Page 110 - Encyclopedia of Philosophy of Language

Language and Mind
for a few minutes, but are there more long-term influ- ences of language on significant areas of our cog- nitions? One fruitful source of evidence is in the presuppositions people bring to the interpretation of utterances. If this author talks of a surgeon many listeners in his culture will assume he is referring to a man, even though this is not explicitly stated, nor is it necessarily true. If he says Jeff is a worse player than George he is suggesting that both Jeff and George are bad players, otherwise he would have used the more neutral George is a better player than Jeff. Dis- entangling the distinctive contribution of language is tricky: in the author's culture most surgeons are male, and this is a fact about the culture, not about language, so it is not easy to tell whether language is relevant here.
One approach is to use problems which are neutral with respect to culture. Noordman (1978) gave Dutch students problems such as A is thefather ofB, A is the grandfather ofC, B is not thefather ofC, What relation could B be to C? The students gave predominantly male answers, and in particular they more often chose the correct answer uncle (50 percent of choices) than the correct answer mother (20 percent of choices). The pattern changed when mother and grandmother replaced father and grandfather in the problem: here there was a bias to give female answers, though it was not so strong as the male bias in the male version of the problem, and in particular the correct answer
father (38 percent) was chosen more often than the correct answer aunt (30 percent). These results show that problem solvers use the language of the problem to create a representation which may be incomplete: a problem in which only one gender is referred to may lead to a representation in which all the possible solutions may have the same gender. However, the greater bias exhibited when all the terms are male than when they are all female may have cultural, not linguistic, roots (one is more used to reading exam- ination problems which refer to he than she).
Gender biases could be viewed as an example of the more general phenomenon of 'markedness.' Here bipolar adjectival pairs such as good/bad, tall/short,
fast/slow are not considered to be symmetrical, but the preferred 'unmarked' member of the pair does double duty, both indicating a particular pole and naming the entire dimension (this 'neutral' aspect of the unmarked adjective can be seen in such phrases as sixfeet tallandHowfast isyourtyping?). Clark(1969) showed that problems involving unmarked adjectives are easier than their marked-adjective counter parts (If John is better than Pete, and Pete is better than Dick, then who is best? is solved faster than If John is worse than Pete, and Pete is worse than Dick, then who is worst?). A full interpretation of this result is controversial, since it is not known whether it derives from some fundamental property of linguistic struc- ture or from the more mundane observation that
marked adjectives have greater general frequency of usage in the language than unmarked adjectives; nonetheless Clark's results show that problem solvers prefer to create representations based on one set of linguistic labels rather than another.
A further example of the influence of labels on thought is the phenomenon of 'functional fixedness.' Subjects generally find it difficult to solve problems requiring them to use objects in novel ways, such as using a spanner to complete an electrical circuit: the normal function of the object appears 'fixed.' This phenomenon is enhanced if the experimenter uses the label spanner in presenting the problem: the label appears to encourage subjects to create an inap- propriate representation for the problem.
The most contentious area concerned with the effects of labeling on thought is that of prejudice. There is a choice of labels to refer to an individual (nigger/negro/colored/black/Afro-Caribbean; queer/ poof/gay/homosexual; chick/crumpet/date/escort/girl-
friend). The label chosen undoubtedly reflects some of the attitudes of the speaker, but one can also ask whether use of a label can shape attitudes—does referring to a person who presides over a committee as a chairman lead one to expect that this person should be a man? In this author's culture, a 25-year old woman is often referred to as a girl, whereas the corresponding term boy is much less frequently used of a 25-year old man—idoes this influence attitudes to people referred to in this way? The evidence suggests that such influences exist. For example, subjects who were asked to describe the images suggested by chap- ter headings in a sociological text were more likely to report images containing only males when headings used generic man (Industrial Man) than when gender was not mentioned (Industrial Society). Kitto (1989) composed short references for hypothetical applicants for jobs. All the applicants were females aged 25. Subjects (mainly university students) preferred appli- cants whose reference referred to them as girl for the low-status job of waitress, but they preferred appli- cants referred to as woman for the higher-status job of personal assistant. Subjects commented that the persons described in the girl references were livelier but less reliable or competent than the persons referred to in the woman references: this was true even for subjects who appeared unaware that their attitudes were being manipulated by the presence of girl or woman. One way of looking at these results is that girl elicits many presuppositions including those associ- ated with the 'proper' use of the term to refer to a female of school age—such females are typically live- lier but less competent than their adult counterparts. Society will not change overnight by banning the use of words like girl (when applied to an adult) and chairman, but it is important to realize how these terms may provoke prejudices which are all the more insidious for our not always being aware of them.