You Know It When You See It — Except When the Internet Tells You You Don’t:

How Research on “Frank” Autism Undermines a Persistent and Harmful Meme

With a few friends, I have started working on a project that may be seen as controversial but feels deeply important to those of us who live with the consequences of how autism is constantly redefined, reframed, and repackaged online[1]. We decided to start small—by analysing the memes that circulate endlessly across social media—and then follow the discussion wherever it leads us.

When I looked at a list of memes titled What Not to Say to an Autistic Person, the very first “not to” made me laugh out loud: “You don’t look autistic”. That instruction has never been applicable to my son. Not when he was a young, nonverbal autistic child. Not now, as a nearly 38-year-old echolalic autistic man.

He does look, behave, speak, move, and live in ways that many people immediately recognise as autistic. In public, strangers often approach us with some version of: “He’s autistic, isn’t he?” This is usually followed by an explanation – someone worked with autistic children years ago, has an autistic child, has an autistic friend, or once shared an office with an autistic colleague. We have had countless friendly conversations like this. The most recent one took place at a Christmas concert, where my son was dancing with his sister’s friend.

“Beautiful dancing,”[2] someone said. – “He’s autistic, isn’t he? My colleague has an autistic daughter…”

This is not rudeness. It is recognition. It is an everyday example of what clinicians refer to – somewhat awkwardly, but precisely—as “easily recognisable autism”. Or, to put it less diplomatically: if it looks like a duck, walks like a duck, and quacks like a duck, it is probably a duck. And if it looks like a goose, walks like a goose, and honks like a goose – it is not a duck.

This meme (of ‘invisible’ autism) circulates endlessly online, often framed as an ethical correction: Autism is invisible. You never know. Stop assuming. The intention may be protective. The claim itself is wrong. Not universally wrong – ASDs are profoundly heterogeneous, encompassing a wide range of developmental trajectories, cognitive profiles, communication styles, adaptive abilities, and lived experiences – but wrong often enough that pretending otherwise obscures both clinical reality and lived experience. And, importantly, research now shows that experienced clinicians are not only making rapid judgments—they are usually correct when they do.

The Concept of “Frank” Autism

In a 2017 paper, de Marchena and Miller introduced the term “frank autism”[3], borrowing frank from medical usage, where it means clinically evident and unmistakable. This phenomenon has been discussed informally for decades, perhaps as “classic autism” (Kanner’s autism). However, there is no unitary “classic” presentation, and classic autism does not seem to correspond to level of functioning. Rather, there are likely several “frank” presentations, when some autistic people exhibit a distinctive behavioural profile that signals autism within minutes—or even seconds—of interaction.

To examine this systematically, de Marchena and Miller surveyed 151 clinicians from a range of disciplines involved in ASD diagnosis. The results were striking:

·         97% of clinicians reported encountering frank autism in their practice.

·         Clinicians estimated that about 40% of autistic individuals have a frank presentation.

·         These impressions are formed rapidly[4]:

o    3% within seconds

o    8% within 1 minute

o    52% within 10 minutes

o    87% within 30 minutes

o    94% within 60 minutes

Importantly, frank autism did not necessarily correspond to level of functioning. Highly verbal, intellectually able, socially motivated individuals could be just as “frank” as those with higher support needs. This alone dismantles another popular meme: that recognisable autism is synonymous with severity.

What Are Clinicians Actually Seeing?

When clinicians were asked what behaviours most strongly contributed to a frank impression, their responses clustered around several domains.

Repetitive behaviours/Interests – motor mannerisms, repetitive language such as echolalia/ scripting, and stimming/sensory-seeking behaviours, restricted interests, repetitive object use (e.g., lining up toys/items), repetitive & ritualised behaviour – were also commonly cited. These behaviours need not be constant; once observed, they can rapidly shape a clinician’s overall impression.

Social reciprocity featured prominently—not merely reduced interaction, but its quality: minimal spontaneous social overtures, atypical engagement, or conversational styles dominated by monologues rather than reciprocity.

Nonverbal communication was another major contributor. Poor or atypical eye contact was frequently mentioned, but so was atypical prosody (unusual rhythm, pitch, or intonation of speech)[5], facial expressions (“flat affect”), social skills.  

Finally, clinicians often mentioned general motor behaviour, including posture, gait, and awkward or unusual movements—features not emphasized in diagnostic manuals but evidently salient in real-world diagnostic reasoning.

Some frank features may be present continuously and thus would be immediately apparent upon meeting an individual (e.g., unusual prosody). Other behaviours may not be present continuously but once exhibited can lead to a rapid impression of ASD (e.g., repetitive motor behaviour or speech). There may be a range of frank presentations based on different specific behaviours or clusters of behaviours.

Crucially, none of these behaviours is unique to autism in isolation. What matters is their specific configuration and qualitative “feel.” As one clinician put it: “It’s really not the severity; it’s those telltale signs that are really only seen in ASD.”

Five Minutes Exactly: When First Impressions Hold Up

Skeptics might argue that such quick impressions are simply bias disguised as expertise. That concern makes the findings of Wieckowski et al. (2021) especially important. In their study, expert ASD diagnosticians observed toddlers aged 12–53 months for just five minutes before completing a full diagnostic evaluation. After this brief observation, clinicians recorded whether they believed the child would ultimately meet criteria for ASD and how confident they were.

The results were unambiguous:

• Initial impressions matched final diagnoses in 81% of cases overall.

• When clinicians suspected autism after five minutes, they were correct 92% of the time.

·         These findings were largely consistent with earlier studies that found similarly high confirmation rates following brief observations (Gabrielsen et al. 2015).

In other words: when autism looks obvious to experienced clinicians, it usually is.

What did not hold was the reverse. When clinicians did not observe autism within five minutes, outcomes were more mixed. This asymmetry matters. Frank autism appears to be highly specific but not universal—present in a substantial subset, absent in others who nonetheless meet full diagnostic criteria.

Wieckowski and colleagues estimate that around 60% of toddlers with ASD in their sample showed a presentation distinctive enough to be recognised almost immediately. This aligns well with de Marchena and Miller’s broader estimate of frank autism across the lifespan.

Experience Changes What You Can See

Both studies agree on another uncomfortable truth for meme culture: expertise matters.

Research on diagnostic reliability shows that agreement between inexperienced clinicians is poor, even when using formal criteria. In contrast, agreement between experienced clinicians making judgments based on clinical impression is excellent (Klin et al. 2000). The implication is not that checklists are useless—but that expertise allows clinicians to integrate subtle behavioural qualities that resist easy quantification.

Even more provocatively, experimental studies show that non-experts can distinguish autistic from non-autistic individuals at above-chance levels after watching video clips lasting only seconds, using gestalt judgments such as awkwardness or approachability (de Marchena & Eigsti 2010; Fusaro et al. 2014; Grossman 2015; Sasson et al. 2017). If autism were truly invisible in all cases, this would not be possible.

Why This Matters (and Why the Meme Fails)

When memes insist that autism is always invisible, they erase:

• autistic people who are visibly autistic,

• families whose experiences contradict the narrative,

• clinicians whose expertise is grounded in observable reality,

• and research showing that rapid recognition is often accurate.

  
  

This does not mean autism is simple. It means that simplicity memes are bad science.

Frank autism does not define the spectrum—but ignoring it distorts both diagnosis and discourse. If we collapse all autistic individuals into a single, undifferentiated category, we risk studying everything at once and understanding very little.

Sometimes, you really do know it when you see it. And pretending otherwise helps no one.

__________________ 

[1]. The project will take the form of a series of blog posts.

[2] I wouldn’t say it’s “beautiful” 😊 – the girl was following his lead, mirroring his movements, while introducing similar but slightly different turns. And they both enjoyed it!

[3] It may be linked to “prototypical autism” (Mottron & Gagnon D 2023).

[4] Categories are cumulative, e.g., “Within 1 minute” includes all participants who responded that they make frankness judgments within 1 minute, as well as all respondents who form impressions within seconds.

[5] Prosody has long been noted as one of the most immediately recognisable features of autism (Mesibov 1992).