The Emergence of Wittgenstein’s Views on Aesthetics in the 1933 Lectures

In this paper I offer a genetic account of how Wittgenstein developed his ideas on aesthetics in his 1933 lectures. He argued that the word ‘beautiful’ is neither the name of a particular perceptible quality, nor the name of whatever produces a certain psychological effect, and unlike ‘good’, it does not stand for a family-resemblance concept either. Rather, the word ‘beautiful’ has different meanings in different contexts as we apply it according to different criteria. However, in more advanced regions of aesthetics the word ‘beautiful’ ceases to play an important role. Instead, we judge things to be more or less correct according to genre-specific standards or criteria, which in an aesthetic discussion are presupposed, rather than argued for. Finally, Wittgenstein came to realise that providing support for an aesthetic appraisal according to some given criteria is not the only and perhaps not even the main focus of aesthetic discussion. More interesting to him became the idea of a puzzle or perplexity in aesthetics, which he discussed in greater detail in his 1938 lectures.

Aesthetic questions, however, are often regarded as comparatively unimportant or secondary. To elevate aesthetics to the same level of dignity as ethics might be at least part of the point of Wittgenstein's bold identification of the two in the Tractatus. In that he would display his cultural background, the high esteem in which aesthetic questions were held both in fin-desiècle Vienna in general and in the Wittgenstein family in particular. 6

II
The Tractatus idea of juxtaposing ethics and aesthetics continues in the lectures of the May Term of 1933 (that is, in Wittgenstein's so-called middle period, characterised by some radical criticisms of his earlier philosophy and increasing emphasis on the manifold uses of language). Although ethics and aesthetics are no longer boldly identified, Wittgenstein repeatedly declares them to be conceptually so similar as to allow them to be discussed jointly and switches freely from one to the other to illustrate parallel conceptual points: 'Practically everything I say of "beautiful" applies in a slightly different way to "good"' (M, p. 339). 7 The philosophical starting point of the May Term lectures is verificationism and the observation that the meaning of a word is 'the way in which it is used' (M, p. 308). Wittgenstein rejects the idea that nouns must stand for 'something at which you can point' (M, p. 313) and then introduces ideas, later to be developed in the Philosophical Investigations, of how the meaning of a word can be far more flexible, less sharply defined than one may be inclined to believe. Sometimes we change the meaning of a word in the course of a conversation, ' as we go along' (see PI,§ 83). Meanings are often vague, so that no precise boundaries can be drawn (M,p. 313;see PI,(76)(77)88). And he introduces the idea of a family resemblance concept, using the same example as in the Investigations (PI, § § 65-71): the concept of a game is not held together by a set of qualities all games have in common. 8 Rather, we explain the concept by giving an open-ended list of examples, each being similar to some others in some respect, although there may be no one similarity shared by all of them (M, pp. 323-24). He then applies these insights to the word 'good', 9 and later also to 'beautiful'. Furthermore, he suggests that the word 'good' may be used attributively: what counts as a (morally) good x, may partly depend on what kind of thing x is (M, p. 325). He also suggests that the meaning of the word 'good' is determined by what in discussion we present and accept as reasons for calling something good (ibid.).

III
After that sketch of his new undogmatic and flexible ideas about meaning (and a digression on Frazer's anthropological explanations), Wittgenstein focuses on two prevalent accounts of meaning, first again with respect to the word 'good', before changing to a more detailed discussion of the word 'beautiful'. The ideas in question are, in the case of beauty, that (i) a certain perceptible quality (or set of qualities) is what all beautiful things have in common, or that (ii) beauty is the causal power to give us certain pleasant sensations. As a case of comparison, Wittgenstein considers the concept of elasticity: If I want to know whether a rod is elastic I can find out by looking through a microscope to see the arrangement of its particles, the nature of their arrangement being a symptom of its elasticity, or inelasticity. Or I can test the rod empirically, e.g., see how far it can be pulled out. The question in ethics, about the goodness of an action, and in aesthetics, about the beauty of a face, is whether the characteristics of the action, the lines and colours of the face, are like the arrangement of particles: a symptom of goodness, or of beauty. Or do they constitute them? a cannot be a symptom of b unless there is a possible independent investigation of b. (

IV
On the first view (i), beauty is a quality all beautiful things have in common: ' an ingredient in beautiful things' (M, p. 332).
Wittgenstein hints that the idea of a common ingredient is mistaken by pointing to the difference between colours and pigments (ibid.). The point is perhaps even clearer in the culinary case, where the word 'ingredient' is naturally used. Two dishes may have a taste note in common -for example, they both taste of vanilla -without having any actual ingredients in common. 'Vanilla flavour' ice cream is typically not made from real vanilla.
But even a certain perceptible quality, produced by whatever ingredients, is clearly not what all beautiful things have in common. Just as not all delicious food tastes of vanilla, or whatever particular taste you care to mention, not all beautiful paintings contain certain colours or shapes. Less crude is the idea that what beautiful things have in common are neither particular ingredients nor a particular perceptible quality, such as a colour, but a second-order quality, for example, a relational quality between first-order qualities. Thus it was suggested (by Aristotle) that beautiful things show order and symmetry or (by Francis Hutcheson) that beauty is 'uniformity amidst variety'. But it is easy to show that such definitions won't do. They are either too vague and inclusive (virtually everything can be said to display some uniformity amidst variety) or fail to cover all instances of beauty. As Kant had already concluded in 1790, no such descriptive definition of 'beautiful' seems possible. Beauty cannot be identified with any particular set of perceptible qualities. So, can I be aware of all the details of a painting and yet not know whether it's beautiful or not? Yes, because 'no arrangement is beautiful in itself. The word "beauty" is used for a thousand different things' (AWL,. 10 The word 'beautiful' is applied in extremely heterogeneous domains (M, p. 333): the human face, a landscape, a painting, a piece of music, even a mathematical proof, can all be beautiful, yet it is hard to see how they might all display the same perceptible quality or feature of qualities. Certain arrangements of shapes and colours are called beautiful, but not in themselves: only in a suitable context. In another context the same arrangement might be rejected as inappropriate.

V
That might suggest the other option: a dispositional analysis of the concept of beauty. As we aren't able to identify beauty with any specifiable (kind of) arrangement of shapes and colours -just as ' elastic' doesn't mean having a specific molecular structure -it would appear that we have to accept the other possibility, namely that such perceptible features are only a symptom of beauty. That is, it may be a well-supported empirical generalisation that whenever something has a certain arrangement of perceptible features it will be found beautiful. But then, there must be an independent criterion of what it actually means for something to be beautiful, corresponding to the possibility of bending a rod as what constitutes its elasticity.
In the following lectures, Wittgenstein considers the idea that beauty is what gives us a certain feeling, say, pleasure (M, p. 339). Then aesthetics would turn out to be a branch of psychology: 'in comparing musical arrangements, for example, one [would be] making a psychological experiment to determine which produces the more pleasing effect' (AWL, p. 38).
Wittgenstein offers six reasons why such a view is mistaken: (i) Aesthetic discussions are about features of the objects perceived, not about our feelings (M, p. 334). (ii) If we were interested in the psychological question as to whether particular features cause certain independently specifiable sensations, it wouldn't be enough to perceive the features and note our sensations (as you can't feel the causation); we would require series of experiments (M, p. 335). (iii) We don't treat works of art as mere causes: means to hedonic ends, but as ends in themselves (M, pp. 339-41). (iv) To the extent to which they can be said to give us pleasure, it is a specific pleasure that cannot be identified independently of our experience of the work in question and is incommensurable with other pleasurable aesthetic experiences. There is no meaningful question of comparing the amount of pleasure derived from entirely different art forms (M, p. 339). One can of course use a locution such as 'it gives me pleasure' in the sense of 'I like (to do) it'. In that sense, any work of art I care to look at or appreciate can be said to 'give me pleasure'. But then to say so is a mere 'tautology' (AWL,p. 38;M,pp. 335,339) and no longer a substantive psychological claim. By the word 'pleasure' in such a locution we don't mean an independently identifiable sensation. As noted above, earlier in the term Wittgenstein had presented his idea of a family-resemblance concept, which he also applied to the word 'good', speaking of ' a transition between similar things called "good"' (AWL, p. 33). However, the discussion of the word 'beautiful' leads him to a different account. A family-resemblance concept applies to different things that do not all share the same set of defining features, and yet that is not a case of polysemy. The word 'game' is applied to football in exactly the same sense in which it can also be applied to chess or hide-and-seek. Indeed, it is exactly the point of a family-resemblance concept that its one meaning has the flexibility to make it applicable to fairly heterogeneous instances. The word 'beautiful', by contrast, is not only applied to different kinds of objects, it is used in different language games, with different meanings (M, p. 338). 12 First, there is semantic changeability due to the word's being used attributively (a point made earlier about the word 'good'): 'The words "beautiful" and "ugly" are bound up with the words they modify, and when applied to a face are not the same as when applied to flowers and trees' (AWL, p. 35). Furthermore, it is not only that the word 'beautiful' functions differently when applied to very different kinds of objects, say, a smell and a mathematical proof; even when applied to the same kind of object it may, depending on the context, have a different role and meaning: 'The phrase "beautiful color", for example, can have a hundred meanings, depending on the occasion on which we use it' (AWL, p. 35). A colour that is beautiful as a feature of a flower arrangement or as the colour of a tie may not be found beautiful on a monochrome wallpaper.

VII
Why, on Wittgenstein's account, do different kinds of objects and different contexts make for different meanings of the word 'beautiful'? Because we apply it for different reasons, according to different criteria, as shown by the way we justify our applications in discussion.
In different cases e.g. beauty of a face, of a flower, you are playing quite different games; & this is shewn by the way in which you can discuss whether the face is beautiful or not.
If you want to know how "beautiful" is used: ask what sort of discussion you could have as to whether a thing is so. (M, p. 333) Here Wittgenstein returns to the key idea with which he began that term's lecture series: verificationism. Possible methods of verification are significant for determining the meaning of a statement, and thus also the meaning of a predicate. To understand it is to understand under what conditions it is correctly applied: the criteria by which its application is to be assessed; so in the case of 'beautiful'. He gives the following examples: To shew ambiguity (& more than this), suppose (1) you are calling a smell beautiful; & can say no more than 'I like the smell of lilac', 'I don't care particularly about it' (2) you are talking about arrangement of flowers in a bed: here you can say much more. This shews that 'beautiful' means something quite different in 2 cases. (M, p. 335) First, there is the most elementary case of an aesthetic judgement: a single sense impression without any discernible structure, a smell, or a colour looked at in isolation (M, p. 337). Where there is no further question of how to combine it with other things, there is little one can say to justify one's judgement. Here, 'This is beautiful' doesn't amount to much more than: 'I like it.' As Wittgenstein later remarked to G. F. Stout, the question of the possibility of verification is illuminating even where it calls forth a negative answer. Far from ruling an unverifiable statement out as nonsensical (as some dogmatic and polemical versions of verificationism suggest), it simply indicates another possible language game, different from that of empirical descriptions. 13 The second example, a flower arrangement, by contrast, has a describable structure and various details, such as colour contrasts or symmetries, to which one can point to say what exactly it is about it that one finds appealing.
In the same lecture Wittgenstein mentions two objects of more advanced aesthetic discussion: musical harmony (M, p. 336) and the design of a door (M, p. 337). Obviously, with works of music or architecture aesthetic discussion can become a lot more sophisticated. Harmony in Western music is a complicated system of rules and conventions that can be invoked to justify an aesthetic judgement, which therefore can be far more thoughtful and richer in content than the praise of a flower arrangement, let alone one's predilection for a certain smell.

VIII
However, at this point, a little inconsistency appears to creep into Wittgenstein's presentation. He promised us different meanings of the word 'beautiful', due to different kinds of discussion of the word's applications, but then, when it comes to more interesting aesthetic discussions, in the realm of music or architecture (and the same could be said for art or literature), he observes that the word 'beautiful' tends not to be used: 'Discussions about the design of a door don't mention "beautiful". They say such things as "It's top-heavy"' (M, p. 337).
Again, a book on harmony is unlikely to say much about 'beauty', rather, it provides rules for what, in a certain system, is deemed correct and what incorrect (M, p. 336). In fact, the word 'beautiful' is scarcely ever used 'in an aesthetic controversy' (M, p. 340; AWL, p. 36). It is typically used to express the kind of aesthetic preference which does not lead to any discussion. Your finding a certain smell beautiful tells me more about you than about the smell. And if you explain to me that you find a certain flower bed particularly beautiful because of the symmetrical arrangement and the contrast of white tulips and purple columbines in it, I may or may not agree; but if I don't -if I can't see much in it myself -I will not on that account begin an argument. In such cases we can give reasons why we like something, by way of explaining what exactly it is about the thing we like, but we couldn't give any reasons to show that different preferences would be wrong. Later Wittgenstein asks: 'Is this flower beautiful or not? What test is there for [the] right answer?' (M, p. 339). The reply to this latter question is clearly: none.
Often, Wittgenstein remarks, the word 'beautiful' serves merely to draw attention to something that appeals to us (M, p. 340; AWL, p. 36), rather than to make any seriously debatable claim.
Serious aesthetics begins where there is room for aesthetic discussion and controversy, and that can only be where we have standards of correctness, such as those introduced by musical theory, and where, therefore, the question of beauty is replaced by the question of correctness (AWL, p. 36). So it was a bit hasty and infelicitous for Wittgenstein to introduce the investigation of the concept of beauty as if it were the key concept of aesthetics. 14 Instead, it would have been better to say that there are different language games of aesthetic appraisal, only some of which, the more primitive ones, involve the word 'beautiful'.

IX
Is beauty a quality? At one point Wittgenstein suggests that it is a confusion to say so, explaining that the attribution of a quality must be a contingent matter. A table has the quality of being brown, as it might have been red instead (M, p. 333). That is to say the bearer of a quality must be identifiable as the same even without it. So being a physical object Wittgenstein would presumably not call a ' quality', for it seems that you cannot make sense of a table still being the same while no longer being a physical object. Or can you? For Macbeth it was a real question whether the dagger he saw was a physical object or a hallucination. Our criteria of identity are very flexible. Even things like daggers or tables can be identified without taking their physicality for granted. And it is certainly possible to identify a beautiful object in such a way that its beauty is not taken for granted. My back garden, for example, might have possessed or lacked beauty depending on my gardening skills and efforts. And it is not plausible to argue that the identity of an object must be determined by the totality of its physical properties, for then even the colour of a table could not be regarded as a contingent quality.
Wittgenstein's concern was whether 'shapes & colours' determined beauty, and he continued to argue that in that case calling them 'beautiful' would not be a contingent matter, hence beauty shouldn't be regarded as a quality (M, p. 333). But the determinate link would only be between certain shapes and colours, and beauty, not between a physical object and beauty. The painting or building or garden might well have lacked beauty, because they might have lacked those specific shapes and colours. So even on Wittgenstein's account, beauty could still be regarded as a (contingent) quality of a physical object, even if it was logically implied by that object's shapes and colours. (Likewise, Jones has the contingent property of being a bachelor, even though it follows logically from his being an unmarried man.) However, as we saw, Wittgenstein denies that an object's shapes and colours determine its beauty (or aesthetic qualities), because aesthetic judgements are context dependent: dependent on what kind of object displays those shapes and colours, but also on what norms and ideals constitute our language game. So, if beauty is not determined by shapes and colours alone, the question arises: 'Is there, when I know what the shapes & colours are like, another investigation as to whether it is beautiful?' (M, p. 333) That sounds a little paradoxical, for what could one possibly investigate in aesthetics if not the perceptible qualities of an object, in the visual case: shapes and colours. But consider the question whether a certain action is legal. Is the answer determined solely by the physical features of the action? Obviously not. So where else do we look? Apart from having to take into account various non-physical circumstances (for example, whose property something was or whether some relevant permission had been granted), we have to consider where the action was performed and investigate what laws were in force there. There is nothing paradoxical in saying that legality is a quality of an action, although it is a relational and conventional quality.
Similarly, there should be no objection to speaking of aesthetic qualities, which are not qualities ' of a distribution of colours in visual space' (M, p. 338). The difference is, however, that the relevant criteria for aesthetic judgements are rarely as straightforwardly codified as laws. There is no book of regulations where I could look up if a given flower arrangement is beautiful or if a certain musical transition is effective. That is why talk of ' another investigation' sounds somewhat odd. But then again, one only needs to remind oneself of our ways of explaining and supporting such judgements to see that there can indeed be ' another investigation', but it would be discursive and clarificatory, rather than empirical (see M, p. 342).

X
What does Wittgenstein say about the standards and criteria that regulate aesthetic language games? On the 15th May 1933 he talks about the ' concept of an ideal' (M, pp. 339-41). An aesthetic ideal is comparable to a law and 'Aesthetic discussion is like discussion in a court of law' (M, p. 351). That is to say, it is not trying to settle the general evaluative question what in a given artistic genre is to count as right or wrong, just as in a court of law there is no discussion as to whether theft is a criminal offence; the question is merely whether in a given case a theft has been committed. Likewise, an aesthetic discussion is not trying to establish norms but to apply them in a given case. Aesthetic norms or criteria are presupposed, like the law in a judicial hearing.
And yet what corresponds to the law in aesthetics; an aesthetic 'ideal' is far more elusive. Typically, it is not laid down anywhere. It does not exist independently of our discussions as a kind of canonical sample that could be produced to settle a dispute. Rather, it is only an abstraction from our practice: 'To find what ideal we're directed to, you must look at what we do: the ideal is the tendency of people who create such a thing' (M, p. 341). An aesthetic ideal has no independent existence and authority, it lies only in a certain consistency of our reactions and preferences, and it changes over time (M, p. 340), especially as there is usually no social enforcement. An example of such an aesthetic ideal is the view Wittgenstein attributes to Bach: that ' a piece mustn't slink away like a thief', meaning that a piece must continue and end in at least the same number of voice-parts (M, p. 352). By contrast, in Bach's aesthetics there appears to be no objection to augmenting the number of voice-parts, as occasionally a piece having three parts at the beginning ends in four.
Naturally, there is no sharp distinction between widely accepted genre conventions and the preferences of only a small group of practitioners and connoisseurs. At the latter end of the spectrum there are individual tastes, for instance: 'you always prefer slightly stronger contrasts, I always prefer slightly weaker [ones]' (ibid.). But then, the taste of an influential individual, an artist or art critic, may appeal to others and become widely accepted.
The concept of an aesthetic ideal or taste is developed further in Wittgenstein's 1938 lectures, where he speaks of a ' cultured taste' (LC,. 15 But already in 1933 he emphasises that such a taste or 'ideal' is presupposed in aesthetic discussion and not itself argued for. Therefore, taking an evaluative stance for granted, aesthetics is descriptive (M, p. 342), rather than a matter of taste (M, p. 346). 'Whenever we get to the point where the question is one of taste, it is no longer aesthetics' (AWL, p. 38; see M, p. 347). Thus, the supreme principle of Wittgenstein's aesthetics turns out to be De gustibus non est disputandum. For a fruitful aesthetic discussion to be possible, a considerable agreement in taste must be presupposed. 16

XI
Another interesting development in the 1933 lectures is this. As we saw, Wittgenstein began the discussion of aesthetics in a fairly conventional way with considering the word 'beautiful' and the traditional question of how to justify calling something beautiful. He soon realised that in more advanced regions of aesthetics the word 'beautiful' ceases to play an important role. Then he observed that interesting aesthetic discussions are largely descriptive, because relevant standards and criteria are presupposed, rather than argued for. Finally, he seems to have realised that providing support for an aesthetic appraisal according to some given criteria was not the only, and perhaps not even the main, focus of aesthetic discussion. More interesting to him became the idea of a puzzle or perplexity in aesthetics.
Typically, our concern is not to demonstrate why something is beautiful or aesthetically successful or correct according to certain standards, but rather to clarify and characterise our distinctive impression of it. It may be a matter of finding the right word to characterise the expressiveness of a melody (M, p. 348) or to hit on a good simile to capture what is striking about a novel (M, p. 356). This idea of an aesthetic puzzle and its explanation -which he thought similar both to philosophical and to mathematical problems (M, p. 358) -was also merely introduced and briefly sketched in 1933 and explored in more detail in the 1938 lectures (LC,. 17