truth in scienceOverview

What is science, and what are its intellectual underpinnings?


To the ancients, scientia meant knowledge and experience: wisdom, in short. But science today implies something else: knowledge collected by following certain rules, and presented in a certain way. Scientists are realists: they believe in the existence of an external reality which philosophers have never been able to prove. The point is worth stressing. Science attempts to make a sharp distinction between the world out there, which is real and independent of us, and the individual's thoughts and feelings, which are internal and inconstant and to be explained eventually in terms of outside realities. {1}

Scientists therefore look for external testimony: they study those aspects of knowledge where there can be overwhelming agreement. And that knowledge they group under laws, which are invariable relations and regularities. Laws of substance are the more descriptive: certain plants have a certain structure: water boils at 100 degrees Centigrade. Laws of function concern cause and effect, the invariable relations that hold between the properties of materials and systems. The social sciences do not have the precision of the physical sciences, of course, and the part played by chance and irreversible processes is being increasingly recognized in all areas of science. {2}

Laws are invariable relationships universally accepted in the relevant scientific community. Theories are more open to doubt and refutation. Hypotheses are tentative theories. But laws do not provide explanations: they simple state the relationship between the relevant variables. Theories give more of a picture, some insight into how it is that a law holds. What then are these theories? Two views. One is that they are convenient fictions, compact reformulations of laws. The other is that they refer to real things — quarks, electrons, the gravitational force — that exist independently of us and our sensory equipment. Scientists themselves tend not to worry about these problems and if pressed might regard theories as compelling understandings of the world, which correspond with observations, fit in broadly with other theories, and make sense. {3}

Must science rest on strong logical foundations? Probably not. Much in quantum theory is contra-intuitive. {4} Randomness enters into relatively simple systems. {5} We deduce consequences from theories so as to check them. And we induce theories from observations, which Aristotle called generalizing. Scientific laws are often best expressed in mathematical form — giving them precise formulation and prediction — but mathematics does not rest on logic: the attempts last century by Russell and Whitehead ended in paradoxes, and the formalist approach of Hilbert was overthrown by Gödel's incompleteness theorem. {6}

The Problem of Induction

Many problems were noted long ago. How much evidence needs to be assembled before a generalization becomes overwhelmingly certain? It is never certain. David Hume (1711-76) pointed out that no scientific law is ever conclusively verified. That the sun has risen every morning so far will not logically entail the sun rising in future. Effect is simply what follows cause: laws of function are only habit. {7}

There are further difficulties with induction. Scientists make a large number of observations from which to generalize. But these observations are made with a purpose, not randomly: they are selected according to the theory to be tested, or what the discipline prescribes as relevant. Then the eye (or any other organ) does not record like a camera, but interprets according to experience and expectation. Theory is to some extent threaded into observation. Finally, there is the reporting of observations, which must be assembled, regimented in accordance with the theory being advanced or refuted.

Does this worry scientists? Not at all. Whatever the philosophic difficulties, science works, and its successes are augmented every day. Besides, the problem can be circumvented by employing statistical relevance. We assemble the factors that might be relevant and see how probability changes as a result. For example: if the probability of Event E given Cause C is changed by Factor A, then A is relevant — matters which can be set out in probability theory. {8}

Karl Popper: The Falsifiability Thesis

But if induction is the weak link in science, why not remove it altogether? Science, claimed Karl Popper (1902-94), proceeds by guesses that are continually tested, i.e. by conjectures and refutations. {9} That is the real essence of science, not that its conclusions may be verified, but that they can be refuted. Metaphysics, art and psychoanalysis can not be so falsified, and they are therefore not science. {10}

An admirable distinction, but is it true? Are scientific theories really formulated so as to expose their potential grounds of weakness? Are they ditched when contrary evidence appears? And is the scientific enterprise conducted this way? The answer to all three questions is generally no.

How Science Works

The first point to be emphasized is the diversity of science. All sciences are objective and empirical, presenting results that can be independently verified by a qualified practitioner. But each discipline in practice, and sometimes each sub-discipline, has its own traditions, ethos and procedures. And these in turn are the product of long training and a communality of views, even to some extent of mentalities: good botanists do not make good astrophysicists.

Then, contrary to Popper's view, most biologists and psychologists do believe that animals form mental pictures, practical working models to guide their activities. They may suffer surprises, and have to adapt and extend their models, but no living organism could survive a flux of continuous uncertainty. {11}

Are scientists objective, carefully considering theories on the basis of evidence, and that alone? Only to some extent. Scientists are human, and their work is fuelled by their interests, career needs and animosities like everyone else's. {12} But independence is claimed for the end product. The scientific paper may not represent the twists and turns of thought and experiment, but aren't the final results objectively presented, earlier workers acknowledged, and arguments for acceptance soberly marshalled? Not really. Papers do not let the facts speak for themselves. The evidence is persuasively presented: there is a rhetoric of science. {13} Papers are refereed, and maverick views excluded. Vetting by peer-groups discounts or expunges work that starts from different assumptions or comes to fundamentally unsettling conclusions, as Velikovsky {14} and Gauquelin {15} both found.

Kuhn's View: The Scientific Paradigm

Science, postulated Thomas Kuhn, employed conceptual frameworks, ways of looking of the world which excluded rival conceptions. These paradigms, as he called them, were traditions of thinking and acting in a certain field. They represented the totality of background information, of laws and theories which are taught to aspiring scientist as true, and which in turn the scientist has to accept if he is to be accepted into the scientific community. Scientific enterprise is conservative. The paradigm legislates. What lies outside its traditions is non-science. And for long periods science proceeds quietly and cumulatively, extending and perfecting the traditions. Anomalies, even quite large anomalies, are accepted for the sake of overall coherence. But when the anomalies become too large, and (crucially) make better sense in a new paradigm, there occurs a scientific revolution. The old laws, the terminology and the evidence all suddenly shift to accommodate the new paradigm. {16}

Kuhn propounded his theory by referring to the history of science, and that history has been much worked over since. The jury is still out. Working scientists are largely happy with Popper, but historians and philosophers of science are less so. Popper didn't deny that much of science proceeded mechanically, but argued that this was bad science. To some extent scientists do act in the appearances-saving manner that paradigms portray, but change is more gradual than Kuhn supposed. Mary Hesse in particular examined the part played by analogy, metaphor and imagination in the creation of theories — none reducible to a method — and was familiar with continental hermeneutics. But science was different, she concluded, though there is space in our mental activities for art and religion. Religious and scientific cosmologies are "collective representations". {17} The controversy has been bitter, but largely restricted to the esoteric areas of theory. Working scientists still broadly see themselves as extending the boundaries of knowledge and are very disinclined to engage in Jesuitical debates on philosophical matters. {18}

Imre Lakatos

The second challenge to Popper came from Imre Lakatos, who grouped theories into "research programmes" and made these the deciding mechanism. Each such programme possessed a hard core of sacrosanct information established over a long period of trial and error. Round the core was a protective belt of auxiliary hypotheses and observations that were being constantly tested and modified. Programmes guided scientists in their choice of problems to pursue, and were attractive (progressive, Lakatos called them) to the extent that they accumulated empirical support and made novel predictions. Above all, programmes protected scientists from inconvenient facts and confusing observations — necessarily, or many eventually successful theories would have been strangled at birth.

Though the auxiliary belt served to protect the research programme core, and was constantly being modified, these modifications could not be made ad hoc, devised simply to get round a particular problem. They had to be falsifiable: Lakatos agreed with Popper that sociology and psychoanalysis were unscientific on this basis. But how is the progressive research programme to be distinguished from the degenerating one, except by hindsight? Kuhn accepted a leap of faith, an intuitive feel for where the future lay, but Lakatos did not. {19}

Paul Feyerabend

Paul Feyerabend initially {20} won a considerable reputation as an historian of science prepared to get down to precise scientific detail. He was a realist in the Popperian sense, and argued that science progressed through proliferating theories, rather than coalescing into a prevailing Kuhnian paradigm. Subsequently, to the horror of colleagues and friends, he took a sociological and anarchistic line, arguing that true science was being stifled by the scientific establishment, an institution as self-serving and undemocratic as the medieval Church. {21} Orthodox medicine, for example, tries to put obstacles in the way of alternative medicine, regardless of the facts. There are no methodologies, he claims, and indeed "anything goes". Feyerabend has been scathing of the philosophy of science, remarking that "almost every journal in the philosophy of science deals with problems that are of no interest to anyone except a small gang of autistic individuals."


Kuhn's views, and more particularly Feyabend's, were seized upon as evidence that the scientific world-view was simply one paradigm amongst many. Despite its prestige and practical triumphs, science was as much a myth as art or literature or psychoanalysis. Kuhn hotly denied this, and backtracked very much from his earlier position. Both he and Popper were dismayed to see their views hijacked by the relativists, as support for the view that each person makes his own reality or concept of truth. {22} Relativism is disliked by philosophers, and the refutation is straightforward. If something is true only within a confined system — one world-view, one person's consciousness — how are we to know whether this has any currency in time or space? Even to record our observations needs a language, and languages cannot be wholly private. {23}

But non-relative truths also have their problems, most notably with language, which is not transparent or logically consistent, as difficulties with meaning all too readily show. We have different conceptions of what we call "truth", moreover, and for Rorty truth is not a property common to true statements. {24} But if Rorty is a maverick despairing of traditional philosophy, Margolis has argued for a relativism involving three-value logics, though not supplied convincing applications. {25} Goodman suggested that artists construct their own worlds, which are "true" when they offend no unyielding beliefs and none of their own precepts. We accept them not by their correspondence to reality, but "rightness of description" — which leads to the question of rightness by what standard, and the usual paradoxes of relativism. {26}

Some Concluding Thoughts

Those who attack science for its remote and reductive nature, its cold-blooded efficiency and elitist decision-making should not forget how well science actually works. Scientific observations may be theory-laden, but those theories are tested in a communality of practice. If once depicted as mechanical and predetermined, science appears less so now that quantum and chaotic processes have been more widely recognized. Science does bring great operational efficiency, and its findings cannot be called myths in the sense understood in anthropology or literary criticism. {28} Science attempts not only to understand nature, but to control nature, and there is hardly an aspect of life today that could be conducted without its help. In short, science does seem essentially different from the arts, and its successes would be miraculous if there was not some correspondence between its theories and "reality", whatever that "reality" may be.

This and other pages in the theory section have been collected into a free pdf ebook entitled 'A Background to Literary Theory'. Click here for the download page.


1. Stewart Richards's Philosophy and Sociology of Science: An Introduction (1983), Chapter 5 of John Passmore's Recent Philosophers (1985), Robert Matthews's Unravelling the Mind of God: Mysteries at the Frontiers of Science (1992), and Chapters 3-5 of Stuart Brown, John Fauvel and Ruth Finnegan's (Eds.) Conceptions of Inquiry (1981). All three have good bibliographies.
2. Many recent publications. J. Briggs and F.D. Peat's The Turbulent Mirror (1989) and R. Lewis's Complexity (1993) are readable introductions.
3. pp. 11-2 in Richards 1983.
4. Quantum physics is covered by many textbooks and popular accounts. P. Da vies and J. Grubbing's The Matter Myth is a handy summary.
5. Briggs and Peat 1989, and Lewis 1993. Also Ilya Prigogine and Isabelle Stengers' Order Out of Chaos (1984).
6. Laymen should refer to the many popular accounts of modern mathematical thought — e.g. John Allen Paulos's Beyond Numeracy: An Uncommon Dictionary of Mathematics (1991), and Ian Stewart's The Problems of Mathematics (1987). Morris Kline's Mathematics and the Search for Knowledge (1986) is more advanced. See pp. 148-52 in Bryan Bunch's Mathematical Fallacies and Paradoxes for a kindergarten introduction to Gödel's theorem.
7. Hume's scepticism is well covered by introductory philosophy texts. See also the Hume, David entry in Ted Honderich's (Ed.) The Oxford Companion to Philosophy (1995).
8. Kenneth Sayre's Cybernetics and the Philosophy of Mind (19766).
9. Karl Popper's The Logic of Scientific Discovery (1959), Conjectures and Refutations (1963) and his Realism and the Aim of Science (1983)
10. Chapter 1 of Popper 1963.
11. Chapter 5 of Walter Mischel's Introduction to Personality (1981), and pp 412-37 in Philip Zimbardo, Mark McDermott, Joroen Jansz and Nico Metaal's Psychology: A European Text (1993).
12. Chapter 9 in Richards 1983.
13. Alan Gross's The Rhetoric of Science (1990).
14. M.J. Mulkay's Science and the Sociology of Knowledge (1979).
15. Percy Seymour's Astrology: The Evidence of Science (1988).
16. Thomas Kuhn's The Structure of Scientific Revolutions (1962) and his The Essential Tension (1977).
17. Mary Hesse's Truth and the Growth of Scientific Knowledge (1976) and her Truth and Value in the Social Sciences (1978).
18. pp. 12-13 and 198-201 of Richards 1983.
19. Imre Lakatos and Alan Musgrave's (Eds.) Criticism and the Growth of Knowledge (1970).
20. Paul Feyerabend's Philosophical Papers (1981).
21 Paul Feyerabend's Against Method (1975) and his Science in a Free Society (1978).
22. pp. 92-97 in Passmore 1985.
23. Relativism, epistemological entry in Honderich 1995.
24. p. 121 in Passmore 1985.
25. Joseph Margolis's The Truth About Relativism (1991).
26. Nelson Goodman's The Languages of Art (1968).
27. Chapter 18 of John Losee's Philosophy of Science (1993).

Internet Resources

1. Identity and Individuality in Quantum Theory. Steven French. Feb. 2000. Metaphysical implications of quantum physics.
2. Ten Myths of Science: Re-examining What We Think We Know. W. McComas 1996. Some popular misconceptions.
3. Dispelling Some Common Myths about Science. Terry Halwes. Mar. 2000. Several linked articles, with references.
4. How Science Works. David L. Hull. 1992. essays/v15p147y1992-93.pdf. Review of The Scientific Attitude by Fred Ginnell.
5. How Science Works. David Goldstein. sciman0d.pdf/$file/sciman0d.pdf. Standard but sensible account.
6. Fashionable Nonsense: Postmodern Intellectuals' Abuse of Science. Alan Sokal and Jean Bricmonts. 1998. First chapter of their book, and review.
7. Planetos. Review of Gauquelin controversy.
8. The Limits of Science. Serghey Stoilov Gherdjikov. 1998. A view similar to Cassirer's.
9. Scientific Revolutions. Henry Folse. 2002. Detailed notes taking this page much further: excellent.
10. Sir Karl Popper (1902-1994). Kelley L. Ross. 2003. Popper's work and beliefs.
11. Popper, Karl. A fuller account of Popper's contribution.
12. Karl Popper. Routledge site with brief biography, articles and links.
13. Karl Popper. Stephen Thornton. Oct. 2002. Extended article in the Stanford Encyclopedia of Philosophy.
14. Outline of Kuhn's Structure of Scientific Revolutions: Normal Science. Henry Folse. 2002. Handy summary.
15. The Structure of Scientific Revolutions. Thomas Kuhn. 1962. philosophy/works/us/kuhn.htm. Excerpt: first chapter and postcript.
16. Guide to Thomas Kuhn’s The Structure of Scientific Revolutions. Malcolm R. Forster: Mar. 1998. A critical synopsis.
17. Thomas Kuhn. Frank Pajares. Detailed biography and excellent listings.
18. Imre Lakatos. Jan. 2004. Introduction, with in-text links.
19. Imre Lakatos. J J O'Connor and E F Robertson. A tribute.
20. Science as Successful Prediction. Imre Lakatos. 1970. Excerpt from his Criticism and the Growth of Knowledge.
21. Lakatos. Biography, article and brief listings.
22. Prediction and Accommodation in Evolutionary Psychology. Malcolm Forster and Lawrence Shapiro. Lakatos's theory in action. (Many papers of interest on this site.)
23. Lakatos and MacIntyre on Incommensurability and the Rationality of Theory-change. Robert Miner. 1998. A critique of Lakatos: somewhat technical.
24. Reconstructing Lakatos. Matteo Motterlini. 1978. pdfgen/motterlini56.pdf. Lakatos's position after the debate with Feyerabend.
25. Paul Feyerabend. Fe. 2004. Useful introduction.
26. Paul Feyerabend. Against Method. 1975. philosophy/works/ge/feyerabe.htm. Sketch of the main argument and the concluding chapter from Against Method.
27. Paul Feyerabend. John Preston. May 2002. Stanford Encyclopedia of Philosophy entry with good bibliography.
28. Richard Rorty. Stanford Encyclopedia of Philosophy entry, with good references.
29. Radical Changes in Aesthetics. Joseph Margolis. Shortened version of Interpretation Radical But Not Unruly: The New Puzzle of the Arts and History, 1995.
30. Vim and rigor: the work of author Nelson Goodman. W.J.T. Mitchell. May 1999.
. Introduction to range of Goodman's interests.
31. 'Languagues of Art. An Approach to a Theory of Symbols' by Nelson Goodman. Sep. 2000. Critical review by Stefan Beyst of Goodman's 1976 book.

      C. John Holcombe   |  About the Author    | ©     2007 2012 2013 2015.   Material can be freely used for non-commercial purposes if cited in the usual way.