Tuesday, June 9, 2009

Was I once a fetus?

Let x be the fetus in my past that grew into me. Here is a valid Aristotelian argument (though Aristotle himself would probably deny (4)).

  1. (Premise) The identity of a bodily organ depends on the identity of the individual whose organ it is, so that if A is c's unshared heart (sharing occurs in the case of Siamese twins), and B is d's unshared heart, and c and d are distinct individuals, then A and B are distinct organs.
  2. (Premise) x has exactly one heart, hx, and it is unshared.
  3. (Premise) I have exactly one heart, hI, and it is unshared.
  4. (Premise) hx=hI.
  5. Therefore, I am x. (By 1-4)
The controversial premises are (1) and (4).


Anonymous said...

What do you think are good reasons for accepting (4)?

David Gawthorne said...

I would target premises 2 and 3, first. Why can't there be heart sharing across time as well as space?

I also think that 1 is vacuous - equivalent to, 'The identity of a bodily organ depends on the identity whose organ it is, except when it isn't.'

Delete the word "unshared" from premise 1 to see what I mean. 'Unshared' is doing all the work.

Alexander R Pruss said...


It's a standard part of Aristotelianism that an organ's identity depends on the identity of the organism. If we allowed diachronic sharing, we'd lose this principle. So in (1) and (2), "unshared" means "not synchronically shared".

To take care of the case of Siamese twins, we need to figure out what should happen then on an Aristotelian view. I think the simplest solutions are (a) to deny that the organ is genuinely shared--it objectively belongs to only one organism, but we can't tell which; or (b) to insist that the organ depends on both organisms--as soon as it ceases to be a part of either one or both, it ceases to exist.


Well, the obviousness of the truth of (5) gives one a good reason to accept (4). But I suppose you were looking for a good reason that is not question-begging in the context of this argument. :-) Well, I suppose one may have the intuition that the fetal heart kept on fulfilling the same function throughout gestation and childhood, and didn't suffer any cataclysmic event, such as a transplant, that would make it cease to exist.

David Gawthorne said...

I don't see how you lose the principle if you allow diachronic sharing. If you can have synchronic siamese twins then why can't you have diachronic siamese twins? What about a heart transplant? Is the heart destroyed when it is implanted in the receipient? This seems counter-intuitive to me.

Alexander R Pruss said...

The Aristotelian view is that the heart is at least in part defined by its telos. Its telos is to pump blood throughout a particular body. When the telos changes, the old heart ceases to exist. When the heart is transplaneted, the telos changes, since the particular body changes.

Andrew said...

Didn't Elizabeth Anscombe have a similar argument? I like the argument, however I agree that (4) is rather controversial... One could accept (4) on the grounds of some kind of 'slippery' slope' reasoning. The heart of this minute is identical to the heart of the previous minute which was identical to the heart of 2 minutes ago, etc.

David Gawthorne said...

Okay, good.

You seem to be biting the bullet in saying that a heart is destroyed during transplant, but maybe you don't share my transplant intuition at all.

This responds to my charge that premise 1 is vacuous in allowing for sharing because you distinguish between synchronic and diachronic sharing, and reject the latter. Then I would say that the distinction between synchronic and diachronic sharing is ad hoc. What reason do you have to differentiate the outcomes of each?

I think that you should be forced to reject all sharing, but then I think that siamese twins are a problem for you.

Alexander R Pruss said...

Here's a reason to reject diachronic sharing. On the Aristotelian view, the identity of a part depends on the identity of the whole. In the case where A is an organ shared by organisms x and y, this seems to imply that the identity of A depends on the identity of x and the identity of y. Consequently, A couldn't exist without x existing and having A as a part, and A couldn't exist without y existing and having A as a part. Suppose you buy this.

Then, consider a case of diachronic sharing. At t0, A is had by x, and at t1, A is had by y. Suppose t1 > t0. Then by the dependency claim, A depends for its identity on x as well as on y. In particular, if y were never to exist or never to have A as a part, A couldn't exist. But this means that whether A exists at t0 depends on what is going to happen in the future, at t1. This is strange--it's a backwards temporal dependence.

Now, backwards temporal dependence is not logically absurd, I think. But it is surely extremely rare and difficult to arrange, and probably beyond our powers. However, organ transplants are not beyond our powers, and are not extremely rare. It would be odd indeed if every case of organ transplant involve backwards temporal dependence.

This shows that diachronic sharing is more problematic than synchronic sharing.

David Gawthorne said...

I don't think that the difference in dependence amounts to much if you take a four-dimensionalist view of time. Temporal parts are just parts temporally related on that view.

However, if you are prepared to reject four-dimensionalism then I will be silenced.

Alexander R Pruss said...

I think the dependence issue is still there given four-dimensionalism. Typical four-dimensionalists do not want to say that there is backwards causation or even this kind of backwards dependence going on in more or less everyday cases. Consider the following weirdness. Suppose that A is Jack's heart at t0, and Fred's heart at t2, due to a transplant around t1. Then on the view under consideration the transplant surgeon around t1 had the power of bringing it about that A never existed. (Because if the surgeon did not do the surgery, then the heart in Jack would not end up in Fred; but the identity of a part depends on the wholes its in; so a heart that does not end up in Fred is numerically different from a heart that does end up in Fred.) And that seems odd.

In fact, I don't see why it makes a difference here whether one is a three- or four-dimensionalist. The question whether three- or four-dimensionalism holds does not seem to force any particular answer to the question whether backwards dependence is possible, and if possible, common.

David Gawthorne said...

Surely the relevant kind of dependence is mereological and not causal when we are talking about the identity of parts of wholes. Four-dimensionalists usually see persons at a time as parts of temporal wholes. I don't see the issue of backwards causation as being relevant, here

Further, that A might have been numerically distinct is weird, but only because the thesis that the identity of A is determined by the whole it is part of is weird. If you believe that the same heart can be transplanted then there is no problem and that is why I find your thesis counterintuitive.

David said...

Would someone who accepted (1) also accept (4) if he didn't already think that (5) is true? The argument given is valid, but I'm not sure that it would have much value in persuading someone who didn't think he was once a fetus to change his opinion.

David said...

On second thought, the argument might have persuasive force for someone who accepted (1)and wasn't sure about (5), or hadn't thought about it, if (4) strikes him as true. He would then have reason to accept (5).