Previous part

5.3.3 Example from Physics: Nature and justification of knowledge. A `test case'[34]

As explained in earlier chapters, Polanyi's concept of Personal Knowledge was presented as a corrective to a deterministic concept of science represented in Polanyi's mind by the positivist-empiricist tradition descended from Mach. As part of the corrective project, Polanyi redefined the meaning of `objective' as well as the meaning of `true' in the context of knowing, infusing both with a personal element.

To review Polanyi's re-definitions:

Knowing is objective in the sense of establishing contact with a hidden reality,[35] while the knower is claiming universal validity (this is universal intent, not universality) for this contact which is the condition for anticipating an indeterminate range of implications.

Truth, for Polanyi is a blending of coherence of a conceptual framework in which we believe, and the claim that what we believe corresponds to reality. This statement can be understood in light of the previous discussions on (1) Polanyi's analysis of `statements of fact' in `The Logic of Affirmation' as consisting of a tacit component of the act of assertion and the explicit component of content, and (2) his conception of `two poles of knowledge,' the personal pole of belief or claim of truth and the external pole of truth itself. The notion of reality, it will be remembered, is tied to the notions of truth and beauty.

Polanyi used examples of discovery from the history of physics to redefine all of the above notions. He interpreted the discovery process as `personal knowledge' activated by Intellectual Passion.

In response to Polanyi's book Personal Knowledge, in which these redefinitions were published, Adolf Grünbaum representing the positivist position attempted to invalidate Polanyi's alternative position in the philosophy of science by leveling charges against two aspects which any epistemological foundation of a philosophy of science must deal with: the nature and justification of knowledge.

For purposes of this `test case,' Grünbaum is taken as representative of the position Polanyi's alternative intended to correct. Grünbaum's position is what Scheffler would call the "standard view" of science,[36] a view which

affirms the objectivity of science ... understands science to be a systematic public enterprise, controlled by logic and empirical fact, whose purpose it is to formulate the truth about the natural world. ... Observation ... supplies the particular empirical facts, the hard phenomenal data which our law-like hypotheses strive to encompass .... [The standard view holds to] the idea that observation supplies us with hard data independent of our conceptions and assertions, data by which, indeed, our conceptions and assertions are controlled. ... [On] the standard view, people with different theoretical beliefs may observe the same things; shared access to a common world is taken for granted.[37]

The basic tenet of the `standard view' is that data is independent of our conceptions, thus fallibility, the cause of error, can be uncertainty of observation or failure of a theory to predict correctly.

Laws or general hypotheses are of two kinds:[38] theoretical laws and observational or experimental laws. (a) Theoretical laws are expressed in abstract language and postulate unobservable elements and functions; they cannot be tested by direct experiment. The function of these laws is to explain observational or experimental laws. (b) Observational or experimental laws generalize data from sense-experience, can be tested directly by inspection or experiment. These laws are expressed in observation language and refer to perceived things and processes.

According to the above distinction between the two kinds of laws, in the `standard view,' fallibility can be dealt with by testing experimental laws, and on the basis of these, adjusting the higher-level theoretical laws.

Grünbaum made two charges against Polanyi's epistemology of Personal Knowledge: (A) on the nature of knowledge --- that Polanyi used intuitionism as an explanation for the discovery process in science, and (B) on the justification of knowledge --- that Polanyi lacks a warrant for deciding whether a theory is true.

5.3.3.A The nature of knowledge

Grünbaum made the following remarks, charging Polanyi's epistemology with `intuitionism' and aprioristic rationalism:[39]

Polanyi seems to believe, said Grünbaum, that he can discount known experimental results in the genesis of Einstein's discovery of Relativity Theory,[40] claiming that Einstein discovered rationality in nature, intuited relativity theory unaided by any observation that had not been available for at least fifty years, and that positivistic textbooks falsely portrayed relativity as a theoretical response to the Michelson-Morley experiment.

Both Polanyi and Grünbaum are referring to Einstein's Autobiographical Notes in which Einstein said on his discovery process:

After ten years of reflection such a principle resulted from a paradox upon which I had already hit at the age of sixteen: If I pursue a beam of light with velocity c (velocity of light in a vacuum), I should observe such a beam of light as a spatially oscillatory electromagnetic field at rest. However, there seems to be no such thing, whether on the basis of experience or according to Maxwell's equations. From the very beginning it appeared to me intuitively clear that, judged from the standpoint of such an observer, everything would have to happen according to the same laws as for an observer who, relative to the earth, was at rest.[41]

Grünbaum embedded this quote in a larger one, tracing Einstein's search for a general principle which he despaired of finding by `constructive efforts based on known facts,' looking at an example of a law (of thermodynamics).

It will be noted that the germ of a solution was already contained in the paradox (and was intuitively clear to Einstein) --- a recognition of the arbitrariness of the axiom (of the absolute character of time) was the clue to the solution of the problem. Einstein's reasoning was aided, as he said in the Autobiography, by studying the critical writings of David Hume and Ernst Mach. But Einstein has not given a step-by-step account of his process; it is Grünbaum's project to reconstruct what he considers to be the genesis of the STR (Special Theory of Relativity), that is, what Einstein needed to write his 1905 proof: the `genesis of the STR' is assumed to be the reasoning in the writing of the proof.

Grünbaum believes that the solution of the paradox (the absolutistic conception of time) needed the proof of the truth of two assumptions;[42] there is very little information offered on these by Einstein, but Grünbaum assumes Einstein used them. Einstein's Autobiographical notes also offer very little information on his actual use of experimental results in formulating the principle of relativity, and this is admitted by Grünbaum.[43] Einstein himself said to Polanyi[44] that the (Michelson-Morley) `experiment had negligible effect on the discovery of relativity.'

Grünbaum, attempting to forestall objections that his reconstruction of the genesis of STR is showing a mere codification of known elements (and not a great discovery), asks:

Can the history of the RT [Relativity Theory] be validly adduced here to prove any more than the untenability of a Machian or Aristotelian-Thomist abstractionist account of theory construction as a mere codification of the results of experiments? And is the untenability of that account not fully recognized by any empiricist conception of scientific knowledge broad enough to accommodate the valid core of the Kantian emphasis on the active, creative role of the scientific imagination in the postulational elaboration of hypothetico-deductive theories? Does such an empiricist conception not allow for the difference between knowing of an experimental result in a narrow sense, on the one hand, and speculatively assigning a wider significance to it, on the other? ... [The] answer ... is indeed `yes' ...[45]

As this last passage more clearly shows, Grünbaum defended his reconstructive effort of the discovery and genesis of STR by invoking Kant. But according to Polanyi, Grünbaum assumed that Einstein used a set of known principles and assumptions he could prove to be true, together with data from experiments proving the validity of certain assumptions as part of his `theory construction.' Grünbaum assumed that such a constructive effort is consistent with an active, creative role of the scientific imagination (elaboration), and that at the same time the discoverer can `speculatively assign wider significance to experimental results.' But is not `speculatively assigning' a psychological act which Grünbaum himself designated as outside the boundary of epistemology?[46]

Although Grünbaum disclaims reconstructing an add-summative or rule-bound procedure for Einstein's discovery process, his focus on `elaboration' and on `speculatively assigning a wider significance [to experimental results] does not agree with Einstein's own account. Grünbaum's reconstruction is of the proof of the solution, not of the discovery context, not of the problem-posing.[47] So what could Grünbaum mean by his claim? That his re-creation of the `genesis of RT' takes into account creative speculation in reaching the solution? Does he consider hypothesis formation outside the province of the genesis of a theory, i.e. of psychological interest only? If he restricts himself to H-D type explanations, he will not be able to explain, as N. R. Hanson has shown, how hypotheses are generated.

Polanyi believed that scientific discoveries are not made either by generalization from induction, nor by deduction from hypotheses, neither are they `constructed':

Specific rules of empirical inference claim (a) to proceed by a prescribed operation from clues to discovery or at least (b) to show how to verify, or at the very least (c) how to falsify an empirical proposition according to some such rules. Claim (a) must be rejected in view of the demonstrable fact that discovery is separated by a logical gap from the grounds on which it is made. It is, ... a travesty of the scientific method to conceive of it as an automatic process depending on the speed of piling up evidence for hypotheses chosen at random. The history of the great scientific controversies teaches us now that claims (b) and (c) are equally unfounded. The reasons are ... [that all] formal rules of scientific procedure ... will be interpreted ... according to the particular conceptions about the nature of things by which the scientist is guided.[48]

N.R. Hanson would agree. As he has shown in the essay `Is there a Logic of Scientific Discovery?'[49] both induction and deduction are partial explanations for how discoveries are made.[50] Hanson calls. The hypothetico-deductive (H-D) type explanation Grünbaum employs, "Logic of finished research report."[51] The H-D method, of which Grünbaum speaks, is according to Hanson the stage of proof.

Commenting on the alleged birth of theories by the H-D process, Hanson remarks:

A hypothetico-deductive system ... describes physical theory more adequately than did earlier accounts in terms of induction by enumeration, for it says what laws are, and what they can do, in the finished arguments of physicists. But it does not tell us how laws come by in the first place; ... [Induction by enumeration] is false. Physicists rarely find laws by enumerating and summing up observables. There is also something wrong with the H-D account ... Physicists do not start from Hypotheses; they start from data [phenomena, explicanda]. By the time a law has been fixed into an H-D system, really original physical thinking is over.... This H-D account is helpful only when discussing the argument of the finished research report ....[52]

To paraphrase Hanson, the H-D account helps to see how experiments can elaborate theory --- but it does not lead to formulation of a hypothesis.[53]

Polanyi disputed Grünbaum's use (in the quote on STR above) of "Kantian emphasis on the active, creative role of the scientific imagination in the postulational elaboration of hypothetico-deductive theories," on the grounds, that Kantian categories have "nothing to do with `speculatively assigning a wider significance' to experience."[54] Polanyi's use of the notion of `scientific imagination' is similar to Hanson's: "The critical moment comes when the physicist perceives that one might reason about the [surprizing] data in such and such a way."[55] The emphasis is on `perceives.'

As the above discussion suggests, the disagreement on what constitutes `genesis of a theory' is deep: while Polanyi speaks of a discovery as the scientist's sudden insight relying on his understanding of the nature of things, Grünbaum speaks of a deliberate construction of a theory.

As Polanyi remarked later, there is a predominant false notion of heuristics as an explicit procedure, a notion which ignores the actual procedure of how ideas are brought up. The false notion of heuristics, according to Polanyi, declares "that rational enquiry [into a problem] is supposed to start with the presence of a hypothesis, which we proceed to test by experience and retain until it is refuted by experience."[56] According to this false notion, heuristics is `to be used for' problem-solving in the sense of explicitly reasoning out a proof which undergirds a theory. In contrast to this interpretation, Polanyi says: "A problem is ... a partial comprehension of a coherent entity, filled with striving toward its completion."[57] The `heuristic` in Polanyi's sense, is the `striving toward completion,' a tacit component. `Heuristic' in this sense implies that we have an insight into what is a plausible hypothesis,[58] a most important act in the advancement of science.

What Polanyi calls the `false notion of heuristics,' is an interpretation of Popper's falsifiability criterion of demarcation for the validity of a hypothesis,[59] that is, the notion of exposing the theory to tests in every conceivable way so it may be refuted by experience. Falsifiability, Polanyi said, is an inappropriate concept applied to the problem-posing aspect of hypothesis generation.[60] Applied to the whole proces of scientific inquiry, it gives a false picture of the advancement of science as the aggregate of hypotheses that have not been refuted by experience.

This picture might claim to present us with a strict procedure for the advancement of science, if strict criteria existed for refuting an hypothesis by adverse evidence. But even if such criteria existed, this procedure would be useless, for unless we have a method for producing hypotheses that are likely to be valid, the chances of discovery would be so slight as to make the procedure impracticable. Such a method does not exist and discoveries are in fact not made by discarding a host of bad ideas, but by having a few good ideas.[61]

The `few good ideas' come from developed talent, an ability enhanced by education.[62] According to Polanyi, talent is a hightened endowment of aspects rooted in evolution --- recognizing patterns, an assumption of uniformity of nature --- and involves an expectation of (reaching) an end.[63] Grünbaum is attempting to see the genesis of Einstein's STR as a reconstructible process, which according to Polanyi's above quoted notes, is `supposed explicit heuristics.' The problem, of course, remains for Grünbaum: how do scientists come to the `having a few good ideas'? Grünbaum's reference to `the valid core of the Kantian emphasis on the active, creative role of the scientific imagination' is not manifest in his reconstruction of the STR.

Polanyi objects, that several factors accounting for creativity are missing in Grünbaum's account. Polanyi in his response to Grünbaum remarked:

My view of scientific discovery and the verification of holding of scientific knowledge can be identified by the following features:
(1) Knowledge of the external world is in general acquired by relying on clues which cannot be fully identified. ...
(2)Discovery differs from routine investigation. While a survey is conducted by explicitly established procedure, discovery cannot be arrived at by the intelligent application of any explicitly established rules of inference. All existing rules of discovery are vague and ambiguous. ...
While they offer useful hints, any attempt to rely on them strictly could merely produce absurdities. The impossibility of formulating adequate rules for discoveries is expressed by ascribing originality to the discoverer. ...
(3) Even the highest degree of originality can operate only by taking for granted some element of an existing interpretive framework. In science, these elements form the premis and the method employed by the scientist. They define his general purpose, inspire his heuristic vision, and eventually bring forth his problem. A good problem is a passionate intimation of a hidden truth; to hit on it is a major scientific achievement. It is impossible to pursue discovery without the capacity to sense our approach to a hidden truth. ...
(4) I recognize the continued operation of these anticipatory powers of verification of discovery and holding of knowledge. The scientist's conviction of having arrived at some true knowledge is akin to the powers by which he recognizes a problem. ... It is absurd to suggest that we should recognize truth by [its fruitfulness]; on the contrary, our recognition of a true piece of knowledge is an anticipation of ... unknown consequences ... [the anticipations] are an expression of the belief that true knowledge is an aspect of hidden reality which ... can yet reveal itself .... I cannot see the relevance of Dr. Grünbaum's observations to my account of the discovery of relativity .... He thinks that I am completely discounting experimental results in the genesis of [STR] whereas I only mean that not all [Einstein's] clues can be specified nor his mode of evaluating them specified....[64]

As Polanyi and Grünbaum saw the debate at the time, neither understood the other. Grünbaum's project was to reconstruct theory generation as a propaedeutic to the science student.[65] He did this from the position of the `standard view' of science, assuming that the creative process can be made explicit. Polanyi saw this effort as lacking what he considered to be two key features of the discovery process --- (a) beauty as a clue to reality, which is part of heuristic vision, and (b) heuristic striving with its anticipatory power which bridges the gap from one conceptual level to a higher one. Both of these are central features in Polanyi's account of scientific discovery as personal knowing, and are aspects of Intellectual Passions. According to him, there is no significant scientific discovery without them. Both of these key features are the distinguishing features operating in the problem posing phase (in insight, in the `hunch') of scientific discovery. These features are considered to be outside the province of epistemology by Grünbaum, in agreement with what Polanyi labeled the `positivist position' originating with Mach.

According to Polanyi's Integrative Philosophy, beauty as a clue to reality can be said to be a Gestalt-perception feature of the `selective aspect' of Intellectual Passions, an initial `seeing through the paradox,' the insight Einstein had at age sixteen. Grünbaum's charge that Polanyi is an `intuitionist' is based on this aspect of Polanyi's interpretation of the context of discovery. The `heuristic striving' to that reality, which the discovery reveals is possible only as an aspect of Intellectual Passion, it is an existential choice of intentional action. This can be said to be the action-guiding feature. It is also the self-transforming act that starts with the problem posing, defines the creative act, and is the vector that reaches toward the solution. The `heuristic striving' is a personal element. This aspect Grünbaum considers subjectivist, a matter for psychology not for epistemology. It is the solution by explicit heuristics, according Polanyi, which Grünbaum wants to call a discovery and a proper concern for epistemology. An account of the nature of knowledge, Polanyi insists, must include the tacit `personal element.' According to Grünbaum, the nature of knowledge is demonstrated by explicit reasoning.

Next part

Notes

34. This section is based on my paper delivered at the 13th Annual Conference in the History, Philosophy, and Sociology of Science, Technology, and Medicine, held at Harvard University on 26-27 Febr. 1994. Stefania Jha, `Polanyi's Theory of Tacit Knowing: A Model of Integrative Philosophy.'

35. The personal pole is joined to the external pole (reality). At the personal pole is the `claim of knowing the hidden reality,' as well as `universal intent.' Briefly, for Polanyi, knowing is grounded in the premises of science (`tradition') and evaluated by the scientific community (`authority').

For grounds for truth conditions, Scheffler says the following: "The truth condition ... purports to reflect an objective commitment of ... [knowledge] attribution, evidenced indirectly in the way specific attributions are critically evaluated ..."; and, "if X is admitted to know, he must be judged not to be mistaken." Israel Scheffler, Conditions of Knowledge, (Chicago: The University of Chicago Press, 1965) pp. 21-24.

Polanyi emphasizes the assertion and certification of truth of a statement about reality (`I assert that P'), and de-emphasizes the concept of meaning of truth of a statement about reality (the statement `P' is true, and in fact P is true.) See `Assertion of Fact, The Logic of Affirmation', PK, pp. 253-255. For a discussion of this difference, see Israel Scheffler, Conditions of Knowledge pp. 52-53.

36. Israel Scheffler, `Objectivity Under Attack', in Science and Subjectivity (Indianapolis: Hackett Publishing Co., Second ed. 1982) p. 8.

37. Scheffler, ibid. pp. 8-14.

38. Paraphrased from Scheffler, `Objectivity' p. 8.

39. Adolf Grünbaum, `The Genesis of Special Theory of Relativity', in Feigl, H. and Maxwell, G. (eds.), Current Issues in the Philosophy of Science, (New York: Holt, Rinehart and Winston, 1961) pp. 43-53. Paraphrased. A fuller version of this article is the chapter `Philosophical Foundations of the Special Theory of Relativity, and Their Bearing on its History' in Adolf Grünbaum, Philosophical Problems of Space and Time. (Second ed., Boston: D. Reidel Publishing Co., Boston Studies in the Philosophy of Science, 1973, Vol XII).

40. For this theory, see for example, Harvey E. White, Modern Physics, 4th ed. (Princeton: D. Van Nostrand Co., Inc., 1962) pp. 380-385; and Albert Einstein, Relativity: The Special and General Theory, Transl. Robert W. Lawson, (New York: Crown Publ., Inc., 1961).

41. Albert Einstein, `Autobiographical Notes', in P.A. Schilpp (ed.), Albert Einstein Philosopher-Scientist (Evanston: The Library of Living Philosophers, 1949) p. 53.

42. For a detailed sequence of Grünbaum's reconstruction and the assumptions he claims were needed by Einstein to solve the STR, see Grünbaum, `Genesis', pp. 44-50.

43. Grünbaum, `Genesis' p. 49.

44. PK, p. 10.

45. Grünbaum, `Genesis', p. 51. [Italics in original]

46. Grünbaum, Philosophical Problems of Space and Time, p. 378.

47. On Einstein's own remarks on scientific thinking in the discovery process in general, see Albert Einstein, `Physics and Reality', Out of My Later Years (London: Thames and Hudson, 1950).

48. PK, p. 167.

49. Norwood Russell Hanson, `Is there a Logic of Scientific Discovery?' in Brody, B.A. (ed.), Readings in the Philosophy of Science (Englewood Cliffs, New Jersey: Prentice-Hall, Inc. 1970) pp. 620-633. Originally published in Feigl, H. and Maxwell, G. (eds.), Current Issues in Philosophy of Science. (Holt, Rinehart and Winston, Inc. 1961).

Compare this to Peter B. Medawar, Induction and Intuition in Scientific Thought (Philadelphia: American Philosophical Society, 1969), especially on the topics of experiments and induction.

50. In Polanyi's working notes (Polanyi Papers 23:14) dated Dec. 1963, Hanson's Patterns of Discovery, page 200 is referenced: (Summary) H-D theorists (1) discuss the reasons for accepting a hypothesis as true, and (2) never raise the question of discussing the rationale behind the proposal of hypotheses as possible explicantia [sic], which is pertinent to what makes us say a hypothesis is plausible.

51. Op. cit. p. 620.

52. Hanson, N.R. Patterns of Discovery: An Inquiry into the Conceptual Foundations of Science (Cambridge: Cambridge University Press, 1958) pp. 70-71.

53. The alternative notion, discovery as hypothesis generation by induction, as Hanson remarks, is also deficient - it cannot lead to generalizations which can be called a hypothesis. (See Hanson 1970. p. 626). The notion of induction by enumeration and summarizing observables sought to describe good reasons for initially proposing hypotheses. However, "there is something wrong. It is false." (Hanson, ibid.). H-D and induction are two ends of the spectrum, not alternative explanations of discovery, as both are deficient.

54. Michael Polanyi's letter to Adolf Grünbaum, dated 26, July, 1961. Polanyi Papers (6:1)

55. Hanson, 1958. p. 88. In his discussion on p. 85, Hanson uses Peirce's term `retroduction,' also called `abduction.' For this important concept, see Charles Hartshorne and Paul Weiss, eds., Collected Papers of Charles Sanders Peirce. (Cambridge: Harvard University Press, 1960), six volumes. Reference is to Vol. 6., p. 469.

56. Michael Polanyi, `Notes on supposed explicit heuristics; material for fourth lecture to be delivered in Chicago, 1967', Polanyi Papers (21:8).

57. Michael Polanyi, `Heuristics; Tacit Knowing, Part 2', Terry Lecture Notes, Summer, 1963, page 2. Polanyi Papers (35:10)

58. See Polanyi's notes on Hanson above.

59. Karl R. Popper, The Logic of Scientific Discovery (New York: Harper & Row, 1959) pp. 40-44. The `criterion of demarcation' problem was also tackled by David Hume. But what was known as `Hume's problem' is that of induction.

- According to I. Scheffler (private communication), Popper's falsifiability criterion was not intended as heuristic. My study shows that Polanyi's usage of the term `heuristics' is not consistent; he re-defined `heuristics,' as he did many terms, duly noted above, yet sometimes he shifts between conventional and newly coined meanings: for example, at times he uses Polya's definition of heuristics (as strategy to discover a solution), but usually he is using the term carrying the connotations elucidated above, as `striving.'

60. PK, p. 47.

61. cf. footnote 56

62. Michael Polanyi, `The Creative Imagination,' Triquarterly 8 (Winter 1967), p. 117. As Polanyi often said, no rules can be given to make a discovery - a specific discovery. Training in the scientific tradition is meant to give general guide-lines; Polanyi himself gave the general outlines of the mental process involved for an individual. His point is that originality means an element of freedom and unpredictability, a surprizse, as he noted in his writing on patents.

63. PK, Ch. 12, in reference to K.Z. Lorenz. It is worth noting, that Polanyi's notion of talent for hypothesis generation resembles Peirce's notion. See C.S. Peirce, `The Logic of Abduction', in Tomas, V. (ed.), Essays in the Philosophy of Science, (Indianapolis: Bobbs-Merrill Co., 1957).

64. Michael Polanyi, `Notes on Professor Grünbaum's Observations', in Feigl, H. and Maxwell, G. (eds.), Current Issues in the Philosophy of Science, (New York: Holt, Rinehart and Winston, 1961) pp. 53-55. [My italics.]

65. Grünbaum, ibid. p. 44.