without rejecting (P1), i.e, without rejecting Looks-indexing. One difficulty in trying to reject (P1) is to do it without weakening the very idea of p-representation. As Wilson says, “the representationalist faces the difficulty of specifying how, if not in virtue of appearances, Availability might be satisfied”204. Nevertheless, I suspect that this is exactly what McDowell wants with his defense of a content-awareness view on perceptual experience. So now it is time to go deeper into McDowell’s notion of conceptual content discussed in the previous chapters, especially regarding his new position.
“Availability question: What makes p-representational content recognizable, or cognitively available, to the subject?”205.
I understand that TFKG is McDowell’s attempt to give his own answer to something along the lines of the Availability question. So my aim now is to unpack McDowell’s thoughts on the issue, in the context of his debate with Travis.
As McDowell’s Travis on Frege, Kant, and the Given (TFKG) title obviously suggests, the text aims to approach the issues involved in his exchange with Travis in light of Frege, Kant, and the Myth of the Given. A good way to make sense of what is at stake in the debate is to start from the different readings of Travis and McDowell of the following remark of Frege:
“Sense impressions are certainly a necessary ingredient of sensory observation, and these are part of the inner world (…). These by themselves do not open the outer world for us. Perhaps there is a being that only has sense impressions, without seeing or feeling things. Having sense impressions is not yet seeing things (...). Having sense impressions is, to be sure, necessary for seeing things, but not sufficient. What must still be added is not something sensory. And it is just this which unlocks the outer world for us; for without this non-sensory thing each of us remains shut up in his inner world”206.
McDowell and Travis dispute this passage as endorsing their own views. My task is to explain how each and one of them interpret what Frege means here, and then clarify how this relates to Travis’s Argument from Looks as well as to McDowell’s rejection of Looks-indexing.
206Cited atTFKG, 30.
205Wilson 2018:217.
Frege establishes a notional difference between sense impressions and seeing.
On the one hand, “sense impressions” should be understood as the sensory awareness of objects, say, a red flower. On the other hand, Frege takes “seeing” as a result of the entering of conceptual capacities that enable one to intelligibly grasp a fact such asthat this flower is red. So the “non-sensory ingredient” - i.e. the actualization of cognitive capacities - would be what enables one not to be “blind” to such facts, as it would happen with the imagined subject who may lack the capacity to properly “see,” and not just “sense” things.
As Travis reads Frege, the only objects of sensory awareness would be sense impressions, such as that of a red flower. As we already saw, that is surely compatible with Travis’s idea that conceptual capacities operate only downstream from perceptual experience. That means for him, in the present context, that related representational contents of judgments are not available to the subject in experience. In the wake of Frege, Travis makes a distinction between what would be two senses of seeing: namely
“O-seeing” and “T-seeing”. O-seeing is seeing objects - something that one can be sensorily aware of; T-seeing is seeing in the sense of one seeing that things are thus and so- what one does when exercising conceptual capacities.
In the same spirit of the “sunset” example, Frege remarks: “But don’t I see that this flower has five petals? One can say that, but then uses the word ‘see’ not in the sense of mere sensing things via light, but one means a thought or judgement connected with that”207. As Travis stresses, what Frege means here is that “what is operative in seeing-T” - what distinguishes it from O-seeing as non-perceptual
207Cited at Travis 2018c:239.
accomplishment - is a non-sensory (nichtsinnliche) ingredient, just the sort of ingredient which, he tells us, ‘unlocks an outer world for us’”208.
McDowell agrees with Travis’s Frege that contents cannot be objects of sensory awareness: “Travis says, correctly, that the thing to think, the thinkable, specified in the
‘that’-clause that is the complement of an expression of T-seeing (...) is not related to sensory awareness in the way things like flowers can be” (TFKG, 34). In effect, as we saw, this is part of McDowell’s new position: the propositional content - what can be expressed by a “that”-clause - no longer figures as an object of sensory awareness.
However, McDowell thinks that his interlocutor mistakenly reads Frege in considering that his distinction between O-seeing and T-seeing would not admit any involvement of thinkables in perceptual experience. To put it another way, Travis’s Frege would think that the only way thinkables may be involved in perceptual experience would be by being objects of sensory awareness, which, in fact, would be unacceptable. In Wilson’s terms, Travis’s point would be that T-seeing (in the present sense, what can be the content of sensory awareness) cannot satisfy Availability, insofar as what might satisfy Looks-indexing could not be something with a non-sensory character. Nonetheless, what McDowell wants is exactly to give expression to the idea that despite the fact that thinkables could not be objects of sensory awareness they could still be contents of sensory awareness.
From that perspective, McDowell sees himself as able to offer a way of rejecting Looks-indexing without rejecting Availability209. But if that is the case, in what sense do contents, with their non-perceptual character, be available to the subject in perceptual
209For a similar view, see Byrne 2009.
208Travis 2018c240, original emphasis.
experience itself, if not in virtue of looks? Well, McDowell thinks that Availability is a condition for avoiding the Myth of the Given. And if it is so, he argues that it is a requirement for one’s “judgments [to] be rational in the light of our sensory awareness [that] one [must] recognizes a seen object as being [so] on the basis of ways it is presented as being in one’s visual experience of it” (TFKG, 34). McDowell’s bet is that avoiding the Myth presupposes Availability, in spite of Travis’s demand that p-representational contents must have a non-perceptual character. That can be stated in the form of an argument:
Argument from McDowell’s Frege
(0) One can avoid the Myth of the Given if and only if one supposes that the contents of perceptual judgments can only make rational connections with other thinkableor judgeable contents.
(1) The object of sensory awareness is expressible as a representational content p[via “a thought or judgement connected with that”].
(2) For the rational connection of the contents of perceptual judgments with the objects of sensory awareness to obtain, sensory awareness must exhibit contents with athinkableor judgeable character [asper(S0)].
(3)Thinkableor judgeable contents cannot be objects of sensory awareness.
(C1) (From P1 through P3) Thinkable or judgeable contents must be contents and not objects of sensory awareness.
(C2) (From C1) Thinkable or judgeable contents can be contents of sensory awareness.
The Argument from McDowell’s Frege is in line with McDowell’s therapeutic stance on philosophical problems, in that specific case “the fundamental problem of perception” which is stated by Travis as follows: “how perception can make the world bear for us on the thing to think”210. McDowell’s therapeutic way of dealing with philosophical problems is inspired by Wittgenstein’s recommendation that they should be dissolved instead of solved, that is to say, that they “should completely disappear”
(PI133, original emphasis). Broadly speaking, Wittgenstein’s idea is that philosophy can be, at best, a description of, not a theory about, its topics. Philosophy, then, should not be meant to give any foundations; for Wittgenstein, philosophy instead “leaves everything as it is” (PI124). From that perspective, recall:
Truism: The way a subjects judges things to be is the way experience Emakes things available tos.
Returning to the Introduction, for McDowell, the conclusion that one judges that things are thus and so because one perceives that things are thus and so reveals not a philosophical theory but a truism. McDowell’s way of thinking about the nature of perceptual experience is meant to illuminate that the “fundamental problem of perception”, once seen from such a Wittgeisntenian perspective, after all, manifests itself as actually not being a problem. His overall strategy to “bypass the anxieties of traditional epistemology” (MAW, 112) is to take off from this kind of a truism in any description of perceptual experience. Putting the point the other way round, McDowell
210Travis 2013c:242, original emphasis.
sees that the relations between mind and world appear as a philosophical problemonly if one supposes that external reality “would have to break out through a boundary that encloses the sphere of thinkable content” (MAW, 39).
One may thus infer that McDowell considers this minimal necessity to avoid the Myth of the Given as a supposition on the part of his opponents, asperS0. As Thornton nicely puts it, McDowell’s way of thinking,
“(...) is primarily addressed to a particular philosophical audience: those who are subject to a particular philosophical discomfort as a result of subscribing to particular philosophical intuitions (...) [and that] [t]o that extent it does not articulate a freestanding context-independent philosophical theory but rather aims to dissolve a particular felt tension”211.
With regard to the Argument from McDowell’s Frege, then, S0 is meant to stress that to avoid the Myth, according to McDowell, one must take sensory awareness as somehow having a conceptual nature. If conceptual capacities imply thinkable or judgeable contents, and if thinkable or judgeable contents cannot be objects of sensory awareness, as per P3, thinkable or judgeable contents must, and therefore can, be contents of sensory awareness. McDowell’s point is that conceptual capacities could be in operation in experience in a non-perceptual way. If so, there would be a way for the Representationalist to reject Looks-indexing: she should take p-representational contents to be recognizable (available) in virtue of the involvement of non-perceptual thinkable contents in sensory awareness. Accordingly, if P1 in the Argument from looks states that “if visual experiences were p-representational then their content would be
211Thornton 2004:211.
recognizable in virtue of how, in experience, things perceptually appear, or look”, Looks-indexingis false.
Now, suppose that McDowell’s way of rejecting Looks-indexing is sound. Even so, Travis contends that S0 in the Argument from McDowell’s Frege incurs a categorial mistake. If conceptual capacities are involved in experience, it must present worldly items as, somehow, already falling under a generality. As Travis stresses:
“[A] concept as such (...) has a certain kind of generality. (...) The key feature of the conceptual, on its present understanding, is that for anything conceptual there is a specific form of generality intrinsic to it. There is then a range which is the range of cases, or circumstances, which would be ones of something instancing that generality”212.
According to Travis, what instances a certain generality is what Frege has called “the particular case”: “A thought always contains something which reaches beyond the particular case, by means of which it presents this to consciousness as falling under some given generality”213. The difference between the particular case and generality is what Travis calls the “Frege’s Line”. For Travis, things that are perceivable, like a red light or a car, fall on the left side of the line. In its turn, those things such asthat I see a red light run fall on the right side. Furthermore, Travis connects this distinction with a distinction between conceptual and non-conceptual things:
“To the right of Frege’s line is the conceptual. What is there to the left? What instances (first-order) conceptual generalities. Such as that piece of meat. A piece of meat is not in the business of being instanced. So treating it would be bad grammar. A fortiori there is
213Cited at Travis 2018a:43.
212Travis 2007:125.
no range of cases which instance it. It is not conceptual. Of course, for any given piece of meat, there is a concept of being it. Being a concept, this does have its range of instances: the meat in the butcher’s case, the meat in butcher paper, the meat on the rug, and so on”214.
This distinction between generality and the particular case is central to the debate between McDowell and Travis. Next, I will discuss that issue in more detail.