2.2 Practical knowledge

Practical knowledge may be defined tentatively as the form of knowledge involved in exercising a skill. It is the ‘knowing how’ which Ryle sought to analyze, providing the following examples: “to make and appreciate jokes, to talk grammatically, to play chess, to fish … [and] to argue” (Ryle, 1949, p. 29). It is the kind of knowledge for which Hubert and Stuart Dreyfus have developed their skill acquisition model (Dreyfus & Dreyfus, 1986) and which they illustrate with examples taken from a range of domains, including car driving, chess playing, medical diagnosis, and flying an airplane. Polanyi offers the further examples of riding a bicycle, using a blind man’s stick to sense the surroundings, and discerning relevant traits on an X-ray picture (i.e. the skill of having certain kinds of experiences). As indicated above, according to Nordenstam, Wittgenstein also argues for the existence of a ‘practical knowledge’ (distinct from experiential knowledge), present in our ability to use language correctly.

This brings me to the second issue where the Scandinavian reception of Wittgenstein – and my appropriation of it – differ from the interpretation most common in the English-speaking world, namely what is involved in his view of adequate language use and, more specifically, what exactly he is trying to convey in his arguments on rule-following. I develop the differences between these readings at some length in Article 5, without however explicitly distinguishing practical knowledge from the experiential aspect of knowledge discussed in the last section. The latter point is added in Article 6 where I also discuss the relationship between the two forms and, in particular, the relationship between them and propositional knowledge, but in turn only touch lightly upon the differences between the reception of Wittgenstein in Scandinavia and in the English-speaking world. In the following, I shall integrate the arguments of the two articles into one exposition.

On both the Scandinavian and the English-world reading, Wittgenstein argues that rules are ‘flexible’ and infinitely interpretable so that any future action could in principle be brought to accord with a given rule, simply by providing some (perhaps fanciful, but not incoherent) interpretation of it. Similarly, both readings agree that Wittgenstein infers as a consequence that practical examples and training are necessary to show what the right application of the rule is. The difference between the Scandinavian and the mainstream English-world receptions lies in the understanding of what the examples actually supply to the rule and, following on from this, how one should construe the relationship between rule and practice. The mainstream English-world reading of him thus is that examples supply the correct interpretation, i.e. the examples indicate which of the infinitely many interpretations which could be given one should actually choose (Baker & Hacker, 1984; Kripke, 1982; C. Winch, 2006; P. Winch, 1990). The premise behind this way of understanding Wittgenstein is that practice is necessarily rule-governed and that the main issue for him was to explain what we actually do when we follow rules, given their ‘flexibility’.

In contrast, this premise does not underlie the Scandinavian reception which emphasizes practice over rules and the doing itself over interpretation. Thus, on this reading, the rule-following arguments of Philosophische Untersuchungen should be understood in the vein of the points later made in Über Gewiβheit that we in our lives – in our actions and ways of dealing with the world – show a certainty which cannot meaningfully be put into words (cf. footnote 10 above). Examples are seen as conveying, not the correct interpretation, but the feel for how to go on – the “weiter wissen” itself. This “weiter wissen” is not an interpretation at all, but rather a tacit, practical, embodied understanding there in the very doing. Rule-following on this view is one way of ‘doing practice’ – of practical knowledge – but not the only possible one. As Wittgenstein says in § 83, we sometimes play games where the rules are made up ‘as we go along’. One can imagine instances of this kind where the rules are not articulated at any point, neither before or after, nor during the game. One can further imagine a game developing only through examples of ‘how to go on’, without it being possible, not even analytically, after-the-fact, to extract a set of rules from the examples, which characterize what distinguishes correct from incorrect ways of proceeding in the game. Not only can we imagine it – according to Wittgenstein, this is actually precisely what we do when we use (and learn to use) a whole range of our everyday words, including his prime example, the word ‘game’ itself: Their meaning cannot be given in a precise definition (i.e. rule), because for all suggested definitions it is possible to find examples which do not fit the definition, but which we still recognize as instances of the concept. Instead, the meaning is given by paradigmatic cases and the ‘family resemblance’ they have to one another as well as to further instances. No one trait runs through the whole ‘family’ of cases exemplifying a concept; just like in a human family, where the children look like both of their parents and each other in some respects, but not necessarily the same ones, and the parents normally do not look alike. In the case of such words, ‘knowing how to go on’ using them in new situations is a practical feel, not for how a rule should be applied, but for what counts as being ‘similar enough’ to the paradigmatic instances. In general, it will not be possible to explicate a set of definite criteria for what constitutes ‘being similar enough’. Rather, it is, again, a tacit, practical understanding present in the very act of discriminating.

An additional example of ‘knowing how to go on’ is provided by Schön (who in turn cites Alexander, 1964; and Vickers, 1978) and concerns the skill of weaving and of distinguishing between good and bad ‘fits’ within familiar patterns (Schön, 1983, p. 52f). Schön points out that the weavers cannot describe what characterizes a good pattern, but that they can immediately tell a bad one from the good ones. He cites Vickers to the effect that deviations from a norm are in general much more easily recognized and described than the norm itself. In the present context, the example illustrates that a norm can take the form of a holistic pattern, rather than of an expressible rule. Having the practical feel for ‘how to go on’ in this case is a question of holistic pattern recognition. It is similar to, but not quite the same as the discrimination of what counts as ‘similar enough’ to paradigmatic instances of a concept.

On the Scandinavian reading of Wittgenstein, then, practical knowledge involves a tacit, practical understanding present in the very doing itself. Some ‘doings’ may be described by rules – some are even constituted by following rules – others are orderly without being rule-governed (Dreyfus, 1979, p. 256ff). In all cases, the ‘feel for’ regulates the way the rule is applied, the pattern recognized, the paradigmatic examples used to discriminate new cases – though of course the ‘feel for’ will in turn also be regulated by the rule/pattern/case use. Still, rules, patterns, and paradigms may all become obsolete over time and be revised or given up accordingly. Obsolescence happens because what we do – our practices as not regulated by the rule/pattern/paradigmatic example – has changed. And when we revise them or give them up, we do so in accordance with what our ‘feel for’ – as something which goes deeper than the rule/pattern/paradigmatic example itself – shows us to be an adequate adaption to the new practice.

Somewhat ironically, philosophical expositions of ‘practical knowledge’ tend to have a distinctively intellectualist flair. This is witnessed in the discussion of rule-following as a practice in the Scandinavian reception of Wittgenstein, referred to in the preceding paragraphs. It is seen in the examples and descriptions which Ryle gave of ‘knowing how’, even as he tried to distinguish this form of knowledge from ‘knowing that’ and to avoid an intellectualist rendering of the former. Thus, the examples he initially provides (cf. above) all, except for fishing, concern so-called intellectual skills. In the course of his examinations, he does mention other ‘non-intellectual’ skills such as lorry-driving, shooting, and boxing. However, he consistently equates ‘knowing how’ with ‘performing intelligently’ and analyzes it as the ‘application of criteria’(Ryle, 1949, p. 29), stressing that reliably performing well in itself is not a sufficient condition for displaying intelligence (so do the clock and the well-trained circus seal, he reasons). The application of criteria need not be executed in conscious explicit thought processes, though – in fact, it is his main claim that in general they are not, and that “to perform intelligently is to do one thing and not two things… to perform intelligently is to apply criteria in the conduct of the performance itself.” (Ryle, 1949, p. 40) Nonetheless, here, as in the discussion of rule-following in Wittgenstein, there is a striking neglect of the fact that certain bodily movements are involved in the performance of the skills mentioned, and that for a person to master the skill it is not enough to ‘apply criteria’, or have a ‘tacit feel for what the correct rule-following procedure is’; the person must master the concrete bodily movements as well (or integrated with the tacit feel, as I shall argue below). The intellectualist bias becomes quite plain in some Ryle commentators who explicitly deny that ‘knowing how’ to do something need involve the physical ability to do it (Stanley & Williamson, 2001), at least when one takes the term in a weak (but not unimportant) sense (Carr, 1981), and who analyze ‘knowing how’ as a form of propositional knowledge, after all (Stanley and Williamson), or as ‘practical discourse’ (Carr) – the latter emphasizing the practical, for sure, but construing the phenomenon as discourse, nonetheless.

Arguably, the intellectualist bent is present even in Dreyfus’ writings, though in a somewhat different way, despite his Merleau-Pontian focus on the significance of the body and his Heideggerian stress on being-in-the-world as phenomenologically and existentially prior to thinking about the world. Thus, in Dreyfus’ articulation of the significance of the body, the primary focus is on its role in perception and sense-making – on its giving the “indeterminate, global anticipation”, as he puts it, required for pattern recognition (Dreyfus, 1979, p. 237) and its supplying “the felt equivalence of our exploratory skills” which “enable us not only to recognize objects in each single sense modality, but… [to] see and touch the same object [i.e. to recognize objects trans-modally].” (ibid, p. 249). That is, his preoccupation with the body is first and foremost as the provider of an (embodied) approach to the world and the things encountered there. He does not concern himself much with what is required for the actual performance of the movements themselves. There are a few places where he does allude to the question of how we master bodily movements, as when he says that once a skill has been learned, “we seem to have picked up the muscular gestalt which gives our behavior a new flexibility and smoothness” (ibid.). However, such comments are made in the context of explaining perception and sense-making, and/or as part of an argument against rule-following as the basis of performing a skill, not as part of analyzing the movements as phenomena in their own right. Similarly, the skill acquisition model developed together with his brother describes how the situation is made sense of and actions decided upon at different skill stages (through deliberation at early stages and intuitively at the last one). It does not touch upon the question of what is involved in making the actual movements, nor does it take into account the fact that some people may be physically unable to undertake what they intuitively perceive to be the right course of action or consider how they actually learn to perform the skilled movements. This is illustrated in their lack of distinction – to the extent that they do not even comment on the differences – between intellectual skills such as chess playing and ones that involve the body more such as car driving. In a sense the skill acquisition model is not about skill acquisition at all, but about the perception of the situation one has once different levels of skill – bodily or not – have been acquired. In sum, though Dreyfus certainly does not neglect bodily actions as Ryle and the Scandinavian interpreters of Wittgenstein do, his interest in them are first and foremost for the understanding they provide and exhibit. The actual acting is taken for granted.

In contrast, on my rendering of practical knowledge, the acting itself is significant. Stressing that the “weiter wissen” is there in the very doing thus not only signifies the ‘tacit feel’ for how to proceed, but the fact that the ‘tacit feel’ is bound up with the carrying out of the movements themselves. Further, because every act involves our body in some way, all skills have a bodily aspect to them which cannot be abstracted away as inessential or extrinsic to what constitutes the skill. In a sense, all skills are bodily skills, though of course the bodily aspect is more pronounced in skills like fishing and boxing than they are in so-called intellectual skills like telling a joke or playing chess. The telling of a joke will be spoiled if one laughs too soon or too much, if the punch line is inaudible or stressed too much, if one’s posture is not supportive of the joke in the appropriate way etc. Part of the skill of telling the joke is getting these bodily aspects right. As is generally the case with practical knowledge, the ‘getting it right’ cannot be explicated fully in a rule, and even if it could, one would still have to learn to actually do it in practice. That is, the ‘tacit feel’ for what is right must be there in the pitching of one’s voice in the appropriate rhetoric manner, in the emphasizing the point in the right way, in the controlling of one’s facial expression and bodily movements so as to support the joke and so on. This is one of the reasons why what is seemingly the same joke (the same exact words) may be funny when told by one person and not when told by another. Similarly, as Dreyfus argues, the expert chess player reacts intuitively to the layout of the board (Dreyfus & Dreyfus, 1986), experiencing the right move as ‘drawn out’ of him by the situation (Dreyfus, 2002, pp. 372, 380). What more precisely constitutes the right physical movements depends on how the chess game is implemented – whether on a physical board where one moves the pieces by hand; on the computer screen where one moves the virtual pieces by striking the keys, clicking the mouse or issuing verbal orders; or perhaps by someone else upon one’s request.

Of course, in one sense – at one level – the chess game will be the same and can be considered and discussed irrespective of its material realization. This is the level of the game itself, within the field of meaning set by chess as an activity. In Articles 7 and 10 I analyze this level as the domain-internal level and contrast it with the activity-internal context level and the activity-framing context level. My point there is that in any situation demands, possibilities and restrictions pertaining to each of the three levels will interact to form a complex whole, and that competence in the situation will consist in the ability to respond appropriately to this complex whole. Translating the point to the issue at hand, the material realization of the chess game is surely not irrelevant to the actual person playing it: Unless he has learned to master computer representations of chess games, the chess expert will not respond intuitively to a move in a virtual game, even if he would be able to do so, had the same move been made on a physical board. The actions ‘drawn out of him’ by the situation will thus not be ‘abstract’ actions within the domain-internal level; they will be the actual actions made in response to the situation’s actual complex of interacting demands, possibilities and restrictions, or “requirement characteristics” as I term them in Article 7. Hence, for the real chess player, the ‘tacit feel’ of how to go about the game is bound up with actual doings, i.e. with the way the domain-internal level is materially realized in the chess games he participates in. Immediate, intuitive response for him is a response in materially realized chess piece movements.

In the same vein, the intellectual skills of reading, writing, solving math problems etc. require bodily skills which differ with the material realization (including virtual ones) of the text or problem13. Though a given math problem (say 343×822, to use Wertsch’s example) at the domain-internal level requires a certain set of abstract calculations which will lead to one specific solution; actually performing the calculations will require very different bodily actions with an abacus, with pen and paper, and with a computer. One may be proficient in dealing with the problem in one type of material realization, but unable to solve it utilizing another one. In one (narrow) sense, the skill of multiplication, like the skill of playing chess, may be said to be constituted solely by abilities at the domain-internal level. However, as in the case of chess, this is a highly abstract sense, since exercising abilities at the domain-internal level will always take place in real situations with actual material realization possibilities. The degree to which one will in point of fact exercise one’s abilities at the domain-internal level will depend on one’s mastery of the realization possibilities. This is so even in the limiting case of performing the multiplication ‘in the head’ where one at the very least has to perform the bodily action of shutting out disturbing noises. In addition, doing a problem ‘in the head’ often mimics doing it physically to some extent – one may for example imagine writing down the numbers or moving the balls on the abacus.

In Article 7 I argue the extended point that displaying competence within the areas tested for by OECD’s Programme for International Student Assessment (PISA) – literacy, numeracy, science, and so-called cross-curricular problem solving – amounts to responding appropriately to the complex whole of interacting requirement characteristics posed by the PISA survey items as they are realized in the context of an international assessment. PISA is therefore not, I argue, testing students’ abilities for applying literacy, numeracy etc. skills to ‘real world problems’ (as alleged by PISA). Rather, PISA is testing how well students exercise competence within PISA focus areas in one very special kind of ‘real life’ situation. My point here is the at once more restricted and wider one that a person’s skills within literacy, numeracy etc. will always be materially realized so that the practical knowledge of for instance multiplication is constituted in response to the way domain-internal requirement characteristics are framed by the activity forms in which they are normally met. The practical knowledge of multiplication will in other words be bound up with the ways it is actually carried out.

Finally, my comments above on Dreyfus’ partial neglect of action, due to his focus on the body’s role in perception and sense-making, do not mean that I for my part take perception to be a relatively passive process which does not involve any bodily skills. Quite the contrary; I think Dreyfus, building on Merleau-Ponty, makes a very clear case that perception is an activity – not just in the intellectualist, Kantian/Piagetian sense of actively apprehending ‘sensory experience’ within pre-given categories or schemas, but in the sense that perception involves moving one’s body in certain ways. One example is Merleau-Ponty’s of feeling and recognizing silk: To learn to do so, one must learn to move one’s hand in a particular way and to have certain expectations of what to feel. Without this skill, the perception of the cloth is less distinct and one may not be able to discern it from nylon or cotton. As Dreyfus notes, the body’s role in auditory, gustatory, and tactile perception is most obvious; yet it is no less significant in visual perception: “…seeing, too, is a skill that has to be learned. Focusing, getting the right perspective, picking out certain details, all involve coordinated actions and anticipations.” (Dreyfus, 1979, p. 249) This is clearly seen in Polanyi’s example, referred to above, of discernment of relevant traits in medical X-ray pictures of the lungs. To the layman and the novice student, all that is visible in such a picture, are the ribs. Making out the lungs, discriminating signs of illnesses and distinguishing such signs from traits due to natural variation, involves learning bodily how to direct, focus and move one’s gaze.

Actually, it is only if one construes perception as a series of discreet units of sensuous stimulation – if one takes what Gibson calls “snapshot vision” to be fundamental to visual perception (Gibson, 1986, p. 1) and correspondingly for the other senses – that the idea of perception as non-related to movement seems credible at all. This is indeed the approach of most traditional psychology and epistemology, in agreement with the ‘mirror of nature’ view of perception and knowledge referred to above (R. Rorty, 1980). However, as Gibson points out, the role which vision plays for animals, including humans, is in the first instance the one of providing continuous guidance during locomotion where the animal moves its head continuously, both as part of moving around and as part of orienting itself to all sides. Very rarely are we waiting passively for pictures to be presented to us. Instead, perception is a natural part of actively engaging in the world. Even when we undertake visual perception as an activity in its own right (as we might do in a museum, planetarium or X-ray lab), we have to move actively towards and around the objects we wish to perceive in order to obtain the right kind of perceptual relationship to them. As Gibson puts it “…we must perceive in order to move, but we must also move in order to perceive” (Gibson, 1986, p. 233) and therefore, the primary form of vision is “ambulatory vision” (p.1). In the case of the planetarium and the X-ray lab, technological artifacts have to be manipulated (under guidance of perception) for us to obtain the images that we wish to perceive. In this respect, Polanyi’s X-ray example is misleadingly simplistic in that it does not take into account the complex of actions necessary for the viewer to get to the ‘snapshot’ situation of confronting the X-ray pictures.

Summing up, my claim in the preceding sections is that practical knowledge is a tacit, embodied understanding of ‘the right way to proceed’ present in the very doing itself and therefore fundamentally bound up with the actual bodily actions required for proceeding in this way, given the material realization of the activity. In this sense, all skills have a bodily aspect to them, because they all involve mastering certain bodily movements. This is not to deny that the bodily aspects are much more pronounced in some skills than in others. On the contrary, some skills quite obviously are immensely more demanding to master than others in respect of pure movement, just as, conversely, some skills involve much more cognitive work than others, i.e. are more intellectual. The point is thus only to counterbalance the intellectualist bent of most philosophical renderings of practical knowledge, precisely by calling attention to the fact that i) analysis of practical knowledge in terms of its intellectual aspect is not exhaustive and that ii) even so-called intellectual skills are inherently exercised by the body in movement in response to actual realization possibilities.

Finally, our practical knowledge constitutes another source of ‘semantic content’ for our words on a par and to some extent intermingled with (cf. below) experiential knowledge. Again, not in the sense of mental representations of the practical knowledge involved – claiming that would be a recurrence both to the representationalism argued against in the last section and to the intellectualist negligence, identified above, of the way knowing is inherently involved in actual doing. Instead, the familiarity of action supplies a motor space of meaning which is drawn upon in understanding propositions set out by others and oneself. To give a simple example: for someone who is able to touch type, the question of where a certain letter is placed on the keyboard will present itself immediately as a question of movements of his hands. Indeed, he may not be able to answer it at all before he has physically simulated the finger movements necessary to locate the letter on the keyboard. Similarly, when someone describes a car incident of e.g. a truck nearly pulling out into her car, the experienced driver will have a much fuller, action-based and action-related understanding of what was at stake than the person who has always only been a passenger. This is because the motor space of driving, including the actual movements necessary, time to respond in critical situations, reaction time of the car, presence of other cars etc., is practically known to the former but not to the latter. Not as a set of mental representations but as a practical sense of the degree of danger present, of what one does in such a situation, and what one’s chances are of avoiding a crash, given the options one realistically has.

A last example, more pervasive in terms of the domains of propositional knowledge affected, are the bodily metaphors which Lakoff and Johnson have analyzed. One of their cases concerns the dependence which several basic axioms in classical logic have on a container metaphor. More specifically, they focus on “Pv–P” (law of the excluded middle); “– – P=P” (law of double negation); and “xRy & yRzxRz” (law of transitivity) (Johnson, 1987, p. 39f; Lakoff & Johnson, 1999, p. 31f). These laws, they argue, build on a basic understanding of spatial containers: either something is inside or outside the container; if it is not outside (=not-inside), then it is inside; and if it is inside a container placed inside another container, then it is inside this container, too. The container metaphor, they further point out, stems from experience: we are ourselves ‘containers’ into which food and air is put and out of which waste is secreted; we live a large part of our lives in containers (houses, rooms, cars, gardens, woods etc.) and very many practical activities involve dealing with containers (pouring milk out of bottles, adding ingredients to pots, putting books in bags, taking clothes out of cupboards etc.). Further bodily based metaphors discussed by them include “More is Up” “Important is Big”, “Change is Motion”, “Understanding is Grasping” (Lakoff & Johnson, 1999, pp. 53-54). Now, in their later writings, they explain these metaphors in a very empiricist-associanist way which seems somewhat at odds with earlier more holistic descriptions at the phenomenological level of metaphorical structuring of a target domain in the light of a source domain14. Be that as it may, in my view these metaphorical structurings are examples of the way the world meets us directly as significance pre-structured on the basis of our bodily engagement in the world and of what is important to us as bodily beings in this engagement. Because our bodily, practical understanding of the world is the pre-reflective outset for whatever we do become aware of and reflect upon, it will form, frame and resonate in this awareness. It is therefore at the heart of our more abstract sense-making. Thus, given for instance the significance of containers for us in our lives and activities, we engage in the world with expectations and anticipations of phenomena meeting us in, having the form of, and structured with the spatial-logic of containers. Further, precisely because of its significance for us, this pre-reflective understanding of spatial-logical relations is then metaphorically mapped onto abstract logical relations so that our dealings with the latter become structured by and resonate with our practical knowledge of the former. In this way, abstract logical propositions make sense to us, we draw logical conclusions, and we posit logical laws on the (metaphorical) basis of our practical sense of relations between spatial objects. Quite generally, the metaphorical structurings pointed out by Lakoff and Johnson on my construal work, not as un- or subconscious schematic representations structuring our understanding (as they claim, especially in their later writings), but by words designating the target area setting practical knowledge within the domain area in resonance and thereby supplying a motor space which metaphorically structures and informs the target domain. Hence, together with the other examples presented above, these metaphorical structurings illustrate how practical knowledge may supply semantic content to our words. They thereby further elucidate my claim in the beginning of the introduction that our words are ‘filled up’ or resonate with immediate, pre-reflective experiences of living and acting in the world.

  1. Wertsch makes a related point in discussing how we learn to use a specific syntax to solve multiplication problems (Wertsch, 1998, p. 28ff). He is, however, more concerned with calling attention to the mediational nature of the tool and its cultural dependency than with our bodily appropriation of it.[]
  2. They write e.g. that “Whenever a domain of subjective experience or judgement is coactivated regularly with a sensorimotor domain, permanent neural connections are established via synaptic weight changes.”(Lakoff & Johnson, 1999, p. 57). This is a neurological version of the classical Humean empiricist associanism according to which we cognitively establish relations between ideas by noticing “constant conjunction” of events. The also classical Merleau-Pontian phenomenological objection is that this explanation begs the question in that it presupposes a prior understanding of relevance and significance according to which two events are classified as ‘the same’ (Merleau-Ponty, 1962, p. 14). Given this prior understanding, there is on the other hand no need for a further explanation in terms of constant conjunction – the relationship has already been established by the former. According to Merleau-Ponty we have this prior understanding in virtue of always already being bodily engaged in the world. In an earlier work, Johnson states his position in a way which seems much more in accord with this Merleau-Pontian view: “Our reality is shaped by the patterns of our bodily movement, the contours of our spatial and temporal orientation, and the forms of our interaction with objects” and “in order for us to have meaningful, connected experiences that we can comprehend and reason about, there must be pattern and order to our actions, perceptions, and conceptions. A schema is a recurrent pattern, shape, and regularity in, or of, these ongoing ordering activities. These patterns emerge as meaningful structures for us chiefly at the level of our bodily movements through space, our manipulation of objects, and our perceptual interactions.” (Johnson, 1987, pp. xix, 29, italics removed). He does ascribe the pattern to Kantian cognitive schemas, though, rather than as ways the world make direct sense to us on the basis of our bodily involvement in it, which would be how Merleau-Ponty would see it (as would Dreyfus, Dreyfus, 2002, p. 373).[]