Are thought experiments a legitimate philosophical method?
Answer by Geoffrey Klempner
I wonder, Andrew, whether you are responding to the previous question from Val, and the answer given by Hubertus Fremerey?
The question was about the statement, “If X might exist but we have no reason to suppose that it actually does exist, then as metaphysicians we should not concern ourselves with X.” I can’t fault Fremerey’s answer but there is an issue which he did not touch, concerning hypothetical questions about contingent non-existents, popularly known as ‘thought experiments’ — or in Daniel Dennett’s words ‘intuition pumps’.
Here’s a thought experiment that has troubled me over the years. Suppose that there existed (as there does not, so far as you or I are aware) a fission ray gun. You point the gun at someone and when you press the trigger, the fission ray causes them to divide like an amoeba, resulting in two identical versions of the person you just shot.
Let’s make this personal. We’ll call the two resulting persons Andy1 and Andy2. Andy1 remembers posting your question on ‘Ask a Philosopher’ and so does Andy2. In every aspect, the mental and physical states of Andy1 and Andy2 are identical, save for the fact that they occupy different positions in physical space, and will from this point onward diverge in their mental and physical properties.
This is a question for you: if you saw the fission ray gun demonstrated and it apparently worked, how would you expect things to turn out when the gun was pointed at you?
‘If all the world was apple pie/ And all the sea was ink/ And all the trees were bread and cheese/ What should we have to drink?’ Why pose questions about things that could not possibly happen? Unless, of course, the simulation hypothesis is true, in which case there could be all manners of violations of the laws of nature, such as vampires, werewolves etc.
So, back to the question. I’ll put myself in your place. Someone would find himself on the left and someone would find himself on the right. We’re ruling out telepathy, so they can’t both be I, can they? Logic says there are three alternatives: both are ‘I’, neither is ‘I’, one is ‘I’ and the other is ‘not-I’. If you want to be exhaustive you can include a fourth alternative, ‘The question makes no sense.’ It’s tempting to go with that last possibility, but I don’t think we have the right to let ourselves off the hook so easily.
You have to do work, ‘real philosophical work’ if I can put it like that, to justify your choice out of the three remaining options. My view would be that the very idea that the ‘I’ is something that persists through time is an illusion. As I have stated before, the existence of I in the world is a deep mystery: ‘I might have not existed but someone exactly like me might have existed in my place.’ But I am talking about I-now, not an I that persists through time.
To ‘prove’ the point here are two more thought experiments. The universe was created five seconds ago, with most of ‘my’ answer already written. Can I still claim authorship? Or there’s a ring on the doorbell and the postman hands me a video showing how the ‘original’ GK was kidnapped during the night and how a perfect physical copy got out of my bed this morning and enjoyed my favourite breakfast meal of peanut butter on toast with coffee and orange juice.
The notion that ‘I-now’ and ‘the world’ are two separate existences is a theory. I have been prompted to make that theory by considering a thought experiment. My theory could be wrong, maybe the thought I have just expressed about my contingent non-existence, everything else remaining the same, makes no sense. Dennett is right to call this kind of thing a mere ‘intuition’. But, ultimately, our notion or sense of what ‘makes sense’ not has to be based on how things seem to us, on our intuitions, once all the logical moves have been gone through. There’s simply nothing else to go on.