storage / computation again

[Joyce Tang Boyland]

This is a bit of a delayed response to Brian
(I'm just coming off of grading mid-terms):

I hope I did not imply that we should use a model of word learning
as the basis for understanding the automatization of syntactic patterns.
If anything, it should be the opposite, that we might use the automatization
of syntactic patterns as the basis for understanding the formation of new
words (like _brokeded_ which does have "slots" so to speak and transitions).

Also, building off Martin Haspelmath's posting which muses about
criteria to use in deciding whether storage or processing is superior,
I think it's worth trying out a stronger claim, namely that
neither storage+retrieval nor computation is intrinsically superior.
In skill learning, you can start either with storage+retrival
of an exemplar (or prototype), or with computation of a sequence.
It depends on the details of the situation which route is used.
But whichever way you learn to use a new construction, the outcome is
not that you have added a new construction (or even a new word)
to your piggy bank, but that you have smoothed a set of transitions
so that such a sequence is more readily assembled and produced.

I agree with Brian that it would be great to have more empirical
psycholinguistic research on these topics, and that it is relevant
both to acquisition and to historical change.

Joyce Tang Boyland

>> Date:         Wed, 14 Oct 1998 17:34:12 -0600
>> Sender: FUNKNET -- Discussion of issues in Functional Linguistics
>> From: Brian MacWhinney <macw at CMU.EDU>
>> Subject:      rote vs rules
>>
>> 3.  I agree with Joyce that language is a skill.  However, the devil is in
>> the details.  If we fail to recognize the fundamental difference between
>> word learning and syntactic automatization, I am worried that we could go
>> down some false paths.  The routinization of the word is supported by a
>> tightly predictive association between audition and articulation.  When we
>> hear a new auditory form, it appears that we use the phonological loop on
>> some level to store it.  As we then attempt to match to this form using our
>> own articulations, we convert a resonant auditory form to an entrenched
>> articulatory form.  Work by Baddeley, Gupta, Cowan and others has taught us
>> a great deal about the details of this process.  Yes, you can use ACT-R to
>> model this, but you will be using a restricted subset of ACT-R and the
>> process of deciding what restrictive subset is applicable is the whole of
>> the scientific process of understanding the neuronal mechanics of word
>> learning.
>>
>> Trying to use a model of word learning as the basis for understanding the
>> automatization of syntactic patterns strikes me as quite problematic.  The
>> central problem is that predicates have open slots for arguments.  Words,
>> as Wally notes, are largely slot-free (of course there are exception, such
>> as infixes etc.).  I tend to think of this level of skill automaticity in
>> terms of Michael Jordan faking out Karl Malone in the last points of the
>> final game of the NBA finals.  Jordan clearly has a flexible set of plans
>> for dunking the ball into the basket against the opposition of a defender.
>> What is automatic in his actions is the move from one state to the next.
>> The skill is in the transitions.  It strikes me that sentence production is
>> like this and that word level articulation is basically not.
>>
>> --Brian MacWhinney