Research Design -- drawing pictures or phtos ?

[Gedeon De ák]

Greetings--
The shape bias for count nouns is a highly
context-dependent phenomenon. This is one line of
evidence for the argument that constraint-based
accounts can't adequately explain word learning
(covered in G. Deák, 2000, Developmental Review, 20,
29-80).
As one example of context-dependency, certain
sentential contexts draw children's attention away
from similarity of shape when generalizing novel count
nouns, as shown in G. Deák, 2000, Journal of Cognition
and Development, 1, 157-192.
One reason SOME studies (not all) find an apparent
shape bias is that most pictorial stimuli tend to
artificially enhance similarity of shape (or a
2-dimensional projection thereof). This enhancement
makes preschoolers seem more shape-biased, as shown by
G. Deák & P. Bauer, 1996, Child Development, 67,
740-767.
The point is not that object shape is unimportant--of
course it IS--but that it is best thought of as one
object regularity that can flexibly be attended-to in
the service of certain kinds of responses or
inferences, e.g. about possible word meanings.
Children will learn to attend to shape in CERTAIN
word-learning situations, if past inferences about
words of the same type, in similar contexts, have
pointed to shape as a diagnostic feature. This does
not mean children represent object word meanings as
shape-based.
Of course, there are many other interesting relevant
sources, including papers by Linda Smith, Susan Jones,
Barbara Landau, Dedre Gentner, Susan Gelman, and
others; references to many of these (through 2000) can
be found in the papers cited above.
Good luck--
Gedeon

--- "UG, NCKU :)" <eugenew45 at yahoo.com.tw> wrote:

> Dear Info-CHILDES members,
>    
>   I'm now conduct a research in which normal hearing
> and hearing-impaired children's cognition will be
> tested respectively based on three constraints of
> lexical learning (shape, whole object, and
> taxonomic) by means of matching.
>    
>   When matching, the chidren have to pick one item
> from the two to match the standard one. One of the
> two items are shape-like as the standard one while
> the other has certain semantic relationship as the
> standard. Therefore, the following items is a set:
> Butterfly-Hair bow-Tiger, in which butterfly is the
> standard one.
>    
>   And my research question is, 
>   In the lexical development for hearing-impaired
> children in Taiwan, is the noun-category bias really
> a noun-shape bias?
>    
>   And the problem I come up with now is which kind
> of following material I sould use when conducting
> such experiments -- drawing picture or real photo?
>    
>   The previous literatures I have read adapt drawing
> pictures as materials.
>   And the current research is mainly based on:
>    
>   Poulin-Bubois, D., Klein, B. P., Graham, S. A., &
> Frank, L. (1993). Is the noun-category bias a
> noun-shape bias? In E. V. Clark (Ed.), The
> proceedings of the twenty-fifth annual child
> language research forum (pp. 221-226). Stanford:
> Stanford Linguistics Association.
>    
>   Many thanks.
>    
>   Best, 
>   Hsin-chin.
>    
>   Hsin-chin Wang
>   Graduate Student
>   College of Liberal Art
>   Foreign Language and Literature Department, NCKU,
> Taiwan
>   eugenew45 at yahoo.com.tw
>   eugenew.languag at msa.hinet.net
> 
> ___________________________________________________ 
> ³Ì·sª© Yahoo!©_¼¯§Y®É³q°T 7.0¡A§K¶Oºô¸ô¹q¸Ü¥ô§A¥´¡I 
http://messenger.yahoo.com.tw/


Gedeon O. Deak, Ph.D.
Department of Cognitive Science
9500 Gilman Dr.
University of California, San Diego
La Jolla, CA 92093-0515

ph:  (858) 822-3352
fax: (858) 534-1127
e:   deak at cogsci.ucsd.edu
http://www.cogsci.ucsd.edu/~deak