“People often speak as if there were a serious philosophical problem about selves. Is there a self? Is the self knowable? How does the self relate to the body? These and other questions are thought to make up something called the problem of the self. I doubt seriously that there is any such problem. Not because the self is unproblematic, or because there are unproblematically no such things as selves. My trouble is that a problem has to be about something: even if there are no selves, there must at least be some problematic idea or concept of a self, if there is to be a problem of the self. As far as I can see there is no such idea. What is a self? For every answer to this question, there is another answer not only incompatible with it, but wholly unrelated. There is virtually no agreement about the characteristic features of selves: depending on whom you believe, selves may be concrete or abstract, material or immaterial, permanent or ephemeral, naturally occuring or human constructions, essentially subjective or publicly observable, the same or not the same things as people. There are not even any agreed paradigm cases of selves, things we could point to or describe and say, “A self is one of those.” But no concept could be so problematic that no one could agree about anything to do with it. For lack of a subject matter, there is no problem of the self. […] In fact the matters discussed under the heading of ‘self’ are so various that it would be a pun to say that they were all about some one thing, the self. Because legitimate inquiries that go under the heading of ‘self’ are really about something else, and can be (and typically are) put in other terms, we can easily do without the word ‘self’.”

[…]

Let us consider some attempts to say what a self might be. It doesn’t matter whether we take these proposals to be definitions of ‘self’ or only as giving essential or paradigmatic or at any rate salient features of selves. Take them as proposed answers to the ordinary question, What is a self?

Account 1. One’s self is that unchanging, simple substance to which one’s impressions and ideas have reference.

[…]

Account 2. One’s self is the inner subject of one’s conscious experiences

[…]

Account 3. One’s self is just that person, himself.

[…]

Account 4. One’s self is that indescribable and unidentifiable private, inner being within one.

[…]

Account 5. One’s self is what one values above all else.

[…]

Account 6. One’s self is the unconscious mechanism responsible for the unity of one’s consciousness.

[…]

Account 7. One’s self is a psychological or behavioural attribute of one.

[…]

Account 8. One’s self is an aggregate of or a construction out of one’s sense experiences

[…]

What, then, of the problem of the self? What is it a problem about? There is clearly nothing that satisfies all of these accounts. There aren’t even two or three similar kinds of things that each satisfy most of them. It should be equally clear that there is no one thing–no single idea–that all of these accounts could reasonably be seen as trying to capture. There is no one sort of thing that some believe is a construction out of senseimpressions and others take to be a mental attribute, a simple substance, an organ, a human being, or in some cases even a house or a hamster. It should also be clear that there are no agreed characteristic attributes of selves, or even any generally accepted cases. (We can’t even pick out a self in a purely relational way, for example as ‘Bertrand Russell’s self’, without controversy, for on some accounts of ‘self’ there are no selves to pick out, while on others Russell might have had any number of different selves.) I conclude that those who use the word ‘self’, if they are saying anything coherent at all, must be talking about completely different things. Thus, there is no such idea as the idea of the self, and therefore nothing for the “problem of the self” to be a problem about.

[…]

Once we have accounted for people, their mental features, their relation to those human animals we call their bodies, and so on, we think we need to say something about “the self” as well. There is no good reason to think so. Or so I claim. Of course, merely putting a number of so-called problems of the self in other terms doesn’t show that the term ‘self’ is superfluous. I may have overlooked legitimate problems or questions that can be put only in terms of the free-standing noun ‘self’ or some equivalent term. In that case I should have to retract my claim that there is no legitimate problem of the self–though the problem so revealed is unlikely to be the problem commonly thought to bear that name. But if the word ‘self’ really has no agreed meaning, and leads us into troubles we could otherwise avoid, and if we can easily get on with our legitimate philosophical inquires without it, there can be no reason, other than tradition, to continue to speak of the self.”

[full text]