The Quiet Erasure of Same-Sex Attraction
How and when sexual orientation was conceptually undermined in Wikipedia
Wikipedia in many ways represents what the internet aspires to be - a vast and ever-growing repository of human knowledge, free and available to all, for the betterment of humanity. It is not perfect, but certainly for straightforwardly factual matters it is a resource that we take for granted - whether directly, or indirectly through other services that depends on its information.
Wikipedia works on a model of consensus, where any edit is presumed to represent the agreed viewpoint until it is disputed, and that this majority view tends towards “truth” over time. Disputes require credible sources to defend contrasting viewpoints, with edits reverted until consensus can be reached in discussions on an article’s “talk page” and then a new, uncontested view put forward. In a worst case, on straightforwardly factual matters alternative viewpoints are presented side by side, letting the reader know as fairly as possible that there is a dispute and leaving them to reach their own conclusions.
While in the vast majority of cases “good enough” is acceptable - and minor errors are harmless - there are situations where the consensus model subtly and dangerously breaks down. Where the demographics of the editors themselves, the bias in the construction of what even constitutes a “credible source”, and the independently contested nature of multiple terms has combined to reflect a false consensus. A consensus which now works against - and ultimately erases - a truth that a minority depend on for their very rights.
On July 7th 2010, the Wikipedia article on “homosexuality” was edited, with no fanfare, from:
This seemingly minor change was virtually the last change made by the user “Gladys Tuffnell” before being permanently banned for edit warring, and the only change to any page relating to sex, sexuality or gender in that account’s multi-year history.
The article that the word “gender” linked to at the time stated in its introduction:
In ordinary speech, it is used interchangeably with "sex" to denote the condition of being male or female.
So this unsourced, uncommented-upon change by a now-banned account will have been unremarkable. The article on gender was linked, but at that point gender was described in ordinary speech as a direct synonym for sex - which the vast majority of readers will have assumed without question. The formulation sex/gender implied an equivalence and it could be argued that linking both was redundant.
To reinforce this perspective, at the time, the opening sentence of the page for “lesbian” stated:
Lesbian is a term most widely used in the English language to describe sexual and romantic desire between females.
This change, however, marked the start of a slow creep in meaning across all pages connected with sex, sexuality and gender, with ideas about gender identity taking precedence over sex eventually becoming an unchallengeable, orthodox position.
So: at what point was there consensus that there was no such thing as same-sex attraction, and how did it come to be that all terms such as homosexuality, bisexuality, gay or lesbian were rewritten to accommodate members of the opposite sex?
On October 9th 2010, user Tommy2010 changed the page on “homosexuality” from “sex/gender” to “sex or gender”. This change was, again, made without any substantiation, and the account responsible made no other edits to any pages on sex, sexuality or gender. It seems an innocuous act of tidying up, yet has the subtle effect of breaking the clarity that sex and gender were regarded as synonyms at this point. “Sex/gender” and “sex or gender“ are slightly different linguistic constructs, and the latter form creates more possibility that sex and gender are not straightforwardly equivalent.
The next edit to that page was on October 11th by user Tijfo098, to change a citation link. This is unremarkable - however, one week later on the “gender” page, over the course of five hours, on October 18 2010 this same user made a series of 29 edits before eventually arriving at an introduction that completely removed the idea that “sex” and “gender” are synonyms, and instead replaced it wholesale with the conceptualisation of gender given by the notoriously unethical sexologist John Money.
This included the statement:
in most contexts, even in social sciences, the meaning of gender has expanded to include sex or even to replace the latter word.
At the time, user Tijfo098 - having started a few weeks earlier on board gaming topics - was highly active on pages such as Transgender, Transvestism, Sissy, Pedophilia, Pornography Addiction, Muscle Worship, Blanchard's transsexualism typology, Kink, Paraphilia and so on. There is nothing in principle wrong with that - someone has to edit these articles - but the issue is that this is the point when the page on “gender” switched to an opinion of sex and gender deeply rooted in essentialist sex stereotypes and transsexualism. It is reflective of a bias towards pornography, kink and fetish that isn’t compatible with a neutral perspective on the reality of same-sex attraction, particularly for lesbians. After a month or two of high activity in these areas, the editor moved on for the subsequent two years of activity on articles mostly around logic and computer programming before the account was abandoned in 2012.
This background change meant that the page on “homosexuality” which linked to it no longer reflected exclusive same-sex orientation, though this was completely invisible to any casual reader who assumed that of course sex and gender were synonymous.
There was no debate on this, nor did it truly reflect a global shift in understanding. There is no sense in which - in 2010 - all of humanity was united in believing that gender was anything other than a plain synonym for sex, or that sexual orientation was towards gender, not sex. Certainly, as a case in point, in the UK Equality Act sexual orientation is protected in law as sex-based.
Not much can be read into choice of username, but as an indication of the level of casual thoughtless arrogance necessary to erase homosexuality as a concept, “tijfo” is an acronym for “Today I Just Found Out”.
In August 2011, some editors had noticed this shift in meaning and its consequences, and debate broke out on the talk page of “homosexuality”. However, attempts to clarify and remove the “or gender” part of the introduction were reverted.
The back and forth in the edits looks like:
Improper use of the word gender. Animals come in sexes. Words and plumbing fittings come in genders. They are not the same and "sex" is not a dirty word. See, e.g. the 19th amendment to the U.S. Constitution
Read the gender article. As a linguist, I'm very familiar with the linguistic use of the term. However, it has another meaning that refers to human beings and is distinct from "sex" (which is why both are included here).
"Homosexualtity" refers to members of the same physical sex only. That's what sex is. A straight man is not, for example, a lesbian no matter is internal gender identity. Using the word gender in other than is actual meaning is confusing.
You still don't seem to understand that a person's sex and gender can be different, which is why the article uses both terms. Please take it to the talk page.
This continued on the talk page, with exchanges like the following:
I will make my original point one last time. A male having sex with a female is NOT engaging in lesbian sexual activity REGARDLESS of his internal gender orientation. The word homosexual relates to persons of the same sex. It has absolutely nothing to do with their internal mental orientation.
And the response:
You keep talking as if somewhere there's a stone tablet on which the definitions of words were carved by God, and on which the word "homosexual" is defined as "romantic and/or sexual attraction or behavior between members of the same sex, regardless of gender". But there's no such tablet, that's not how language works, and I'm pretty sure that's not how most relevant researchers understand the term "homosexual". Now, I'm sure that many people understand the term to exclude couples where each member is of a different sex, regardless of their gender; you're clearly one of them. That's fine. You can use words however you want in all sorts of spheres of life. But Wikipedia is not one of those spheres. Here you have to find consensus. And, as I understand it, the consensus is that a person with an X and Y chromosome could certainly be called a lesbian. So if you want to change the article, you're going to have to provide reliable up-to-date sources to support your claim to the contrary
This editor had decided that the academic consensus was overwhelming that men can be lesbians, and if you want to say otherwise you’ll have to back that up with a source.
Anyone familiar with the current sex/gender debate will recognise exchanges like this. One person insists that the idea that men can be lesbians erases the very concept of lesbianism, and the response amounts to “get over it, men can be lesbians”, with deference to academic gender studies.
Essentially, from this point on, same-sex attracted people would be unable to assert their own existence or have a word in language for themselves without building up a significant body of academic work in justification.
So to recap:
Before July 7th 2010 - Homosexuality means same sex attraction
July 7th 2010 - Homosexuality means same sex/gender attraction, where gender still means “sex”.
October 11th 2010 - Homosexuality means same sex or gender attraction, but again where gender still means “sex”.
October 18th 2010 - Homosexuality means same sex or gender attraction, where gender now means whatever sexist stereotypes the unethical quack John Money decided it means.
August 30th 2011 - You can’t change homosexuality back to meaning exclusive same-sex attraction without an academic source.
I think overwhelmingly people were unaware that this shift had taken place, and that the plain meaning of homosexuality had been erased in this way - and few reading it at the time would have predicted this as an outcome - but as of October 2010, the consensus that Wikipedia purported to represent was that there was no word for exclusive same-sex attraction in humans anymore, whether anybody realised it or not.
And this consensus seems to have been arrived at as a result of two long-dead accounts making their solitary, unsourced contributions to this topic, and one editor deeply enmeshed in niche, contentious and pornographic subject matter at the time.
If one - probably male - person taking a break from board gaming to spend a few weeks immersed in sexology, pornography and pedophilia decides that John Money’s ideas of gender supersede sex without even considering the impact this has on those who are same-sex attracted, there is no adequate counterbalance to this bias. The self-selecting nature of those with the time and capability to edit these articles, and the impact of this sort of information being curated in this way is immeasurable.
Obviously this societal-wide erasure of meaning cannot be boiled down to a single action, but this particular sequence of events is indicative both of the timing, and the attitudes that have driven it. It should not be the case that men whose primary experience of lesbianism is as a category in pornography can ever be in a position to dictate to lesbians what they are.
With Wikipedia such an important part of the machine learning corpus that drives technology platforms like Google, such hidden assumptions and biases will simply propagate unchecked through all levels of our technological infrastructure, compounded by the fact that the people responsible for this process tend towards the same biases. Online content surfaced by search results will adhere to these biases, gain in popularity, and feed back into future machine learning, reinforcing the same bias continually. What is in Wikipedia is what is returned by search results, and therefore accepted by the casual reader as true, or at least “true enough on average”. Any error in the data is assumed to be cancelled out by sheer volume of information, such that mistakes are minor and self-correcting. So what if fact checking AI trained on this data ends up flagging social media comments by people claiming homosexuality means same-sex attraction as “misinformation”?
There is no sense in which fundamental aspects of human existence are treated separately and shielded from this process. Our technological infrastructure becomes a relentless machine for driving the cultural viewpoints and blind spots of its architects, seemingly not by design, but by thoughtlessness.
And this infrastructure is not somehow separate from our culture - it is our culture. Our mode of communication shapes our mode of thought, and its bias alters what we are even able to say - which, again, feeds back into the same machine learning processes. Sexist ideas of gender supersede sex, so now platforms like Twitter will not allow you to report sexist abuse, only attacks based on “gender identity” and “misgendering”.
So we end up trapped in a false consensus, created by the inability to communicate any other reality. When you cannot name it, you cannot measure it, and it ceases to exist.