"Y'know, I'm fairly confused by your reaction here to Tannen's work, on several levels:
1) I'm remembering your talk about metaphors and how, for example, many men use the "football" metaphor, which is not very resonant to women (and non-sports-oriented men), which seems to me a pretty gender-based
language differential. ..."
When I do talks or seminars on crossgender communication, I do always bring up that metaphor difference -- most adult men in the U.S. who are native speakers of American English operate out of the Football metaphor in their non-intimate language interactions; their adult female counterparts tend to operate out of the Traditional Classroom metaphor. That information clarifies many of the standard male/female communication breakdowns. [It may be that this is changing for the younger generations, but that doesn't make it less useful; the principles hold and the strategies remain applicable, even if the specific metaphors are not the same ones.]
However, there is not in my opinion any credible evidence that the reason men lean toward the Football metaphor and women lean toward the Traditional Classroom metaphor is because of their biological gender. From the moment infants arrive in this culture at birth, people start herding them toward the "appropriate" metaphors. I can talk about this at more length if necessary, but will just sum up here by saying that in my opinion the metaphor split is caused by the process called "socialization." Boys are raised in the Football metaphor framework; girls are raised in the Traditional Schoolroom metaphor framework; they behave accordingly as adults. I'm not denying that the difference exists -- it certainly does, and it's important. But I don't think it has anything at all to do with biological gender; I believe that it's learned behavior.
"2) Your development in the Native Tongue books of a women's language, made up of words for concepts that are not only foreign to men, but inexpressible in their language. That would seem to pre-suppose a pretty deep language divide between genders. .... "
In the very small population of people who are aware of that language, there's a persistent idea that all it takes to make a language a "woman's language" is a different vocabulary, and that all I did was construct one. That's not accurate; I did a lot of things, in specifying the structure of the language, to tailor it for women. (There's some information about that at http://www.sfwa.org/members/elgin/Laadan.html and in the other sections on the language that are at that site.) But the "language divide" isn't linked to biological gender.
What human cultures do in the development of their languages is make it convenient to talk about some topics and cumbersome to talk about others. Whatever group has the most power over the language development process will determine which topics fall into the convenient category and which fall into the cumbersome one. It happens that -- for American English -- males have, over the course of history, had far more power to shape the language than women. It is therefore not surprising that topics that interest most men tend to be convenient to talk about, while topics that most men aren't interested in tend to be cumbersome to talk about.
Which leads to the stereotypical AE dialogue in which the woman is trying to make something that matters to her clear to a man, and the man is saying "Well, would you just get to the point, for god's sake? Why do you have to keep going on and on and on about it??" And the harder she tries to make her message clear, the more annoyed he gets (and usually, the less he listens). Our culture comes up with all sorts of reasons for this. "Men are logical; women are emotional." "Men just say what they mean; women talk everything to death." Tannen says men do "report" talk, women do "rapport" talk. (As you know, I disagree.)
It's not that what the woman wants to say is "inexpressible" in American English; it's just that it can't be said conveniently and efficiently, because the language doesn't provide the necessary resources.
This situation isn't the result of biological gender; it's the result of power and status differences. Languages can't pay attention to everything; choices have to be made. And those choices are made in accordance with the desires of the members of the culture having the most power. In American English, the language power has always belonged to men. For most of our history, it has been men who owned the publishing houses and printing presses, compiled the dictionaries, wrote the influential books, gave the influential speeches, ran the government, determined the structure of education, filled the presidency, and so on ad infinitum. If if had been the other way around -- if women had been the more powerful gender, in the same sense -- American English would have developed differently.
The idea that there are differences between male language behavior and female language behavior in American English isn't controversial. The controversy is over why those differences exist. One hypothesis is that they exist because of biological gender differences; the other (and the one I believe to be valid) is that they exist because of power differences.