Brandom’s Woodbridge Lectures. More on Autonomy

Following up from my last post, I just read (on Pete’s advice) Brandom’s Woodbridge lectures (available here). This is all really a digression, and I want to get back to Making It Explicit – but I was worried that if I didn’t post something on the lectures they’d drop entirely from memory. So very quickly:

The core of the lectures is a reconstructed narrative of the shift from Kant to Hegel in German Idealism – Brandom sees both as (idealist) pragmatists interested in the active creation of concepts and norms in the synthesising activity of subjects. The shift from Kant to Hegel is, basically, a shift from understanding this practical activity as rooted in the individual subject, versus understanding it as rooted in community or society. Brandom endorses a (reconstructed) Hegelian position – and I agree with most of his conclusions, even though I don’t agree with the idealist way they’re framed. Brandom has a very nifty interpretation of the Kantian synthetic unity of apperception, and a nicely pragmatist/deflationary read of Hegel’s ‘march of reason through history’. (The march of reason can only be seen as such retrospectively, because it only is such retrospectively – that is, from the social-temporal perspective of a specific community’s norms and commitments and understanding of ‘reason’). I don’t have enough Hegel to know how forced Brandom’s read of him is – but (from what I remember of the First Critique, which I admittedly read a very long time ago) I thought Brandom’s discussion of Kant seemed really astute and sound – Brandom seemed overly apologetic for his departure from standard interpretations.

Anyway, hopefully I’ll be able to revist some of this content in relation to Brandom’s other works. I’ve already complained about some of the lectures’ ideological nonsense (in a comment here) so I’ll leave that aside. In this post I want mainly to revisit the question of ‘autonomy’, which I discussed in my last post.

Now unfortunately Brandom’s voicing throughout these lectures is rather unclear – he’s giving his reconstructed versions of Kantian and Hegelian positions – these reconstructions clearly owe something to Brandom’s sense of what these figures should have said, given their commitments, rather than simply what they did say – but that doesn’t mean that Brandom himself necessarily endorses his reconstructions. I’m therefore unclear just how much of the lectures’ content expresses Brandom’s own positions (except towards the end of the last lecture, where Brandom is clearly articulating positions that he himself endorses).

I’m going to do this at a pretty high level of abstraction, therefore, and with somewhat tinker-toy versions of the claims. Obviously I’m not at all abreast of the broader debates on these issues, and would be more than happy to be informed of further material I should engage with, or arguments that I should take account of.

So – The claim we started off with was the ‘thesis of autonomy’. In Pete’s words: “This is the claim that any authority is binding on us only insofar as we accept that authority. This includes the authority that norms have over us.”

Now at one level this is trivially true, if what it concerns is our experience or conviction that norms are binding on us. If what we mean when we say that norms are binding on us, is that we are experiencing them or understanding them as binding on us – then, obviously, our experience/understanding of norms as binding involves some level of accepting them as binding – otherwise they wouldn’t seem binding. I don’t know if this is tautologous, but it’s certainly pretty close.

If this is the claim then I have no problem with it. But I think the claim is meant to be stronger than this – norms can be binding on us that we don’t experience as binding, but have nonetheless accepted as binding, without being aware of it. This is also, of course, true. This is one of the areas where Brandom’s emphasis on inference pays dividends – in committing oneself to proposition A one has also committed oneself to propositions B, C, D, E, F, G, etc. – because commitment to these propositions follows from commitment to proposition A, according to the norms of concept use deployed in the specific linguistic game of asking for and giving reasons in which one is presently engaged. (Commitment to these further propositions is just what commitment to proposition A means, on Brandom’s inferentialist account of propositional content – but one need not be aware of every aspect of A’s meaning, in using A.) These norms are not determinable solely by the individual subject, but rather by the series of ongoing (open-ended and of course potentially contradictory) social interactions that constitute the community/communities to which one belongs by virtue of one’s participation in such linguistic interaction.

All this is fine and dandy. Brandom expresses it in terms of the difference between one’s commitment to a norm (which one can control) and the content of the norm to which one commits oneself (which one cannot control – because it is socially, rather than individually, determined). But it seems to me that this set-up leaves the putative principle of autonomy vulnerable to the argument I articulated in my last post. Once our theory has acknowledged that one’s deontic status – the norms one is obliged to follow – is determined not just by one’s own actions, but also by the actions (deontic attitudes) of others, there seems no meta-theoretically non-arbitrary way to limit the determination of deontic status by others to the content of the norms one is obliged to follow, as opposed to the matter of which norms one is obliged to follow in the first place. In fact, there seems to be no way to coherently distinguish between these two situations, at the level of abstraction at which Brandom’s basic theoretical apparatus is pitched (for one of the many features of Brandom’s Hegelian modification of the basic Kantian normative pragmatics is to collapse the conceptual distinction between appearance and thing-in-itself, at the most general discursive level available to analysis.)

[That last sentence is pretty compressed and the thinking behind it therefore probably unclear – this is sort of the core of Brandom’s third lecture, and is perhaps a somewhat complicated argument. I’d like to post on this a great deal more, but ideally I’ll finish Making It Explicit before I do so, since my assumption is that Making It Explicit is where Brandom gives the most developed articulation of this argument. In a word, this is essentially another version of Brandom’s embedding of the concept of objectivity within his inferentialist semantics, as opposed to the grounding of the latter in the former. But that requires a post or posts of its own. In any case, I don’t think that this point is necessary for the argument of this post to work – though it probably strengthens it.]

The point is that I don’t see on what grounds a Brandomian apparatus could coherently and non-arbitrarily distinguish, at the most general theoretical level, between (legitimate) obligations one has taken on oneself, and (illegitimate) obligations that one repudiates oneself, but that are imposed upon one by others. The move from a Kantian to a Hegelian understanding of the constitution of normative commitments seems to preclude such a distinction (at, let me again repeat, the ‘transcendental’ level at which Brandom’s argument is pitched).

This has various consequences. It means, for one thing, that there’s nothing intrinsically meta-theoretically wrong about attributing obligations to someone, even if that someone utterly repudiates those obligations. Thus, there is no intrinsic meta-theoretical problem with my saying of a person “You are obliged to do exactly as I say”, even if they completely reject the legitimacy or reality of that obligation: this obligation can still exist, in the terms of a specific social-temporal normative perspective (to which I presumably belong). (Though in this instance that social-temporal perspective probably wouldn’t do me much good, w/r/t getting someone to do anything; a bit of hard power would be necessary too.) Similarly, there is nothing intrinsically meta-theoretically wrong with me saying “torture is abhorrent – your action in torturing that person was abhorrent”, even if the torturer in question is in every respect willing to endorse the moral legitimacy of torture. I can still be right about the ethical monstrousness of torture – again, in terms of the normative perspective constituted by a specific community of discursive and ethical practice, to which I but (presumably) not the torturer belong.

And the principle of autonomy would, on this approach, be a normative principle that must be constructed and considered as binding on people at this downstream level of Brandom’s normative pragmatics – rather than at the ‘transcendental’ level at which Brandom is laying out the conditions of possibility of any rational discourse.

There’s as usual much more that could be said to develop this line of thinking… but I guess I’ll stop there for now. Provisional conclusion: the principle of autonomy cannot be derived as a transcendental condition of rational discourse (one is not bound by the principle of autonomy simply in virtue of being a subject). And, although I’m willing to be convinced, I’m not totally sure (mainly due to the difficulty of tracking voicing in these lectures) that Brandom thinks it can be. Whether he does or not, I certainly don’t think his theoretical apparatus gives him the right to compellingly and consistently make this claim.

Now back to Making It Explicit!

Promissory note: Brandom on autonomy

I’m still using the blog to take reading notes on Brandom; this post is essentially a promissory note for a specific critique and repurposing, both of which I’ll want to unfold in more detail further down the road, once I’m much more familiar with Brandom’s work. I’ve already articulated some of the reasoning behind this line of thought in conversation at ktismatics – this post takes a slightly different approach to the same set of issues.

Brandom’s system (which is exceptional in its scope, power and insight – Making It Explicit is one of the most impressive works of systematic philosophy I’ve ever read) is built upon very minimal foundations; the resources Brandom needs to unfold his apparatus are extraordinarily slight, given the reach of his analysis. One of the aspects of the system that seems, to me, however, to be serving a ‘grounding’ role, is Brandom’s version of the Kantian thesis of autonomy. Pete has discussed this in relation to what he calls ‘the primary bind’. Pete’s idea is that certain minimal normative commitments are a precondition of rational subjecthood; that participating in any discourse at all means committing oneself to these principles. One of those principles, Pete writes, is “the insight into normativity that Brandom finds in Kant, namely, the thesis of autonomy. This is the claim that any authority is binding on us only insofar as we accept that authority.”

I think Pete is right to see this claim as something like a transcendental condition of rationality in Brandom’s work. I also, for what it’s worth, agree with Pete that Brandom’s apparatus is sufficiently powerful to escape the obvious conceptual difficulties such a transcendental principle seems to pose (principally the consequence that, in Pete’s words, “we need only accept correction when we want to accept it, and if we don’t want to accept it then it isn’t really correction.”) Nevertheless, I think that Brandom is wrong to give this principle the status that he does; I think that doing so significantly weakens the power and reach of his system, and is, further, inconsistent with his system’s own best practice.

I don’t plan on trying to cash out these claims for some time yet (though it helps my thinking to articulate them, which is why this post exists) – but as a gesture towards a more complete analysis, let me add the following: Much of the work of Brandom’s system pivots around the distinction between deontic attitudes and deontic status. The former (on my interpretation of Brandom’s argument) produces the latter; the latter can be analysed in terms of the former (subject to some potentially confusing nuances around the nature of the intentional stance, which I will address in a future post). The two most basic deontic attitudes are those of undertaking and attributing a commitment (a commitment being a deontic status); of these two, attributing a commitment is fundamental. As Brandom puts it on page 196 of Making It Explicit:

The fundamental concept of the metalanguage employed in specifying the model of assertional practice is that of the deontic attitude attributing a commitment. For the deontic attitude of undertaking a commitment is definable in terms of attribution: undertaking a commitment is doing something that licenses or entitles others to attribute it. … The attitude of acknowledging a commitment is in effect that of attributing it to oneself.

This last sentence articulates very clearly and succinctly Brandom’s order of explanation. Attributing a commitment is fundamental, and attributing a commitment to oneself is a subspecies of attributing a commitment in general: in Brandom’s theory of normative social practice, it is the reflexive ability to relate to oneself in the same way in which one relates to another (for the practice of attributing a commitment is the same, no matter to whom the commitment is attributed) that enables the complex socially mediated relation to self of which sapience principally consists.

This leveling of the social playing field of deontic-status-attribution means, however, that Brandom’s insistence that a commitment can’t really be a commitment unless it is (in some very broad sense) accepted as such by the self, cannot be justified at the most basic explanatory level of his system. At this very abstract (‘transcendental’) level of the system, the claim is an arbitrary imposition – an intruder from a more downstream and concrete level of normative practice. The attribution of deontic status to self has (at this level of meta-theoretical abstraction) no features that would substantively differentiate it from the attribution of deontic status to others, such that the principle of autonomy (and the privileging of self-relation on which it depends) could be grounded in this difference.

I’ll expand on this thought, suggest other reasons for reaching a similar conclusion, and further analyse Brandom’s use of the principle of autonomy in future posts. This post as stands is just a promissory note.