Saturday, June 28, 2008

Go Ask Alice

In this paper I aim to address McDowell’s claim that in the Sellarsian picture of intentionality sense impression look like ‘idle wheels’, particularly the view which deVries attributes to McDowell that sense impressions are ‘causally idle’. I aim to offer an argument against McDowell and in defense of Sellars that sense impressions are in fact necessary for the complete picture of how conceptual states get to be related to and accurately represent physical objects; I will do so via deVries argument in McDowell, Sellars, and Sense Impressions in addition to relying upon Sellars’ own works in Empiricism and the Philosophy of Mind, Science, Perception and Reality, and of course, the text with which this discussion is principally concerned, Science and Metaphysics.

I will proceed by first laying out the core of Sellars’ argument in Science and Metaphysics, then examining McDowell’s counterargument in Having the World in View: Sellars Kant and Intentionality. Lastly I will move to explicate and defend deVries’ counter against McDowell and defense of Sellars and the validity of sense impression in McDowell, Sellars, and Sense Impressions. To be clear though, deVries, as he takes some pains to point out, does not exactly offer a concrete and specific defense of Sellars’ theses in SM, more accurately, his concern is to erode and reduce to a level of conjectures and unfounded assumption the argument that McDowell makes against Sellars so that Sellars’ piece may be left standing for independent critique by his readers. deVries believes that McDowell attributes several theses to Sellars, some of which I will discuss later in this paper. deVries’ own conclusion is that:

none of these theses are compatible with Sellars’s philosophy; that Sellars’s conception of the transcendental is thoroughly misunderstood by McDowell, and that McDowell consequently misunderstands the role of sense impressions or sensations in Sellars’s thought. Because of this, the critique McDowell offers of Sellars’s relatively robust conception of sensation, which is also a vital part of Sellars’s corrected Kant, does not hit the mark (MSSI; 183).

he goes on to say “I won’t worry about whether Sellars or McDowell is the better Kantian, nor about whether Sellars, McDowell, or Kant is, in the long run, right.” (p. 183, MSSI). McDowell incorrectly believes sense impressions are posited to explain conceptual representation in general when actually they are posited to explain the minimal conceptual reports we have of objects. (190, MSSI). Clearly, deVries’ focus is relatively narrow especially when compared with that of Sellars or McDowell. He wants to deconstruct and identify the problems with McDowell’s interpretation of Sellars, and, once he has done so “Since McDowell’s criticisms of Sellars do not hit their target, they leave Sellars’ complex and profound reflections on sensation standing for your independent appraisal.” (p. 183, MSSI). Now let me begin by breaking down Sellars’ argument in SM.

In “Sensibility and Understanding”, the first chapter of his work Science and Metaphysics, Sellars aims to indicate where, in his belief, Kant failed to adequately explain the relationship between two modes of representation. These modes are sensibility, which is the receptivity of the mind for impressions and which does not involve concepts, and understanding, which through the representations received by sensibility produces concepts and enables the perceiver to “know an object” (SM, p. 2). Kant maintains that these modes of representation are distinct from each other and opens the Transcendental Doctrine of Elements with the following statement:

“Our knowledge springs from two fundamental sources of the mind; the first is the capacity of receiving representations (receptivity for impressions), the second is the power of knowing an object through these representations (spontaneity [in the production] of concepts). Through the first an object is given to us; through the second the object is thought in relation to that [given] representation…The understanding can intuit nothing, the senses can think nothing” (Smith; p. 92, A 50/B 74; p. 93, A51/B75).

Given their different roles and the inability of each to function in the manner of the other, sensibility and understanding are clearly construed as separate from each other. Sellars finds, however, what he believes to be a discrepancy concerning the conception of the two faculties as distinct within Kant’s own philosophy. Kant explains that there is a process by which we come to attain knowledge that involves receptivity, understanding, and something which he labels synthesis operating within the imagination. By virtue of sensibility we receive representations of individual objects, intuitions; these representations are not related to each other, what we have thus far is a “manifold of pure intuition”. It is the job of the imagination operating through synthesis, to isolate that which is common to various representations. In the final stage towards the production of knowledge, the understanding unites the pure synthesis in concepts (Smith, p.112, A 78 – A 79). It is in the curious activity of the imagination, acting on intuitions, that Sellars finds a blurring of the distinction between “sheer receptivity” (not concept-involving) and understanding (conceptual). He writes:

“Kant clearly commits himself to the view that some representations of individuals are intuitions and yet involve a ‘synthesis which, if not a function of the understanding in its role of subsuming representations under general concepts, is certainly no matter of sheer receptivity, but rather of that interesting meeting ground of receptivity with spontaneity which is the ‘productive imagination’… it turns out, most clearly in the second edition (B 151-3), that [productive imagination] is the understanding functioning in a special way.” (SM, p. 4)

Intuitions, in representing individual objects, are not originally supposed to employ concepts. How then, are we led to derive concepts from intuitions? How is it that we can consider these independent representations of individuals in comparison with each other, and pick out their common qualities if we do not yet have concepts? If understanding is separated from sensibility, how does it so to speak reach into the intuitions themselves and grasp those qualities of them which it can then, through the synthetic activity of the productive imagination, unite under general concepts, such as (to use an example from “Sensibility and Understanding”) that of “cube”. As Sellars indicates, metal spinning Kant uses the term “intuition” to refer to both “the representations which are formed by the synthesizing activity of the productive imagination and the purely passive representations of receptivity which are the ‘matter’ (A86; B108) which the productive imagination takes into account” (SM; p 7). Intuitions grasp only individual objects, but, since we do not yet have concepts with which to properly make sense of these individuals, we cannot categorize them as “cube” etc, nor can we make judgments about them (this latter point will be elaborated upon in a moment). On the other hand, these intuitions are meant to guide the understanding in the production of concepts – they are the ‘matter’ for the productive imagination to pick out the qualities common to the manifold of representations. In order for these intuitions of “sheer receptivity” to guide concept-formation in the understanding, we must be able to know something about them. Insofar as this is required, Sellars theorizes that those intuitions which are to be taken up by the productive imagination must have the form of the demonstrative “this” attached to a predicate: “this-cube”, for example. The “this” tells us that what is intuited is an individual object; the “cube” part here does not function as a general concept insofar as it is only used to describe the one object. However, to the extent that there must be something about the representation that enables the productive imagination to discriminate between representations which are and are not alike, do and do not share common qualities, this predication of the object (i.e. as “cube”) is necessary. This is very problematic. Questions are raised along the lines of: how can one label something “this-cube”, “this-book” etc, without first having the concept of “cube”, “book”? The difficulty here, which Sellars identifies, is in holding the general concepts to be ‘genetically posterior’ to the individuated intuitive representations (SM, p. 6).

Sense impressions belong under the heading “sheer receptivity”. They are non-conceptual states of consciousness in which we entertain received representations of objects, and can be exemplified by the expression, to use Sellars’ example: “an impression of a red rectangle” (SM, p. 9) or “an impression of a blue triangle”. What is included seems to be, in Sellars’ view, along the lines of shape and color and not something like a descriptive object term like ‘book’ which implies other characteristics. This is I think why Sellars is careful to explain that in having an impression of a pink ice cube, the term “ice” is actually not used descriptively. This seems to be because the term, the concept of ice itself involves a number of implications and causal properties that are not readily available upon receiving the impression of the pink cube. Ice implies certain things or properties about the cube that are not available to us at first glance and when we are not apperceiving, when the experience is an immediate state of consciousness. We have a manifold of representations in receptivity, but not a representation of a manifold. The ice gets dropped out because it is conceptually rich in itself, with other implications not explicitly apparent in the sense impression we have of it which has the character of not being ‘of’ anything complex.

The fundamental problem here is in locating the link between sense impressions, states of consciousness of a purely non-conceptual nature, and conceptual representation that occurs within the understanding. How do we come to accurately conceptually represent objects having particular qualities in the presence of the objects with these qualities? How does this correspondence come about if our initial representations of those objects do not involve concepts, with which we can represent to ourselves the objects in a particular and sophisticated way? There must be something which is immediately (in the sense of not being mediated by concepts) communicated to us about these objects, which allows us to identify them and class them as potentially being described by way of certain general concepts. Sellars proposes that when we non-conceptually represent these individuals to ourselves, the representation is not completely devoid of all content. On the one hand we already have the knowledge that what we are representing is an individual thing; but on the other hand we can in some sense know something about this object, even if not yet in relation to other objects. We have already posited that the representations of sensibility guide the formation of the conceptual representations of individuals. But exactly how is it that we can correlate objects with their conceptual representations? This is need of explication.

“The answer would seem to require that all the possible ways in which conceptual representation of colour and shape can resemble and differ correspond to ways in which their immediate non-conceptual occasions, which must surely be construed as states of the perceiver, can resemble and differ from each other. 45. Thus, these non-conceptual states must have characteristics which, without being colours, are sufficiently analogous to colour to enable these states to play this guiding role.” (SU, p. 18)

These non-conceptual states of consciousness, sense impressions, are what is to be the explicans for the problem of how we come to conceptually represent objects in the world; through them, we are able to infer things about objects. The qualities they represent are capable of being mapped onto the actual physical qualities of the objects. Sellars urges that these must not be confused with what he refers to as minimal conceptual representations, representations which are minimized both with respect to their content, which is descriptive of the object of perception, and with respect to the weakness of the judgment being made about the object. With regard to this latter point, the verbal exemplification of a minimal conceptual representation (a minimal conceptual report) would be qualified by phrases such as looks or seems which indicate that the perceiver entertains some doubt as to the faithfulness of their representation to the physical object they hold in view (SM, p. 15). An example of a minimal conceptual report might be “It looks to me as though there is a blue and triangular physical object in the vicinity”.

Let us now explore the relationship between sense impression and minimal conceptual reports. Since part of central question concerns what the sense impression inference is posited to explain, this discussion will be helpful because the relationship between these two is defined by this inference. In SM Sellars specifically says “Rather [the sense impression inference’s] primary purpose is to explain the occurrence of certain conceptual representations in perceptual activity. The representations I have in mind are those which are characteristic of what we have called ‘minimal conceptual representations’.” (SM 17, Sellars). So, instead of being identifiable with sense impressions, minimal conceptual representations as such are to be explained by the positing of sense impressions in perception. As Willem deVries puts it in “McDowell, Sellars, and Sense Impressions”, the structures of minimal conceptual responses “can be adequately accounted for only by positing ranges of sensory impressions, the structures of which ranges are isomorphic to the qualitative spaces specified by our minimal conceptual responses and the concepts of which are analogous to the concepts of sensible qualities we report on in our minimal conceptual responses to the world” (deVries, p. 190). His use of the concept of isomorphism in the description is, I believe, very on point and revealing of what Sellars has in mind when he relates these to each other.

McDowell, in Mind and World, and again in Having the World in View takes issue with what he thinks Sellars originally wanted the sense impression inference to explain. McDowell thinks the sense impression inference is posited to explain the following:

How is it that the same claim could be contained in, say, each member of a trio of possible experiences of which one is a case of seeing that there is a red and triangular physical object in front of one, one is a case in which something in front of one looks red and triangular although it is not, and one is a case in which it looks to one as if there is something red and triangular in front of one although there is nothing there at all? (McDowell 1998: 443)

So in McDowell’s view, sense impressions are posited in EPM to explain why these three experiences could contain the same claim. This is a mistake though. In EPM Section 7 Sellars states that in each case where someone has the experience of seeing a red triangle, whether there is actually one or not, s/he has the sensation of a red triangle. But later Sellars says that when analyzing the similarity of these three cases, there are two elements which come into play and seem to be required for their explanatory power: 1) they each involve the idea, the proposition, that the object is red, and 2) “the aspect which many philosophers have attempted to clarify by the notion of impression or immediate experience.” (Sellars EPM, 89). So it seems that while sense impressions are necessary for explaining the similarity of these three experiences, they do not do all the work and cannot alone explain the similarity of the three experiences, the similarity of he propositions involved is needed as well. Futhermore it seems that Sellars never states that what the sense impressions are supposed to explain is the similariy of the claims or propositions, rather they work in concert with the claims made in each case to allow for the similarities of the experiences overall. What is similar in each case seem to be the propositions and the impressions.

Let us move on now to examine why in the broader picture of explaining the similarity between the three experiences McDowell thinks that sense impressions can “drop out”. As McDowell points out, in EPM Section 7, Sellars says that the “proximate cause of such a sensation is only for the most part brought about by the presence in the neighborhood of the perceiver” of whatever the object in question is (having characteristics corresponding accurately to the ‘claims’ involved in the experiences of it). So he implies that there is something other needed to cause a sensation and that the object may not even be present. The ultimate goal is to explain what the cause is for us having conceptual epsidoes, and McDowell does not think it necessary to posit sensation as a necessary step along the way to conceptual episodes. He has us assume similarity at the level of the proximate cause of the sensation which he construes as “impacts from the environment on a perceiver’s sensory equipment” (McDowell 1998: 443-44). So in his picture now we are still at the physical level – impacts from the object in question on sensory apparatuses of the perceiver. deVries qualifies this as a retinal image. McDowell says: “If the impacts are suitably similar, there is nothing puzzling about a similarity between the conceptual episodes they trigger” (Mcd 1998: 443-4). So there is assumed to be a line or link directly from physical episodes or interactions between the environment and the perceiver and on the other end, conceptual episodes, whose content is filled in by the impingement of, let’s say with deVries, light on the retina. However, deVries points out an important hole in this assumptive argument. In some cases exceptions to the rule seem to overturn the whole line of argumentation and here I think an important exception can be found. I do not think that deVries is mistaken in characterizing McDowell’s view of ‘proximate causes’ as retinal images. So let us see how this plays out:

Here the proximate cause would be a retinal image, triggered by some object. Let us say that there are three observers, calling them A, B and C. In the case of observers A and B, they are each presented with a different red object. In each case the object brings about an impact on the observer manifested as a retinal image of red. The retinal image of each observer (separately) induces visual representations of red which provide the foundation for conceptual representations of red, hence, the retinal image is the proximate cause of the conceptual representation. Observer C also ends up with a conceptual representation of involving the quality ‘red’. But in his case, though he also has a visual representation of red, as the other observers do, he does not have as its proximate cause a retinal image of red. Using one of deVries example, let us say that he has instead received a blow to the head which in turn produced a sensation of red. The stimulus to the brain need not be an actual physical object possessing the characteristics that are present in or correspond to the conceptual representation of any given characteristic of the object. It need not even be a retinal image of red. Experiments have been performed whereby certain sensations, even emotions (although this is not really relevant, though interesting in its own right), are produced in a subject by manually stimulating various parts of the brain CITE. Sensations of red can be produced in a subject without even involving the visual system, which is as much to say that the process may be entirely internal. This is not to say that deVries subscribes to the brain-in-a-vat image of representation. He grants that in most cases these proximate causes described by McDowell may in fact be similar and therefore produce similar results, similar conceptual representations. But this does not account for each case. So in McDowell’s view, if the impacts, which here we have exemplified as retinal images, of the environment on the perceiver are similar, then this can explain why conceptual episodes would be similar. But as I have explained, what would be construed as an impact of an object on a perceiver does not have to even come from the surrounding environment of the perceiver. The imagined (as physically existent) object may be retroactively posited by the subject themselves simply because they have a conceptual representation of some characteristic that is usually brought about by an object impacting on the sensory apparatuses of the observer. We can assume that there is some physical stimulus responsible, and usually we are correct, but where there is a level of rigor such as in the present discussion required, we cannot take it for granted that it covers every experience of conceptual representing. deVries also notes, and I think that this is worth mentioning, that sensations and conceptual representations can also be produced by ingestion or other influence of chemicals. There are situations or states which produce sensations of cold which are not actually cold i.e. he tendency of someone with bad circulation to feel coldness and ‘pins and needles’ when there are no such cold or needlelike objects affecting their limbs. Chemicals such as psilocybin and hyoscyamine (found in belladonna) are widely acknowledged to be productive of visual and auditory hallucinations. In these cases, there is not ‘proximate cause ‘ of the brand McDowell describes involved in the process at all. There may be a visual sensation but the retinal image is absent. The brain is impressionable not just by objects of sensory experience, but also by chemicals which induce experiences of objects which do not actually exist but are hallucinatory in nature, the representations of which are brought about by chemicals that in a sense mimic something which would in normal circumstances bring about the same reaction as the chemically induced hallucination. Summarily, brain activity and specifically conceptual representation is not solely brought about by the actual visually perceived objects, but by things (blows, chemicals) which act on the brain in other, perhaps subtler ways. As deVries says, and here he explicitly links the discussion up with conceptual episodes: “The proximate cause of sensations can themselves be many and various…What they all have in common is only that in given circumstances they cause one and the same effect: a sensation of red. And it is sensations of red that tend to evoke conceptual representations involving the concept of red.” (deVries 2006; 191).

Now that we have disposed of the objection that sensations look like ‘idle wheels’, let us move to that which McDowell believes to be Sellars’ new argument for the indispensability of sense impressions. McDowell says, in Having the World in View (Woodbridge Lectures) that Sellars, in Science and Metaphysics, posits sense impressions for a different reason – on transcendental grounds. And he believes he does this to avoid making what he thinks was the mistake he just pointed out. We must get clear about what this means. Sense impressions are seen to be part of the structure of any perceiver in any given historical, whatever, context relating to their environment. This is something that transcends any given linguistic foundation or schema and the particularities of the context it falls within/is integrated into but it common to the structure of all contexts where there is a relationship – involving linguistics – between perceiver and environment. This is posited on/ is supposed to tap into epistemological norms that are present in any of the described contexts. In Having the World in View McDowell writes: “Sellars’s ‘sense impression inference’ is a piece of transcendental philosophy in the following sense: it directed at showing our entitlement to conceive subjective occurrences as possessing objective purport… he undertakes to vindicate the objective purport of conceptual occurrences from outside the conceptual order.” (HWV; 17). deVries does not agree with this view of Sellars’ attempted task

But if we could hold alongside each other a conceptual representation of a car – a particular car –as blue, and then be simultaneously presented with the car but as it really is, which is not blue but, rather, green, then could not the sense impression we receive of the car as its true color do the corrective work which would counter my mistaken, inaccurate, conceptual representation of the car? I think there is something here that can lead us to say that sense impressions are not phenomenologically idle. deVries quotes Sellars in his rebuttal of McDowell’s position and I think this quote in demonstrative of the difference that Sellars does in fact believe exists between conceiving of and actually perceiving: “…imagining is an intimate blend of imaging and conceptualization, whereas perceiving is an intimate blend of sensing and imaging and conceptualization” (IKTE in KTM, 423). Why does McDowell, as deVries writes, attribute to Sellars the idea that conceptual episodes can be phenomenologically colorful? Is it really off the mark to suppose that conceiving of something, or imagining it, must, under normal circumstances stand trial to the state of the imagined object when the perceiver is presented with it. I was thinking of a song constructing the notes and the progression as such when the song came on the radio and I was alerted to the error of my conception of the song. There is a difference between conceiving of and actually perceiving but it is difficult to locate. deVries thinks the difference hinges on the notion of conceptions as being phenomenologically colorless, etc.


the problem, as Sellars explicitly says, that the sense impression inference is used to explain why/how we come to have minimal conceptual representations. In a broader sense, sense impressions can explain how we come to conceptually represent objects at all, how our states of consciousness, conceptual episodes about objects have any content at all. McDowell conceives Sellars more narrowly. His is an interesting problem in its own right.

It seems that what McDowell is guilty of here is an, either conscious or unconscious, oversimplification of Sellars’ reason for postulating sense impressions.

We need sense impressions because we need something that is similar enough to objects of our environment to allow us to accurately represent to ourselves these objects. This is what Sellars means when he says that sense impressions, in terms of their characteristics, are analogous to those things which they represent. Without themselves being spatial or temporal – what Sellars means when he says that we have impressions (this seems obvious I suppose but in the interest of clarity…) is not actually meant to be understood to say that impressions are physical things which are located inside us – sense impressions represent the spatio-temporal characteristics of objects. When I perceive a textured object my impression of it is not actually textured; but the fact that when I conceptually represent this object to myself, or to word it differently, when I have a conceptual episode involving this object, I represent the object to myself as textured needs to be accounted for. Sense impressions do not involve concepts, the content of my sense impressions is non-conceptual, yet the nature of my conceptual representations and the fact that they employ concepts which both accurately represent and categorize objects – this fact provokes the question: how are these conceptual representations or episodes shaped or formed so as to represent as they do? Here Sellars posits non-conceptual sense impressions which play a guiding role in shaping and forming these conceptual representations. Minimal conceptual representations are given by impressions and provide a very basic grounding/foundation for ‘full-blown’ conceptual episodes. Minimal conceptual representations are in fact “conceptual” precisely because they do include content which is descriptive of actual physical things. For example, a minimal conceptual representation might be of a “red and octagonal surface with white patches”, where as a full-blown conceptual representation might be of a “stop sign”. ??? In the case of sense impressions however we can have an impression of a red rectangular object but not, as Sellars give the example “impression of a man lurking in the corner”; this later employs concepts, concepts of ‘man’ and ‘lurking’ etc which are themselves conceptually rich. Man implies certain other conceptual characteristics such as bipedal, cordate, etc. red and rectangular are irreducible to anything else. Red and rectangular in this instance do not refer to anything else, only the individual in question, the one instantiation. The general concepts of redness and rectangularity are not actually employed because we refer only to the unique instance. In this sense, impressions are narrowly concerned with their subject; they so to speak have the blinders on to anything outside of their singular instance of concern.

music and arts blog The last thesis, of the five, which deVries believes McDowell has attributed to Sellars (wrongfully/on a way which is inadequately supported is the claim that “Dispensing with sense impressions costs us nothing in our ability to understand experience; they are phenomenologically idle” (deVries; 183).

McDowell does not believe that Sellars has justified the necessity of positing sense impressions which actually do some work which neither sensory-physical interactions nor conceptual representations, on their part, are capable of doing. McDowell thins that because conceptual episodes already contain the notion of, for example, something being colored (a table being brown and the like), sense impressions do not need to be added to the picture to ‘color’ the conceptual episode or provide along those lines. The whole process beginning with receptivity towards physical stimuli and ending with complex concept formation or employment of concepts can be explained simply by these two aspects of coming to know. We do not need anything between the object’s impingement on our sensory apparatuses – a physical process – and conceptual episodes. As McDowell writes “In [Sellars’s] view, conceptual episodes of the relevant kind are already, as the conceptual episodes they are, cases of being under the visual impression that such-and-such is the case. It is not that as conceptual episodes they are phenomenologically colorless, so that they would need to be associated with visual sensations in order that some complex composed of these conceptual episodes and the associated visual sensations can be recognizably visual. These conceptual episodes are already, as the conceptual episodes they are, shapings of visual consciousness” (McDowell 1998: 442). This whole business of conceptual episodes being phenomenologically colorful (he denies that they are phenomenologically colorless), deVries interprets as McDowell’s claiming that they are “intrinsically qualitative because of and through their conceptuality” (deVries; 195). In protest to this, deVries calls attention to parts V and XI of EPM where, he says, it becomes very clear that Sellars believes conceptual episodes to have no ‘particular phenomenology’. Maybe this can be found here…PSIM

But I think a stronger example might be the one which deVries moves on to cite “Roughly[,] imagining is an intimate blend of imaging and conceptualization, whereas perceiving is an intimate blend of sensing and imaging and conceptualization.” Though this gets at the matter rather indirectly. deVries goes on to quote Philosophy and the Scientific Image of Man, in which, he argues that the “difficulty…which stands in the way of the identification of thought with cerebral process, arises from the mistake of supposing that in self-awareness conceptual thinking presents itself to us in a qualitative guise” (PSIM; 32).

In trying to get at what is the common descriptive content between the trio of experiences, seeing and qualitative and existential lookings, which has been labeled immediate experience, Sellars entertains one possible way out which involves assimilating “Jones has a sensation of a red triangle” to “Jones believes in a divine Huntress”. But there is a problem with the drawing a parallel between having a sensation of an (ostensibly) extended object and having a belief in a nonextenstional object.

Now, I think that most contemporary philosophers are clear that it is possible to attribute to the context …sensation of… the logical property of being such that “There is a sensation of a red triangle” does not entail “There is a red triangle” without assimilating the context “…sensation of…” to the context “…believes in…” in any closer way. For while mentalistic verbs characteristically provide nonextensional contexts… not all nonextensional contexts are mentalistic. (EPM; 55).

“Indeed there is no reason why [sensations] should be assimilated to any of these [nonextensional contexts]. “…sensation of…” or “…impression of…” could be a context which, though sharing with these others the logical property of nonextensionality, was otherwise in a class by itself” (EPM, 56). Sellars seems to be alluding the idea that sensation involves something beyond mere belief, because, though they may be directed toward or about equally nonextensional objects, the sensation actually does refer to physical objects (most of the time) and has to do with them on a regular basis, and therefore, the intentionality of each is different. Belief in a divine Huntress is something which cannot possibly be proven or held to the testimony/tribunal of a commonly experienced physical reality. In believing it is assumed, first of all, with reference to fact that it is a belief we are dealing with here, that it cannot be proved or disproved. Secondly, the object of the belief is not something as concrete and demonstrably indicated as an object or quality such as redness of an object. In sensations, we can be held to the tribunal of physical reality – the objective environment. We can see that Sellars considers this when he carefully describe how “the fact that x, over there, looks to Jones to be red would be a seeing, on Jones’ part, that x, over there, is red, if its propositional content were true” (EPM; 54) In having a sensation of red, Jones is referring to some object possessing this quality in the objective environment; in order to prove or disprove the truth of the what the perceiving is ostensibly claiming we must appeal to that objective reality and there we will find our answer. In the case of believing, first of all there is no intention or necessity to prove or disprove, furthermore, the indicated object is not something that would necessarily be revealed to us through our senses and therefore, through inspection of the physical environment.

… I think this harkens back to the distinction Sellars made in PSIM between imagining and perceiving. There are simply different elements and demands for truth made in each case.

So sense impressions or immediate experiences as he here labels them, are reducible to neither conceptual episodes NOR beliefs. Because they do not have the quality of being mere beliefs, abstract, and do have this special relationship to the external world which beliefs do NOT enjoy, sense impressions are both not phenomenologically idle and are needed in the process which goes on between physically registering physical objects (lights patterns on retinas…) and having conceptual episodes involving the objects. They do enjoy a relation to the external world while being nonconceptual. But we still must explore what the conceptual aspect, within the understanding, does leave out that needs to be accounted for with sense impressions.

Quotes from Sellars:

“…the correlation of the correct conceptual response with objects perceived is as much in need of explanation as the correlations of conceptual responses with abnormal perceptual situations. It is not enough …to explain the latter by saying that the proximate stimulus caused, e.g., by a black object in normal circumstances, is the same as that caused by a red object in abnormal circumstances.” (SM 18, Sellars)

McDowell does not seem to acknowledge that the sense impression inference, as Sellars states in Science and Metaphysics, is put forth to explain minimal conceptual representations. In the past sense impressions have been confused with minimal conceptual representations as Sellars points out. This is because neither of them involve physical experiences or states directly. Minimal conceptual representations however, as the name suggests, involve concepts while sense impressions do not. Nothing can be assumed in sense impressions either and this in part is why Sellars is careful to explain that in having an impression of a pink ice cube, the “ice” is actually removed, or what is better/clearer, simply not there. This seems to be because the term, the concept of ice itself involves a whole mess of implications and causal properties that are not readily available upon receiving the impression of the pink cube. Ice implies certain things or properties about the cube that are not available to us at first glance and when we are not apperceiving, when the experience is immediate – an immediate state of consciousness. We have a manifold of representations in receptivity, but not a representation of a manifold. The ice gets dropped out because it is conceptually rich in itself, with other implications not explicitly apparent in the sense impression we have of it which has the character of not being ‘of’ anything complex.

“Objects are given to us by mean of sensibility, and it alone yields us intuitions…” (Kant p. 65, A 19)