Lessons from a Reductionist Account of Individual Agency

Carol Rovane

In this chapter, I will be concerned to address the question of my title by taking on three some­what more specific questions:

(1) In what sense is it correct to say that individual agents set the boundaries of their own responsibility?

(2) How does the issue of responsibility sort out when a group of human beings qualifies as an individual agent in its own right?

(3) To what extent, and why, are individual agents responsible for the outcomes of collective endeavours to which they contribute, but which do not lie within their power to bring about on their own?

Before proceeding to answer these questions, I want to register a preliminary point about respon­sibility For various reasons, we are predisposed to focus on backward looking responsibility — on holding someone to account for something already done.

With these preliminary points about responsibility in hand, I can now outline the answers that I will be developing in this chapter to my three questions:

(1) The sense in which it is correct to say that individual agents set the boundaries of their own responsibility is supplied by a reductionist account of individual agency, according to which agents are self-constituting in a quite literal sense. When an individual agent constitutes itself, it sets boundaries within which it thinks and acts as one. These boundaries are at once the boundaries of the rational point of view from which it thinks and acts, and the boundaries that mark it off from other agents. In setting these boundaries, the agent not only sets the boundaries of its own thought and action, but also, the boundaries of its own responsibility. According to the reductionist account, these boundaries of individual agency need not coincide with the boundaries that mark one human life off from another. Thus, while the condition of individual agency can be realized within the boundaries of a single human life, it can also be realized within a group of human lives, and within different parts of a single human life. In other words, there can be group agents and multiple agents as well as agents of human size.

(2) When the condition of individual agency is realized within a group of human lives, this is not a case of collective agency in which many distinct agents of human size exercise their agency together, by thinking and acting from their distinct points of view. It is a case of genu­inely individual agency. The group agent constitutes itself, by thinking and acting as one, from its own rational point of view.

(3) In contrast, collective agency is a social phenomenon in which many distinct individual agents bring about shared ends by acting together. Individual agents do this by thinking and acting from their own separate rational points of view, even as they coordinate their thoughts and actions in order to realize a collective outcome together. This means that the collective does not qualify as an individual agent in its own right, for it does not think and act as one, and indeed it does not think and act at all. Therefore, the collective itself cannot take responsibility for the outcomes of the collective endeavour; only the individual agents who comprise the collective can do that.

Before going on, a side note about terminology. For many philosophical purposes, and indeed, in many real life settings (personal, moral, legal, political), it makes sense to equate the concept of a person with the concept of an individual agent. But there are also many purposes and settings in which the custom is to equate the concept of a person with the concept of a human being. It should already be apparent from what I have said so far that if we accept the former equation then we cannot accept the latter equation — because we will have to accept that there can be group persons and multiple persons as well as persons of human size. Elsewhere I have argued that there is good reason to accept all of this, and to regard the reductionist account of agency as an account of the condition of personal identity (Rovane 1998).

The reductionist account of personal identity bears a significant resemblance to Parfits psychological reductionist account of personal identity, which claims that the existence of a person consists in nothing but the existence of certain sorts of events standing in certain sorts of relations — in other words, it doesn’t consist in any further fact such as a Cartesian ego (Parfit 1984). Yet it could fairly be objected that Parfit’s reduction left out of account the fact that persons are agents.² For on his description, the events that make up a person’s life just happen in more or less the same sense that the events that make up the rest of the course of nature — events from which agency is entirely missing — just happen. I propose to rectify this shortcoming by characterizing the reductive base of the reductionist account of individual agency in fully agen­tial terms.³ The events that I claim make up the life of an agent are not mere happenings, but intentional activities of a certain kind, in the form of both thinkings and doings; and the relations in which such thinkings and doings must stand in order to constitute the existence and identity of an individual agent are normative relations; through these relations a certain kind of unity is achieved, which defines individual rationality. All of these terms of the reductionist account of agency — intentional activities, normative relations, rational unity — are irreducibly normative in the sense that they cannot be reduced to the terms in which the natural sciences describe nature.⁴ Yet the account is nevertheless reductionist in the following sense: it reduces the identity of any particular agent to the various intentional activities — the thinkings and doings — that constitute its life, and does not equate it with any further fact such as a Cartesian ego, or any other meta­physical condition that is allegedly distinct from the various intentional activities in which the life of an agent consists.⁵

Now that I have clarified the sense in which the reductionist account of individual agency is reductionist, I want to clarify next why it entails the possibility that the conditions of individual agency can be realized at the level of a whole group of human beings.

Consider a philosophy department that wants to review and revise its requirements for the PhD in philosophy. We would normally describe such a department as constituted by its human ‘members’, and so, to start with, I shall use this terminology. But we will soon see that the ter­minology can be misleading, precisely because when we refer to the human ‘members’ of a group agent such as a philosophy department, we normally presume that they must necessarily function as individual agents in their own right, who then constitute the group agent through social activities — in other words, the presumption is that group agency is a special case of col­lective agency. Once I have shown how and why this presumption is mistaken, I will no longer use the terminology of ‘members’, and refer instead to the human constituents of group agents, or more simply, to human beings. It will be important to bear in mind that whenever I employ this other terminology, of‘human constituents’ and ‘human beings’, I myself do not presume that human beings always necessarily qualify as individual agents in their own right. But let me start by referring to the human ‘members’ of a department, while leaving the presumption to the side for the moment, without further comment.

Suppose that the ‘members’ of the department come to believe that they ought not to argue and vote on the requirements for the PhD, because they recognize that, then, if students were to ask about why they must satisfy certain degree requirements, the only available response would be, Well that's the way the vote cut. Thus, the ‘members’ of the department recognize that what a responsible department should want to do instead is to work out what set of degree requirements would be best, all things considered. But in order to do that, the department must identify the things- to- be- considered.

When this kind of departmental deliberation takes place, each of the human ‘members’ is a site of intentional activities — thinkings and doings — that are carried out from the depart­mental point of view. I want to emphasize that this point of view is necessarily distinct from the points of view of its human ‘members’ — whom we ordinarily think of as individual human beings, who are also faculty members. In order to bring this feature of the case, concerning the distinctness of the department’s point of view from the points of view of its ‘members’, into better relief, it may be helpful if I admit that it is a real case and not a manufactured example. I am a faculty member of a philosophy department in which I persuaded my colleagues that we should not argue about and then vote on the requirements for the PhD, but rather, the depart­ment should deliberate about them as a single body, working out the all-things-considered significance of everything that all of the faculty members might ‘bring to the table’ for con­sideration. As it happens, my own personal view was that the only requirement that the depart­ment should impose for the PhD in philosophy is a completed and successfully defended dissertation, and that everything else should be viewed as mere preparation for this one requirement, and furthermore, whatever additional preparation any given student should be required to make should be dictated by the nature of their particular philosophical interests and projects, and previous training. So I advocated individual programmes of study, rather than uniform requirements such as distribution requirements, logic requirements, foreign language requirements, comprehensive and qualifying examinations, third year essays, etc. But, predictably, all of those other requirements, and various reasons for and against them, were also put on the table for consideration. This would ordinarily be an occasion for arguments among the ‘members’ of the department, in which each would try to persuade the others of their own personal view. But instead of launching into such arguments, what my depart­ment actually did in this case was consider all of the possible requirements that had been put on the table for consideration, and the various reasons for and against them, and then work towards a deliberated conclusion about which combination of them would best serve our PhD students and why. The conclusion that the department actually arrived at imposed many more requirements than a completed and successfully defended dissertation, and so it clearly was not a conclusion that followed from my personal point of view — that is, the department’s con­clusion did not follow from what I believed about what would be best for our students. The conclusion followed from a set of considerations that included beliefs that I didn’t happen to share, but which were definitely on the table for consideration by the department. And the same held for each and every ‘member’ of the department: the conclusion that the depart­ment arrived at did not strictly follow from what they individually believed from their own personal point of view, but rather from a larger set of considerations that had been put on the table for consideration by the department, which constituted the department’s point of view.

I’m now ready to clarify why the reductionist account of individual agency takes the case of group agency as a model for all cases of individual agency. Insofar as the department deliberated in the way I just described — aiming to work out what would be best for it to do, by taking into account all of the thoughts that figure in its point of view — it aimed to realize a rational ideal that defines what it is for an individual agent to be fully, or ideally, rational. The ideal is to achieve overall rational unity within oneself by arriving at and acting upon all-things-considered judgements about what it would be best to do, that take into account all and only one’s own thoughts about what is true and good. It is because the department aimed to meet this ideal of individual rationality that it makes sense to say that it thought and acted as one, and thereby qualified as an individual agent in its own right, who can be engaged and held responsible by others. The reductionist account of individual agency moves from this insight about group agency to a general claim about all cases of individual agency: whenever and wherever there is a commitment to realizing the ideal of overall rational unity, by arriving at and acting upon all-things-considered judgements, there is an individual agent that thinks and acts as one, in the sense that it deliberates and acts from its own point of view, in accord with the requirements of individual rationality. And here is why this general claim qualifies as reductionist: The existence and life of the philosophy department, qua group agent, consists in nothing but the intentional activities through which it thinks and acts as one, by striving for the kind of overall rational unity among them that is characteristic of individual rationality; and so it is with all individual agents, that their existence and life consists in nothing but such intentional activities, through which the kind of unity that is characteristic of individual rationality is achieved.⁶

The reductionist account allows that the basic agential capacities through which rational unity is achieved belong by nature to human beings. But there is no necessity that such unity must be achieved within the whole of a human life, for it can also be achieved within different parts of a human life, thereby giving rise to multiple agents within it. This is perhaps the most controversial implication of the reductionist account.⁷ And yet, we often come close to rec­ognizing it in ordinary life. For example, we are all aware that the considerations from which we deliberate for the purposes of our work life may be quite separate from the considerations from which we deliberate for the purposes of our home life. Similarly, it is almost a cliche that someone whom we regard as our close friend may suddenly seem like ‘another person’ when we encounter them in their role of loan officer at the bank, because of the way in which, in that context, everything they say and do reflects bank policy rather than their attachment to us — and so we say that the ‘bureaucrat’ we encounter at the bank is not recognizably the same person as our friend at all. It may seem an unnecessary exaggeration to construe these cases as ones in which a human being is actually host to multiple agents, as opposed to multiple projects. And it is important that the reductionist account can distinguish cases in which there are many distinct agents within a single human life from cases where there is one agent within that life, with many distinct projects. It does so in terms of whether there is an overarching commitment to overall rational unity within that life. It should not be surprising that such an overarching commitment to unity within the whole of a human life tends to be present in most of the cases of individual agency that we know. After all, we agents of human size tend to teach one another, and urge one another, to approach life in a unified way, by asking What are you going to do with your life? and by declaring the worth of so-called life projects that generally do take up the whole of a human life, such as marriages and careers. But we should not overlook the ways in which our social conditions may sometimes actually promote multiplicity over unity within a single human life. Consider, for example, the social conditions of many immigrant children in New York City who come from highly traditional immigrant homes, and who must also navigate the social space of an elite public high school. In order to navigate each of these spaces effectively, it may well be that a ‘switch’ is required as they go back and forth between home and school, between the mind-set and values that would enable harmonious and dutiful life in a traditional home, and the mind-set and values that would afford a suitably ambitious and autonomous response to the educational opportunities afforded by the culture of liberal individualism.⁸

I am not arguing that the best human life would be one that was marked by the kind of div­ision, or rational fragmentation, that would go together with the existence of multiple agents within it. Perhaps the best use of a human life would be to achieve the kind of overall unity within it that would go together with the existence of a roughly human-sized agent. But if that were so, then it would still be important to acknowledge the difference that it makes when we regard such overall unity within a human life as an option that is worthy of our pursuit, rather than as our metaphysically given condition. Correlatively, it is important to acknowledge that such overall rational unity within a whole human life is not a rational requirement any more than it is a metaphysically given condition. All there is here, is the following constitutive truth: any rational agent will be committed to achieving overall rationality within itself, because that is what it is to be a rational agent. It might be tempting to construe this commitment as a rational response to a categorical imperative to be rational. I myself am not so tempted — I think rational beings might find reasons to cease to be rational, even if acting on those reasons amounted to a kind of suicide. But in any case, even if we were tempted to think that there is a categorical imperative to be rational, it could be obeyed by any rational agent, no matter what its size — it could be obeyed equally by multiple agents, by agents of human size, and by group agents too. And there is an important corollary point, which is that according to the reductionist account of individual agency such an imperative, that simply said Be rational! would inevitably be incom­plete: it cannot be followed until the boundaries are set, within which the demand for rational unity is to be realized!

And so the question arises: How do such boundaries get set? The answer supplied by the reductionist account is: Through a commitment to a unifying project that cannot be pursued without arriving at and acting upon all-things-considered judgements within particular boundaries. In my chapter so far, I’ve referred to various such unifying projects: a philosophy department’s project of setting requirements for the PhD degree in philosophy, which mandates thinking and acting as one at the level of the whole department; life projects, such as marriages and careers, that mandate thinking and acting as one within a whole human life; smaller projects that mandate thinking and acting as one within one or another of many different social spheres that a single human being might move among.

I said at the outset that, according to the reductionist account, individual agents are self­constituting. Let me now add that part of what I mean is that they are not constituted by anything else. In particular, an individual agent is not constituted by any metaphysically given facts, such as the existence of a given human body and consciousness — for it comes to be only with and through the intentional activities, the thinkings and doings that constitute its agential life. An agent is not constituted by other agents either — for to repeat, it comes to be only through its own intentional activities. If an individual agent comes to be only through its own intentional activities, and if its existence consists in nothing else besides these intentional activities, then it follows that it is literally self- constituting.

I also said at the outset that when an individual agent constitutes itself it sets boundaries within which it thinks and acts as one. And then I added: these boundaries are the boundaries of the point of view from which the agent thinks and acts; and they are the boundaries that mark the agent off from other agents; and they are the boundaries of the agent’s own responsibility. Let me elaborate what I meant by these further claims by returning to the example about the philosophy department.

In my descriptions so far, I have taken care to emphasize that the department’s point of view is distinct from my point of view, because the department deliberates from considerations that I don’t embrace from my point of view. But what I have not yet emphasized is that I emerge to be an individual agent of somewhat smaller than human size, in a way that is very much like what I described above in connection with multiple agents. That is, the human life that is the site of my existence is also a site of rational fragmentation in the following sense: some of the inten­tional activities that go on within that human life proceed from the point of view of the department, while other intentional activities that go on within the same human life proceed from my point of view. This is what I had in mind when I said at the outset that the boundaries of an agent’s rational point of view are also boundaries that mark its life off from the lives of other agents. Although it may seem a strange way to put it, the department’s unifying project, of setting the requirements for the PhD in philosophy, set boundaries within which the depart­ment deliberated and acted as one, supplied the reasons for setting those boundaries in such a way that thoughts and actions spanning many different human lives figured in, and constituted, the life of the department rather than its members.

So let me turn now, finally, to the issue of responsibility. As I’ve just explained, when the department thinks and acts as one, the thinkings and doings through which it constitutes itself belong to it and to no one else — including even me, and even in the case of its thinkings and doings which are located in the same human organism that my thinkings and doings are located in. It follows that I stand in an interpersonal relation to my department. It also follows that I do not bear any personal responsibility for what the department does — such responsibility lies squarely with the department. This makes for a profound difference between a group agent’s responsi­bility and what we ordinarily think of as collective responsibility.

Collective agency is the agency of many, not one. It is a social phenomenon, in which many distinct individual agents exercise their agency together in order to do something that does not lie within the agents’ power to do all on their own. There is an important ambiguity here, about what does and does not lie within the power of the individual agents who together comprise a collective. Although it is true by definition that a collective action, which is intended and undertaken by many individual agents, does not lie within the power of any one of them to do on their own, it is sometimes the case that one or another of them has the power to bring about the same outcome on their own. Consider a gruesome case in which many individual agents of human size intentionally exercise their agency together in order to kill a human being, by each inflicting just one of a thousand cuts. It may be that any one of those agents has the power to bring about that human being’s death on their own. But all the same, it does not lie within their individual power to bring about the death in the way that it is actually intended in the collective case, which is that it be brought about by each member of the collective inflicting only one of a thousand cuts. When I refer to collective ends from here on, I will be referring to them in this way that takes due account of the description under which they are intended, as collectively achieved ends. And I repeat that by definition, such collective ends do not lie within the power of any one individual agent to bring about on their own. The puzzle about collective respon­sibility, then, is who is in a position to take responsibility for such a collective end. It is in the nature of the case that the collective itself has no point of view of its own from which it thinks and acts as one. It is by hypothesis and definition, the agency of many agents who each act from their own individual points of view. One natural first thought is that these individual agents can only take responsibility for what lies within their power to do individually, and so the extent of their responsibility is confined to just their own contribution to the collective end. Thus in my gruesome example, each individual would be responsible for just one cut. A natural second thought, though, is that this is a completely unacceptable conclusion, because each individual agent intends their cut as a contribution to a death by a thousand cuts, and so each is really trying to take responsibility for the intended collective end as well as their own individual contribution to it.⁹ Whichever one of these thoughts we find more plausible, I take it we need to embrace one or other of them when we seek to locate responsibility for collectively achieved ends — for there simply is no other agent available to hold responsible for them, if what we mean by an agent is someone who can be held responsible for what they do in the engaged way that presupposes that the agent can keep up their end of the engagement by taking responsibility for what they do.¹⁰

I have argued that the situation is entirely different when we find that the conditions of individual agency are realized at the level of a group of human beings. Then there is someone besides the human ‘members’ of the group to hold responsible, namely, the group agent itself. I have also argued that when we embrace the reductionist account of individual agency, not only is it the case that groups of human beings can be sites of individual agency, but also, the human ‘members’ of such groups may not qualify as individual agents of human size, because those human lives will be marked by rational fragmentation, owing to the ways in which some of the intentional activities within those lives both constitute, and proceed from, the group agent’s point of view and not from any other points of view — they are the thinkings and doings of the group agent and no one else. And for this reason, insofar as the human ‘members’ appear to be sites of agents of human size, the truth is that they are agents who are somewhat smaller than human size, whose lives are constituted by the other intentional activities that proceed from their own smaller points of view, which exclude those of the group agent.

Most philosophers who are prepared to allow that a group of human beings may qualify as an individual agent in its own right would not agree with these latter claims, because they view the human ‘members’ of such group agents as natural agents, who must necessarily remain the individual agents that they are, even when they combine their efforts so as to forge a group point of view from which thought and action proceed, and which can be engaged and held responsible.¹¹ To take this view is to reject the reductionist account of agency — it is to say that individual agents are not self-constituting, via the thinkings and doings through which they live their lives as the individual agents they are, but rather, they are human beings, and as such their lives are metaphysically given with the biological facts of human life. When we suppose that the human ‘members’ of a group agent are natural agents, we will conceive the group agent not as natural but as artificial — as brought into being by the individual efforts of its human ‘members’; and then group agency will emerge as a special case of collective agency, in which the human ‘members’ of the collective must necessarily always be thinking and acting from their individual, human-sized, points of view — which are also metaphysically given with their biological exist­ence. As I’ve made clear, one primary reason why I am convinced that the reductionist account of agency is correct is that I am convinced that there can be thinkings and doings on the part of a group agent that proceed from a point of view that is distinct from any such human-sized points of view.

I realize that matters may not appear this way to the ‘members’ of a group agent — even in a case like my own philosophy department, in which the ‘members’ had become convinced that it would be better if the department arrived at a reasoned conclusion about what PhD requirements to set, by thinking and acting as one, rather than having a social process of debate and argument followed by a vote. The problem was not that the ‘members’ of the department did not appre­ciate that there is a real difference between a process of deliberation that would proceed from a single set of pooled considerations that comprise the department’s point of view, and a social process of debate and argument that would be carried out from the many different points of view of the department’s ‘members’. The problem was that they found it hard to reconceive both the boundaries of their own agency and responsibility, and their relation to the depart­ment, as these things would emerge to be if the former condition were realized.

One reason why this reconception is so difficult derives from the fact that the method by which a group agent like a philosophy department achieves the sort of rational unity that is characteristic of individual agency involves speech and communication, which we take to be social phenomena, and therefore very unlike individual deliberation. Thus, when the ‘members’ of my department filed into their meeting, the way in which the consider­ations that constituted the department’s point of view were pooled together involved each of the ‘members’ making ‘utterances’ that probably felt to them as though they were being made from their own points of view. And if they felt that way, then perhaps that is what they were — utterances that expressed the individual viewpoints of the ‘members’ of the depart­ment. But they were also acts of communication, through which the department came to be aware of the many considerations that it ought to take into account as it deliberated. And insofar as those deliberations involved certain utterances that issued from the mouths of different ‘members’ of the department, they counted as steps in the department’s deliberations because they articulated steps in reasoning that was aimed at working out the joint norma­tive significance of all of the pooled considerations that constituted the point of view from which the department was deliberating. Insofar as such utterances did articulate steps in such reasoning, they did not express the separate views of the separate ‘members’ of the department, but rather the view of the department itself. This is compatible with such an utterance giving voice to a point of agreement between a given faculty member and the department. But all the same, insofar as a step in reasoning really was taken, which aimed to work out what that utterance entailed, when considered together with the other considerations that constituted the department’s point of view, it was a step in the department’s deliberations, as opposed to the deliberations of one or another of its ‘members’.

Thus, when ‘I’ initially put on the table, for the department’s consideration, the proposal that there be no requirements for the PhD besides a completed and successfully defended disser­tation, it is not correct to say that ‘I’ was making a causal contribution to a collective outcome that I was jointly pursuing with other ‘members’ of the department. It was not like laying one brick alongside other bricks in the joint activity of building a house together. Once I had put my consideration on the table, I had ‘conveyed’ it to an active process that took on a life of its own, of working its way to conclusions from a deliberative base, which is to say, a point of view, that was not mine. Of course, if I were to build a house together with others by laying down some bricks, neither the bricks nor the house would strictly speaking be mine either, and nor would the collective process through which the house that got built. But building a house with others is a collective effort in which my intentional efforts combine with those of others; and what makes it a collective process is not the fact that many distinct human beings are involved, but the fact that the process is carried out from many different rational points of view. And this is not going on in the case where the department comes to function as an individual agent. When a department does this, it comes to function as an individual agent in its own right, and then whatever its ‘members’ do in relation to it is not a contribution to a shared collective process, but an interpersonal effort to influence another individual agent, which is the department itself. Surely, some of the utterances that come out of the mouths of the department’s ‘members’ during departmental deliberations will have this social character or function, of trying to influ­ence another agent, which is the department. But for the most part, this will not be so in the sort of case I am describing, in which the department is deliberating and acting as one from a departmental point of view. For the most part, such utterances either articulate, or are, steps in deliberations that are being carried out by the department. This is why they are not properly thought of as contributions made by department’s ‘members’ to a ‘collective’ process — for such contributions would have to be made from separate points of view, whereas these steps proceed from the department’s point of view.

What follows about responsibility for a group agent’s actions? I take it to follow that the human constituents of a group agent are no more responsible for what a group agent thinks and does, and also, no less responsible for what a group agent thinks and does, than any individual agent is ever responsible for what another individual agent thinks and does. We might suppose that one agent may and should take responsibility for another agent’s attitudes — by, say, sharing information with them about what is really true and good. We may suppose that one agent may and should take responsibility for another agent’s actions — by, say, using any means available to get the other agent to do good things, and to prevent the agent from doing bad things. But if we believe in individual autonomy, then we may be consigned to leave others to think and act as they judge best, once we have done due diligence by informing them of what we think. And if this belief in individual autonomy is coherent, then there is room for carrying it over to our relations to group agents, and this is true even in those cases where the human lives that are the sites of our own existence are also sites of intentional activities that constitute the life of a group agent. Thus, I may leave my department to think and do what it judges to be best, even though if it were up to me, the department would not be imposing all of the requirements that it cur­rently does impose for the PhD in philosophy; and I do not bear any kind of personal respon­sibility for what the department does, of the sort that I would bear in cases of collective agency, where there was no group agent acting from its own point of view.

I have found that when I share these claims with others, many of them may follow me a good part of the way to my own conclusions, but then they stop short of following me all the way. Here are some of the points along the way that they might be willing to accept: A human being does not begin life already able to exercise its agential capacities. Rather, a human being begins life as a creature that is buffeted about by its perceptions and desires, and it isn’t able to take responsibility by acting in a truly considered way until fairly well on in its development. At some point, there is a stepping back from desires, and an evaluation of them in the light of what is really good and what is really better than what; and at some point, there is a recogni­tion of more complicated and interesting projects that would require significant coordination of thought and effort over time; it is when such unifying projects are embraced that there is a real need to deliberate before acting — to consider what would be best in the light of many con­siderations taken together.¹² Here is what the reductionist account makes of these undeniable facts of human life: It is only when such unifying projects are embraced, which actually require coordinated thought and effort, that the human being becomes the site of agency in the full sense, which involves forging a point of view from which deliberation and action proceed, by arriving at and acting upon all-things-considered judgements. It is only once such a deliberative point of view has been forged that there is an agent whom others can hold fully responsible, as an agent who is taking charge of its own priorities, choices, and actions. Of course, long before a human being develops sufficiently to have such a deliberative point of view, it has a point of view in other senses. It has a bodily point of view from which it perceives and moves, and it also has a phenomenological point of view from which there is something it is like for it to experience and feel. Points of view in these latter two senses are both biologically given with the existence of a normally functioning human organism — and for this reason, I count them as metaphysically given, as opposed to products of effort and will. What is also biologically given are basic agential capacities.¹³ But according to the reductionist account of agency, what is metaphysically given in this way with human life does not suffice to make the individual human being an individual agent. What is both necessary and sufficient for individual agency is that certain thinkings and doings actually occur, in a way that is directed at achieving the kind of rational unity that is characteristic of individual agency, by arriving at and acting upon all-things-considered judgements. According to the reductionist account, what an individual agent ought to take into account in such judgements are all of its own thoughts, both about what it can do, and also about what it would be best to do in the light of its thoughts about what is good and true. Thus it equates the scope of an agent’s all in all-things-considered judgements with the boundaries of that agent’s deliberative point of view — whatever falls within the one falls within the other. And these boundaries that circumscribe the agent’s own deliberative point of view are not set by the boundaries of its body or its consciousness. Possessing the kind of bodily point of view from which an animal perceives and moves, and possessing the kind of phenomenological point of view from which a subject is directly aware of its own subjective states in consciousness, is nei­ther necessary nor sufficient for possessing a deliberative point of view from which to deliberate and act — what I have elsewhere called a rational point of view. We generally assume that these things all coincide in human life — that is, we assume that deliberation is a conscious process, which is directed at options that are metaphysically given to us by virtue of possessing a body over which we have direct intentional control. But according to the reductionist account of individual agency this assumption is mistaken twice over: firstly, by overlooking how group agents do not require a single, unified consciousness in order to reason as one or a single human body through which to act as one; and secondly, by overlooking how a single unified consciousness housed within a single human body does not suffice for such reasoning and acting as one, owing to the possi­bility that there can be multiple agents within a single human life.

As I’ve said, many philosophers will follow me a good part of the way towards the conclusions of the reductionist account, but then they persist anyway in viewing human beings as natural agents. Their view seems to be that even if it requires effort and will to become an individual agent each human life must necessarily become the site ofjust one such agent. Sometimes it is claimed in support of this view that a human agent can do only one thing. This claim is patently false, for right now I am typing these words on my laptop while stroking my dog with my foot. But what is more important here is this: even if it were true that actions involving the human body can occur only one at a time, this would not suffice to undermine my claim that a human being can be the site of non-pathological rational fragmentation. This claim does not turn on whether only one thing can be done at a time through any one human body (which as I’ve said I deny anyway); it turns on whether a human being must be the site of a single delibera­tive point of view at any given time, or whether it can be the site of two (or more) co-existing deliberative points of view. I am claiming that the test of whether this is so is not how many things are being done at once, but whether the point of view from which such things are done must be a point of view of human size, or one of several points of view within the same human being. Take speech as such a test. Let me assume for the sake of argument that words cannot come out of a given human being’s mouth in such a way that they simultaneously express two such co-existing points of view. This assumption does not rule out the possibility that some of the words that come of that human being’s mouth can express a smaller-than-human-size point of view while others express the departmental point of view, and that both points of view exist simultaneously. And so it is with all actions that might be performed with or through that human being’s body.

I think the better way to object to the reductionist account of individual agency is not to try to argue for the false claim that there is a practical necessity of overall rational unity within a whole human life, as there would be if human beings were natural agents. The better way would be to try to argue from the true claim that there is at least a practical possibility of such overall rational unity within a whole human life, and to try to make the case that this possibility has important implications concerning who is ultimately responsible for the actions of a group agent. For such an argument would be proceeding from a premise that the reductionist account acknowledges to be true, which is that it is not only possible, but often close to actual, that a human life is the site of overall rational unity within it. The argument might begin by suggesting that, in any case I would describe as a case of rational fragmentation, in which a smaller-than-human-size agent co-exists within the same human life as intentional activities that figure in the life of a group agent, it is always possible for that smaller-than-human-size agent to exert intentional control over all of the intentional activities within that human life; and then the objection would pro­ceed by suggesting that the mere fact that this is possible renders that smaller-than-human-size agent personally responsible for all that goes on within the human life in question, including the intentional activities that belong to the group agent. To put it another way, the argument would be that the smaller-than-human-size agent would have ceded control over certain parts of that human life to the group agent, and that the smaller-than-human-size agent could always regain such control, and that this would be a morally important thing to do in cases where the smaller agent finds that the group agent reasons and acts in ways they do not approve of.

The problem with this line of objection is that it fails to reckon with the true nature of intentional control. There is no thing — the agent — that stands apart from the intentional activ­ities through which it exists, that is standing ready to take control, and therewith, responsibility, for everything that falls within its own metaphysically given sphere of intentional control. Such intentional control as exists, enters the world through intentional activities themselves. So if the kind of rational unity that is characteristic of individual agency ever comes to be, it is because there are intentional activities that include a recognition of ends (unifying projects) for the sake of which it is worth achieving such unity, along with efforts to achieve such ends via such unity. When such efforts are directed at achieving unity within a philosophy department, those efforts do not somehow intrude upon ‘my’ metaphysically given sphere of intentional control and personal responsibility. They are just what they are, efforts directed at ends, and it is in part through those efforts that a departmental point of view gets formed. They do not belong to me, the philosopher writing this chapter. What I do is constitute myself And that is what the phil­osophy department does too, insofar as it comes to function as an individual agent in its own right. Each of us is setting the boundaries within which we are taking responsibility for what happens, by exercising our agency in the unified way that is characteristic of individual agency. But this just means that the efforts through which we come to be are directed at achieving unity within those different boundaries. Each of us takes personal responsibility for our own thoughts and actions. Each of us might bear some limited interpersonal responsibility for what the other does, insofar as we have opportunities to influence each other for the better, or even to promote or hinder the other’s actions. But the situation is very unlike the situation of collective agency, in which many distinct agents act together, but from their own points of view, for the sake of a collective end, which is then brought about by them — and not by any further agent who is distinct from them.

I think it is correct to say that in the collective case, the individual agents who make up the collective each aim to extend the reach of their personal responsibility, by aiming to help bring about the collective end to which they are contributing. And one of the important things for such individual agents will be to assess whether and why they should contribute vs. refuse to contribute — this would be, as we like to say, a matter of individual conscience. Because what these individual agents are working out is what they should do, by exercising their own individual agency. And my argument in this chapter is that this is not the right model on which to under­stand cases in which a group agent emerges, who functions as an individual agent in its own right. Individual agents are never made by, or made up of, smaller agents, as we see pictures in the great frontispiece to Hobbes’s Leviathan. This is the truth about self-constitution, as it emerges via the reductionist account of individual agency.¹⁴

Notes

1 This assertion may not answer everyone’s idea of what it means to take responsibility for what happens. Some may construe the idea more narrowly, perhaps in connection with undertaking obligations of one kind or another. But what I am getting at here is that agents are not by-standers when it comes to their own actions, waiting to see what happens, because they aim to bring it about that something or other happens. In such cases, what happens is their doing — that is why we hold them responsible — and if it is their doing then they have already taken a kind of responsibility for what happens, because they are not merely standing by, but acting.

2 This is an objection that Christine Korsgaard raised early on in response to Parfit. See her ‘Personal Identity and Unity: A Kantian Response to Parfit’, Philosophy and Public Affairs. 18, no. 2: 101—132.

3 For a fuller elaboration of the reductionist account see Bounds of Agency, Princeton University Press (Princeton, NJ: 1998), and for further discussion of what it means to say that agents are literally self­constituting, see ‘Reductionism, Self-Constitution and The Moral Significance of Personal Identity’ in Reading Parfit, Andrea Sauchelli, ed. (Routledge, forthcoming).

4 I have deliberately left out of my description of the reductive base a term that has figured centrally in all of my other discussions of the reductionist account of agency, namely, ‘commitment’. I use that term mainly to refer to intentional attitudes, insofar as they are conceived entirely in terms of their contents — thus I would describe a belief that P as a commitment to the truth of P, and I would describe an evaluative attitude that registers the value of Q as a commitment to goodness of Q. Commitments so conceived are irreducibly normative in the following sense: they are not dispositions to reason and act in accord with them. What they register is the agent’s own recognition that it ought to reason and act in accord with them. But it is possible to recognize an ought without necessarily conforming to it — so long as one views failures to conform to it as grounds for self-criticism. So, according to the reductionist account of individual agency, the life of an agent consists in undertakings of commitments (intentional attitudes), and efforts to live up to those commitments by reasoning and acting in accord with them. This is what I am referring to in the text above, when I refer to the intentional activities — the thinkings and doings — in which the life of the agent consists.

5 Of course, it may be wondered, doesn’t there have to be a further fact — the agent — who is carrying out all of the intentional activities that I am claiming constitute the agent’s life. My answer is no — see the work cited in note 3 for further discussion of why.

6 Here I mean to be speaking, specifically, of the existence and life of the philosophy department qua group agent. Most philosophy departments do not qualify as individual agents in this sense. Yet even when they do not, they do exist nevertheless, as some kind of thing — a thing that has a certain life and function within the institutional setting of a university, and in the academic world at large. I thank an anonymous reviewer for inviting this clarification.

7 This is a crucial point of disagreement with Korsgaard, which I press in Rovane, op. cit.

8 For further elucidation of the possibility of multiple agents within a single human life, see my works cited in note 3. See also, ‘A Nonnaturalist Account of Personal Identity’, in Perspectives on Naturalism, ed. M. deCaro, Harvard University Press (Cambridge, MA: 2004).

9 This line is convincingly pushed by Kirk Ludwig. See his contribution to this volume, ‘From Individual to Collective Responsibility’.

10 As an aside, I will confess that I myself feel the pull of both of the natural thoughts about the extent of an individual agent’s responsibility for collective ends, on which it is either confined to individual contributions or extended to the intended collective end. On the one hand, I’m inclined to think that an individual agent who inflicts one of a thousand cuts while aiming to contribute to a death should bear full responsibility for that death. But, on the other hand, when individuals make small contributions to an extremely large fund that will pay for something like ending extreme poverty, I’m not so sure they should get much credit for anything beyond their own individual contribution — espe­cially in cases where they can afford to give much more than they do. This makes me wonder if there is any general truth about the extent of individual agents’ responsibility for the collective ends to which they contribute, or whether it might be a matter of context.

11 The most influential account of group agency along these lines is Christian List and Philip Pettit, Group Agency: The Possibility, Design and Status of Corporate Agents (Oxford University Press, 2011).

12 This transition from being an animal that is merely pushed around by its desires, to an agent that can step back from its desires and evaluate them is one that Christine Korsgaard makes much of in her account of self-constitution. But she makes many claims that I do not accept. She claims, for example, that the fundamental act is one of laying down a law for oneself, whereas I claim it is undertaking commitments of various kinds — beliefs, evaluative attitudes, unifying projects, all-things-considered judgements. She also claims that anything with agential capacities will necessarily regard itself as bound by categorical imperatives, first to be rational by laying down laws for oneself, and then by respecting a Kantian style universalizability constraint, whereas I do not think there are any categorical imperatives to be rational or moral. She claims that the human organism cannot be the site of multiple agents within it, whereas I claim it can be. See her Self-Constitution, Agency, Identity, and Integrity (Oxford University Press, 2009).

13 I do not think we should equate these basic agential capacities of human beings either with motor capacities or with powers to perform basic actions in Arthur Danto's sense. They go beyond mere motor control, because they include whatever capacities make it possible for a human being to develop full intentionality and rationality; and they fall short of powers to perform basic actions, because basic actions are fully intentional actions that require self-conscious knowledge of one's power to perform them, and it is clear that human beings are not born with such self-known powers. (For a classic statement of Dantos conception, see his ‘Basic Actions', American Philosophical Quarterly, 2, no. 2, April 1965.)

14 I would like to thank many colleagues, students, and friends for their help as I thought through and wrote up this chapter, and to especially mention: Akeel Bilgrami, Michael Della Rocca, Francey Russell, the editors of this volume, and an anonymous reviewer.

References

Danto, A. (1965) “Basic Actions”, American Philosophical Quarterly 2, no. 2: 141—148.

Korsgaard, C. (1989) “Personal Identity and Unity: A Kantian Response to Parfit”, Philosophy and Public Affairs 18, no. 2: 101—132.

Korsgaard, C. (1996) The Sources of Normativity, Cambridge University Press.

Korsgaard, C. (2009) Self-Constitution: Agency, Identity, and Integrity, Oxford University Press.

List, C. and Pettit, P. (2011) Group Agency: The Possibility, Design and Status of Corporate Agents, Oxford University Press.

Parfit, D. (1984) Reasons and Persons, Oxford University Press.

Rovane, C. (1998) Bounds of Agency: An Essay in Revisionary Metaphysics, Princeton University Press.

Rovane, C. (2003) “A Nonnaturalist Account of Personal Identity”, in Perspectives on Naturalism, ed. M. de Caro, Harvard University Press.

Rovane, C. (2017) “Is Group Agency a Social Phenomenon?” Synthese DOI 10.1007∕s11229-017-1384-1. Rovane, C. (forthcoming) “Reductionism, Self-Constitution and The Moral Significance of Personal Identity” in Derek Parfit's Reasons and Persons: An Introduction and Critical Inquiry, ed. Andrea Sauchelli, Routledge Press.