UBC Theses and Dissertations
The use of contextual information and the understanding of subordinate and superordinate elements as a function of young children’s comprehension of syntactic and semantic factors Blenner-Hassett, Marnie
Children of ages two, three and four years and adults (N=24 at each age level) were given a comprehension task which involved pointing to the picture, from an array of four pictures, that matched a verbally given sentence. There were two sets of sentences, one consisting of probable and improbable sentences, which could be both active and passive with the improbable sentences forming a hierarchy of improbability depending on the abstractness of the sentence objects. This set was designed to examine young children's understanding of subordinate and superordinate elements expressed as functional relations. The second set of sentences were reversible and non-reversible, probable and improbable, and active and passive. This set was designed to investigate interactions among the three factors of reversibility, probability and voice and, to determine whether young children were able to handle semantic change in sentences more or less easily than syntactic change. Both sets of sentences were employed in examining children's use of contextual information in the understanding of sentences. It was found that subjects (Ss) at all age levels were able to distinguish probable from improbable sentences, implying that children as young as two years are able to employ contextual information in the comprehension of sentences. It was also found that there was a greater difference in performance between probable and improbable sentences at the two most abstract levels of the improbability hierarchy indicating that the subject and object in these sentences are in separate subjective categories. This supports Anglin's (1970) position that children in the early stages of language acquisition have a large number of semantic elements, and as the child's semantic knowledge increases, these are combined into larger and more abstract categories. One interaction was found on both hierarchical and non-hierarchical sets of sentences, between a syntactic and a semantic factor, which indicated that in the case of probable sentences, active voice is easier than passive, and for active voice, probable sentences are easier than improbable. This suggests that the syntactic factor is only stronger or more difficult than the easier semantic element and the semantic factor is only stronger or more difficult than the easier syntactic element, indicating that these two factors are of much the same strength. Other findings regarding the syntactic factor of voice and the semantic factors of probability and reversibility, are extremely tentative, as they are based only on particular subgroups of sentences and often found only for males or females at particular age levels. The errors made were examined and it was found, generally, that children most often reversed the actor and object, and infrequently chose a sentence with a different verb or a sentence with a different verb and a different actor. It was also found that the performance of two-year-olds, both males and females, was particularly affected by whether the sentence had ‘boy' or 'girl’ as actor and as first noun in the sentence, such as occurs in passives. They performed well on sentences with 'boy' as actor and first-mentioned and poorly on those with 'girl’ as actor and first-mentioned. It was concluded that either a subtle design feature contributed to this finding or it indicates that two-year-olds actually are better able to handle the notion of 'boy' as actor than 'girl’.
Item Citations and Data