LINGUIST List 3.480

Thu 11 Jun 1992

Disc: Rules, Innateness

Editor for this issue: <>

Directory

1. Swann Philip, 3.473 Rules, Innateness
2. Rick Wojcik, Re: 3.473 Rules
3. benji wald, Re: 3.473 Rules, Innateness
4. , Innateness and Linguistic Theory

The penultimate issue of Behavioral & Brain Sciences contains two
articles on the innateness issue with extensive peer commentary and
bibliographies. Highly recommended!

Philip Swann
University of Geneva

Message 2: Re: 3.473 Rules

Vicki Fromkin writes:

>...When checking as to what
>people consider speech errors we obviously have to test with speakers of
>the same dialect unless we are interested in the differences in dialecta
>grammars. My point is simply that in order to judge something as an
>error, one must have a set of principles/rules/constraints in one's
>mental grammar which determine what is or is not ill formed according
>to that grammar (not someone else's).
?1;2c
It is possible that judgments about well-formedness can come from a variety of
sources--not just a single system of principles/rules/constraints. For example,
you might judge something ill-formed because you can't imagine yourself saying
such a thing. On the other hand, you might judge something well-formed because,
even though you wouldn't say it, you've heard others say it. When linguists
distinguish between 'acceptability' and 'grammaticality' judgments, they
explicitly recognize the possibility that not all judgments about linguistic
form have a monolithic origin. It is possible that well-formedness judgments
are really epiphenomenal--based on mental operations that play some other role
than just to supply the speaker with knowledge about good form. So the question
is whether there exists a coherent psychological 'grammar' in the generative
linguist's sense. Do we need to define a set of principles that gives rise
to grammatical judgments directly, or do we need to define knowledge sources
about several things--e.g. language production, expected production from others,
social dictums, etc.--that give rise to grammatical judgments indirectly? Is
my judgment that somebody is speaking with a southern accent (or fake southern
accent) based on the same type of knowledge as my judgment that the
speaker is not speaking my dialect? The former is based on what I expect to
hear, but the latter on what I know I can say.

Message 3: Re: 3.473 Rules, Innateness

issue 3.473. In response to McConvell's reservations about my last comment,
I guess I have to make it clear that I think it is most unlikely that speech
error is ever a source of change. My intention was to distinguish speech
errors which may be so for one speaker, but be "rule-governed" norms for
another, from those which cannot be "rule-governed" in a linguistic sense
for anybody -- because language "doesn't work that way". Examples of
speech errors which have no parallel in any empirically observed grammar are
such things as those long distnace metathese I mentioned in a "cu/f/ a
co/p/ee". No language has a grammatical rule "to form the past tense,
plural, or whatever, first you need two words, and then you must take
a specified segment of the two words and exchange them". The question
had earlier been raised about whether errors were strictly linguistic
or not, I think, or maybe it was whether the rules which they violate are
based on some innate linguistic capacity, or from some more general
capacity which the particular linguistic capacity also happens to exemplify.
I don't really know what speech errors as a global phenomenon have to do
with this question -- but the question suggested to me that different
kinds of speech errors may have different motivations, some more obviously
nonlinguistic than others. Vicky's blends are borderline, because
dialect geography shows us that in some cases blends can become linguistic
changes NOT!!!! THAT THEY STARTED OUT AS SPEECH ERRORS, please understand.
That is, there is a blending process in some types of linguistic change.
On the other hand, I'm sure Vicky's intention in the examples to which
I was responding was to exemplify planning/execution errors, resulting in
reordering of intended activities -- something which occurs quite apart
from language in other activities, and which I expect also occurs in the
behavior of other animals as well. Does it, Vicky? and if not, is that
part of your point about what speech errors reveal about ling-cognition?
 Problem with blends is that in some cases more than one possible
explanation for the source of the behavior may be possible, "by any
chance". In any cases, clearly nonlinguistic and problematic cases should
be distinguished for methodological purposes from cases in which
a constraint is lost or "relaxed" on a linguistic "rule". The problem
remains to my mind whether a speaker's reaction should have any bearing
how we analyse the behavior. As opposed to how we analyse reaction to the
behavior. Socially, of course, it is of interest that
the speaker, or some other speakers, describe the behavior as a mistake.
But in terms of "heavy" stuff like human cognition, innate
capacities, psycho logical reality, and so on, I think what need
other ways of testing
what the behavior means in terms of linguistic rules and human capacities.
Beyond that, I have no idea what the discussion, which I entered late,
is all about, and what is disputed. Whether "speech error" is a viable
concept?

Vicki Fromkin suggests, in effect, that one can buy
mental realism in linguistics, without buying innate
structure. At first glance, this is a reasonable idea.
In fact, it's one I was wont to hold. But I now wonder
whether the two really are independent.

The central problem, I think, is what we're to count as
knowledge of language if we don't posit an innate language
faculty. This issue has, of course, been raised by Chomsky
many times. I'm coming to think that he's right.

Consider the following cases of "knowledge":

1. Anaphors must be bound in their governing category
2. Poodles are a kind of dog
3. "The President of the U.S." refers to George Bush
4. In Montreal, it's illegal to display English-only signs
outdoors
5. It's rude to say "shit" in a formal gathering
6. The French pronoun "vous" is more formal than the pronoun "tu"
7. A brown house is brown on the *outside*
8. Ideas are not the sort of things which are colored
9. A person's handwriting style is not related to their
philosophical abilities
10. In conversation, one should have sufficient evidence
for one's statements

Which of these are "knowledge of language"? I presume this is
something we discover, by doing research. It's not
just a matter of decision. Now, if there is a language faculty,
and if it incorporates information about some of 1 - 10, but not
about all of them, then *that* is what determines the boundaries
of linguistic knowledge. But if there is not innate structure
that develops, and ends up in a certain state, then I at least
can't imagine how we'd *discover* which of 1 - 10 is "truly"
knowledge of language, rather than just knowledge about the world.

The lines we draw between syntax, semantics and pragmatics;
between competence and performance; between language faculty
and the conceptual system; all can be drawn if the language
facutly is an innate structure. But how are they to be even
sketched if knowledge of language is distributed across
cognitive systems?

Best,
Rob Stainton