Perhaps since Wittgenstein, we’ve become a bit cautious about taking language as a guide to reality, but here is, I suggest, one place where reflection on language provides a profound metaphysical suggestion about the nature of selfhood.
Amongst many of the realists in India, it is common to think of the self as a unique kind of substance, and its states (affective, cognitive, volitional) as properties. As such, my current state of visual perception of a computer screen would be schematized as my (the self’s) possessing the property of visual awareness of a computer screen. And this particular state of visual perception (dharma) does not hang in space, it is grounded in my self, the property bearer (dharmin).
In The Character of Logic in India, Matilal reminds us that predication in India often uses the locative case, where here, the primary relation isn’t expressed as “property—possessor”, but “locatee—location”. In the context of stock examples of inference, smokiness is located on the mountain, and therefore, fieriness is also located there.
We may note right away that the two ways of expressing predication often coincide. Using the genitive of possession or the locative of placement is then a matter of choice, and they are often used to express the same things. (E.g., for Rāmānuja, the selves and materiality are “located in” God, and also owned by him.)
I am not arguing that the “locatee—location” model is always better or radically different, but rather that the notion of the self as a location seems to me to make sense of what a self is, and what it does, in a profound way. On this idea:
A self is a unique location of mental states; a unique “place” where such states may be located. The existence of a self also allows for the co-location of varied states, the cohesion of which provides the basis of personhood.
What do we mean by location? I mean that in the “space of selves”, my self would be the unique place where the cluster of mental properties that comprise my personality are co-located (as well as the deeper subsurface states which support it). And this is the difference between my pains and yours. Mine are located “here”, yours “there”. You would say the opposite, correctly.
This is why there exist subjects and not simply a world of objects. Subjects are locations of experience, and hence suffering, pleasure, harm, and the rest. This is why they have moral worth, and why we must navigate our dealings with them with more care than with mere objects, which are unfit to be the location of such states.
In terms of personal identity, the self as location is not sufficient for diachronic or synchronic personal identity, but—at least the realists would argue—it is necessary, as it is needed to co-locate mental states, doing so for a host of states both synchronically and diachronically.
The self as unique location also make sense of notions of haecceities or uniqueness (sometimes mirrored by talk of viśeṣa in India). You can strip away all of my current properties and I’d cease to be a person, by my deep self is still a unique location of possible states and relationships. It need not be seen as an “essence” in a robust sense, but an irreducibly unique place that may uniquely locate possible subjective states.
One nice feature of this way of thinking of mental predication is that one need not hold that a self is a substance in a deeply metaphysically laden way other than noting that it is a place where one can locate and co-locate psychological properties.