The Dharma of Gestalt Processing: Hidden in the Evidence

Older evidence often outlives the theories that first explained it. Gestalt processing reveals observations preserved for decades—not as new discoveries, but as phenomena waiting for a more coherent architecture.

Baron-Cohen, S., & Staunton, R. (1994). Do children with autism acquire the phonology of their peers? An examination of group identification through the window of bilingualism. First language, 14(42-43), 241-248.

Kremer-Sadlik, T. (2005). To be or not to be bilingual: Autistic children from multilingual families. In Proceedings of the 4th International Symposium on Bilingualism (pp. 1225-1234).

For years I believed I was searching for evidence.

I no longer think that is what I have been doing.

Evidence was never the scarce resource. The scarce resource was an architecture capable of recognising what the evidence was already holding.

This became clear to me whilst reading papers that predate the language of gestalt processing itself. Again and again I found careful observations embedded inside theoretical conclusions that seemed almost unable to see what they had just described. Researchers documented autistic children preserving intonation, carrying whole phrases across contexts, speaking through familiar voices, retaining accents that their peers had long abandoned, organising language around repetition rather than analytic recombination. The observations were often meticulous.

The interpretations belonged to another world.

Seen through the lens of dharma, I began asking a different question.

Not whether these papers had “found” gestalt processing.

But what they had been upholding all along.

The dharma of evidence is not to be correct.

Its dharma is to preserve the phenomenon.

Interpretations change. Diagnostic categories change. Entire theories rise, dominate a generation of research and quietly disappear. The phenomenon itself remains remarkably indifferent to these intellectual fashions. Children continue developing. Families continue noticing patterns. Clinicians continue recording behaviours. Something real continues unfolding even when we lack the language to understand it.

The evidence carries that reality forward.

This became particularly striking when I revisited Baron-Cohen and Staunton’s 1994 study of bilingual autistic children. They observed that autistic children often retained the phonological patterns of their families rather than converging on the accents of their peers. The interpretation was straightforward within the dominant framework of the time: autistic children were thought to identify less strongly with peer groups and therefore assimilated less readily.

It was a coherent explanation.

It was not the only one.

From the standpoint of gestalt processing, the same observation becomes visible through a different architecture. The family voice is not merely an accent awaiting replacement by the socially dominant variety. It is part of the relational whole through which language first became coherent. The melody of speech belongs to the people who regulated distress, established routines, laughed, reassured, comforted and welcomed the child into language itself.

The accent is not simply preserved.

The relationship is.

Nothing about the observation changes.

Only the architecture through which we understand it.

Kremer-Sadlik’s work on multilingual autistic families performs another quiet service. Rather than treating bilingualism as a problem to eliminate, she documented the lived realities of multilingual homes and questioned clinical advice that routinely encouraged families to abandon their heritage languages. Once again, the observations exceeded the theories surrounding them. The data were richer than the questions researchers believed they were asking.

This pattern appears everywhere once it becomes visible.

Echolalia was documented long before it was understood as communicative.

Prosody was measured long before it was understood as relational.

Delayed language was described long before anyone asked whether development might be organised around wholes rather than parts.

Researchers kept recording the phenomenon.

The framework kept translating it into deficit.

This is why I have become increasingly uncomfortable with claims that there is “little evidence” for gestalt language processing.

There is little evidence if evidence means papers that explicitly use today’s vocabulary and today’s theoretical commitments.

There is an extraordinary amount of evidence if evidence means careful observations waiting for a more adequate conceptual home.

The distinction is methodological rather than rhetorical.

Every scientific observation contains at least two layers.

The first is what was seen.

The second is what the observer believed it meant.

Those two layers are often treated as inseparable.

They are not.

The observation may survive the collapse of the theory that first interpreted it. Indeed, this is how science often progresses. New theories do not always require new phenomena. Sometimes they simply allow old phenomena to become intelligible in ways they never were before.

Gestalt processing has taught me to read the literature the way a gestalt processor reads language.

The whole matters.

But wholes can also be opened.

The original papers deserve respect because they preserved something genuine. Yet they need not remain permanently bound to the conceptual frameworks that first held them. Just as a Stage 2 gestalt can loosen without losing its coherence, an observation can become separated from the explanatory structure that once surrounded it.

The phenomenon survives.

The theory does not always.

This has become the central methodological insight behind Hidden in the Criteria. I am not arguing that generations of researchers somehow overlooked the existence of gestalt processing because they failed to observe autistic children carefully enough.

I am arguing something much stranger.

They often observed it with remarkable care.

They simply inherited categories that made other interpretations appear inevitable.

The evidence was never absent.

It was organised differently.

Perhaps this is the dharma of older evidence.

Not to prove the future.

Not to anticipate concepts that did not yet exist.

But to carry the phenomenon farther than the theory that first interpreted it.

Like a gestalt, it preserves more than its surface appears to contain.

It waits.

Until another architecture makes it speak differently.