Nagel on an expansionist conception of mind

"Conceiving the Impossible and the Mind-Body Problem"

Nagel says his position is similar to Colin McGinn but less pessimistic. I see that his position is compatible with Searle's biological naturalism, perhaps he is just dissatisfied with Searle's vague comments about emergent phenomena. He accepts supervenience, but claims that is not yet an explanation.

My own explanation has been evolving for years, basically trying to apply Barwise's concept of a scheme of individuation, where I see the scheme for mental states to be intrinsically first-person as well as third person. This accounts for the explanatory gap.

Below are some rough notes I compiled while reading Nagel's article, perhaps I will have time to clean them us some day. If I need to clarify the concept of a "verbal scheme of individuation" as part of the foundation of lexical semantics (for my planned thesis work), maybe I will need to set aside some time to come up with a clearly argued essay.

Notes on Mind:

Elements of a solution to the mind-body problem

1. Nagel: expansionist conception of mental entities.
2. Peirce: type-token distinction
3. Frege, Russell, early Wittgenstein, Barwise-Perry: atomism of situation-types
4. Searle: biological naturalism, introspection does not have complete access, epistemological-subjective can be ontological-objective
5. Perry: indexical reference?

Physicalists want an explanation of mind in terms of microparticles and their interactions. We don't need to invent any new particles, but we may need to expand the concept of interaction to allow for First-Person-Third-Person (1-3P) Schemes of Individuation.

A percept (available to any animal with eyes) is a (realist) type linking two glances at the same external object.
- below the level accessible to consciousness is the unity of neural activation patterns between saccades
- at the level of consciousness is the scanning of a scene, and integrating the successive visual impressions into a single recognition experience
- integration between perceptual modalities highlights the synesthetic experience, highly e-subjective

An occurrent concept is the interaction (within a type-level shared scheme of individuation) of two brain situation-types picking out "the same" shared information state.
- this generalizes the type-equivalence of two saccadic situation-types into a larger abstract situation-type which classifies the external token situations picked out by the two saccades, as well as future possible situation-tokens into the same type. The regularity of the external world is the wider type, but the narrow type is grounded in the regularities of a brain or several brains.
- social communication between mammals establishes externally communicated "concepts", including the precepts of a shared immediate situation.
- eye gaze ...


The token microparticles of a particular mental experience are primarily the spacetime-located firing brain cells of several individuals who shared the information states the caused the constituent concept-type elements of the experience, and secondarily the spacetime-located environment (both perceptible and enabling action) that is causally linked with those elements. The type of the mental experience is a complex relation classifying the firing processes of those brain cells and the relevant material environment via a shared scheme of individuation. The scheme is itself a biological (innate) and cultural (acquired) regularity grounded in the brain cells of an individual.

What is the gap between physical description and 1st-and-3rd-Person mental description? Physical description uses a 3P scheme, and the relations are non-dynamical and deterministic. Mental description uses a 1-3P scheme, which crucially references the 1P qualia of the mental entities.

A 1-3P scheme individuates mental entities that are literally unseen, but are readily inferrable from the observed behavior of a relevantly similar conspecific. The mental states are seen to cause the observed behavior, much as the observers own ocurrent mental states cause self-behavior via intention-in-action. The ability to cognize the unseen causal states in the minds of others is innate (in non-autistic individuals) in humans, and those mental states have intrinsic 1P properties (qualia) that are inherent in the use of a 1-3P scheme of individuation.

Mental causality is behavior caused by unseen mental states individuated through a 1-3P scheme.

Attunement to caused behavior, and the unseen mental causes that explain it, engages the hominid 1P understanding of the subject's ("my") mental will (states of intending-in-action) causing the kinesthetically accessed behavior. The kinesthetics of intention-in-action are 1P properties of the mental type, and thus are associated with an externally cognized token of that mental type as well as an introspectively cognized self-token.

The verbal-mental scheme is also 1P. The words for describing mental action (propositional attitudes) associate the "kinesthetics" of thinking a mental state with any external or internal token of that verbal type. Mental signifieds have 1P properties.

In this model, there are two layers. A mental event is a relation between nerve cells (not just the brain, the ennervated body is important) and the real or possible intentional-object of the event's mental "state". That Level One relation is part of a scheme of individuation, which could potentially be a private scheme (let us say for music at a high pitch that most humans can't hear). Then we ask what is the ontological grounding of that scheme? It turns out that there are two types of scheme, depending on an epistemological issue. Some schemes are strictly 3P, while schemes for classifying the mental states of others are 1-3P schemes, and attribute qualia to any token of the type. The qualia associated (felt, what-it-is-like-to-be) with a mental token of the observer is a neurophysiological "kinesthetic" access and very real. The qualia associated with the mental state of others is inferred, with the presupposition that they are physiologically relevantly similar in structure. The schemes are types that are grounded in species-innate regularities as well as acquired cultural concepts (notably verbal concepts), so cannot be reduced to the structure of one brain.

So occurrent mental events are tokens of a relation (a type of parametrized state of affairs where a hominid cognizes a real or possible situation). This relation exists as part of a scheme of individuation. A mental event of an actual physical object/event engages a strictly 3P scheme, so it is a ontologically objective brain state that is 3P cognized (thus epistemologically objective as well). If a mental event is about a possible situation, then it o-objective existence is in doubt, although that could be resolved to true with more information. Mental events about fictional 3P situations have no o-objective existence, their ontological status is purely o-subjective, although these cultural fictions may play an important role. So we can have 3P e-objective cognizing of o-subjective fictional situations. Now what if the observer is cognizing mental states in a conspecific? They cannot see mental states, but they can reliably infer the o-objective existence, using their 1P neurophysiological "kinesthetic" qualia to support a feeling of certainty about the existence of souls and their causal mental states. This is at some level a e-subjective cognizing of an o-objective mental state in others. We have intuitions of certainty about the o-objective existence of that unseen state, which we know using a 1-3P scheme. When a hominid cognizes its own mental state, it uses its 1-3P scheme to e-subjectively cognize, but their is even greater certainty of the o-objective existence of "clear and distinct ideas" with "kinesthetic" qualia. Of course even this 1P certainty can be mistaken, such as optical illusions, or confabulations, or delusions.

So a mental state does not require substance dualism, it is merely a token of a type in a scheme which may be 3P or 1-3P. It does not require property dualism either, since the gap between between 3P and 1-3P schemes can be explained by an expansionist conception of mind. Our expanded conception of "interaction of microparticles" only needs to encompass cognitive schemes of individuation. The explanatory gap about mental qualia emerges from the difference of 1-3P schemes (e-subjective, but usually o-objective if the mental state being picked out is actual) from 3P schemes (o-subjective, and usually o-objective if the intentional-object is actual).

Level One invokes Brentano's (and later Searle's) insight about Intentionality as the mark of the mental. At Level 2, we need Barwise's insights about schemes of individuation, and the recognition that some schemes (for recognizing mental states) are inherently 1-3P schemes. Level One is not controversial, but we need Peirce's insight about the type-token distinction, for use at level 2. The ontology of a scheme is controversial, since the 3P schemes lend themselves to a functionalist-physicalist reductionism. But I would take a neuro-chauvinist view that functionalist role-relation type is not adequate, a silicon realization would lack the physiological grounding to be physically the same. This is highlighted when we consider schemes of classifying mental states, where the phenomenology would be unavailable to funcationally similar robots, since the absence of physiology would make any "kinesthetic" qualia implausible, and two functionally similar states would not be the same mental state in a 1-3P scheme.

The explanatory gap can now be understood, and strong AI recedes as a prospect. Zombies turn out to be a confusion of reductionism, since a physiological zombie would not be functionally identical if it did not have a 1-3P scheme. The phenomenology cannot be separated from the physiology, and the e-subjective certainty about the identity of mental states is species-specific. Even if chimpanzees in a space colony were to evolve into human like intelligence, we wouldn't know what it is like to be a chimpanzee smelling a rose or eating a tea cake.