18/09/2026
My take on this: Iโm not sure that what is written in books and newspapers actually represents the preferences of the people with lived experience. It simply reflects the way that OTHER PEOPLE talk (write) ABOUT US.
How should we refer to people with mental ill health, disability, or neurodivergence?
Terminology should be respectful, apt and avoid carrying stigma. Itโs also important to listen to peopleโs preferences about how they refer to their condition. But how much consensus is there and how has terminology changed?
Our new study โ the largest of its kind โ explores how this language has evolved over the past 45 years to reflect our shifting views of mental illness and disability.
We found preferences can depend on the condition, and whether we are talking about individuals or groups.
๐๐๐ฆ๐ข๐ง๐ ๐ฆ๐, ๐ฐ๐ก๐๐ญโ๐ฌ ๐ฉ๐๐ซ๐ฌ๐จ๐ง-๐๐ข๐ซ๐ฌ๐ญ ๐ฅ๐๐ง๐ ๐ฎ๐๐ ๐?
In the 1980s, disability activists argued common expressions such as โdisabled personโ equated the person with their disability.
Instead they proposed โperson-firstโ language. Referring to someone as a โperson with disabilityโ puts their personhood in the foreground and doesnโt define them by their impairment.
Professional organisation soon adopted person-first language as preferred terminology. Guidelines for academic publishing and media reporting followed.
๐๐ก๐ ๐ฆ๐จ๐ฏ๐ ๐ญ๐จ ๐ข๐๐๐ง๐ญ๐ข๐ญ๐ฒ-๐๐ข๐ซ๐ฌ๐ญ ๐ฅ๐๐ง๐ ๐ฎ๐๐ ๐
More recently there has been a backlash. Critics argue person-first language presents disability or mental illness as a devalued and peripheral aspect of the person.
US psychologist Louis Sass expresses this beautifully:
"๐๐ด ๐ด๐ค๐ฉ๐ช๐ป๐ฐ๐ฑ๐ฉ๐ณ๐ฆ๐ฏ๐ช๐ข ๐ฃ๐ฆ๐ด๐ต ๐ค๐ฐ๐ฏ๐ค๐ฆ๐ฑ๐ต๐ถ๐ข๐ญ๐ช๐ป๐ฆ๐ฅ ๐ข๐ด ๐ข ๐ฅ๐ช๐ด๐ฆ๐ข๐ด๐ฆ ๐ฆ๐ฏ๐ต๐ช๐ต๐บ ๐ต๐ฉ๐ข๐ต ๐ค๐ฐ๐ฎ๐ฆ๐ด, ๐ข๐ด ๐ช๐ต ๐ธ๐ฆ๐ณ๐ฆ, ๐ง๐ณ๐ฐ๐ฎ ๐ฐ๐ถ๐ต๐ด๐ช๐ฅ๐ฆ, ๐ข๐ฏ๐ฅ ๐ธ๐ช๐ต๐ฉ ๐ธ๐ฉ๐ช๐ค๐ฉ ๐ต๐ฉ๐ฆ ๐ด๐ถ๐ง๐ง๐ฆ๐ณ๐ฆ๐ณ ๐ฎ๐ถ๐ด๐ต ๐ค๐ฐ๐ฏ๐ต๐ฆ๐ฏ๐ฅ, ๐ข๐ด ๐ธ๐ช๐ต๐ฉ ๐ข ๐ฃ๐ถ๐ณ๐ฅ๐ฆ๐ฏ ๐ฐ๐ณ ๐ข๐ฏ ๐ฆ๐ฏ๐ฆ๐ฎ๐บ? ๐๐ณ ๐ช๐ด ๐ช๐ต ๐ฃ๐ฆ๐ต๐ต๐ฆ๐ณ ๐ด๐ฆ๐ฆ๐ฏ ๐ข๐ด ๐ข ๐ฎ๐ฐ๐ณ๐ฆ ๐ช๐ฏ๐ต๐ช๐ฎ๐ข๐ต๐ฆ ๐ง๐ข๐ค๐ต๐ฐ๐ณ: ๐ด๐ฐ๐ฎ๐ฆ๐ต๐ฉ๐ช๐ฏ๐จ ๐ต๐ฉ๐ข๐ต ๐ฆ๐ฎ๐ฆ๐ณ๐จ๐ฆ๐ด ๐ง๐ณ๐ฐ๐ฎ ๐ธ๐ช๐ต๐ฉ๐ช๐ฏ, ๐ฑ๐ฆ๐ณ๐ฎ๐ฆ๐ข๐ต๐ฆ๐ด ๐ต๐ฉ๐ฆ ๐ท๐ฆ๐ณ๐บ ๐ค๐ฐ๐ณ๐ฆ ๐ฐ๐ง ๐ฐ๐ฏ๐ฆโ๐ด ๐ฃ๐ฆ๐ช๐ฏ๐จ, ๐ข๐ฏ๐ฅ ๐ฎ๐ถ๐ด๐ต, ๐ฑ๐ฆ๐ณ๐ฉ๐ข๐ฑ๐ด, ๐ฃ๐ฆ ๐ถ๐ฏ๐ฅ๐ฆ๐ณ๐ด๐ต๐ฐ๐ฐ๐ฅ ๐ข๐ด ๐ข๐ฏ ๐ข๐ด๐ฑ๐ฆ๐ค๐ต ๐ฐ๐ง ๐ต๐ฉ๐ฆ ๐ด๐ถ๐ง๐ง๐ฆ๐ณ๐ฆ๐ณโ๐ด ๐ท๐ฆ๐ณ๐บ ๐ด๐ฐ๐ถ๐ญ?"
Proponents of โidentity-firstโ language say expressions such as โautistic personโ recognise that some conditions are not external visitations but integral aspects of the individual.
The American Psychological Association endorsed identity-first language in 2020, and surveys show many neurodivergent people prefer it.
But more broadly, is person-first or identity-first language preferred?
๐๐ ๐ฌ๐ญ๐ฎ๐๐ข๐๐ ๐ก๐จ๐ฐ ๐ฉ๐ซ๐๐๐๐ซ๐๐ง๐๐๐ฌ ๐๐ก๐๐ง๐ ๐๐
In our new study, we assessed the relative popularity of person-first and identity-first expressions for 15 mental health, disability and other conditions over 45 years.
We examined two vast English-language databases: the Google Books database (1980โ2019) and the 23-billion-word global News on the Web collection of newspapers and magazines (2010โ2025).
For each condition and year, we counted the frequency of four expressions: person-first or identity-first terms for individuals or groups of people.
For autism, for example, we counted โperson with autismโ, โautistic personโ, โpeople with autismโ and โautistic peopleโ.
We then computed the relative popularity of person-first versus identity-first language when referring to individuals or groups.
๐๐ก๐๐ญ ๐๐ข๐ ๐ฐ๐ ๐๐ข๐ง๐?
Despite increasing advocacy, there is currently no widespread preference for identity-first language.
At most, advocacy has slowed the rise in person-first language promoted by an earlier generation of advocates.
๐๐ก๐๐ญ๐ก๐๐ซ ๐ญ๐๐ซ๐ฆ๐ฌ ๐ซ๐๐๐๐ซ ๐ญ๐จ ๐ข๐ง๐๐ข๐ฏ๐ข๐๐ฎ๐๐ฅ๐ฌ ๐จ๐ซ ๐ ๐ซ๐จ๐ฎ๐ฉ๐ฌ ๐๐ฅ๐ฌ๐จ ๐ฆ๐๐ญ๐ญ๐๐ซ๐ฌ
When describing individuals, identity-first expressions have generally been more common. But person-first expressions are now equally popular.
When describing groups, person-first expressions now more common.
Mental health problems and disabilities are typically understood as aspects of personal identity rather than group identity. So itโs not surprising identity-first language is preferred when referring to individuals.
It makes sense to use โpsychotic personโ rather than โperson with psychosisโ but also use โpeople with psychosisโ rather than โpsychotic peopleโ if we see psychosis as a personal experience, not a basis for shared identity.
๐๐ข๐๐๐๐ซ๐๐ง๐๐๐ฌ ๐๐ฆ๐จ๐ง๐ ๐๐จ๐ง๐๐ข๐ญ๐ข๐จ๐ง๐ฌ
These findings, based on averages, conceal large variations across our 15 conditions.
Identity-first language was consistently preferred for some conditions, such as alcoholism, anxiety and obesity.
Person-first language was consistently preferred for others, such as anorexia, bipolar, diabetes and epilepsy.
The conditions also show diverse patterns of change. In recent years, identity-first terminology has risen for autism and dyslexia, but person-first terminology has risen for bipolar, anorexia and obesity.
These findings indicate that terminological trends are quite specific to the condition.
๐๐ก๐ฒ ๐ฆ๐ข๐ ๐ก๐ญ ๐ข๐ญ ๐๐๐ฉ๐๐ง๐ ๐จ๐ง ๐ญ๐ก๐ ๐๐จ๐ง๐๐ข๐ญ๐ข๐จ๐ง?
Some conditions have become the focus of identity claims more than others.
The neurodiversity movement, for example, has promoted the view that neurodevelopmental conditions are inherent and valued aspects of the person. The rise in identity-first language for autism and dyslexia supports this possibility.
More broadly, whether we favour identity-first or person-first language for different conditions may depend on our beliefs about them. Do they represent identities or attributes: who we are or what we have?
Conditions that people see as biologically caused and enduring are more likely to be viewed as part of someoneโs โtrue selfโ. We might expect identity-first language will be preferred for these conditions.
The language we use in this domain is not a collection of inert labels. Itโs contentious and ever-evolving.
โ๏ธ by Jemima Kang, PhD Candidate, School of Computing and Information Systems, The University of Melbourne
Alsa Wu, PhD Candidate, School of Psychological Sciences, The University of Melbourne
Naomi Baes, PhD Candidate and Computational Social Science Researcher, The University of Melbourne
and Nick Haslam, Professor of Psychology, The University of Melbourne
๐ Read more from The Conversation AU: https://theconversation.com/au
๐ฌ Comment "newsletter" and we'll send you the link to sign up for our free wrap of the latest insight and analysis.
