Scientific Realism

Scientific realism is the view that science aims to describe a world that exists independently of us, and that it often succeeds: the unobservable entities its best theories posit—electrons, genes, quarks, gravitational fields—are real, and the theories that describe them are at least approximately true.

13 frameworks, from 1906 to today · Philosophy Of Science · Philosophy

still a working view ended · each leads to its entry below

Overview

Scientific realism is the view that science aims to describe a world that exists independently of us, and that it often succeeds: the unobservable entities its best theories posit—electrons, genes, quarks, gravitational fields—are real, and the theories that describe them are at least approximately true. The stakes are large. If realism is right, scientific progress is genuine discovery, and the success of science is evidence that it has latched onto how things really are1. If it is wrong, science may be a superb instrument for predicting experience while telling us nothing about what lies beneath it. Much of the history of philosophy of science since the early twentieth century can be read as a long argument about this claim.

Logical Empiricism and the Demotion of Theory

The debate begins with the logical empiricists of the Vienna Circle and Berlin group in the 1920s and 1930s—Moritz Schlick, Rudolf Carnap, Hans Reichenbach, Carl Hempel, and their allies. They inherited the empiricist conviction that knowledge must be grounded in experience, but they faced a new problem: modern physics was saturated with unobservable entities. How could a theory that spoke of atoms and electromagnetic fields be meaningful if meaning came only from observation?

Their answer was a sharp distinction between observational and theoretical language1. Theoretical terms were to be tied to observation through correspondence rules, and the theoretical part of science was treated as a calculating device for organizing predictions. This was not quite instrumentalism—Carnap and Hempel insisted theories had cognitive content—but it pushed toward the view that the observable was the epistemically secure base. The logical empiricists also developed the deductive-nomological model of explanation and the hypothetico-deductive account of confirmation, tools that shaped the field for decades.

The internal difficulties mounted. The observational/theoretical distinction proved hard to draw: is "the pointer reads 5" observational while "the current is 5 amperes" theoretical1? And the correspondence rules never fully captured how theoretical terms function in practice.

Instrumentalism and Its Critics

Against this background, instrumentalism—the view that theories are instruments for prediction rather than descriptions of reality—found its most influential statement in Pierre Duhem's The Aim and Structure of Physical Theory (1906) and, later, in the work of the logical empiricists' successors. Duhem argued that physical theory is a system of mathematical propositions whose purpose is to save the phenomena, not to reveal hidden causes. His holism—that hypotheses are tested in bundles, never singly—also undercut the idea that individual theoretical claims could be confirmed as true2.

The realist counterattack came from several directions. Grover Maxwell's 1962 paper "The Ontological Status of Theoretical Entities" argued that the observable/unobservable distinction is a continuum, not a sharp line: we can see through a window, through glasses, through a microscope, and there is no principled point at which "seeing" stops. If we accept the entities we observe through instruments, consistency demands we accept the entities posited by theory. Meanwhile, the causal theory of reference developed by Saul Kripke and Hilary Putnam in the 1970s and 1980s gave realists a way to say that "electron" refers to the same thing across theory change even if beliefs about electrons change—reference is fixed by causal contact with the world, not by a description.

The Pessimistic Induction and the Realist Response

The most serious challenge to realism was formulated by Larry Laudan in 1981 as the "pessimistic meta-induction." The history of science is a graveyard of theories once regarded as true: phlogiston, caloric, the luminiferous ether, the crystalline spheres. If past theories were false despite their success, why believe current successful theories are true? The induction from past failure to present fallibility is the pessimistic induction3.

Realists responded by refining what they claim. The no-miracles argument, associated with Putnam and Richard Boyd, holds that the predictive success of science would be a miracle if theories were not at least approximately true—realism is the best explanation of success. Against the pessimistic induction, realists developed selective realism: we need not claim that whole theories are true, only that the parts responsible for success are. Philip Kitcher and Stathis Psillos argued that the working posits of successful theories—the entities actually doing explanatory work—tend to be retained, while the idle metaphysical baggage is discarded. The ether was not doing the work; the electromagnetic equations were.

Structural Realism

A more radical response came from John Worrall in 1989, drawing on Henri Poincaré's earlier suggestion: structural realism. Worrall observed that in the transition from Fresnel's wave theory of light to Maxwell's electromagnetism, the equations survived even as the ontology (the ether) was abandoned. What persists across theory change is mathematical structure, not the nature of the entities. Worrall's "epistemic" structural realism holds that we can know the structure of reality but not its intrinsic nature. Later, James Ladyman distinguished this from "ontic" structural realism, the view that structure is all there is—there are no objects with intrinsic natures underlying the relations. The debate over whether structure can be known without knowing what it is structure of remains live.

Entity Realism

A different route to realism was taken by Ian Hacking in the 1980s4. In Representing and Intervening (1983), Hacking argued that the strongest evidence for realism comes not from theorizing but from doing: when physicists spray electrons at a niobium ball to change its charge, or use them to probe other particles, they are manipulating electrons as tools. You can't spray something that doesn't exist. This entity realism—realism about entities rather than theories—holds that experimental intervention gives us grounds to believe in the entities we manipulate, even if our theories about them are imperfect4. Nancy Cartwright developed a related position, arguing that the fundamental laws of physics are not true but that the causal capacities of entities are real.

Constructive Empiricism

The most systematic anti-realist position of the late twentieth century is Bas van Fraassen's constructive empiricism, set out in The Scientific Image (1980). Van Fraassen accepts that theories should be literally interpreted—they say what they say—but argues that acceptance of a theory requires only that it be empirically adequate: that it save the phenomena. Belief in the unobservable entities a theory posits is not required. The position is distinguished from earlier instrumentalism by its insistence that theories are genuine representations, not mere calculating devices, and by its use of the semantic view of theories (theories as families of models). Van Fraassen's challenge forced realists to say precisely why empirical adequacy is not enough—why success at the observable level should lead us to believe in the unobservable.

Kuhn, Paradigms, and the Turn to Practice

Thomas Kuhn's The Structure of Scientific Revolutions (1962) reshaped the terrain without directly settling the realism question. Kuhn's account of paradigms, normal science, and revolutionary paradigm shifts raised the specter of incommensurability: if successive paradigms cannot be fully translated into one another, the realist claim that later theories are closer to the truth becomes hard to state56. Kuhn himself resisted the relativist reading, and later work—by Kuhn and by others—softened incommensurability into localized translation failure. But the historical turn he inaugurated shifted attention from logical relations among statements to the practices, communities, and instruments of science, and this shift fed directly into the entity realism of Hacking and the practice-oriented philosophy of the 1980s and 1990s.

The Present: Pluralism and Open Questions

Contemporary scientific realism is less a single doctrine than a family of positions. The no-miracles argument and the pessimistic induction remain the twin poles around which debate turns. Selective realism and structural realism compete as responses to theory change. Entity realism and Cartwright's capacities approach keep attention on experiment and causal powers. Van Fraassen's constructive empiricism remains the benchmark anti-realist position, though it has been joined by more radical forms of empiricism and by deflationary and pragmatic accounts that question whether the realism/anti-realism dichotomy is well-formed.

New fronts have opened. The realism debate has been extended to the special sciences—biology, economics, psychology—where the question of what the "entities" are and whether theories aim at truth looks different7. The status of models and computer simulations has raised questions about what representation and truth even mean for a model that is deliberately idealized. And the "new experimentalism" that followed Hacking has made the laboratory, rather than the theory, the central site of philosophical attention.

What unites the field is the question that opened it: whether the astonishing predictive and technological success of science is best explained by supposing that our theories are tracking a mind-independent world, or whether that success can be accounted for without that supposition. The answers have multiplied, but the question has not gone away.

References

  1. Scientific realism (Main features, No miracles argument, History), Wikipedia
  2. Duhem–Quine thesis, Wikipedia
  3. Pessimistic induction, Wikipedia
  4. Entity realism (introduction, Criticism), Wikipedia
  5. Thomas Kuhn, Wikipedia
  6. Commensurability (philosophy of science), Wikipedia
  7. Philosophy of science, Wikipedia

The frameworks of Scientific Realism

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1906-1980

    Instrumentalism

    Are theories descriptions or prediction tools?

    Fly to its tower in Nousfarer

  2. 1920-1960

    Logical positivism

    How can theoretical terms be meaningful?

    Fly to its tower in Nousfarer

  3. 1962-1990

    Kuhnian Paradigm Theory

    How does scientific change actually happen?

    Fly to its tower in Nousfarer

  4. 1970-2000
  5. 1975-Present

    No-Miracles Argument

    Why does science predict so well?

    Fly to its tower in Nousfarer

  6. 1980-Present

    Constructive Empiricism

    When should we believe in unobservables?

    Fly to its tower in Nousfarer

  7. 1980-Present

    New Experimentalism

    What does experiment contribute to knowledge?

    Fly to its tower in Nousfarer

  8. 1981-Present

    Pessimistic induction

    Does past failure doom present theories?

    Fly to its tower in Nousfarer

  9. 1983-Present

    Causal Capacities Realism

    Are laws or capacities what is real?

    Fly to its tower in Nousfarer

  10. 1983-Present

    Entity Realism

    Does manipulating entities prove they exist?

    Fly to its tower in Nousfarer

  11. 1989-Present

    Structural Realism

    Is only structure retained across theory change?

    Fly to its tower in Nousfarer

  12. 1990-Present
  13. 1990-Present

    Selective Realism

    Which parts of theories survive revolutions?

    Fly to its tower in Nousfarer

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

InstrumentalismLogical positivismKuhnian Paradigm TheoryCausal Theory of ReferenceNo-Miracles ArgumentConstructive EmpiricismNew ExperimentalismPessimistic inductionCausal Capacities RealismEntity RealismStructural RealismScientific PluralismSelective Realism

Beyond the field

Where Scientific Realism's ideas came from in other fields, and what they led to there: the strongest 12 of 38 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Scientific Realism is a planet in Nousfarer

Its city has a tower for each of these 13 frameworks. Land, walk in, and climb one concept at a time.

Land on Scientific Realism

Related fields

Fields its ideas are linked to

More in Philosophy Of Science

See where Scientific Realism sits on the Map of Human Knowledge