In my earlier books, I developed a critique of the dominant view of the mind, according to which attitudes like beliefs are in the first instance brain states, and I offered an alternative, more pragmatic, approach. See Saving Belief: A Critique of Physicalism (Princeton, NJ: Princeton University Press, 1987) and Explaining Attitudes: A Practical Approach to the Mind (Cambridge: Cambridge University Press, 1995). On my alternative, attitudes — like believing, desiring, and intending — should be understood not primarily as brain states but as states of whole persons. Such a view raises the questions What is a person? and What is the relation between a person and her body? These are the questions that I hope to answer in this book. The answers require a very rich and detailed theory that I call the ‘Constitution View’ In this book, I set out the Constitution View and defend it against criticism and rival views.

I have tried out much of the theory and argument that appears here at departmental colloquia and at conferences where I have given papers recently: Yale University, Notre Dame University, York University (Ontario), Texas Tech, Texas A&M, University of Oklahoma, University of California (Santa Barbara), Whittier College, Utrecht University (Holland), Conference on Lynne Baker's Theory of the Attitudes (Tilburg University, Nijmegen University, Dutch Research School in Philosophy [the Netherlands]), American Philosophical Association (Central Division), the Twentieth World Congress of Philosophy, the Conference on Naturalism (Humboldt University [Berlin]), and the Conference on Epistemology and Naturalism (University of Stirling [Scotland]).

In Choice: The Essential Element in Human Action Alan Donagan argued for the importance of “will” to our shared understanding of intelligent action. By “will” Donagan meant a complex of capacities for forming, changing, retaining, and sometimes abandoning our choices and intentions. (Choice is, for Donagan, a “determinate variety of intending.”) Our capacity to intend is to be distinguished both from our capacity to believe and from our capacity to be moved by desires. And Donagan thought that intentions involve what, following Austin, he called “ ‘as it were’ plans.”

I am broadly in agreement with these main themes in Donagan's book, and I will pretty much take them for granted in what follows. I will suppose that intention is a distinctive attitude, not to be reduced to ordinary desires and beliefs; that intentions are central to our shared understanding of ourselves as intelligent agents; and that “the study of intention” is in part the “study of planning.” My hope is that these common elements in our views about intention can serve as a basis for reflection on the phenomenon of shared intention.

That we do sometimes have intentions that are in an important sense shared seems clear. We commonly report or express such shared intentions by speaking of what we intend or of what we are going to do or are doing. Speaking for you and myself I might say that we intend to paint the house together, to sing a duet together; and I might say that we are going to New York together.

We are planning agents. Our purposive activity is typically embedded in multiple, interwoven quilts of partial, future-directed plans of action. We settle in advance on such plans of action, fill them in, adjust them, and follow through with them as time goes by. We thereby support complex forms of organization in our own, temporally extended lives and in our interactions with others; and we do this in ways that are sensitive to the limits on our cognitive resources. These facts are, I believe, an important key to an adequate philosophical treatment of (1) the very idea of intention, (2) basic features of our agency, (3) important forms of shared agency, and (4) important forms of responsible agency.

I discussed planning agency, and its significance to (1) and (2), in my 1987 book, Intention, Plans, and Practical Reason. Since then I have tried to elaborate and to deepen this approach, to extend it to (3) and to (4), and to explore its relations to the work of others. The present volume of previously published work includes eleven essays that were the result of this effort, as well as a pair of earlier critical studies that also seemed useful to include.

At the heart of my theory is a model of our planning agency and the attempt to use it to understand intention. I call this the planning theory of intention. The main idea is to see intentions as elements of stable, partial plans of action concerning present and future conduct.

REASONING: THEORETICAL AND PRACTICAL

In theoretical reasoning our concern is with what to believe, in practical reasoning with what to intend and to do. In the former we are trying to find out how the world is; in the latter we are concerned with how to (try to) change it. These can seem to be two different – though no doubt importantly related – enterprises. J. David Velleman argues otherwise. His view is “that practical reasoning is a kind of theoretical reasoning, and that practical conclusions, or intentions, are the corresponding theoretical conclusions, or beliefs” (p.15). These twin identifications – of practical reasoning with a kind of theoretical reasoning, and of intention with a kind of belief – are at the heart of this wide-ranging, imaginative, and important study of autonomous agency, rationality, and value. Let us call this pair of identifications “cognitivism about practical reason.” Velleman's cognitivism about practical reason lies at the foundations of his book. I want to know if it is defensible.

SPONTANEOUS SELF-KNOWLEDGE

Velleman's route to his main conclusions begins with reflection on the special kind of knowledge and understanding we normally seem to have of our own intentional conduct. When I act intentionally I usually know what I am doing and I usually know at least a rudimentary explanation of why. Further, I seem to know this in a distinctive way. As Velleman puts it, there is a kind of “spontaneity” in my knowledge: I do not need to step back and observe what I do and infer what my motives are in the way you would need to for you to have such knowledge about me.

Much of our behavior is organized. It is organized over time within the life of the agent, and it is organized interpersonally. This morning I began gathering tools to fix the bicycles in the garage. I did this because I was planning to go on a bike trip tomorrow with my son and knew I needed first to fix the bikes. Just before I gathered the tools, I was on the telephone ordering tickets for a trip to Philadelphia next week. I did that because I was planning to meet my friend in Philadelphia then. If all goes well, each of these actions will be part of a distinct coordinated, organized sequence of actions. Each sequence will involve coordinated actions, both of mine and of others. These sequences will also need to be coordinated with each other. Such coordination of action – between different actions of the same agent and between the actions of different agents – is central to our lives.

How do we accomplish this organization? It seems plausible to suppose that part of the answer will appeal to commonsense ideas about planning. It is part of our commonsense conception of ourselves that we are planning agents (Bratman 1983, 1987). We achieve coordination – both intrapersonal and social – in part by making decisions concerning what to do in the further future. Given such decisions, we try to shape our actions in the nearer future in ways that fit with and support what it is we have decided to do in the later future. We do that, in large part, by planning.

In several recent papers I have sketched a general approach to phenomena of shared agency that do not involve relations of authority (Bratman, 1992; 1993). I focused, in particular, on what I called shared intention, shared intentional activity, and shared cooperative activity. The basic idea was that at the heart of these phenomena is shared intention – a shared intention, for example, to paint the house together. Shared intentional activity, in the basic case, is activity suitably explainable by a shared intention and associated forms of mutual responsiveness. Shared cooperative activity requires, further, the absence of certain kinds of coercion, and commitments to mutual support in the pursuit of the joint activity.

What are shared intentions? My strategy here was two-pronged. I tried to specify roles distinctive of shared intention: roles such that it would be plausible to identify shared intention with what plays those roles. I argued, in particular, that our shared intention to J plays three interrelated roles: It supports coordination of our intentional activities in the pursuit of J, it supports associated coordination of our planning, and it structures relevant bargaining (Bratman, 1993, p. 99 [this volume, p. 112]). I then argued that a certain kind of public, interlocking web of intentions of each of us would play those roles. This supported my conjecture that shared intention could be identified with that web of intentions of the individuals.

Shared intentions are intentions of the group. But I argued that what they consist in is a public, interlocking web of the intentions of the individuals.

The Sources of Normativity derives from the 1992 Tanner Lectures and Seminar in Cambridge, England. It consists of a brief introduction by Onora O'Neill, a “Prologue” and four lectures by Christine Korsgaard, separate discussions of these lectures by G. A. Cohen, Raymond Geuss, Thomas Nagel, and Bernard Williams, and a concluding “Reply” by Korsgaard. It is a major work of the first importance.

We have beliefs and intentions, and we perform actions. In each case we make ought judgments that purport to say what to believe, what to intend, what to do. In making claims on what we believe, intend, or do such judgments are normative, and we can ask what the source, the justification, of this normativity is. Morality, in particular, makes normative claims – sometimes quite demanding – on (at least) what we intend and do. We can ask what the grounds are for the normativity of such moral demands.

Korsgaard's main question – which she labels “the normative question” (10) – is this last question about the “claims morality makes on us.” (10). But her answer involves a general approach to grounds for normative claims on what we intend and do. Her answer – which she sees as Kantian, and notes (99n) has similarities to views of Harry Frankfurt – is that the basic ground of practical normativity is “the reflective structure” of our minds (100). We – normal, mature human beings – have the capacity to be reflectively aware of the desires and inclinations that tend to move us to intention and action.

FRANKFURT'S CHALLENGE

In his 1991 presidential address to the American Philosophical Association, Harry Frankfurt describes the “notion of identification “ [as] fundamental to any philosophy of mind and of action.” This is a striking claim. Standard philosophies of action tend to be rather minimalist. Some are extremely minimalist and include only belief, desire, and action. Others introduce distinctive forms of valuation. Yet others insist further on the need to include, at a basic level, intentions and plans, and the decisions which are their normal source. Frank-furt's effort to focus our attention on “identification” poses a twofold challenge: We need to know what identification is, and we need to know if recognizing this phenomenon requires yet a further, fundamental addition to our model of our agency.

Frankfurt emphasizes that an agent may sometimes see her motivation as “external” even though it is in one straightforward sense hers. This may be the attitude a drug addict takes toward her overwhelming desire for the drug, or a person takes toward his “jealously spiteful desire to injure” an acquaintance, or someone takes toward a “spasm of emotion” that “just came over” him. Seeing one's motivation as external may frequently involve characteristic feelings of estrangement, though Frankfurt does not seem to see these as essential. In contrast, one may sometimes on reflection “identify” with one's motivation; one sees it as grounding action that is, in a sense that needs to be clarified, fully one's own.

We see ourselves as both responsible agents and planning agents. How are these two kinds of agency related?

Begin with planning agency. We frequently settle in advance on prior, partial plans for future action. We then proceed to fill in these plans in a timely manner, and to execute them when the time comes. Such planning plays an important organizing role in our lives, both individual and social. It helps us organize our own activities over time, and it helps us organize our own activities with the activities of others.

We typically see present actions as elements in planned activities that extend over time. Frequently, it is only when seen in this light that our present activities make the right kind of sense to us. You will not understand why I am now typing this sentence unless you see this action as embedded in a larger planned activity, one that includes elements both past and (I hope!) future.

We also see many of our actions as shaped by intentions and plans that are in an important sense shared. When you and I are talking with each other our activity is normally seen by each of us as embedded in a shared activity of having a conversation, or trying to solve a problem, or trying to arrange a lunch meeting, or. … There is a sense in which we intend this shared activity – a sense in which we have a shared intention that structures our planning and action.

SHARED COOPERATIVE ACTIVITY: THREE FEATURES

We have a recognizable and important concept of a shared cooperative activity. This concept picks out a distinctive kind of interpersonal interaction, one that many of us see as important in our lives. You and I might sing a duet together, paint a house together, take a trip together, build something together, or run a give-and-go together in a basketball game. In many such cases ours will be a shared cooperative activity. Such shared cooperative activities can involve large numbers of participating agents and can take place within a complex institutional framework – consider the activities of a symphony orchestra following its conductor. But to keep things simple I will focus here on shared cooperative activities that involve only a pair of participating agents and are not the activities of complex institutions with structures of authority.

Shared cooperative activity (SCA) involves, of course, appropriate behaviors. If you and I successfully engage in the SCA of painting the house together then, of course, we paint the house together. But we might paint the house together without acting cooperatively. Perhaps neither of us even knows of the other's activities, or though we each know of the other's activities neither of us cares.

Given appropriate behaviors, what else is needed for ours to be a SCA? Suppose that you and I sing a duet together, and that this is a SCA. I will be trying to be responsive to your intentions and actions, knowing that you will be trying to be responsive to my intentions and actions.

In practical reasoning and action we seek to realize our intentions and satisfy our desires in the light of what we believe. Or so we are taught to say. In this essay I question the last clause and suggest that the cognitive attitudes guiding practical reasoning and action go beyond our beliefs. I begin by locating my problem within the planning conception of practical reasoning I have developed elsewhere (Bratman 1987).

PRIOR PLANS AND THE BACKGROUND OF DELIBERATION

As I see it, prior intentions and partial plans play central roles as inputs into the practical reasoning and action of intelligent agents like us: agents with serious resource limitations and with basic needs for coordination, both social and intrapersonal. In particular, prior plans structure and guide further practical reasoning in two major ways. First, one's plans are subject to demands for coherence, given one's beliefs. Prior plans are typically partial, but as time goes by they must be filled in in appropriate ways in order to avoid incoherence. Second, one's plans need to be consistent both internally and with one's beliefs. These consistency constraints create a filter on options to be considered in further deliberation as potential solutions to problems posed by threats of incoherence in one's partial plans.

This gives us a three-stage model of practical reasoning. First, one's prior partial plans generate problems, given threats of incoherence. Second, one tries to specify options that would be at least partial solutions to the problems posed and would satisfy the cited consistency constraints.

In a variety of papers and books – most especially in his recent work Thinking and Doing ([6]) – Hector-Neri Castañeda has developed and refined one of the most subtle and thorough philosophical theories of the relation of thought to action to be published in recent years. I begin this paper by laying out some of the main features of this theory. I then raise some questions about this theory, which lead up to a challenge to Castañeda's basic conception of practical thinking.

A central case in which thinking issues in intentional conduct is the case in which one's practical reasoning concludes with an appropriate intention and, as a result, one acts accordingly. Castañeda focuses on this central case. He wants a theory of practical reasoning which explains how various mental elements enter into such reasoning and how they are related to its conclusions.

Castañeda supposes that among the basic elements in such reasoning are one's beliefs and intentions, and that both believing and intending can be understood as the endorsement (acceptance) of a certain thought-content (noema). For example, to believe it will rain is fully to endorse the proposition that it will rain. Analogously, to intend to run is also fully to endorse an appropriate noema. It is not, however, to endorse the proposition that one will run, but rather what Castañeda calls the intention canonically expressed by ‘I to run’. The difference between intending to run and believing one will lies not in a difference in attitude towards the same proposition (that one will run) but in the different contents of the same generic attitude of endorsement.

In an important and fascinating series of papers, spanning a decade and a half, Donald Davidson has sketched a general theory of intention, a theory that tries to explain what it is to do something intentionally, what it is to intend to do something later, and how these two phenomena are related. In this paper I say what this theory is, argue that it faces a pair of serious difficulties, and diagnose these difficulties as rooted in an overly limited conception of the role of intentions and plans in practical reasoning.

DAVIDSON'S THEORY: MAIN IDEAS

Davidson begins with intentional action. In his classic paper ‘Actions, Reasons and Causes’, he sketches the following general view. Intentional action is action that is explainable, in the appropriate way, by appeal to the agent's reasons for action. The reasons that explain intentional actions are appropriate pairs of the agent's desires (and other ‘pro-attitudes’) and beliefs. When one acts for a, certain reason an appropriate desire–belief pair causes one's action. Suppose, for example, that I intentionally go to Davies Hall because I want to hear Pavarotti sing and believe going to Davies Hall would be a way of doing this. Since I act for this reason, the pro-attitude and belief cause my action.

This conception of intentional action makes no essential appeal to any distinct state or event of intending to go to Davies Hall that intervenes between my desire–belief pair and my action.