SITUATIONAL VARIABILITY IN THE SPEECH OF PRESCHOOL CHILDREN *

A common observation among psychologists, dating back at least to the 1920s (cf. Issacs, 1930; Piaget, 1923; Vygotsky, 1978), is that children appear to talk in a more sophisticated way and to accomplish more complicated intellectual acts in the course of spontaneous interactions with their social and physical environments than they do when they are being interrogated by adults. There are, generally speaking, two lines of explanation for this observation. Arguing from somewhat different developmental perspectives, Piaget and Vygotsky both claimed that spontaneous problem-solving was, in general, easier than problem-solving at the instigation of adults, which requires that the child respond in terms of special rules that restrict her response options in various ways. Although she did not take up this topic in a formal fashion, Issacs interpreted performance differences associated with spontaneous and elicited problem solving differently. She believed that the young child performed better in spontaneously generated, everyday interactions because the contents of those interactions are things the child knows and cares about (in contrast with the content of adult questions used for cognitive assessments, which are arbitrary, uninteresting, and perhaps alien in content as well). This issue has taken on considerable contemporary importance for both practical and theoretical reasons. Practically, the search for cognitive assessment devices that would be valid across different ethnic and language groups has led to repeated reassessments of the standardized cognitive and linguistic tests used to screen young children entering educational programs or for evaluation of those programs (cf. Raizen and Bobrow, 1974). A common claim by critics of standardized tests echoes Issac’s dissatisfaction with Piaget many years ago: in the modern context, it is claimed that minority-group youngsters know less and care less about the contexts of standardized tests than do their Anglo, middle-class counterparts. In effect, the argument goes, tests are biased against minority group youngsters because of test content and language. There is also widespread feeling that the social interactions embodied in standardized testing situations are themselves a source of performance differences. This latter claim has been made most forcefully by William Labov, who stated that failure of Black English Vernacular speakers is not the result of a deficit in the children but of failures in the educational system itself: (1) the failure to take into account the child’s dialect; (2) the failure to assess Black children’s intellectual competence accurately; and (3) the failure to implement policies to remedy (1) and (2).