Brandom as a reader of Kant: A revised account of key Brandomian themes

If anyone can stomach yet another tract on Brandom, I’ve been trying to come at the themes of autonomy and objectivity from a different angle. The results are somewhat lengthy, again, so I’ve put them below the fold.

This post has grown out of dissatisfaction with my earlier accounts and criticisms of Brandom. In them, I presented Brandom as someone who may be well-placed to answer the core problem of my thesis, which asks how we can reconcile our authority over ourselves with the notion that we are answerable to a world beyond ourselves. To that end, I gave a general account of his project, paying special attention to what he says about the source of normativity being autonomous action and the historical background to this position, as well as his claim that, “Our cognitive attitudes must ultimately answer to … attitude-transcendent facts.”[1] Then, as the beginnings of a critical response, I formulated an objection to the model of autonomy which Brandom adopts and suggested that he cannot do justice to a robust sense of normative constraint from objects since he gives experience a subsidiary role in his account.

There are a number of things lacking in the story as set out so far though. Firstly, I fail to do full justice to the systematic connection between many of the topics raised. For example, I do not adequately reflect how what Brandom says about autonomy is bound up with his inferentialism, how this relation is itself shaped by his critiques of regulism and regularism, and how all these topics do not merely cohere with but are integral to his social conception of objectivity. So too, when raising concerns over the roles given to autonomy and experience in Brandom’s system, I fail to demonstrate how these roles in fact stem from the same argumentative move and therefore that each must be understood in the light of the other. Secondly—and not coincidently—I do not make sufficient reference to the Kantian and Sellarsian roots of many of these topics as they feature in Brandom. So, for example, I do not place enough emphasis on the ‘proto-inferentialism’ that Brandom takes from Kant and the ‘representationalism’ that he rejects, nor on the pivotal role for material inference in avoiding regularism and regulism that comes out of his reading of Sellars.

Here then, I want to retell parts of the story told so far but in a different key and hopefully in a more accurate way. I think that the best way to do this is by placing Brandom’s account within a Kantian frame.[2] This is because it is in response to different strands in Kant’s work that we can best see the connection between the topics Brandom discusses as well as coming to understand his strategy as a whole. It is with reference to such a Kantian problematic that we will now begin.

I – Brandom and Kant

At the heart of Kant’s account in the first Critique is a strict separation between two sorts of representation: intuitions and concepts.[3] Intuitions belong to sensibility, which is a receptive faculty, whereas concepts belong to the understanding, which is a spontaneous faculty.[4] At least, intuitions are paired with receptivity and concepts with spontaneity in finite beings like ourselves. Thus, while one sort of representation is passively received, another is actively employed by us such that through deploying the latter in combination with the former we come to know objects.

Brandom endorses this connection between concepts and spontaneity, as well as an additional Kantian claim about the role of concepts. This claim is that concepts are only cognitively relevant in virtue of their contribution to judgement, that this is the only use of concepts that the understanding can make.[5] At this stage then, Kant and Brandom can agree that it is the judgement, conceived as an exercise of conceptual capacities in a free (because spontaneous and thus uncoerced) manner, that is the proper locus of concepts.

However, Kant goes further than Brandom. For him, in judgement, concepts are synthesised in such a way as to relate them to a further thing: an object as presented in intuition. This has implications for Kant’s understanding of the properties of concepts. Kant takes this to show that the content of concepts must be fixed through their potential relations to objects. Still, the form of concepts, understood as their role in determining the formal validity of judgements to which they contribute, is simply a matter of the relation between judgements rather than their relations to objects. While Brandom admires what he calls Kant’s ‘proto-inferentialist’ treatment of the formal features of conceptuality, he thinks that Kant mars his overall picture by supposing conceptual content is fixed in relation to intuitive objects. Why is it marred? Because to suppose that the non-conceptual deliverances of experience can provide a representational link between objects and ineliminably conceptual judgements is, thinks Brandom, to fall into the Myth of the Given.[6] Thus, for Brandom, we cannot appeal to objects to do the work of fixing conceptual content that Kant’s theory of judgement demands of them. For him, any relation between conceptual content and objects must proceed from an antecedent specification of this content, not mediated by its relation to experience.

This is a very compressed outline of Brandom’s relation to Kant. However, it contains many of the seeds of Brandom’s final position. By fleshing it out, we can come to understand the relation between the major elements in Brandom’s philosophical system. We will focus on three notions introduced here: freedom as grounded on the exercise of conceptual capacities; the relation between judgements as a correlate of features of concepts; and the rejection of experience as a rational constraint upon judgement. In pursuing the connection between these three ideas, we will also see the social inflection that Brandom gives to them and which ultimately binds them together.

II – Freedom

Brandom takes himself to be developing a fundamentally Kantian conception of freedom. This is a notion of freedom as autonomy, understood as a sort of self-binding. The idea is that freedom consists in the authority to undertake a type of commitment, and thereby acquire a rational responsibility to act in certain ways. The type of commitment in question is discursive—one arising from and pertaining to the use of concepts.[7] Brandom follows Kant in taking concepts to express norms, and so to use a concept is to commit oneself to the norm that it expresses. So, in the conceptual activity of forming a judgement or practical maxim, one acquires responsibilities. These responsibilities are self-imposed because, as we have seen, concepts belong to spontaneity, and are thus something that we have to actively bring to bear rather than conceptual matters being deliverances that we are simply lumbered with. What we are left with is a positive conception of freedom then, because it consists in a power to do something: to employ concepts so as to make a commitment to the norms they express.[8]

Brandom has two criticisms of the Kantian model of autonomy though. Firstly, he thinks that it overly intellectualises the process of coming to follow a norm. Secondly, he thinks that, without introducing a social dimension, it cannot maintain a distinction between taking oneself to be following a norm and actually following a norm. Examining these criticisms will allow us to see how Brandom’s invocation of material inference and an inferentialist model of conceptuality is integral to his conception of autonomous action. Having introduced these latter ideas, we can then go on to see how Brandom exploits them to reject experience as a normative constraint upon us, replacing it with a social conception of objectivity.

Brandom’s first criticism of Kantian autonomy is Sellarsian.[9] Kant thinks that what distinguishes us as autonomous agents from the natural order is the fact that we act according to our representations of laws rather than the mere natural necessity of laws.[10] On Brandom’s understanding of this claim, Kant is thought to be a regulist. In other words, the position ascribed to him is one in which following a norm is always a matter of applying an explicit rule in order to regulate your behaviour. But if this is so, then it is not clear what it is to follow an explicit rule to which one has bound oneself. The complaint is this:

Norms explicit as rules presuppose norms implicit in practices because a rule specifying how something is correctly done (how a word ought to be used, how a piano ought to be tuned) must be applied to particular circumstances, and applying a rule in particular circumstances is itself essentially something that can be done correctly or incorrectly.[11]

The idea here is that an explicit rule can always be interpreted in such a way as to allow or disallow the particular action assessed, where appeal to other explicit rules to settle the application of the original rule are themselves open to the same problem as to how to settle their correct application. To prevent a regress of rules, Brandom supposes that we must appeal to a different form of norm that does not stand in need of interpretation: namely, an implicit norm grasped in practical activity. This is what Brandom takes Wittgenstein to be appealing to when he says that “there is a way of grasping a rule which is not an interpretation, but which is exhibited in what we call ‘obeying the rule’ and ‘going against it’ in actual cases.”[12]

The converse danger is regularism. If we only appeal to features of our actual behaviour, like regularities in it, to explain what it is to follow a norm, then we risk finding ourselves with a descriptivist position that puts the very notion of correct norm-following under threat. This is because all behaviour displays many regularities, and as such there is no such thing as a performance being regular tout court. Specifying a particular regularity that ought to be conformed to simply postpones the problem. We are thus left without an explanation of normativity. This would be equally inhospitable to Brandom’s account of freedom because it would leave it mysterious how—through the exercise of conceptual capacities—we could commit ourselves and so make ourselves responsible in an act of self-binding.

The dilemma is this. Concepts express norms. Our use of concepts in judging and formulating practical maxims makes us responsible to these norms. Our becoming so responsible is what freedom consists in. So, to understand freedom we must understand what it is to be responsible to a norm, where this question is equivalent to asking how we can follow norms. If we appeal to explicit rules, we are faced with a regress. But if we appeal to behavioural regularities, we leave normativity unexplained.

The Brandomian-Sellarsian solution is to look to assessments of correct norm-following. This approach tries to steer a course between the excessive rationalism of regulism and the excessive naturalism of regularism, neither of which can explain the relation between norms and empirical practice. This way of understanding what it is to follow a norm, ties in neatly with the account of autonomy, which is our interest here. This is because Brandom thinks that the notion of self-binding depends upon separating whether we bind ourselves from what we are bound to. If we get to determine both, then he thinks that we are not really bound at all.[13] So, in keeping with the appeal to social assessment introduced by the critiques of regularism and regulism, Brandom takes autonomy to require a division of conceptual labour. We determine whether to undertake a commitment, which we do by using concepts through judging or willing. But others determine what we are committed to, or what is the same, they determine what it would is for us to correctly follow the norm in question. Autonomy is thus conditional upon sociality.

III – Inferentialism

We have seen that on Brandom’s account of autonomy, the individual only gets to determine whether they undertake a commitment but not what they are thereby committed to. This invites the question, how is what they are committed to determined? One answer would be to say that other people determine the content of the commitment. While this is correct, it does not serve as a full explanation. Giving a full explanation will enable us to show the connection between Brandom’s account of autonomy and his appeal to inference as an essential theoretical concept.

It would be equally correct to say that the content of the commitment that one undertakes by using a concept is determined by the content of that concept.[14] But how is the content of the concept determined? To grasp Brandom’s account of this, we should start by recalling his reaction to Kant. He praises Kant’s version of the context-principle, which makes the cognitive relevance of a concept conditional upon its role in judgement. So too, he approves of Kant’s focus on the relation between judgements as the key to understanding formal properties of the concepts those judgements feature in. However, he rejects the idea that the content of concepts is determined through their role in judgement insofar as this content is derived from their relation to the objects of intuition. Brandom follows Sellars, radicalising Kant’s handling of the logical form of concepts by generalising it to content. In place of “the familiar notion (Kantian in its origin, but present in various disguises in many contemporary systems) that the form of a concept is determined by ‘logical rules’, while the content is ‘derived from experience,'” Sellars claims that “there is nothing to a conceptual apparatus that isn’t determined by its rules.”[15]

For Brandom, the content of a concept is its inferential role.[16] What does this mean? The idea is that what a concept expresses is a function of the inferential relations between judgements in which it features. Brandom understands this claim in a distinctive way though.

Two different ways to read the claim would be logical correlatives of the regulist and regularist approaches respectively.[17] On the regulist reading, we would suppose that reasoning was enthymemic. So, the claim ‘Lightning is seen now, therefore thunder will be heard soon’ is implicitly construed as ‘If lightning is seen now, then thunder will be heard soon’ along with the suppressed premise. On the regularist reading, we would suppose that reasoning was habitual association. This model would suppose that what is happening in the case of lightning and thunder is that upon seeing lightning we just expect thunder, or rather upon judging ‘Lightning is seen now’ we simply form the judgment ‘Thunder will be heard soon’, as a result of a causal regularity. But again, Brandom wants to steer a course between the excessive rationalism and naturalism of these two approaches. So, he rejects interpretations of inferential roles as products of explicit principles, such as modus ponens, that determine allowable inferences between judgements. Equally, he would reject interpretations of them as habitual associations between judgements, which threatens to rob them of their normative ground—the idea that one is entitled to conclude ‘Thunder will be heard soon’, for example.

Brandom’s approach is to look to the notion of material inference to specify inferential role. Like the logical regulist, he insists on beginning from an appeal to a notion of appropriate inference between judgements, thus holding on to the normative dimension of reasoning. However, against the logical regulist, he does not think that this should be understood in terms of formal rules. Instead, an appeal is made to a nonlogical, content-based notion of correct inference.[18] We start with the notion of treating a case of reasoning as good or bad, whether or not we can attribute deductive validity to it. It is only after starting with the implicit practice of treating the inferential moves between judgements as correct or not that we then codify some of them as deductively valid and others as materially valid (i.e. not formally valid but appropriate just because of the conceptual content involved, like ‘Birmingham is south of Sheffield, so Sheffield is north of Birmingham’). The notion of inferential role is specified simply at the former level though, without being dependent on deductive validity; the more primitive, practical notion of correct inference is prior to it.

Since, for Brandom, the content of a concept is a function of the inferential roles of judgements in which it features, then conceptual content is determined by this primitive nonlogical notion of inference. But what settles what a good inference of this type is? That is, what determines whether a judgement is a good one in virtue of what it means? This is where sociality re-enters the account. For Brandom, we are to understand what a good inference is with reference to the practice of treating someone as committed and entitled to other judgements on the basis of judgements that they are antecedently committed and entitled to. As Redding puts the point:

Brandom then makes the move of conceiving of the semantic content of an utterance as constituted in terms of the changes of such commitments and entitlements that can be attributed to the speaker by an interlocutor. Language games need ‘score-keepers’ to keep track of the speaker-player’s deontic commitments and entitlements at any time, but of course for the most part there are no separate umpires of scorekeepers – we keep score on each other, holding each other to our particular commitments and entitlements as we go along.[19]

Now we are in a position to see that there is no tension between the two claims with which we began. We wanted to know, if autonomy consisted in deciding whether to undertake a commitment but not what we are thereby committed to, what determines what we are committed to. The Brandomian answers given were that what we are committed to is determined by other people and by the content of the concepts we use in committing ourselves. These answers are compatible because the content of concepts is determined by what people (in the guise of scorekeepers) treat as good inferences between judgements in which these concepts feature.[20]

The previous section claimed that, for Brandom, autonomy is conditional upon sociality. We should now have more of an idea what the role of sociality is in such a picture. The model is not one where I make a judgement and then others separately determine what this judgement commits and entitles me to do. Rather, since deploying a concept is essential to making a judgment, the very act of judging presupposes the social game of giving and asking for reasons that determines the content of the concepts that I use to bind myself. On this story, without the social-inferential articulation of concepts, secured by the practice of treating judgements as following from one another, there would be no discursive norms to work with in the first place.

IV – Objectivity

Brandom radicalises the nascent inferentialism he finds in Kant. In doing so, he replaces Kant’s appeal to experience as something that mediates between concepts and objects. For him, conceptual content cannot be determined by the non-conceptual content of experience. Nor can judgements be responsible to experience, which lacks the form of judgement, because “nothing nonjudgmental … could serve to justify perceptual judgments, rather than just to cause them.”[21] For Brandom, to think otherwise is to succumb to the Myth of the Given: the idea that raw, unconceptualised items are members of the rational order necessary to support normative relations in addition to causal ones.

We have seen how individuals can exercise authority over themselves to undertake responsibilities, which is what their freedom consists in. But, absent experience, where in this picture is there room to do justice to the authority held by how things stand in the world—what the facts are—to have a grip on us? Brandom thinks he can meet this challenge and explain how his account meets what he calls the ‘rational constraint constraint.’ This constraint is one that imposes an obligation to account for how “the world imposes not merely causal, but rational constraints on thinking.”[22]

Once again, Brandom’s solution to this problem is to appeal to features of the social practice of treating each other as committed and entitled to certain claims. Unlike someone like Rorty, who rejects the idea that we are answerable to the world rather than to each other, Brandom thinks the idea of responsibility to how things are in the world is intelligible.[23] Nevertheless, it is through our relation to our fellows that we discharge such a responsibility: “the responsibility of noninferential reports to the corresponding facts is essentially a social phenomenon”.[24] Keeping this social dimension in mind, Brandom’s response to these matters can be posed in terms of two claims. Firstly, that objectivity is just what Brandom calls “a structural aspect of the social-perspectival form of conceptual contents.” [25] Secondly, that the rational constraint constraint can be met by showing how perceptual judgments can be justified and criticised by scorekeepers.

The first story is complicated and I shall only hint at the details of it here. The basic idea is that a scorekeeper can attribute both de dicto and de re attitudes to the scoree. De dicto ascriptions, attached to that-clauses, track what by the scorekeeper’s lights the scoree takes themselves to be committed and entitled to. De re ascriptions use the scorekeeper’s set of commitments and entitlements to specify objects that the commitments of the scoree are answerable to by the scorekeeper’s lights. The de dicto ascriptions do not specify such objects, since they only specify the proposition that the scorekeeper thinks that the scoree takes themselves to be committed to. This introduces the permanent possibility from the scorekeeper’s perspective of a distinction between what the scorekeeper takes the scoree to take themselves to be committed to and what objects the scorekeeper takes to settle whether the scoree is correct to do so. This is what from the scorekeeping perspective someone acknowledges and what they are committed to, respectively.

Brandom resists trying to specify a privileged perspective which by itself settles whether someone is correct (whether their acknowledgements really do match their commitments). Instead, it is a commonality to all perspectives that he wants to highlight: “What is shared by all discursive perspectives is that there is a difference between what is objectively correct in the way of concept application and what is merely taken to be so, not what it is—the structure, not the content.”[26]

This distinction between what is and what is merely taken to be correct is written into the very process of social scorekeeping. Through making it in practice, we hold each other responsible to what we take to be the facts. The rational constraint constraint is met in this way. On Brandom’s story it is enough that we be able to do the following:

If a suitable story is told about how [non-inferentially elicited perceptual judgments] are rationally criticizable by those who key their correctness to their correspondence to the facts reported, and about their entitlement to the reliability of the noninferential process that elicits them, then rational constraint by how things actually are is secured.[27]

For Brandom, this can be done from within the game of giving and asking for reasons. As such, we do not have to appeal to the deliverances of experience to supply a normative constraint upon our activities, since this is part and parcel of our ordinary activities of discussing, assessing, criticising and challenging beliefs. Brandom still thinks that “perceptual experience is the medium through which our thought becomes answerable to how things are.”[28] But this is only through its causal role in ‘wringing’ perceptual judgments from us—judgments that derive any justification from our treating them as warranted, not from the objects they put us in touch with.

[1] MIE, p.137.[2] The benefits of approaching Brandom via Kant were suggested to me through the readings given by Macbeth, D. ‘An Antinomy of Empirical Judgement: Brandom and McDowell’ [ref?] and Redding, P. (2007) Analytic Philosophy and the Return of Hegelian Thought, Cambridge: Cambridge University Press, ch.2.[3] Although, of course, much post-Kantian hay has been made from those sections of the second edition Transcendental Deduction where Kant seems to hint at ways in which (pace the ‘official’ story) intuitions and concepts might be brought back together.[4] CPR, A50=B74.[5] Such a claim is a precursor to Frege and Wittgenstein’s context-principles that state that names only have meaning in the context of sentences or propositions.[6] ‘No Experience Necessary’, p.4 [ref?][7] As Macbeth (loc cit. p.2) says, there is sense in which “Brandom takes exercises of conceptual capacities just to be exercises of freedom, of the spontaneity of judgment.”

[8]Spontaneity, in Kant’s usage, is the capacity to deploy concepts. Deploying concepts is making judgments and endorsing practical maxims. Doing that, we have seen, is committing oneself, undertaking a distinctive sort of discursive responsibility. The positive freedom exhibited by exercises of our spontaneity is just this normative ability: the ability to commit ourselves, to become responsible. It can be thought of as a kind of authority: the authority to bind oneself by conceptual norms. That it is the authority to bind oneself means that it involves a correlative kind of responsibility. That the norms in question are conceptual norms means that the responsibility involved in exercising that sort of authority is a rational responsibility. […] Kantian positive freedom is the rational capacity to adopt normative statuses.” ‘Autonomy, Community and Freedom’, Woodbridge Lecture, p.8-9.

[9] See Sellars’ ‘Language, Rules and Behaviour’ [ref?], whose contents do much to structure chapter 1 of Making It Explicit.

[10] Kant, Grundlegung, Ak.4.412.

[11] MIE, p.20.

[12] PI, §201 quoted in MIE, p.21.

[13] This seems wrong, at least when construed as Brandom does. We would only fail to be robustly bound if we could rescind our commitment at any time, not simply because we could choose both whether to enter a commitment and what we are bound to.

[14] This is simpler to grasp at the level of judgement as a whole. For example, what I am committed to by asserting that ‘The cat is on the mat’ just depends upon what this sentence means. The concepts employed in asserting it are what specify the responsibilities are thereby undertaken. Brandom seeks to give a non-traditional account of this.

[15] ‘Inference and Meaning’, Mind [ref?] p.337. Another version of this claim is put forward earlier in the article: “the meaning of a term lies in the materially and formally valid inferences it makes possible.” p.317.

[16] MIE, p.105.

[17] On the other hand, perhaps these approaches are just regulism and regularism simply seen from a different vantage point.

[18] MIE, p.101.

[19] Redding, loc cit. p.78.

[20] It might seem that there is a slide here from conditions of intelligibility of norms to the conditions of the constitution of norms. That is, we are now talking about what determines the content of a norm rather than how we must understand the content of a norm, which we might think are distinct issues. At some level, Brandom seems happy to license such a move though: “Norms are in an important sense in the eye of the beholder, so that one cannot address the question of what implicit norms are, independently of the question of what it is to acknowledge them in practice.” MIE, p.72. What he want to resist, it seems, is not scorekeepers determining what we are committed to, but their activity being the ultimate arbiters of whether in any particular case we have discharged that commitment.

[21] ‘No Experience Necessary’ p.4.

[22] ‘Perception and Rational Constraint’, p.369.

[23] See Rorty, R. ‘The Very Idea of Human Answerability to the World: John McDowell’s Version of Empiricism’ in Truth and Progress [ref?]

[24] ‘Replies’ (1997) Philosophy and Phenomenological Research, n.1, p.191.

[25] MIE, p.597.

[26] MIE, p.600.

[27] ‘Perception and Rational Constraint’, p.372.

[28] ibid. p.369.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s