I think that some traditional positions in ontology and epistemology are too ambitious for addressing the mind-body problem.
Since Einstein, the conception of spacetime allows us to think of ordinary objects as having worldlines, like worms in four dimensions. But as far as visual objects are concerned, this is only accessible at the granularity of saccades. The impression of an individual saccade is not accessible to consciousness. The visual object recognition faculty involves cognizing a perdurant object across saccades.
This granularity warns us that ordinary cognition is quite different from instrument-aided scientific observation. We cognize macro-bodies and the relations they stand in, not the micro-particles and interactions of traditional physicalism. So in studying the schemes of individuation underlying ordinary cognition, cognition researchers are interested in the spatial granularity that can be discerned by unaided vision, and the temporal granularity that binds successive saccades.
Primates can track moving objects, and have some innate and acquired schemas for classifying them. The motion of an object is one of the things that gives it interest value. When do social primates develop a shared acquired scheme? If we refer to a shared acquired schema (for a type of visual object) as a visual concept, how are visual concepts related to other concepts? And how do these various concepts fit into the shared scheme of individuation of cognizing hominins? In what way are these concept-level schemes a pre-adaptation for the lexical-level schemes of language?
What mechanisms are there for acquiring a shared concept schemas in the absence of language? Deaf cognition, both in acquiring sign language lexis and the concept schemas of linguistically-isolated deaf children, provide interesting evidence.
How do hominins individuate the mental causes of the macro-behavior of other hominins. What can we say about the imputed psychic-substance (in the Aristotelian sense of ousia, not in the Cartesian sense of res cogitans) that bears beliefs and desires? For Aristotle, the psyche is the form of a human being that is hylomorphically inseparable from the body. Descartes had reasons for insisting on a separation are historically interesting, but no longer contemporarily relevant (I suspect they have to do with a compromise with theology about the domain of scientific truth, after the lesson of Galileo's house arrest; and Descartes' Platonic focus on Reason as a precursor to the enlightenment). Today, cognitive researchers focus on the embodied mind.
One hypothesis is that there is a mental-state-recognizing faculty, perhaps not so different from the face-recognizing faculty, an innate competence that acquires specific content in performance. It undergirds a capacity to individuate macro-objects. There is a relatively limited set of dimensions of classifying faces (certainly male vs. female, child vs. adult; but most faces are sui generis), what about the classification of mental states, in hominins with and without language?
Faces are directly perceived (for the blind, using hands not eyes), while a mental state is a "causal object" that is readily-inferred and not directly perceived. In a sense, mental-state recognition is more like phonology or melody recognition, which is layered above the impressions of direct acoustic perception. Recognizing beliefs and desires is related to causal explanation of hominin action. Here I mean explanation in a pre-linguistic sense, accounting for the constraints relating a motivating situation to an action result.
Even non-linguistic hominins need to classify hominin behavior, and recognizing the beliefs and desires (among other mental states) is clearly selective for the species. It would be interesting to review the data from research on ape language. We can be quite skeptical about whether chimp behavior is language, but we would still like to understand the capacity of chimps to cognize the beliefs and desires of other hominins (chimp and human). This is related to their own repertoire of facial, gestural and vocal displays, and how they recognize intended displays in others. Unlike the colors in a butterfly wing, primate displays are muscle-controlled, and a caused by intention-in-action. How does a primate discern the underlying intentions-in-action, and what is the "vocabulary" (which may be infra-lexical concepts) that are shared between intentions-in-action, desires, and beliefs?
What makes two mental states "the same"? Let us start with recognizing that two visual objects are the same object and the same type of object. If an object is visually tracked, they our unconscious visual processing and visual object recognition individuates a single object with identity over time. If we lose sight of an object, or go to sleep, how does a hominin identify a later sighting as the same object, or the same type of object? We may distinguish between the categories of objects that receive proper names (people, places, pets), and those that don't. Even those that are referred to only by common nouns, some may be relevant to action as particular objects, and hominins may refer to them indexically (and deictically) as this fruit here, or my handaxe.
To the extent a hominin's cognitive ontology recognizes two sightings of a visual object as the same particular object, there is a moderate risk of error. If it is sufficient to guide action to recognize an object as of the same type (a mouse, a vegetable), the hominin may not distinguish particular instances.
The identity of persons is different from that of most animals (pets being a transitional case). We attribute a psyche to persons, and we recognize and remember details of that psyche (what does that person believe, or typically desire) as a basis of classifying their macro-behavior in terms of occurrent causal mental states.
When are two mental states "the same" mental state? We can distinguish between an occurrent thought, and a remembered mental state. Generally, there is no relevance to action associated with the continuous identity of an occurrent thought, a belief held yesterday and today is a single belief, for purposes of classifying observed behavior in terms of causal intention-in-action. When humans think of the same type of mental state, they use a discrete scheme of individuation at the lexical or sub-lexical concept level. These concept schemals (including lexical schemas, if present) are shared by a social group.
The shared concept schemas of mental states are unique relative to other cognized entitites, because they can be shared between the observed hominin and the observing hominins. Perhaps a new concept is acquired at first as a commonality between two observed hominins. However, it seems inevitable that the observing hominin will also experience that mental state in sympathy. At any rate, it seems most mental states have a first-person character or qualia. The entire scheme of individuating mental states is distinguished from other parts of the cognitive scheme of individuation because of this intrinsic first-and-third person character, while non-mental entities have only a third-person character. This accounts of Descartes' intuition (and Ewing's) that res cogitans is a different substance from res extensa. However, with our improved understanding of biological matter and processes, we can in fact provide a hylomorphic account of mental entities, and the lead to substance dualism is unjustified. But mental properties are different from non-mental properties, but not in a way that requires supernatural intervention. We can naturalize our scientific explanations of mental states, and folk psychology of mental states, with a natural (but dual aspect) theory of cognition.
The material basis for a mental entity is to some extent the occurrent neural pattern that fires during an occurrent thought, but also the dynamical memory and innate genetic-ontogenic dynamical structures of the innervated body. This neurophysiological characterization suffices for explaining the third-person characteristics of mental states. To account for the first-and-third person characteristics, we need to layer on top of this the shared schemes of individuating the mental, which don't exist within a single innervated body, but are a social construct (over a shared species competence) of a group.
