Hide/Show Apps

Overextension as evidence for deductive word learning : a theoretical examination

Ünal, Erdem
This thesis is a philosophical assessment for the status of deductive abilities in infants’ object representation and naming. In the speech of children between 1:0 and 2:5, the application of early acquired words temporarily occurs in a broader range of referents i.e., more or less than conventionally accepted referents. This short period of time, namely overextension period, gives substantial clues in regard to our understanding of children’s linguistic knowledge during one-word stage and their cognitive developments. In cases of overextension, the partial meaning of the word can somehow intersect with the semantic domain of the object labeled by that word in such a way as shape, color, taste, texture or function. Overextension on the basis of the perceptual properties of objects signals the possibility that deductive (logical) inference may play a significant role in the extensions of acquired words. Within the context of overextension as a path of development in the acquisition of language, this study is spirited by two proposals about the earliest word learning in children younger than 3 years of age. One proposal is empirically testable: the first strategy of early word learning is referential that is based on the relationship between the perceptual properties of physical objects and infants’ conceptual knowledge; for creating a general hypothesis about the object-name relation, infants have to attend to basic perceptual features of physical objects with regard to their immaturity on visual perception, attention and semantic memory. Second proposal is theoretical: infants can be viewed as deductive learners, that is, they formulate an overview of word meaning on the basis of one or two perceptual properties of objects and assign names to objects by deductive inferences.