I don't think it is. What I'm talking about is the
Sexual-Orientation Continuum theory. According to the APA (via Wikipedia):
Sexual orientation exists along a continuum that ranges from exclusive heterosexuality to exclusive homosexuality and includes various forms of bisexuality. Over at Razib's there was some discussion around Tim Pawlenty's answers on Meet the Press regarding homosexuality-as-a-choice. I thought he (Pawlenty) deflected it expertly, but some
commenters on
Razib's post referred approvingly to a rather rambling dissertation on the sexuality-as-continuum theme on another blog.
Now such a continuum would suggest that sexuality is normally distributed, with exclusive homosexuality and heterosexuality at the extremes, with intermediate behavior (i.e., bisexuality) the norm, taking up the vast middle of the distribution. Of course we know this is not the case - the vast majority of people are heterosexual. So the distribution is skewed - fine. Still, if it's a continuum, we should expect that the intermediate state should be also be intermediate in frequency between the dominant state and the rare state - i.e., we should expect bisexuality - which represents a state intermediate between the (predominate) heterosexual state and the (rare) homosexual state to have an incidence rate somewhere between the two. Does it?
Let's look at the
GSS. Since 1991, the GSS has asked participants about their sexual partners over the last 5 years(variable=SEXSEX5). The choices are Exclusively Male, Exclusively Female, and Both Male and Female. These are the cumulative results since 1991:
As you can see, rather than being somewhere between exclusive homosexuality and heterosexuality in incidence, bisexuality is actually
less common than homosexuality. That is certainly unusual behavior for an intermediate value. In case you might think having results from the early 90's might distort the results somehow, here are the results since 2000:
No real difference - bi-sexuality is the least common sexuality (results from individual years can be found
here if you don't wish to reproduce results yourself in the GSS). Now why would this be? Well I suppose i'm not qualified to speculate - but the results don't seem consistent with a continuum of sexuality, but more with a binary condition, with some switching between conditions as the situation might require in rare cases.
What about self-reporting issues? These results - sub-3% for homosexuality - are lower than we often hear. Claims of 10% are silly, but I've often heard 4 - 5%. Would GSS respondents under-report their homosexuality? Perhaps, though I don't know why they would do so here - year-in and year-out (only 2000 and 2002 for men and 1991 for females showed >3%) - and not elsewhere. But regardless, the point is that bisexuality shows a lower incidence of homosexuality - the opposite of what you'd expect if there were a heterosexual-bisexual-homosexual continuum. And it is certainly not plausible that bisexuality would be under-reported vis-a-vis homosexuality.