Abstract
Finding a consensual term for persons diagnosed with autism spectrum disorder has recently been debated in the scientific literature. Considering the stigma associated with using terms deemed offensive, it is paramount to address autism respectfully and consensually. As of now, this study is the first to consult French-Canadian participants beyond autistic people themselves. This mixed-method study aimed to document and understand the preference and offensiveness of terms used to refer to persons living with autism. Participants (N = 327) were adults who self-identified as part of the autism community (i.e., autistic person, family or friends, professionals, or clinicians). By means of an online survey, they rated and ranked six terms used to designate an autistic person. Participants also explained their ranking. Results show no consensus for the use of any one specific term. A clear dichotomy appears between autistic adults’ and professionals’ preferences in terminology. The latter prioritized terms related to the medical model, whereas autistic persons preferred using identity-related language. Among all respondents, Autistic person was the most preferred and least offensive term. Thus, we suggest asking for the concerned person's preference whenever possible or using terminology preferred by the majority when this cannot be done.
Lay abstract
Recent studies and editorials by autism researchers suggest distancing ourselves from the medical terms used to name autism. Considering the stigma associated with using terms deemed offensive and the influence culture has on language, it is paramount to address autism respectfully and consensually in a culturally sensitive way. As of now, data on the French-Canadian population has yet to be collected. Participants (N = 327) who self-identified as part of the autism community (i.e., autistic person, family or friends, professionals, or clinicians) completed an online survey. They rated and ranked six terms used to designate an autistic person. Participants also explained their ranking. Among all respondents, Autistic person was the most preferred and least offensive term by most respondents. An apparent dichotomy appears between autistic adults’ and professionals’ preferences in terminology. The latter prioritized terms related to the medical model, whereas autistic persons preferred using identity-related language. However, considering the French syntax, the latter justified their preference based on identity-first and person-first principles. Because results show no consensus for one term, we suggest asking for the concerned person's preference whenever possible or using terminology preferred by the majority when this can’t be done.
Introduction
Vocabulary and social representations shape each other. Naming and addressing the population concerned respectfully and consensually reduces the stigma perpetuated or maintained by using terms deemed offensive. Using “person-first” before a given diagnosis is well-intentioned. However, some began arguing that, to the contrary, person-first language emphasizes the disability, thus stigmatizing those with a diagnosis (Botha et al., 2023; Gernsbacher, 2017; Sinclair, 2013).
Stigmatization, defined as “the act of treating someone or something unfairly by publicly disapproving of them” (Cambridge Dictionary, 2024), can cause psychological distress (Frost, 2011), thus leading to increased chronic stress and poor mental health (Botha & Frost, 2020). Sometimes flagrant, other times insidious, social representations of autism in the media contribute to this stigmatization. While it is encouraged to look towards scientific papers for valid and reliable knowledge on clinical characterizations of autism, it is sometimes the case that the language used in research may be considered dehumanizing and stigmatizing. (see Botha et al., 2022 for a detailed enumeration of terms used in past and recent papers). Now, awareness is raised in response to these terms and damaging social representations. Members of the autism community, including autistic researchers, are taking a stance away from terms related to the medical model and person-first language (Dunn & Andrews, 2015; Pellicano & Heyworth, 2023; Sinclair, 2013). The re-appropriation of language and reframing of societal conceptions are steps toward rehabilitating autism terminology and managing stigma (Botha & Gillespie-Lynch, 2022).
Several studies have empowered the autistic community by depicting the terms that should be favored to describe autism or to refer to the individuals in a respectful and destigmatizing manner. Most of these studies surveyed Anglo-Saxon populations. Different stakeholders of the British and American autism communities expressed their adherence to person-first (PFL) or identity-first language (IFL) for describing autism (Kenny et al., 2016; Taboas et al., 2023). The results of the language preference task showed that autistic adults preferred identity-first terms (e.g., autistic), and professionals preferred person-first terms (e.g., person with ASD). Bury et al. (2023) surveyed a sample of Australian autistic adults. In this study, terms that did not refer to a disorder (e.g., person on the autism spectrum, autistic, autistic person) were most preferred. A thematic analysis of comments left by the respondents offers some important considerations in choosing or avoiding a term, including (a) expressing one's identity, (b) respecting diversity within the spectrum, and (c) desiring to communicate pragmatically (Bury et al., 2023). Most importantly, participants expressed their desire to avoid vocabulary related to the medical model, no matter their preferred term. Rather, the participants emphasized the person's “shared humanity” with those without autism.
Only a handful of studies outside the Anglo-Saxon populations have been published, all conducted within Europe (Buijsman et al., 2023; Garcia-Molina, 2019; Geelhand et al., 2023). To our knowledge, the first and only study to have surveyed French-speaking autistic adults was completed primarily by participants from France and Belgium (3,3% of their sample were from Canada). The authors concluded that identity-first terms were preferred by their French-speaking sample (Geelhand et al., 2023). It is imperative to note that in French, Autistic person, considered in English to place identity first, translates into “personne autiste”, the adjective (autiste) being in second place and person first. Therefore, Geelhand et al. (2023) coded “personne autiste” as an identity-first term, not a person-first term. Another European survey was conducted in Spanish (Garcia-Molina, 2019). It could have similarities with French since their grammatical rule puts “persona” in front of “autista” when referring to an Autistic person, just like the French “personne autiste”. Unfortunately, Garcia-Molina's multiple-choice question merged the PFL terms person with autism and the IFL into the same response choice, “persona con autismo/persona autista”, hence interpreting both as PFL. Another limitation to comparing Garcia-Molina's (2019) results with the studies cited above is that this sample included teachers, doctors, psychologists, students, and unemployed individuals from the general population without identifying participants who might have a diagnosis of autism. Additionally, studies on term preferences have yet to take into account how long ago the person received the diagnosis of ASD. Recent evidence suggests that individuals who are diagnosed with autism as adults are likely to have developed a more positive autistic identity (Fletcher-Watson, 2024; Overton et al., 2023). Kiehl et al.'s (2024) meta-synthesis revealed that the experience of being diagnosed in adulthood led to an increased self-understanding, a sense of relief, and improved self-esteem. This may contribute to a preference for IFL. In contrast, individuals who received the diagnosis during childhood may have internalized the label of ASD given to them many years ago and continue to use PFL. Thus, experiencing autism first-hand and age at the time of diagnosis may be associated with term preference (Kiehl et al., 2024; Kenny et al., 2016).
The present study adds to the body of research on terminology preference by surveying French-speaking Canadians, who use French somewhat differently in syntax and vocabulary than French-speaking Europeans. In addition, French-speaking Canadians have a distinctiveness in their culture, reflected in their language. Since the culture and semantics of a specific language can bring specificities to one's choice and use of preferred terms (Bottema-Beutel et al., 2021; Keating et al., 2023), it is essential to survey the French-Canadian population to document the terms used to refer respectfully to persons from the autistic community.
Hypotheses
Based on current literature, we expect that 1) no unique term will be unanimously preferred in the broader autism community or within each group of stakeholders.
2) Most autistic adults will prefer identity-first terms more than parents, clinicians, and professionals. The latter will prefer a person-first term more than parents and autistic adults. 3) We expect that autistic respondents diagnosed in adulthood will prefer significantly more identity-affirming terms and be more offended by traditional medical terms compared to those diagnosed before the age of 18. Finally, rather than trying to confirm the themes provided by previous researchers (Bury et al., 2023; Geelhand et al., 2023; Kenny et al., 2016), no hypotheses are formulated for the qualitative explanations of personal vocabulary choices.
Methodology
Recruitment and procedure
Prior to recruitment, the Université du Québec en Outaouais Research Ethics Committee approved the study (#2023-2333). Over 167 organizations throughout Canada were contacted between December 2022 and March 2023 by email or telephone. These varied from local, regional, and provincial French-speaking associations concerned with autism to specialized schools and camps, autism research teams, Facebook groups, respite homes, specialized clinics, and non-profit foundations. Fifty-one shared the recruitment poster through their social networks and newsletter. They also displayed our recruitment poster in their organization and shared the information with employees. To diversify the process, we posted an announcement in a rural newspaper in the Canadian province of Quebec and displayed posters in public places in the largest city in this French-speaking province, Montreal.
Participants were asked to complete an online survey that included demographic questions, different tasks assessing participants' preferences and views of offensiveness of six terms, and open-ended questions to explain their answers for the different tasks. The LimeSurvey questionnaire was online from November 16th, 2022, to March 9th, 2023.
Participants
To be eligible for this study, participants had to self-reportedly confirm that they met three inclusion criteria: 1) be part of the autism community (i.e., autistic person, family or friends, professional, or clinician concerned by autism), 2) be 18 years or older, and 3) be a French-speaking Canadian. Nearly 500 individuals consulted the online questionnaire, whereas 327 people consented to participate and completed the term ranking task and level of preference and offensiveness tasks.
Table 1 presents demographic information for the entire sample and by group of participants. The participants ranged in age from 18 to 85 and older (the majority between 25 and 54). Participants were mostly autistic adults (42.8%). Half of the autistic adults identified as women, 35.7% as men, and the remaining as trans men, trans woman, non-binary, or “other’’, reflecting the greater gender diversity typically observed in the autistic community (Kallitsounaki & Williams, 2023). The high number of women professionals and mothers influenced the female ratio in the total sample. The “other” category (n = 26, 8.0%) was composed of 7 (2.1%) participants identified as friends or family members other than parents, 16 (4,9%) participants who had dual-membership (i.e., parents who are also professionals) and 3 (0,9%) participants who specified that they were researchers or service managers.
Demographic information for the entire sample and by group of participants.
The percentage is based on the total sample of each subgroup.
Autistic participants were also asked to disclose their diagnosis. Nearly half the autistic adults identified as having an autism spectrum disorder (ASD; n = 50; 35.7%) or self-identified as meeting autism diagnostic criteria, including Asperger syndrome, Pervasive Development Disorder – Not Otherwise Specified (PDD-NOS), and Autistic disorder (n = 45; 32.1%). The remaining participants in this group reported their diagnosis as Asperger's Syndrome (n = 36; 25.7%), PDD-NOS (n = 6; 4.8%), or Autism (n = 2; 1.5%). Nearly two-thirds (65%) received their diagnosis in adulthood.
Measures
Terminology preference and perceived offensiveness
We translated the survey employed by Bury et al. (2023), published online in 2020, to document the level of preference and offensiveness of six current terms employed to refer to autistic persons (see Table 2). As per these authors, we did not include terms related to Asperger's in coherence with recent nomenclature and to avoid limiting a term pertaining to a specific group of individuals (Kenny et al., 2016). We translated the survey to French using a forward-backward procedure (Brislin, 1970). Specifically, SMF translated the questionnaire into French. Then, V.D. (see acknowledgment) and C.L.N. each translated the items back to English without knowledge of the original version. Any discordance with the original survey was discussed until an agreement was reached.
Terms included in the questionnaire and their English translation.
Following consent and demographic information, participants were asked to complete four tasks (two assessing preference and two assessing offensiveness). First (task 1), we asked the participants to rate each term individually regarding their level of preference on a five-point Likert scale (1 = I dislike it “n’aime pas du tout”; 2 = Indifferent “indifferent. e”; 3 = I somewhat like it “aime peu”; 4 = I mildly like it “aime modérément”; 5 = I strongly like it “aime fortement”). Answering “do not know/ prefer not to answer” was also possible. Participants then could specify any other term they preferred in a text box. This task was repeated (task 2) for the level of offensiveness (1 = Extremely offensive “extrêmement offensant”; 2 = Offensive “offensant”; 3 = Slightly offensive “peu offensant”; 4 = Indifferent “indifferent.e.”; 5 = Not at all offensive “pas du tout offensant”).
Next, participants were asked to rank these six terms from most preferred to least preferred (task 3) and from most offensive to least offensive (task 4). Contrary to tasks 1 and 2, tasks 3 and 4 forced the participants to choose a unique term by discriminating it from the others. They were then asked to specify why they placed the first term as their preferred and most offensive in two text boxes. The survey ended with a general open-ended item where participants could share any comments concerning the questionnaire or the subject of terminology in autism.
Data analysis and data cleaning
Quantitative data was imported into IBM SPSS Statistics software (version 29.0). Because of the non-normality of distribution and non-equal intervals in the Likert scales, we proceeded with nonparametric Mann-Whitney U tests to identify differences between groups (Sullivan & Artino, 2013). Medians characterized each group's terminology preference and offensiveness.
A small proportion of respondents chose “do not know/prefer not to answer” at any multiple-choice for tasks 1 and 2 (from .9 to 3.1%). Higher percentages were not specific to a precise term but to a task. Ratings of the level of offensiveness (task 2) showed higher percentages of “do not know/prefer not to answer” for the six terms addressed (1.8% to 3.1%). This answer was coded as a missing value for analytical purposes. Missing values for each task varied from .3% to 9.5%, with term ranking regarding offensiveness (task 4) having the highest missing values. Missing values were not imputed and thus left out of the analytical procedure.
Regarding the qualitative analysis, the open-ended items whereby participants explained their ranking (tasks 3 and 4) were coded using thematic analysis (Braun & Clarke, 2019). Four authors (authors’ initials – including one research partner with lived experience) contributed equally to the thematic analysis using NVivo 1.7.1. We proceeded in pairs to analyze answers left by parents, professionals, and autistic adults separately to name and define each theme. Two authors coded all comments from the sub-sample of autistic adults, and two others coded all comments from the sub-samples of parents and professionals. They discussed the themes extracted in pairs, refining their definitions and labels and identifying which comments fit best as exemplary of each theme. In the next step, the four authors discussed similarities and differences between the themes found among autistic adults and stakeholders (non-autistic in the majority). This led to further refining of labels and definitions until a consensus was reached regarding each theme. An additional effort was made to group these five themes into broader categories. Again, these were discussed until a consensus was reached that these categories adequately reflected a common thread between themes.
Community involvement
This study is led according to the principles of participatory research. Four autistic adults and four researchers identified the research interest, questions, and objectives. Each member was involved throughout the different phases of the study, including the publication of this article.
Results
The quantitative results are presented first, followed by the qualitative results of the thematic analysis. The quantitative results are organized by objective and task, with results about preference (tasks 1 and 3) presented first and the results on offensiveness (tasks 2 and 4) presented subsequently.
Missing values
A fair proportion of the total sample did not specify their identity. Therefore, responses from 8.6% of the sample could not be analyzed when differences between categories of respondents were addressed. We found differences in preferences between these respondents and those of parents (significant Man-Whitney-U test for terms: Autistic, Autistic person, and Person with ASD), as with preferences and offenses of professionals (preference of terms: Person with autism, Autistic, Autistic person and Person with ASD; offensiveness of terms: Autistic, Autistic person and Person with ASD). This was also the case for term ranking. Undetermined respondents differed from autistic adults in the offensiveness ranking of Autistic person. Differences with parents were also observed in preference and offensiveness rankings of this term and Person with ASD. In contrast to professionals, undetermined respondents’ preference ranking differed for the terms Person on the autism spectrum, Autistic, Autistic person, and Person with ASD; and offensiveness ranking for Autistic, Autistic person, Person with ASD, and Person with ASC. Because of limitations in interpreting this data, we excluded the subsample of undetermined respondents in further analyses of differences in response across categories of respondents.
Preference ratings for the entire sample
Preference for the terms varied significantly (Zs = 7.27 to 13.99, p < .001). Overall, participants rated Person with ASC (“Personne avec une condition du spectre de l’autisme”) as the least preferred term, with a median score of 2, followed by Person with autism (“Personne avec autisme”) (Mdn = 3) and Person on the autism spectrum (“Personne sur le spectre de l’autisme”) (Mdn = 3). The preferred terms were Person with ASD (“Personne avec un TSA”) (Mdn = 4.00), Autistic (“Autiste”) (Mdn = 4.00), and Autistic person (“Personne autiste”) (Mdn = 4.00). The term Autistic person was the most frequently “strongly liked” (n = 145, 44.34%) and the lowest number rating of “strongly disliked” (n = 21, 6.42%). Person with ASC was “strongly disliked” by the highest number of participants and “strongly liked” (n = 127, 38.8%) by the lowest number of participants (n = 24, 7.34%).
Differences in terms of preference ratings across categories of respondent
Figure 1 presents the median score on the preference scale for each term according to each category of respondent. A Kruskal-Wallis test was used to verify if participants’ preferences for each term differed across categories of respondents (i.e., autistic adults, parents, professionals, or others). Groups differed on the level of preference for the terms Person with autism (H(3) = 26.99, p < .001), Autistic (H(3) = 28.94, p < .001), Person with ASD (H(3) = 65.40, p < .001), and Person with ASC (H(3) = 17.41, p < .001). Most autistic participants preferred the terms Autistic person and Autistic, while professionals preferred Person with ASD. Parents mostly rated Autistic, Autistic person, and Person with ASD as their preferred terms. Most autistic participants “disliked” the terms Person with autism and Person with ASC. Parents, professionals, and other respondents did not have a median under 2 for any of the terms, suggesting that none of the terms were generally “disliked” by these participants. No significant differences were observed for the terms Person on the autism spectrum or Autistic person.
Median preference for each term by respondent category.
Thus, pairwise analyses using the Mann-Whitney tests were conducted for the four terms that differed significantly to better understand group-level differences. Figure 1 summarizes the results of the pairwise analyses for the six suggested terms and specifies the median preference of each term.
Although significant differences were observed between parents, professionals, and other participants, below we emphasize the difference in contrast with autistic participants as these showed the largest differences (differences in median ≤ 2). Autistic respondents scored the term Autistic significantly higher than the three other categories of participants (parents: U = 2946.5, nautistic adults = 135, Mdnautistic adults = 5, nparents = 58, Mdnparents = 4 p < .01; professionals: U = 2928.5, nprofessionals = 73, Mdnprofessionals = 3, p < .001; others: U = 1172.5, nothers = 26, Mdnothers = 3, p < .01). In contrast, autistic adults rated significantly lower than the other participants the terms Person with autism (parents: U = 3104, nautistic adults = 136, Mdnautistic adults = 1, nparents = 58, Mdnparents = 3, p < .05; professionals: U = 3005.5, nprofessionals = 74, Mdnprofessionals = 3, p < .001; others: U = 1279, nothers = 26, Mdnothers = 3, p < .05), Person with ASD (parents: U = 2502.5, nautistic adults = 138, Mdnautistic adults = 2, nparents = 58, Mdnparents = 4, p < .001; professionals: U = 2035.5, nprofessionals = 74, Mdnprofessionals = 5, p < .001; others: U = 930.5, nothers = 25, Mdnothers = 4, p < .001) and Person with ASC (parents: U = 3040, nautistic adults = 135, Mdnautistic adults = 1, nparents = 57, Mdnparents = 3, p < .05; professionals: U = 3441, nprofessionals = 74, Mdnprofessionals = 3, p < .001).
We proceed with the Mann-Whitney U test to compare the preferences of adult-diagnosed participants to those who received their diagnosis in childhood. Results suggest that preferences for certain terms differed if participants were diagnosed in adulthood. Mainly, they preferred the term Autistic compared to those diagnosed in childhood or adolescence (U = 2049, nadulthood = 91, Mdnadulthood = 5, nchildhood = 36, Mdnchildhood = 4, p < .05) and did not like the terms Person with autism (U = 1294.5, nadulthood = 91, Mdnadulthood = 1, nchildhood = 36, Mdnchildhood = 2, p < .05).
Forced ranking of preference for the entire sample
We asked participants to rank the six terms in order of preference. This exercise forced a choice by comparing each term to one another. The most preferred term was placed in the first rank. Autistic person was the most preferred, with a median rank of second place, and the least preferred was Person with ASC with a median rank of fifth place. Rankings of each term differed as confirmed by Kolmogorov-Smirnov tests (Zs = 7.79 to 14.49, all ps < .001).
Differences in preference ranking by category of respondent
We then proceeded to analyze ranking preferences regarding the categories of respondents. Kruskal-Wallis test showed that groups ranked each term differently except for Person with ASC as the least preferred.
Autistic was ranked most preferred by 42.1% of autistic adults, and Autistic person by 30%. Mann-Whitney test identified which term ranking differed between pairs of respondents. Compared to parents and professionals, autistic adults ranked higher in the terms Person on the autism spectrum (U = 2287, nautistic adults = 133, nparents = 56, p < .01; U = 3466, nautistic adults = 133, nprofessionals = 71, p < .001), Autistic (U = 2865, nautistic adults = 134, nparents = 56, p < .01; U = 1888.5, nprofessionals = 71, p < .001), and Autistic person (U = 2776, nautistic adults = 133, nparents = 56, p < .01; U = 3487.50, nprofessionals = 72, p < .001) and lower the term Person with ASD (U = 2388.5, nautistic adults = 129, nparents = 56, p < .001; U = 1597.50, nprofessionals = 73, p < .001). Term preference ranking also differed between autistic adults and professionals for the terms Person with autism (U = 3811.50, nautistic adults = 132, nprofessionals = 72, p < .05), with autistic adults ranking this term in lower preference.
Offensiveness ratings for the entire sample
Participants rated how offensive to them each term was on a scale of 1 to 5, where 1 was “extremely offensive” and 5 was “not at all offensive”. The level of offensiveness differed significantly across terms (Zs = 11.77 to 14.07, all ps < .001). The frequency scores suggest that the term Person with autism was the most offensive, with 54.43% (n = 178) of the participants scoring it as 3 or less on the scale (i.e., slightly to extremely offensive). This result is somewhat coherent, with participants rating this term as one of the least preferred.
On the other hand, the term Autistic person was considered not offensive by most participants, with 27.21% rating it as ≥ 3 (Mdn = 5). The remaining terms had a median of 4, suggesting that they were “indifferent” to their use, or in other words, they did not find it offensive or inoffensive.
Differences in term of offensiveness ratings by category of respondent
Figure 2 presents the median score on the offensiveness scale for each term by category of respondent. Of interest, all groups of respondents reported a median varying between 3 (slightly offensive) and 5 (not at all offensive) for each term. Thus, none of the terms were “offensive” or “highly offensive” according to participants in each category of respondent.
Median offensiveness for each term according to respondent category.
Kruskal-Wallis tests suggest a significant difference in the level of offensiveness regarding respondent categories for five out of six terms. Specifically, groups differed in the level of offensiveness for the terms Person with autism (H(3) = 17.71, p < .001), Autistic (H(3) = 39.01, p < .001), Autistic person (H(3) = 14.18, p < .01), Person with ASD (H(3) = 51.58, p < .001), and Person with ASC (H(3) = 33.26, p < .001).
Mann-Whitney U tests were performed for each pair of respondents for these five terms. As per the preference rating, we emphasize the difference between autistic participants and the three respondent categories, as these are the largest differences observed. The results suggest that the term Autistic was considered significantly less offensive by autistic adults than by professionals (U = 3 490.00, nautistic adults = 138, Mdnautistic adults = 5, nprofessionals = 72, Mdnprofessionals = 2, p < .001) and others (U = 1 199.00, nothers = 23, Mdnothers = 2, p < .05). In contrast, the term Person with ASD was considered more offensive by autistic adults than by parents (U = 2 718.50, nautistic adults = 138, Mdnautistic adults = 3, nparents = 57, Mdnparents = 5, p < .001), professionals (U = 2 274.00, nprofessionals = 72, Mdnprofessionals =5, p < .001), and other respondents (U = 1 026.00, nothers = 24, Mdnothers = 5, p = .002).
We proceed with the Mann-Whitney U test to compare the perceived offensiveness of adult-diagnosed participants to those diagnosed in childhood. Results suggest that they are less offended by the terms Autistic (U = 2019.5, nadulthood = 90, Mdnadulthood = 5, nchildhood = 35, Mdnchildhood = 4, p < .01), and Autistic person (U = 1998.5, nadulthood = 90, Mdnadulthood = 5, nchildhood = 35, Mdnchildhood = 4, p < .01) and more offended by the term Person with an autism spectrum condition (U = 1142.5, nadulthood = 89, Mdnadulthood = 2, nchildhood = 35, Mdnchildhood = 4, p < .05). Finally, those who received a diagnosis in childhood were less offended by the term Person with an ASD (U = 1998.5, nadulthood = 90, Mdnadulthood = 3, nchildhood = 35, Mdnchildhood = 4, p < .01).
Forced ranking of offensiveness for the entire sample
As with preference, we asked participants to rank the terms according to offensiveness, placing the most offensive term first. Term rankings differed significantly, confirmed by Kolmogorov-Smirnov tests (Zs = 8.24 to 12.3, all ps < .001). Coherent with the preference ranking, the least offensive term was Autistic person with a median rank of fifth place for the entire sample. Based on the median ranking, not one term can be considered most offensive.
Difference in offensiveness ranking by category of respondent
Kruskal-Wallis test of term ranking based on offensiveness showed a difference between groups. Ranking of terms by the four groups of respondents differed for the terms Person with autism (H(3) = 10.34, p < .05), Autistic (H(3) = 45.71, p < .001), Autistic person (H(3) = 20.01, p < .001), Person with ASD (H(3) = 35.25, p < .001), and Person with ASC (H(3) = 15.89, p < .001).
Person with ASD and Person with ASC were ranked the most offensive terms by 27.9% and 26.4% of autistic respondents. Comparing ranking by autistic adults with those of parents, professionals, and others, Mann-Whitney U tests suggested that autistic adults ranked as more offensive the terms Person with autism and Person with ASD compared to parents (U = 2 664.50, nautistic adults = 130, nparents = 53, p < .05; U = 2 539.50, nautistic adults = 128, nparents = 53, p < .01) and professionals (U = 3 425.5, nprofessionals = 68, p < .01; U = 2 180, nprofessionals = 67, p < .001). Autistic adults ranked as less offensive the terms Autistic and Autistic person compared to parents (U = 2 372.50, nautistic adults = 126, nparents = 53, p < .01; U = 2 467.5, nautistic adults = 127, nparents = 53, p < .01), professionals (U = 2 006, nprofessionals = 68, p < .001; U = 2 824.50, nprofessionals = 68, p < .001). Autistic adults and professionals also differed in the ranking on Person with ASC (U = 2 839.50, nautistic adults = 129, nprofessionals = 67, p < .001) as the latter placed this term at a lesser offensiveness rank. Ranking of terms based on offensiveness by autistic adults and other respondents only differed for the terms Autistic (U = 926.50, nautistic adults = 126, nothers = 25, p < .001) and Person with ASD (U = 1 050.50, nautistic adults = 128, nothers = 53, p < .01).
Thematic analysis of term preference and offensiveness
After ranking their most preferred and offensive terms, participants were asked to explain their selection in open-text questions. Three categories of themes emerged from the coded data to justify the rankings. These are: 1) to be respectful, 2) to be understood, 3) to be accurate. Each category and its representative themes will be presented, supported by exemplary citations from respondents, justifying their preference or offensiveness.
To be respectful
The main reason to prefer a certain term over others was justified by a desire to be respectful when referring to or addressing an autistic person. However, this intent was not restricted to any “better” term over others. It applied to many different terms, depending on individual choices.
Places the person first
Five of the six response choices included “person first”, since the English IFL Autistic person translates into French as “Personne autiste”. Thus, Autistic person was preferred because “a person is not only autistic, it is a person with blue eyes, a person with temperament, an intelligent person, a person with passions” (Parent, 1st rank preference: Autistic person). When Autistic was used as a noun (i.e., without “person” preceding it), it could be rated as most offensive by some because it was said to reduce the person to their condition. “The identity of an autistic person includes many facets, not only autism.” (Autistic adult, 1st rank offensiveness: Autistic). Autistic alone “implies that the person is just that, autistic, not a whole person” (Parent, 2nd rank offensiveness: Autistic). Some professionals also justified their choice of term based on respect, even though their preference was not in line with those of the majority of autistic respondents: “Because I find that it puts the person at the forefront of their disorder, they have a disorder, they are not a disorder [a trouble]” (Professional, 1st rank preference: Person with ASD).
Cannot be used as an insult
In everyday language, Autistic is sometimes heard with the intent to insult one another, for example, on the school playground. Consequently, any other response choice was preferable, considering it could not be used as an insult. The term Autistic “has a strong negative connotation within my son's social circle. Children label others ‘autistic’ when referring to someone slow, just like “retard” was used in my youth. However, I noticed that this negative connotation is rarely present when you add the term ‘person’ [after] autistic. As though the addition of the word ‘person’ were more humanizing.” (Parent, 1st rank preference: Autistic person). Although few autistic respondents chose Person with ASD as their most preferred, one participant wrote that it had the benefit that “it can’t be used to insult someone” (Autistic adult, 1st rank preference: Person with ASD).
Person's choice
To be respectful could mean asking the autistic person how they want to be called or referred to. “You will not find homogeneity in [their] answer. Every person is different… Labels are used to define oneself within a process of acceptance of one's difference from the norm.” (Autistic person, 1st rank preference: Autistic). Parents reported asking their children or checking on social media what were the most appropriate or appreciated terms to use. Another said she “wanted to adopt the correct vocabulary to talk about [her] child's identity or particular traits with respect and kindness” (Parent, 1st rank preference: Autistic person).
An autistic participant noticed that non-autistic persons tend to choose a different terminology from the autistic community, and “the choice of term should depend on whom you are talking to” (Autistic adult: general comment).
Reject the pathology paradigm
The rejection of medicalized or pathologizing language was justified to avoid stigma. For some, response choices containing the diagnostic jargon Autism Spectrum were rated as most offensive, especially when they contained the words disorder or condition. It is worthy of note that “disorder” is translated into French by the word “trouble” in the DSM. This term adds another negative dimension of being “troubled” or a “trouble”. Hence the recurring comment “je ne suis pas un trouble” (I am not a trouble) (Autistic adult, 1st rank preference: Autistic person). “It's like saying first and foremost that the person has a problem.” (Parent, 1st rank offensiveness: Person with ASD). Trading “trouble” for “condition” did not remove this negative, pathologizing connotation. “It sounds like autism is an incurable disease” (Autistic adult, 1st rank offensiveness: Person with ASC). One parent “never liked autism to be called a disorder. My son does not have a disease, he is simply different” (Parent, 1st rank preference: Autistic person). Nevertheless, one could have mixed feelings regarding the word “trouble”. Without the label that autism is a disorder or a condition, “there's a risk that society will discharge itself from its responsibility” (Autistic adult: 1st rank offensiveness: Person with ASD) of providing services.
A neurominority in a neuro normative world
When asked why a certain term was rated as most offensive, some respondents explained that what has been labeled a disorder, disability, or condition is created by society. Thus, autistic participants justified the ranking of their selected most offensive term as a lack of respect for their reality. The challenges they face stem from the fact that they “have a neurological difference which makes [them] less adapted” (Autistic adult: 1st rank offensiveness: Person with ASD) to a society designed by and for individuals with predominant neurotype. Certain respondents stated that no particular term was offensive to them. What they found offensive was the lack of understanding from society, mostly composed of non-autistic people. “I don’t have any problem per say. The society is not adapted to my needs, nor those of our [the autistic] community. Neurotypical individuals are ignorant about the reality of autism. That's the problem!” (Autistic adult: 1st rank offensiveness: Person with ASD).
To be understood
Preferences were also justified by a desire to be understood. Therefore, the choice of vocabulary used to designate oneself or an autistic person depended on the social context or interlocutor.
Short and simple
Although Autistic was feared to become an insult by some, it was most preferred by those who wanted to use simple and short terms to be understood, as it was better known and needed no further explanation. Referring to Person with autism or Person on the autism spectrum, an autistic adult expressed that these nominations were unnecessarily long and ambiguous; the shorter, more “direct term Autistic [has] exactly the same definition” (Autistic adult, 2nd rank preference: Autistic) and “it says it all.” (Autistic adult, 1st rank preference: Autistic). Respondents added that such convolutions create misunderstandings and may promote further prejudice. In that sense, Autism spectrum condition, translating to “Personne ayant une condition du spectre de l’autisme”, was especially problematic. “It's way too long, it is currently unknown… and ordinary people will never be able to adopt it.” (Autistic adult, 1st rank offensiveness: Person with ASC). One professional went further, saying she found it “disturbing.”
In the end, using a long string of words to be politically correct was considered “hypocritical” as though an attempt to hide the condition from people who do not know the term or make the diagnosis of autism easier to accept by using many “intelligent sounding” or “pompous” words (e.g., spectrum, condition). “The situation remains the same”, for the autistic person and their entourage. “Stop trying to be politically correct… I’m autistic. Nothing is insulting in [saying] that.” (Autistic adult: 1st rank preference: Autistic person).
To be accurate
Some respondents explained their number one choice as a desire to be accurate and precise, which also applied to the desire to be true to one's identity.
Autism is an identity, not an accessory
The terms Autistic and Autistic person were most preferred by autistic respondents and justified by comments that reflect it as part of one's identity, not an accessory. At times, they explicitly referred to identity-first language “the same way one could say she is right or left-handed.” (Autistic adult, 1st rank preference: Autistic). Respondents referred to autism as an identity, not something one chooses “to leave behind” or “put down” like an object you carry along when referring to the most offensive terms. These arguments applied to Person
Autism is diversified
Although “autism spectrum” framed as a disorder or a medical condition, was rated as most offensive, pathologizing, and stigmatizing, it also had its proponents as the preferred way to refer to autism. One reason presented above was that it puts the person first. Another reason was that the word “spectrum” conveyed the great variability of expression of autistic traits and “levels of functioning” among the autistic community. At one end of the spectrum, “one could have minor difficulties up to major impacts on social participation” at the other end (Professional, 1st rank preference: Person on the autism spectrum). For that same reason, the term Autistic could be rated most offensive since it lacked nuance and was too restrictive. It referred “to the most severe cases and categorizes my child as a dysfunctional, disturbing person in society.”
Back to the future with asperger?
When asked what other terms respondents used to refer to autism, “Aspi,” “Asperger,” and “Neurodivergent” were the most cited. They mentioned that Asperger is the term they use amongst themselves. Some were rooting for the “return of the [diagnosis of] Asperger in the DSM-6 because [they] are pretty different, [them] Aspis, compared to other autistics” (Autistic adult, general comment). “Compared to us, most autistic persons are much more challenged. I don’t feel comfortable calling myself Autistic… I have much less severe limitations… It's like calling myself a burn victim when my finger barely touched the flame of a candle.” (Autistic adult, general comment).
Some participants with a diagnosis of Asperger Syndrome prefer to present themselves as such since they have been confronted with denial or disbelief from certain people when disclosing that they have autism. In other instances, they reported having had to explain the lesser-known notion of Spectrum. Others thought using the former diagnosis in popular culture was a disfavor to themselves and others with what once was considered “high-functioning autism.” It denied the recognition of the challenges they face in society and caused unnecessary division within the autism community.
Discussion
The choice of words is paramount for developing a positive identity and the process of social acceptance. Autistic people, having a minority neurotype, experience intersectionalities and stigma that cause them stress and affect their mental health (Botha & Frost, 2020; Shaw et al., 2022). Therefore, it appears essential for people wishing to be allies to identify and employ the correct terms when addressing them.
Confirming our first hypothesis, no unique term was unanimously preferred by our sample of the French-Canadian autism community, nor inside each group of stakeholders. When considering only quantitative results, one may be prompt to conclude that autistic adults prefer identity-first terms, such as Autistic and Autistic person. However, supported by the comments respondents offered to justify their term ranking, such “a move completely away from PFL may be premature” (Botha et al., 2023, p.871). Furthermore, our results support the association between term preference and experiencing autism first-hand, as suggested by Kenny et al. (2016). In fact, quantitative results show a clear dichotomy between which term professionals prefer and what autistic adults consider offensive. Person with ASC was ranked lowest in preference and most offensive among autistic respondents, replicating results from Bury et al. (2023). Recent studies, including this one, have shown that within the autistic community itself, there is no unanimous preference for identity-first language (Buijsman et al., 2023; Bury et al., 2023; Geelhand et al., 2023; Keating et al., 2023; Kenny et al., 2016; Taboas et al., 2023). Since they are the ones who will have to live with the terms chosen to identify them, it seems evident that autistic people themselves need to be questioned on an individual basis to find out how they want to be addressed.
In accordance with our third hypothesis, age at the time of diagnosis plays a role in one's choice of words. As predicted, those who received their diagnosis in adulthood preferred and were less offended by the terms in line with identity-first language (e.g., Autistic and Autistic person) and manifested a dislike or were offended by terms related to pathologizing or person-first language (e.g., Person with autism, Person with an autism spectrum disorder). Supported by recent research (Kiehl et al., 2024; Overton et al., 2023), those diagnosed in adulthood may have different, more positive experiences in their autistic identity development than those diagnosed in childhood. This could be because they have recently self-identified as autistic, lived without a medical label for a longer period of time, and have developed a more positive autistic identity (Fletcher-Watson, 2024; Overton et al., 2023). On the other hand, individuals who have received the diagnosis in childhood may have internalized the ASD label and are consequently less offended by PFL (Fletcher-Watson, 2024). Additionally, they may have been called “autistic” in a derogatory way in childhood, which could explain their reluctance towards this term and their preference for the term ASD (Trundle et al., 2022). Thus, reference to IFL terms for the majority of our sample is coherent with the inclusion of autism in their identity and the rejection of references to pathology, being, or having a problem.
Respect should Come first
The rejection of medicalized language was apparent among French-Canadian autistic respondents. The label “disorder” being translated as “trouble” in the French DSM-5, could assign an additional negative connotation of being a “trouble for society” or “being troubled.” Professionals working with autistic individuals seem unaware of this aversion. Instead, they favored using Person with ASD, as reported in other studies (Garcia-Molina, 2019; Kenny et al., 2016) and this one. Some even justified their preference for this term based on respect. They are thus well-intentioned but seem to miss the mark. We suspect this strong preference reflects a desire by professionals to be accurate and use the appropriate revised diagnosis chosen by the DSM-5-TR (APA, 2022) as they were taught and employed in their work environment.
Being a person first is not a problem
The thematic analysis suggests that the debate may not be between “person-first” and “identity-first” language (Vivanti, 2020) but rather pathologizing/ableist language (Bottema-Beutel et al., 2021; Kapp et al., 2013) where autism is considered a disorder or a deficit versus a neurodiversity-affirmative language where autism is conceived as a difference (Shaw et al., 2022). As qualitative results from our study suggest, it may not be the “person first” that precedes the appellation of the diagnosis that people reject, but “with” “having” and “on” that emphasizes the person's impairment. Despite autistic activists rooting for adopting IFL over PFL, two studies have shown that Person on the autism spectrum was the most preferred term by their autistic participants (Bury et al., 2023; Kenny et al., 2016). This fits well with our interpretation that “person-first language” is not as problematic or offensive as the terms “disorder” and “condition” (Botha et al., 2023) following “autism spectrum”. Both IFL and PFL arguments justify the preference for Autistic person by our French-Canadian sample. One cannot assume that this specific term is IFL – as Geelhand et al. (2023) have done – nor that it is PFL – as assumed by Garcia-Molina (2019) for the Spanish “persona autista”. In accordance with previous studies, respondents’ comments made it clear that it is both an identity (Botha et al., 2022; Bury et al., 2023; Geelhand et al., 2023; Keating et al., 2023) and a part of one's identity (Bury et al., 2023; Keating et al., 2023) which explains the desire of putting the person first.
Furthermore, advocates for IFL have argued that separating autistic individuals from their autism, in expressions such as Person with autism, Person with ASD or Person with ASC leads to stigmatization, dehumanization, and violence (Botha et al., 2022, 2023; Botha & Gillespie-Lynch, 2022; Bottema-Beutel et al., 2021; Gernsbacher, 2017; Vivanti, 2020).
In sum, based on the present findings, it seems that the pathology paradigm exemplified by ableist language is more central to the debate than using language that places the person first (Bottema-Beutel et al., 2021; Bury et al., 2023; Kapp et al., 2013; Kenny et al., 2016; Walker & Raymaker, 2021). The overall preference and justifications for Autistic person and clear rejection of Person with ASC by French-Canadians calls for considering autism as part of one's identity, which cannot be “left behind” or “treated”. Our results highlight the nuance required in choosing terminology, namely, selecting a clear, comprehensible term that accepts, weighs, and contextualizes the person's abilities and challenges (Sinclair, 2010). Because the term Autistic person was one of the most preferred terms and the least offensive for the main stakeholders, we conclude that this term responds to what is considered respectful, clear, and accurate.
Limits
Appreciation of the results must be done in consideration of certain limits. We did not ask participants to specify their region of residence in Canada. As the province of Quebec is considered a distinct nation, one may infer that its culture is specific to this region. Moreover, because culture impacts vocabulary, it is plausible to expect slight differences in terminology from one region to another inside this vast country. Therefore, caution is warranted if one wishes to generalize the results to all French-speaking Canadians.
We chose not to include Asperger's syndrome and its affiliated terms in the selected terminology for the survey. We recommend including these terms in future studies, as many respondents commented on the matter.
Future research
As mentioned previously, preference in terminology is not a stable concept. As awareness, acceptance, and understanding of autism evolve, terminology will continue to adapt to reflect these changes. We encourage future efforts to track which terms are considered offensive and which are preferred to keep up with evolution and avoid causing harm. In further consultations with the autism community, we recommend adding terms not necessarily validated by the DSM, such as Asperger syndrome or neurodivergent (Kapp et al., 2013), since our current results tend towards a distancing from the medical terminology.
Conclusion
Terminology in autism has rapidly evolved in the last decade. At this time, a consensus in the French-speaking community of Canada has not been reached, nor should it. Our study sheds light on a clear dichotomy between autistic adults’ wishes and professionals’ use of terminology. Although professionals intend to be kind and respectful when using PFL, it does not come through that way because of the overbearing stigma related to medical terminology. We conclude that, when possible, it is best to ask which terms the person prefers we employ, thereby creating an opportunity for dialogue and bringing down a barrier to inclusion. When this is not possible, because, for instance, a group of individuals is concerned, the term Autistic person (“personne autiste”) is to be favored because it is rated as the least offensive, it represents IFL and PFL in French, and is one of the most preferred terms by autistic participants of our study. Furthermore, using this term is coherent with a person-centered approach favored in professional settings. Thus, we invite professionals to reflect on the coherence of their preferred terminology and intervention approach.
Our study adds to the growing body of literature in favor of moving away from terminology conveying the attitude that autism is a disorder in need of a cure in favor of adopting the viewpoint that Autistic persons form an ethnocultural minority group in need of adaptations in recognition of their rights.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Fonds institutionnel de développement de la recherche et de création- Université du Québec en Outaouais, (grant number 3451035).
