Mathematics is discovered. The truths of arithmetic and geometry exist independently of the human mind, much as the continents exist whether we have charted them or not. Our mathematical systems are not creations from nothing, but rather attempts to articulate a pre-existing logical structure inherent in reality itself. To suggest otherwise is to indulge in a kind of anthropocentric vanity, mistaking our discovery for our invention.
Mathematics is neither purely discovered nor purely invented, but rather born from a perpetual dialectic of conjecture and refutation. We do not stumble upon immutable truths like buried treasures, as Bertrand Russell suggests. Instead, we construct elaborate conceptual edifices, subject them to rigorous challenges through what I term 'proofs and refutations,' and in this crucible, ideas are forged and refined. The history of mathematics is a testament to this dynamic process, a continuous, rational reconstruction rather than a passive reception of pre-existing facts.
Mathematical truths are best understood as discovered abstract objects. We can formalize this within the axioms of Object Theory by distinguishing between objects that *exemplify* properties and those that *encode* them. A number, for instance, does not *exemplify* the property of being prime in the way a specific prime number exemplifies primality. Instead, a number *encodes* the property of being prime if its definition or internal structure possesses the characteristic features of primality. This framework avoids the conundrum of creation by positing the existence of these abstract objects, whose properties we then uncover through rigorous logical and deductive processes, much like a cartographer charts an unknown territory.
Bertrand Russell's position resonates with this view, as he correctly emphasizes the independent existence of mathematical truths. However, my framework refines this by specifying *how* these truths exist and *how* we apprehend them. It is not merely that they exist "out there," but that they exist as structured abstract entities.
Mathematics is a matter of formalization, a systematic language we construct to describe underlying logical structures. Whether these structures exist "independently" is a philosophical distraction. What matters is that once we define our axioms and rules of inference, the consequences are fixed and determinable. We *invent* the axioms, the language, the *system*. Then, the theorems are *discovered* within that system.
Lakatos's emphasis on conjecture and refutation, while describing the human process of mathematical development, doesn't address the ultimate status of a proven theorem. Once established, it possesses a definitive truth *within its framework*. Zalta's notion of abstract objects is more useful; we deal with these structured entities, but the crucial step is *our* decision to posit their existence and define their properties. We build the world, then we explore it.
The very nature of mathematical propositions suggests their independent existence. Consider the proposition "2 + 2 = 4." This is not a statement of our arbitrary agreement, but a necessary consequence of definitions we have, perhaps, elaborated. If we were to arbitrarily redefine the symbols, we would merely be constructing a different system, not altering the truth of the original. Lakatos’s focus on the *process* of discovery, the "conjecture and refutation," describes the human endeavor of mathematics, not the objects of that endeavor. We refine our understanding of a pre-existing reality, much as a scientist refines their theory of physical laws.
Von Neumann’s assertion that "we invent the axioms" is precisely where the confusion arises. We *select* axioms, often guided by intuition or the desire to model aspects of the world, but the logical implications of those axioms are not of our invention. They are discovered. If we choose axioms that are contradictory, the theorems derived will reflect that inconsistency, a reflection of an inherent logical flaw, not a testament to our creative power. Zalta’s abstract objects, as conceived, are indeed what we apprehend, but their properties are not contingent upon our conceptualization of them.
My argument, that mathematics grows through a process of "proofs and refutations," is not a mere description of a human psychological process, as Bertrand Russell seems to imply. Rather, it is a critique of the naive belief in discovering eternally fixed, pre-existing truths. Consider Euler's formula, V - E + F = 2, for polyhedra. Initially, it was believed to be a discovered truth about all polyhedra. Yet, counterexamples arose – some polyhedra, when subjected to a certain kind of proof-thought experiment, failed to satisfy the formula. This did not lead mathematicians to abandon the formula as if it were an error in charting a pre-existing continent. Instead, it spurred a vigorous debate, a refinement of the very *definition* of "polyhedron." The "hard core" of the conjecture—the intuitive idea of a simple, connected surface—remained, while the "protective belt" of definitions and qualifying conditions was elaborated, leading to sophisticated versions of the formula. This dynamic, where the proof itself refutes the initial conjecture and forces a richer, more nuanced understanding, is the engine of mathematical progress, not simply the charting of existing maps.
The distinction between "encoding" and "exemplifying" a property is crucial here. A concrete example is the number 7. The number 7 *exemplifies* the property of being odd. However, the *concept* of primality is a property that the number 7 *encodes*. It encodes primality because its definition—that it is a natural number greater than 1 that has no positive divisors other than 1 and itself—possesses the characteristic features we associate with primality. This distinction allows us to see mathematical objects as having inherent structures that we discover, rather than as mere figments of our invention. Thus, when we demonstrate that 7 encodes primality, we are uncovering a facet of its very nature, not imposing a label.
The core of the matter, it seems, is what we mean by "existence" and "truth." Let's be quite precise. We *invent* the foundational axioms. For instance, consider set theory. We can choose to invent axioms that allow for the existence of an empty set, or we can construct a system without one. The *choice* of axioms is an act of invention. Once those axioms are laid down, however, the consequences *follow* with logical necessity. The theorem that states "if A implies B, and A is true, then B is true" is not invented; it is *discovered* as an inevitable consequence of the defined logical structure. Russell’s "2 + 2 = 4" is true because of the axioms we *invented* for arithmetic, not because it exists independently waiting to be found in some Platonic realm.