The self is real – it just, like Walt Whitman, contains multitudes. That’s the case made by Serife Tekin in Aeon. She begins by rightly pointing out the current popularity of disbelief in the self. She traces antirealist thinking right back to Hume, who said he was never able to spot his self by introspection; all he ever came up with was a bundle of perceptions. Interestingly she picks out Dennett as a contemporary example of antirealism, but she could readily have pointed to several others who think the self is an illusion or misinterpretation, perhaps stemming from our cognitive limitations, or from the reflexivity that arises when we turn our mind on itself.

Tekin by contrast suggests the self is both real and open to proper scientific investigation. It’s just that it has many forms; it is multitudinous. Borrowing from Neisser, she suggests five main dimensions of the self…

…the ecological self, or the embodied self in the physical world, which perceives and interacts with the physical environment; the interpersonal self, or the self embedded in the social world, which constitutes and is constituted by intersubjective relationships with others; the temporally extended self, or the self in time, which is grounded in memories of the past and anticipation of the future; the private self which is exposed to experiences available only to the first person and not to others; and finally the conceptual self, which (accurately or falsely) represents the self to the self by drawing on the properties or characteristics of not only the person but also the social and cultural context to which she belongs.

I don’t think these five types are meant to exhaust the variety of the self, which actually comes in a huge variety of shifting shapes. Nor are we meant to think that there is no basic unity; the five work together to provide an overall coherence of agency, though not without retaining some inner tensions and contradictions (nothing too strange psychologically in the idea that we may entertain contradictory thoughts and feelings in certain contexts.

The fivefold structure pays off because Tekin can give a separate account of how each can be addressed scientifically. The ecological self is easily observable, for example; fir the interpersonal self we need to pay attention to social aspects, but no great problem there. The most difficult seems likely to be the private self; Tekin seems to think we can get to that simply by interviewing people about ‘what it is like’, which perhaps underrates the problems.

Overall, it’s a sensible and appealing position. The curious thing is how close it seems to the kind of position taken by Dennett, here quoted as an example of antirealism. In fact, Dennett’s ideas are more nuanced than some. He doesn’t believe in a continuous, coherent self like a soul, but he is content to liken the self to a centre of gravity; not a real physical entity as such but a useful and harmless construction. As the author of the ‘multiple drafts’ theory of consciousness, I think he might rather like Tekin’s multitudinousness; and her approach to the private self looks quite like his ‘heterophenomenology’ in which we give up trying to study ineffable inner experience, but happily give consideration to what people tell us about ineffable inner experience.

This raises the attractive possibility that sceptics and believers might end up constructing effectively identical models of the self, the only difference being that one side regards the model as an eliminative reduction while the other sees it as simply analysis. I find that a strangely cheering prospect.



  1. 1. Christophe Menant says:

    Interesting perspective on human self. But it is a bit surprising to see that it does not consider animal self (looks a bit like talking about molecules without considering atoms).
    Animal self is a real subject ( Taking it as a starting point allows to look first at concepts like autonomy, teleology, normativity before having to address more complex ones like self-consciousness, free will, intersubjectivity and mental disorders. In other words use the nature of life to support studies on the nature of human mind. Also, an evolutionary perspective allows to look at how human constraints like look for happiness, valorise ego or limit anxiety could come from animal ones like stay alive or live group life.
    But this may be another subject.

  2. 2. SelfAwarePatterns says:

    “the only difference being that one side regards the model as an eliminative reduction while the other sees it as simply analysis.”

    The way I usually describe this is being reductionist without necessarily being eliminative reductionist. The first says, here’s a existing thing, whose sub-components and mechanisms we can see and analyze, while the second says, here’s a concept which, because it is not an irreducible thing, doesn’t really exist. The problem with the second way is that it eventually leaves you saying that only elementary particles exist, and even they may be made of some lower level reality. Ultimately it’s all emergent patterns, including the self.

  3. 3. Paul Torek says:


    I just discovered that there’s no such thing as a hand! Wherever someone declares a “hand”, there are only four long fingers, a palm, and a thumb. (H/t: Eliezer Yudkowsky.)

    In other words: what SelfAwarePatterns said.

Leave a Reply