The Matched Guise Technique: a Classic Test for Formal Measurement of Language Attitudes, by Marina Solís Obiols

The study of language attitudes is of special importance for sociolinguistics; Formal measurements of these provide us with results that can be used to predict the linguistic behaviour of members of a given social group in terms of their use of linguistic varieties in bilingual and bidialectal situations.
Within the context of the methodology used in the study of language attitudes, this paper focuses on the description and bibliographical criticism of the renowned matched guise technique. Broadly-speaking, this technique involves asking interviewees to evaluate the personal qualities of speakers whose voices are recorded on tape, whereby the same speaker uses different linguistic varieties. This technique has been widely used in studies and research on language attitudes, in both international and Catalan sociolinguistics. (1)

PDF printing version. 87 k

Summary

1. Introduction

2. Language attitudes and linguistic behaviour

3. Direct and indirect methods

4. The matched guise technique

5. Conclusions

6. Bibliography

1. Introduction

As stated above, the study of language attitudes is important for sociolinguistics because it can ‘predict’ a given linguistic behaviour: the choice of a particular language in multilingual communities, language loyalty, language prestige...

However, due to paradoxes of science, sociolinguistics has not undertaken serious research, using either theoretical or methodological approaches, on language attitudes, which are closely linked to the social psychology of language (2). Moreover, the exchange of theoretical and methodological knowledge on language attitudes between these two disciplines has been meagre and insignificant, since sociolinguistics and the social psychology of language have followed parallel paths, except for when obligatory exchanges between the two disciplines have taken place (see Agheyisi & Fishman, 1970; Cooper & Fishman, 1974). (3)

2. Language Attitudes and Linguistic Behaviour

A number of theories have been developed on the study of language attitudes. (4) The two most important include: the mentalist theory and the behaviourist theory, which differ in their understanding of attitude. (5) On the one hand, the mentalist approach sees language attitudes as being mental and neural states of disposition (Allport, 1967) that cannot be observed directly, but that can be inferred using the right stimuli; on the other hand, the behaviourist current considers attitudes to be behaviours or responses to a given situation (Agheyisi & Fishman, 1970).

Nevertheless, authors such as Bierbach (1988) conclude that the differences between these two schools of thought are minimal when it comes to empirical research. (6)

However, one aspect that does differentiate the two currents from a theoretical point of view is the multicomponential or unicomponential conception of language attitudes: for behaviourists, attitudes only have one component – the affective – while for mentalists, attitudes have three components: the affective, the conative and the cognitive. The studies of Lambert and his team (pioneers in the development and application of the matched guise technique) at McGrill University, Canada, on the social psychology of language are based on the multicomponential theory.

The behaviourist approach has a serious scientific disadvantage because the affective component alone cannot predict verbal conduct (or anything else for that matter) (López Morales, 1989), whereas this is not the case with the mentalist approach. As mentalist conceptions are able to predict linguistic behaviour, they have become first choice for developing theoretical models on language attitudes.

3. Direct and indirect methods

Interest in the study of language attitudes as a sociolinguistics variable stems from the work of Lambert – already a classic – who, as we said earlier, along with his collaborators, used the matched guise technique (indirect methodology), in the context of Quebec. Texts recorded by bilinguals in French and English were evaluated by ‘judges’, whose L1 (first language) was French or English. The aim was to reveal the inclinations and preferences as to the personality of the latter – or, quite the opposite –aspects determined by the linguistic variety used each time by the speakers on tape (Lambert, 1960; 1967). (7)