The "unfortunate, familiar" line of thought that "scarcely anyone" would accept sounds pretty good. And the response:

Rather they would take the contrapositive of the argument, and reason that since the foundation of empirical knowledge is the non-inferential knowledge of such facts, it does consist of members of a class which contains non-veridical members.

While in some sense logically unassailable, has the flavor of an appeal to consequence. That is, I can't understand the motivation for taking this line, except from deciding a priori that a certain conclusion is unacceptable and so rejecting any argument that reaches it.

though for it to strike them as it does, they must overlook the fact that if it makes sense to speak of an experience as veridical it must correspondingly make sense to speak of it as unveridical.

Hmm. Why does Sellars think this?

He's assuming that for the concept of sense data to throw light on the foundations of empirical knowledge, sense data must itself be knowledge — Terrapin Station

No, he's saying that if sensation itself does not constitute knowledge, then it's appropriate to ask in what sense sensation grounds knowledge. He acknowledges a couple times that having a sensation might be a logically necessary condition to coming to empirical knowledge, and thus serve as a 'foundation' in some sense other than being a 'cognitive fact' (i.e. sensation constituting knowledge in of itself).

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. Thus, the non-inferential knowledge of particular matter of fact might logically imply the existence of sense data (for example, seeing that a certain physical object is red might logically imply sensing a red sense content) even though the sensing of a red sense content were not itself a cognitive fact and did not imply the possession of non-inferential knowledge.

Although I do think it's odd for him to equate the notions of 'being a cognitive fact' and 'constituting knowledge.'

Hmm. Why does Sellars think this? — The Great Whatever

Here, I'll take a swing at it. I think Sellars means something like this:

"I had experience x."
"I had experience x, and it was veridical."
If the former is the same as the latter, then the second clause in the second statement is vacuous and veridicality adds nothing to the discussion. If, on the other hand, the second clause in the second statement is not vacuous, then these are two different statements. It follows that "Having a veridical experience of P" is something over and above "Having an experience of P." But this implies that veridicality is "optional" - it makes sense to speak of the experience as lacking veridicality.

That can't be right, though, because the sense datum theorist doesn't claim that all things of the form 'I had experience X' have no duality between veridicality and non-veridicality, only that there's a certain class of experiences that it makes no sense of to call non-veridical.

The question is why Sellars thinks, of that class, why this must be sensical. One might think (quite reasonably – this seems to be a recurring trend in philosophy for thousands of years) that claims grounded in experience that are about that experience itself self-verify, and so it makes sense to speak of them as veridical but not as non-veridical. The whole point of non-veridicality, if you like, is that there was a seeming that led you to believe something that wasn't so. But if what 'was so' was merely the seeming, this possibility seems to become incoherent. Sellars seems to want to say that this makes veridicality incoherent too, but I'm not sure what leads him to think this.

He explains this in this passage:

Now it is, of course, possible to delimit subclasses of ostensible seeings, hearings, etc., which are progressively less precarious, i.e. more reliable, by specifying the circumstances in which they occur, and the vigilance of the perceiver. But the possibility that any given ostensible seeing, hearing, etc., is non-veridical can never be entirely eliminated. Therefore, given that the foundation of empirical knowledge cannot consist of the veridical members of a class not all the members of which are veridical, and from which the non-veridical members cannot be weeded out by 'inspection,' this foundation cannot consist of such items as seeing that the facing surface of a physical object is red and triangular. — Wilfrid Sellars

I don't at all buy his "therefore" in that passage, though. It seems like a non-sequitur to me.

Right, it occurred to me after typing that that I may have fallen into the exact same error as Sellars. Let me look a little further into this...

You are saying that the claim, "I saw a red triangle," is self-verifying if it's grounded in the experience of seeing a red triangle. My first instinct is to assert that we can be wrong about our experiences, but I sense that this would miss your point. You're saying that, if claim A is solely about experience B, and claim A is solely grounded in experience B, then A can't be non-veridical. That is to say, if I have experience B, I can't infer a falsehood about experience B solely from that experience. Do I have you right, or at least, mostly right?

EDIT: d'oh, seems like you edited your post! Lemme read it again real quick and re-assess.

As I read this passage Sellars is speaking in the voice of the sense datum theorist, and has not yet reached the sense data punch line, viz. that the foundations of empirical knowledge must depend upon a class of things that don't have veridical and non-veridical members that are in principle indistinguishable, and that sense data or appearings are such an appropriate class. It's only afterward that he makes the objection, that the possibility of veridicality implies the possibility of non-veridicality, and this, so far as I can see, he does not explain or justify, but merely says. I don't expect to find an answer for why he believes it in the paper, because I don't think he gives one. I'm just speculating on why he might think that.

Yeah, that sounds about right. I think you could make a case for delusion, linguistic incompetence, etc. creeping in even at this level, to get non-veridicality back in the door, but this seems to rely on people not being within their wits, which is not usually what epistemologists have in mind in these sorts of scenarios. The point is supposed to be that with claims about objects, there's no way to tell the difference between veridicality and non-veridicality in any particular case, not that one might slip up and miss the difference, misunderstand their own language, etc. In other words, even a perfectly alert and rational person could never tell the difference in principle between a veridical and non-veridical experience of an object in at least some cases, whereas it seems absent such lack of faculties, that kind of mistake isn't possible in the experience-oriented case.

Well, but the objection just turns out to be the old "we can't be 100% certain that any given sense datum is veridical," as if 100% certainty should be the goal, should be required, or is even possible.

Again, I don't think Sellars thinks that; I think he's speaking in the voice of the sense datum theorist.

And in fairness to the sense datum theorist or skeptic generally, the point is not 'you can't be 100% sure,' but 'your epistemological commitments force you to say you can't even be 1% sure.' So it goes if you open the door to nonveridical perceptions that can't be distinguished from veridical ones, but believe experience is foundational to knowledge. The sense datum theorist sees a way to rescue knowledge via experience.

I don't at all agree with your interpretation, and I don't think the passage makes sense on your interpretation. He's specifying that sense datum theorists point out that they can achieve progressively more reliable claims, but the problem in his view is that they can't rule out the possibility of non-veridicalness.

The whole point of non-veridicality, if you like, is that there was a seeming that led you to believe something that wasn't so. But if what 'was so' was merely the seeming, this possibility seems to become incoherent. Sellars seems to want to say that this makes veridicality incoherent too, but I'm not sure what leads him to think this. — The Great Whatever

Okay, I'm gonna conjecture a little hypothetical that might illustrate Sellars' intuition on this matter. I'm gonna ramble a little, but I'm still in the "clear the area" phase here. Let me see what I can do here...

Suppose I make this claim: "I experienced a red triangle." Where is the self-verifying experience here? The experience of the red triangle is not self-verifying, since I very well could have seemed to see a red triangle even if there wasn't one (bad lighting or whatever, to use one of Sellars' examples). You could say that it's not the experience, but the claim, "I experienced a red triangle," that is self-verifying, but that claim is grounded on my experience of remembering a former a experience.

But that isn't being completely charitable to what you've put forth (also with deference to your point about delusion/dream/linguistic skepticism not being relevant here). Let's try and push it a little further: while having the experience of seeing a red triangle, I say, "I seem to see a red triangle right now." You're saying I can't be wrong with that, yeah? Provided that the claim I'm making is totally grounded in just that experience, and nothing else, then I can't be wrong.

I can't think of a problem with this offhand, so let me fiddle with it a little. What would it take for me to be wrong here? I would need to have experience B, and infer something false about B, with only B as grounds for my claim. Obviously, I can lie about it, but I take it as implicit here that it's self-verifying for me only. I guess the only way out for Sellars here is to ask for an account of what it means for a claim to be grounded in an experience. Perhaps he could say that there is something a little fishy about saying, "For any experience I have, it is self-evident for me that I am having that experience." Does one have an experience of self-evidence here? Or is this more of a logical thing, inferred from what it means to have an experience?

Again, in that section he's speaking in the voice of the sense datum theorist, and hasn't reached the conclusion (of the sense datum theorist) yet, which is that sense data constitute precisely such an empirical domain that rule out the possibility of non-veridicality.

His objection is then in one sentence, implying that the possibility of veridicality implies the possibility of non-veridicality.

Again, in that section he's speaking in the voice of the sense datum theorist, — The Great Whatever

I don't agree with this. He's explaining how a sense datum theorist might approach the problem, but he's critiquing the approach in that same paragraph. The comment about not being able to remove the possibility of non-veridicalness is part of Sellars' critique. It's not what a sense datum theorist would say.

That objection of not being able to remove the possibility of non-veridicalness is the old cliche of focusing on not being able to achieve certainty. Despite the fact that the sense-datum theorist is saying that increased reliability etc. is possible, Sellars is focusing the fact that it's still possible to be wrong.

I didn't agree with your earlier contrary interpretation about Sellars assuming that sense data being knowledge is necessary for it to throw light on sense data as a foundation for empirical knowledge, either, but I don't want to argue about every single interpretation, because probably that's all we'd be doing in this thread in that case.

"That seeming led me to believe that I had that same seeming." I could question this claim, but, if it's true, then the belief (that I had that seeming) can't be false.

How's this: "It makes sense to talk of that belief as being unveridical because it makes sense to talk of the claim, "This belief was grounded in that seeming" being false."?

Alright, man, but I think this is less an issue of interpretation and more of reading comprehension. That is, I think you're misreading the paper on a very surface level.

That is, I think you're misreading the paper on a very surface level. — The Great Whatever

Well, and you should be able to guess that the thought is mutual.

With the second section, "Another Language," I don't have any major objections yet, at least barring a few comments in the last few paragraphs that I'm not sure I understand. The point of most of the second section is completely lost on me at the moment, though. . . it seems like a very laborious way to not say much beyond the (maybe kind of sarcastic/spoofy) idea that we could read sense data sentences as being akin to logical variables (with quantification). But Sellars promises that it's leading up to something, so we'll see as we go on.

According to Sellars, sense-data are not epistemic elements. Empiricists thought that sense-data are directly 'given' to epistemic subjects. Sellars thought that sense-data should be constituted by linguistic ways to be epistemic. This is called Sellars' psychological nominalism. According to psychological nominalism, language is prior to sense-data.

Well, regarding the first point of disagreement, I've provided what I think is pretty damning textual evidence against you. To repeat, this:

He's assuming that for the concept of sense data to throw light on the foundations of empirical knowledge, sense data must itself be knowledge

Cannot be right, given that he says this:

Yet it would be hasty to conclude that this alternative [that is, alternative (a), that sensing is not itself knowing] 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. Thus, the non-inferential knowledge of particular matter of fact might logically imply the existence of sense data (for example, seeing that a certain physical object is red might logically imply sensing a red sense content) even though the sensing of a red sense content were not itself a cognitive fact and did not imply the possession of non-inferential knowledge.

That gets as close to a straight denial of your claim from the text itself as I can think of. He also says this:

He can abandon A, in which case the sensing of sense contents becomes a noncognitive fact -- a noncognitive fact, to be sure which may be a necessary condition, even a logically necessary condition, of non-inferential knowledge, but a fact, nevertheless, which cannot constitute this knowledge.

Now being a logically necessary condition of non-inferential knowledge clearly is a way in which something 'throws light' on the foundations of empirical knowledge. But this is something that he allows of sense data, even if sense data is not itself knowledge.

So your claim that he thinks the only way something can 'throw light' on the foundations of empirical knowledge is for that thing to actually constitute knowledge, is wrong.

---

Regarding the second disagreement, you seem to be implying that Sellars is critiquing the sense datum theorist for allowing uncertainty in their empirical foundations, which is just not the point of this first section. The point is rather that sense datum theorists have according to him been confused about whether or not sensing something is an inherently epistemic or cognitive fact.

Furthermore, the two paragraphs in which he talks about 'progressively less precarious' subclasses of sensations are in quotations, followed by these words:

This unfortunate, but familiar, line of thought runs as follows:

This signals that Sellars is taking the viewpoint of the sense datum theorist in order to criticize it, not expressing his own view as you imply above. He then concludes by saying that the sense datum theorist is subject to confusions, not that he fails to secure some sort of apodicticity in his position. The latter is simply not what this section of the paper is about.

Okay, if you want to endlessly argue with me about interpretation, let's do that. That will be fun.

(1) you'd need to argue that the non-inferential knowledge of particular matter of fact might logically imply the existence of sense data is sufficient to amount to noncogntive sense data throwing light on the foundations of empirical knowledge, especially in light of the fact that he says, "It would seem, then, that the sensing of sense contents cannot constitute knowledge, inferential or non-inferential; and if so, we may well ask, what light does the concept of a sense datum throw on the 'foundations of empirical knowledge?"

So your argument that Sellars counts non-inferential knowledge of particular matters of fact might logically imply the existence of sense data is sufficient for noncognitive sense data throwing light on the foundations of empirical knowledge is ________?

Re this:

He can abandon A, in which case the sensing of sense contents becomes a noncognitive fact -- a noncognitive fact, to be sure which may be a necessary condition, even a logically necessary condition, of non-inferential knowledge, but a fact, nevertheless, which cannot constitute this knowledge. — Wilfrid Sellars

The "a logically necessary condition" is simply a reference/restatement of what you'd just quoted. Again, that reason that makes it sufficient to throw a light on the foundations of empirical knowledge in Sellars opinion, in your view, is _______?

you seem to be implying that Sellars is critiquing the sense datum theorist for allowing uncertainty in their empirical foundations, — The Great Whatever

I didn't say that. I said that he's critiquing the approach arguing for increased reliability, etc. on the grounds that it's still (merely) possible for sense data to be wrong about "sense content" (a horrible term for what he's talking about there on my view).

I'm also not saying that my comments about this are in the context of the "point" of the first section.

Regarding the second disagreement,

Actually rereading that section, I'd agree that Sellars is not speaking for himself in the comment about the possibility of being wrong about sense content, but I still don't agree that he's presenting the view of sense data theorists either. In fact, I'm not sure whose view he's presenting--he might not be accurately presenting anyone's view in a way that anyone would agree with. He says right before that:

Unfortunately, the idea that there are such things as sensations of red triangles . . . seems to fit the requirements of another . . . line of thought so well that it has almost invariably been distorted to give the latter a reinforcement without which it would long ago have collapsed. — Wilfrid Sellars

Right after which he says:

Thus baldly put, scarcely anyone would accept this conclusion. — Wilfried Sellars

So who is he supposedly paraphrasing "baldly" anyway?

though for it to strike them as it does, they must overlook the fact that if it makes sense to speak of an experience as veridical it must correspondingly make sense to speak of it as unveridical.

Hmm. Why does Sellars think this?

He wants to formally separate sense datum from our conception of sense datum, the former non-cognitive and non-intelligible and the latter both cognitive and intelligible.

So:

"I had experience x."
"I had experience x, and it was veridical."

Where "I had experience x", is a report. The actual experience, at its own level of immediacy has no valuation, no intelligible content. We may be unable to be wrong about having an experience, but that does not give it any value, only its report can be judged true or false. To claim that immediate sense experience has any true or false value is to commit the naturalistic fallacy.

Again, I think this is a matter of reading comprehension, not interpretation, which is important to a reading group.

So your argument that Sellars counts non-inferential knowledge of particular matters of fact might logically imply the existence of sense data is sufficient for noncognitive sense data throwing light on the foundations of empirical knowledge is ________? — Terrapin Station

There's no argument to be had: Sellars outright admits that even if sensation of sense data doesn't constitute knowledge, it might nonetheless shed light on the foundations of empirical knowledge (i.e. by being a logically necessary condition for empirical knowledge), contrary to your claim. You essentially are saying the opposite of what Sellars says in the paper.

Again, that reason that makes it sufficient to throw a light on the foundations of empirical knowledge in Sellars opinion, in your view, is _______? — Terrapin Station

Are you asking me how it is that isolating a necessary condition for the existence of something sheds light on its foundations?

So who is he supposedly paraphrasing "baldly" anyway? — Terrapin Station

The sense datum theorist. Hence why the very paragraphs end with conclusions about sense data, and right after he says that not all sense datum theorists have been subject to certain confusions, but that they have formed an integral part of the tradition.

Sellars outright admits that even if sensation of sense data doesn't constitute knowledge, it might nonetheless shed light on the foundations of empirical knowledge (i.e. by being a logically necessary condition for empirical knowledge), — The Great Whatever

Hahaha--just reword your interpretation as if that's what he said, and then have the balls to lecture someone else on reading comprehension. Nice.

Also, reading comprehension is interpretation.

Also, reading comprehension is interpretation. — Terrapin Station

Interpretation can't happen until reading comprehension is complete. A 'criticism' that misreads a paper on the most basic level is not even a criticism; it hasn't yet understood well enough to be in a position to criticize.

Such is, for example, the 'criticism' that Sellars' 'therefore' in his quoted paragraphs is a non sequitur; it misreads the text in such a fundamental way, in thinking that Sellars is there making an argument, and not speaking in the voice of the sense datum theorist, that it is not a cogent criticism. To think that Sellars is here criticizing the sense datum theorist for not being rigorous enough in excluding the possibility of non-veridicality is not yet to engage with this paper, because that is not what's going on.

I don't know when we're moving on to part III, "The Logic of 'Looks'" . . . one big problem with that section is that Sellars explains the difference between "x looks green to S" and "S sees that x is green" so that "sees" is referring to inferential knowledge and an endorsement of the claim "x is green," in the manner that one would say "S has come to believe that x is green," while "looks" is being used in the phenomenal sense.

However, he then goes on to say:

Thus, when I say "X looks green to me now" I am reporting the fact that my experience is, so to speak, intrinsically, as an experience, indistinguishable from a veridical one of seeing that x is green. — Wilfrid Sellars

But given the difference he just established between "looks" and "sees," it's not at all the case that as an experience, "looks" is indistinguishable from "sees." In fact, in most cases, for most individuals that is, that's surely wrong.

Interpretation can't happen until reading comprehension is complete. — The Great Whatever

Reading comprehension can't happen without interpretation.

Mmm. I think we're done.

You always tease with that promise, but so far haven't followed through.

Also re part III, this is not at all a necessary truth:

"x is red .<--> . x looks red to standard observers in standard conditions"

The left-hand side is about the objective properties of the object in question. The right-hand side is about the subjective experience of the object in question, or that is, the perception, of the object in question. It could easily be the case that x has a property that, given the metaphysics of a possible world, always appears differently to observers than it is, and this could be the case so that the property in question, for some existents, is perceived accurately. In other words, it could merely be a quirk of that particular (type of) object.

Of course, we could say that we're going to use "red" on the left-hand side stipulatively so that whatever it looks like to us under standard conditions is the property that we're going to assume it veridically has, but there's nothing necessary about that.

As our friend John becomes more and more sophisticated about his own and other people's visual experiences, he learns under what conditions it is as though one were seeing a necktie to be of one color when in fact it is of another. Suppose someone asks him "Why does this tie look green to me?" John may very well reply "Because it is blue, and blue objects look green in this kind of light." And if someone asks this question when looking at the necktie in plain daylight, John may very well reply "Because the tie is green" -- to which he may add "We are in plain daylight, and in daylight things look what they are." — Wilfrid Sellars

The same thing would go just as well in both cases.

"Why does this look green to me?" "Because it is blue, and blue objects look green in this kind of light."

"Why does this look blue to me?" "Because it is green, and green objects look blue in this kind of light."

Neither sort of light is ontologically privileged. What would actually be the case with the object is that in light 1, it looks green, and in light 2, it looks blue. Both would be accurate.