Art and Expression
is an outline of Carroll, Philosophy of Art, Chapter 2
My personal comments are in red, like this, and you can see that they
have a different style of lettering.. They elaborate on Carroll or react
The expression theory of art (an attempt to define art)
Representation theories and expression theories agree that art communicates, but they disagree about what is communicated. The
expression theory emphasizes emotions, not ideas or thoughts. Advantage: an idea or thought must be about something, but some
emotions do not. So artworks that lack denotation can still be art.
Two types of expression theory: arousal and cognitive (non-arousal conveyance)
We will concentrate on Tolstoy's
version of the arousal theory.
x is a work of art if and only if x is
(1) an intended (2) transmission to an audience (3) of the self-same (type identical) (4) individualized (5)
feeling state (emotion) (6) that the artist experienced (himself/herself) (7) and clarified (8) by means of lines, shapes, colors, sounds,
action, and/or words.
(1) rules out cases where one's ordinary behavior conveys a sympathetic response in others, e.g., your loss of a job makes you
behave in a way that makes the observer feel sad. But sympathetic response doesn't
make something into art. The action must be intended (performed (?) intentionally) to move an audience.
(2) builds in the assumption that art communicates. Combine #2 with #5
and you have the core of the arousal theory.
(3) puts a restraint on what counts for successful communication (an identity condition).
(4) is an originality requirement; it rules out generic expressions, such as mass-produced greeting cards.
(6) is the experience condition. Add it to condition #3 and we have a
(7) rules out mere venting/letting off steam. It is the clarity
(8) restricts art to expressions that arise in a publicly-accessible medium, suggesting that some skill must be acquired by artists for
exploring public media.
(7) is central to Collingwood's
theory, but it only receives a very brief mention in Chapter Fifteen
of Tolstoy's What Is Art?, where it seems part of his ideal of
universal-accessibility. To actually have Tolstoy's theory, we would
replace "and clarified" with "and made universally
Carroll proposes that expression theories are more comprehensive than representational theories, so expression theories are a step
forward in theorizing about art.
Objections to the expression theory
We could reject condition #2 and have a SOLO expression theory: the artist is not interested in communicating with others, but only
wants to clarify her (Emily
Dickinson) or his (Franz
Kafka) emotions to herself/himself. And we want to allow this
"solo" expression to be art.
Response to the solo version: If the artist does not intend to communicate to others, why
"fix" the clarification in a publicly-accessible medium?
In principle, a work of art is capable of transmitting emotion to others. This variation allows
"lost" works and practice works to count as art.
Another reason to reject the solo version:
If the artist really wants to keep it private, she/he could have developed a private language (an idiolect) that would prevent others from
Reply: There is no such thing as a private language.
A better reason to reject the solo version: Each artist is her/his first audience. The artist has a dual perspective on the artwork, both as
creator and as critic. Otherwise, the artist would not know how to revise, or even stop working on, a work in progress. Therefore each work
of art really is intended for an audience, even if only an audience of one.
(Kant's views in
section 50 endorse this
last position. Without the critical perspective of taste, genius might produce original nonsense.
Genius individualizes, but taste "clips its wings" by insisting on clarity and order.)
response to the solo version: In Chapter 8 of What is Art?,
Tolstoy argues that the more we restrict the audience, the more we
trivial the artwork. To try to create "private" art is to
treat art as something of limited consequence. "Solo" art is
another species of counterfeit art.)
Problems with the experience condition
first challenge to the identity condition:
If the actor playing Iago in Shakespeare's Othello really felt
Iago's self-hatred, then the actor would not recall all the words, etc.
Therefore it will not often be of the self-same type.
Carroll takes the identity condition to require that the artist feels the emotion at the time of the creation of the artwork.
Romantics have a more complex view than this. For example, Wordsworth:
"I have said that poetry is the spontaneous overflow of powerful feelings; it takes its origin from emotion recollected in tranquility; the emotion is contemplated till, by a species of
reaction, the tranquility gradually disappears, and an emotion, kindred to that which was the subject of contemplation, is gradually produced, and does
itself actually exist in the mind.") A good portrayal of Iago will summon up, from the past, real feelings the actor has felt.
Carroll's second objection to the identity condition: Each art form develops standard formulae or conventions for
generating emotions, and an artist can communicate emotions by learning these
(Aristotle's Poetics analyzes these for ancient Greek
tragedies). Once we have rules in place, we don't need the artist
to feel anything.
Carroll's third objection:
a "cynic can make a moving artwork." And a lot of art is commissioned, and there is no reason to suppose
that every artist carrying out a commission feels the emotions she/he is paid to convey.
Reply to Carroll's second and third objections: Both Tolstoy and Collingwood are aware that this happens, but both argue that these
cases are not genuinely art. They are insincere and fake: like forgeries, they are not really art.
Problems with the experience condition
Carroll finally asks whether an artist must
"at some time or other" have had the emotion that is communicated. Carroll thinks not.
Psychopaths can manipulate people despite their lack of feelings. Can't there be psychopath artists?
(Tolstoy and Collingwood would answer that a "scam" is not a genuine work of
Problems with the clarity condition.
Some art is raw and unprocessed. Beat poetry, and punk art.
Reply: Punk art? Would that be punk music and fashion? Nothing was less spontaneous and more calculated than punk! If a punk band
plays a song that was written previously --whether a "cover"
version of someone else's song, as when the Sex Pistols covered the
Monkees' song "I'm Not Your Stepping Stone," or they perform
their own song for the 500th time, as when the Sex Pistols sang
"Anarchy in the U.K." at show after show-- it can hardly
count as raw and unprocessed. Furthermore, improvising (what the beats did) does not
demonstrate a lack of clarification.
different (and better) attack on the clarity condition: Surrealism and Symbolist art aim at elusive feelings. Clarification is avoided. Aleatoric
(chance) art intentionally adopts strategies to eliminate the clarification process.
Problems with the originality/individualized condition
The histories of
Christian or Hindu art show traditions full of very generic emotions.
(Hindu aesthetics specifically values generic emotions above individualized ones! The more generic, the more the experience leads to a
negation of the self.)
THE BOTTOM LINE: does all art express emotion?
No. A lot of recent conceptual art conveys ideas, but no emotions.
"It is cognitive, not emotive." Look at Escher, Warhol, etc.
Defenders of expression theory could reply that humans CANNOT HELP BUT express emotions when they communicate. Carroll's
reply: The accountant who adds up a column of figures and communicates the results can do so without conveying emotion (the result will
be the same if a second person, in a different emotional state, does the same thing).
Some pure/absolute music is beautiful and pleasing, without conveying emotion. The audience feels pleasure, but the music does not
express pleasure. It stimulates it.
Do we even need a public medium?
Couldn't it just be done "in" the artist's head? And then reported on? A lot of conceptual art seems to work this way. The
public object reports on the work of art rather than transmits it.
(Couldn't we say the same about a music score and a music performance?)
Carroll concludes that none of the necessary conditions are really necessary.
Worse, they are not even sufficient.
The nasty letter/speech examples satisfy all the requirements, but they are not works of art.
But Carroll does not engage with Tolstoy's response to this point: Carroll is talking about
"counterfeit art." For Tolstoy, lack of individuality gives us a counterfeit. Lack of clarity gives us a counterfeit. Lack of sincerity gives us a counterfeit. So generic Christian art is
no good. Symbolist art is no good (in fact, Tolstoy goes out of his way to attack Symbolism). Finally, insincere art is no good.
problem with AROUSAL theory
distinguishes expression from mere propaganda? Transmitting an emotion
to an audience does not distinguish art from propaganda. To deal with
this, Collingwood has a substitute condition for condition 2 as
discussed by Carroll. The emotion is not transmitted to the audience.
Instead, it is clarified for the audience. Instead of
"arousing" the emotion in the audience, the audience becomes
aware of their existing emotions.
How is it possible for art to be
Although some art is not expressive, a lot of
art is. However, there are puzzles concerning this. PEOPLE (and some other
animals) are expressive in the literal sense of "expressive,"
having emotional and other mental properties that they reveal to
But works of art are physical objects,
repeatable structures, and other artifacts.
The challenge: If something has no emotions, it cannot be expressive. So
works of art cannot be expressive.
challenge rests on two central claims: If artworks (and/or their parts)
possess expressive properties, they must be capable of possessing mental
properties, but these are not the kind of things that can bear mental
Two ways to respond to
- Deny that artworks literally express
anything: they only metaphorically do so.
- Fight the challenge directly by
attacking the two claims.
thesis says that artworks (and their parts) literally possess properties
which we describe metaphorically with expressive labels. (E.g., when it
is slow and in a minor key, we are likely to say that music is sad.)
notion of exemplification tells us that the artwork possesses,
metaphorically, some of the same properties as another thing. We say
that it is done metaphorically because it doesn't look as if artworks
can possess these properties literally (see the challenge
Problem: what system
allows us to consistently map our metaphors onto the literal properties?
Metaphor approach has no clear answer to this question.
the reasons to deny literal exemplification
apply expressive labels to a wide range of representations, both
fictional and non-fictional. We label both the characters represented
and the representations in which they appear. But we do not do so by
attributing these expressive properties to the person who actually
created the representation.
Fox Mulder is deadpan. The X-Files series is deadpan.
I can write a letter that expresses love, why can't I write a fictional
one with the same quality? And why can't I create a fictional character
who expresses the same mental property? If a real person can express it,
both a fictional character and a representation can do so.)
analysis rests on our ability to distinguish between the mental
properties of the author (often called the
historical author), the narrative persona (the point of view
adopted), and the characters. E.g., Mark Twain
(or Samuel Clemens, the historical author) wrote Huck Finn in the voice of Huck (the
narrative persona), a fictional character, and in the narrative, the
characters of Tom and Jim have display various emotions.
things have expressive properties literally, without having mental
The expressive persona of the sad-looking St. Bernard dog is different
from the actual feelings of the dog, yet the dog has the expressive
property of looking-sad without being sad. Obviously, this is due to the
dog's facial configuration. But other configurations might also
have a sad look, e.g., a willow tree. A storm can be furious, etc.
apply the terms anthropomorphically. Is this metaphorical in
Proposal: metaphors become dead metaphors, at which point they are
that Carroll's claim limits expressive properties to configurations or
structures. Is this true?