Escribir el título del artículo



Aesthetic Normativity and Knowing How To Go On


Hannah Ginsborg*

University of California, Berkeley, U.S.



This paper addresses a problem about aesthetic normativity raised by Kant. Can aesthetic experiences be appropriate or inappropriate to their objects? And, if so, how is that possible given that, according to Kant, aesthetic experience is not objective? Kant thought the answer to the first question was yes. But his official answer to the second question, in terms of the free play of the faculties, is obscure. The paper offers a clearer answer, inspired by Kant, which invokes Wittgenstein’s notion of “knowing how to go on.” Aesthetic normativity is problematic only on the assumption that claims to the normativity of one’s responses to things must be based on the recognition of objective properties of those things. However, Wittgenstein’s discussion shows that we need not accept that assumption. There can be legitimate claims to the normativity of one’s responses which do not rely on those responses’ reflecting appreciation of objective facts.



aesthetics, normativity, Kant, Wittgenstein




§1 Kant on the normativity of aesthetic response

            This paper addresses a problem about aesthetic normativity, and specifically about whether there are norms governing aesthetic experience or aesthetic response. Is there such a thing as an appropriate or inappropriate aesthetic response to an object, and if so, how is that possible?[1] This is a question which was thematized by Hume in “On the Standard of Taste” and discussed more systematically by Kant in his Critique of Judgment, specifically the “Analytic of the Beautiful” and the “Deduction of Taste.” In the Kantian formulation the paradigm aesthetic experience is the feeling of pleasure in the beautiful, whether in nature or art, and this is the kind of aesthetic response I will have in mind throughout. When I come upon a cherry tree in bloom, or see a vase in the window of an antique shop, or hear music on the radio as I walk into a room, I may have a certain experience of pleasure which is naturally expressed by thinking or saying “That’s beautiful.” That is the paradigm aesthetic experience for Kant, and the question I want to raise, following Kant, is whether we can think of such an experience as one which we not only do have, but in some sense should have, when confronted with the object. Is there such a thing as how I should respond aesthetically to an object, where the “should” reflects something over and above how I am naturally disposed to respond? Can I claim, of someone who is left cold by the cherry tree, that they are mistaken in how they are experiencing the object, that they are failing to see something in it which calls for a response of pleasure? Or is my response to be compared to liking or not liking the taste of goat cheese or “peaty” Scotch whisky, where talk of how one ought to respond seems out of place, and where there doesn’t seem to be room for criticism of someone whose tastes differ from mine?

            Kant himself wanted to defend the idea of aesthetic normativity in this sense. He held that we can indeed say of my response to the cherry tree or the vase that it is appropriate. He makes the point in terms of a sharp distinction between judgments of beauty and judgments of the agreeable. As he puts it, judgments of beauty have what he calls universal validity and necessity, where the necessity he has in mind is clearly normative necessity. When I judge something to be beautiful I claim that everyone —myself included— should judge it as I do. But this is missing in the case of judgments of the agreeable; these more or less correspond to expressions of my own liking for something. When I judge that an object is agreeable, saying for example that Lagavulin whisky is delicious, I don’t judge that everyone should agree with me—I accept that others might not share my taste for peaty Scotches. I’m happy to accept, as a reformulation, “I like the taste of Lagavulin.” This is not the case for judgments of beauty, according to Kant: in their case it seems to us that we are picking out an objective property of the object, one whose presence should be recognized by everyone. In fact, as we shall see, he denies that judgments of beauty are in fact objective: beauty is not an objective property of the cherry tree and in judging that the object is beautiful we are not bringing it under a concept, at least not in the ordinary way. But judgments of beauty carry the same claim to their own appropriateness as objective judgments, and for that reason it can seem as though they involve the same normativity as objective empirical judgments. 

            Now it might seem here that in talking about Kant on aesthetic judgments I have changed the subject. I began by raising the question of the normativity of aesthetic response or aesthetic experience: isn’t there a difference between my pleasurable response to the cherry tree and my judgment that it is beautiful? On my reading of Kant, the answer to this last question is “actually no; for Kant the pleasure just is the judgment.” Saying of the cherry tree “That’s beautiful” just is a way of putting into words the pleasure one feels, so the pleasure is just the taking, or finding, of the tree to be beautiful. This is a controversial reading of Kant.[2] If it is accepted, then saying that the judgment of beauty involves a claim to its own appropriateness to the object amounts to saying that the pleasure involves a claim to its own appropriateness: it is part of my feeling of pleasure that I feel the rightness of that pleasure as a response to the object. There are reasons, which I won’t go into here, to ascribe that view to Kant; and it is also an appealing view in its own right. But you do not need to accept my controversial view on the identity of the pleasure and the judgment to find in Kant a defence of the normativity of aesthetic response as well as judgment. For he makes clear that, in judging something to be beautiful, I claim not only that everyone should agree with my judgment, but that everyone should share the pleasure I feel. So even if we do make a distinction between feeling pleasure in an object’s beauty, and making some kind of considered judgment about the object on the basis of that pleasure, we still have to allow that the pleasure itself, for Kant, is normatively governed. And this, I think, has some plausibility in its own right. If somebody I respect disagrees with me aesthetically—for example if they dismiss Bill Evans as mere lounge music, or Andrea del Sarto’s religious paintings as listless and vapid—it is my aesthetic response that I will feel defensive about, not any explicit judgment I might have made on the basis of that response. A criticism of my taste criticizes my liking what (in the view of the critic) doesn’t deserve to be liked, or my failure to like what I should like. My own inclination is to see this as grounds for identifying aesthetic response and aesthetic judgment: to be a good judge of aesthetic value in some domain to have the appropriate aesthetic responses to objects in that domain.[3]  But we do not need to go down that route to see that Kant’s view at least includes the idea that we take our aesthetic responses to be subject to normative constraint, and that that idea has some initial plausibility.


§2 The problem of aesthetic normativity and Kant’s answer to it

            Why is there a problem about the possibility of this kind of aesthetic normativity? Why shouldn’t it be appropriate for me to experience pleasure in the flowering cherry tree just as it is appropriate for me to see the blossoms as white rather than yellow, or to see the cherry tree as having a tree-like shape rather than the shape of a house? In other words, why shouldn’t we treat the perception of an object’s beauty just as we do the perception of any other property of the object, that is, as subject to norms of veridicality? On that way of treating aesthetic experience, my feeling pleasure in something which arguably doesn’t deserve it, or failing to feel pleasure in something which does, is simply a matter of mistaken perception: it is a matter of my seeming to discern an objective property of the object which is not in fact there, or failing to discern one which is. (This is the approach which is suggested by Hume’s parable, in “On the Standard of Taste,” of the key with the leathern thong.) On that approach there is no special problem about aesthetic normativity, any more than there is a problem about how ordinary perceptions can be veridical, or fail to be veridical, with respect to the object perceived. Kant’s reasons for thinking that there is a problem are bound up with his denial that judgments of beauty can be assimilated to cognitive judgments. In his terminology, although the judgment of beauty makes a claim to universal agreement, it is not objective. Relatedly, when we feel pleasure in a beautiful object, or find it beautiful, we are not perceiving a feature of it which is there independent of the particular experience we are having of it. This has to do with the difference between the feeling of pleasure or displeasure, or “subjective sensation,” and the objective sensation associated with, say, the perception of colour (CJ §3, 5:206). In the case of the feeling of pleasure our experience is never of the object as such, but always of what he calls its relation to the subject: how the object is affecting us in the experience we are having of it. This is true both of pleasure in the agreeable and of pleasure in the beautiful.[4] 

            The subjective status of the feeling of pleasure is reflected in a number of other features of judgments of beauty to which Kant draws attention. One is that judgments of beauty cannot be proved; there are no criteria for determining whether or not something is beautiful. Another is that I cannot judge that something is beautiful on the basis of someone else’s feeling of pleasure. I cannot, for example, learn by hearsay that something is beautiful; I have to see the thing myself in order to be able to make a judgment of beauty about it. Moreover, if I do find something beautiful, and someone else looking at the same object fails to share my pleasure in it, that is not a reason for me to change my judgment. “The approval of others in no way provides. . . a valid proof for judging beauty . . . what has pleased others cannot serve as the ground of a judgement of beauty’ (CJ, §33, 5:284). The point that it is only my own feeling that is relevant to judging beauty, and hence that I must myself perceive the object in order to judge it to be beautiful, is brought out by Kant through a comparison with judgments of the agreeable. “Someone may recount to me all the ingredients of a dish, and remark of each of them that it is agreeable to me... Yet I am deaf to all these reasons; I try the dish with my own tongue and palate, and lay down my judgement accordingly.” The same general point, according to Kant, applies to judgments of beauty. “Although critics, as Hume says, may reason more plausibly than cooks, they still share the same fate” (CJ, §34, 5:285). It is my feeling of pleasure alone which is decisive in determining whether I can or cannot judge something to be beautiful. This distinguishes judgments of beauty from what might seem to be their nearest cognitive analogues, nonevaluative judgments of sensory properties like colour and taste. For in these cases I can rely on the reports of others to determine whether or not such judgments are legitimate.

            This lack of objectivity, according to Kant, is what makes aesthetic normativity, in the sense I have been discussing, problematic. In the kind of paradigm experience I have described, of finding a flowering cherry or a passage of music beautiful, I take my own pleasurable response to the object to be appropriate to or called-for by the object. And as long as I do not think that the response is due to some personal interest that distinguishes me from other human beings—for example happy memories of flowering cherries from a childhood visit to Paris—I take it to be called for from all human beings. So I make a normative claim on the agreement of others, that is, a normative claim that my aesthetic response should be shared. And I do so, Kant says, in just the way that I demand agreement from others regarding the perceptual states relevant to cognition. He compares the claim to that made by someone who perceives a movable drop of water in a rock-crystal and “rightly demands that everyone else must find it just as he does.” (CJ Introduction VII, 5:191). The demand, he says, is the “very same” in the case of a judgment of beauty. Presumably in the case of a judgment of beauty I am also demanding that everyone have the same experience as I do, in respect of my finding it beautiful, or feeling the corresponding pleasure in it. One might here wonder why Kant uses the somewhat arcane example; I believe that it has something to do with debates going on at the time about whether rock-crystal, that is quartz, was formed from hardened ice. Presumably if someone sees a drop of water in a rock-crystal that might be evidence for the ice hypothesis. What is relevant here is that the demand bears not on a theoretical judgment but on the experience of seeing the rock-crystal: I rightly demand that everyone should see it as I do, where the seeing it as I do is described as seeing a movable drop of water in it. My entitlement to do that is a condition of our going on to have a theoretical debate about the origin of rock-crystals—it is needed if we are to have common data on which to base our subsequent hypotheses.

            But there is also an obvious disanalogy. In the rock-crystal case I can cite a ground for my claim in terms of a property of the object. There is a fact about whether or not there is a movable drop of water inside the crystal, and I can appeal to that fact as offering an objective, albeit uninformative justification for the claim. (Spelled out, the justification would be something like: Everyone ought to see it the way I do, because I see it as having a drop of water in it and it does have a drop of water in it.) There are also facts which might be seen as warranting or providing evidence for the claim, and so providing a kind of subjective justification: for example that it looks as though there is a movable drop of water, or that it looks as though there is a clear liquid in it, and an arbitrary clear liquid in these circumstances is most likely to be water. And there are means of finding out whether I have been experiencing the situation correctly. If there is doubt about whether the liquid is water, we can crack open the crystal and find out. Or if my perceptual capacities are in question, I can have other people, with better vision, take a look. But in the aesthetic case, there are no corresponding sources of justification. Since beauty is not an objective property of things I cannot claim that my finding the flowering cherry beautiful is appropriate to the flowering cherry just because it is, in fact, beautiful. Relatedly, there are no tests or criteria I can employ to justify my claim. And, as Kant is at pains to emphasize, I cannot appeal to how other people experience the object. Whether or not I am entitled to claim that something is beautiful, and so that everyone else ought to share my pleasure in it, is independent of whether anyone else feels pleasure in it: the only basis for my judgment can be my own pleasure. So how can I possibly be entitled to demand universal agreement for that pleasure? As Kant puts it in a canonical statement of the problem: “how is a judgement possible which, merely from the subject’s own feeling of pleasure in the object, independent of the concept of it, judges this pleasure to attach to the representation of the same object in every other subject, and does so a priori, that is without being allowed to await the agreement of others?” (CJ, §36, 5:288).



§3 Kant’s answer to the problem

            Kant offers an apparent answer to this question in terms of his notion of the free play of the faculties of imagination and understanding. He claims that pleasure in the beautiful consists in a state in which the imagination apprehends the object in a way which accords with the requirements of understanding, but does so freely, that is, without being governed by concepts of the understanding. This state of free play, he says, is one for which we can claim universal agreement, he says, because it manifests what he calls the “subjective formal condition of a judgment in general” (CJ §35, 5:287). The possibility of cognition requires that we be able to claim universal agreement for our cognitive states, for example experiential states like that of seeing the movable drop of water in the rock-crystal. This yields an entitlement to claim universal agreement which carries over from our cognitive states to the state of the free play. Because of the free play’s status as a subjective condition of cognition, we are entitled to claim agreement not just for particular cognitions, but for the free play, and hence for the feeling of pleasure in which it is manifested. This answer, viewed at the most general level, attempts to justify the normativity of aesthetic judgment by making a connection between aesthetic and cognitive experience. Although aesthetic experience is not itself cognitive, its relation to cognition is such that we cannot deny the claim of aesthetic experience to universal agreement without also undermining the possibility of cognition. But it is notoriously difficult to make sense of Kant’s argument here. Those who attempt a charitable interpretation are typically faced with a dilemma regarding the relation of the free play to ordinary cognition. If the free play is a necessary condition of cognition then it would appear that every object we cognize should be experienced as beautiful. But if it is not a necessary condition of cognition then it is not clear why the claim of cognitive experience to universal validity should carry over to aesthetic experience. And even aside from this specific obscurity in the argument, we also have to contend with the more general obscurity of the transcendental-psychological language in which it is framed. So if we take seriously the question Kant raises, and want to address it, we either need to find a way of clarifying Kant’s answer and making it persuasive, or adopt another approach.

            Philosophers who have taken the problem seriously but rejected Kant’s answer to it have tended to address it either by rejecting aesthetic normativity or by insisting on the objectivity of aesthetic experience. The latter can seem like the more tempting approach, especially given that, in many ways of using the term “objectivity” it is interchangeable with Kant’s notion of universal validity. It can seem like a contradiction in terms to assert that I am entitled to claim universal agreement for my judgment that something is beautiful, and yet to deny that the judgment is objective. Or it can seem as though Kant must be operating with an unduly demanding notion of objectivity, one which demands, say, that judgments must be provable to count as objective, or that it must be possible to formulate them in terms which make no reference to specifically human responses. But I think that this approach misses an important commonality between the experience of something as beautiful, or positive aesthetic experience more generally, and the experience of just liking something, as in Kant’s judgments of the agreeable. We do not want to lose sight of the fact that ordinary experiences of beauty are in the first instance experiences of pleasure, liking or enjoyment. They play a different role in our psychological economy, so to speak, from experiences which serve to register the features of objects, to tell us what kind of things they are. If I find the flowering cherry beautiful and you do not, it seems strained to describe the difference in our experiences by saying that I notice a feature of the cherry which you fail to notice. Assuming that the two of us have an equally clear view of it and that we come to it with the same conceptual background, we see it as having all the same features. The difference lies in how we see it, our way of seeing it, where “how we see it” does not translate into “how we see it as being.” Unlike you, I see it in a way which is pleasurable; but this does not mean that I see it as being a certain way, that I see it as having a property which you fail to see it as having.[5] In not seeing it the way I do you are not cognitively worse off. What you are missing is not some item of information about the cherry tree, but rather the pleasurable way of seeing it. If we insist in this way on the subjectivity of taste, dissociating aesthetic experience from cognition and insisting on the connection with pleasure, then the question of aesthetic normativity remains pressing. 

            In the remainder of this paper I want to propose an answer which is at the very least inspired by Kant’s own answer, and arguably an interpretation of it. I have in fact argued previously that Kant’s account of aesthetic normativity should be understood roughly along the lines of the answer I’m now about to sketch,[6] but in this paper I will not try to defend that point, but will simply present the answer in its own right. Putting it very roughly: it is a misapprehension to think that there is a conflict between the normative character of aesthetic judgments and the fact that they are not objective. The thought that there is a tension rests on the mistaken view that normativity in one’s judging or more generally in one’s cognitive or perceptual response to an object, depends on the judgment’s being objective or on one’s response to the object amounting to appreciation of an objective feature of it. We can avoid this mistaken view by acknowledging a kind of normativity applicable to our psychological responses to things that does not depend on those responses registering objective facts but that is, rather, a condition of objective cognition. Once we recognize the possibility of taking one’s response to an object to be normative—to be appropriate to the object—in a way that doesn’t stand in need of justification in terms of the objective features of the object, then there is no longer any mystery about aesthetic normativity. The fact that we can take our psychological response to a beautiful object to be one that everyone should share is simply an instance of a more general entitlement to take our natural psychological responses to things to be appropriate to those things. This answer explicitly follows Kant in accounting for the normativity of aesthetic response in terms of the conditions of cognition, and more specifically in terms of the normativity of the perceptual responses which enter into cognition. But it aims to clarify the kind of normativity at issue in a way which makes clear that it is not incompatible with the denial of objectivity. 


§4 Knowing how to go on and primitive normativity

            I want to explain the notion of normativity I have in mind by looking at an idea which appears frequently in Wittgenstein’s later philosophy, the idea of “knowing how to go on.”[7] This idea appears in the context of Wittgenstein’s rule-following considerations. Here Wittgenstein is concerned with the general question, what is it to grasp a rule, where it is clear that his concern is also with grasping the meaning of an expression or grasping a concept. Part of his strategy in approaching this question is to consider how children come to master linguistic expressions or acquire concepts. His answer is that they do so by a kind of training in which the teacher presents them with examples of behaviour which they are encouraged to follow. A normal child presented with a sufficient number and range of examples will come to “know how to go on” from them, and this knowledge makes possible the grasp of a corresponding rule or concept. In the Philosophical Investigations, Wittgenstein illustrates this knowing how to go on with the example of mastering basic arithmetical concepts. Learning the decimal system, for example, is a matter of multiple stages of learning how to go on from examples of behaviour modelled by the teacher. A child must first learn how to write the individual numerals 0 through 9 by first perhaps having her hand guided by the teacher, or tracing the shape that the teacher has given her, before she can go on to write each number on her own. She must then learn to get the numerals in the right order, again going on from examples she has been given. Following that she must learn to continue the pattern of numerals exemplified by the series of natural numbers in the teens, twenties, thirties and so on, and then into the hundreds. At some point we can say she knows how to go on: she has, we might say, got the hang of, or come to understand the principle of decimal notation, even if she is not able to give a good explanation of it. This teaching might employ verbal explanation, but Wittgenstein makes clear that it does not have to. Describing it in the Investigations he talks about “[using] particular emphases, underlin[ing] figures, writ[ing] them one under another in such-and-such ways, and similar things” (Investigations §145). And in Zettel he says that the learning may happen “only through examples,” without any rule being given (Zettel §295). Once a child has learned the principle of decimal notation, we can teach her the meaning of expressions for simple arithmetical operations. Wittgenstein gives the example of an expression he calls ‘+2’ which corresponds to developing the series of even numbers (Investigations §185). We can teach a child the meaning of this expression, and, correspondingly, the concept of adding two to an even number, by giving her the example of the series of 0, 2, 4, 6, 8... and so on, and showing her how to continue it. If she is able to continue the series for high enough numbers, we can say of her that she knows how to go on, and this, for Wittgenstein, is equivalent, or at least comes close, to her having come to grasp the rule or concept underlying it.

            Against this background, Wittgenstein introduces his famous example of the aberrant pupil who has apparently learned to develop the sequence of even numbers, and who does so competently up to 1000, but then goes on with 1004, 1008, 1012 instead of, as expected, 1002, 1004, 1006. When the pupil is challenged, he says: “Yes, isn’t that right? I thought that was what I was supposed to do” (Investigations §185). We try to explain to him that this is not the right way to go on, but he misunderstands our explanations in just the same way in which he seems to have misunderstood what we were trying to teach him.  Wittgenstein suggests that it is as though it comes naturally to him to understand the command “+2” as we would understand the order “Add 2 up to 1000, 4 up to 2000, 6 up to 3000” and so on. And he compares the case to one in which a person responds to a pointing gesture by looking in the direction of fingertip to wrist rather than the other way around (ibid.). The aberrant pupil, we’re inclined to say, has got the wrong end of the stick: he has not caught on to the rule we intended him to catch on to. But, as in the case of someone who misunderstands the pointing gesture, we do not seem to have any way of correcting him. In the case of the pointing gesture there is literally nothing we can do to show the person which way to look: anything we provide (more pointing, arrow signs, even physically turning him around) will be understood in the opposite way from the way we intended. Something similar is the case for the aberrant pupil.  Imagine how we might try to show him that he is going on wrong. We might have him write the numbers vertically in columns and compare the unit column for the numbers he’s given with the corresponding columns for all the other numbers. Look, we might say, this last column is different! And here we have to imagine that he insists that it’s the same: that the column 0, 4, 8, 2, in the context where those numbers appear in the units column of the four digit numbers, is the same as the column 0, 2, 4, 6 where the numbers appear in the units column of the three digit numbers. Perhaps then we have him compare just the number 4 which he has written in 1004 and compare it with the 2 he wrote in 992, and he again insists that they’re the same. Let’s suppose now that we just cut out the two numbers -- the 4 he wrote in 1004 and the 2 he wrote in 992 and put them side by side.  “Now they’re the same,” he says; but when we put them back in their places he says, “Now they’re different.”[8] It is as though something went wrong at the very beginning, when we were teaching him just to write the numerals. We thought he had learned how to go on in the sequence of writing the same numeral again and again, that he had got the hang of what the numeral 2 is. But actually he hadn’t: he had apparently picked up an alternate way of going on in which the correct numeral to write after 2 if you are, say, copying numbers from the units column of a column of three digit numbers to the units column of a column of four digit numbers, is 4. 

            The example of the aberrant pupil is relevant for our purposes because it brings into relief a certain kind of normativity associated with knowing how to go on, and it is in terms of this kind of normativity that I want to address the question of how aesthetic normativity is possible. Notice that, when we challenge the aberrant pupil, the first thing he says is “Isn’t that right?” He not only says 1004, he thinks that it’s the right thing to say in this context, that it’s the appropriate response to being shown the series of numbers he’s already written and being asked to “go on.”[9] This isn’t a peculiarity of the aberrant pupil; a normal child who writes 1002 also probably thinks that what she is doing is correct, although she is not likely to express this thought unless she is challenged. And we can assume that this is true also in the more basic cases of going on, like copying individual numerals or writing them in an order which matches the teacher’s. The child is not simply producing the numerals mechanically like an automaton, but is rather conscious of herself as attempting to meet a normative demand set by the teacher’s examples. The case of the aberrant pupil gets us to focus on this normativity by having us imagine a context in which it is called into question. Consider again the aberrant pupil’s claim that going on with 1004 is correct. I think the natural reaction to this claim is to reject it: writing 1004 is going on wrong from the previous sequence and it is 1002 which is right. Yet there is, as we have seen, no way to persuade the pupil of this. And this isn’t just because he is stubborn or stupid. We could imagine him, indeed, as highly intelligent and capable of sophisticated rational argument, but still unconvinced. The point is that there is in fact no rational argument to be given for the correctness of his writing 1002 as opposed to his writing 1004. If we could assume that he had been previously been applying the add-two rule, and that in asking him to go on from the sequence we were asking him to apply the same rule he had applied previously, then we could argue that 1002 is correct by appealing to the obvious objective fact that the add-two rule as applied to 1000 yields 1002. But his previous behaviour is equally compatible with his having applied any number of rules, including the rule “add 2 up to 1000 and then add 4.” Once we drop the assumption that the pupil’s previous behaviour manifested his application of the add-two rule, then we have no objective ground for rejecting his claim that writing 1004 is just as good a way of going on as writing 1002. 

            One way in which Wittgenstein’s example has been understood is as calling into question the very possibility of rule-following and meaning. The impossibility of giving reasons for the correctness of 1002, or of citing an objective fact in virtue of which writing 1002 is correct, has implications not just for the pupil but for ourselves; it implies that we have no grounds for our own conviction, under normal circumstances, that writing 1002 is the right way to go on from a series which begins with 0,2,4,6,8... and goes up to 1000. We might think that at least in our own case, if not in the pupil’s, we know what rule we are applying when we write such a series, but Wittgenstein gives convincing reasons for supposing that we are no better off epistemically in our own case than in that of the pupil. I know, for instance, that I have been following the rule that I have been taught by being given examples like those that the pupil has been given, or with words like “count by twos” or “write out the even numbers.” But knowing that I learned the rule from being shown the examples does not tell me any more about what I should write than I know about what the pupil should write. And regarding the verbal instructions I was given, I learned the meaning of the words in them by learning how to go on from a finite set of examples I was shown, so the question of what I should write in the light of those instructions remains just as open as the question of how I should go on from the sequence taken in isolation. The upshot, according to this way of understanding Wittgenstein, is that “knowing how to go on” is a misnomer. We may all be disposed to continue the sequence in the same way and to do so with a feeling of conviction that we are going on appropriately, but there is no genuine recognition of how to go on. And if, as Wittgenstein suggests, knowing how to go on is the foundation of our grasp of rules, concepts and meaning, then that yields the skeptical conclusion that there are no such things, and correspondingly no such thing as truth-evaluable judgment.[10]

            However, as I see it, the moral we should draw from the example is quite different. The fact that there is no rational argument to be given for the rightness of 1002 as opposed to 1004 should not be taken as undermining our conviction that writing 1002 is the right way to go on and that writing 1004 is not. Rather it should be taken as showing that our claim to the correctness of 1002 can stand without rational justification. More generally, in the kind of learning situation which Wittgenstein describes, the sense of correctness in which 1002 and not 1004 is the correct thing to write, is not a matter of objective truth. The appropriateness of 1002 at that point in the sequence neither amounts to, nor depends on, the truth of the arithmetical judgment that 1000+2=1002. It is, I want to say, more primitive, both in the conceptual sense that we need it to make sense of their being such things as arithmetical judgments, and in the ontogenetic sense that our capacity to grasp it precedes our capacity to recognize arithmetical truth. 

            I understand Wittgenstein’s examples of learning how to go on as intended at least in part to draw our attention to this primitive appropriateness. We cannot explain the correctness of writing 1002 after 1000 in terms of the accordance of that behaviour with a previously grasped rule. But that does not mean to say that we cannot make sense of 1002 being the correct thing to write in those contexts. The same goes for the more basic cases of knowing how to go on illustrated by a child learning to write individual numerals following the teacher’s example. If a teacher writes the number 2 several times and tells the pupil to go on, the correct thing for the pupil to do is to keep writing the number 2; she should not, for example, start writing 4 after 500 occurrences, or if the place she is asked to write it is in the units place of a four-digit number. Moreover, although Wittgenstein illustrates this kind of correctness in a context of learning arithmetical expressions and concepts, it is applicable also to the acquisition of everyday concepts like dog and chair. A child comes to grasp the concept dog, in a way which allows her to make judgments like “That’s a big dog” or “Grandma’s dog is in the park,” through mastery of sorting activities in which she learns to respond in certain specific ways to dogs in contrast to other things in her environment. Such activities might form part of everyday interactions with actual dogs: the child learns to treat unfamiliar dogs with caution, to pat them if allowed to do so, to say “hello doggie,” and so on. Her learning how to behave around dogs can be seen as a kind of learning how to go on: her parents have encouraged her to say “hello doggie” to various dogs she has encountered and at a certain point she finds it natural to do so spontaneously with new dogs which cross her path. She can also learn the concept through more structured activities, for example activities of sorting different kinds of toy animals into different bins. The child who is able to add more dogs to the bin where she has seen the teacher putting them, and to leave on side the cats and the horses, is “going on” in a way analogous to that of the child continuing the add-two series.

            As in Wittgenstein’s examples we can speak of knowledge of how to go on, and not a mere disposition to go on, because the child recognizes, as we do, that what she is doing fits the context. We might imagine her, like the aberrant pupil, protesting if her behaviour is challenged. If we try to put the spotted dog with the cats rather than with the other dogs she might rightly object that it “doesn’t belong there.”[11] So there is a recognition of her behaviour as normatively governed, both by her and by us: we all recognize that the spotted dog should be sorted with the dogs and not the cats. But, again as in Wittgenstein’s cases, the “should” here is not a matter of the truth of an objective judgment. We cannot justify the claim about where the spotted dog belongs by appealing to the fact that it is a dog, or that it has other objective features in common with the dogs in the bin. We can see this by noting that the child must be able to recognize the “should” before having acquired the concept dog or other concepts of features which dogs have in common; the sorting activity is one which makes possible, rather than depending on, possession of the concept dog. Another way is to note that, as the rule-following considerations suggest, there are any number of objective features to which we could appeal to justify sorting the spotted dog with the cats rather than the dogs already in the bin. For example the spotted dog shares with the cats the feature of being either a spotted dog or a cat; and assuming that none of the other dogs in the bin have spots, this is a feature which it does not share with the dogs that have previously been sorted together. The upshot is that our grasp of ordinary empirical truths about dogs and chairs, like our grasp of arithmetical truth, depends on the recognition of a kind of normativity more primitive than that associated with objective judgment. There is such a thing as sorting an object correctly or incorrectly with others, where the correctness does not depend for its ascription on the idea that the object possesses some general feature in virtue of which it should be sorted that way. 


§5 Conclusion

            Let us return now to the problem of aesthetic normativity. As we saw, that problem as Kant construes it arises because of a seeming conflict between the normative character of aesthetic judgments and the fact that they are not objective. I proposed to address the problem by showing that it is based on the misapprehension that claims to the normativity of one’s responses to things must be based on the recognition of objective properties of those things. There can be legitimate claims to the normativity of one’s responses which do not rely on those responses’ reflecting appreciation of objective facts. My discussion of the normativity involved in knowing how to go on was intended to show how this can be. In order to avoid skepticism about rules, meaning, and content we have to allow a kind of normativity in our ways of going on—imitation of others’ behaviour, continuing patterns, sorting objects based on examples—whose recognition does not depend on the appreciation of objective truth. If this kind of normativity is accepted, then the supposed problem of aesthetic normativity disappears. 

            It is important to note that this solution does not depend on simply identifying aesthetic responses to objects with the kinds of perceptual responses to things which make it possible to know how to go on. I am not claiming, for example, that the child’s knowing to sort the spotted dog with the other dogs rather than the cats is due to some kind of aesthetic sensitivity. Although we might think of her behavioural response to the dog as reflecting her way of perceiving it—a way that allows her to see it as more similar to the other dogs than to the cats—this is a quite different way of perceiving than the pleasurable way in which we perceive the flowering cherry. Conversely, when one feels pleasure in a thing, one is precisely not having the kind of response which allows one to classify it as belonging in some general kind rather than another. That distinction is central to Kant’s distinction between objective and subjective sensation as well as to his point that judgments of beauty are not conceptual. The point is rather that we can appeal to the normativity involved in knowing how to go on in order to show that there need not be a special problem about how we can take our aesthetic responses to be appropriate to the things we perceive. Rather than suppose that any claim to appropriateness or universal agreement requires a justification for its legitimacy, we can, so to speak, shift the burden of proof. If someone wants to deny that our aesthetic responses can be appropriate to their objects, then they need to come up with a reason. They need to argue, for example, that aesthetic pleasure is invariably “interested” (in Kant’s sense of the term), that is, that it is always rooted in the individual’s particular desires or preferences. That would require showing that my pleasure in the flowering cherry is not, as it seems to be, something which is obviously called for by the object and so something which any human being ought to recognize, but rather that it reflects idiosyncratic preferences due to my particular history and education. I have not been claiming that aesthetic normativity cannot be challenged on those or similar grounds. What I have been claiming is that the mere absence of what Kant calls objectivity in our aesthetic judgments is not sufficient to motivate such a challenge. The impossibility of providing rational justifications for our aesthetic judgments is not in itself a reason for questioning our intuitive commitment to the normativity of our aesthetic responses.[12]


Ginsborg, H. (forthcoming), “Conceptualism and the Notion of a Concept”, Forthcoming in Christoph Demmerling and Dirk Schröder (eds.), Concepts in Thought, Action and Emotion: New Essays, Routledge.

Ginsborg, H. (2020), “Wittgenstein on going on”, Canadian Journal of Philosophy, 50, pp. 1-17. 

Ginsborg, H. (2017), “In Defence of the One-Act View: Reply to Guyer”, The British Journal of Aesthetics, 57 (4), pp. 421-35.

Ginsborg, H. (2015), The Normativity of Nature.  Oxford: Oxford University Press.

Guyer, P. (2017), “One Act or Two? Hannah Ginsborg on Aesthetic Judgement”, The British Journal of Aesthetics, 57 (4), pp. 407-19.

Kripke, S. (1982), Wittgenstein on Rules and Private Language.  Cambridge, MA: Harvard University Press.

Langer, J. (2001), “The mosaic evolution of cognitive and linguistic ontogeny”, in Melissa Bowerman and Stephen C. Levinson (eds.), Language acquisition and conceptual development, Cambridge: Cambridge University Press.

Langer, J. (1986), The origins of logic: One to two years. Orlando, FL.: Academic Press.

McDowell, J. (1998), “Wittgenstein on Following a Rule”, in Mind, Value and Reality, Cambridge, MA: Harvard University Press, pp. 221-62.

Sugarman, S. (1983), Children's Early Thought: Developments in classification. Cambridge: Cambridge University Press.

Wittgenstein, L.  (2009), Philosophical Investigations.  Translated by G. E. M. Anscombe, P. M. S. Hacker and J. Schulte.  Revised fourth edition by P. M. S. Hacker and Joachim Schulte. Oxford: Blackwell.

Wittgenstein, L. (1984), Remarks on the Foundations of Mathematics. Edited by G.H. von Wright, R. Rhees, and G.E.M. Anscombe. Cambridge, MA: MIT Press.

Wittgenstein, L. (1967), Zettel.  Edited by G.E.M. Anscombe and G.H. von Wright.  Translated by G.E.M. Anscombe.  Oxford: Blackwell.

Wright, C. (1980), Wittgenstein on the Foundations of Mathematics. London: Duckworth.Descripción: kant_ctk_TRANSPARENTE1.png

* Academic affiliation: University of California, Berkeley, U.S. Email :

[1] Note that I am not here concerned with whether aesthetic experience is subject to, say, moral norms, for example, whether it is morally inappropriate to feel pleasure in art with racist or sexist content, or which one knows to have been produced with morally objectionable intentions.  Rather, my concern is about how there can be normativity intrinsic to aesthetic experience.

[2] See for example my recent exchange with Paul Guyer, framed in terms of an opposition between “one-act” and “two-act” approaches to Kant’s account of aesthetic judgment, in Guyer (2018) and Ginsborg (2018).

[3] This is not to deny that someone could be good at discerning what is likely to cause aesthetic pleasure in most people, even if they don’t feel the pleasure themselves. (We could imagine an antique dealer who dislikes eighteenth-century British china but is skilled at predicting what will please collectors.) It is just to say that such a person wouldn’t be skilled at aesthetic judgment (wouldn’t “have good taste”) but would, rather, be skilled at predicting the aesthetic judgments of others.

[4] Here we should keep in mind that to speak of the “object itself” in this context is not to speak of the thing in itself, in the sense of the first Critique, but rather the spatio-temporal object, that is, what in the first Critique Kant calls an appearance. The objects and subjects which figure in the third Critique are ordinary spatio-temporal objects and individual embodied human beings; for the purposes of the present discussion we can disregard Kant’s transcendental idealism and simply take for granted the empirical realism which he also takes himself to have established in the first Critique.

[5] For this distinction, see Ginsborg (2006, p. 358), and Ginsborg (2015, pp. 181-182).

[6] See Essays 5 and 8 of Ginsborg (2015).  The answer presented here differs from the answer presented earlier in ascribing a more central role to Wittgenstein. 

[7] I develop the idea in its own right in Ginsborg (2020).

[8] Compare the woodsellers at Remarks on the Foundations of Mathematics,  I §149. “How could I show them that —as I should say—you don't really buy more wood if you buy a pile covering a bigger area?—I should, for instance, take a pile which was small by their ideas and, by laying the logs around, change it into a 'big' one. This might convince them—but perhaps they would say: "Yes, now it's a lot of wood and costs more"—and that would be the end of the matter.”

[9] Here I diverge from my earlier reading of Wittgenstein in Essay 5 of my Ginsborg (2015), where I claimed that examples like that of the aberrant pupil leave out, or at least fail to highlight, the normativity involved in knowing how to go on (Ginsborg 2015, p. 127).

[10] Saul Kripke reads Wittgenstein in roughly this way (see his 1982, ch. 2). So does Crispin Wright; see e.g. chs. 2, 11 and 12 of Wright (1980). John McDowell comments as follows on the skeptical, or at least counterintutive, consequences of Wittgenstein’s view as interpreted by Wright: “If Wittgenstein's conclusion, as Wright interprets it, is allowed to stand, the most striking casualty is a familiar intuitive notion of objectivity. The idea at risk is the idea of things being thus and so anyway, whether or not we choose to investigate the matter in question, and whatever the outcome of any such investigation” (McDowell 1998, p. 222).

[11] Jonas Langer describes a 30-month old child in a sorting experiment “rebuking” the tester with “No belongs this way” when she was given a mismatched set of coloured blocks (Langer 2001, p. 22). Langer’s descriptions  of younger children’s nonverbal “self-correcting” activity, starting around 15 months (see e.g. Langer 1986, p. 57), as well as descriptions of self-correcting behaviour in 18-month-olds in Sugarman (1983, pp. 90-91), suggest that these attitudes predate the capacity to express them verbally. I discuss the relevant normative attitudes further in Ginsborg (forthcoming).

[12] This is a lightly reworked version of a talk presented at a workshop on aesthetic normativity at Harvard University in May 2019. I am grateful to the audience on that occasion for discussion, and especially to Francey Russell for her excellent comments.  

Enlaces refback

  • No hay ningún enlace refback.
Utilizamos cookies propias y de terceros para realizar análisis de uso y de medición de nuestra web para mejorar nuestros servicios. Si continua navegando, consideramos que acepta su uso.

Creative Commons by-nc 3.0 Logo

ISSN: 2386-7655


  Scimago Journal & Country Rankscopus logo


DOAJ LogoErih Plus LogoCitefactor logoredib Logo

LatIndex LogoISOC Logo MIAR Logo
MLA LogoZenodo Logo
ESCI LogoEBSCO LOGOWeb of Science