CHAPTER 8

THE GIVEN

      The linchpin of Sellars' materialistic system is his claim that the given (in some sense) is a myth. He is quite aware of the importance of his views on the given to the rest of his philosophy when he writes: "The views [about the myth of the given] I expressed are so central to my way of thinking that if they were to fall apart the result would be a shambles."{1} He is also aware of the importance of the given to most other philosophies. When he reflects on the history of philosophy, he finds that the belief in a given has been a pervasive presupposition:

This framework [of the given] has been a common feature of most of the major systems of philosophy, including, to use a Kantian phrase, both 'dogmatic rationalism' and 'sceptical empiricism'. It has, indeed, been so pervasive that few, if any, philosophers have been altogether free of it; certainly not Kant, and, I would argue, not even Hegel, that great foe of 'immediacy'.{2}

      Most philosophers are faulted for embracing one or another form of the myth, and Sellars' own philosophical constructions are devised to skirt it.

      The view that the given is a myth has been popularized by Sellars, and it may well be one of his distinct contributions to philosophy. This is acknowledged both by his sympathizers and critics. The following comment by Richard Rorty is probably shared by most students of Sellars: "Sellars's attack on the Myth of the Given seemed to me to render doubtful the assumptions behind most of modern philosophy."{3}

      The function of the given, as Sellars understands it, is to provide a "foundation" for knowledge:

the epistemological category of the given is, presumably, to explicate the idea that empirical knowledge rests on a 'foundation' of non-inferential knowledge of matter of fact.{4}

      As an epistemological category, the given is meant to be a piece of elementary information playing an incorrigible evidential role in epistemic justification. But before I examine this role, we must be clear about the nature of the given.

I. DISAMBIGUATION OF THE GIVEN IN SELLARS

      When we look for a clarification of the concept of the given in Sellars, we do not get one clear unambiguous concept; but what amounts to a vacillation between a cluster of concepts. Something like this is admitted by Sellars: "in general, the scope of the concept of the given was ill-defined."{5} My first task, therefore, will be to distinguish the different senses of the "given" in Sellars' writing.

      Two widely different treatment of the given occur first in dealing with types of judgments, and second with the nature of concept acquisition, which in its turn contains a further ambiguity. But despite these ambiguities, all the discussions about the given have relevance to the foundational theory of knowledge. Sellars, in talking about the given, has at least two targets in mind: he is interested in rebutting particular claims about the certainty of some judgments, and in rebutting claims about concept acquisition. There is thus this initial ambiguity in Sellars. Is the given a theory of concept acquisition or a theory of the justification of judgments? These two problems are, as I see it, independent. A thesis about concept acquisition may have no implications about the certainty or uncertainty of judgments. If some concepts are given, this may mean that they are either innate or abstracted; and if some judgments are given, this means that they are non-inferentially known in some way and are certain or not.

      Sellars position on these two issues is that no concepts are either innate or abstracted, and that no judgments are known with certainty. My position, by contrast to Sellars', is that there is a given in both these senses. The difference between us concerns the nature of cognition and its relation to a conventional language. Specifically the question before us is whether the possession of a conventional language is necessary to cognition. The superficial answer is that Sellars affirms this, and I deny it. It is a superficial answer because of the need to distinguish between rs-cognitions and conceptual cognitions. My claim is that conventional language learning presupposes a pre-conventional representational system (Lewis' 'sense meanings') which I have been abbreviating as RS. RS presupposes the existence of an innate capacity to abstract and to interrelate experiences. This is my alternative to Sellars' theory of concept acquisition. As concerns the given of judgments, this is a more complicated matter requiring, as I have shown, a transcendental argument. Put succinctly my claim is that rationality presupposes the reliability of some judgments. But, again, as I said, the two issues are independent. The first concerns the theoretical foundations needed to explain the possibility of learning; the second concerns the epistemic problem of justification of claims. If an RS exists -- as Sellars has conceded in his latest writings -- then it can provide an epistemic given, but not in the sense of self-justifying information, but in the sense of information that does not normally require justification, though which can receive a transcendental justification.

A. THE GIVEN IN JUDGMENTS: PARTICULARS OR FACTS?

      Sellars understands a foundationalistic view of knowledge as entailing a belief in a non-inferential incorrigible type of knowledge. This type of knowledge has normally been identified with phenomenal knowledge. And if that is what the given is, then Sellars poses the following dilemma for any foundationalistic approach that relies on this type of given in knowledge. He says that either particulars or facts are given. If particulars are given, then

(a) It is particulars which are sensed. Sensing is not knowing. The existence of sense data does not logically imply the existence of knowledge, or (b) Sensing is a form of knowing. It is facts rather than particulars which are sensed.{6}

      My view is that alternative (a) does not provide a foundation to knowledge since sensing a particular is a non-epistemic occurrence. So if foundationalism is to be a viable possibility, some version of (b) must be defended. If 'S senses K' is to be an epistemic fact by implying 'S knows K' (in the so called knowledge by acquaintance), the sense of 'knows' in this context, Sellars states, must be stipulated so that, for example, 'S knows K' means 'S knows that this here now is K'. Sellars' position is that the given, if it exists as knowledge, must be a linguistically formulated self-justifying information, having the form 'S knows that this here now is K'. But, according to Sellars, there is no such linguistically formulated certain information; therefore, there is no given in this sense. On the other hand, if the given existed in some other way, it would be certain information occurring non-linguistically. And since it is impossible, for Sellars, that there be such information independent of language, the given, in this sense too, is a myth.

      I am in agreement with the formulation as presented. But this formulation is misleading by talking about knowledge of kinds (sortals) and omitting talking about a perceptual knowledge of phenomenal properties and relations. Let us use 'P(ph)' for phenomenal properties, and 'R(ph)' for relational properties, then the foundationalist claims could be expressed as 'S sees that this here now is P(ph)' or 'S sees that this here now which is P(ph) is next to that there now which is P'ph'. And though these claims are expressed linguistically, they correlate to beliefs which occur non-conceptually -- but representationally.

B. THE GIVEN OF CONCEPTS

      At other times Sellars identifies the given with a view about concept acquisition. In formulating the common presupposition of Rationalism and Empiricism, he writes:

Actually there are various forms taken by the myth of the given in this connection, depending on other philosophical commitments. But they all have in common the idea that the awareness of certain sorts -- and by 'sorts' I have in mind, in the first instance, determinate sense repeatables -- is a primordial, non-problematic feature of 'immediate experience'.{7}

and,

Locke, Berkeley, and Hume . . . all take for granted that the human mind has an innate ability to be aware of certain determinate sorts -- indeed, that we are aware of them simply by virtue of having sensations and images.{8}

Sellars calls this position 'concept empiricism', which consists of the following dogma:

(CE) All awareness of universals is derived from the awareness of instances.

      Thus, for example, according to this dogma, a word such as 'red' is learned by associating it with an awareness of redness. His own alternative is called 'psychological nominalism', which, by contrast, consists of the thesis that:

(PN) All awareness of sorts, resemblances, facts, etc., in short all awareness of abstract entities -- indeed, all awareness of even particulars -- is a linguistic affair.{9}

      Analogously to 'concept empiricism', Sellars could have coined the phrase 'concept rationalism' to label all those views which believe in innate concepts, such as the views of Descartes, Leibniz, Spinoza, and Kant. And by contrast to 'concept rationalism', he did write consistently with (PN) that "there is no awareness of logical space prior to, or independent of, the acquisition of a language." {10}

      The common element of both 'concept empiricism' and 'concept rationalism' is the view that man possesses, as Richard Rorty put it, a mental eye through which the mind is impressed with empirical and abstract concepts. According to Sellars, the proponents of the given, have maintained that this ability to be conceptually impressed, "presupposes no learning, no forming of associations, no setting up of stimulus-response connections."{11}

      I am in agreement with Sellars on all these points. We have to be cautious in understanding that the word 'awareness' is being used in the sense of 'conceptual awareness'. Once this is understood, Sellars' theses are all corollaries of what it means to have a concept.

C. FURTHER AMBIGUITY IN THE CONCEPT OF THE GIVEN

      Having stated in a general way Sellars' position and my disagreement, I will give a more detailed account of what Sellars understands by 'the given' because there is further ambiguity in his explanations, and I will indicate which of these views about the given I accept and which ones I don't.

      The ambiguity is in his view about concept acquisition. I will distinguish, what I will call, an explicit and an implicit version. I will agree with Sellars that the explicit version is a myth, but I will disagree that the implicit version is a myth as well. And if, as I will argue, the implicit version can serve foundational ends, then Sellars has not refuted one form of foundationalism.

II. NARROW (EXPLICIT) SENSE OF THE GIVEN

      Sellars explicit (self-conscious) attacks on "the given" are attacks on what he also calls the propria persona principle, which reads:

(PP) If a person is directly aware of an item which has categorial status C, then the person is aware of it as having categorial status C.{12}

He writes: "This principle is, perhaps, the most basic form of what I have castigated as 'The Myth of the Given'"{13}and adds,

To reject the Myth of the Given is to reject the idea that the categorial structure of the world -- if it has a categorial structure -- imposes itself on the mind as a seal imposes an image on melted wax.{14}

      If we recall that ontological commitment is for Sellars determined by names (logical subjects) and not by the predicate terms, the question of the categorial status required by (PP) should be the question of finding the category of the subject of predication. This is clear when we take note that his nominalistic endeavors are to paraphrase sentences in which the subject terms have an apparent references to abstract entities in such a way that the seeming reference disappears. Predicates have no referential function for him, but are tied conceptually to the nature of the subject phrase. This dependence of predicates comes forth in the following passage in which Sellars, trying to spell out the "descriptive content" of perceptually reductive reports, claimed that at most what can be said within the Manifest Image about this content is this:

(PP1) When one ostensibly sees an object which is red and triangular on the facing side, something, in some way red and triangular is in some way present to the perceiver other than as thought of.{15}

      In this passage Sellars is trying to express a categorial neutrality about the nature of the subject. The "something" is simply a place holder for an unspecified subject. It is the substitutional existential quantifier. The predicates 'red' and 'triangular' as is implied by the qualifier 'some way' depend on the subject term. In (PP1), all Sellars seems to be concerned with is to deny that we can know inspectionally or innately that this something is a physical object, a state of ourselves, an event, or whatnot. I agree that such knowledge is not given, as follows from the rejection of (PP). James Cornman seems to have a similar reading of Sellars as presented in the following passage:

Thus he [Sellars] can consistently maintain that color is given in some form or other, but in what form it actually occurs -- whether as properties of entities (such as physical objects, light, or sensa). or as internal specifications of sensings, or as something else -- is to be decided by what explains best scientifically. On this view, then, science must interpret sensuous features of the world, such as sounds, smells, and tastes; it cannot eliminate them or transform them so much that there remains nothing significantly like them. Such a Sellarsian can, then, eliminate sensuous color from the external world in a Lockean way, but if he does so he must locate something much like it in perceivers. In this respect, Sellars accepts one sort of given.{16}

      But if this is what the denial of (PP) and givenness amounts to, then (PP) is misleading. It does not explicitly distinguish between the concept of a subject and predicate. To prevent this possible misreading, I would initially amend (PP) to read:

(PP') If a person is directly aware of the fact, (Ex) Px, where x has categorial status C, i.e., x = C, then the person is aware of it as having categorial status C, i.e., as P(C).

      And I have no difficulty in rejecting this principle. The reason for my agreement with Sellars here is that the x is for me an 'inferred' theoretical entity whose categorial status is uncertain. This needs some explanation. Sellars would agree that x may be viewed as a theoretical entity, but he would not use the word 'inferred' in this context. He would rather speak of the x as a 'caused' conceptual entity. This distinction is correlative with his Manifest-Scientific Image distinction generally. The Manifest Image is a caused conceptual framework; the Scientific Image is an inferred conceptual framework. This distinction will be discussed in the chapter on Animal Realism.

      However I have an objection to (PP') as a paradigm of identifying something, and consequently to (PP1) as well. (PP') is formulated using a quantified variable; I would prefer to use indexical expressions. The paradigm should be:

(PP'') If a person is directly aware of the fact, P(this here now), where this has categorial status C, i.e., this = C, then the person is aware of it as having categorial status C, i.e., as P(C).

      The difference between (PP') and (PP'') is that the latter entails the former, i.e., 'P(this)' entails '(Ex)Px'; but the former does not entail the latter. That is, if I see this thing which is in front of me to be P, it follows that there is something or other which is P; but from the fact that something is P, it does not follow that this something is the thing before me now.

      I would modify (PP1) to reflect the use of indexicals:

(PP2) When one ostensibly sees an object which is red and triangular on the facing side, this here now, in some way red and triangular is in some way present to the perceiver other than as thought of.

1. STATUS OF PROPERTIES

      We must avoid talking at cross-purposes about properties. To do so we have to be aware of the distinction between objective and phenomenal properties.

      The conceptual status of P in 'Something (physical object) is P' is different from 'Something (transcendental) is P'. In the former case P is an objective, physical property governed by the meaning postulate:

x is P. <--> .x would appear P to standard observers in standard conditions

A specific example of this is,

x is red. <--> .x would look red to normal observers (who are, for example, not color blind) under normal conditions of illumination (such as daylight or in- candescent lights)

Sellars accepts this as a necessary truth, but not as a definition or as an analysis of physical redness.{17} The reason for this is that, for Sellars, physical redness is an undefined, i.e., a primitive term in the Manifest Image.

      Now this is puzzling since such a formula with a biconditional seems like a prima facie case of an analysis. The only reason the appearance statement could not be an analysis is if the appearance statement itself presupposes the objective statement. And this is indeed the line of reasoning that Sellars takes. The appearance statement is conceptually tied to the contrastive objective statement. The question which must be raised, then, and which was raised by R. Firth in his reply to Sellars' Carus Lectures, is whether the contrastive appearance statement presupposes in turn a non-contrastive appearance statement. To this Sellars replied that the contrastive appearance statement presupposes a non-contrastive objective statement.

      My quick reply to such a move is to claim that there is no problem for a sense-data theorist to accept the following identities:

(ID) Sense-datum statement = non-contrastive appearance statement = non-contrastive objective statement

      However, in describing the non-contrastive objective statement, e.g., 'This is red', Sellars recategorizes the predicate 'red' from being an attributive predicate to being a mass term.{18} 'This is red' is transposed to 'This is an expanse of red stuff', "where 'stuff' carries with it implications concerning the causal role of determinate portions of stuff in the physical world."{19}

      My reply to this is that if this is meant to be a description of the phenomenological world of a child, then it seems plausible. But if this is intended as a phenomenological reduction to the rock bottom, then I think it fails. To say that the phenomenological experience has to be such as to contain a belief in the causal properties of something is to commit the propria persona fallacy (i.e., to accept the myth of the given). On other occasions Sellars has recognized that in a phenomenologically reduced experience the causal properties are not experienced.

Of course in seeing it as a pink ice cube we are seeing it as having the causal properties characteristic of ice. But do we see of the object these causal properties? The question is, to be sure, an awkward one. But if we consider certain other questions of a similar form, the answer seems to be no.{20}

      So, if Sellars' considered judgment is that causal properties should be excluded from a phenomenological reduction, then I believe the identity thesis (ID) can be accepted. And it seems to me that, at least on the line of reasoning I have followed, Sellars' position entails a phenomenological reduction of perception of physical objects to the awareness of sense data.

III. BROAD (IMPLICIT) SENSE OF THE GIVEN

      The broad formulation (which also has a right to be recognized as a version of the given) is the claim that people have a pre-linguistic awareness of categories. We find in Sellars the rejection of the following principle:

(P) X can be experienced under some category C prior to the possession of a conventional language.

(P) claims that there is a primordial, pre-linguistic awareness of kinds.

      Although Sellars does not explicitly call this principle a version of "the given", he does explicitly reject it -- calling the rejection of such a principle "psychological nominalism." Since this principle is a denial of a necessary condition for (PP), it has the right to be called a primordial version of the given.

      This principle applies to subjects and predicates alike. I would, of course, also be a foe of the given if the thesis of the (broad) given were formulated as:

(P') X can be experienced under some category C prior to the possession of a representational system.

My disagreement with Sellars is over the question whether besides conventional languages there is a representational system. I affirm this while Sellars -- what can I say? -- forgets it.

The narrow and the broad forms of the myth are not equivalent. The denial of the broad sense (P) entails the denial of the narrow sense (PP), but the acceptance of the broad sense does not entail either the truth or falsehood of the narrow sense. The denial of the amended broad sense (P') is compatible with the possibility of accepting the following amended version of the narrow sense (PP).

(RS-PP) If a person is directly rs-aware of an item which has categorial status rs-C, then the person is rs-aware of it as having categorial status rs-C.

      Now Sellars believes this to be a form of the given:

To reject the Myth of the Given is to reject the idea that the categorial structure of the world -- if it has a categorial structure -- imposes itself on the mind as a seal imposes an image on melted wax.{21}

      And this is exactly what I want to claim. The only qualification that I make is that the idea of a 'categorial structure' has to be disambiguated. A distinction has to be made between a 'conceptual structure' and a 'representational structure'. And my position is that, in a manner of speaking, the world does through evolutionary development of our sensorial and cognitive processes at the animal level impose on us, what may be called, an RS-Manifest Image.

      I think my approach is compatible with the Aristotelian-Thomistic view about species impressa, but disagrees with it about species expressa. In "Being and Being Known"{22} Sellars confronts the Thomistic tradition and rejects the species impressa hypothesis as entailing a thesis that the senses are in the intentional order; and he rejects the species expressa hypothesis as entailing an abstractive theory of concept formation. The only feature of the Thomistic tradition Sellars accepts is the idea of a concept as a verbum mentis.

      My disagreement with Sellars here is about the intentionality of sensing. He fails -- forgetting? -- to make the distinction between non-conceptual intentionality (representation) and conceptual intentionality (representation). Sensing is a type of non-conceptual intentionality. Sellars agrees with this in "Mental Events" when he at- tributes to animals representational systems.

      It seems to me that if the world did not impose a representation of itself (i.e., a map) at the animal level where learning is necessary to survival, I don't see how animal survival would be possible. As I see it, even if different Gestalts are imposed on different species, still the causal connections between variable Gestalts must be invariant.

      The denial of (P) is, however, incompatible with these possible scenarios. According to the denial of (P), there cannot be any categorial structure whatsoever prior to the acquisition of a conventional language.

      Conventional languages have for Sellars the character of more or less random structures which compete for survival. And ultimately a conceptual structure is accepted for two reasons. The first is that it is useful for explaining and predicting; the second, because of its ability to "picture" the world more adequately than other competing structures.

      Sellars' polemics about the myth of the given should be read with the narrow/broad, and conceptual- /non-conceptual representational systems distinctions in mind; otherwise the issue of the polemics may not be clear; in fact, the various polemics may be arguments at cross-purposes.

      A case in point is Sellars' running debate over the years with R. Firth. Sellars argues against Firth that the explicit (narrow) version of the given is a myth without realizing that this is not where the real disagreement with Firth lies. Firth in fact agrees with Sellars' verdict that the narrow sense is a myth. In his comments on Sellars' Carus Lectures, Firth wrote:

I also agree with Sellars (par. 44)[p. 11] that the following principle embodies a fallacy (one that I think might well be called, in the present context, the Fallacy of Conceptual Retrojection): "If a person is directly aware of an item which has categorical status C, then the person is aware of itashaving categorical status C.{23}

Firth's real disagreement is over the broad (implicit) sense of the given.

      The disambiguation of "the given" can also throw new light on the disagreements with C. I. Lewis, R. Chisholm, W. Robinson, J. Cornman, R. Bernstein, T. Russman, A. Marras, H-N. Castañeda, R. Clark, W. Rottschaefer, and Alan Goldman. They all agree, in opposition to Sellars, in positing a pre-linguistic awareness of kinds, i.e., they all accept some version of (P).

IV. WILLIAM ROBINSON ON THE GIVEN

      William Robinson's essay "The Legend of the Given"{24} is possibly the most ambitious attempt to reconstruct Sellars' reasons for believing the given to be a myth. His conclusion is that Sellars has not presented an acceptable refutation of the given in knowledge and, therefore, the given is not a myth, i.e., definitely false, but remains a legend, i.e., possibly true. His procedure is to reconstruct from Sellars' principles the best possible argument against the given, and then refute the argument. However, I contend that his endeavor at refutation fails because of fundamental misunderstandings, and his alleged counterexample, I will argue, is not a counterexample.

A. MISUNDERSTANDING OF THE GIVEN

      Robinson starts out by using the following passage from Sellars as a premise:

(F) the point of the epistemological category of the given is, presumably, to explicate the idea that empirical knowledge rests on a 'foundation' of non-inferential knowledge of matter of fact.{25}

      From this premise in conjunction with others, which are supposedly implicit in Sellars, he tries to conclude that "we are, therefore, entitled to assert the possibility of a primordial awareness of sorts that is not knowledge."{26}

      But this attempt right from the beginning is completely misdirected. It is obvious from (F) that the given is meant to be a type of epistemic claim; whereas Robinson intends to argue that the given is not epistemic. At best, Robinson is arguing at cross purposes. In other words, there are two concepts of the given being conflated -- Sellars' concept, given(s); and Robinson's concept, given(r).

1. EXPLAINING THE MISUNDERSTANDING

      Assuming that Robinson has misunderstood Sellars, my first task will be to explain how this occurred. The best place to start is with Robinson's own difficulty in deciding whether the following formulation represents Sellars' view (I maintain Robinson's numbering and lettering of theses):

(22) The explication of a foundation view requires that the primordial awareness of sorts used in the explication be non-inferential knowledge of matters of fact.{27}

      It seems clear to me that (22) follows from (F): both are referring to a kind of foundational non-inferential knowledge. But this reference to knowledge is not clear to Robinson.

      He writes: "It is clear that if (22) is true there is no hope of analyzing the given non-epistemically."{28} This is correct, and Robinson should have terminated his search for a non-epistemic given. But he doesn't. And in a footnote he writes:

There may be a temptation to cut off this possibility by holding that Sellars is assured of the truth of (22) because he makes it true by definition. (This is tantamount to including 'being non-inferential 353 knowledge' in the definition of givenness.){29}

      Yes, this is exactly what Sellars intends. The given is, by definition, an epistemic notion! So how are we to explain Robinson's failure to see this? Suppose (22) is Sellars' considered position. In that case, Robinson says: "Such a move would indeed simplify the present discussion, as also several other in this essay."{30} What it really means is that his discussions are pointless since he is trying to establish claims not really denied by Sellars.

      So what are his reasons for thinking that Sellars would deny (22)? He gives two. The first reason is that the acceptance of (22)

would also have the effect of making Sellars' results irrelevant to some traditional views. Price, for example, is at some pains to distinguish sensing from knowing, in his introduction of sensing sense-data. Vide H.H. Price, Perception (London: Methuen, 1932): 5, 17-18, and esp. 7.{31}

      The second reason is that "section I, para. 3 of [Sellars'] EPM ["Empiricism and the Philosophy of Mind] (128-29) is barely intelligible on the assumption that Sellars is taking (22) as definitionally true."{32}

      Let us examine these reasons. The first reason Robinson gives is that according to Price the given of sense-data is equivalent to the sensing of sense-data, and the sensing of sense-data is not a form of knowing; therefore, the given is not a form of knowing.

      In the passages referred to in Price, Price is making a distinction between two types of 'intuitive apprehensions', namely, 'apprehension that' and 'apprehension of'. These are elsewhere identified with 'knowledge of facts about' and 'knowledge by acquaintance' respectively.{33} Robinson is simply mistaken in thinking that Price is making a distinction between sensing and knowing; rather he is making a distinction between two kinds of knowing -- sensing as knowing by acquaintance, on the one hand, and knowing of facts about, on the other. Sensing, for Price, is a type of knowledge. This is even suggested by looking at the index of his Perception under the heading "sensing" which has the subheading "is a form of knowledge". So Robinson's reasoning that if Sellars accepts (22), he will not do justice to Price is caused by a misreading of Price. The second reason Robinson has for thinking that Sellars would reject (22) is based on his reading of the following passage from Sellars:

      On alternative (a) the fact that a sense content was sensed would be a non-epistemic fact about the sense content. Yet it would be hasty to conclude that this alternative precludes any logical connection between the sensing of sense contents and the possession of non-inferential knowledge. For even if the sensing of sense contents did not logically imply the existence of non-inferential knowledge, the converse might well be true. {34}

      Robinson's comment on this passage is this:

Sellars does not criticize this view, but goes on to set out alternatives. If he thought that what is required to explicate a foundation view must be non-inferential knowledge, we might expect that he would reject the alternative just mentioned immediately. Thus, the fact that he does not do so gives us some reason for thinking that Sellars does not believe (22).{35}

      But, contrary to Robinson's reading, Sellars does reject alternative (a) immediately by pointing out that it would be a non-epistemic fact. We could deduce this position from a general principle of Sellars' epistemology, namely, that epistemic facts cannot be explicated by non-epistemic facts, as is clearly stated in the following passage:

Now the idea that epistemic facts can be analyzed without remainder -- even 'in principle' -- into non-epistemic facts, whether phenomenal or behavioral, public or private, with no matter how lavish a sprinkling of subjunctives and hypotheticals is, I believe, a radical mistake -- a mistake of a piece with the so-called 'naturalistic fallacy' in ethics.{36}

      By pointing out that something is a non-epistemic fact immediately precludes it from serving in an explication of an epistemic fact. Paradoxically, Robinson seems to be aware of this principle in Sellars when he attributes to him the claim that "epistemic facts cannot be analyzed without remained into non-epistemic facts."{37} So it is puzzling why Robinson persists in seeking a non-epistemic given. Perhaps Robinson is not clear on how Sellars is using the notion of 'analysis' or 'explication'. In "Empiricism and the Philosophy of Mind," Sellars distinguishes between an analysis and an explanation. The form for an analysis, Sellars tells us, is this: if q is the correct analysis of p, then it is impossible for someone to believe p without believing q.{38}

      Yet Robinson's effort is to conclude that "the given can be analyzed non-epistemically. For it can be analyzed as a non-subsumptive use of an ur-concept" (italics mine).{39} Let us ignore for the moment the notion of 'non-subsumptive use of an ur-concept' and concentrate on the claim that 'the given can be analyzed non-epistemically'. This is not how Sellars uses the term 'analysis'. There is no such thing for Sellars as a non-epistemic analysis of an epistemic fact. If there were, it wouldn't be an analysis. Robinson has compounded his misunderstanding. Not only has Robinson misunderstood Sellars notion of the given, but he has also misunderstood Sellars's notion of analysis as well.

B. ROBINSON'S ARGUMENT

      Starting then with these two misunderstandings, Robinson reconstructs the following alternatives from Sellars' writing:

(1) Either (a) the givenness of sense contents is defined in terms of non-inferential knowledge of sense contents; or (b) the givenness of sense contents is a "basic or primitive concept of the sense-datum framework."{40}

(2) Either (c) sensing is a "unique and unanalyzable act"; or (d) sensing is analyzable.{41}

(3) Either (e) sensing is analyzable in non-epistemic terms; or (f) sensing is analyzable in epistemic terms.{42}

      Robinson notes that if we identify sensing with being given, then substituting 'given' for 'sensing' in the above formulas, we get the following counterparts:

(c') Something's being given is a unique and unanalyzable occurrence.{43}

(e') Giveness is analyzable in non-epistemic terms.{44}

(f') Giveness is analyzable in epistemic terms.{45}

      He interprets Sellars' rejection of the given as "the rejection of (a), (c'), (e'), and (f')";{46} and he believes if any of these theses has not been refuted, then the given has not been refuted. Robinson evidently agrees with these rejections save for (e') which he wants to defend. He points out that (e') "contains the term 'analyzable', a term which might lead to quite extraneous difficulties."{47} The difficulties this does lead to are not extraneous, but are, I believe, as I have already mentioned, a fundamental reason for Robinson's misunderstanding of Sellars. Sellars points out that various entailment relations may exist between propositions which do not amount to an analysis, and that analysis is not to be confused with explanation. However, Robinson ignores these precautions and provides an account which is not an analysis by Sellars' stipulations. Therefore, Robinson's so called 'analysis' is at cross-purposes with Sellars'.

1. RECONSTRUCTION OF SELLARS' ARGUMENT

      Robinson reconstructs a Sellarsian argument for the conclusion (7) as based on the following implicit premises in Sellars.

(10) 'x has primordial classificatory consciousness' entails 'x makes a primordial use of a concept'.

(11) 'x makes a primordial use of a concept' entails 'x immediately subsumes something under a concept'.

(12) 'x immediately subsumes something under a concept' entails 'x has non-inferential knowledge'.

Therefore,

(9) 'x has primordial classificatory consciousness' entails 'x has non-inferential knowledge'.

      From (9) and the premise:

(8) 'x has a primordial awareness of sorts' entails 'x has primordial classificatory consciousness'.

it can be concluded:

(7) 'x has a primordial awareness of sorts' entails 'x has non-inferential knowledge'.

      The argument is valid. It has the form of a series of hypothetical syllogisms. But according to Robinson the argument is not sound. Since Robinson wants to deny the truth of (7) while accepting (8), he is forced to reject (9). To reject (9) he must reject at least either (10), (11), or (12). Robinson does not dispute either (10) or (12). As to (11), he writes "with (11) we have come to a 'rock bottom' principle of Sellars' system" and, he adds, "(11) is false."{48}

      To make his case, Robinson combines premises (8) and (10) to draw the conclusion:

(13) 'x has a primordial awareness of sorts' entails 'x makes a primordial use of a concept'.

      He then defines the idea of a 'primordial use of a concept' as the thesis:

(14) 'x makes a primordial use of a concept C1' entails 'C1 is independent of physical concepts'.

      Combining (13) and (14) we get:

(15) 'x has a primordial awareness of sort C1' entails 'C1 is independent of physical concepts'.

      He adds a terminological point: "I will call concepts which are independent of physical concepts ur-concepts."{49} As an example, Robinson gives the concept of red as an ur-concept. He goes on to claim that Sellars would deny the existence of ur-concepts in his sense. This is true. There are various sensorial experiences, such as an after-image, which Sellars would say are conceptualized on the basis of perceptual experiences of physical objects. What Sellars denies is that perceptual reports entail ur-concepts. But Robinson thinks that if Sellars granted the existence of ur-concepts, then he would grant the following thesis:

(16) 'There are ur-concepts' entails 'descriptive contents [of perceptual reports] can be intrinsically characterized by using ur-concepts'.

      I don't think this is true for Sellars. Sellars can grant that ur-concepts (in Robinson's sense) are genetically secondary as conceptually derivative, but still hold that they now have an independent application in experience. But even if we grant the current independent use of ur-concepts, it does not follow that perceptual reports of physical objects are analyzable in terms of ur-concepts.

      Not mindful of this objection, Robinson's aim is to demonstrate the possibility of ur-concepts. His reasoning is this. Robinson points out that (11) entails:

(11a) 'x makes a primordial use of a concept' materially implies 'x immediately subsumes something under a concept'.

      This can be granted. And he writes: "(11) is false if the falsehood of (11a) is possible."{50} He adds that the denial of (11a) "can be made intelligible . . . if non-subsumptive uses of concepts can be made intelligible."{51} Since Robinson agrees that subsumptions under a concept are thoughts, it follows that

(a) If there are primordial awarenesses of sorts, they have to be thoughts.

      Robinson's refutation boils down to the claim that there is a use of concepts which is not subsumptive and is not an awareness of sorts. To make his point, he distinguishes between subsumption and the counting use of concepts.

      But before I pursue his reasoning further, we need a clarification of the concept of a concept. As Robinson proceeds to construct his counter-argument to Sellars, he introduces the following definition of concept:

Definition III. A concept is a mind dependent ability to be in a state which corresponds to a property.

      He goes on to make the following theses about concepts: "a use of a concept is the actualization of an ability to be in a state corresponding to a property"; and "a subsumptive use of a concept is the actualization of an ability to be in a state corresponding to an assertion." {52}

      By contrast, for Sellars a concept is a word that has an essential use in inferences. And since, for Sellars, there are mental words that are functionally analogous to physical linguistic words, it is proper to talk about mental concepts as well. Given this understanding of concepts, Robinson's definition of concepts would be, with qualifications, at best a sub-category of Sellars' concepts. The translation into Sellars idiom would be something like this: A mental predicate concept corresponds to a linguistic predicate concept.

      Robinson's understanding of concepts, though not quite Sellars', is really innocuous and not at cross-purposes. However, when Robinson introduces the idea of ur-concepts he opens room for misunderstanding. The term 'ur-concept' is used by Sellars and R. Firth in their polemics with each other.{53} Sellars and Firth use it in a sense in which

(23) x is an ur-concept = df. x is a concept in which the appearance-reality distinction has collapsed.

Robinson, however, offers the following definition:

(24) x is an ur-concept = df. x is a concept independent of physical concepts.{54}

      Sellars denies (24). In his debate with Firth, Sellars insists that ur-concepts are applicable only to physical objects, and, therefore, they are conceptually dependent on the concept of a physical object. Robinson is therefore correct in claiming that Sellars denies the existence of ur-concepts in Robinson's sense. Sellars can agree with Firth that the appearance-reality distinction collapses, but not that it collapses into an ur-concept of an appearance, as Firth claimed in his reply to Sellars, but rather that it collapses into an ur-concept of a physical object; and, by the same token, Sellars can reject Robinson's definition of an ur-concept as independent of the concept of a physical object.

      If Robinson is to make his case, he has to come up with an example of a concept which is independent of physical object concepts, and which is not subsumptive (classificatory), and which does not involve a thought (proposition). To do this he cites the existence of a primordial use of counting concepts. He claims that a classificatory use of concepts presupposes a counting use of concepts.

If I count the books on my shelf, I may say, as I point, "One, two, three, . . ." These words reflect a state of mind corresponding to the properties 'first book counted by me on this occasion', 'second book counted by me on this occasion', etc. But I have not subsumed these books under concepts corresponding to these properties. Nor, indeed, could I have correctly done so until I had counted them. Even if I had said "This is the first book, this is the second book, . . . ," this would only look like subsumption, but would actually be counting. Of course, my counting does provide a ground for subsumption, by myself or another, of these books under such concepts as "first book counted by me on this occasion. " But this is not to say that the original counting is itself a case of subsuming. {55}

      Evidently Robinson wants to conclude that counting is a type of use of concepts that is independent of subsumption:

a use of a concept is the actualization of an ability to be in a state corresponding to a property, while a subsumptive use of a concept is the actualization of an ability to be in a state corresponding to an assertion. {56}

      It is not clear that Robinson is providing a distinction, since it is evident that counting can have the appearance of an assertoric subsumption, e.g., 'This is the first book'.

      However, Robinson is overlooking a necessary condition required for counting to take place. In order to count, items have to be distinguished in some manner. The manner in which anything is distinguished is by way of a concept. For Sellars this is a fundamental point: "Observing that the situation is thus and so already involves the use of a conceptual frame.";{57} and "to have the ability to notice a sort of thing is already to have the concept of that sort of thing."{58} It is only relative to this differentiation through concepts that counting makes sense. Suppose we are looking at two squares: one green and the other red, and two circles: one green and one blue. Imagine the following dialogue about counting these things.

Jack: Count.
Jill: Count what?
Jack: The things before you.
Jill: What am I to take as a 'thing' here?
Jack: Enclosed figures.
Jill: There are four.
Jack: Now count the number of green things.
Jill: Two.
Jack: How many red circles are there?
Jill: None.

      As we can see, counting makes sense relative to subsuming that which is to be counted under concepts. Robinson presupposes the subsumptive concept by saying "If I count the books on my shelf . . ." After all, he could have also said "If I count the books in my room . . ." And then possibly he would get a different count by including books on a table, on the floor, etc. The point is that by setting out to count books, he already has made a subsumptive use of concepts by differentiating his environment through the concept of a book.

      In conclusion, Robinson believes that Sellars allows for a given in a non-epistemic sense, which by his understanding is a non-subsumptive use of concepts, and he believes that he has provided for a non-subsumptive use of concepts in the case of counting. Robinson errs on both counts. Sellars does not allow for a non-epistemic given; and counting, rather than being different from subsumption, presupposes subsumption.

      But perhaps I have missed something in Robinson's argument. If I have then my excuse is the same as C. Echelbarger's, who wrote in a criticism of Robinson's essay, "I am still in the dark about this notion of a primordial awareness involving nonsubsumptive uses of ur-concepts."{59}

V. SELLARS AND LEWIS ON THE GIVEN

      Talk about a given element in knowledge is a prominent aspect of C.I. Lewis' Mind and the World Order, in which he wrote:

There are, in our cognitive experience, two elements; the immediate data, such as those of sense, which are presented or given to the mind, and a form, construction, or interpretation, which represents the activity of thought. Recognition of this fact is one of the oldest and most universal of philosophic insights. {60}

      Sellars agrees with Lewis that this bifurcation of cognitive experience into a given and an interpretative element is probably a universal assumption of philosophers; but Sellars does not believe that this assumption is an insight, it is rather a universal mistake!

      Sellars views his polemics with Chisholm and Firth as an extended polemic with Lewis. This is clearly expressed in the following passage:

Firth's paper ["Coherence, Certainty, and Epistemic Priority"] is such a rethinking of Lewisian themes. It is not so much on Lewis as it is Lewis himself in that mode of immortality with which Plato invested Socrates not by writing about Socrates, but by being Plato. If Firth is Lewis, however, so, too, is Chisholm, who will also play a major role in these pages. And if Firth is not Chisholm, nor Chisholm Firth, so much the worse for identity.{61}

      That Sellars makes an identification between Lewis and Firth is significant for the following reason. I believe that Sellars has misinterpreted Lewis, and the nature of this misinterpretation comes to the surface in Sellars' polemics with Firth. Sellars' uneasiness with his interpretation of Lewis is expressed by the fact that he has tried to come to grips with Firth's paper "Coherence, Certainty, and Epistemic Priority" four times: in "Giveness and Explanatory Coherence" (1973); in the Appendix to the first lecture of "Structure of Knowledge" (1974); in "More on Giveness and Explanatory Coherence" (1979); and finally in the first Carus lecture "The Lever of Archimedes" (1981).{62} Firth replied to the last.

      References to Lewis are found as early as "Is There a Synthetic 'A Priori'?"{63} -- written before his monumental "Empiricism and the Philosophy of Mind", and here Sellars presented his first interpretation of Lewis:

If we put this implication is a slightly different way, we immediately establish contact with a characteristic contention of Professor Lewis. All classification of objects, however confident and preemptory, is a venture, a venture which at no point finds its justification in a pre-symbolic vision of generic and specific hearts on the sleeves of the objects of experience. Classification resembles the grasping tentacles of an octopus, now tentative, now confident, rather than a salesman's selection of a suit for a customer after a glance at his build. I am afraid, however, that our agreement with Lewis is more shadow than substance. For while he writes in this manner of the interpretation of the given, he also holds that the sensible appearances of things do wear their hearts on their sleeves, and that we do have a cognitive vision of these hearts which is direct, unlearned, and incapable of error -- though we may make a slip in the expressive language by which these insights are properly formulated. In other words, the assumption to which we are committed requires to extend to all classificatory consciousness whatever, the striking language in which Lewis describes our consciousness of objects.{64}

      And the last reference to Lewis appears in the Carus Lectures, where Sellars still charges him with believing in the Given: "In the grip of the Myth of the Given, a C.I. Lewis . . ."{65}

      Sellars, in his rejection of the Given, is criticizing, what he takes to be, a fundamental thesis of Lewis' philosophy. This is the view "that we do have a cognitive vision of these hearts [sensible appearances] which is direct, unlearned, and incapable of error -- though we may make a slip in the expressive language by which these insights are properly formulated."

      There is an ambiguity here which is the source of a misinterpretation. Reference is made to "sensible appearances". Given Sellars' distinctions as we find them in "Mental Events" (1981),{66} we are in a position to ask: Are sensible appearances conceptual or non-conceptual? And if they are non-conceptual, are they representational or non-representational?

      Now it is not clear to me that these distinctions were always in the foreground of Sellars' thinking. They are in the late seventies, and I suspect that this is why "More on Giveness and Explanatory Coherence" was written. But in his interpretation of Lewis, Sellars seems to have been working with a dichotomy whereby either something is conceptual or it is not -- meaning it is not even representational. And since Lewis clearly states that the given is non-conceptual, Sellars takes this to mean that it is non-representational.

      To understand Sellars' criticism of Lewis, we first need to know what Lewis is claiming. I believe that Lewis makes the following four claims:

(I) All empirical cognitive experiences are composed of two cognitive elements: the given (in intuition) and an interpretation of the given.

(II) All meanings are divided into sense meanings (independent of language) and linguistic meanings.

(III) The given and the interpretation can be found in sense meaning independently of linguistic meaning.

(IV) The given is needed to provide foundations for empirical knowledge.

      Sellars, as I understand him, in his initial interpretation of Lewis would have denied either my interpretation of Lewis or the truth of these claims. Against the first claim he would have said either -- and I am not certain which -- that all cognitive experience is conceptual or that (I) is a misinterpretation of Lewis: the claim should be that (I') all empirical cognitive experiences are composed of a non-cognitive (given) and a cognitive (interpretational) element.

      Against the second claim, he would have said that sense meaning (if there is such a thing) can exist only in the ambience (in some sense) of linguistic meaning.

      Against the third claim, he would have said that there is no conceptual meaning independent of linguistic meaning. Against the fourth, he would have said that since the given is non-cognitive, it cannot provide epistemic foundations. The given for Lewis is a direct cognition through sense meaning (representation) which is incorrigible, can be true, and foundational for knowledge. Does Sellars address himself directly to this complex claim? No. In his polemics, Sellars is working, I believe, with the following argument.

(1) For a person to be aware of something requires the use of concepts.
(2) Concepts arise only through the learning of a conventional language.

Thus,

(3) to be aware of anything at all requires the possession of a conventional language.

      Since (1) and (2) are presupposed by (3), their denial would presumably amount to the acceptance of what we can call a primordial sense of the given. But I don't think that this is consistent with Sellars considered judgment. (1) could be taken as an analytic statement. In that case it explicates Sellars' use of 'aware'.

      To confront Lewis on his own ground, the question we must ask is: Is there anything analogous to conceptual awareness? And Sellars should answer, animal "awareness."

      I believe that Sellars' (1) was denied by Lewis when he introduced the concept of 'sense meaning'.

A. THE OBJECTIVE AND SUBJECTIVE

      The classification of any indeterminate awareness as subjective is already interpretational. In Lewis' terminology, the given, if such an interpretation is used, is the object of subjective awareness which does not contain concepts; whereas objective awareness is anticipatory of experiences and does contain concepts. If we agree that the necessary condition of any awareness is classificatory consciousness, then Lewis' distinction amounts to the following. Subjective awareness classifies without an anticipatory or predictive component in the awareness; whereas objective experience is precisely, and contrastingly, pre-occupied with anticipatory and predictive classification. Subjective awareness consists of the ability to see similarities, continuities, differences, and discontinuities. Objective awareness consists of the ability to anticipate and imagine forthcoming experiences. This classification of present experiences as signs of anticipated experiences Lewis calls 'concepts'.

      The subjective-objective distinction could be made on the basis of the type of awareness involved. Thus, the awareness of similarities and dissimilarities and relations, understood as expressing how things appear to one, is a subjective affair. Expecting determinate appearances, however, is an objective matter. 'Knowledge', as Lewis uses this word, is applicable only to such objective judgments in which something is anticipated or predicted. Since subjective awarenesses are non-anticipatory and non-predictive classifications, Lewis refrains from applying to them either the word 'judgment' or 'knowledge'.

      The distinction between subjective and objective experiences is not a distinction to be found in a single experience, a single experience is neutral between subjectivity and objectivity. The "distinction between the subjective and the objective," writes Lewis, "is irrelevant to givenness as such."{67} In the case of a single experience a person is conscious of the contents of the experience, but is not reflectively conscious about the connection of this experience to other experiences, except in terms of similarities and dissimilarities.

      From a Sellarsian point of view all of this can be put it terms of his description of a Representational System (RS) as consisting of entrance positions, moves within the representations, and exist transitions. Lewis' 'given' is equivalent to Sellars 'entrance position', and Lewis' 'sense meaning' is equivalent to Sellars' 'Humean inference'. The proper correlations between Lewis' and Sellars' views are the following:

Sellars	Lewis
RS-entry =	given
RS-inference =	sense meaning (= concept =Humean inference)
RS-exit
L-entry =	proposition
L-inference =	syntactic and semantic meaning (Aristotelian inference)
L-exit

(RS = Representational System; L = Language)

Now we can agree that in some way these two representational systems are connected in human beings. What is the proper way to talk about them? For Lewis, both systems are phenomenologically accessible. For Sellars, this is not the case for the Humean RS. It is, for Sellars, part of an explanatory framework. However, this is not an impasse. After all, Lewis writes -- paradoxically -- that "the given" is not given: "The given is in, not before, experience."{68}

      To make sense of this, we can borrow the apparatus of Firth's article "Sense-Data and the Percept Theory" and distinguish that which is straightforwardly given as described correctly by, what Firth calls, the Percept Theory; and that which is given through a phenomenological reduction, which is described by the Sense-Data Theory. Lewis' claim is that "the given" is given only in phenomenological reduction. And phenomenological reduction is a technique of "bracketing" and abstraction.

      Sellars' position is a form of the Percept Theory, which can be formulated as the denial that a statement expressing the perception of a property of a physical object can be analyzed by a statement expressing the sensing of a property of a sense datum. Now it is assumed that there are such states of consciousness as described by the Percept and the Sense Datum Theories. It is simply denied by the Percept Theory that the state of consciousness described by the Sense-datum Theory is a 'constituent of' or 'present in' (even implicitly) in the state of consciousness as described by the Percept Theory: the two states of consciousness are distinct. Now Firth agrees with this. But he believes that a connection between the statements expressing these two states can be made through a 'definition is use'.

We can say that the statement "These two perceptions are different interpretations of the same sensory core", should be understood to mean: "If these two perceptions were perceptually reduced exactly similar states of direct awareness would be produced in the two cases."{69}

      In view of this "definition in use," the issue between Sellars and Lewis (= Firth = Chisholm) becomes whether Sellars' 'sense impressions', which are theoretical posits, are accessible to consciousness through phenomenological reduction. The issue becomes an issue about the possibility of phenomenological reduction. It is this question which Sellars tackled in his first lecture of the Carus Lectures. And though Sellars did go some distance in his phenomenological reduction, it was not far enough -- as Firth argued in his reply.

      I will discuss this in greater detail in the next chapter on perception.

---

[Go to Chapter 9]

---