Mark Liberman goes over the research on language and gender and, looking at one particular language feature, asks whether the empirical evidence really stands up to assertions that women use more “tag questions” in their speech than men. He concludes that the evidence does not support the claim, in fact the opposite is true; however, while men may use more tag questions than women, it does seem that women do use a higher percentage of “affective tags.” These are phrases which often serve to move a conversation along. In this respect he seems willing to concede Deborah Tannen’s assertion that, as he puts it, “women are saddled with a higher proportion of ‘interactional shitwork.’”
Personally, I think the real question is whether the number of tag questions used by women increases when they are trying to act more feminine? In other words, if we don’t treat language and gender stereotypes as being accurate depictions of how people actually speak, but as social norms. William Labov famously mapped such shifts in the use of “r” sounds by different classes in New York City. The important finding being that r-lessness was seen as lower class (think of cab drivers saying “toidy toid street”), so people would add an “r” sound to their speech in order to sound more “standard.” This resulted in some people adding “r” sounds where they didn’t belong, such as saying “terlit” for “toilet.” (This phenomenon is called “hyper-correction.”) But he also found that women used the standard form more than men from similar social backgrounds.
Penny Eckert, in her book Jocks and Burnouts, found that high school girls are likely to be either the most standard or the most non-standard in their phonology. That is, women’s speech was not just more standard than was men’s, as Labov found, but when it was non-standard it would be more non-standard than men’s as well! In other words, it was more marked as either local “Chicago” speech, or standard speech, while men’s speech fell more in the middle. It is also important to note that these shifts were most noticeable for those phonological features which the speakers themselves were aware of as markers of regionalism. There might not be any noticeable shift in phonological features which linguists know to be non-standard, but which the speakers themselves are unaware of.
So, going back to tag questions, I wonder how aware speakers are of these as markers of gender, and how much of a shift we might observe when speakers are trying to sound more or less “feminine”? For instance, when out on a heterosexual date? Or when spending time with other women (or men), without any men (or women) present?