linkedinbloggertumblr
facebooktwitterreddit

A Historical Introduction to the
Philosophy of Science

Ch. 18: The Debate over Scientific Realism

Book cover: A Historical Introduction to the Philosophy of Science by John Losee

The following is a summary of the eighteenth chapter of John Losee's book, A Historical Introduction to the Philosophy of Science (fourth edition), with some ancillary notes.

Philosophers discussed in this chapter: Richard Boyd (1942—); Ian Hacking (1936—); Bas C. Van Fraassen (1941—); Arthur Fine (1937—)

(p. 252–3) Realists and instrumentalists debate over:

  1. what kinds of facts scientists are aiming at
  2. how to account for the progress of science throughout history

Truth Realism

(p. 253) The realist answers:

  1. scientists should aim to devise theories that describe the structure of the universe
  2. scientific progress shows (a) the universe has a structure that is independent of human thought and (b) theories are increasingly accurate

In the 1970s, realists pointed to the success of plate tectonics and the theory of DNA structure.

Putnam argued the increasing predictive success of theories shows greater approximation to truth and that posited theoretical entities (e.g., electrons, genes) really exist ('No Miracle' Argument).

Boyd emphasized how, for example, the methodological principle of assigning properties to theoretical entities yields increasing predictive success/instrumental reliability.

(p. 254) Boyd offered an 'abductive' argument (inference to the best explanation) for realism:

  1. If successive theories converge upon truth, then principles of scientific method are instrumentally reliable.
  2. Principles of scientific method are increasingly more instrumentally reliable.

Therefore: It is probable that successive theories typically converge upon truth.

Non-realists like Laudan, in contrast, point to how the long-term predictive success of a scientific theory is no guarantee of its truth (e.g., Ptolemaic planetary models, phlogiston theory, electromagnetic ether).

Laudan also complained that realists have not explained what they mean by 'approximate truth'/'progress toward truth'. Universal generalizations may be true, but per Hume, can't be proved to be true.

Entity Realism

(p. 254–5) Harré sought to overcome the problems with 'truth realism' by identifying three kinds of entities:

Realm 1: observable entities (e.g., Mars, Atlantic trench)

Realm 2: possibly observable entities with amplified human senses and posited by 'iconic theories' (e.g., micro-organisms, X-ray stars)

Realm 3: not possibly observable entities even with amplified human senses (e.g., neutrinos)

Book cover: Worldviews: An Introduction to the History and Philosophy of Science by Richard DeWitt

(p. 255) While Realm 2 entities easily satisfy scientists' existence criteria, Realm 3 entities are only detected indirectly by triggered events and so are more open to question in specific cases.

Hacking supported 'entity realism' by pointing out how scientists manipulate unobservable entities to produce new phenomena and to investigate unrelated areas of enquiry.

(p. 256) For example, the electron microscope uncovers the optically invisible structure of some proteins.

Contrariwise, Leplin points to entities that cannot satisfy Hacking's test of manipulability. For example, quarks act as triplets and cannot be isolated individually.

For Leplin, physicists elevate unifying explanatory power over empirical confirmation via novel predictions.

Losee observes that even though 'entity realism' does not require that every theoretical entity refer to an object in the real world, none the less, its failure to support existence claims for the entities postulated by today's fundamental theories is a problem.

Van Fraassen's Constructive Empiricism

(p. 257) Instrumentalists hold that:

  1. theories are calculating devices that facilitate the organization and prediction of observations
  2. only observation statements are true or false
  3. theories are only 'useful' or 'not useful'

With Van Fraassen's 'constructive empiricism', the aim of science is not the 'truth' of a theory, but its 'empirical adequacy'. A theory is empirically adequate if it saves the relevant phenomena.

For Van Fraassen, a statement about an observable is true or false if its value can in principle be determined by the unaided human senses. Craters on Neptune counts as an observable as humans can, in principle, travel to Neptune.
[LA: But does this distinction hold in cases like the craters on Neptune? Astronauts arriving at Neptune would require many aids to compensate for the lack of oxygen on Neptune and its enormous gravitational pull. Viewing its craters would require a highly controlled environment just to keep them alive. Furthermore, theoretical entities cannot be dispensed with. Even an astronaut looking through her specially-designed visor assumes the veracity of the laws of optics that describe how photons reflected by the craters travel through her visor to arrive at her eyes. If these theories of optics cannot be regarded as 'true', then how is the astronaut's observation of the craters corroborated as 'true'.]

Van Fraassen conceded that statements about the motions of electrons are both empirically adequate and capable of being true or false. But, for van Fraassen, scientists ought to remain agnostic about their existence.

Sober objected that van Fraassen's 'constructive empiricism' gives contradictory advice, depending on how a theory is stated (e.g., as 'food web' or 'eat but not eaten').

(p. 258) Hacking opposed van Fraassen's restriction of observables to what is perceived unaided with counter-examples of grids used in microscopic observation.

Hacking's pointing to the successful detection and manipulation of theoretical entities to investigate unrelated phenomena further refutes the instrumentalist account of these entities (e.g., electrons).

Fine's Natural Ontological Attitude

Fine agreed with Hacking that scientists posing truth claims about theoretical entities promote progress in science.

Fine distinguished two kinds of realism:

  1. 'local' realism – hypothesizes existence of theoretical entities in specific cases
  2. 'global' realism – assumes some theories truly represent the structure of the world

(p. 259) 'Convergent realism', as a variant of 'global' realism, posits that successive theories better approximate the truth.

Global anti-realists insist that no scientific theory is true or approximates the truth.

Fine's  'Natural Ontological Attitude' (NOA) puts the epistemic and ontological conclusions of scientists on par with common sense claims, while admitting that such claims are neither incorrigible nor invariably progressive in all respects.

For Fine, the endeavour to uncover globally 'the aim of science' requires therapy. Realism is outer-directed to correspondence with the world, while anti-realism is inner-directed to human relations with concepts and truth.

The 'Natural Ontological Attitude' accepts the evolving standards for judging truth-claims among scientists.

Cartwright on Truth-claims about Causal Mechanisms

(p. 259–60) Realists remain persuaded by the 'No Miracle' Argument. While anti-realists remain persuaded by the historical 'Pessimistic Meta-Induction' Argument.

(p. 260) Cartwright capitalized on how idealised theoretical entities are known not to exist (e.g., massless charges, ideal pendulums, absolute vacuums), thereby making the fundamental laws of physics false.

Cartwright regarded low-level phenomenological laws, on the other hand, as true. She also agreed with Hacking that such laws are evidence for theoretical entities where such entities cause the known law-like regularities (e.g., curves in a cloud chamber).

Structural Realism

Book cover: Epistemology: A Contemporary Introduction to the Theory of Knowledge by Robert Audi

(p. 261) With his 'Structural Realism', Worrall steered a middle path between Truth Realism/Entity Realism and anti-realism. Structural Realism eschews claims to the truth/approximate truth of theories and the existence of theoretical entities.

Worrall claimed only a mathematical mapping between theories and physical structures.

For example, although Fresnel and Maxwell had different ontological commitments in their two competing theories (Maxwell's had no ether), their mathematical structures are the same.

Worrall noted that with some competing successive theories, mathematical form is shared under limiting conditions (e.g., Special Relativity and Newtonian mechanics).

(p. 262) Psillos objected that Worrall assumed without argument that a persisting structure is an indicator of real relations between objects.

As objective structural relations require entities to be so related, Chakravartty concluded that Structural Realism assumes Entity Realism. Conversely, if theoretical entities exist, there must be structural relations binding them for them to be detectable reliably. Thus, Entity Realism assumes Structural Realism.

Ladyman pointed out a further problem for realism: in modern physics, the ontological status of subatomic entities is ambiguous (wave-particle duality). Ladyman's solution is to grant existence to invariant structures only (e.g., elementary particles with invariant quantities).

Ladyman's solution is more inclusive than Hacking's in allowing for the existence of neutrinos and quarks.

Harré and Madden see the atomistic metaphysical position favouring Entity Realism, while modern 'Great Field' theory favours Structural Realism. Physicists' preference today for the Great-Field metaphysic lends more weight to Structural Realism.

Questions to Consider:

  1. If universal generalizations cannot be proved, then how would a realist make sense of 'closer approximations to the truth'?
  2. Is Entity Realists' inability to indentify every theoretical entity with a successful novel prediction a serious problem for Entity Realism?
  3. Does van Fraassen's sharp distinction between theoretical entities on the one hand and in principle observable entities on the other withstand scrutiny?
  4. Does the fact that there are no idealised theoretical entities in nature make the fundamental laws of physics false?
  5. Is Structural Realism a more defensible version of realism compared with Truth Realism and Entity Realism? If so, why?

Copyright © 2022–3

PDF Download Download this review

You will be interested in
Book cover: Language, Truth and Logic by Alfred J. Ayer
Book cover: Key Philosophical Writings by Rene Descartes
Book cover: Berkeley's Philosophical Writings by George Berkeley
Book cover: Clavis Universalis by Arthur Collier
Book cover: What Is This Thing Called Science? by Alan F. Chalmers
Book cover: Meditations and Other Metaphysical Writings by Rene Descartes

Share This

  • twitter
  • facebook
  • linkedin
  • googleplus
  • gmail
  • delicious
  • reddit
  • digg
  • newsvine
  • posterous
  • friendfeed
  • googlebookmarks
  • yahoobookmarks
  • yahoobuzz
  • orkut
  • stumbleupon
  • diigo
  • mixx
  • technorati
  • netvibes
  • myspace
  • slashdot
  • blogger
  • tumblr
  • email
Short URL: