In Why Is There Something Rather Than Nothing? I admitted to an instinct I have never quite been able to shake: that the laws of physics are discovered rather than invented, and that mathematics might be genuinely fundamental - not a human language we lay over reality, but part of the bedrock. I said that if we ever reach base reality, maths is the thing most likely to get us there. I left it as a feeling. This post is me taking that feeling and seeing how far a serious physicist has been willing to push it.

That physicist is Max Tegmark, who sits at the top of my favourite physicists list, and the idea is his Mathematical Universe Hypothesis. It is one of the boldest claims in modern physics, and it is easy to half-hear it as a poetic flourish - “the universe is deeply mathematical, isn’t that lovely.” That is not what Tegmark means. He means it literally. He is not saying reality is described by mathematics. He is saying reality is mathematics, that there is no other ingredient, and that you and I are self-aware substructures inside a mathematical object that exists in the same timeless way the number seven does.

TL;DR

  • The Mathematical Universe Hypothesis (MUH) is the claim that our external physical reality is not merely described by mathematics but is a mathematical structure. Tegmark lays it out formally in his 2007 paper The Mathematical Universe and at book length in Our Mathematical Universe.
  • It starts from a softer, widely shared puzzle - the unreasonable effectiveness of mathematics in describing nature - and pushes it to its most extreme conclusion.
  • A consequence is the Level IV multiverse: if every consistent mathematical structure has the same claim to existence ours does, then all of them exist, and physical reality is the whole space of mathematical structures.
  • Tegmark later proposed the Computable Universe Hypothesis (CUH), restricting reality to computable structures, partly to tame the infinities and the trouble that Gödel’s incompleteness theorems cause for the strong version.
  • The idea offers a genuinely different angle on why there is something rather than nothing: mathematical structures do not need to be created, so maybe existence never had a “switching on” moment at all.
  • The strongest objections are real and I take them seriously: the leap from mathematical consistency to physical existence, the measure problem, Gödel, and the awkward question of where conscious experience fits in a universe made of pure structure.
  • I am drawn to it - more than I am comfortable admitting, given how neatly it fits what I already believe - so I am holding it at arm’s length rather than treating the fit as evidence.

A note before I start

I should say up front, as I always do with this material, that I am an interested outsider and not a physicist or a mathematician. Tegmark has spent a career building this argument with the technical machinery to back it up, and I am working through it as a fascinated reader. Treat everything below as me thinking out loud, and treat my reservations as questions rather than verdicts. This is also a topic where I am unusually aware of my own bias, because the conclusion lands close to something I already lean toward, and “it agrees with me” is the opposite of a reason to trust it.

The soft version almost everyone accepts

There is a mild form of this idea that is hard to argue with, and it is worth starting there because the strong version grows out of it.

Mathematics works on the physical world far better than it has any right to. In 1960 the physicist Eugene Wigner wrote a famous essay on the unreasonable effectiveness of mathematics in the natural sciences, pointing out that concepts dreamed up by mathematicians for their own abstract reasons keep turning out, decades later, to be exactly the tools physics needs. Complex numbers, non-Euclidean geometry, group theory - invented as pure mathematics, then found waiting at the heart of quantum mechanics, general relativity, and particle physics. Wigner called it a miracle we neither understand nor deserve.

You can stay very moderate about this and still feel the pull. Predictions in physics are made in the language of equations, and the universe keeps obeying them to absurd precision. The magnetic moment of the electron has been predicted and measured to twelve significant figures. That level of agreement is not the behaviour of a world we are merely approximating with a convenient notation. It feels like the notation is catching something that is genuinely there.

The usual move is to say: of course maths describes the world well, because we built maths to describe the world. But that does not really account for Wigner’s point. The most useful mathematics was usually not built for the job. It was built for its own internal reasons and turned out to fit anyway, often a century early. Tegmark’s hypothesis is, in essence, the most radical possible answer to why that keeps happening.

The strong version: not described by, but made of

Tegmark’s answer is to remove the gap entirely. If our universe is described so perfectly by a mathematical structure, he asks, what is the actual difference between the universe and the structure? What is the extra ingredient - the “stuff” - that the equations are supposedly painted onto?

His claim is that there is no extra ingredient. The structure is all there is. We imagine that mathematics is an abstract description and the physical world is the concrete thing being described, but Tegmark argues that distinction dissolves under inspection. A mathematical structure is a set of abstract entities and the relations between them, with no properties beyond those relations. And when you strip the physical world down to what physics can actually say about it - particles defined entirely by their properties and how they relate and interact - you are left with relations between entities that have no further intrinsic nature either. The two descriptions meet in the middle. There is, on this view, nothing left over in the physical world that is not captured by the structure, which means the physical world simply is the structure.

This is what he calls the Mathematical Universe Hypothesis, and he sets it out formally in The Mathematical Universe (2007). One way he frames it: a sufficiently advanced description of reality should be expressible without any “human baggage” at all - no words, no notation tied to our particular minds, just the bare mathematical structure. A truly complete theory of everything would not be a story about the universe. It would be a mathematical object, and we would be inside it.

The part that genuinely reorders your thinking is what this does to existence. We do not usually think the number seven was created, or that it is waiting to wink out of existence, or that it exists more in some places than others. It just timelessly is, as a feature of mathematical reality. Tegmark’s claim is that our universe has exactly that kind of existence. There was no moment when the mathematical structure that is our universe was switched on. It exists in the same atemporal sense seven does, and what we experience as the flow of time and the unfolding of history is the view from inside it.

Level IV: if one structure, why not all of them?

Here is where the hypothesis stops being a reinterpretation of our universe and starts predicting an enormous amount more.

If our universe exists because it is a consistent mathematical structure, then the obvious question is what is special about that particular structure. Tegmark’s honest answer is: nothing. There is no principled reason to grant physical existence to the structure we happen to live in while denying it to every other consistent structure. So he bites the bullet. Every mathematically consistent structure exists, in exactly the same full-blooded sense ours does, and each is a universe (or a whole class of universes) unto itself. This is the Level IV multiverse, the top of the four-level hierarchy of parallel universes he has mapped out.

It is worth pausing on how far this goes, because it dwarfs the more familiar multiverse ideas. Level I is just space beyond our cosmic horizon, more of the same kind of universe. Level II is the bubble universes of eternal inflation, where the constants of physics can vary from bubble to bubble. Level III is the many-worlds branching of quantum mechanics, which I wrote about in A New Universe All Throughout The Day. Levels I to III still share our basic mathematical framework - the same fundamental equations, different solutions or different regions. Level IV throws even that out. Different structures mean genuinely different mathematics, different laws at the deepest level, governed by equations that need have nothing in common with ours.

I will admit this is the point where the hypothesis stops feeling abstract for me, because it rhymes hard with an intuition I have had for a long time and set down in that earlier post: that reality is far larger than the single universe we find ourselves in, that there are very likely parallel universes and an unbounded run of big bangs, and that the laws of physics themselves may not be the same from one universe to the next. Tegmark’s Level IV is the most rigorous attempt I have seen to take that exact feeling and give it a foundation. That resonance is precisely why I am being careful with it.

The computable retreat, and why Gödel forces it

The strong hypothesis runs into a famous wall, and Tegmark to his credit walked straight up to it rather than around it.

The wall is Gödel’s incompleteness theorems. Gödel showed that any sufficiently rich formal system contains true statements it cannot prove, and cannot prove its own consistency from within. That is awkward for a hypothesis that wants to identify physical existence with mathematical consistency, because for most interesting structures we cannot even establish that consistency in the clean way the picture seems to need. There is also a flood of related trouble: many mathematical structures involve uncomputable quantities, real numbers that can never be calculated, infinities stacked on infinities, and it is not obvious any of that should correspond to anything physically real.

Tegmark’s response, in a 2007 follow-up with the working title of the Computable Universe Hypothesis (CUH), is to shrink the claim. Maybe physical reality is not every mathematical structure, but only the computable ones - those whose relations can be generated by a halting computation, with no uncomputable functions and no actual infinities anywhere in the description. That neatly dodges the Gödelian mess, because computable structures are exactly the well-behaved corner where the pathologies do not arise. It is a real cost, though, and Tegmark knows it: it amputates almost all of standard mathematics, including the continuum that nearly all of current physics is written in. The honest status is that nobody, Tegmark included, has shown how to rebuild working physics on a strictly computable, infinity-free foundation. It is a promissory note, and I think it should be read as one.

I find this the most telling moment in the whole programme, and oddly the most reassuring about Tegmark himself. The easy thing would have been to wave Gödel away. Instead he let the objection reshape the hypothesis, even at the price of making it smaller and harder to satisfy. That is the behaviour of someone treating the idea as a real theory to be tested rather than a creed to be defended.

What it does to “why is there something rather than nothing”

The reason I wanted to write this post is that the Mathematical Universe Hypothesis offers a distinctive line of attack on the question I find deepest, the one I circled for a whole essay without landing.

In that piece I floated the possibility that absolute nothing might not be a genuine option - that being could be the default and nothingness the thing that would actually require an impossible explanation. The Mathematical Universe Hypothesis gives that intuition a concrete mechanism. Mathematical truths do not look like the sort of thing that needs to be created. Two plus two equals four whether or not a universe exists to count anything. If physical reality is mathematical structure, then asking why there is something rather than nothing becomes a bit like asking why there are prime numbers rather than no prime numbers - the question may simply not have the bite we assumed, because mathematical existence was never the kind of thing that could have failed to obtain.

I want to be careful not to oversell this, because it is exactly the kind of argument that can feel more conclusive than it is. Saying “mathematical structures exist necessarily” is doing a lot of quiet work, and a sceptic can reasonably reply that mathematical existence and physical existence are just different things wearing the same word, and that no amount of necessity in the abstract realm puts an actual sky over our heads. That is the real fault line in the whole hypothesis, and it deserves its own section.

The objections I cannot wave away

I take this idea seriously, which means taking its problems seriously too. There are four that I keep returning to.

Consistency is not existence. This is the heavy one. The hypothesis trades on the claim that mathematical existence and physical existence are the same thing, but that is asserted more than it is demonstrated. Stephen Hawking put the worry memorably when he asked what it is that breathes fire into the equations and makes a universe for them to describe. Even if our universe corresponds perfectly to some mathematical structure, the structure on paper and the world we actually inhabit, with its fire and weather and Tuesday afternoons, still feel like they might differ in some respect that the formalism cannot see. Tegmark’s reply is that the felt difference is itself an artefact of being on the inside - that “fire” is just what some mathematical relations feel like from within - but I cannot tell whether that dissolves the problem or relabels it.

The measure problem. Once you have all consistent structures existing, you need some way to say what a typical observer should expect to see, otherwise the theory predicts nothing. But there is no obvious, non-arbitrary way to put a probability measure over an infinite space of mathematical structures. This measure problem is not unique to Level IV - it haunts inflationary multiverses too - but it bites hardest here, and without it the hypothesis struggles to make the testable predictions Tegmark wants it to make. He argues we should at least find ourselves in a structure that is mathematically simple and generic among observer-supporting ones. Whether that is a genuine prediction or a hope is, I think, still open.

Gödel, again. Even with the computable retreat, there is a lingering discomfort that a reality identified with mathematical structure inherits all of mathematics’ foundational uncertainty about what is even consistent. The fix works, but it works by shrinking the claim to the safe corner, and the safe corner may be too small to contain physics as we actually find it.

Where is the experience? A universe of pure structure is a universe with no obvious room for the felt quality of being conscious. Tegmark says self-aware substructures - us - are simply complex patterns within the mathematical object, and that our sense of an inner life is how certain self-referential structures process information about themselves. But this lands us right back at the hard problem of consciousness, and a purely structural story is exactly the kind of account that critics say leaves the felt redness of red unexplained. I do not think the Mathematical Universe Hypothesis solves consciousness. At best it relocates the mystery, and that relocation is where it brushes up against everything else I keep thinking about.

The seam where this meets consciousness

That last objection is the one I cannot leave alone, because it sits on the open question I genuinely have not settled - whether consciousness is fundamental or computational.

If reality is nothing but mathematical structure, then consciousness has to be a pattern within that structure - which is essentially the computational answer, the view that a rich enough arrangement of relations simply is an experiencing mind. The Mathematical Universe Hypothesis is, in that sense, a deeply structuralist and broadly computational picture of reality, and it has no obvious place for consciousness as a separate fundamental ingredient.

That is exactly the tension with the other thread I have been pulling on. When I wrote about Donald Hoffman, I was working through the opposite hypothesis - that consciousness is the bedrock and the physical, mathematical world is something it experiences rather than something it is made of. Hoffman and Tegmark are both trying to formalise the foundation of reality rather than just assert it, which is the quality I admire in each of them. But they are reaching in opposite directions: Tegmark puts mathematics at the floor and asks consciousness to emerge from it, while Hoffman puts consciousness at the floor and even argues that mathematics is part of the structure of mind rather than the other way around. I find I cannot fully commit to either, and the place they disagree is precisely the place I am most undecided. If consciousness really is fundamental, the strong Mathematical Universe Hypothesis is probably incomplete, because it would be missing an ingredient that does not reduce to structure. If consciousness is computational, the hypothesis has a clean home for it. I do not know which way that falls, and I have stopped pretending the not-knowing is a temporary state.

A note on the simulation question

It is worth heading off a natural confusion, because this looks superficially like the simulation hypothesis and is not the same thing at all.

A simulation needs a computer, a substrate, a base reality running the code. Tegmark’s picture has no substrate and no programmer. The mathematical structure is not being computed by anything in a deeper physical world - it just exists, statically and self-sufficiently, the way a mathematical object exists. There is no machine humming away underneath. In that respect the Mathematical Universe Hypothesis is closer to the projected picture I keep being drawn back to than to the engineered-simulation one: a reality that is self-existent rather than manufactured, with no operators standing outside it. The difference is what does the projecting. Where I have wondered whether a single underlying consciousness might be living out every possible life from the inside, Tegmark would say there is no projector and no audience - only structure, all the way down, with the feeling of being someone a pattern the structure contains. Those are very different bedrocks, and which one you prefer comes back, again, to the consciousness question I cannot close.

My own view

I land where I half-expected to, which is partway in and openly suspicious of how much I want it to be true.

What I am most confident about is the soft version. The effectiveness of mathematics in physics is real, it is strange, and “we just invented maths to fit” does not honestly account for the structures that fit a century before anyone needed them. Something about the deep regularities of the world is genuinely mathematical, not merely dressed in mathematical clothing. I have believed for a long time that the laws are discovered rather than invented, and nothing in working through Tegmark has moved me off that. If anything it has sharpened it.

The strong version - that reality is mathematics and nothing else, that the number seven and our universe exist in the same way, that every consistent structure is its own world - I hold much more loosely, and I notice I want to believe it more than the evidence licenses. It fits my standing intuition that maths is fundamental and our best route to base reality. It fits my sense that the multiverse is real and that the laws differ from one universe to the next. It even offers the cleanest version I have found of the thought that nothingness might be the truly impossible option. All of which is a reason to be more careful, not less. When a grand hypothesis flatters everything you already lean toward, that is the moment to keep your hand on your wallet. So I am filing the strong Mathematical Universe Hypothesis as the most rigorous expression I know of an instinct I trust, which is not at all the same as filing it as true.

The honest sticking point, for me, is consciousness. A universe made of pure structure does not obviously have room for the felt fact of experience, and on the days I find the hard problem most pressing - which is most days - I suspect a complete account of reality will need mind at the ground floor rather than as a pattern that emerges late. On other days I can see how a rich enough structure might simply be an experiencing thing, and then Tegmark’s picture closes up neatly. I go back and forth, genuinely, and the Mathematical Universe Hypothesis has not resolved that so much as shown me exactly what it would cost to come down on either side. If maths really is the bedrock, mind has to be made of it. If mind is the bedrock, maths is not the whole story. I would love to know which, and I am content, for now, not to.

The standing caveat, which matters more here than almost anywhere: this is my thinking as it stands today, and “today” is carrying real weight in that sentence. I revise this stuff constantly, sometimes from one week to the next, as I read further or run into an argument that pushes me somewhere new, so the version of me writing this will very likely have shifted by the time you read it. I do not find that troubling - it is the part of all this I enjoy most. A theory is only ever a theory, something to be tested and pressed on and actively tried to be proven wrong, and on the question of whether reality is made of mathematics I would be genuinely glad to be shown which way it goes, whichever way that is.

Further Watching

Our Mathematical Universe | Max Tegmark | Talks at Google

Our Mathematical Universe with Max Tegmark

Our Universe Is A Math Problem! Max Tegmark’s Theory of Reality