LINGUIST List 4.994

Wed 24 Nov 1994

Misc: Linguistics as psychology, Canadian raising, Implicature

Editor for this issue: <>

Directory

1. Vicki Fromkin, Re: 4.967 Linguistics as psychology
2. , "Canadian" Raising
3. Rob Stainton, Literal Meaning and Implicatures

Message 1: Re: 4.967 Linguistics as psychology

re Koenig's comments on the aphasic patient. It is not true that all
linguistic theories and or processing models distinguish between
syntactic and semantic aspects of the grammar. See for example some recent
papers on connectionist theories of processing which specifically deny
such a distinction in comprehension. And it is interesting that it
is not necessary to know the meaning of 'kick' nor of the agent/patient,
in order to determine who did what to whom when the syntax makes
this clear. However, it is true that this patient studied by Saffran and
Breedin does not provide evidence for or against a particular theory which
does make this distinction in the syntactic assignment of thematic roles.

re the original question which started this whole discussion -- is
linguistics a subbranch of pyschology? Does it really matter? If we wish
to lump all areas of inquiry concerning the nature of the mind and cognition
as subbranches of psych then fine. But it seems to me that what is more
important are the questions one is interested in and how one views linguistic
acquisition and representation and processing. The processing aspects or
linguistic knowledge in use (i.e. production and comprehension) must
involve psychological mechanisms. The question which divides many in the
field is what those mechanisms access. Some of us believe that underlying
all linguistic processing (or performance) there is a real mental grammar
and a major area of linguistic research is to determine the nature of this
grammar, what the universal constraints are which to some of us are
genetically determined, and the extent to which there is an interaction
in production and comprehension between the grammar and other aspects of
cognition including non-linguistic conceptual knowledge, and general
mechanisms which may not be specifically linguistic.

Vicki Fromkin

David Wigtil is mistaken, I think, in the same way that a whole
generation of phonologists (incl. Chomsky and Postal) were about
the so-called Canadian Raising phenomenon being the same thing
as the slight lengthening of vowels before voiced obstruents
esp. in monosyllables (or shortening before voiceless ones, I guess).
Canadian Raising involves, depending on dialect, only /ai/ or
both /ai/ and /au/ and typically produces lexical (or if you
prefer the term, phonemic) contrasts, whereas the lengthening
(shortening) rule applies to all vowels and produces no
contrasts. Using  to represent a vowel higher than [a], there
are speakers who have a contrast between the noun and verb 'hide',
one being /haid/, the other /hid/, for example. I believe this
occurs in many speakers from Chicago, and perhaps nowhere else.
Now, as far as I can tell, the /i/ in /hid/ is longer than the
/i/ in /hit/ 'height', the same way that the /ae/ in 'fat' is
shorter than the /ae/ in 'fad'. Likewise, there are speakers who
contrast the verb and adverb 'like', using /ai/ in one and /i/ in
the other, and again, the /ai/ in /laik/ will be short, like the
/ae/ in 'lack' as opposed to in 'lag'.

(I cannot at the moment recollect which is which in these minimal
pairs. That's why I have phrased things the way I have above. Sorry.)

John Lawler's note about "God, I love people of who signal" reminded me of
two other interesting examples of "non-literal" speech.

I. First Example

Many people -- myself included -- are tempted by the idea that "non-literal
meanings" are properties of *utterances*, not of expressions. (Let's call
this "Donald Davidson's view", for lack of a better name.) So, for
example, on "Donald Davidson's view" the *sentence* (1) has only a single
reading. (A reading on which it is evidently false).

(1) Jim is a bulldozer

But speakers can use (1) to communicate something which is true. However,
goes this line of thought, what speakers communicate is no part of the
meaning of the *sentence*.

Here, however, is an example which is at least puzzling given such a view.

(2) Is anyone seated here?

Given "Donald Davidson's view", it might seem necessary to claim that (2)
has only a single meaning (whose answer is inevitably "no", because it's
*obvious* that no one is currently occupying the seat). But, the story
would go, (2) can be used by speakers to ask a perfectly reasonable
question, namely:

(3) Can I sit here?

The difficulty is that the *standard use* of (2) (in so far as there is
such a thing) is to ask (3). Hence the claim that (3) is not part of the
meaning of (2) is not immediately plausible.

I find this example interesting because of the issue it raises with
respect to conventions of usage and conventions of meaning. There
certainly seems to be some sort of convention governing the use of (2) to
ask (3). But does this convention contribute to the *literal meaning* of
(2)? Ah, there's the rub!

II. Example Two

The standard line on conversational implicatures seems to be that a speaker
must at least appear to be violating a maxim in order for a conversational
implicature to kick in. (Whether this was Grice's view is another
question...)

But consider the following case. I was talking with one of
my undergraduates, saying that I really enjoyed playing bridge. He replied:

(4) My grandparents play bridge

His saying this gave rise to certain implicatures. (E.g. "Only old people
play bridge".) But, at first glance, he didn't even *appear* to violate any
maxims. After all, telling me that his grandparents play bridge is a
perfectly cooperative step in a conversation about bridge.

How, then, does the implicature arise? (I suspect it has something to do
with *degrees of relevance*. Cf. Sperber and Wilson)

Anyway, neat examples to chew on this rainy Saturday afternoon.

Best,
Rob Stainton

 -- Robert Stainton -- Philosophy -- Carleton University
rstaintoccs.carleton.ca