Re: RT list: question to D Wilson

Dear Alessandro,

I am pretty sure that my knowledge of modularity issues is far too
limited to respond to your specific question, but as far as I
understand the massive modularity thesis, I see a paradox with what
you are saying.

My understanding of the thesis (which has been 'officially' taken up
in RT since Sperber & Wilson 2002), is that if you have a massively
modular mind you do not need to have central inferential systems (at
least for the purpose of comprehending ostensive stimuli). Now,
regarding the comments you received with respect to the issue of
innateness, I again believe that the official RT line is that this
ostension-driven submodule of our mindreading ability is an
adaptation, a mental structure with which we are genetically endowed
and which adopts an all the more efficient strategy for interpretation
at specific points during our cognitive development (Sperber 2000 and
the Wilson paper you refer to) - and if it were not genetically
prespecified these stages of sophistication would vary from individual
to individual. This, to my mind, means that we do not have any
developmental stage at which central inferential processing is taken
over by modular inferential processing, but rather that the module
exists right from the beginning and its operation gets refined/ more
sophisticated with time. The whole point of having a massively modular
mind, at least to me, is that in this way we manage to overcome the
otherwise slow e' (along with the mysticism that surrounds any attempt
to spell out the exact workings of central inferential processes that
operate across several domains - a point that Fodor himself underlined
ever since 1983)... In this sense, e' is redundant if you have e'',
otherwise what's the point of having two mental structures doing the
same thing?

So, and this is only mere stipulation on my behalf, when it comes to a
module for written symbols, the answer I assume you would get would be
that there is a dedicated module for this purpose, since modules in
the Sperberian sense come in all sizes and formats, even the size of a
single concept. This, however, does not mean that there is no
intricate organisation of submodules that belong to a more general
module that operates over writing and reading or whatever have you.
Neither does it mean that there are no mental structures (call them
modules in Carruthers' sense) whose domain of operation is to link
symbols together and feed them to the comprehension processor... Maybe
then, your friends condition might not mean that she has some
cognitive deficiency with processing inference as such, but that her
damage has caused her to not recognise written language (which perhaps
is a whole other domain on its own). This might help explain why her
communicative competence in spoken language has not been otherwise
compromised.

Again, these are only semi-educated guesses over a whole lot of
literature that I am not sufficiently aware of. I only hope they bring
about some stimulating discussion that will also help me clarify
things a bit better.

Best,
Stavros

Quoting Alessandro Capone <alessandro.capone@istruzione.it>:

> I have been reading
>
> New directions for research on pragmatics and modularity.
>
>
> A propos of dedicated modules, a friend of mine has had a stroke and
> received damage in the brain:
>
> as a result (among other things), she can no longer read or write written
> texts - yet we can speak fluently and understand as any other normal human
> being (with the difference that she gets tired ).
>
>
> So, should we deduce there is a module or submodule for written symbols?
>
>
> Now a question to D Wilson:
>
> suppose that interrpetation e of utterance U in context C can be obtained
> through two different strategies:
>
> Inference through central systems: we call e'
>
> Inference through a dedicated inferential module : we call e''
>
>
> In order to avoid having two different processes for the same inference,
> we need to say that e' is slower than e'' and that once e'' is ready, the
> mechanims resulting in e' is blocked.
>
> Presumably cognitive effort is involved in this.
>
> But can we exclude cases when the two processes converge and one process
> checks what the other has done?
>
>
> Also we need to specify when the central system processes prevail - that
> is when conscious inferential mechanisms are activated.
>
> alessandro
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

-- 
Stavros Assimakopoulos
Visiting researcher, University of Edinburgh
http://www.ling.ed.ac.uk/~stavros
-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.