Филологические науки. Языкознание

Старший преподаватель Рябикина Е. В.

Ростовский Государственный Экономический Университет «РИНХ», Россия

HEDGES AND BOOSTERS IN WOMEN’S AND MEN’S SPEECH

1. Introduction

Though Robin Lakoff’s (1973, 1975) claims about the linguistic forms she considered characteristic of ‘women’s language’ have been attacked, misrepresented, qualified, refuted and constantly criticized over the last 15 years, no one can say they have been uninfluential. As a result of her hypotheses about the way women speak and why, we now know a great deal more about the speech behavior of women compared to men than we did in 1973.

My own research on epistemic modality led very naturally to the collection of data which could appropriately be used to examine Lakoff’s claims. Many of the features she identified as characteristic of women’s language, including hesitations, rising intonation, tag questions, hedges and intensifiers, are linguistic devices which may be used to express epistemic modality or degrees or certainty about a proposition (see Holmes, 1983)

With a rich data base of speech collected to explore the ways in which native speakers of English express epistemic modality, I was well placed to examine Lakoff’s claims about ‘women’s language’ in at least some contexts. I therefore decided to explore in some detail the distribution of a number of linguistic forms in women’s and men’s speech. In order to avoid at this stage the semantic assumptions encoded in labels such as ‘hedge’ and ‘intensifier’ I will refer to these forms initially as pragmatic particles. In this paper I intend to summarise the findings on each of the following:

(1) the tag question;

(2) three pragmatic particles usually regarded as hedges: sort of, you know and I think;

(3) a pragmatic particle usually regarded as an intensifier: of course.

The first step in the analysis of the distribution of these forms in women’s and men’s speech involved rectifying two weaknesses in much of the research which has attempted to investigate Lakoff’s clams. Firstly it was necessary to indentify carefully the relevant linguistic forms and their functions; secondly it was important to devise a methodology to protect against avoidable bias in the data collection and analysis.

2. Forms and functions.

2.1. Identifying the relevant forms

In an earlier article (Holmes, 1984a) I discussed in some detail the misleading effect obtained by simply counting linguistic forms without taking account of their function in context. Since then I have examined a number of linguistic forms widely recognised as potential ‘hedges’ in Lakoff’s sense (i.e. attenuators or mitigators of the strength of a speech act) and the results have been consistent: linguistic forms are complex and the functions they express cannot be identified in a social and textual vacuum.

Let me illustrate with two of the forms on which I am focusing in this paper.

2. 1. 1. Tag questions. There is a rich literature on the complex syntax and phonology of tag questions (e. g. Arbini,1969; Holmes 1982b; Hudson, 1975; Millar and Brown, 1979; Östman, 1982). This research demonstrates that tags differ in polarity, in intonation, in syntactic derivation and in lexical form. Yet many of the researchers investigating Lakoff’s claims ignore this variation and treat tags as invariant forms, assuming that the relative frequency of use by women and men, regardless of variation and treat tags as invariant forms, assuming that the relative frequency of use by women and men, regardless of variation in form and function and function, is itself significant.

Two examples will illustrate the point. Firstly, when one is concerned to identify tags functioning as hedges (i. e. as ways of attenuating the strength of a speech act), then it is clear that the range of forms must include not only what might be described as canonical tags, such as are you, isn’t she and can’t they. The list must also include forms such as eh, with falling or rising intonation, as well as non-standard forms such as in I and in’t (Cheshire, 1981, 1982, p. 165), since these forms may serve exactly the same function as canonical tags in some dialects, as (1) and (2) illustrate. (I have supplied with each example a very brief indication of the context of the utterance in order to facilitate interpretation of the illocutionary point and approximate strength of the speech act. The slashes / and // indicate two lengths of pause. Intonation is also marked, where relevant, by left- and right – sloping slashes above words.)

The speaker signals with the use of of course that he assumes the facts associated with Watergate are common knowledge.

CONFIDENTAL of course also signals the status of knowledge, or beliefs as shared by the speaker and hearer, but the assumed knowledge or beliefs are personal and specific to a particular social network. The assumptions are based on previous contact, shared information and membership of an in – group. The information or attitudes referred to are often introduced as an ‘aside’ in a narrative, with of course functioning to signal the status of the proposition as mutual pre – existing knowledge.

CONFIDENTAL of course may be glossed ‘as you know on the basis of information or experience we have shared’ or ‘as you might deduce, given our shared attitudes’. In this meaning it generally precedes and is part of the same tone unit as the proposition. The example illustrates its use to refer to previously imparted personal information.

3. Methodology

Turning now to the second problem with post – Lakoff language and gender research, there are a number of major weaknesses in the methodology used to compare women’s and men’s usage. I have discussed the problem of identifying the forms to be counted. Another problem involves an appropriate universe of discourse for analysis and ensuring that the language samples from each sex are carefully matched.

3.1. Contextual factors and discourse type

3.1.1. Medium and formality. In her discussion of women’s language forms, Lakoff (1973, 1975) suggested that they were far more likely to occur in spoken than in written contexts ‘or at least in highly informal style’ (Lakoff, 1975, p. 59). Subsequent investigations have confirmed this claim for many of the particular pragmatic particles she identified, though with some qualifications: e. g. you know is very decidedly most frequent in relaxed casual spoken interaction (Holmes, 1986, p. 12), as are soft of and kind of (Holmes, 1988b); of course, on the other hand, is more frequent in writing than most particles, but even of course occurs more than three times as often in speech as in writing, and, like DELIBERATIVE I think, it is most frequent in semi – formal speech contexts such as interviews (Holmes. 1985, 1988a).

3.1.2. Sex of addressee. In addition to the medium of interaction and the formality of the context the sex of the addressee appears to be a relevant factor in accounting for the frequency of pragmatic particles in women’s and men’s speech (e.g. Brown, 1980; Bell, 1984; Holmes, 1986, 1988b). The effects of addressee sex are far from clear, however (see Bell, 1984).

3.2. Quantification

A major problem which bedevils a great deal of the research on language and gender is the question of how to quantify the data. This involves two related issues:

(i) Providing a valid data base;

(ii) Operationalizing the relationship between form and function.

3.2.1. Providing a valid data base. Most of the research on language and gender compares the number of forms used by men with the number used by women in the same contexts during a particular period of time. In providing this information most of these researchers control for the number of female and male contributors to the talk, some control for the role and statuses of participants, but information on the total amount of speech contributed by each sex is rarely supplied. The number of instances of a particular form produced by women and men is being compared, therefore, without any information on how it relates to the total amount of speech collected from each sex. Such information is meaningless.

The information, for instance, that in the discussion following the papers at a conference men used 33 tag questions and women used none may simply reflect the fact that women contributed very little to the discussion. There is abundant evidence that in formal mixed – sex context women do not speak as often or for as long as men (e. g. Swacker, 1979; Edelsky, 1981) and some of it relates quite specifically to the conference context (Swacker, 1979; Holmes, 1988c). If men dominate the available talking time then it is scarcely surprising if they produce more of any linguistic form than do women in the same time.

4.4. I think

The form I think is included in Lakoff’s list of hedges alongside sort of, you know, and tag questions, and like them it proves on further analysis to be complex in form, function, and distribution (Holmes, 1985).

As described in Section 2. 2. 1, I think may express two distinct and contrasting functions: TENTATIVE I think may be used to express uncertainty (epistemic modal meaning), or as a softener expressing politeness (affective meaning), while DELIBERATIVE I think expresses certainty (epistemic modal meaning) and reassurance (affective meaning). While there are prosodic and syntactic features (see Section 2.1.2. and Holmes, 1985) which correlate to some extent with these contrasting functions, there also cases where formal features alone are insufficient for a definitive classification. As with other particles, these cases require careful consideration of the context of utterance, including such features as the relationship between the participants, the topic, and the formality of the interaction.

Tables 4 shows that while the overall frequency of I think of in women’s and men’s speech is not very different, there is a clear contrast in the functional distribution between the sexes. Women use DELIBERATIVE I think more frequently than they use TENTATIVE I think (62 vs 31% respectively), while the reverse is true for the men. Men use TENTATIVE I think more frequently than DELIBERATIVE I think (59 vs 36%respectively).

Moreover, Table 4 also makes clear that women use I think significantly more frequently in its DELIBERATIVE function than men do, thus refuting Lakoff’s claim that women use I think as an uncertainty marker more often than men do. So when one compares the distribution of I think in a corpus of women’s and men’s speech which is matched for quantity and context, there is no support for Lakoff’s claims about women’s language. Indeed the evidence directly refutes her preditions, suggesting that, at least in my corpus, women use language considerably more assertively than Lakoff allows for.

5. Conclusion

I have examined the form, function and distribution of a range of pragmatic particles which have been suggested as features of women’s language. It is clear from the analysis that they are complex forms with sophisticated functions. All function at the discourse – planning level as verbal fillers (Brown, 1977), devices which facilitate the smooth flow of the discourse by providing the speaker with planning time, or as conversational lubricants in interaction, encouraging easy turn – taking between participants. At a more specific level they consistently express epistemic modal meaning (or the extent of the speaker’s certainty about a proposition), and affective meaning, or the speaker’s attitude to the addressee in the interaction (Holmes, 1982a, 1983). It is these meaning in particular which distinguish between different particles.

Future research in the area of language and gender must consistently take account of contextual and quantitative factors if the results are to be useful. And there is good reason to extend the scope of analyses well beyond the particles which have preoccupied researchers to date.

REFERENCES

1. AIJMER, K. 1987 ‘Sort of’ and ‘kind of’ in English conversation. Studia Linguistica 38, 118 – 128.

2. ARBINI, R. 1969 Tag – questions and tag – imperatives in English. Journal of Linguistics 5, 205 – 214.

3. BEll, A. 1984 Language style as audience design. Language in Society 13, 145 – 204.

4. BROWN, P. and LEVINSON, S. 1978 Universals in language usage: politeness phenomena. In Goody, E. N. (Ed) Questions and Politeness: Strategies in Social Interaction, pp. 56 – 280. Cambridge University Press, Cambridge.

5. BROWN, P. and LEVINSON, S. 1987 Politeness: Some Universals in Language Usage. Cambridge University Press, Cambridge.

6. BROWN, P. 1980 How and why are women more polite: some evidence from a Mayan community. In McConnell – Ginet, S., Borker, R. and Furman, N. (Eds) Women and Language in Literature and Society, pp. 111 – 136. Praeger, New York.

7. CAMERON, D. 1985 Feminism and Linguistic Theory. Macmillan, London.

8. CHESHIRE, J. 1981 Variation in the use of an urban British English dialect. Language in Society 10, 365 – 381.

9. CHESHIRE, J. 1982 Linguistic variation and social function. In Romaine, S. (Ed.) Sociolinguistic Variation in Speech Communities, pp. 153 – 166. Edward Arnold, London.

10. COATES, J. 1986 Women, Men and Language. Longman, London.

11. COATES, J. 1988 Gossip revisited: language in all – female groups. Mimeograph. In Coates, J. and Cameron, D. Women in Their Speech Communities, pp.94 – 122. Longman, London .

12. EDELSKY, C. 1981 Who’s got the floor? Language in Society 10, 383 – 421.

13. HOLMES, J. 1982a Expressing doubt and certainty in English. RELC Journal 3, 9 – 28.

14. HOLMES, J. 1983 Speaking English with the appropriate degree of conviction. In Brumfit, C. (Ed.) Learning and Teaching Languages for Communication: Applied Linguistic Perspectives, pp. 100 – 113. Centre for Information on Language Teaching and Research, London.

15. HOLMES, J. 1984a ‘Women’s language’: a functional approach. General Linguistics 24, 149 – 178.

16. HOLMES, J. 1984b Hedging your bets and sitting on the fence: some evidence for hedges as support structures. pp. 47 – 62.

17. HOLMES, J. 1985 Sex difference and miscommunication: some data from New Zealand. In Pride, J. B. (Ed.) Cross – cultural Encounters: Communication , pp. 24 – 43. River Seine, Melbourne.

18. HOLMES, J. 1988a Of course: a pragmatic particle in New Zealand women’s and speech. Australian Journal of Linguistics 8, 49 – 74.

19. HUDSON, R. A. 1975 The meaning of questions. Language 51, 1 – 31.

20. LAKOFF, R. 1973 Language and woman’s place. Language in Society 2, 45-79.

21. LAKOFF, R. 1975 Language and Woman’s Place. Harper & Row, New York.

22. MILLAR, M. and BROWN, K. 1979 Tag questions in Edinburgh speech. Linguistiche Berichte 60, 24 – 45.

23. ÖSTMAN, J. - O. 1982 A functional approach to English tags. Studia Anglica Posnaniensia 13, 3 – 16.

24. SWACKER, M. 1979 Women’s verbal verbal behaviour at learned and professional conference. In Dubois, B. L. and Crouch, I. (Eds) The Sociology of the Languages of American Women, pp. 155 – 160. Trinity University Press, San Antonio, Texas.