home
events
contacts
mailing list
directions
Linguistics
Department
Stanford
University |
Stanford
Humanities Center
Mellon
Foundation
Graduate
Research Workshop Program
Stanford
Semantics and Pragmatics Workshop:
THE
CONSTRUCTION OF MEANING
Tuesday, May 11, 5:30pm in 460-126:
Causative Alternation Errors and Innate
Knowledge: Consequences of the 'No Negative Evidence' Fallacy
Jean-Philippe Marcotte (Stanford)
Pinker (1989) argues that children's causative alternation errors ("Don't
fall me down", "The ball didn't throw in") must be a manifestation of
innate grammatical knowledge if there is to be an explanation for the fact
that children eventually attain correct knowledge despite the absence of
negative evidence. I argue in my dissertation that negative evidence is no
more or less prevalant in children's environment than positive evidence,
and that the knowledge Pinker claims must be innate can in fact be
attained through the interaction of the nature of the target knowledge and
the nature of child/adult interactions.
Children's ability to detect shared meanings between their own utterances
and adult ones is an unspoken but crucial precondition to obtaining even
positive evidence. Both positive and negative evidence are the outcome of
a process of comparison between the child's parse of an adult utterance in
its context, and a child-generated representation expressing the same
meaning in that context. Matches yield positive evidence, mismatches yield
negative evidence.
A construction paradigm approach, in which linkages between verbs and the
constructions in which they can enter are determined by the meanings of
verbs, accounts for both the semantics of argument alternation errors and
their bidirectionality, whereas Lord (1979) fails in the former respect
and Bowerman (1974, 1982) in the latter. I provide a theory of language
use in which linguistic representations of meaning are divorced from, and
refer to, conceptual representations of events (Talmy 2000). It emerges
from this that children acquire erroneous verb meanings (Behrend 1990) in
just such a way that they hypothesize incorrect construction paradigms,
yielding a possibility of error. A verb is likely to be used erroneously
in a construction when properties of the event to be denoted favor the use
of that construction (Wolff 1999, 2003).
Preliminary results are consistent with this approach: naturally occurring
causative alternation errors (n=222), gathered from published and
unpublished diaries, are restricted to verbs for which it is predicted
that erroneous meanings are acquired, and arise when the construction with
which the error is made is favored by the context.
Please contact one of the workshop organizers
if you have suggestions for presentations or the workshop in general.
Back to the workshop homepage.
This workshop is sponsored by
the Stanford Humanities Center, and funded by a grant from the Mellon
Foundation.
|