Section 2A: Mathematics and the Individual
The case for Frame A in mathematics, argued from within.
The Natural Picture
Mathematics, on first approach, looks like the strongest possible case for Frame A. Numbers are the paradigm individuals: discrete, self-identical, clearly bounded. The number seven is what it is. It does not shade into eight, borrow properties from its neighbours, or depend on its context for its identity. You can count with it in any order, in any language, in any culture, and it remains seven. If anything in the universe has intrinsic identity, it is the natural numbers.
This intuition is not naive. It connects to one of the oldest and most persistent positions in the philosophy of mathematics — Platonism — and to serious alternatives that share its individualist structure even when they reject its metaphysics.
Platonism and the Reality of Mathematical Individuals
Mathematical Platonism holds that mathematical objects — numbers, sets, geometric figures, functions — exist independently of us, independently of our notation, and independently of any physical instantiation. They are abstract individuals, fully constituted, waiting to be discovered rather than invented.
This view has remarkable staying power, and not without reason. The experience of mathematical discovery feels like exploration rather than construction. Mathematicians consistently report encountering necessity: not that they chose to define things a certain way, but that the structure forced a particular result. Euler did not decide that e^(iπ) + 1 = 0. He found it. The theorem was there before he arrived. If that feeling of discovery tracks something real — and it is very difficult to explain mathematical practice without assuming it does — then mathematical objects have a prior existence that the mathematician’s activity merely makes contact with.
Platonism gives Frame A its strongest mathematical foothold. If numbers and sets are genuine individuals existing in their own domain, then relations between them are secondary: defined on top of individuals that already have their identity. Two is even not because of its relations to other numbers but because of what it intrinsically is. Its relational properties follow from its individual nature, not the other way round. The individual is prior.
The most sophisticated modern defence of this position comes from Gödel himself. Gödel argued that mathematical intuition — our ability to perceive mathematical truth — is analogous to sensory perception: a faculty that puts us in contact with a domain of mind-independent objects. His incompleteness theorems, far from undermining this view, seemed to him to confirm it: the fact that formal systems cannot capture all mathematical truth implies that mathematical truth outruns any human construction, which is exactly what Platonism predicts. (See Gödel, What is Cantor’s Continuum Problem?, 1947.)
Formalism and Nominalism
Platonism is not the only position available to Frame A. Two important alternatives share its individualist structure while rejecting abstract objects entirely.
Formalism, associated with Hilbert, holds that mathematics is the manipulation of symbols according to rules. Mathematical objects are not abstract individuals existing in a Platonic realm — they are formal marks, individuated by their syntactic identity. The number seven is the symbol ‘7’ governed by the rules of arithmetic. Relations are formal operations defined over these symbols. The individual mark, with its formal identity, is still prior.
Nominalism holds that only concrete, particular individuals exist. Mathematical objects are either eliminated in favour of concrete inscriptions, or reconstructed from patterns among physical objects. Either way, the individual particular — the concrete token, the specific physical arrangement — is the basic unit. Abstract individuals are replaced by concrete ones, but the individualist structure is preserved.
These positions differ significantly in their metaphysics. What they share is the commitment that the basic units of mathematics — whether abstract objects, formal symbols, or concrete particulars — are individuals whose identity is fixed prior to their relations. Relations are defined over them, not constitutive of them.
What Frame A Asks of Mathematics
The Frame A picture of mathematics is coherent and philosophically serious. What it asks of mathematics is that the objects be, at some level, prior to the relations: that there is a stable inventory of mathematical individuals whose identity can in principle be specified independently, with relational structure erected on top.
Whether mathematics actually delivers this — whether the deepest mathematical structures reward an individualist approach or resist it — is what Section 2B examines.
Section 2B: Mathematics and Structure
The case for Frame B in mathematics, argued from within.
What Mathematics Keeps Finding
Frame B does not begin by imposing a philosophical thesis on mathematics. It begins by asking what mathematics, pursued on its own terms and driven by its own internal pressures, keeps discovering about the nature of its objects.
The answer, repeatedly and from different directions, is this: the identity of mathematical objects is constituted by their structural relations, not by any intrinsic properties they possess independently. When mathematicians probe the deepest and most constrained structures — the ones that turn out to be necessary rather than chosen — they find that relations are primary and individuals are secondary. This is not a philosophical preference imported into mathematics. It is what the mathematics says.
Category Theory: Identity as Relational Position
The clearest formal expression of this is category theory, introduced by Eilenberg and Mac Lane in 1945 and now one of the most foundational frameworks in mathematics.
Category theory defines mathematical objects not by their internal constitution but entirely by their relations to other objects — by the morphisms, or structure-preserving maps, between them. Two objects that stand in the same relational pattern to everything else in the category are, from the category-theoretic point of view, identical — regardless of what they are “made of.” Internal structure is irrelevant. Relational structure is everything.
This is not merely a convenient way of organising mathematics. It is a substantive claim about mathematical identity. The Yoneda lemma — one of the central results of category theory — formalises this precisely: a mathematical object is completely determined by its relations to all other objects. There is no residue, no further fact about the object beyond its relational position. The object just is its pattern of relations. (See Mac Lane, Categories for the Working Mathematician, 1971.)
The scope of this result turned out to be enormous. Category theory unified previously separate areas of mathematics — algebra, topology, logic, geometry — by looking past the objects to the relations between them. The unifications were not superficial. They revealed that apparently different mathematical domains were instances of the same abstract relational pattern. The pattern was primary; the objects were instantiations of it.
The Division Algebras: Constrained Structure
A second and independent witness comes from the division algebras — the number systems in which division is always possible. There are exactly four: the real numbers (ℝ), the complex numbers (ℂ), the quaternions (ℍ), and the octonions (𝕆). This was proved by Hurwitz in 1898. You cannot construct a fifth. The sequence is exhaustive.
What is striking is the structure of the progression. At each step, an assumption about the independence of individual elements is relaxed — and the resulting system is richer and more constrained:
- Real numbers are fully ordered and commutative: elements behave as independent units, interchangeable in any operation
- Complex numbers lose real ordering but gain rotational structure in the plane
- Quaternions lose commutativity — a × b ≠ b × a — the order of relation now matters; elements are no longer interchangeable
- Octonions lose associativity — (a × b) × c ≠ a × (b × c) — even the grouping of relations matters; no element can be treated independently of its full relational context
At each step, what is given up is the assumption that individual elements can be treated as prior to their relations. At each step, the structure that results is more constrained, more surprising, and — as the Physics section will show — more physically significant.
The sequence terminates at the octonions. This termination is not a modelling choice. It is a hard mathematical result: there is no fifth coherent structure of this kind. What the sequence exhaustively maps is not a set of systems we invented but a set of limits we found — the complete space of ways coherent relational structure of this type can exist. (See Baez, The Octonions, Bulletin of the American Mathematical Society, 2002.)
Bott Periodicity and the Fano Plane
The octonions do not merely terminate the sequence. They organise it. Bott periodicity, proved by Raoul Bott in 1957, establishes that the homotopy groups of the classical Lie groups repeat with period 8 — a period that reflects the 8-dimensional structure of the octonions. The algebraic landscape does not just end at 𝕆. It wraps around, with the octonions as the organising principle of the repetition. (See Bott, The Stable Homotopy of the Classical Groups, Annals of Mathematics, 1959.)
The Fano plane makes the relational character of the octonions visible geometrically. It is the unique projective plane of order 2: seven points, seven lines, three points per line, three lines per point. It encodes the multiplication structure of the octonions — each line corresponds to a quaternionic triple, and the geometry determines which combinations are valid.
What is remarkable about the Fano plane is its complete symmetry. There is no privileged point and no privileged line. Every element stands in exactly the same relational pattern to all the others. Identity is constituted entirely by position within the whole. It is, in the most literal sense, a diagram of pure relation — a structure in which the individual elements have no properties beyond their relational position, and in which that position is everything.
Gödel: The Whole Exceeds the Individual
A fourth line of evidence comes from Gödel’s incompleteness theorems (1931). Any formal system powerful enough to express basic arithmetic contains true statements that cannot be proved within that system. The system cannot simultaneously be complete and consistent.
Read in the context of Frame B, this result has a specific significance. Any attempt to capture mathematical truth within a closed individual system — a single, self-contained formal structure — will fail. The relational fabric of mathematical truth outruns any individual representation of it. There is always more to the whole than any part can contain.
This does not prove relational ontology. But it is exactly the pattern Frame B would predict: the individual system, however powerful, is insufficient. The relations to the wider structure — to the mathematical territory that exceeds any map — are not incidental. They are where the truth lives.
What Frame B Finds in Mathematics
The picture that emerges from inside Frame B is not one of objects with relations defined on top of them. It is one of relational structures within which objects are identifiable as positions. The objects are real. But their identity is constituted by the structure, not prior to it.
This is not a philosophical gloss applied to mathematics from outside. It is what category theory says, what the division algebra sequence implies, what Bott periodicity confirms, what the Fano plane makes visible, and what Gödel’s theorems suggest about the limits of any closed individual representation. The mathematics keeps finding relational structure at its deepest levels — not because mathematicians went looking for it, but because it is there.
Section 2C: Mathematics as Witness
What the mathematical evidence establishes — and what it leaves open.
The two sections above are not symmetric, and it would be dishonest to present them as if they were.
Frame A’s mathematical case rests on philosophical positions — Platonism, formalism, nominalism — that are serious and well-developed, but that are positions about mathematics rather than findings within it. Platonism explains the phenomenology of mathematical discovery. Formalism preserves consistency. Nominalism avoids ontological extravagance. These are genuine virtues. But they are frameworks imposed on mathematical practice in order to make sense of it. They do not emerge from the internal development of mathematics itself.
Frame B’s case rests on results that emerged from within mathematics, driven by internal pressures and producing surprising discoveries. Category theory was not developed to vindicate relational ontology — it was developed to solve concrete problems in algebra and topology, and the relational character of its foundations was a finding, not a starting assumption. The division algebra sequence was not constructed to illustrate Frame B — it was discovered as an exhaustive constraint on a certain kind of algebraic structure. Bott periodicity was not anticipated. The Fano plane’s complete symmetry was not designed.
This asymmetry — Frame A offering philosophical frameworks, Frame B offering mathematical discoveries — is itself significant. It is what Frame B would predict: that the deepest mathematical structures, when found rather than constructed, reveal relational character. And it is what Frame A struggles to account for, since on Frame A’s picture the relational structure of category theory should be a secondary feature of mathematics, not its most unifying and powerful framework.
This does not constitute proof. Frame A can respond. A Platonist can argue that the abstract individuals exist prior to any structural description we give of them, and that category theory is simply a particularly efficient way of talking about them — not evidence that the objects themselves are relational. A formalist can argue that the elegance of relational frameworks reflects something about the tractability of formal systems, not about the nature of mathematical objects. These responses are available.
What they require is an explanation of why the most constrained mathematical structures — the ones we found rather than chose — are precisely the ones where individual elements cannot be treated independently of their relational context. That explanation has not been given. Its absence is not a proof of Frame B. It is an open question that mathematics hands to physics, where the same structures reappear — and where the question of whether they are merely useful descriptions or genuine maps of the territory becomes harder to avoid.