed into the further channels of agency despite an " o f f " i n the form of a b e l i e f that acting so as to s a t i s f y the desire w i l l be, on the whole, damaging to the system, where t h i s means that more important d e s i r e s of the system w i l l be f r u s t r a t e d . A compulsive action i s what re s u l t s from t h i s sort of aberration i n the system. Since agency has broken down i n t h i s way at that point i n the process where desires are r a t i o n a l i z e d , compulsions do not acquire the status of intentions, and compulsive actions are not i n t e n t i o n a l actions. The schema for RAAG explicates the phenomenon of compulsion. There are two related phenomena which can occur at the point i n the process where compulsions v i o l a t e RAAG. The f i r s t i s the case i n which a desire makes i t s way past the vetting process at step 2 not because i t i s too powerful to hold i n check but because the other desires and/or b e l i e f s of the system are too weak. The second case i s one i n which a desire becomes e f f i c a c i o u s , not through r a t i o n a l i t y , but because the system lacks further desires with which the f i r s t competes. These may be two ways i n which a system can be immature. Since they are both a b e r r a t i o n s of RAAG, they may j u s t i f y the d e n i a l of i n t e n t i o n a l i t y and r e s p o n s i b i l i t y . 4. Involuntariness As I remarked i n the previous c h a p t e r 1 5 the theory of r a t i o n a l agency i s also a theory of autonomous r a t i o n a l agency. Autonomy i s f o r RAAG to function standardly unimpeded from without. In p a r t i c u l a r , i t i s for the f i r s t stage to occur, where d.y "drops down" to engage the rest of agency, without interference, either from the environment i n which the system 1 5 See above, p. 67. 97 finds i t s e l f , or from another agent. With such interference we have a case of involuntariness, of which we may d i s t i n g u i s h two types. The defense of necessity applies to cases i n which the environment i s such t h a t i t f o r c e s a d e s i r e to become e f f i c a c i o u s which would otherwise not have been. Given the set of desires and b e l i e f s you have, the world may l i n e up i n such a way that you are forced to do something you would otherwise not have done, i n order to preserve some greater good, for example your l i f e or that of another. Throwing the cargo overboard i n a storm i n order to keep from sinking i s a case of necessity, and would be excused as such. The second type of involuntariness i s for another agent to force you to do something that you would otherwise not have done. We may d i s t i n g u i s h t h i s from necessity by c a l l i n g i t a case of coercion. To be coerced i s to be forced to have a desire become e f f i c a c i o u s , given the rest of your desires and b e l i e f s , which would otherwise not have been e f f i c a c i o u s . For example, the gunman forces you to desire to give him your wallet, and you do desire t h i s , but only because you believe that i f you do not comply, you w i l l be harmed. Given your desire to l i v e , and your desire not to be harmed, you come to desire to hand him your wallet since you believe that that i s a means to s a t i s f y i n g those more important of your desires. I t i s important to notice that i n contrast to compulsion, necessary actions and coerced actions are intentional actions, despite being involuntary. They count as in t e n t i o n a l , for RAAG, 98 since the system functions as i t should. You intend to throw the cargo overboard, and you intend to hand over your wallet. Given the s i t u a t i o n s i n which you f i n d yourself, these are the r a t i o n a l things to do. There i s no diminution of agency due to f a l s e b e l i e f s or compulsive desires, but only that your options are so severely limited, that you are forced to act i n these p a r t i c u l a r ways. Your defense i s that you would not have so acted had the probable consequences of not doing so not have been so d i r e . D. THE ACTIVE - PASSIVE DISTINCTION So far, RAAG has provided an account of what i t i s for agency to function standardly i n the causal genesis of some state of a f f a i r s : we have standard actions. RAAG has also demonstrated some of the ways i n which agency can function non-standardly i n the causal o r i g i n of an event: we have tryings, mistaken actions, accidents, etc.. These cases have themselves been distinguished by the d i f f e r e n t ways i n which agency has functioned non-standardly. There are several other ways for agency to be involved i n the ancestry of an event which i s not an action but for which an agent may be held responsible to some degree. These cases are sometimes i d e n t i f i e d by the active-passive d i s t i n c t i o n . In t h i s section I w i l l analyze several ways of being passive. As w i l l become apparent, the term "passive" i s misleading. Some of the cases I w i l l describe which do not count as actions include several of the components of agency, and others w i l l include agency, although not of the events i n question. The purpose of the analysis i s that with 99 RAAG we can d i s t i n g u i s h several ways of being "passive". Since we can d i s t i n g u i s h several ways i n which agency can function i n the ancestry of an event which i s not an action, we have a way of a s s i g n i n g v a r y i n g d e g r e e s o f c u l p a b i l i t y f o r the corresponding event. We should expect that for every d i f f e r e n t mens we can diagnose, we could assign a d i f f e r e n t degree of r e s p o n s i b i l i t y , and do so with good reason. The case I w i l l use i s the f a m i l i a r k i l l i n g - l e t t i n g die d i s t i n c t i o n . According to RAAG, i n order to count as a k i l l i n g , i t i s both s u f f i c i e n t and necessary that the agent desire the death and believe that doing such and so w i l l cause the event of the death, where the intention which i s comprised of these desires and b e l i e f s i s e f f i c a c i o u s , i n the r i g h t way, of the event of the death. To begin to set up the contrast with l e t t i n g die, we s h a l l c a l l k i l l i n g the d e l i b e r a t e p r o v i s i o n of c a u s a l s u f f i c i e n c y , i n the circumstances, for the death. It i s necessary i n order to count as a case of l e t t i n g die that the agent be able to prevent the death and that he believe t h i s or should have believed i t . Suppose that an agent could have prevented a death but believed, on the best available evidence, that he could not, having met the required standards of care with respect to hi s b e l i e f s . If i t i s not the case that he should have believed i t possible to prevent the death, then we would not c a l l t h i s a case of l e t t i n g die. 100 A case of l e t t i n g die may be c a l l e d an act of omission. To ascribe to an agent an act of omission i s to describe h i s agency i n terms of a r e l a t i o n that i t should have had and didn't. That i s , i t i s to say of the agent that he should have acted i n some way which he did not and should have done so because of the appropriate desires and b e l i e f s which he should have had. We can d i s t i n g u i s h two broad categories of l e t t i n g die according to the presence and absence of the following desire and b e l i e f : (1) the desire for the death (2) the true b e l i e f that there i s , or w i l l be or could be a dying and so a death To count as a case of Intentional letting die, (1) i s necessary. Without the desire for the death, a l l other cases of l e t t i n g die are cases of negligent letting die. Within these two categories, we can d i s t i n g u i s h differences based on the four possible combinations of t h i s desire - b e l i e f pair, (1) and (2). 1. Intentional L e t t i n g Die a. The straightforward case (1) desire for the death (2) b e l i e f that death i s imminent This i s the case we may c a l l i n t e n t i o n a l l e t t i n g die where 101 the agent desires the death and believes that the s i t u a t i o n as i t i s i s s u f f i c i e n t f or the death. Since he desires the death, the agent does nothing to i n t e r f e r e with the causal s u f f i c i e n c y , e i t h e r by standing fast with the purpose of not i n t e r f e r i n g or by doing something e l s e . In such cases, the agent may be c r i t i c i z e d f or having bad desires, or for f a i l i n g to intervene to prevent the death. To take an example 1 6, suppose that you are canoeing downstream toward the drowning man. You would l e t him die in this way i f either of the following i s true: you are paddling and you believe that one more stroke w i l l bring you within grasp and so, since you desire h i s death, you do that with your body which w i l l not i n t e r f e r e with the present causal s u f f i c i e n c y . Or the above i s true, and you paddle i n some way other than the way which would have l e t him save himself. This case of l e t t i n g die i s distinguished from the next by the b e l i e f here that non-interference i s necessary i n the circumstances for the death. b. The hard case (1) desire for the death (2) b e l i e f that death i s not quite imminent In t h i s case, the agent desires the death but believes that causal s u f f i c i e n c y does not yet ex i s t for the death, and that agency i s required i n order to achieve the s u f f i c i e n c y . In our example, such a case would be where you believe that should you not paddle, or act i n some other relevant way, the drowning man w i l l come within reach of your canoe and so save himself. Since 1 6 This may be Dick Sikora's example. 102 you desire the death, you would l e t him die i n t h i s way i f you, for example, take one more paddle stroke which you believe i s necessary i n the circumstance i n order for there to be a causal s u f f i c i e n c y for the man's death. Thus, t h i s type of l e t t i n g die d i f f e r s from the former i n that i n the former non-interference i s necessary, while i n the l a t t e r interference i s necessary. In t h i s case, the agent i s c r i t i c i z a b l e for bad desires, and for acting so as to contribute to the cause of the event. To be g u i l t y of t h i s i s to be g u i l t y of a more serious offence than i n the former case, since agency i s employed here i n a way necessary for the death. 2 Negligent Let t i n g Die a. Pure Indifference (1) no desire for the death (2) b e l i e f that death i s imminent The type of l e t t i n g die which we may c a l l pure Indifference i s characterized by the b e l i e f that, e.g., the death w i l l occur, where such a b e l i e f does not " c a l l up" or activate a desire to prevent the death by employing agency. That i s , the b e l i e f does not function instrumental l y for the agent. One g u i l t y of pure in d i f f e r e n c e i s c r i t i c i z a b l e , not for having bad desires, but f o r not having the r i g h t d e s i r e s , given h i s b e l i e f s , and therefore i s c r i t i c i z a b l e for not acting so as to prevent the death. b. (1) no desire for the death (2) f a l s e b e l i e f that death i s not imminent 103 It may seem that i n order to count as a case of l e t t i n g die, i t i s necessary that the agent believe that the death i s imminent. But an agent can be g u i l t y of a negligent l e t t i n g die when he does not meet some standard of care with respect to his b e l i e f s which i s required of him. That i s , he does not believe that death i s imminent and should have so believed, perhaps because he has been charged with that r e s p o n s i b i l i t y . In our example, an i n a t t e n t i v e l i f e g u a r d or river-keeper would be c r i t i c i z a b l e f or the state of h i s cognition, for not having the ri g h t b e l i e f s , the assumption being that, had he known of the drowning, he would or should have acted so as to prevent i t . c. (1) desire the death (2) f a l s e b e l i e f that death i s not imminent In t h i s case of l e t t i n g die, the agent i s negligent with respect to the state of his cognition since he f a l s e l y believes that death i s not imminent and should have known. In contrast to the previous case, the agent desires the death and so had he t r u l y believed that death was imminent he would have approved and not acted so as to prevent the death. He i s therefore c r i t i c i z a b l e f or not having the r i g h t b e l i e f and f o r having the wrong desires. The presence of the desire for the death makes cases of type 3 prima facie more morally repugnant than those of type 1 and 2. We may i d e n t i f y a variant of type-a negligence which counts as a defence against the charge of l e t t i n g die. Where an agent does not d e s i r e the death and does believe that death i s 104 imminent but does not act so as to prevent the death, then depending on h i s reasons, he may be excused. Where the cost to the agent would be so great that he could not be expected to intervene to save the drowning man, he may be excused from the charge of l e t t i n g die. Where the agent f a i l s to intervene out of fear, c u l p a b i l i t y may be reduced. In a l l cases of l e t t i n g die the ro l e which agency plays i n the ancestry of the death i s less contributory to the death than i s a case of k i l l i n g . Therefore, l e t t i n g die i s less morally repugnant than i s k i l l i n g . Even i n the problematic case, where the agent acts so as to prevent an event which would have prevented the death, the agent's behaviour i s less repugnant than a case of k i l l i n g not because of any difference i n desire between the cases, but because he employed agency i n a very d i f f e r e n t way. To k i l l i s to contribute the causal s u f f i c i e n c y for the death. To l e t die i n the problematic case i s to do what i s n e c e s s a r y i n the circumstances i n order t h a t c a u s a l s u f f i c i e n c y e x i s t s . The difference can be brought out i f we imagine two people, the f i r s t of which goes about producing harmful events, and the other which goes about looking for imminent harmful events and either ensures that they are not i n t e r f e r e d with or ensures that they occur by removing an impediment. The c e n t r a l notion of RAAG, which i s that action i s a diagnostic concept, has allowed us to di s t i n g u i s h several types of l e t t i n g die with t h e i r accompanying differences i n moral 105 repugnancy. In a l l of the above non-standard cases, we have been able to make and maintain fine-grained d i s t i n c t i o n s i n the mental c a u s a t i o n i n v o l v e d . Since a c t i o n s are cases of efficacious m e n t a l c a u s a t i o n , w i t h t h e s e f i n e - g r a i n e d d i s t i n c t i o n s we have a foundation for inferences to what was done, and therefore to what one i s responsible f o r . BIBLIOGRAPHY Anscombe, G.E.M.. Intention. Oxford: B a s i l Blackwell, 1957. A u s t i n , J.L.. "A Plea f o r Excuses." Philosophical Papers, edited by J.O. Urmson and G.J. Warnock. Oxford: Oxford University Press, 1961. Bratman, Michael. "Intention and Means-end Reasoning." The Phil. Review 90 (1981): 252 - 265. . "Taking Plans Seriously." Social Theory and Practice 9 (1983): 271 - 287. . "Two Faces of I n t e n t i o n . " The Phil. Review 93 (1984) : 375 - 405. . "Davidson's Theory of Intention." In Essays on Davidson, ed. Bruce Vermazen and M e r r i l l B. Hintikka. Oxford: Carendon Press, 1985. B u t l e r , Ronald J. . "Report on Analysis 'Problem' No. 16." Analysis 38 (1978): 113 - 118. Chisholm, Roderick M. . "The Structure of States of A f f a i r s . " In Essays on Davidson, ed. Bruce Vermazen and M e r r i l l B. Hintikka. Oxford: Clarendon Press, 1985. C h u r c h l a n d , P a u l . "The L o g i c a l C h a r a c t e r o f A c t i o n -Explanations." The Phil. Review 79 (1980): 214 - 236. Coval, Sam and D.D. Todd. "Adjusters and Sense-Data." American Phil. Quarterly 9 (1972): 107 - 112. Coval, S.C. and J.C. Smith. Law and Its Presuppositions. London: Routledge and Kegan Paul, 1986. Danto, Arthur C. . "Basic Actions." American Phil. Quarterly 2 (1965): 141 - 8. Davidson, Donald. Essays on Actions and Events. Oxford: Oxford University Press, 1980. Fodor, J.A.. "Fodor's Guide to Mental Representation." Mind 94 (1985) : 76 - 100. 106 107 Goldman, A l v i n J . . A Theory of Human Action. New Jersey: Prentice - H a l l , 1970. G o r r , M i c h a e l , and Terence Horgan. " I n t e n t i o n a l and Unintentional Actions." Philosophical Studies 41 (1982): 251-262. Grice, H.P.. "Intention and Uncertainty." Proceedings of the British Academy 57 (1971): 263 - 279. Harman, G i l b e r t . " P r a c t i c a l Reasoning." Review of Metaphysics 29 (1976): 431 - 463. . "Rational Action and the Extent of Intentions." Social Theory and Practice 9 (1983): 123 - 141. . " W i l l i n g and Intending." Philosophical Grounds of Rationality, e d i t e d by R i c h a r d E. Grandy and Richard Warner. Oxford: Clarendon Press, 1986. Hess, Peter H.. " A c t i o n s , Reasons, and Humean Causes." Analysis 41 (1981): 77 - 81. Hornsby, Jennifer. Actions. London: Rout ledge and Kegan Paul, 1980. Locke, Don. " B e l i e f s , D e s i r e s and Reasons For A c t i o n . " American Phil. Quarterly 19 (1982): 241 - 249. McGinn, C o l i n . The Character of Mind. Oxford: Oxford University Press, 1982. Montmarquet, James. "Actions and Bodily Movements." Analysis 38 (1978): 137 - 140. Prichard, H.A.. "Acting, W i l l i n g , Desiring." Chap, i n Moral Obligation: Essays and Lectures. Oxford: Clarendon Press, 1949. Searle, John R.. "The I n t e n t i o n a l i t y of Intention and Action." Inquiry 22 (1979): 253-80. . Intentionality. Cambridge: Cambridge U n i v e r s i t y Press, 1983. . Minds, Brains and Science. Cambridge Mass.: Harvard University Press, 1984. Smith, Michael. "The Humean Theory of Motivation." Mind 96 (1987): 36 - 61. Smith, Peter. "Hess on Reasons and Causes." Analysis 41 (1981): 206 - 209. 108 Smith, P e t e r , and O.R. Jones. The Philosophy of Mind. Cambridge: Cambridge University Press, 1986. Strawson, P.F.. "Causation and Explanation." In Essays on Davidson, ed. Bruce Vermazen and M e r r i l l B. Hintikka. Oxford: Clarendon Press, 1985. T a y l o r , Denise Meyerson. " A c t i o n s , Reasons and Causal Explanation." Analysis 42 (1982): 216 - 219. Thalberg, Irving. Enigmas of Agency. New York: Humanities Press, 1972 "How Does Agent Causality Work?" In Action Theory, ed. M. Brand and D. Walton. Holland: Dordrecht, 1976. van Inwagen, Peter. "On Two Arguments f o r Compatibilism." Analysis 45 (1985): 161 - 163. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0097898/manifest

Comment

Related Items