Relocation, Quality Spaces, and Grain

Relocation,
Quality Spaces,
and Grain
David Rosenthal Sellars Centenary Conference
CUNY Graduate Center 8 June 2012
Philosophy and Cognitive
Science
http://tinyurl.com/dr-cuny/

Introductory
Few of Sellars‟ views are as challenging
to interpret and evaluate as those about
sense impressions and their distinguishing
mental qualities.
I‟ll take up some of his views about those
matters, and also urge some adjustments
of his view—knowing he would not wholly
approve of what I‟ll suggest.
But I believe that where I‟ll depart from
him, my departure is not only defensible,
but also in the spirit of much though not
all of what he held about those issues.
Relocation, Quality Spaces, and Grain Sellars Centenary Conference, 8 June 2012 2

Since Galileo proclaimed that the book
of nature “is written in the language of
mathematics,” philosophers have been
struggling to make our commonsense
picture of reality fit with the dictates of
science.
Thus Sellars‟ effort in all his discussions
of sense impressions to fuse the manifest
and scientific images.
Much of what Sellars says about these
topics is, I believe, important and correct.
But I‟ll urge that his view also embodied
a mistaken strand—though happily one we
can readily detach from the rest.
And I‟ll sketch how things look once we
do detach that strand.

OVERVIEW
I. Relocation and
Consciousness
II. Quality-Space Theory
III. Grain and Analogy


I. Relocation and
Consciousness
Galileo‟s dictum casts apparent doubt on
the nonmental reality of proper sensibles
such as color and sound,
since such properties appear to resist being
“written in the language of mathematics.”
How might the red an apple seems to us
to have succumb to scientific treatment?
The usual move is to relocate the manifest
redness to the mind leaving behind mere
reflectance properties, and they do readily
lend themselves to scientific formulation.
Similarly for the other proper sensibles.

Such relocation isn‟t free. One cost has to
do with such spatial properties as shape,
size, and location.
We see the sizes, shapes, and locations of
objects by seeing boundaries among colors
(cf. “SRII,” 356; “Phenomenalism,” 75n., citing Berkeley).
So the properties of states that represent
visible (not, e.g., tactile) shape and size are
inseparable from those that represent color.
But whatever the temptation to relocate
colors and other proper sensibles,
there‟s none to relocate spatial properties.
That‟s a puzzle if the two types of property
are inseparable in sensing: If we relocate
colors, must we not also relocate their
boundaries? I‟ll return to this in II-III.

Another cost of relocation is that whatever
seems problematic about manifest physical
color will then show up in the mind. On
Sellars‟ view, that‟s ultimate homogeneity.
Sellars adopts the relocation move, though
in an unusually subtle way.
Still, that move has a consequence that
has not generally been noted.
The proper sensibles resist mathematical
treatment only as we are aware of them.
Sellars‟ pink ice cube “presents itself to us
as—we are aware of it as—over there”
(Carus III 15; quoted emphasis is always Sellars‟ own).
And it‟s only as we are aware of it that its
manifest pinkness resists the treatment we
can give to wavelengths and molecules.

Here‟s the unnoticed consequence:
When we think of the manifest pinkness of
the ice cube, we‟re thinking of a property
as we‟re aware of it, not as it is in itself.
So if we relocate manifest physical pink
to the mind, we‟ll also conceive of that
relocated property as we‟re aware of it.
Moreover, a sense impression that we‟re
aware of is a conscious state.
So conceiving of sense impressions on the
relocation picture automatically results in
conceiving of those states as conscious.
Sellars wouldn‟t contest this: “The esse of
cubes of pink is … sentiri” (Carus III 66)—
whether as manifest physical cubes of
pink or as sensing states of ourselves.

Sellars sees this relocation as simply a
“recategoriz[ing]” of the physical cubical
volumes of pink that we see “as sensory
states of the perceiver” (Carus III 44; cf. 47).
We come to think about sense impressions
of pink cubes by thinking about volumes of
pink we seem to see out there as instead
being states of our perceiving selves.
This has implications for mental qualities—
and so for Sellars‟ grain argument.
I‟ll argue that the better path is simply to
accept that not all sensations of color are
conscious: We see color both consciously
and nonconsciously.
We may then need not see color terms as
mass nouns for color “stuff” (Carus I 47; 46, 62).

Does the relocation picture require that we
conceive of the manifest physical property
as we‟re aware of it?
Mightn‟t we relocate that manifest property
independently of how we‟re aware of it?
Then we wouldn‟t need to conceive of the
relocated mental qualities as conscious.
No. The physical property we relocate is
problematic for the particulate natures of
the scientific image—and so mathematical
treatment—only insofar as that property
is manifest, i.e., only as we‟re aware of it.
There‟s nothing about these properties in
themselves— i.e., independent of how they
are present to conscious perceiving— that
requires relocation to the mind.

Conceiving of mental qualities as relocated
physical properties results in seeing those
mental qualities as intrinsically conscious.
But if a mental quality is conscious, one
has first-person access to it.
So if being conscious is intrinsic to mental
qualities, it‟s also intrinsic to them that we
have first-person access to them.
Third-person access to such states is of
course not intrinsic.
So if first-person access is intrinsic, it will
trump any third-person access.
We then can‟t block allegedly pretheoretic
intuitions about the inversion or absence
of mental qualities—and hence the so-
called explanatory gap and “hard problem.”

There are also implications for the Given:
A state‟s being conscious means having
subjective access to that state.
But conceiving of the mental qualities as
relocated versions of perceptible properties
as we‟re consciously aware of them results
in seeing mental qualities as intrinsically—
i.e., directly— accessible.
And direct accessibility would be in respect
of specific qualities— so they‟d be Given.
Sellars has an ingenious way to avoid this
standard conclusion:
First-person access to sense impressions
relies on concepts that “are analogical
extensions of concepts pertaining to the
public or intersubjective world” (“BBK,” p. 48).

Thus the subtlety of Sellars‟ relocation story.
“[V]isual impressions are prior in the order
of being to concepts pertaining to physical
color,” since manifest physical colors are
projections of mental qualities.
That‟s the ontological relocation of manifest
physical colors to the mind.
But our concepts of physical colors “are
prior in the order of knowing to concepts
pertaining to visual impressions” (“SRII,” 357).
And that‟s the rejection of the Given, since
concepts of color sense impressions rest
on concepts of manifest physical colors.
Distinguishing ontological from conceptual
priority does thereby avoid the Given—
but not intrinsic consciousness.

Seeing physical colors as projected from
mental qualities means that we‟ve relocated
those manifest physical colors to the mind.
The only reason to do so is that manifest
physical colors as we‟re aware of them
resist the scientific image.
Since we start by conceiving of manifest
physical colors as we‟re aware of them,
we‟ll also conceive of sense impressions
as we‟re aware of them.
Conceiving of sense impressions in terms
borrowed from concepts for manifest
physical colors doesn‟t avoid that.
So we‟ll conceive of sense impressions as
being automatically conscious—and being
conscious as intrinsic to their nature.

On any relocation picture, Sellars‟ included,
sensations likely must be conscious. But
did Sellars himself draw that conclusion?
He never mentions subliminal cases—and
“[t]he esse of cubes of pink is … sentiri.”
He does say there‟s no contradiction in
states of consciousness that aren‟t noticed
or apperceived (SM I ¶24). But that may well
mean only that sensations, though always
conscious, need not be introspected.
One can be aware of something but not its
categorial status (Carus I, 44). So we could
be always aware of sensations—resulting
in their being always conscious—but not
aware of them in respect of their being
sensory states of oneself (Carus I 153).

Would being aware of what are in reality
sensations of color—but being aware of
them as manifest physical colors—result
in those sensations‟ being conscious?
One might think not. One wouldn‟t then
be aware of oneself as having a sensation.
But being aware of something as a colored
physical object in front of one is, according
to Sellars, being aware of something that
is in reality a sensation of a volume of
color, which one wrongly conceptualizes
as a colored physical object. One would
be aware of a sensation, though not as
such, which would be conscious. That
would help explain Sellars‟ apparent view
that sensations are invariably conscious.

Sellars‟ goal was to explain “how we could
have arrived at th[e] idea” of states with
the explanatory power sense impressions
must have (“EPM,” 60, p. 190).
The danger was that this might require
conceiving of sense impressions as Given.
His solution is that we conceive of them
using concepts derived from our concepts
for, e.g., manifest physical colors.
But as long as we conceive of them as
relocated from those manifest colors, we‟ll
see them as intrinsically conscious—since
relocation is motivated by conceiving of
manifest colors as we‟re aware of them.
It‟s useful to compare Sellars‟ account of
our first-person access to intentional states.

After Jones teaches his contemporaries to
posit intentional states, “it now turns out—
need it have?—that [they] can be trained
to give reasonably reliable self-descriptions,
using the language of the theory, without
having to observe [their] overt behavior. …
What began as a language with a purely
theoretical use has gained a [first-person]
reporting role” (“EPM,” 59, p. 189).
Such subjectively noninferential access to
one‟s own thoughts results in their being
conscious, since subjectively noninferential
reportability is the accepted commonsense
standard of a state‟s being conscious.
Need it have turned out that we come such
access? There are reasons to think so.

Jones‟s contemporaries posit thoughts in
their own case as well as others‟, and in
time gain facility in such self-ascribing.
So they come to be disposed to say „I think
it‟s raining‟ whenever they‟re disposed to
say „It‟s raining‟. They come to be—like
us—pretty much equally likely to use one
type of location as to use the other.
This is not something built in. It‟s just that
applying Jones‟s theory to oneself comes
to be automatic and second nature.
Jones‟s contemporaries—like us—need just
to be able to become habituated to apply
his theory to themselves in a way that does
not subjectively rely on any inference or
self-observation (Rosenthal, CM, 301-305).

Sellars gives a parallel story for sense
impressions (“EPM”, pp. 194-5, 62).
Jones‟s theory recategorizes manifest red;
so training is needed to report colors thus
recast as mental qualities.
But if the esse of sense impressions is
sentiri, why is training needed for one to
be able to report such states?
Sellars would reply that pre-Jonesians
don‟t yet have the concepts needed to
report their sense impressions—as sense
impressions.
So Sellars can have sense impressions be
always conscious—though not as such.
For pre-Joneseans, they‟re conscious only
as manifest perceptible properties.

II. Quality-Space Theory
Our concept of visual impressions derives
from concepts for perceptible objects:
“The essential feature of the analogy is that
visual impressions … resembl[e] and [differ]
… [in ways] structurally similar to the ways
in which the colours and shapes of visible
objects resemble and differ‟‟ (“EPM,” 61, p. 193).
Such similarities and differences underwrite
the way our concepts for sense impressions
derive from those for perceptible objects.
But we can rely on discrimination that isn‟t
conscious, rather than conscious perceiving.

Consider colors. Building on a person‟s
reports of whether two hues match or are
just noticeably different (JND), we can
construct a quality space of colors that
reflects the similarities and differences
among all hues, for that person.
Averaging over people, we get a quality
space like this (CIE 1931):
And people can reliably
guess about matching
or JNDs even when
they do not see stimuli
consciously— because,
e.g., the stimulus is
masked experimentally
or otherwise degraded.

Nonconscious discrimination of perceptible
properties is experimentally demonstrable.
Guessing about stimuli seen nonconsciously
due to masking and other techniques is
highly accurate (>80%), and nonconscious
seeing primes subsequent psychological
reactions. Also, confidence in one‟s guess
decreases far more rapidly than accuracy
as the stimulus is degraded.
So we can conceive of sense impressions—
and the mental qualities in virtue of which
they resemble and differ—not by analogy
with perceptible properties as we are
consciously aware of them,
but rather as we can discriminate among
them, whether consciously or not.

Sense impressions are states in virtue of
which we are able to discriminate among
perceptible properties, such as colors.
So every difference we can discriminate
between two perceptible properties
requires states that differ in that way.
So the same quality space that reflects the
differences among properties perceptible
by us fixes the sense impressions
by virtue of which we can make
those perceptual discriminations.
Sense impressions, as Sellars urged, do
“resembl[e] and [differ] … [in ways]
structurally similar to the ways in which
the [perceptible properties of perceptible]
objects resemble and differ” (“EPM,” 61).

This captures how concepts of sensations
derive from those of perceptible objects.
How about the other half—the ontological
dependence Sellars sees of manifest colors,
e.g., on sense impressions of color?
Do we “project or transpose the attributes
of sense impressions [of color] into the
categorial framework of physical things
and processes” (“SRII,” PP, 357)?
We must adopt that relocation picture only
if colors resist scientific treatment,
and they‟ll do so only if we conceive of
them as consciously perceived. So if the
analogy for rests on discrimination that
needn‟t be conscious, physical colors need
not be projections from sense impressions.

Since colors and sounds as nonconsciously
perceived don‟t resist the mathematical
treatment of the scientific image,
we‟ll have no reason to see them as
relocated to the mental realm.
As we nonconsciously discriminate among
them, they‟re simply wave phenomena to
which we respond differentially.
And though we categorize them by appeal
to our discriminative abilities, they are real
independent of being so discriminated.
Once we no longer conceive of colors and
sounds only as we are consciously aware
of them, they are no more problematic for
the scientific image than any other physical
properties. We can be color realists.

Perceptible colors in the scientific image
would be spectral-reflectance profiles. But
for purposes of conceiving of sensations
by analogy with perceptible properties, we
want to stay in the manifest image, which
knows nothing about spectral reflectances.
So the manifest image identifies colors—
independent of how we‟re consciously
aware of them— with whatever properties
of visible objects allow us to discriminate
them in respect of color. A manifest-image
account of what colors are in themselves
must be schematic in that way.
But that does fix the manifest-image colors
as they are in themselves, independent of
our being consciously aware of them.

So we can conceive of sensations by an
analogical extension from concepts for
properties we can sense—understanding
the sensing to be not always conscious.
And since the relevant sensing needn‟t be
conscious—dispelling a need to relocate
perceptible properties such as color—
we also won‟t see sense impressions as
relocated versions of properties as we
consciously perceive them. Sensations will
sometimes occur without being conscious.
Concepts of sensations can derive by
analogy from concepts for perceptible
properties, but sensations themselves
need not be recategorized or relocated
versions of perceptible properties.

Sellars does insist that “sense impressions
… are common sense theoretical constructs
introduced to explain the occurrence not
of white rat type discriminative behavior,
but rather of perceptual propositional
attitudes ... .”
So these commonsense posits are “bound
up with the explanation of why human
language contains … words for perceptible
qualities and relations” (“Identity Approach,” ¶48).
But the appeal is to actual discrimination,
not mere discriminative behavior.
And perceptual propositional attitudes can
be nonconscious no less than sensations.
So we need not be consciously aware of
the relevant perceptible properties.

Has something been lost? What about the
way physical colors appear to us when we
do perceive them consciously? How do
we account for those conscious seemings?
Perceptual sensations aren‟t intrinsically
conscious; so they won‟t always result in
conscious subjectivity.
Such sensations are subjectively present
for us only when they are conscious.
But if sense impressions aren‟t intrinsically
conscious, what explains their sometimes
being conscious? What is it that does
result in those conscious seemings?
No state—whether an intentional state or
a sense impression—is conscious if one is
wholly unaware of it.

So a state is conscious only if one is aware
of it—only if one has some higher-order
awareness (HOA) of that state.
Indeed, that‟s the upshot of Sellars‟ “need
it have?” remark in “EPM” ( 59):
It turns out—need it have?—that Jones‟s
contemporaries can be trained to report
their own intentional states.
But reports—like all speech acts—express
intentional states with the same content
as the reports (e.g., “Notes on Intentionality”).
So reporting one‟s thoughts requires one
to have thoughts about those thoughts—
higher-order thoughts (HOTs).
The HOAs in virtue of which we‟re aware
of our conscious thoughts are HOTs.

As noted, Sellars gives roughly the same
account for sensations.
Indeed, without such a story he couldn‟t
claim that being aware of sensations as
sensations requires Jones‟s new concepts.
But sensing is sometimes nonconscious;
they won‟t ever be conscious without
some HOA of them.
Still, can HOTs—purely intentional states—
give rise by themselves to the conscious
“lighted up” character of our conscious
sense impressions?
Yes. Strikingly, simply learning new words
for one‟s sensations does sometimes—as in
wine tasting—result in subjectively more
fine-grained qualitative experiences.

The new words typically have no prior
semantic associations—as with, e.g.,
„tannin‟ for the new taste of a wine.
But how can simply learning new words for
our sensations result in one‟s having a new
type of subjective awareness of them?
Learning new words enables us to have
new thoughts—in this case new thoughts
about our sensations, i.e., HOTs.
And new HOTs make one aware of one‟s
sensations in new ways. HOTs do make a
difference in what it‟s like for one.
So HOTs must also result in there being
anything at all that it‟s like for one. They
make the difference between sensations‟
being subjectively conscious and not.

Where does all this leave us with sense
impressions, manifest perceptible
properties, and relocation?
Sellars holds that our concepts of sense
impressions of color derive from concepts
for manifest-image perceptible colors.
The physical perceptible colors, however,
are projections from conscious color
sensations.
Banned from the physical realm, color is
relocated as sense impressions of color.
But if we understand perceptible colors in
respect of nonconscious discrimination—
and so as lacking the character they
appear to have in conscious perceiving—
there‟s no need to relocate.

We can still explain manifest seemings—
but not by appeal to the intrinsic nature
of sensations, but rather to the way we‟re
subjectively aware of our sensations.
The spatial aspect of sensing—sensing
size, shape, and location—also allows
treatment in terms of quality spaces.
A sensation of a red triangle exhibits a
mental quality of color that resembles and
differs from other such qualities in ways
that parallel the discriminable similarities
and differences that physical redness has
from other physical colors.
And we can also give just such an account
for the triangular aspect of a sensation of
a red triangle.

The mental qualities of shape, size, and
location resemble and differ from other
mental qualities in their families in ways
structurally parallel to the discriminable
similarities and differences among the
relevant physical spatial properties.
In both cases the perceptible properties
are fixed by relative discriminability.
So we have a uniform treatment of mental
colors and the mental spatial properties we
visually access as boundaries among
mental colors.
We‟ll conceive of both by analogy with
relevant similarities and differences—
with no more need to relocate physical
colors than with visible spatial properties.

As already noted, Sellars stresses that the
geometrical and content properties of
perceptible objects must go together
(“SRII,” PP, 356; “Phenomenalism,” 75n.).
That‟s of a piece with his insistence that
sensations are modeled on perceptible
physical objects— the wafers of “EPM” (pp.
192-3), since their size, shape, and location
are visible as color boundaries.
That‟s all in the manifest image. But the
uniform treatment of mental colors and
of mental spatial qualities of vision would
allow them to remain together even were
we to move to the scientific image.
And having an account of spatial properties
will help as we turn to the grain problem.

A few quick points. Mental qualities do
represent—not as intentional states do,
but by tracking corresponding perceptible
properties.
And they do that by reflecting one‟s ability
to discriminate a particular perceptible
property from its nearby neighbors.
Mental qualities thereby connect with the
intentional content of perceiving, since
that content is about the objects that the
qualities represent.
And though mental qualities are not holist
in the way concepts are, there is a limited
holism: Mental red, e.g., is the property
with the right set of similarities to and
differences from other mental colors.

Jones plainly relied on colors as they
appear to us; so on his theory sensations
will always be conscious.
But we can adjust his view of colors and
sensations to accommodate physical colors
as they are in themselves and sensations
independent of being conscious.
Nonconscious discrimination is responsive
to colors themselves, whereas conscious
visual sensations reveal how they appear.
So in Sellars‟ spirit we can say that in the
order of knowing conscious sensations and
colors as they appear are primary—
but that nonconscious sensations and
colors independent of how they appear
are primary in the order of being.

Can the manifest image accommodate
sensations that aren‟t conscious? The
manifest image, “by stipulation [!], …
does not include … the postulation of
imperceptible entities … to explain the
behaviour of perceptible things” (“PSIM,” 7).
Sensations aren‟t perceptible, nor are they
entities, strictly speaking. They also aren‟t
posited to explain perceptible things.
Still, this stipulation about the manifest
image might lead Sellars to rule out
sensations we‟re wholly aware of—despite
ample evidence available to him for the
occurrence of sensing that‟s subliminal.
And this wouldn‟t apply to thoughts, which
are modeled on speech acts, not objects.

Still, Sellars acknowledged that intentional
states often occur without being conscious,
though he seldom used „conscious‟ to
apply to mental states, as against applying
it to creatures.
Why, then, his resistance to qualitative
states‟ occurring without being conscious—
given that intentional states can and do?
The pre-Jonesean concept of states that
are actually sensations is the concept of a
perceptible property—as we‟re aware of it.
But since, Sellars holds, those properties
are projections of states of ourselves,
in being aware of those properties we are
in reality aware of states of ourselves—
though not as states of ourselves.

III. Grain and Analogy
There have long been exegetical questions
about Sellars‟ grain problem—often tied
to other issues of Sellars interpretation.
I‟ll take for granted a relatively standard
version, which Sellars sees as “essentially
the same” as Eddington‟s two-table puzzle
(“PSIM,” SPR, p. 36/¶102).
And I‟ll argue that the problem is met by
adjusting Sellars‟ appeal to quality spaces—
to qualitative similarities and differences—
so as to accommodate sensations and
mental qualities that fail to be conscious.

In “Philosophy and the Scientific Image of
Man,” Sellars asks: “[C]an we define, in
the framework of neurophysiology, states
which are sufficiently analogous in their
intrinsic nature to sensations to make …
identification possible?” (SPR, p. 35/¶99).
The answer,” he continues, “seems clearly
to be „no‟”—because of the ultimate
homogeneity of “the perceptible qualities
of things” (p. 35/¶100), which neural systems
lack. (Cf. “Identity Approach,” PPME 179, ¶45.)
“[C]olour expanses in the manifest world
consist of regions which are themselves
colour expanses”—however tiny (p. 35/¶100).
Can we reconcile manifest-image ultimate
homogeneity with the scientific image?

A property is manifest if one is aware of it.
Still, the properties of the manifest image
may not be in themselves as we are aware
of them (Benham‟s disk and Fechner colors).
Things in the manifest image have natures
independent of being perceived—
consciously or not.
Appearance differs from reality within the
manifest image —and so nonconscious
discrimination of perceptible properties may
mislead less than conscious perceiving.
So the appearance of perceptible colors as
being ultimately homogeneous (“dissective”—
Goodman 1951), does not by itself establish
that those properties really are in
themselves ultimately homogeneous.

Sellars‟ use of quality spaces—similarities
and differences—to fix the perceptible
properties appeals to conscious perceiving.
So it restricts us to perceptible properties
as those properties appear to us in
conscious perceiving.
But we can accommodate perceiving that
isn‟t conscious by appeal to quality spaces
built from discriminative ability—matching
or JNDs.
Since discrimination needn‟t be conscious,
we‟re no longer confined to perceptible
properties as they consciously appear.
So we‟d need an additional reason to
credit the appearance that perceptible
properties are ultimately homogeneous.

When an object of the manifest image
appears to have a color it doesn‟t actually
have, we can explain that illusion—
in terms suitable to the manifest image.
That‟s a reasonable way to test whether
a particular case of distinguishing reality
from appearance lies within the manifest
image—or is instead a contrast of the
manifest with the scientific image.
And we can explain, in terms distinctive
to the manifest image, why perceptible
colors—and corresponding mental qualities
of color—would appear to be ultimately
homogeneous,
even if they aren‟t that way independent
of how they appear subjectively.

An expanse is ultimately homogeneous if
every part, however small, exhibits the
color of the whole expanse.
We conceive of an expanse of perceptible
color and mental color qualities in terms
of position in a quality space—
in terms of how similar the perceptible or
mental color is to neighboring properties
in its quality space.
Indeed, we conceive of perceptible and
mental colors exclusively in such quality-
space terms.
And since there are also mental qualities of
size and shape, there‟s no problem about
how we conceive of a mental expanse of a
mental quality of color—or its mental parts.

A sensation of a red triangle has a mental
quality of red, and also a mental quality of
an expanse with triangular borders.
Mental red occurs in a mental triangular
expanse when it occurs in a way that
resembles and differs from other mental
expanses in ways that parallel the ways a
visible triangular expanse resembles and
differs from other visible expanses.
But the mental red itself is solely a matter
of relative position in a quality space of
mental color qualities.
So any mental part of that mental expanse
will also appear to have a mental color
that‟s solely a matter of relative position
in a quality space of mental color qualities.

Every mental part of a mental expanse of
mental red will seem to exhibit mental red.
One might object: If we use location in a
quality space to fix mental qualities, that
location fixes not only their appearance,
but their intrinsic reality as well. With
“sensations,” Sellars claims, “the analogy
concerns the quality itself”—the “intrinsic
character [of] sensations” (“PSIM,” p. 35/¶99).
Since mental qualities are manifest-image
posits, perhaps their intrinsic nature is fixed
by the analogy that underwrites the posit.
With thoughts, the analogy fixes only their
roles— “leav[ing] open … their intrinsic
character (¶99). But the sensations analogy,
Sellars insists, fixes their intrinsic nature.

Thoughts are fixed by their conceptual
roles—conceived in relational terms.
But location in a quality space is no less
relational: It‟s how much a particular
perceptible or mental color, e.g., resembles
or differs from others in its family.
So it‟s unclear why an analogy built on that
relative location fixes intrinsic nature more
than an analogy based on conceptual role.
Why would Sellars regard mental qualities
as fixed in respect of intrinsic natures?
Presumably because of consciousness:
It‟s tempting to see both mental qualities
and perceptible properties as subjectively
presenting themselves to us as they
actually are intrinsically.

If we relied just on how mental qualities
subjectively present themselves,
we might see the subjective appearances
as fixing their intrinsic natures.
But that would hold only insofar as the
mental qualities are conscious.
In themselves they‟re fixed by location in a
quality space—relatively, not intrinsically.
And the temptation to see consciousness
as fixing the intrinsic natures of sensations
is in any case mistaken:
The HOAs in virtue of which sensations
are sometimes conscious represent them
in terms of their mental qualities—
and so in terms of those qualities‟ relative
location in a quality space.

Indeed, it‟s clear even subjectively that
consciousness represents mental qualities
in relative terms. Thus we consciously
distinguish hues more finely when other
hues nearby in the quality space are
consciously available for comparison.
Conceiving of mental qualities on analogy
with the discriminable similarities and
differences among perceptible properties
does not appeal to the intrinsic nature of
either type of property.
So the quality-space model, independent
of consciousness, explains why sensations
appear to be ultimately homogeneous—
and also leaves it open that, in themselves,
they are not ultimately homogeneous.

Thank you for
your attention


Relocation, Quality Spaces, and Grain

More Related Content

What's hot (19)

Recently uploaded (20)

Relocation, Quality Spaces, and Grain