The Mind and Metaphysics of Agency

University of Texas at El Paso DigitalCommons@UTEP Open Access Theses & Dissertations 2011-01-01 The Mind and Metaphysics of Agency Gareth Fuller University of Texas at El Paso, [email protected] Follow this and additional works at: https://digitalcommons.utep.edu/open_etd Part of the Philosophy Commons Recommended Citation Fuller, Gareth, "The Mind and Metaphysics of Agency" (2011). Open Access Theses & Dissertations. 2486. https://digitalcommons.utep.edu/open_etd/2486 This is brought to you for free and open access by DigitalCommons@UTEP. It has been accepted for inclusion in Open Access Theses & Dissertations by an authorized administrator of DigitalCommons@UTEP. For more information, please contact [email protected]. THE MIND AND METAPHYSICS OF AGENCY GARETH PAUL FULLER Department of Philosophy APPROVED : ______________________________ John Symons, Ph.D., Chair ______________________________ Luciana Garbayo, Ph.D. ______________________________ Bruce Louden, Ph.D. _________________________ Patricia D. Witherspoon, Ph.D. Dean of the Graduate School Copyright © by Gareth Fuller 2011 Dedication To the usual suspects, those with us and those not. The Mind and Metaphysics of Agency By GARETH PAUL FULLER, B.A THESIS Presented to the Faculty of the Graduate School of The University of Texas at El Paso in Partial Fulfillment of the Requirements for the Degree of MASTER OF ARTS Department of Philosophy THE UNIVERSITY OF TEXAS AT EL PASO May 2011 Acknowledgements I would like to acknowledge the faculties of the departments of Philosophy and Political Science at Guilford College. I owe a debt of gratitude to each of them, as well as the entire college, for providing an excellent undergraduate experience. I would like to acknowledge all of the members of the faculty at UTEP for providing an academic environment where I have been able to grow exponentially. Whether it was making it painfully clear that I need to work harder, pointing out the philosophical importance of stories, and providing the basic skills to research on my own. I would like to acknowledge the members of my thesis committee, Dr. Louden, Dr. Garbayo, and Dr. Symons, for the help that they provided. I would like to give an extra nod to Dr. Symons for being the committee chair for my thesis, as well recommending the book that started me on the path to this topic. I would also like to thank Dr. Tim Cleveland for the excellent comments he provided at a conference on a paper that ultimately became the majority of Chapter II. And last but not least, I would like to thank all of my friends and fellow students who have made my academic life as fruitful as it has been. The many intellectual debts I owe to the many authors I have read will be apparent throughout this thesis. v Table of Contents Page Acknowledgements.........................................................................................................................v Table of Contents..........................................................................................................................vi\ CHAPTER INTRODUCTION...........................................................................................................................1 I. Historical Overview.............................................................................................................1 II. Thesis Overview..................................................................................................................6 1. Action and Agent.......................................................................................................................11 1.1 Agents................................................................................................................................11 1.2 The Cognitive Theory and Intentionality ..........................................................................13 1.3 Action.................................................................................................................................20 1.4 Conclusion.........................................................................................................................23 2. The Ontology of Events ............................................................................................................24 2.1 Ontological Status of Events..............................................................................................25 2.2 Event Recurrence and Event Identity................................................................................28 2.3 Rough and Fine-Grained Theories.....................................................................................30 2.4 Against Fine-Grained Theories..........................................................................................32 2.5 Against Rough-Grained theories........................................................................................33 2.6 The Componential View ...................................................................................................35 2.7 Conclusion ........................................................................................................................37 3. Intentions in Action....................................................................................................................38 3.1 Intentions...........................................................................................................................39 vi 3.2 The Simple View...............................................................................................................47 3.3 Solutions In Place of the Simple View..............................................................................49 3.4 Intentional Action..............................................................................................................51 3.5 Brand’s Intentional Action.................................................................................................52 4. Synthesis....................................................................................................................................55 4.1 Intentions and Parts of Actions..........................................................................................56 4.2 Causation and Parts of Actions..........................................................................................59 4.3 Conclusion.........................................................................................................................61 5. Concluding Thoughts................................................................................................................62 List of References .........................................................................................................................63 Curriculum Vita............................................................................................................................67 vii Introduction This thesis is concerned with agency, in particular, what it is to be a fully functioning, human agent. What distinguishes human agency is a set of conceptual, cognitive, and social skills that are unique to humans.1 I believe that agency is, in some form, ontologically prior to one’s society, (which is far from uncontroversial), and because of this I will not try to place agency in any sort of societal role.2 The purpose of this introduction is to give a brief historical review of recent work on agency, covering approximately the last century, and then give an outline of what is to be taken up in more detail in the later chapters. I. Recent Historical Work in Action and Agency Starting with some of the earliest philosophers, agency has posed many troubling questions. Plato and Aristotle tangled with the idea of autonomous, self-movers, which has provided material for about twenty-five hundred years of discussion.3 Although the Greeks would often bring arguments about self-movers into cosmological discussions, it has importance to our concept of human agency. However, a more important contribution from these two, and particularly from Aristotle, is the connection between reasons and action, as will be discussed later in this chapter. It is with these two in mind that much of the discussion has been framed. A significant work that has raised many enduring questions about agency is William James’, The Principles of Psychology. James’ massive work, about twelve hundred pages in 1 This is said with ignorance as to the totality of life in the universe. It is possible that there is other life that shares these attributes. 2 See Martin, Sugarman, and Hickinbottom for a take on agency that is defined through one’s society. 3 A good collection of essays is Self-motion: From Aristotle to Newton, edited by Mary Louise Gille and James G. Lennox. 1 total, gives an extended analysis of many different psychological phenomenon, including emotion, perception and, most importantly for my purposes, that of action. In his section on the will4 he provides interesting approaches to several of the significant questions about agency, but what has been of the most apparent influence is James’ lengthy discussion of kinesthesia.5 James points out that if an agent’s arm is moved by someone else there is still fairly intimate knowledge of the movement, making the agent aware of the motion and location of his arm, without the use of visual or other forms of confirmation. He would then be able to reproduce this movement, James claims.6 However, should there be some sort of inhibition of the nerves in one’s arms, if it has been anesthetized for instance, then the movement of the patient’s arm by someone else does not provide the same information. This discussion of kinesthesia is particularly interesting because it points out a clear difference between voluntary and involuntary behavior. There is an extra sense attached to the movement when the limb is voluntarily moved by the agent. 7 As stated above, if a limb is anesthetized the agent will not know if it was moved by someone else, however the agent will still get the feeling the he is moving the arm if he is asked to do so voluntarily. Pointing out research done on anesthetized patients, James brings up the fact that the agent has knowledge that he is moving his appropriate limb without looking at it, but no knowledge of whether or not his movements are what he intends them to be. This makes 4 Chapter XXVI James 488 6 James 489 7 James calls actions and movements voluntary and involuntary. I will use his terminology while discussing his work; however it is not the terminology I will use in the majority of my thesis. Intuitions about something being voluntary can conflict in cases of coercion or something similar. It can be claimed that the agent does not voluntarily perform the action because he is threatened, however he does voluntarily perform the action because he is in control of the movements. I will use intentional and unintentional for I believe that this allows a quick and easy way around this possible confusion. 2 5 apparent a particular privileged belief in connection with action.8 Other research showed that an anesthetized patient can be tricked into believing that he is moving his arm when it is actually being held still.9 This led James, as well as many others who have followed after him, to the claim that there is some extra mental capacity that leads to voluntary acts. One knows, or believes, that he is acting by something other the visual or other sensory feedback.10 The fact that voluntary actions provide a certain extra sensation of doing something is of great influence on how many later philosophers discuss action.11 Another influential philosopher, perhaps more directly so, is Ludwig Wittgenstein.12 Not only did he influence much discussion through his writing, some of which contributes ideas similar to James’, he also taught several very influential philosophers of action and agency. Much of Wittgenstein’s influence stems from one simply stated, yet intriguing question, which is, “what is left over if I subtract the fact that my arm goes up from the fact that I raise my arm.”13 For Wittgenstein, there is an important distinction between an arm raising and an arm rising. The arm rising is merely the motion of the arm, and can result from many different causes. It could be that there is a machine designed to lift an anesthetized patients arm. This would be an arm rising but not an action. However, the occurrence of an arm raising is 8 James 490 Ibid 10 I find that this sort of account of action of knowledge to be very apparent I the writings of Brian O’Shaughnessy and Jennifer Hornsby, where actions are identified by mental “tryings.” See O’Shaughnessy 1973 (reprinted in Mele 1997) and Hornsby 1980. 11 It can be seen very easily in many volitional or mental act theories, for these mental actions and events can be linked very easily to the knowledge of action. In his book Theory of Action, Lawrence Davis posits this extra knowledge as something that must be explained by every successful theory of action. 12 Clearly Wittgenstein is influential to all of philosophy, but I believe his influence in action theory is underestimated. 13 Wittgenstein 1958 Section 621. 3 9 specifically an action, and can only result from the appropriate connection to an agent. Arm raisings therefor are arm risings, but with something extra to make them actions. It is an answer to this question that is the purpose of much literature on action and this question led Wittgenstein to look towards kinesthetic research and the will for an appropriate answer. More influential to the philosophy of action than Wittgenstein is G.E.M. Anscombe. She was a student of Wittgenstein, as well as a friend. Her book Intention brought a new approach to the discussion of agency because it focused on analyzing the different meanings and grammatical forms of intention. She pointed out three different forms, taking three different grammatical positions 1. Xing intentionally [adverb] 2. Xing with the intention of doing Y (noun), and 3. A intends to Y (verb)14 It was the third form that she found to be of direct importance to understanding what an intention is, although she gave attention to all three. She also provided arguments for a belief/desire psychology of intentions, something that has become the standard theory but still provides material for considerable debate. To gain a deeper understanding of the belief/desire psychology of intentions she compared them to other mental states that she thought were similar, specifically predictions. She saw them as being similar since both are future oriented and require a belief about the occurrence of some future event. However, honest predictions are in no way connected to the predictor’s desires for the future event. Intentions do imply a desire for the intended event, and from this she argues that intentions require a desire. 14 Anscombe 1963 pg. 1 4 Another key point she makes involves reasons. This comes in two parts, one is that intentional action is acting for reasons, so that to understand what it is to act intentionally requires understanding what it is to have a reason. She believed that this could be discovered simply by asking why something was done. The second aspect was discovering what the connection between a reason and its action is. She did not believe that reasons could be the causes of actions. This was the generally accepted conception of reasons due to arguments similar to the following 1. Causation can only occur between logically distinct things 2. Reasons and actions are conceptually the same and not logically distinct 3. Therefore reasons cannot be causes15 Accepting this general thesis, Anscombe argued that someone may be startled by a face in the window and knock a coffee cup off of the table. Under these conditions, spilling the coffee would not be an intentional action. What explains the event by answering the why question is a cause and does not denote any intention on the part of the coffee spiller. While the necessity of intentions seems fairly obvious, it presents a very strong case against many other theorists of her time who did not consider intentions as the significant determiner of actions.16 Around the time of the publication of Anscombe’s book came several essays by Donald Davidson. The most influential of these was his “Actions, Reasons, and Causes.” Claiming Aristotle as an influence, Davidson argued that reasons are to be considered causes and took on the conceptual connection argument from above. Davidson pointed out that an intention lacks all 15 This is not explicitly stated and certainly not in this form, but it can be found in much of the literature from the 50s and early 60s, as in Anscombe 1963. 16 What were important were things like volitions and the will. Mental events that were not content oriented like intentions. 5 of the specific characteristics that the action has.17 One may have an intention to flip a light switch and this intention lacks all of the specifics of the action, such as the exact time and way in which the action is carried out. Using this argument Davidson pointed out that actions and reasons are logically distinct, and therefore a causal connection can be drawn between the two. Not satisfied with just arguing that reasons can be causes, Davidson then posed a challenge.18 Davidson challenged non-causal theorists to explain how a reason explains an action without causing the action. If there is no causal connection, then positing a specific reason as one’s reason for acting becomes problematic. This was not a problem when actions and reasons were not considered logically distinct but after arguing that they are, a connection between the two became necessary.19 Both Davidson’s and Anscombe’s writings are still of great importance.20 They have not only influenced discussions of intention and action, but the metaphysics of events and actions, the logic of action and event sentences, and many other issues of action theory. II. Outline of the Thesis Part of what makes human agency unique is the ability to engage in modal reasoning. This is reflected in the considerable work provided in attempting to develop a modal logic of agency.21 Much of the work is concerned with how to draw the appropriate connection between a future state of affairs and the current state of the agent. The logical operators are used to make clear 17 Davidson 1980 25 Mele has given it the name “Davidson’s Challenge.” 19 This challenge posed by Davidson was most explicitly stated in “Actions, Reasons and Causes” but appears in several of his subsequent works. 20 Not that it has been very long since either one was writing. 21 Belnap and Perloff 1992, Chellas 1992, Huffer 2007 6 18 whether a sentence such as, “Jones knocked over the bike,” is ascribing an act of agency to Jones or not. These agential operators are beyond the scope of this thesis, for they require argumentation on subjects outside of my goals. 22 I will examine two competing theories of agency, ignoring the modal implications. The first theory of agency is what I will call the causal theory. This holds that the relation of the agent to the event is simply a causal connection.23 The second theory of agency is what I will call the cognitive theory. This is the theory that has garnered the most defenders recently, particularly with the development of agent sensitive modal logic, but finds strong proponents not working in agential logic. This theory of agency holds the relation between an agent and an event, what makes the event an action of the agent, is whether the event can be given a description under which it is intentional. I will accept the cognitive theory of agency because it satisfies many of the underlying intuitions about what it is to be a human agent. Accepting the cognitive theory brings up several important issues that will then need to be satisfied. One is extending Anscombe’s discussion of intentions. Specifically, what are intentions and what is their role in intentional action? I will take these two parts separately. The first part will be a defense of folk psychology. Accepting something similar to a belief/desires psychology, beliefs and desires are then given a prominent role in constructing other mental states, such as intentions. In Chapter three I will take up a discussion of the exact content of intentions and what it is to do something intentionally. The cognitive account also points towards another question that will be taken up in chapter two, which is, what is it to give an event alternate descriptions? How does one find out that two descriptive sentences refer to the 22 The logic of agency is also beyond the scope of this thesis because it requires not only a study of modal and deontic logic, but also an understanding of temporal semantics. See Belnap, Perloff, and Xu 2001 for an extended look at all of these issues. 23 See Thomson 1977 Chapter III and Cornman 1971 pg. 26 for discussions of this theory of agency 7 same event? The first major issue to be tackled is the defense of folk psychology, therefore defending the existence of propositional attitudes. Although the existence of propositional attitudes seems phenomenologically obvious, there have been many arguments presented against folk psychology. I will examine arguments commonly put forward by eliminative materialists denying the existence of propositional attitudes. What eliminative materialists argue is that appealing to propositional attitudes (folk psychology) is relying on an ancient, static, and prescientific understanding of the mental lives of humans and possibly other higher animals. What they propose to replace folk psychology with is a perfected neuroscience. What is needed for this to work is a theory of inter-level bridge laws of reduction that provides a reason to call folk psychology into question.24 Inter-level bridge laws attempt to reduce theories of a higher level, such as psychological states, to lower level theories, such as brain states and neuroscience. Lower-level explanations of folk psychology may be the case, but if no evidence is provided that these lower-level explanations support a different description on the same level as folk psychology then the eliminative materialists have proven nothing. Eliminative materialists claim that neuroscience has now provided evidence that folk psychology is mistaken, and that when neuroscience has advanced far enough, there will be an explanation for phenomena on the level of folk psychology. I will argue that the radical inter-level reduction of eliminative materialists is misguided and that folk psychology should be retained. Part of the defense of folk psychology will rely on the work of Daniel Dennett which has some instrumentalist leanings.25 Full 24 There are two types of bridge laws, inter-level and intra-level. Inter-level is between explanations on different scales, such as neuroscientific explanations and psychological explanations. Intra-level bridge laws are used to reduce explanations of similar scale, such as phlogiston to oxidation theory. Both can be used to eliminate other explanations. 25 Where he fits into the instrumentalist brand is of some debate, but there are certainly instrumentalist leanings in many of his writings. 8 instrumentalism is the view that folk psychology plays a functional role in everyday discourse, but it is not actually a description of the brain’s content. Just as saying that the sun is rising is perfectly fine for everyday use; one knows that it is not a description of how the solar system actually functions. Ultimately I will not accept instrumentalism, for I hold that the folk psychological states employed in describing actions are the actual causes themselves and therefore exist as more than patterns. The discussion of propositional attitudes is important for the rest of the thesis because it outlines the cognitive representation of the metaphysical issues to be discussed next. Propositional attitudes fulfill several important roles, one of which is laying the foundation for intentions, which have as their content a specific action. I will assume that actions are events and argue that they are to be understood as particulars, somewhat similar to physical objects. Part of the problem of considering events to be particulars is setting up an event identity. It seems counterintuitive to consider events to be exactly like physical objects for the reasons to be discussed later. As well as arguing that events are particulars, I will defend the claim that actions are to be understood as componentially constructed, meaning that they are made up of many smaller events and actions. This allows one to appropriately describe actions that are temporally extended, as well as the apparent fact that actions and events are often the combination of smaller events. This raises another issue, discussed in the following section, because not all of the parts of an action are controlled by the agent. This leaves a question as to the connection between intentions, action components, and complete actions. I will argue that the content of an intention is a whole action, and that to perform the whole action intentionally it must satisfy several conditions, to be laid out later. Important to 9 performing actions intentionally is the role of act plans in managing and monitoring intentions, as well as having the appropriate skill. To satisfy the requirements of performing something intentionally, the parts of the action must fit into the overall act plan appropriately, and cannot violate the skill requirement, as will be made clear later. With this brief introduction I have described the path that I will take through this thesis. To begin I will discuss the use of the terms action and agent. The purpose of this section will be to explain exactly how I am using these terms throughout the thesis, as well as to discuss the two theories of agency from above. 10 Chapter 1 Action and Agent The purpose of this section will be to outline how the terms action and agent are to be used. I take these two terms to be very closely connected, for a complete concept of one is central to the other. The outline in this section will be fairly skeletal, for it is the purpose of the rest of this project to fill in the details. The overall intention is to outline what it is to be a fully functioning human agent, particularly focusing on the mental capacities. Most important to this discussion of agency will that of intentionality. However, I will begin by examining two competing, general theories of agency. One holds agency to be a causal connection between the agent and the event. This I will call the causal theory of agency. I will argue that this account of agency is mistaken because it ignores the cognitive and conceptual abilities that are part of being a human. Ultimately this theory of agency does not allow for moral appraisal, and lacks the ability to distinguish something the agent does from something that happens to him.26 This will lead me to the second theory, which I call the cognitive theory of agency. This theory of agency holds that someone is the agent of an event if, under some description, it can be considered intentional. What it means for two different descriptions to be of the same event is something that will be explored more thoroughly in chapter two. 1.1 Agents The two theories of agency, as discussed above, are the causal theory and the cognitive theory. The cognitive theory has more defenders than the causal theory, but both have had strong 26 Some argue that our common understanding of moral judgment is mistaken. This is not a claim that I am going to get into, but I believe that the cognitive account allows for the possibility of libertarian and compatibilist accounts of moral responsibility. 11 proponents. John Cornman and Judith Thomson have both provided a defense of the causal theory, although Thomson does not want to take a side so she presents a possible defense for the cognitive theory as well. Donald Davidson has provided one of the most explicit defenses of the cognitive theory of agency, and it is from his that I draw my influence. The causal theory of agency connects agents to events directly through causal connection. At first glance this seems to be something that is quite obvious, for who would want to claim that agents do not cause anything? A common example provided is of a person who kills someone while in an epileptic seizure. The causal theory holds that this person is the agent of this event, for he is the cause of the killing. While it is certainly acceptable to posit the epileptic person as the cause of the death, to call him the agent of such an event betrays many of the intuitions we have about agency; intuitions I find no good reason to give up. One intuition that the causal theory has trouble supporting is how we distinguish humans as agents and other substances as agents. A tree may fall and kill a person just as an epileptic person may kill someone else in the midst of a seizure, and in this way they are similar in causal agency. However the person is capable of deliberating his killing while the tree is not. In order to clearly define types of agents, ad hoc clauses must be added to the causal account. These clauses would be such things as “is animate,” “is human,” or “has the mental capacity to...” These distinctions are inherently made in the cognitive account. Another issue with the causal account is that it is confusing to distinguish who is the agent of what. If someone were to spill his coffee, he would clearly be the agent of the spilling if he intentionally turns the cup over and pours out the coffee. We would not consider him the agent if he were bumped and spilled his coffee. However, the causal connection may not readily 12 distinguish who is the agent in such cases of bumping, for how is this causal connection to be defined? The most apparent explanation is a counterfactual, as in “if X had not bumped Y then the coffee would not have been spilled.” This would then put X as the agent of the spilling and not Y. However such counterfactual analyses of causation are far from accepted as true designations of causation. A third intuition is that human agency is connected with a sense of moral responsibility. If one is the agent of an action, one can be held responsible for that action. I am not saying that everything humans do is worthy of moral appraisal, because we would not want to assign any serious moral value to someone eating a bowl of cereal. We would not hold the epileptic killer morally responsible for the killing, or at least greatly reduced responsibility. Our concerns with moral responsibility require an appeal to intentions, and ultimately the cognitive theory of agency. In order to distinguish moral agency, and maintain our general intuitions about humans as moral agents, it is required to provide a further mental distinction, rather than just a causal connection.27 1.2 The Cognitive Theory and Intentionality The cognitive theory of agency captures some basic intuitions about human agency, in that it readily distinguishes humans from other objects on the basis of mental abilities, and allows for moral distinctions to be made. This requires being able to attribute intentions and propositional attitudes to people. Not only does the cognitive theory easily distinguish morally responsible behavior, it also easily distinguishes actions from non-actions. The causal theory 27 Causal theory will be used again in chapter III. This is a different causal theory about how an agent’s reasons cause his actions. This will be a theory that is accepted, unlike the causal theory of agency just discussed. 13 would need to appeal to propositional attitudes to distinguish actions from events involving the agent, because ascribing agency does nothing to make either one explicit. Common sense moral responsibility relies on attributing propositional attitudes to agents. As well as allowing for moral worth to be determined, propositional attitudes provide an important layer to understanding explanations of actions. This is something that has been understood since Plato, and something that was very important to Aristotle. Much of current action theory has been greatly influenced by both; more directly by Aristotle. The resurgence of action theory, particularly in the arguments presented by Anscombe and Davidson, draws much from the writings of Aristotle, and more indirectly from Plato. In the Phaedo, Plato says, “if one should say that Socrates does with intelligence whatever he does, and then, in trying to give the cause of the particular thing I [Socrates] do, should say first that I am sitting here because my body is composed of bones and sinews…[which] by relaxing and contracting, make me able to bend my limbs now, and that is the cause of my sitting here…and [he] should fail to mention the real causes, which are, that the Athenians decided it was best to condemn me, and therefore I have decided that it was best for me to sit here, and that it is right for me to stay and undergo whatever penalty they order. For, by the dog, I fancy these bones and sinews of mine would have been in Megara or Boetia long ago, carried thither by an opinion.”28 Plato is pointing to the fact that a person’s actions are explained and understood by appealing to their propositional attitudes. He also makes it clear that Socrates’ having bones, 28 Quoted from Donagan pg.1 14 sinews, and a body in general, are what allows him to sit as he does, but they cannot be considered the ultimate cause of his sitting where he does.29 Understanding actions through propositional attitudes was taken up by Aristotle, (although it is probably not unique to either of these two philosophers for their time period), and it is his analysis of the connection between propositional attitudes and actions that has been most influential in current debates about human action and agency today. Aristotle was more specific in his discussion of how one’s propositional attitudes influence her actions. Aristotle believed that in cases of reasoned actions the agent uses what is now known as the practical syllogism 30, which uses one’s desires and beliefs to lead her to act. It starts with one’s desire for a specific action as the base premise and then uses one’s belief about how best to satisfy the desire to lead to a conclusion. Written out it would look like; P1. I want to cool down, (desire premise) P2. Turning on my fan is the most efficient way to cool down, (belief premise) C. I should turn on my fan This is a very general example of the practical syllogism, and the exact wording of the premises and conclusion is a topic of much debate. Whether the conclusion should be in the form of “I should turn on my fan” or “I will turn on my fan,” is of great importance, but I do not intend to take a stance one way or the other. There could also be some deliberation between beliefs, perhaps between turning on the air conditioning, the fan, and going to the pool. These beliefs 29 Ibid Notice how this is quite different from Anscombe who believes that such propositional attitudes cannot be causes. 30 Taken from Davidson 1963. 15 could also be used in a comparison between which desires to fulfill, if there are no acceptable ways to satisfy the desire to cool down this would probably lead one to try and satisfy a different desire. Practical reasoning is a concept that requires a very detailed discussion which I will not give it. It does highlight the importance of propositional attitudes in explaining actions.31 These propositional attitudes that Plato appealed to are of great use in predicting and explaining the behavior of agents. However, their place in scientific discourse about the nature of agency and action has come under attack recently. There have been several theories that want to limit or eliminate folk psychological terms from truly scientific discourse, two of which are Eliminative Materialism and Instrumentalism. Eliminative materialists (EM) are those that want to eliminate folk psychology. Championed by Paul and Patricia Churchland32, EM argue that folk psychology is an ancient, static, and false theory of how cognition works. EM often cite the fact that folk psychology is unable to account for many issues of mental life, such as dreaming. 33 As a scientific study of mental life, EM often compare folk psychology to other failed scientific explanations, such as phlogiston. EM believe that what is required is a deeper understanding of neuroscience and the appropriate bridge laws and then folk psychology can be explained away.34 How it is that one formulates bridge laws between higher and lower level theories is of great importance to the success of EM. The purpose of a bridge law is to connect lower level explanations, such as explanations at the level of neurons, with higher level explanations, such as 31 This discussion of practical reasoning has covered very little of what there actually is to be covered. The issue of whether or not the conclusion of it causes action, as is attributed to Aristotle by many philosophers, is a topic of great debate. There are some who argue against the use of the practical syllogism (Shueler) as well as many other aspects of practical reasoning. There is a lot of literature on the topic of practical reasoning, and it is far too much to cover here. 32 See in particular their joint book On the Contrary 33 Braddon-Mitchel and Jackson 267 34 Haselager 15 16 explanations at the level of folk psychology. What these bridge laws do is provide the appropriate means for translating higher level explanations into terms of the lower level explanations.35 The intuition behind eliminative materialism is that the bridge laws will show that folk psychological notions cannot be reduced to the lower-level explanations, and that they should be disregarded as a theory about what is actually happening in cognition. Eliminative materialists want to replace folk psychology with a perfected neuroscience, believing that this is what provides the true explanation of what mental life is. There are several responses to eliminative materialism. Several come from instrumentalists and provide reasons for maintaining folk psychology. Daniel Dennett is considered to be an instrumentalist of some sort, particularly in his early work, and he defends the claim that folk psychology has its uses.36 Instrumentalism holds that, while folk psychology is not a description of what is actually going on mentally, it does provide an accurate way of predicting and explaining actions.37 Dennett holds a stance that is less anti-realist about folk psychology and it is useful in disarming some of the arguments presented by eliminative materialists. Dennet believes that this dichotomy between realism and anti-realism about propositional attitudes is a false one, and has tried to do so through several analogies. He laments, “this conviction prevails in spite of my best efforts over the years to undermine it with various analogies: are voices in your ontology? Are centers of gravity in your ontology?”38 The point that 35 Haselager 15 It is claimed that his views changed from instrumentalism to something that posits a more real stance on propositional attitudes. See Yu and Fuller 1986 (this author’s name is Gary Fuller) 37 Lyons 35 38 Dennett “Real Patterns”, 288 17 36 Dennett is trying to make in this paper is that such a black and white dichotomy is not necessarily the way things must be. Dennett argues that folk psychological explanations may not be an exact description of what is causally efficacious in terms of mental aspects, but there is a close connection between them and what they explain, such that they are quite accurate.39 What he is claiming is that patterns of behavior are real and that creatures who are experts in picking out these patterns notice them very quickly. He provides an example of expert chess players “who can instantly perceive (and subsequently recall with high accuracy) the total board positions in a real game [of chess], but are much worse at recall if the same chess pieces are randomly placed on the board.”40 These patterns are easily discernable to the expert chess player, but not to the novice. He then goes on to argue that these real patterns are noticeable in the behavior of humans and can be picked up easily by humans. What is particularly important about arguments presented by Dennett is his distinction between the role of personal and sub-personal levels.41 To take things at the personal level is to define things in terms of a fully functioning social animal. To try to provide explanations of actions in terms of sub-personal levels requires defining the person at this sub-personal level. What happens on the personal level cannot be completely explained by the mechanisms that underlie the personal level of explanation.42 This causes particular problems for the attempt to create bridge laws because it is argued that lower level explanations lack full explanatory power of higher level theories. 39 Dennett admits that they are not completely accurate, but are very accurate. It is this small chance of being incorrect about an agent's beliefs and desires that makes games like chess fun. 40 Dennett1 293 41 Dennett2 “Personal and Sub-Personal Levels of Explanation” reprinted in Bermudez 2006 42 “Introduction” Bermudez 2006 pg. 5 18 Although Dennett provides many defenses against eliminative materialism, I am unsatisfied with his theory because I take folk psychological states to be causally efficacious aspects of mental life, making my stance very much so a realist stance.43 That beliefs, desires and intentions are the causes of actions, not simply descriptions of patterns of behaviors is what I intend to argue. This leads us towards a functional explanation of folk psychological concepts. 44 An example of the use of psychological concepts present in other animals will shed some light on human folk psychological concepts.45 In ducklings, there is the concept of imprinting, where the baby imprints on the first animal that it sees and follows it. What type of animal it is that the duckling imprints upon does not matter. What imprinting tells us about folk psychological matters is that certain things are to be defined through functional roles. Imprinting allows us to know that there is a “trace” placed in the ducklings mind, so that it follows the same animal. To claim that imprinting will be replaced by a perfect neuroscientific understanding of ducklings is misguided. Neuroscience may provide an explanation of how imprinting occurs, however it cannot replace the concept of imprinting due to its functional role. The analogy is to be drawn over into the folk psychological domain. Neuroscience will provide an explanation of how beliefs and desires come about, but there is a functional role that cannot be replaced by such lower level theories.46 This may look like Dennett’s arguments, and it is very similar in ways, however this argument due to functional roles posits that psychological states are causally 43 It could be that the real patterns of behavior do have some causal ability, or help determine behavior in some way, but I want a more realist stance on folk psychology. 44 Functional explanations are of great use in action theory. Intentions tend to be understood in functional terms as will be seen in chapter three. 45 This example is taken from Braddon-Mitchell and Jackson pg. 268 46 Braddon-Mitchell and Jackson 287 19 efficacious. Just as imprinting is what causes a duckling to follow a specific researcher around, so too are our folk psychological states causally efficacious in our actions. With this defense of folk psychology in place, I will now move on to how to define actions. These folk psychological concepts will be of great importance in chapter three, when a discussion of intentions is brought up. 1.3 Action Defining action goes hand in hand with the cognitive account of agency discussed above. It posits very clear distinctions for what are actions and what are mere events. A very clear explanation of this is given in Davidson’s essay “Agency”. As he commonly does, Davidson starts his essay off with a question. He asks, “What events in the life of a person reveal agency; what are his deeds and his doings in contrast to mere happenings in his history; what is the mark that distinguishes his actions?”47 He wants to distinguish actions as a distinct type of event, and how one can tell actions from events. Davidson presents a brief scenario where he is woken up in the morning by someone playing violin, dozes, gets up, washes, turns off a light, trips, and several other things. Some of them are clearly not acts of his, such as being woken up the violinist; others clearly are actions of his, such as getting up. However, some of the events may appear to be actions, such as turning off the light, but with further information it could be discovered that it is not an act of his, because he accidentally brushed against the light switch on his way down the hall. He discusses some failed attempts to determine actions by examining the grammatical form of sentences and verbs, and how they distinguish actions as connected to the agent. 47 Davidson 1980 43 20 However this fails to be useful because the agent could be in either the active subject or the passive subject and neither one consistently designates agents in connection to actions. Another error is that the connection between the agent and action is inherent in some verbs used, and their placement in sentences. However Davidson points out that, while this is true of some verbs, it is not true of all. Such statements as, “He blinked,” and, “He rolled out of bed,” do not clearly designate an action. More information is needed than what is provided by the verb, such as the example of the light switch.48 This leads Davidson to look at other possibilities for distinguishing actions. He then looks at the possibility of incorporating intentions in defining actions. Turning off the light is an action if he intentionally did it, not an action of his if he brushed against the switch, he claims. However intentionality does not cover all possibilities of describing actions. Someone may unintentionally spill his coffee, while intentionally spilling the contents of his cup, thinking that his cup is full of tea and not coffee. From this Davidson makes three kinds of distinct cases of his spilling his coffee. One is where he does it intentionally, the second is where he mistakes the content of his cup for tea and spills it, therefore not spilling his coffee intentionally but the contents of his cup, and then a third where it is clearly not an act of the agent, such as being bumped by someone and accidentally spilling it.49 He moves on to cases of mistakes, such as misreading signs, misinterpreting orders, and miscalculating sums. These are things that cannot be done intentionally, Davidson points out. One can pretend to misread a sign, but one cannot “intend to fail.” Even if one could intend to fail, there are times when one would unintentionally misread, misinterpret, or miscalculate. 48 49 Davidson 1980 44 Davidson 1980 45 21 However, even if one does make a mistake he is still performing an action. From this Davidson concludes that actions are more than just intentional, for unintentional events may still be actions.50 What Davidson then concludes to be the designator of actions is, if under some description the event is intentional, then it is an action. In the case of spilling one’s coffee thinking it is tea, spilling the coffee is not intentional, the spilling of the contents of the coffee cup is. Therefore spilling the coffee is an action, for it can be re-described in a way that makes the event intentional. Even when one makes a mistake one is still acting. Misreading a sign is an action because it is an intentional reading that went awry.51 Davidson defends against the criticism that if someone were to trip on the rug while walking through her dining room, it would be because of an intentional action. However, Davidson points out that it is necessary to keep the cause, (one’s intentional walking), separate from the unintentional affect, (tripping). Just because an intentional action is the cause of something, it does not mean that the affect can be brought under the same intentional description.52 Providing this account of distinguishing actions allows for easy distinction between actions and other events. There are several issues still to be discussed, such as what it is to provide alternate descriptions of an event and what it is to do something intentionally. These descriptions of action and agency provide many advantages, and allow our intuitions about human agency to be maintained. 50 Davidson 1980 46 Ibid 52 Ibid 51 22 1.4 Conclusion In this section I tried to clearly define what it is to be a human agent and how I am going to use the term action.53 I define a human agent as something more than just a cause. Human agents rely on certain cognitive and conceptual abilities that require the use of propositional attitudes. It is in these proposition attitudes that agents represent the actions and form the intentions that cause their actions. These issues will be taken up later, for now I need to provide a discussion of the ontology of events. 53 From now on I will use action and agent to refer to my definitions of them as laid out in this section, unless noted otherwise. 23 Chapter 2 The Ontology of Events and Actions An ontology and theory of events will have an important role to play in filling out our theory of agency. Important to a theory of agency, as pointed out in the earlier section, is distinguishing which events that involve an agent are mere happenings or should be considered actions. In this section, the ontological status of events will be considered, as well as the issue between rough and fine-grained theories of event description. The ontological issue concerns whether events are particulars, properties, or propositions. I will argue that events are particulars as Davidson and Brand have. As has been pointed out, the deficiency of particularist theories is that they do not readily lend themselves to a concept of event identity and the issue of recurrence. 54 I will examine the identity conditions proposed by Davidson. I will then present a solution to the issue of recurrence of events. Discussion of the ontological status of events leads fairly naturally into the issue between rough and fine-grained theories of action individuation. Taking actions as a class of events, the ontological status of events affects how actions are to be attributed to agents, whether they are properties of agents or whether there is a three place association between the agent, action, and time. Rough-grained theories tend to combine what could be separate events into one action based on similar causal chains and agential origins. A fine-grained theory is far stricter in how it distinguishes actions, relying on the properties exemplified by a substance. For instance, when one pumps water into her house she would move her arm up and down, move the water pump 54 Brand 1976 24 handle up and down, and pump water into her house.55 A rough-grained theory would hold all of these to be the same action while a fine grained theory would argue that all of these are separate actions. For a rough-grained theory, each of these would be the same action under different descriptions, for the focus would be on similar causal chains and consequences. Fine-grained theories would separate these instances due to the fact that pumping water into one’s house exhibits a different act property than moving one’s arm, even if the two are intimately connected. I will argue that neither type of theory properly distinguishes actions, and that a componential view, one that holds actions as being made up of many smaller events and actions, is to be accepted. I will then use this viewpoint to alter the account of event identity so as to incorporate this specific stance on action individuation. 2.1 The Ontological Status of Events Discussions of events tend to take one of two approaches, that of basic ontology or that of semantics, logical form, and language. While each has an impact on the other, for whatever ontological category one places events into will affect how one analyses sentences and one’s semantic theory and parsing of event sentences affects whether or not one includes events as existing entities, one may consider each separately. However I find the interplay between them to be important in argumentation for and against a specific metaphysical stance.56 Generally, ontologies consider three distinct categories, those of particulars, properties, and propositions. Not all ontologies admit the existence of all of these categories, some 55 Taken from Anscombe 1963 The many writings of Davidson, Casteneda, and Higginbotham (to name a few) tend to take a semantic approach, while those of Goldman, Kim, Montague, Brand, and others take a more ontological approach 25 56 preferring to omit the categories that are considered universals (properties and propositions). I do not hold such a reductive ontology, and find no problem including universals into a discussion of ontology.57 Although I will argue for a particularist theory of events, I find it necessary to include these other ontological categories. Considering these ontological categories, we can see theorists who have placed events into all three. Roderick Chisholm presents events as a type of proposition, Richard Montague has argued that they are properties, and Brand has argued that they are particulars.58 On top of these theorists, Kim, in some papers, and Goldman have argued that events are properties exemplified at certain times by an object, making events a type of particular.59 I will save the discussion of Kim/Goldman type of events for later and focus on the theories of Chisholm and Montague, providing reasons for rejecting their theories of events. As stated above, Chisholm holds events to be a proposition like entities. More precisely, Chisholm holds propositions and events to be different types of states of affairs, but admits that what he considers to be states of affairs are commonly called propositions.60 The propositional theory of events has a very easy time solving the issue of recurrence and identity conditions. Events literally recur, for the state of affairs may be the case, or as Chisholm says “concretized,” and then not be the case, only to be concretized again. This makes event identity very easy because it is associated with the state of affairs that is concretized. For instance, two different sun risings are the same event just being concretized at two different times. 57 Perhaps this requires some argumentation, but I will not do this here. This is certainly not an exhaustive list but they are among the most prominent who have argued for each ontological status of events. 59 See Castaneda 1979 for a quote from Goldman saying so, and Steward 1999 Chapter 1, Horgan 1978 for a discussion of Kim’s view of events and its changes. 60 See Chisholm 1976 chapter V. 26 58 Events for Montague are properties of moments or intervals of times.61 So, for Montague a certain event is a property P being exemplified at a time t. What separates his property theory of events from the Kim/Goldman type is that general and particular events are not in different categories for Montague. Kim and Goldman rely on the use of event properties and event tokens to distinguish a general event from a particular event, such as a sunrise (event property) and the sunrise we saw this morning (event token). Montague argues that the difference is the particularity of the property being exemplified. So the sunrise we saw this morning exemplifies a different property then a sunrise, namely being seen this morning. This makes issues of recurrence and identity simple as well because the property of a sunrise recurs in every general case of a sunrise just as the property of being a chair recurs in every chair.62 While both theories have similar advantages, they are counterintuitive at best. Both theories at first glance promote common ways of talking about events. For instance, we can say that Kant takes the same walk every evening. If Kant were to be so particular as to start every walk with the same foot, follow the same path, and count his steps so as to make sure he takes the same amount on every walk, it would seem that he does. However, we are not literally saying that Kant takes exactly the same walk every evening (numerically singular), which would be the case for Chisholm and Montague.63 Many authors have argued that such a statement can be paraphrased as “there is a route along which Kant takes a walk every evening.”64 What this makes reference to is event types and particulars. As well as betraying linguistic usage, which is not a complete reason to reject a theory, events under the theories of Chisholm and Montague are 61 Montague 1969 pg. 160 Ibid 63 Higginbotham, Pianesi, and Varzi page 7. 64 Adapted from Davidson 1970 62 27 very thin and would be considered multiplier theories, which I will argue against in a later section. 2.2 Event Recurrence and Event Identity I have already alluded to a solution for event recurrence above. Considering events as particulars requires some work to provide appropriate accounts of recurrence and identity, but it can be done with the use of distinguishing types and particulars. Although Montague did something similar, with his generic and particular event properties, a particularist theory of events holds that the generic types and the particular tokens are of different categories. Particular events, such as an individual occurrence of the sun rising, is of the type “sun rising,” but it is to be considered a particular having the property that makes it of this type. This holds recurrence to be not literal, for each sun rising is its own individual event just as each chair is its own individual chair, but it is of the same type, and it is this type that is exemplified when it recurs. Event identity is a little more complicated. Davidson originally argued that events are to be identified through common causes and effects.65 He argued, that for events x and y, x = y iff x and y have all of the same causes and effects.66 Therefore, remembering the scenario of the person pumping water, moving one’s arm up and down and pumping water into the house are the same event because they share the same causes and effects. The charge of circularity is that such an account of identity relies on already having a determined identity of events to distinguish the 65 He later changed his stance due to charges from several philosophers that his causal account is circular. In particular see Quine 1988 and Davidson’s response in the same book. 66 Davidson 1980 179 28 causes and effects of events x and y.67 How can one distinguish cause and effect without already knowing how to identify different events? As well as this, events that have no causes and effects would all be the same event. Although there is no strong reason to believe that such events exist, the only arguments against them are epistemic, and our epistemology should not be the measure of our metaphysics. If a proper event identity can be presented that would avoid such an issue then it should be considered superior to the causal account. One issue with providing identity conditions for events is that, though they are particulars just like physical objects, they do not seem to share in many of the defining properties of physical entities. Most importantly is that it is not clear whether events fill the entire spatiotemporal region in which they are situated. Brand argues that it is clear that events do not, and this requires some fancy work setting up identity conditions. Quine argues that events and physical objects are not to be distinguished, for “each comprises simply the content...of some portion of space-time, however disconnected or gerrymandered.”68 If events are to be taken as a type of particular maintaining a constant theory of identity seems rather intuitive. Claiming that events completely fill spatiotemporal regions leads to some difficult issues with event identity. If one were to catch a cold while swimming the English Channel, then the catching of the cold and the swimming of the channel are to be considered the same event, occupying the same spatiotemporal region.69 However, this certainly seems strange to claim, because we naturally seem to want to distinguish the crossing from the cold. Also, defining the exact spatiotemporal region of some events becomes very difficult. It is clear where the swimming of the English 67 This is assuming that event causation is true, which Davidson certainly did, and I find to be very plausible. 68 Quine 1960 pg. 131 69 Higginbotham, Pianesi, and Varzi page 8 29 Channel took place, through whatever path the swimmer took across. However, it seems intuitively wrong that the catching of the cold took place in the path that the swimmer took across the channel, for this ignores the role of the immune system and other locatable things. Also, another example provided by Davidson makes the point more clearly. 70 If a metal ball spins and at the same time heats up, are these to be considered the same event? Intuitions point towards these two being separate events occupying the same spatiotemporal region. I think that it is clear that events do occupy a certain necessary place, even if it is hard to define, but do not completely fill the region they are in. Clearly, the event of swimming the English Channel has a particular space, involving the English Channel, the path taken across it, and the body of the swimmer. This can then be distinguished from the catching of the cold even if the swimming of the channel and the catching of the cold occur at the same time because they do not share the same necessary parts. Catching a cold in no way requires the English Channel, for one can catch a cold without swimming across it.71 Therefore an event may be distinguished by necessary conditions implied by the type of event it is, so the swimming and the catching of the cold are distinguished by necessary aspects of the types that they are tokens of. Then to differentiate various performances of swimming and catching a cold the spatiotemporal location is the distinguishing factor. 2.3 Rough and Fine-Grained Theories Irving Thalberg presents a useful example of how different rough and fine grained theorists, approach events, and actions in particular. There are nine occurrences 70 71 Davidson 1980 pg. 178 This is not exactly the claim made by Brand, but it draws much inspiration from Brand 1984. 30 1. 2. 3. 4. 5. 6. 7. 8. 9. My right foot changes position Tendons contract in my leg and foot Feedback impulses travel to and from my brain I push my foot downwards I depress the accelerator of my roadster I speed up I careen past a red light I commit two traffic offenses I provoke a highway patrolman72 A multiplier (fine-gained theorist) would hold all of these to be different events, and the ones that are easily distinguishable as acts, namely 4-9, are to be considered different actions. Multipliers tend to hold a theory of event identity that is; Event a = Event b iff a and b exemplify all of the same properties.73 Multipliers usually rely on the use of act properties and tokens of these acts, and make distinctions by what properties are exemplified by what. A unifier, philosophers like Davidson and G.E.M Anscombe, would bring up the fact that these actions, and in fact all of 1-9, have a common “origin”, with Davidson providing a statement of event identity that is; Event a = event b iff a and b have exactly the same causes and effects.74 This statement clearly provides a very different identity of events, so that a unifier would hold that his pushing his foot downwards is the same action as his depressing the accelerator and his speeding up. 72 Thalberg 1971 pg.782 Goldman 1971 pg.766 74 Davidson1969 pg.178. I do not believe that this statement of event identity is completely correct, but I am sure it can be appropriately bolstered so as to be defensible. For arguments against it see Goldman. 31 73 2.4 Against Fine-Grained Theories The argument that I have against fine-grained theories relies on an argument from actions that require temporal parts, especially ones that are spread quite distantly in time. Let us consider an assassin (call him Sirhan) who has been hired to kill a prominent political figure (call him Kennedy), and does so by shooting him. It is fair to assume that the assassin pointed his gun, flexed his finger, pulled the trigger, fired the gun, and our assassin shot his target. However, lets us say that his target did not die immediately, but died several hours later strictly because of the gunshot wound. Nothing else was the cause or can be linked to his death. Let us also assume that the assassin killed himself immediately after shooting his target. A fine-grained theory would hold that the flexing of the finger and the firing of the gun would be separate events because the flexing of the finger is a property exemplified by the assassin, while firing is a property of the gun. The fine-grained theorists must also hold that the killing of the victim is its own separate action, (it would be an action under both theories of agency provided in chapter 1) for it exemplifies a very different event property from anything else in the story. This attributes an action performed by the assassin after he has died, and presumably not capable of acting. Thomson points out that “For Sirhan to have killed Kennedy by shooting him was for Sirhan’s shooting of Kennedy to have been Sirhan’s killing of Kennedy. After all, once Sirhan had shot Kennedy, he needed to do nothing more in order to kill Kennedy…after the shooting, Sirhan himself could have done anything or nothing, he could have stood on his head.”75 75 Thomson 1977 pg. 47 32 The fine-grained theorist must then attribute an action to someone who is no longer acting, perhaps dead, or doing something completely unrelated, standing on his head, if they are to individuate actions in such a way. 2.5 Against the Rough-Grained View The rough-grained theory of events provides some answers to the temporal issues that plague the fine-grained theory. I will first provide a criticism against rough-grained theories concerning temporal issues of action and then present a solution. Then I will argue that roughgrained theories are susceptible to other criticisms, mainly an issue with the by relation. Goldman argues that rough-grained theories fall to temporal issues of actions, however in a way quite different from fine-grained theories.76 While fine-grained theories can be criticized by attributing too many actions to one agent, Goldman argues that rough-grained theories run into issues in their reduction of actions. In the case of the assassin, a rough-grained theory would hold that the pulling of the trigger, and the shooting and killing of Kennedy are all the same action. However, there clearly is temporal distance between the pulling of the trigger, the bullet entering Kennedy, and the death of Kennedy.77 What Goldman argues is that if the killing and the trigger pulling are the same action, and in fact rough-grained theorists argue that the trigger pulling is the killing, then the killing of Kennedy happened before the bullet even entered Kennedy’s body. This, to Goldman as well as Thomson, seems very clearly wrong, for the bullet must hit Kennedy before he is killed. The response to this criticism shows the reliance on causal dependency in distinguishing actions. If actions are to be causally distinguished then the pulling 76 77 Goldman 1971 pg. 765 Goldman 1971 pg. 765 33 of the trigger can very easily be considered the killing, for it is the basic action of pulling the trigger, or perhaps flexing the finger, that caused the bullet to hit Kennedy, causing the hemorrhaging, and so on until his death. To call an action a killing is to refer to the bodily movement that caused the death.78 Therefor the most basic act that can be performed by an agent, pulling the trigger, is the cause of the other events that led to the death of Kennedy, and so this is to be identified as the killing of Kennedy. While it is true that it was not a killing of Kennedy until he died, this seems to be more of an epistemic issue rather than a metaphysical one. The birth of Beethoven was the same birth as the birth of a famous, deaf composer. This was not known to be true until he was famous, deaf and a composer. Events may acquire descriptions as time progresses, but once again this is an epistemic issue, not a metaphysical one. Although rough-grained theories provide a way to solve the temporal issues presented in fine-grained theories, there are other reasons to reject the rough-grained view. The most prominent complaint is that of the logic of the “by” relation.79 The by relation is when one claims to do something by doing something else. In the case of the killing of Kennedy we would say that Sirhan killed him by shooting him, and shot him by pulling the trigger. We could then go on to say that Sirhan killed Kennedy by pulling the trigger. However, the converse is not true. To say that Sirhan pulled the trigger by killing Kennedy does not even seem to make sense. If the pulling of the trigger and the killing are the same action, then they should share all of the same properties, unless one is willing to reject Leibniz’s Law.80 What this means is that the by relation 78 Davidson 1980 pg. 179 Mackie pg. 40 80 Tim Cleveland argues that the causal individuation of events is a way of supporting Leibniz’s law. See Cleveland 2011. 34 79 makes it apparent that it refers to two distinct events, or many distinct events, because it is a relation that is not shared by both descriptions. 2.6 The Componential View To solve the temporal issues above, and to incorporate the by relation, I suggest accepting a componential view. The componential view holds that events are made up of many smaller events, eventually getting down to indivisible “micro-events.” The componential view lends itself quite well to comparison to physical particulars, which is a comparison I argued for earlier. The solution to the temporal issues of the fine-grained theories is quite apparent with the componential view. The killing is to be considered all one action that includes the pulling of the trigger and the bullet striking Kennedy. The killing is not a completely separate action from the pulling of the trigger, but it is something that includes it. Thalberg makes an apt comparison when he points out, “the frame that is part of your reading glasses is not something you wear in addition to your glasses,”81 so the pulling of the trigger is not something that is done in addition to the killing. This solves the issue by attributing only one action, that of killing, which is made up of smaller acts and events. How the action is to be defined, as in what makes this a killing instead of a pulling of the trigger, is how the action is represented in the intention of the agent, which will be discussed later. I believe doing this represents how agents plan, as well as pointing towards a solution to the issue of basic actions, which is something I will not deal with in this thesis. The componential view solves the issues of rough-grained theories because it allows for the by relation to be asymmetrical. Since the pulling of the trigger is not the same as the killing, 81 Thalberg 1971 pg. 786 35 just a part, Sirhan can claim that he killed Kennedy by pulling the trigger of his gun without running into issues with Leibniz’s Law. Pulling the trigger is a part of the act of killing, and in particular a part that Sirhan did, so that he did kill Kennedy by pulling the trigger. Once again, separating actions in this way lends itself to discussions of basic acts, for one must perform a greater action by performing a basic act.82 The next issue however is how acts and events are to be constructed. Just like roadsters, killings and all events, are only what they are by how their parts are put together. One may have all of the pieces of a roadster strewn across a garage floor, but it seems very strange to say that he therefore has a roadster. One may take all of the pieces of his roadster and place each one into a separate non-roadster car that he owns, and it certainly would not do to say the he has a roadster.83 So, I take it that the roadster is something more than just the parts; it is in the construction as well. Events are very similar, except the parts of an event take a causal connection. Therefore, I hold a killing as all of its parts in an appropriate causal sequence, for we could not call Kennedy’s death Sirhan’s killing of Kennedy if Kennedy died before Sirhan fired his gun. The way that the constituent events are put together is of great importance to the identity of the overall event. The overall structure of the meta-event determines its identity, and this takes 82 I know I have been very vague about basic acts but it is a major issue that I do not want to approach. A brief definition of basic acts would be the most basic part of an action that is under the direct control of the agent. In the case of throwing a ball there are certain parts of the act that are not under the direct control of the agent, but are very important to the action, such as the activity of ones nerves and neurons in producing the action. However, there are other aspects that are under the control of the agent, such as the movement of the arm. What is to be considered the basic act is a large debate, whether it is a volition, mental try, or a physical act is just a small part of it. For very good discussions of basic acts see Brand and Walton 1976. 83 This may point me towards a theory on the problem of the Ship of Theseus, but I do not want to take any particular stance here. 36 into account the causal and temporal relations of all constituent events. If Sirhan’s gun had fired before he pulled the trigger, however this may be possible, we would not want to consider this an intentional killing of Kennedy by Sirhan. Not every constituent event plays such a drastic role in defining the event. For instance, if we consider two killings of Kennedy by Sirhan, killing A and killing B, with the only difference being a single neuron that fired in causing the pulling of the trigger in A did not in B we would still consider B Sirhan’s killing of Kennedy. However, A and B would be a different actions because they inhabit distinct spatiotemporal regions. 2.7 Conclusion A componential view provides advantages over both rough and fine-grained views of action individuation due to temporally extended actions. This also allows this view to avoid complaints presented by some theorists who argue that actions are not to be considered events. 84 Most importantly, it leaves open an easy discussion of how to describe intentionally performed events. A major issue with accepting the fine-grained view is that two descriptions of a murder, such as “the murder of Caesar,” and “the vicious murder of Caesar,” could be considered different actions, particularly when events are as thin as Chisholm holds them. Goldman would say that such a distinction would not be made by his theory, for he allows some alternate descriptions, but his theory still fails for being too thin. Davidson’s events were too roughgrained and did a proper account of event individuation. The componential view allows for such a distinction to handle temporally extended actions. 84 See Lowe 2002, Bach 1980, and Alvarez and Hyman 1998. I will not get into these arguments here, but they focus on similar complaints I provided towards both rough and fine-grained theories and I believe that their criticisms can be answered by the theory of events I present. 37 Chapter 3 Intentions in Action To complete the discussion of agency, we must now look at what it is to do something intentionally. For this, I will begin by discussing what is known as the Simple View. The Simple View states that whenever something is done intentionally it must follow from an intention to perform such an act. This seems to be obvious at first, but as will be shown, there are many good arguments against it. After rejecting the Simple View, it will be necessary to find a clear definition of what it is to do something intentionally. I will then consider an account presented by Alfred Mele and Paul Moser.85 This account builds on previous theories presented by Carl Ginet86 and Brand87, but adds a skill requirement for acting intentionality. I will consider some issues involved with this and present an argument in favor of maintaining a skill requirement. However, before getting into a discussion of the simple view debate we will look at Davidson’s causal theory of action, as it is spelled out in “Actions, Reasons, and Causes,” which holds that the reasons for one’s actions are the causes of that action. The importance of this paper on the history of philosophy of action is that it brought a causal theory back to prominence, for Davidson cites Aristotle as a source of inspiration and the first causal theorist. The main interest I take in “Actions, Reasons, and Causes,” is to set the background for a causal theory and a discussion of intentions. 85 Mele and Moser 1994 reprinted in Mele 1997 Ginet 1990 pg. 87 87 Brand 1984 pg. 28 38 86 3.1 Intentions Davidson’s causal theory has been massively influential in action theory, being the theory that has garnered the most supporters in the past fifty years. I will briefly walk through his essay, “Actions, Reasons, and Causes,” (from here on referred to as ARC) and discuss several points that I believe are the most important in understanding his theory. The causal theory outlined in this paper, and the others that followed, has become what is now called the “standard story” of action.88 Although his theory changed some over time I am focusing on ARC first because it laid the foundation, and was expanded on in his later papers. Davidson begins ARC by asking, “What is the relation between a reason and an action when the reason explains the action by giving the agent’s reason for doing what he did?” Davidson calls this relationship a rationalization, saying that the reason rationalizes the action. He also holds that this connection, between rationalization and action, is a causal connection. Davidson also points out that this essay is a redeployment of an ancient account that many of his contemporaries felt had been laid to rest. 89 To Davidson a reason rationalizes an action if it “leads us to see something that the agent saw, or thought he saw, in his action,”90 that led him to perform the action. A reason is more than simply stating that a certain action appealed to the agent, it must explain what it was about the action that the agent liked. With this foundation, Davidson’s account of doing something for a reason requires two mental states of the agent; (a) a pro-attitude (understood very broadly to 88 Hornsby 2003 pg. 4 He is referring to Aristotle in this case, and in fact some important aspects of Aristotle’s account are central to his later discussions. 90 Davidson 1980 3 39 89 include wantings, desires, obligations, moral stances, aesthetic view, as well as life-long stances or fleeting desires) towards actions of a certain kind, and (b) that the belief (understood very broadly to include knowing, perceiving, noticing, remembering...) that the action to be performed is of that certain kind.91 When an agent gives a reason it may be the pro-attitude, the belief, or both, and this is to be called a primary reason. Davidson then provides two theses to give the paper direction. (1) In order to understand how a reason of any kind rationalizes an action it is necessary and sufficient that we see, at least in essential outline, how to construct a primary reason. (2) The primary reason for an action is its cause.92 He then begins arguing the first point. To begin, Davidson points out that the two aspects of primary reasons discussed before (belief and pro-attitude) are not sufficient for connecting the reason to the action. Davidson presents his now famous example of someone (a) flipping a light switch, (b) turning on a light, (c) illuminating a room, and (d) alerting a prowler. An agent may provide a reason for flipping the light switch; say to turn on the light. This is okay under some frames of questioning, however, Davidson holds the rough grained view to action individuation, which would take this scenario to be one action under four different descriptions, instead of four closely related, separate actions (fine grained view), or four components of one actions (componential view). What this means is that, under the current account of acting for a reason, the reason given for 91 92 Ibid Davidson 1980 pg. 4 40 flipping the switch can be provided as a reason for all of the outcomes of this scenario.93 This does not separate the reason from other descriptions of the action, leaving it possible to sight the agent’s reason for alerting the prowler to be that he wanted to turn on the light. Davidson then refines his account of primary reasons to; (C1) R is a primary reason why an agent performed the action A under the description d only if R consists of a pro-attitude of the agent towards actions with a certain property and a belief of the agent that A, under the description d, has that property.94 This makes primary reasons description specific so that the reason given for flipping the switch is not to be assumed as the reason for other possible descriptions of the action. Wanting to turn on the light is no longer a possible reason for alerting the prowler. Davidson then turns to one possible criticism. He points out that the phrase “I turned on the light,” has a particular event as its object, and one might contest that the phrase “I wanted to turn on the light,” has the same object, therefore not making them logically independent. If these two phrases are not logically independent then using one as a reason for the other would be redundant. However, Davidson points out that such a critic would be mistaken because the event that is the object of the phrase, “I turned on the light,” has every detail fixed, while the object of “I wanted to turn on the light” can be fulfilled in any number of ways. This means that the object of “I turned on the light,” cannot be the object of “I wanted to turn on the light.” Also, the event of turning on the light must have occurred for the phrase “I turned on the light” to be true but no 93 Unknowingly alerting a prowler may not be considered an intentional action, but Davidson points out that it should not be inferred from this that it is a separate event. 94 Davidson 1980 pg. 5 41 such event must have happened for the phrase “I wanted to turn on the light” to be true, Rendering the two statements logically independent. Next, Davidson discusses the causal relation between rationalization and an action. He begins by pointing out that reasons, as he has laid them out so far, are not sufficient for his account. A primary reason cannot simply be the agent having a certain pro-attitude and belief relevant under a certain description, for an agent may have a reason, perform the action, and not have performed the action for that reason. Davidson points out that “central to the relation between a reason and an action it explains is the idea that the agent performed the action because he had the reason.”95 What Davidson believes is that we need an understanding of the role that “because” plays in connecting an action with its reason. He then looks at teleological explanations of action, focusing on an example given by A.I. Melden.96 The theorists that subscribe to such accounts of action explanation claim that, when explaining an action, the agent simply re-describes the action so as to fit it into a coherent pattern. Two problems with this account of action are then pointed out. The first issue is that these theorists take the action as separate from its causes, and therefore the re-description is also separate. What this would mean is that, in explaining an action, the agent does not need to cite a cause. However, Davidson points out that this is a mistaken belief and that explaining an action may simply be an appeal to its cause. Davidson leaves this section simply by saying that the “because” in a reason is to be taken as a causal account, since there is no better, or even equally successful way in which to understand the role that it plays. He then spends the rest of ARC defending it from several possible criticisms, leaving the reader with this account of his causal theory of action. 95 96 Davidson 1980 pg. 9 Melden 1961 42 With this it should be clear how this theory answers the problem of connecting reasons to actions, for they are the cause of actions, and how we have knowledge of our actions, for they have the appropriate beliefs and desires. The advantages of a causal account are made fairly apparent in this paper, being that it very easily connects one’s reason for acting with one’s action. While this is in no way a definitive account of action explanation, it is very influential and easily defended. However, there are several issues that I take with his account of reasons. First I will make clear a few things that are apparent in Davidson’s theory. Davidson focuses on the explanatory power of a reason, its ability to make clear what the agent wanted to do. In doing this, he sets up what is called the reductive account of intentions. He explicitly states that there is no separate state of intention, and to say that someone has an intention is the same as saying that he has a desire to perform a certain action and a belief of how to accomplish this action.97 This reductive view has been called into question lately by several theorists, claiming that a reductive account places great restrictions on accounts of practical reasoning and the function of intentions in action.98 One very important claim against the reductive accounts is that intentions are very different from other states, particularly desires, because of an implied constraint of rationality. If an agent were to hold two opposing intentions, where, should the agent satisfy one intention it would be impossible to satisfy the other, this agent is open to criticisms of irrationality. If someone were to have the intention to be on campus for a two o’clock seminar but also had the 97 This is a claim of his that changes throughout his writings. Other reductive theorists are Monroe Beardsley 1978 and Robert Audi 1973. 98 Enc 2003, Bratman 1981, 1983, and 1984, to name a few. 43 intention to be across town at the golf course for a two o’clock tee time, this would certainly show them to be irrational. A Davidsonian reductive account does not readily lend itself to an explanation of why someone may not have opposing intentions. It is no problem to have competing pro-attitudes, the same agent may have equal desires to go to both the seminar and to the golf course, but the issue arises with the belief. Beliefs share the rational constraints of intentions, but as Davidson employs beliefs, they do not provide the proper rational support. Davidsonian beliefs are simply about how to accomplish the desired action, so one may have a desire to be at both the seminar and the tee time, and a belief about how to accomplish both of these, and therefore have an intention for both of these actions. The role of beliefs needs to be altered if a reductive account is to solve this issue. However, reductive accounts have taken on different forms, as defended by Monroe Beardsley and Robert Audi.99 The change in the belief is not simply how to satisfy a desire, but that the agent believes he will satisfy the desire. More simply, Agent A has an intention to B iff (1) A has a desire to B and, (2) A has a belief that he will B.100 The exact wording of the second part may vary, for some people believe that an agent must believe that he will do B and others hold that the agent must believe that he will not, not B. I 99 Beardsley 1978 and Audi 1973. Other reductive accounts come from Casteneda and Sellars, both reducing intentions to desires, Sellars saying desires are “long term, dispositional intentions.” (cite) 100 Adapted from Brand 1984 44 know this is a confusing distinction, but it is one that can make a difference in what one considers an appropriate intention.101 This reworked version of the reductive account provides a solution to the previous complaints. However, it is still open to several criticisms. The first, coming from Brand102, is that desiring is not a requirement for having an intention. Even the broader term pro-attitude does not work for Brand. Brand’s argument is that the conative aspect of action, which is usually attributed to the desire, comes in many different forms, whether they are desires, obligations, or some other attitude which he believes should not necessarily be lumped together. He also finds the general term pro-attitude to be of no use because the conative state can either be a positive attitude towards something, say a desire to leave the room, or a negative attitude towards something, perhaps a desire to not be stabbed by the murderer in the room. Brand argues that one can have both the positive and negative attitude at the same time and that they could hold different positions in the brain, making them distinct.103 This type of reduction is quite controversial, and so it is not an argument I am willing to adopt. The argument I believe to be stronger concerns the functional roles of intentions. Intentions play a distinct functional role in the monitoring and control of actions.104 When one forms an intention, this intention will guide his action through to its completion. As McCann points out, if he forms the intention to go to the Burger Deluxe Bar, and it comes time to enact his intention, this intention will guide him on his path to Burger Deluxe. If the road he 101 For good discussion of this type see Wayne Davis 1984, (reprinted in Mele 1997), and Mele 1989b 102 Brand 1984 pg. 121 103 Brand 1984 Chapter 4 104 McCann 1991 and Bratman 1984 45 commonly takes to get there is closed, this intention will help guide and correct his route. A reductive account of intentions does not lend itself to an explanation of how intentions are successful monitors of action. Reductive accounts easily provide rationalizations for actions but do not explain the role of intention after playing its initiating role. It may be that reductive accounts do not want to posit a role for intentions beyond the initiation of action, but this then leaves an explanatory gap between the initiation of an action and its completion.105 Another functional role of intentions is in future planning.106 If one decides to go to the Burger Deluxe then, not only has one come to a decision about what one will do at lunchtime, but he now has a reason to go to the bus station at lunchtime to get to Burger Deluxe.107 Once formed, the intention has a bootstrapping affect for other associated actions and intentions. An example provided by Bratman makes the issue more apparent.108 If I am stuck in choosing between going to Law School or Med School, and my beliefs and desires make the internal debate come out in a draw, simply deciding, say by the flip of a coin, provides a reason to make many other decisions. If I decide to go to med school, I now have a reason to take the MCATs and to fill out the appropriate application. While the original reasons for applying to med school may have been what Bratman calls desiderative reasons, it now adds a further non-desiderative reason for taking the appropriate steps to accomplishing my goal of applying. These intentions for the appropriate steps take other intentions into account. This is an aspect of the rational constraint of intentions. The fact that a decision has been made plays an important role in planning and reasoning, because intentions provide a reason for creating further intentions about 105 McCann 1991 pg.207 Ibid 107 Adapted from Bratman 1981 108 Ibid 106 46 intervening steps. This role in planning is beyond the explanatory power of the reductive accounts.109 It can be argued that this bootstrapping can be accounted for by a desire to avoid such incoherence, however, Bratman points out that such a higher-order desire is not required to account for coherency of plans, and is therefore superfluous.110 Bratman says that “if I am going to plan at all I have a reason to have coherent plans, just as I have to reason to have coherent beliefs if I am going to have beliefs at all.”111 Plans must be rational if they are to serve any useful function, and adding the desire to be coherent is unnecessary. These two criticisms point towards a separate mental state of intending. Keeping the arguments for the functional role of intentions we will now look at what is known as the Simple View. This view holds that all cases of intentional action issue from an appropriate intention. 3.2 The Simple View The simple view, as stated above, is the claim that all intentional actions issue from an intention to perform that specific act. Exactly how one defines intention, whether it is reductive, a positive belief of accomplishment, or any other variation is not important, all that matters is that one holds to the claim that all intentional actions are preceded by intentions.112 The defenders of the Simple View argue that only a full-fledged intention, meaning an intention that has a specific action as its content can issue in an action.113 109 Michael Bratman 1981 and 1984. Bratman 1981 111 Bratman 1981 pg. 253 112 As we will see, the arguments against the Simple View lead towards a non-reductive account of intentions 113 This will become important later. 47 110 While the Simple View seems to be self-evident there are several arguments that bring up issues. One of them plays on the belief content of an action. It is generally agreed that one cannot intend to do what he believes he cannot. The first scenario involves someone trying to back his car out of his driveway, but the night before a fierce storm knocked a large branch across his driveway. The branch looks too heavy for the man to move by himself, and so he figures he will probably need to call a tree service to have it removed. However, before calling the service, he decides to try to move the branch. He actually believes that he cannot do so, but will try. To his surprise he is able to move the branch. We would certainly say that he moved the branch intentionally, but such a scenario runs afoul of appropriate belief content. Why is it that this action is considered intentional, even though the agent believes he will fail, while other scenarios, such as shooting a full court basketball shot where one believes he will fail are not intentional. I will return to this discussion after the second scenario.114 The second scenario involves an ambidextrous video game player who plays two of the same game simultaneous, each with one hand. The game involves controlling a missile and leading it into a target. There is a prize for hitting either one of the targets and the games are linked so that, if either target is hit both games will end. Also both games will shut down if both targets are about to be hit simultaneously, causing the player to lose. Despite the fact that the player may only hit one of the targets she is of the skill level where it is more sensible for her to play both games, even with the risk of shutting down both games. If she hits the first target this would certainly be intentional, owing to her skill in the game and not luck. However, the same is true if she hits the second target as well. Since she cannot hit both targets she cannot have an intention to hit either one, because that would require having incompatible intentions, which 114 This scenario and the following on were originally presented by Bratman 48 would violate the rationality criteria from above. To intend to hit either target, while sincerely playing both games, would require our agent to be irrational. However, we do not consider the agent irrational in playing both games and we consider a victory to be intentional, regardless of which target she strikes. 3.3 Solutions in Place of the Simple View Those who oppose the Simple View do not claim that there is absolutely no intention involved in the above scenarios. The point of contention is over what it is an intention to do. Those who support the Simple View would say that to perform A one must intend to A. Bratman and Mele provide two different answers to what exactly the intention may be, and why we would call the successes in the above scenarios intentional. Bratman believes that intentions, not only being important in issuing actions, have what he calls a motivational potential.115 Motivational potentials are a part of intentions that actually cover a broader base of actions. If one has an intention to perform an action, there is something similar to a one to one correlation between the action and intention. However, the motivational base of an action incorporates the performing of similar or related actions, so that these actions may be considered intentional. In the case of moving the log, the agent may have the intention to pull on the log with all of his strength, and included in this motivational potential is moving the log. This accounts for why we consider the action to be intentional. As well as positing motivational bases for intentions, Bratman discusses guiding desires, which he claims replace the functional roles of intentions in cases like the video game scenario. These guiding desires are to provide the same action directing and planning, yet, being desires and not intentions, are not 115 Bratman 1987 pg. 119 49 subject to the same rationality constraints. However, seeing as how Bratman argues for the necessity of intentions in acting and planning, arguing affectively against reduction, he does not provide a clear explanation of what a guiding desire is to be, and why it is able to take the place of intentions.116 Being a desire leaves it open to many of the arguments provided above.117 Also there is not clear reason given as to why guiding desires are not the standard case for all actions as opposed to intentions, since they serve the same function. Mele provides a solution that maintains the use of intentions, so as to avoid issues involving guiding desires, but the content of the intention is altered. The Simple View requires that intentionally doing A requires an intention to do A, but Mele argues that intention must be to try to A. Mele does not want to posit this for all actions because he holds that we do intend to do some things directly. Intending to try is simply for the cases where there is a belief that the agent will fail but an attempt is made anyway. This preserved the role of intentions in issuing actions. Mele presents another scenario, of two basketball players standing at half-court. One is about to take a half-court shot, but before doing so the other asks if she intends to make the shot. Her reply is simply “No, but I intend to try.”118 For Mele this is to be taken literally. The shooter has an intention to try. McCann argues that trying is not something specific to any action but is a part of all actions. For this, McCann rejects trying as an appropriate thing to intend to do, for it is not actually an action. For McCann, intentions to try are either vacuous, or to intend to try to make the half-court shot needs to be supplemented with an intention to make the half-court shot which brings us back to Simple View. Mele contends that the vacuity of an intention is based on the 116 Chan 1999, pg. 4 See brand 1984 118 Mele 1989 117 50 goal and act plan of the intention.119 An intention to try has an appropriate act plan, which is to aim carefully and throw with all my might, but the goal is slightly different, namely to try and not to necessarily make the basket.120 Mele’s solution of intending to try is a way of getting around the complaints of the simple view, while maintaining intentions as what issues in actions. Positing desires to take the functional role of intentions either requires these guiding desires be subject to the same rational constraints as intentions, for it is these rational constraints that allow intentions to factor into planning and monitoring actions, or these are subject to the same criticisms that Bratman leveled against reductive theories of intentions.121 Incorporating intentions to try into intentional action parries the issues involved in both the Simple View and with Bratman’s theory of guiding desires. Getting rid of the Simple View leaves us with some unresolved issues however, most importantly what it is to do something intentionally. How are we to distinguish one’s intentional moving of the branch (from above) and the unintentional straining of his back while doing so? 3.4 Intentional Actions Defining what it is to do something intentionally is quite difficult, but I believe that the following discussion, including Davidson, points us in the right direction. We started with the Simple View, which told us that to A intentionally requires an intention to A. This could be expanded further to say that to A intentionally is to have an intention to B, where B is a means to A, and where B is the basic action required to A. However, this was found to be unacceptable 119 Mele 1989. For a good discussion of the necessity of act plans in intentions, one that Mele adopts, see Brand 1984. 120 Mele 1989 121 Chan 1999 pg.5 51 because it required agents to violate the rational roles of intentions. This leads us to a further discussion of what it is to do something intentionally. In this section, I will examine an account of intentional action that is presented by Myles Brand, consider some objections and conclude on an account of intentional action that includes a skill requirement. First, let us look at Brand’s account of intentional action. 3.5 Brand’s Intentional Action Very important to Brand’s account of action is the role of act plans in intentions. He is adamant about the role of act plans in forming and enacting appropriate intentions. For instance, if someone has a very rich uncle and is his only family, he may decide to kill him in order to inherit his wealth. This decision, forming an intention, leads the nephew to create a plan on how to do this. He decides that the best way to kill his uncle and get away with the murder is to use his key to get into his uncle’s mansion while he is out on his nightly walk, which is always at eight o’clock, wait for him to return, and shoot him as he comes in the back door, as he always does after his walk. This intention has formed a complete act plan, although minor details may not be completely filled in. Now let us say that the nephew arrives at his uncle’s mansion and finds out that his uncle has changed his locks, for whatever reason. The nephew now has to sneak in through the window and wait for his uncle to return. Upon shooting his uncle, the murder would still be considered intentional even though it did not follow the act plan exactly. However, if the nephew were to lose control of his car and run over his uncle on his way to the mansion, this clearly would not be an intentional murdering of his uncle, even though it satisfies his intention to kill his uncle. With these scenarios in mind, we can look at Brand’s account of intentional action. 52 Brand holds that intentional actions satisfy the following criteria, (BD) S’s Aing during t is an intentional action iff (1) S’s Aing during t is an action; and (2) either S has an action plan P to A during t such that his Aing is included in P and he follows P in Aing or S’s Aing during t satisfies (prin. S, prin. Im, or prin In).122 The three principles noted at the end are the principles of action substitution, improvisation, and interruption. The only one of importance for our purposes is that of substitution, which is, (S) A’ may be substituted for A if (1) A’ing is an action, (2) S’s A’ing is a substitutive action for S’s Aing in regards to P, and (3) S is not able to A. This was shown in the case of the window entry instead of entering through the door.123 The principle of improvisation usually comes with the principle of substitution. In the case of entering through the window, it satisfied the principle of improvisation because it was not part of the original act plan. Due to unforeseen issues with the locks, climbing through the window was incorporated to ensure the completion of the overall act plan. The principle of interruption allows for one to intentionally end the act plan in case of an unforeseen obstacle that makes it more reasonable to stop. This account of intentional action ties what it is to do something intentionally in with one’s act plan. Mele and Moser point out an issue with this account of intentional action. Suppose that someone is working at a nuclear reactor that is in the middle of a meltdown, and he needs to shut it down. He looks at the code given to him in case he has to shut down the reactor, but misreads 122 123 Brand 1984 pg. 28. The previous discussion has been taken from Brand 1984, particularly pg. 28. 53 it. He then goes to put in the misread code, but with great luck the code was written down incorrectly and in misreading what was written down he entered the right one. The shutting down of the reactor would then be intentional under Brand’s theory, even though a great amount of luck played a role. It followed the action plan perfectly, but required far too much luck to actually be considered intentional. This leads Mele and Moser to provide a skill requirement, which requires that an action not diverge too far from the act plan, that S’s action contains a suitably reliable skill in S’s act repertoire, and that S’s action is a predicatively reliable skill. The second two additions are pretty easy to fill out so as to satisfy many of the requirements of action but the first needs some further explanation and expansion. This addition is used to stop such issues as presented to Brand’s theory. Simply it is to state that S’s Aing, as it actually is needs to be relevantly close to how it is that Aing fits into P for S. Therefore, such issues of luck are to be excluded from intentional action, for they are deviations from the actual act plan. 54 Chapter 4 Synthesis Now it is important to synthesize all that we have discussed so far, so as to bring it together into a coherent account of agency. I will briefly go over each discussion again so as to bring it together in a clearer light. To begin, I will recap the discussion of agency. Our discussion of agency focused on the cognitive capacities of agents. Most importantly, I argued that humans are a type of agent that can represent the world in terms of propositional attitudes, and that these propositional attitudes are real, not just accurate descriptions of patterns of behavior. I relied on some of the instrumentalist arguments against eliminative materialism to posit the use of folk psychological descriptions. Then, I defended the reality of propositional attitudes as causally efficacious entities in the mental life of agents. After a discussion of folk psychology, I provided an account of distinguishing actions from events. This went hand in hand with the theory of agency outlined earlier, in that actions are things that are intentionally done by agents. This account of actions requires the ability to provide different descriptions of actions, which was taken up in the next chapter. After this, I moved on to a discussion of event and action individuation. This is important because the ontological status attributed to events determines how we can describe them. For instance, if events are considered to be propositions like Chisholm argues, they would be very thin, and the descriptions of an event like “the murder of Caesar” and “the vicious murder of Caesar” would be considered different actions, even if they shared all other properties. I argued that events are particulars and that like particulars they can be made up of several parts. From 55 this I argued that a componential view of action individuation would be the best because it can handle the issues presented by temporally extended actions. Finally, I looked at what it is to consider something an intentional action. I rejected the simple view because it would require an agent to hold incompatible intentions, which is something I find unacceptable. Then, I looked at Brand’s theory for distinguishing intentional action, and found that it lacked the appropriate skill requirement. With that I added a requirement so that one’s actions are a reflection of his skill. 4.1 Intentions and Parts of Actions The remaining issue of concern is how to apply the discussion of intentional action to that of the discussion of action individuation. If actions are to be made of parts, some of the parts may not be something that the agent actually does. Davidson points this out with his example of throwing a ball through a window.124 Some things we do, the rest we leave up to nature, for when a ball is thrown through a window, after it leaves the agent’s hand the forces acting on the ball are not under his control. I will look at Davidson’s solution to this issue. Even though he has a different account of action individuation his thoughts on the matter can be of some use. First, I will go back to the discussion of Davidson’s paper “Actions, Reasons, and Causes. In this paper Davidson makes the claim that one’s reasons for acting are the causes of his actions. Most importantly, he pointed out the mistake which had been at the base of action theory which was that reasons and actions were logically connected, and therefore were not causally connected. Davidson relied on the Aristotelian argument that the result of practical reasoning is 124 Davidson 1980 56 an action, and that the reasoning caused the action. Having one’s reason cause an action is important in the action being intentional.125 Although Davidson originally insisted on a reductive account of intentions, he later changed his view to include intentions as a mental state all to themselves. This has become the most generally accepted stance on intentions. However, with intentions being considered a mental state all to themselves, it is now intentions that are the causes of actions. This is a key aspect of an action being intentional, that it is correctly caused by one’s intention. This claim is not without its issues though. Exactly defining what it is for an intention to correctly cause an action is a difficult thing to do. Consider our murderous nephew from before. Let us say his plans are different and that he is going to drive over to his uncle’s and run him over while he is mowing his lawn.126 On his drive over to his uncle’s mansion his intention, and the thought of what he is about to do, agitates him in a way so that he loses control of his car and runs over his uncle, as he had planned. With the other requirements for doing something intentionally, involving act plans and skill, this would not be intentional. The claim made by Chisholm, and many others, is that the intention caused the nephew to run over his uncle, yet this is clearly not something that we would call intentional. The difficulty is differentiating this case from one where the nephew’s intention caused the killing of his uncle appropriately. There are several responses to this counterexample. One is that it assumes a transitivity of causation, which is not accepted in many cases of causation. Irving Thalberg presents an 125 126 Davidson 1980 pg. 3-19 Adapted from Chisholm 1966 57 example adapted from Dorothy Mitchell.127 This scenario involves a water skier who is bitten by a mosquito, which causes him to become distracted and run into a boat launch. The resulting trauma from the collision causes the water skier to die.128 The claim made here is that to say the mosquito bite caused the death of the water skier is mistaken. An example that is more convincing is that of begetting.129 It would not be acceptable to say that one’s grandparents caused one to be born. This reply requires taking a stance on the issue of transitive causation, which is not something that I intend to do here. For another reply we turn to Brand. Brand argues that the problem is that causal theories allow for a space between the mental causes and the action. Davidson, and similar theorists, felt that discovering how intentions caused actions was a job for psychologists or experimental scientists130. Brand argues that simply adding a clause for proximate causation will solve the issue. Instead of a psychological examination of what it is for the mental events to cause the action, Brand presents a metaphysical answer. He defines proximate causation as E proximately causes F iff (1) E causes F and (2) there are no events G1, G2,…Gn (where n is less than or equal to 1) such that E causes G1, and G1 causes G2,… and Gn causes F.131 Adding the requirement of proximate causation settles the issue for causal theories, for actions can now be defined as those events that are proximately caused by the antecedent mental events. 127 Thalberg 1984 pg. 250 Ibid 129 Mitchell 1982. She uses the term begetting. 130 Brand 1984 Chapter 1 131 Brand 1984 pg. 20 128 58 4.2 Causation and Parts of Actions If an action is to be proximately caused by the intention, what parts of an action require causation? As stated above, not all parts of actions are something that we necessarily do, so what must be caused to make a complete event an action? Davidson provides an answer, and this is where I will start. Davidson’s claim is that all of our actions are the movements of our bodies. When a sniper shoots his victim, the action is limited to the flexing of the finger. Part of this is a result of his account of action individuation; however, it has some credence in a componential view as well. In cleaning one’s house, a cleaning product may be used that one is to let sit while it does its job.132 If asked whether or not he is cleaning, it is certainly acceptable for the agent to respond affirmatively, however, there is very little that he is actually doing. It would seem reasonable to limit his action to just the application of the cleaning product, limiting actions to what exactly it is that the agent does. Therefore, intentional actions could be defined as the intentional causing of the bodily movement. However, this account of actions brings us back to Davidson’s unifier account of action individuation and the issues with the by relation will persist. We can now turn to a suggestion by Thomson.133 In her discussion of the various accounts of agency, she discusses and sets up a system for both causal and cognitive accounts. She claims that a causal account can be set up as an appropriate group of causings. She argues 133 134 Adapted from Thomson 1977 Thomson 1977 pg. 255 59 that this formulation will not work for the cognitive account, which may be true, but there is something to be taken from this.134 Considering intentions as the cause of actions and, as argued in a previous section, the role of actions plans in creating and maintaining intentions, we can incorporate Thomson’s causal account. We can simply state that an action is an intentional group of connected causings, where every event that makes up the action fits into the act plan or satisfies one of the substitutions clauses. For instance, in the case of an assassin killing his victim, after pulling the trigger, the assassin no longer does anything to shoot his victim. However, in the act plan it requires that the bullet fly through the air and strike his target causing the appropriate trauma. These being part of the act plan must be fulfilled for the action of killing his victim. The assassin may be successful in shooting his victim, but not killing his victim. In this case there would still be an intentional action performed, namely that of shooting his victim, even though he did not complete his intended action, that of killing. This is a case of performing B in order to achieve A, for the assassin thought it best to shoot his victim to kill him. In this case, part of the intentional action of killing requires the intentional shooting of the victim. The act plan does not need to be exactly fulfilled as laid out, however. If the assassin fired his gun at his victim thinking he was firing through an open window at his target but it turned out that the window was closed and the assassin shattered the glass before hitting his victim, this does not make the action unintentional. This provides an unintentional consequence, but does not make the action unintentional. Also, the assassin may have a terrible knowledge of human anatomy, and thinks that shooting the victim through the heart will cause a heart attack 135 Ibid 60 instead of hemorrhaging. Even though the belief about what is going to happen, and even though the assassin intends to cause a heart attack, the killing is still intentional. This is because these deviations from the act plan do not affect the skill required in the action of killing. While it can be said that the assassin unintentionally caused hemorrhaging, it cannot be said that he unintentional caused the death of his victim. The skill requirement also relies on the act plan, for the events that make up the action must fit into the action plan.135 However, parts of actions are not under the control of the agents, such as the actual flight of the bullet. To fit this into the skill requirement it must be amended that those parts of the action that are under the agents control, aiming the gun and pulling the trigger, must have an appropriate effect on the parts of an action that the agent does not directly control, so that his skill will affect the complete action. 4.3 Conclusion For an action to be intentional in must be caused by an intention, which cannot be reduced to other mental events or states. Creating and managing intentions are act plans and for an action to be intentional, its parts must fit into the act plan or satisfy one of the substitution rules. An unforeseen consequence does not make an action unintentional as long as it does not affect the skill requirement the action is intentional. 136 I keep saying the events that make up the action. This may be misleading because an action can have other actions as constituent parts but this would require a discussion of basic actions. I use events because actions are a type of event. 61 Concluding Thoughts This thesis has been an attempt to shed light on what it is to be a human agent. What it takes to present a complete theory of agency is quite massive, and it was not my intention to do so. However, there are several aspects of agency that I find to be of great importance, particularly those relating to the cognitive functions of humans. In discussing such cognitive functions there are several issues that need explication, namely the ontological status of events and how they are represented in the agent’s propositional attitudes and plans. This is what I attempted to cover in this thesis. 62 List of References 1. Alvarez, Maria and Hyman, John. “Agents and Their Actions.” Philosophy (1998) 73; 219-245. 2. Anscombe, G.E.M. 1963. Intention. Second Edition. Ithaca: Cornell University Press. 3. Audi, Robert. “Intending.” Journal of Philosophy (1973) 70; 387-402 4. Bach, Kent. “Actions are Not Event.” Mind (1980) 89; 114-120. 5. Belnap, N. and Perloff, M. (1992). “The way of the agent.” Studia Logica, 51, 463–484. 6. Belnap, N., Perloff, M.,and Xu, M. (2001). Facing the Future: Agents and Choices in Our Indeterministic World. Oxford: Oxford University Press. 7. Bermudez, Jose Luis; ed. 2006. Philosophy of Psychology: Contemporary Readings. New York, Routledge. 8. Binkley, Robert; Bronaugh, Richard; Marras, Ausonio; eds. 1971. Agent, Action, and Reason. Toronto, University of Toronto Press. 9. Beardsley, Monroe. 1978. Intending. In Values and Morals, ed. A. Goldman and J. Kim. Reidel. 10. Braddon-Mitchell, David and Jackson, Frank. 1996. The Philosophy of Mind and Cognition Oxford, UK ; Cambridge, MA : Blackwell Publishers. 11. Brand, Myles. 1976. “Particulars, Events, and Actions.” In Action Theory. Eds. Myles Brand and Douglas Walton. pp. 133-58. 12. _____ 1984. Intending and Acting; Toward a Naturalized Action Theory. Cambridge, Mass. MIT Press. 13. Brand, Myles and Walton, Douglas; eds. 1976. Action Theory. Boston. D. Reidel Publications. 14. Bratman, Michael. “Intention and Means-End Reasoning.” The Philosophical Review (1981) 90; 252-265. 15. _____ "Taking Plans Seriously," Social Theory and Practice (1983) 9; 271-287 16. _____ “Two Faces of Intention.” Philosophical Review, (1984) 93; 375-405. 17. _____ 1987. Intention, Plans, and Practical Reason. Cambridge, MA: Cambridge University Press. 18. Casteneda, Hector-Neri. “Intensionality and Identity in Human Action and Philosophical Method.” Noûs, (1979) 13; 235-60. 63 19. Chan, David K. “A Not-So-Simple View of Intentional Action.” Pacific Philosophical Quarterly (1999) 80; 1-16. 20. 517. Chellas, B. “Time and modality in the logic of agency.” Studia Logica (1992) 51, 485– 21. Chisholm, Roderick. (1966) "Freedom and Action", in K. Lehrer, (ed.) Freedom and Determinism (NY: Random House), pp. 11-44. 22. ¬¬¬_____ 1976. Person and Object: A Metaphysical Study. La Salle, Illinois. Open Court Publishing Company. 23. Paul M. Churchland and Patricia S. Churchland. 1998. On the Contrary: Critical Essays, 1987-1997. Cambridge, Mass.; MIT Press. 24. Cleveland, Tim. 2011 Comments on a paper presented to the New Mexico West Texas Philosophical Society annual meeting. 25. Cornman, James. 1971. “Comments.” In Binkley, Robert; Bronaugh, Richard; Marras, Ausonio; eds. 1971. 26. Davidson, Donald. “Actions, Reasons, and Causes.” Journal of Philosophy (1963) 60; 685-700. (Reprinted in Davidson 1980, pp. 3-19.) 27. _____ 1969. “The Individuation of Events.” In N. Rescher, ed., Essays in Honor of Carl G. Hempel, Dordrecht: Reidel, pp. 216-34; reprinted in Davidson (1980), pp. 163-80. 28. _____ 1970a. “Events as Particulars.” Noûs, 4, 25-32; reprinted in Davidson (1980b), pp. 181-87 29. _____ 1970b. “How is Weakness of the Will Possible.” In Joel Feinberg ed., Moral Concepts, Oxford. OUP. Reprinted in Davidson 1980 30. _____ 1971. “Agency” In Binkley, Robert; Bronaugh, Richard; Marras, Ausonio; eds. Reprinted in Davidson 1980. 31. _____ 1978. Intending. In Y. Yovel ed. Philosophy of History and Action. Dordrecht, Holland: D. Reidel, pp. 41-60. (Reprinted in Davidson 1980, pp. 83-102.) 32. ¬¬¬_____ 1980. Essays on Actions and Events, Oxford. OUP. 33. Daniel Dennett “Personal and Sub-Personal Levels of Explanation.” In Content and Consciousness, New York, NY. Routledge 1969. Reprinted in Bermudez 2006 34. _____ “Real Patterns” Journal of Philosophy (1991) 88; 27-51. Reprinted in Bermudez 2006. 35. Davis, Wayne. “A Causal Theory of Intending.” American Philosophical Quarterly (1984) 21; 43-53. Reprinted in Mele 1997. 64 36. Donagan, Alan. 1987. Choice: The Essential Element in Human Action. New York, NY. Routledge and Kegan Paul. 37. Enc, Berent. 2003. How We Act: Causes, Reasons, and Intentions. Oxford; OUP. 38. Fuller, Gary and Yu, Paul. “A Critique of Dennett.” Synthese (1986) 66; 453-476. 39. Ginet, C. 1990. On Action. Cambridge: Cambridge University Press. 40. Mary Louise Gille and James G. Lennox eds. 1994. Self-motion: From Aristotle to Newton. Princeton, N.J.; Princeton University Press. 41. 774. Goldman, Alvin I. “The Individuation of Actioin.” Journal of Philosophy (1971) 69; 761- 42. Haselager, W.F.G. 1997. Cognitive Science and Folk Psychology : The Right Frame of Mind. London ; Thousand Oaks, Calif. : Sage Publications. 43. James Higginbotham, Fabio Pianesi, Achille C. Varzi; eds. Speaking of Events. New York : Oxford University Press, 2000. 44. Sarah Hickinbottom, Jack Martin, Jeff H. Sugarman. 2010. Persons : Understanding Psychological Selfhood and Agency. New York ; London : Springer, 2010. 45. Horgan, Terrence. The Case Against Events, Philosophical Review (1978) 87, 28-47. 46. 226. _____ 1986. “Folk Psychology is Here to Stay.” Philosophical Review (1985) 94, 197- 47. Hornsby, Jennifer. 1980. Actions. London. Routledge & Kegan Paul. 48. _____ 2003 “Agency and Actions.” In John Hyman and Helen Steward eds. Agency and Action. Cambridge, Cambridge University Press. 49. 265. Huffer, Beth. “Actions and Outcomes: two aspects of agency. Synthese (2007) 157; 241- 50. William James ; Frederick H. Burkhardt, general editor ; Fredson Bowers, textual editor; Ignas K. Skrupskelis, associate editor. The Principles of Psychology. Cambridge, Mass.; Harvard University Press, 1981. 51. Ernest LePore and Brian McLaughlin eds. Action and Events : Perspectives on the Philosophy of Donald Davidson. Oxford : Basil Blackwell, 1988. 52. E.J. Lowe. A Survey of Metaphysics. Oxford ; New York : Oxford University Press, 2002. 54. William Lyons. Approaches to Intentionality. Oxford : Clarendon Press, 1998. 65 55. Mackie, David. “The Individuation of Actions.” The Philosophical Quarterly (1997) 47; 38-54. 56. McCann, Hugh “Settled Objectives and Rational Constraints.” American Philosophical Quarterly (1991) 28; 25-36. Reprinted in Mele 1997. 57. Melden, A. I. Free Action. London, Routledge & Paul; New York, Humanities Press 1961 58. Mele, Alfred. “Intention, Belief, and Intentional Action,” American Philosophical Quarterly (1989a) 26; 19-30. 59. Mele, Alfred. “She Intends to Try,” Philosophical Studies (1989b) 55; 101-106. 60. Mele, Alfred; ed. 1997. The Philosophy of Action. Oxford, OUP. 61. Mele, Alfred and Moser, Paul K. “Intentional Action.” Noûs (1994) 28; 39-68. Reprinted in Mele 1997 62. Mitchell, Dorothy. "Deviant Causal Chains", American Philosophical Quarterly (1982) Vol. 19; 351-53. 63. Montague, Richard. “On the Nature of Certain Philosophical Entities.” The Monist (1969) 53; 159–94. 64. O’Shaughnessy, Brian. “Trying as the Mental Pineal Gland.” The Journal of Philosophy (1973) 70; 365-386. Reprinted in Mele 1997. 65. Quine, W.V.O. 1960. Word and Object. Cambridge, Mass., M.I.T. Press. 66. _____ 1988 “Events and Reification.” In Ernest LePore and Brian McLaughlin eds. 67. Shueler, G. F. 2003. Reasons and Purposes: Human Rationality and the Teleological Explanation of Action. New York, OUP. 68. Steward, Helen. 2000. The Ontology of the Mind: Events, Processes, and States. Oxford, OUP. 69. Thalberg, Irving. “Singling Out Actions, Their Properties and Components.” Journal of Philosophy (1971) 68; 781-787. 70. Thalberg, Irving. “Do our intentions cause our intentional actions?” American Philosophical Quarterly (1984) 21; 249-260. 71. Thomson, Judith. "Individuating Actions." Journal of Philosophy (1971) 68; 774-781 72. Thomson, Judith. 1977. Acts and Other Events. Ithaca, N.Y.; Cornell University Press. 73. Wittgenstein, Ludwig.1968. Philosophical investigations. Translated by G. E. M. Anscombe. New York, Macmillan 66 Curriculum Vita Gareth Fuller was born and grew up in Hyattsville, Maryland. He graduated from Queen Anne High School in 2005 and entered Guilford College, in Greensboro, North Carolina, that fall. He began his studies as a music major, but ended up with a double major in philosophy and political science. While at Guilford he was a member of the collegiate lacrosse team as well as the Jazz band and he continues to pursue both of these as interests. He graduated from Guilford in the spring of 2009 and started his MA at UTEP that fall. His philosophical interests are varied, covering philosophy of action, metaphysics, philosophy of mind, and a budding interest in the philosophy of language. Other areas of academic interest include Linguistics, particularly Psycholinguistics, and the various cognitive sciences. Permanent Address: 6505 44th Avenue University Park MD. 20782 67