With monkeypox in the news, some people are encountering the phrase “men who have sex with men” (“MSM” for short) for the first time. I’ve even seen some hot takes positing that it’s a new invention being pushed by woke moralists.

Screenshots of some tweets describing 'men who have sex with men' as a new or unfamiliar phrase

First off, it seems odd to read this plain English phrase which means more or less what it says on the tin as something sinister or euphemistic. Why not just say “gay” (or “gay or bisexual”)? Because the Venn diagram of these sets has non-empty “lobes” – there exist celibate gay men, and, most importantly, there exist men who have sex with men but who do not identify as gay or bisexual. The potential reasons for this are many, but consider for example a man who…

  • Is attracted to men, but denies this attraction out of shame
  • Acknowledges his attraction to men, but eschews the label “gay” because he does not identify with the cultural stereotypes of this category (the British physique photographer John S. Barrington is a good example of this)
  • Engages in homosexual acts when confined to an all-male environment (such as a boarding school or prison), but would not do so otherwise (“situational sexual behavior”)

When it comes to messaging about diseases like monkeypox or AIDS, a person’s risk profile is more a function of their behaviour than their orientation or tribal identification.

Secondly, “men who have sex with men” is definitely not a new term or concept. Many of the replies to the above tweets correctly point out that “men who have sex with men” has been in use for a long time, though the commentariat is vague or inconsistent about exactly how long (“a few years at least”, “decades old”, “since at least the 90s”, since “around 1994”, “it dates to the 1980’s”…).

This post is the result of a weekend spent on a mission to more precisely pinpoint the origin and spread of this concept.

Prior work

Wikipedia claims that men who have sex with men has been in use “since 1990 or earlier”, and that the initialism MSM was coined in 1994 by dentist Michael Glick, in an article about periodontal issues afflicting HIV patients. These dates are cited to a 2005 article from the American Journal of Public Health, “The Trouble With “MSM” and “WSW”: Erasure of the Sexual-Minority Person in Public Health Discourse”.

Another source, a 2011 paper by Tom Boellstorff, “But Do Not Identify As Gay: A Proleptic Genealogy of the MSM Category”, gives two 1988 quotations as the first published references to the concept (“to my knowledge”), an antedating which has since been repeated in a number of places (e.g. here, and here).

The Oxford English Dictionary has an entry for “men who have sex with men”, with its earliest quotation also from 1988. Though, whereas the quotes given by Boellstorff are from academic papers, the OED’s 1988 quote is from a newspaper, the New York Amsterdam News:

The NYC Department of Health is requesting..applicants..who have the ability to communicate with the following populations: men who have sex with men [etc.]

The OED also has an entry for “MSM”, with two citations that predate the supposed 1994 coinage claimed by the paper cited by Wikipedia. The earliest is from 1990:

The Gay Men’s Health Crisis (GMHC) convened focus groups with black men who have sex with men (MSM).

men who have sex with men before 1988?

Google Books turns up many instances of the 6-gram “men who have sex with men” well before 1988. Here’s an example from March 3, 1978, from Jim Fitzgerald, a columnist for the Detroit Free Press (available through the Internet Archive library):

Anita Bryant recently said if government protects the civil rights of men who have sex with men, and women who have sex with women, it should do the same for “people who have sex with dead people and St. Bernards.”

But we’re not just interested in when these 6 words were put together in a particular order. We’re interested in when it was first used with a certain meaning – as a category that deliberately includes men who engage in homosexual behaviour but don’t identify as homosexual, bisexual, etc. In the quote above, Fitzgerald seems to be using the phrase merely as a synonym for “gay men” (perhaps to create a natural rhetorical parallel with “people who have sex with dead people and St. Bernards”) – in fact, the full Anita Bryant quote he references is: “Making homosexuality a civil rights issue would mean you have to give minority-group status to prostitutes, to people who have sex with dead people or St. Bernards.”

The earliest uses of the phrase in the modern public health sense that I was able to find on Google Books were in a couple US government documents from 1987, a year before our earliest known citations. Here’s one example:

Evaluation of California’s AIDS Community Education Program, a government publication from 1987 (Google Books preview):

Men who have sex with men but are not gay-identified had the lowest levels of AIDS-related knowledge and were the least likely to modify the behaviors. These men, who typically relate to the exclusively sexual subculture, may keep their homosexual contacts hidden from other parts of their lives.

However, newspapers.com turns up a handful of earlier occurrences stretching all the way back to an article from January 1983(!) in the Springfield News-Sun, early in the AIDS crisis (the article describes AIDS as “a mysterious lethal disease that attacks the body’s own immune system”).

The article, “Blood centers screen area donors”, deals very directly with the problems motivating the MSM/gay distinction. It reports on some local blood centers which have begun “asking donors whether or not they are homosexuals before taking their blood” as a precaution, given the existence of some evidence linking AIDS to gay men. The article’s author interviewed a gay activist, Howard Getz, who objected to the formulation of the screening question:

Getz called the question inaccurate because AIDS had not been diagnosed among any lesbians. Also, men who have sex with men but do not consider themselves homosexual would be permitted to give blood and could taint the supply as easily as an admitted homosexual, he said.

(Not “admitted homosexual”.)

Getz’s position would soon be vindicated. In 1985 we see an AP story, “Men who have had sex with men urged not to give blood”, and a 1987 article in the LA Daily News quotes “Are you aware that men who have sex with other men run a high risk of contracting acquired immune deficiency syndrome?” as one of the standard questions nurses are asking prospective blood donors.

Some other early examples from newspapers…

From a 1984 letter to the editor in The Sydney Morning Herald:

It is now recognised that a highly significant danger in spreading AIDS outside the gay community is through heterosexually identified men who have sex with men as well, sometimes in parks/toilets, often in steam-baths.

“Health officials ignore AIDS threat in prisons”, a 1985 article in The San Bernardino County Sun:

There’s little question that jails and prisons are potential breeding grounds for AIDS. A large number of the 700,000 Americans behind bars are drug users and men who have sex with other men. […] “Sexual activity inside prisons is the by-product of a number of factors that bear no relationship to inmates’ sexual orientation prior to arrest,” New York City Police Commissioner Benjamin Ward wrote the New York Times recently.

The 1984-85 San Francisco AIDS survey

In the early years of the AIDS epidemic, the San Francisco AIDS Foundation commissioned a telephone survey, with a first wave of interviews in August 1984, and follow-ups in April 1985. The foundation sought information about men’s sexual practices, their awareness of AIDS, and their consumption and acceptance of public health messaging about AIDS prevention. Though a year later than the earliest quote in the above section, this is a great early example of the MSM paradigm being successfully used to a practical end.

Headlines about the study characterized it as a survey of “gay and bisexual men”, but the full report’s methodology section is more nuanced:

To be qualified, a respondent had to report that he at least occasionally had sex with other men, or to identify himself as gay or bisexual.

Here’s an excerpt from the script used by the interviewers, who phoned up a random sample of SF households:

For the rest of this survey, we are interested in speaking with one group of people who are at highest risk for AIDS: men who have sex with other men, or who identify themselves as gay or bisexual. Would you include yourself in one of these groups?

(IF HESITANT, CONTINUE:) We are also interested in speaking with men who may only occasionally have sex with other men. If you fall into one of these categories, we would very much appreciate it if you could complete the survey. As I said in the beginning, in no way will your name ever be associated with this survey and your answers will help the AIDS Foundation in their prevention campaign.

In the 1984 survey, 81% of respondents described themselves as gay, 6% as bisexual, and 13% as “other” – which is to say, enlarging their criteria to “men who have sex with other men” had a very real benefit in this case.

Incidentally, the whole report is a fascinating read, peppered with surprising and occasionally moving details. For example:

The general impression of interviewers was that heterosexual men were remarkably sympathetic to the survey and its importance – 85% of non-qualified male respondents were noticeably supportive, while the remaining 15% expressed the range from indifference to what is typically construed as a homophobic response.

Of the 19 men who conducted the first survey, all but one were gay or bisexual, and their “gaydar” seemingly proved very useful:

We also devised a system to keep track of respondents who initially refused to indicate whether they were qualified or not but whom the interviewer sensed probably were qualified. As the survey period progressed, we approached these refusals again with successful results. We completed interviews with 25 of the 62 refusals contacted, or 40%.

Meta commentary

The first instances I was able to find which commented directly on the phrase “men who have sex with men” itself were from 1988, and it’s worth quoting a couple of them…

From some 1988 US Senate hearings related to AIDS:

Our [The Latino Aids Project’s] target groups are men who have sex with men, Latino(a) intravenous drug users, and Latino(a) youth who are at risk for HIV infection[…] The reason I have used the term “men who have sex with men” is because in our community we have men who participate in sexual activities with other men, however, they don’t identify as being gay, homosexual or bisexual.

This anecdote from a 1988 research article, “Integrating Safer-Sex Counseling into Social Work Practice”, by Michael Shernoff powerfully drives home the practical value of this framing:

In attempting to assess whether a client is at risk for AIDS, clinical social workers must ascertain both what the client’s current sexual practices are as well as what they were in the past. Simply asking ”Are you gay?” is not sufficient. Health care professionals cannot assume that a client who is not openly gay has not engaged in sex with other men. For example, a married man with numerous symptoms of AIDS-related complex (ARC) who had never been transfused, who reported no history of shared needle use or other risk factors for exposure to HIV, and who stated he was not gay baffled his physician. However, when the man was questioned by a social worker as to whether he had ever had sex with other men, the client stated that he had a long history of sexual activity with men.

Thus many men who have sex with men do not label themselves as homosexual and certainly do not identify with the gay community. Questions regarding sexual practices must be asked in an accepting, nonjudgmental, and gentle way that does not incorporate the use of labels. For instance, the practitioner may ask: ”As an adult have you ever had any sexual contact with another man?” If the answer is ”Yes,” then asking ”When was the last time?” can provide useful and pertinent information about the client’s possible exposure to HIV.

MSM initialism before 1990?

OED’s earliest quotation for the initialism “MSM” is from 1990 (handily beating the article cited by Wikipedia which claims a 1994 coinage). But it turns out we can also push that figure back slightly:

Here’s an excerpt from a 1989 Senate report:

AIDS incidence is growing more quickly among intravenous drug users (IVDUs) than it is among men who have sex with men (MSM):

  • In 1988, for the first time, more new AIDS cases were reported among IVDUs than among MSM (43.4% vs. 43.2%).
  • From June to December of 1988 MSM fell from 57.9% to 51.4% of […]

As I wrote in my earlier post about the history of “pro-life” and “pro-choice”, it’s important not to lose the forest for the trees when searching for the earliest occurrences of a particular term. We should also consider when a term was popularized, and how its usage has changed over time. The Google Books Ngram Viewer is helpful here:

This roughly indicates that “men who have sex with men” began to see use in the mid-to-late 80s, increasing sharply in frequency through the early 90s, and more gradually since then. Popular use of the initialism “MSM” in print has trailed this trend by about 10 years.

(Technical notes: the ngram viewer maxes out at 5-grams, which is why I used a “shingling” approach of plotting “men who have sex with” and “who have sex with men”. The fact that their lines track each other so closely (I set smoothing to 0 for this plot) suggests they both mostly occur as part of the same “men who have sex with men” 6-gram. The occassional use of the phrase “men who have sex with other men” may explain why the “men who have sex with” shingle consistently floats slightly above the other.)

To get a rough sense of the absolute numbers involved, Google Scholar turns up 1,080 results for the phrase “men who have sex with men” in 2000, 4,920 in 2010, and 11,700 in 2020 (though these numbers may or may not be fuzzed).

Yeah, but what about in the media?

At this point, a skeptic might concede that MSM has been widely used for a long time in technical publications by public health experts, but how come all of a sudden the terminology is being used in the mainstream media (the other MSM)?

Well, here’s a chart showing the annual number of articles using this phrase in the New York Times. “men who have sex with other men” first appears in a March 1987 article (headline: “4 New York Bathhouses Still Operate Under City’s Program of Inspections”), with the shorter form debuting in 1993:

Number of NYT articles containing "men who have sex with (other) men" per year from 1993-2022

The Times did use the term during the first two decades of the AIDS epidemic, but only rarely. Its usage picks up rapidly in the early 2000s. 2022 is a record year by a large margin, but this doesn’t imply an increased push to use this term. It’s more easily explained by the fact that a disease disproportionately affecting MSM is a major news story. If we excluded articles that include “monkeypox” from the above chart, the bar for 2022 would be at a mere 3 so far.

Tagged: Linguistics