Language Learning Principle(4)
언어 학습 원리
https://brunch.co.kr/@eduinsight/55
https://brunch.co.kr/@eduinsight/57
https://brunch.co.kr/@eduinsight/59
In Experiment 2, we investigated whether infants could identify regularities in two speech streams composed of the same sounds. The infants were divided into two groups: one group was in the "Consistent Speaker condition," where the same speaker presented both streams, while the other group was in the "Variable Speaker condition," where different speakers presented the streams.
In this experiment, the syllables were identical, making it impossible to distinguish the sounds. We hypothesized that the additional information from a "Consistent Speaker" would help infants recognize patterns in the two streams. Therefore, we predicted that infants would need this information and that only those in the Consistent Speaker condition would discover the regularities. Conversely, there was also a possibility that infants relied on auditory cues and would struggle to differentiate the two streams without this additional information.
Ultimately, this experiment aimed to shed light on how infants understand language, particularly what information is important for identifying patterns within complex sounds.
The results of the ANOVA (Test Type × Condition) for Experiment 2 showed that, unlike in Experiment 1, there was no significant main effect of test type (F(1, 38) = 0.68, p = 0.42). The interaction between test type and condition was also not significant (F(1, 38) = 0.32, p = 0.58). Infants did not look significantly longer at Familiar versus Unfamiliar items in either condition. In the Consistent Speaker condition, infants looked at Familiar items for an average of 6.1 seconds and Unfamiliar items for 5.5 seconds (t(19) = 1.09, p = 0.29); in the Variable Speaker condition, they looked at Familiar items for 6.7 seconds and Unfamiliar items for 6.6 seconds (t(19) = 0.17, p = 0.87).
In the Consistent Speaker condition, 12 infants and in the Variable Speaker condition, 11 infants showed a preference for Familiar items. When the Target and Nontarget speech streams consisted of the same sounds, infants did not learn the Target regularities. Even when the speaker information was perfectly aligned with the structure in the Consistent Speaker condition, infants failed to utilize this association. This indicates that indexical cues alone were insufficient to highlight the underlying structure in the two speech streams.
References
Potter, C. E., & Lew-Williams, C. (2019). Infants' selective use of reliable cues in multidimensional language input. Developmental Psychology, 55, 1-8.
