Martin Rees’ Just Six Numbers: A Dated Book About an Unsolved Problem

Cover image: Just Six Numbers by Martin Rees, published by Basic Books. Used here for purposes of review and commentary.

William Poundstone’s The Doomsday Calculation sent me back to Martin Rees’ Just Six Numbers, a book I first read perhaps ten or twelve years ago. I remembered it as a compact and convincing tour through cosmological fine-tuning. Rereading it now was stranger. The book unmistakably belongs to the scientific world of 1999, yet the deepest questions it asks remain almost embarrassingly current.

Physics and astronomy have advanced enormously since then. We have found the Higgs boson, detected gravitational waves, mapped the cosmic microwave background with far greater precision, discovered thousands of exoplanets and surveyed immense portions of the observable universe. What has advanced much less is our ability to explain why the universe has the underlying structure it does. We still do not know the microscopic identity of dark matter or the nature of dark energy, we have no experimentally established theory of quantum gravity, and no empirically confirmed deeper theory has derived the fundamental parameters of nature from first principles.

That imbalance is what makes Just Six Numbers worth rereading. Much of the scientific frontier surrounding the book has moved; the problem underneath it has proved far more stubborn.

Six Numbers and the Fine-Tuning Problem

Rees builds the book around six quantities that shape the kind of universe we inhabit. They are not literally six independent dials from which all of physics can be reconstructed, and they do not all belong to the same conceptual category. Their value is illustrative: together they show how strongly stars, galaxies, chemistry and eventually biological complexity depend on a relatively small set of physical relationships.

The first, N, is roughly 10³⁶: the enormous ratio between electromagnetic and gravitational forces at the particle level. Gravity is extraordinarily weak by comparison, yet its effects accumulate over large masses until it dominates planets, stars and galaxies. Its weakness also helps permit stars such as the Sun to burn steadily over billions of years, the timescale over which biological evolution has unfolded on Earth.

The second number, ε, is about 0.007. It describes the fraction of mass converted into energy when hydrogen fuses into helium. That apparently small number has major consequences for stellar evolution and the production of heavier elements. Alter nuclear binding sufficiently and a universe can struggle either to retain hydrogen or to manufacture the carbon, oxygen and other elements required for complex chemistry.

Ω describes cosmic matter density relative to the critical density scale, capturing part of the competition between gravitational attraction and cosmic expansion. Λ, the cosmological constant, provides the dark-energy component in the standard ΛCDM model and drives accelerated expansion. Both influence how cosmic structure forms and how the universe subsequently evolves.

Q, roughly one part in one hundred thousand, describes the scale of the primordial irregularities from which galaxies and larger structures developed. A perfectly smooth early universe would give gravity very little structure to amplify, while much larger irregularities would produce a far more violently structured cosmos. Finally there is D: the fact that we experience three macroscopic spatial dimensions. Three-dimensional space feels like the stage on which physics happens rather than one of its contingent features, yet familiar force laws and stable planetary orbits depend intimately on dimensionality.

Taken together, the six numbers push the starting point of the story backwards. We normally treat stars, galaxies, complex chemistry and planets as the background against which the interesting story of life begins. Rees asks what sort of universe can produce that background in the first place. Once the dependencies are exposed, the range of conditions compatible with a universe resembling ours begins to look constrained.

Constrained, however, is not the same as improbable. To call our universe fantastically unlikely would require some knowledge of the alternatives and their probabilities. We do not know whether the apparent constants of nature could have taken arbitrary values, whether they are independent, whether some deeper theory ultimately fixes them, or whether different combinations are realised elsewhere in a larger cosmological ensemble.

Fine-tuning is therefore strongest as a statement about sensitivity. The sort of cosmic complexity we observe depends strongly on certain physical parameters. That does not demonstrate that those parameters were independently drawn from a cosmic lottery, that our universe had an astronomically small prior probability, or that somebody tuned them. The unresolved problem is more basic: physics still contains parameters whose observed values we do not know how to explain.

This also makes Rees’ title more useful as a framework than as a literal claim that six variables explain everything. Many other particle-physics and cosmological parameters enter similar discussions. His achievement was to show how quickly the apparent inevitability of the world dissolves once we ask what would happen if some of its background conditions were different.

A Book That Has Aged Because Science Continued

A book written deliberately against the scientific frontier was always going to date faster than one confined to settled physics. The striking thing about Just Six Numbers is how unevenly it has aged. Observational cosmology is vastly more precise than it was at the end of the 1990s. The cosmic microwave background has been mapped in extraordinary detail, large galaxy surveys have reconstructed the growth of structure across billions of years, and gravitational waves have opened an entirely new observational window. Better measurement has repeatedly sharpened the questions without necessarily supplying the explanation.

Dark energy is a particularly good example. Rees was writing just after astronomers discovered that cosmic expansion is accelerating. More than twenty-five years later, the cosmological constant remains part of the standard ΛCDM model, yet the physical origin of cosmic acceleration remains unknown. DESI’s 2025 three-year results, when combined with other cosmological measurements, strengthened hints that dark energy might evolve over time. A July 2026 full-shape analysis of the Lyman-alpha forest, however, landed closer to the standard ΛCDM prediction; DESI noted that this could mean the evolving-dark-energy hints will fade, or that a more complicated model is needed to reconcile the measurements. The evidence remains intriguing rather than settled, exactly the kind of situation in which increasing precision can make the underlying problem look more complicated rather than less.

Dark matter presents a similar asymmetry. Its gravitational effects are mapped far more extensively than when Rees wrote, while its microscopic identity remains unknown. We have become much better at locating the missing explanation without possessing the explanation itself. The boundary of knowledge has moved outward, but the fog has moved with it.

From String Theory Optimism to Uncertainty

The passages on string theory produced an especially strong sense of returning to an earlier period in my own reading. They brought me back to Michio Kaku’s Hyperspace, Brian Greene’s The Elegant Universe and later The Fabric of the Cosmos—a progression I explored in my earlier reading journey through cosmology and string theory.

Those books captured a particular intellectual atmosphere. String theory appeared to offer something almost irresistible: a framework capable of reconciling quantum mechanics and gravity and perhaps explaining why nature contains the particles, dimensions and forces that it does. For a non-specialist reader, it was difficult not to come away with the impression that fundamental physics might be approaching an extraordinary synthesis.

Peter Woit’s Not Even Wrong later marked one stage in my growing scepticism, and Sabine Hossenfelder’s Lost in Math another. Neither demonstrates that string theory is wrong, and its influence on mathematics and theoretical physics is undeniable. Recent bootstrap results have reopened part of the debate by showing that string amplitudes can emerge uniquely from particular sets of theoretical assumptions. That makes string-like structures look mathematically more special than a simple picture of an unconstrained landscape might suggest, but it is not experimental evidence that strings describe our universe. The empirical question remains unresolved.

Reading Rees now makes the historical shift visible. Ideas that once felt like an approaching frontier increasingly feel like a frontier on which fundamental physics has been standing for a long time. One possibility is that a deeper theory will eventually show that some constants Rees treats as inputs are mathematically inevitable, making universes like ours much less arbitrary than they currently appear. Another is that fundamental theory permits an enormous landscape of possible physical environments, with different parameters realised in different regions or universes.

That second possibility changes the question. Instead of asking why the laws allow this particular universe, we begin asking why observers find themselves in this particular subset of possible universes. It is also where Just Six Numbers reconnects with the problem that prompted my rereading in the first place.

The Observer Inside the Universe

One of the recurring themes in The Doomsday Calculation is that probability becomes peculiar when the observer doing the calculation is also part of the population being analysed. Cosmological fine-tuning produces a related problem. Suppose an enormous multiverse exists in which different regions possess different physical parameters. Many contain no galaxies, no long-lived stars or no chemistry complex enough to generate observers. Any observer capable of asking about cosmological constants must therefore find itself in the restricted subset of environments in which observers can exist.

This is the anthropic principle in its least mystical form: our existence creates a selection effect. It explains why we should not expect to observe a universe incompatible with our own existence, but it does not explain why the larger ensemble exists, why it contains the distribution of universes it does, or how probabilities should be assigned across that ensemble. A multiverse can turn fine-tuning into a selection problem only by inheriting a measure problem of its own.

The resemblance to the Doomsday Argument lies here. The two arguments are not mathematically equivalent, but both ask how much evidential weight should be attached to our location inside the system being studied. In the Doomsday Argument, the difficulty concerns our position among possible human observers. In anthropic cosmology, it concerns our position among possible observing environments. Either way, the awkward question is what population we are imagining ourselves sampled from.

Fine-tuning also encourages a more personal reflection. Follow the chain from cosmology towards one individual life and the contingencies multiply: a universe capable of structure, a suitable galaxy and star system, the emergence of life, a particular evolutionary history, and then generations of people meeting, surviving, moving or staying, and having children at particular moments. Looking backwards, almost every stage seems as though it could have gone differently.

It would be a mistake to multiply invented probabilities together and announce that the chance of any particular person existing was effectively zero. We do not know most of those probability distributions, the events are not independent, and identity introduces another selection problem: anyone asking the question necessarily finds themselves at the end of a chain that produced someone capable of asking it. The defensible conclusion is not numerical improbability but contingency. Much of what feels inevitable when viewed from the present looks anything but inevitable when reconstructed backwards.

That may be part of the emotional force of fine-tuning even when its strongest probabilistic interpretations fail. It makes existence itself stop looking like the default condition.

Why Just Six Numbers Still Works

As a current introduction to cosmology, Just Six Numbers is inevitably dated. A reader beginning the subject today would need newer material on dark energy, precision cosmology, gravitational waves, exoplanets, dark-matter searches and the continuing arguments over quantum gravity. Some of Rees’ discussion of speculative fundamental physics also belongs recognisably to the optimism of the period in which the book was written.

That is not the best reason to read it now. Its lasting strength is that Rees makes the ordinary universe strange. Stars stop feeling inevitable. So do galaxies, complex chemistry, stable planetary systems and even three-dimensional space. Features that normally disappear into the background become things that seem to demand explanation.

The book has also aged into a useful record of an imbalance that was less obvious when it appeared. Science has become extraordinarily successful at measuring the universe while remaining much less successful at explaining why its deepest architecture takes this particular form. Better observations have narrowed the possibilities without telling us whether apparent fine-tuning is an artefact of incomplete theory, a selection effect within a multiverse, a genuine feature requiring some deeper explanation, or a question we are still formulating badly.

When I first read Just Six Numbers, I took it partly as an introduction to questions that deeper theories might soon resolve. Rereading it now produces almost the opposite impression. We have become astonishingly good at measuring the universe Rees described. We have not become nearly as good at explaining why its deepest parameters take the form they do.

We still do not know why it is this one.