The relevant issue is whether perceptions of objects is direct or indirect, not whether awareness of perceptions/sensations is direct or indirect. — Luke

If awareness of sensations is direct and awareness of objects is indirect then perception of objects is indirect.

The cake emits the odour molecules, presumably. — Luke

So? Why is the object of perception not the specific thing that stimulates the sense receptors? Why do you get to go back a step in the causal chain and say that it's the cake?

Earlier you seemed to be saying that smelling a cake and "smelling" odour molecules were equivalent, just like watching pixels/light and watching Joe Biden on television. — Luke

Yes, you can describe what happens as "smelling cake" or you can describe it as "smelling odour molecules", much like you can describe it as "seeing fireworks" or you can describe it as "seeing coloured light".

That's not the same as saying that "perceptions are equal to their causes" (which I can't even make sense of), so your prior question is misguided.

Secondly, if perceptions are not equivalent to their causes, then we can ignore the causes, which are irrelevant to the question of whether or not our perceptions are directly of objects or not. — Luke

Okay, let's ignore causes. What does it mean to say that some sensation is the "direct" perception of some distal object? What conditions must be satisfied for some distal object to be the "direct" object of perception? By your own remarks you cannot defer to some causal explanation.

The relevant issue is whether perceptions are direct or indirect, not whether awareness is direct or indirect. — Luke

If we are directly aware of sensations and not directly aware of distal objects then we do not directly perceive distal objects.

The smell of cake is not a property of the cake either; it’s an interaction between the cake and the perceiver. That doesn’t mean the perception is not of the cake. — Luke

The odour molecules in the air are the more proximal cause. So why is it that the interaction between the odour molecules in the air and the sense receptors in my nose is the (direct) perception of a cake in the oven? What does it even mean to say that the interaction between the odour molecules in the air and the sense receptors in my nose is the (direct) perception of a cake in the oven?

Is the perception of smelling cake equivalent to the cause of the perception? — Luke

No. The cause of the sensation is odour molecules in the air stimulating the sense receptors in my nose. The perception is (the rational awareness of) the subsequent sensation.

What's the connection between either of these things and the cake in the oven?

You provide an accurate account of the mechanics of perception (odour molecules stimulating the sense receptors in my nose, leading to a sensation) but then just throw in the non sequitur "therefore it's the direct perception of a cake in the oven" at the end with no explanation.

And what if, say, the cake has since been taken away and eaten, but the smell lingered. What am I (directly) smelling now? Nothing? The contents of my family's stomachs? Odour molecules in the air?

If the causal chain of odour molecules, olfactory system, etc. is equivalent to the perception of smelling cake, then what’s the intermarry? The causal chain can’t be both the perception and the intermediary. What’s between the perception and the cake? — Luke

Sensations are the intermediary that sit between rational awareness and distal objects. A sweet smell is not a property of some distal object but an olfactory sensation. I am directly aware of a sweet smell and through that smell indirectly aware of some food stuff that contains caramel. Hence the epistemological problem of perception.

Your intentionalism seems to accept the existence of such sensations but nonetheless wants to say that we are directly aware of the distal object, and even though something like odour molecules are the more proximal cause of the sensation. I can't make sense of what you mean by "direct". Grammar notwithstanding, with respect to the ontology of perception it seems like indirect realism to me (and to Robinson).

I read it, and nothing about it conflicts with indirect realism.

I don't believe so. I directly smell the cake. I do not smell an intermediary. — Luke

You're just reasserting the irrelevant argument about grammar.

The indirect realist says that the painting is just paint. The intentionalist says that the painting is of a flower. There's no disagreement.

The causal chain of odour molecules entering the nose, interacting with the olfactory system, converting to brain signals, etc. can explain its effect: our smelling cake. But molecules entering the nose is not equivalent to smelling molecules, and molecules entering the nose, by itself, is insufficient to cause us to smell anything. Therefore, we don't smell odour molecules. The effect of this causal chain (the sensation of smell) cannot be its own cause. Moreover, it doesn't work the other way: the sensation is not an explanation for its distal cause. That is, smelling cake isn't an explanation for why odour molecules enter the nose, etc. So, I don't believe these are equivalent. — Luke

It means that we don't perceive things directly in the naive realist sense of taking physical objects directly into one's mind (somehow). It is just as I am describing: a perception (including representation) is the end result of a causal chain; for example, taking odour molecules into the olfactory system and converting them into brain signals, etc. The output of this causal chain is a perception such as a smell... — Luke

Everything you say here is consistent with indirect realism. Sensations/sense-data/qualia are (usually) caused by stimulation by some distal object. These sensations/sense-data/qualia are (at best) mental representations of that distal object. That distal object and its properties are not directly present in experience and so the epistemological problem of perception remains.

The semantic argument over whether we should describe perception as "seeing a distal object" or "seeing a mental representation" is as irrelevant as arguing over whether we should describe what I do as "watching TV" or "watching Joe Biden's inauguration".

Are the pixels the perception or the cause of the perception? In your previous example you said that the odour molecules were the cause of the smell. Here you appear to imply that the perception and its cause are equivalent. — Luke

I'm simply explaining that the "semantic" approach seems to miss the point. You say you smell a cake. I say I'm watching Joe Biden's inauguration. These are both perfectly ordinary ways to describe what happens. But this ordinary way of describing what happens does not entail direct realist perception.

Even though I describe what I'm doing as "watching Joe Biden's inauguration" (rather than, say, "watching pixels" or "watching light"), my perception of Joe Biden's inauguration isn't direct. Even though you describe what you're doing as "smelling a cake" (rather than, say, "smelling odour molecules"), your perception of the cake isn't direct.

I think I am using language in an ordinary way when I say that you can smell the cake directly. — Luke

The ordinary way of speaking is not an accurate account of the ontology of perception. The ordinary way of speaking developed according to our naive, pre-scientific understanding of the world.

It's odd, then, that Intentionalism was included in the SEP article you were quoting. — Luke

Even the SEP article adds:

Thus, like sense-datum theorists and adverbialists, intentionalists reject Direct Realist Presentation, and admit that we are not ever directly presented with ordinary objects, not even in veridical experience.

It's not clear to me what the intentionalist means by "we directly perceive ordinary objects that are not directly presented to us". It seems hopelessly confused. At best they're equivocating and mean two different things by "direct", at worst they are straight up contradicting themselves.

And I'll refer once again to Semantic Direct Realism:

The most common form of direct realism is Phenomenological Direct Realism (PDR). PDR is the theory that direct realism consists in unmediated awareness of the external object in the form of unmediated awareness of its relevant properties. I contrast this with Semantic Direct Realism (SDR), the theory that perceptual experience puts you in direct cognitive contact with external objects but does so without the unmediated awareness of the objects’ intrinsic properties invoked by PDR. PDR is what most understand by direct realism. My argument is that, under pressure from the arguments from illusion and hallucination, defenders of intentionalist theories, and even of relational theories, in fact retreat to SDR. I also argue briefly that the sense-datum theory is compatible with SDR and so nothing is gained by adopting either of the more fashionable theories.

To put it bluntly:

The perception is: the smell (of cake).
The causes of the perception are: the odour molecules in the air stimulating the sense receptors.

What you perceive/smell is the cake.
What you don’t perceive/smell are the causes of the perception.

The perception is the final product; the smell. All you smell is the cake. You don’t smell the causes of the perception. — Luke

Your account is akin to saying: I'm not watching pixels activate on my television screen, I'm watching Joe Biden's inauguration.

This "semantic" directness is so far divorced from the phenomenological directness that concerns the epistemological problem of perception and the dispute between naive and indirect realists that it seems entirely misplaced in these discussions.

I don’t know how you could smell the cake more directly. — Luke

That it's "as direct as it can be" isn't that it's direct. The point made by indirect realists is that you can't smell the cake directly. Direct perception of a cake would require naive realism to be true, which it never is. This non-naive sense of "directness" is a misnomer.

You seem to be describing enactivism?

I don't understand why you don't understand what I'm asking and I don't know how to explain it in any simpler terms.

One difference could be that qualia exist and sense data don't. — RussellA

Which is like saying "one difference could be that bachelors exist and unmarried men don't".

For example, the relational view of color does a good job explaining how the properties of the object perceived, the ambient enviornment, and the perceiver all go into the generation of an experience. Could an adverbial description do the same thing? Maybe, but not easily. And it's hard to see what the benefit would be. — Count Timothy von Icarus

So how do you make sense of this:

Phenylthiocarbamide tastes bitter to 70% of people and doesn't taste bitter to 30% of people.

What does the word "bitter" mean/refer to? What does the phrase "tastes bitter" mean/refer to?

Only if sense data exist. The Adverbialist doesn't need them. Why do you think sense data exist if they are not needed? — RussellA

I don't understand what you're asking.

I'm saying that the terms "sense data" and "qualia" refer to the same thing. Therefore, if qualia exist then sense data exists. According to the SEP article, adverbialists accept that qualia exist.

Your comments are like saying "I believe that bachelors exist but I don't believe that unmarried men exist".

So if I'm wrong and there is a difference between sense data and qualia then what is that difference?

The Adverbialist may accept qualia but don't need sense data. — RussellA

That's what I don't understand. As I understand it, sense data and qualia are the same thing.

The thesis of Direct Realism (at least, according to the SEP article) is that "we can directly perceive ordinary objects". Some of us believe this thesis but disagree with naive realism. We are also direct realists. I genuinely disagree that we always perceive an intermediary and that we cannot directly perceive ordinary objects. Call that a semantic disagreement if you will, but we can't both be correct. — Luke

In what sense is an olfactory sensation caused by odour molecules in the air stimulating the sense receptors in my nose the "direct" perception of a cake in the oven?

Any non-naive sense of "direct" seems to stretch the meaning of "direct" into meaninglessness, and does nothing to resolve the epistemological problem of perception.

But again, if what naive direct realists mean by "direct" isn't what non-naive direct realists mean by "direct" then it's possible that perception isn't direct in the naive sense but is direct in the non-naive sense. Indirect realists argue that perception isn't direct in the naive sense. Indirect realists argue that perception is indirect in the naive sense.

Indirect realism is compatible with intentionalism, even if intentionalists refer to themselves as being direct realists. Each group simply means something different by "direct". This is the argument made by Robinson in Semantic Direct Realism.

The Adverbialist rejects sense data. Sense data should go the way of the aether, of historic interest only. — RussellA

According to the SEP article adverbialists accept qualia. If sense data and qualia are the same thing then according to the SEP article adverbialists accept sense data.

Maybe there's a distinction between accepting the existence of sense data and accepting the sense datum theory of perception. Perhaps it's a semantic distinction; an argument over whether or not "I see sense data" is correct grammar.

I don't think there is a distinction. But the quote you were quoting also wasn't making that distinction. — flannel jesus

Russell was saying that adverbialism rejects the sense data theory but the SEP article says that adverbialism accepts (and even requires?) qualia.

Hence my confusion.

That's a very interesting piece of information, but I think it's still the case that most people talking about qualia here are talking about the experience, and not the data. — flannel jesus

I think you're misunderstanding what is meant by "sense-data". From here:

The technical term “sense data” was made prominent in philosophy during the early decades of the twentieth century by G. E. Moore and Bertrand Russell, followed by intense elaboration and modification of the concept by C. D. Broad, H. H. Price, and A. J. Ayer, among others. Although the promoters of sense data disagreed in various ways, they mainly agreed on the following points:

1. In perceiving, we are directly and immediately aware of a sense datum.
2. This awareness occurs by a relation of direct mental acquaintance with a datum.
3. Sense data have the properties that they appear to have.
4. These properties are determinate; in vision, we experience determinate shapes, sizes, and colors.
5. Our awareness of such properties of sense data does not involve the affirmation or conception of any object beyond the datum.
6. These properties are known to us with certainty (and perhaps infallibly).
7. Sense data are private; a datum is apprehended by only one person.
8. Sense data are distinct from the act of sensing, or the act by which we are aware of them.

Whereas you seem to be suggesting that "sense data" is something involving light striking the rods and cones in the eyes?

I don't think it's that simple. From here:

Historically, the term ‘qualia’ was first used in connection with the sense-datum theory by C.I. Lewis in 1929. As Lewis used the term, qualia were properties of sense-data themselves.

So much like we might say that mass is a property of physical objects, he says that colour qualia is a property of sense data.

But also from that article:

As noted, adverbialism is committed to the view that experiencing something white, for example, involves your experience being modified in a certain way: experiencing whitely. A natural way to understand this is in terms of the idea that the experience is an event, and the modification of it is a property of that event. Since this property is both intrinsic (as opposed to relational or representational) and phenomenal then this way of understanding adverbialism is committed to the existence of qualia.

I don't get the distinction between sense data and qualia. To me it's all just sensations, which are a mental phenomenon. Distal objects are not constituents of sensations. There is nothing more than a causal relationship (with physical intermediaries) between distal objects and sensations.

However, if all knowledge is necessarily "indirect," and "direct" knowledge is an impossibility because of what knowledge is, then it doesn't seem like the adjective does any lifting at all, regardless of if you think it should be "direct" or "indirect." — Count Timothy von Icarus

The knowledge that I am in pain and am tasting something sweet is direct. The knowledge that I stood on a nail and am eating something that contains a lot of sugar is indirect.

According to the SEP article, direct realism is the thesis that "we can directly perceive ordinary objects." It doesn't say only in the "direct" sense of naive realism. — Luke

So if "direct" in the naive sense doesn't mean the same thing as "direct" in the non-naive sense then there are two different meanings of "direct", and so two different meanings of "we directly perceive ordinary objects". Which meaning of "we directly perceive ordinary objects" do you think the indirect realist is arguing against?

If the indirect realist is arguing against the naive sense of "we directly perceive ordinary objects", and if we do not directly perceive ordinary objects in the naive sense, then indirect realism is correct.

It is this concept of an "unmediated awareness of objects" that I consider to be incoherent. Do indirect realists only hold the negative view that this concept is incoherent? Or do they also hold the positive belief in their position that we cannot directly perceive ordinary objects? — Luke

The naive realist believes that we are directly aware of the constituents of experience and that ordinary objects are the constituents of experience, and so that we are directly aware of ordinary objects. There is no intermediary between awareness and ordinary objects.

The indirect realist believes that we are directly aware of the constituents of experience and that sensations/sense-data/qualia are the constituents of experience, and so that we are directly aware of sensations/sense-data/qualia. Sensations/sense-data/qualia are the intermediary between awareness and ordinary objects.

This is explained in more detail here.

What I find strange about your position is that you seem to accept the existence of something like sensations/sense-data/qualia, seem to accept that we are (directly?) aware of sensations/sense-data/qualia, but also claim that we are directly aware of ordinary objects. I just don't understand what you mean by "direct" in this final claim.

What is the difference between claiming that awareness of sense-data is direct awareness of ordinary objects and claiming that awareness of sense-data is indirect awareness of ordinary objects?

As I suggested here, naming these non-naive direct realisms as being "direct" realisms seems to be a misnomer. At the very least they seem to mean something different by "direct" than what is meant by naive and indirect realists.

On this point it is worth reading Robinson's Semantic Direct Realism:

The most common form of direct realism is Phenomenological Direct Realism (PDR). PDR is the theory that direct realism consists in unmediated awareness of the external object in the form of unmediated awareness of its relevant properties. I contrast this with Semantic Direct Realism (SDR), the theory that perceptual experience puts you in direct cognitive contact with external objects but does so without the unmediated awareness of the objects’ intrinsic properties invoked by PDR. PDR is what most understand by direct realism. My argument is that, under pressure from the arguments from illusion and hallucination, defenders of intentionalist theories, and even of relational theories, in fact retreat to SDR. I also argue briefly that the sense-datum theory is compatible with SDR and so nothing is gained by adopting either of the more fashionable theories.

So, naive realists argue that perception is direct₁, indirect realists argue that perception is not direct₁, and intentionalists argue that perception is direct₂.

The claims that perception is not direct₁ and that perception is direct₂ are consistent, and so indirect realists and intentionalists can both be correct.

And as I've mentioned before, the core of the issue is the epistemological problem of perception, and if perception is not direct₁ then even if it's direct₂ the problem remains.

It follows that your criterion for a direct perception is to have the distal object somehow be physically present in one's phenomenal experience. In other words, your criterion is that the object is identical with one's phenomenal experience. — Luke

It's not my criterion. I'm summarising the various views as explained here:

Direct Realist Presentation: perceptual experiences are direct perceptual presentations of ordinary objects.

...

Direct Realist Character: the phenomenal character of experience is determined, at least partly, by the direct presentation of ordinary objects.

...

On [the naive realist] conception of experience, when one is veridically perceiving the objects of perception are constituents of the experiential episode. The given event could not have occurred without these entities existing and being constituents of it; in turn, one could not have had such a kind of event without there being relevant candidate objects of perception to be apprehended. So, even if those objects are implicated in the causes of the experience, they also figure non-causally as essential constituents of it... Mere presence of a candidate object will not be sufficient for the perceiving of it, that is true, but its absence is sufficient for the non-occurrence of such an event. The connection here is [one] of a constitutive or essential condition of a kind of event.

It is not enough that some distal object causes some sensation (even a "representative" sensation) for perception to be in any meaningful sense direct.

How would that work? How is that kind of perception possible?

It's not, which is why direct realism is false.

As I understand it, indirect realism asserts that we perceive representations (of objects). My position is not that we perceive representations (or some other intermediary), so my position is not indirect realism. My position is that perception involves representations. Representations are not the object of perception, as indirect realism asserts; instead, representations are formative in having perceptions. Or, as you put it earlier, representations are part of the "mechanics of perception". — Luke

I think you're just reading too much into the grammar. I see a mountain when I dream. What sort of thing is the object of perception when I dream? The indirect realist claims that whatever sort of thing is the object of perception when I dream or hallucinate is also the sort of thing that is the object of perception when awake and not hallucinating. The only difference is that when awake and not hallucinating the experience has an appropriate distal cause.

That's my point. Michael was asserting that a direct perception must be when a perception is identical with its object. My reply was that this isn't a perception at all, because it excludes any representation (and, more simply, because objects are not identical with perceptions). You can't have a perception without a representation, yet Michael calls this a direct perception. — Luke

I don't quite understand what you're suggesting I'm saying, but the representational theory of perception is indirect realism.

What is the distinction between direct and indirect awareness? The dispute is not over our (direct or indirect) awareness of our perceptions. This talk of "awareness of perceptions" is just another of your attempts to push our perceptions back a step; to create a gap between ourselves and our perceptions (much like your earlier talk of "experience of perceptions"). We do not perceive our perceptions; we perceive the world. — Luke

I'm aware of colours. Colours are not properties of distal objects. Colours are features of phenomenal experience alone. They are something like sensations/sense-data/qualia. Therefore, I'm aware of something like sensations/sense-data/qualia. The same with smells and tastes and pain and all other so-called "secondary" qualities.

When I dream, I see things. The things I see are not distal objects. The things I see are features of phenomenal experience alone. Even the so-called "primary" qualities in dreams are something like sensations/sense-data/qualia.

The indirect realist argues that the "primary" qualities of veridical experience are of the same kind as the "secondary" qualities of veridical experience and the "primary" qualities of dreams and hallucinations, and that the sorts of things that are the (direct) objects of perception when I dream and hallucinate are the sorts of things that are the (direct) objects of perception when awake and not hallucinating.

This is the common kind claim.

The difference between veridical and non-veridical experience is only that veridical experience has the appropriate distal cause.

One argument against the common kind claim is that distal objects are not just causes of but actual constituents of (veridical) experience (in lieu of something like sensations/sense-data/qualia). That’s the naive realist view.

If there’s such a thing as non-naive direct realism that can avoid the common kind claim without arguing that distal objects are actual constituents of experience then it needs further explanation.

For Locke a primary quality belongs to the object — Leontiskos

If primary qualities belong to an object then nothing in experience is a primary quality, because objects and their properties are not constituents of experience.

and it seems obvious that one can interact with the same spatial property via both sight and touch. Some humans can interact with spatial properties via hearing, but there are other species which tend to be better at that. — Leontiskos

If by this you just mean that there is a causal relationship between an object's properties and our experience then indirect realists would agree.

The relevant question is whether or not (and how) the relationship between experience and an object's properties is "direct".

Do you view shape the same way you view color? — Leontiskos

Yes. Vision is not fundamentally different to any other sense. Are there "primary" taste qualities? Are there "primary" smell qualities? Are there "primary" sound qualities? I don't think so.

With regards specifically to shape, studies have shown that those born blind who are later able to see cannot recognise shapes by sight even though they can recognise them by feel. They have to learn the connection. So, shapes-as-seen are different to shapes-as-felt. When you say that shapes are primary qualities, is that shapes-as-seen or shapes-as-felt?

And, again, related to this is the thought experiment here addressing visual orientation. Neither group can be said to be seeing things "correctly" (such that the other group is seeing things "incorrectly").

Well, I think that there is no “resemblance” between a thing’s appearance and a thing’s (objective) properties. The common example is colour. I reject primitivism. It is true that certain surfaces reflect light of certain wavelengths, and that certain wavelengths are usually responsible for certain colour sensations, but that relationship is nothing more than causal. There’s nothing like “resemblance” or “representation” involved. The same with smells and tastes. A sweet taste sensation does not “resemble” or “represent” any property of sugar.

And, as examined here, the same is also true of so-called “primary qualities” like visual geometry.

The world “behind” appearances is just a mess of quantum fields. I don’t know if this is what counts as “insuperable unreliability”, but in any case I can’t see how anything about this can correctly count as direct realism - and indirect realism in its simplest form is simply a rejection of direct realism.

The argument from illusion is indeed one of the arguments against direct realism, much like the argument from hallucination and the common kind claim that follows. (2) is how some avoid the common kind claim, but this entails naive direct realism.

But you didn’t really answer my question(s). How does non-naive direct realism differ from naive direct realism? Does it reject (2) and so also (1)? If it rejects (1) then how does it complete the premise “something is an object of perception iff …”? And how does its version of this premise maintain the “directness” that was intended to resolve the epistemological problem?

We perceive a tree. A tree is a concept. Therefore we perceive a concept. — RussellA

I think adverbialism provides a better account of what I think you're getting at. It's not that when we see a tree we see a concept but that when we see a tree we are "seeing treely", which is a mental state. The grammatical distinction between the verb "seeing" and the noun "a tree" doesn't accurately represent the facts about perception. Rather, "seeing a tree" is more properly understood as a verb and an adverb.

Both hypericin and @Michael keep adverting to naive forms of direct realism — Leontiskos

What is the difference between naive and non-naive direct realism?

Taking my earlier comment, the naive view is that:

1. Something is an object of perception iff it is a constituent of experience
2. Distal objects are constituents of experience
3. Therefore, distal objects are objects of perception

The indirect realist accepts (1) but rejects (2). Instead their view is that:

1. Something is an object of perception iff it is a constituent of experience
4. Sense data is the constituent of experience
5. Therefore, sense data is the object of perception

Assuming the non-naive direct realist rejects (2) and (5), it must be that they reject (1) and/or (4).

If they reject (4) but accept (1) then something other than sense data and distal objects is the object of perception. This wouldn’t be direct realism but a different kind of indirect realism.

If they reject (1) but accept (4) then, at the very least, they accept the existence of sense data. They must then provide an alternative to (1) to explain what it takes for something that isn’t a constituent of experience to nonetheless be an object of perception.

If they reject both (1) and (4) then, again, they must provide an alternative to (1), but also an alternative to (4) to explain which things are the constituents of experience.

But my own take is that being a constituent of experience is the only meaningful account of “directness”, and so if (2) is false then experience of distal objects is not direct, even if they are the objects of perception. In other words, if (1) is false then “we experience X directly iff X is the object of perception” is false, and so non-naive “direct” realism isn’t direct realism at all.

“Directness” is intended to resolve the epistemological problem of perception such that if perception is direct then there is no problem, but if (2) is false then the common kind claim is true and disjunctivism is false, the epistemological problem of perception remains, and so perception isn’t direct.

Can you cite your source? — Leontiskos

Sorry, it was referenced in an earlier comment. It’s from https://plato.stanford.edu/Entries/perception-problem/ quoting M.G.F. Martin’s defence of naive realist disjunctivism. It’s how to define the difference between a veridical experience and a subjectively indistinguishable hallucination.

And I think this is where demands to define "indirect" in terms of physical interactions becomes relevant. Perhaps there is some way to demarcate direct and indirect physical processes, although I am skeptical of this. — Count Timothy von Icarus

Using the examples from the SEP article, we can say that the experience of a distal object is direct iff the distal object is a constituent of the experience.

If we then say that indirect realism is the rejection of direct realism, we can say that the experience of a distal object is indirect iff the distal object is not a constituent of the experience.

Does the science of perception agree with or disagree with the claim that distal objects are constituents of experience? I think it disagrees with it. It certainly shows a causal connection, but nothing more substantive.

Direct realism would seem to require a rejection of scientific realism, perhaps in favour of scientific instrumentalism, allowing for something like colour primitivism and for experience to extend beyond the body, both of which are probably what was believed by the direct realists of old (and which is my uncritical, intuitive view of the world in everyday life).

Sokolowski's real focus in thought and language though, not perception. He goes on to elucidate how names, words, and syntax, are used intersubjectively to present the intelligibility of objects. Roughly speaking, the intelligibility of an object is exactly what we can truthfully say about it, what can be unfolded through the entire history of "the human conversation." — Count Timothy von Icarus

I think that's the sort of approach that many here are taking when they claim to be direct realists, even though whatever they're saying has nothing to do with the actual mechanics of perception, the relationship between perceptual experience and distal objects, or the epistemological implications thereof.

Howard Robinson calls this the retreat from phenomenological direct realism to semantic direct realism, and argues that semantic direct realism is consistent with indirect realisms like the sense-datum theory.

The patterns thus have a representational character in the sense that they disperse a representative form of the thing into the surrounding media. — Count Timothy von Icarus

I think the notion that they are representational is questionable. Phenylthiocarbamide is a chemical that some taste as bitter and some don't. For the sake of argument, let's assume that some taste it as sour. Which of "sourness" and "bitterness" is a representation of phenylthiocarbamide? Does it make sense to suggest that either is a representation? I think it makes much more sense to simply say that each is just an effect given the particulars of the eater's bodies.

Or perhaps "representation" is something that only works in the case of visual geometry? I think my thought experiment here brings even that into question. I don't think there's reason to treat sight as fundamentally different to any other mode of experience.

So, even if [1] those objects are implicated in the causes of the experience, [2] they also figure non-causally as essential constituents of it — SEP

This is the important part.

Indirect realists agree with [1] but disagree with [2], and if [2] is false then the epistemological problem of perception remains.

The idea is that there is some alternative vantage point which is more fundamental than phenomenal experience, and which makes inferences based on the phenomenal experience. — Leontiskos

There is. There's rational interpretation. There's the "thinking" self. See the duck-rabbit above. Sometimes I see a rabbit, sometimes I see a duck, even though nothing about the phenomenal experience has changed.

Much like a homunculus isn't required for self-reflection, a homunculus isn't required for indirect perception.

Let's take the SEP article.

Direct Realist Presentation: perceptual experiences are direct perceptual presentations of ordinary objects.

...

Direct Realist Character: the phenomenal character of experience is determined, at least partly, by the direct presentation of ordinary objects.

...

On [the naive realist] conception of experience, when one is veridically perceiving the objects of perception are constituents of the experiential episode. The given event could not have occurred without these entities existing and being constituents of it; in turn, one could not have had such a kind of event without there being relevant candidate objects of perception to be apprehended. So, even if those objects are implicated in the causes of the experience, they also figure non-causally as essential constituents of it... Mere presence of a candidate object will not be sufficient for the perceiving of it, that is true, but its absence is sufficient for the non-occurrence of such an event. The connection here is [one] of a constitutive or essential condition of a kind of event.

Perhaps you could explain how to properly interpret the parts in bold.

Under any ordinary reading, the flower is not "directly presented in" or "a constituent of" the photo. The photo is just a photosensitive surface that has chemically reacted to light.

And by the same token, the flower is not "directly presented in" or "a constituent of" phenomenal experience. Phenomenal experience is just a mental phenomenon elicited in response to signals sent by the body's sense receptors.

So given the above account of direct/naive realism, direct/naive realism is false.

If I take a photograph of a flower, then the flower is in the photograph. — Luke

No it's not. The flower is on the ground. The photograph is in my pocket. The photograph is just a photosensitive material that has chemically reacted to light.

Distal objects are present in phenomenal experience in the same sense.

Which is not a direct sense. It's an indirect sense. The photograph is a representation of the flower and phenomenal experience is a representation (at least, perhaps, with respect to primary qualities) of distal objects.

By that standard, no perception can be direct. — Luke

Phenomenal experience is directly present in conscious awareness.

You really are just describing indirect realism but refusing to call it that.

In what sense are they not? — Luke

Phenomenal experience doesn't extend beyond the body. Distal objects exist beyond the body. Therefore, distal objects are not present in phenomenal experience.

Distal objects are a cause of phenomenal experience, but that's it.

This is even more apparent in the case of the stars we see in the night sky. Some of them have long since gone. A thing that doesn't exist cannot be present.

Otherwise, it just boils down to an ambiguity in the meaning of "perceive", with one camp taking it to refer to perceiving real objects and the other camp taking it to refer to the way those objects are perceived and the contents of our phenomenal experience. — Luke

Yes, that's something that I have argued many times before, and is why I keep saying that arguing over the grammar of "I see X" misses the point entirely.

Regarding the dress, for example, there is a sense in which we all see the same thing and there is a sense in which different people see different things. When considering the sense in which different people see different things, the thing they see, by necessity, isn't the distal object (which is the same for everyone).

The relevant issue is the epistemological problem of perception; the relationship between phenomenal experience and distal objects. Distal objects are not present in phenomenal experience and the features of phenomenal experience are not the properties of distal objects. That is indirect realism to me, as contrasted with the direct realist view that distal objects are present in phenomenal experience and that the features of phenomenal experience are the properties of those distal objects.