Exploring the role of input and exposure on infant word segmentation

Schreiner, M. S. & Mani, N.

Georg-August-University Göttingen

While most research on infant word segmentation has investigated the extraction of words from IDS in standard laboratory settings, the current series of experiments examined the role of different kinds of input and different kinds exposure on infants? word segmentation abilities. The first experiment suggests that infants are able to successfully segment words from fluent IDS and ADS provided they were familiarized with these words over an extended-exposure period of six weeks at home. These 9-month-old infants, however, seem to be unable to segment words from fluent IDS during a standard laboratory-familiarization. Therefore, the second experiment examined whether German infants might require a more exaggerated IDS exposure similar to the one American English infants are addressed with. Here, 9-month-old, but not 7.5-month-old, infants successfully segmented the words presented in an exaggerated IDS register. Using neurophysiological measures, the third experiment further explored 7.5-month-old infants? segmentation abilities and revealed that these infants successfully segmented words from both, fluent exaggerated and standard German IDS. Hence, the current paper provides new insights into infant word segmentation from different kinds of input and different kinds of exposure. Furthermore, it supports the idea of attention as being a central mechanism in early language acquisition.