کانت و فلسفۀ علم

by Michael Friedman, سعید جعفری (Translator)

کتاب حاضر گزیده‌ای است از مقالات مایکل فریدمن، استاد فلسفه دانشگاه استنفورد که به دوره‌ای از تاریخ و فلسفه علم مربوط می‌شود که از نیوتن آغاز و به اینشتین خاتمه می‌یابد. فریدمن با دیدگاهی نوکانتی به مسائل فلسفه علم می‌نگرد. اهمیت دستگاه کانتی در نظر فریدمن تا به آن حد است که هسته بنیادین مباحث زیربنایی فلسفه قرن بیستم را همگی، به نحوی از انحا به این دستگاه مرتبط می‌داند. در خصوص فلسفه علم فریدمن معتقد است بعد از همه انقلاب‌های علمی، که با هریک از آن‌ها مبانی علوم دقیقه به لرزه افتاده‌اند، آن‌چه هنوز پابرجاست، نگرشی است بازنگری شده از مفهوم پیشینی کانت، منتها در شکل نسبی شده و تاریخمند.

Genres: Philosophy, Science, Physics

414 pages, Paperback

First published January 1, 1992

About the author

Librarian Note: There is more than one author in the GoodReads database with this name.

Michael Friedman was an American philosopher who was Emeritus Patrick Suppes Professor of Philosophy of Science and Professor, by courtesy, of German Studies at Stanford University. Friedman was best known for his work in the philosophy of science, especially on scientific explanation and the philosophy of physics, and for his historical work on Immanuel Kant. Friedman has done historical work on figures in continental philosophy such as Martin Heidegger and Ernst Cassirer. He also served as the co-director of the Program in History and Philosophy of Science and Technology at Stanford University.

Reviews

[Dayi Behrad · Rating 3 out of 5]

کتاب روایت نداشت. انشا و ترجمه برای من خوش‌آیند نبود، ولی محتوای جالبی داشت.
نویسنده آشنایی مخاطب با نیوتون، اینشتین، کانت، هیوم، کارنپ، کون (Kuhn که مترجم در اقدامی عجیب آن را کیون ترجمه کرده بود!)، کواین و رایشنباخ را فرض گرفته. کتاب دلچسبی نبود. متن در واقع یک کتاب نیست، مجموعه‌ای از مقالات سنگین است که به هم چسبانده شده.

[Ehsan Ebrahimian · Rating 4 out of 5]

نویسنده یک انسان بسیار دقیقه که متون اصلی فلسفه علم رو به دقت فراوان کند و کاو و بررسی کرده و از روی همین بررسی ها روایتی نسبتا منسجم از فلسفه و تاریخ علم داره، گرچه آشکارا پایبندی خودش به کانت رو اعلام میکنه و این سوگیری واضحش بعضی جاها از اصالت نظراتش کم میکنه اما جایی که داره از متون و تاریخ حرف میزنه دقت و اشرافش خیره کننده است. ترجمه متن خوب نبود و کلمات تخصصی رو خوب معادل سازی نکرده بود و فهم متن رو سخت می‌کرد

[William Bies · Rating 3 out of 5]

Michael Friedman herein maintains it to be profoundly mistaken to read Kant as if he were independent of the scientific context of his day, as commonly done. Why? First off, he offers a model of fruitful engagement; but moreover, the present situation in the aftermath of logical positivism, is determined by the fact that Kant fits so well with the eighteenth century and hence not so well with the twentieth. In the present work, Kant and the Exact Sciences (Harvard University Press, 1992), Friedman follows the development of Kant’s thought on the question through all three of his major periods: pre-critical, critical, post-critical (Opus postumum). He wishes to investigate how the context, especially in mathematical sciences, shapes but does not indeed determine Kantian philosophy. Friedman’s thesis: the previous generation of scholarship tends to dismiss Kant’s work on exact sciences because the knowledge of his day, such as it was, has since become outmoded; but instead, in fact, Kant’s lifelong engagement with them was formative for his whole philosophical career and one of his great strengths.

The pre-critical period is subject of Friedman’s relatively brief introduction. The critical, of the more substantial Part I (chapters one to four). Here, Friedman engages Kant’s two innovative ideas: a sharp delineation between intellectual and sensible faculties of mind, on the one hand, while one the other, the thesis that the intellectual has genuine content only when applied to possible experience of phenomena in space in time (it has no grasp of the noumenal realm). Lastly Part II takes up the Opus postumum, in which (to follow Friedman) Kant despairs of understanding chemistry in terms of the Newtonian physics so well justified by the Metaphysische Anfangsgründe. Hence, Kant’s transition project, to follow Friedman motivated by Lavoisier’s chemistry: to show how the imponderable caloric fluid or aether has an a priori and not hypothetical status.

There is not much to remark about the introductory review of Kant’s pre-critical period, competent but of concern only to specialists. For most readers, the exposition will really pick up in Part I. Here, what is salient is Friedman’s continual effort to relate the proper subject of this study, Kant in the context of the eighteenth-century, to concerns current among philosophers of science nowadays. For instance, a discussion of monadic versus polyadic logic [pp. 58-59]; comments on why Kant needs intuition to do geometry while we do not anymore (allegedly) [p. 70]; and an analysis of motion in mathematics and Kant’s use of Kästner’s fliessende Größen [pp. 74-77]. Another example in point is a mention of how Koch-type fractal curves (discovered in the first decade of the twentieth century) exceed the scope of Kantian kinematic interpretation [p. 80].

Concerning the nature of different branches of mathematics, geometry has axioms whereas arithmetic does not [pp. 87-88]. In what may strike the reader as unexpected, Friedman contends that there is no room in critical philosophy for an anti-Russellian pure intuition [p. 92]. Euclidean geometry for Kant is to be compared not with Hilbert but with Frege’s Begriffsschrift [p. 94]. To understand in chapter two what Kant is up to, one needs to distinguish between general logic versus transcendental logic [p. 97ff].

Like Frege, he is interested above all in the objective meaning of mathematical signs and in objective content or truth of mathematical assertions. [p. 99]

The example of the two-sided plane figure naturally leads to a further, and very striking, idea: the existence of consistent systems of non-Euclidean geometry not only fails decisively to refute Kant’s conception of mathematics, it actually conforms to that conception perfectly and in fact provides rigorous confirmation of Kant’s view. For the existence of such systems shows precisely that Euclidean geometry is not logically necessary and that there are indeed logically possible spaces that do not satisfy Euclid’s axioms. Yet Kant, as we know, is not interested in mere formal systems; he is interested in objective meaning and objective truth. A synthetic source must therefore be found that provides objective reality for the concepts of geometry and which, as it were, constitutes a model for one logically possible system of geometry in preference to the others. Such a model is provided by the space of pure intuition, which, for Kant, is necessarily Euclidean. [p. 100]

Friedman disagrees; for him, only empirical not pure intuition is capable of providing a model for the truths of mathematics [p. 101]. He presses on to a critique of Kant’s logical but not real possibility, cf. Parsons [p. 103]. Another point Friedman brings home that ought to be more widely appreciated is that Kant does not say arithmetic:time::geometry:space – rather, the science of time is pure kinematics [p. 105].

Chapter three moves from the domain of pure mathematics to that of Newtonian science. Friedman’s objective in this chapter is to explain why the disagreement between Kant and Newton does not take place at the level of matter theory. Rather, Kant concludes that Newtonian science stands in need of a properly metaphysical foundation. Just what this may entail we learn better in chapter four. Mathematical principles are constitutive with respect to intuition while dynamical principles of pure understanding are merely regulative [p. 163]. Newton presents laws of motion as facts, Kant as conditions [p. 171]. For Kant, the first three laws of motion are a priori but not the law of gravitation [p. 179]. Mathematical principles are constitutive of experience while dynamics is concerned with existence of appearances and their interrelation [p. 180]. Therefore, against a Newtonian mechanical natural philosophy Kant prefers to frame a dynamical natural philosophy. The following two passages from Friedman suggest why:

When Kant claims that the understanding of prescribes laws a priori to nature, and that the understanding is accordingly the ground of the law-governedness of nature, he has in mind primarily dynamical principles of the understanding, and, in fact, the analogies of experience [p. 183]

The notion of nature pertains to the dynamical categories and principles because it explicitly involves the concept of existence. And, since the latter concept cannot be constructed, the notion of nature belongs to metaphysics or philosophy rather than to mathematics. [pp. 188-189]

More specifically, space as form of intuition versus as formal intuition [pp. 197-198]: in other words, the unity of the manifold of experience demands an explanation that goes beyond mere geometry. Kant’s view of metaphysical foundation of physics during his critical period is summarized well at the close of Part I:

Now Kant never had any doubt that the Newtonian natural philosophy comprises the true physics, but he always insisted that this physics nonetheless requires a metaphysical foundation. The question was what precisely does such a metaphysical foundation for physics involve. In the critical period Kant answers this question in terms of the doctrine of the schematization of the pure concepts of the understanding: metaphysical concepts – the concepts of substance, causality and community – provide a foundation for natural knowledge in virtue of, and only in virtue of, their application to our spatio-temporal form of sensibility, ‘which realizes the understanding in that it simultaneously restricts it’ (A147/B187). Such metaphysical concepts are thereby seen as the most general rules and conditions of ‘time-determination’. In the present context, I suggest, this doctrine is to be understood as pointing towards the essential limitations of an exclusively geometrical approach to the foundations of physics. What is missing from a merely geometrical approach, that is, is a recognition of the unique and indispensable role of dynamical structure: the need for an explicit construction of the spatio-temporal relations subsisting in a collection of different particular spatial manifolds occurring at different moments of time. Newton’s fundamental contribution was to fashion the appropriate dynamical concepts – concepts of mass, force and interaction – which, together with the law of motion governing these concepts, first make it possible to employ pure geometry in the construction of such a dynamical spatio-temporal structure. Kant’s fundamental contribution, however, was to see further than anyone else into the philosophical implications of the Newtonian achievement – its implications, in particular, for the nature and future of metaphysics. [pp. 209-210]

To begin our review of Part II, let us reproduce Kant’s own words on why chemistry cannot be a science, as quoted and translated by Friedman:

A whole of cognition that is systematic can already, for precisely that reason, be called science, and, if the connection of cognition in this system is a complex of grounds and consequences, even a rational science. But if these grounds or principles in it, as, e.g., in chemistry, are still in the end merely empirical, and the laws from which the given facts are explained through reason are mere empirical laws, then they carry no consciousness of their necessity (are not apodictically certain) and conseqently the whole [of cognition] does not deserve the name of a science in the strict sense – and therefore chemistry should rather be called systematic art than science. [p. 217]

Friedman explains why there’s no hope of deriving laws of chemistry from phenomena [p. 241]: the microscopic laws that generate chemical reactions involve movement on a scale too small for us to observe directly. Therefore, the case is not analogous to celestial mechanics, where Newton in Book III of the Principia can deduce the inverse-square law of gravitation from observations in the solar system. What, however, would Kant make of modern computational quantum chemistry? True, not all regularities in the chemical domain have been reduced to matters of calculation, owing to the intractable nature of large-scale problems, but certainly many aspects of chemistry, such as the kinds of chemical bond, have been rendered intelligible through the application of quantum mechanics. More on this below.

But we are getting well of ahead of what Kant himself could have known. His transition project embraces physics problems involving the ‘specific variety’ of matter and seeks to show why aether is not hypothetical [p. 221]. In an extended exposition [pp. 222-257], Friedman traces the development of Kant’s post-critical thought. The key idea – at the risk of oversimplification – seems to be just that the possible moving forces can be enumerated a priori, whereupon one looks into the world to find which are realized empirically [p. 247]. To this recensionist, however, E. Förster in Kant’s final synthesis: An essay on the Opus postumum (Harvard University Press, 2000) appears to be clearer on the gap that necessitates the transition.

Friedman isolates two main discoveries that contribute to chemical revolution [pp. 268-271]: i) the discovery that air is a mixture, and ii) the distinction between temperature and heat. Latent heat was hard to explain on basis of the kinetic theory of heat [p. 271]. Lavoisier in 1780 finds that solidity, liquidity and gaseous states reflect three phases of the same thing rather than represent three altogether different elements [p. 272]. Lavoisier’s chemical discovery of oxygen was not sufficient by itself to overthrow phlogiston [pp. 274-275]. An excellent account of the chemical revolution follows [pp. 264-279] but it is only loosely connected with Kant. Friedman documents Kant’s growing awareness of the chemical revolution [pp. 280-290] but it remains only a hypothesis that this is primarily what stimulated his transition project. Kant tries to inject transcendental content into caloric theory [p. 297]; i.e., why does it amount to more than just an empirical hypothesis?

Friedman now comes around to stating his thesis in Part II:

My suggestion, then, is that after the execution of the Metaphysical Foundations and the articulation of reflective judgment as an autonomous faculty, it becomes clear – from the point of view of the critical philosophy itself – that the absolute dichotomy between regulative and constitutive principles cannot be maintained. It becomes clear, that is, that the critical system requires completion via the Transition project, whose task is precisely to establish principle that are at the same time both regulative and constitutive. Hence, it is not unexpected when the aether-deduction of Transition 1-14 attempts to forge and explicit connection between the distributive unity of experience and its collective unity….These considerations – together with the circumstance that the object of the aether-deduction, unlike the transcendental ideal, is a phenomenal or empirical object – make it perfectly intelligible, it seems to me, that the aether-deduction should now aim to establish the actual objective existence of its object. [p. 305]

Friedman characterizes the post-critical Kant’s turn by appealing to the difference between an analytic versus a synthetic universal = cannot be a product of discursive understanding. Thus, the critical Kant treats the systematic unity of the laws of nature as a regulative ideal but then the synthetic unity itself would go unexplained; therefore, we need a constitutive representation to ground it and the late Kant determines that there is one such, and only one, available to us, viz., the aether. To follow Kant, aether can be deduced as the condition of the possibility of realized or hypostatized space [pp. 310-311]. Here, Friedman tells us why the alternative atomistic program is insufficient for Kant: it ‘appears...clearly to lack the crucial element of collective or synthetic universality’ [p. 311]. One must distinguish matter = Materie (necessarily singular) versus material = Stoff (could be plural), [pp. 312-313]. Using Newton’s grounding of space through universal gravitation as a foil, Friedman argues to this fate: all of the many versions of the aether deduction are unfortunately fatally flawed [pp. 325-326] – an instance of how Friedman allows his imagination to be too limited by current-day scientific prejudices.

Before wrapping up, let us throw out a few comments for the reader’s reflection:

1) Friedman’s interest in Kant is not purely antiquarian, rather he views him from a late twentieth-century vantage; i.e., wants not just to perform an internal critique but also to throw light on general and topical philosophical problems by confronting Kant with what we have since come to know. As a good illustration of this tendency, see pp. 80-95, starting off with this passage in particular:

The present approach to Kant’s theory of geometry follows Russell in assuming that construction in pure intuition is intended primarily to explain mathematical proof or reasoning, a type of reasoning which is therefore distinct from logical or analytical reasoning. Again following Russell, we have sought an explanation for this idea in the difference between the essentially monadic logic available to Kant and the modern logical of quantification theory….Yet Russell’s assumption has been vigorously debated. [p. 80]

2) What about any potential impact of modern empirical science in moral philosophy? Is there any bearing on the postulates of practical reason, faculty of judgment [Urteilskraft], categorical imperative or metaphysics of manners [Metaphysik der Sitten]? Friedman doesn’t address questions like this in the present work, but given how central seventeenth-century mechanical philosophy is to Hobbes’ atheistic political program in the Leviathan, it could be of great interest to look for a parallel in Kant, a wide-ranging thinker who is ultimately keener on moral philosophy than on natural science per se.

3) Assuming one is sympathetic to Kant’s dynamical philosophy of matter, what becomes of it in light of quantum mechanics? For one of Kant’s guiding themes is infinite divisibility of the continuum. But we today know that the Planck length (a tiny 10^(-33) cm) represents a limit to the process of subdivision, a scale at which laws of nature fundamentally different from those we currently conceive must intervene and change everything. Many theoretical physicists these days, indeed, suppose that space, or whatever it is that underlies space, must become discrete at the quantum-gravity scale. Then, quantum physics poses a far-reaching challenge to the whole Kantian enterprise as summed up in the transition project of his late years – for aren’t Kant’s ideas dependent on a classical world-view now superseded? As far as this reviewer is aware, current literature in the philosophy of science neglects questions such as these, probably owing to the mainstream tendency Friedman here combats to view Kant as but a dim star as far as natural science goes.

Conclusion: Kant and the Exact Sciences serves well as a standard commentary on an important aspect of Kant’s oeuvre, which is to say, as the latest iteration in the ever-repeating round of scholarship. It could be of interest in this connection to consult Erich Adickes’ old work from the 1920’s (Kant als Naturforscher in zwei Bänden, recently reprinted by de Gruyter), for this resource might be more informative on the physics of Kant’s contemporaries (Friedman engages mainly Newton and Lavoisier in their relation with Kant).

[Julia · Rating 3 out of 5]

Friedman is a historian of analytic philosophy, and reads Kant's 3 periods--pre-critical, critical, post-critical--as a development towards the telos of the foundations of mathematics in the early 20th century and debates concerning this in contemporaenous philosophy. It's a very useful book for references and for getting an idea of how a certain type of argument about the history of analysis might be constructed. But it doesn't push issues and implications far enough, given, I suppose, its lack of interest in philological concerns. E.g. I haven't yet seen how the moral and aesthetic philosophy fits into his picture of "Kant", and Lambert, the non-Euclidean geometer to whom Kant was to dedicate his first Critique, is not mentioned at all. Author is influenced by Ricketts, Hintikka ...

[Louna]

55-95 skolematization

[Brian Beakley · Rating 4 out of 5]

A beautiful book, whether or not you agree with Friedman's conclusions. So densely researched -- it would take me 5 lifetimes to write such a book. Really brilliant -- like, Bas van Fraassen-league brilliant.