Existence Is Evidence of Immortality

Here, I prove the reality of reincarnation.* This is my favorite argument that I’ve ever thought of.

[ *Based on: “Existence Is Evidence of Immortality,” Nous 55 (2021): 128-51, DOI: 10.1111/nous.12295. ]

1. Overview

Claim: Every person has been incarnated infinitely many times in the past, and will be reincarnated (live another lifetime) infinitely many times in the future.

Summary argument: Time is infinite in both the past and the future directions. Given infinite time, whatever conditions (including disjunctive and conjunctive conditions) caused you to be born will repeat, to any desired degree of approximation, infinitely many times. A sufficiently close approximation to the conditions that brought about you (i.e., your current incarnation) will count as literally bringing you to life again. There are some theories of persons that would reject this, but these theories also imply that your present existence has probability zero. Since you are presently here, those theories are wrong.

2. The Infinitude of Time

Everyone seems to agree that the future is infinite. Why think the past is infinite?

2.1. Intuitive considerations

If time didn’t exist, nothing could bring it into existence, since causation requires time. Nor could it come into existence by chance or uncaused, because for something to come into existence means this: first (that is, at one time) the thing does not exist, then (at a later time) it exists. But there could not be a time when time did not exist. So time could not come into existence. So it must have existed forever.

2.2. Philosophical objections

Some people say that the past can’t be infinite because one cannot have an “actual infinity” or a “completed infinity”, and if the past were to be infinite, then the series of all past events would be an actual, completed infinity.

Reply: One can have an actual, completed infinity. When you drop a ball, it goes half the distance, then half the remaining distance, etc. This is an infinite series. Yet it gets completed when the ball hits the ground.

2.3. Modern cosmology

Some cosmologists think that time itself began with the Big Bang, so literally nothing happened before it. The first thing that ever happened was that there was a huge concentration of energy in a tiny region, moving outward, for no reason.

If you buy this as a plausible theory, I have a better theory for you: The universe began in 1950, already in the state that it was in then, and then things just evolved forward from there. There’s no reason why it was in that state.

My point: If the Big Bang theory with the beginning of time is reasonable, then so is the 1950 theory. But the 1950 theory is completely unreasonable. So the Big Bang theory is also unreasonable.

3. Infinitude Yields Eternal Recurrence

Subject to certain reasonable assumptions, if you have infinite time, any state of the universe must approximately repeat, to any desired degree of approximation (i.e., the universe will come arbitrarily close to repeating any given state). Roughly speaking, the assumptions are that there is a fixed amount of mass/energy in the universe, and there is a limited range to the ways it can be configured.

Aside: you do not ever have to exactly repeat a state, because there are infinitely many (continuum many) possible states of the universe, so you can keep going through distinct states for an infinite time.

Objection: The range of configurations could be unlimited, because the universe may continue expanding outward forever.

Reply: This isn’t compatible with the infinitude of the past. More generally, given the two-way infinitude of time, there cannot be any overall direction in which the universe is headed across time (e.g., larger size, higher entropy). It has to be cyclical.

Objection: That assumes that there is only one universe. We might be part of an infinite and constantly-expanding multiverse. Each universe within the multiverse moves in one direction (e.g., toward higher entropy, until thermal equilibrium), but new daughter universes periodically spawn (e.g., starting in low-entropy states).

Reply: You’re right, I haven’t ruled that out. However, that will be okay for my argument, which only requires that certain qualitative states recur. They need not recur “in the same universe”. In the multiverse theory, the state that produced the current incarnation of you should recur in other universes.

4. Recurrence Yields Reincarnation

We can distinguish two kinds of views about persons and personal identity:

Permissive views: These are views that would allow the same person to live multiple lives (they permit reincarnation), with nonzero probability. For instance, suppose that there is a certain brain configuration such that, whenever a brain with that configuration comes into existence, the person with that brain counts as you. Or suppose there is an immaterial soul that, after leaving a body, is capable of going into another body. These are both permissive views.

Restrictive views: These are views that exclude reincarnation. For instance, the view that personal identity depends on spatiotemporal continuity, or the view that the identity of a person depends on the particular token cause of their conception.

I take it that both kinds of view have a nonzero initial probability.

On a restrictive view of persons, you could only be born once. Now, whatever conditions were required for you to be born the first time, those conditions have either probability zero or a nonzero probability. If they have probability zero, then you shouldn’t be here. If they have a nonzero probability, then they should have occurred before (like, in a previous century), given that there has been infinite past time. On the restrictive view, however, that would prevent you from being here now. So again you should not be here now. I.e., the probability that you would be here now is zero.

On the other hand, on a permissive view of persons, there is a nonzero probability that you would be here now. Since you are in fact here now, that is evidence that a permissive view of persons is correct.

Bayesian calculation:

Let H = the hypothesis that persons can live more than once.

E = the evidence that you are here now.

From Bayes’ Theorem:

P(H|E) = P(H)*P(E|H) / [P(H)*P(E|H) + P(~H)*P(E|~H)]

As discussed above, P(E|~H) = 0. So:

P(H|E) = P(H)*P(E|H) / [P(H)*P(E|H) + P(~H)*0]

= P(H)*P(E|H) / [P(H)*P(E|H)]

= 1,

provided P(H) and P(E|H) are nonzero.

So the probability of a reincarnation-friendly theory of persons being correct, given that you’re here now, is 100%.

5. Questions & Objections

5.1. The heat death of the universe

According to thermodynamics, the universe is destined to increase in entropy until it eventually reaches thermal equilibrium (“heat death”), after which it will remain in that state. This is a state in which life, and indeed anything at all interesting, is no longer possible. So there will be no way to be reincarnated after that.

Reply: The 2nd Law of Thermodynamics is only probabilistic. It is merely highly improbable (on a human timescale) for the entropy of any macroscopic system to spontaneously decrease, not impossible. Given infinite time, arbitrarily large spontaneous entropy decreases will occur.

Anyway, given that the universe at the time of the Big Bang was in an extremely low-entropy state, and given that (as argued above) the past is infinite, there may be some presently-unknown process that systematically decreases the entropy of the universe.

5.2. The physicalist objection

Suppose you believe in physicalism (everything in the world, including humans, is purely physical). That seems to rule out the possibility of reincarnation.

Reply: Not so. Maybe you are a gappy spacetime worm. In other words, there is a human organism, O1, that is alive now. In some future century, there will be another organism, O2, that will be physically very similar to O1. Then another one, O3, still later. And so on. A physicalist could identify you with the fusion of these organisms (the thing that has O1, O2, O3, and so on as temporal parts and no other parts that don’t overlap with those). So reincarnation is still conceptually possible on a physicalist view.

5.3. Recurrence of material objects

If the argument works to show that persons are reincarnated, what about inanimate objects? Can a parallel argument show that the table I’m now using will exist again at some future time after its current incarnation is destroyed? If so, you might consider this an implausible implication of my theory.

E.g., suppose we argue: this table is here now. If tables could only exist once, the probability that this table would be here now would be zero. So tables can be “reincarnated”.

Reply: It’s not clear to me that this would be a bad consequence, if tables could be “reborn”. But in any case, there are two important differences between tables and people:

a. The identity conditions of tables, plausibly, are purely conventional. So if you dismantle a table, then later reassemble a table from the same parts, swapping the locations of two of the legs, there is no objective fact about whether the result is “the same table” as the original. That is a semantic question. As such, we can just stipulate that tables don’t recur after the first time they are destroyed, in which case that proposition gets probability 1.

By contrast, the identity conditions of persons, I take it, are non-conventional. If a person dies, then later his body gets reanimated, there is a fact (though perhaps unknown) as to whether the reanimated person is “the same person” as the original. It is not a semantic question. (Consider that if you thought this sort of question was merely semantic, you could devise a plan for making everyone immortal just by changing some linguistic conventions.)

b. I claim that your basic evidence contains direct references to two particulars: yourself (that is, your individual mind) and the present time. It does not contain direct reference to a particular table. (Argument: If the particular table were replaced with a qualitatively indistinguishable table, you would have the same evidence that you currently have.)

Thus, hypotheses about tables repeating or not repeating make no differential predictions about your basic evidence. Your basic evidence would be the same regardless of whether table T repeats or only gets succeeded by a qualitatively indistinguishable table.

5.4. Space & time

This isn’t in the original paper, but every time I present this argument, someone in the audience will ask, “Can you give an analogous argument, but substituting space for time? Just as you argued that given the infinitude of time, you must live at other times, can you also argue that given the infinitude of space, you must now be living in different places?”

Reply: There are two important differences between place and time.

a. First, one is not directly aware of any place in the sense that one is directly aware of the present time. This is why, e.g., you can’t tell whether you’re moving through space (along with all the objects in your environment), nor which direction you might be moving. All you really observe is the material objects that are positioned in space. If you (and all the objects presently around you) were in a different place, you would have the same evidence you currently have.

b. Anyway, you have direct introspective access to all of your present conscious experiences, regardless of where in space (if anywhere) they are occurring. If you were now also living other lives in other places, you should be experiencing whatever is happening in those other places. But you know that you aren’t; you’re only experiencing one stream of experiences appropriate to one location.

By contrast, you do not have direct access to all of your conscious experiences across time. If you had (or will have) conscious experiences at other times, you would not introspectively detect that.

---

TL;DR: The past & future are infinite. Given this, if persons could only live once, then the probability that you would be alive now would be zero. But you are alive now. So persons can live more than once.