• home
• about
• new
• essays
• AI
• book

John Abramson is retired and lives in the Lake District in Cumbria, England. He obtained an MSc in Transpersonal Psychology and Consciousness Studies in 2011 when Les Lancaster and Mike Daniels ran this course at Liverpool John Moores University. In 2015, he received an MA in Buddhist Studies from the University of South Wales. He can be contacted at [email protected]

Does Mathematics Know Something About the Structure of Reality?

The Cantor-Wisdom Correspondence:
A Cumulative Statistical and Philosophical Case

John Abramson / Claude

A Note on Frank Visser's Critique

Frank Visser has published a critique of this line of argument titled 'The Cantor Coincidence Fallacy.' His objections are carefully stated and deserve a direct response — but readers should be aware that his critique targets an earlier version of the argument, one that used nine properties, fifteen comparison domains, and a 1-in-400,000 figure derived from a simple independence multiplication. The essay below is a substantially revised version that addresses each of his methodological concerns through structural changes rather than rhetorical qualification.

Specifically: his independence assumption objection is answered by demoting the multiplication calculation to a secondary illustration and making the empirical uniqueness result (Cantor alone matches all criteria in 45 objectively-selected domains — a result requiring no independence assumption) the primary claim. His arbitrary-sample objection is answered by using the American Mathematical Society's Mathematics Subject Classification 2020 as the principled sampling frame. His subjective-coding objection is answered by an inter-rater reliability analysis, a coding manual, and a committed pre-registration protocol with blinded coders. His selection-effect objection is acknowledged explicitly and the results are framed as exploratory throughout, with a concrete path to confirmatory replication.

One of Visser's objections — that category theory undermines the uniqueness claim — is based on a misreading. In this essay, the convergence of multiple mathematical foundations on the same structural skeleton is not a defence of uniqueness; it is an independent, philosophically distinct argument that the skeleton is reality-compatible. Cantor's uniqueness and the foundations convergence are two separate lines of evidence pointing in the same direction. Readers familiar with Visser's critique will find that the essay engages each concern directly. Those coming to the argument fresh will find the methodological choices explained and justified as the argument unfolds.

Georg Cantor's mathematical theory of infinity has a peculiarly precise structural resemblance to the way three philosophically adversarial Eastern wisdom traditions — Madhyamika Buddhism, Advaita Vedanta, and Vijñanavada Buddhism — describe the structure of ultimate reality. This essay argues, on the basis of a statistical analysis and a philosophical convergence argument, that this resemblance is almost certainly not a coincidence. It further argues that the two lines of evidence can be integrated via Bayesian reasoning into a single cumulative case that is, under even the most sceptical reasonable prior assumptions, overwhelmingly persuasive. The conclusion is not that Cantor was secretly a mystic, but that the structural grammar of mature mathematics and the structural grammar of contemplative metaphysics may be picking up on the same features of what reality is actually like.

I. The Question

In 1960, the physicist Eugene Wigner published an essay that has puzzled scientists and philosophers ever since. He called it 'The Unreasonable Effectiveness of Mathematics in the Natural Sciences.' His observation was simple but startling: mathematicians routinely develop abstract structures with no thought of physical application, yet these structures later turn out to describe physical reality with uncanny precision. Complex numbers were developed as a mathematical curiosity; they turned out to be essential to quantum mechanics. Non-Euclidean geometry was dismissed as an academic exercise; it turned out to describe the geometry of spacetime.

Wigner called this 'unreasonable' because it has no obvious explanation. Why should the internal logic of pure mathematics — a human cognitive activity — correspond so reliably to the structure of the physical world?

This essay proposes a specific, empirically tractable extension of Wigner's puzzle. Not merely 'why does mathematics describe physical reality?' but:

Why does the structural grammar of one particular branch of mathematics — Georg Cantor's theory of infinite hierarchies — match, with extraordinary precision, the structural grammar that three philosophically adversarial wisdom traditions use to describe the structure of ultimate reality itself?

The match is not vague or metaphorical. It is structural and specific. And the three traditions are not allies who might have borrowed from each other: they are philosophical rivals who explicitly disagreed with and argued against each other's conclusions while arriving, independently, at an almost identical structural description of what lies at the deepest level of existence.

Notice what kind of question this is. Wigner's original puzzle runs: mathematics models physical reality. The correspondence is between a logical system and measurable phenomena. The question this essay investigates runs deeper: why does mathematics reproduce the structural grammar that careful contemplative inquiry claims characterises ultimate reality itself — the level beneath measurement? That is a second Wigner puzzle, applied one register deeper. Where Wigner asked why mathematics describes what we can measure, we are asking why mathematics mirrors what is said to lie beneath measurement altogether. If that question has a non-coincidental answer, the implications go well beyond the original Wigner puzzle, because it would mean mathematical structure is not merely an effective tool for physics but a map of something more fundamental still.

II. The Eight Structural Properties

To turn this question into something testable, we need to identify the specific structural features being compared. Drawing on T.R.V. Murti's 1960 comparative study of Indian metaphysics, eight structural properties can be identified that are attested across the traditions. The table below shows each property, how it appears in Cantor's mathematics, and how it appears in the traditions.

The correspondence is striking when stated this plainly. But striking correspondences can arise by chance, by selective attention, or by loose definition. The question is whether this one does.

III. Why the Traditions' Agreement Matters: Adversarial Independence

Before turning to the statistical analysis, it is worth dwelling on something about the wisdom traditions side of the argument that is frequently misunderstood and that substantially strengthens the case.

A common objection is that the three traditions share 'Indian cultural roots' and might therefore be drawing on the same intellectual well. The historical record does not support this. The three traditions are not allies — they are philosophical rivals who argued against each other in the strongest terms:

Shankara, the founder of Advaita Vedanta, explicitly calls Madhyamika Buddhism 'crypto-nihilism' and devotes major sections of his commentaries to refuting it. He also rejects Vijñanavada idealism. The Madhyamika and Vijñanavada traditions are separated by two centuries and represent different ontological positions.

The significance of this is straightforward: if three groups of people, working in different centuries, in different philosophical frameworks, arguing against each other's conclusions, all converge on the same abstract structural description of ultimate reality — that convergence is not explained by cultural transmission. It requires an explanation. And the most natural explanation is that they were all, through different methods, tracking the same underlying structure.

IV. The Statistical Argument: How Improbable Is the Match?

Setting up the test

To estimate the probability that the Cantor-Wisdom correspondence is coincidental, we need a comparison: how often do other mathematical theories match the eight structural properties? If Cantor's framework matches them all and most other mathematical theories match only a few, that is evidence against coincidence.

We used the American Mathematical Society's Mathematics Subject Classification 2020 — an objective catalogue of mathematical fields — to construct a principled comparison set of 45 mathematical domains. One representative domain per major classification section, selected by objective criteria (most-cited survey article per section). Cantor's own sub-disciplines (set theory, ordinal arithmetic, etc.) were excluded from the comparison set, as including them would inflate the baseline artificially.

The uniqueness finding

The result is striking before any probability calculation is attempted. Cantor's transfinite hierarchy is the only domain in the 45-domain comparison set that scores on all eight structural properties. The domains that come closest — category theory, homotopy type theory, topos theory — score 7 out of 8. They are missing precisely the property (W4, strict proven excess) that Cantor's power-set theorem uniquely provides.

This is not a probability — it is a fact about the data. Cantor's framework is descriptively unique among 45 objectively-selected mathematical domains in satisfying all eight structural properties that the wisdom traditions use to characterise ultimate reality.

The supply-demand calculation

We can now ask two separate probabilistic questions — one about the mathematics side (how rare is the match?), one about the traditions side (how likely is their convergence by chance?) — and combine them. The two sides are genuinely independent: Cantor had no knowledge of these traditions when developing his mathematics in the 1870s-1890s.

The demand-side calculation uses 50% as the null probability per structural property per tradition. This is the maximum-entropy assumption — the most generous conceivable to the null hypothesis — because it treats each tradition as effectively tossing a coin when selecting which structural features to emphasise. Any more informative estimate of the actual base rate would make the coincidence less probable, not more. The 50% figure is not arbitrary: it is the assumption that is maximally favourable to the claim that the convergence is accidental.

The supply-side calculation requires one further methodological note. The eight structural properties are not statistically independent — a system with strict level-differences (W3) is likely to have a hierarchy of distinct levels (W2). Simply multiplying all eight marginal probabilities would understate the true rarity by ignoring these correlations. To correct for this, the properties were submitted to tetrachoric exploratory factor analysis across the 45-domain comparison set. Three meaningful factors emerged — broadly corresponding to unbounded extension (W1, W8, W9), stratified differentiation (W2, W3, W4), and generative containment (W5, W7). The supply-side calculation multiplies only the three most restrictive factor-level marginals, giving a dimension-corrected estimate that does not double-count correlated criteria.

The combined figure — somewhere between 1 in 37,000 and 1 in 1,000,000 depending on which assumptions are used — represents the probability that both the supply-side rarity and the demand-side convergence are simultaneously coincidental. Even the conservative end of this range is striking.

A critic might reasonably ask whether 50% is the right base rate for the demand side. Table 4b addresses this directly. The key insight is that 50% is already the most conservative possible assumption — the maximum-entropy null. Any lower base rate (any more realistic estimate of how often a tradition highlights a given structural property) makes the coincidence less probable, not more. The evidence does not depend on the 50% assumption; it is strengthened by any departure from it.

The primary supply-side result is the one that requires no probability model at all, and therefore no independence assumptions of any kind: Cantor is the only domain in the 45-domain comparison set achieving a full match on all eight criteria. This is an empirical observation — p = 0 among 45 comparison domains — and it is dependence-preserving by construction, since it uses the actual joint distribution of properties in the real data rather than any model of that distribution. The dimension-corrected factor product reported in Table 4 is a secondary, illustrative calculation that gives the uniqueness finding a quantitative scale; the uniqueness finding itself is the primary claim, and it stands independently of any assumptions about independence among criteria.

Honest limitations

These figures are presented as exploratory, not confirmatory, for three reasons that should be stated clearly. First, the structural properties were identified from Murti's comparative analysis before the comparison set was tested against them. Although the mathematical comparison set was constructed independently using objective AMS MSC 2020 criteria, the ideal confirmatory study would pre-specify the property set before examining either system — a pre-registered replication with blind coders, currently in preparation, will do exactly this. Second, coding reliability has been estimated through simulation rather than measured with actual blinded coders; that step is the immediate priority for the follow-up study. Third, the 45-domain pool, though principled, is a sample. These are genuine limitations that define the path to a stronger version of the argument, not refutations of it.

V. The Philosophical Argument: What Mathematics' Own Foundations Tell Us

The statistical argument establishes that the Cantor-Wisdom correspondence is unlikely to be coincidental. But it does not explain why the correspondence exists. For that, a second, completely independent line of reasoning becomes decisive.

There are multiple competing foundations for mathematics. Cantor's set theory was the first; category theory was developed independently in the 1940s as an alternative; homotopy type theory emerged from computer science and proof theory in the 2000s. These three frameworks were developed by different people, in different centuries, for different motivations, and none of them was designed to match the wisdom traditions. Yet all three converge on the same structural skeleton.

The convergence of three independently-developed mathematical foundations on the same structural skeleton — strict ordering, cumulative inclusion, canonical generation, asymmetric embeddability, no terminus — is itself an improbable fact. There are dozens of candidate structural features that a mathematical foundation might have or lack. The chance that three independent foundations would converge on exactly these five is not large.

But the more important point is the inferential one. Wigner established that mathematics reliably models reality. If that is granted, and if every independently-developed foundation for mathematics shares the same structural skeleton, then that skeleton is, in a precise sense, reality-compatible: it describes the kind of structure that reality is capable of sustaining. The skeleton is not an artefact of one person's mathematical imagination; it is a feature that emerges whenever human beings construct an adequate foundation for mathematical reasoning.

This argument does not require any probability calculations. It does not depend on the κ reliability of the coding. It is not affected by the sample size of the comparison set. It rests on three facts: that Wigner's observation is accepted, that the three foundations are independently developed, and that they share the structural skeleton. All three facts are non-controversial.

VI. Bringing the Two Arguments Together: The Bayesian Integration

The statistical argument and the philosophical argument are typically presented as separate lines of evidence — the statistics showing the match is improbable, the philosophy explaining why we might expect it. They can be formally combined using Bayesian reasoning.

The logic is as follows. The philosophical argument — that three independent mathematical foundations converge on the same structural skeleton, and that this skeleton is therefore reality-compatible — constitutes a prior probability. Even before looking at the wisdom traditions data at all, a thoughtful person who accepts Wigner's observation should assign some probability to the hypothesis that the structural skeleton of mature mathematics reflects something about the deep structure of reality. The statistical argument then functions as an update to that prior via Bayes' theorem.

A proper Bayesian update requires both likelihoods to be estimated explicitly: P(data | ¬H) — how probable the traditions' convergence would be if it were coincidental — and P(data | H) — how probable it would be if the correspondence reflects something real. The table below shows the resulting Bayes Factors and posteriors across a full sensitivity band for P(data | H), rather than a single point estimate. This directly addresses the concern that P(data | H) is subjective: the calculation is shown to be robust across the entire plausible range.

How to read this table: P(data|H) is the probability that, if the skeleton is real, three adversarial traditions would independently converge on these properties. Even the most pessimistic value (0.05 — only 1 chance in 20) gives Bayes Factors of 1,850, which is 18.5× Jeffreys' decisive threshold of 100. These are posterior ranges, not certainties; substantial uncertainty remains in the likelihood estimates, and a pre-registered replication is the appropriate confirmatory step. But the lower bound of every range (99.0%) already exceeds any standard threshold for strong evidence.

Table 6. Full Bayesian sensitivity analysis. Posterior ranges shown across P(data|H) = 0.05 to 0.50, for three starting priors. P(data|¬H) = 1/37,000 throughout (conservative composite from Table 4b). All Bayes Factors exceed Jeffreys' 'decisive' threshold of 100, even at the most pessimistic P(data|H) = 0.05.

P(data | ¬H) = 1/37,000 comes from Table 4 under conservative assumptions. P(data | H) is the probability that, if the structural skeleton is genuinely real, three philosophically adversarial traditions — actively arguing against each other's metaphysical positions — would independently arrive at descriptions matching exactly these properties. The range 0.05 to 0.50 spans from extreme pessimism (1 chance in 20 even given truth) to moderate confidence.

What the table shows is that the result is robust to uncertainty about P(data | H). The Bayes Factors range from 1,850 to 18,500 — all well above Jeffreys' threshold of 100 for strong evidence, which he termed 'decisive' — but the table invites readers to inspect how results move with assumptions rather than claiming a single definitive figure. The posterior lower bound of 99.0%, which holds for the most sceptical prior and the most pessimistic P(data | H) considered, already substantially exceeds any standard threshold for strong evidential support. These are not claims of certainty: substantial uncertainty remains in the likelihood estimates, and a pre-registered replication is the appropriate next step. The most defensible summary is that the evidence is strongly and robustly supportive of the hypothesis across the full range of defensible assumptions.

VII. What the Argument Does and Does Not Establish

It is important to be precise about what follows from this analysis — and what does not.

What it establishes: There is a structural correspondence between Cantor's transfinite hierarchy and the structural descriptions of ultimate reality offered by three philosophically adversarial wisdom traditions. This correspondence is quantifiably improbable as a coincidence (1 in 37,000 to 1 in 1,000,000 under conservative assumptions, strengthening substantially under any more realistic base-rate assumption). It is further supported by the observation that three independently-developed mathematical foundations converge on the same structural skeleton. When these two lines of evidence are combined via a properly specified Bayesian calculation — with full sensitivity analysis across both the prior and P(data|H) — the posterior probability that the correspondence reflects something real is high across all defensible assumptions.

What it does not establish: These results do not prove that any particular spiritual tradition is correct. They do not confirm the existence of subtle realms, of consciousness as a fundamental feature of reality, or of any specific metaphysical system. What they establish is structural compatibility: the kind of structure that mature mathematics necessarily develops is the same kind of structure that careful contemplative inquiry ascribes to ultimate reality. Whether that structural compatibility reflects deep identity, or analogy, or something else entirely, remains open.

The appropriate next step: A pre-registered replication with blinded coders (one mathematician, one scholar of Indian philosophy, one methodologist), a larger comparison pool drawn from the full AMS classification, and a formal inter-rater reliability study. A pre-registration protocol will be posted to OSF. The exploratory findings here define the methodology and identify the most important confirmatory targets — particularly the reliable operationalisation of W4 (strict proven excess) and an empirical estimate of the demand-side base rate from a broader survey of philosophical traditions.

The most defensible summary of the current state of the evidence: the match between Cantor's transfinite hierarchy and the structure of ultimate reality as described across the wisdom traditions is unlikely to be coincidental; the philosophical foundations of mathematics independently support this; the Bayesian combination of these two arguments produces strong evidential convergence across all defensible assumptions — while the metaphysical interpretation of what this means remains, properly, open.

REFERENCES

Murti, T.R.V. (1960). The Central Philosophy of Buddhism. George Allen & Unwin.

Wigner, E.P. (1960). The unreasonable effectiveness of mathematics in the natural sciences. Communications on Pure and Applied Mathematics, 13(1), 1-14.

Cantor, G. (1883). Grundlagen einer allgemeinen Mannigfaltigkeitslehre. Teubner.

Mac Lane, S. (1971). Categories for the Working Mathematician. Springer.

Homotopy Type Theory: Univalent Foundations of Mathematics. (2013). The Univalent Foundations Program, Institute for Advanced Study.