Abstract
Most traditions present knowledge as transmitted down the ages through diverse institutions and practices, especially religious. These traditions hardly allow for the growth of knowledge. (They characteristically ascribe new ideas to ancient sources.) By contrast, philosophical tradition presents knowledge as constantly challengeable and always able to stand up to all challenge because it has been proved. Yet there is little or no tradition of discussing proof. What traditional discussion there is relates mainly to competing claims for thought or for experience as the basis for proof; and critical examinations of these claims abound. The last century saw a stupendous growth of proof theory, but it pertains only to logic and to mathematics, distinctly not to science. There still is no theory of proof in science. Today the received mode of philosophical research is to skip this stage and discuss some proof-surrogate. A criterion for what such a proof-surrogate should satisfy to be a serious contender is lacking.
Michael Polanyi bravely ignored all this. He offered instead a compromise between tradition and philosophy: the received stock of knowledge—science for short—is largely transmitted yet it grows in small measures; individuals in charge of science (of the received stock of knowledge)—the leaders for short—sort through the new ideas that turn up as contenders (for the status of new additions to science) and recognize some of them from time to time. This reduces the problem of knowledge drastically—which is Polanyi’s target—as the demand to prove received ideas gives way to the question: how do leaders decide which new idea to recognize? Polanyi’s answer: they cannot say.
This sounds very disappointing. But wait. Listen to great artists explain how they behave as art producers, art critics, and art teachers. Plato noted (Ion, conclusion) already that what they say is not serious, although this does not detract from their art. Aware of what they do, they are still unable to articulate it, says Polanyi. What they say is due to their surrendering to the demands of excessively rationalist philosophers. Wise artists simply refuse to respond to the challenge: they recognize the fact that their knowledge is tacit. By now the label “tacit knowledge” is Polanyi’s brand name.
This discussion is intriguing. Polanyi invited critical discussion of it. During his lifetime he was almost entirely ignored. He was not bitter about it. He was a physical chemist who for his doctorate early in the twentieth century invented a bold theory of solids adsorption of gases, and later he offered a bolder theory of catalysis. The first won recognition about half a century later; he took this as an example of the wisdom of the leadership: when he developed his theory of adsorption it did not chime with the intellectual framework of the time. The framework changed and his theory was vindicated. But why was he awarded his PhD for a theory that the leaders deemed inadequate? He offered a personal answer: his adviser was nice. This is too modest. Einstein showed interest in his theory and corresponded with him, yet he was no acquaintance. Quite generally, Polanyi appealed to the authority of scientific leaders with no theory of leadership, least of all scientific. He noted that when a scientific leader goes out on a limb the crowd may join in or change their leader. Which is which? No answer. Traditions do offer an answer, however inadequate: the origin of our culture is supernatural and the leadership is transmitted by the authority of the first leader supernaturally inaugurated. This does not hold for science. There we have a slightly more complicated picture. We do not know the truth, yet we survive because we do possess some knowledge. We acquire knowledge before we acquire language and so this knowledge is nonverbal—tacit—and so not given to the usual skeptical critique of knowledge. Now some individuals have more or better tacit knowledge, and we all know this tacitly. This is how we acknowledge leadership, in science as elsewhere.
Thus stated, this theory cannot possibly be true; as an anti-Stalinist Polanyi would have rejected it out of hand. So what distinguishes Einstein from Stalin? No answer. Perhaps Polanyi found it too obvious to state. Yet Stalin fooled sufficiently many scientists to provoke Polanyi to start a magnificent large-scale campaign against Stalinism in science. There is no way around this trouble: no matter how admirable Polanyi’s system is, it is a nonstarter because of his not having discussed critically the matter of scientific leadership. We should admit to him that we have tacit knowledge and that it is important. Already Hume noted that, and indeed for the same reason that Polanyi repeated it: the insoluble problem posed by the critique of science, of induction, of knowledge, of what-have-you. Polanyi added the idea that some of us have more of it than others and that they are often leaders, even scientific leaders, and rightly so. We should admit to him that this is important too. Yet he stressed that we do not live in Utopia, which is admirable, and he applied it to the scientific leadership of his day, which is brave. Hence, as a solution to the problem at hand it is a nonstarter. Harry Collins now elaborates on Polanyi’s idea of tacit knowledge. He accepts it but finds Polanyi’s application of it insufficiently broad. Collins has another fish to fry: he wants to explain social life with no reference to social institutions (see below). His project is a nonstarter too, and for similar reasons.
The project that Collins undertakes here is more ambitious than Polanyi’s in that he discusses not only individual tacit knowledge but also, and mainly, collective tacit knowledge (pp. 2, 4-6, 119, 135, 138). As this is the novelty of this book, allow me one paragraph about it. It is not mysterious, says Collins (p. 132), and so is not “collective consciousness.” Is tacit knowledge subconscious then? Collins faces this question (p.5), saying it is hard:
The hard part of this [question] is the initial repertoire of the drives that I might have. If this repertoire of drives is partly provided by the way I am embedded in society then the psychiatrist will not know how to make explicit how that repertoire was initially formed. (pp. 79-80)
So yes, subconscious knowledge is tacit. And yes, although this quote adumbrates Freud’s (refuted) conjecture that the cure of a neurosis depends on the discovery of its origin, it is not Freudian; it is much less Jungian. Perhaps it is Wittgensteinian. Indeed, Collins aims to wed Polanyi the philosopher of tacit knowledge with Wittgenstein the philosopher of forms of life (p. 169). Be that as it may, he declares forms of life pregnant with collective tacit knowledge (pp. 3, 19, 27-28, 23, 26, 34, 59, 65, 79, 87, 91, 135, 144, 146-47, 178). Collins’ reference to Wittgenstein and to collectives is significant: he recognizes language as a carrier of many aspects of the social, but he does not recognize language as an institution. With Wittgenstein he recognizes artificial and natural language, not language as an institution, and so not the fact that in literate societies language as an institution interacts with other institutions and often alters by decree; the word “institution” or any of its cognates (other than “society”) is practically absent from this book. It repeatedly alludes to social philosophy, distinguishing between Cartesian Cartesianism and social Cartesianism and alludes to the authority of Émile Durkheim. This is an appeal to authority and to a distinction. We may always accept distinctions as they do not impose ideas; at best they help wording them well and at worst they are mere ballast. The authority of Durkheim has nothing to do with the current discussion and unqualified respect for him is rare even among fans; Collins is no fan: unlike Durkheim, he ignores institutions.
As Collins notes, it is not too clear what tacit knowledge is: we can easily be bogged down with questions like, does a tree know how to be, a cat how to hunt, and a calculator how to calculate (p. 78). They are no good: they are cases of “language going on holiday,” as Wittgenstein would say (p. 78). The error is this: the idea of tacit knowledge is born in contrast with explicit knowledge and so it cannot be extended to cases that have no room for language. This is the weakness of Collins’ methodology: he knows when to extend a discussion when not, but would not tell us how he decides—very much like Wittgenstein. In the field, however, people do not wait for license from Wittgenstein or from Collins: they try to extend any discussion they want any way they fancy and if they are lucky and come up with a good solution to a good problem, we all jump for joy.
Collins says, ascribing tacit knowledge to a tree, a cat, or a computer is in violation of a rule. He declares that discussions of tacit knowledge have to start with reference to explicit knowledge. Yet whether computers use language is under dispute: it is a hypothesis, not a rule. Kripke says, as cats and dogs are of the same ilk, in ordinary language “cat” and “dog” are substitutable; not so “robots”: they are a different kettle of fish. He makes up this rule. (Logic allows all sorts of substitution-classes. Determining them so as to settle disputes is too high-handed.) Collins does better elsewhere, as he discusses the view that robots do not think, and expresses admiration for Hubert Dreyfus, indefatigable campaigner for human superiority over machines. Collins dissents from Dreyfus, incidentally: their joint mythical mentor that Dreyfus calls Heidegger, Collins calls Wittgenstein (p. 169).
Keeping within Polanyi’s problematique, the obvious questions for us to ask are, “Could all tacit knowledge be made explicit” (p. 86) and, more importantly, “Could all collective tacit knowledge be made explicit” (p. 138). An affirmative answer to these questions will render the exercise trite. The trite exercise of rendering the tacit explicit signifies in logic and by extension in mathematics. Thus, when Pasch discovered that an axiom was tacitly used in Euclidean geometry, it was explicitly added to the Euclidean axiom system. More interestingly, when Stephen Kleene showed the tacit assumptions behind the fan theorem of L. E. J. Brouwer, he caused a furor as the smuggled tacit idea is one that Brouwer had banned. Polanyi’s idea of tacit knowledge is still more intriguing. He was tackling the problem of induction: he saw his theory of tacit knowledge as a way to circumvent Hume’s criticism of science by dispensing with proof, by allowing faith into science. (He thus also resolved to his satisfaction the tension between two traditions, science and religion.) So he insisted that tacit knowledge will never be rendered fully explicit; he explicitly disparaged undertaking research into ways of doing so.
This was a grave error. There is a boundary between knowledge and ignorance, and efforts proceed to push the boundary and reduce our ignorance, in full knowledge of the endlessness of this task. Polanyi drew attention to a similar boundary within knowledge, between the explicit and the tacit. This boundary deserves the same treatment. His view of tacit knowledge as interminable is thus right, yet his disdain for efforts to render explicit as much tacit knowledge as possible is erroneous. It borders on obscurantism.
Consider then Collins’ collective tacit knowledge. Those who doubt that it exists might consider texts that no one remembers now but that may later be retrieved. Particularly striking are here machine-produced texts, such as logarithm tables or overlooked parts of research output. No less striking is the repeated expression of amazement by great artists and by great thinkers as they look afresh at their own output. However one sees it, the difference is clear between the knowledge that dwells in an individual’s brain (or desktop computer) and the knowledge that sits patiently in a library awaiting retrieval. These are all clear cases of collective tacit knowledge, or rather institutional, since libraries are social institutions. Collins offers many instances but he does not explain specific cases, as this requires reference to institutional settings, and these he evades.
Collins dwells on his differences with Polanyi. Here (on p. 148) we find a strange panegyric to him: he is the inventor of the term and he fought for the idea, but he lived in an excessively rationalist world, so that “it is no surprise that Polanyi was tempted to make tacit knowledge into something mystical and inspirational.” Mystical. On the blurb of this slim volume we find praise that Trevor Pinch lavishes on Collins: he “neatly turns Polanyi on his head by showing us that the really deep mystery is how knowledge ever becomes explicit in the first place.” One person’s mystery is another person’s obviousness.
Polanyi’s worst weakness is his “stress on the personal element of tacit knowledge.” It is “harmful” and Collins, we should know by now, has come to rectify it with his concept of collective tacit knowledge. This seems to me a return to collectivism that imposes vagueness on an otherwise admirably clear text. Reading it as students is more enjoyable than reading it with a critical eye and making an effort to place it in context. The most important tacit idea here is anti-individualism: I cannot cite a single sentence that would sum up the moral or social philosophy that Collins advocates, yet I take it as obvious that it is not one I share.