No Givenness Please, We’re Sellarsians

Dan Sacilotto over at Being’s Poem has just put up an excellent post discussing some issues that myself and Ray Brassier have been working on, in the light of a comparison between the two titans of Hegelianism in contemporary philosophical world: Badiou (the paragon of mathematical ontology) and Brandom (the paragon of inferentialist semantics). As Dan was so generous in the complements with which he opened his post, I feel I should say a little something in return. The pleasure in our correspondence has been entirely mutual. Dan is an incredibly enthusiastic and sincere interlocutor, and he’s consistently challenged me to improve both the content of my ideas and their form of expression. He’s also patiently and valiantly attempted to explain Badiou to me, and has been very helpful, in spite of my persistent inability to grasp what Badiou means by ‘presentation’. Dan exemplifies a lot of the virtues of a good philosopher: he’s intensely autodidactic, philosophically omnivorous, he doesn’t pull his discursive punches, and he refuses to write about things unless he thinks he understands them. All in all, a top chap.

Now that that’s out of the way, I’d like to address a few of the aspects of Dan’s post. I’m not going to cover everything, as it’s filled to the brim with interesting content. However, I do think that I can present my own point of view on several issues in a bit more detail, and provide some additional context for those who aren’t aware of the way mine and Ray’s Sellarsian projects have been developing of late. To this end, I’m going to carry on my recent practice of quoting from my own correspondence, and post a few snippets from my correspondence with Ray.

However, before I get down to this it’s useful to quickly summarise the central point of Dan’s post. His basic idea is that, although their rejection of the primacy of phenomenal givenness is highly laudable, both Badiou and Brandom end up going too far in minimising the role of experience, especially in their rejection of the role that sensation plays within it. Although the way this happens within each philosophical system differs, he takes it that they both seem to collapse back into something like Hegelian idealism, albeit from opposite directions. He sees myself and Ray as attempting to avoid this danger by championing the work of Sellars, ameliorating the Hegelian dangers of Brandom and Badiou by returning to a more Kantian approach to the relation between thought and Being. The aim here is to give experience its due, without collapsing back into the Myth of the Given, and thereby establish both the principled separation and effective connection between mind and world. However, Dan also suggests that Ray’s greater interest in Sellars’ account of sensation (and the associated notion of picturing) keeps him safer than my own more Brandomian proclivities. Needless to say, I’ve got a few points I’d like to make about this.

1. Brassier, Rationalism, and Me

Before I start addressing specific bits of Dan’s post, there’s an additional point I’d like to make about the relation between Ray’s work and my own. I cannot overstate Ray’s influence on my own thinking. He’s responsible for my increasing commitment to philosophical rationalism, and more recently for forcing me to return to the treasures of Plato’s work (which I’ll say more about below). We have some importantly different influences (Laruelle and Badiou on his side, and Deleuze and Heidegger on my own), and some equally importantly overlapping ones (Kant and Hegel). However, we came to Sellarsian philosophy independently and from different directions at about the same time. Ray has taken the more direct route by leaping headlong into Sellars’ own work (something which I’m lagging behind on), whereas I’ve taken the more indirect route, approaching it by way of Brandom’s eminently post-Sellarsian philosophy of language. This has lead to certain differences of expression and emphasis between us, but these only serve to strengthen our distinct lines of attack upon the conceptual citadel that is Sellars’ legacy. There are obviously disagreements between us, but even I’m not entirely sure we’re they all lie yet.

However, there is one helpful way of thinking about the difference in emphasis between our approaches, which reveals both how they differ and how they complement one another. Ray has famously spearheaded the charge to bring epistemology back into the fold of ‘continental philosophy’, with reasonable success, and I see myself as trying to lead a similar charge on a different front, namely, to bring semantics back into the fold. The systematic study of the content of thought is something that was central to the work of Hegel and Husserl, but the filtration of their legacy through Heidegger and Derrida (and even the later Wittgenstein) has left a philosophical landscape which is exceedingly hostile to the idea that there is a universal structure of thought, let alone that we can describe it in any level of detail. This hostility is one of the central props holding open the chasm between the analytic and continental traditions, and it must be conceptually demolished with as much pyrotechnic flair as we can muster. In this respect, the return of rationalists like Badiou and Meillassoux to the centre of mainstream continental discourse has been entirely positive. This also explains why some of the attempts to find common ground between the traditions by way of the legitimate parallels between Heidegger and Wittgenstein have been so unsuccessful, insofar as they try to use this prop as a bridge. Anyway, it’s important to see that my own and Ray’s charges are not opposed, but are part of a single rationalist campaign to return reason to its rightful place at the heart of philosophy. As Brandom is fond of saying: “semantics is the soft underbelly of epistemology”, and this makes my own concerns entirely congruent with Ray’s.

2. Perceptual Challenges

In order to keep this as brief as possible, I’m going to quote a bunch of passages from Dan’s post and then provide my own comments on them. I’m going to start by tackling the questions he raises about perception, sensation and metaphysics:-

Now, I think that Pete and Ray are prepared to agree in that Sellars’ own account of perception and his account of picturing are not obviously reconcilable with the minimalist account that Brandom provides. A discussion of these two positions merits a full scoped investigation which I intend to carry out in coming work. My impression is that the restriction of perceptual experience to judgments, with no intermediary states of the sort McDowell’s polemical reading argues for to play any part, creates problems for any claims to realism. Specifically, the flattening of perceptual judgments to the general capacity to have the appropriate RDRDs deflates sensible experience in the way that vitiates the way perception serves to anchor us in a causally autonomous world. For one of Sellars’ most important insights is that while perception is conceptual, the ontological constitution of sensation, while remaining epistemically mute, permits us to rehabilitate a notion of correspondence and a theory of picturing in which concepts and so perceptual judgments are causally knit to  physical objects, thus exceeding a purely semantic account of truth.

I’m afraid I disagree here. I think that Sellars’ account of picturing (and the post-Sellarsian accounts inspired by it, such as Ruth Millikan’s work) can be reconciled with Brandom’s minimalistic account of perception. The crucial Sellarsian point is that, in order to play the discursive role they do (i.e., language-entry moves), observation claims must be subject to a certain special kind of challenge on the basis of causal facts about the mechanisms that produce them. What Brandom does is to crystallise and condense this point. His minimalism consists in his insistence that these challenges, and the responses to them, need involve nothing like phenomenological descriptions of the content of the states our perceptual mechanisms are in during their production of observation claims. Put another way, there need be no deployment of my own special capacities to observe my own perceptual states, and a fortiori I need not even have such special capacities to count as perceiving something. This is not to say that the relevant arguments cannot involve such descriptions, or that we needn’t tell a story about creatures who do have the special capacities to produce them, just that none of this is required in order for the relevant epistemic features to be retained.

To see why, we must just apply the same strictures set in the example of the physicist to a different situation, of the sort proposed by McDowell: a man learns to reliably report that their neighbor is home when he sees that his car is parked in the driveway. For Brandom, assuming the man knows that he is reliable, this counts as observational knowledge and he sees that the neighbor is home. However, the difference between the direct knowledge that is involved in seeing that the neighbor is home by having the neighbor before his eyes, and the knowledge that would be obtained in seeing the same fact when seeing the car, is obscured thereby. Both instances would count as cases of direct observational knowledge, having the same underlying fact as their reported, propositional content, i.e. both report that ‘the neighbor is home’. And since they are both non-inferential states, one cannot appeal to the fact that the man ‘arrives’ at such knowledge by a prior consideration of the knowledge that the car is there, since then one must explain how this latter fact motivates the former in situ, which starts sounding a whole lot like inference again. Even if both cases could be construed as examples of non-inferential knowledge, it seems as if what Brassier describes above attests to a complicity between perception and sensible experience, i.e. natural linguistic objects are connected to physical events by the externalist requirement to be properly caused relation to environmental conditions. Picturing describes a non-semantic relation between perceptual states and the world which, while enveloped conceptually, retain autonomy.

The real issue here is what we mean by ‘non-inferential’ commitments. This has two dimensions: i) the normative sense in which entitlement to the commitment is not inferentially inherited from other commitments, and ii) the causal sense in which the commitment is not inferentially produced as the result of some process of reasoning. Both dimensions are important. On the one hand, if we are to stave off the Pyrrhonian skeptical regress, then we need to be able to tell a normative story about how at least some of our commitments can have non-inferential default entitlement. On the other, if we are to stave off challenges of idealism, then we need to be able to tell a causal story about how the world constrains us by directly disposing us to make certain claims about it. However, these dimensions don’t overlap seamlessly. One can become non-inferentially entitled to the same proposition by a number of different causal routes, and some of these causal routes can involve reasoning processes, even if not all of them can.

For example, I can tell that a glass of water is at room temperature by sticking my finger into it, or I can do so by sticking in an electronic thermometer. In each case, I gain a defeasible entitlement to the same claim, but the differences between the causal mechanisms through which the claim is produced mean there are different potential defeasors of the entitlement (e.g., my contracting leprosy, or the shoddy manufacture of the thermometer, respectively). This means that the way in which the observation claim could be challenged in each case is different, even if its consequences are strictly the same. Similarly, although it’s obvious that there is nothing like reasoning going on in the relatively immediate transmission of temperature stimuli from my finger to brain, the electronic thermometer has been designed to take one quantitative reading (e.g., mercury volume) and systematically translate it into another quantitative reading (e.g., degrees celsius). This is a very simple computation, but it is a kind of reasoning process. Modern science is built upon a panoply of much more complicated instruments that automate lengthy series of calculations which we previously would have had to wind our own inferential path through.

So, the normative issue of which observation claims we grant default entitlement to is essentially a matter of which inferential shortcuts we allow within the specific discursive context we’re in. This is why a physicist can count as non-inferentially observing a mu-meson flying through a bubble chamber, rather than inferring its presence from the vapour trail, regardless of whether the causal process that produces his commitment incorporates a quick little chain of reasoning or is merely reflex. It doesn’t matter which causal process is involved in taking the normative shortcut, what matters is whether the normative shortcut is permissible within that discursive context, i.e., in the context of trained physicists in a laboratory setting. This shortcut wouldn’t have been present when the bubble chamber was first invented, as the theories and practices upon which it was based would still be contentious enough to preclude such default entitlement. We can then even take the human element out of these shortcuts, and construct more elaborate observational devices on the basis of them. This is precisely what has happened in the development of increasingly sophisticated particle detectors over the last half-century or so.

Quickly parsing McDowell’s example then, yes, the man can ‘see’ that his neighbour is home by seeing that his car is in the drive, insofar as there is a reliable link between the two that has been transformed into an acceptable discursive shortcut. If asked by his wife whether his neighbour is home, he can simply look out the window and legitimately say ‘yes’, rather than ‘his car is in the drive, therefore he is home’. However, his claim is subject to challenges that the more reliable process of responding to the neighbour’s bodily presence is not, such as the assertion of the defeasor: ‘remember, honey, that he’s had car trouble of late, and has been biking to work a lot’. In conclusion, talking about picturing relations is fine, but this is a matter of understanding the ways that the functional structure of perceptual mechanisms underwrite challenges to the observation claims they produce, and this is something that extends beyond our own biological sensory systems to the experimental equipment we’ve built to augment them.

And Sellars’ supplementary account of picturing is not clearly reflected in Brandom, particularly since it is not clear that he would allow for the ontological role sensation plays within our understanding of perception.

Again, I think that an account of picturing can be slotted in to Brandom’s minimalist account. However, the real problem here is the idea of ‘ontological role’. The whole point of my approach is to derive the very idea of something like ‘ontological role’ from the structure of objective discourse itself, and particularly the way it necessarily involves perceptual constraint. This precludes me from using any ontological claims to describe the nature of perception. This is why I find Brandom’s rigorously non-metaphysical account so compelling. I thus think we have to be very careful in talking about the ontological or metaphysical valence of sensation not to put the metaphysical cart before the epistemological horse.

My own position on this matter is that Brandom’s strict inferentialism seems to reproduce the pragmatic conflation of thinking and being which already plagued Quine, and which ends up undermining materialism, not unlike Badiou, in lacking the sufficient resources to disambiguate between form and content. Content dissolves into propositional content, and what Pete calls a ‘thick’ sense of reality falls out the window. I think with Pete that a thick notion of reality is necessary, and that such a notion needs to be advanced without reintroducing the metaphysical dualism of thought and matter. Thus the key moment for Brassier remains the Kantian juncture between the non-metaphysical normative space of reasons and the ontological-natural-causal domain of natural scientific research, while for Pete’s more Brandomian, and by extension Hegelian, position (although I agree with Zizek in that Pittsburg Hegelianism is a misreading of Hegel) the primary task is to eviscerate the myth of phenomenological content in favor of the primacy of logical relation in a deflationary account of thought.

I’m pretty much behind this, but I take issue with being described as being more Hegelian because I am more Brandomian. I aim to stand with Ray in the Kantian juncture, I simply try to do it by cultivating the Kantian themes implicit within Brandom’s work against his own more Hegelian intuitions. There is sometimes a tension between Sellars/Kant and Rorty/Hegel in Brandom’s work (e.g., in his more recent work on modal semantics and empirical content), and here I try to push Brandom back towards the former in opposition to the latter.

3. The Value of Aesthetics

I’m now going to tackle what Dan has to say about aesthetics, perception and value. To begin:-

Here recent questions about aesthetics raised by Peter and Brassier become peculiarly interesting, since I think they can allow us to see how perception continues to play a role in the story in a way that illuminates the peculiarity of Sellars’ position. As I take it, Brassier endorses the modern severance of the beautiful from the sensible, thereby advocating the former’s allotment to the conceptual. Here I agree with Pete in that the intrication between the Beautiful, the Good, and Value brings Plato to the context of a rationalist epistemology with an inferentialist bent. The basic idea is that we can distinguish Beauty as a species of Value, with varying scales of Universality. Thus Pete distinguishes a broad sense of Beauty akin to that of Value-in-itself, or its pure form, which is independent of all rational interests, and a narrower sense in which different aesthetic values are pitted against each other within the conceptual norms of the sensus communis and which make possible the negotiation of aesthetic judgments.

What he’s discussing here are ideas that I’ve been developing in correspondence (principally with Ray) that I haven’t yet posted on the blog. I’m now going to post some of that correspondence in order to contextualise what he says. I’m principally going to quote myself, but I will include a small quote from Ray in order to make the best sense of what is said. I’m never quite sure of the ettiquette of quoting from emails on blogs, and I try to err on the side of caution, but I don’t think Ray would mind in this case. Here is the relevant part of the first email from me on the matter:-

I’ve had another idea that I think you might find interesting. I’ve told you about my attempt to provide an alternative categorical imperative before: the idea that one must preserve the institution of rationality itself above all things [see here]. This is of course part of a more complex structure of imperatives, but it provides the crucial link between the abstract transcendental norms of rationality (e.g., the obligation to divest oneself of incompatible commitments), and the concrete structural norms through which these are instituted within a rational community (e.g., the norms through which we individuate rational subjects and the acts through which they undertake commitments). It commits us to having to deal with the actual way in which rationality is instituted in the world, and thus forces us to engage in the kind of complex instrumental reasoning regarding our practices that Kant’s categorical imperative eschews. The big problem I’ve had so far is justifying this idea. I now think I may have found this justification in the most unlikely of places – Plato’s account of the Beautiful.

I’m starting to think that Plato’s claim that the Beautiful itself is what is most beautiful is correct. This is normally interpreted as a matter of some archetype of beauty, which is itself in some sense a beautiful thing. This obviously doesn’t work. However, if one instead interprets it as the claim that the very fact that there is beauty in the world is what is most beautiful, then it makes a lot more sense. This beauty is something which shines through in every beautiful thing, rather than something to be opposed to them as a further thing with which they could be compared. In addition to being valuable for its specific features, each beautiful thing is valuable insofar as it manifests beauty as such. I think this is a pretty interesting idea in itself, but it becomes far more interesting if we expand it to the genus of Value as such. This would mean that what is most valuable is that some things (be they objects such as artworks, or actions such as good works) are valuable. This is a kind of transcendental value that all valuable things manifest. Given this, we could categorically prioritise the value of anything that is a condition of the possibility of imbuing things with value. Given that the institution of rationality is the fundamental condition of this, we can thereby justify the categorical imperative I’ve proposed.

What’s even more fascinating is that this provides an interesting way of thinking about the various different species of Value, such as Beauty, Goodness and Truth. What is most good is that there is goodness (this prioritises the institution of an ethical community), and what is most true is that there are truths (this is a nice counter to relativism). Moreover, it lets us think about the relationships between them in interesting ways.

To begin with, Beauty is that which is supposed to be valuable in itself independently of any practical or instrumental worth. There are some tricky issues in working out this claim, such as how to make this independence compatible with the idea that it isn’t intrinsic (i.e., that it is imbued), and the distinct between Art and natural Beauty. I think the former question can be handled reasonably easily, insofar as we can explain Moore’s proof of intrinsic value (i.e., imagine two universes without people, but one has an artwork in it, which one is more deserving of existence?) in terms of the way hypothetical reasoning functions. The underlying phenomenon is essentially the same as Kripkean rigidity: the normative properties we ascribe to things in counterfactual situations are the same as those we ascribe to things in actual situations, with all this implies, unless stipulated otherwise. This should be sufficient to let us separate independence from intrinsicality. The latter question is a lot more tricky, and I need to think about it more. It’s exceptionally important given the connection between Art and the Foucauldian theories of Freedom and Justice I’ve been developing, in which the imperative is to construct ourselves (as individuals and as collectives, respectively) as works of Art, in accordance with the ideal of rational autonomy [see here]. I suspect there’s a strong connection here.

Following this, Goodness is that which is the condition for Beauty, insofar as it is the norm of acting in accordance with the categorical imperative of maintaining the institution of rationality, without which there could be no Beauty. There is thus a sense in which Beauty is a higher ideal than Goodness, insofar as it is not absolutely independent in the same way. However, this subordination also conceals a prioritisation, insofar as because Goodness is necessary for Beauty, we must follow the dictates of ethics over those of aesthetics. This means that the categorical imperative is truly categorical even though it derives its force from elsewhere. We must perpetuate the institution of rationality even when faced with the transitory destruction of Beauty. This can be true even when it is true that a universe containing an artwork is in some sense better than one without. I still need to connect all this to the principle of autonomy, and thus to the ethics of Freedom and the politics of Justice, but I think it can be done, and that it’s natural to see these as species of the Good.

Finally, I think we can place Truth within the same hierarchy as the Good and the Beautiful. This is because we can see that the fundamental norms of discourse which constitute the ideal of truth are conditions of the possibility of the institution of rationality. In a sense, they are the most fundamental layer of the transcendental cake. There can’t be any effective ascription or revision ethical claims, let alone aesthetic claims, without them. However, it seems to me that the Good still mediates the ideal of truth, insofar as we can imagine legitimate situations in which the quest for truth would be truncated in favour of ethical demands. The Good deals with the way in which the ideal of truth is instituted in practice, and thus retains its categorical status within the practical domain. This is something that requires working out in detail though. When we consider the relation to Beauty, we actually end up with the classic question of whether Truth is beautiful. This then gets broken down into a series of complex issues: Is the sheer fact that there are truths valuable in itself? (I’d suggest yes, but this needs working out); Is every specific truth valuable simply qua truth, or is there a further non-instrumental sense in which specific truths can be valuable? (again, I’d suggest yes in some but not all cases, such as some of the truth of mathematics, which may be profoundly beautiful); What is the hierarchical relation between the value of Truth/truths and Beauty/beautiful things? (this one is going to be especially complex, I’d suggest that (i) Truth is subordinated to Beauty in the same way as the Good is, but that it’s own Beauty puts it above all other specific Beautiful things/states of affairs, (ii) both Truth and Beauty rank above any particular truth or beauty, and (iii) that there is a free for all between specific truths and specific beautiful things.) Regardless, it’s fantastic to be able to resurrect this kind of question.

Needless to say, this is all pretty rough, but I really like the shape it’s taking. I never would have suspected I’d become a Platonist, but I suppose you’ve been a bad influence on me. I’m hoping to start a Plato reading group up here with the intention of reading through the collected works. I don’t know if it’ll come together, but it’d really give me the incentive to familiarise myself with it all. I’m also now convinced that I need to read Lotze, as I suspect this might not be too far away from his normative reading of Plato.

Ray responded to this by challenging my use of the notion of the Beautiful, and raising a number of really interesting questions in relation to it:-

Your thoughts below on Plato are fascinating, as always. I guess it’s the idea of the beautiful that is most problematic and most difficult to rehabilitate for me, since I take it it’s precisely the category of the aesthetic, in its alleged contradistinction to the conceptual, that artistic modernism necessarily subverted, and once the bond between beauty and sensation has been severed—as I think it must– then the distinction between the beautiful thing and the beautiful as such, or “unthinged beauty”, becomes more difficult to articulate than in the cases of truth or goodness, precisely insofar as the characteristic marks of the beautiful become wholly conceptual (as when mathematicians speak of a “beautiful proof”). Indeed, ironically enough, I note that for a long time now, my mistrust of the category of the beautiful as applied to art has been accompanied by a tendency to characterize abstract conceptual structures as beautiful in themselves: Sellars’ for instance! But must the obverse of learning to conceptualize feeling be the aestheticization of thinking? Is the aestheticization of the conceptual the unwitting complement to the de-aestheticization of sensation? Perhaps Hegel had already worked some of this out in his account of the death of art? Is the reified dualism of reason and feeling overcome when the beautiful is reconceived as the “self-showing” or “exhibition” of the concept? I’d be very keen to learn if is this is part of what Hegel meant, and I suspect it may be quite close, or at least pertinent, to what you are trying to work out with Plato.

This forced me to articulate some of my ideas in greater detail, particularly on aesthetics and the place of the notion of the Beautiful within it. It’s these comments that Dan was alluding to in his talk about the split between the two senses of Beauty above:-

I’m glad you found my thoughts interesting. You certainly raise some difficult questions in response. My own thinking on the nature of Beauty and Art has been undergoing a lot of interesting transformations of late. I’ve always been a big fan of both Kant’s aesthetics and Deleuze’s aesthetics (with Heidegger sitting somewhere in the middle), but I’ve had great difficulty in reconciling them, despite certain obvious affinities (e.g., in the account of the Sublime). I most certainly need to read Hegel, and see whether he can help me resolve some of these issues. What has always attracted me to Kant’s aesthetics (other than my obsession with his transcendental psychological analysis of the experience of the Sublime and the Beautiful) is his emphasis upon the subjective universality of aesthetic judgements and his insistence upon their disinterestedness, both of which capture important features of the structure of arguments about aesthetic value, in my opinion. These correspond to what I called in the last email non-intrinsicality and independence. By contrast, what has always attracted me to Deleuze’s aesthetics is his insistence that Art is not a matter of communication, but one of composition. It is not something principally concerned with the artist’s intentions, but with the artwork itself. I support this even in the face of Deleuze’s notorious hostility to conceptual art, insofar as I think it’s entirely possible to analyse conceptual art as using conceptual materials to compose affects that deploy our conceptual capacities, rather than our sensory capacities. The real problem with both Kant and Deleuze is accounting for the distinction between Nature and Art. Kant treats Art as for the most part a pale imitation of the Beauty of Nature, and Deleuze’s focus on the independence of the artwork prevents him from drawing the distinction from the other side, as it were.

My Platonic musings have opened up a possible path out of these difficulties. This is because I’ve been forced to consider Beauty as a species of Value independent of the nature of the aesthetic experience, which is something that both Kant and Deleuze are very focused upon. I still haven’t settled on fixed terminology here, but I’m tempted to distinguish between a broad and a narrow sense of ‘beauty’. The broad sense is precisely what I was trying to define in the last email: that value which is entirely unconditional, insofar as it is entirely independent of all interests. It is crucial that it is even independent of the formal interests of reason itself. It is the limit-case of Value: what is purely and simply valuable. In this it exemplifies both of the Kantian characteristics above: subjective universality and disinterestedness. However, it’s important to recognise that this simplicity does not completely isolate it from rational assessment, leaving it open only to revelation (becoming its degenerate theological form: [the holy]). It is perfectly licit for us to argue about why something is valuable in this sense, and whether something is more or less valuable than something else, in virtue of a variety of other qualities it possesses. However, I suspect that such discourses are to some extent free-floating, and can never become anchored in any sort of fundamental reason why any given thing is valuable in this way (in contrast to ethics). They will thus always be grounded in something like a Kantian sensus communis, which establishes the consensual premises from which aesthetic discourse can proceed. It is important that this need not be understood entirely in terms of common sensory capacities, but equally involves common conceptual capacities. It’s simply that these conceptual capacities play a role in the production of affective experience. It is at the level of these capacities and the way they constitute our sensus communis that we start requiring more specific aesthetic categories, and this is where the narrow concept of Beauty is to be found, describing one kind of aesthetic value as opposed to another (e.g., as opposed to the Sublime, which is also a species of the Beautiful broadly construed). There is plenty of room to integrate the insights of Deleuze’s aesthetics at this level.

The question is then how to distinguish Beauty in Art and Nature. It’s possible to think that this is a non-problem. Art is simply just Beauty that contingently happens to be composed by rational agents for rational agents, but this has no important baring upon the nature of its Beauty. I don’t think that this answer is satisfactory, and I suspect this is where Hegel will perhaps turn out to be helpful. The difficulty for me here is reconciling the seemingly intentional nature of Art with what I think is Deleuze’s correct assessment of its non-communicative character. I think that the secret is that Art must engage our communicative capacities, without thereby being a communication. Art is framed in such a way as to produce affective experiences that are dependent upon our communicative understanding, without thereby having anything like a determinate semantic content. This makes Art essentially conceptual in a way that natural Beauty is not, even in cases of Art that are principally dependent upon our sensory capacities. This might be the basis of something like Hegel’s hierarchy of arts, insofar as one might argue that the highest forms of art are those that exemplify this conceptuality, but I’m not sure I’d be willing to endorse this yet. At the very least though, it enables us to distinguish forms of man-made Beauty that do not count as Art insofar as they are not framed in such a way as to involve our communicative capacities. There may be some sense in which my house is painted beautifully, even though it is not Art in any sense. There is equally the possibility of the communicative framing of Nature, which feeds into the degenerate theological forms of aesthetic value discussed above.

Finally, this feeds into some of the themes of my Spinoza post [again, see here], insofar as I’ve been trying to present politics as the practice of constructing the state as a work of Art in accordance with the ideal of Justice, in much the way that Foucault presents ethics as the practice of constructing oneself as a work of Art in accordance with the ideal of Freedom. You’ll see more of the details of my account of Freedom and Justice in the post, but I think it’s interesting to add here the way in which the notion of Art functions to supplement both of them. The first point is that it unbinds the process of construction from desire, in such a way that desire itself is something that can be acted upon. The second point is that although it does not unbind us from Freedom and Justice, which place fundamental limitations upon the process of self-construction, it nonetheless exceeds them. In essence, we are to build ourselves in accordance with values that exceed both our own desires on the one hand, and the enabling structures of Freedom and Justice on the other. The third point is that this makes the construction of a sensus communis an important aspect of the process of construction itself. I haven’t thought through this in detail, but it is worth thinking about. I believe Arendt was interested in the potential of this as a political concept. The final point is that there is something interesting about the idea that Art involves a communicative framing, and that we relate to ourselves as works of Art. In what sense are we framing our relations to ourselves communicatively here? I’m particularly interested in the sense in which something like a state could be seen to be relating to itself in a communicative fashion, especially when there is the possibility that there is nothing outside it to get the message.

All of this work on value theory and aesthetics is still in a very early stage of development (though there is a precursor: here), but I think it looks very promising. I’m also having a great deal of fun with it, and I’m aiming to submit something to the Thinking the Absolute conference at Liverpool Hope that draws on some of these ideas (my provisional title is ‘Absolute Spirit as a Work of Art’). Dan has riffed on these ideas in a number of interesting ways. He’s taken up the concern with the link between perception and the rough account of aesthetics I’ve presented in more detail than I’ve done, and he’s tried to clarify my idea of Art as communicatively framed without being communicative by introducing the concept of semantic indeterminacy.

With regard to the former riff, he’s mentioned something that we’ve discussed privately in more detail, namely, the incompleteness of Sellars’ account of ‘looks’ talk. He says:-

The basic idea is to disambiguate a sense of looks-talk which is not merely emphasizing the epistemic withdrawal of endorsement before a proposition about the world, i.e. in which the function of ‘x looks y to S’ is not reducible to ‘x withdraws endorsement from x is y’. This can be exemplified by using predicates such as ‘looking fuzzy’, where it is clear that the role of ‘looking’ therein is to make a report about a fact concerning the functioning of our perceptual mechanisms, and not the epistemic withdrawal of endorsement. This means that we can accept that there is a role to be played for experientially specific judgments about perception which provide the anchoring on sensibilia without rehabilitating the valence of sense datum transparency, and which would thus be continuous with the conceptual envelopment of the aesthetic.

This is a connection I had not yet made, and I think there’s a lot of very interesting work to be done here connecting up this expanded form of ‘looks’ talk with the structure of aesthetic discourse more generally. This is somewhere where I can provide some more additional context from a much older bit of correspondence I’ve been meaning to clean up and post here for a while. This is from a discussion of Sellars’ theory of perception between Ray and myself that came out of his comments on my Essay on Transcendental Realism, which in turn arose out of conversations we had at the Transcendental Realism Workshop. It expands on some of the ideas mentioned in my comments on Dan’s remarks on perception:-

Getting on to your points about Sellars’ account of perception, I should initially say that you’re right that I went beyond the Sellarsian position in talking of getting rid of intuitions entirely (and not merely pure forms of intuition). I’ve since had to opportunity to read James O’Shea’s book on Sellars, which I very much enjoyed, and thus now have a slightly better frame of reference from which to judge my proximity to Sellars. Now, I agree with you entirely that we must be able to make a place for sensation which does not collapse it into conception, and that doing so is both paramount to completing the eradication of givenness and more difficult than it seems. However, this is the part of Sellars’ account of perception that I can’t yet bring myself to endorse. I pretty much endorse the account of [Reliable Differential Responsive Dispositions (RDRDs)] (as modified slightly by Brandom) and the account of looks talk [see here], but I am unmoved by the grain argument, which seems to provide most of the motivation for his account. Similarly, as much as I have process philosophical leanings, I simply don’t think that such concerns about the roll of sensation can mandate the adoption of a process philosophical approach. Indeed, the distinction between physical1 and physical2 that Sellars proposes strikes me as reminiscent of ad hoc emergentism. I must emphasise that this is all tentative, as I have not read much of Sellars himself, but these are my first impressions on the matter.

To go into a bit more detail, I’m of the opinion that we need to tell a story about the role of sensations within perception that is analogous to (and importantly related to) the story we tell about reliable [differential] responsive dispositions. This is to say that it must be a story that is sufficiently abstract to be independent of the empirical study of the particular perceptual mechanisms we possess, while also making space for such empirical study. To this end, although I think that we should indeed talk about ‘sensations’, I think we should be wary of talking about ‘sensory content’, insofar as this encourages us to treat sensation as analogous to the propositional/conceptual content of seeing that/as. To explain further, what justifies talking of content in either case is the notion of identity of content. There are propositional contents just insofar as two different utterances can express the same proposition, and similarly there are sensory contents just insofar as two people (or the same person at different times) can have the same sensation. On this basis, there are two different ways in which one could articulate a notion of sensory content, lets call them content1 and content2.

Content1 is general enough that we can talk of identity (along with similarity and difference) relations between sensations in a way that does not depend upon any functional specification of the perceptual mechanisms in which they’re involved. Content2, by contrast, is dependent upon such functional specifications, so that we can only say that two people have the same sensory content only in the same way that we can say that two video tapes have the same content. It’s important to note that such functional specifications don’t require that the content be stored in the same ‘format’ as it were. We can talk about a VHS and a Betamax tape having the same contents because there is a functional mapping between them, much as we can talk of two different file formats storing the same content on a computer. What is important is that anything like identity relations are dependent upon such functional mappings. This means that whether or not two creatures can share something like a sensory content2 is dependent upon facts about the causal structure of their perceptual mechanisms that are to be determined by empirical study. I don’t have any problem with the idea of ‘sensory content’ if we’re talking about content2, but if we’re talking about content1, which I think most people are, then I have issues.

The only way in which it is possible to tell this kind of generic story about sensory content independent of the specific functional systems involved (content1) is to make it parasitic upon propositional content, which is independent in precisely this way. This is exactly what happens when we take ‘looks’ and ‘seems’ talk to describe internal states, because such talk is itself parasitic upon ‘is’ talk (as Sellars has shown). This is what leads us to talking about sensations as having properties analogous to those we ascribe to the things that cause them, such as colour and shape. This is also what seems to lead Sellars to talking as if there is ‘something’ that is pink and cubic. However, if we abandon content1, then we can retain the idea that there are sensations while nonetheless denying that these sensations have properties like ‘pinkness’ and ‘cubicness’. We can equally deny that they have properties like ‘homogeneity’, and thereby reject the grain argument. We still have to show that content2 can do all the jobs that people want content1 to do, which involves providing a more nuanced story about how ‘looks’ talk allows us to describe our sensations, but I think this is possible.

Let’s start with ‘looks’ talk. Following Sellars, we know that it’s ordinary use is essentially to express dispositions to make certain claims whilst nonetheless withholding assent from those claims. The most simple way in which this happens is when we have some knowledge which acts as a defeasor, for example, when I say ‘the ball looks green’ because I know that ‘the lightbulb illuminating the scene is blue’. This is not a matter of any change within my perceptual mechanisms, but a matter of my understanding of how my perceptual mechanisms work under various circumstances. However, ‘looks’ talk can also be used to describe the content of my sensations, such as when I describe the pink elephant I’m currently hallucinating. In this case we have to treat ‘looks’ talk as using concepts which ordinarily ascribe occurrent properties to perceived entities to indirectly describe internal states of our perceptual mechanisms. The best way to do this is by introducing what we might call second-order reliable [differential] responsive dispositions to respond to internal states, analogous to those dispositions Sellars posits for responding to our own internal thought episodes. These second-order [RDRDs] develop out of first order [RDRDs], insofar as my ability to describe my sensations as ‘coloured’ and ‘shaped’ is dependent upon my ability to describe things in the world as coloured and shaped. However, this does not mean that my sensations are coloured or shaped. Rather, our second order observations (or introspections) tell us about specific features of our internal states that are dependent upon the particular constitution of our perceptual mechanisms, and they do so by feeding into theories about how our perceptual mechanisms are structured. It just so happens that the primitive functional categorisation of our perceptual mechanisms we start with is modelled on the things we perceive (much as Sellars says). This does not mean that this theory, and our second order [RDRDs] can’t develop so as to model and respond to these states in more nuanced ways. It is both possible for us to develop more complex theories which allow us to take our own and others’ ‘looks’ talk to indicate certain states (even when the others in question do not understand the theory) and to develop new [RDRDs] to respond to features of our sensations that are not easily modelled on the things we perceive (e.g., the many counter-intuitive aspects of our sensory apparatus that Dennett is so fond of). This is somewhat sketchy, but I think it’s along the right lines.

The other major thing that needs addressing is the common factor problem. I think the content2 approach is not only perfectly adequate to this, but that it allows for quite an interesting position that incorporates aspects of both highest common factor accounts of perception and disjunctivist ones, insofar as it allows for a whole spectrum of similarity in content2. On the one hand, we can follow disjunctivism in allowing for situations in which the sensory content of a hallucination is distinct from that of a veridical perception, while maintaining that we cannot tell the difference between the two (our second order [RDRDs] being insufficiently sensitive or otherwise impaired), and on the other hand follow the highest common factor view in allowing for situations in which the sensory content is functionally identical despite the absence of the object perceived (e.g., perfect simulation through stimulation of the optic nerves). The interesting thing is that the extent to which there is a common factor in perception is something delegated to empirical study of the perceptual mechanisms involved. So, on this view, our sensations are things we have unique, but not privileged access to. We can be wrong about the character of our experience, but we can equally be trained to introspect interesting aspects of it that are not readily apparent.

As a final point, I also think that insofar as this approach bequeaths the study of sensations to empirical science, it opens us up to understanding the nature of sensations in far more complex ways (as opposed to classical pictorial approaches). To illustrate why this is interesting, it’s important to remember that sensations are functional states bound up in the causal mechanisms in which first order [RDRDs] consist. This means that, despite our rejection of sensations as possessing anything resembling conceptual content, they must be in some sense proto-conceptual insofar as they are involved in the process of eliciting conceptually articulated responses. Sensations do not need to be understood as manifolds of raw sensory input, but include the functional features of the processing of this data (which ultimately results in those conceptually articulated responses). How complex this processing is, and thus how the second order [RDRDs] fit into it is a matter for empirical study, but it would be reasonable to surmise that introspection can be involved at different levels of this process, rather than being limited to one.

I must reiterate that this is all very tentative. A large part of this tentativeness stems from my general discomfort with the use of ‘functional’ vocabulary insofar as I take this to be a species of normative vocabulary (and thus broadly interpretational discourse on my characterisation). This ties back into the brief point I made earlier about my opinion that functional discourse should be understood along the regulative lines one finds in the 3rd critique. In order to properly square the account above with my other commitments I thus need to complete this analysis of functional discourse, and secure my entitlement to functional concepts. I think this can be done, precisely because the above account does not give sensations any special metaphysical status, but places them squarely within the domain of empirical science. I think this provides the necessary disconnect between the normative terms used to discuss sensations and those used to discuss concepts, insofar as the former are regulatively deployed within objective discourse and the latter are constitutively (?) deployed within interpretational discourse. This requires more work though.

There’s a lot to process here, but you should be able to see the way it fleshes out what I earlier called our ‘special capacities to observe our own perceptual states’, in terms of what I call second-order RDRDs. What’s interesting about Dan’s suggestion is that it makes the sort of empirical study of the relationships between first-order RDRDs, second-order RDRDs, and our conceptual capacities I discuss here into a valuable resource for aesthetics. It may be that we can wean aesthetics from it’s historical dependence upon introspective phenomenology in favour of a more complicated and potent extrospective phenomenology.

Dan sums a lot of this up when he says:-

The idea would then be that the semantic indeterminacy introduced in artistic creation would be the (dialectical) interplay in the production of new perceptual judgments and relations, and which include (albeit not exclusively) statements about how things produce affect as states relative to the functioning of our sensory organs, i.e. perceiver-relative facts.  These would constitute the specifiable content which relativizes aesthetic judgment to perceptual judgments, though not wholly, without losing grip on participation in the generation and negotiation of value. The trick here is to coordinate properly aesthetic judgments in art with perceptual judgments (whose content is determinate) to explain how the indeterminacy of artistic works themselves is to be understood relative to the articulation of conceptual norms within the sensus communis. The obvious question is whether this requires that we make aesthetic judgments in nature subject to the same sort of dialectics, and how the intentional stance ultimately weighs in.

I don’t have anything to add to this other than to say that I agree entirely.

Finally, Dan poses a couple additional quandaries for the account of Value and Art laid out above:-

 As a provisory note, I would remark that the notion that value is ‘independent’ becomes quite difficult to cash out. On the one hand, I think that the distinction between the projection of value (which is our prerogative), and its construction (which we don’t make) is opaque for the moment. It also doesn’t seem clear to me what natural value consists in, beyond the trivial assertion that nature is not made by or for us. If by projection we simply mean that we need to deploy concepts to make aesthetic judgments, then it is not clear how these judgments are proper to art because of its intentional inflection. Clearly, some sort of projection in that sense would be necessary for natural judgment. This is ultimately a tangential matter.

There are two different registers that are getting mixed here: the difference between independence and intrinsicality, and the difference between communicative and non-communicative forms of Beauty. The former concerns the nature of Value as such (including Beauty) and the latter deals with the difference between Beauty in Art and Nature.

The first idea is that whether something is valuable (and how valuable it is) can be independent of our attitudes about it, which is to say, there is some important sense in which our attitudes about value can be wrong, without this being grounded in some intrinsic value property possessed by the thing, analogous to its various empirically describable properties. The means that we can have genuine arguments about value, without thereby thinking that these arguments are resolvable by the same means as arguments in the natural sciences. The justification of this idea is twofold. On the one hand, we acknowledge that (in accordance with Brandom’s objectivity proofs, Moore’s thought experiments, and Kripke’s ideas about rigid designations) we can have meaningful arguments about the value of things in counterfactual conditions in which our attitudes are different (e.g., if I’d been brought up to hate Captain Beefheart’s music, he’d still be a fucking genius) or non-existent (e.g., if humans had never evolved, the sunset would still be beautiful). On the other, we acknowledge that the essence of value (and all normative statuses) is its role in rationally motivating action (see here). The normative force in which value consists is impossible without the conditions facilitating rational action that the institution of rationality provides. It is in this sense that value is projected upon the world by us: it makes no sense outside of the framework of practical reason.

The second idea is that what makes something Art is the fact that it is framed in such a way as to stimulate not only the capacities that we use for engaging with non-intentional features of the world (which as Kant understood are conceptual as well as intuitive: the experience of beauty involves the freeplay of the imagination and the understanding), but also our capacities for engaging with its intentional features, namely, other rational agents, their actions, and their dispositions. Not all use of concepts is communicative. It’s not deploying our ability to use concepts in causing affective experience that makes something Art, it’s deploying those same abilities as if they were being used to interpret intentional phenomena (such as straight up linguistic communication) that does it. In essence, Art is the contribution to the Beauty of the world that we make, or that we see ourselves in. This maybe a bit Hegelian, but I need to do more reading to be sure.

4. Conclusion: The Road Ahead

That’s all I have to add to Dan’s musings, but I’d like to finish with a further quote from him that sums up the task we are faced with in very concise and powerful terms:-

The idea is finally to defend a sense of objectivity as part of metaphysics, without for this reason endorsing a neo-Scholastic metaphysics of objects, and a suitable notion of objective knowledge and so an epistemology, without for this reason endorsing a metaphysical divide between transcendence and transcendent.  And all of this while describing the nature of experience as involving a conceptually circumscribed role for perception as that which anchors our relation to the external world.

Wonderful.