One gets used to anticipating the right questions in a subject from broad (not necessarily deep) experience with it. For me, it's fun to be listening to a podcast or reading an article and immediately asking the right questions, which is nearly always answered with a quick check "Yes someone saw the same thing i did, and there is subsequent research on it." There is some satisfaction in feeling smart about it, but more in knowing the question I am asking has some kind of answer.
Today's example: Reading about whether there is an identifiably gay voice. Short version: there is, but it doesn't much match the stereotype and its usage is complicated. The article started off with some history about the stereotypical lisp, and that it might have derived from the old Freudian idea of male homosexuality of arrested psychosexual development, as children are more likely to lisp. They included links to movies and TV comics in the 50s where that was the primary feature of the supposed gay man's speech, followed by some mention that previously that homosexuality wasn't referred to in popular entertainment at all, and perhaps even being mocked was a step up of recognition from invisibility and "the sin that dare not speak its name." That this took place in Hollywood, where some entertainers were known to be gay - and no big deal - but the information was kept from the public for PR purposes was considered evidence for that idea. Then a small study in Minnesota was discussed in which people could somewhat reliably tell which voices were gay and which straight while reading identical texts (more on this later).
All the while I am thinking code-switching. The answer is code switching. Gay men will exhibit more identifiably gay speech when they are in situations where it is accepted or even encouraged, but less of it when it is irrelevant to their task or even a hindrance, such as at work. Hasn't anyone studied this? Heck, even I know that's what you should be doing.
Enter Rob Podevsa of Stanford who did speech as an ethnography, following people around with tape recorders in a variety of situations and evaluating the results for his dissertation in 2006. I will note that this is rather late in the day for someone to figure that out, but at least it happened, and things were as I predicted. Gay men had more distinctive speech when they were together in informal situations than when they were explaining things to people at work, to take one example. This is hardly surprising, because this is what we all do all the time. Our regional accents become more pronounced when we are with the people we grew up with. We are more careful in pronunciation when we are in more formal settings, or when we are expected to be more authoritative. Every child knows that the situation has changed when Mom or Dad suddenly starts saying consonants more crisply. Yipes. Podevsa was not measuring just gay speech but many people after having learned their characteristics at the outset and deciding who to follow. Some people are very susceptible to this and their speech varies considerably over situations, while others are more constant. African Americans and Hispanics engage in enormous amounts of code-switching, and can usually size up exactly how much ethnicity to show before they have even heard anyone speak.
I was very aware of this as far back as college, before I knew what code-switching was. I had zero awareness of anyone being gay while I was in high school, but entering a college theater department, and briefly even a stint as a dancer, I learned the territory quickly. Then sophomore year I ended up by luck of the draw an a floor that was about 40% gay, and thus got to see these peers in a variety of environments: at home, at work, and at play, as it were. My roommate pointed it out with great merriment about another econ major he had classes with the year before and also that semester, noting that he never picked up he was gay before, by then saw him going out with platform shoes and purple eyeshadow on a Friday night and said "He even talks differently! You know, flame on!" And it put words to what I had only dimly noticed but now could see everywhere. I even kidded a couple of friends about it when they asked "How did you know Marty was gay? He's pretty discreet about it?" I chuckled at them, saying "Because you two started speaking gay to him the moment he came into the room." College age would be particularly fertile ground for all this, as young people in general are sexually advertising more than others, yet also have to start finding their places in a sober adult world.
Interesting features of gay male speech as recorded (others have now done more extended versions of that ethnographic study). It is not a lisp, but overprecise "s's" that are characteristic. In fact, more precise pronunciation is another characterisation of gay male speech that holds up empirically. People both gay and straight find it harder to distinguish female sexual preference from vocal characteristics alone, but one clue with both is the adoption of more coastal, urban speech patterns, dropping regional ones. That was first picked up in that Minnesota study, where it was noted they displayed less of the distinctive Minnesota vowel sounds and used hipper, more modern slang. This was confirmed with Southern accents later. It's not that those disappear, but they become more subdued. It makes an intuitive sense that there would be a perhaps unconscious declaration "these are more my people than those." There is more variation in tone and pitch, a greater expressiveness that is also associated more with females. Yet as far as I could find there is not a corresponding move to the middle or more masculine speech by lesbians that was picked up, and one study expressly contradicted that. Real life is different than stereotypes, but stereotypes usually capture something vaguely accurate.