‘But what, then, am I?’, Descartes famously asked. Although many of us today reject Descartes's equally famous answer – ‘I am an immaterial mind’ – Descartes was right, I believe, to identify himself with a thinking thing, a thing who ‘doubts, understands, affirms, denies, wills, refuses, and which also imagines and senses’.Footnote 1 But neither an immaterial mind nor a material brain is the thing that thinks. The thing that thinks is the person. Just as your legs and feet are the limbs by means of which you walk, you the person – not your legs and feet – are the walker; so too the brain is the organ by means of which you think, but you the person – not your brain – are the thinker.
Where Cartesians see a relation between minds and bodies, I see a relation between persons and bodies. Understanding ‘person’ to refer to entities like you and me, it is obvious that persons exist. And just as clearly there are bodies. So, the important philosophical question – whose answer cannot be read off neurophysiology or scientific psychology – is this: What is a person? What is the relation between a person and her body?Footnote 2
On the answer that I shall propose – I call it ‘the constitution view’ – persons are not identical to their bodies, nor to parts of their bodies (e.g. brains), nor to their bodies plus something else (e.g. immaterial souls). In ‘logical space’ there is room for another possibility, which I shall develop and defend. I shall explore the idea that a person is constituted by a body, where constitution is not identity. On such a constitutional account of persons and bodies, it is necessary that human persons are embodied; but it is not necessary that they have the bodies that they in fact have. Thus, the view that I shall develop shares with the Cartesian dualist the claim that persons are not identical to their bodies (I could have a different body from the one that I do have), and it shares with the classical materialist the claim that, necessarily, human persons are embodied. After setting out this constitution view, I shall turn to the metaphysics of resurrection.
First, let me comment on the term ‘human being’. Some philosophers use ‘human being’ to denote a biological kind.Footnote 3 Others use it to denote a partly psychological kind.Footnote 4 I use ‘human being’ in the latter way, to name a partly psychological kind, a human person. All human persons are human beings, and vice versa.
The constitution view of human persons
What makes a human person a person is having what I'll call a ‘first-person perspective’. What makes a human person human is being constituted by a human body.
A first-person perspective is the defining characteristic of all persons, human or not.Footnote 5 From a (robust) first-person point of view, one can think about oneself as oneself and think about one's thoughts as one's own. In English, we not only use first-person pronouns to refer to ourselves ‘from the inside’ so to speak (e.g. ‘I'm happy’) but also to attribute to ourselves first-person reference (e.g. ‘I wonder whether I'll be happy in ten years’). The second occurrence of ‘I’ in ‘I wonder whether I'll be happy in ten years’ directs attention to the person per se, without recourse to any name, description, or other third-person referential device to identify who is being thought about. The first-person perspective opens up a distinction between thinking of oneself in the first person and thinking of oneself in the third person. Once someone can make this distinction, she can think of herself as a subject in a world of things different from herself. And since human persons are necessarily embodied, a person can think of her body, as well as her thoughts, from her first-person perspective.
A being may be conscious without having a first-person perspective. Non-human primates and other higher animals are conscious, and they have psychological states like believing, fearing, and desiring. They have points of view (e.g. ‘danger in that direction’), but they cannot conceive of themselves as the subjects of such thoughts. They can not conceive of themselves in the first person. (We have every reason to think that they do not wonder how they will die.) So, being conscious, having psychological states like beliefs and desires, and having a point of view are not sufficient conditions for being a person.
To be a person – whether God, an angel, a human person, or a Martian person – one must have a first-person perspective. Person is a non-biological genus, of which there may be several species: human, divine, bionic, Martian, etc. It is in virtue of having a first-person perspective that an entity is a person. So, what makes something a person is not the ‘stuff’ it is made of. It does not matter whether something is made of organic material or silicon or, in the case of God, no material ‘stuff’ at all. In short, person is an ontological kind whose defining characteristic is a first-person perspective.
Babies are not born with the kind of robust first-person perspective that I have been describing, but they are born with what I call ‘rudimentary first-person perspectives’: they are sentient; they imitate; they behave in ways which require attribution of beliefs and desires to explain. An organism comes to constitute a person when it develops a rudimentary first-person perspective, provided that the organism is of a kind that normally develops a robust first-person perspective. Human babies are persons in virtue of having rudimentary first-person perspectives and of being members of the human species. Members of the human species – unlike non-human animals who may also have rudimentary first-person perspectives – normally develop robust first-person perspectives as they mature and learn a language. A human organism that has a rudimentary or a robust first-person perspective at time t constitutes a person at time t.Footnote 6
At the other end of human life, a person who becomes demented still has a first-person perspective. Patients who are severely mentally handicapped (e.g. with late Alzheimer's disease) still can conceive of themselves as ‘I’. If you think that you don't exist (Cotard's syndrome), you have a first-person perspective. Your existence on earth comes to an end with the permanent and irretrievable loss of the ability to think of yourself from the first person. As long as it is physically possible for a patient (even in a coma) to regain the ability to think of herself in the first-person way, there is a person. When the physical possibility of that ability is forever lost (as in the case of Terry Schiavo), but the brain stem is still functioning, then there is no person there, but only an organism.
A first-person perspective is the basis of all self-consciousness. It makes possible an inner life, a life of thoughts that one realizes are one's own. The appearance of first-person perspectives in a world makes an ontological difference in that world: a world populated with beings with inner lives is ontologically richer than a world populated with no beings with inner lives. But what is ontologically distinctive about being a person – namely, a first-person perspective – does not have to be secured by an immaterial substance like a soul.
Human persons differ from non-bodily or immaterial persons (if there are any) in that human persons are not just pure subjects; they do not exist unembodied. So, myself includes my body. And persons' bodies are the objects of first-person reference. If Smith wonders whether she has cancer, she is wondering about her body from a first-person perspective. She is not wondering whether there is a malignant tumour in some particular body identified by a third-person demonstrative pronoun or description; she is wondering whether there is a malignant tumour in her own body, considered as herself. This is different from wondering about a material possession, say. If Smith wonders whether her car will run, she wonders about a particular car, which she identifies by a description or a third-person demonstrative reference. Without a third-person way to think about the car, she could not wonder about its battery. But if Smith is wondering how she will die, she can think of her body as her own without recourse to any name or description or second- or third-person demonstrative pronoun. And reference without recourse to the familiar third-person devices is the mark of first-person reference.
Human persons – who, like all persons, have first-person perspectives – are distinguished from other kinds of persons in that human persons are constituted by human bodies that are the objects of their first-person thoughts. A human person is a person who is constituted by a human body during some part of her existence. (I say ‘is constituted by a human body during some part of her existence’ to avoid issues raised by the Incarnation. The orthodox Christian view is that the eternal second person of the Trinity is identical with Jesus Christ, who is both fully human and fully divine. How this could be so is ultimately a mystery that requires special treatment far beyond the scope of this paper.)
Putting that issue aside, a human person is constituted by a biological entity – an organism, a member of the species homo sapiens – that is physically able to support first-person intentional states.Footnote 7 (It is up to neuro-scientists, not philosophers, to determine the biological conditions under which a human being is able to support first-person intentional states.)
A human person – Smith, say – must have a biological body that she can think about in a first-person way. Smith can think of a biological body in the first-person way if she can entertain thoughts about that body without aid of a name, or description, or third-person pronoun. Even if she is totally paralysed, Smith has a first-person relation to her body if she can entertain the thought, ‘I wonder if I'll ever be able to move my legs again’. To put it differently, Smith can think of a biological body in the first-person way if she can conceive of its properties as her own. For example, Smith's thoughts about how photogenic she (herself) is, or her worries about her (own) state of health – thoughts that she would express with first-person pronouns – make first-person reference to her body as her own. Since a body constitutes a person, a first-person reference to one's body is ipso facto a first-person reference to oneself.
So, what makes a particular body Smith's, rather than someone else's, is that it is the body that Smith can think of and refer to in a first-person way, ‘from the inside’. The body to which Smith has a first-person relation is the body some of whose parts she (normally) can move without moving anything else, the body that she tends when she is in pain, and the body that expresses her intentional states. States like pain, longing, sadness, hope, fear, frustration, worry, effort, and joy as well as states like believing, desiring, and intending are expressed through posture, facial expression, sounds, and other bodily motions.
The body that expresses Smith's intentional states is the body to which Smith has a first-person relation. Smith's first-person relation to her body at t does not imply that Smith is actually thinking of her body at t; indeed, Smith may believe at t that she is disembodied. The body to which Smith has a first-person relation is the body whose sweaty hands manifest the fact that Smith is nervous, and the body whose stomach's being tied in knots expresses the fact that Smith is frightened, or the body that would move if Smith carried out her decision to leave the room. Smith's body at time t distinguishes Smith from all other persons at t. What distinguishes me now from all other coexisting persons – even physical and psychological replicas of me, if there are any – is that at this time, I have a first-person relation to this body and to no other; and any replica of me at this time has a first-person relation to some other body, but not to this one.
The body to which I have a first-person relation constitutes me. But what is constitution? Elsewhere,Footnote 8 I have a more rigorous account of the relation of constitution, but the general idea of constitution is this: when various things are in various circumstances, new things – new kinds of things, with new causal powers – come into existence. Every concrete object is of (what I call) a primary kind. A thing has its primary-kind property essentially. So, kind membership (or species membership) is not contingent. The relation of constitution unites things of different primary kinds, and hence things with different essential properties. For example, a human organism is essentially a member of the human species; a person essentially has a first-person perspective.Footnote 9 A human person is a person constituted by a human organism.
Constitution is everywhere: pieces of paper constitute dollar bills; strands of DNA constitute genes; pieces of cloth constitute flags; pieces of bronze constitute statues. Constitution is never identity: the piece of cloth that constituted the first Union Flag could exist in a world without nations; hence that piece of cloth could exist without constituting a flag, and the first Union Flag is not identical to the piece of cloth that constituted it. Similarly, the piece of bronze that constituted Myron's statue ‘Discobolus’ could have existed in a world without art; hence that piece of bronze could have existed without constituting a statue.Footnote 10
The non-identity of persons and their bodies may be seen in another way – in a way that has no parallel for statues. Despite the similarities between persons and statues, there is a major difference between them: Persons have bodies that change drastically over the course of a person's life, but pieces of marble that constitute statues change very little. To put it the other way around: if the piece of marble that constitutes Michelangelo's ‘David’ were to change significantly, the statue ‘David’ would no longer exist; but Smith's body alters radically while Smith endures.
Leaving aside the analogy between persons and statues, consider another argument against the person/body identity theory, based on criteria for individuating bodies and persons. Criteria of individuation may be vague, but they are not totally elastic. Smith's body is a human body in virtue of being a member of the species homo sapiens. What makes something a human body are its biological properties; its career may be followed from beginning to end without respect to whether or not it is any person's body. Similarly, its persistence conditions are independent of whether or not it is any person's body. The identity of a human body is independent of whether it is Smith's or any other person's body.Footnote 11
In the natural course of things, our organic bodies undergo full atomic replacement over some years, and we persons survive this total replacement without interruption in mental functioning. It seems possible that we could equally survive gradual replacement of organic cells by bionic cells – until finally the body that sustains us is no longer an organic body. Exactly how much replacement of parts a human body may undergo and still remain a human body is somewhat vague, but if a body is mostly made up of inorganic material and is not sustained by organic processes, it is not a member of the species homo sapiens. The non-organic body that ends up constituting Smith now is a different body from the organic body that was a member of the species homo sapiens.
Consider the organic body that Smith was born with. Call it ‘OB’. Suppose that the organs of OB were totally replaced over a period of time by bionic parts, until what remained was a fully bionic, non-biological body that resembled OB in appearance, that moved in ways indistinguishable from OB, that emitted sounds that we took to be English sentences that reported memories of things that had happened to Smith, and indeed that we took to be professions that this person was Smith. Is the bionic body the same body as Smith's biological body OB? No. OB was a carbon-based body that was a member of the species homo sapiens. The bionic body is not a member of any biological species. Would Smith still exist? Of course. Otherwise Smith's possessions and property should be taken from the bionic-body-Smith and distributed to Smith's heirs. After the organ replacement, Smith would still exist but would no longer by constituted by OB; rather, Smith would be constituted by a bionic body. (I really do not like bizarre thought-experiments, but I think that we are actually close to bringing this thought experiment to fruition. There are now devices implanted in brains that allow paralysed people to operate computers by their thoughts, cochlear implants allow deaf people to ‘hear’, and so on. Moreover, it's easy to imagine billionaires seeking ‘whole-body’ replacements to prevent ageing.)
The point is that this is a realistic example that shows that a single person may be constituted by different bodies at different times: Smith had a first-person relation to a biological body at one time, and to a bionic body at a later time, and a biological body is essentially organic, and is not numerically identical to any bionic body. Note that spatio-temporal continuity in general does not signal sameness of entity: Very slowly, atoms could be added or taken away from Smith's biological body until it was indistinguishable from a turnip or a bookcase. In that case, it would no longer be the same body, and presumably Smith would no longer be with us. Indeed, there may be a period of time during which it is indeterminate whether there is a human body or not. I have argued elsewhere that everything that we encounter in the natural world comes into existence gradually; hence, everything that we interact with has vague temporal boundaries.
To sum up: on the constitution view, a human person is constituted by a particular biological body, but the person is not identical to the body. What distinguishes persons from all other beings is that they have first-person perspectives essentially. The persistence conditions of a human person are determined by the property in virtue of which she is a person – viz. the property of having a first-person perspective: a human person could cease to have an organic body without ceasing to exist. But she could not cease to be a person without ceasing to exist.
On the constitution view, then, a human person and the organic body that constitutes her differ in persistence conditions without there being any actual physical intrinsic difference between them. The persistence conditions of animals – all animals, human or not – are biological; and the persistence conditions of persons – all persons, human or not – are not biological.
On the metaphysics of resurrection
All the great monotheistic religions – Judaism, Christianity, and Islam – have doctrines of an afterlife. These are religious doctrines, whose grounding in scripture and tradition leaves open how they should be understood metaphysically. I want to focus on the Christian doctrine of resurrection, and to find the best metaphysics to support it.
To begin, consider three features that characterize the Christian view of resurrection.
First, identity: the very same person who exists on earth is to exist in an afterlife. Individuals exist after death, not in some undifferentiated state merged with the universe, or with an eternal mind, or anything else. Not only is there to be individual existence in the resurrection, but the very same individuals are to exist both now and after death. ‘Survival’ in some weaker sense of, say, psychological similarity is not enough. The relation between a person here and now and a person in an afterlife must be identity.
Second, embodiment: resurrection requires some kind of bodily life after death. Post-mortem bodies are different from pre-mortem bodies in that they are said to be ‘spiritual’, ‘incorruptible’, or ‘glorified’. Even if there is an ‘intermediate state’ between death and a general resurrection, in which the soul exists disembodied, those who live after death will ultimately be embodied, according to Christian doctrine.
Third, miracle: life after death, according to Christian doctrine, is a gift from God. Christian doctrine thus contrasts with the Greek idea of immortality as a natural property of the soul. The idea of miracle is built into the Christian doctrine of life after death from the beginning. Since resurrection, if it occurs, is miraculous, we cannot expect a full philosophical account or explanation of it. There will always be some mystery left. The best that we can hope for is a metaphysics consistent with and congenial to the doctrine.
The task for a metaphysics of resurrection is to present a view of human persons whose persistence conditions allow, by means of a miracle, for post-mortem as well as pre-mortem life. The best that metaphysics can do is to show how resurrection is metaphysically possible. That is, any candidate for a metaphysics of resurrection must conceive of human persons in such a way that it is metaphysically possible (even if physically impossible) that one and the same person whose earthly body is corruptible may also exist with a post-mortem body that is incorruptible. That is the task. I shall argue that the constitution view fares better than its competitors in fulfilling that task.
There are a number of candidates for a metaphysics of resurrection:
(1) Immaterialism: sameness of person is sameness of soul both before and after death.
(2) Animalism: sameness of person is sameness of living organism before and after death.
(3) Thomism: sameness of person is sameness of body/soul composite before and after death.
(4) The memory criterion, according to which pre- and post-mortem persons are the same person if and only if they are psychologically continuous.
(5) The soul-as-software view, according to which sameness of person is analogous to sameness of software.
(6) The soul-as-information-bearing-pattern view, according to which sameness of person is sameness of pattern of information.
(7) The constitution view, which I explained earlier.
Let's consider each of these.
Immaterialism
Although souls in this world are linked to brains, there is no contradiction, according to Richard Swinburne, in the soul's continuing to exist without a body. Indeed, the soul is the necessary core of a person which must continue if a person is to continue.Footnote 12 Since, on Swinburne's view, no natural laws govern what happens to souls after death, there would be no violation of natural law if God were to give to souls life after death, with or without a new body. Swinburne solves the problem of personal identity for this world and the next by appeal to immaterial souls.
There is a metaphysical problem with immaterialism: in virtue of what is a soul the same soul both before and after death? Perhaps the best answer is that souls are individuated by having a ‘thisness’ or haecceity. This is an intriguing suggestion that I cannot pursue here. An haecceity view, if otherwise satisfactory, may well be suitable as a metaphysics of resurrection – if it did not leave dangling the question of why resurrection should be bodily.
However, I believe that immaterialism should be rejected. My reason for rejecting immaterialism has less to do with resurrection than with the natural world. Immaterial souls just do not fit with what we know about the natural world. We human persons evolved by natural selection (even if God actualized this world on the basis of His foreknowledge of the outcome). Immaterial souls would simply stand out as surds in the natural world.
Someone may object: ‘If you dismiss immaterial souls on the grounds that they would be surds, then you should dismiss resurrection too. Resurrected persons would surely be surds if immaterial souls are.’ This objection can be met: My opposition to souls concerns their putative existence in the natural world. Resurrected persons, by contrast to immaterial souls, would not be surds in the natural world, because resurrection is not part of the natural order in the first place. Resurrection involves miracles, and miracles require God's specific intervention. We human persons – who, as I mentioned, evolved by natural selection – are part of the natural order, but immaterial souls are not. At least, I do not see how immaterial entities (unlike first-person perspectives, whose evolutionary roots can be seen in chimpanzees) could have evolved by natural selection.
Animalism
According to animalism, a human person is identical to a human animal. Therefore, animalists hold, a human person has the same persistence conditions as a human animal. If animalism is correct, then the story about Smith's having a biological body at one time and a distinct bionic body at another time is incoherent: on the animalist conception, no human person can have numerically distinct bodies at different times. I believe that this disqualifies animalism as part of a metaphysics of resurrection. Here's why.
If any sort of animalism is true, then a human person has her human body essentially. Her body changes cells, size, and shape, but the human person is nothing but that (changing) body. If her body went permanently out of existence, then that person would go permanently out of existence. Here is a simple argument to show that a biological body is not identical to a resurrection body.
Let h be your human biological body, the one that you have now. Let b be your spiritual body, the one that you have in the resurrection. Then:
(1) h is corruptible.
(2) b is incorruptible.
(3) Whatever is corruptible is essentially corruptible.
So,
(4) h≠b.
Both the second and third premises may seem open to challenge. Consider the second premise. Someone may hold that resurrection bodies are not really incorruptible; they remain corruptible, but God just prevents them from actual decay.Footnote 13 I have a couple of responses.
First, the suggestion that your resurrection body is the same body as your corruptible earthly body raises the well-known problems of reassembly of earthly bodies that, prior to resurrection, have burned to ashes or decayed or been eaten by animals.Footnote 14 I have been convinced by Peter van Inwagen that God could not restore a particular body by reassembling the particles formerly in the body.Footnote 15 And the other suggestions about how an earthly body could survive to be a resurrection body without reassembly (Dean Zimmerman's and Van Inwagen's)Footnote 16 seem to me much less plausible than the constitution view.
The next response to the claim that resurrection bodies are not incorruptible comes from Paul, who in I Corinthians 15, calls resurrection bodies ‘incorruptible’ or ‘imperishable’ or ‘spiritual’, depending on the translation. In The New English Bible, Paul says: ‘What I mean, my brothers, is this: flesh and blood can never possess the kingdom of God, and the perishable cannot possess immortality’ (I Corinthians 15.50). Although I am leery of proof-texts, Paul's words clearly suggest that resurrection bodies are not identical to earthly bodies – despite the tradition to the contrary. So, I stand by the second premise: resurrection bodies are incorruptible.
Now consider the third premise. You may think that God, in His omnipotence, could transform a corruptible body into an incorruptible body. I agree. But the transformation would be what Aristotle and Aquinas call a substantial change. The incorruptible body would not be identical to the corruptible body from whence it came. Why not? A corruptible body has different persistence conditions from an incorruptible body. A corruptible body would go completely out of existence under different circumstances from an incorruptible body. Since things have their persistence conditions essentially, a single body cannot change its persistence conditions; so, a single body cannot be corruptible at one time and incorruptible at another time.Footnote 17
To put it another way: earthly bodies are organisms, and organisms are essentially carbon-based. Anything that is carbon-based is corruptible. So, anything that is incorruptible is not carbon-based, and is not an organism, not a human biological body. Since resurrection bodies are incorruptible, they are not carbon-based and hence not identical to organisms, human biological bodies.
God could transform your human body into a resurrection body in the same way that He transformed Lot's wife into a pillar of salt. The pillar of salt, which is not organic, is not identical to Lot's wife's body, which is essentially organic. (Nor, of course, is the pillar of salt identical to Lot's wife.) Nothing that is a pillar of salt is identical to Lot's wife's body. Similarly, if God changed your human biological body into a resurrection body, the resurrection body would not be identical to your human biological body. So, if animalism (or Thomism, for that matter) is true, you would not exist in the resurrection.
If my argument here is correct, then no view of human persons (like animalism or Thomism) that construes a person's corruptible body to be essential to her is consistent with the doctrine of bodily resurrection.
Thomism
Thomism takes over Aristotle's notion of a human being as a substance for which the body supplies the matter and the soul supplies the form. According to Thomas, then, a human being is a composite of a rational soul (form) and a body (matter). The human being is a substance; the rational soul is not – it is a substantial form that nonetheless can ‘subsist’ on its own. Before the general resurrection, people who have died are in an ‘intermediate state’, during which the human being (the substance) does not exist. What continues through the intermediate state is the rational soul that subsists (disembodied) until reunited with the body, at which time the human being is recovered.
I think that there are two difficulties with Thomism, considered as a metaphysics of resurrection. The first is the same as with animalism: Thomas requires that a person's resurrection body be numerically identical to his or her earthly body. But (as we just reflected) resurrection bodies and earthly biological bodies have different persistence conditions, and are thus not numerically identical.
The second difficulty is how to individuate disembodied souls. In the case of immaterialism, we could appeal to haecceities, because according to immaterialism, the soul itself is a substance. But according to Thomas, the soul is not a substance. Disembodied souls are individuated by the bodies that they long for and desire to be reunited with. Smith's soul is the one that longs for and desires reunion with a certain body. But what makes a body (mere potency, the matter of which the soul is the form) the body that Smith's soul longs for? It can only be that Smith's soul longs for ‘it’. But since the body is mere potency, there is no ‘it’ for Smith's soul to long for. Hence, what makes a soul Smith's soul cannot be the body that it longs for. As Caroline Bynum says, ‘God can make the body of Peter out of the dust that was once the body of Paul’.Footnote 18 If this is the case, then disembodied souls cannot be individuated at a time by their yearning for certain bodies – because the identity of the body (Smith's, say) will depend upon the identity of the soul. It is difficult to see how Aquinas can combine the Aristotelian view that matter individuates with his view that the soul is a substantial form that can ‘subsist’ – and experience God – apart from a body.
Let me pause here and say that I realize that there is scriptural basis for the view that resurrection bodies will be identical to human biological bodies. There are puzzling metaphors in I Corinthians 15 and in II Corinthians 5, as well as the post-resurrection appearances of Jesus, in which he still seems to have his wounds. On the other hand, the fact that he can walk through locked doors and disappear into thin air may lead us to suppose that resurrection bodies are not identical to human biological bodies. But I don't think that such passages wear their meanings on the sleeve.
The memory criterion, the soul-as-software view, and the soul-as-information-bearing-pattern view
These may be considered together. The memory criterion is familiar from Locke (and his Scottish opponents). What I am calling the soul-as-software view takes seriously a computer metaphor: the soul is software to the hardware of the brain; if persons are identified with souls (software), they can be ‘re-embodied, perhaps in a quite different medium’, as D. M. Mackay put it.Footnote 19 Another materialistic view of the soul (this one from Polkinghorne) conceives of the soul as an ‘information-bearing pattern, carried at any instant by the matter of my animated body’.Footnote 20 At death, God will remember the patterns and ‘its instantiation will be recreated by him’ when at the resurrection.Footnote 21
These views share a widely recognized defect: the duplication problem. The problem is that two people (B and C, say) may both be psychologically continuous with (or run the same software, or exhibit the same information-bearing pattern as) a single earlier person, A. If B and C bear exactly the same relationship to A, and if B and C are distinct, then the relation that they both bear to A cannot be identity. A cannot be identical with two distinct objects, and it would be arbitrary to suppose that A is identical to one but not the other. Identity is a one–one relation, but person A's (quasi-)memories, software, information-bearing pattern, etc., could be transferred to more than one person. So, sameness of (quasi-)memories, software, or information-bearing pattern cannot suffice for sameness of person. To avoid this problem, defenders of the memory criterion and the like usually add the (ad hoc) requirement that there be no duplication.
However, there is a theological argument, suggested in conversation by my colleague Gareth B. Matthews, that supporters of the memory criterion etc. need not worry about duplication and need not appeal to ad hoc stipulations. I'll call the argument ‘the Matthews argument’. The premises of this argument are explicitly religious. They appeal to God's necessary attributes – viz. that God is essentially just – and to the notion of a judgment after death. If God is essentially just and God judges everyone, then it is metaphysically impossible for God to let a person A branch into persons B and C.
The reason that it would be metaphysically impossible for A to branch into B and C is this. Assume that everyone except Christ deserves punishment. God is essentially just and judges everyone. Suppose that person A branched into persons B and C: both B and C had A's (quasi-)memories (caused in the right way etc.). Whom does God punish? If God punished B but not C, or C but not B, then God would not be essentially just: B and C are related to A in exactly the same way; it is impossible to be just and to judge B and C differently. On the other hand, if God punished both B and C, then there would be twice the punishment that A deserved, and again God would not be essentially just. Either way, supposing that B and C both had A's (quasi-)memories (caused in the right way), violates God's essential justice in judgment. Since God is essentially just, if A deserves punishment, it is metaphysically impossible for B and C both to have A's (quasi-) memories. So, God's essential justice rules out the metaphysical possibility that A could have a duplicate in the afterlife.
The Matthews argument relies on weighty theological assumptions; but it does rescue the memory criterion from the duplication problem. And it works equally well to save the soul-as-software view and the soul-as-information-bearing-pattern view. So, if the memory criterion (or the soul-as-software view, or the soul-as-information-bearing-pattern view) could be developed in ways that avoid other problems (besides the duplication problem), any of them would be suitable candidates for a metaphysics of resurrection.
The constitution view
Now let me turn to the constitution view, according to which sameness of pre- and post-mortem person is sameness of first-person perspective. In the first place, the constitution view avoids some of the pitfalls of the other candidates for a metaphysics of resurrection. Since human persons are essentially embodied, the constitution view avoids the problem of individuating disembodied souls – a problem that afflicts Thomism. Since a person's identity depends on her first-person perspective, the constitution view avoids the problem of the numerical identity of corruptible and incorruptible bodies – a problem that afflicts both animalism and Thomism.
Still, the constitution view is not home free. What is needed is a criterion for sameness of first-person perspective over time. In virtue of what does a resurrected person have the same first-person perspective as a certain earthly person who was born in, say, 1800? In my opinion, there is no informative non-circular answer to the question: ‘In virtue of what do person P1 at t 1 and person P2 at t 2 have the same first-person perspective over time?’ It is just a primitive, unanalysable fact that some future person is I; but there is a fact of the matter nonetheless.
We can see this by means of an argument from providence. Now, according to the traditional doctrine of providence, God has two kinds of knowledge – free knowledge and natural knowledge. God's free knowledge is knowledge of contingent truths, and His natural knowledge is knowledge of logical and metaphysical necessities. (I'm disregarding the possibility of middle knowledge here.) Again, according to the traditional doctrine of providence, the obtaining of any contingent state of affairs depends on God's free decree. Whether the person with resurrected body 1, or body 2, or some other body, is Smith is a contingent state of affairs. Therefore, which if any of these states of affairs obtains depends on God's free decree. No immaterial soul is needed for there to be a fact of the matter as to whether Smith is the person with resurrected body 1. All that is needed is God's free decree that brings about one contingent state of affairs rather than another. If God decrees that the person with body 1 have Smith's first-person perspective, then Smith is the person with body 1.Footnote 22 So, there is a fact of the matter as to which, if any, of the persons in the resurrection is Smith, even if we creatures cannot know it. On the Christian idea of providence, it is well within God's power to bring it about that a certain resurrected person is identical to Smith.Footnote 23
Notice that the argument from providence provides for the metaphysical impossibility of Smith's being identical to both the person with body 1 and the person with body 2 in the resurrection. For it is part of God's natural knowledge that it is metaphysically impossible for one person to be identical to two persons. And according to the notion of God's natural knowledge, what is metaphysically impossible is not within God's power to bring about. Hence, there is no threat from the duplication problem. Indeed, this argument from providence may be used to support, not only the constitution view, but also immaterialism, the soul-as-software view, the soul-as-information-bearing-pattern view, and the memory criterion, to guarantee a fact of the matter about which person is you in the resurrection. The only views of persons that receive no aid from the argument from providence are those (like animalism and Thomism) that require that incorruptible resurrection bodies be identical to corruptible biological bodies.
The relative merits of the constitution view
The constitution view can deliver the benefits of immaterialism and Thomism without having to postulate immaterial souls, which would be surds in the natural world. In light of the Matthews argument, the memory criterion, the soul-as-software view, and the soul-as-information-bearing-pattern view may be saved from the duplication problem, but none of these is really a fully developed metaphysical theory. The constitution view of persons is superior in that it is integrated into a comprehensive unified view of the natural world.
But the real advantage of the constitution view, at least for Christians, is over animalism. In contrast to animalism, the constitution view does not take being a person to be just a contingent and temporary property of beings that are fundamentally non-personal (organisms). On animalism, being a person has no ontological significance at all.
Indeed, on the animalist view, our having first-person perspectives (or any mental states at all) is irrelevant to the kind of being that we are. But the Christian story cannot get off the ground without presuppositions about first-person perspectives. On the human side, without first-person perspectives, there would be no sinners and no penitents. Since a person's repentance requires that she realize that she herself has offended, nothing lacking a first-person perspective could possibly repent. On the divine side, Christ's atonement required that Christ suffer, and an important aspect of his suffering was his anticipation of his death (e.g. the agony in the Garden of Gethsemane); and his anticipation of his death would have been impossible without a first-person perspective. This part of Christ's mission specifically required a first-person perspective. What is important about us (and Christ) according to the Christian story is that we have first-person perspectives.
Also, of course, there is Genesis 2.26, according to which God said: ‘Let us make man in our image, after our likeness’. A natural reading of this verse is that we were made to be persons, to be capable of reflective thought about ourselves – in short, to have first-person perspectives. On the animalist view, our first-person perspectives are just contingent features of us. On the constitution view, they essential to us.
Given how important the first-person perspective is to the Christian story, Christians have good reason to take our having first-person perspectives to be central to the kind of being that we are. Hence, Christians have good reason to endorse the constitution view.Footnote 24
