|
Participants were eighteen four-year-old subjects (mean age 4;5) from the
Champaign-Urbana area. All subjects were native speakers of English.
30
Part I: Language Processing and First-Language Learning
Table 2.5.
Participants performing significantly above chance (
N = 16)
Nonsense
Vague
Verb
Noun-Absent
Noun-Present
Number
15/16
15/16
15/16
1/16
1/16
% of Total
93.7
93.7
93.7
6.25
6.25
Content made available by Georgetown University Press, DigitalGeorgetown,
and the Department of Linguistics.
Materials
The illustrations and story from Experiment One were modified so that only
four objects appeared in each illustration, and all four objects were named in the cor-
responding story. The answer sheets included eight objects, as in Experiment One.
Explanation of Conditions
Experiment Two included noun, verb, and nonsense conditions
that were identical to the noun-absent, verb, and nonsense conditions in Experiment
One. In addition, there was a foil condition in which a noun was named that did not
appear in the illustration of the story. Order of word type was counterbalanced across
trials. The answer sheet included the three novel objects or creatures that were re-
ferred to in the first three conditions, an object representing the primary meaning of
the noun homonym, the object named in the foil condition, and three novel objects
that did not appear in the illustration.
Procedure
The procedure was the same as with Experiment One except that children
were instructed to first determine whether the object Tommy saw in the story ap-
peared in the illustration. For example, children heard the following portion of the
story:
Next to the slide, Tommy saw a pink tree. Way up in the tree, Tommy saw a
[keyword]. “Maybe my baseball is stuck up in the tree too,” Tommy thought.
Tommy climbed the tree and looked all around, but still didn’t find his
baseball.
The experimenter then covered the illustration of the story and said, “Now, Tommy
saw a [keyword] in the story. Did you see one too?” If the child answered yes, the
child was presented with the answer sheet and asked to point to it. If the child re-
sponded no, the experimenter marked the answer as a no-go and proceeded to the
next part of the story.
Results
Statistics were first run to compare the number of no-go answers in each of the four
conditions. Repeated measures ANOVA with post hoc comparisons were performed
using word type (foil
⫻
noun
⫻
verb
⫻
nonsense) as the within-subjects factor. Tests
show that all conditions differ significantly at or below the .01 level. In particular, a
comparison between the noun (M
⫽
2.89, SD
⫽
1.23) and the verb (M
⫽
1; SD
⫽
0.69) conditions reveals a significant difference (F(1,17)
⫽
32.32;
p
⬍
.001), indicat-
ing that children rejected noun homonyms more readily than verb homonyms. The
mean number of no-go answers for each condition is reported in figure 2.5.
Statistics also were run to compare participants’ success at mapping each of the
word types to their object referents. For the purposes of scoring, no-go responses as
well as responses that indicated an object other than the intended referent of the story
were considered incorrect. Again, repeated measures ANOVA with post hoc compar-
isons were performed, using word type (noun
⫻
verb
⫻
nonsense) as the within-sub-
jects factor. The tests found that all conditions differed significantly at or below the
p
⬍
.001 level. Results are reported in figure 2.6.
HOMONYMS AND FUNCTIONAL MAPPINGS IN LANGUAGE ACQUISITION
31
Content made available by Georgetown University Press, DigitalGeorgetown,
and the Department of Linguistics.
The line graph in figure 2.6 shows the trend of increased difficulty in mapping
each word type. As with Experiment One, children are significantly less willing to
map a new meaning to a noun homonym than to a verb homonym. Unlike Experi-
ment One, however, this experiment suggests that children do experience more diffi-
culty mapping verb homonyms than nonsense words (F(1,17)
⫽
17,
p
⬍
.001).
Do'stlaringiz bilan baham: |
