Grice and autism

In "HOT theories of thought, autistic utterers and the concept of truth
and falsity" M. Plug (PhD, King's College London) refers to K. Gluer and P.
Pagin's attack of Higher-Order-Thought ("HOT") theories of meaning, such as the one offered by Grice back in 1948.

They do so by highlighting psychological research on a certain subset of autistic children: autistic children who have a high verbal mental age, as measured by standard vocabulary tests, and yet have trouble attributing thoughts to others
(as measured by so-called "false-belief" tests).

A HOT theory of meaning such as Grice's posits that in order to mean something with one's words (in order to assert something, for example), one has to
attribute certain complicated beliefs to one's addressee.

These attributions seem outside the capability of those who can't pass false belief tests, it is argued, and yet autistic children do seem to MEAN *something* with their words.

Some have criticised Gluer and Pagin (Reboul, 2006) because data on the communicative abilities of autistic children is widely taken to be confirming Grice's account of linguistic communication (the theory of conversational
maxims).

Why, Reboul wonders, do Gluer and Pagin take the data to be disconfirming when
it is widely taken to be confirming Grice's theory (this puzzlement is the basis on which Rebould build further arguments against Gluer and Pagin)?

Plug argues that Reboul's criticism (and further arguments) miss their mark because Gluer and Pagin attack Grice's theory of meaning, not his theory of communication.

Reboul does not properly distinguish between the two.

However, distinguishing properly between the two theories also highlights
the connections that exist between Grice's theory of meaning and theory of communication, and then it becomes clear that the Gricean theory of meaning
can't be refuted in the way Gluer and Pagin attempt to do, by simply referring to the high verbal mental age (based on vocabulary measures) of autistic children.

Gluer and Pagin assume that a verbal mental age of between 8 and 10 years old,
based on vocabulary measure alone (and considerations around language development), is proof of the capability of making the speech act of assertion (See also Pagin (2007)) but this is just something that Grice could deny, given the fact that autistic
children have severe problems in communicating, and given his argument that assertions are essentially communicative, an argument which comes
online before the analysis in terms of intentions which is under attack.

In addition to referring to the psychological data, one needs to give a separate
argument why the linguistic performance of autistic children counts as making assertions, among other speech acts.

Plug eventually argues contra Grice (and Strawson and Recanati and even Pagin (2004)
who take assertion to involve an intention for a social effect, albeit that this intention does not determine the content of the assertion) that speech acts, especially the speech act of asserting, are not essentially communicative.

To proof that speakers are making assertions it is su cient (but probably not
necessary) to proof that the representational character, the difference between a true sentence and a false sentence is understood (the speaker possesses
the concept of objectivity)|which requires more than just applying the terms `true' and `false' correctly and that the behaviour surrounding an utterance showed that the speaker had the intention to utter a sentence with a truth value (and
not a mock sentence).

To prove that a subset of autistic children have this ability before passing
false belief tests requires looking into more than just vocabulary development.

One has to look at grammar development and other specific linguistic
performance, all in relation to false belief test performance.

Plug reviews the data and concludes that Grice's theory of meaning, if not as Grice conceived it, is indeed refuted by data on autistic children. Or not.