Seeing is sensation — Terrapin

Contextually, I think it's clear that Sellars is discussing seeing qua perception (and perception is not the same thing as sensation.)

And also if we're talking about veridicality versus non-veridicality, we're talking about how sensations link up with something that's not the sensation, which is the same, functionally, at least, as objects external to you. — Terrapin

Yep.

perception is not the same thing as sensation — csalisbury

What's the difference when we're talking about veridicality versus non-veridicality?

Re your "yep," hence we're talking about sensation.

What's the difference when we're talking about veridicality versus non-veridicality? — Terrapin Station

The difference is that sensations are not the kinds of things that can be veridical or non-veridical. Unlike beliefs (of which perceptions are a species), they have no normative import.

One of the main points of EPM is to show that sense-datum theorists fail to adequately distinguish between the causal and normative senses of "immediacy". They either simply assume that the one implies the other, or equivocate the two senses entirely. According to Sellars, once the distinction is properly understood it becomes clear that causal immediacy can't underwrite the kind of epistemic immediacy that sense-datum theorist would like it to - it simply isn't up to the task - and any brand of foundationalist epistemology that depends on that presumption is invalidated as a consequence.

That's my take, anyway.

Okay, but again, he says both

(1) "These philosophers believe that it doesn't make sense to speak of unveridical sensations" (paraphrased, according to the interpretation folks stressed above)

and

(2) "The seeing that the facing surface of a physical object is red and triangular is a veridical member of a class of experiences -- let us call them 'ostensible seeings' -- some of the members of which are non-veridical."

With respect to (1), you're then saying that Sellars is agreeing with them that it doesn't make sense to speak of unveridical sensations?

And then with (2), you're saying that "seeing" isn't being used to refer to the sensation of seeing? (Re the above comment from csalisbury, I'd agree that perception isn't identical to sensation, but because I'd say that perception is a subset of sensation--so perceptions are still sensations, but sensations are not necessarily perceptions.) Perhaps you're saying that he's using "seeing" in the quirky apparent distinction I noted above?--however, how would that make sense with how he's using the word "hearing" in the same passage?

Happy to stumble into this reading group today. It sounds like the group kicked off with something like a plan to close-read two sections per week? What are we up to?

I like this online version of the essay. It has easily copyable text.

For Sellars, raw sensory content is epistemologically inert – it has no propositional content and it has no inferential consequences. Sensory content consists of particular “things” – colors, shapes, etc. By contrast, perceptions add propositional content – they make claims about sensory contents.

So when Sellars talks about “seeing that X”, he is talking about perception insofar as X is a claim or proposition (e.g. "the table is green", etc.). The modality (hearing, tasting, etc.) is irrelevant.

I'm struck by the resemblance between Sellars's first sentence, and Hume's first sentence in the the second section of the Enquiry.

EVERY one will readily allow, that there is a considerable difference between the perceptions of the mind, when a man feels the pain of excessive heat, or the pleasure of moderate warmth, and when he afterwards recalls to his memory this sensation, or anticipates it by his imagination. — Hume

I PRESUME that no philosopher who has attacked the philosophical idea of givenness or, to use the Hegelian term, immediacy has intended to deny that there is a difference between inferring that something is the case and, for example, seeing it to be the case. — Sellars

Each begins by locating a bit of common ground in the conversation, by characterizing a particular claim as indicative of a consensus view, and taking the allegedly noncontroversial claim for granted as a starting point in discussion. Hume beings with a point he supposes "everyone will allow", while Sellars begins with a point he presumes no critic of the concept of givenness "has intended to deny."

Each employs a distinction between perception and another sort of activity. Hume draws a distinction between perception on the one hand, and memory and imagination on the other. Sellars begins with a distinction between perception and inference -- or at least between "seeing that p" (presumably a species of "perceiving that p") and "inferring that p".

Take the section-heads into account: Hume stakes out common ground in speaking about perception, memory, and imagination, to orient his discussion "of the origin of ideas". Sellars stakes out common ground in speaking about perception and inference to orient his discussion of "an ambiguity in sense-datum theories".

Perhaps it's reading too much into the texts, but I'm amused that Hume mentions perception first, while Sellars mentions inference first -- as if, I'd want to say, in each case the first-mentioned term is the one held to be less mysterious by the contemporaries addressed by the author, those interlocutors invited to begin speaking together on this common ground.

Another reach on audience: Hume seems to position himself, at least rhetorically and in principle, to be speaking with "everyone". Sellars seems to think he's speaking primarily to philosophers, or professional philosophers, or professional epistemologists concerned with the concept of givenness... or most narrowly, to professional epistemologists who are, like Sellars and Hegel, critical of the concept of givenness.

Would anyone like to give us a bit of background on the history of the "sense-datum theories", and theorists, that Sellars takes aim at here?

Perhaps it's reading too much into the texts, but I'm amused that Hume mentions perception first, while Sellars mentions inference first -- as if, I'd want to say, in each case the first-mentioned term is the one held to be less mysterious by the contemporaries addressed by the author, those interlocutors invited to begin speaking together on this common ground. — Cabbage Farmer

Gold! It's true, Hume was using skepticism to push the British Empiricist project as far as it could go. Sellars seems to be dialing the clock back on that, in some sense. There does seem to be an implicit dialectic here.

With respect to (1), you're then saying that Sellars is agreeing with them that it doesn't make sense to speak of unveridical sensations?

Also, I'm curious if you'd say that Sellars is claiming that propositions are perceived in some sense. Or would he agree that "seeing <a proposition>" is a very loose, metaphorical way of speaking--that is, we're not literally seeing a proposition?

I PRESUME that no philosopher who has attacked the philosophical idea of givenness or, to use the Hegelian term, immediacy has intended to deny that there is a difference between inferring that something is the case and, for example, seeing it to be the case. — Sellars

Do any of us here deny that there is a difference between inferring and seeing that something is the case? Or more generally, between inferential knowledge and noninferential knowledge, or likewise between judgments based on inference and judgments based on perception?

Or short of denying the difference, do we find anything objectionable or confusing about the distinction drawn that way?

If the term "given" referred merely to what is observed as being observed, or, perhaps, to a proper subset of the things we are said to determine by observation,... — Sellars

"what is observed qua being observed": What do we suppose this phrase means? Perhaps a weak sense of "observation", as in, S observes that p, though it's possible that not-p? (In other words, "S observes that p" is consistent with not-p; in this respect "observes" would resemble "believes" rather than "knows"...)

similarly, "a proper subset...": as in, the subset of "true observations"?

...the existence of "data" would be as noncontroversial as the existence of philosophical perplexities. But, of course, this just is not so. — Sellars

I presume this is the "data" of the "sense-data theorists".

Sellars coordinates the terms "given" and "sense data", and associates them with the term "observation". If philosophers used the term "given" merely to refer to the observed qua observed, or perhaps to a proper subset of what may be determined by observation, then the existence of "sense data" would be as noncontroversial as the existence of philosophical perplexities.

Dense rhetoric! It's the very existence of sense-data that Sellars puts into question here. Do sense-data exist, or are they mere creatures of fiction, misinterpreted marks in a logician's notebook? In what sense are philosophical perplexities said -- by epistemologists -- to "exist"? Which perplexities? Shop-talk about nonexistent objects? Shop-talk about justified true belief? About riddles of "believe"-talk and "know"-talk in the first person? Perhaps all that "exists" in such cases is talk and perplexity, sound and fury -- perhaps shop-talk about "sense-data given in observation" is another such case....

Sellars suggests there is an ordinary-language, common sense, use of this family of terms -- "given", "(sense-)data", and "observe" -- that is philosophically unobjectionable, but also philosophically unproductive: The unobjectionable uses only give us ways of stating (or restating) traditional epistemological problems, but don't entail solutions to those problems.

The phrase "the given" as a piece of professional -- epistemological -- shoptalk carries a substantial theoretical commitment, and one can deny that there are "data" or that anything is, in this sense, "given" without flying in the face of reason. — Sellars

Sellars alludes to arguments made by defenders of the epistemologist's "given", who argued that to deny there is a "given" in the relevant sense is to "fly in the face of reason". Not so, says Sellars: The defender's argument relies on a conflation of the ordinary, common sense, unobjectionable and uninformative sense of "givenness", and the highly contentious sense of "givenness" that figures prominently in epistemological shop-talk.

I wonder what prominent defender had suggested that some critic's denials would entail that there is no difference between "inferring" and "seeing".

Of course there is a difference between "inferring" and "seeing", replies Sellars; none of us has intended to deny such a claim. Nor would we deny that there is an ordinary, noncontroversial sense in which "data" is "given" in "seeing" and in "observation". That much we have in common, Sellars insists, planting himself firmly on the common ground of human experience and ordinary language, and taking aim at a brand of epistemological shop-talk associated with a particular way of speaking about the given.

With respect to (1), you're then saying that Sellars is agreeing with them that it doesn't make sense to speak of unveridical sensations? — Terrapin

Yes, Sellars will ultimately agree that claim, though in the context of the quote that you’ve been discussing he’s not mentioning it in order to agree or disagree with it.

Also, I'm curious if you'd say that Sellars is claiming that propositions are perceived in some sense. Or would he agree that "seeing <a proposition>" is a very loose, metaphorical way of speaking--that is, we're not literally seeing a proposition? — Terrapin

He’s saying that perception takes the form of a proposition.

In the first paragraph Sellars positions himself in the conversation of contemporary epistemologists, and seems perhaps to align himself with Hegel. He stakes out a common ground, characterizes a contentious theory he intends to critique, and rebuts one sort of objection to his proposed line of criticism. In the second paragraph, he provides more context while giving us an idea just how much he intends to bite off and chew:

Many things have been said to be "given": sense contents, material objects, universals, propositions, real connections, first principles, even givenness itself. And there is, indeed, a certain way of construing the situations which philosophers analyze in these terms which can be said to be the framework of givenness. — Sellars

What is it, in this sentence, that "can be said to be the framework of givenness": "a certain way of construing", or "these terms"?

Suppose he means the former: The framework of givenness is or consists in a certain way of construing situations that philosophers have analyzed in terms such as "sense contents, material objects, universals, propositions, real connections, first principles, even givenness itself."

This framework has been a common feature of most of the major systems of philosophy, including, to use a Kantian turn of phrase, both "dogmatic rationalism" and "skeptical empiricism". — Sellars

Would anyone please provide a characterization of "dogmatic rationalism" and "skeptical empiricism" as these terms are intended here?

Why are these two "philosophical systems" given special mention here?

We might even ask: What is a "philosophical system" in the sense intended here?

It has, indeed, been so pervasive that few, if any, philosophers have been altogether free of it; certainly not Kant, and, I would argue, not even Hegel, that great foe of "immediacy". — Sellars

The framework, the custom of construing situations and analyzing them in terms of certain bits of traditional epistemological shop-talk, has influenced every epistemologist in the tradition Sellars is concerned with here, or perhaps every epistemologist simpliciter.

Again there's mention of Hegel's special historical role as a critic of the epistemological concept of immediacy or givenness. Again we anticipate that Sellars is aligning himself here with Hegel, though in this case it seems he's set to tackle a job he believes Hegel's left unfinished. It also seems that, at least with respect to the question of givenness, Sellars considers himself to be more closely aligned with Hegel than with Kant.

Often what is attacked under its name are only specific varieties of "given." Intuited first principles and synthetic necessary connections were the first to come under attack. — Sellars

These first attacks on intuited first principles and on synthetic necessary connections: When and where were they located? Who were the attackers, and who the defenders?

He's talking about pre-Kantian modern-Western philosophy? Was it Hume who first attacked synthetic necessary connections? Was it an empiricist or rationalist who first attacked intuited first principles? Or are we going back behind the West and before Christianity, to ancient philosophical traditions?

And many who today attack "the whole idea of givenness" -- and they are an increasing number -- are really only attacking sense data. For they transfer to other items, say physical objects or relations of appearing, the characteristic features of the "given." If, however, I begin my argument with an attack on sense-datum theories, it is only as a first step in a general critique of the entire framework of givenness. — Sellars

Sellars criticizes his contemporary allies: Many of them miss the deeper point, and "transfer... the characteristic features of the 'given'" to their way of construing situations they analyze in terms of physical objects, relations of appearing, or other bits of shop-talk.

What are the "characteristic features" of "the given"? How are they related to characteristic "ways of construing" situations analyzable in terms of the items of shop-talk historically associated with the "framework of givenness"? Something I'd look for as the essay proceeds.

At the end of section I.1, Sellars gives an idea how the essay will proceed: He'll begin his argument with an attack on sense-datum theories, and move on to a general critique of the whole framework of givenness.

He’s saying that perception takes the form of a proposition. — Aaron R

Does he ever say why he believes that or give an argument for it?

2. Sense-datum theories characteristically distinguish between an act of awareness and, for example, the color patch which is its object. — Sellars

This must be the "act-object conception" that Shoemaker loves to wrestle with. I always got the feeling he never quite shakes himself free of it.

An "act of awareness" and an "object" of the act. What object? Not an apple or a chair or a man, but a "color patch". Perhaps something like the smudgy strokes in an impressionist painting?

I take it Shoemaker's line of criticism is still in fashion nowadays: The analyst wants to analyze perception so that the perceptual object is the thing we ordinarily say we "perceive" ("see", e.g.): The apple, the chair, the man -- not some color patches.

The act is usually called sensing. — Sellars

Is there a difference, for the sense-data theorist, or for his critic, between sensing and perceiving?

Classical exponents of the theory have often characterized these acts as "phenomenologically simple" and "not further analyzable." But other sense-datum theorists -- some of them with an equal claim to be considered "classical exponents" -- have held that sensing is analyzable. — Sellars

Who are the more or less classical exponents mentioned, on either side of this divide?

On one side, they say particular acts of sensing are "phenomenologically simple". On the other side, they say particular acts of sensing are "further analyzable" -- phenomenologically, or in some other way?

Is it the act, or the object, the analyzability of which is here contested? Is the object counted as "part" of the act; i.e., we identify an individual act of sensing-awareness, analyze it into two parts, act and object, and then either find that we can go no further, or find that we can continue analyzing (act, or object, or both) into component parts?

And if some philosophers seem to have thought that if sensing is analyzable, then it cannot be an act, this has by no means been the general opinion. — Sellars

Who seems to have thought this, and why wasn't it a popular view?

Why should anyone think that an act -- an act of awareness, an act of sensing -- must be unanalyzable, phenomenologically simple, etc. Do these thinkers object even to the analysis into act and object?

There are, indeed, deeper roots for the doubt that sensing (if there is such a thing) is an act, roots which can be traced to one of two lines of thought tangled together in classical sense-datum theory. — Sellars

"if there is such a thing": i.e., if traditional shop-talk in terms of "sensing" holds up under scrutiny.

What deep tangled roots?

Who doubts that sensing is an act?

Well, sensing in what sense, and act in what sense? I might doubt that "sensing unanalyzable color patches" or "sensing phenomenologically simple sense-data" is ordinarily (or ever) an act of awareness. For instance, it may be that whatever the simplest elements of sensation happen to be, they are processed at a preconscious, prephenomenal, stage of cognition, and that from a phenomenological point of view, it makes more sense to say that we see things and places in the world around us, than that we see color patches.

For the moment, however, I shall simply assume that however complex (or simple) the fact that x is sensed may be, it has the form, whatever exactly it may be, by virtue of which for x to be sensed is for it to be the object of an act. — Sellars

According to sense-data models, the form of "the fact that x is sensed", whatever exactly that form may be, entails at least that if x is sensed, then x is the object of an act [of sensing].

What is the purport of this clunky way of speaking? I suspect there is some scribble in a logician's notebook at issue here behind the scenes, when philosophers like Sellars begin to speak of the "form" of a "fact", and begin to lay down implications this way. I suppose the scribble might include marks that stand for acts of sensing, marks that stand for objects of those acts, marks for variables like x, and some way of packaging such marks to express "facts" about states of affairs in which objects are sensed.

According to Sellars the model-builders agree along these lines: there are facts with the form "x is sensed", and if x is sensed, then x is the object of an act [of sensing].

According to Sellars the model-builders disagree about whether a fact of the form "x is sensed" is simple or complex.

One bit of complexity seems missing from the account so far: Whose act? What senses x? What mark in the model for this?

Being a sense datum, or sensum, is a relational property of the item that is sensed. — Sellars

I suppose "being a sense datum" is the same as "being the object of an act of sensing". It's a relational property, according to which an object is associated with an act of sensing in a specific way and in a particular "fact".

To refer to an item which is sensed in a way which does not entail that it is sensed, it is necessary to use some other locution. — Sellars

Much as, to refer to an item which was thrown by Jim at noon today, in a way that does not entail that the item was thrown by Jim at noon today, it is necessary to use some other locution.

Sensibile has the disadvantage that it implies that sensed items could exist without being sensed, and this is a matter of controversy among sense-datum theorists. — Sellars

Of course it's controversial, if the sensum is something like a "color patch".

Sense content is, perhaps, as neutral a term as any. — Sellars

Oh my, are we going to start speaking about "content"? Was the term in use already in this connection by the time Sellars composed this essay?

What does Sellars introduce the term "sense content" for here? As a substitute for "sense datum" or "object of sense-act", a substitute that doesn't imply that the object could exist without being sensed; and that doesn't entail that the object is (has been, will be) in fact sensed

What kind of possibly nonexistent and possibly not-sensed object is a particular "sense content" supposed to be?

I suppose it's something like an "object of possible experience" -- or how do the Kantians and phenomenologists put it? We use a term like "sense content" to talk about the sense data that would be sensed -- in hypothetical cases, for the purpose of analysis, for the sake of this strange conversation -- if things were in fact the way we have characterized them in our notebooks.

There appear to be varieties of sensing, referred to by some as visual sensing, tactual sensing, etc., and by others as directly seeing, directly hearing, etc. — Sellars

Who are those who speak in each way indicated?

But it is not clear whether these are species of sensing in any full-blooded sense, or whether "x is visually sensed" amounts to no more than "x is a color patch which is sensed," "x is directly heard" than "x is a sound which is sensed" and so on. — Sellars

Good point. Notice how the point is developed in terms of the act-object conception: do the varieties of sensing sort out into different sorts of act, or into different sorts of object for the same sort of act.

I presume this is an important difference in the logician's notebook. It might also lead us to inquire into the bases of the distinctions between sorts of sensing-acts (seeing, hearing, etc), or the bases of the distinctions between sorts of sense-contents (color patches, sounds).

Either way it begins to seem there may be room for analysis beyond the form of a fact "x is sensed", if every such fact entails that x is some sort of sense content, or that the sensing of x is some sort of sensing, or both. Or is this not the sort of "analysis" at issue?

In the latter case, being a visual sensing or a direct hearing would be a relational property of an act of sensing,... — Sellars

In the latter case -- the case in which the varieties of sensing are not "species of sensing in any full-blooded sense", but only various sorts of sensing distinguished in terms of their different sorts of object -- being a visual or auditory sensing would be merely a relational property of an act of sensing. In that case, a place for this relation is built into the form of a fact that x is sensed, but is (I presume) determined in each case by some property of the object of the act: it is a color patch, or a sound patch, one or another sort of sense-content.

I suppose to approach the matter with less bias, we might leave it an open question whether this relation is determined by some feature of the sense-content, or instead by some mediating factor, some other term in the analysis.

...just as being a sense datum is a relational property of a sense content. — Sellars

"just as": i.e., the first is a relational property, no less than the second. Is the analogy much stronger than this?

Being a sense datum is a relational property of a sense content. In other words, it is a property of a sense content when the sense content is related to an act of sensing in the right way. But this seems strange: What is a sense content, when it is not a sense datum? Perhaps merely a figure of speech....

Earlier I read Sellars as peeling off the "existence" and the facticity of "being sensed" that belong to the sense datum (in the model), in order to produce a highly abstract theoretical construct of "sense content", which I tentatively glossed as something like a hypothetical "object of possible experience" characterized somewhere in a logician's notebook for the purpose of analysis and conversation. Is this line of interpretation still open?

Are there "color patches" floating around somewhere, sometimes taken up in acts of sensing, other times slipping by unsensed? Are these sense contents, sensed and unsensed, inside heads, or outside heads, or mere possible-sensums marked down in philosophical narratives assembled according to a certain cumbersome framework?

Being a sense datum, or sensum, is a relational property of the item that is sensed. — Sellars

Like a living organism may have a disposition to sense items in a certain way, also items may have a disposition to be sensed in a certain way. Perhaps this is what Sellars means by relational property?

Some reflections on sections 1-2.

Is it the act, or the object, the analyzability of which is here contested? Is the object counted as "part" of the act; i.e., we identify an individual act of sensing-awareness, analyze it into two parts, act and object, and then either find that we can go no further, or find that we can continue analyzing (act, or object, or both) into component parts? — Cabbage Farmer

By the end of section I.2, it seems the object is not counted as "part” of the act, at least in the logician's notebook. One way to draw up the account: Two terms, act and object, figure in a predicate relation of "sensing" that can be used to express "facts" involving acts and objects in that relation. So it’s both acts-of-objects and objects-of-acts that figure as logical objects of sensing-relations used to express facts. Some marvel along these lines:

There is some a and some o such that: SENSE (a, o).

I take it this would have implications for what logicians and model-builders call the "ontology" of the model. There are two sorts of entity implicated here, two sorts of entity-in-the-model: i) acts of sensing-relations and ii) objects of sensing-relations. You might call it a "hard" sort (or what's a more convenient phrase): We might expect that what counts as an object in one sensing relation cannot figure as an act in another sensing relation; and that what counts as an act in one sensing relation cannot figure as an object in another sensing relation; and that act and object in the same sensing-relation cannot be identical to each other. (Though these rules may be contested, depending on what kind of “sensing” it turns out we’re discussing here.)

Along these lines, it would seem I strayed from the mark earlier by using the language of entailment to paraphrase Sellars remark that "the fact that x is sensed... has the form... by virtue of which for x to be sensed is for it to be the object of an act." For once we start packaging acts and objects in the sort of formal predicate language I've indicated today, it seems there just is no way to say "x is sensed", or to express a fact that x is sensed, without calling up something like the relation SENSE (a, x). This is an example of a "form by virtue of which for x to be sensed is for it to be the object of an act" -- which seems now a far more binding condition than the implication I used to paraphrase in my reading notes: "if x is sensed, then x is the object of an act [of sensing]." Given the ordinary conventions of predicate modeling, and the terms I've drawn up today, I'm not sure how one could express that conditional statement without the most blatantly redundant tautology.

According to Sellars the model-builders agree along these lines: there are facts with the form "x is sensed", and if x is sensed, then x is the object of an act [of sensing]. — Cabbage Farmer

In light of today’s shift in interpretation, I might want instead to say:

According to Sellars the model-builders agree along these lines: There are facts with the form “x is sensed”, and by virtue of this form, x is the object of an act.

According to Sellars the model-builders disagree about whether a fact of the form "x is sensed" is simple or complex. — Cabbage Farmer

Could this mean that for some of the model-builders, the fact that x is sensed is not analyzable as a fact involving both an act and an object? Sellars seems to rule this out. For instance:

[...]however complex (or simple) the fact that x is sensed may be, it has the form, whatever exactly it may be, by virtue of which for x to be sensed is for it to be the object of an act. — Sellars

So for clarity’s sake I might want to say:

According to Sellars the model-builders disagree about whether a fact of the form "x is sensed" is simple or complex; but no matter how simple or complex a fact of this form is said to be, the model-builders agree that the form is such that, for x to be sensed is for x to be the object of an act.

I wonder if Sellars’s way of carving up the fact of sensing is supposed to shed light on “characteristic features of the given” and on “the framework of givenness”. Or does this carving pertain only or primarily to the special case of sense-data theory, while the implications for the more general framework are not yet in view in the narrative?

Continued reflections on sections 1-2.

The most arcane thing in the essay so far is the introduction of the term “sense content” in the last two paragraphs of section 2. I’m not clear on the motive or the justification for this move. And I’m not sure there’s parity in the treatment of act and object along these lines.

Sellars wants two names for the sort of item that enters into sensing-relations as the object of an act of sensing: When the item figures in a sensing-relation, call it a sense-datum. When the item does not figure in a sensing-relation, call it a sense-content.

Sellars notes it's “a matter of controversy among sense-datum theorists” whether “sensed items could exist without being sensed” -- i.e., whether a sense-content could exist without being identical to a sense-datum, without existing as a matter of fact in a sensing-relation as the object of a sensing-act.

Perhaps this is the only motive for the distinction: He wants a language that’s neutral with respect to this particular controversy. This interpretation is reinforced when he settles on the term “sense-content” as perhaps “as neutral a term as any”.

Fair enough. So long as the controversy’s undecided, we use a language unprejudiced with respect to the outcome, with resources sufficient to accommodate advocates on either side. Though I'll want to remember to look for alternative motives for this abstraction as the essay proceeds.

Is there an analogous controversy about the act-side of the sensing relation? Does anyone in this story claim or wonder whether the sort of item that enters into sensing-relations as the act of an object-sensed could also exist independent of any such relation? Perhaps there’s no analogous controversy on this point. That would account for the disparity in Sellars’s account, if it turns out there is a disparity: Either everyone’s agreed, the act in the sensing relation can exist outside that relation; or everyone’s agreed, it cannot. If everyone’s agreed that it can, or if it rather is a controversial matter, we should expect Sellars to offer us two names to refer to the thing called the act, just as he took pains to provide two names for the thing called the object of the act of sensing: One name for the item when it exists in sensing-relations; and one name when, if ever, it exists without involvement in any sensing-relation.

In any case, it may be that by giving two “locutions” to refer to the item that stands as object in sensing-relations, Sellars has defined two distinct predicates. One seems, thus far, a two-place relation, like SENSE (a, x). The other is perhaps a one-place predicate, as for instance, S-CONTENT (x).

If SENSE (a, x) then S-CONTENT (x).


Perhaps Sellars spins the analogy -- between act and object, with respect to independent existence of the item -- in another direction?

He says: “In the latter case” -- the case in which “varieties” of sensing-relations and sensing-facts are sorted according to the varieties of sensed-object, and not by distinguishing “full-blooded species” of sensing --

being a visual sensing or a direct hearing would be a relational property of an act of sensing, just as being a sense datum is a relational property of a sense content. — ”Sellars”

But this seems to jump ahead of the question, can the item that enters into a sensing-relation as an act, also exist independent of any such relation, and do we want our language to have the resources to refer to such items by way of “another locution” that is prima facie logically independent of any such relation? For instance, a one-place predicate, S-ACT (a), perhaps determined in part by the condition

If SENSE (a, x) then S-ACT (a).

Hasn’t he jumped ahead, isn’t there a disparity? Having distinguished in principle, and in the language of the model, sense-data from sense-contents, Sellars neglects the question of whether there is an analogous distinction for sense-acts, and moves on to discuss varieties of sensing-relation.

Just as, he says, being a sense datum is a relational property of a sense content, so being a “visual sensing” would be a relational property of an act of sensing. This does not seem to address the question of existence independent from the sensing-relation; it only tells us what sort of sensing-relation, on the basis of the sort of object in the relation -- “x is a color patch which is sensed”. So not quite “just as”: For there’s no mention of a “locution” by which we might refer to the item that enters as act into the sensing-relation apart from any such relation. A “visual sensing-relation” would be a relation of a sensing-act and a visual sense-content; but what, if anything, are we to call a sensing-act that has no object, or at least a sensing-act as an item logically distinguishable from any sensing-relation?

One way we might characterize the “former” case -- the case of various full-blooded “species” of sensing-relation: Define distinct predicate-relations for each species of sensing, each correlated in the same way with the generic sensing-relation. Say for visual-sensing:

If V-SENSE (a, x) then (VS-ACT (a) and VS-CONTENT (x) and SENSE (a, x)); and
If VS-ACT (a) then S-ACT (a); and
If VS-CONTENT (x) then S-CONTENT (x).

And the same song and dance for each sensory mode assimilable to this form.

Boiled down my notes on sections 1-2.

1. Two tasks in the essay

• attack sense-datum theories as a special case
• criticize the whole framework of givenness

2. Framework of givenness (framework of immediacy)

• “a way of construing” situations analyzable in terms of characteristic bits of shop-talk, with characteristic theoretical commitments, common to most “major philosophical systems”
• some of the terms: given, (sense-)data, observation
• some things that have been said to be given: sense contents, material objects, universals, propositions, real (aka synthetic necessary?) connections, first principles
• “characteristic features of the given”: tbd
• inference vs. "givenness"
  
  • e.g., inferring vs. seeing that something is the case
  • inferential vs. noninferential knowledge acquisition
  • analysis of observational judgments

3. Sense-datum theories (SDTs)

• act-object model
  
  • an act of awareness called “sensing”
  • an object sensed (e.g., a “color patch”)
• the fact that x is sensed, however complex or simple that fact may be, has a form by virtue of which for x to be sensed is for x to be the object of an act.
• “relational properties” of sensing-acts and objects-sensed
• sense-datum distinguished from sense-content
  
  • being a sense-datum is a relational property
  • sense-content is “another locution” to refer to the same item as logically distinct from any sensing-relation or sensing-act
  • two-place predicate-relation vs. one-place predicate?
  • introduced for the sake of neutrality with respect to controversy #3; any other motives?
  • no analogous distinction yet considered for the sense-act. (Does Sellars, or SDT, tend to conflate (or identify) the sensing-act with the sensing-relation?)
• SDT controversies
  
  • Controversy #1: are sensing-acts “phenomenologically simple” or “further analyzable”?
  • Controversy #2: If sensing is analyzable, can it be an act?
    
    • “deep roots for the doubt that sensing (if there is such a thing) is an act”, roots traced to one of two lines of thought tangled together in classical SDT
  • Controversy #3: Can sensed items exist without being sensed?
  • Controversy #4: are varieties of sensing distinguished by sort of object-sensed, or in terms of “full-blooded species” of sensing-act or sensing-relation?
    
    • no mention of “both” as an option?
    • Is this a controversy within SDT, or is it a bit of analysis from Sellars?
  
  4. Look for
  
  • an ambiguity in sense-datum theories: What’s the ambiguity?
  • inferring, seeing, “sensing”: How will the distinction be drawn?
  • play between professional epistemological shop-talk and the common ground of ordinary language and experience
  • hints about motives for sensum/sense-content distinction
  • the third term: Where is the subject of the sensing-act in this story?
  • further definition of SDT
  • further definition of the framework of givenness, its “characteristic features” and theoretical commitments
  
  5. References or allusions to trace through the literature
  
  • Kant and Hegel
  • “dogmatic rationalism” and “skeptical empiricism” (major philosophical systems)
  • What philosophers have attacked the philosophical idea of givenness or immediacy?
  • What defenders have claimed that critics of givenness “fly in the face of reason”, perhaps charging they imply there’s no difference between inferring and seeing (sensing, perceiving)?
  • Who has called each “given”: sense contents, material objects, universals, propositions, real connections, first principles, givenness itself
  • first attackers of “intuited first principles”?
  • first attackers of “synthetic necessary connections”? (Hume)
  • contemporaries who claim to attack “the whole idea of givenness”, but who “are really only attacking sense data”
  • “classical exponents” pro and contra “phenomenologically simple”, “not further analyzable” (controversy 1)
  • pro and contra: if sensing is analyzable, it cannot be an act (controversy 2)
  • deeper roots for the doubt that sensing is an act
  • pro and contra: items sensed can exist independently of being sensed (controversy 3)
  • Who uses each idiom to distinguish varieties of sensing: i) visual, tactual sensing; ii) directly seeing, directly hearing.
  • Is there prior discussion of the bases for distinguishing varieties of sensing (“controversy 4”), or is this Sellars’ analysis?

3. Now if we bear in mind that the point of the epistemological category of the given is, presumably, to explicate the idea that empirical knowledge rests on a 'foundation' of non-inferential knowledge of matter of fact, we may well experience a feeling of surprise on noting that according to sense-datum theorists, it is particulars that are sensed. — Sellars

Need Sellars presume that the category's purpose is to explicate the idea of such an epistemological "foundation"? Perhaps it is presumptuous: The ordinary, unobjectionable sense of "givenness" is arguably prior to the Gothic quest for that foundation. It may be only in tying ordinary intuitions about "what is immediately given" to anxious thoughts about foundation, that the epistemologist constructs the "epistemological category of the given".

What's put in this category: matters of fact grasped without inference, paradigmatically on the basis of "being sensed".

Why does the presumption, or the point presumed, occasion surprise that, for the sense-datum theorist, "it is particulars that are sensed"? Is there some further prejudice by which "objects of knowledge" are characterized as general rules, as opposed to particulars? Does the arguably generic character of our grasp on sensed particulars militate against the claim that it's in fact particulars we thus grasp? Who claims it is not particulars that we sense?

For what is known even in non-inferential knowledge, is facts rather than particulars, items of the form something's being thus-and-so or something's standing in a certain relation to something else. — Sellars

What sort of wedge is this? If a fact involving particulars is counted as known: when, and in what sense, and on what grounds are the corresponding particulars counted as not known? Perhaps this very point is at issue in the controversy over whether "the fact that x is sensed" is analyzable?

Or is the point that, if sensing is merely the sensing of particulars, then it's insufficient for knowledge, for it's facts that are known, and particulars are not in themselves sufficient to constitute facts. This seems reasonable. Though it would be a strange sort of "sensing" by which particulars were grasped -- or indeed not grasped, but "barely sensed" -- bereft of all facticity. Even the fact of being sensed, or being sensed by S, or being sensed at time t, or being visually sensed -- surely some whiff of fact would accompany any such "bare sensing" of particulars?

Does anyone raise the objection, that we may call each fact a particular; and then say, the particulars that are sensed are facts involving non-fact particulars?


A fine bit of shop-talk, "it's facts rather than particulars, that are known". When did this way of construing situations analyzable in terms of "facts" and "particulars" turn up in the tradition? Does Sellars really stand by the slogan, as he seems to here?

I presume the sense-datum theorist, at least, will not object to it.

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?' — Sellars

If the "sensing" of sense-datum theory is mere sensing of particulars, it is insufficient to constitute knowledge. But "sensing" in this theory is associated with the role of providing what's "given", and the purpose of the category of givenness is to provide a foundation for empirical knowledge in noninferentially grasped matters of fact. The theorist must show how his "sensing" can yield knowledge of facts.

The sense-datum theorist, it would seem, must choose between saying:

a. It is particulars which are sensed. Sensing is not knowing. The existence of sense data does not logically imply the existence of knowledge.

or

b. Sensing is a form of knowing. It is facts rather than particulars which are sensed. — Sellars

I'll reiterate my reservations about (conscious, phenomenologically available) "sensing" of particulars without any whiff of fact. It seems a legacy of strange shop-talk that one should feel compelled to make (or feel justified in proposing) this hard choice, either facts or particulars, as if we don't get them both together in perception.

On alternative (a) the fact that a sense content was sensed would be a non-epistemic fact about the sense content. — Sellars

Alternative (a): It is particulars which are sensed. Sensing is not knowing.

The fact that a sense content is sensed -- and perhaps (analytically) related facts pertaining to the architecture of the fact of sensing, that it was sensed by S at t, and so on -- would be nonepistemic facts about the sense content. I suppose they'd be epistemic facts about the sensing-act? But nonepsitemic about the sense content. Of course the fact of sensing gives the sense content as related to the act of sensing, but facts involving relational properties of this sort are "nonepistemic facts about the sense content".

Given as, taken as. It seems we must take things as they're given, however else we may take them.

What is "nonepistemic fact" supposed to mean? Does it mean: to grasp facts nonepistemic about a sense content is not to have knowledge about the sense content -- but only knowledge about a fact of sensing, in which a sense content figures as sort of placeholder, an empty term, until further notice? Do we, in effect, grasp a sense-datum as such, without reaching here the corresponding sense content?

Or do we have perhaps beliefs, but no knowledge, about the sense content, given the existence of sense data?

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. — Sellars

Even if "the existence of sense data does not logically imply the existence of knowledge", there may be another logical connection between them. The denial of that particular implication doesn't rule out the alternative that the existence of sense data is necessary but insufficient for the existence of (at least some sorts of) noninferential knowledge -- another way of putting the "converse" case.

"not itself a cognitive fact": is "cognitive fact" here synonymous with "epistemic fact", or does it have, in addition, some phenomenological implication or connotation? Is the thought perhaps that the fact of sensing is not (ever) directly available to consciousness, to the thinker, to the knower.... That the fact of sensing, its act and its sense-datum, are theoretical constructs, perhaps in principle verifiable by empirical research, but not ordinarily available from the first-person perspective?

What sort of constructs? Theoretical conditions of noninferential knowledge acquisition.

On the second alternative, (b), the sensing of sense contents would logically imply the existence of non-inferential knowledge for the simple reason that it would be this knowledge. But, once again, it would be facts rather than particulars which are sensed. — Sellars

Alternative (b): Sensing is a form of knowing. It is facts rather than particulars which are sensed.

On this alternative, the fact of sensing is sufficient for the existence of noninferential knowledge. Presumably this is knowledge of the sense content. The fact that is sensed is a fact "epistemic" about the sense content. The fact that is sensed is a cognitive fact.

Perhaps this is what Sellars means by relational property? — jkop

See, e.g., SEP on "Intrinsic vs. Extrinsic Properties"

Whenever I hear philosophers go on for long about properties, I get a little dizzy and recall Duns Scotus.

Better to let the world teach us how to speak, than to pretend we understand heaven and earth by practicing grammar.

Glad to have you on board. I envisioned this thread as a kind of free-form reading group, but unfortunately I've since become too busy to participate regularly. As it stands, I suppose this thread could either become simply a 'place' to discuss Sellars' essay, or, if someone else wants to take the reins, could still remain a week-by-week, section-by-section discussion.

You've brought up plenty of good points/avenues of exploration, but, for now, I want to respond to just one. I, too, am a little confused by Sellars' distinction between sense content and sense data - or at least confused by the importance of making such a distinction. The possibility that he draws this distinction simply to remain neutral within a larger controversy makes sense, tho, if that is the case, I wish I understood that controversy and what was at stake (or believed to be at stake) in it.

That said, the distinction between data and content is put into action in section 2 ('Another Language?'):

No one, of course, who thinks -- as, for example, does Ayer -- of the existence of sense data as entailing the existence of "direct knowledge," would wish to say that sense data are theoretical entities. It could scarcely be a theoretical fact that I am directly knowing that a certain sense content is red. On the other hand, the idea that sense contents are theoretical entities is not obviously absurd -- so absurd as to preclude the above interpretation of the plausibility of the "another-language" approach. For even those who introduce the expression "sense content" by means of the context ". . . is directly known to be . . ." may fail to keep this fact in mind when putting this expression to use -- for example, by developing the idea that physical objects and persons alike are patterns of sense contents. In such a specific context, it is possible to forget that sense contents, thus introduced, are essentially sense data and not merely items which exemplify sense qualities. Indeed, one may even lapse into thinking of the sensing of sense contents, the givenness of sense data, as non-epistemic facts. — Sellars

This analysis leads me to think that the importance of the distinction is that sense contents are something like qualities, something universal, while sense data are something fundamentally immediate (keeping in mind the Hegelian point that to deal with 'qualities' is already to deal with universals and mediation.) (Perhaps too, the sense of 'content' in 'propositional content' is analogous? We might say that a proposition expresses some propositional content (we can understand why one wouldn't want to reduce the propositional content of some proposition to that particular proposition, or the particular relation whereby a particular proposition expresses some propositional content)

What do you think? I'm still shaky on this, but that's the best way I've found to understand it so far.

What is "nonepistemic fact" supposed to mean? — CabbageFarmer

I have a pet theory, but I don't know how good it is. So:

(a) It is particulars which are sensed. Sensing is not knowing. The existence of sense data does not logically imply the existence of knowledge. — Sellars

&

On alternative (a) the fact that a sense content was sensed would be a non-epistemic fact about the sense content. — Sellars

I take this to mean that while it is a fact that the sense-content was sensed (in sensing act x and/or by senser y), this does not imply that the senser gained any knowledge about the sense content. If you like, it's a fact in-itself, but not (necessarily) a fact for anyone (any knower.)

"not itself a cognitive fact": is "cognitive fact" here synonymous with "epistemic fact", or does it have, in addition, some phenomenological implication or connotation? — Cabbage Farmer

I think they're synonymous; Sellars himself appears to equate the two.

Finally, I will say of a sense content that it is known if it is sensed (full stop), to emphasize that sensing is a cognitive or epistemic fact. — Sellars

You've really analyzed the hell out of this paper. Nice work, man! Like Calisbury, I'm short on time otherwise I'd participate a little more.

I take this to mean that while it is a fact that the sense-content was sensed (in sensing act x and/or by senser y), this does not imply that the senser gained any knowledge about the sense content. If you like, it's a fact in-itself, but not (necessarily) a fact for anyone (any knower.) — csalisbury

I think this is spot on. My take is that he's ultimately pressing a distinction between two senses of "immediacy": the immediacy of the act of sensing from the immediacy of the content of what is sensed. Or, if you like, between the non-inferential nature of the act of sensing and the non-conceptuality of the content of what is sensed (ugh, there's got to be a better way of saying this).

Sellars' dialectic crescendos in the realization that immediacy of act does not imply immediacy of content, contra what the proponents of "giveness" would have us believe. Just because I know some content via a non-inferential act does not imply that the content thereby known is not inferentially related to other contents, espeially in the sense of being open to challenge for justification.

Glad to have you on board. I envisioned this thread as a kind of free-form reading group, but unfortunately I've since become too busy to participate regularly. As it stands, I suppose this thread could either become simply a 'place' to discuss Sellars' essay, or, if someone else wants to take the reins, could still remain a week-by-week, section-by-section discussion. — csalisbury

Glad to have stumbled in. I've been meaning to work through this text for years, and the group's given me a nudge. I'll chime in when I can. Time's short and you see how slowly I trudge through it.

I, too, am a little confused by Sellars' distinction between sense content and sense data - or at least confused by the importance of making such a distinction. The possibility that he draws this distinction simply to remain neutral within a larger controversy makes sense, tho, if that is the case, I wish I understood that controversy and what was at stake (or believed to be at stake) in it. — csalisbury

Sellars mentions the controversy in (I.2) during his search for "another locution" by which to refer to sense-data. He rejects the term "sensible" for the role, as having "the disadvantage that it implies that sensed items could exist without being sensed", adding that this is "a matter of controversy among sense-datum theorists". So Sellars settles on the term "sense content" as a neutral choice.

It's a strange passage. He seems to want a term that

i) refers "to an item which is sensed in a way which does not entail that it is sensed"; and

ii) does not imply "that sensed items could exist without being sensed" (for this is a matter of controversy)

At this point I don't see what motive Sellars has for criterion (i). Perhaps in his circles, it's a commonplace that one needs something more than a relational predicate to refer to an item in good faith?

Being a sense datum, or sensum, is a relational property of the item that is sensed. To refer to an item which is sensed in a way which does not entail that it is sensed, it is necessary to use some other locution. Sensibile has the disadvantage that it implies that sensed items could exist without being sensed, and this is a matter of controversy among sense-datum theorists. Sense content is, perhaps, as neutral a term as any. — Sellars

I can only guess, and hope the motive's cleared up in the course of the reading.

On the other hand, in light of a controversy about whether sensed items can exist without being sensed, it may be that Sellars is aiming at neutrality, preparing his terms so as to be able to range the discussion over both sides of this particular dispute. Or perhaps he's biased in favor of the claim that sensed items can exist without being sensed.

What's at stake in the controversy, for the sense-data theorist? I suppose it depends in part on what sort of things sense-data and sense contents turn out to be. Another point that's quite unclear.

This analysis leads me to think that the importance of the distinction is that sense contents are something like qualities, something universal, while sense data are something fundamentally immediate — csalisbury

You're way ahead of me in the reading but I'll try a tentative reply.

So far as I can see, a sense-datum just is (for the sense-data theorist) a sense content in an appropriate relation to an act of sensing. Though this is an awkward way of speaking if one denies that sense content can exist without being sensed.

Oddly enough, it seems the "immediacy" of sense-data has something to do with their place in a relation. To be a sense-datum is to be a sense content given in an act of sensing.

If a sense content can exist without being sensed, is there some other way for it to be the object of an "act of awareness"? For instance, can sense contents be imagined, remembered, or thought when they are not sensed? I don't know whether sense-data theorists speak this way. Does this show the lines of the controversy Sellars mentions?

Perhaps one way to avoid speaking that way would be to say: Any act of awareness that has a sense content as an object is or involves an act of "sensing", including the acts we call imagining, remembering, and thinking sense contents.

Do any sense-data theorists want to say: If noninferential knowledge is given in an act of awareness, then that act is an act of "sensing"? Suppose the same sense content can figure both in an ordinary act of sensory perception, and in a "vivid and forceful" imagining that closely resembles perception, like a dream or hallucination. Arguably, in being aware of the imagining, I acquire noninferential knowledge about a matter of fact, that a particular imaginer is imagining a particular sense content. Analogously, in being aware of the ordinary perceiving, I acquire noninferential knowledge about a matter of fact, that a particular perceiver is perceiving a particular sense content. The analogy only goes so far; but that doesn't make it unreasonable to call the acquisition of noninferential knowledge in each case a sort of "sensing".


I hope that's not too far afield. In any case, talk of "immediacy" and "givenness" of sense-data seems closely connected to the idea of noninferential knowledge acquisition, noninferential knowledge of "facts." If sense contents are said to exist without "being sensed", we’ll need to hear more about how else they can "exist", whether they're involved in other acts of awareness that are not called "sensing", and whether they can exist without any involvement in an act of awareness. That should help us figure out whether they differ from sense-data with respect to "immediacy" and "givenness".

(keeping in mind the Hegelian point that to deal with 'qualities' is already to deal with universals and mediation.) — csalisbury

Can you develop the Hegelian point?

I'm never sure what's meant by "qualities" and "universals." I recall Kant speaks of "sensible qualities" -- is that the right ballpark?

I can think of two sorts of "mediation" to watch out for: mediation by inference and mediation by concepts. I expect the given to play the role of foil in this narrative, while Sellars develops an argument supporting a view in which nothing is given in noninferential knowledge acquisition without some contribution from concepts -- or however he puts his version of the Hegelian theme developed in our time by Sellarsians like Brandom and McDowell.