Cap'n Arbyte's
Main page

Zany stuff
Blog archives




Brains, Souls, and Thought Experiments

Thought experiments can be dangerous. They can often illuminate interesting issues but they can also distract and confuse and be absolutely irrelevant. Thought experiments have limits and it's important to recognize when they're not helpful.

I'd like to talk about two thought experiments. The first I heard from a philosophy professor in college, and the second from a recent post of Steven den Beste's.

I don't remember most of the details of that day's discussion in the philosophy class, so I'm not sure exactly how this thought experiment came up, but it had something to do with the subjective experience of perception and whether different people experienced perceptions in the same way. Anyway, I was asked to consider the thought experiment of waking up one day to discover that the colors red and green had been switched.

I said it was absurd and that I couldn't consider it — the thought experiment was in irreconcilable conflict with facts I knew about the real world. The professor offered two implementations to help me along: That aliens from outer space (those sneaky aliens…!) visited while I was sleeping one night and either painted red everything that was green and green everything that was red, or that they had implanted a device in my eyes to accomplish the same effect. The point was to create a scenario where I would have to accept the change in my experience of color, unable to prove that anything had "happened". This is a poor exposition of the matter, but it was years ago and I don't remember more details that would clarify it. Just play along. The reasons we were talking about the issue aren't relevant.

I dismissed the paint possibility immediately by noting that the first time I cut myself, my blood would be red — the same color as the (painted) tree leaves. But blood and tree leaves aren't the same color, and I know it, so the painting would be exposed. Besides, paint would flake off many surfaces, and I might accidentally spill some paint thinner in the yard. Or the grass would grow and would be green by the roots. Or the paint would kill all the plants. In any case, it would be too easy to detect.

I rejected the eye-implant possibility by explaining that it would also quickly be detected. The physical properties of light are not malleable, and if I looked at a red laser and saw green, I'd know something had changed. I could easily re-measure the wavelength of the laser light so I would know that my perception and not the real world had changed. A person more familiar with optics could no doubt take advantage of the properties of chromatic response in the human eye to figure out the truth, too. They could probably even distinguish which side of the retina the change was on — they could distinguish a neurological change from a device implanted in the orb.

The point is that it is possible to distinguish a change in the real world (paint) from a change in our perception of it (implant) because I know facts about the world that allow me to uncover the truth. The thought experiment does not work because any implementation my professor offered — and I insisted on one; I refuse to consider the totally arbitrary — would lead me to discover some incongruity with preestablished facts.

My point in being so argumentative that day was to show that reality is an interconnected whole and you cannot make thought experiments willy-nilly. Even a supposed "small" change like switching the colors red and green is not plausible because a cursory effort to integrate the new condition with the rest of one's knowledge exposes the difficulties.

"Is the fact that water ice floats on water liquid a contingent or necessary truth?" Aaaagh! For ice to sink would require changes in the laws of physics that would render the universe totally unrecognizable. You can't just have sinking ice, or swap two colors, and leave everything else the same. That's not an option, physics forbids it, and it's clear as soon as you try to integrate sinking ice with the rest of your knowledge. I can not and will not consider an alternate universe that bears no resemblance to the real one. And that's why I reject the distinction between "necessary" and "contingent" truths. In a very real sense I cannot conceive of sinking ice, not in the real world, because it would change everything. I could conceive of it in a comic book, and I fear that's how people who accept the necessary/contingent distinction view the world — as a comic book, not as an integrated whole.

Steven's thought experiment was about the theological implications of brain transplants (I kid you not):

One can propose thought experiments which such believers might find troubling. For instance, suppose that a mad scientist kidnaps a pair of identical twins and performs a mutual brain cross-transplant, moving the brain of each into the body of the other. Then assume that one of them is hit by a bus and dies. Presumably someone's soul went to heaven, but whose soul was it? Was it the soul which was originally associated with the body, or the soul originally associated with the brain? A different way to ask the question is this: During the mad scientist's operation, when the brains were moved, did the souls move with them? What part of the body is the soul actually hooked to?

He also considers the variation of swapping memories instead of tissue:

Suppose that instead of swapping the brains, the mad scientist copied and somehow stored Alan's memories, copied and stored Bob's memories, wiped clean all the memories in both brains, and then copied Alan's memories into Bob's brain and copied Bob's memories into Alan's brain. The only thing moved would be information; no tissue would be transplanted.

… and copying memories without erasing the ones already present:

What our mad scientist does this time is to copy and store Alan's memories, copy and store Bob's memories, write Alan's memories into Bob, and write Bob's memories into Alan. But the mad scientist does not wipe the previous memories in either brain before putting the new memories in. When they awaken, Alan and Bob will both be deeply confused. Alan will remembers being Alan but will also remember being Bob. And he won't be able to tell which set of memories was original and which was written in.

These thought experiments trouble me. They don't evoke the same level of disdain as the color-swapping or ice-sinking examples I already discussed, but I think that's primarily because I'm significantly more ignorant of neuroscience than I am of physics. (I'm nursing the suspicion that it's primarily ignorance that makes thought experiments seem plausible.)

In the brain transplant case I'm struck by problems at the interfaces between tissue that is transplanted and tissue that is not. There will not be a 1:1 mapping of these interfaces even between identical twins. What if one person had 2% more axons between one eye and the brain, or from the brain to a muscle — perhaps due to an injury? Even if the numbers were the same, what if the behaviors mapped onto them differently, so that some example like flexing the thumb required two groups of signals in one body but three smaller groups of signals in the other body? I'm hypothesizing these problems; I'm sure someone knowledgeable could state them more clearly (or tell me I'm full of crap), but I feel confident in saying that problems at the interfaces would lead to difficulties in at least perception and muscular activity. And those difficulties make me prone to reject the thought experiment for the same reason I rejected the color-swapping one: it's unrealistic.

The "software-only" examples of a memory swap or a memory addition are troubling because we know that memories are encoded in part in the physical structure of the brain. Memory swap would require tremendous "hardware" changes and would only be successful if it reached approximately the same level of change as the brain transplant case — and then we're back to those problems. Also, the brain has some finite storage capacity and it's not clear that it could hold two complete sets of memories. On a finer level of detail, some brain structures may be closer to their capacity limits than others, and would have problems even if the overall storage capacity of the brain wasn't reached. Memory recall would also be interesting — if both people had a memory linked to some particular scent, would both memories be recalled simultaneously? Can the brain even do that in parallel, or does it need to task swap? Would the person have both personalities, or just one? If they spoke, whose characteristic choices of words would they use?

These thought experiments raise a lot of questions that, no doubt, Steven will say are unimportant to his basic point. And I agree. That's why these are examples of bad thought experiments. There are plenty of theological issues to grapple with even for things that we do know are possible, such as twinning and cloning. I think it's much better to use those real cases to approach the theological issues rather than use thought experiments that people like me (heh) will get hung up on because we view thought experiments with great suspicion.

Better questions are: What, exactly, is a soul? When does it "enter" the embryo? How do you know? Why do identical twins have different souls? If it's because the souls "enter" after the embryos separate, why do clones also have different souls — because there, the separation is clearly after the soul is present? (I assume that human cloning will eventually happen and that the clone will be considered to have its own soul.) When do souls "leave" the body? If at brain death, what about people in unrecoverable vegetative states with atrophied brains — do they still have souls? What if that condition is reached through gradual decay? Do brain injuries that alter a person's personality also alter their soul, or is a person's behavior only a weak reflection of their soul? How do you know?

Granted, these questions do not cover the same areas as those Steven asked through his thought experiments, but they have the advantage of being based on things we already know are possible. They don't require imagination — or suspension of disbelief — to ponder.

Tiny Island