[ CogSci Summaries home | UP | email ]

Hunt, E. & Agnoli, F. (1991) The Worfian hypothesis: A cognitive psychology perspective. Psychological Review, 98 (3). pp 377-389.

Author of the summary: Debajyoti Pati, 2000, gte811q@prism.gatech.edu

During the 1920s and 1930s, Benjamin Lee Whrof, and amateur linguist propounded his linguistic relativity hypothesis. The stronger version of this hypothesis argues that language controls both thought and perception. There is a weaker version whose argument was that language influences thought. Both the versions have, ever since, been rejected as false or unprovable. More particularly, Chompsky"s trend towards study of linguistic universals resulted in little attention to this area.

Making use of the research works in various areas since the 1930s, Hunt and Agnoli attempt in this paper to take another look at the linguistic relativity hypothesis. One of the major criticisms of the Whorfian hypothesis focused on the issue of intertranslability. It is generally agreed that statements could be translated from one language to another, thus making the Whorfian hypothesis weak. Similarly, the weaker version suggests that thoughts easily expressed in one language has the possibility of not being produced at all in some other. While both the possibilities seem weak , a separate line of approach appears to produce evidence in support of the hypothesis. This approach takes into calculation the number of decisions a person has to make while choosing a word or to construct an utterance (an analogy of computational models). One case in point is the coding conditions, which place a demand on the user"s psychological capacity depending on the language used. Recognizing and selecting lexical terms, and analyzing structures place certain demand on the long term and short term memory. This suggests that the language user thinks most efficiently about topics which has efficient codes provided by the lexicon. Whorf believed that the influence of the grammar of a language is a more important determinant of thought than the categorizations of the lexicon. While many argue that the complexity argument cannot be tested, the authors believe that it is possible if a minimal size effect could be chosen and tested. To view it differently, if it is possible to find cross linguistic effects are as large as intralingual effects, the Whorfian hypothesis could be tested.

In order to look at the possible effect of language on thought, the authors first concentrate on representational level thinking, where two sources of information seem particularly important for this area of study; the lexically identified concepts, and the culturally developed schema. They argue that people consider the cost of computation when they reason about a topic, and different languages involve different cost for transmission of messages, and thus language influence cognition.

In the area of lexicon, they argue that differences in lexicon in two different languages influences how users structure the same experience. They identify two effects; the direct effect (the discrimination that a person must make while choosing or comprehending a word) and the indirect effect (the semantic relation between the word chosen and the other words in the speaker"s lexicon). In direct effects, based on research works by Carroll and Casagrande (1958), and Luci and Shweder (1979) the authors point out that memory is not to language. It is partly based on linguistically based records of the people"s description of the events to themselves, and because they are coded by language, memory is subject to bias depending on the person"s language. Palmer"s (1974) study concerning eye-witness memory displays such a bias. Similarly, Lay and Kempton"s (1984) study illustrated the effect of language on color perception (variations in categorical perception across languages). The indirect effects could be observed in polysemous words, where the current studies have shown that all meanings of a word are activated, and a correct meaning is selected based on context. Citing example studies of English and Italian words, where the former were shown to be significantly more polysemous, the authors argue that it exerts influence on thought. Similarly, anaphoric references seem to have varying influence across languages.

In another area of language, the syntactical-semantic effects are shown to be potential candidates in support of the Whorfian hypothesis. One good example cited is the structure of the English sentences as compared to Italian an German. Several studies suggest that different language use different cues for parsing, and speakers of different language show wide variations in their use of different cues. The studies of bilinguals provide additional evidence for the differential strengths of cues. The authors argue that the difference in ambiguity among languages could be forcing different styles of reasoning on the speakers. Yet another area contributing to similar ambiguities is parsing. Here too different languages seem to differ in utterance ambiguities. The examples show that languages impose different cognitive burden on speakers, and to support the hypothesis, that languages affect how speakers think about the nonlinguistic world, is the fact that linguistic and nonlinguistic reasoning often occur concurrently and affect each other.

The Whrofian argument that languages affect ones interpretation of the world around us, could be observed in the study of schema in different languages. The differences in number schema as well as spatial schema show the possibility of such differences having an impact on thought. Moreover, the differences produced by different coding systems, social ranks, social perception, etc, across languages seem to suggest considerable support to the Whorfian hypothesis.

Back to the Cognitive Science Summaries homepage
Cognitive Science Summaries Webmaster:
JimDavies (jim@jimdavies.org)
Last modified: Wed Apr 26 11:23:15 EDT 2000