Fandom and identity science

The International Anthropomorphic Research Project (IARP), written up for the public as Furscience, is the main empirical program on the furry fandom. Across convention and online samples it treats a fursona, an anthropomorphic animal persona, as a typical practice. Participants rate that persona as carrying more desirable traits than they assign to themselves (more attractive, confident, energetic, playful; less shy and disorganized, on the project's own wording of the figure). They also describe the fursona as a way to try something they do not get in ordinary life.

In a separate sample of 1,113 furries, sexual attraction to parts of the interest was present, and some fan behaviors were sexually motivated. Non-sexual motives, including social belonging, were larger and correlated more strongly with furry identification (Brooks et al., 2024).

Species choice is measured on the same project. Summer 2020, international online volunteers, let people mark more than one species, so the shares sum past 100 percent: other 23.9%, wolf 20.7%, dragon 16.0%, fox 15.8%, hybrid 14.0%, dog 10.6%, housecat 8.2%. Wolf was the most common single named species in that wave. Dragon outranked fox for second place. Gerbasi and colleagues (2008) had already found dragons, canines, and felines, and no nonhuman primate. The shape of that list is the specificity question, treated on its own in the integration.

Furscience, section 3.12, "Fursona as ideal self," section 3.1 on species popularity (Summer 2020), and the Anthrocon 2017 summary on the same site. Brooks, Bennett, Myhre, Plante, Reysen, Roberts, and Gerbasi (2024), Journal of Sex Research, 61(2), 324-335. Roberts, Davies-Kneis, Gerbasi, Fein, Plante, Reysen, and Côté (2024), Deviant Behavior, 45(1), 50-79, describe the program's scope: on the order of dozens of studies and tens of thousands of participants, by the authors' own count, reached through conventions and online volunteers.

E2. Several large convenience samples, more than one wave, and a within-program replication of the ideal-self pattern. The species percentages are E2 for that multi-select item in that wave, not a census of fursonas. There is no probability sample of the general population, so nothing here is a prevalence of furries in a country. The project itself says the causal arrow between an ideal-self fursona and wellbeing is still open. The explanation of why wolves and dragons lead is E3 and lives with the specificity hypotheses.

People who attend a convention or finish a long survey differ from people who only lurk. Trait ratings are self-report. A person who is already doing well may build a brighter persona, rather than the persona producing the wellbeing. Sexual motivation is real in the Brooks sample and is the wrong thing to delete. It is also the weaker correlate of identification in that same sample.

Does changing the fursona toward valued traits change later wellbeing, or does wellbeing come first? The project has pointed to a longitudinal study and has not, in the pages opened for this site, published a settled causal result.

• A3 Fursona theory
• C14 Proteus effect
• I50 Erotic art and identity
• B11 Belonging
• Integration, specificity

Gerbasi and colleagues surveyed 217 furries and 29 non-furry attendees at a convention, plus 68 college students. Furries named dragons and canine and feline species as an alternate-species identity. None named a nonhuman primate. Two yes/no questions ("do you consider yourself to be less than 100% human" and "if you could become 0% human, would you") made a four-cell typology. The abstract calls the axes self-perception (undistorted versus distorted) and species identity (attained versus unattained). One quarter of the furry sample said yes to both, the cell the authors called Distorted Unattained.

The abstract then says that this type "has certain characteristics paralleling gender-identity disorder," and that the typology and a proposed "Species Identity Disorder" need further research. That is the paper's framing. It is a proposed construct, attached to a convention sample, not a demonstrated disease.

Gerbasi, Paolone, Higner, Scaletta, Bernstein, Conway, and Privitera (2008), Society & Animals, 16(3), 197-222. Probyn-Rapsey (2011), Society & Animals, 19(3), 294-301. Gerbasi, Scaletta, Plante, and Bernstein (2011), Society & Animals, 19(3), 302-304. Roberts, Plante, Gerbasi, and Reysen (2015), Health & Social Work, 40(2), e42-e50.

E2 for the survey counts and the species names. E3 for Species Identity Disorder, because the authors marked it as a proposal. The gender-identity parallel is their interpretation of item endorsements, including a feeling of being a non-human species "trapped in a human body," which the 2011 rebuttal says was the basis of the analogy.

Probyn-Rapsey argued that the typology assumes a settled meaning of "human," that gender identity disorder was already a contested diagnosis, and that the design smuggled a pathologizing frame into a first description. She reports the paper's "distorted" group, people identifying as less than 100% human, as 46% (her reading of p. 220). That is a different cut from the abstract's one quarter, who said yes to both questions. Both numbers are about this sample. Neither is a population rate.

The same research group answered that the point of the 2008 paper was to publish the first systematic data on a misrepresented subculture. By 2015 their clinical note tells health workers to offer unconditional positive regard and to treat stigma and concealment as the practical problem. The arc is part of the evidence: a pathologizing proposal, a published critique, a rebuttal, and a later clinical stance that does not ask clinicians to diagnose the identity.

Which of the 2008 items, if any, predict later distress or impaired functioning, once stigma is measured separately? That study has not been done in the sources opened here.

• D18 Clinical lycanthropy
• D19 Species dysphoria
• D20 Clinical guidance
• H48 The transgender analogy

A fursona is a character a person maintains: name, species, look, sometimes a suit or a tag. The IARP's trait data say the character is psychologically close to the person and, on average, a bit "better" on desirable traits. Markus and Nurius (1986) called images of who one might become "possible selves." Higgins (1987) separated the actual self, the ideal self, and the ought self, and treated gaps between them as a motor of emotion and effort.

Read together, the fursona is a possible self that a person can inhabit in art, chat, and a convention floor. The theory predicts play, practice, and sometimes self-change. It does not predict that the person believes they are, in species terms, a wolf.

Furscience 3.12 and 3.11 (self-fursona similarity: in one study, 35% called the fursona's personality virtually identical to their own; psychological and behavioral similarity outran physical similarity). Markus and Nurius (1986), American Psychologist, 41(9), 954-969. Higgins (1987), Psychological Review, 94(3), 319-340. The last two are standard references. This build did not re-read the original pages.

E2 for the similarity and ideal-self ratings. E3 for the possible-self and self-discrepancy reading. Those theories were not built on furries, and the IARP pages do not report a direct test of Higgins's discrepancy scales.

Some furries say the fursona is a game, a brand, or a sexual character, and the ideal-self average hides them. A persona can also be a mask that increases the gap rather than closing it. Physical dissimilarity is the usual case: the body in the mirror stays human, which is a point in favor of "persona" and against reading every fursona as a species identity.

Do fursona traits move toward the person's everyday behavior over years, as a possible-self model would hope, or do they stay a costume? Longitudinal trait data would decide this. They are not in hand here.

• A1 IARP
• C14 Proteus effect
• B9 Adolescent identity
• Integration, persona

"Furry" names a fandom: interest in anthropomorphic animals, shared art, meets, and argot. Social identity research treats a fandom as a group a person can identify with, gain esteem from, and hide when the group is stigmatized. Reysen's line of work distinguishes fanship (love of the thing) from fandom (tie to the other fans). The IARP sits in that line and uses identification as a variable.

Identification predicts time, friends, disclosure, and sometimes wellbeing. It is a maintenance engine for the persona. A person can hold the group identity tightly and still have an ordinary human self-concept.

Reysen and Branscombe's fanship/fandom distinction (2010, Journal of Sport Behavior). This build did not re-open that article. Title and volume are a standard citation, not a page check. Roberts et al. (2015) on stigma and concealment. Furscience section 10.1 on disclosure. Roberts et al. (2024) cite Reysen and Shaw (2016): among a set of about 40 fan groups, furries were among the most negatively viewed. The 2016 article itself was not opened for this build.

E2 for concealment and stigma findings inside furry samples. E3 when a general fandom scale is asked to explain why a persona is an animal rather than, say, a knight or a starship captain.

Group identity explains staying. It explains arrival only if something else made the animal content attractive first. High identification is also compatible with distress: a valued group can be a costly group. Negative public ratings (Reysen and Shaw, as cited second-hand) show the stigma is real as a perception among outsiders. They do not show that the fandom is harmful.

A comparison that holds media genre constant (anthropomorphic animals versus other niche fandoms of similar stigma) would show how much of the psychology is "fandom" and how much is "animal."

• A7 Stigma
• B11 Belonging
• B12 Minority stress
• G42 The litter-box hoax

The IARP's samples are young-adult samples. Winter 2011: mean age about 23 after the team removed minor respondents (about 19 if those respondents were left in). They were not permitted to analyze minors. Summer 2020, an international online survey: mean age 28.30 (SD 8.65). In that wave, 59.2% said their first contact with the fandom was under 18. That is a memory from adults, not a child study.

Sex and gender depend on the question. One artists-page table of furries: sex male 72.4%, female 27.4%; gender man 67.1%, woman 23.3%, genderqueer or non-binary 10.0%. Summer 2020 used "check any that apply": male 73.2%, female 10.1%, trans 12.5%, non-binary 12.5%, and smaller shares for other labels, so the boxes overlap. An older public blurb ("What's a furry?") said about 84% male, 13% female, 2.5% transgender. The honest summary is: majority male in these samples, a substantial minority of women on the sex item that forced a binary, and transgender and non-binary identification well above ordinary population estimates the project cites. There is no single percentage to memorize.

Sexual orientation also moves with wording. Summer 2020 (shares sum to 100, so one choice): gay or lesbian 28.8%, straight 10.1%, bisexual 23.4%, pansexual 16.5%, asexual 10.5%, unsure 5.8%, something else 4.9%. Other IARP pages say exclusively heterosexual was still the single most common category in one comparison, about one quarter in the 2011 online survey, under a quarter in the 2014 three-fandom study, and "only about 20%" when open labels were used. Furries in these samples are much less exclusively heterosexual than the general-population comparisons the project uses. They are not a homosexual fandom. The modal label is not stable across instruments.

Ethnicity: majority white in North American and English-language samples. Summer 2020, which included China at 11.1% of respondents, reported white 78.6% and East Asian 15.3%. Country list in that wave also included Canada 16.5%, the UK 9.1%, Germany 8.2%, Finland 4.5%. This is still a volunteer list, not a world census.

A stable religious-affiliation figure was not on the pages opened for this build. It is omitted on purpose.

Furscience: Winter 2011 online survey; Summer 2020 survey; section 1.3 and the artists demographic table in 6.2; section 5.1 on orientation (including the staging copy); the 2014 three-fandom study; "What's a furry?" All are project pages, checked November 2026. Where two pages conflict, both are reported above.

E2, and only as a description of who answers these surveys. Do not promote any figure to "furries are X percent of the population" or even "furries are X percent female" without the instrument and the year.

English-language and convention recruitment under-counts scenes that do not use those channels. Gender boxes that overlap cannot be compared to a single census category. Orientation items that force "heterosexual / homosexual / bisexual" erase labels the later surveys found in double digits. Self-selection into a research project run by known furry researchers may raise candor about sexuality. See the measurement note. Grey (2019/2020) made a related point about a different survey's recruitment text.

A probability sample, in more than one language, with one pre-registered gender item and one pre-registered orientation item. Until then the demographic sentence stays qualified.

• A6 Neurodivergence
• G43 Global scenes
• I50 Sexuality
• G44 Minors

At Anthrocon 2018, just over 1,200 convention participants were asked whether a professional had diagnosed them on the autism spectrum. 11.7% said yes. The project compared that with general-population rates and then listed its own cautions: the fandom is disproportionately male, and autism diagnoses are more often given to males; intense special interests are part of the autism phenotype, so fandoms as a class may gather autistic people. The same page says earlier IARP work did not find a general excess of mental illness, and sometimes pointed the other way.

Anthrocon 2013: 16.1% reported current or former depression or bipolar disorder, set by the authors next to a 17% general-population figure they were using; no relationship, in that summary, between being furry and being prescribed psychiatric medication; a small association between strength of furry identification and an Asperger's label (B = .083, p = .023). "Asperger's" is the term in that summary. Diagnostic manuals have since folded it into autism spectrum disorder. About 8% of that sample were therians, by the project's count.

A separate, non-IARP study of therians (Clegg, Collings, and Roxburgh, 2019) found the odds of a high Autism Spectrum Quotient score were nearly six times those of the non-therian comparison group (OR 5.93, 95% CI 2.04-17.31), in 112 therians and 265 non-therians. Therians scored lower on wellbeing items tied to positive social relationships. Therian identity also moderated links between autism, introverted anhedonia, and autonomy, which the authors read as a possible protective pattern. That is one convenience comparison, not a population odds ratio.

Furscience, Anthrocon 2018 study page; Anthrocon 2013 page. Clegg, Collings, and Roxburgh (2019), Society & Animals, 27(4), 403-426. Roberts et al. (2024) note convention resources the team built for autistic attendees, which is practice, not a new prevalence.

E2 for the self-report rates, with the sample attached. E3 for every causal story: social-space (the fandom is a tolerable place), special-interest (the art is the interest), sampling artifact (who comes to Anthrocon and who will say yes to a researcher), and diagnostic bias (who gets assessed). ADHD and anxiety are discussed in the project's mental-health writing. This build did not extract a single ADHD percentage it is willing to print.

Self-report of a past diagnosis is not a reassessment. A convention that is loud, scheduled, and full of in-jokes selects a particular autistic subgroup and scares others off. Elevated autism traits among therians may be "people who will join an online identity study," which is a known bias in internet research. Lower social-relationship scores can be the effect of stigma rather than of the identity. The protective-autonomy finding is statistical moderation in one sample. It is not a prescription.

Clinician-confirmed diagnoses in a furry sample and in a matched non-furry fandom sample, plus a general-population item on therian identity. Design in the gaps.

• B10 Animal affinity
• D20 Clinical intersection
• C16 Absorption
• B12 Minority stress

Furscience 10.1: furries disclose less than comparison fan groups. In the studies summarized there, about 40-45% said no one in family, work, school, or day-to-day life knew. Only about 35% were out to some or most of their family. Friends, often themselves furries, were told more often: about 55% said many or most friends knew. An older public blurb gave a harsher pair (about 65% had told almost no family; about 70% almost no one in daily life; about 60% agreed that society was prejudiced against furries). The two write-ups do not match exactly. Both say concealment is common.

Goffman (1963) described stigma as a spoiled public identity, managed by passing and by covering. The furry pattern fits the description: a valued identity, a hostile audience, selective disclosure. Roberts et al. (2015) add that media exaggeration of rare behavior makes disclosure to clinicians feel dangerous.

Inside the fandom, therians and otherkin are a minority and are sometimes the target of the same "not really us" boundary that outsiders draw around furries. Roberts et al. (2015) define therians as a small subset. Anthrocon 2013 put them near 8% of that sample. The older "What's a furry?" blurb said about 20%. Intra-fandom hierarchy is widely described by participants. A single peer-reviewed prevalence of anti-otherkin sentiment inside the fandom was not extracted for this build, so the hierarchy is reported as a community pattern (E5) plus the subset sizes above (E2).

Furscience 10.1; "What's a furry?"; Anthrocon 2013. Roberts et al. (2015). Goffman (1963), Stigma. Reysen and Shaw (2016), as cited by Roberts et al. (2024), on outsider dislike.

E2 for concealment rates in named samples. E4 for Goffman as the vocabulary. E5 for the texture of intra-fandom boundary policing, until a counted study is cited.

Concealment can be privacy, not fear. Some furries are fully open and pay little visible cost. Stigma research that recruits through the fandom over-represents people for whom the identity matters. Outsider dislike of a label is not evidence about the mental health of the people who hold it.

A measure of within-fandom stigma toward therians and otherkin, predicting their disclosure and wellbeing, separate from stigma that comes from outside.

• B12 Minority stress
• G42 Litter-box hoax
• D20 Clinical guidance
• F36 Modern religious reactions