|
INCIDENTAL
WORD LEARNING BY TWO-YEAR-OLDS
Deena
Skolnick* and Anne Fernald
[*Honors student at Stanford University; now a graduate student
at Yale University]
Adults
encountering a novel word are frequently able to infer the
referent of the new word using their knowledge of the meanings
of other words in the sentence in combination with other contextual
cues. For example, an adult directed to "use the flogger
on top of the varnish can" to complete a particular painting
task would probably choose the correct brush without further
clarification. At the early stages of language acquisition
such indirect learning is too difficult, and infants benefit
from the ostensive definitions of novel words frequently provided
by caretakers. However, by the end of the second year many
observations suggest that children begin to learn new words
indirectly, using linguistic and contextual cues other than
ostension to identify a new word's referent. Most experimental
studies of infant word learning have focused on how children
learn words taught ostensively. While such studies give us
valuable insight into children's early learning abilities,
they reflect only one way that children learn words "in
the wild." Here we explore the ability of young 2-year-olds
to learn the name of a novel object that is mentioned incidentally,
in a sentence directing the child's attention to a familiar
named object.
The
goal of the study was to determine whether 28-month-olds (n=32)
can learn the word "noopa" when it is only presented
incidentally in association with a known object name, as in:
"Look at the doggy on the noopa". Participants were
tested in a looking-while-listening procedure, in which they
saw one or two pictures accompanied by a recorded voice labeling
objects in one picture. Children's eye movements were videorecorded
during the test session and coded frame-by-frame off-line
to a resolution of 33msec. Four teaching, 4 distracter, and
10 testing trials were interspersed with 10 filler trials.
Trial
types:
Teaching:
Children saw a picture of a familiar object on or next to
a novel target object, which was labeled indirectly with stress
on the familiar noun: e.g."Look at the baby by the noopa".
Distracter:
Children saw a familiar object on or next to a second novel
object; the familiar object was labeled but the novel distracter
was never named, e.g. "Look at the doggy over there."
Testing:
Children saw both novel objects on two screens and were asked
to "Find the noopa."
Filler
trials: No novel objects were present
Speed
and accuracy of novel word recognition were measured in terms
of latency to shift from distracter to target picture and
proportion of time spent looking at the target picture.
Preliminary
results indicate that 28-month-olds are able to map a novel
word onto a novel object presented in conjunction with a familiar
named object, although the novel object is neither singled
out nor labeled directly during teaching. This study extends
our knowledge of early word learning beyond the common learning-through-ostension
paradigm by exploring incidental learning of new vocabulary,
a process that becomes increasingly important in lexical development
beyond the second year.
|
