The Persuasion Engine, Sixty-Four Years Later

Kuhn's book has a strange fate for something so precise: it became famous for a word — paradigm — that now means almost nothing, deployed by management consultants and TED speakers to describe any change of mind whatsoever. This is not incidental damage. It is the book's own thesis eating its tail. Kuhn spent the postscript, written just seven years after publication, trying to claw back "paradigm" from readers who'd already flattened it into a synonym for worldview. He failed. The disciplinary matrix he proposed as a replacement never caught on; the loose word did. Reading this in 2026, after "paradigm shift" has been used to sell everything from cryptocurrency to protein powder, you understand that Kuhn diagnosed something about how ideas travel that his own vocabulary then demonstrated in real time — a term escapes its technical cage the moment it's useful for something other than the truth it was coined to protect.

What reads differently now is the chapter on the invisibility of revolutions — the argument that textbooks systematically disguise scientific rupture, presenting cumulative brick-laying where there was actually rupture and reconstruction. In 1962 this was a claim about pedagogy. In 2026, with science communicated in real time through preprints, Twitter threads, and press conferences held mid-crisis, the textbook's slow, retrospective smoothing has been replaced by something faster and more exposed: the public now watches normal science fail to resolve anomalies in real time — mask guidance reversing, dietary consensus flipping, replication crises airing in public — without the decades of quiet textbook revision that used to metabolize the mess into a clean narrative afterward. Kuhn assumed a lag between crisis and its historical laundering. That lag is gone. We watch the crisis and the persuasion campaign simultaneously, which means we watch the "complex judgment comparing old and new frameworks" happen as spectacle rather than settled history — and it turns out publics do not process an open crisis the way Kuhn's insulated professional community does. They don't wait for a better paradigm before abandoning the old one; they often just abandon paradigm-thinking altogether.

This is where the book's most defensible move — insisting that incommensurability doesn't mean irrationality, that theory-choice still runs on shared values like accuracy and scope — became, against Kuhn's explicit and repeated protest, ammunition for people arguing the opposite. The postscript's careful non-relativism, the argument that science is "unidirectional and irreversible" in problem-solving power even without approaching Truth-with-a-capital-T, was precisely engineered to block the reading that any two paradigms are equally valid social constructions. And yet that is exactly the reading that escaped into the wider culture and now furnishes climate denial and anti-vaccine rhetoric with a borrowed vocabulary: consensus is just persuasion, paradigms are just what a community happens to believe, so who's to say. Kuhn built the fence and watched, or would have watched, the fence get repurposed as a gate. The chapter on progress — arguing science advances because its communities are insulated from external social pressure — reads almost tender now, a description of a condition (professional insulation, unbothered puzzle-solving) that funding politics, social media, and platform-mediated publics have mostly dissolved.

What it anticipated better than it gets credit for: the trading-zone problem, the way genuinely different specialties develop mutually opaque vocabularies and must construct a pidgin to collaborate at all. This describes, almost exactly, the current relationship between machine learning researchers, neuroscientists, and philosophers of mind trying to discuss whether a language model does anything resembling cognition — three communities using the same words (representation, understanding, learning) to mean structurally different things, each convinced the others are equivocating. Kuhn thought this kind of incommensurability was a mark of healthy specialization, akin to speciation. It's less clear it's healthy when the disagreement concerns whether we've built something with interests. Which leaves the real question the book raises now, unintentionally: Kuhn's whole apparatus — crisis, anomaly, professional insecurity, conversion — depends on a community of people who can feel embarrassed, threatened, persuaded. If normal science increasingly means pattern-matching systems generating and testing hypotheses without ever experiencing an anomaly as a personal or professional crisis, does the revolution still happen, or does puzzle-solving just industrialize forever, anomaly-free?