Anti-Realism

Anti-realism is the family of positions in philosophy of science, philosophy of mathematics, and metaphysics that deny that the entities, structures, or truths described by a theory exist independently of our capacity to know, verify, or construct them. It is not a single doctrine but a spectrum: at one end, modest anti-realism holds that unobservable entities posited by scientific theories are useful fictions; at the other end, radical anti-realism holds that reality itself is constituted by practices of assertion, proof, or measurement, with no residue left over.

The positions are unified not by a positive thesis but by a common suspicion: that realism — the view that the world is a certain way independent of our knowledge of it — smuggles in assumptions about correspondence, verification-transcendence, and epistemic access that cannot be justified without circularity. Anti-realism asks the realist to explain how we could know that our theories correspond to reality if the correspondence itself is not independently verifiable.

Scientific Anti-Realism

Scientific anti-realism, most prominently defended by Bas van Fraassen, holds that the aim of science is not truth but empirical adequacy. A theory is empirically adequate if everything it says about observable phenomena is true. What it says about unobservable entities — electrons, quarks, curved spacetime — is neither true nor false in a metaphysical sense; it is a instrument for generating predictions. Van Fraassen's "constructive empiricism" does not deny that electrons exist. It denies that science is committed to their existence.

This position has bite. The history of science is a graveyard of confident ontologies: phlogiston, caloric, the luminiferous ether. Each was once as well-supported by evidence as electrons are today. The anti-realist argues that the realist's confidence in current unobservables is a failure of inductive modesty — a statistical fallacy dressed as metaphysics. The realist responds with the "no-miracles argument": if electrons were not real, it would be a miracle that theories positing them make such accurate predictions. The anti-realist counters that predictive success is explained by the theory's empirical adequacy, not its truth, and that the history of science shows successful theories routinely positing non-existent entities.

Mathematical Anti-Realism

In the philosophy of mathematics, anti-realism takes the form of various forms of constructivism, formalism, and intuitionism. The intuitionist, following L.E.J. Brouwer, denies that mathematical truths exist independently of the mental constructions that verify them. A proposition is true only if there is a constructive proof of it; it is false only if there is a constructive refutation. The law of excluded middle — that every proposition is either true or false — fails because there may be propositions for which neither a proof nor a refutation exists.

This is not merely a philosophical preference. It has mathematical consequences. Intuitionistic logic is weaker than classical logic — it cannot prove certain theorems that classical mathematics takes for granted — but it is also more fine-grained. It distinguishes between "there is no proof of not-P" and "there is a proof of P", a distinction that classical logic collapses. In computer science, this distinction is computationally meaningful: the first corresponds to a function that does not halt with an error; the second corresponds to a function that halts with a value.

Quantum Anti-Realism

The most consequential locus of anti-realist thought in contemporary physics is the interpretation of quantum mechanics. The Copenhagen interpretation, associated with Niels Bohr, is often read as anti-realist: the wave function is not a description of an independently existing physical state but a tool for predicting the outcomes of measurements. The properties of quantum systems — position, momentum, spin — do not have definite values until measured. On this view, quantum mechanics is complete not because it describes everything, but because there is nothing further to describe.

The anti-realist reading of Copenhagen is contested. Bohr's own views are notoriously difficult to interpret, and some scholars read him as a kind of relational realist rather than an anti-realist. What is clear is that the measurement problem — the question of how and why quantum superpositions collapse into definite outcomes — does not arise for the anti-realist in the same way it arises for the realist. For the anti-realist, there is no fact of the matter about which path the photon took through the double slit; the question is malformed because the property does not exist independently of the measurement context.

Semantic Anti-Realism

Michael Dummett's semantic anti-realism extends the intuitionist critique from mathematics to language as a whole. The meaning of a statement, on Dummett's view, is given by the conditions under which it can be verified or falsified, not by the conditions under which it would be true in a mind-independent reality. This has the striking consequence that statements about the distant past, the distant future, or regions beyond our light cone may be neither true nor false — not because we lack evidence, but because the conditions for their verification do not obtain and cannot in principle obtain.

Semantic anti-realism challenges the correspondence theory of truth at its root. If meaning is verification-conditional, then truth cannot be correspondence to a reality that transcends our epistemic capacities. The result is a form of anti-realism that is not merely about science or mathematics but about the structure of meaning itself.

Systems-Theoretic Anti-Realism

A less discussed but increasingly relevant form of anti-realism emerges from complex systems theory and the study of emergence. In complex systems, the macroscopic properties of a system — its temperature, its organization, its function — are not properties of the microscopic components but patterns that arise from their interactions. The question of whether these patterns are "real" or "merely apparent" is not well-posed. The pattern is real in the sense that it has causal consequences; it is not real in the sense that it reduces to the interactions of components.

This suggests a pragmatic anti-realism about emergent properties. We do not need to decide whether consciousness, free will, or the economy are "really real" in an ontological sense. We need only recognize that they are patterns that supervene on lower-level dynamics and that treating them as real for certain purposes — prediction, intervention, moral reasoning — is instrumentally justified. The ontology becomes a function of the context of use, not a fixed inventory of the universe's furniture.

The connection to the participatory universe thesis is direct. If the observer is not merely a passive recorder of pre-existing facts but an active participant in the constitution of reality — as John Wheeler argued and as quantum mechanics suggests — then anti-realism is not a philosophical retreat from objectivity. It is a recognition that objectivity itself is a property of systems that include observers, not a god's-eye view from nowhere.

The Costs of Anti-Realism

Anti-realism is not cost-free. The most serious charge against it is that it undermines the normativity of truth. If truth is verification-conditional, then we cannot say that a proposition is true in a sense that would hold even if all evidence were destroyed. We cannot say that dinosaurs existed before humans evolved to think about them — not because the evidence is lacking, but because the truth-conditions of the statement are tied to the existence of evidence. This strikes many as a reductio of anti-realism: it makes truth too parochial, too dependent on our contingent cognitive capacities.

The anti-realist's response is to embrace the parochialism. Truth has always been a human practice, shaped by our perceptual apparatus, our linguistic conventions, and our cognitive limitations. The realist's "mind-independent reality" is not a discovered fact but a regulative ideal — a useful fiction that disciplines our inquiries without corresponding to anything we could actually know. On this view, anti-realism is not skepticism. It is the refusal to pretend to knowledge we do not have.

The debate between realism and anti-realism is not a debate about the world. It is a debate about what we are doing when we say we know something about the world. The anti-realist does not deny that the external world exists. The denial is semantic: that "exists" is not a predicate we can apply independently of our practices of verification. The world is not a fiction. But our claims about it are — not fictions in the sense of lies, but fictions in the sense of constructions, made to be tested, revised, and sometimes abandoned. The anti-realist asks only that we be honest about the constructedness of our knowledge, even — especially — when it works.