Laying the groundwork for the DELV (Precursor literature, dissertations, joint work of the Working Groups prior to the conceptualization of the DELV)
Designing a test battery for language that is appropriate across dialects is difficult. Many of theitems traditionally studied to assess the sophistication of preschoolers’ speech include inflections such astense and plurality that behave differently across dialects, resulting in the mis-assignment of clinical statusto normal speakers of dialects such as African American English. One strategy towards solution of thisdilemma has been to choose items that involve tests of linguistic knowledge that have no established dialectdifferences. Recent work on articles in English suggests there is promise in this area, since a) there are no established dialect differences in the use of definite and indefinite articles between AAE andSAE.b) SLI children, on the other hand, show significant difficulties using articles appropriately (Ramos, 1999) Given those two premises, it is clear that the article items may be responded to in the same way by childrenregardless of dialect but dependent on their clinical status. The purpose of this study was to establish abaseline with normally-developing children and confirm that there are no dialect differences for 3-7 yearolds.Previous studies of article use have been fraught with methodological difficulties, and considerablevariation depending on the methods and materials. Our analysis of the inadequacies of earlier experimentson this topic (see Cziko 1986) led us to design the task so that no objects relevant to the elicitation werepresent during the experiment. That is, we modified a recent procedure by Schafer and de Villiers (2000),which gets around the materials problem by using none! In effect, the children are asked questions that theyanswer “out of their heads”, a procedure that might also lessen the disadvantage of children who are lessused to test-type book materials. The purpose of the study was to elicit from children noun phrasescontaining the articles a and the under conditions that controlled for the maximum number ofextralinguistic contributions to familiarity and uniqueness.