The Normative/Agentive Correspondence

1 Introduction

The basic aim of Robert Brandom’s Making It Explicit (1997), along with most of his subsequent work, is to give an account of rational mindedness. According to Brandom, the way to answer the question of what it is to think that, judge that, or know that something is the case, is to ask what it is to make a corresponding assertion, and the way to answer the question of what it is to make an assertion is by thinking of the making of an assertion as the making a “move” in “the game of giving and asking for reasons”. These moves, on Brandom’s account, can be understood entirely in terms of the normative relations that they bear to other moves, as these normative relations are assessed from players of the game who keep score on the other players who make them. The result is a systematically third-personal account of the mind: an account articulated from the perspective of an attributor of commitments and entitlements, a scorekeeper, who takes others to think, judge, and know. This contrasts, of course, with a first-personal account of the mind: an account articulated from the from the perspective of the agent of mental acts, who herself thinks, judges, and knows, and, in taking herself to do so, takes herself to be entitled to undertake commitments. If one thinks that the mind is something that is essentially comprehended first-personally, in and through actually engaging in mental activity—thinking through, in thinking, judging, and knowing, what it is to think, judge, and know—then one will be inclined to think that Brandom’s “account of the mind,” in virtue of its systematically third-personal methodology, essentially leaves out the very thing for which it aims to account: the mind. This, I think, is what the class of philosophers whom Brandom regards as his deepest critics—most notably John McDowell (1994, 2010), but also Sebastian Rödl (2010), James Conant (2020), Irad Kimhi (2018), Mathew Boyle (2005), and several others—think, though they wouldn’t all put it in quite these terms.^[1] My aim here is to provide the basic materials needed to resolve this dispute, articulating an account of the mind that incorporates both the third-personal perspective on judgment, articulated by Brandom, and first-personal perspective in judgment, articulated by his critics, one that both Brandom and his critics ought to be happy endorsing.^[2]

My point of departure is Brandom’s articulation, in Between Saying and Doing (2008) and A Spirit of Trust (2019), of a correspondence between alethic modal relations of entailment and incompatibility, principally articulated in terms of the alethic modal statuses of possibility and necessity, and deontic normative relations of entailment and incompatibility, which Brandom articulates in terms of his core normative statuses of entitlement and commitment. On Brandom’s account, these two sorts of entailment and incompatibility relations respectively articulate the objective and subjective poles of the intentional nexus, the nexus that obtains between the objective world about which minded agents think and the minded agents who, in their subjective acts of thinking, think about the objective world. I claim that Brandom’s failure to incorporate the first-person perspective in his account of the mind is owed to his failure to recognize that the intentional nexus is to be understood in terms of 3-way, rather than 2-way, modal correspondence. More specifically, the subjective pole of the intentional nexus has two fundamentally distinct modes of articulation—a normative one and an agentive one. Brandom, in his systematically third-personal methodology, only has the former in his purview. His critics, in their principally first-personal methodology, principally have the latter in theirs. To spell out the full account of intentionality of which Brandom’s account is a part, I articulate this missing link: the correspondence between, on the one hand, Brandom’s core normative statuses of entitlement and commitment, which index what an agent can do, in the sense of their being entitled to do something, and what an agent must do, in the sense of their being committed to doing something, and, on the other hand, the agentive modal statuses of ability and compulsion, which index what an agent can do, in the sense of their being able to do something, and what an agent must do, in the sense of their being compelled to do something (Mandelkern, Boylan, and Schultheis, 2017). These modals have application not only in action (as has been the focus in the literature) but also in perception and inference, and, in both of these cases, there is a direct mapping between the normative statuses that one bears towards various claims, on Brandom’s account of discursive practice, and the agentive modal statuses that one bears towards various judgments, on the agentive account of the mind that is the natural complement of Brandom’s account of discursive practice.

To illustrate the elucidatory and explanatory power of the account of the mind that emerges once the normative/agentive correspondence is brought to light, I consider specifically the case of perception, showing how, by exploiting the normative/agentive correspondence, we can resolve the dispute between Brandom and McDowell over the necessity of appealing to perceptual experience in accounting for perceptual knowledge. The correspondence enables us to articulate an agentive account of perceptual experience, in the same vein as but superior to McDowell’s (1994) conceptualist account of perceptual experience, which is the natural compliment of Brandom’s experience-free account of perceptual knowledge. On this account, the content of perceptual experience is articulated in terms of two dynamically evolving sets of judgments: those that having that experience enables one to make and those that having that experience rationally compels one to make. These two sets of judgments, those in potentiality and those in actuality through rational necessity, are articulated, in any given case, through first-personally thinking through what it is to be having the perceptual experience that one is having. However, they directly correspond to the two sets of claims, articulated from a third-personal scorekeeping perspective that another would take on one, according Brandom’s account of perceptual knowledge: those to which another would take one to be entitled and those to which another would take one to be committed. After highlighting the normative/agentive correspondence in this particular case, I’ll articulate a general account of it that promises to enable us to bridge the rift in Pittsburgh Philosophy separating Brandom from those whom he takes to be his deepest critics.

2 Brandom’s Normative/Alethic Correspondence

Brandom’s Making It Explicit is, in its broadest possible characterization, an account of content, both mental and linguistic—the content of acts of thought (fundamentally, beliefs or judgments) and the contents of acts of speech (fundamentally, assertions). Brandom’s explanatory strategy is to give an account of the contents of beliefs and judgments as derived from the contents of acts of assertion, and to give an account of the contents of acts of assertion by thinking of such an act as the making of a “move” in what he calls “the game of giving and asking for reasons,” (1994, xviii). We understand the significance of such a move, on his account, by thinking about how it changes the “score” of the game, the set of normative statuses that have been assigned to the players of the game by those very players (1994, 190; see also 2000, 165).^[3] The two basic normative statuses that Brandom employs in articulating this account of content are entitlement and commitment. Brandom characterizes these statuses as deontic, as relating to duty or obligation, introducing them by saying that they “correspond to the traditional deontic primitives of obligation and permission,” (1994, 160). Very roughly, to be entitled to a move is to be permitted to make it, and to be committed to a move is to be obligated to make that move if (appropriately) prompted and to be obligated to respond to (appropriate) challenges to the making of that move (Brandom 1994, 173).^[4] It is this deontic characterization of the statuses of entitlement and commitment in virtue of which they can be said to be distinctively normative statuses; they are used to specify what moves in the game of giving and asking for reasons can, can’t, or must be made in the sense of those moves being permissible, impermissible, or obligatory.

To characterize the statuses of entitlement and commitment as normative in this way is to contrast them with a different class of modal statuses, alethic modal statuses, statuses concerning what can, can’t, or must be the case in the sense of something’s being possible, impossible, or necessary. The few places in Making It Explicit where Brandom brings up the statuses alethic modality, he does so only to contrast them with his core normative statuses. However, in more recent work, Brandom (2008, 2019) has worked to articulate not just a contrast between his class of normative statuses and the class of alethic modal statuses, but also an important connection between these two classes of statuses. He does this first by articulating two distinct flavors of the relations of incompatibility and entailment, an alethic flavor and a normative flavor, and then going on to say how these two flavors of entailment and incompatibility relations relate to each other.

Consider first how states of affairs that might or might not obtain in the world and properties that things in the world might or might not instantiate stand in alethic modal relations of entailment and incompatibility. For instance, something’s being red entails its being colored in the sense that if something is red, it is necessary that it be colored; in this sense, something that is red must be colored. Likewise, assuming for simplicity that all objects in the domain of discourse are essentially monochromatic, something’s being red is incompatible with its being green in the sense that, if something is red, it is impossible for it to be green; in this sense, something that is red can’t be green. The “must” and “can’t” in the these sentences are the “must,” and “can’t” of alethic modality. They express what is necessary or impossible. Entailment and incompatibility relations of the variety just expressed have this particular flavor: the flavor of alethic modality.

Now consider how acts of applying concepts to objects stand in normative relations of entailment and incompatibility. Following Kant, Brandom thinks of concepts “in terms of the rules that make them explicit, rules that specify how the concepts are properly or correctly applied and otherwise employed,” (1994, 10). To apply a concept, on this conception, is to bind oneself by a certain class of rules, the class of rules that articulate what one can, must, and cannot do insofar as one is to be counted as properly applying that concept. These rules, for Brandom, are to be understood, in the first instance, as linguistic rules, rules for the correct use of linguistic expressions. To articulate those rules for a given linguistic expression is to make explicit the concept that one applies to something when one applies that expression to it. Consider how we might do this for the expression “red”, To make the concept red explicit just is to articulate the rules for the correct use of the expression “red,” for this concept just is the concept that one applies to something when one says of it that it is “red.”^[5] So, articulating these rules, we might say, for instance, that someone who applies the concept red to something is committed to applying the concept colored to it; if one applies the former to something, one must at least be prepared to apply the latter to it. Likewise, someone who applies the concept red to something is precluded from being entitled to apply the concept green to it; if one applies the former to something, one can’t apply the latter to it. Here, the “must” and “can’t” have a distinctively normative flavor, expressing what one is committed or precluded from being entitled to do.

Brandom conceives of the alethic and normative flavors of the relations of entailment and incompatibility as corresponding to two poles of the intentional nexus (2008, 181; 2019, 80–85). The intentional nexus is the relation that obtains when a subject represents things in the world as being certain ways. The objective pole of this nexus is the structure consisting in the things that are represented, the properties that they are represented as having, and the relations that they are represented as standing in. This pole, Brandom claims, is what is articulated with the use of alethic modal vocabulary. When we say, for instance, that if something is red, then it is not possible for it to be green, we are articulating (in part) what it is for something to have the objective property of being red. So we are articulating an alethic modal relation that obtains between properties, things that belong to the objective pole of the intentional nexus. On the other hand, the subjective pole of the intentional nexus is the pole that concerns the activity of subjects representing the things in the world as being certain ways. So, to concern ourselves with this pole is to concern ourselves with our subjective acts of representing, not the objective world that is represented. This pole, Brandom claims, is principally articulated with the use of normative vocabulary. When we say, for instance, that if someone applies the concept red to something, then they are precluded from being entitled to apply the concept green to it, we are (in part) articulating what it is to perform an act in which someone represents something as being red. Brandom’s key claim concerning the relation of these two poles is that what it is take oneself to be representing an object as instantiating properties and relations that stand in objective entailment and incompatibility relations just is to take oneself to be bound, in one’s subjective applications of concepts to that object, by corresponding normative entailment and incompatibility relations.

3 Introducing the Vocabulary of Agentive Modality

What I now want to do is introduce a bit of new modal vocabulary, never explicitly considered by Brandom, that I hope will enable us to articulate the full account of intentionality of which Brandom’s account is a part: the vocabulary of agentive modality. Agentive modals are a class of expressions that index what an agent “can,” “cannot,” and “must” do in the sense of their being able, unable, and compelled to do certain things. This class of modal expressions has only very recently been brought to philosophical attention as a unified class since it has been pointed out that, like other classes of modals that are of interest to philosophers and logicians, the agentive modals of ability and compulsion are logical duals (Mandelkern, Schultheis, and Boylan, 2017).^[6] For some modal operator ◇ , its dual, □ , can be defined as ¬ ◇ ¬ , where ¬ is a formal negation operator.^[7] In the logic of alethic modality, p is necessarily the case just in case it is not possible for p not to be the case. In the logic of deontic modality, someone is obligated to φ just in case they are not permitted not to φ. In the logic of agentive modality, we can conceive of the statuses as ability and compulsion as logical duals in just this way; an agent is compelled to φ just in case they are not able not to φ. In English, a natural way to express this status such that it preserves its double negativity is to say of someone that they “cannot not” φ or, perhaps more naturally, that they “cannot but” φ.^[8] According to the thesis that ability and compulsion are logical duals, to say such things is to express the same status that one is expressing when one says of someone that they “have to” φ or “must” φ, where “have to” and “must” here express the agentive modal status of compulsion.

The first thing that I’d like to point out with respect to these agentive modal statuses, as they relate to Brandom’s project, is that, just as there are alethic entailment and incompatibility relations and normative entailment and incompatibility relations, there are agentive entailment and incompatibility relations. An act φ agentively entails an act ψ just in case φ-ing compels one to ψ in the sense that, if one φs, one cannot help but ψ. For instance, if I say “If I eat one Pringle, I just have to eat another one,” I express two acts standing in a relation of agentive entailment. This is a consequence relation that has neither an alethic flavor nor a normative flavor, but, rather, a distinctively agentive flavor. Certainly, it’s possible for me to eat one Pringle and then not eat another, and, if I eat one Pringle, I’m certainly not obligated to eat another one. Rather, what I’m saying here is that eating one Pringle compels me to eat another; if I eat one, preventing myself from eating another is not within my power.^[9] This is a distinctively agentive entailment relation that I’m expressing here. There is, likewise, an agentive incompatibility relation. An act φ is agentively incompatible with an act ψ just in case φ-ing precludes one from being able to ψ.^[10] Patting my head is agentively incompatible with rubbing my tummy in the sense that, insofar as I’m doing the first thing, I’m precluded from being able to do the second. Once again, this is a distinctively agentive modal relation.

Now, in illustrating the flavor of agentive modality here, I’ve been using examples of the sort that are common in the literature on them, examples that concern actions like eating a Pringle or rubbing one’s tummy. However, this class of modals not only has application in speaking of “actions” properly so called, but also in speaking of activities like inference, perception, and other members of the class of activities the main aim of which is not to act in the world, but to understand it—in Kantian terminology, activities that principally belong to the domain of theoretical rather than practical reason.^[11] These are activities that issue in acts of belief or judgment. To “believe” or “judge,” at the level of abstraction at which I am using the term, is to take something to be the case. Without making any distinction here, I’ll simply use the term “judgment” to speak of an act or activity of taking something to be the case.^[12] As subjects who represent an objective world, we bear, at any given point, various agentive modal statuses to various judgments. There are some judgments we are able to make (perhaps if we do some inferential work or look around a bit), some judgments that we are compelled to make (perhaps given the other things we judge or what’s right in front of our eyes), and some judgments that we are precluded from being able to make (once again, perhaps given the other things we judge or what’s right in front of our eyes).

We can point out now that the agentive modal entailment and incompatibility relations that obtain between judgments largely correspond to the alethic and normative entailment and incompatibility relations just considered. Consider again the example involving something’s being red. Judging that something is red, compels me to judge that it is colored; the first act agentively entails the second. Likewise, judging that something is red, precludes me from being able to judge that it is green; these two acts are agentively incompatible. To bring out this correspondence a bit more, it is worth considering not only material inferential relations like these ones, but also formal deductive relations.

First, consider deductive entailment. One can alethically characterize the entailment relation that obtains between a set of premises and a conclusion that is deductively entailed by them by saying that, if all of the premises are true, the conclusion must also be true. One can normatively characterize such an entailment relation by saying that, if one is committed to all of the premises, one is thereby committed to the conclusion.^[13] One can also, however, characterize a deductive entailment relation between a set of premises and a conclusion agentively, by saying that someone who accepts the premises is, in a certain sense, compelled to accept the conclusion. For a clear expression of this of the sort of compulsion, consider Lewis Carroll’s (1895) classic article “What Achilles Said to the Tortoise”, Towards the end of the article, the Tortoise asks Achilles what happens if he accepts A and B and C and D, a set of premises that deductively entail a conclusion Z, but “refuses” to accept Z. Achilles replies to the Tortoise as follows:

Then Logic would force you to do it! […] Logic would tell you ‘You can’t help yourself. Now that you’ve accepted A and B and C and D, you must accept Z!’ So you’ve no choice, you see, (280).

Achilles’ “must” here is the “must” of agentive modality. It is a compulsion modal. When Achilles says to the Tortoise “you must accept Z,” he aims to be expressing the fact that the Tortoise is compelled to accept Z, and he makes this clear by glossing what he says by saying “you can’t help yourself”. Nevermind that the Tortoise pretends, in order to make a philosophical point, that he isn’t so compelled; the relevant point here is that someone who grasps the premises of a clearly valid deductive argument is generally, as a matter of fact, compelled to accept the conclusion. This, on the proposal here, is one mode of articulation of the logical nexus that obtains between the premises and the conclusion of a deductively valid argument of which the normative mode, articulated in terms of a relation of commitive consequence, is another.^[14]

Consider now formal incompatibility, or contradictoriness. Alethically, if p is case, it impossible for not-p to also be case. Normatively, if one is committed to p, then one is precluded from being entitled to not-p. Agentively, if one judges p, one is precluded from being able to judge not-p. One cannot, in a distinctively agentive sense, consciously perform both acts of judgment at one time. This observation, which traces back at least to Aristotle, is the starting point of Irad Kimhi’s (2018) recent book Thinking and Being. Kimhi articulates the capacity for thinking as a unity of a manifold of capacities for thinking determinate thoughts. He calls each element of the manifold, each capacity through which one can think some determinate thought, a “propositional capacity”. The core metaphysical thesis of Kimhi’s work is that each of these propositional capacities is what Aristotle (Met, Θ -2) calls a “two-way capacity.” which can be in act positively, in thinking that something is the case, or it can be in act negatively, in thinking that something is not the case, but cannot be in act in both ways at once.^[15] Kimhi claims, on the basis of this account, that the psychological principle of non-contradiction, that one cannot think a thought of the form (p and not-p), is no less of a logical principle than the ontological principle of non-contradiction, that there cannot be a state of affairs of the form ( p   a n d   n o t ‐ p ) . Readers have struggled to understand exactly how this claim is supposed to follow from Kimhi’s account.^[16] I suggest that if we hear Kimhi’s “cannot” as fundamentally an expression of agentive modality rather than alethic modality, far from being mysterious, the psychological principle of non-contradiction follows straightforwardly from his account of an act of thinking as the actualization of a two-way propositional capacity. Minimally, if one actualizes a propositional capacity positively, then one cannot, in that very same same act of thinking, actualize that very same capacity negatively. This principle, expressed with the “cannot” of agentive modality, is an expression of the modal core of the concept of a two-way capacity for thinking. What it is for a capacity to be such that it can be in act in either of two opposite ways at once but not both ways at once just is for one’s continuing to actualize it in one way to preclude one from being able to actualize it in the other way. Now, on Brandom’s account of negation, not-p is the weakest claim such that, commitment to it precludes one from being entitled to p, and p is the inferentially weakest claim such that commitment to it precludes one from being entitled to not-p. So, what it is to make the claim not-p is to make a claim such that commitment to it precludes one from being entitled to the opposite claim, p. Thus, once again, on the proposal here, Kimhi’s agentive articulation of the logical nexus that obtains between a proposition and its negation is one mode of articulation of this nexus, of which Brandom’s normative articulation, in terms of commitment and preclusion of entitlement, is another.

So, not only is there a correspondence between the alethic entailment and incompatibility relations that obtain between states of affairs and the normative entailment and incompatibility relations that obtain between claims, but also between these two classes of entailment and incompatibility relations and the agentive entailment and incompatibility relations that obtain between judgments.^[17] The main claim I want to make here is that when we understand the modal correspondence as a 3-way rather than 2-way correspondence, we bring into view another dimension along which the intentional nexus can be articulated. Brandom is right that the vocabulary of alethic modality is the fundamental modal vocabulary to be employed to articulate the objective pole of the intentional nexus, but he is wrong to think that the only flavor of modal vocabulary required to adequately articulate the subjective pole of the intentional nexus is the normative flavor. On the contrary, both normative and agentive modal vocabulary have a fundamental role to play in articulating the subjective pole of the intentional nexus. The distinction between the two aspects of the subjective pole that are articulated by these respective modal vocabularies corresponds to Brandom’s perspectival distinction between attributing a commitment to someone else and undertaking a commitment oneself, between the third-personal perspective on judgment and the first-personal perspective in judgment. To put things a bit more provocatively, the three flavors of modality—agentive, normative, and alethic—respectively articulate self, other, and world.

To illustrate the work that the inclusion of the agentive modal statuses of ability and compulsion are able in articulating the full account of the mind of which Brandom’s account is a component, I’ll show how the statuses of agentive modality can be brought to bear on an aspect of Brandom’s theory that has been most heavily criticized for leaving out the first-personal perspective on judgment: his account of perceptual knowledge. First, however, let me bring out why Brandom has the problem with the first-person that he does.

4 Brandom’s Problem with the First-person

One fundamental feature of the account of content that we get in Making It Explicit is that it is, in the first instance, from the perspective the attributer of commitments and entitlements to claims, rather than from the perspective of the undertaker of commitments to claim. The notion of undertaking a commitment, on Brandom’s story, is treated as conceptually derivative with respect to the notion of attributing a commitment. Brandom tells us,

[F]or someone to undertake a commitment, according to this story, is to do something that makes it appropriate to attribute the commitment to that individual. […] It is by reference to the attitudes of others toward the deontic status (attributing a commitment) that the attitude of the one whose status is in question (acknowledging or undertaking a commitment) is to be understood, (1994, 161).

Though this feature of Brandom’s account is what gives it much of its explanatory power, it has left many readers of Brandom with the sense that something is essentially left out of Brandom’s account of content, namely, what might be called “the first-personal perspective” on thought or judgment, the perspective “from within” the activity of thinking or judging. The sense that Brandom’s theory leaves out something of this sort is, I believe, responsible for one of the major rifts in the Pittsburgh School of philosophy.^[18]

To get this rift into view, compare how Brandom characterizes his philosophical methodology with how Irad Kimhi characterizes the tradition of “philosophical logic,” a tradition which he takes to run through Aristotle, Kant, and Wittgenstein and to which he takes his own work to belong:

Philosophical logic […] is a first-personal engagement from within the activity of thinking, one which allows the articulation and comprehension of thinking to emerge from out of itself, (2018, 2).

This way of thinking about philosophical logic (or philosophy more generally) is characterized at greater length by James Conant (2020) who characterizes Kant’s philosophical methodology as one of comprehending what it is for the capacity for thinking and judging to be in act through self-consciously actualizing the capacity. Understanding the form of our thought, Conant writes, “is a matter of understanding the character of what we are doing, in thinking and judging, in and through the very act of doing it,” (774). Brandom, by contrast, in virtue of thinking of the act of judgment as an act of undertaking a commitment and thinking of what it is to undertake a commitment in terms of what it is to attribute a commitment, seems to systematically leave this perspective on thinking and judging out of the story. On Brandom’s account, we do not understand what it is to judge by thinking through what it is that we’re doing in judging. Rather, we understand what it is to judge by thinking through what it is that we’re doing in taking someone else to judge, attributing a commitment to them.

With his critics on the other side of the rift, I maintain that Brandom’s systematically third-personal methodology renders his framework incapable of delivering an account of the first-person perspective in inference, judgment, and perception, the activities that are absolutely central to his philosophical project.^[19] Accordingly, insofar as these activities are essentially known self-consciously—first-personally comprehended through doing them and thinking through what it is to be doing what one is doing—there is a sense in which Brandom’s framework is incapable of delivering an account of these notions at all. What Brandom’s framework is capable of delivering are accounts of the third-personal correlates of those notions: scorekeeping, the attribution of commitment, and the attribution of non-inferential entitlement. So, any account that he might give from his framework, as it stands, is at best a partial account, systematically leaving out an essential aspect of the phenomena for which he aims to be accounting. The main claim of this paper is that, once we supplement the Brandomian framework with the vocabulary of agentive modality and specify its place in our general account of discursivity, the upgraded framework is able to deliver an account, in the sense of an explanation, of the perspective from “within” the capacity for thinking, albeit one from “outside”, I will work through one case in particular in order to make this claim, exploiting the normative/agentive correspondence to make progress on settling a dispute with Brandom and one of his critics, entering into Brandom’s dispute with John McDowell on perceptual knowledge and showing how it can be resolved through an appreciation of the normative/agentive correspondence, as it applies in this case. It should be emphasized from the outset, however, that I take this to be a single instance of a general strategy that can be deployed to bridge the gap between Brandom and his critics.

5 A Person-sized Hole in Brandom’s Account of Perception

Nowhere does the third-personal nature of Brandom’s philosophical methodology come out more strongly than in his account of perceptual knowledge. Accordingly, that is the aspect of Brandom’s theory that has perhaps been most criticized by those who worry about Brandom leaving out the first-personal perspective on judgment.

Brandom’s account of perceptual knowledge is, properly speaking, an account of the attribution of perceptual knowledge. The account, which is articulated from “our” perspective as the attributors of perceptual knowledge, goes like this. We know that members of our discursive practice are reliably disposed to respond to certain features of our shared environment with the inclination to affirm certain claims, claims whose significance in our discursive practice they grasp.^[20] We attribute entitlement to them, when we recognize them to be in the relevant circumstances for response, to these claims that they are inclined to affirm in these circumstance. For instance, we know that a member of our discursive practice has a disposition such that, in response to red things, she reliably acquires the inclination to say “That’s red,” a claim whose significance in our discursive practice she grasps, as, for instance, committing her to “That’s colored,” precluding her from being entitled to “That’s green,” and so on. Accordingly, when we recognize her to be in the relevant circumstance for response (i.e., looking at something red), we’ll take her to be entitled to the claim “That’s red”. Additionally, though this is not an element of Brandom’s account that he explicitly brings out, when someone’s clearly in the relevant circumstances for response (i.e., looking squarely at something clearly red), we’ll attribute to them commitments to claims. Only if we attributed commitments in these circumstances could we, on Brandom’s account, attribute perceptual knowledge at all. According to Brandom’s official definition to attribute knowledge to someone is to attribute to them commitment to a claim, to attribute to them entitlement to that claim, and to undertake a commitment to that claim oneself (1994, 202). To attribute perceptual knowledge to someone, then, is to take them to be committed and entitled to a claim in virtue of exercising a reliable differential responsive disposition, and to undertake a commitment to that claim oneself.

The most striking feature of this account of perceptual knowledge is that there is no mention at all of any notion of perceptual experience. Brandom claims that the only role that a notion of experience might be able to play in the account, if it is to play any role at all, is a merely causal role, functioning to help explain how it is that members of a discursive practice are reliably differentially responsive to features of their environment that they are (2010, 323). Perceptual experience is not to play any essential epistemological or semantic role. So we are not to conceive of someone as being entitled to make observational claims about features of the world in virtue of those features being perceptually manifest to them. Critics like John McDowell (2010) have claimed that this is a serious problem with the account. Only if our entitlement to make claims is owed to our having perceptual experiences in which features of the world are made perceptually manifest to us, McDowell thinks, can we speak of our entitlement to make claims as distinctively perceptual entitlement, and so only in such a case could we speak of ourselves as having distinctively perceptual knowledge. Only if we can, for instance, see that things are certain ways, where seeing that things are certain ways essentially involves things being manifestly certain ways to one, can we know, perceptually, that things are certain ways. As such, Brandom’s account of “perceptual knowledge” is no account of perceptual knowledge at all.

In pursuit of his case, McDowell considers the case of a chicken-sexer, who knows himself to be reliably differentially responsive to the sex of a chick but the features to which he is reliably differentially responsive are not manifest in his experience.^[21] So, when confronted by some chicks, he feels inclined to say “male” and, when confronted by others, he feels inclined to say “female,” and he knows that, as a matter of fact, his inclination to say “male” is reliably correlated with a chick’s being male and his inclination to say “female” is reliably correlated with a chick’s being female. Furthermore, he knows what the discursive significance of saying “male” and “female” is—he understands the saying of “male” or “female” as the application of the concept male or female, and he grasps the norms that he binds himself by in applying either of these concepts. Given Brandom’s account, this chicken-sexer should qualify as having perceptual knowledge. However, something crucial seems to be lacking. Though he knows he is differentially responsive to a chick’s having either the property of being male or being female, and he might even know what detectable feature of the chick that is reliably correlated with its having one of these two properties to which he is differentially responsive, this feature is not first-personally manifest to him. McDowell puts his epistemic situation as follows: “though he can account for his inclinations in that external way, in another way the inclinations are unaccountable to him; he simply finds himself with them,” (2010, 140). The basic issue with Brandom’s account is that the only way that one can account for perceptual judgments is “that external way”—through a reliability inference of the sort that one would make concerning another agent who one takes to be reliable reporter. Brandom gives an account of the sorts of grounds that one has for attributing perceptual knowledge to someone else, not an account of the sorts of grounds one has for having perceptual knowledge oneself.

McDowell raises an issue with Brandom’s account of perceptual knowledge in which, as Brandom says, “Perceptual experiences, if any, are only contingently involved,” (2010, 322). Maintaining his commitment to this claim of the contingency of perceptual experiences, Brandom makes a very startling admission in response to McDowell’s case of the chicken-sexer. He writes, “Could the responsiveness to reality of some discursive practitioners—no doubt, ones quite different from us—be ‘chicken-sexing all the way down’? I do not see why not,” (2010, 325). We would, on Brandom’s account, be able to attribute (what Brandom calls) perceptual knowledge to these discursive practitioners for whom responsiveness to reality is “chicken-sexing all the way down”. But do they actually perceive anything at all? I don’t see how we could answer anything but “No”. From their perspective (if they can be said to have one at all) the lights are completely out. Though Brandom says that these practitioners would have to be “quite different from us,” their difference, on his account, would have to be a contingent empirical matter. There are no a priori grounds to rule out beings just like us for whom it is “chicken-sexing all the way down”. So, ruling out conceptual possibilities on only a priori grounds, Brandom commits himself to the conceptual possibility of discursive practitioners who would be nothing other than what David Chalmers calls “philosophical zombies,” (1996, 84). By emphasizing the third-personal perspective of the attributor of entitlements and commitments rather than the undertaker of commitments, Brandom’s account leads him to admit of the possibility of discursive beings for whom the first-person perspective in perception is eliminated entirely.

6 Reviving Perceptual Experience

What is it that Brandom’s hypothetical “chicken-sexers all the way down” are lacking? According to McDowell, it is what Kant calls “intuitions,” that aspect of experience with respect to which one is passive through which one has objects in view so as to be able to actively apply concepts to them in acts of judgment.^[22] Brandom rejects the applicability of a Kantian distinction between the passive having of intuitions and active application of concepts to his work. In the concluding chapter of Making It Explicit, he claims that Kant’s distinction between intuitions and concepts is essentially dualistic, saying that “A distinction becomes a dualism when its components are distinguished in terms that makes their characteristic relations to one another ultimately unintelligible,” (1994, 615). Intuitions and concepts, Brandom claims, cannot be articulated in a set of terms that makes their relation intelligible. Given how Brandom is thinking about what it is that these two terms function to pick out and the set of commitments that he undertakes in Making It Explicit, this claim is not surprising. He thinks of Kant’s distinction between intuitions and concepts along the model provided by C.I. Lewis (1929) where we have, on the one hand, a non-conceptual given element in experience, and, on the other, conceptual activity that orders or structures this non-conceptual given (Brandom 1994, 616).^[23] McDowell (1994), however, has suggested an alternate way of thinking about the Kantian distinction between intuitions and concepts in which our conceptual capacities are already in play in our having of intuitions, where this is still importantly distinct from our actually making judgments, actively applying concepts to objects that we have in view in experience.^[24]

Now, McDowell has wavered over the years on how exactly the distinction between intuitions and concepts is to be understood.^[25] Here, I’ll offer my own way of articulating this distinction, principally employing the statuses of agentive modality.^[26] What I want to suggest is that not only can we frame a distinction between intuitions and concepts in terms of the statuses of agentive modality, thereby specifying their relation in a way that Brandom should find perfectly intelligible, but that, when we do so, we see that there’s a direct mapping between the statuses we bear towards judgments on the agentive modal articulation of perceptual experience and the normative statuses we bear towards claims on Brandom’s account of perceptual knowledge. If we can then articulate an explanatory dependence relation going from the latter perspective to the former, we can show that, despite Brandom’s claims to the contrary, Making It Explicit actually implicitly contains an account of perceptual experience, distinction between intuitions and concepts and all.

So, imagine that you’re looking at a complex arrangement of 3-dimensional solids of various shapes, sizes, colors, and patterns. How should we articulate what it is in which your visual experience consists? The suggestion is that we articulate it in terms of the agentive modal statuses that you bear towards various judgments in having that experience. Start with the status of ability. Sebastian Rödl (2007b) suggests that is through the deployment of this modal status that one can understand the contribution of sensibility in perception, speaking of “acts of sensibility as something by which one is in a position to think of an object,” (176).^[27] Here, we should understand the talk one’s being “in a position to think of an object” in terms of one’s being able to think about it. The core thought here is that experience puts us in a position to make judgments about the things in world, enabling us to make these judgments.^[28] For instance, if there is some visibly red cube in my field of view, I’m able to judge that the cube is red. Now, though it’s hard to draw a determinate line, the set of judgments that I’m able to make, in having a particular experience, extend beyond the set of judgments that I actually do make. There is a whole bunch of things in my field of view right now, visibly instantiating a whole bunch of very determinately specifiable properties, and so I’m able to make a whole bunch of judgments, attending to various features of my visual experience and taking things to be how they visibly are. I don’t, however, actually make all of these judgments; I couldn’t possibly do so. Indeed, one feature of perceptual experience that is often pointed out in debates about the nature of perceptual content is that experience is “rich” in the sense that it enables one to make a potentially infinite set of judgments.^[29] For instance, I can judge that the cube on top of the blue cylinder is slightly more orange shade of red than the cube under the purple sphere. What enables me to make this judgment? Well, I have these things in view, and I can see that they are these ways. The notion of “having something in view” here is not to be understood apart from the acts of conceptual capacities that are in potentiality in having it in view; having something in view, and having it in view as being certain ways, just is to be able to make judgments about it, judging that it is these ways. Intuitions are thus characterized in purely agentive modal terms, and the only “contents” here are conceptual contents, the contents that figure in acts of judgment.^[30]

Consider now the status of compulsion. I’ve just said that when there’s something visibly red in one’s field of view, one is able to take it to be red. But additionally, when there is something clearly red right in front of one’s nose, not only is one able to take it to be red, but one is compelled to do so, in the sense that one cannot but take it to be red. There have been various ways of describing this compulsion in the literature on perception, but one of the more picturesque ones is from Wilfrid Sellars, who speaks of a claim as being “wrung from a perceiver by the object perceived,” (1956/1997, 40). This phrase is echoed by Brandom in summarizing his own account, saying “when we are properly wired up and trained, the perceptible facts wring from us perceptual judgments,” (1996, 372). Though Brandom is clearly moved by this phrase, he never gives any official account of what is meant by it, and it’s not clear at all what the place of this vocabulary can actually be on his account of perceptual knowledge. The talk of an object or fact “wringing” a judgment from someone brings forth the imagery of the judgment’s being, in some way, forced out of them. However, taken literally, the only sort of “force” that mere objects and facts themselves exert is merely causal, and as Sellars (1956/1997) and McDowell (1994) have powerfully argued, the constraint placed upon us by objects in perception cannot be merely causal.^[31] The way to give precision to this idea of a perceived object or fact “wringing” from us a judgment, I am suggesting, is in terms of the agentive modal status of compulsion. If there is a big red cube right in front of my nose, the judgment that the cube is red is “wrung” from me in the sense that I cannot but make this judgment, given the clear presence of this manifestly red cube in my perceptual experience. The compulsion here is not merely causal. My experience makes it clear to me that this judgment is to be made, and, it being made clear to me that I am to make this judgment, I cannot but make it. So, I am not merely causally, but rationally compelled to make this judgment, given my experience.^[32]

So, when we imagine a rich scene before us, we can articulate two sets judgments to which we bear the agentive modal statuses of ability and compulsion in having the perceptual experience we have; our perceptual experience enables us to make a certain set of judgments and compels us to make certain set of judgments, where the second set is a proper subset of the first. Furthermore, for any judgment that we are able to make, there is something that we can do such that, if we do it, we are compelled to make that judgment, namely, attend to the relevant features of our experience. Now, imagine a Brandomian scorekeeper who knows what our reliable differential responsive dispositions are, attributing to us normative statuses. They’ll take us to be entitled to the claim that the cube on top of the blue cylinder is slightly more orange shade of red than the cube under the purple sphere. Given that these things are in our line of sight, this is a claim that, normatively speaking, we are in a position to make. They won’t, however, take us to be committed to this claim, not, at least, until they point it out to us. They will, on the other hand, take us to be committed to the claim that the bright red table is red. So, we’ll be attributed two sets of normative statuses to claims, a set of claims to which we are entitled and a set of claims to which we are committed, where the second set is a proper subset of the first. Furthermore, for each member of the first set that is not a member of the second set, there is something this attributor can do, such that, if they do it, we’ll come to be committed to that claim, namely, point it to us. So, the experience we have can be articulated, first-personally, in terms of the set of judgments that we are able or compelled to make, given our experience, or third-personally, in terms of the set of claims to which we’re entitled or committed, given our reliable differential responsive dispositions. There is a direct mapping between what is articulated by these two modes of articulation.

What are we to make of this correspondence? Let me now sketch, at a very high level of abstraction, the basic structure of a general account, one that promises to bridge the rift in Pittsburgh Philosophy that I spoke of earlier.

7 A General Account

On the account of discursive mindedness offered by Brandom, and Sellars before him, one’s being discursively minded is a product of one’s having been inducted into a discursive practice, a practice which essentially involves practices of giving and asking for reasons. It is only in virtue of being inducted into a discursive practice by those who are already competent participants of that practice that one comes to be minded at all. The process of being inducted into discursive practice is a process in which one, prior to being a competent participant, is held in speaking and acting to the norms of the practice by those who are already competent participants, positively sanctioned when one says and does what they do say and do and negatively sanctioned when one says or does what they don’t say or do.^[33] Through being held to the norms in this way, one’s activity gradually starts to come into conformity with them. Eventually, one starts, in speaking and acting, holding oneself to the norms of the practice to which one is being held by others, coming to act under a conception of them, doing and saying things in virtue of thinking that those things are to be said and done, and not doing or saying things in virtue of thinking that those things are not to be said or done. Since sayings and doings are activities that take place in a world with others, to be able to think about what is to be said or done is at once to be able to think about the world in which these things are to be said and done and the others with whom one shares it who are able to evaluate these sayings and doings as correct or incorrect. In this way, discursive mindedness—the capacity for thinking about oneself, the world, and others—is a product of one’s having been inducted into a discursive practice.

Now, on the account of discursive mindedness offered by Conant (2020), which I take to be representative of the members of the Pittsburgh School on the other side of the rift, the form of the capacity for thinking is something that one is able to comprehend only by self-consciously thinking through what it is for the capacity to be, as Conant puts it, in act.^[34] This comprehension that one has of the what it is for the capacity for thinking to be in act, this comprehension of the “form” of the capacity, is essentially first-personal in its source: it is a comprehension of what it is for the capacity to be in act that one comes to have in and through one’s own actualization of that capacity. Nevertheless, the structure that one is actually articulating, in articulating the form of one’s capacity for thinking, is no other than the structure of the norms of the discursive practice into which one has been inducted, for it is through the induction into these norms that one has the very capacity that one is articulating. Now, once again, the norms that constitute this structure are the norms to which members of that practice hold one another in their discursive activity. As such, the fundamental mode through which these norms are to be articulated is third-personal; they are to be articulated in terms of the attribution of normative statuses to someone else whom one is holding to the norms of the discursive practice. So, even through the explicitation of the form capacity for thinking must come in a first-personal vocabulary, the explanation of the form of capacity for thinking can still come in a third-personal one. Let me finally spell out, using Brandom’s game-playing model of discursive practice, just what this explanation might be.

In order to play the game of giving and asking for reasons, a player must keep track of their own score. They have to be able to keep track of the claims they’ve made, the claims to which they’re committed, the claims to which they’re entitled, and the claims to which they’re precluded from being entitled. They do this, in the first instance, not according to anyone else’s principles for keeping score, but according to their own. So, they keep a scorecard for themseslf, and they do this according to their own scorekeeping principles. When a player, for instance, takes themself to be committed to p, and, in being so committed, to be precluded from being entitled to n o t ‐ p , this is in virtue of their applying a general scorekeeping principle to themself, keeping the scorecard they keep for themself in accordance with this scorekeeping principle. On this account, the capacity for thinking just is that capacity through which discursive practitioners keep track of their actual and potential places in the normative structure in which they are capable of placing themselves, as it appears from “within”, So, if this player were to first-personally think through what it is for them to think p, this activity of applying a general scorekeeping principle to themself—the principle commitment to p precludes entitlement to n o t ‐ p —would be something on which they would get a grip. This principle shows up in some way in their first-personally thinking through what it is for the capacity to think p to be in act, but it does not show up in normative terms, as a scorekeeping principle. Rather, it shows up to them in agentive terms: they take that they think p, and, insofar as they continue to do so, this precludes them from being able to think n o t ‐ p .^[35] The agentive modal fact on which they first-personally have a grip is, articulated from “without,” their self-application of basic scorekeeping principle that they’ve acquired through their induction into discursive practice, as they, along with anyone else belonging to a discursive practice with a negation operator, maintain the scorecard that they keep for themself by way of it. Now, it is not wrong to say that, in having a grip on this scorekeeping principle, they have a grip on a basic principle articulating what it is to be a thinker. Indeed, they do. It’s just that what it is to be a thinker just is to have internalized the basic norms of the discursive practice into which one has been inducted and to have one’s capacity for cognitive activity constituitively enformed by these norms.

This account of the capacity for discursive thinking yields a philosophical conception of ourselves that is “formally idealist” in Conant’s (2020) sense, adhering to the following thesis:

[T]here is no understanding the concept of a capacity for knowledge from a standpoint that seeks to comprehend what knowledge is from the outside—that is, apart from the form of self-conscious self-understanding that a subject who possesses the concept of such a capacity exhibits precisely in virtue of her knowing the capacity through its exercise—where the capacity one seeks to understand is the very capacity one must, at the same time, exercise in order to achieve such understanding, (776).

Conant is right that we cannot comprehend what it is for the capacity for knowledge to be in act from the outside; we are to comprehend what it is for the capacity for knowledge to be in act through self-consciously actualizing it. Nevertheless, we can, from the outside, explain how it is that we’ve come to have it, articulating, in third-personal terms, the normative structure into which we’ve been brought in virtue of which we have the capacity for knowledge that we have, which we comprehend first-personally. We can maintain that we know about the structure of discursive practice in and through actualizing the capacity for knowledge while also maintaining that the structure of the practice that we thereby know about in fact explains the form of the capacity for knowledge through which we know about it. There is no problem here insofar as we remember, as Sellars has taught us, that priority in the order of knowing and priority the order of being, or, as I’d put it here, priority in the order of comprehension and priority in the order of explanation, can go in opposite directions.^[36] My hope is that exploiting the correspondence between deontic and agentive modality we will be able to bridge a major rift in Pittsburgh Philosophy, synthesizing the work of Brandom, on the one hand, who understands our discursivity in terms the structure of the norms of the discursive practice to which we belong, and thinkers such as Conant ( 2020], [Rödl (2007 2013), and Kimhi (2018) on the other hand, who understand our discursivity in terms of the form of our capacity for thinking, judging, and knowing, which we comprehend in and through self-consciously actualizing that capacity. My impression is that many thinkers on the both sides of the rift think that it’s unbridgeable. I think that the correspondence I’ve highlighted here, and the account of its significance that I’ve just sketched, give us the resources to bridge it.

8 Conclusion

Brandom conceives of the intentional nexus as fundamentally articulated in terms of two flavors of modal vocabulary: alethic and normative. I have suggested here that we add another flavor of modal vocabulary to the mix: the vocabulary of agentive modality. Adding this bit of modal vocabulary to the mix enables us to officially incorporate into Brandom’s story something that several of his critics have claimed he cannot: the first-personal perspective on thinking, judging, and knowing. I have considered here specifically how this vocabulary can be brought to bear in giving an account of the first-personal perspective in perceptual knowledge. In doing this, I hope I have demonstrated a general method for exploiting the correspondence between agentive modality and discursive normativity that can be pursued in order to fill in various aspects of the agentive theory of mind that emerges from Brandom’s normative theory of discursive practice. My hope is that, by actualizing the potentialities for philosophical theorizing that are afforded to us by having this tool for thinking ready at hand, we can bridge the gap between Brandom and his deepest critics.