An Answer to Prof. S. Watanabe’s Paper titled “NEEDED: A Historico-Dynamical View of Theory Change”
Up until now, I have published papers and books on the theory of “theory change” in science, all of which were written in Japanese. Here, I will present to English readers an outline of my theory on the basis of a part of my paper appeared in KAGAKUSHI (History of Chemistry), 1988, pp.185-190.
Ken-ichi KARAKIDA
2021年8月13日追記：この記事の日本語訳は8月4日付の当ブログ（”渡辺慧教授の論文「求む：理論変化の歴史的・動的見解」に答える”）に掲載しました。
1. INTRODUCTION
Prof. Satosi Watanabe published a paper[1] having a title “Needed: A Historico-Dynamical View of Theory Change”, where he criticized the most authors who had been discussing the problem of theory change. He chose mainly Thomas S. Kuhn[2] as the target of his criticism, but he added that many of his objections could be addressed equally well to some other authors, including some of Kuhn’s critics. According to Watanabe, the disputants seem to have had forgotten that they have been discussing the actual theory changes in history. He, then, specified four separate, but interrelated, complaints about the discussions of the subject.
2. Watanabe’s Four Points
(1) One-wayness of Change
Kuhn’s theory does not reflect any trace of one-wayness of history. When A and B are two successive theories, the most important aspect of the relation of A to B: f(A,B) should be characterized by its asymmetry: f(A,B)≠f(B,A). Any theory which does not take seriously this asymmetry does not reserve the name of a historical study of theory change. Kuhn compares theories A and B as if the two were just alternatives of equal validity. Although he abruptly introduced the idea of progress in the last chapter of his book, this caused a direct contradiction to what he preached in the preceding chapters.
(2) Change and Continuity
Kuhn claims that successive theories are incommensurable. He compares two theories, new and old, to two languages and stresses that languages are basically untranslatable. However, in spite of what philosophers of language and linguists may say, the existing human languages are mutually translatable to a degree that is perfectly useful in ordinary human life. If we force the analogies of languages to consecutive theories in science, they should not be compared to, say, Hopi and English, but to adult’s English and child’s English. The language of children is translatable into that of adults, but not vice versa. Similarly, the language of Newtonian mechanics is perfectly understandable in, and derivable from, the language of Einsteinian mechanics, but not vice versa.
(3) Dynamics of Change
Kuhn emphasizes the role of anomaly or crisis that causes the scientific revolutions. However, if one follows his line of thought, one cannot explain why sometimes a scientific revolution (theory change) happens without actually some empirical facts being discovered to be in conflict with the old theory. In addition, he is completely silent about the influence of external factors on the production of crises and in some cases also on the concept formation. These defects reveal that Kuhn’s theory lacks a basic historico-dynamical understanding of the process of theory development.
(4) Uniqueness of Theory
Kuhnians believe that there are many alternative theories possible to cover the same field of experience. However, those who face the actual history of science with intellectual honesty discover that in reality the one theory which has been adopted by history is practically the only theory possible. For instance, given the domain of mechanics usually covered by special relativity, can anybody produce an alternative theory? The Kuhnian theory cannot cope with this simple historical fact.
3. About the Scheme of Theory Change A→B
In the scheme of theory change A→B, where B is a new-born theory, it should be noted that theory A can mean a combination of existing theories as well as an existing theory. About the Copernican revolution, theory A is a Ptolemaic system, i.e., typical A→B. In the case of special relativity, Einstein combined Newtonian mechanics (Galilei transformation) with Maxwell’s electromagnetic theory in order to study “the electrodynamics of moving bodies”. As the result, Newtonian mechanics underwent significant change, while Maxwell’s theory remained unchanged. The problem of black body radiation, which finally generated quantum mechanics, needed various disciplines including thermodynamics, statistical mechanics, wave theory, electromagnetics, etc. as well as Newtonian mechanics. The establishment of quantum mechanics brought great changes on all of those disciplines.
4. Newly Proposed Theory of Theory Change
A key factor of theory change A→B is internal contradictions (inconsistencies) found in theory A. That is, new theory B is generated by surpassing the contradictions in theory A. In other words, being different from the claim of Kuhn, A and B are not two competing theories, but B comes from A.
Copernicus was induced to his system by the knowledge that Mathematicians (Ptolemaic astronomers) are inconsistent in the investigations[3]. In the case of special relativity, Einstein found the “asymmetry” (inconsistency) in the theoretical interpretation of Maxwell’s electromagnetics when it was applied to moving bodies. The wave-particle duality (contradiction) is well-known in the establishment process of quantum mechanics.
Observed data contribute to the theory change only when they bring about contradictions in the theory through some theoretical treatment. Einstein’s theoretical interpretation on the photoelectric effect (“the light quantum hypothesis”) caused contradictions in existing theories. In contrast, the anomaly of perihelion precession of Mercury was known since mid-19th century, and general relativity (1915) of Einstein revealed that the phenomenon was beyond the scope of Newtonian mechanics, but the anomaly itself didn’t cause the theory change.
5. Relation between Theories A and B: “Semi-incommensurability”
The theory change is a jump from A to B, and
(a) Theory A constitute a framework of the jump. However,
(b) Theory B can never be deduced from nor reduced to theory A. On the other hand,
(c) The real significance of Theory A is first given by theory B as the ultimate significance of Newtonian mechanics was first comprehended in the light of special relativity and quantum mechanics.
I call this relation “semi-incommensurability”, i.e., there is a logical gap from A to B, but A can be understood with B.
I derived the above relation from the study of the theory change in science. However, afterwards I found that the relation is equivalent to the structure of Project (Pro=jet in French) of Jean-Paul Sartre[4]. I found further that the relation is also exactly the same with “life’s irreducible structure” of Michael Polanyi[5]. Thus, the relation “semi-incommensurability” between A and B has more general significance than initially considered.
6. An Answer to Watanabe
The relation of “semi-incommensurability” answers to both points of “(1) One-wayness” and “(2) Change and Continuity” of Watanabe (see section 2). That is, there is a logical gap from A to B, but A can be understood with B. This also means that the relation of A to B is characterized by its asymmetry.
The reason why sometimes a scientific revolution happens without actually some empirical facts being discovered to be in conflict with the old theory〔the point of “(3) Dynamics of Change”〕 is explained in section 4 in the case of special relativity as an example. Einstein found the contradiction theoretically.
According to my theory, a key factor of theory change A→B is internal contradictions found in theory A. Therefore, the discoverer who contributed to the theory change belonged to the old theoretical framework. This explains the dynamics of change. A theory change is a theoretical problem, and the biggest theoretical problem for a theory is the presence of its internal contradictions, which is serious especially for the person who is deeply committed to the theory. In addition, a fundamental contradiction has the nature of becoming more and more obvious in the trials to resolve it. This motivates the change of the theory.
The influence of external factors is essentially understandable in Sartre’s structure of Project (the pyramid of a hierarchized multiplicity of significations). However, I don’t discuss it further here.
As for “(4) Uniqueness of Theory”, if there were two independent theories possible to cover the same field of experience, it only means that contradictions are present in the theoretical field which should be surpassed.
7. Concluding Remarks
From my point of view, although there are no a priori criteria for the evaluation of theories, the “asymmetry” between the theories A and B or the progress in the theory can be recognized. In the process of history, we have no “objective marks” which assure us of the correctness of our choice, but the result of the choice is significant. According to the concept of Project, a person is defined by his/her choice in the situation. This is also the focus point of our research on science history.
K. Karakida’s Publications on Theory Change (in Japanese)
Papers:
KAGAKU-KISORON-KENKYU, 16, No.3 (1983), pp.17-21.
KAGAKUSHI, 1985, pp.186-192; 1988, pp.185-190; 16 (1989), pp.49-54; 27 (2000), pp.169-175; 28 (2001), pp.171-174; 31, (2004), pp.215-224.
Books:
Riron no Sôzô to Sôzô no Riron, AsakuraShoten (1995).
K. Nishimura et al. ed., Bunsûga dekinai Daigakusei, TôyôkeizaishinpôSha (1999), pp.37-58; ChikumaShobô (2010), pp.61-85.
[1] S. Watanabe, Synthese, 32 (1975), pp.113-134.
[2] T. S. Kuhn, The Structure of Scientific Revolutions, The University of Chicago Press (1962, 1970, 1996).
[3] The Preface of Copernicus’ De Revolutionibus Orbium Caelestium (1543). See also K. Itakura, Kagaku to Hôhô, KisetsuSha (1969), pp.81-133 (in Japanese).
[4] J.-P. Sartre, Critique de la raison dialectique (pécédé de Question de méthode), Tome I, chapter III, Librairie Gallimard (1960).
[5] M. Polanyi, Knowing and Being (ed. by M. Grene), The University of Chicago Press (1969), 14.