The method
How a gender term gets an entry
The typical page on this subject is a list of forty terms with one sentence each: no etymology, no first use, no source, and no line drawn between an identity, a way of dressing and a pronoun set. It is the same forty terms on every site, lifted from each other, and the sentences get shorter each time they are copied.
GENDERSpec writes the other page. Every entry says what the word means, who first used it, when and where, which communities use it now, and where the record stops. Those are all claims about documents, so every one of them can be checked, and every one of them can be wrong in a way somebody is able to point at.
What an entry is filed as
One type per entry, decided by what the term names rather than by what it is about. Separating identity from expression is the most useful structural decision on the site: butch is not an identity in the way non-binary is, and the conflation drives most of the confused searching in this subject.
- 01Identity
- What a person is. The term names somebody's own sense of their gender.
- 02Cultural
- A gender category belonging to one people, named in their own language.
- 03Expression
- How gender is presented. A style and a set of codes, never an identity.
- 04Pronoun
- A pronoun set, with its full declension and its first attested use.
- 05Concept
- The vocabulary the rest of the catalogue is written in.
Every entry also records where its word came from: scholarship, a community, a dated post, or a living culture that had the category long before English had a word for it. That field is descriptive and it is never a ranking. A term coined in 2014 describes people exactly as real as one attested in 1858, and where the record behind a word is thin the entry says the record is thin. It never implies that the people are.
How an entry is written
The sources for a term are gathered and saved before anything is drafted, and the entry is then written from that material and nothing else. A fact that is not in it is left out rather than supplied from memory. Everything else here is a preference; that one is the rule the catalogue rests on.
Sources are read for what they can settle: the dictionary entry that dates a word for the date, the publication a term first appeared in for the attribution, the scholarship of a people for their own category. Where a term is widely repeated with no traceable first use, the entry says so, and no source earlier than this one has been found is one of the most useful sentences on many of these pages. Attribution sits inside the sentence that makes the claim, with the name and the year, because that is what survives being quoted somewhere else.
A culture-specific category is written from that culture’s own scholarship, and the entry names the specific people rather than a continent. Where a category resists being mapped onto English categories, the resistance is the most interesting thing on the page and the entry says so rather than smoothing it over.
Before an entry can be published, a checker has to pass on it: the answer in the first paragraph, the headings, every internal link, the punctuation, the third person, and the absence of any advice. It fails the build rather than leaving a warning. What it cannot check is whether a fact came from the material, which is why that rule is enforced by the method rather than by a script.
Entries are written in British English, revised rather than reissued, and each one prints the date it was added and the date it was last revised. The changelog records what changed across the catalogue, and the last revision was 2026-09-07.
What this site will not do
No advocacy, and not its opposite. The entries report what is documented: what a term means, who first used it, what the literature says, and where sources disagree. They do not argue a position on a contested question and they do not pretend one is settled. A reference work that smuggles a verdict in through the phrasing is worse than one that states it outright.
No medical content and no legal content. A term like gender dysphoria gets an entry saying what the term means and which document defines it. Nothing here discusses transition, hormones, surgery or eligibility for anything, and there is no register of what the law says in any country.
No self-assessment, and no score attached to a person. No quiz, no percentage of legitimacy, no ranking of terms by how real they are. An algorithm handing somebody a verdict about their own gender is indefensible at any traffic number. The one step an entry does carry, P1 to P6, measures how thoroughly the word has been counted and nothing else; the section below sets out what each step takes.
No tracking, no comments and no byline. There is no analytics script, no pixel, no advertising network, no account and no cookie, and every page is a static file, so no server holds a record of what anybody read. There is nothing for a reader to submit and therefore nothing to moderate. The catalogue publishes as an organisation rather than under a name, because what it asks a reader to trust is the sources and the method rather than a person.
How thoroughly a term has been counted
Every entry carries a step from P1 to P6, and the step is not a rating. It says how much of a population anybody has actually measured the word across, which is a fact about the evidence rather than about the word or the people who use it. A term can be widely used and never counted, and several here are.
The steps are defined by the kind of evidence each one takes, so two entries on the same step are there for the same stated reason. An entry names the dataset its figure came from, and where the material holds no figure at all the entry says so instead of estimating one.
- P1General-population scale
Identity measurable as a share of the general population in representative surveys (>=~0.5% or the population default).
- P2Major community term
Selected by >=10% of gender-diverse respondents (multi-select) in the Gender Census 2022-2025, or comparable share in Trevor Project nonbinary data.
- P3Established minority term
1-10% of gender-diverse respondents in community surveys (checkbox or write-in).
- P4Rare / microlabel
<1% in community surveys, write-in only, or no quantitative record at all; documented mainly in community wikis.
- P5Culturally localized
Prevalence meaningful only within a specific culture or region; population estimate given where one exists (census, academic, NGO).
- P6Historical / doctrinal
Category from religious-legal texts; no living population to measure.
- NMNot measured (umbrella/descriptor)
Institutional descriptor or umbrella concept not used as a self-identification label in surveys.
The samples are not the same kind of thing and the entries say which is which. Gender Census and USTS are self-selected community samples (shares of the gender-diverse community, multi-select, checkbox lists change yearly), not population prevalence. Only Williams Institute and Ipsos provide general-population estimates. Culture-specific figures mix census counts, academic studies, and advocacy estimates - each row cites its source.
How to quote an entry
Quoting is expected. Attribute the entry rather than the site, since the claim belongs to one page: GENDERSpec, the entry name, its URL and the date it was last revised. Carry the dates with the claim, because a first attestation quoted without its year has lost the thing this catalogue is for, and treat a sentence saying that a term has no traceable origin as a finding rather than as an omission.
Start anywhere: the A–Z index lists every entry on one page, the categories group them by theme, and the collections are the arguments, from the words with a known author to the categories that do not translate.