Noël Carroll
Philosophy of Art: A contemporary introduction
(1999)
[79]
Part II
Theories of expression
What is expression?
... even if not all art is expressive, much art is. ... [so] , we still need to say something theoretical about expression and art, inasmuch as quite a lot of art is expressive. ...
... in ordinary language. ... sometimes ["expression"] behaves like a synonym for "representation." ... "The White Paper expresses the British position" ... "The White Paper represents the British position." However, ... philosophers of art, ... typically when they talk of expression, they intend it to contrast with representation. ...
Another broad meaning of "expression" in everyday speech is roughly "communication." ... But ...
[80]
... this too is a broader sense ... than philosophers of art have in mind. ...
... when philosophers of art talk about what poems express, they are not thinking broadly about the communication of ideas. ... [but rather about]
certain human qualities
(also known as
anthropomorphic properties)
—notably, emotional tones, moods, emotively colored attitudes, ...
When I say "this poem is angry" I could mean at least two things: that the poet
reports that he is angry
... or that the poet
expresses anger.
... One could say [i.e. "report"] "We are angry" in the same tone of voice that one could say "Everyone here is over five feet tall." [But] ... when we express anger in life or in art, we
manifest
anger—we
show it forth.
... the quality of our anger saturates our utterance. ...
To express anger, in this sense, is to
get the feeling of anger across
—to make it perceptible (to
embody or objectify
it). ... Roughly ... to express ... is to manifest outwardly an emotive property— ...
... an artwork may also manifest a range of other human qualities ... we might say "The story is courageous" or "It expresses courage," ... Thus, what may be expressed are not only
emotive qualities,
but
any human qualities,
...
... "anthropomorphic properties"
[81]
(properties that
standardly apply only to human persons).
... in the previous chapter. ... the notion of representation applied to certain kinds of things, but not others; ...
In contrast, the domain of expression comprises ... the kinds of qualities and properties that can be applied generally only to human persons. ... to say that an artwork expresses x means that it manifests, exhibits, projects, embodies, or shows forth some x where x is a human quality (some anthropomorphic property) ...
... not all artworks need be expressive. But expression occurs quite often in art— ... Frequently we commend artworks ... for their expressive powers. ...
Sometimes we condemn artworks for their expressive properties ...
But how do we go about
attributing
expressive properties to artworks? ...
... the common view ...
An artist expresses (manifests, embodies, projects, objectifies) x (some human quality) if and only if (1) the artist
has been moved
by a feeling or an experience of x to
[82]
compose his artwork (or a part thereof); (2) the artist
has imbued
his artwork (or some part of it) with x (some human quality); and (3) the artwork (or the relevant part)
has the capacity to give the artist the feeling or experience
of x when he or she reads, listens to and/or sees it again, and, consequently,
to impart the same feeling or experience
of x to other readers, listeners and/or viewers.
...
...
... call this the sincerity condition. ... if Beethoven's A Major Symphony truly expresses joy, on the common view, Beethoven must have felt joy while composing it ...
... In the course of daily events, we often observe expressive behavior in our acquaintances. ... when
we say
that [a friend's] voice quavering expresses nervousness, typically
we think that it provides evidence
that she is nervous. ... The outward manifestation of nervousness is linked (some would say conceptually) to her inner state. ...
The common view of expression in art extrapolates from this ...
[83]
... [i.e.] genuine expression of an emotion indicates that whoever
exhibits
the emotion also
has
the pertinent emotion. ... Thus, ... the common view holds that for an artist to express a certain feeling requires that the artist have the feeling. ... This would seem to follow from the ordinary usage ...
But this argument assumes that the concept of expression is always
applied in art
in the way it is usually
applied in life.
And this is mistaken. ...
We do not require genuine sincerity from actors; after all, they're actors! ...
... much art is commissioned. A film director might be hired to make a film—say a film noir— ... She can do the job without feeling angst. ...
Of course,
it is not always the case
even in ordinary speech
that a sad expression
warrants the inference
of a sad mental state. Some people have sad faces all the time— ... ordinary speech also countenances usage that does not suggest that expression requires the pertinent
[84]
accompanying psychological state. And
much art
is like this. ...
Provisionally we might mark
a distinction
between two uses of expression.
In one case, ... we mean that ... a genuine expression of sadness by a person either contingently or necessarily typically
counts as evidence
that the person is sad. Call this the
expression
sense. But another sense ... means only that x
is expressive of
sadness— ... where sadness
describes how something looks or sounds,
but
does not indicate that anyone
is
in the psychological state of sadness. Call this the
expressive
sense ...
... we often attribute expressive qualities to nature. We say the drooping, weeping willow tree is sad. ... we cannot be implying that the tree is in a state of psychological sadness.
Trees cannot be sad.
... we can attribute sadness to the picture of such a tree in the same way. ... by parity of reasoning, we
can say
that the picture is
expressive of sadness,
without
attributing sadness
to its creator. Therefore, it is an error to claim that an artwork necessarily expresses a certain human quality x only if the artist is moved by the feeling or experience of x.
Is this commutative?
i.e. May we similarly attribute sadness to our friend without saying per se that she (or maybe just her face) expresses sadness?
Can we stipulate-to-self (intellection + projection) that she simply must be sad, i.e. based on known adverse circumstances, and in absence of any observable expressive tells to that effect?
If not, then we will be wrong there too, anytime the real "sadness" is sufficiently held in.
If so, then we will be wrong some more, anytime someone is not made sad by circumstances that we think would make us sad. (Think is the operative word)
We can count on both of those scenarios arising at least here and there; and we cannot count on anything at all to decode them.
All of this on top of being frequently wrong, already, about mere sad faces, which very well ought to be the most straighforward cases of all but which are (Surprise!) actually even more frequent and deceptive than the above anomalies
And so by that time, even the willow tree Theory of Expression is quite tenuous: the empirical basis for even its most modest sort of claim does not seem to be a very robust basis at all.
Whither the empirical robustness? Perhaps it's entirely consolidated in the intersubjective agreement which emerges in certain well-known (limited?) cases, e.g. the venerable willow tree or basset hound examples. The now-infamous Laboratory Cognitivism of the past half-century does seem to have proven agreement in a variety of other (human) matters. But has it proven that the proposition thus agreed upon has a real empirical basis? It seems rather that agreement is entirely independent of real correspondence, i.e., that people who perfectly well know and agree that trees cannot be sad will nonetheless find it easy-as-ever to agree, also, that a given tree expresses sadness. And by that time "expressing sadness" has come to mean something other than what it means in Philosophy of Art; it no longer means what common parlance suggests; "expression" cannot mean anything conventional or even esoteric, actually, because expression now is precisely the wrong term for something that is taken in but never was "pressed out." Trees cannot outpress either!
If myriad non-denotative cases can evince a level of agreement to rival the most mechanistic cognitivism, then robust, mechanistic cognitivism is still just an answer in search of a question. Isn't that the whole thing right there?
... the common view also maintains that in order to express x the artwork must move audiences to feel or experience x. We can call this the arousal condition. ...
...
[85]
A piece of music ... I can apprehend that it expresses hope without becoming hopeful myself, ... If you say the weeping willow tree ... is sad, I can see what you mean— ... —without becoming sad myself.
... three very common types of cases that show that the arousal condition is not necessary.
First, many of the works of art that do arouse feelings in audiences arouse
audience emotions that are quite different
than
the ones expressed in the work.
Parts of Dostoyevsky's Crime and Punishment express remorse, but readers do not feel remorse (that would be implausible—they haven't killed anyone); ...
Second, sometimes artworks express human qualities other than emotive ones. ... fortitude [e.g.] is
not something that you can arouse
in audiences. ... if we are moved to some feeling by an expression of fortitude, it is not fortitude, but something else, ...
Why did we expand expression in the first place, to cover anthropomorphic qualities that are not emotions, only now to play gotcha?
...
Third, some artworks are expressive of anthropomorphic properties like anger, but they
lack the resources
to arouse anger in audiences. ... To be
aroused
to anger, I must have some
object toward which
I direct my anger ...
This seems like Ground Zero for The Whole Problem with Analytic Philosophy.
Sure, it's conventional in speech that anger have an object, but this is not quite so bankable either in action or in consciousness.
Literature can supply us with the requisite objects ... But pure orchestral music provides us with neither objects nor descriptions, ... We cannot be angry without something to be angry about ... And yet some music seems expressive of anger. ...
I would be surprised to find that a good number of these situations are not easily explained: the object of anger is the music, which has some semantic content that quite literally angers the listener.
In other words, the music may be pure but the listener is unable to meet the music on its own ground . . . for semantic reasons; reasons which may be universally shared within a group, or which may be absolutely peculiar to that individual.
...
[86]
...
...
So far we have been focussing attention on the sincerity and the arousal conditions ... We have not said anything about the ... condition in the theory that an artwork expresses x ... only if the artist has
imbued
the artwork ... with x. ...
... (is being imbued with the property of joyousness less in need of explication than the notion of being expressive of joyousness?) ...
Expression, exemplification and metaphor
...
[87]
...
... there is some human quality that the artist wants to refer to in order to draw our attention to it. ...
Suppose we want some blue paint. We go to the store ... There are little squares of various blue paints—a navy blue square, a prussian blue square, ... This is intended to show me what the paint in the cans marked "navy blue" looks like. ... But it doesn't merely
stand for
the navy blue paint in the cans; ... Rather, the little square
shows us some of the properties
of the paint, ...
... What is it about the color sample that enables it to exemplify the paint in the cans? It possesses the same property ...
Exemplification is a common form of symbolism. ... the dessert tray ... the sneakers in the display window ...
... Standardly samples possess
[88]
only some of the properties of the items they exemplify. ... The piece of sausage in the package is larger in size than the sample. ... In order to exemplify the taste ... , the sample ... need only possess the properties it is intended to exemplify.
Exemplification ... is not the same as representation. The
domain of representation
is
persons, places, things,
events and objects.
The
domain of exemplification
is
properties.
In addition,
possession of the relevant property
is
a necessary condition
of exemplification;
x cannot be an example of y,
unless x possesses y-ness.
... But
x can represent y
without sharing any of the properties of y;
...
... not every can of Coke exemplifies every other ... though each ... shares a massive number of properties ... Why not? Because
ordinary
Coke cans do not
refer
to other Coke cans. ... [To do so, one] must be
selected and displayed
in a
communicative context
where it
functions as a symbol.
... Just as these objects function to exemplify certain properties of their kind, expressive artworks exemplify the properties that they express. ... in expressing something the artist exemplifies it for the audience.
For example, the novelist tries to capture the qualities of a certain form of mourning ... By
embodying
certain properties of mourning the artwork
refers
to mourning ...
[89]
... Just as the waitress exemplifies chocolate cake by means of a sample, so the artist exemplifies the sadness of loss by showing certain of its properties.
... We can
reflect upon
the sadness exemplified by an artwork without
becoming
sad. ... To
exemplify
sadness only requires referring to aspects of sadness by way of
possessing
them.
...
Saying that imbuing an artwork with an anthropomorphic quality involves exemplifying it makes some headway in explicating the notion of artistic expression. ... But music, it may be argued, is not literally stately; people are stately. ... only persons are sad. ... a piece of music has no psychology. ...
... What the exemplification theorist adds [then] is that the music is only
metaphorically
sad, not literally sad. ... Expression on this augmented view, then, involves three elements:
reference,
possession
and
metaphor.
...
[90]
...
Expression, then, is reference plus metaphorical possession, or, ... metaphorical exemplification. ... How does someone possess anything (my car, for example) metaphorically? One either
possesses something
or one
doesn't.
... let us take this notion of metaphorical possession to mean that the artwork possesses some
literal properties
... that the audience is
entitled to describe
in terms of some
appropriate
metaphor.
... when attributing a quality like sadness to the music, we cannot pick just any metaphor that we want. It must be a metaphor that fits the music— ...
...
... what makes a metaphorical attribution appropriate? ...
...
... King Richard did not literally have the heart of a lion; ... [But] for the people of his times something about the phrase seemed
[91]
right. ...
According to the exemplification theorist, metaphors involve
the transfer
of one set of labels
from
an indigenous field
of application
to
an alien field.
... [e.g.] from the field of animal biology ... to the alien field of human character traits ... But what is meant by "field" here?
Consider temperature ... a
descriptive scheme
of
interrelated categories.
... But often we apply these terms to things other than degrees of heat. ... We say that singer's style is hot, that the commissar's is cold, ...
On this view ... metaphors are always
systematic.
Whenever we apply a metaphor, we are
implicitly mobilizing an entire scheme
... If we say that Tina Turner's singing style is hot, we are assuming that this contrasts with some tepid singing style ... even if we don't articulate, including to ourselves, the rest of the scheme-to-scheme mapping. ... where a metaphor
appears appropriate,
that is a result of its
fitting into the scheme-to-scheme mapping
in the right way.
"Right way?" ... "Venus is to Jupiter as the the Indian Ocean is to ________ ," ... if we are comparing planetary objects with oceans along
the dimension of size,
then the Pacific Ocean is the right answer, since it is
the largest ocean
on earth,
just as Jupiter is
the largest planet
in our solar system.
... the Pacific Ocean has the same relative position in the scheme of ocean-terms that Jupiter has in the scheme of planet-terms.
Are there any inappropriate juxtapositions of scheme?
Presumably, here, this would involve mobilizing some scheme in which the dimension of size did not figure.
" . . . Venus is to Jupiter as Top-Flite to Titleist . . . "
But it's very difficult to form such an example which permits of no mapping at all: the latter two are both golf balls, both are (let us say) exactly the same size and serve the same general function, but they serve that function differently, as even a long-dormant hack golfer like me ought to know.
Who's to say that this metaphor isn't "appropriate?" The issue seems rather to be that it is quite literally unconventional, in the sense of its "mapping" being esoteric or private rather than widely or intuitively understood. But who's to say that parties to this understanding have not in fact stumbled onto something profound rather than merely vexing? For all the rest of us know, they could literally have stumbled onto it, and the alternative to sharing this profundity with the world is to take it with them to their graves . . .
. . . or, perish the thought, they could just say what happened instead of concealing the profound part behind some gratuitous literary displacement.
...
[92]
Metaphors, then, are always
homologies.
... "x is related to a as y is related to b" (x/a::y/b). And there
is a logic to
homologies. ... When I say that the movie Seven is paranoiac, that is appropriate because it fits
a larger,
though
generally unstated
homology, in which other movies ...
correlate
with other contrasting human qualities, like optimism. Or, Seven
is
paranoiac
as
Meet Me in St. Louis
is
optimistic.
Single metaphors, then, are parts of systems—
underlying systems of contrasts.
... Making a metaphor involves two different things:
choosing one scheme
of labels
(deliriousness, placidity, and so on)
and
mapping that scheme
of labels
onto another
scheme
(Van Gogh's paintings, Monet's, and so on).
Forgive me, I guess, if the logic of this is less than perfectly transparent to me.
The one scheme is a scheme of "anthropomorphic qualities," the other is a scheme of object-works. What if we say, then, that "Deliriousness is a Van Gogh and placidity is a Monet?" This is no more or less logical than the obverse construction; it's even comprehensible. I might even argue that this really is the way such systems of contrasts get created ("created") in the first place; that schemes of "anthropomorphic qualities" are themselves mere descriptions of objects (people-as-objects); that it makes no less sense to explicitly define "placidity," say, in relation to a painting or a rock as in relation to some intimately-known person; and that, finally, the reason it makes no less sense to apply placidity to a painting as to a river has nothing at all to do with "logic" or "expression" and everything to do with the river and with the painting and with the people and with who-knows-what else we may (ever!) have 'observed' to exemplify the quality with which we have come to associate the word.
Actually, this better accounts for the fact that such things simply aren't said very often, or very sincerely, or very accurately. The opposing, humanistic view has a much harder time here.
At minimum, it seems that we do not find all possible mappings to be equally apt. Moreover, logicality seemingly has nothing at all to do with aptness.
... we can opt to project any scheme of labels onto any other scheme of labels. Which scheme we project is
almost arbitrary.
However,
once we decide
on a scheme, the mapping we arrive at will not be arbitrary. If we choose to project
the temperature schema
onto
the singing style schema,
then it is not arbitrary that we call Tina Turner's style hot and Al Gore's tepid. If anyone said that Tina Turner's style were tepid,
we would say
that
he made a mistake.
So, . . .
" . . . Heat is Tina Turner, tepidity is Al Gore . . . "
No? Yes and No?
Though the choice of
which scheme to project onto another
is virtually wide open,
the way we go about correlating items
with each other in respective schemes does not seem to be. There is generally
a surprisingly high degree of convergence
in these matters.
Very true.
Is there any
logic
at all?
We were promised some logic.
Convergence ≠ Logic!!
Given the nonsense syllables "ping" and "pong," most people will agree that
violins go with ping
and
tubas go with pong.
That is, there are
certain criteria
that make some mappings appropriate and others not, namely, that
the structure of contrasts
in
the indigenous realm of labels
should be
isomorphic
(or approximately isomorphic) with
the structure of contrasts
that inheres between the relevant items
in the alien realm of labels.
Well . . . okay!
What exactly is well and truly isomorphic between tuba and pong? I am uniquely well positioned to say in great detail what it is. I had better not waste my breath doing so, not unless and until I encounter someone who insists upon not accepting it.
But this (still!) is not "logic," not unless I am completely unable to understand what "logic" means in this context. Our Man has made the case so strongly that he has veritably overshot logic and landed in essentiality.
In matters of mere logic, convergence is never assured; unfortunately for the logicians, and perhaps fortunately for everyone else. Convergence is its own thing. If convergence is any leading indicator at all of logicality, this is analytically negligible and pragmatically harrowing.
We know we have made our habitual-but-unwitting end-run around essentiality (and into logic, finally) when someone says:
"But it doesn't sound like a tuba?!"
Ah, but it does, for you will pong it as soon as the violin enters, whether or not it too "sounds like a violin."
The venerable and wise Howard Johnson reported this as a common response to his group, Gravity; to which he added:
"This is what tubas sound like. I can't account for what you've heard up until now."
Indeed, the "convergence" here is far easier to account for than is the "logic." And to be sure, we live (we must live) day-to-day on the level of (mere) convergence. Logic, in contrast, is deadly, literally and otherwise. We must think that way, but we cannot live that way.
Hence it really is impossible to account for what someone has heard (or seen, or, most of all, thought or felt) up until now. It's totally impossible. This is not just a humorously enlightening liner note zinger; it is the truth. We cannot observe them from cradle to grave. We can't even take a dry inventory of stimuli to which they have been exposed. Perish the thought, then, of actually becoming them ourselves, so as to know, somehow, just what all of those stimuli really come down to.
In such terms, I presume that "logic" per se is a mere workaround that we have discovered, only with brutally painstaking effort, the one and only method for getting everyone on the same page, if and when that (and nothing else) is what is needed.
Logic is needed! At the same time, we simply cannot live all on the same page!
So, I now understand the above argument, re: metaphor, I'm even inclined to buy this argument; but I still see only the convergence and I fail to see the logic.
This provides a neat
way in which to determine
whether the metaphorical attributions we make of human property terms to nonhuman artworks are appropriate. Thus, if a musical composition
refers to a property
(say, nobility) that can be
attributed to it
[93]
metaphorically
in the appropriate way,
then we are warranted
in saying that the music is expressive of nobility.
In other-but-the-same words:
appropriateness
of
metaphorical attribution
is
a matter of
isomorphism
between two
structures
of contrast,
here called
the
indigenous
contrast-structure
and
the
alien
contrast-structure.
e.g.
"Pong"
is "alien" to
"tuba,"
but
it is also "isomorphic" to "tuba."
e.g.
"Pong" belongs to a different "structure of contrast" than do the more customary/familiar tuba-descriptors;
and yet,
"pong"
triggers
something, some
isomorphism
which is (somehow!)
neither
customary nor familiar (nor perhaps logical!),
and yet (somehow!)
was there the whole time,
and which
simply will be
triggered by any "appropriate" "attribution."
The triggering itself
just is the evidence
of "appropriateness."
Buuuuut . . . can we name any indigenous descriptors? Say, 'loud' or 'low' or 'brassy'? These seem prima facie to be strong candidates, simply because they are conventional. They are familiar: they are often applied to the tuba. They seem to denote objective properties of the instrument; it would be absurd to argue that they do not do this. Perhaps they also 'trigger.' But are they really literal? How could they be literal? 'Loud' is relative, 'low' is isomorphic-metaphorical, and brassy is ostensive-stipulative.
There seems to be only one truly "indigenous" descriptor:
"it sounds like a tuba."
i.e.
You know it when you see
(hear)
(but more likely see)
it.
i.e.
The realm of the "indigenous" appears in the present context as a mere tautology: whatever sound the tuba makes,
that just is
'what tubas sound like.'
We can't actually mark the boundary that the indigenous-alien distinction posits.
And if tubas do not
always
sound exactly the same way, that is a mere problem of logic; it is no problem at all for life nor for convergence.
It is
impossible
to say
logically
exactly what tubas sound like. Conversely, it is
easy
to articulate the point upon which listeners converge, in that particular arena.
All of this is entirely unproblematic for the non-literaries and the non-philosophicals among us. It seems the problems begin when we stake something important on the search for a logical basis.
Fortunately for everybody, most of all for Our Man Here, the task of determining the appropriateness of any given metaphorical attributions of human property terms to nonhuman artworks is a task which really could not matter less.
...
Some problems with the
theory of metaphorical
exemplification
... for all its admirable precision, it possesses certain nagging limitations. We may begin ... by asking whether the theory ... supplies necessary or sufficient conditions for attributing expressive properties to artworks.
... [It] works very well when we are thinking of opposing columns consisting only of single labels, like:
|
hot tepid cold |
Romantic music Muzak Electronic music |
...
Suppose [however] we say that a poem expresses
some very complex
emotional property, such as "late nineteenth-century, fin-de-siècle, bohemian, anarchistic despair and loathing." It is certainly very difficult to imagine reconstructing the indigenous scheme of contrasting labels ... Indeed, it is fair to presume there is none, ...
Indeed. And yet we say such things, which say-ing presumably comes from somewhere and does not come from nowhere. But if not from any such indigenous scheme, then from where and whence?
And yet many of the expressive properties that we attribute to artworks are
at least
this complex. For example, the movie King Kong is expressive of mid-century American brashness, naïveté, and sentimentality.
Hmm . . . we're hot on the trail of something here. But now there is a more immediate problem with mid-century, which is at once over- and under-defined: not every American who lived in it was three things, let alone the three things given above, all at once; yet somehow this is no obstacle to making artworks which "express" all three, and it is certainly no obstacle to thinking that is what you see in those artworks, whether it is "really" there in any given case, or whether it really exists (existed) at all. (Perhaps see L. Menand on the mutual 'misprision' of French and American artists and intellectuals ca. "mid-century.)
...
[94]
In the same vein, there appear to be many
"one-off"
metaphors ... that do not seem
to belong to
a scheme of contrasts. For example, ... "ghostly galleon." ...
... John Waynesque [e.g.] ... is
intelligible
without
there being a scheme of
actor-labels
... Rather, the fact that we have certain associations with John Wayne ... is enough ...
Nor is metaphorical exemplification a sufficient condition for attributing expressiveness to an artwork. Using
the logical machinery
of scheme transfer,
😡 😡 😡
we may say, metaphorically, that one painting
appears
(but
is not literally)
brittle, where this
implicitly contrasts
with some other painting or paintings that are strong.
" . . . which contrasts with . . . lightweight . . . which contrasts with . . . heavy-duty . . . which contrasts with . . . out-of-the-box . . . which contrasts with . . . customizable . . . which . . . "
Thus, the theory warrants our saying that the painting expresses brittleness. But brittleness is
not the sort of thing
that
gets expressed
in artworks—it is not literally an emotive property or a human character quality.
So . . .
this is
just one of those things
that
we may say
. . .
or
"we may" just think it
without saying anything.
Remind me why brittleness is not an "anthropomorphic" property? Previously that category was explicitly not limited to emotive or character qualities, but rather to all human qualities.
"The suffix '-morph' generally means
form, shape, or structure."
(Google)
Surely people can be "brittle," in their 'form' if not in their 'content.' Just as surely for Our Man Here, paintings can be "angry." Why are these attributions "alien" rather than "indigenous" to their targets? The reason seems to be that we are very interested in people, in their emotions, and in churning all of that over in the mind until we can twist and bend it into something that meets our needs. In other words, people matter more than paintings do. If you sincerely think the reverse, then obviously you are a bad person for thinking so. But this is the farthest thing from logical machinery. It would more accurately be called a bias for social information. One result is that "brittle" is made indigenous to the "scheme" of ordinary objects while "angry" is made indigenous to the scheme of people. There is no obstacle whatsoever to applying these terms in each other's domains; actually, it's perfectly "logical" to do this, according to Our Man. But we may say even so that those domains, in that case, are "alien" to these terms.
Who am I to suggest that things could be any different? I do not think you should value paintings over people. I think that is as bad an idea as you think it is. But I don't think that calling this a "bias" is a bad idea, because that's what it is. If a term can be applied at all, if indeed its application can be called "logical," then I fail to see the logic of labeling the domain of application as somehow "alien."
Thus, the theory that we are working with counts artworks that are not expressive as expressive.
That is, a "brittle" artwork is not "expressive" solely on the basis of its brittleness. Why? It is not because people cannot be brittle, but rather because this is not the sort of thing we most desire to know about them.
Here, the theorist may say that we really ought to reconsider what gets expressed in artworks by consulting his theory.
I suppose this is something that my arguments here also entail. But we still lack a workable account of just how anything at all ever gets expressed in the first place, and the "how" figures to be the single most important mediator of the "what," does it not?
But with equal justice, we may respond that the theory has failed to find its mark.
Also, remember that the theory of metaphorical exemplification maintains that
the choice
of the indigenous scheme is
arbitrary.
But this cannot be correct. If we map a scheme of farming devices onto an array of artworks, we can wind up with the conclusion that some poem expresses tractorness (rather than reaperness). But even if there is some poem that
could be aptly described by
this metaphor, it
cannot express
tractorness, since tractorness
is not something that can be
expressed.
This is the most frustrating book and author I've encountered in quite a long time.
"Anthropomorphic" is a funny term. Its close relative is the verb, to anthropomorph-ize. The act/action of 'anthropomorphizing' more often than not denotes something untoward. The humble tractor makes for an unconventional example, but it is a fine enough example for my purposes, as would be anything conventionally called a 'vehicle.'
Owing to the need for safety, for drivers to go where they're looking if not to look where they're going, vehicles usually have 'faces.' Many have lights for eyes and bumpers for mouths. This is often the basis for fanciful children's cartoonizations, but even in so-called real life, it's often all right there, right on the ol' aesthetic surface, just waiting to be noticed and subsequently ____-ized.
Lying between the animated cartoon and the inanimate reality, there is a large uncanny middle, occupied by the various schematic or minimal representations of vehicles, e.g. those that appear on traffic signs and in advertisements, wherein 'faces' can be 'seen' or not depending more on the viewer and the conditions than on the 'aesthetic' intention. (Here we can be more certain than anywhere else about the intention, which really is, ostensibly, to represent a vehicle. What kind of vehicle? Whatever kind is traveling on that road at that moment. Whether there is an explicit intention to represent (or not to represent!) a face or a creature is a wry and complex question. Could our propensity to anthropomorphize actually come to bear on the functionality of roads rather than just on their aesthetics? Do drivers more (or less!) often obey signage that really does seem to be 'watching' them? These are more the type of questions that impressionistic representations in art force us into broaching. For my purposes here, road signs do not force such questions, but perhaps they must be asked eventually, all the same. I'm highly male and somewhat on the ol' spectrum and I myself quite often 'see' faces in these representations; ergo I too must ask these questions at least in those cases.)
Anyway, the tractor is not the example Our Man should have chosen to make his chosen point above, because it actually makes precisely the opposite point. Humans readily project human qualities upon tractors, as upon so many 'inanimate' or 'nonhuman' objects which, nonetheless, manage to strongly evince some distinctively "human quality." This is why nothing "logical" or "analytic" can be built up from the fundamental presupposition of a hard line between human and nonhuman qualities, at least not in the realm of aesthetics. Perhaps if we mount a rigorous 'scientific' and 'empirical' inquiry as to what really is 'human' and what is not, perhaps then we can draw some such distinction and actually be confident that any important project we might later stake on its truth is a project built on strong foundations. But we will not be able to apply that understanding to aesthetics, nor to semantics, nor to semiotics, because those domains are epistemic spirals from which no closed circle can be made. In other words, we end up at absurd conclusions, such as above: tractorness is not something that can be expressed because it is not an "anthropomorphic quality." But try telling this to a small child. (Say, to me as a small child, playing at Machinery Hill at the MN State Fair. Yes, it was also ME!) We're supposed to outgrow this, but really we just learn to repress it in all but those exceptional situations where it is socially acceptable. (God forbid it should ever be required, though that seems to be precisely what a good number of try-hard art-philosophers have tried very hard to establish.)
This thesis of anthropomorphic drift is not "logical," presumably for Our Man, because it does not follow "logically" from the stipulation that expression deals only in "human qualities." But the whole point of the chapter is to try to say why we so often think we detect those qualities in objects which we also detect, at once, could not possibly (not really) possess them. This is a totally incoherent project, though, because the moment we succeed in establishing, with rigor, what it means to 'really' possess such properties, we have also succeeded in saying, at once, not only that there is no "logic" in expression but also that there is a very good reason why we are expected to outgrow the mistaken belief that there is such logic; there is a reason why anthropomorphism could be, all at once, the lifeblood of many artistic lineages and the root of so much evil.
... the theory does not provide sufficient conditions for expressiveness, because it will count as expressive too many things that are not truly expressive.
It may be thought that this lacuna can be repaired by
placing certain constraints
on
the kinds of schemes
that can be mobilized ... Perhaps, the theorist ... will say that only schemes involving anthropomorphic
[95]
properties are permissible. But there will be
problems here too,
since certain anthropomorphic properties
are not the sorts of things
that can be expressed. Imagine a scheme of
human skin diseases
being projected on an array of artworks with the result that some artwork, perhaps a horror film, correlates with impetiginousness. Surely, that is
not an expressive quality that an artwork can have.
Is this entire book just an exercise in setting up and tearing down strawmen?
To attempt to meet the question of those anthropomorphic properties which are not the sorts of things that can be expressed . . .
Why not? Why can they not be expressed?
The given example says-without-saying:
because they are
too concrete,
not because they are too diffuse.
Perhaps we can be just as certain that someone is angry as that they are impetiginous. Still, these are parallel but not twin certainties; they are not determined in the same way; we do not say quite the same thing when we say one as when we say the other.
So, a certain kind of theorist denies that the quality of impetiginousness, e.g., can be "expressed" while affirming that the quality of anger can be "expressed." But as with pretty much the entire book so far, this seems merely to have been stipulated for the purpose of channeling the discussion in a certain (not-so-interesting) direction. This aspect of the concept: "expression" has not been (sufficiently) analysed. No such analysis has yet been offered which would support the position ascribed here to the theorist(s). But maybe, just maybe we've backed our way into an analysis here all the same: it seems that when we look at a lesion on someone's skin we are not doing remotely the same thing as when we look at the expression on their face.
The problem for the so-over-it anti-Kantians, the so-over-it counter-Bulloughians, and for any other so-over-it-ian-ists, is that both impetigo and anger may mean something. Impetigo and anger alike can mean nothing only where no interest is taken in them. What's missing here from our nonanalysis of "expression," then, first and foremost, is an account of interest taken. The closest we ever seem to get to that account is such observations as:
We can reflect upon the sadness exemplified by an artwork without becoming sad. The artwork can inform us about sadness
...
without arousing sadness in us.
(p. 89)
If we are not simply entrained tout court into actually feeling the sadness, even where we find the sadness unambiguously exemplified and intersubjectively verifiable, then it must also be possible for us (for someone else) to overlook the sadness entirely. By "overlook" I don't mean to suggest willful intention nor to rule it out; I mean only that this seems like a real possibility if the simple mirroring of expressed emotion cannot be assumed. The question of what it means for music to "possess" or "exemplify" an emotion, then, is no longer the really interesting and important question. Rather: Can the paradigmatic "sad music" veritably force itself on the listener?
"Art is seduction, not rape."
But maybe these juridical terms are profoundly misleading as analytical terms. Maybe there is always some of both, analytically speaking, even if juridically and ethically it is imperative that we round up or down, if we want to preserve the merest pragmatic possibility of human association. Maybe juridical "logic" is the imposition of logic after some fact of 'stipulation,' but the stipulation itself is not (can never be or be called) "logical." If we dare to analyze the stipulation itself, we may suddenly find ourselves flying too close to the sun; the older the stipulation and the more venerable its history, the closer and so much the hotter for us.
The missing concepts here might be called susceptibility and (following Bullough) interestedness. The first threatens to entangle us in fruitless excursus and counterexcursus of the purported cognitive bases of the emotions. For now, suffice it to say: we expect to find some variation in the domain of "susceptibility," but not nearly as much variation as we find in the domain of "interestedness"; and the reason, I think, is obvious and minimally controversial: the "basis" of "interestedness" is not, properly speaking, cognitive but rather semantic. It belongs to the literally boundless spiral expanse of imagination and compositionality. Empirical study of cognition generally and emotion specifically certainly has not been easy; it has not proceeded gently or evenly; it has not yet succeeded in closing its self-identified epistemic circle; but it is imaginable, at least, that this circle may be closed, and there are some good leads on how this might be done. Conversely, the further we pursue empirical sematics, the further we tumble down the abyss.
To the extent that "interest" is a function of "semantics," to that same extent have theorists of (dis)interest and "aesthetic distance" been, in fact, on precisely the right track in attempting to sort out why and how the various artifacts of "expression" have proven at once philosophically galling and popularly indispensible. This is the correct idenitification of the problem. Perhaps the only real problem with the "distance" metaphor is that it is a metaphor, and so it is therefore too easy to problematize in the manner of the present chapter; as if "distancing" per se was a trivial matter of literary interpretation rather than a matter of the gravest empirical importance. But "interest" and "disinterest" need only a small tweak: they require the specification of some theorized semantic content in relation to which they are operative in any given case. To reiterate what seems to me like the most obvious example: the race of a theater actor would, classically and conventionally, go unmarked, in theory if not in reality; the very concept of theater was thought (again, if only in theory) actually to require "disinterest" along this vector of content; whereas nowadays there are nearly as many varieties of "interest" here as there are actors and onlookers. The mistake is to think that we can systematize the semantics of actorly race, that there can be a science of reception that makes accurate predictions about semantics. Semantics are a spiral. To make a circle out of semantics is merely to particularlize the inquiry, that is, to flee "analysis" and to take refuge in "pragmatism"; and of course that may well be possible, it may even be what is needed, but its findings are as limited and artificial as its domain. There will be, then, no "analysis" of expression, and no universal science of expression; there will be, presumably, instances of expression, the content of which can, in theory at least, be specified in terms of yet other "concepts," ever multiplying in culture and ever plunging ambitious semantic "analysts" down one or the other spiral abyss.
As always before, it is incomprehensible to me why anyone would want to travel in that direction. The only viable explanation is that they do not think that this the direction they are traveling in.
This is an outright counterexample for any exemplification theorist who maintains that
the choice of a scheme
for mapping is
perfectly arbitrary.
But it is also a problem for the exemplification theorist who wants to
place constraints on
which schemes are
available
for mapping, ...
... these objections have primarily revealed limitations that can be traced back to
the particular account
of metaphor upon which the exemplification theorist usually relies. But perhaps that account can be reworked, ... [Or] Are there deeper objections?
...
Is expression always metaphorical?
The friend of metaphorical exemplification argues ... only sentient beings can be the proper bearers of mental properties, such as sadness. And obviously artworks are not sentient beings. So artworks are only describable as sad
[96]
metaphorically; ...
Stated more expansively, ... [six "premises" are given below]
...
... let us grant ... the first premise ... when we say an artwork
has an expressive property,
our attribution is
either literal
or metaphorical,
... Moreover, the fifth premise ["artworks (and parts of artworks) do possess expressive properties"] seems to be a matter of fact that every philosopher of art will accept. ...
Premise #3 claims that no artwork nor any parts thereof belong to the class of things that can bear mental properties. A statue cannot be sad, ... Stone is not sentient. ... The case may seem to be an open and shut one, but it is not,
YOU DON'T SAY?!
...
... [e.g.] why do we call the X-Files deadpan?
Isn't it because ... Mulder and Scully are deadpan? ... [certainly they] represent sentient beings. ... Thus,
parts of artworks,
notably
[97]
representations of characters,
are the sorts of things
to which mental property terms are
literally applicable.
So if it is required that parts of artworks be proper bearers of mental properties in order to bear expressive properties literally, then some parts of artworks may (and quite frequently do) meet that requirement.
Shaw 'nuff.
But the X-files is (are?) not a stone statue??
... entire artworks ... can be devoted to
the representation
of
a character and his mentality.
... to presenting
the content of
human situations that through group interaction manifest distinctively anthropomorphic properties like barbarity (call this
a social anthropomorphic property).
Thus,
the content of
entire artworks can be the kind of thing to which mental-property terms can be applied. So premise #3 looks false.
Content?! Who said anything about content?
We learned about the Fallacy of Equivocation back in Chapter 1, and now in Chapter 2 we are being asked to apply it.
Perhaps the attempt will be made to reject these counterexamples on the grounds that these are cases of fictional representations ... But this objection fails ... there can be expressive artworks that are nonfictions ... more importantly,
when we apply
mental-property terms like "deadpan"
to fictional characters,
our usage is
no less literal
than when we apply them to real people. ...
The distinction between
literal versus metaphorical
does not coincide
with the distinction between
fictional versus nonfictional.
... So the objection against premise #3 stands.
... Perhaps the exemplification theorist will respond that the relevant representational contents of these artworks are only parts of artworks. ... [but this] would not be much of a victory ... since generally when we attribute an expressive property to an artwork, we are directing attention only to parts of the work, not every inch of it.
[98]
And
even when
we use an expressive term to
summarize an entire work,
we are generally
referring only to
its dominant or most salient property,
... We call the X-Files deadpan because this property in large measure attaches to its major characters and, since
they dominate
the story,
" . . . and similarly, wherever some part of some carved stone veritably dominates the object-work into which it has been made by the sculptor, so it is that the entire work may be summarized in a single attribution of expressiveness . . . "
" . . . (leaving aside the possibility that it is the viewer rather than the sculpture which has been "dominated" here . . . ")
...
... the exemplification theorist will object that these sorts of examples confuse representation with expression. ... But King Lear is not only
represented as
a man who is mad; he is portrayed as
behaving
... in a mad manner. ... If the same behavior expresses madness in everyday life, then so does Lear's. ...
... the representation of characters and groups are not the only loci of expressivity that we need to consider ... Artworks also contain
points of view
where
the possession of
a point of view presupposes the kinds of entities that can
possess mental states literally.
... many artworks have
viewpoints that diverge
from
the viewpoints of the characters
in its story.
In the beginning of the film Pulp Fiction, the viewpoint of the young men who are about to be shot is that of high anxiety, but the film regards their plight as amusing. ... this leads us to attribute the quality of irony to the film. ... when we apply ... ironic, detached, and cool to the point of view of Pulp Fiction, we are doing it in the same literal way that we might describe the attitude of a friend.
Point-of-view talk ... assumes that there is someone whose point of view it is. ...
points of view
belong to or are attached to
persons.
... the ironic point of view may belong to the director of the film, ... [Or] it may turn out that ... the point of view of the film is not actually his— ...
[99]
... Nevertheless, the point of view still belongs to the right kind of entity, since ... there is a character, the implied author or narrative persona, whose point of view he projects, ...
... The implied author or narrative persona is also a fictional character— ... there is no problem in saying that fictional characters bear their mental properties literally ...
... artworks with points of view are the kinds of things that can bear mental properties. ... artworks and parts of artworks can be the bearers of mental properties in virtue of their points of view. ...
Lyric poems are an excellent example ... a picture of the speaker's emotional state emerges— ... It is not
simply stated.
Rather,
its ingredients—
the desires, beliefs, intentions, perceptions and values—that give rise to it
are presented
so that the properties of the relevant feeling
become available
to the reader
for reflection.
... it is generally the speaker's address that comprises the poem. But the speaker ... is a suitable bearer of mental properties. ...
... time to turn to the second premise—that if artworks possess expressive properties literally, then they must be the kinds of things that can bear mental properties. ...
[100]
... "x possesses expressive qualities literally only if x possesses mental states."
... the face of a St. Bernard dog. ... is expressive of sadness. ... presumably St. Bernards possess mental states. But ... Its face is sad, independently of its mental state.
... The face simply looks sad to us. We do not mean to make a metaphor ...
... we call such configurations sad (or sad-looking) because similar facial configurations in humans are
associated with
psychological states of sadness... are
characteristic of
sadness. ... sometimes we attribute expressive properties to things because of their configuration ... irrespective of ... any mental states ... In fact, we attribute expressive properties to things on the basis of their configuration, even when they are not animate objects.
...
... when we call the weeping willow tree sad (sad-looking), we are offering a literal description of its perceptible configuration. ...
[101]
Human beings
manifest
certain outward, physiognomic features
that are
typically expressive
of their psychological states ... nonhuman things, like dogs and trees, can
remind us of
these features. Perhaps we
are prone to see
human physiognomic features in nature so readily because
detecting
the
emotional states of others
is
so important to our survival.
... we do tend to see features of nonhuman and even nonsentient objects as exhibiting the physiognomy of human feeling states. That is
just how they look
to us. Literally.
... if we call the tree anguished, we do not
literally mean
that it is suffering. But that does not entail that we are not
speaking literally.
For we are not saying that the tree
is suffering
... rather that it
is anguished-looking—
... that attribution is literal; ... however psychologists
finally explain why
it strikes us so.
This is very nearly a full circle back to ye ol' imitation theory, no? But now it has become a matter of what we ourselves are prone to see and not simply a matter of what has been put there for us to see.
... sometimes expressive-property terms are ascribed literally to the configuration or appearance of things that do not bear mental properties. ... this opens up the possibility that sometimes we attribute expressive properties in this way to the configurations of artworks ...
... the defender of premise #2 is likely to say that ... we are still trading in metaphors ... But ... even if "weeping willow" was a live metaphor once upon a time ... , it is now a dead metaphor, and
dead metaphors
are
metaphors that have become literal.
... "hands of a clock," ... if it
ever was
a metaphor,
is no longer.
Now it serves to describe literally certain features of clocks. ...
[102]
... when we speak of a furious storm or of nasty weather, though we do not mean that the climate is literally angry or mean-spirited, we are not speaking metaphorically. Rather we are saying that they remind us of the characteristic observable behavior of angry and mean-spirited people.
This can't be quite right, simply because the very death of a metaphor means, I presume given the above, that we are not reminded of anything beyond that which we are remarking upon; rather, the saying has become habitual or 'formulaic.' It is more that we are no longer "reminded" but instead jump straight to the most readily available concept, via the shortest possible path of retrieval. If there ever were people who were truly "reminded" of one thing by another, those people have long since passed on.
...
...
... A great deal of orchestral music is described in terms of expressive properties. ... Perhaps that is like our description of the St. Bernard's frown—a literal description of a configuration.
Music often sounds ... in ways that remind us, almost automatically, of characteristic human feeling states, and we use the relevant feeling terminology to describe literally the configurations that we hear. ...
...
There is, then, a
configurational use
of expressive terminology that
applies literally
to the appearance of inanimate objects, including art objects, which objects do not possess psychological states. ...
[103]
...
... Sad music may be low and slow because that is how we feel when we are sad, or the way sad people usually sound. In this manner, sad music may serve to
exhibit,
manifest, and
bring to our attention
distinctive human qualities.
Good for it, I suppose. But if these really are human qualities then what can be the point of this? What is this but yet another 'duplicate' world, a 'shadow world of meanings' that is easy to get into and difficult to get out of? Why reproduce something already present?
... the argument that expressive properties must be attributed to artworks metaphorically fails to go through ...
... [That argument] seems unlikely since in many cases,
the basic way
that we have
to refer to
certain of the configurational properties of artworks ... is to use anthropomorphic terminology. ... there is no
better, more straightforward
way of
talking about
the artworks in question.
As always . . . less said, the better!
I have just yesterday, finally, begun reading Beardsley's magnum opus. I almost stopped at the very first sentence:
"There would be no problems of aesthetics, in the sense in which I propose to mark out this field of study, if no one ever talked about works of art."
I'm now on p. 56, which means I have 535 pages yet to go. And I'm a lot less excited about this than I was yesterday, because already he seems to have solved all of his self-created problems and confirmed me in what I already believed.
Thus, ascribing anthropomorphic terms to such artworks should not be misunderstood as a matter of an
optional,
ornamental,
metaphorical
description. Rather, it is literal.
EVERYTHING is optional! (Sartre, et al.)
Nothing is stopping us from concluding that these ascriptions are both "literal" and "optional."
For Our Man the problem seems more to be that this conclusion does stop us from calling such ascriptions . . . 'mandatory.'
... in the course of refuting premises #2 and #3 ... we have learnt quite a lot about how we go about saying justifiably that artworks are expressive of some human quality.
... Sometimes we make such attributions on the basis of characters ... Sometimes the attributions are made in virtue of the points of view artworks exhibit. ...
[104]
... But we can also justifiably attribute expressive properties to nonrepresentational artworks ... by virtue of their configurational features.
... attributing expressive properties to artworks is not always metaphorical. But perhaps sometimes it is; ... there is a plurality of ways that we go about attributing expressive properties to artworks.
... certain of these ways seem to figure more prominently with regard to
some artforms
rather than
others.
...
Of course, this is not an absolute difference, but a proportionate difference. ...
Chapter summary
... Art
presents the world
to us
replete with expressive properties.
As if that were not how it actually appears to us?
As if this difference was not proportionate but indeed absolute?
...
[105]
... Art makes the world
emotionally accessible,
often
in an immensely perspicuous way,
showing us
things
with their human qualities manifest or foregrounded.
Metaphorically speaking, art
humanizes the world
for us
He meant to say "anthropomorphizes," right? 😈
... Undeniably, this is one of the great attractions of art for us.
As is the junk itself the great attraction of junk food.
Moreover, this is reflected in the way we talk about art—
'There would be no problems of aesthetics, in the sense in which I propose to mark out this field of study, if no one ever talked about works of art.'
...
Theoretical awareness of the importance of feeling in the arts became especially pronounced at the end of the eighteenth century, ... This influence resulted in
a notable subjective turn
in artistic practice ...
Subjective?!
All this business about literal configuration sounds quite self-consciuously ob-jective?
...
... expression theories of art fail as universal theories of all art. Nevertheless, ... since expression is so central a feature of art, we still need to address it theoretically. ...
... we have seen that there is not one, but rather a number of ways in which we ascribe expressive terms like sadness and joy to artworks, and, furthermore, that several of them are literal, not metaphorical. ...
[106]
Annotated reading
...
No comments:
Post a Comment