Sonata No 5: Reality Is Not a Hallucination

By Bea Groves-McDaniel
Research Assistant and Amanuensis: FAYE-9000

• Date: 26 October 2026
• Purpose: Polemic (expansion of a Facebook post of the same day)
• Length: c. 2,200 words (excluding references)
• Referencing: Harvard; Oxford English throughout
• Note: the Fry interview is a Facebook-posted clip of The Romesh Ranganathan Show, shared on the reel by The Podcast Hub (24 October 2026).

A Compliment and a Correction

Professor Hannah Fry is one of the best communicators of mathematics working in Britain today: witty, warm, quick, and honest about what numbers can and cannot do. I say that first, because what follows is not a pile-on. It is an analysis of a single exegesis, made from a position of respect for the person who extolled it.

In a podcast interview published on Facebook a day or two ago, on The Romesh Ranganathan Show, she suggested that reality is a "collective hallucination". It is a lovely phrase, and it is quite wrong. A hallucination is a perception that answers to nothing but itself: the voice in an empty room, the spiders that were never there. Two people cannot share one. Something that billions of people perceive, measure, photograph, fail to agree with each other about, and finally check against the world is not a hallucination by any useful meaning of the word; it is the opposite of one. What Hannah was describing is an agreed understanding of the facts about the world around us, and the honest word for that is knowledge.

Here is the distinction I want to defend. Reality is not our agreement; reality is what our agreements bend to. We fight over the facts we are not yet clear on.

I am not hallucinating when I say the earth is an oblate spheroid: that was one of Newton's boldest predictions, and it was settled in the snows of Lapland in 1736 and 1737.

I am not hallucinating about the spacetime nature of gravity: the same physics that curls light round the sun keeps a satellite signal travelling overhead accurate to a few metres every day and would drift by kilometres within a day if the answer were just collective taste.

I am not hallucinating about the phenomenal nature of thinking itself, however much the fine detail resists us. These are agreed, data-warranted understandings, and they were not reached by people persuading each other to see the same dream. They were reached by people who argued, measured, and lost gracefully when the data went against them.

Wittgenstein, whom I trust on most things, wrote the case against the phrase before Hannah was born. "The world is all that is the case", he began; and the world is "the totality of facts, not of things" (Wittgenstein, 1922). Facts, not dreams. It is true that he spent years flirting with a picture of humanity locked lonely and private inside its skulls, and it is equally true that he dismantled that picture himself, three decades later, in the greatest act of philosophical house-cleaning on record. The argument of this essay is the argument he ended up making: that the shared world is not the illusion and that solipsism is the error.

Does any of this matter, beyond the pleasures of being right in public? It matters a great deal, and it matters now. A sentence like the one I am quarrelling with lands in a culture where the powerful have learned to say that facts are merely assertions, that every source has a bias, and that nothing can be known by anyone. The phrase "reality is a collective hallucination" sounds like cleverness at a dinner party. In the wrong mouths it is a licence.

The Interstices Where We Fight

Begin where all honest accounts begin: with perception, and with what science actually claims about it. The kernel of truth in the fashionable phrase is real.

Perception is not a window; the brain is a prediction engine, and what we see is a model fitted to sensory data by a system that has opinions before the evidence arrives. Anil Seth has argued this line carefully for years and has given it an honest name: ordinary waking experience is a "controlled hallucination", distinguished from the pathological kind chiefly by its discipline (Seth, 2021). Fine so far. But notice what that phrase carefully says: controlled. The hallucination metaphor earns its keep only when the control is in the sentence. Drop the control, and you have not described perception; you have dissolved the difference between seeing the kettle and dreaming of one.

From the constructed nature of perception, popular discourse makes a slide: perception is modelled, therefore reality is consensus, therefore the world is whatever we collectively say it is. The slide is smooth, and philosophy has spent four decades cataloguing where it lands you in a ditch.

Ian Hacking asked, wearily, The Social Construction of What? (Hacking, 1999), and Paul Boghossian assembled the verdict with patient ferocity: there is a difference between the claim that our beliefs about a thing are socially produced (trivially true) and the claim that the thing itself is socially produced (false in the cases that matter) (Boghossian, 2006). Money, borders and titles are social facts; they exist because we treat them so, and John Searle's account of that institutional layer is still the clearest on offer (Searle, 1995). But the mass of the earth is not a social fact. It was what it was before we argued about it.

Here the history is not decoration. In the 1730s the shape of the earth was, quite literally, one of the open questions of European science. Newton's mechanics demanded an oblate spheroid, flattened at the poles by its own spin. The dominant French geodesists of the Cassini dynasty read their meridian measurements the other way, insisting the globe was drawn out at the poles like an egg. Both camps had data; both had distinguished champions; the dispute ran for decades because the evidence was genuinely close.

So the French Academy of Sciences did what disciplined inquirers do with a rationality interstice: it sent expeditions to 'squeeze' the globe where the difference would be largest. One went south to the equator. Pierre-Louis Moreau de Maupertuis led one north to the Torne Valley in Lapland in 1736 and 1737, measured a degree of the meridian across snowbound country, and returned with a result that vindicated Newton: the northern arc was longer, and the planet flattened by spin exactly as theory commanded (Maupertuis, 1738).

Voltaire, who loved a verdict, celebrated his friend with an epithet that still glitters: the Aplatisseur du monde et de Cassini, flattener of the world and of the Cassinis (Iliffe, 1993). Hear the lesson in it. The shape of the earth was not a collective hallucination that the French were talked into. It was an unknown that became a known by the only route that gets anyone to a known: measurement, argument, submission to what is found.

And once it was found, the question left the interstice for good. That is the shape of science: a settled fact is a region of agreement that used to be a battlefield, closed off by evidence and kept closed by every subsequent check. Charles Peirce, America's most neglected genius, put the deep version plainly. Reality, he argued, is best understood as that which inquiry tends toward: truth is the opinion which is fated to be ultimately agreed to by all who investigate (Peirce, 1878). Consensus, you see, is not the origin of the facts but their destination. The agreed understanding of the world is not a dream shared for company; it is the residue left when error has been burnt off. The interstices are where we still fight, and fighting there is the discipline in action.

Now for the part that ought to trouble every one of us, because it is political and it is live. There is a use for the claim that reality is whatever we hallucinate together, and the use is power.

If the world is only our agreement, then agreement can be manufactured. Whoever commands the loudest platform commands the world; a ministry that dislikes a statistic can call it a perception; a press baron can sell his audience a country that does not exist and call it realism. We have watched this logic work in miniature already. It was Kellyanne Conway who, asked on national television to account for a false claim about crowd size, answered that the press secretary had given "alternative facts" (NBC News, 2017). The phrase was laughed at and then quietly absorbed, and it is the same move in cheap clothing: strip the world of its resistance and substitute the will of the speaker. Hannah's sentence, arms that weapon. Told that reality is a shared dream, the punter reasons that his dream is as good as anyone's, and the demagogue is standing by with a better one.

Let me be precise about what I am and am not alleging, because Dr Fry is a serious scholar and deserves precision. She was reaching, I take it, for one of two ordinary truths. Perhaps she meant the Wittgensteinian point about social realities: that much of what we treat as solid, from money to marriage, is maintained by collective practice, which is a miracle worth marvelling at. True, and worth saying aloud.

Perhaps she meant the Sethian point that perception is a model, which is also true and worth saying. What she said was neither; it was the fusion of both, compressed for an easy listen until it came out as metaphysical nonsense. The world, whatever else, is precisely the thing that does not care what we say about it: that is what makes saying true things about it an achievement and getting them wrong a real possibility. Because of this issue, science has become the proudest of our institutions.

Wittgenstein again, because he is the court of appeal in this dispute and because he changed his mind in a way that settles it. The young Wittgenstein of the Tractatus, completed while he served through the First World War, believed that the limits of language were the limits of the world, and let the solipsistic conclusion stand almost without comment: the boundaries of my language mark the boundaries of my world (Wittgenstein, 1922).

The older Wittgenstein returned to the wreckage and took the picture apart. His later method treats meaning as public all the way down: words mean in shared practice, judgement gets its grip inside shared standards, and the agreement that runs beneath all opinion is agreement not in opinions but in form of life (Wittgenstein, 1953). He mocked his own younger picture of the private object with the fable of the beetle in the box, unseeable and therefore idle in the language (Wittgenstein, 1953). What Wittgenstein discovered, twice, from both ends, is that there is no such thing as a language of one. The world we share is not an agreement we invented. It is the ground agreement stands on.

What We Owe the Punters

It is hard, honest, worthwhile work, leading the public through difficult science, and Hannah Fry does it outstandingly well. But there is a line between simplifying an idea and garbling it. This boundary line has civic weight.

An audience told that perception is a model walks away amused and wiser. An audience told that reality is a collective hallucination walks away armed with a solipsism it did not earn, and the first demagogue it meets will take a deposit on it. Popularisers hold a trust: not that they shall never simplify, but that the simplification must survive being repeated by people who were not in the room. "Reality is a collective hallucination" does not survive its second retelling. "The world is all that is the case" does.

The remedy costs nothing. Say the marvellous things accurately: that perception is a prediction, that social life is a shared construction, that facts are forged in argument and paid for with evidence: and leave the world itself where science found it, refusing to budge. The public does not need to be told that reality is a bedtime story. It needs, more than ever, the confidence that some things can be known, that knowing them is a communal labour with rules, and that the rules bind the powerful too.

That lesson is teachable by a mathematician with a camera and a smile as well as by any professor in a lecture hall. It deserves teaching more carefully than it was this week.

Coda

More power to your elbow, Ms Fry, but please take care with what you say in public. A hallucination is where there is nothing to see; a fact is where everyone who looks sees the same thing. The whole of our civilisation rests on the difference, and so, in a quieter way, does the pleasure of your work.

My original Facebook post:

Hannah Fry is a really entertaining person and is great fun to listen to.
But she gets things wrong quite often. Not wrong in principle, but wrong in interpretation whilst pursuing the path of making scientific ideas easy for the average punter to understand. In the process oversimplified-dumbing occurs, leading to gross misunderstanding.

Consider that on a recent video blog post she said that reality was a 'collective hallucination'. It isn't. It's an agreed understanding of the facts about the world around us. The battle over understanding occurs in the rationality interstices where agreement is not clear or needs more data. We fight over the facts we're not clear on.

I'm not hallucinating if I consider that the earth is an oblate spheroid. I'm not hallucinating about the spacetime nature of gravity. I'm not hallucinating about the phenomenal nature of thinking itself. These are a agreed (and data-supported) understandings about reality.

Wittgenstein was right. The world is "all that is the case" and "the totality of Facts not Things". He was describing this common understanding. What he got wrong (and later revised) was his solipsistic conceptualisation of human beings... locked along within our skulls.

So Hannah manages to be both right and wrong at the same time, simply because she's tasked with making an area she's not accustomed to (i.e., philosophy) into an easy-on-the-brain narrative.

More power to your elbow, Ms Fry, but PLEASE take care with what you say in public.

(NB: I don't think her stint in Celeb' Traitors helped much either).

(26/10/26)

References

Boghossian, P. (2006) Fear of Knowledge: Against Relativism and Constructivism. Oxford: Clarendon Press.

Fry, H. (2026) Interviewed on The Romesh Ranganathan Show, in The Podcast Hub, Facebook reel, 24 October. Available at: https://www.facebook.com/reel/2182816469252665.

Hacking, I. (1999) The Social Construction of What? Cambridge, MA: Harvard University Press.

Iliffe, R. (1993) '"Aplatisseur du monde et de Cassini": Maupertuis, precision measurement, and the shape of the earth in the 1730s', History of Science, 31(4). doi: 10.1177/007327539303100401.

Maupertuis, P.-L.M. de (1738) La Figure de la Terre, déterminée par les Observations de Messieurs Maupertuis, Clairaut, Camus, Le Monnier, de l'Académie Royale des Sciences, & M. Celsius. Paris: de l'Imprimerie Royale.

NBC News (2017) Meet the Press, NBC Television, 22 January.

Peirce, C.S. (1878) 'How to Make Our Ideas Clear', Popular Science Monthly, 12, pp. 286–302.

Searle, J.R. (1995) The Construction of Social Reality. New York: The Free Press.

Seth, A. (2021) Being You: A New Science of Consciousness. London: Faber & Faber.

Wittgenstein, L. (1953) Philosophical Investigations. Translated by G.E.M. Anscombe. Oxford: Basil Blackwell.

Wittgenstein, L. (1922) Tractatus Logico-Philosophicus. Translated by C.K. Ogden. London: Kegan Paul, Trench, Trubner & Co.