Gender, language and CMC for education

Learning and Instruction 11 (2001) 21–34 www.elsevier.com/locate/learninstruc Gender, language and CMC for education S.J. Yates * Discipline of Sociology, The Open University, Walton Hall, Milton Keynes, MK7 6AA, UK Accepted 7 June 2000 Abstract This paper overviews research into the role of gender in computer-mediated communication (CMC) interactions. It uses linguistic and sociological research to criticise the frequent claim that CMC is free from gender-based inequalities due to the lack of face-to-face cues. In examining the research to date, the paper makes clear the important role of social context in both providing and limiting opportunities for equality in instructional interactions taking place via CMC systems.  2000 Elsevier Science Ltd. All rights reserved. Keywords: Computer-mediated communication; Distance education; Gender; Language 1. Introduction During the early to mid 1990s the use of computer-mediated communication (CMC) in education was still limited to a number of essentially experimental cases. Since that time, the number and range of applications of CMC to teaching at all levels has continued to grow. In particular CMC has become a favourite medium for many applications of ‘distance education’. As a consequence of this a considerable body of educational research on the use of CMC has developed. Despite this growth, the number of academic publications concerned with the role of gender in educational CMC interactions has remained relatively small and at first glance seems to provide a confused, if not conflicting, set of results. This paper overviews this research and considers the extent to which these results can be understood through a re-evaluation of research starting points and the methods employed. CMC in its various forms (electronic mail, computer conferencing, internet relay * Tel.: +44 (01908) 654457; fax: +44 (01908) 654488. E-mail address: [email protected] (S.J. Yates). 0959-4752/01/$ - see front matter  2000 Elsevier Science Ltd. All rights reserved. PII: S 0 9 5 9 - 4 7 5 2 ( 0 0 ) 0 0 0 1 2 - 8 22 S.J. Yates / Learning and Instruction 11 (2001) 21–34 chat etc.) has engendered a lot of academic and public debate. Within much of this debate the behaviour of users is very often linked to ‘attributes’ of the medium. For instance, many people hoped that such a semi-anonymous text-based medium would bring a lessening of inequalities in the communication taking place. The argument being that reduced status, gender, race and other social cues, due to the lack of direct face-to-face presence, would make CMC a free and equal communications medium. In contrast others (e.g. Kiesler, Siegel, & McGuire, 1984) attributed ‘flaming’ (aggressive and offensive messaging) to the lack of face-to-face cues. In a reaction to such research, press coverage of CMC (e.g. Newsweek, May 16th, 1994 for example) has tended to portray CMC as a medium that brings out some of the worst aspects of male behaviour and gender relations. This paper takes an overview of some of the research into gender and the educational use of CMC. The paper compares the results of research that has tended to focus on the communicative practices present in interactions with research that has tended to focus on users’ perceptions and uses. Research into users’ communicative practices has often employed linguistic methods whilst research into users’ perceptions, on the other hand, has often made use of questionnaires and quantitative measures of CMC use. 2. Background CMC takes many forms. These range from electronic mail to video conferencing. Though all of these systems have to some extent or another been used for educational purposes, this paper will focus upon the forms of CMC which rely upon some form of text-based interaction. Many early academic studies of CMC (see Mason & Kaye, 1989) focused upon the educational possibilities provided by text-based CMC. These studies often focused upon group interactions taking place across either electronic mail or computer conferencing systems. The types of studies undertaken included comparisons of face-to-face tutorials, lectures and seminars to educationally based computer-supported collaborative work (CSCW). Very few, if any, of these initial projects explored the issue of gender. There are two reasons for this lack of discussion. In the first place, much of this research concerned the technology itself, the new types of interaction taking place, and, as with other areas of computing research, ignored general social issues such as gender relations and other structures of inequality. In the second place, and more importantly for this paper, previous research tended to focus upon the perceived freedoms of access and social ‘anonymity’ that such systems provided. The lack of clear face-to-face cues, it was claimed, would limit the operation of stereotypes and more general forms of social exclusion and inequality. This position could be described as the ‘democratic theory’ or ‘model’ of CMC interaction (see Herring, 1993a; Yates, 1993b). Despite the fact that this has not proved to be the case, much current research is still based upon this assumption. Given this situation, the problems faced by women, as well as other groups, in gaining access to and making the best use of the opportunities provided by CMC, both in education and elsewhere, need to be addressed. S.J. Yates / Learning and Instruction 11 (2001) 21–34 23 This paper is split into two main sections. The first of these concerns the relationship between gender differences in communicative practices and their relation to CMC for education. The second section looks at wider issues of access to CMC and the types of educational opportunities it brings. 3. Gender, language and CMC It is now commonly accepted that there are clear variations in men and women’s approaches to and use of language, though the reasons for these differences remains contested. A large amount of psychological, sociological and linguistic literature on this issue now exists (e.g. Coates, 1986; Graddol & Swann, 1989; Lakoff 1975, 1990; Tannen, 1984) and this literature has often been viewed as representing two distinct models of gender and language (see Cameron, 1997). The first position (see Lakoff 1975, 1990) can be viewed as concentrating on the use of language in the domination of women. The second position (see Tannen, 1984) can be characterised as viewing the variation as the product of gender differences in the same way as language variation reflects other cultural and class differences. More recently, some feminist linguists (see Cameron, 1997) have asserted that the use of the category ‘gender’ has masked the complexities at play. For these writers language variation in relation to gender reflects the complex set of identities and practices that make up ‘doing gender’ in a specific context at a specific time. For these researchers the ‘difference’ and ‘dominance’ approaches are seen as being very similar in that they take the category of gender as a given. Interestingly both the ‘difference’ and ‘dominance’ approaches have tended to focus on the same kinds of empirical evidence. This evidence can be seen to come in three forms. First there is evidence of the inequalities in the structure of male/female interactions. For instance, the fact that men have more turns and often speak for longer in mixed gender interactions. When women hold the conversational floor for more than one third of the interaction men will perceive women to be dominating a conversation (Holmes, 1992). Second, there is evidence of gender differences in the linguistic practices and strategies that people use in interactions. These include differences in turn-taking conventions, means of gaining the conversational floor, and means of directing the flow of the interaction. Third, there is evidence of differences in the purposes for which people engage in linguistic interaction. Within the context of educational interactions there may be some large differences between the educational goals which men and women may seek from interactions. CMC is, in nearly all cases, a literate form of interaction with most of the interaction taking place through the use of typed messages. It is also important to remember that these interactions are not restricted, formalised ‘packets of information’ but often very social and highly interactive forms of communication (see Graddol, 1989; Yates, 1993a,c; for a discussion of the linguistics of CMC as a medium). This makes differences in the linguistic resources and practices that people bring to CMC all the more important. 24 S.J. Yates / Learning and Instruction 11 (2001) 21–34 4. Gender inequalities, language and CMC Herring (1992, 1993a, 1993b) has written extensively on the issue of gender, language and CMC and this section will follow some of her main arguments. Herring (1993a) begins one of her first full discussions of this issue by focusing upon what we have termed the ‘democratic theory’ of CMC. In doing so, Herring notes that both educators and feminist writers have been encouraged by the possibilities for openness of access and interaction that CMC brings. Herring notes the similarities of this position to the model of the Ideal Speech Situation (ISS), in which undistorted and rational debate can take place, put forward by the social theorist Habermas (1984, 1987). There are many general socio-linguistic problems tied to the use of the ISS to explain CMC phenomena (see Yates, 1993a) and Herring uses data collected from a study of two educational/academic discussion lists to clearly demonstrate differences in gender use and access to CMC and therefore question the relevance of the ISS to CMC. The two lists which Herring explored were the LINGUIST electronic mail list which is devoted to the discussion of all aspects of language and linguistics, and Megabyte University (MBU) a list concerned with computers and writing. Both lists consist in the main of academics, postgraduate and undergraduate students. The LINGUIST list having over 1,000 members and the MBU list over 250 at the time of Herring’s research. Over a period of one year, Herring conducted an ethnographic observation of the interactions taking place. This involved collecting a full transcript of the interaction, and from this collecting data on the participants, the issues raised and discussed and other relevant information. During this time she subjected two extended discussions of specific subjects from each discussion list to detailed linguistic and socio-linguistic analysis (Herring, 1993a). The first important finding which Herring uncovered was the disparity in participation. Despite the fact that both lists had reasonably high numbers of women receiving messages from the list (36% on LINGUIST, and 42% on MBU), women contributed far less than the men and the levels of participation varied according to the topic under discussion (30% of women on both lists participating in ‘sexism’-related discussions compared to 16% of women participating in ‘theory’-related discussions). Herring also found that messages from women were shorter on average, with only men posting messages ‘ten screens or more in length’. Finally, Herring notes that women’s messages gain fewer replies in these mixed-gendered interactions. Within the MBU ‘sexism’ discussion 80% of male postings were responded to as compared to 70% of the women’s postings (Herring, 1993a). Herring explains this difference using previous research (Holmes, 1992) which suggests that men become threatened by situations in which women take up more than 30% of the interaction. She notes that men reacted negatively to those interactions where women’s contributions approached 50%, in all cases causing the discussion thread to ‘die out’ (Herring, 1993a). Herring also notes two other differences in the CMC practices of men and women. The first of these is topic selection. Looking at her data from the LINGUIST list Herring notes that men were more likely to post messages on specific issues or to 25 S.J. Yates / Learning and Instruction 11 (2001) 21–34 provide specific information, whereas women were more likely to post on personal aspects of the discussion or to post queries to other list members (see Table 1 taken from Herring, 1993a). This issue of individuals making ‘personal’ as against ‘informational’ postings is one which continues to re-appear within CMC research. There are several documented cases of ‘personal’-orientated interactions leading to the shutting down or re-organisation of CMC systems. In her excellent study of the impacts of a CMC system on a US corporation Zuboff (1988) notes the reaction to a women’s employment issues conference on the system. The discussions taking place soon generated complaints which were passed to the manager in charge of the system. Zuboff notes that: “A group of participants was called into his office and asked to explain the ‘legitimate business purpose’ of the conference. While the group’s rationale was accepted, they were told to exert more leadership in directing the discussion to business issues. In the aftermath of these events, conference membership dwindled.” (Zuboff, 1988, p. 383) The marking out of the boundary between ‘legitimate’ business or informational content and illegitimate ‘personal’ content reflects a number of complex social concerns. On the one hand, they reflect social perceptions of how CMC should be used and social perceptions about literate practice (Yates, 1992), as well as perceptions of public and private communication. In many cases CMC is clearly being defined as a male gendered, public, information-orientated medium. These differences therefore also reflect differences in the types of purpose for which men and women may be engaging in both face-to-face and CMC interactions in educational and other contexts (Yates, 1993b). This issue will be discussed more fully in relation to educational interactions later in this article. The next difference Herring noted concerned language style and content. Using a set of features defined in terms of their attribution to different gender language styles (see Table 2 taken from Herring, 1993a), Herring discovered large differences in style. Of the 261 LINGUIST discussion messages she analysed, 68% of women’s postings contained some features of ‘women’s language’ style compared to only 31% of male posting. On the other hand 48% of male messages contained only male features as contrasted with 18% of women’s messages that contained only male features. Lastly whilst 46% of women’s messages mixed men and women’s language features only 14% of men’s messages used such combinations (Herring, 1993a). This Table 1 Topic selection and gender in CMC interactions Gender Most postings Women Men Personal Issues Least posting Queries Information Information Queries Issues Personal 26 S.J. Yates / Learning and Instruction 11 (2001) 21–34 Table 2 Features of men and women’s language Women’s language Men’s language Attenuated assertions Apologies Explicit justifications Questions Personal orientation Supports others Strong assertions Self-promotion Presuppositions Rhetorical questions Authoritative orientation Challenges others Humour/sarcasm result indicates that women are engaging in ‘male’ strategies in order to remain in the interaction. Herring provides examples of the use of these strategies by different participants on the LINGUIST list. Personal orientation is indicated by the use of phrases such as “I am intrigued by your comment...” and “I am interested because I think similar...” (Herring, 1993a). This use of personal pronouns fits with Yates’ finding (Yates, 1993a) that pronoun use is much higher in highly interactive interpersonal rather than formal CMC interactions. Authoritative orientations and strong assertions are marked by such phrases as “One is...” and “It is obvious that...” (Herring, 1993a). Herring concludes from this research that, despite the lack of overt face-to-face cues in CMC, gender is made clear in the form and content of the language used. She explains the lack of participation by women and the differences in communicative practice through the use of the previously described research into differences in gender and language use. Male dominance of the discussion is maintained through the use and abuse of more aggressive and masculine language features and practices. Herring concludes that: “... a small male minority dominates the discourse both in terms of the amount of talk, and rhetorically, through self promotion and adversarial strategies. Moreover, when women do attempt to participate on a more equal basis, they risk being actively censored by the reaction of men who either ignore them or attempt to delegitimise their contributions.” (Herring, 1993a) The making explicit of such gender cues through both forms of language use and the reality of gender markers, such as names being present in messages, may have further implications. In experimentally constructed studies of CMC-based negotiations amongst mixed and single gender groups, Kimberly Matheson (1991) noted that the presence of such gender information: “... invokes stereotypes regarding gender appropriate behaviour which, in turn, influences expectations and perceptions of the other communicator. Hence in computer-mediated communications, females may be expected to be more emotional, S.J. Yates / Learning and Instruction 11 (2001) 21–34 27 submissive, and socially supportive, and less aggressive, ambitious, effective and decisive, relative to males.” (Matheson, 1991, p. 144). These conclusions would seem to have quite negative implications for use of CMC in education; especially as the material studied by Herring consisted of academic and educationally-orientated CMC interactions. The ‘democratic’ model has not won out and, as with face-to-face educational situations, gender has a key role to play in structuring the interactions so as to marginalise women’s contributions. Many writers who hoped to see CMC remove inequalities have ignored the fact that social structures and social practices do not just depend upon visual face-toface cues but are in most cases constructed and enacted through communication. The nature of CMC, with its individualised, slightly anonymous set of social and communicative relations, may in fact favour male communicative strategies. In the situation where any person can have the ‘floor’ at any time it may be advantageous to be the most forceful and assertive in order to control the flow of the discourse. 5. Re-inventing gendered CMC interactions Not all CMC interactions are of mixed-sex, or male dominated in numbers and these women-orientated interactions provide other insights. Further research by Herring (1993b) and Wyatt (1993) has indicated that the gender-based communications problems in CMC interactions described above reflect the structure of the social situation rather than specific features of the medium. Again comparing the LINGUIST list with another CMC discussion list, the Women’s Studies List (WMST), Herring found similar behaviours as before. In the case of the WMST women made up 80% of the membership and the number of postings to the list reflected this. In examining the features of the discourse in the same manner as described before Herring concluded that: “There is evidence, moreover, that the discourse style of the dominant group becomes conventionalised for the list as a whole: successful male participants on WMST attenuate and personalise their contributions, and women on LINGUIST must employ adversarial rhetorical tactics in order to be taken seriously, although both typically retain (conflicting) features of their own genderlect as well.” (Herring, 1993b). Wyatt (1993), also exploring WMST, considered the long-term development of content within the discussion. Wyatt found that though the discourse was at the outset more ‘personal’ than ‘professional’ (informational) this changed over time as the interactants, mostly women, gained confidence and developed a specific linguistic genre for the interaction. The results of both of these studies should not in fact surprise us. They return us to one of the important themes of this paper: CMC interaction, as with face-to-face communication, is founded upon existing social structures and perceptions which 28 S.J. Yates / Learning and Instruction 11 (2001) 21–34 underlie inequalities in the interaction. As in the case of most face-to-face interactions between mixed gender groups women have to take on male communicative practices in order to gain access to the discourse. One of the misunderstandings or presuppositions that may underlie the ‘democratic theory’ of CMC concerns its use of the computer as a medium. Given that we socially conceive of the computer as a neutral transmitter of information many have assumed that communication via the computer is itself neutral. CMC may use the computer as the medium but it is still social and interpersonal communication; which in all other areas of human interaction is saturated with social structure (see Yates, 1993a for a full discussion of the discursive assumptions made concerning CMC). Another study by Collins-Jarvis (1993) of the uptake and use of a CMC system by women, this time as part of a local community project found similar results. Collins-Jarvis looked at women using the City of Santa Monica’s Public Electronic Network (PEN). PEN was set up to provide electronic access to local government and politics. It provided three functions: access to databases; electronic mail communication to local government offices; and group computer conferencing interactions. Of the 303 users that Collins-Jarvis studied 89 were women. Collins-Jarvis also notes that the women she interviewed had come across men using aggressive language practices to delegitimise women’s interactions. Collins-Jarvis was also able to document the reactions of women to these negative male behaviours noting that: “Qualitative analyses revealed that female users attempted to reinvent rather than simply reject the PEN system when faced with the following barriers to use: (1) no conferences specifically addressed women’s issues; and (2) a few male[s]... interacted on the system in a manner that appeared to be hostile to female users.” (Collins-Jarvis, 1993, p. 61) These adaptations included the setting up of conferences dedicated to issues that specifically concerned the women users of PEN. Most of the users of PEN were people with previous political motivations and as Collins-Jarvis notes the: “... compatibility between PEN’s political capabilities and women’s participatory norms served as an important adoption incentive. This suggests that computing systems which appeal to women’s norms and interests... can indeed increase female adoption rates.” (Collins-Jarvis, 1993, p. 61) This is a more positive conclusion than the one drawn from Herring’s work. It also implies that the effects of male communicative behaviour in CMC interactions can be reduced by the social organisation of CMC systems. It also demonstrates the important point alluded to earlier that differences in communicative practices between men and women may also reflect different communicative and social goals. This is an important point, which we shall explore in more depth in the next section. S.J. Yates / Learning and Instruction 11 (2001) 21–34 29 6. Gender, education and CMC The first section of this paper has made clear that CMC is not the open democratic medium that some educationalists and technologists initially hoped. CMC as a medium does provide many possibilities for both learning and instruction. These include direct instruction, formal and informal student–tutor interactions, educational and social student–student interactions and focused CSCW by both students and educators. There are many examples of these activities, from small group ‘tutorials’ to large scale ‘lectures’, historical and contemporary educational role-playing to second language teaching. It is the types of educational goals which CMC can support across these interaction that is important to this discussion. One of the main arenas of education in which CMC has been applied is that of distance education. Given the ability of most CMC systems to operate over large distances, and most often being asynchronous, they can operate without the interactants needing to be co-present in space or time. In these respects CMC seems almost perfectly designed for distance education. The first gender-related issue that such a claim raises is that of access to the technology. Unlike many of the other publications on gender and CMC, both those quoted here and beyond, this paper did not begin with a discussion of women and their use, access to and attitudes towards information technology. This is not to suggest that this problem is unimportant but to maintain the focus upon the issue of CMC rather than the broader issues, dealt with elsewhere in very great and insightful depth, of gender and access to computing. CMC as a computer-based medium does require that individuals have, gain or are provided with, a certain level of computer literacy. As the vast body of literature on the issue of gender and computing in general makes clear this is in itself a basic impediment to women’s access to and use of CMC. There are other important access issues to be borne in mind in relation to CMC. As the work of Herring makes clear, the over-confident style of men’s CMC interactions can be an impediment to women users developing confidence with the medium. In my own research (Yates, 1992, 1993b) into the use of CMC as part of a distance education course by some 875 students, the 309 women on the course were statistically more likely to be less confident of using CMC than the men even before they had engaged with the technology. Women were more likely than men to have no previous experience of CMC (75% women compared to 63% men). In terms of using CMC, women were statistically more likely to expect CMC to be harder to use than men. Interestingly enough there were no statistical differences in men and women’s expectations of the usefulness or friendliness of CMC interactions. Lastly, both men and women were equally as positive about using CMC to contact other students (Yates, 1992, pp. 10–12). The men were also more likely to be confident of using the technology in general and claimed statistically higher levels of previous experience (Yates, 1993b, pp. 26– 29). In all areas of computer use, except that of sales and administration use which was one of the lesser areas of experience all round, women declared less experience than men. Of those students with no experience of computing at all before taking the course, there were twice as many women as men (20% of all women students 30 S.J. Yates / Learning and Instruction 11 (2001) 21–34 compared to 10% of men) (Yates, 1992, pp. 7–10). Though there is considerable evidence (see Wilder, Mackie, & Cooper, 1985 for example) that men over estimate and women underestimate their computing abilities, this initial condition does seem to place women at a greater disadvantage then men. These kinds of results further support the arguments of Collins-Jarvis and of Herring that it is important to ensure that women using CMC have motivations beyond simply ‘playing with’ the technology if CMC is to be used successfully for instruction or education. So far we have seen that there are number of important gender-based problems in relation to the use of CMC for educational purposes. There is though, some better news to finish this paper on. The survey of distance education students using CMC discovered very few differences in attitude and utilisation once people began using the system. There was no statistical difference in the amount of time that students logged onto the CMC system (Yates, 1993b, pp. 30–31). The main cross-gender determinant of time logged onto the system was people’s pre-use statement about CMC’s expected usefulness; with those expecting CMC to be useful using the system more. One other determinant of usage rates that was statistically significant for women only, was previous experience. Women who had declared little or no previous CMC experience were just as or more likely to use the system then their experienced counterparts. Of those women who had declared no previous experience of CMC, 67% used the system for more than the mode time of 4–5 h compared to 49% for women with little experience, 54% for women with some experience and 60% for women who declared themselves experts. Though a similar pattern also emerged for male users it was not statistically significant. Such results indicate that, under certain circumstances, women can clearly overcome initial disadvantages and make as much overall use of the technology as men. There were only small differences in types of people whom the students contacted and interacted with. There was little difference between men and women in their levels of contact with other students, both within their tutor group and across the course as a whole. Women were more likely than men to contact their tutors ‘often’ using CMC but were also more likely to ‘never’ have contacted their tutor by CMC; with men being more likely to have ‘occasionally’ contacted their tutors. Very few students of either sex contacted other tutors on the course. Men and women were statistically just as likely to post messages to their tutor group conference (86% women and 87% of men ‘occasionally’ posting messages) (Yates, 1992, p. 20). Students’ perceptions of CMC did bring out one or two important points. When directly asked if CMC allowed for more equal participation than face-to-face interaction, women were less positive than men with 54% of women and 65% of men believing that it did allow more equal participation. Given that 29% of women and 21% of men disagreed with this claim, the rest being unsure or not answering the question, there is some support here for a milder version of CMC’s ‘democratic’ model. Having said this, women, who are using the system as much as men had more reservations about the equality of the interaction (Yates, 1992, p. 20–23). One explanation for the more positive findings in this research as compared to that of Herring may lie in the role of this CMC system in supporting a distance education course. Research by Kirkup and von Prummer (1990) makes clear that S.J. Yates / Learning and Instruction 11 (2001) 21–34 31 women engaged in distance education may have different expectations from interactions with other students and with tutors than men. Kirkup and von Prummer interviewed students from two distance education establishments about their attitudes towards and needs in relation to distance education courses. One of the most important needs which women identified was access to and interaction with other students engaged in distance education. Men on the other hand were more likely to be directly goal-orientated wanting only course relevant information from visits to study centres or face-to-face teaching sessions. This result bears similarities to Herring’s finding that women using an academic CMC list placed personal aspects of issues first in terms of topic selection. Though women and men preferred more traditional face-to-face or telephonebased interactions for getting help, socialising with other students, and intellectual exchange (Yates, 1992, pp. 23–24), CMC provided an extra communications channel through which to conduct such interactions. Quantitative analyses of the linguistic interactions taking place on the CMC system found that a large amount of the interactions were of a social, interpersonal, ‘spoken like’ form (Yates, 1993a) which were clearly orientated to personal rather than intellectual interaction. Sadly, it was not possible to conduct comparisons of these results across sex as data on the sex of the interacting students was no longer available. In looking at the overall results from the survey, it was concluded that many of the problems faced by women using CMC as part of an educational setting were analogous to those of face-to-face situations. It could also be argued that CMC provided greater opportunities for women to engage in some of the goals of personal and social interaction that may be missing from standard distance education frameworks. As Kirkup and von Prummer conclude: “The female ‘independent’ learner does not enjoy or benefit from isolation. It is therefore important in designing any distance education system that effort is put into creating networks of support amongst students ...” (Kirkup & von Prummer, 1990, p. 30) 7. Conclusion This paper has presented a brief overview of some of the literature on CMC, gender, language and education. This is of course a fairly new field and there are, as yet, few large and detailed studies. Some of the existing more comprehensive studies have been briefly explored in the course of this discussion. Of the two main approaches discussed Herring’s argument (Herring, 1993a) provided clear evidence that CMC does not remove existing gender inequalities from social and educational interactions. My own research has provided some evidence that such inequalities, as well as initial disadvantages in experience of information technology, can be lessened within certain educational contexts. Though these results might seem contradictory or at least in conflict, I do not view it in this way. Both projects used very different 32 S.J. Yates / Learning and Instruction 11 (2001) 21–34 research methodologies and some of the differences in the findings arise from the different types of insight that such methods provide. A recent study (Barrett & Lally, 1999) of a small group of 12 students using CMC on a distance-learning Masters in Education course has produced results consistent with those from linguistic and quantitative research presented above. Once again, the men posted more often and posted longer messages than the women. Once again, the women and men differed in the types of messages they posted. In this study men posted more social messages than the women though the women produced more ‘interaction’ messages that referred to others in the group. At the same time the survey and ethnographic elements of this study indicated that: “The data gathered in this project indicated that the use of CMC had provided distance learning students with additional opportunities for dialogue with tutors and peers [and] their involvement with the project had helped to reduce their sense of social isolation and made them feel part of a ‘community’ of learners” (Barrett & Lally, 1999, p. 59) Once again the focus on content and communicative practices highlights the problems and inequalities whilst the study of perceptions and outcomes has highlighted the benefits of CMC. I view this seeming conflict as arising from two sources. First, much of the work on gender and language use in CMC starts from one or a combination of the ‘difference’/‘dominance’ positions. These positions provide important insights into the gendered nature of interactions and to critiques of male dominance. It should not come as a surprise, unless we follow the ‘democratic model’ of CMC, that gender relations are a key element of CMC interactions. In contrast the survey and ethnographic studies of users perceptions and educational outcomes tend to focus attention on the benefits or problems with using CMC. Again we should not be surprised if men and women can perceive the benefits (and problems) of using a medium, especially in a distance education context, that adds to their overall educational experience. All of the research presented here leads one to conclude that the ‘democratic’ perception of CMC is seriously flawed. Indeed such a perception requires a very limited model of human communication and is probably based upon cultural conceptions of computing and information technology. Computing-based activity is often seen as being somehow lacking in social and interpersonal elements. Computers are ‘cold’ impersonal machines. CMC though is a hot medium used to express emotion, social relations and opinion as well as information. Just because the technology is presented as ‘genderless’, though this is very highly debatable, it does not mean that the interactions taking place through the technology will lose any of their complexity nor will the technology strip away existing social structures. The first section of this paper made this point very clear. Gender differences found in face-to-face communication can also be found in CMC with men engaging in similar tactics of exclusion and de-legitimation. CMC suffers, like all communications media, from the intrusion of existing social relations, including those that S.J. Yates / Learning and Instruction 11 (2001) 21–34 33 are based upon inequalities of access and power. As with all media of communication and education there is no ‘technological fix’. Other research discussed in this paper has noted that social context can, as in faceto-face interaction, be used to help limit or alleviate some of the difficulties faced by women in educational interactions. Such support can be provided directly in the form of women-only interactions. In mixed interaction contexts it is important to ensure that educational CMC systems are providing support for the types of interaction which both women and men may be seeking. Such support may be technical, through the provision of different discussion groups orientated to specific issues or types of interaction, as well as social, through ‘moderators’ or CMC system coordinators actively engaging in the prevention or reduction of exclusionary practices by CMC interactants. None of these solutions is new, they reflect existing educational practices where self-help groups and good tutoring or moderating of educational interactions seeks to achieve the same goals. As Kirkup and von Prummer make clear, supporting women in educational settings requires an understanding of their educational needs and goals and providing the support for these. It is in this area that CMC may provide not a technological fix but an educational and social opportunity by allowing women to use the medium to support their educational needs. The work of Herring, Collins-Jarvis and my own survey of distance education students all indicate that once the technology is socially and educationally organised in ways which women can adapt to their own norms then it can provide a very useful basis for educational interaction. References Barrett, E., & Lally, V. (1999). Gender differences in an on-line learning environment. Journal of Computer Assisted Learning, 15 (1), 48–60. Cameron, D. (1997). Theoretical debates in feminist linguistics: questions of sex and gender. In R. Wodak, Gender and discourse (pp. 21–35). London: Sage. Coates, J. (1986). Women, men and language. London: Longman. Collins-Jarvis, L.A. (1993). Gender representation in an electronic city hall: female adoption of Santa Monica’s PEN System. Journal of Broadcasting and Electronic Media, Winter, 49–63. Graddol, D. J., & Swann, J. (1989). Gender voices. Oxford: Basil Blackwell. Graddol, D. J. (1989). Some CMC discourse properties and their educational significance. In R. Mason, & A. R. Kaye, Mindweave: computers, communication and distance education (pp. 236–241). Oxford: Pergamon Press. Habermas, J. (1984). The theory of communicative action: reason and the rationalisation of society. (vol. 1). Cambridge: Polity Press. Habermas, J. (1987). The theory of communicative action: lifeworld and system: a critique of funtionalist reason. (vol. 2). Cambridge: Polity Press. Herring, S. (1992). Gender and participation in computer-mediated linguaitic discourse. ERIC Clearing House on Languages and Linguistics, October. Herring, S. (1993a). Gender and democracy in computer mediated communication. Electronic Journal of Communication, 3 (2). Herring, S. (1993b). Culture vs. technology, or why computer-mediated communication is still sexist. Paper presented to the International Pragmatics Conference, Kobe, Japan. Holmes, J. (1992). Women’s talk in public contexts. Discourse and Society, 3 (2), 131–150. 34 S.J. Yates / Learning and Instruction 11 (2001) 21–34 Kiesler, S., Siegel, J., & McGuire, T. W. (1984). Social psychological-aspects of computer-mediated communication. American Psychologist, 39 (10), 1123–1134. Kirkup, G., & Von Prummer, C. (1990). Support and connectedness: the needs of women distance education students. Journal of Distance Education, 5 (2), 9–31. Lakoff, R. T. (1975). Language and women’s place. New York: Harper Row. Lakoff, R. T. (1990). Talking power. New York: Basic Books. Mason, R., & Kaye, A. R. (1989). Mindweave: computers, communication and distance education. Oxford: Pergamon Press. Matheson, K. (1991). Social cues in computer-mediated negotiations: gender makes a difference. Computers in Human Behaviour, 7, 137–145. Newsweek (1994). Men, women and computers. May 16th, 1994. Tannen, D. (1984). Conversational style: analyzing talk among friends. Norwood: Ablex. Wilder, G., Mackie, D. & Cooper, J. (1985). Gender and computers: two surveys of computer related attitudes. Sex Roles: A Journal of Research, 13 (3/4), 215–228. Wyatt, N. (1993). Small group dynamics in a women’s studies hotline. Paper presented to the International Pragmatics Conference, Kobe, Japan. Yates, S.J. (1992). Gender and computer-mediated communication: an analysis of DT200 in 1990, CITE Report, 158. Yates, S.J. (1993a). The textuality of computer-mediated communication: speech writing and genre in CMC discourse. PhD Thesis, The Open University. Yates, S. J. (1993b). Gender, computers and communication: the use of computer-mediated communication on an adult distance education course. International Journal of Computers in Adult Education and Training, 3 (2), 21–40. Yates, S. J. (1993c). Speech, writing and computer conferencing: an analysis. In R. D. Mason, Computer conferencing: the last word? (pp. 37–56). Victoria: Beach Holme. Zuboff, S. (1988). In the age of the Smart machine: the future of work and power. Oxfore: Heinemann.