The evolution of 'philosophical psychology': from Ryle to Grice

If you try to study the 'philosophy of mind' (or philosophical psychology, as Grice prefers) of the Oxford Play Group to which Grice belonged (Grice, Austin, Strawson, Hampshire, Hare) you fail --. I do have Pears, "Questions in the philosophy of mind" (He was also a member of the group) and Hamsphire, "Philosophy of mind" volume, and a few others --. The opus magnum by Grice here is "Method in philosophical psychology". But there is nothing really that UNIFIES THEM.

In "Method" Grice refers to Ryle. Ryle, as per "The concept of mind", was labelled a 'behaviourist' or 'analytic behaviourist'. No ghost in the machine, and the rest of it. But what are Grice's criticisms?

Grice's criticisms are subtle. He does not find a good argument FOR 'analytic' behaviourism. It's not that Grice wants to rehabilitate dualism. Rather, Grice proposes to treat 'psychological concepts' as theoretical concepts. His approach follows proposals by D. K. Lewis, and before him, Aristotle ("De anima").

So, the behaviourist (under which label we can also place Wittgenstein, that Grice quotes quite a bit in "Method" -- "no psychological predicate without the behaviour that manifests it" -- from "Philosophical Investigations") is claiming that there is no need to EXPLAIN psychological concepts because there are no such.

Alla Paul, "Is there a problem about sense data?". Is there a problem about 'the concept of mind'. Only if you are confused.

The type of proposal by Grice is labelled 'functionalist', rather, alla D. K. Lewis and Aristotle. It is based on the Turing black box thing. There are

perceptual input ------ i.

and

behavioural output ---- o

In between, there is the black box, which Grice symbolises by

Ψ

---

"Ψ" is a psychological predicate. The logic is dyadic. Sort of. Consider:

"... believes ..."

or as Grice prefers, '... accepts ...' (he wants a concept that is general enough to stand for 'believes' and 'desires').

'... accepts ...' holds between an agent, A,

so that Ψ becomes

Ψ_A

but that's one argument. If it's dyadic there must be another argument. Here Grice introduces the 'that'-clause. What follows "... acccepts". John accepts that the cat is on the mat.

This yields the form of a psychological ascription (the ascription of a propositional attitude or psychological attitude in Russellian parlance) to be:

Ψ_Ap

But this is the 'propositional' logic level. We need to consider 'predicate logic'. "The cat is on the mat" (Ex)Cx & Mx.

So, the minimal propositional or psychological ascription becomes:

Ψ_A(Ex)Cx & Mx.

But you knew that!

----

Once we are able to work with these formulae, we see the corollaries. If an utterer utters, "The cat is on the mat" (or the king of France is bald) he presupposes or implicates that he believes there is a cat (or a king of France). So we can play with these sort of 'folk-psychological' laws:

Ψ_A(Ex)Cx & Mx ---> Ψ_A(Ex)Cx

and so on.

Or consider

A accepts that p
A accepts that if p, q
Therefore, A accepts q.

The way to go here is via principles of that sort alla Loar, Mind and Meaning, or Harman, Principles of Reasoning -- both Griceans, they.

---

At this point, we see that the larger ontological issue is only of mere academic concern for Grice. His proposal requires that we are able to work with psychological attitudes of a certain rich, or rather fine-grained, type.

Skinner's, or Watson's, or Ryle's type of behaviourism would not foot the bill in that respect.

Grice was enough of a naturalist NOT to be want to disassociate 'mind' from the rest -- so anything like Chomsky's "Cartesian Linguistics" and dualisms as such were a no-no.

By the time Chomsky tried to keep up with philosophers of mind, philosophers of mind were into very convoluted types of type-token identity theories alla Smart -- that Grice also considers. In this respect, Grice comes out as a non-reductionist. And so, his criticisms of Ryle's and other types of behaviourism strikes back with a vengeance.

But Grice's rejection of 'monism' (or behaviourism) remains methodological. He wants to say that if a folk-psychological theory 'instantiates' generalisations where the psi-operator occurs, this means that there IS a level of psychological explanation that is NOT, merely, a physicalist explanation of the type that the behaviourist is committed to.

Grice will also want to say that most behaviourists are NOT physicalist enough. He would mention that most descriptions of 'behaviour' are already 'loaded' with 'psychological' notions. "He approached me". "He moved towards me". -- is one of his examples. From Wittgenstein. Even our 'behaviourist' talk is imbued with psychologis. Or so Grice thought.

The panorama vis a vis Morris, Peirce, and "Unified Science", would perhaps present yet another problem with Grice. And he would ultimately recourse to issues like the ones mentioned by J -- we don't want 'robots'. We need free-willing agents and so on. At this point he would use 'practical arguments' to show that a mechanical reduction is the standard rejection of libertarianism. But we need libertarianism. Practical arguments are for Grice metaphysical arguments. Or rather, metaphysical arguments deal with the practical side of our thought products. Just because we don't want to commit to mechanism, we prove, eschatologically, that mechanism is a sort of minimalism that we need not buy. Supposing they were selling it. Or something.