Almost a decade in the making, this book explains why more time, more effort, and more detail do not reliably improve certain forms of communication. Beyond a point, returns diminish sharply. In some domains, they collapse altogether.
The manuscript focuses on English, but the hypothesis has already been extended to French (published separately), and I am continuing work on other ontological barriers. If you’re interested in testing or extending the framework in your own language, feel free to get in touch.
Over the coming weeks, I’ll be unpacking aspects of the Language Insufficiency Hypothesis in more depth here. The book’s role is deliberately limited: it defines the problem, establishes the structure, and offers grounding examples. The real work happens in the consequences.
For now, the important thing is simple: the book is finally available.
I’ve been working on A Language Insufficiency Hypothesis since 2018. At least, that’s the polite, CV-friendly version. The truer account is that it’s been quietly fermenting since the late 1970s, back when I was still trapped in primary school and being instructed on how the world supposedly worked.
Social Studies. Civics. Law. The whole civic catechism. I remember being taught about reasonable persons and trial by a jury of one’s peers, and I remember how insistently these were presented as fair solutions. Fairness was not argued for. It was asserted, with the weary confidence of people who think repetition counts as justification.
I didn’t buy it. I still don’t. The difference now is that I have a hypothesis with some explanatory power instead of a vague sense that the adults were bluffing.
But that outsider position had consequences. It led me, even then, to ask an awkward question: Which peers? Not because I thought I was superior, but because I was plainly apart. How exactly was I meant to be judged by my peers when no one else occupied anything like my perspective?
Later, when I encountered the concept of fundamental attribution bias, it felt less like a revelation and more like confirmation. A peer-based system assumes not just similarity of circumstance, but similarity of interpretation. That assumption was dead on arrival.
Then there were reasonable persons. I was assured they existed. I was assured judges were trained to embody them. I had never met one. Even as a teenager, I found the idea faintly comical. Judges, I was told, were neutral, apolitical, and dispassionate. Writing this now from the United States, one hardly needs to belabour the point. But this wasn’t prescience. It was intuition. The smell test failed decades ago.
Before LIH had a name, I called these things weasel words. I still do, as a kind of shorthand. Terms like fair, reasonable, accountable, appropriate. Squishy concepts that do serious institutional work whilst remaining conveniently undefinable. Whether one wants to label them Contestables or Fluids is less important than recognising the space they occupy.
That space sits between Invariables, things you can point to without dispute, and Ineffables, where language more or less gives up. Communication isn’t binary. It isn’t ‘works’ or ‘doesn’t’. It’s a gradient. A continuous curve from near-certainty to near-failure.
Most communication models quietly assume a shared ontology. If misunderstanding occurs, the remedy is more explanation, more context, more education. What never sat right with me, even as a child, was that this only works when the disagreement is superficial. The breaking point is ontological.
If one person believes a term means {A, B, C} and another believes it means {B, C, D}, the overlap creates a dangerous illusion of agreement. The disagreement hides in the margins. A and D don’t merely differ. They are often irreconcilable.
Image: Venn diagramme of a contested concept. Note: This is illustrative and not to scale
Fairness is a reliable example. One person believes fairness demands punishment, including retributive measures. Another believes fairness permits restoration but rejects retribution, citing circumstance, history, or harm minimisation. Both invoke fairness sincerely. The shared language conceals the conflict.
When such disputes reach court, they are not resolved by semantic reconciliation. They are resolved by authority. Power steps in where meaning cannot. This is just one illustration. There are many.
I thought it worth sharing how LIH came about, if only to dispel the notion that it’s a fashionable response to contemporary politics. It isn’t. It’s the slow crystallisation of a long-standing intuition: that many of our most cherished concepts don’t fail because we misuse them, but because they were never capable of doing the work we assigned to them.
(or: Why Neither Humans nor AI Create from Nothing)
In the endless squabble over whether AI can be “creative” or “intelligent,” we always end up back at the same semantic swamp. At the risk of poking the bear, I have formulated a response. Creativity is either whatever humans do, or whatever humans do that AI can’t. Intelligence is either the general ability to solve problems or a mysterious inner light that glows only in Homo sapiens. The definitions shift like sand under the feet of the argument.
Audio: NotebookLM podcast on this topic
Strip away the romance, and the truth is far less flattering: neither humans nor AI conjure from the void. Creativity is recombination, the reconfiguration of existing material into something unfamiliar. Intelligence is the ability to navigate problems using whatever tools and heuristics one has to hand.
I cannot be causa sui, and neither can you.
The Causa Sui conceit, the idea that one can be the cause of oneself, is incoherent in art, thought, or physics. Conservation of energy applies as much to ideas as to atoms.
We metabolise them through cognitive habits, biases, and linguistic forms.
We output something rearranged, reframed, sometimes stripped to abstraction.
The AI process is identical in structure, if not in substrate: ingest vast data, run it through a model, output recombination. The difference is that AI doesn’t pretend otherwise.
When a human produces something impressive, we call it creative without inspecting the provenance of the ideas. When an AI produces something impressive, we immediately trace the lineage of its inputs, as if the human mind weren’t doing the same. This is not epistemic rigour, it’s tribal boundary enforcement.
The real objection to AI is not that it fails the test of creativity or intelligence; it’s that it passes the functional test without being part of the club. Our stories about human exceptionalism require a clear line between “us” and “it,” even if we have to draw that line through semantic fog.
My Language Insufficiency Hypothesis began with the recognition that language cannot fully capture the reality it describes. Here, the insufficiency is deliberate; the words “creativity” and “intelligence” are kept vague so they can always be shifted away from anything AI achieves.
I cannot be causa sui, and neither can you. The only difference is that I’m willing to admit it.
Philosophics 