The third gender in twentieth-century America

Journal of Social History, Winter, 1996 by Randolph Trumbach

George Chauncey's brilliant and often persuasive study of male homosexual relations in early twentieth-century New York was published two years ago on the twenty-fifth anniversary of the Stonewall riot that inaugurated the recent gay liberation movement. The world that he describes was the product of a major shift in western sexual behavior that had begun two hundred years before, around 1700. And his book is in dialogue with the scholars who over the past twenty-five years have tried to analyze that shift. The nature of the problem to be discussed can be indicated by asking whether homosexuality and heterosexuality are biological categories that divide the world into a majority and a minority that can be found in all times and places. To such a question most western people today would reply yes. And while they would probably wonder why a minority should be homosexual, they would simply accept without question that most people are heterosexual. Since the 1970s, however, the work of some historians and sociologists has radically challenged these presumptions. Mary McIntosh in a classic article in 1968 began the discussion by proposing that homosexuality in modern society was a deviant role into which some men were socialized beginning around 1700. Nine years later in 1977 Jeffrey Weeks and myself, under McIntosh's influence but independently of each other, rephrased McIntosh's proposal. Weeks maintained that the modern homosexual role emerged in the late nineteenth century when the concepts of homosexuality and heterosexuality were invented.(1)..... click here to read more.

Sounding gay

At first blush, schoolyard taunts might not seem a fruitful source of truisms about the human condition, but the epithet “it takes one to know one” does have some basis in reality, at least for gays and lesbians, who are often attuned to the special somethings that subtly and covertly distinguish gay from straight. In gay vernacular, the ability to identify who is gay and who isn’t is termed gay radar, or gaydar. But exactly which signals does gaydar pick up? And what differentiates the signals given off by gays and straights?

The voice offers a few clues. Though popular stereotype holds that gay men lisp, lisping is quite rare and is often evidence of a diagnosable speech impediment. The most famous lisper alive today, boxer and convicted rapist Mike Tyson, hardly fits into the “gay” category. But the lisping stereotype is not altogether removed from reality: In fact, some North American gay males do pronounce sibilants (s, z, sh, and the like) in a distinctive way – by adding more sibilation, hissing, or stridency, a phenomenon phonologists call assibilation.


Here, of course, the dangers of stereotypes become apparent. Gay men are not the only group whose members sometimes speak with assibilation. A habit of assibilating “stops” like t and d is a prominent feature of Quebec French, for example, and the source of much derision from national French speakers. A word like térébentine (“turpentine”) in certain Quebec French dialects is pronounced something like tsérébentsine. Many New Yorkers of all persuasions, and some American Jews, also assibilate in ways similar to Quebec French or stereotypical gay speech. Moreover, gay men who speak with what a North American newsreader would consider an “accent” – such as British, Australian, or even Texan gays – rarely assibilate at all. Nailing down just what makes a gay voice gay is as vague and slippery as human sexuality itself.

That’s not to say the problem hasn’t been studied. In one experiment, Rudolf Gaudio, an openly-gay linguistics student at Stanford University, asked four gay and four straight men to read two passages into a tape recorder. The first text was a dry excerpt from an accounting volume, the second a dramatic passage from Harvey Fierstein’s play about gay life, Torch Song Trilogy. A group of 13 subjects of both sexes listened to selected snippets of those recorded passages and ranked each one according to a “semantic differential” technique, i.e., on a seven-position scale between opposite terms: straight and gay, effeminate and masculine, reserved and emotional, affected and ordinary.

As Gaudio noted, “listeners’ guesses about speakers’ sexual orientation were largely accurate: with ‘straight’ at the left pole of the continuum and ‘gay’ at the right pole, all the straight speakers rated on the ‘straight’ side, and all the gay speakers were to their right (i.e., sounded ‘more gay’).” That pattern held true for both the accounting and dramatic passages.

Gaudio’s research was not concerned with gaydar per se; rather, his interest was in correlating pitch measurements with the listeners’ ratings. Oddly, though, in a range of pitch measurements taken from the actual sound waves of the four gay and four straight men’s voices, there was no significant correlation with the listeners’ judgements. The experiment, then, could provide no quantifiable reason why the listeners’ perceptions about gay and straight speakers were correct.

Gaudio explains this anomaly by noting that his experiment considered only a narrow range of measurements; gay and straight men’s speech might well differ according to criteria Gaudio did not measure. A leading openly-gay linguist, Arnold Zwicky of Ohio State University, echoes that interpretation and adds that gay men’s speech can differ from straight in a number of ways; listeners might pick up on only one or some combination of those factors – and not necessarily the ones Gaudio measured. Still, the likelihood of further research in this area, according to Gaudio and Zwicky, is remote due to the political touchiness of studying gay speech.

[Originally published in the Economist, 1995 ¶ Posted 2001.05.07; updated 2006.07.26]