The basic problem of segmentation is readily apparent to anyone listening to speakers conversing in an unfamiliar language: what one hears is a babble lacking any readily identifiable words. Thus, an initial task facing infants acquiring language is to learn how to recover individual linguistic units, such as words or phrases, from continuous speech. This task in nontrivial because these units are not physically separated from one another by pauses, but rather flow into one another in the speech stream. The development of segmentation skills is prerequisite to further acquisition of language, for if a language learner cannot break input utterances into their constituent parts, it will be impossible to learn how these parts fit together or what individual parts mean. Units emerging from segmentation serve as input to further stages of analysis, and the nature of analytic capacities recruited in language acquisition is dependent upon the nature of these units. Despite the clear importance of segmentation, however, little is known at present about the nature or development of requisite perceptual capacities. The research proposed here utilizes two novel extensions of the conditioned head turning technique widely used in studies of infant speech perception to provide converging evidence on the nature of infants' auditory perceptual units. The proposed studies comprise three groups. First, a set of studies will examine contributions of individual rhythmic and intonational cues to 6- to 10-month-old infants' segmentation of continuous speech. Second, studies using cue-conflict paradigms will examine relative contributions of multiple cues to infants' segmentation of continuous speech. Finally, a pair of studies will examine infant segmentation of natural speech samples drawn from familiar and unfamiliar languages, providing information pertaining to the role of experience with specific languages and to the external validity of the above studies. Results of the proposed studies will contribute baseline and comparative information pertinent to future investigation of further questions, including: How do perceptual abilities recruited for segmentation develop across the first year? Are the abilities recruited for segmentation specific to speech or are instances of more general auditory capacities? To what extent do modifications characteristic of child-directed speech aid in solving problems of segmentation? Does partial sensory loss (e.g. due to chronic otitis media) adversely affect infants' abilities for segmenting speech? Do individual differences in early segmentation abilities predict aspects of later language development?
|Morgan, J L; Saffran, J R (1995) Emerging integration of sequential and suprasegmental information in preverbal speech segmentation. Child Dev 66:911-36|