In a new(ish) paper (first published a year ago), Lorcan Kenny of the Centre for Research in Autism and Education asks “Which terms should be used to describe autism?“. The paper provides a much-needed empirical analysis of this question which has beset the autism community for years. Conflict over the appropriate terminology to describe both autism itself and specific elements of autistic experience is rife. This semantic debate can sadly create barriers between groups who would otherwise be aiming for common ground – as articulated by James Cusack of Autistica at a recent seminar “to help autistic people live longer, happier, healthier lives“. Different stakeholder categories including people on the spectrum, parents of children with autism, practitioners from various sectors (healthcare, education, charities) and researchers should ideally be working together to make progress towards this goal, and instead, all too often, the effort is hampered by disagreements over language.
But, now that we have this evidence-base on the best terminology to describe autism, we can put all that behind us, right?
Sadly, no. Don’t get me wrong – this paper is excellent and really helps us understand the lie of the land with regard to talking and writing about autism. For anyone who hasn’t read it, the paper surveyed almost 3500 people in the UK who were connected to autism: “autistic people, parents and their broader support network”. The authors asked people to choose words from a list of different ways to describe autism and to rate a series of terms on a scale from strongly dislike to strongly like. Respondents were also asked to select the one term they would use to describe autism. This methodology isn’t perfect (when is it ever?) – for example, the respondents had to choose from pre-defined options rather than offer their own forms of language. But it is a solid piece of social science and, most importantly, directly addresses an key question with clear empirical data.
However, what it doesn’t do, and no research study really ever can do, is definitively dictate the “correct” form of language to use.
I’ve got three reasons.
- A wide range of language is preferred to describe autism by different individuals.
Selecting the ‘correct’ form from the list is basically impossible because there’s so much variability in opinion. In first place, there are differences between stakeholder categories. For example, professionals are less likely to endorse “autistic” and more likely to endorse “autism spectrum disorder (ASD)” than the other groups surveyed. Parents, professionals and other supporters are all roughly equally likely to endorse “has autism” (about 45%-50% do so) but only about 35% of autistic respondents select this phrase as an acceptable form of language.
This pattern may reflect the fact that different forms of langage are appropriate to different scenarios, readers or listeners. This issue isn’t only about using the form of language which is most acceptable to a particular audience, but also about the context. For example, many practitioners doubtless use “autism spectrum disorder” to describe autism because that is the most correct terminology for a healthcare or education context. If they are writing reports which have the function of securing the support that an indivdual needs – such as classroom assistance or specialist mental health services * – they need to use the terminology which is most likely to serve that purpose. Taking context into account is a crucial filter for determining the ‘correct’ language to use.
However, we might also legitimately state that, all other things being equal, the most important voice in the debate over the ‘correct’ language to describe autism is the autistic voice. Unfortunately, there are also big differences between individuals within that category. One often-cited figure is that 60% of autistic people endorse the term “autistic” to describe autism (Figure 1 in the original paper). However, slightly more endorse “autism” and more than 50% in this stakeholder category endorse “on the autism spectrum” and “Aspie” as well. When asked what specific term they would use to describe themselves the most popular options are “Asperger’s syndrome” or “has Asperger’s” – and these are just not appropriate for everyone with an autism diagnosis. When asked to select just one term to describe autism from a list of options, “autistic” is the most popular, but only receives about 20% of the vote. Close contenders include “on the autism spectrum”, “has autism or has Asperger’s” and even the more medical “autism spectrum disorder” which has about 12% of the vote.
2. This is only a single result
Going back to that 60% figure – the number of autistic respondents who endorse the term “autistic” to describe autism. First, let me re-emphasise that this 60% figure comes from the section of the survey where the respondents were able to select a number of different acceptable forms of language. So some of the 60% who selected “autistic” as a preferred option also selected other options as well. “autistic” in this context is just one of a few different terms which receive more than 50% support.
But even if we take that as the ‘preferred’ term, what if we replicated this survey again with a new sample – what might happen to that statistic? It could go up – we might have a larger percentage of respondents on the spectrum who select “autistic” as a preferred term to describe autism. But it could also go down. What if that figure dropped below the 50% mark in a second survey study, so it wasn’t even endorsed by a straight majority? What if a survey was designed that allowed more autistic people who also have intellectual disabilities to respond (i.e. a survey less reliant on the ability to read) – what would happen to the ratings then? ^
In a nutshell, it is premature at best, inaccurate at worst to take any isolated statistical finding as fact and this is no exception.
3. We need to take into account how strongly people feel about the use of “incorrect” language
Looking at Table 2 in the orginal paper, we can see modal scores for each of the language options presented – the modal score is the score most commonly given by a specific respondent group for a specific item. Interestingly, the modal scores for the respondents with autism never drop below 3, which represents “neither like nor dislike”. In other words, while this group do show a preference for some specific terms to describe autism, there’s no evidence here that there are any terms which are powerfully disliked by this group as a whole. In contrast, the other respondent groups show fairly marked dislike for the term “Aspie” (a favourite among autistic people with a modal rating of 5 = strongly like). Professionals also largely dislike the term “autistic person”.
From these data we might construe that, if our goal is to find a form of language which is minimally disliked by the largest number of people from the autism community, we should be using the terms “has autism”, “has Asperger’s” and “on the autism spectrum”. All of these get modal scores of 4 = like in every respondent group. This reveals a very pleasing level of consensus between groups – in fact there’s a positive message to be gleaned from this piece of research, which is that there is more overlap than disagreement in our thinking about autism terminology than at first meets the eye.
So, what language should I use?
I hope I’ve shown you that being dogmatic about language just isn’t the way to go. In my opinion, this is a case where more research is not going to offer us any answers either. Things will shift as societal perceptions of autism are updated and upgraded, but for the time being I think we should be comfortable with the fact that there is no single, correct form of language to describe autism. This can be very liberating – if we can move beyond language, accepting that different preferences exist, we may make better progress towards our shared goal: making a positive difference in the lives of people on the autism spectrum, and their families. For the time being, I will continue to do what I have been doing for some time (and what I have done in this post), which is varying my language from sentence to sentence. I hope that by doing so, when I write or speak about my research, attention to my specific choice of language will diminish and the true message will more powerfully reach my audience.
One final point…
There is one other facet to this paper which I haven’t managed to cover in detail and wanted to mention before I finish this post. The authors also did a qualitative analysis on answers to an open-ended question about why people selected the language they did. These reveal a consistently positive and powerful underlying attitude that pervades the entire community. Respondents from all groups recognise the power of language, the importance of an individual’s personal preference in reinforcing their identity, and the value of listening to each other and tailoring terminology to the context. They acknowledge the negative impact of language which emphasises deficit or disorder, an strive to put strengths and differences in the foreground. To me, there is an overwhelming sense of an entire community focused on putting language to good use in providing positive recognition and appropriate support to people on the spectrum. And that’s something we can all endorse.
[…an addition responding to excellent debate on twitter, 16th August 2016…]
There’s been a lot of healthy debate on twitter (see my feed for 15th and 16th August 2016 for details) since I posted a link to this piece yesterday. I wanted to collect together a few thoughts on that and highlight some places where other perspectives on this issue can be found.
First, I think the point which I need to make most clearly is that this blog post is about interpreting a specific piece of research to see if there is empirical evidence for a ‘correct’ way to talk about autism. My conclusion is that the research doesn’t support one single preferred option, even when we look only at the responses of autistic people. However I simultaneously think that:
- There are other ways to approach this question, such as arguing for a particular form of language from a theoretical position. Lots of people make the very effective case that ‘autistic person’ is the best option because it places autism at the core of the individual rather than as an implicitly-undesirable add-on. Thank you to @spacedoutsmiles and @soundcube and @soniaboué among others for making this point. Others on the spectrum feel the precise opposite – actively preferring forms of language which place autism as a part of them, but not their entire identity. This post by former colleague @autismsim describes this perspective, with evidence from a number of people with autism, extremely well.
- Whatever the research may or may not say and whatever theoretical position you adopt on terminology, nothing (in my personal opininon) should be more important than the personal preference of the individual you are speaking to. For this reason, it is always important to check with a person about their preferred language and adopt that in conversation. Of course, this principle doesn’t help when we are speaking to / writing for a large group whose opinions may differ.
The second thing I want to do is link through to some other places where you can find interesting debate on this piece of research but also on the richer issues around it – such as the question of why differences in terminology arise (thank you @kabieuk and @milton_damian for highlighting this point in particular). If you scroll down to July 2015, this blog post by @autismage discusses the legitimacy of incorporating professionals into the autistic community and placing their language preferences on an equal footing to the preferences of people on the spectrum. The same point was also highlighted on twitter by @the woodbug and others. Other posts which have infuenced my thinking on the language used to talk about autism include a piece by Judy Endow which argues for a person-first attitude which transcends choice of language. And another influence is this heartfelt plea by Autistic Not Weird, to move beyond disagreements about language.
If you would like to respond to the original post, or these additions, by sharing your perspective please consider commenting below.
* chance would be a fine thing…
^ It is important to note here that I very powerfully believe in the importance of a strong autistic self-advocate movement to represent the unheard voices of people on the spectrum who are excluded from many traditional modes of communication and, consequently, from research and from societal and policy debates. Respecting the voices of autistic people and responding to their messages confers benefits on the entire a/Autistic community by undermining disabilist attitudes. However we cannot presume that those voices – of autistic people with and without intellectual disabilities – would or will always want to transmit the same message.