The Metaphysics of Sensation; Psychological Nominalism and the Reality of Consciousness

3 The Metaphysics of Sensation: Psychological Nominalism and the Reality of Consciousness Ray Brassier, American University of Beirut (LE ) For philosophers of a realist stamp, psychological nominalism, understood as the claim that all awareness is a linguistic affair (EPM 160), is of a piece with philosophical behaviorism, which allegedly denies the reality of consciousness. If psychological nominalism is complicit with this denial—which strikes many philosophers as absurd—some would say it deserves to be relegated to the dustbin of philosophical history along with its behaviorist sibling. Wilfrid Sellars first coined the expression “psychological nominalism” and defended the doctrine throughout his long philosophical career. But he also distinguished between two aspects of the mind, thinking and sensing, and laid claim to realism about both. This distinction between thinking and sensing (which can be traced back to Kant) essentially prefigures David Chalmers’s more recent distinction between the functional-psychological and the phenomenalexperiential aspects of mind (see Chalmers 1995 and 1996). The crucial difference, however, is that while Chalmers endows phenomenal-experiential states with a cognitive authority equal to if not greater than that of functionalpsychological states, Sellars reserves cognition for functionally characterized thinking alone and characterizes sensation as a non-cognitive state that plays a causal but not justificatory role in empirical knowledge. Yet Sellars cannot be accused of downplaying the significance of the phenomenal-experiential aspect of mind. Indeed, his account of sensory consciousness leads him to make the controversial claim that sensation has a metaphysical purchase insofar as it (indirectly) reveals a fundamental aspect of physical reality: its intrinsic, qualitative aspects, which, as he sees it, cannot be reduced to the extrinsic or dispositional properties of things. Sellars’ Rortyan heirs have made much of his psychological nominalism, but tend to dismiss his suggestion that sensation has a metaphysical purchase as an 59

60 Wilfrid Sellars, Idealism, and Realism aberrant relapse into dogmatic rationalism. I want to argue that Sellars’ account of the metaphysics of sensation should not be dismissed as a dogmatic regression; rather, it offers a way of resolving the deadlock between realists whose equation of consciousness with knowledge leads them to embrace dualism and antirealists whose reduction of consciousness to conceptual awareness threatens to sever the causal link between mind and nature. The transcendental difference and the rejection of givenness Thinking and sensing are both modes of awareness. How can Sellars insist that all awareness is a linguistic affair and still claim to be a realist about both thinking and sensing? To answer this question we need to understand what “realism” means for Sellars. Sellars is a transcendental philosopher and the transcendental perspective alters the conditions in terms of which the issue of realism is framed. Transcendental philosophy can be contrasted with dogmatic rationalism on the one hand and skeptical empiricism on the other. The dogmatic rationalist assumes that the mind enjoys a priori cognitive access to mind-independent reality and believes that reason can deduce the fundamental features of that reality. The skeptical empiricist, for her part, insists that all knowledge is rooted in but limited by sensory experience and denies that reason can deliver a priori knowledge of mind-independent reality. The transcendental philosopher rejects both dogmatism and skepticism. She rejects the rationalist’s assumption that the mind enjoys a priori cognitive access to reality just as she rejects the empiricist’s claim that all knowledge derives from sensory experience. Both dogmatism and empiricist skepticism remain beholden to what Sellars calls “the framework of givenness”1: dogmatism because it takes the correlation between thinking and being as given; skepticism, because it takes the intelligibility of sensory experience as given. At the heart of the framework of givenness is the assumption, common to empiricism and rationalism, that mental states are self-intimating. Rejecting the framework of givenness, Sellars refuses the assumption that the mental is selfintimating. This means that minds do not necessarily know themselves. There is a fundamental difference between thinking and knowing what is thought. By the same token, there is a fundamental difference between sensing and knowing what is sensed. The awareness of something is not the awareness of something as something. This difference—between thinking and thought, or sensing and sensed—follows from the rejection of givenness. The core feature of the

The Metaphysics of Sensation 61 framework of givenness is the premise that the fusion of thinking and thought or of sensing and sensed is guaranteed, either by intellectual intuition in the case of rationalism or sensory intuition in the case of empiricism. Awareness and thought Sellars’ insistence that all awareness is a linguistic affair is a transcendental claim, not an empirical or metaphysical one. The awareness in question is cognitive awareness, i.e. thinking. It is thinking as cognitive awareness that is a linguistic affair. Sellars defines psychological nomimalism as follows: “[A]ll awareness of sorts, resemblances, facts, etc., in short all awareness of abstract entities—indeed all awareness even of particulars—is a linguistic affair” (EPM , 160). Cognitive awareness is a linguistic affair because it involves being aware of something as something. Such awareness requires conceptualization and concepts are linguistically instantiated rules. We would not be aware of ourselves as thinking were it not for language. Once one rejects the premise that thoughts are essentially self-intimating, or the assumption that to think is to know that one is thinking, it becomes possible to understand the concept of thinking as modeled on the concept of speaking. Sellars’ “Myth of Jones” is a philosophical reconstruction of the process through which minded beings come to recognize themselves as minded.2 Thoughts are understood as inner episodes modeled upon overt public utterances. These inner episodes are theoretical entities postulated as structurally analogous to linguistic utterances, yet devoid of the physical characteristics of utterances: they are silent, invisible, intangible, etc. The claim is not that thoughts are the causes of utterances or that utterances are the expressions of thought. Although thoughts are (implicitly) understood to be categorially different from linguistic utterances—principally insofar as they do not involve any of their physical characteristics, i.e. vocalizations, gestures, inscriptions, etc.—they are not confined within some withdrawn inner recess inaccessible to publicly perceptible space. They are not located anywhere at all; certainly not in speakers’ heads. They are introduced as provisionally unobservable theoretical entities, not “immaterial” entities. Jones’ theoretical innovation proceeds by identifying overt speech with thinking in the same way in which gases are identified with populations of molecules. Thus, thoughts “are ‘in’ language using animals as molecular impacts are ‘in’ gases, not as ‘ghosts’ are in ‘machines’ ” (EPM , 187). Empirical observables

62 Wilfrid Sellars, Idealism, and Realism are identified with theoretical unobservables, without the former being the overt manifestation of the latter’s covert existence. Linguistic utterance is not the exteriorization of thought and thought is not the interiorization of utterance; rather, utterance itself is a “thinking-out-loud” and thinking is a “silent saying.” But once our language has been enriched with these theoretical terms, they can acquire an observational role through which they can be used in perceptual reports. Thus we learn to perceive each other and ourselves as thinking (believing, hoping, fearing, wishing, etc.). Meaning The meaningfulness of overt public utterance provides the model for understanding the meaningfulness of the thought-episodes with which they are theoretically identified. The intentionality of thought is modeled upon the intentionality of language, where this is understood to depend upon the development of metalinguistic resources that allow us to speak about speaking. Speaking about speaking is the condition for thinking about thinking. Semantic vocabulary such as “means,” “designates,” “refers to,” “expresses,” provides us with the conceptual resources we need to think about meaning as a phenomenon exemplified by language. But the question then is whether the meaning of words derives from the meaning of thoughts or vice versa. Sellars’ claim is not that we learned how to think only after we had learned how to speak. Sellars allows for a proto-conceptual language-of-thought (“mentalese”) through which animals represent their environment. Mental representations are not linguistic or conceptual in Sellars’ sense, although they are endowed with propositional form, i.e. a referring and characterizing function.3 Nor is Sellars claiming that thinking depends upon language in an ontological sense.4 This is a metaphysical misinterpretation of psychological nominalism. The Ryleans in Sellars’ mythic parable certainly think but they do not think of themselves as thinking until Jones provides them with a concept of what it is to think. By embracing Jones’s theory of thoughts as inner-episodes of saying, the Ryleans acquire the concept of thinking as what underlies meaningful utterance. They become aware of themselves as thinking. The concept of “meaning to say,” and of “intention” in the psychological and semantic sense, is ushered in with the shift from conscious to self-conscious thought. This is precisely what Jones’s theory of thoughts as inner-episodes is supposed to explain. But it entails that

The Metaphysics of Sensation 63 the intentionality of public language is the condition for the intentionality of thought: [T]he concept of thought essentially involves that of intentionality in the following sense. To say of a piece of verbal behavior that it is a thinking-out-loud is to commit oneself to say of it that it means something, while to say of it specifically that it is a thinking-out-loud that-p, is to commit oneself to say of it that it is a piece of verbal behavior which means p. Thus, at the primary level, instead of analyzing the intentionality or aboutness of verbal behavior in terms of its expressing or being used to express classically conceived thoughts or beliefs, we should recognize that this verbal behavior is already thinking in its own right, and its intentionality or aboutness is simply the appropriateness of classifying it in terms which relate to the linguistic behavior of the group to which one belongs. LTC , 527, my emphasis–RB The intentionality of public discourse is a matter of appropriately classifying linguistic behavior according to standards encapsulated by collectively accepted norms. These norms involve the rules of criticism, or “ought-to-bes,” to which speakers are encouraged to conform, and the rules of action, or “ought-to-dos,” which speakers are expected to realize. Since ought-to-bes imply ought-to-dos, conforming to the former is indissociable from accepting the latter. Linguistic action and linguistic behavior are both norm governed; they cannot be opposed to one another since the capacity to conform is the precondition for the ability to act. If psychological nominalism identifies thinking with verbal behavior, this is not in the same sense in which psychological behaviorism (allegedly) identifies thinking with behavioral dispositions. Verbal behavior expresses thought only insofar as thinking is understood as a kind of doing cognitively modeled on linguistic utterance. Thinking is a kind of doing because it involves inferring, which in turn involves representing—although thinking is not representing and concepts are not representations. Concepts are inferential roles and these roles are defined by networks of implication across webs of linguistic assertion. To say that intentionality is the appropriateness of classifying a linguistic expression relative to the behavior of a linguistic community is to say that intentionality is a functional characteristic. Sellars’ dot-quoting device is the exhibition of this linguistic function. To dot-quote a linguistic expression is to show what it does in the language in which the dot-quoting occurs. This of course presupposes familiarity with the dot-quoted expression’s proper function.

64 Wilfrid Sellars, Idealism, and Realism This familiarity is the background of tacit understanding against which the exhibition of the meaning of particular expressions can stand out as an instance of explicit understanding. Exhibiting the meaning of a specific expression is only possible against the backdrop of a language and a form of life within which the use of words is necessarily embedded. But this exhibition, rather than stating, of complex rule-governed functioning, which presupposes familiarity with the language in which the dot-quoting device is deployed, re-embeds language within a world of practical involvements. Thus Sellars’ psychological nominalism, and his claim that cognitive functioning is linguistically instantiated, envelops awareness within language only insofar as language itself is enveloped in a world comprising a variety of (human) practices and purposes. That this is a world whose structure is constellated around human interests, both practical and cognitive, is the necessary corollary of Sellars’ transcendental approach, which situates itself within the correlation between representing and represented and seeks to draw out its underlying conditions. Drawing out those conditions exposes the chain of nested mediations from awareness to awareness-as, from awareness-as to conceptual role, from conceptual role to linguistic practice, and from linguistic practice to a world of practical involvements. What is exposed through this chain of mediations is the way in which the content of any single state of awareness is bound up with a context of significance that includes a whole practical life-world. Psychological nominalism is usually reproached for denying the reality of mental states. But if anti-realism in this context means denying that mental states belong in the world, then this reproach is misguided, for on Sellars’ account, mental states are as real as anything else in the world of which they are a part. The fact that cognitive awareness of mental states is conceptual through and through in no way compromises their reality, since on Sellars’ transcendental account this is as true of tables, zebras, and galaxies as it is of thoughts and sensations. If the issue is one of realism, the more fitting reproach to psychological nominalism, and to the transcendental stance from which it ensues, would be that its realism, whether about mental states or non-mental phenomena, comes at the price of an exorbitant holism which is obliged to take as given a total context of significance; i.e. a form of life in which the transcendental perspective is necessarily embedded. Why should this be so? On Sellars’ account, the “ought-to-be”s and “ought-todo”s which condition all our thinking and acting constitute the linguistic framework within which we live. This framework furnishes the resources for all justification and evaluation; although claims made within the framework can be

The Metaphysics of Sensation 65 assessed, criticized, or rejected, there is no higher authority or court of appeal with regard to which the framework as a whole could be judged and found wanting. As Wittgenstein put it: “You must bear in mind that the languagegame . . . is not based on grounds. It is not reasonable (or unreasonable). It is there—like our life.” (Wittgenstein 1969 No. 554). If our linguistic framework is simply “there, like our life,” it is not susceptible of rational revision; it is just something to be accepted as it is. Sellars’ emphasis on holism—both semantic and epistemic—would seem to point towards a quietist acceptance of the form of life within which we find ourselves as the ineluctable horizon of all thinking and doing. If the rejection of the empiricist and rationalist versions of givenness ultimately entails accepting the givenness of a form of life as the ultimate horizon circumscribing reflection, this may prove too high a price. Transcendental reflection would find itself constitutively incapable of critically interrogating the legitimacy of the practices—and specifically of the cognitive practices—within which psychological states and conceptual roles are necessarily embedded. Sellars’ conception of philosophy as a theoretical endeavor requiring conceptual construction as well as conceptual revision stands opposed to such quietism. The root of this opposition lies in Sellars’ bifurcated holism. Our understanding of the world is split into two competing conceptual structures, the manifest and scientific images of man-in-the-world.5 Both are autonomous categorial schemes, neither of which can be wholly subsumed within the other. Moreover, Sellars accords to the scientific image the peculiar privilege of describing and explaining why the world appears to us as it does. This is something the manifest image cannot do: transcendental reflection within the bounds of the manifest image reveals normative practice as the fundamental horizon of understanding, but cannot explain why the world manifests itself to us as it does or why we engage in these particular practices but not others – indeed, such questions make no sense so long as transcendental investigation is confined to the manifest image. However, they are perfectly legitimate if one understands the task of transcendental reflection to be that of articulating the two images. Ultimately, “the common sense picture [manifest image] of the world, in spite of its delicate coherence, poses problems which it lacks the resources to resolve” (SK , 310). The topic of sensation is crucial in this regard because it lies at the interface between the two images: it is the juncture between the causal nexus investigated by science and the normative space of reasons within which we conduct our daily lives. Thus, “sensations are essential to the explanation of how we come to construct the appearance which is the manifest world” (PSIM , 36).

66 Wilfrid Sellars, Idealism, and Realism In response to the charge that holism entails quietism, it is worth pointing out that Sellars’ bifurcated holism makes room for conceptual revision because it allows for the possibility of diagnosing categorial dysfunction, not only within each image but also across the two images. This is perhaps the decisive difference between Sellars’ critical holism and the quietist varieties of holism embraced by Wittgenstein or Rorty. Sellars’ controversial proposal for a metaphysics of pure processes exemplifies his critical-constructive conception of philosophy because it is motivated by concern over the malfunctioning of the category of sensation, rather than by dogmatic deference towards the epistemic authority of qualia.6 In order to understand this we have to clarify why the awareness involved in sensation is not the same as the awareness involved in intuition, or more simply, why sensation is not intuition. Intuition and sensation Sensing is not thinking because it does not involve conceptual classification: it is not awareness of sorts, resemblances, facts, or any other sort of abstract entity. This much seems relatively uncontroversial. But where Sellars breaks with empiricist orthodoxy is with his claim that sensory awareness does not qualify as awareness of particulars. To be aware of a particular understood as a “this such,” e.g. “this red brick facing me edgewise,” is a conceptual affair, requiring the intervention of conceptual intuition. What is conceptual intuition? On the traditional account, intuitions deliver particulars which are perceived as being thus and so. Consider the following perceptual report: I see this as a red rectangular brick The traditional view maintains that my beliefs about the brick are distinct from my seeing the brick: I see this as a red rectangular brick and I believe that this a brick with a red and rectangular facing surface Here we have a distinction between intuitive perceptual taking, or “seeing as,” and belief proper, which has propositional form. My seeing this as being thus and so and my belief that it is (or is not) thus and so are distinct, with the latter presupposing the former. This traditional account falls prey to the Myth of the Categorial Given: the assumption that to be aware of X is to be aware of it as X. In this version, the myth

The Metaphysics of Sensation 67 fuses thinking and sensing: it assumes that things present themselves to sensory consciousness already endowed with categorial form. We sense something as something before superimposing onto it our belief that it is thus and so: I see this as a red rectangular brick and I believe it is too big for the job at hand. What is wrong here is the assumption that objects cause us to be in certain sensory states, and these sensory states are already endowed with the categorial form that allows them to play a justificatory role in empirical knowledge. The object causes me to see it as what it is and this justifies my subsequent beliefs about it and its relations to other objects. Causation and justification are illegitimately fused. By separating them, we distinguish between the sensory states which objects cause perceivers to be in, and the perceptual states in terms of which perceivers respond to their sensory states. For these perceptual responses to play a justificatory role in empirical knowledge they must already be endowed with categorial form: they must be seeings-as (or hearings-as, tastings-as, touchings-as, smellings-as). This is to say that they must involve categorially-formed conceptual intuitions of sensible particulars. Once we acknowledge the fundamental role of conceptual intuition in empirical perception, we can subsequently distinguish between what we see of objects, and what we see objects as. What we see of an empirical object is a function of our embodied, perspectival relation to it. But this perspectival relation already presupposes the conceptual intuition of the object as something thus and so. What we can see things as is not limited by what we see of them: we see the red exterior of the apple as containing white flesh, and the white flesh is imagined as co-present along with the red exterior, yet we do not see of the apple its white flesh.7 What is intuited is never a bare particular; rather, it is a condition of our ability to intuit particulars that they be conceptually intuited as internally complex. They are endowed with conceptual content and grammatical form. In his reconstruction of Kant, Sellars explains why conceptual intuition is an epistemically irreducible type of representation: Consider the statements: This is a pyramid This pyramid is made of stone The first has the explicit grammatical form of a sentence. So does the second. But notice that the grammatical form of a sentence is lurking in the subject of the second sentence. From the standpoint of transformational grammar we would think of it as derived from the deep structure: This is a pyramid and it is made of stone

68 Wilfrid Sellars, Idealism, and Realism One might be tempted to think of “this” as a pure demonstrative having no other conceptual content than that involved in being a demonstrative. Kant does think of an act of intuition as a demonstrative thought, a mentalese “this.” However, he does not think of this mentalese demonstrative as a bare mentalese “this.” An example of an act of intuition would be the mentalese counterpart of: This cube facing me edgewise where this is not to be understood as, so to speak, a mentalese paraphrase of: This is a cube which faces me edgewise The role of an intuition is a basic and important one. It is the role of bringing a particular object before the mind for its consideration. Thus, though there is a close relationship between: This cube facing me edgewise . . . and This is a cube which faces me edgewise the former is an irreducible kind of representation. It is a demonstrative representation which has conceptual content and grammatical form. As noted above it contains the form and content of the judgment “This is a cube.” Thus for Kant, intuitions are complex demonstrative thoughts which have implicit grammatical (and hence categorial) form (IKTE , 428–9). Intuitions are conceptually formed “this-suches”: “this cube facing me edgewise,” “this stone pyramid,” etc. To say that conceptual intuitions are irreducible representations is to say that they provide the fundamental data for perceptual experience and that they deliver the ultimate subjects of predication for empirical judgments.8 It is because intuitions are representations endowed with conceptual content and grammatical form that they can play this fundamental role in empirical knowledge. Thus perception cannot be decomposed into the sensing of bare particulars coupled with propositionally structured beliefs about those bare particulars. What is intuited is categorially determined and thus already available for propositionally structured belief. How then do intuitions relate to sensations or what Sellars calls “senseimpressions”? Our perceptual reports are primarily about physical objects; they are not about the sensory states caused by those objects and responded to by our reports. We perceive physical objects as thus and so and we deliver perceptual reports about those objects, although those reports are in part responses to the

The Metaphysics of Sensation 69 sensory states caused by those physical objects. But according to the Jonesean theory, just as our psychological vocabulary is modeled on candid public speech, so our sensory vocabulary is modeled on the publicly perceptible properties of physical objects.9 We must have mastered that vocabulary before we could subsequently learn to deploy it to make perceptual reports about our own sensory states. Thus the ability to perceive our own sensations presupposes the ability to perceive publically accessible objects and the properties of senseimpressions are the postulated counterparts of the properties of physical objects. Whereas physical objects are particulars, sense-impressions are postulated as states of sentient organisms. They are postulated to explain the experiential content common to veridical perceptions and the kinds of “ostensible perceiving” exemplified by perceptual illusions and hallucinations. Sense-impressions are what perceivers are responding to when they ostensibly perceive things that are not physically present. As states of sentient organisms, sensations cannot be decomposed into a relation between sensing and sensed. The notion that the sensation of a red rectangle involves a relation to a red and rectangular sense datum follows from the philosophical reification of sensory states into hybrid particulars comprising physical and mental characteristics. Qualia and sense data are reifications of sensory states; a reification that results from conflating the analogical properties of theoretical postulates with the empirical properties of physical entities. Sellars rejects reification and opts for a non-relational account of sensation. Rather than postulate sense data as immaterial relata for acts of sensing, Sellars adopts an adverbial account whereby the sense impression of a red rectangle is not a relation between a sensing and a red rectangular sensum but a sensing red rectangularly. This adverbial characterization of sensation allows sensings to be described as ways in which perceivers sense, rather than as relations between perceivers and non-physical particulars. Since sense impressions as states of perceivers are the analogical counterparts of the physical particulars that serve as their models, the model features a commentary specifying the disanalogies between model and copy. The sensation of a red rectangle can be characterized as a “sensing red rectangularly” even though it is understood that strictly speaking there is no red and rectangular state of the nervous system in the organism having the sensation. In what sense then can sensations be adverbially characterized as sensing red rectangularly? This is to ask whether it can ever be literally true within the manifest image to say of someone that they are having a sensation of a red rectangle. Since sensations play a causal role in explaining our perception of the world, they must be accounted for. Characterizing sensations as states of perceivers

70 Wilfrid Sellars, Idealism, and Realism explains why we seem to perceive or believe we are perceiving something in cases where the physical objects standardly responsible for those perceptions are not present. If sensations play a causal role in explaining what perceivers say or believe within the manifest image, it must be possible to make true claims about them within the confines of that image. Once we reject the Myth of the Given, together with the philosophical reification of sensory states and the dualism that ensues from it, the problem is to explain what the counterparts of redness and rectangularity are when we sense “red rectangularly.” Two options present themselves: either identify the properties of sensations with physical properties of the central nervous system (which means accepting that neurophysiology tells us what sensations really are); or deny that such physical identification is possible and embrace mindbody dualism. It is important to note that thoughts do not present the same obstacles to reductive identification as sensations. Since thoughts can be characterized in terms of their functional roles, Sellars insists that “there is no barrier in principle to the identification of conceptual thinking with neurophysiological process” (PSIM , 34). Thus the conceptual role of “red rectangular” is identified through dot-quoting as the metalinguistic sortal •red rectangular•, and this linguistic function is (at least in principle) identifiable with the set of physical states in which its particular tokenings are realized. But although a sensation can be functionally characterized as a “sensing red triangularly,” it cannot be identified with its functional role. Perceptual form and perceptual content To understand why sensation cannot be identified with its functional role we must distinguish between the formal and content properties of objects. Formal properties include all of an object’s causal, dispositional, and structural properties: they are conceptual. Content properties comprise the object’s sensible properties in the strict sense: whatever can only be directly seen, touched, heard, tasted, or smelled of an object. While the former are categorial features, the latter are purely occurrent qualities: [W]e do not perceive of the object its causal properties. What we see of it are its occurrent sensible features. This can now be generalized as follows. We do not perceive of the object its character as a substance having attributes, its character

The Metaphysics of Sensation 71 as belonging with other substances in a system of interacting substances, its character as conforming to laws of nature. In short, we do not perceive of the object what might be called “categorial” features. IKTE , 427 What we see of an object are its sensible, content properties; what we see an object as are its conceptual or categorial features. The latter are intuited, not sensed. While formal properties can be functionally characterized and physically identified, content properties possess intrinsic qualitative features that resist functional characterization and physical identification.10 Thus sensation is endowed with a metaphysical significance because it is keyed in to what Sellars calls the “ ‘occurrent content” of physical objects: It should be noted that if physical objects are genuine individuals, they can scarcely have only iffy properties (powers, propensities, dispositional properties, and the like). They must have some non-dispositional or occurrent attributes. Nor, as Whitehead reminds us, will it do to limit their occurrent attributes to such “primary” qualities as shape and size; for, to use an Aristotelian turn of phrase, geometrical qualities are “formal” qualities and presuppose a “content” or “matter”, thus color. Things which had “primary” qualities without “content” qualities would have “vacuous actuality.” That Whitehead construed the occurrent content of physical objects in terms of feeling, rather than color, is a symptom of the revisionary character of his metaphysics. SK , 302 Whitehead construes occurrent content in terms of “feeling” because he is a panexperientialist for whom actuality divorced from subjectivity is necessarily vacuous. A vacuous actuality in Whitehead’s sense is a non-subjective actuality, i.e., an un-perceived actuality wholly independent of subjective experience. At the root of Whitehead’s pan-experientialism lies a conflation of sensing with perceiving, awareness with awareness-as. Sellars is too much of a Kantian to accept Whitehead’s revisionary recategorization of occurrent physical content as an instance of feeling. He endorses the traditional characterization of the metaphysical task as the identification of ultimate categories understood as summa genera. “Categories in general,” says Sellars, “are classifications of conceptual roles” (FMPP, 19). Like Kant, Sellars views categories as ways of classifying conceptual representings rather than attributes of unrepresented things. Thus he insists that metaphysics must accurately describe our extant categorial framework, i.e. our extant ways of

72 Wilfrid Sellars, Idealism, and Realism representing things, before engaging in the revisionary task of postulating new categories or re-categorizing entities. From the viewpoint of descriptive metaphysics, the category of actuality is not subordinate to the category of subjectivity as species to genus; they are distinct genera. Occurrent actuality manifests itself to us through sensation, and sensations are occurrent states of perceiving subjects, but they need to be conceptually apperceived, i.e. represented, to be known by the subject whose states they are. Although occurrent contents are revealed through sensation, sensation itself is not the cognitive awareness of occurrent contents. This is what prevents Sellars from embracing Whitehead’s pan-experientialism, as well as the kind of property dualism championed by Chalmers. It is also why Sellars’ “grain argument” for the irreducibility of content qualities is not phenomenological; at least not as that term is used by analytic philosophers of mind.11 The grain argument The grain argument can be summarized as follows. According to Sellars’ “principle of reducibility”: If an object is in a strict sense a system of objects, then every property of the object must consist in the fact that its constituents have such and such qualities and stand in such and such relations or, roughly, every property of a system of objects consists of properties of, and relations between, its constituents. PSIM , 27 If perceivers are ultimately identical with systems of microphysical particles, then every property of perceivers must be explicable in terms of the properties and relations of their microphysical constituents. Now, perceptions are caused by sensations, which are states (i.e. properties) of perceivers. But sensations are occurrent content properties. (Since the content properties of manifest physical objects can occur without being perceived, so can sensory content properties, which is why sensory awareness is not awareness that one is sensing or of what one is sensing. Actuality is not bound to subjectivity—this is Sellars’ rejoinder to Whitehead.) Consequently, the properties of sensation cannot be explained in terms of the properties and relations of the microphysical constituents of the nervous system since (ex hypothesi) these are all formal properties. Occurrent content properties feature an “ultimate homogeneity” which is not explicable in terms of the formal

The Metaphysics of Sensation 73 properties and relations of the constituents of the nervous system. A manifest pink ice cube, writes Sellars, “presents itself to us as something which is pink through and through, as a pink continuum, all the regions of which, however small, are pink. It presents itself to us as ultimately homogeneous” (PSIM , 26). Since the homogeneity of occurrent pinkness is not explicable in terms of the formal properties and relations of the ice cube’s microphysical constituents, it is irreducible. Moreover, since Sellars appears to hold that ultimate homogeneity is a characteristic of all occurrent properties, and since he insists that the properties of sensation are occurrent properties, he concludes that sensations are irreducible to physical states. Stated in this way, the argument is as perplexing as it is straightforward. The claim that occurrent pinkness “presents itself ” as ultimately homogeneous seems to be a particularly egregious regression to the Myth of the Given. The argument also appears to hinge on the presumptive identification of properties that are only supposed to be analogically correlated: the properties of sense-impressions were supposed to be modeled upon the content properties of manifest physical objects, not identified with them. Last but not least, the argument seems to require that imperceptible entities can only have formal properties, not content properties. Since Sellars rejects the (Whiteheadian) claim that occurrent content properties must be perceived, it is hard to see why he should think imperceptible entities lack occurrent properties. To see why he does not, and why the “grain argument” is neither regressive nor careless, we must draw upon Richardson and Muilenberg’s patient and illuminating reconstruction of it. In their reconstruction, Richardson and Muilenberg distinguish formal analogies from material analogies.12 Whereas material analogies correlate properties of individuals, formal analogies correlate properties of properties, i.e. second-order properties. Clearly, the Jonesean theory of sensation proposes a formal analogy between manifest physical particulars and senseimpressions. This is to say that it establishes a correlation between the second-order properties of perceptible physical objects and sensations. Thus the respect in which a sense-impression is like a physical particular is not that of first-order object properties, such as redness or rectangularity, but the properties of these properties. If the analogy between manifest physical particulars and sense-impressions is formal, rather than material, then the homogeneity that Sellars ascribes to the ice cube’s occurrent pinkness is a second-order property of the colored pink expanse; it is not itself an occurrent property somehow present in the pink; a property that could be perceived in the same way in which one perceives the pinkness of the ice cube. The ice cube, says Sellars,

74 Wilfrid Sellars, Idealism, and Realism presents itself to us as ultimately homogeneous; and an ice cube variegated in color is, though not homogeneous in its specific color, “ultimately homogeneous”, in the sense to which I am calling attention, with respect to the generic trait of being colored. PSIM , 26, my emphasis–RB Homogeneity pertains to the generic, i.e. categorial, trait of being colored, not to occurrent pinkness as such. The pink is seen as homogeneous insofar as it is a color, but homogeneity is not something one sees of the ice cube. More precisely, it is the coloredness of the ice cube that is seen as homogeneous, rather than its occurrent pinkness. Homogeneity is not an occurrent content property but a formal-generic property of pink as a species of color. And as we saw earlier, Sellars insists that we do not perceive the categorial features of objects. Crucially, although content properties are categorially different from formal properties, the properties of content properties need not be content properties. Moreover, while content properties can have formal properties, the reverse does not hold because content properties are necessarily first-order properties. This is why Sellars believes they provide a clue to the fundamental material or “stuff ” of the world, even if they cannot be straightforwardly identified with it. The occurrent sensory qualities present in perception (what we see of things) provide the epistemic link between the empirical properties of perceptible things and the content properties of their imperceptible constituents. If there is a sense in which the grain argument is phenomenological, it is the classical Husserlian sense in which phenomenological description is a transcendental exercise rendering visible the invisible categorial structure latent in everyday perceptual experience. Sellars’ claim about homogeneity does not rely upon an empirical description of how things seem to the subject of experience. What motivates the grain argument is not antipathy to reductionism but a worry about categorial dysfunction both within the manifest image and across the two images, i.e. manifest and scientific. The occurrentness of sensations must be preserved because it is what allows them to exercise a causal function and bring about perception and action. Within the manifest image, sense-impressions are invoked to explain otherwise puzzling features of our behavior. It is important to account for them without resorting to mind-body dualism; not only because such dualism is the result of a philosophical misprision of the categorial structure of the manifest image, but also because it is the biggest obstacle to the stereoscopic integration of the manifest and scientific images. Sensation is destined to play a key role in

The Metaphysics of Sensation 75 the synoptic fusion of the manifest and scientific images: as the link between the mental and the physical (or reasons and causes), it can help science explain why the world appears to us as it does. Sellars’ argument is not anti-reductionist per se; rather, it is directed against a peremptory reduction which, because it elides the connection between sensible qualities and content properties, threatens to jeopardize the cognitive integration of the two images. Homogeneity and process The elimination of homogeneity as a categorial feature of sensory content vitiates our understanding of the occurrentness of sensation and hence of its causal efficacy. Having identified what he thinks is omitted in the attempt to reduce sensory qualities to the properties of microphysical particulars, Sellars believes he has a rationale for a revisionary metaphysical proposal. This proposal is intended to correct the transcendental dysfunction afflicting the category of sensation. As Sellars sees it, what is required is a theoretical counterpart for occurrent sensory content among the properties and relations of the imperceptible constituents of organisms. Thus Sellars postulates “sense” as the imperceptible counterparts of occurrent sensations. It is important to note that this is a postulation within the manifest image; albeit one that Sellars hopes will be eventually corroborated within the scientific image. Nevertheless, the introduction of sensa requires amending one fundamental component of the scientific image: its commitment to a particulate paradigm of fundamental physical entities. If sensa are to be understood as properties of fundamental physical entities, the latter cannot be elementary particles as currently conceived within the Standard Model since these are devoid of homogenous content properties. Sellars fastens upon the manifest category of “process” as a model for the physical entities in which sensa will feature as “ingredients.” “Pure” or “absolute” processes are distinguished from “object-bound” processes in that they cannot be attributed to objects. While running has an attributive status in expressions like “Socrates runs,” thundering is not attributed to anything when we say “It thunders.” Absolute processes are said to be “subjectless” insofar as they cannot be attributed to an underlying substratum. Their other significant characteristic for Sellars’ purpose is that they are characterized in terms of an intrinsic qualitative aspect that, although associated with their typical causes, cannot be identified with those causes. Thus while

76 Wilfrid Sellars, Idealism, and Realism buzzing is commonly associated with bees and can be attributed to their activity, the qualitative aspect proper to buzzing is independent of the activity that generates it. Buzzing can be generated by many different causes, but the qualitative aspect that makes of it a buzzing (rather than a fizzing or a roaring) remains the same: [W]hat is primary in the various senses of the verb “to buzz” is the concept of the intrinsic character of a certain kind of process which can be identified in terms of its typical causes. The verb “to buzz”, then, would have a sense in which processes of that intrinsic kind would be buzzings, even when they were not being brought about by one of these typical causes. Thus, in this sense of the verb “to buzz” we could say that a buzzing is going on without implying that some object, e.g., a bee, is buzzing. FMPP, 60 This is to say that buzzing as an occurrent content quality is irreducible to the formal properties of its constituents (e.g. the activity of bees). The concept of an absolute process is the concept of an intrinsic quality occurring independently of the properties of objects or relations among objects. This occurrent aspect is decisive not only because it allows us to conceptualize content qualities independently of any objects that might bear them, but also because it provides us with a paradigm of an absolute occurring; an occurring that is not attributable to something that begins and ceases to occur: In the case of absolute processes we can speak of absolute coming to be and ceasing to be, because when a sounding, e.g., a C#ing, begins, there is nothing which begins—in the relevant sense—to sound. FMPP, 51 Absolute processes combine three basic characteristics: intrinsicality, occurrentness, and homogeneity. Unsurprisingly, these are precisely the characteristics Sellars also ascribes to sensa. Yet the two are distinct: while sensa occur as components of absolute processes traversing the nervous systems of sentient organisms, not all absolute processes involve sensa; moreover, the characteristics of processes are defined without any reference to organisms. But can sensa be stipulated to occur only “in” the nervous systems of sentient organisms without lapsing into dualism? Only if this localization is explicable in physical terms. Sellars distinguishes between a generic and a specific sense of the adjective “physical,” for which he reserves the terms “physical1” and “physical2” respectively. An entity is said to be physical1 if it exists in the framework of space and time; an entity is said to be physical2 if it is “definable in terms of theoretical

The Metaphysics of Sensation 77 primitives adequate to describe completely the actual states though not necessarily the potentialities of the universe before the appearance of life” (CE , 252). The sense in which entities are described as physical2 is obviously a restriction of the broader generic sense in which any entity existing in spacetime is physical1. Thus physical2 predicates are not sufficient to characterize all the properties of living things. Sensa and absolute processes are obviously physical1; but if the distinction between these two senses of physical is not to turn into an alibi for vitalism or pan-experientialism it is necessary to insist that physical2 entities are also composed of absolute processes. Sellars calls the latter ϕ2 processes and jokingly proposes “electronings” and “quarkings” as possible examples. In a similar vein, the absolute processes constituent of sensation, such as reddings, buzzings, and C#ings, can be called σ-ings. With these distinctions in place, it becomes possible to sketch the relation between sensa and physical1 processes in terms of the ingredients of σ-ings within physical2 objects: [W]hereas the objects of contemporary neuro-physiological theory are taken to consist of neurons, which consist of molecules, which consist of quarks – all physical2 objects—an ideal successor theory formulated in terms of absolute processes (both ϕ2-ings and σ-ings) might so constitute certain of its “objects” (e.g., neurons in the visual cortex) that they had σ-ings as ingredients, differing in this respect from purely physical2 structures. FMPP, 124 Reddings, buzzings, and C#ings would be construed as constituents of nervous systems alongside neurons, and the particulate entities of which neurons are composed—i.e. molecules and quarks—would be understood as objects abstracted from underlying physical2 processes such as quarkings and electronings. In the final analysis however, both quarkings and reddings would be physical1 processes. Concluding remarks While vitalism insists that living things are fundamentally different in kind from and irreducible to non-living things, Sellars’ monism of absolute processes proposes a common basis for living and non-living things. It also suggests that both living and non-living things are composed out of this fundamental element. The ingredience of physical1 processes within physical2 objects, and of σ-ings within physical1 processes, is supposed to stave off vitalism and

78 Wilfrid Sellars, Idealism, and Realism pan-experientalism. However, certain questions remain as to the definition of absolute processes, particularly the relation between homogeneity, intrinsicality, and occurentness. As Richardson and Muilenberg point out, there is a distinction to be made between the reducibility of structural properties, e.g. being a ladder, and the reducibility of content properties, e.g. being colored.13 The property of being a ladder is reducible because it consists in the fact that the parts of a ladder—the rungs, the frame, etc.—have certain properties and are related to each other in certain ways. Obviously, not every part of a ladder is a ladder. By way of contrast, the property of being a red brick wall consists in the fact that each part of the wall, i.e. each brick, is itself red. In this instance, being a red brick wall is a reducible property only if each of its parts, i.e. each brick, is also red. Content properties require homogeneity from whole to part in order to be reducible. However, in Sellars’ account, while occurrent pinkness must be taken as a content property, its homogeneity seems to be a structural property (structural features are sufficiently akin to formal properties to allow this equivalence). According to the grain argument, the reason why we cannot apply the principle of reducibility to occurrent pinkness considered as a content property is because it cannot be attributed to each physical constituent of the organism, since we know that the physical particles composing organisms cannot be colored. But if homogeneity is in fact a structural-categorial feature of occurrent sensation, then it need not be attributable to the organism’s component parts to be explicable in terms of their properties and relations. Thus while the reducibility of occurrent pinkness qua content property requires its attribution to the constituent parts of the object bearing the property, the reducibility of the homogeneity of occurrent pink does not. There seems to be an equivocation between occurrentness and homogeneity lurking in Sellars’ account of sensation: while homogeneity is a property of occurrent properties, such as sensations, it is not itself an occurrent property, i.e. an intrinsic feature of sensation. To qualify a property as occurrent is also to qualify it as homogeneous, since the property of homogeneity is entailed by the property of occurrentness; but the converse does not hold: to qualify a property as homogeneous is not to say that it is occurrent. The intrinsic qualitative particularity that is central to Sellars’ characterization of absolute processes is clearly tied to the property of occurrentness; but its connection to homogeneity is less clear. Thus while the formal nature of homogeneity can be used to rebuke those critics of Sellars who mistake it for a perceptible property, it is also what prevents it from being inherent in occurrentness in the way that Sellars’ argument seems to require.

The Metaphysics of Sensation 79 Notes 1 2 3 4 EPM , 128. See EPM , sections X–XVI , 175–96. See ME , 336. The postulate of a language of thought—to which Sellars is not unsympathetic, and with which his own verbal behaviorism is surprisingly compatible—claims that thinking has a linguistic structure, but that this structure is not that of a natural language. Verbal behaviorism in Sellars’ sense does not rule out the possibility that the capacity for overt public speech depends upon the existence of an inner language of thought. What it does claim is that our initial conceptualization of thinking is modeled upon overt public speech, even if subsequent investigation of the nature of thinking should reveal that competence in a natural language depends upon the operation of an inner language of thought. Familiarity with the structure of public language will still be prior to familiarity with the language of thought in the order of knowing, even if the language of thought (or mentalese) is prior to public language in the order of being. 5 Note that the bifurcation is in our conceptual grasp of the world, not in the world as such. Sellars’ transcendental dualism of images should not be taken to be incompatible with ontological monism. 6 Dennett’s rejection of Sellars’ case for the irreducibility of sensation seems to assume that the argument is empirical-phenomenological. See Dennett 1981. I side with Seibt, as well as Richardson and Muilenberg, in taking Sellars’ argument to be transcendental-categorial. See Seibt 1990, 233–70; and Richardson and Muilenberg 1982. The issue is not about what is or is not empirically obvious but about whether and how occurrent sensory qualities can have causal powers. 7 See Sellars IKTE §12–24, 421–3. 8 This is not to say that conceptual intuitions deliver ultimate subjects of predication in the metaphysical sense. Sellars is careful to distinguish between complex particulars, exemplified by perceptual “this suches,” and the simple or ultimate particulars sought for in metaphysical discourse. See Sellars LCP. 9 See EPM §60–2, 190–5. 10 This is essentially how Chalmers characterizes the basic difference between states of awareness and states of consciousness: see Chalmers 1995 and chapter 1 of Chalmers 1996. 11 See Sellars PSIM , V–VII , 25–38. 12 Richardson and Muilenberg 1982, 189–90. 13 See Richardson and Muilenberg 1982, 177–8.

80 Wilfrid Sellars, Idealism, and Realism Bibliography Chalmers, David (1995), “Facing up to the Problem of Consciousness,” Journal of Consciousness Studies vol. 2 no. 3: 200–19. —— (1996), The Conscious Mind: In Search of a Fundamental Theory, New York and Oxford: Oxford University Press. Dennett, Daniel (1981), “Wondering Where the Yellow Went,” The Monist 64: 102–8. Richardson, R. C. and G. Muilenberg (1982), “Sellars and Sense-Impressions,” Erkenntnis vol. 17 no. 2: 171–211. Seibt, Johanna (1990), Properties as Processes: A Synoptic Study of Wilfrid Sellars’ Nominalism, Atascadaro: Ridgeview Publishing Company. Sellars, Wilfrid BBK —— (1991), “Being and Being Known,” in Science, Perception and Reality., Atascadero: Ridgeview Publishing, 41–59. CE —— and P.E. Meehl (1956), “The Concept of Emergence” in H. Feigl and M. Scriven (eds.), Minnesota Studies in the Philosophy of Science: Vol. I: The Foundations of Science and the Concepts of Psychology and Psychoanalysis, Minneapolis: University of Minnesota Press. 239–52. EPM —— (1991), “Empiricism and the Philosophy of Mind,” in Science, Perception and Reality, Atascadero: Ridgeview Publishing Company, 127–96. FMPP —— (1981), “Foundations for a Metaphysics of Pure Process,” The Monist 64: 3–90. IKTE —— (2002), “The Role of Imagination in Kant’s Theory of Experience,” in J.F. Sicha (ed.), Kant’s Transcendental Metaphysics, Atascadero: Ridgeview Publishing Company, 419–30. LCP —— (2005), “On The Logic of Complex Particulars” in J. Sicha (ed.) Pure Pragmatics and Possible Worlds: the Early Essays of Wilfrid Sellars, Atascadero: Ridgeview Publishing Company, 140–67. LTC —— (1969), “Language as Thought and as Communication,” Philosophy and Phenomenological Research, vol. 29, no. 4: 506–27. ME —— (1981), “Mental Events,” Philosophical Studies, vol. 39, no. 4: 325–45. PSIM —— (1991), “Philosophy and the Scientific Image of Man,” Science, Perception and Reality, Atascadero: Ridgeview Publishing Company, 1–40. SK —— (1975), “The Structure of Knowledge,” in Hector-Neri Castañeda (ed.), Action, Knowledge, and Reality: Critical Studies in Honor of Wilfrid Sellars, Indianapolis: Bobs Merrill Company, 295–346. SM —— (1968), Science and Metaphysics: Variations on Kantian Themes, London: Routledge & Kegan Paul.

The Metaphysics of Sensation SRLG 81 —— (1991), “Some Reflections on Language Games,” in Science, Perception, and Reality, Atascadero: Ridgeview Publishing Company, 321–58. TTC —— (2002), “Towards a Theory of the Categories,” in J.F. Sicha (ed.), Kant’s Transcendental Metaphysics, Atascadero: Ridgeview Publishing Company, 321–39. Talhouk, Omar (2014), “Natura Naturans: The Concept of a Nature for Representation,” MA thesis, American University of Beirut. Wittgenstein, Ludwig (1969), On Certainty, G.E. Anscombe and G. H. von Wright (eds.), Oxford: Basil Blackwell.