2.1 Experiential knowledge


The aforementioned §78 of Philosophische Untersuchungen runs as follows:

Vergleiche: wissen und sagen:

wieviele m hoch der Mont-Blanc ist –

wie das Wort „Spiel” gebraucht wird –

wie eine Klarinette klingt.

Wer sich wundert, daβ man etwas wissen könne, und nicht sagen, denkt vielleicht an einen Fall wie den ersten. Gewiβ nicht an einen wie den dritten.7 (Wittgenstein, 1984a, § 78)

According to Tore Nordenstam, Wittgenstein is here pointing out the existence of three fundamentally different types of knowledge, namely ‘propositional knowledge’/‘theoretical knowledge’, ‘skill’/‘practical knowledge’ (or ‘procedural knowledge’), and ‘knowledge of familiarity’(Nordenstam, 1983, p. 21). Nordenstam develops his epistemological position – and his take on Wittgenstein – from this outset, claiming that only the first type of knowledge may be articulated in words (is expressible) whereas the latter two are tacit (inexpressible). Other Scandinavian readers follow him in this interpretation (e.g. Göranzon, 1987; Rolf, 1989, though Rolf stresses that the three forms do not make up a strict typology). Now, depending on how one construes the ‘knowledge of familiarity’ postulated here, there might seem to be a straightforward contradiction between the articulation of this concept and Wittgenstein’s contention that we cannot refer to private experiences. Thus, one way of understanding it is as referring to mental states representing the phenomenal quality of experiences (the ‘what it sounds like’ in the example given); to the “quale” of the experience to use the philosophical jargon (e.g. Churchland, 1985; Dennett, 1991, 1992; Jackson, 1982; Nagel, 1974, 1986; the term was first coined by Peirce in Peirce, 1866/1982). Judging from the further examples Nordenstam provides, he interprets Wittgenstein this way. This is perhaps most evident in the description which he includes from Peter Gullers of what constitutes the latter’s professional knowledge as a photographer and how putting this knowledge to use allows him to take better pictures with a manual camera than it is possible with an automatic one. Gullers – and Nordenstam with him – here refers to “pictures in the memory”, (minnesbilder) and “concrete experiences in instances of picture-taking” (konkreta upplevelser og erfarenheter vid fotograferingstillfället) as ‘knowledge of familiarity’. Gullers describes how he (without being consciously aware of it) compares present light conditions to the ‘pictures’ he has in his memory of previous ones. Such ‘pictures in the memory’ would seem to be exactly the type of mental states which Wittgenstein has been taken to argue elsewhere in Philosophische Untersuchungen do not have any significance for meaning ascription and rule following (e.g. §§ 151, 152, 153, 180, 191, 196, 293). Understood in this way, that is, Wittgenstein is in § 78 proposing as an example of inexpressible knowledge precisely the kind of ‘beetle in a box’ which he denies the meaningfulness of in §293: The ‘quale’ of the clarinet sound appears to be a ‘thing’ of which each person has his own in a private box that no one else can look into – which might therefore be different from person to person or even non-existent for some. Wittgenstein’s point in § 293 is that since there would be in principle no way to ‘get at’ the thing for the language community, it could have no role to play – could not supply reference and meaning – in the language. To the extent that the word ‘beetle’ was in point of fact used in the community, it would be for something else, perhaps the box itself or certain activities involving it. In sum, this reading of Wittgenstein entangles him in self-contradictions.

As I said, I do think there is a tension present in Wittgenstein’s writings on this issue – i.e. between advocating the significance of tacit experiential knowledge and arguing against mental representationalism. That is, I agree with Nordenstam that Wittgenstein is referring to a form of knowledge which is inexpressible in words and which is characterized by being a ‘knowledge of familiarity’ – or a ‘knowledge by acquaintance’ as I prefer to call it, to situate the term within other philosophical debates (Churchland, 1985; Jackson, 1986; Russell, 1910-11, 1912; Williams, 2001; Russell introduced the term) and to indicate that one can have experiential knowledge within areas with which one is not all that familiar. However, I do not think Wittgenstein is committed to the view that experiential knowledge consists of mental states, nor do I find it appropriate to construe it as such, though I don’t question that mental representations may be involved in specific instances of it. Conversely, I do not read Wittgenstein in the vein of some interpreters (Baldwin & Baldwin, 1978; Malcolm, 1971) who see him as an ontological behaviorist proposing the claim that there exists no such thing as mental states. In several paragraphs he quite clearly acknowledges that we may mentally represent things and states of affairs to ourselves (Wittgenstein, 1984a, e.g. §§ 313, 386, 398). His point is rather that it does not make sense to speak of ‘inner processes’ as objects, and furthermore that mental representations aren’t private and that they do not have the role in constituting language which they have been accorded in the philosophical tradition from Descartes and onwards. I argue this point in Article 5.

The first thing to note here is that experiencing the sound of a clarinet is not having an inner feeling, sensation, or mental representation. It is attending to something in the world. Knowing what it sounds like is anchored in this relation; it is not the possession of a certain ‘memory picture’ of the sound. In particular, it is not the possession of a ‘data file’ stored in the ‘library program’ of the mind, which is brought out for inspection upon someone mentioning the clarinet or the sound of it. Actually, I may not be able to conjure up a ‘mental picture’ or do a ‘mental replay’ of the sound at all. Being able to do so is not a necessary condition for knowing ‘what it sounds like’ or even a reliable indicator of knowledge.

These statements of course run counter to the tradition starting with Descartes, carrying on through classical empiricism, over Kant and into logical positivism and phenomenalism of portraying perception as having an ‘idea’, ‘percept’, ’sense-datum’, ’representation’ or the like in one’s mind (R. Rorty, 1980, provides an overview of this development). They concur with Rorty’s point that this tradition is fundamentally in the grips of a metaphor of perception as the reflection in the ‘mirror of the mind’ of the phenomena of the ‘outside world’ and of knowledge as the possession of non-distorted reflections (R. Rorty, 1980). In contrast to Rorty, however, the dismissal of this metaphor and the epistemology it implies does not lead me to give up on ‘knowledge by acquaintance’. Nor do I think Wittgenstein did, though Rorty seems to claim otherwise (cf. e.g. p. 169)8. Rather, it leads me to look for another interpretation of what Wittgenstein could be referring to in § 78.

Here, Wittgenstein would in all probability reiterate his comment of “…denk nicht, sondern schau!” (Wittgenstein, 1984a, § 66)9, to make us look at how the term “the clarinet’s sound” and the phrase “wissen wie eine Klarinette klingt” are used in our language games, as an alternative to speculating philosophically about the possible ontological realization of ‘knowledge by acquaintance’ as mental representations. If we do indeed take such a look, we come upon phrases such as “the sound is somewhat like that of the oboe, though with a fuller timbre”; “No, I don’t know what a clarinet sounds like – I have never heard it played. Can you play it for me to hear?”; “the clarinet sounded a bit out of tune”; “if you buy a very cheap clarinet, it’ll have a very shallow sound compared to an expensive one” etc. That is, we notice that in our everyday language perception of the sound is on the one hand treated as an integral part of our language game, as publicly accessible to all whose hearing is not impaired (i.e. not as a private ‘beetle in a box’), but on the other hand in such a way as to recognize that having been in and experienced certain situations is necessary for one to know what is being referred to (the reference here being to something in the world, not in the head). That is, the knowledge is ‘tacit’ in the sense that you cannot communicate it to someone who has not been in and experienced these situations. If you have never heard the clarinet played, you do not know what it sounds like, and others cannot really convey it to you, except by comparing it to other sounds (like the one of the oboe) which you are then in turn required to ‘know by acquaintance’. Perceiving, sensing, and feeling are, I submit, to Wittgenstein integrally a part of participating in the language games of one’s form of life – and that we can perceive, sense and feel are (under normal circumstances) something which it does not make sense to doubt((For which reason we also according to Wittgenstein cannot – under normal circumstances – say that “I know that I can perceive/sense/feel”. But, as he says concerning “knowing that there is a chair”: “Mien Leben zeigt, daβ ich weiβ oder sicher bin, daβ dort ein Sessel steht, eine Tür ist usf. Ich sage meinem Freunde z.B. “Nimm den Sessel dort”, “Mach die Tür zu”, etc., etc.“ (Wittgenstein, 1984b, § 7) (“My life shews that I know or am certain that there is a chair over there, or a door, and so on. – I tell a friend e.g. “Take that chair over there”, “Shut the door”, etc. etc. (Wittgenstein, 1969/1979). That is, though it does not under normal circumstances make sense to state one’s knowledge as knowledge because ‘having knowledge’ presupposes the meaningfulness of ‘being in doubt’, our actions show a certainty which we cannot express. A similar certainty of the fact that we and others can perceive, sense, and feel is displayed in our actions towards one another.)). It is pragmatically presupposed in what we do and say. Though the tension between the acceptance of tacit experiential knowledge and the rejection of the linguistic role of mental representations is definitely there in Wittgenstein, I think it can be resolved in this way by on the one hand stressing that experiential knowledge is not constituted by mental representations but by practical involvement in/attentive interactions with given situations – that experiential knowledge is knowledge of the phenomenon in the world not of a mental representation of it. And on the other hand pointing out that such practical involvement/attentive interaction, though in principle ‘publicly accessible’ (because they are of phenomena in the world), will not de facto be or have been undergone by everyone: Depending on the distribution and nature of the phenomenon in question, some or many people may not have been in the situations required for experiencing them, or they might lack the ability to perceive them (e.g. because of hearing impairment). Given this interpretation, the linguistic role which Wittgenstein denied to mental representations can furthermore without self-contradiction be accorded to the practical involvement/attentive interaction which makes up the experiential knowledge: Hearing the clarinet supplies meaning – semantic content as I call it in Article 6 – to the phrase “what the clarinet sounds like”. It is on this interpretation, finally, that I can claim, as I did above, to take a basically Wittgensteinian outset when arguing that our words are ‘filled up’ or resonate with meaning from our immediate, pre-reflective experiences of living and acting in the world.

Still, a more positive description of what experiential knowledge is has yet to be provided. In developing such a description, as I do in Article 6, I proceed well beyond what Wittgenstein would probably agree to, and draw on my integration of his views with phenomenological considerations. It is suggestive to observe that in contrast to the ability to ‘mentally replay’ the sound of the clarinet, being able to recognize the sound upon hearing it (under normal hearing and listening conditions) is both a necessary condition and a reliable indicator of this form of knowledge, though it isn’t constituted by this ability. This signifies that knowledge by acquaintance is rooted in the individual’s relation to the world, not in his/her mind, and certainly not solely there. As Dreyfus puts it: “… my memories are inscribed in the things around me… My memories are stored in the familiar look of a chair or the threatening air of a street corner where I was once hurt.” (Dreyfus, 1979, p. 266) Of course, at the level of what Dreyfus calls “the energy exchange” where “physical processing of physical energy” takes place (Dreyfus, 1979, pp. 266, 268), experiencing something new (like hearing a clarinet for the first time) may result in the establishment of neuronal connections in the brain. But the point is that this is indeed a physical, not a psychological process, and that therefore, in particular, it is not one of information processing or mental representation. Whatever the neuronal preconditions, the result of the ‘energy exchange’, as Dreyfus stresses, is the well-known world of meaning in which we are always already immersed; not an inside middle one of mental representations. This claim in turn is an elaboration of the basic Merleau-Pontian view that our experience of the world is first and foremost one of active bodily involvement in it, not of mental representation of it and that mental representation is only a secondary, derivative attitude towards the world, dependent on the primary, involved one. As Merleau-Ponty puts it, my body is “open on to the world, and correlative with it” (Merleau-Ponty, 1962, p. 143)((”… ouvert sur le monde, corrélatif du monde.” (Merleau-Ponty, 1945/1993))). On this view, listening to clarinet music is not in the first instance a matter of taking mental note of its sound, it’s a matter of engrossing oneself in it, i.e. to give oneself over to the music and “be one with it”, rather than experience it as the subjective side of a subject-object-divide. One could almost say that experiencing the sound is ‘being out there with it’, rather than ‘in here with it’ – except that this terminology seems to fundamentally accept the subject-object dichotomy, but just places the experience on the other side of it. Instead of acknowledging the Heideggerian-Merleau-Pontian point that immediate, pre-reflective experience is a precondition of the secondary move of reflectively distinguishing between subject and object – between, in this instance, the listener and the clarinet sound listened to.

Building on this phenomenological elaboration of the Wittgensteinian outset, my claim (as put forward in Article 6) is that experiential knowledge of a phenomenon should be understood as a ‘felt holistic bodily responsiveness’ encoded in one’s bodily handling of the situation in which the phenomenon is met. The ‘bodily responsiveness’ may involve or conjure up one or more mental states, but it doesn’t necessarily do so, and in any case will not be reducible to these states. Thus, upon recollecting a former incident, one’s whole body is ‘set in motion’ as an attitude towards the world, including whatever desires, aims, emotions etc. it might involve((These aims, desires, emotions etc. will be the ones which one now senses as the ones one had at the time of the incident. They need not coincide with the ones which one would now have towards this incident, but conversely they may also be transformed from the ones one actually had at the time.)) – the incident is ‘there with me in my going about the world’ so to speak, rather than ‘in my head’. Equally, coming across a phenomenon which one has experienced at an earlier time of life may transform one’s take on and experience of the world here and now. Hearing a piece of music from one’s childhood, for example, will ‘take one back’, not in mental representation, but in one’s being-in-the-world, i.e. in one’s emotional and perceptual relation to the situation of now as if it were the past.

Of course, the degree of emotional involvement will vary with the phenomenon concerned and with what happened at earlier occasions. It will vary, one might say, with its scope and the degree to which one’s experiential knowledge of it is nested within one’s experiential knowledge of other phenomena. Thus, for many people, knowing what the clarinet sounds like, what mauve looks like, or what kangaroo tastes like will not involve many emotions. For others, they might be the phenomena which ‘take them back’ to childhood birthdays, make them ‘relive’ their divorce, or feel again the frustration of quitting tobacco. In such cases, the simpler experiential knowledge of e.g. the clarinet’s sound will be nested within the much more complex knowledge of ‘what it is like’ to celebrate birthday as a child, live through a divorce, or fight to quit one’s addiction to smoking. These latter examples on the other hand illustrate quite clearly that knowledge by acquaintance cannot be reduced to mental representations. In living through a divorce, the complex mixture of feelings of anger, sorrow, remorse, and (perhaps) freedom and relief; transformations in interpretations of past events in the light of the marriage’s failure; changes in personal, social, economic status and perspectives etc. is a heterogeneous whole where each aspect delimits and is itself delimited by the significance of the others. Mental representations may certainly be invoked in actual situations of remembrance, but instead of constituting the remembrance, they will then be part of ‘living the emotions’, as when one for instance broods over the recalled image of one’s former partner as part of letting the feeling of anger fill one up. Rather than be constituted by mental representations, the experiential knowledge resides as a multifaceted feeling in the body as a whole, in one’s interactions with the world, showing up at later stages of life e.g. in the ways new situations meet one as having or not having ‘romantic potential’, new partners as being reliable or not, etc. When directly called upon in remembrance, the experiential knowledge may color the present situation positively or negatively, giving it – and one’s actions – an ‘air’ of e.g. desolation or freedom, and invoking corresponding feelings of e.g. restlessness, anger, or animation.

As should be clear from the above, experiential knowledge is not a matter of ‘percepts’ of the world, acquired through relative passive ‘opening one’s mind’ to it. On the contrary, as shown by Polanyi’s example, referred to in the last section, of discriminating traits on an X-ray picture of the lungs, developing experiential knowledge may be a highly specialized process, requiring prolonged education of perception. It may thus take years of training to get to know certain rare and illusive signs of illness ‘by acquaintance’. I shall return to this in the next section where I discuss the skills associated with perception.

  1. Compare knowing and saying: how many feet high Mont Blanc is – how the word “game” is used – how a clarinet sounds. If you are surprised that one can know something and not be able to say it, you are perhaps thinking of a case like the first. Certainly not of one like the third” ( Wittgenstein, 1958). []
  2. Incidentally, this claim from Rorty is illustrative of the way the Scandinavian stress on tacit knowledge in the reading of Wittgenstein differentiates this reception from the one he has had in the Anglo-American world. []
  3. “…don’t think, but look!” (Wittgenstein, 1958) []