Extract on the transmission of tacit knowledge

Collins on the capriciousness of tacit knowledge

In order to consider the transmission of tacit knowledge we will begin by drawing on two discussions from Harry Collins whose work we discussed in chapter 3. In both cases, we will highlight how assumptions Collins makes about the nature and transmission of forms of tacit knowledge put the idea that it can be communicated under threat. One virtue of the analysis offered in preceding chapters is that it does not, implausibly, suggest that tacit knowledge is particularly difficult to share.

First then, the empirical findings. In Changing Order, Harry Collins summarises a piece of sociological inquiry he carried out in the 1970s. He visited six of seven UK laboratories who were attempting to build a working laser which, although of a new design (a Transversely Excited Atmospheric pressure CO₂, or TEA, laser), had already been successfully built in other laboratories in the US, five of which he also visited. In one case, a scientist – Bob Harrison – who had already built one working model aimed to replicate it so as to have two working models. Despite this limited problem – a clear case of Kuhnian  ‘normal science ‘ – and despite the availability of explicit instructions, Collins discovered a surprising difficulty.

[N]o scientist succeeded in building a laser by using only information found in published or other written sources. Thus every scientist who managed to copy the laser obtained a crucial component of the requisite knowledge from personal contact and discussion. A second point is that no scientist succeeded in building a TEA-laser where the informant was a ‘middle man’ who had not built a device himself. The third point is that even where the informant had built a successful device, and where information flowed freely as far as could be seen, the learner would be unlikely to succeed without some extended period of contact with the informant and, in some cases, would not succeed at all. [Collins 1985: 55-6]

Collins concludes that because it has three significant features or properties, the knowledge involved in laser construction is tacit knowledge.

In sum, the flow of knowledge was such that, first, it travelled only where there was personal contact with an accomplished practitioner; second, its passage was invisible so that scientists did not know whether they had the relevant expertise to build a laser until they tried it; and, third, it was so capricious that similar relationships between teacher and learner might or might not result in the transfer of knowledge. These characteristics of the flow of knowledge make sense if a crucial component in laser building ability is ‘tacit knowledge’. [Collins 1985: 56]

On Collins’ view, tacit knowledge is mysterious. Although, on this early view, it can only be passed on by accomplished practitioners, that is necessary but not sufficient for its transfer. It is capricious in that similar relationships between teacher and learner are sometimes sufficient and sometimes not. And it is invisible in its flow. Even the scientist whose ongoing attempts to build a laser Collins particularly studied, Bob Harrison, is unaware of the details of his own knowledge how to build the laser.

Harrison would not have been a lot of use as an informant at the beginning of his attempt to build Jumbo [his first working laser]; there is no way that he could have informed anyone about the necessity of having the leads from the capacitor to the electrodes as short as possible, for example, since he did not realise the importance of this himself. But, he did not know that he did not know. [ibid: 73]

Whilst much of the time Collins suggests that the knowledge involved in laser construction is practical, skill-like knowledge, it is clear, from the descriptions he gives, that it also includes elements that appear more like matters of fact or knowledge-that. Summarising this early empirical work more recently, Collins explicitly connects the ineffability of the scientists’ knowledge to Polanyi’s slogan.

[Y]ou may not know what you need to know and I may not know what I know. Thus, in the early days of TEA lasers scientists did not necessarily know that the inductance of the top lead was important but by copying existing designs they built in successful short top leads without knowing why. [NS]

Collins summarises his early view of tacit knowledge in a number of propositions which begin:

Proposition One: Transfer of skill-like knowledge is capricious.
Proposition Two: Skill-like knowledge travels best (or only) through accomplished practitioners...
Proposition Three: experimental ability has the character of a skill that can be acquired and developed with practice. Like a skill, it cannot be fully explicated or absolutely established.
From the three studies it seems firmly established that laser-building is something you do not know whether you possess until you have built a laser. Thus, laser building is invisible in its passage...
Proposition Four: Experimental ability is invisible in its passage and in those who possess it. [ibid: 73-4]

This presents a picture of the sharing of tacit knowledge which looks to be particularly difficult. Perhaps influenced by his view of Wittgenstein’s regress argument according to which it is not possible fully to specify a rule and thus there must be something more to a rule than its specifiability to underpin our mysterious ability to know what accords with it, Collins suggests that empirical tacit knowledge is similarly invisible to those involved.

If this account were true, it would suggest that teaching or sharing such tacit knowledge is, of necessity, fraught with difficulty. If such knowledge is invisible, even to those who possess it, then it cannot be demonstrated. As Wittgenstein puts it, the student has to guess the essential drift of the teacher’s practical demonstration. No wonder then that ‘ it cannot be fully explicated or absolutely established’. But if that is the case, the prospects for improving teaching methods are also dim. There will be no way to compare the results or potentially more and potentially less successful forms of demonstration and pedagogy.

But as we have argued in preceding chapters, there is no reason to think of tacit knowledge in this pessimistic way. If, for those with eyes to see at least, context-specific but still conceptually structured practical knowledge can be articulated and demonstrated in practice, there is no need to think of it as essentially invisible or essentially silent, to change sensory modality.

Collins’ account generalises from what seem to have been particularly difficult cases to all cases of tacit knowledge. Even thought as a matter of physical theory a piece of Kuhnian normal science, the actual construction of a complex laser from basic components was clearly a difficult and challenging task. Further, unlike everyday structured abilities, such as the ability to swim front crawl or butterfly, there seem to be difficulties in testing the component abilities. This is a particular feature of the context in which Harrison works and merits a separate summarising proposition.

Proposition Five: Proper working of the apparatus, of parts of the apparatus and of the experimenter, are defined by the ability to take part in producing the proper experimental outcome. Other indicators cannot be found. [ibid: 74]

In other cases, this is not true. One might have long-standing knowledge of how to make Yorkshire pudding: a structured capacity to assemble, mix and cook the basic ingredients. If so, a run of failures given a change of oven (the puddings fail to rise or are over cooked, for example) is evidence, at least, for one of a range of failures. Perhaps it is the oven or the particular ingredients which will also change from time to time. But if the same ingredients are, in the same period, yielding fine pancakes, that is evidence, at least, against a problem with the ingredients. If cognitive and practical knowledge, basic ingredients and methods all change or are introduced simultaneously then the predicament may be as Collins describes in proposition five. But that is not an essential feature of tacit knowledge.

Just because it is sometimes difficult to test whether someone has a practical ability does not mean that it is always (any more than parallel claims apply to knowledge-that). The transfer of skill-like knowledge need not be capricious. Practical teaching – such as of bike riding, or instrument playing – can be quite predictable. And in such cases, there is no reason to think that mastery is invisible.

McDowell's reply to Mind and World as transcendental anthropology?

I have received a copy of John McDowell’s response to my paper ‘Mind and World as transcendental anthropology?’, a version of which is here.

“I am grateful for Thornton’s serious and longstanding engagement with Mind and World, here and elsewhere.

In this essay, he suggests that there is a tension between my avowal of therapeutic aims, on the one hand, and my positive “anthropological” offerings, on the other.

I think the suggestion of a tension is overblown. Thornton asks: “how … can an anthropological perspective be reconciled with a therapeutic aim?” But if this question arises about me, it arises already about Wittgenstein, who is, of course, the inspiration for my therapeutic conception of how to deal with the philosophical anxiety that I identify. It is surely undeniable that Philosophical Investigations is full of anthropological remarks. Consider, for instance, what Wittgenstein says in §25: “Commanding, questioning, recounting, chatting, are as much part of our natural history as walking, eating, drinking, playing.” And surely this feature of Wittgenstein’s text stands in no tension with a therapeutic conception of the task of philosophy, provided only that the anthropological remarks can be understood to have the character of reminders (cf. §127): not expressions of disputable philosophical doctrine, but statements of things everyone knows but some people perhaps forget, so that they fall into a frame of mind in which there seems to be a need for positive philosophical doctrine to address well-posed philosophical questions.

It is true that my contributions to philosophical anthropology — to go on talking in Thornton’s terms — are unlike Wittgenstein’s, in at least two respects that might seem relevant here. First, my contributions have a somewhat systematic character. And second, as Thornton notes, they draw on the philosophical canon. The main element in my attempt to dissolve the anxiety I identify about the very possibility of content is a roughly Kantian positive characterization of the perceptual experience of rational animals.

But — surprising though this may be — I want to claim that, even though there are those differences, my anthropological remarks are like Wittgenstein’s in having the character of reminders. To understand my characterization of the perceptual experience of rational subjects, one may need to learn a new vocabulary, or at any rate a new way of using some vocabulary with which one is already familiar. But I want to claim that if one really understands the language in which I give my description, it can only be as a result of some philosophically generated confusion — for instance what I try to unmask as a misconception of the intellectual obligations of naturalism — that one can find the description disputable.

My fragment of a philosophical anthropology — again, to continue talking in Thornton’s terms — might be called “transcendental”, in a narrower sense than the one Thornton cites from Jonathan Lear: its purpose is transcendental, in the roughly Kantian sense of being directed at alleviating an anxiety about the very possibility of thought’s being directed at objects. The only point of calling it “non-empirical”, which is what Lear extends “transcendental” to mean, is that it is not intended as a possibly disputable theory, something for which evidence might be demanded. I think that goes for Wittgenstein’s anthropological remarks also.

As I said, my anthropology is transcendental in purpose. I mean that to stand in contrast with saying that it is transcendental in content. My transcendental purpose is discharged by saying things that are meant to be in themselves mundane, such as to be obvious if it were not for philosophically generated blind spots. Lear’s contrast between “empirical” anthropology and a kind of philosophical anthropology that points to unsayable insights gets no grip on anything I do. I think this goes for Wittgenstein also, but arguing for that would take me too far away from Thornton’s essay.

Thornton suggests I am threatened with “slippage into a merely Quinean picture” — a picture in which there is no room for any idea of analyticity. He thinks this threat impinges on me because I reject “a two factor model with endogenous and exogenous elements”, and he thinks I should avert the threat by rehabilitating an idea of an endogenous factor, rescuing that idea from seeming to amount to an idea of an endogenous Given.

This seems off key to me. What I reject is a conception according to which endogenous and exogenous factors are supposedly intelligible independently of one another, but combine to account for the form and content of world views. The Given, conceived as in what Sellars exposes as a Myth-encumbered version of empiricism, would be an exogenous factor supposedly intelligible independently of any appeal to our intellectual capacities. As Thornton in effect notes, one can discard the Myth of the Given without discarding the idea that things are given — from outside, in an innocuous sense — in perceptual experience. And he is certainly right to urge that that innocuous acknowledgment of an exogenous factor should have as a counterpart an innocuous acknowledgment of an endogenous factor, rehabilitated by not being conceived as intelligible independently of anything exogenous. But I do not see why he thinks meaning, as it figures in the idea of truth by virtue of meaning that we lose if we fall into “a merely Quinean picture”, would need to be understood as wholly determined by the endogenous factor, as opposed to what I suggested, that it needs to be understood in terms of both factors. I do not see why Thornton is not satisfied by the picture I gave when I wrote the following (Mind and World, 157/L’esprit et le monde, 197 — a passage Thornton actually quotes):

When we reject the dualism of scheme and world, we cannot take meaning to constitute the stuff of schemes, on the dualistic conception of schemes. But that does not deprive us of the very idea of meaning. So if I am right that Quine’s insight is really a glimpse of the unacceptability of the dualism, perhaps we can rehabilitate the idea of statements that are true by virtue of their meaning without flouting the real insight.”

Retrospective expectation?

In an introductory lecture today to give UCLan philosophy undergraduates a taste of a real problem and the sorts of options available to philosophy to address it, I put up the following passages from the Philosophical Investigations.

I see someone pointing a gun and say “I expect a report”. The shot is fired. - Well, that was what you expected; so did that report somehow already exist in your expectation?
Or is it just that there is some other kind of agreement between your expectation and what occurred; that the noise was not contained in your expectation, and merely supervened when the expectation was being fulfilled? –
But no, if the noise had not occurred, my expectation would not have been fulfilled; the noise fulfilled it; it was not an accompaniment like a second guest accompanying the one I expected.
Was the thing about the event that was not in the expectation too an accident, an extra provided by fate? - But then what was not an extra? Did something of the shot already occur in my expectation? - Then what was extra? for wasn’t I expecting the whole shot?
“The report was not so loud as I had expected.” - “Then was there a louder bang in your expectation?” [Wittgenstein 1953 §442]

I’m not sure in a brief lecture how many students will immediately have seen the problem here but I had three excellently philosophical responses. So, starting with the first question “did that report somehow already exist in your expectation?” there was a degree of agreement very quickly that the noise could not have existed in the expectation because it – that very noise – had not yet happened. (I did not expect them to see this as such a problem so easily.)

That pushed the discussion to the second option: that there is some other kind of agreement between the expectation and what occurred. Here I had no volunteers to make a suggestion for how it might work and didn't have enough time to try to draw one out. What I think of is something like a design specification, an advance blueprint, which might lay down some sort of condition that the actual noise meets and it would have interesting to get them to reflect on whether the Wittgensteinian worry that any such partial specification (missing out the accidental details of the actual event) would fail to capture the whole of the shot is really a good worry. (There are other regress-related reasons to be suspicious of the suggestion, of course.)

But in the end, there were three positive suggestions (from people who have done no university level philosophy and will not have looked at this topic before).

1: “It is something to do with past experience. A solution to the problem of induction will also solve this problem.” That seems a fine suggestion at this stage (second day! of three years) of the programme, though of course it does not really address the worry not of the justification for an expectation but rather how one can come to entertain it (its content) at all. Still, it was good to have a connection to another area of philosophy and maybe a proper solution to the problem of induction would indeed presuppose an account of intentionality.

2: “Maybe we don’t actually entertain expectations, we just talk as though we do.” This came right at the end and so I didn’t have time to draw out the underlying ideas. But being prepared to entertain such a revisionary possibility in the face of the philosophical puzzlement in the face of the ‘how-possible?’ question was rather bracing. (The speaker cannot have known of the Kripkean direction of travel here.)

3: “Perhaps it is only possible to form expectations in retrospect?” This struck me as brilliant (bloody brilliant, even). Grasping the nettle that the very noise cannot exist in the expectation because it has not happened yet but by contrast can after the fact – because one can think of that noise – the suggestion is that expectations are always formed retrospectively because we can see how they – at least – would be possible. I see a promising philosophical future for this student.

Phenomenology at the INPP

It seems to me that one of the repeated topics of the INPP conferences is phenomenology and its application to psychiatry. Sometimes it is not the direct topic, which may be, instead, the nature of the person, conceived of as the primary subject matter for psychiatry. (In so saying one is denying, I take it, that sub-personal states such as disease states are the subject. Though what that in turn amounts to needs – in each context – some unpacking.) But, to return, in this latter context, phenomenology swiftly returns as what a concentration on the person as the proper focus for psychiatry also requires. If we are to think of the person as the proper focus then we had better embrace phenomenology. Well that, at least, has been a claim I have often heard at INPP sessions in the past.

(The contrast I have in mind is something like this. At a session on the nature of psychiatric taxonomy I do not expect to hear much about the way in which we might reason about it: what methods and tools we might use to frame an analysis. I expect, instead, just the – contested – analysis.)

So it was interesting to hear after the opening talk at this year’s conference by David Woodruff Smith, which articulated both a conception of phenomenology and how empathy might be both its method and its topic or subject matter (though I must admit to not getting much of  a feel, in a short presentation, for what his conception of empathy was: an ability to grasp the nature of other people’s mental states and experiences or a particular experiential route to that) a question from Derek Bolton. If one does not assume a Kantian framework governing a priori constraints on how minds experience the world; and if one does not simply follow empirical scientific results; what standards of correctness, what particular routes of inquiry, underpin phenomenology?

If my notes are at all reliable, the answer started with the fact that Husserl himself had had a pre-Kantian phase whilst still using a phenomenological method and so Kantianism can’t be necessary. The approach should be ‘plainly’ phenomenological and thus does not need such a high level construction (as Kantianism, I inferred). Still experience is structured by interpretation and just as one has no need of empirical tests to test linguistic theories (eg about subject and verb order) so one can appeal to our pre-empirical grasp of a kind of basic mental vocabulary to ground the interpretation of experience.

I fear this does not do justice to Peter's answer. But it does leave the question: what is this low-level, non-hifalutin, non-Kantian structure on the assumption that the appeal to language was merely a metaphor. What is left to be the grammar of experience? I seem to have blind-spot every year to this issue: to what the short answer should be.

(Of course, the question of how philosophy ever works remains a problem pretty much in every approach. How can it be both distinct from empirical inquiry and yet purport to tell is anything about the world? With the death of conceptual analysis it is not as though phenomenology's rival approaches have much better answers. But this is less evident at INPP conferences because philosophy as such is rarely the topic of inquiry outside the repeated focus on phenomenology.)