Death and the Afterlife

Two quite different claims travel under the heading of life after death, and treating them as one doctrine is the first mistake available.

One says that a person has a part which is not the body and which continues when the body stops. The other says that persons are not divided that way, that death is the end of them, and that they are brought back. These are not two descriptions of the same hope. They rest on incompatible views of what a person is, and the traditions that hold them have been in tension about it for a very long time.

It is also worth noticing at the start that the question is separable from the rest of this subject in both directions. Survival without God is a coherent position and has been held. God without survival is also coherent, and has been held by fewer people than one might expect.

The soul that continues

Plato's Phaedo is the founding text, and the argument that gave the view its long life is about simplicity. Composite things fall apart because they have parts to come apart. The soul, being what does the thinking, is not composite in that way, so there is nothing in it to disintegrate. Death is a separation of two things that were joined, and only one of them was capable of ending.

Descartes gave the modern version, on which mind and body are different substances, each of which can be conceived without the other. The argument is different but the payoff is the same: the mind's continuing does not require the body.

The standard objection is the one from dependence. Every mental capacity we can examine can be disrupted by damage to the brain: memory, personality, judgement, the capacity for language, the emotional responses that make someone recognisable. That is not a proof that thinking requires a brain, since a dependence in this life is consistent with independence after it, and the dualist has answers about the instrument being required for the performance without being the performer. But the pattern is very general and it points the other way, and the argument from simplicity does nothing to explain it.

The simplicity argument also has a structural difficulty. If it works, it proves too much and too little at once. Too much, because whatever it establishes about the soul's indestructibility applies as much to an animal's soul as to a human's, which few of its users want. Too little, because surviving is not the same as surviving as someone: an indestructible simple thing that retains no memories, no character, and no continuity of outlook has not delivered what anybody wanted from the doctrine.

The person who is raised

This is the doctrine of the New Testament rather than of the Academy, set out at length by Paul in the fifteenth chapter of the first letter to the Corinthians, where the body that is raised is insisted on and is also said to be transformed rather than merely reanimated.

It differs from the previous view at every point that matters. There is no part of the person that escapes death. There is a genuine interval during which the person does not exist. What comes back is a person with a body, not a mind that has shed one. And the return is an act of God rather than a natural property of anything.

The two pictures were fused in later Christian thought, producing the familiar composite in which a soul separates at death and is later reunited with a body. That composite inherits the problems of both and is not obviously the doctrine either source was teaching.

For the philosophy of religion the important point is that resurrection makes the survival of persons depend entirely on divine action. Nothing in the person's nature secures it. That is a theological advantage, since it removes any suggestion that God is obliged by the metaphysics, and it is a philosophical liability, since it puts the whole weight on a question about identity that the next section takes up.

Whether it would still be you

Three difficulties come up in order, and each is harder than the last.

Reassembly is the crudest. If a person is their body and the body has decayed, dispersed, and had its material incorporated into other things, then reconstituting it from the original matter is not available and reconstituting it from other matter looks like building a copy. Nothing about a collection of atoms makes it the original collection once it has been scattered and reused.

The gap is worse. On the resurrection view the person does not exist for some interval. Ordinary identity over time runs through a continuous existence, and it is not obvious that anything can survive an interruption. A thing that stops and a thing that starts are two things, unless something connects them, and the connection is what is missing.

Duplication is worst, because it defeats the natural repair. Suppose the answer is that God creates a being with the right memories, character, and apparent continuity. God could do that twice. If both would be the original person, one person is now two, which is not how identity works. If neither would be, then the single case was not identity either, since nothing about the single case is different from what each of the two would have. And if only one would be, something other than the memories and character must be doing the work, and the account has not said what.

Peter van Inwagen faced this squarely in "The Possibility of Resurrection", International Journal for Philosophy of Religion 9, no. 2 (1978), pp. 114-121. His proposal is that God removes the body at the moment of death and substitutes a simulacrum, which is what is buried or burned, so that the person's actual body is preserved and continuity is never broken. He offered it as a demonstration that resurrection is possible rather than as an account of what happens, which is the defence structure from the theodicies article used in a new place. Even sympathetic readers regard it as an uncomfortable proposal, and its point is precisely that the uncomfortable proposal is what it takes to secure continuity on a materialist view of persons.

Aquinas answered differently, treating the soul as the form of the body rather than as a second thing joined to it. The same form informing matter again yields the same person, and the identity is carried by the form rather than by the matter. This is a genuine third option and it avoids both the reassembly problem and the crude dualism. Its difficulty is what happens in the interval: a form without matter is not a human being on this account, so either the person does not exist between death and resurrection, which reintroduces the gap, or the separated soul is doing more than the theory allows it to do.

The identity problem is not special to religion. It arises for teleportation, for gradual replacement of neurons, and for any case where continuity is interrupted or duplicated, and the subject of personal identity treats it in its own right. What is specific here is that a doctrine of survival needs an answer and cannot borrow one from the case where nothing surprising happens.

Return rather than survival

The Indian traditions supply the developed versions, and the difference between two of them is philosophically the interesting part.

Hindu accounts typically posit a persisting self that passes through successive lives, in which case the identity problem is answered in the same way the immortality view answers it, by having something that never stops.

Buddhist accounts deny that there is any such persisting self to be reborn, which makes what continues a causally connected series rather than a thing. That is a much more austere position and it is closer to some contemporary views in the philosophy of mind than either of the Western options here. It also gives up what most people wanted, since a successor causally derived from me is not me, and the tradition treats that consequence as a correction of the wanting rather than a cost.

The standard objection is memory. If nothing carries over that the person can access, the claim that the later life is the same person has nothing to fasten on, and the Buddhist version accepts the point by denying the claim. Where memories are reported, the reports face the same evidential questions the previous article raised about testimony, with the additional difficulty that details of a past life are exactly the sort of thing that can be acquired in ordinary ways and misattributed.

Reincarnation also inherits a version of the distribution problem from the soul-making article, and answers it in a way that should be noticed. On a karmic account the suffering in a life is explained by conduct in earlier ones, which supplies exactly the reason theodicy struggles to supply. The cost is a commitment that every present misfortune is deserved by its bearer, which many people find harder to accept than the problem it solves.

Hell

This is the problem of evil in its sharpest available form, and the previous articles' responses do not transfer. The suffering is unending, so appeals to a greater good later have nowhere to point. It is imposed rather than permitted. And it falls on people whose fate is settled, so no soul-making can be in progress.

Thomas Talbott sets the problem out as three claims that cannot all be true: that God's love extends equally to all, that God's love will finally prevail, and that some are separated from God forever. Each of the three main Christian positions is produced by rejecting one.

The proportionality objection is the most pressing. Finite beings, acting in ignorance and under illusion, in lives of limited length, do things whose gravity is bounded. Unending punishment is not bounded. Anselm's reply is that the gravity of an offence depends on whom it is against, and an offence against an infinite being is accordingly infinite. The counter is that this makes culpability independent of the offender's understanding, which contradicts how responsibility is assessed everywhere else, and that the traditional doctrine additionally has to explain how anyone could deserve this for a defect they inherited, when inherited defects normally reduce culpability rather than raise it.

Three responses divide the field.

The Augustinian view accepts that some are separated forever and denies that God's saving action extends equally to all. It preserves the tradition at the cost of the claim about universal love, and it has to say that God could have saved everyone and did not.

The free-will response, developed by Jerry Walls in Hell: The Logic of Damnation (1992) and in different form by Richard Swinburne, keeps universal love and holds that some freely and finally refuse it. Hell is then chosen rather than inflicted, and God's respecting the choice is a form of respect rather than of cruelty. The difficulty is making the choice coherent. It has to be free, fully informed, and irrevocable, and a fully informed rejection of one's own good is very close to a description of someone who is not competent to choose, in which case the choice does not carry the weight put on it.

Universalism keeps both claims about love and denies that anyone is separated forever, holding that God's purpose eventually reaches everyone. It answers the problem completely, and this is the view Hick's theodicy was driven to for independent reasons. Its costs are exegetical rather than philosophical for the most part, and it faces the free-will response's own question in reverse: if refusal is possible, what guarantees it will not last.

This is the fourth option and it is often left out of the list. It fits the resurrection picture more naturally than the immortality picture, since on the resurrection view nothing about a person continues by nature and there is nothing to be preserved in torment unless God preserves it.

It answers proportionality directly. Ceasing to exist is a finite loss, terrible but bounded, and no infinite penalty is imposed for a finite life. It also removes the worst version of the problem of evil that the traditional doctrine generates.

Its cost is precise, and it is the reason the position does less work than it appears to. Any response to the problem of evil that relies on things being put right afterwards is unavailable for those who are annihilated. Their suffering in this life is not compensated, not redeemed, and not made part of any larger good, because there is no longer anyone for it to be good for. The doctrine buys its answer about punishment by giving up the answer about suffering.

Next

The previous objection raised a question and left it open. If a belief's having a comforting or biological explanation does not show it false, what would follow from such an explanation, and is there any version of the inference that works? The last article of the tier takes up the natural explanations of religious belief and asks exactly that.