Introduction
Traditionally, corporate realism attempts to assign to the state and other associations the ontological status commonly ascribed to the individual. That is to say, associations are depicted as intelligent organisms with minds and intentional states of their own.Footnote 2 Few contemporary theorists believe that this analogy makes much sense. I do not disagree. In this realist form, the idea of a corporate person should be rejected. The view that I discuss in this paper, however, puts the analogy between association and individual on its head. I want to begin the case for corporate personality right at the opposite end of the spectrum. Granted, associations are not individuals ‘writ large’. However, it could be argued that individuals are associations ‘writ small’. The reversal of this analogy, as I shall present it here, treads a fairly thin line between two affiliated claims. The first claim is that individuals have the same ontological status as associations. It is based on the extrapolation of the standard form of ontological reductionism to a form that applies tout court, that is, to associative and individual persons alike. The second argument claims that, in consequence, individuals need to be awarded the same ethical status as associations. Together, both arguments show that ontological reductionism forms a double-edged sword for ethical individualism. On the one hand, it provides the tool that exposes the weakness of corporate realism. Collectives, of course, are not a free-floating consciousness. They should be broken down into individually conscious parts. But on the other hand, a more thoroughgoing ontological reductionism uncovers that individuals are in turn like miniature associations. In other words, there is the possibility that reductionism is pointed at its own conceptual heart, namely that of the individual person as the apex of our rational and moral universe. This possibility has been most notably advanced by Parfit in his Reasons and Persons.Footnote 3 In Part Three of that book, Parfit sets out to achieve two main ambitions. First, he seeks to develop a particular ontological conception of the nature of personal identity. Second, he aims to show how this particular conception of identity undermines all person-based conceptions of rationality and morality, also under attack in the other parts of the book, and which claim or imply that personal identity is something that fundamentally matters. To avoid any confusion, I should pause to point out in advance that I do not support Parfit’s views across the board. I rely on them only in part. To be more precise, his second conclusion, that we should liberate morality and rationality from the idea of personality as such, cannot be sustained. But the premise of this conclusion is correct. In other words, personal identity of individuals is reducible to a bundle or web of constitutive parts in a way that mirrors the plight of associations. I propose to elaborate this latter claim first. In Section 1, I sketch the Lockean and Humean background to Parfit’s reductionism tout court. Next, in Section 2 I discuss the conflict between this view and the Ego theory. Section III claims that the reductionist criterion of personal identity corresponds in important ways to that of associations. Ontologically, it is argued there, individuals and associations are on a level pegging. Finally, in Section III, I explore Parfit’s ideas – arguably his most daring – on the ethical implications of this ontology of persons.
1 The psychological principle of personal identity
In order to set out the backdrop to the argument, I want to start by drawing a broad distinction between two schools of thought in the philosophy of personal identity. The first school is inspired by the ideas of Locke and Hume, while its modem members include Grice, Parfit, Perry and Quinton. It advances what is often described as the psychological criterion of personal identity. According to this criterion, personal identity over time consists in the relations between a series of mental states at different moments. It is also referred to as the bundle theory. The second school is mainly inspired by the ideas of Descartes. It maintains that personal identity is grounded, not in a bundle of psychological states, but in the identity of the underlying substance of an Ego.
It is in accordance with Locke’s own viewpoint that we should start with an account of what the concept or idea of a person essentially means.Footnote 4 To be a person, according to Locke, is to be something that can only be described as a particular mental faculty or state of being. Briefly, it is to be a thinking and self-conscious system or: ‘a thinking intelligent Being, that has reason and reflection, and can consider it self as it self, the same thinking thing, in different times and places; which it does only by that consciousness, which is inseparable from thinking, and as it seems to me essential to it’.Footnote 5 Locke focused on mental states and faculties because in his view the notion of a person was essentially a moral one. According to Locke, the term ‘person’ ‘is a Forensick Term, appropriating Actions and their Merit; and so belongs only to intelligent Agents capable of a Law, and Happiness and Misery’.Footnote 6
Consequently, the crux of the Lockean concept of the person is that it severs the psychological aspects of personality from the substance or stuff that does the thinking. On the one hand, there are thoughts, beliefs, memories, or as Hume put it, ‘perceptions’, while, on the other hand, there is the being of which we might say that it ‘owns’ these mental states. There are, of course, distinct views as to what the nature of this being or stuff might be. Descartes, in his Meditations, proposed that there exists purely spiritual substance.Footnote 7 Other views hold that there is only material substance. Locke believed in spiritual substance. In any case, he did not seek to dismiss this idea. Parfit and other Lockeans seem more partial to the materialist position. But whatever our view of the nature of substance might be, the decisive point for all Lockeans is that the idea of a person is separable from the idea of thinking substance. Ultimately, the person is constituted purely by conscious thought, by reflection and self-awareness, and emphatically not by whatever stuff does the thinking.Footnote 8
Locke, then, first distinguished between the ‘Idea of a Thinking Substance’, the ‘Idea of a Person’ and the ‘Idea of a Man’. By ‘person’ is meant a self-consciousness, intelligent, reasonable and accountable being. By ‘substance’ we mean the stuff that ultimately renders us capable of consciousness and thought. And by ‘man’ we mean a living organism. Consistent with this tripartite division of ideas, Locke devised three distinct criteria of identity, each ‘suited to the idea’. These criteria are not meant to settle what sort of things substance, man or person are, i.e. what separates them from each other. Rather, they decide what separates one substance from another, one man from another man and one person from the next. What is at stake here is the principle of individuation, ‘which determines a Being of any sort to a particular time and place, incommunicable to two Beings of the same kind’.Footnote 9 To be more specific, Locke wanted to address how substance, man and in particular persons retain their identity through time. To say that a thing is identical in this sense is to point to a specific unity relation between that thing at T and at T1. And Locke argued that substance, man and person each have a different unity relation.
(1) Perfect identity. The identity of substance is identity in what Butler called ‘the strict philosophical sense of the word’.Footnote 10 A singular substance at T is identical at T1 if, and only if, it continues to exist as numerically the same, unchanged and discrete particle. This criterion presumes that there is some indivisible substance, either of material or immaterial nature. But applied to a number of monadic particles, it implies, according to Locke, that if one particle were removed or added, it was no longer the same body or mass. A mass or body remains identical, however, if the same parts were ‘differently jumbled’.Footnote 11
(2) Organic identity. The idea of man refers to a living body of which the parts are organised so as to sustain a continued life. In this sense, the idea of a man is not different from the idea of animal or vegetable. Both are living and growing organisms. For Locke, the identity of a man is the continuation of the same life ‘by constantly fleeting particles of matter, in succession vitally united in the same organized body’. So, while the parts can be replaced, a living body remains identical as long as the body parts sustain the same life.
(3) Personal identity. The identity of a person refers to the ‘sameness of a rational being’. But in what consists this sameness? According to Perry, we first need to distinguish between the two subquestions to fully appreciate the complexity of the problem. Where K stands for person, we have to determine:
(i) What relation obtains between simultaneous K-events that are events belonging to the same K;
(ii) What relation obtains between K-stages that are stages of the same K.Footnote 12
The first question is less important. It probes into the synchronic nature of a person at any one moment in time. What makes me one person at this very moment? As we saw, for the Lockean, a person is constituted by a variety of psychological experiences, or what Hume called perceptions. In one view, we could have just one such experience at a time, so the question does not arise. We are a thought, desire, perception or some other ‘person-event’. In another view, such events first need to be bundled in a synchronic unity relation. But both views eventually arrive at the notion of a temporal part of a self or a ‘person-stage’.Footnote 13 The second question examines the diachronic unity of a person over time. What historical relation should hold between two or more person-stages such that we could say that these person-stages form one person with one personal history? Is there a special connection such that if and when person-stages are so connected, certain psychological experiences at T and T1 are the experiences of the same person?
According to Locke himself, the faculty of consciousness was the key to both questions. It is through consciousness, he argued, that simultaneous sensations, experiences and thoughts are identifiable as mine. If I am thirsty and tired these experiences are mine because I am aware of them. These events are suitably linked into the same person-stage, that is, because they figure in the same state of consciousness. More importantly, Locke went on to claim that personal identity is extended through time by means of memory: ‘This personality extends it self beyond present Existence to what is past, only by consciousness, whereby it becomes concerned and accountable, owns and imputes to it self past actions, just upon the same ground, and for the same reason, that it does the present.’Footnote 14
If I remember visiting New York in 1996, then I am consciously aware that I then visited New York. In other words, I know this experience to have been mine.Footnote 15 Thus, my person-stages are linked into the same personal history if a later person-stage contains or could (if I made a proper effort) contain a memory of an experience of an earlier person-stage.
This memory criterion invited much early critique, most notably by Joseph Butler and Thomas Reid.Footnote 16 In response to this, others have extended and fine-tuned Locke’s argument in diverse and elaborate ways. While it goes beyond the scope of this paper to address all these developments, it is important to mention the following two. First, as Hume points out, personal identity should include connectedness between all kinds of psychological states.Footnote 17 Personal identity does not just consist of memory, but also of thoughts, beliefs, intentions, desires, character traits and so on. This extension is important because the idea that memory is a necessary rather than a sufficient criterion for personal identity is in itself implausible. I do not now remember many of my experiences of last year. Consequently, the basis of my personal identity would seem indefensibly weak. On the extended version of the criterion, however, my identity is more robust. I may not remember much about last year, but I have the same character traits and mostly the same beliefs and long-term plans. Hume argues that two person-stages are linked into one personal history whenever they sufficiently resemble each other. So if a desire or some other mental characteristic at t is still present at T1, then both person-stages are connected in the relevant way. He also suggests that person-stages are relevantly linked by causal connections. As Shoemaker puts it, two person-stages are linked ‘if the later of them contains a psychological state (a memory impression, personality trait, etc.) which stands in the appropriate relation of causal dependence to a state contained in the earlier one’. So even if a desire or a conviction is not retained in a later person-stage, it may still have inspired or formed the source of a later desire or conviction. In this view, then, person-stages are linked in the way of a person, not just if they are connected by memory, but also if they resemble each other or if a person-stage contains a ‘successor state’ of an earlier self.
Second, building on the work of others, Parfit makes a distinction between psychological connectedness and psychological continuity. ‘Connectedness’ refers to the direct psychological links between two person-stages. If I could remember an experience I had twenty years ago, then there is one such link. Locke seems to have thought that connectedness is a necessary condition of personal identity through time. However, as his critics noted, psychological connectedness is not a transitive relation.Footnote 18 Neo-Lockeans have since tried to amend this criterion in order to overcome this difficulty. The solution usually advanced grounds the criterion in the continuity of mental states rather than just the direct connectedness between them.Footnote 19 Parfit, however, points out that continuity refers to an essential overlapping chain of strong psychological connections. In this view, each connection is like a strand in a rope. While most strands do not touch each other directly, they are all indirectly connected into one rope. Hence, Parfit defines the criterion of personal identity – or what he refers to as ‘Relation R’ – in terms of ‘psychological connectedness and/or psychological continuity’.Footnote 20
Briefly, I now want to turn to the second school of thought. At the deepest level, this school rejects the Lockean distinction between the identity-criteria for person and substance. The identity of a person, in fact, is not given by the temporal continuity of purely mental states such as thoughts, intentions and memory, but rather by the persistence of an underlying substratum, an Ego, to which these thoughts and intentions are said to belong. Personal identity through time is not the identity of thought and perception, but the identity of the subject that thinks and perceives. As Reid expressed the difference between Locke’s view and his own: ‘My personal identity […] implies the continued existence of that invisible thing which I call myself. Whatever, this self may be, it is something which thinks, and deliberates, and resolves, and acts, and suffers. I am not thought, I am not action, I am not feeling; I am something that thinks, and acts and suffers. My thoughts, and actions, and feelings change every moment; they have no continued, but a successive, existence; but that self or I, to which they belong, is permanent, and has the same relation to all the succeeding thoughts, actions and feelings which I call mine.’Footnote 21
In other words, thoughts must belong to a Thinker, something that exists separately from our mind-states.Footnote 22 The most important upshot of this theory is that personal identity is irreducible. Persons are not bundles of the temporal parts of a person, tied together as the strings in a rope, but indivisible. As Reid put it: ‘The identity of a person is a perfect identity: wherever it is real, it admits of no degrees; and it is impossible that a person should be in part the same, and in part different; because a person is a monad, and is not divisible into parts.’Footnote 23
2 Person and state
I do not want to concern myself here with the debate between the two theories. I shall simply assume that the reductionist account along psychological lines is the more plausible theory.Footnote 24 I now want to flesh out the psychological approach by looking at some of its most remarkable upshots. That is, if the bundle theory were correct, individuals would after all be comparable to states and other associations. I have already remarked that this (inversed) analogy has two different aspects. First, it claims that the identity-relations of persons are like the identity-relations of associations. Second, it argues that we should consequently treat persons like we normally treat associations in normative theory. I want to concentrate on the issue of identity first.
Reductionism about personal identity turns the analogy between person and state upside down. Like Hume, Parfit argues:
‘When considering nations, most of us are Reductionists. We believe that the existence of a nation involves nothing more than the existence of its citizens, living together on its territory, and acting together in certain ways. In contrast, when considering persons, most of us believe the Non-Reductionist view. We believe that our identity must be determinate. This cannot be true unless a person is a separately existing entity, distinct from his brain and body, and his experiences. Most of us are thus Reductionists about nations but not about people. It is the difference between these common views which explains the two comparisons. The claim that X is like Y typically assumes the common view of Y. We shall therefore say, “People are like nations” if we are Reductionists about both. If we are Non-Reductionists about both, we shall say, “Nations are like people.” The belief in super-organisms may be a Non-Reductionist View about nations.’Footnote 25
The nub is that psychological reductionism cannot account for personal identity in the strict or perfect sense of the term. Personal identity thus becomes comparable to the identity of political parties, churches, nations, states and other relational facts. Like the identity of associations, the idea of personal identity must be applied in the imperfect sense.
As we saw, perfect identity is identity of the kind that admits neither of degrees nor change. Whether a thing is perfectly identical at another time is a question that has a determinate answer. A is identical to A1 in the perfect sense if, and only if, they are numerically the same thing. In the Ego theory, personal identity is perfect, a matter of all or nothing, not of degree. However, if the person is constituted by psychological events, this is no longer true. Reid wrote: ‘Our consciousness, our memory, and every operation of the mind, are still flowing like the water of a river, or like time itself.’Footnote 26 Thus, psychological events form a stream of incessant change and movement. My thoughts, feelings and ideas now are never numerically identical to my thoughts at the next moment. They might be matching in content, or they might be connected in a causal chain, but they are not literally the same thoughts. So once the bifurcation between the ideas of substance and a person goes through, personal identity, in this perfect sense, must be a fiction.
We have to turn, therefore, to a relaxed understanding of identity, a notion more forgiving of change and degrees. As Hume noted, many of our everyday judgments on the identity of objects are relaxed in this sense. For example, when I shave off my beard, I do not conclude that my body is no longer the same. We believe that such a partial change is simply too insignificant to warrant this conclusion. As the bulk remains intact, so my body endures. Hume mentions the example of a river. The water molecules that form that river are forever in flux, hurtling down to the sea, eventually merging with it within days or weeks. Yet we do not hesitate to refer to it as the same river. Even rapid motion may not affect our sense of the identity of things. What these examples show is that our sense of identity survives in the face of change as long as this change is ‘identity-preserving’.Footnote 27 And this means that changes have to respect certain constraints. It matters, for instance, that change is gradual or insignificant in proportion to the whole. Usually, it also matters that the replacing parts perform the same function or are causally related to the previous parts. Such constraints form the basic elements of a theory of imperfect identity.
Essentially, Hume and Parfit apply this understanding of imperfect identity to persons.Footnote 28 Relation R refers to the extent to which some person-stage Jack at T is psychologically connected and/or continuous with some other person-stage John at T1. But this entails that the identities of persons are now comparable to those of political parties, states and relational facts. For example, we know that the identity of groups and associations is indeterminate. Is France the same country as it was fifty years ago? While some things have changed, other things have not. Is this change identity-preserving? Here the problem is that there is no non-arbitrary cut-off point. Whether France is the same cannot be decided with accuracy. We know that there is some measure of historical continuity. But to say that France is or is not the same country is arbitrary. The same is now true for personal identity. According to Parfit, the question of whether or not some person is identical through time has no determinate answer.Footnote 29 The psychological connectedness between person-stages tends to vary. Remote person-stages are often less connected. Indeed, they may no longer be connected at all. As he puts it:
‘On my proposed way of talking, we use “I”, and the other pronouns, to refer only to the parts of our lives to which, when speaking, we have the strongest psychological connections. When the connections have been markedly reduced when there has been a significant change of character, or style of life, or of beliefs and ideals we might say, “It was not I who did that, but an earlier self”.’Footnote 30
Do two such person-stages still belong to the same person? Contrary to what people tend to believe, the answer is not simply ‘Yes’ or ‘No’. The only answer available is that they are to some degree psychologically connected. We could say that they are either weakly or strongly connected, but we cannot conclusively decide that they are identical or not.
We also know that associations often split up. Belgium may divide into a Flemish part and a Walloon part. In which part, if at all, will Belgium continue? Or does it continue in both? These seem hard questions. But they are equally hard for personal identity. One consequence of the psychological criterion of personal identity is that persons could branch or divide in two.Footnote 31 In such cases, a person-stage Jack at T has strong psychological connections with John at T1 and with Jim at T1. Personal division may be technically impossible, but it is not inconceivable. Parfit imagines a world where surgeons could separate the two halves of Jack’s brain, each half retaining consciousness. Both halves could then be transferred to different bodies resulting in the birth of John and Jim. Both John and Jim are psychologically connected with Jack in the way of a person. They have memories of Jack’s previous experiences; they share his beliefs and values, and so on. Now, in the non-reductionist theory, persons cannot divide into parts. However, on the basis of the reductionist theory, Jack can be seen to survive as both John and Jim.Footnote 32 According to Parfit, personal division is ‘almost as good as survival’ because R-relatedness is what fundamentally matters. So he claims that: ‘Since my relation to each of the resulting people is about as good as if it were identity, it may carry most of the ordinary implications of identity.’Footnote 33 For example, John and Jim may both be held responsible for Jack’s actions. The key is that Jack and John and Jim are all psychologically connected. There is simply nothing we could sensibly add.
In the reductionist account, in short, the identity of a person is pulled together from temporally separate parts. Hence, it should be obvious that a person cannot be a monadic unity. The unity of a person is a measure of organisational integrity. Comparing persons to associations, Parfit writes:
‘In 1881 the French Socialist Party split. What happened? Did the French Party cease to exist, or did it continue to exist as one or other of the two new Parties? Given certain details, this would be an empty question. Even if we have no answer to this question, we could know just what happened.’Footnote 34
And elsewhere he argues:
‘Talk about successive selves can easily be misunderstood, or taken too literally. It should be compared with the way in which we subdivide a nation’s history. We call this the history of successive nations, such as Anglo-Saxon, Medieval, and Tudor England.’ Footnote 35
Persons, in other words, should not be contrasted with groups. We are merely associations of a particular kind.
3 The extreme claim
Parfit’s analysis puts associations and individuals back on a par. Associations are constituted by numbers of individuals bound together by a particular relation. And this means that associations have no identity in the strict sense. However, this fact does not differentiate associations from persons. For, as we have seen, persons are in turn constituted by a number of person-events and person-stages bound together like the strings in a rope.
Now of course, this is not the kind of analogy that corporate realism had hoped to uncover. Rather than raising the moral stature of associations, the analogy may only show that we ought to demote the moral standing of persons across the board. It may prove, in other words, that individuals ought to be treated like we normally treat associations, rather than the other way around. To put it differently again, this analogy could now be used to argue that ethical individualism is just as old-fashioned as corporate personality.
Locke’s eighteenth-century critics thought that this would inevitably be the outcome. Reductionism about personal identity, that is, would inevitably result in moral bankruptcy. They therefore thought that Locke had to be wrong. ‘A part of a person’, Reid wryly remarked, ‘is a manifest absurdity.’Footnote 36 Even the briefest of introspective moments or the tiniest morsel of common sense would surely indicate that much. For Reid, strict personal identity was the ‘foundation of all rights and obligations’.Footnote 37 For if Locke were right, Reid thought, ‘no man could be responsible for his actions’.Footnote 38 Butler pointed out that Locke’s theory was equally hostile to prudential reasoning: ‘for if the self or person of today, and that of tomorrow, are not the same, but only like persons, the person of today is really no more interested in what will befall the person tomorrow, than in what will befall any other person’.Footnote 39 Put differently, Locke threatened to explode all the familiar reasons that guide human action. Concepts like desert and punishment would, sure enough, lose their last shred of authority. The choice we faced, according to this objection, was between reason on the one hand and moral anarchy on the other.
Now, whatever he meant to say, Locke did certainly not want to leave morality in ruins. However, note that for the hardnosed reductionist these practical objections merely beg the question. According to Parfit, the objection only shows how much practical reason is tainted by ‘Cartesianism’.Footnote 40 Parfit claims that we should simply revise our theories of rationality and morality. Nonetheless, the practical upshots of reductionism are by no means self-explanatory. Parfit makes an important distinction between the ‘extreme claim’ and the ‘moderate claim’. Both claims are essentially at odds over how to appraise the moral and rational status of R-relatedness. The extreme claim bluntly denies that relational facts could have moral and rational significance. Thus, it denies that the interests and intentions both of past and future person-stages are in any sense relevant to my present self. And by doing so it denies exactly what the moderate claim wants to affirm.
According to Parfit, each claim is in principle defensible, and thus, his own positioning remains obscure, straddling both views.Footnote 41 When confronting the effects of reductionism on prudence, responsibility, punishment and obligations of commitment he adopts a decidedly moderate stance. However, there is less trace of this moderate stance in his treatment of justice. While it is clear that Parfit’s overall agenda is to show that practical reason ought to become more impersonal, it is not entirely transparent whether these implications must be understood in an ‘extreme’ or ‘moderate’ way. As I hope to make clear in the rest of this paper, much rests on this distinction. Hence, in the remainder, I am going to ignore Parfit’s own somewhat awkward positioning in between these opposing claims. I shall concentrate on the extreme claim.
3.1 Prudence and future selves
I want to begin by exploring the consequences of reductionism for prudential or self-interested reasoning. To illustrate the problem, let us assume that Jack is a heavy smoker. Jack is well aware that as a result he might die twenty years from now. But why should he care? The standard reply rests on two basic arguments. First, Jack has prudential reason to care because it is his future. The reason for him caring is agent-relative. Second, his future wellbeing is equally his wellbeing. The reason why he should care is time-neutral. His future interests are no less important than his present interests.Footnote 42
There is, of course, one further assumption on which the standard reply depends: it must really be Jack who will die in agony. Now, in the Ego theory, this condition may (perhaps) be met. But in the bundle theory, this is much less evident. There are two ways in which this is less evident. First, it is not inconceivable that the psychological relatedness between Jack and John will be weak. So it is unclear to Jack that he will suffer terribly later. Second, even if Jack and John were strongly linked, we could still deny that their psychological relatedness provides the proper agent-relative reason for prudential concern.
The extreme claim advances the latter version of the argument. It asserts that, if reductionism were true, Jack could never have a genuinely agent-relative reason to care about a future person-stage of his.Footnote 43 According to this claim, there could be such a reason only if Jack and John were perfectly identical. Only this kind of identity would guarantee that he would be the same self in the appropriate, reason-giving sense. It could be said that the extremist argues for what Parfit describes as the ‘full relativity of reason’.Footnote 44 Standard prudential reason is agent-relative but neutral across time. The extreme claim argues that reasons should also be relative across time. Not only does Jack have an agent-relative reason to care for himself rather than for others, he also has reason to care for his present self rather than for the wellbeing of a future self. To put it differently, his future self is like another person, whose interests or preferences do not figure as a basic element in his prudential calculation.Footnote 45
What is the argument for temporal relativity? The argument ultimately revolves around the normative appreciation of relational facts. The standard view supports a very sharp and far-reaching distinction between a person on the one hand and the social relations between persons on the other. If John were a different person, his suffering would carry no direct or non-derivative weight for Jack. Nor would it be of ultimate significance, in this view, if Jack had been closely related to John. Interpersonal relatedness does not matter in the sense that it facilitates agent-relative reasons. However, if this last claim were true, why still be neutral across time? What is so special about my future person-stages in contrast with my friends or relatives or neighbours? After all, we know now that personal identity is itself merely one kind of relatedness among many. Put differently, the extremist applies the agent-relativity of reason not just to the separateness between persons, but also to the separateness between person-stages.
Notice that in this view the strength of the relatedness is entirely inconsequential. Jack may be psychologically connected to John, but this is still no reason to take his suffering more seriously than the suffering of others. Granted, he may have long-standing ideals, plans and hopes, and John may have many memories of Jack’s earlier experiences. Indeed, all these connections may give his life the relational coherence of a Humean bundle of experiences. But psychological relatedness, however strong, will still not suffice to give him an agent-relative reason to prevent his future misery. Thus, although Jack may be psychologically linked to John, both remain separate agents. Consequently, it is not imprudent or irrational for Jack to pay no heed to his future suffering.Footnote 46
Notice further that the extreme claim could still condemn imprudent behaviour by shifting towards moral reasons.Footnote 47 According to Parfit, the standard view of prudential reason is flawed because it is a hybrid.Footnote 48 It is agent-relative, but time-neutral. Alternatively, there are two pure theories. First, we could resort to the full relativity of reasons. This is the pure theory of rationality. Second, we could advocate the full neutrality of reasons. This is the pure theory of morality. Moral reasons are agent-neutral as well as neutral to time. According to morality, our future selves could be compared to future generations. Whereas most of us believe that we are one and the same person throughout our biological lives, and possibly even afterwards, reductionism permits that one biological body may sustain several persons. Consequently, we have to say that someone will suffer terribly as the result of my actions. That person will not be me, but there will be terrible suffering at some point in the future. In other words, while the extreme claim rules out an appeal to prudence, it nevertheless unlocks the door to morality. Morality suggests that Jack has reason to prevent John’s agony, not because it is rational, but because he has a duty to avert terrible future experiences, regardless of whose experiences they are. If the extremist is correct, we may have a moral duty to act ‘prudentially’.
3.2 Desert, responsibility, obligation and past selves
Essentially, the extreme claim asserts that the relatedness between temporal person-stages is of no non-derivative significance. From the practical point of view, the fact that person-stages are connected and/or continuous in the ways that Locke and Hume describe bears no genuine weight. That is, these relations ought to have no greater impact on our rational and moral judgments than the relations we have with other people. We have seen what this implies for prudential reason. The interests of my future person-stages do not count as reasons for me now. What, then, are the implications for backward-looking concepts such as obligations of commitment, desert, responsibility and punishment?
Locke himself suggests one astonishing answer. According to him, a person cannot deserve punishment, in principle at least, for a crime committed in the past if that person has no recollection of that crime.Footnote 49 Reid suggests, even more radically, that no one could ever be held responsible for any of his past deeds.Footnote 50 I cannot, after all, deserve punishment or reward for something I did not do. Desert and responsibility, that is, can only be allocated to those entities that meet the proper identity conditions. Clearly, I may well be related to an earlier person-stage. And I may even remember doing this or that. But that cannot justify that I should now be held responsible for these actions. Stone makes the point as follows:
‘If I say that I feel remorse because I murdered my brother, you would be right to be puzzled. For all this comes to is that I happen to stand in certain causal relations, to the man who killed my brother and no one else does, and why feel remorse over that? Why feel guilt? And if identity is merely psychological continuity it becomes impossible to account for the fact that people bear responsibility over time. “I am the man who did the deed,’ the criminal confesses, “but all this comes to is that I have the misfortune to stand in certain causal relations to the man who did it and no one else does. I’m not to blame for this situation; I find myself this way. Punishing me for his crime is no better than punishing the son for the crimes of his father.”’Footnote 51
Reductionism, in this extreme view, is simply incompatible with notions like responsibility and desert.
Similar implications follow for commitments and promises. A promise retains its validity over a period of time only if ‘promisor’ and ‘promise’ retain their perfect identity.Footnote 52 John may be related to Jack, but this cannot bind him to a commitment or a promise originally assumed by Jack. Their psychological relatedness is, in this sense, simply irrelevant. Jack may have promised to meet you for lunch. But that promise cannot obligate John. Jack’s promise is not merely a past commitment; it is the commitment of another person.
This reasoning may be applied to any intentional action since the intention-former is different from the intention-executor. As has been shown in Section 3.1, I cannot bind my future self and he (my future self) has no compelling reason to take ownership of ‘my’ intentions – and so intentional, and therefore moral, action is impossible.
Desert, responsibility and obligation are backward looking in the sense that they presuppose a view on which present and future selves could be bound by the actions, intentions and commitments of an earlier self. Past intentions, promises and actions, that is, must carry normative weight such that I can later be said to deserve, owe, be obligated and so forth. The extreme claim, however, denies that personal identity has these normative aspects. Psychologically relatedness has no normative upshots.
3.3 Justice and the separateness of persons
There is yet another way in which reductionism might have important implications. Parfit argues that it lends credibility to impersonal versions of utilitarianism. It is in arguing for this position that Parfit’s own view becomes almost indistinguishable from the extreme claim.Footnote 53 It should be noted that rights-based theories of justice, of the kind advanced by Rawls, commonly claim overriding importance for persons and personal identity.Footnote 54 This claim is often described in terms of the separateness of persons. The appeal to this separateness revolves around what Parfit calls the ‘compensation claim’. This states that the benefits of one person cannot compensate the burdens of another. So we cannot justify imposing certain costs on Ralph, merely by claiming that this will greatly benefit Paul. To do so would be unfair. Extreme cases aside, this is true even if the benefits outweigh the costs. It is for this reason that, apart from an increase in benefits, these theories are also interested in finding just distributions of burdens and benefits. Thus, the notion of the person puts constraints on consequentialist morality. About the tradition of classical utilitarianism, Rawls remarks that it ‘does not take seriously the distinction between persons’, and mistakenly adopts ‘for society as a whole the principle of rational choice for one man’.Footnote 55 Utilitarianism, that is, ignores the crucial distinction between social justice and rational prudence. As he puts the criticism: ‘if we assume that the correct regulative principle for anything depends on the nature of that thing, and that the plurality of distinct persons with separate ends is an essential feature of human societies, we should not expect the principles of social choice to be utilitarian’.Footnote 56 And Nagel writes: ‘To sacrifice one individual life for another, or one individual’s happiness for another’s is very different from sacrificing one gratification for another within a single life.’Footnote 57
Notice that Rawls and Nagel assume compensation within one life to be fairly unproblematic. That is, a burden for someone now could be compensated if benefits would accrue to that person later. We generally do not think it wrong to incur certain burdens presently if this will pay off later on in life. Trade-offs over time are not objectionable, then, as long as they occur within the life of one person. Thus, it is evident how much rights-based theories rely on sharp boundaries between persons. Within the unity of a life, compensation and trade-offs are permitted. Here, we can just maximise. But as soon as costs and benefits are assigned to separate persons, maximisation should be abandoned.
However, in the reductionist account of personal identity, there are no fixed and clear boundaries. According to Parfit, reductionism has a dual effect on theories of just distribution. First, it leads to an extension of the scope of justice. That is, benefits and burdens ought also to be distributed fairly between the temporal person-stages that make up our lives. As Parfit observes, the question ‘When?’ should be compared to the question ‘Who?’Footnote 58 As a result, the maximising principle should be constrained not less, but further. Incurring a present cost in order to reap a later benefit is precisely comparable to the case of Paul and Ralph. Future benefits do not make up for present costs because my future self and I are not strictly the same person. We may be psychologically related, but compensation still constitutes an injustice. After all, Paul may also be psychologically related to Ralph, but this fact does not justify compensation either.
However, Parfit goes on to show that the principles of fairness and equality lose weight at the same time. In fact, the entire notion of distributive justice becomes blurry at this point. Over what should burdens and benefits still be distributed? Why is it bad that some bundles of experiences are miserable and others happy? Surely, what is more important than the distribution of misery and happiness over such bundles, is that there are as few as possible miserable experiences. The problem, Parfit suggests, is that the units over which we distribute sink to the micro-level of temporally discrete experiences. But at this level, justice as fairness is simply on a par with utilitarianism.Footnote 59
Rawls argues that classical utilitarianism ‘extends to society the principle of choice of one man’, and that it does not recognise the separateness of persons as ‘an essential feature of human societies’. Parfit points out that Rawls is making a similar mistake. He extends to man the principle of rational choice for one temporally discrete person-stage, and he fails to see that the separateness of person-stages is an essential feature of human life.
The bottom line can be expressed in terms of the analogy between individuals and groups. The extreme claim maintains that individuals are like associations not just from the ontological point of view, but also from the point of view of justice. Justice as fairness presupposes that the wellbeing of persons has non-derivative and reason-giving value. But Parfit claims that persons are of no greater moral importance than nations. From the point of view of justice, the boundaries between people are like boundaries between countries. They delineate nothing of ultimate significance. Individualists agree when he claims:
‘If there is nothing more to a nation than its citizens, it is less plausible to regard the nation as itself the primary object of duties, or possessor of rights. It is more plausible to focus upon the citizens, and to regard them less as citizen, more as people. We may therefore, in this view, think a person’s nationality less morally important.’Footnote 60
However, because persons are really only groupings of experiences, we should focus more on ‘experiences themselves’ rather than the ‘subject of experiences’. He goes on to argue analogously:
‘It becomes more plausible, when thinking morally, to focus less upon the person, the subject of experience, and instead to focus more upon the experiences themselves. It becomes more plausible to claim that, just as we are right to ignore whether people come from the same or different nations, we are right to ignore whether experiences come from within the same or different lives.’Footnote 61
Of course, Parfit is always careful to couch his argument in terms of relative plausibility, but on a more bold statement, we might say that individuals, like nations, are no more than the receptacles of what really matters. In fact, moral individualists are exactly like extremists when considering relatives, friends and nations. They do not think that such ties and relations are themselves reason-giving. To use Godwin’s famous words, there is no magic in the pronoun ‘my’.Footnote 62 But the extremist’s point against ethical individualism is precisely that persons, too, are associations.
Conclusion
This paper argues for the inversion of the analogy between individual and group. We no longer believe in Gierkean collective Egos. Reductionism about personal identity suggests that we should also stop believing in the Self. On this basis, individuals and groups are once again on a level pegging. Associations are not super-individuals, but individuals are miniature associations. This reasoning has radical implications, for example, in the way we describe criminal groupings. We may delineate the ‘criminal’ groups (professional bank robbers, for example), with their loose bonds but tight sense of intentionality, from ‘criminogenic’ groups (street gangs, for example), with their tight bonds and loose sense of intentionality; or we may use some other typology to do the same job. However, my argument still applies. One cannot simply ascribe intentionality so easily nor use the concept of the group so freely.
This analogy has two different aspects. The bundle theory of personal identity provides the ontological side of the argument. It is a common view that groups consist merely in a plurality of individuals who are related or associated in a specific sense. According to the bundle theory, something similar applies to the individual itself. Personal identity is based on the psychological relatedness between intentional episodes or person-stages. A continuous person is an association between beliefs, desires and intentions at different points in time, not some ‘further fact’ that exists underneath them.
The practical side of the argument claims that, as a consequence, our theories of rationality and morality need serious revision. According to the extreme claim, reasons should become more impersonal. We should regard persons more like ethical individualists regard nations and other associations. Ethical individualists believe that the value of nations and states derives from the extent to which they contribute to individual wellbeing. The extreme claim holds that the value of personal identity is similarly derivative. Like associations, the good of an individual is not valuable in a fundamental sense, but only insofar as this contributes to the wellbeing of person-stages. Moreover, the extremist claims that we should reject notions of desert, responsibility and obligation.
The analogy between groups and associations, then, appears to hold up in spite of everything. But, it will be objected, does it actually strengthen the case for corporate personality? For although individuals are like groups, it is also obvious that this is a Pyrrhic victory. The problem is that we have levelled down rather than up. We may only have shown that the idea of a person should be expunged from rationality and morality altogether. For though one can show that the person breaks down into a sequence of either experiences, or stages, by utilizing Parfit’s work it is, nonetheless, worrying that this process may be without end and also that, perhaps, personal identity does not matter so much as had been ordinarily thought. This line of reasoning has considerable practical problems for the way we conceive of our actions in the world and for the way we organise ourselves legally, morally and socially. However, in being honest about the nature of persons, we have also exposed an implicit dishonesty in the way we have traditionally conceived of the structures those ‘persons’ inhabit – legally, morally and socially.
