Adapting to the male majority
One other result from Herring's study is especially fascinating. Women on the LINGUIST list posted more opposed variants compared to women on WMST. One possible explanation might be that women linguists are just more irascible and argumentative than women involved in women's studies programs. Another, though, and one that may be more plausible, is that women - as the minority gender on LINGUIST - are adapting their posts to the dominant male style. When they are in the majority, the dominant male style that favors the expression of views through disagreement is not particularly in evidence, and norms evolve that favor a more supportive discussion environment. I'm tempted to conclude that those pioneering female Internet users I mentioned in an earlier chapter, who were doing more flaming in the newsgroups were adapting to the male-dominated online world, and going them one better.
Herring points out also that the minority gender on WMST - the men - may also have modified their style. One male poster, for example, sent in an opposed variant to WMST, arguing that a biological basis for certain sex differences should be considered, but he was rather indirect about it and included lots of hedges such as "it seems to me" and "perhaps." Recall that these hedges are slightly more common in women's speech compared to men's.
Victor Savicki, Dawn Lingenfelter, and Merle Kelley confirmed that the gender composition of online groups is significant, and it affects how we act. They analyzed 2,692 messages sent by 1,208 different people to 27 different newsgroups, looking for signs of gender differences in style, and comparing these to the proportion of women in the groups. Unlike the mailing lists, where you can obtain a list of all the subscribers and get some notion of gender ratios and size of the group, these discussion forums were wide open with no subscription needed. No doubt they had many lurkers whose gender remains unknown. The gender ratios in these groups are based only on those who actually posted a message.
First, and to the surprise of these researchers, all of the groups had more men than women, and almost 75% of the posts came from men. Of the remaining posts, about 13% came from women and the others were from people whose gender could not be identified. Even though women were always in the minority, the pattern of interaction styles differed depending on the size of that minority. For example, groups with a higher proportion of males contained more messages with language in which the author made opinionated statements disguised as facts, linguistically speaking. An example was, "The government is loaded with freeloaders," leaving out the "I think" at the beginning. The groups with more men also contained posts with more calls to action. Groups with a higher percentage of women had posts with more self-disclosure in them, and also more attempts at tension prevention and reduction.
As you can see, we are affected by the characteristics of the group around us, and some of the subtle gender differences we see online may diminish as the gender ratio evens out. The proportion of women online shot up from 31% to 38% between 1997 and 1998, and for the 10- to 18-year-olds, the sex ratio is even closer to even (46% to 54%).