Wikipedia has a really good summary of Hume’s views about the self, and I strongly suggest reading that short and sweet summary of what he thought (not least because it references the mind-blowing suggesting that Hume may have held a Buddhist view of the self). Here are Hume’s own words concerning his thoughts on the self:
“There are some philosophers who imagine we are every moment intimately conscious of what we call our self… For my part, when I enter most intimately into what I call myself, I always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. I never can catch myself at any time without a perception, and never can observe any thing but the perception. When my perceptions are removed for any time, as by sound sleep, so long am I insensible of myself, and may truly be said not to exist. I may venture to affirm of the rest of mankind, that they are nothing but a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement.”
I don’t think does justice to what a self is.
Now if you know philosophy, you know that the identity of the self is a really tangled knot of a problem that is hard to undo. It seems like every definition of the self that one can provide can be shown false by some thought experiment in which that attempted definition is satisfied but your gut tells you that self-identity hasn’t been maintained, indicating that you need to modify or discard your definition. A good overview of these problems and self-theories can be found here.
I suspect that the word ‘self’ is a lot like the word ‘kind’ used by creationists. Creationists say that one ‘kind’ of animal cannot change into another ‘kind’ of animal, thus meaning that evolution can’t happen. I’ve written a humorous dialogue to convey my thoughts on this:
Mr. Sensible: Tell Me, Mr. Creationist, what is a kind?
Mr. Creationist: Look at the Great Dane, the Collie, the Chihuaha, and the Poodle. They’re very different, but we can detect that they ought to be grouped together due to their great similarity. I call that grouping a ‘kind.’
Mr. Sensible: Tell me what the defining characteristics are for that group. Just so we can have some scientific precision.
Mr. Creationist: A dog walks on four legs, it has fur, two eyes, two ears, you get the picture.
Mr. Sensible: But what about the Peruvian hairless? Is it still a dog even though it has no hair, which you said was a defining characteristic?
Mr. Creationist: Okay, I’ll take fur off my list. How about I just use the other items?
Mr. Sensible: Not so fast. Time for a thought experiment: What if my pet dog gets pregnant and has a mutant baby with no ears. Things like this have been known to happen from time to time. Would that still be a dog?
Mr. Creationist: Sure. I see where you’re going with this, so I’ll drop ears off my list too.
Mr. Sensible: But we could do the same thing with just about any and every characteristic you’ve listed.
Mr. Creationist: Uh-oh.
Mr. Sensible: So the word ‘kind’ refers to nothing, at least given the way that you’re using it.
Mr. Creationist: I grant that you’ve got a good philosophical point, but you use ‘kind’ and other similarly vague words all the time in your own day to day life. Surely you don’t think you’re talking nonsense when you do that? I mean, these words must have meaning, even if there are some philosophical puzzles about them that we don’t know how to solve?
Mr. Sensible: Of course. However, the way that we deal with vagueness is not going to be friendly to your argument. Let me explain: suppose that we define the word ‘dog’ to include all the characteristics you mentioned, but instead of the definition being rigid and inflexible, let’s say that we treat each characteristic like a check box. Every animal that can meet most of those checkboxes counts as a ‘dog.’ We may not be able to say exactly how many checkboxes the animal in question has to have in order for it to be a ‘dog’, but we can push that problem aside if we just allow that enough checkboxes in order to match our intuitions about what the word ‘dog’ means. If we adopt this framework, the you’re whole contention that “one kind cannot change into another kind” is dissolved. If a “kind” just represents a set of checkboxes, and an animal in any given “kind” doesn’t have to meet all of them in order to be considered a member, then it’s safe to say that no evolutionist believes an animal ever gave birth to an animal of a completely different kind.
Mr. Creationist: But I thought they did. Don’t evolutionists believe that the whale kind came from the land animal kind?
Mr. Sensible: Sure they do, but the belief that whales evolved from land mammals isn’t at odds with this framework. We evolutionists think that a population of land mammals changed a little bit a time over many generations into whales. Every single generation gave birth to animals that you would say were “of the same kind.” But every few generations a characteristic or two changed. If a scientist had been able to study these whales over tens of thousands of generations, and he kept a checkbox list to describe the species, every few hundred generations he would’ve had to add a new check box or cross one off. Eventually, the check box list he would end up with would look a lot different from the one he started with. So you see, evolution can co-exist with idea that no animal ever has a child that doesn’t at least mostly match the parent’s checkbox description. All you have to do is say that the radical changes in the checkbox took place over lots of generations and not just one. [Note: A longer version of this dialogue was posted here.]
To return to my point, I suspect that word ‘self’ doesn’t really have an ultra-precise definition. Rather, the ‘self’ might be much easier to define if we define it with a list of check boxes. If we’re going to identify a person as “the same person” from one moment to the next or one decade to the next, we first need to come up with a working understanding of what a self is in one moment. A self is not just a collection of memories or present experiences, although these may certainly qualify as aspects of a self. A self is a set of personality traits, it is a complex action/reaction system which are ultimately made manifest by the brain (although the brain need not be the medium). While personality traits are probably ultimately defined by observable behaviors or patterns of behavior, and as such are ultimately reducible to sensory data, I don’t see this as a vindication of Hume’s thesis. There’s a self-explanatory difference between this concept of the self and the sensory bundle theory.
To return again to the point, if we created a list of characteristics of a self, the thing that makes a self identifiable as the same from one moment to the next would be something like the sameness or near sameness of body and psychology from one moment to the next.
Sameness or near sameness of body and psychology from one moment to the next… Let’s take the time to unpack that: “Sameness or near sameness” is added because we are constantly changing in various ways from one moment to the next. A personality trait can appear, disappear, or change in various ways over years or in the course of a moment. Sam might resolve to never drink again after suffering from alcohol poisoning, but no one would say that Sam literally became a different person ten minutes after he resolved not to drink, even if he kept his and never drank again. Same is still the same person because their is continuity over time between his body before he began to drink and his body after he sobered up, and there is probably great continuity in Sam’s psychology, even if several there are several noticeable differences (for instance, if Sam liked Dorito’s and counrty music when he was a drinker, he will probably still like them in his new life as a sober man).
Sameness or near sameness of body and psychology from one moment to the next. I believe that psychology is ultimately reducible to the brain. However, the psychological properties of the brain might be able to be exemplified by certain arrangements of matter. For example, one day in the future it might be possible to create sentient robots. And it might also be possible to upload the contents of your brain to a computer. If someone were to induce your body into a coma-like state so that you lost consciousness, and they uploaded your thoughts and all your brain content onto a computer whilst killing your body, would you still be the same person? Suppose, for the sake of argument, that your stream of conscious awareness continued as usual, and being a computer upload was just like being a brain. I’d say that you are the same person. So a self in one moment can be the same as the self in another moment if there is a great degree of continuity in the psychology of the self through time, regardless of whether the physical parts that make-up the self are the same. We may not be able to say precisely what degree of continuity makes a past self identical to a present or future self, but so what? No one ever promised that there has to be a super-precise definition of the word ‘self,’ and if a somewhat fuzzier definition fits, we need to go with it.
That brings me to another important thought experiment: what if, in the process of uploading your thought patterns to a computer, you suddenly wake up from the coma? Is the computer uploaded self you or is flesh-and-blood body self you? Are both of them you?!? Assuming we have to pick one, my view is that the flesh-and-blood self would be you in that situation (but not the situation where your body doesn’t survive) on account of the fact that the body would have a greater degree of continuity with your past self than the computer uploaded self would. Identifying a present self as being the same as a past self is not just a matter of having a sufficient degree of continuity with the past self, but also having the greatest degree of continuity than any other contender.
Sameness or near sameness of body and psychology from one moment to the next. I’ve heard that every molecule that makes up your body is replaced every seven years. I have no idea if this is true (I’m pretty suspicious of it) but for the sake of argument let’s suppose that it is. Let’s also suppose that it is possible for all of someone’s personality traits, likes and dislikes, and so on, to be changed dramatically or even replaced over the course of seven years. If that were true, would it pose a problem for identifying a self throughout time? I don’t think so. Even though the self might differ completely in terms of the atoms it is made of and the way those atoms are arranged at the beginning of the seven year period and at the end, as long as there are intermediate steps in between that don’t differ so dramatically (if the person is always at least 95% the same from one day to the next at every point in the seven year period) then I’m totally comfortable identifying the person as being the same in both the beginning and end point of the seven years. I wouldn’t be so comfortable making that identification if all of the atoms and the arrangements of atoms constituting personality traits, memories, etc. were replaced in a single second. But if the transition is more gradual, my intuition says that it is the same self. And that’s all that matters, since finding the right explicit definition of a word is only determined by our intuitions of what the word refers to and how we use it.
So, those are my thoughts on the self. In case you’re interested, my views on the self are what Robert Nozick called ‘The Closest Continuer Theory’ which he discussed in his masterpiece Philosophical Explanations.
