Gibbs 1987

Reference:

Under construction

Review

Excerpts

'My purpose in this article is to assess the role of mutual knowledge and beliefs in a psychological theory of conversation inference. I will attempt to support what I call the mutual knowledge hypothesis, which assumes that listeners use the knowledge and beliefs they share with speakers in the process of interpreting utterances in conversation' (562).

Domains: Under construction |

Key Terms: | conversation | hypothesis | inference | knowledge | purpose | role | theory |



'This original definition of mutual knowledge raises a serious paradox. The infinite series of beliefs statements is seen as necessary, although it is highly unlikely that listeners can compute an infinite series of these propositions in a finite period of time' (565).

Domains: Under construction |

Key Terms: | definition | knowledge | paradox |



Gibbs introduces the following example:
'He: Can Susan drive a Buick?
She: She can drive any car.' (566).

Domains: Under construction |

Key Terms: |



'The issue of what kinds of experience constitute "knowing" or "assuming" as opposed to merely being "manifest" is a delicate one. Yet making a distinction between something being "known" and other things being "manifest" to someone might artificially create a difference where none exists' (568).

Domains: Under construction |

Key Terms: | difference | distinction |



'My main contention, then, is that Sperber and Wilson are "sneaking" mutual knowledge in the backdoor of their theory of conversational inference by appealing to the idea of mutual cognitive environments which can be manifest but not known. At a psychological level, it appears that Sperber and Wilson have adopted a framework for describing verbal communication which crucially depends on the very concept that they wish to abandon' (569).

Domains: Under construction |

Key Terms: | communication | concept | contention | framework | idea | inference | knowledge | level | theory | verbal |



'The first task facing a listener in understanding an utterance, according to Sperber and Wilson's relevance hypothesis is to identify its propositional form.... They assume that the initial parse of a sentence begins in some sort of input module ... which results in its literal, context-free, representation' (569-70).

Domains: Under construction |

Key Terms: | form | hypothesis | listener | literal | relevance | representation | sentence | understanding | utterance |



Gibbs claims that 'These results [of Ortony et al. (1978) on comprehension of metaphors] strongly demonstrate that context plays a crucial role in the interpretation of metaphoric language. With sufficient linguistic and social context, people understand the nonliteral interpretations of metaphoric utterances directly without first analyzing their putative literal meanings' (571-2).

Domains: Under construction |

Key Terms: | comprehension | context | interpretation | language | literal | role |



'Even if one assumes that mutual knowledge is not necessary for comprehension, there is much psychological evidence which is not in accord with many of the basic assumptions underlying the relevance hypothesis. The weight of this evidence suggests that it may be premature to accept the relevance hypothesis as a reasonable model of the psychological processes used in working out conversational inferences' (574-5).

Domains: Under construction |

Key Terms: | comprehension | evidence | hypothesis | knowledge | relevance |



'Clark and Marshall (1981)... have attempted to show that mutual knowledge can be established in practice by arguing that the apparent paradox of mutual knowledge is base on two incorrect assumptions. The first is the assumption that mutual beliefs must be represented in a mental model as an infinite series of belief statements'. What they propose is that 'if A and B make certain assumptions about each other's rationality, they can use certain states of affairs as a basis for inferring the infinity of conditions all at once' (575).

Domains: Under construction |

Key Terms: | assumption | basis | infinity | knowledge | paradox |



'Perhaps the most fundamental difference between the mutual knowledge hypothesis and the relevance hypothesis concerns the constraints each places on the kinds of inference generated during utterance interpretation. Consider the following exchange:
'He: Are you going to the party tonight?
'She: I hear Jack's coming.' (579).

Domains: Under construction |

Key Terms: | difference | hypothesis | inference | interpretation | knowledge | relevance | utterance |



'Although Sperber and Wilson (1986) comment that it is easy enough to modify [their] definition [of communication] ... and make intentionality a defining feature of communication, it is not at all clear how this modification can be made without acknowledging the specific role mutual knowledge has on the determination of ... inferences' (581).

Domains: Under construction |

Key Terms: | communication | definition | determination | intentionality | knowledge | role |



'When listeners are addressed by someone, they usually assume the speaker has done his/her best to enable them to understand him/her' (582).

Domains: Under construction |

Key Terms: | speaker |



'People will design their questions in light of the mutual knowledge which exists between themselves and their listeners' (584).

Domains: Under construction |

Key Terms: | knowledge |



'My arguments is favor of the mutual knowledge hypothesis and against the relevance hypothesis are grounded in five interrelated observations. First, I have suggested the Sperber and Wilson's proposal that mutual cognitive environments constitute the true context for comprehension is not sufficiently clear and distinguishable from the concept of mutual knowledge. As such, the relevance hypothesis seems to make use of the very idea that it attempts to replace. Second, mutual knowledge is possible to determine in a finite period of time via Clark and Marshall's (1981) mutual knowledge induction scheme without resorting to an infinite set of beliefs statements usually viewed as a consequence of establishing mutual knowledge. Third, it appears that mutual knowledge is indeed a necessary prerequisite for the comprehension of many kinds of utterances in conversation. This is particularly true if listeners are to distinguish between inferences that are ostensively intended or "authorized" by speakers from inferences that are "unauthorized". Conversations are only cooperative to the extent to which speakers specifically intend and listeners specifically recognize "m-intended" messages. Part of my thesis here is that how listeners are able to distinguish "authorized" versus "unauthorized" inferences must be part of a cognitive theory of conversational inferences. Fourth, parts of the processing model underlying the relevance hypothesis are not supported by contemporary psycholinguistic research. Specifically, there is little empirical evidence in favor of the idea that listeners must first decode an utterance into some propositional representation before choosing a context in which that proposition is viewed as most relevant. Finally, there is some recent psycholinguistic evidence demonstrating that speakers formulate their utterances precisely to satisfy the amount of knowledge they share with their listeners. This shared knowledge is also directly utilized by listeners when interpreting utterances in everyday discourse. These findings appear most congruent with the predictions of the mutual knowledge hypothesis' (585).

Domains: Under construction |

Key Terms: | comprehension | concept | consequence | context | conversation | discourse | empirical | evidence | hypothesis | idea | knowledge | proposition | relevance | representation | research | theory | thesis | utterance |



Last Modified: July-12-96 9:20:24

Reply to randy_radney@sil.org

[A Lexicon of the Humanities | SIL Home Page | Contributions]