1 Feminism and language
There is no doubt that feminism has been and continues to be one of the main social movements of this century. Its impact is felt in many societies around the world and in many spheres of life. The women’s or feminist movement strives, amongst other things, for the elimination of gender discrimination and for the greater recognition of women’s contributions to society as well as aims to change many cultural and social practices which perpetuate patriarchal value systems. Language was and is seen by many feminists as a powerful instrument of patriarchy: for example, the feminist Dale Spender, spoke of the English language as being ‘manmade’ and as being an important contributor to women’s oppression (Spender 1980). It is therefore not surprising that language and discourse practices were and are subjected to feminist scrutiny, often leading to elaborate and detailed descriptions of sexist practices affecting language use. 2. Feminism and linguistic reform
Feminists, at least in western societies, also expressed a desire to change the patriarchal and sexist ‘nature’ of language and therefore engaged in various types of linguistic reform or language planning. Although many feminists shared the belief that changing linguistic and discourse practices is an important element in women’s liberation, this did not result in a uniform approach to linguistic reform (see e.g. Pauwels 1998). The social, cultural, political and philosophical diversity which characterizes members of the feminist movement is also reflected in the approaches to and aims for feminist language reform. For example, not all forms of feminism, interpret women’s liberation as a question of achieving mere equality of the sexes. Similarly, not all linguistic reform proposals have as their main aim the achievement of linguistic equality of the sexes. Some reform initiatives primarily aim at exposing the sexist nature of ‘patriarchal’ language by causing linguistic disruptions.
The strategies used to achieve linguistic disruption frequently involve experimentation and creativity with all parts of speech. The word ‘herstory’ to refer to history which is not only about men, is an example of linguistic disruption: a morphological boundary has been reconstituted to + on semantic grounds. Creating a women-centred language capable of expressing reality from a female perspective is another prominent objective of some forms of feminist language planning. Proposed changes range from the creation of new women-centred meanings for words like ‘witch’, ‘hag’ and neologisms such as ‘malestream’, ‘femocrat’, graphemic innovations including ‘womyn’ or ‘wimmin’ and ‘LehrerIn’ (German), to developing women-focussed discourses and even creating an entirely new language.
An example of the latter is the Láadan language created by the science-fiction writer and linguist, Suzette Haden Elgin ‘for the specific purpose of expressing the perceptions of women’ (Elgin 1988:1). Despite this diversity in reform initiatives and objectives for feminist language planning, it is the ‘linguistic equality of the sexes’ approach which has become synonymous with feminist language planning in the eyes of the wider community. This is in part due to the prominence of liberal feminist approaches in the public arena which focus on achieving sex/gender equality. Linguistic discrimination is seen as a form of sex discrimination which can be addressed in ways similar to other forms of sex discrimination (e.g. in employment). In fact the question of gender bias in occupational nomenclature is directly linked to gender discrimination in the employment arena. The prominence of the linguistic equality approach is also due to the media’s attention to non-sexist language guidelines, the main instrument of promoting this type of feminist language reform.
Advocates of the linguistic equality approach use the strategies of gender-neutralisation (sometimes gender abstraction) and/or gender-specification (feminisation) to attain their goal of creating a language system which allows for a balanced representation of the sexes. Gender-neutralisation involves minimising or eliminating gender-specific expressions and constructions. It entails ‘that any morphosyntactic and lexical features marking human agent nouns and pronouns (or other parts of speech) as masculine or feminine are ‘neutralised’ for gender, especially in generic contexts’ (Pauwels 1998: 109).
Examples for English include the elimination of gender-suffixes of -ess, -ette, -(tr)ix in relation to human agent nouns (e.g. hostess, aviatrix, usherette), the creation of compound nouns involving -person (e.g. chairperson, tradesperson), and the avoidance of generic ‘he’. Gender-specification (also known as feminisation) is a strategy used to achieve linguistic equality by making the ‘invisible sex’ (in most cases, women) visible in language through systematic and symmetrical marking of gender.
Although English does not use this strategy much (it is found more often in languages with grammatical gender), the use of ‘he or she’, and of phrases such as ‘police women and men’, ‘actors and actresses’ in generic contexts exemplifies the gender-specification strategy. Underlying the linguistic equality approach to reform is a belief that making changes to linguistic forms will contribute significantly to the promotion of non-sexist meanings. 3 Evaluating feminist linguistic reform
In the previous section I indicated that there are several approaches to feminist language reform and that the linguistic equality approach is the most prominent and possibly, the most widespread one. In this paper my focus is on the evaluation of the linguistic equality approach. Evaluating the outcome (a result or an effect of an action) is a crucial aspect of any form of language planning. Language planners together with the interest groups, agencies or institutions which encouraged, demanded or sanctioned (allowed) the reforms are usually keen to assess the impact of planning on the linguistic behaviour of the individuals, groups or communities targeted by the reforms. Whereas advocates and/or opponents of linguistic reform are primarily interested in the extent to which the linguistic reform proposals have been adopted or rejected, for language planners the evaluation exercise also provides valuable information on the process of language planning, the factors which facilitate and/or obstruct change.
A further interest for language planners who are also linguistic scholars is the possibility of comparing the process of the spread of so-called ‘planned’ vs ‘unplanned’ linguistic change thus contributing to a better understanding of linguistic change. Here I wish to explore two major aspects of the evaluation of feminist language planning: (1) Evidence of the (successful) adoption of feminist linguistic proposals; (2) Insights into the ways feminist language changes spread throughout the community.
The adoption and spread of feminist linguistic reform are examined in relation to a prominent feature of feminist linguistic reform of the ‘linguistic equality’ type: the use of gender-neutral and/or gender-inclusive occupational nouns and titles. Data for this discussion come mainly from English, although reference is also made to Dutch, French and German studies. The discussion of linguistic spread is very preliminary as most data have not yet been subjected to a thorough analysis: i.e. only trends will be noted. 4 Adopting feminist linguistic reform: success or failure?
4.1 Occupational nomenclature
In many western societies feminist concerns about gender bias in occupational nouns, professional titles and terms attracted attention primarily through its link with Sex Discrimination Acts and other legislation aimed at eliminating gender-based discrimination in employment. Feminists and women activists in a range of professional bodies highlighted the fact that occupational and professional nomenclature used in employment-related contexts displayed bias in favour of men leading to women’s invisibility in this area of language use. For example, linguistic practices found in many job classifieds assumed applicants to be male. Male-stereotyped language was used to describe applicants (e.g. aggressive, dynamic, virile). The use of ‘masculine’ generic nouns and pronouns (e.g. the applicant – he; storeman, tradesman, cameraman – he) further reinforced the ‘maleness’ of the desired applicant.
Research in the 1970s and 1980s (e.g. Bem & Bem 1973, Hamilton 1988, Kidd 1971, Mackay & Fulkerson 1979, Martyna 1978, Pincus & Pincus 1980, Schneider and Hacker 1973, Wilson & Ng 1988) found evidence that masculine generic nouns and pronouns were seldom interpreted in a generic, gender-neutral sense. Instead they were associated with male-specific images in many language users. Two major strategies emerged to eliminate this gender bias in occupational nomenclature: gender-neutralisation and gender-specification (feminisation). Selecting one strategy over another seems partly linked to linguistic typology. Gender-specification as a main strategy is more likely to occur in the case of grammatical gender languages (e.g. German, French, Italian, Spanish) which still have productive gender suffixes (e.g. German).
Gender-neutralisation is more likely to be applied to languages with a natural gender system (e.g. English) or languages in which gender suffixes are less or no longer productive (e.g. Danish, Swedish and Dutch). However, the choice of the main strategy is also influenced by extra-linguistic or social arguments. Gender-neutralisation is clearly aimed at ‘taking gender out of the occupational arena’. In other words, the aim is to have a society in which a person’s sex has no relevance or significance for their occupational status. Proponents of the feminisation strategy, on the other hand, argue that it is socially more effective to achieve linguistic equality by showing that there are an increasing number of women in all areas of the paid work force, i.e. women’s participation in the work force needs to be made more visible through the strategy of gender-specification or feminisation.
In order to demonstrate successful adoption of feminist linguistic reform in this area of language use, evidence needs to be found that the feminist alternatives are used increasingly in preference to the gender biased forms and that the actual use of the feminist alternatives is in line with their promoted use. In language planning terms , successful feminist linguistic reform entails evidence that the feminist alternatives move from a status of ‘discouraged’ or even ‘disapproved’ use to that of ‘tolerated’, and eventually ‘preferred’ or ‘promoted’ use (Kloss 1968). Findings from Dutch, English, French (see especially Burr in this volume) and German research into the adoption of non-sexist occupational nomenclature confirm that feminist linguistic alternatives are (increasingly) used, although adoption rates vary substantially from language to language and vary according to linguistic context/genre. For the purposes of this paper I will confine the presentation of evidence to that found in relation to the print media (mainly newspapers).
English speech communities seem to lead the way in the adoption of feminist linguistic alternatives for occupational terms. Cooper (1984) studied the impact of feminist language planning on the use of masculine generic pronouns and nouns (including occupational nouns) on a corpus of 500000 words taken from American newspapers, current affairs and women’s magazines covering the period 1971 – 1979. He found a dramatic decline in the use of masculine generics, especially of generic ‘man’ and generic ‘he’: their use fell from 12.3% per 5000 words in 1971 to 4.3% in 1979. In New Zealand Meyerhoff (1984) analyzed changes in the use of masculine generics in a corpus of 150000 words taken from five newspapers with a different audience (i.e. a national and a regional daily, a student newspaper, a TV magazine and a women’s magazine as well as a monthly publication of the New Zealand’s journalists’ union).
Her study found evidence of a significant reduction in the use of masculine generic nouns and pronouns with the decrease being most pronounced for the student newspaper and the journalists’ union publication. The only publication to support ‘- person’ compounds was the student newspaper. Holmes’ analysis of the occurrence of ‘-person’ vs ‘-man’ and ‘-woman’ compound forms in the Wellington Corpus of Written New Zealand English covering the period 1986 – 1989 found that most such forms occurred very seldom (1 per 1 million words) with the exception ‘spokesperson’ and ‘chairperson’ (Holmes in press). The use of these two forms, however, was considerably lower than that of their masculine generic alternatives: ‘spokesman’ and ‘chairman’. The corpus revealed 6 instances of ‘chairperson(s)’ vs 109 for ‘chairman/men’ and 2 for ‘chairwoman/women’. ‘Spokesperson(s)’ occurred 4 times in the corpus, ‘spokespeople’ once, ‘spokeswoman/women’ twice and ‘spokesman/men’ 36 times.
Holmes (in press) did note that the ‘overwhelming majority of the instances of chairman were identifiable as male, a sad reflection of the social reality that it was men who held this position most often, even in 1986’. She found only 4 instances of ‘chairman’ being used to refer to a woman. My own study which comprised a corpus of 200000 words taken from two national Australian newspapers in 1992 and in 1996 similarly found an overall low incidence of -person, -man and -woman compound forms. The number of occurrences of ‘chairman/chairwoman/ chairperson’ revealed the continued predominant use of ‘chairman’, although a breakdown of the numbers according to referents showed that ‘chairman’ was predominantly used to refer to male referents.
The few occurrences of ‘chairperson’ and ‘chair’ (see Table 1) do not allow for an interpretation of emerging trends. In the case of ‘chairman’ I would have to agree with Holmes’ comment that its continuing, frequent use reflects the fact that far more men than women continue to occupy this position. It should also be said that newspaper articles are not an ideal source to establish generic uses of this term, as most references to this position specify the incumbent.
In the case of ‘spokesman/spokeswoman/spokesperson’ a more substantial change can be noticed: although 38 instances of ‘spokesman’ were recorded, ‘spokesperson’ appeared 32 times. A breakdown in terms of referents showed that 47% of ‘spokesman’ uses referred to a male and that ‘spokesman’ was never used to refer specifically to a female. Most uses of ‘spokesperson’ had no specific referent. There is also some indication that ‘spokesperson’ is being used in connection with male as well as female referents, hence avoiding the trend that the ‘-person’ compound is used as a mere substitution for the ‘-woman’ compound form.