Home > Positioning > Persons > van Fraassen

Bas van Fraassen (1941–)

Van Fraassen proposed that science does not owe us the truth about what cannot be observed. Constructive empiricism, the position of The Scientific Image (1980), holds that science aims at empirical adequacy — theories that get the observable phenomena right — and that accepting a theory requires believing only that much. A scientist can accept atomic theory, work inside it, build careers and instruments on it, without believing that its unobservable posits are real; acceptance is a commitment to confront further phenomena within the theory’s conceptual framework, not assent to a picture of the invisible. The move re-founded empiricism after logical positivism’s collapse, on new ground: no verificationist theory of meaning, no reduction of theory to observation-talk — theories are true or false as they stand; the empiricist simply declines to believe the part that outruns the phenomena. Published at the high tide of scientific realism, the book became the position realists measure themselves against, and forty years of the realism debate have been organised around it.

Life

Born 5 April 1941 in Goes, in the German-occupied Netherlands; his father, a steam fitter, was conscripted to factory work in Hamburg while the family remained in the Netherlands; reunited after the war, they emigrated to Edmonton, Canada, in 1956. BA at the University of Alberta (1963); PhD at Pittsburgh (1966) under Adolf Grünbaum, with a dissertation on the causal theory of time. He taught at Yale, the University of Southern California, and Toronto before Princeton, where he was McCosh Professor of Philosophy; after retiring from Princeton in 2008 he joined San Francisco State University. Alongside the philosophy of science he did foundational work in philosophical logic — free logic, the semantics of presupposition, and supervaluations, a technique he introduced that became standard well beyond its origin. A practising Catholic, he has written on the relation between empiricism and religious life — an unusual pairing he treats as a matter of stance rather than contradiction.

Constructive empiricism

The claim. Science aims to give us theories that are empirically adequate: what a theory says about the observable things and events is true. Acceptance of a theory involves belief only in its empirical adequacy — plus a pragmatic commitment to live and work inside the theory’s conceptual world. Truth about the unobservable is not the aim; belief about it is supererogatory. The position is agnostic, not atheist, about electrons: the theory may be true, and the constructive empiricist simply does not take that further step.

Observable is a scientific distinction. The load-bearing line — observable versus unobservable — is drawn by science itself, not by philosophy: observable is what beings like us, with our actual sensory endowment, can detect unaided. Moons of Jupiter are observable (fly closer); electrons are not, for any creature like us. Critics from Grover Maxwell onward pressed the continuum objection — glasses, binoculars, microscopes, where does observation stop? — and Paul Churchland asked why our sensory endowment should carry epistemic weight at all. Van Fraassen’s reply is that vagueness does not erase a distinction, and that the empiricist is describing our epistemic situation, the only one we have: the limits of observation are contingent facts about us, and empirical adequacy is adequacy to what is within those limits.

Against inference to the best explanation. The realist’s central engine — the theory best explains the phenomena, therefore it is probably true — van Fraassen attacked twice. Explanatory goodness, first, is a pragmatic virtue: explanation is an answer to a why-question, interest-relative and context-bound (his account of explanation in The Scientific Image develops this), and a theory’s being lovely is no mark of its being likely. Second, the “bad lot” argument (Laws and Symmetry, 1989): inferring the best of the available explanations assumes the true one is among them — the best of a bad lot is still a bad explanation. The realist replies — Boyd’s, Psillos’s — hold that the reliability of explanatory inference is itself empirically vindicated; the argument continues, and the miracle argument (realism as the only philosophy that doesn’t make science’s success a miracle — Putnam’s formulation) meets van Fraassen’s Darwinist deflation: theories that weren’t empirically adequate didn’t survive; no miracle, and no truth about the invisible required.

Laws, symmetry, and models. Laws and Symmetry argues that the concept of a law of nature — a necessity in the world answering to our best generalisations — should be given up: the philosophical accounts of laws all fail, and science can be understood without them, through symmetry principles and the structure of models. The semantic view of theories, which van Fraassen championed, carries the same shift: a theory is not a set of sentences but a family of models, and a theory is empirically adequate when the appearances are isomorphic to the empirical substructures of some model of the family. Scientific Representation (2008) completes the arc: models represent for users, in contexts — representation is a practice, not a mirror-relation.

The empirical stance. The Empirical Stance (2002) turns the method on empiricism itself. If empiricism were a doctrine — a thesis like “experience is the sole source of knowledge” — it would be self-defeating, itself unwarranted by experience. Empiricism is instead a stance: a cluster of commitments, attitudes, and values — respect for experience, suspicion of metaphysical explanation-demands — that one can adopt without contradiction and without proof. Rationality, on his voluntarist epistemology, is permission rather than compulsion: what reason requires is consistency, and much that we believe is decision, not derivation. The stance idea has had a career of its own beyond his position — developed and debated by Anjan Chakravartty and Paul Teller among others — giving the philosophy of science a vocabulary for positions that are neither theses nor mere tastes.

Quantum mechanics. Quantum Mechanics: An Empiricist View (1991) develops a modal interpretation — his “Copenhagen variant,” distinct from the versions of Kochen and Dieks: the quantum state describes what is possible, and measurement outcomes are actual without the state’s collapse being a physical process. The empiricist refusal to demand a picture behind the phenomena is the orientation he brings to relational quantum mechanics: “Rovelli’s World” (Foundations of Physics, 2010) works through Rovelli’s programme and presses its hardest question — whether relative facts can be coherently stated at all, given that any statement of them seems to require a vantage no relative fact provides — before concluding that the consistency questions resolve when the relational character of the description is taken fully seriously. The paper is one of the standing philosophical treatments of RQM, exact about what the relational picture does and does not owe an account of.

Where van Fraassen stops

The programme is deliberately austere about what philosophy may demand, and its boundary is the boundary it polices: the refusal to answer questions it regards as badly posed. Constructive empiricism does not say what the world is like behind the phenomena — not because the question is forbidden but because answering it is not science’s aim and not belief’s obligation. What the position owes on its own terms — an account of observability’s epistemic significance, a story about how acceptance-without-belief works in practising scientists, the defence of voluntarism against the charge that permission-rationality permits too much — van Fraassen has supplied in instalments across the corpus, and each remains argued. What he declines to supply, on principle, is the metaphysics: why the phenomena are as they are, what grounds the regularities, which interpretation of quantum mechanics is true. The declining is the position — empiricism as the stance that such demands may be set aside — and whether that is discipline or evasion is precisely what his critics and he dispute.

Key works

See also: Putnam · Rovelli · Carnap · Relational quantum mechanics