Openness to the World and External Constraint

According to Kant, sensibility is characterized by passive receptivity. It apprehends that which is given in intuition, both pure and empirical. The understanding, on the contrary, is characterized by spontaneity. It provides concepts as instruments of thought and judgement. In claiming that only the unity of receptivity and spontaneity may result in knowledge, Kant clearly does not give any preference to either of the two:

Contrary to Kant, McDowell claims that there is nothing that is passively given in experience. In empirical thinking our total freedom finds its expression (5). There is no gap between the impact of external reality and our conceptual response to it. Although thinking is characterized exclusively by spontaneity and freedom, sensibility, according to McDowell, is not something that is entirely receptive and passive. It is wrong, McDowell says, to see thought and sensibility as opposing extremes or distinctly different capacities: “When we enjoy experience, conceptual capacities are drawn on in receptivity, not exercised on some supposedly prior deliverances of receptivity” (10). Spontaneity, he insists, is operational all the way out to the conceptual contents that “sit closest to the impacts of the world on our sensibility” (11).

McDowell’s idea of our openness to the world in empirical thinking emerges directly from his thesis about the joint involvement of receptivity and spontaneity in perceptual experience: if we agree that experiences are always already equipped with conceptual content, we should conclude that in veridical experiences we know how things really are (25).^{Footnote 6} The world itself, which is transparently and directly open to us in veridical perception, in its transparency provides guidance for our thought.^{Footnote 7} The thesis that mind and world share a conceptual nature allows McDowell to arrive at a strong structural isomorphism: I may think that spring has begun, and it also may be that spring has begun (27). He is careful to stipulate that the identity between facts and conceptual contents, which expresses the gist of his direct realism, should not be understood in the sense that he does not care to distinguish between the fact of, say, Jim sitting on a chair right now and Jim’s thinking or perceiving that he is sitting on a chair right now. Although the fact of the world really is also the content of Jim’s experience, what is precisely not identical, even in veridical experience, is the fact of the world and Jim’s act of thinking, in which the fact is being revealed to him (28).

The distinction between facts and acts of thinking about those facts helps McDowell to introduce the idea of external rational constraint. The constraint, he says, should be from outside any particular act of thinking, but it need not be from outside thinkable conceptual contents that our acts of thinking embrace. Conceptual contents thus open us to something that is factual, qualitative, and has certain haecceity. In order to understand how the constraint works, McDowell says, we need to go as far back in our reasoning as necessary to reach the contents that are “ultimate in the order of justification” (29). He then clarifies that “when we trace justifications back, the last thing we come to is still thinkable content; not … a bare pointing to a bit of the Given” (Ibid.). Alluding to “Sense-certainty, or This and Meaning,” the first chapter of G.V.F. Hegel’s Phenomenology of Spirit, McDowell adds that we should “reject the idea that tracing back the ground for a judgement can terminate in pointing to a bare presence” (39).^{Footnote 8} What McDowell is trying to do in the corresponding passages is not to stop the regress of knowledge, which supposedly can be terminated only by something that is in any sense simply given, but rather to undermine the basic assumption that leads to the regress, the assumption being that no intentional state can put a perceiver directly in contact with reality.^{Footnote 9} Accordingly, in stressing our openness to the way the world really is, we must be able to register our experience in empirical judgements. There should be a conceptual structure shared by experiences and judgements and, therefore, every perceptual experience should already have a propositional form. To extend McDowell’s example used earlier, it may be that spring has begun, I may think that spring has begun, and I may report that spring has begun.

This, however, is not the only way the notion of external constraint is introduced in Mind and World. There is a second way, which is related to McDowell’s ideas of tradition and second nature. When analyzed in relation to the first one, it raises two concerns. The first concern is the circularity between having conceptual capacities as they are described in the ‘factual’ scenario above, and participating in learning practices that habituate us into those capacities, as described in the ‘social’ scenario, to be discussed below. The second concern is that McDowell does not explain how exactly Bildung—the term that he introduces in Lecture IV, and that presupposes an appeal to social practices—can ground our openness to facts. In what follows, I will first describe the way the notion of external constraint is introduced within the framework of second nature, and I will then discuss the concerns.

Just like Sellars and Brandom, McDowell believes that every concept is understood only as forming a part in a larger whole. Even a simple colour concept cannot be grasped autonomously, without understanding how it fits in to a perceiver’s worldview that rationally links our conceptual capacities into a network (31).^{Footnote 10} Correspondingly, any judgement, which endorses experiences framed by a concept, becomes intelligible only in a variety of interrelated social and cognitive contexts, in which the concept is constantly reappraised. McDowell needs this social dimension because without it our openness to the world in experience conveys nothing about us as knowers. Only our learning history can tell how we actually ascend to the space of reasons. McDowell provides an account of the idea of learning in Lecture IV by introducing the notion of ‘second nature.’ He first derives the notion from Aristotle’s account of moral character, and then extrapolates it to his own theoretical account of initiation into custom and, more broadly, into human conceptual capacities, including responsiveness to rational demands beyond those of ethics.^{Footnote 11} In brief, the idea of second nature is that we are not born with conceptualized perceptual capacities, and that we must learn to see and hear the world by mastering the use of words and participating in social discursive practices. However, having learned to see and hear the world by mastering the use of words, what we see and hear is precisely this: the world (27). McDowell’s central thesis is that our learning history, which constitutes our second nature, is not the history of building a mediating screen of appearances between us and the world, or some sort of conceptual scheme that provides us with access to the world by means of organizing sense data. The openness of the world to us as possessing this nature is unmediated, and so there’s no ontological chasm between world and thought, because second nature, which is acquired in our upbringing, does not add anything to our first one, but simply—in the distinctly Hegelian sense—sublates it.

As a Wittgensteinian, McDowell holds that a practice is a habituated modus operandi that matches a type of response to an appropriate type of circumstances. Therefore, it cannot be reduced to a set of explicitly formulated rules. Likewise, what language acquisition results in is not just a finite set of particular skills, but a new, second nature that ‘discovers’ the world as it really is and predetermines possible outcomes of our actions and perceptions. By sublating first nature, second nature provides us with the possibility of rational response and, therefore, with the very possibility of outer experience. Referring to Hans-Georg Gadamer’s Truth and Method, McDowell writes that “mere animals” cannot transcend beyond their immediate needs, that their existence is restricted by immediate biological imperatives that predetermine particular successions of problems, and that “the life structured in that way is led not in the world, but only in an environment” (114-115). There is not much out there for a creature that lives within itself, affected only by immediate challenges. The outer, as a description of experience, in turn, refers to the reflective stance specific to human consciousness. This consciousness expresses our sapience, which cannot be derived from mere first-natural sentience. This stance makes us capable of deliberating and choosing between different incentives and, therefore, of having reasons to act one way rather than another. This establishes a strong connection between being out there in the world and the idea that the world is entirely open to us, because it contains nothing extraconceptual.^{Footnote 12} McDowell’s overall objective in this set of arguments about second nature is to restate his reinterpretation of Kant in different terms and to put the discussion about external constraint and the relationship between receptivity and spontaneity into a wider philosophical focus.

In Lecture V of Mind and World McDowell finally introduces the concept of ‘Bildung’—the concept that I take to express the result of fusing together the ideas of proper human upbringing and initiation into a custom, and which is the keystone in McDowell’s overall reinterpretation of the idea of nature. Whereas Aristotle’s idea of proper ethical upbringing describes an autonomous space of ethical reasons, and Ludwig Wittgenstein’s idea of initiation into a custom allows us to understand perception within the framework of language games, Bildung represents the autonomous space of reasons in the most general sense. Characteristically, though, McDowell keeps his readers in the dark as to what part of the rather long history of the term he refers to.^{Footnote 13} For this reason, McDowell’s ideas of second nature and Bildung are often deemed to be undetailed and ambiguous.^{Footnote 14} Bildung and second nature represent a cluster of ideas that are only vaguely defined, and initiation into a custom sounds more like a magical transformation of sorts.^{Footnote 15} McDowell only tells us that Bildung signifies the transformation of human beings, who are born mere animals, into “thinkers and intentional agents in the course of coming to maturity,” adding that Bildung also represents “a central element in the normal maturation of human beings,” of which practices of language acquisition are the most important part (125).

Social Practices and Conceptual Contents: Two Scenarios

Whereas, in the case of our openness to the facts of the world, the conceptual is realized as our knowledge of how things are, in the case of Bildung, which refers to our learning history, things get more complicated. One particular worry is McDowell’s interpretation of Aristotle in which McDowell uses the formation of ethical character as a model for initiation into conceptual capacities in general. Unlike McDowell, Aristotle makes an explicit point that our perceptions and irrational impulses form a part of human nature just as much as our reasoning does. The passionate/perceptive part is transformed by deliberation into choices, and so both the rational and the passionate/perceptive sides of the soul together take part in the formation of moral character through acquiring habits of action. David Forman claims that, due to an intricate dialectics between potentiality (i.e., capacity to do something habitually) and activity, which Aristotle develops in Nicomachean Ethics, the way consecutive steps of habituation are described by Aristotle cannot serve as a model for a broader theory of how our basic conceptual abilities are acquired.^{Footnote 16} In Nicomachean Ethics, Aristotle says that habituation, which results in acquiring a virtuous skill—say, playing the lyre, or building a house—begins by things we actually repeatedly do in order to play the lyre or build a house. Meanwhile, in the case of our openness to things being thus and so in sensual perception, as McDowell implies, we first have to acquire the capacity to see or hear, and only then exhibit the activity. The unwelcome result is that, “if we possess the conceptual ability before we have the habit, then the habit cannot itself be what distinguishes a conceptually informed perception from a merely natural response to the environment.”^{Footnote 17} It appears that an attempt to model perceptual habits on habits that constitute virtuous behaviour lead to an impasse: we seem to need to have conceptual capacities, which ground the isomorphism between experience and language, before habituation can result in acquiring them. On this view, our first and second natures work quite differently, and the second one can actually be called ‘nature’ by weak analogy only. Besides, McDowell, who in the first three lectures of Mind and World treats perceptual judgement as a model for all judgements, upon the introduction of the idea of second nature changes his strategy and begins to treat Aristotelian moral judgements as a model for perceptual judgements.

Although no explanation for this change in priorities is provided in Mind and World, some sort of a justification for it may be found in McDowell’s earlier works. In particular, in his early article “Virtue and Reason,” which stresses Aristotle’s holistic approach to virtues, McDowell claims that “[w]e do not fully understand a virtuous person’s actions—we do not see the consistency in them—unless we can supplement the core explanations with a grasp of his conception of how to live.”^{Footnote 18} According to Aristotle, virtues are interconnected, and no single virtue can be enacted unless a person possesses all of them. A virtuous person has an overall conception of how to live, and this conception can be reduced neither to separate virtues, nor to a particular set of explicit rules of conduct. It seems obvious that this early text of McDowell’s provides a link that is actually missing in Mind and World; namely, the link between McDowell’s interpretation of Aristotle’s and Wittgenstein’s theses that we gradually acquire understanding of language as a whole rather than acquiring a complete and clear understanding of certain parts of it, or particular situations a language game may put us in. For example, in On Certainty Wittgenstein says: “When we first begin to believe anything, what we believe is not a single proposition, it is a whole system of propositions. (Light dawns gradually over the whole.)”^{Footnote 19} Thus, Aristotle’s notions of habituation and human upbringing are interwoven with Wittgenstein’s idea of initiation into a custom. According to Aristotle, I cannot express how to live in a maxim, but I have habits and a capacity to get involved in certain practices. Likewise, according to Wittgenstein, following any rule—whether epistemic or ethical—is a public practice and, in following this practice, I simply act as I have been trained to. The middle course between a performance understood as a blind reaction to a situation, on the one hand, and interpreting a rule by an explicitly stated formula, on the other hand, is initiation into a certain custom.

It seems that, even if we agree with Forman in that McDowell’s reading of Aristotle leads to an inconsistency in terms of the relationship between conceptual capacities and the process of habituation, Wittgenstein’s idea of initiation into a custom somehow remedies the situation. Not only having virtues, but also having perceptual experience—which is indissolubly related to having the capacity to use correctly appropriate linguistic expressions—is a matter of habituation. Meanwhile, we also need to keep in mind that McDowell’s initial claim is that we are ultimately capable of achieving an agreement between each other based on shared contents. It’s not enough to share a form of communication, normative presuppositions, social practices, or a charitable disposition toward others. It is reasonable to conclude that, on McDowell’s view, every natural language is attuned to the standpoint that reveals the way the world really is, or the ultimate conceptual structure whose every element can be retrieved from the repository by going sufficiently far back in an appropriate line of reasoning. Meanwhile, Wittgenstein’s idea of habituation through initiation into a custom presupposes that any ‘epistemic constraint’ or ‘normative friction’ is a matter of answerability to a way of life rather than to facts or contents.^{Footnote 20} Moreover, it is tempting to think that siding with any existent reading of Wittgenstein’s rule-following paradox is to agree that appealing to past usages can tell us very little of substance about meaning.

The second concern about the relationship between the two scenarios (and the two corresponding ways of introducing the idea of external constraint) is this. McDowell’s contention is that conceptualized perceptual experience can be considered both relational and contentful because having propositional experiential content itself puts the perceiver in a position to have associated, socially grounded conceptual knowledge about her environment.^{Footnote 21} For instance, some visible features of this red rectangular object in front of me are inseparable from the fact that it is a book. When I see what is usually called the ‘front cover,’ I think of pages filled with letters, even though I don’t see those letters just by looking at the red rectangular object. At the same time, what I see can also be described as an ‘introspectively knowable content.’ The upshot of this argument is that we simply cannot separate what we take to be the content of our experience from our relatedness to the object of it, because, in McDowell’s own jargon, “if an experience warrants one in knowledge of the sort that is an act of rationality, the warrant-constituting status of the experience must be part of the content of an at least implicit self-consciousness that belongs to one’s cognitive state in knowing what one does.”^{Footnote 22} In other words, the way I can justify my perceptual experience is inevitably a part of the content of that experience. From this, it follows that the only way to pose a gap between the epistemic significance of experiences in terms of content and a social-relational account of the cognitive role of perception is “to conceive the content as intervening between the subject and her environment.”^{Footnote 23} And this, obviously, is a bad result, as it leads to the direct contact of mind with the world being lost. In the concluding lecture, Lecture VI, of Mind and World, this seemingly non-contradictory unity of meaning and justification is supplemented by the new role of language. The principal function of language now is not to register our experience in empirical judgements, but to serve as “a repository of tradition, a store of historically accumulated wisdom about what is a reason for what” (126). According to this interpretation, that things are thus and so is due not only to the way a given content is related to other contents, but also to the way it is maintained, modified, and stored by our tradition-preserving language.

But how exactly does McDowell’s conceptual holism, re-described in view of the idea of language as a repository of tradition, work? For one thing, McDowell is as far as possible from endorsing inferentialism of the Brandomean stripe. Particular haecceities, which, according to McDowell, one gets to be opened to at every step of tracing her justifications back, are nothing like Brandomean conceptual contents that are nods in inferential networks defined entirely in terms of material incompatibilities and consequences of judgements in which they are used. These haecceities are something we are open to in the way one cannot be open to inferential networks. Brandom is quite explicit about his disagreement with McDowell on this score. In Brandom’s case, what particular instances of using an expression reveal within such networks is based on the interplay between explicit judgements and norms implicit in discourse. On this view, we can go back and forth changing the configuration of our current commitments and entitlements ad infinitum, but we cannot play it all back to the contents that, as McDowell has it, can “sit closest to the impacts of the world on our sensibility” (11).^{Footnote 24} According to Brandom, to be rationally constrained by the world is to have a set of reliable differential responsive dispositions to a fairly large number of things and situations. The dispositions mutually define and coordinate individual perspectives in such a way that excludes a unique irreducible community perspective above and beyond the aggregated actual perspectives of all participants. Consequently, Brandom’s way of defining the external constraint on what we think and do is to say that the objectivity of the norms we apply has something to do not with contents made available in experiential thinking, but with a structural feature of discourse common to each individual practitioner’s perspective captured by the non-coercive authority of reasons interlocked by mutual accountability of all participants.^{Footnote 25} While McDowell ultimately links the answerability of knowledge to experience, Brandom confesses that “experience” is simply not one of his words, and insists that using this word implies succumbing to the myth of the given.^{Footnote 26}

Let me summarize. According to McDowell, conceptual activity belongs to the realm of freedom. Freedom demands constraint, which involves the idea of externality. The constraint, McDowell claims, can be provided only by and within experience—hence the temptation to appeal to the experiential given. McDowell’s solution is to show that experience is conceptual through-and-through, and that the constraint need not be from outside thinkable contents. But when the constraint is introduced, we are told two different stories. On the one hand, we have what McDowell calls ‘tradition.’ On this account, our experiential openness to the world is an outcome of the coherence of different stories of personal upbringing and initiation into customs, as well as of the history, throughout which tradition as a whole is developed and critically reassessed. On the other hand, we have contents that are something more than simply nods in inferential networks. They are grounded in experience and are not defined exclusively in terms of the roles they play in a game of defining what is a reason for what. Most importantly, they allow us to travel back and forth along the order of justification and ground our agreement in the way social practices, interlocked individual perspectives, charitable attitudes, and shared competences are unable to ground it. Given that McDowell doesn’t give enough detail on how being constituted by social practices and being constituted by conceptual contents are related, the goal is to understand how concepts can be interpreted as both socially and linguistically charged and, at the same time, genuinely involving the world as it really is. If the above analysis makes sense, it shows that the relationship between the two scenarios is, at the very least, underexplained—which leads to the two problems discussed above. When we move from the structural isomorphism between facts of the world, conceptual contents, and linguistic reports to the idea of language as a repository of tradition, we face two issues: first, the circularity between possessing conceptual capacities and being habituated into those capacities; second, the need to provide an additional explanation as to how exactly Bildung, which presupposes an appeal to social practices, can ground our experiential openness to the world. My suggestion is that the point of entrance for the social in the overall architectonics of Mind and World might be clarified in light of the Sellarsian idea of sense impressions. I don’t mean to define the social by means of sense impressions, but only to claim that this idea can help us understand how the social and the conceptual coexist in Mind and World.

Sense Impressions and the Social

Like McDowell, Sellars aims to show that the choice between the myth of the given and the frictionless spinning of our coherent beliefs in the void is a forced one. In describing the idea that the authority of our reports about states of affairs depends on the authority of self-authenticating non-verbal episodes, which result in our immediate awareness of something being the case, Sellars intentionally uses the language of the obviously nonsensical ‘flat earth mythology.’ He says that the episodes of which we are allegedly immediately aware “constitute the tortoise on which stands the elephant on which rests the edifice of empirical knowledge.”^{Footnote 27} However, in spite of this metaphor, almost at the outset, Sellars introduces the idea of sensory impressions, or sensations. He stipulates that sensations neither represent the traditional empiricist sensory given, nor constitute empirical thinking in the McDowellian sense of the term. As Michael Williams puts it, according to Sellars, a sensory impression like seeing a red triangle is not so much something that opens us to the world of facts as it is a “state of the perceiver that is common to his seeing that the object over there is red and triangular, that object’s looking to him to be red and triangular, and its looking to him that there is a red and triangular object over there.”^{Footnote 28} Sellarsian impressions thus mark what exactly (a) the fact that the perceiver must be there, (b) the fact that the object might not be red, and (c) the fact that there might be no object at all—have in common. And while McDowell suggests that the sameness of the claim that ‘x over there is red’ in the three experiences is already quite enough, Sellars insists on adding impressions in the mix.^{Footnote 29}

Sellars points at two ways to explain the common role of impressions in (a), (b), and (c) above that must be avoided. First, it is the empiricist explanation of it by postulating immediately experienced sense-contents that are just given. Second, it is the conceptualistic interpretation of impressions as thoughts. In the first case, we arrive at the unwelcome idea that our ability to sense sense contents is unacquired. In the second case, there’s much more going on in conceptually loaded observational reports than what is provided by particular sensual impressions. Impressions, Sellars insists, are states of a perceiver rather than thoughts or particulars perceived. Moreover, it seems wrong to talk about having impressions ‘on purpose,’ so impressions should also be “purged of epistemic aboutness.”^{Footnote 30} In describing the idea, Sellars mentions “a temptation to assimilate ‘having a sensation of a red triangle’ to ‘thinking of a celestial city’ and to attribute to the former the epistemic character, the ‘intentionality’ of the latter.”^{Footnote 31} He insists that we should be wary of considering ‘having a sensation of a red triangle’ readily recognizable as a cognitive and epistemic fact (he adds, though, that we do not need to remove impressions from the domain of epistemology entirely). Impressions are necessary in order to distinguish between ‘red X, here, now’ and a unified conceptual structure that contains, say, an image of a red surface covering a volume of white, the taste of a cool juicy fruit, and the general idea of an apple.

Importantly, Sellars thinks that we are led to believe in a non-conceptual character of impressions not because we must postulate something outside the network of concepts to connect the network to the outer reality, but because impressions do the work of causal guidance in connecting perceivers and objects perceived, whether the connection represents illusions, hallucinations, or something one really encounters in experience.^{Footnote 32} Once all of this is taken into account, Sellars says, we can establish that the fundamental link between language and the world does not require unacquired sense contents, and we can also claim that sensual impressions causally guide our perceptual reports towards the objects these reports are about, “without being committed to the mistaken idea that it is ‘really’ sensations of red, rather than red physical objects, which are the primary denotation of the word ‘red’.”^{Footnote 33}

As Williams notes, according to Sellars, on the level of sensory impressions, experience may be presented as a succession of “glimpses of reality,” where the idea of such glimpses suggests “the momentary registering of a passing phase in the career of an object: a recording of the here and now.”^{Footnote 34} This recording differs from ordinary observational reports in that ordinary observational reports “characterize objects in terms that go way beyond anything strictly available to momentary inspection,” and are “extensively ‘iffy’, attributing causal powers that cannot in any obvious sense be ‘glimpsed’.”^{Footnote 35} Observational reports are conceptually loaded; they occupy the final stage in “the career of an object”—quite unlike impressions, that are “minimal representations, in that they register the austere cues that guide the rich conceptual episodes exemplified in perceptual takings.”^{Footnote 36} As non-cognitive states, impressions are, in some sense, like Kantian ‘appearances’—they are transitive perceptual states that are already apprehended as a part of a larger succession, but for all parts of the succession to hold together, they are yet to be run through, in order that the unity of intuition might arise out of their manifold.^{Footnote 37} These impressions are already not just passive intakes of sensibility, but not yet full-blown ‘phenomena’ in the Kantian sense of the word. To become fully determinate, they must be organized under concepts, which, according to Kant, is done at the final stage of the synthesis of imagination.

Now, why does Sellars need sensations, thus defined? In McDowell’s jargon, he needs them in order to avoid a theoretical leap between the “disenchanted” interpretation of nature as totally devoid of meaning and values (73), and our second-natural capacity to use reason—which, as McDowell insists, opens us to the world as it really is. On this interpretation, Sellars, the Hegelian, purports to identify the minimal difference between being simply trained to respond in a certain way, on the one hand, and being able, in thus responding, to take a position in the space of reasons, on the other hand. But how exactly does this Sellarsian reinterpretation of the role of impressions help in McDowell’s case? First, I suggest that it is possible to accept the Sellarsian idea of sense impressions, but reject both McDowell’s concern over the transparency of sense impressions, and his claim that the non-conceptual character of sense impressions would prevent our openness to the world. Second, the Sellarsian interpretation of the role of sense impressions, I think, can do some useful work in properly delineating the idea of the social in its relation to perceptual judgements in McDowell.

But let’s restate the problem first. McDowell’s task is to conceive of experience in a way that it could figure in the order of justification, i.e., in the way it could stay within the space of reasons. What is foundational for McDowell’s overall project is both postulating our radical openness to the world in empirical judgements and demonstrating the conceptual character of experience all the way down. He believes that the idea of non-conceptual sense-impressions betrays the purpose and that dispensing with sense impressions would cost us nothing in our ability to understand experience. Sense impressions are phenomenologically idle and can never be something in which we find ourselves open to the world (144). Moreover, as sense impressions are transitive, opaque, and result from passive perceptual takings, accepting them would strip perception of the spontaneity McDowell ascribes to it. However, characterizing experience as conceptual all the way down comes at a price. McDowell’s holism requires that meaning and justification go hand-in-hand. But although, according to McDowell, justification presupposes social practices that frame our conceptual skills, the role the very notion of the social plays in creating a proper framework for the idea of meaning is yet to be clarified. In Lecture V, McDowell does stipulate that “the point is not that the social constitutes the framework for a construction of the very idea of meaning: something that would make the idea safe for a restrictive naturalism, the sort that threatens to disenchant nature.” (95) But, by his own admission, this is only to criticize the way Wittgenstein’s Philosophical Investigations are read by Crispin Wright and Saul Kripke, according to whom, on McDowell’s own interpretation, everything we need to know about meaning “can be reconstructed out of independent facts,” the facts being “social interactions, described in a way that does not presuppose the material to be reconstructed” (92). McDowell may well be right in dismissing this reading, but this dismissal by itself does not constitute an answer to the question of how sociality is related to meaning. Given this, it remains unclear in what sense conceptuality is ascribed both to experience as it is nurtured by the ever-changing social environment, and to our individual experiences each of which opens us to the factuality of the world.

An impression, as described by Sellars, is not phenomenologically independent of “the career of an object,” but, at the same time, it is not, as logical atomism would have it, a part of this career that can be considered in separation from any other part. What a sense impression, as a passing phase of this career, registers, is prior to the full-blown conceptuality of the fact one embraces in reporting that ‘x is red.’ But it neither constitutes a private episode, nor can it be considered a member of the class of private red particulars connected with language by means of associative ties. It is not given—in the sense that the myth of the given implies. Acquiring a concept, in Sellars’s own words, “involve[s] a long history of acquiring piecemeal habits of response to various objects in various circumstances.”^{Footnote 38} Impressions provide causally framed guidance in this—essentially social—history of concept use and concept formation. As samples of a perceiver’s momentary inspection, they are minimal steps in formation of piecemeal habits of response that are at work not only when a concept is being acquired, but also every time an observational report is made. Every instance of an observational report, as it were, carries with it the full history of habitual responses that made it possible—the history, parts of which the report shares with other reports, as well as with other uses of expressions it includes. From this perspective, every act of perception exemplifies openness to the world understood as an act of re-creation of this history, guided by sense impressions. Every new match, as well as every new mis-match provoked by the way certain impressions are appropriated in reporting about richer conceptual episodes (perceptual facts), contribute to the way tradition as a whole is developed and critically reassessed. On this view, observational reports, as well as the conceptual contents they reveal, are constituted by social practices and cannot mean what they mean without those practices. And what lies below the line of the social construction of meaning is not conceptual contents, but non-conceptual, transitive sense impressions that result from our perceptual takings.