No idea how or why, but is the above correct? Shouldn't it be something like:

X looks red to S = There is a class of red sense data that belongs to S

or better

X looks red to S = There is sense-data that belongs to S. S has a theory that the sense-data belongs to the class of red sense data. S attaches the label "red" to X.

See section 9 in Empiricism and the Philosophy of Mind. Sellars says '(X looks red to S) = There is a class of red sense-data which belong to X, and are sensed by S'.

I think the claim is that if something looks red in the veridical case then it is red (i.e. the object itself has the property of being red). Therefore, if redness is sense-data then this sense-data is a property of the object itself.

Although my understanding is just based on a brief skim of the paper.

See section 9 in Empiricism and the Philosophy of Mind. Sellars says '(X looks red to S) = There is a class of red sense-data which belong to X, and are sensed by S'. — quine

My argument is not with you. I'm sure you are quoting correctly. It just seems obviously wrong.

I think the claim is that if something looks red in the veridical case then it is red (i.e. the object itself has the property of being red). Therefore, if redness is sense-data then this sense-data is a property of the object itself. — Michael

What colour is the dress?

I see white and gold, but I'm told that it's actually blue and black. As such, one might say that my experience isn't veridical.

I'm not really sure the relevance of the question, though. I'm simply providing the reasoning. So if you have an issue, then it must be with one of the premises. Either redness isn't sense-data or looking red in the veridical case and being red are different things (in which case there's likely conflation on the term "red").

Just out of curiosity, could I ask why you wish to do this?

I'm only loosely familiar with first-order logic so I could be mistaken here, but my first guess would be:

If we approach this from the perspective that:

- There are "sense data" belonging to X, which we will call set A.
- There are "sense data" that S is capable of perceiving, which we call set B.

Then we could say:

∃x ∈ A ∈ B

Which translates roughly to, "There exists some property x which belongs to A (i.e. is a sense-datum of X) and belongs to B (i.e. can be sensed by S)."

It may be appropriate to replace B with a formula, but this may just be a matter of preference.

Again, I would urge you to check this with someone else before basing any life decisions on it.

As for the dress: The colour information stored in the image file is that of white and gold. Some people claim that their brains choose to reinterpret this as blue and black - and the originator claims that the dress was in fact blue and black.

The sense data is yours and not the dress's.

The sense data is yours and not the dress's. — tom

Except we say that the dress really is blue and black even though I see it as white and gold. So is the blueness and the blackness that the dress really has sense-data? Or would you say that it doesn't make sense to talk about the "real" colour of the dress; only the colour we see it to be (in which case it doesn't make sense to distinguish between veridical and non-veridical colour experiences?)

The implication is that either a) the essence of the colour (the qualia, if you will) is stored in X (which I would personally disagree with) or b) some property of X is perceived as red by S. It feels like the first option is the one intended by the author.

As for the dress: The colour information stored in the image file is that of white and gold. Some people claim that their brains choose to reinterpret this as blue and black. — Efram

That's interesting. The dress is actually blue and black, the colour in the image is actually gold and white (well, a bluish silver really), and some people see it as blue and black. The wrong colours in the image allow some people to recognise the real colours of the dress.

As for the dress: The colour information stored in the image file is that of white and gold. Some people claim that their brains choose to reinterpret this as blue and black. — Efram

No it's not. The image file is black and blue.

Center of the neck


Center of the left lapel

The image file is black and blue — tom

So the black and blue sense-data is in the image file? Or is the black and blue in the image file not sense-data?

So the black and blue sense-data is in the image file? — Michael

OK, so the file is actually more brown and blue, which weirdly I am beginning to see. The image file contains binary data, so no, there is no sense data stored there, just numbers.

The image file contains binary data, so no, there is no sense data stored there, just numbers. — tom

And what about in the dress (which is said to "really" be blue and black)?

Hey,
This thread is about correct formulation of sense-datum statements... Don't forget about it.

I didn't! See my first reply. ;)

Except we say that the dress really is blue and black even though I see it as white and gold . . . — Michael

What I thought Tom was getting at is that it doesn't have anything to do with sense/sensing/etc. until it gets to you.

What I thought Tom was getting at is that it doesn't have anything to do with sense/sensing/etc. until it gets to you. — Terrapin Station

Then I guess the "red" in "the dress is red" is different to the "red" in "the dress looks red", as the "red" in "the dress looks red" is sense-data but the "red" in "the dress is red" isn't? In which case the premise that looking red (even in the veridical case) and being red are the same thing is false?

Right. I don't believe there's anyone who says that the red of a dress, say, is literally, identically in anyone's head. Thinking that anyone would be saying that seems like a weird, Aspie-ish misunderstanding of language.

I haven't read enough of the paper to know what Sellars' thoughts are on the matter. But I doubt he picked the definition out of thin air. Presumably there are people (naive realists?) who in some sense think that the red of the dress is "like" red sense-data.

I'm a naive realist, by the way. What I always point out is that naive realists are not saying that perceptions are not perceptions. And we're not saying that external stuff is perceptions.

I'm not tied to the notion that naive realists believe this. I'm just suggesting that there are likely people who do. It seems like a perfectly understandable thing to believe prior to any in-depth scientific analysis.

Set theory is a good treatment.

If you have or know of an alternative, I'd be interested to (eventually) hear it. At first I wondered if this was a homework assignment, but now I suspect you may already have some thoughts and just wanted to see others?

This wasn't a homework assignment. I have tried something as follows:
'The king of France looks red to Russell'
= (∃x)(Fx & (∀y)(Fy → x = y) & (∃z)(Gzx & Hza))
In set theory, I have tried:
(∃x)(x ∈ A & x ∈ B)
That's it.