Conversational Implicature in Children's Referential Communication

77 p.Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1980.Contemporary theories of language comprehension emphasize the importance of listeners' active, cognitive contributions to message meaning. A major class of contributions is the listener's knowledge of the conventions of language use (Rule Knowledge). H. P. Grice has described some abstract conventional structures which serve as both a productive and interpretive framework for conversation: the Cooperative Principle (CP) and its maxims of Quality, Quantity, Relation, and Manner. That children's referential communication deviates from the CP is very well-documented: the Quantity Maxim, which requires speakers to give as much information as the listener needs but not more, seems lacking in both the speaking and listening skills of grade school children. As speakers, children often give descriptions which do not provide the listener with enough information to make unambiguous choices, and as listeners, they seem to be unaware that they lack important information. The dominant view of children's comprehension is that they remain unaware of message ambiguities and so respond idiosyncratically. This thesis attempts to show that grade school children do in fact apply conventional stuctures like the Quantity Maxim in their efforts to make sense of poor messages.First- and third-grade children from six schools played a referential communication game with an adult in which the object was for one partner to provide a message adequate for the other to pick one specific object from a six-object array. The adult experimenter gave two messages, both of which were purposely left indefinite, so that they applied to either two or three of the six objects (Quantity Maxim violations). One of the two messages was constructed in such a way as to encourage children to make interpretations based on conversational implicatures, while the other was constructed so as to preclude similar implicatures. Children in both grades treated the implicature-yielding message as significantly less ambiguous than the non-implicature-yielding message. The particular object choice predicted by the implicature was significantly and overwhelmingly preferred to the other two possible choices, for all but the first-graders from two schools. Children's explanations of their choices reflected awareness of conventional structures like the CP.Children's Rule Knowledge seems surprisingly adult-like by early grade school. Children's comprehension deficiencies seem attributable to undeveloped World Knowledge or to the sort of general processing constraints described by Marilyn Shatz.U of I OnlyRestricted to the U of I community idenfinitely during batch ingest of legacy ETD