We Made Language Anyway: How the Language Acquisition Device Fails GLPs, and What Comes Next

Why Chomsky’s theory failed us, what it cost, and how gestalt language processors are reclaiming the story of how we speak.

Chomsky’s LAD never described how I or other GLPs came into language. We weren’t broken—we were just never included. This piece explores why the theory endures, what it erases, and how we speak anyway.

Introduction

It’s a strange thing, realising—years or decades after the fact—that you didn’t acquire language in the way you were told all humans must. That you never had some innate device quietly unfolding syntax in your brain like a flower in spring. No neat grammar rules emerging from your biology, no universal grammar lighting the path. Just fragments. Chunks. Gestalts. Scripts. Echoes. You learned language the way a scavenger learns shelter: by piecing it together from what you found, what felt safe, what sounded like meaning. Each language I now operate in some form or another—five in all—I learned. Not absorbed. Not decoded. Learned. Constructed. Rehearsed. Re-assembled. That fact alone unseats decades of linguistic theory—particularly the model that still, stubbornly, claims to describe all humans: Noam Chomsky’s Language Acquisition Device (LAD).

This isn’t just a theoretical difference. It’s an existential one. Because if the LAD was supposed to explain how we come into language, then what does it mean when it doesn’t explain me? Or the autistic Moroccan boy who learned fluent English from YouTube videos, without live human interaction (Kadiri & Anasse, 2023)? Or the echolalic child whose scripts hold more truth than any sentence a therapist might try to coax from them? What does it mean when the theory doesn’t make room for us—not just as exceptions, but as entire ways of being?

This piece is not just a rejection of the LAD. It’s an exploration of its limits, its persistence, and the quiet violence it enacts—on the minds it misreads, the speech it pathologises, and the lives it renders illegible. For GLPs, the consequences are not abstract. We are marked early, framed as delayed or disordered, denied communicative agency unless our language arrives in the form the model expects. Our gestalts are labelled as scripts. Our echolalia is dismissed as noise. The depth of our meaning—emotional, relational, embodied—is flattened into deficit. The LAD doesn’t simply fail to describe us; it disqualifies us from full recognition as language users. And still, somehow, we make language anyway.

Thus we begin with a simple recognition: that for many of us, especially autistic and disabled gestalt language processors, language was never acquired—it was survived, remembered, repeated, and eventually made our own. And so I ask: if Chomsky’s theory fails to describe us, why does it still dominate the field? What institutional machinery keeps it alive? And more importantly—what frameworks might begin to honour how we actually come into language, in all our plurality and difference and rhythmic, recursive truth?

Part I: How the LAD Fails GLPs

The LAD was introduced by Chomsky as a kind of theoretical fix. A placeholder for something innate, internal, and quietly omniscient—an internal module that allowed children to acquire language rapidly, despite the so-called “poverty of the stimulus.” The idea was elegant in its way: if children don’t hear enough clean, consistent grammatical input to explain their fluency, then something within must be doing the heavy lifting. Something universal, rule-based, recursive. Something that knows how to generate language.

And yet—so many of us are living proof that this model doesn’t hold. Because language didn’t arrive in us that way. Not through generative rules. Not as a syntax-first phenomenon. Not as an automatic unfolding of grammar inside the brain. For gestalt language processors—GLPs—language arrives differently. Not as structure, but as scene. Not as discrete units, but as wholes. It is remembered, not constructed. Stored, not assembled. Unpacked over time through repetition, resonance, and relational safety.

We are not broken examples of typical development. We are living counterexamples to the theory itself.

What the LAD Claims

Chomsky’s LAD rests on a few key assumptions, each of which collapses when held up against GLP experience:

• That all humans share an innate Universal Grammar, preloaded in the brain.

• That language is acquired rapidly and naturally in early childhood through exposure to spoken input.

• That language development is driven by recursive syntactic structure—the embedding of clauses within clauses.

• That deviation from this pattern implies disorder, delay, or dysfunction.

This framework doesn’t just miss the GLP experience. It actively erases it.

What GLPs Actually Do

For GLPs, language isn’t constructed from smaller pieces like Lego bricks. It’s more like a quilt—stitched from patches of remembered experience. What we call “gestalts” are often whole scripts, intonational contours, or emotionally-laden fragments recalled from films, songs, caregivers, or moments of deep resonance. These gestalts are not meaningless parroting. They are layered, often intentional, and carry rich internal logic.

Language for us is episodic before it is grammatical. We often use full phrases before we can use a single word reliably. We echo before we can generate. We blend languages or voices when that’s what we’ve learned to do to make sense of the world. Meaning is primary; grammar, if it comes at all, comes later.

And crucially—we do not rely on innate rules. We rely on contextual repetition. The same phrase might be used again and again until its rhythm feels safe enough to reshape. Over time, those gestalts may be broken down, re-combined, layered into new expressions—but the process is entirely different from the one the LAD presupposes.

Alternative Pathways: Evidence the LAD Ignores

You don’t have to look far to find empirical cases that undermine the LAD’s universalist claims—particularly when GLPs, autistic people, and non-normative language learners are centred rather than excluded.

• The YouTube English case (Kadiri & Anasse, 2023) offers a particularly striking example: an autistic boy in Morocco who acquired fluent English by watching YouTube, with little to no live human interaction. This undermines both Chomsky’s LAD and Vygotskian interactionism. Language, in this case, arrived through pattern, repetition, and affect—not through dialogue or innate syntax.

• A Japanese study on narrative development in autistic children (Peng, 1988) found that discourse-level language could be taught and developed even when social-pragmatic skills were limited. The children learned through structured storytelling, not recursive grammar.

• In the everyday lives of GLPs, we see language developing not through “triggered” grammatical structures, but through stored wholes. A child might use a full script from a cartoon for months before any single word from that script can be flexibly used. And yet, it comes—through repetition, not recursion.

• Echolalia, long pathologised in clinical frameworks, is increasingly understood by lived-experience researchers as meaningful, strategic, and communicative. It serves regulatory, expressive, and narrative functions. It is not a glitch—it is a method. And it is one the LAD completely fails to account for.

• Observations emerging from the Natural Language Acquisition (NLA) community further challenge LAD assumptions—not as doctrine, but as grounded descriptions of what’s happening when language development is gestalt-based. NLA offers not a new universalism, but a documentation of difference. It notices what the LAD erases.

What the LAD Gets Wrong

The LAD model doesn’t just misrepresent GLP development. It frames it as broken—as a deviation from the norm that must be remediated. But the real deviation lies in the theory itself: its presumption of linearity, its insistence on universality, its demand that language emerge in only one way, or else be disqualified.

GLPs process differently. We learn differently. And our language is not less rich, less human, or less worthy of study. It is simply outside the frame Chomsky drew—and maybe that’s the point.

The model isn’t descriptive. It’s prescriptive. It says: this is what counts as language. And everything else? Noise.

Consequences of That Erasure

And so we are misunderstood. Misdiagnosed. Mislabelled. We are taught to feel shame about scripting, to suppress the very structures through which we first made meaning. Echolalia is labelled disordered. Late emergence is equated with deficit. Interventions are imposed not to understand us, but to bring us into compliance with a model that never made space for us.

We are not learning the “wrong” way. We are simply doing what the model could not imagine.

Summary: The Model Fails

If GLPs do not fit the LAD, the problem is not the GLPs. It is the model. It fails not just descriptively but ethically. It positions our difference as defect. It treats our emergence as delay. It encodes a colonial, neurotypical ideal of language and calls it human nature.

And yet, we exist. We speak. We echo. We remember. We remix. We make language anyway.

Part II: Why the LAD Remains Dominant

You’d think, wouldn’t you, that a theory so thoroughly challenged—empirically, experientially, philosophically—would have faded by now. That once you could point to the autistic Moroccan child who learned English through YouTube, or to the echolalic speaker whose scripts bloom into nuanced, poetic truth, or to the GLP adult whose multilinguality emerges from gestalt, not grammar—that the model that erased us would be retired. Refined. Replaced.

But theories don’t die just because they’re wrong. They die when they’re no longer useful to power.

The LAD, as it turns out, is very useful.

A Model That Serves Systems, Not People

The LAD continues to thrive not on its scientific robustness, but on its institutional convenience. It fits neatly into the architectures of Western academia, especially in disciplines like linguistics, cognitive science, and AI. A tidy, internal, rule-based model can be coded. Quantified. Simulated. It aligns beautifully with machine learning metaphors—language as code, input as trigger, recursion as elegance.

And it offers something else systems love: a universal standard. One fixed developmental arc. One idealised speaker-listener dyad. One language acquisition pathway that defines what is normal, and therefore what is not.

It is from that binary—normal vs. impaired—that entire economies emerge: diagnostic tools, therapeutic protocols, special education eligibility rubrics. All predicated on a model that defines the ideal by erasing everything else. For GLPs, that erasure becomes lived. Institutionalised. Material. It becomes IEP goals written in someone else’s voice. It becomes withheld accommodations. It becomes funding denied because one did not acquire language “on time,” even though one acquired it deeply.

Epistemologies of Control: Feminist and Postcolonial Critiques

If Part I addressed the descriptive failures of the LAD, Part II exposes its epistemic violence—the way it smuggles in ideology under the banner of science.

Deborah Cameron, in her work on feminist linguistics, argued that the very idea of “standard language” is a construct—defined and maintained by institutions of patriarchal power (Yukawa, 1999; Hayes, 2010). Linguistic correctness is not neutral—it is political. And the LAD, with its assumption of an internal grammar processor functioning identically across all humans, quietly encodes a particular gendered, classed, and racialised speaker: fluent, male-coded, monolingual, analytic. Everyone else becomes deviation.

Alastair Pennycook takes this further. He writes not just about the content of linguistic theory, but about its methods—how “scientific” approaches to language obscure their own ideological scaffolding (Pennycook, 1991; Martin & Degollado, 2022). For Pennycook, the LAD is part of a larger colonial project: it universalises Western speech norms, then disciplines all other language users against that norm. Method itself becomes control. A way of sorting whose speech gets recognised, and whose gets treated as background noise.

Even the idea of language acquisition becomes suspect under this lens. Acquired from whom? In what context? At what cost?

The Human Line: Posthumanist Disruption

In later work, Pennycook moves into posthumanist linguistics, critiquing the deep-seated human exceptionalism embedded in LAD-style thinking (Sealey, 2018). The LAD imagines language as the defining feature of the human species—a singular cognitive function that separates us from animals, machines, and “less developed” humans.

But what if language isn’t something that only lives in the brain? What if it also lives in gesture, rhythm, pause, co-regulation? What if it’s relational, not internal? Shared, not generated? Then the LAD—like so many Western cognitive theories—collapses under the weight of its own anthropocentric assumptions.

For GLPs, this shift matters. Our language is often multimodal. It may live in repetition, movement, affective resonance. It may arrive through pattern, not parsing. We might use AAC. We might script. We might wait in silence until the moment is safe enough to speak. Posthumanist thought offers a frame that can finally see these not as deficits, but as forms of fluency.

A Colonial Standard Disguised as Science

At its core, the LAD is a colonial artefact. It declares one path—analytic, syntactic, monolingual—as universal, and renders all others primitive, delayed, or in need of correction. It cannot see Mayan children acquiring language through culturally-situated story cycles. It cannot see Deaf children signing narrative poetry in full spatial grammar. It cannot see the GLP child building a language from fragments of Tiny Planets, street noise, and bedtime rituals.

Because the LAD does not want to see difference. It wants to control it.

Difference, under LAD logic, becomes failure. Non-normative language becomes pathology. Multilingualism becomes interference. Echolalia becomes disorder. The entire landscape of linguistic diversity is reframed as a problem to solve, rather than a richness to learn from.

And so the model persists—not because it explains us, but because it disciplines us.

Summary: The Power Beneath the Theory

The LAD survives not by accuracy, but by utility. It gives institutions a metric, a threshold, a binary. It reinforces normative timelines. It shores up whiteness, neurotypicality, and analytic cognition as defaults. It is the linguistic equivalent of a ruler pressed against a child’s body—not to measure their growth, but to remind them they are out of line.

It builds an idea of language that excludes those whose voices—and silences—fall outside the canonical form. And because it is useful to power, it is slow to die.

But we are no longer waiting to be included. We are speaking anyway.

Part III: Other Ways of Knowing Language

If we accept that the LAD cannot describe how all humans come into language—if we’ve seen, and lived, the failures of that model—then the next question mustn’t be: What’s the better model? That, too, would be a trap. A search for the next tidy theory to explain us, contain us, simplify us. But we are not looking for a new device. We are looking for permission to exist as we are.

For GLPs, language is not acquired like a skill or triggered like a circuit. It is relived, remixed, and recalled. Not decoded, but reconstructed. Not parsed, but felt. Not universal, but particular—layered, relational, recursive. Our language does not emerge from some internal engine of grammar. It emerges from memory. From moment. From meaning.

Language as Gestalt: Memory Before Grammar

One of the most consistent realities of GLP communication is that words are rarely just words. They are scenes. Soundscapes. Emotional snapshots. We don’t start with single units and learn to combine them. We start with wholes—scripts, lines from films, phrases heard over and over again in emotionally charged moments—and we later deconstruct them. If we do.

Sometimes we don’t. Sometimes we build entirely new expression by blending multiple gestalts. Language becomes collage.

In this way, meaning always comes before grammar. A scripted line isn’t a deficit. It’s a remembered resonance—something that meant something once, and means something again now, re-deployed in context. Syntax is not the foundation. Context is.

Language as Relational and Co-Regulated

GLPs are not linguistic islands. Our ability to access language is profoundly affected by our surroundings. Safety matters. Sensory conditions matter. Emotional attunement matters. For many of us, language doesn’t “develop” in a vacuum—it emerges in co-regulation, through repetition, pacing, and presence.

We might go silent in one space, and flourish in another. That’s not inconsistency—it’s attunement. Language may appear delayed, but it is never absent. It is often waiting—for resonance, for rhythm, for invitation. And once those arrive, we often surprise even ourselves with what we are able to express.

Language as Sensory, Rhythmic, and Poetic

GLP language is often more musical than mechanical. We speak in cadences. We hear rhythm before meaning. We attach feeling to syllables, emotion to intonation. A sentence may hold its weight not in the words it contains but in how those words feel in the mouth, or how they once felt when someone else said them.

This is why scripted language and echolalia hold such staying power. Not because they are empty. But because they fit—into the body, into the scene, into the need. Gestalt language is often poetic by nature. It loops, echoes, layers. It builds meaning not through clarity, but through accumulation. It tells the truth slant.

Emerging Frameworks: Beyond the Device

There are those beginning to notice what we’ve long lived. Some clinicians and researchers have begun documenting these patterns—describing how autistic language sometimes begins in gestalts, how echolalia can serve communicative and regulatory functions, how scripting is not mere mimicry but message. These insights, found in frameworks like Natural Language Acquisition, are welcome—but they must not become new orthodoxies.

We are not looking to replace one rigid model with another. What we need is space—to describe what we are actually doing. To narrate our own emergence. To have our language understood on its own terms.

In my own framing—as a GLP, an autistic person, as trans, as femme, a teacher and witness to others—I return often to three anchoring metaphors:

• Relational memory: Language is carried in relationship. It is tied to moments, scenes, safety.

• Recursive narrative: We don’t speak in straight lines. We loop. We return. We layer meaning through story.

• Embodied knowledge: Language is not just in the mind. It is in the body. In breath, pause, motion, and sensory texture. As a trans femme, some of the scripts that once kept me safe no longer fit—they fall flat, feel foreign, or carry a resonance I’ve since outgrown. And sometimes, it takes time—months, even years—to find the ones that speak true again, where an analytic processor might find words instantly and move on.

This isn’t a return to the so-called poverty of stimulus. It’s a rejection of stimulus as the root metaphor. GLPs are not triggered. We are tuned. And when the world is too loud, too fast, too wrong—we tune out. But the language is still there, waiting for the right key.

Summary: Language on Our Terms

GLPs teach something that can’t be diagrammed: that language isn’t always generated—it’s grown. It doesn’t always emerge through rules—it can be received. It doesn’t always begin with output—it can begin with resonance.

We are not failures of acquisition. We are not delays. We are not missing devices.

We are stewards of another linguistic tradition—one built on repetition, memory, emotion, rhythm. One that resists easy translation but never lacks meaning.

The question is no longer whether our way of speaking fits their model.

The question is: how much meaning has been missed by those who never thought to listen differently?

Final Thoughts …

Chomsky’s LAD was never meant to describe us. It was designed to explain a narrow slice of human experience, and to universalise that slice as if it were the whole. It imagined a world tidy with rules, where language emerged cleanly and on schedule. But that’s not the world I live in. That’s not the world my students live in. And it’s certainly not the world GLPs speak from.

We made language anyway. We gestured and echoed and storied our way into being. We carried scripts like life rafts until we could build boats of our own. And even now, we speak—differently, but clearly—for those willing to hear.

The LAD’s time has passed. Not because we’ve found the correct replacement, but because the lives it erased are speaking for themselves—and because the costs of its dominance are becoming impossible to ignore.

The LAD didn’t just shape academia. It gave rise to whole industries—entire diagnostic systems and pedagogical frameworks—that pretend to be neutral while punishing difference. Its logic echoes through capitalistic scams like the so-called “Science of Reading,” which promises literacy through phonics-first drilling, whilst ignoring the embodied, relational ways many of us come into language. In my practice, whenever I meet a student who’s been failed by that system, I usually find a high-masking GLP—often an undiagnosed autistic, though not always—who’s been left behind, not by their own mind, but by a model that never made room for it.

That’s why I wrote Holistic Language Instruction—to offer an alternative for native English speakers navigating literacy outside the analytic norm. And it’s why I followed it with Decolonising Language Education—to make space for non-native speakers, often multiply marginalised, whose gestalts are shaped by migration, multilingualism, and cultural memory.

I didn’t write these books to correct Chomsky. I wrote them because his theory never saw us. And now, we are writing ourselves in.

And we are many.