Chapter 2

Defining Art: Intension

and Extension
George Dickie
2. 1
Philosophical t heori zi ng arose, at least in t he West, in ancient Greek times. There
and t hen, t he t wo main compet i t ors for explaining t he order t hat we experience
were t he atomists and t he Platonists. The atomists explained this order as being
t he result of t he mi cro-st ruct ure of individual things - gold has a different mi cro-
st ruct ure from iron, tigers from lions, and so on. The atomists were t heori zi ng
about t he essence of things, and they did so by talking about spatio-temporal,
physical micro-structures. The Platonists also explained t he order that we experi-
ence in t he world, but they said t hat individual things fall into t he types that they
do because they participate in various Forms, these being non-spatial, non-
t emporal abstractions. The essence of t hi ngs resides, for t he Platonists, in t he
Forms. The atomists and t he Platonists differed about t he nat ure of t he essences
of things. They also differed about how t he essences are responsible for t he order
we experience: for t he atomists, causality
1
was responsible, whereas for t he Platon-
ists, it was participation.
2
Thus, at t he very begi nni ng of t heori zi ng t here was a
radical disagreement over what t he t heori zi ng was about - about what is real.
Once t he atomists had enunciated their thesis t hat t he micro-structures of
things differ from one another, they t hen had little else to say about t hose struc-
tures. They spoke of invisible micro-structures but had no means of knowi ng any-
t hi ng about these essences, al t hough they did speculate about t he movement and
wei ght of atoms. They were so far ahead of their times t hat they were speechless;
they were barred from inquiry into what they regarded as real by a lack of
t echnol ogy and developed theory. The Platonists, however, were able to speak
volumes, for in addition to t heori zi ng about t he order and essence of things, they
also focused on words and their meanings, which were available in great supply.
They had a philosophy of language, which t he atomists lacked. The atomists were
as speechless about words as they were about invisible micro-structures. The
45
George Dickie
Platonists spoke of non-visible things t oo, but they claimed we could know
t hem because we underst and t he words whose meanings were constituted by t he
non-visible Forms. The Forms were t hus alleged to be not only responsible for
t he order of things but also t he meanings of words. The knowi ng of a Form is
allegedly demonst rat ed when a word is underst ood. Compl et e knowi ng of a Form
is supposedly demonst rat ed when an adequat e definition of a word - "figure,"
"justice," or t he like - is successfully achieved.
The voluble Platonists, with a philosophy of language integrated with and
underwri t t en by their metaphysics, won an easy victory over t he speechless
atomists for control of subsequent philosophizing. Theori zi ng about t he essences
of things and t he definitions of words - including art and "art" - t hus had its
devel opment within t he Platonic vision of language and reality, with reality being
underst ood to be a hierarchically ordered, rational st ruct ure of non-spatial, non-
t emporal Forms that give order to t he world of experience and constitute t he
intensions of words in a language. The metaphysical st ruct ure is rational in t he
sense t hat it has t he form t hat woul d be given by a rational arranger, al t hough no
arranger is envisaged within t he system. The st ruct ure of t he Forms is taken to be
such t hat t here is a genus-species relation i nherent in every intension. Within this
vision, t he essences of all sorts of different things - gold, water, tigers, justice, art,
whatever - are taken to be of t he same sort and t he essences themselves are subject
to dialectical analysis that can yield t he intensions - necessary and sufficient con-
ditions - of t he words t hat apply to t he things with t he essences. For t he Platon-
ists, inquiry about what is real was conceived of as a search i nt o t he intelligible
realm of t he Forms for essences and meani ngs or intensions; empirical inquiry was
regarded as a pursuit into t he illusion of sense. The intellectual agenda was t hus
set by t he Platonists, and philosophical inquiry became an at t empt to produce t he
essences of things and t he intensions of words. Philosophical inquiry continues to
have somet hi ng of a Platonic flavor today.
An apparent advantage t he Platonists' approach has over t he atomists' is t hat
while in principle t he atomists have a way of explaining physical phenomena, t here
is no obvious way for t hem to explain how non-physical characteristics (linguistic,
moral, cultural, and t he like) arise from atomic configurations. On t he ot her hand,
t he Platonists' metaphysics and philosophy of language supposedly deal with all
kinds of phenomena from physical to moral.
As is well known, in recent times t he search for t he essences of art and ot her
not i ons has been challenged by t he claim t hat t here are no such essences, and t hat
"art" and ot her words apply to t he t hi ngs they do in virtue of intensions that take
not e of t he overlapping similarities among those things. This is a challenge to
whatever is left of t he Platonic tradition in philosophy. Ther e are t hree main dif-
ficulties with this untraditional development. First, it is unclear how t he similar-
ities are to be specified - how similar do t wo characteristics have to be to count
as linking t wo works of art under t he word "art" (or any t wo objects under t he
same word), and how many such similarities are required to make t wo objects fall
under t he word "art" (or any word)? Second, t he reliance on similarities threatens
Defining Art: Int ensi on and Extension
to draw i nt o and collect into t he class of art, or any ot her not i on specified in t erms
of similarities, every object in t he universe because, in some way, everything resem-
bles everything else. Thi rd, focusing now only on art, if one tries to contain t he
collecting tendencies of similarities by specifying t hat t he similarities must be to
prior-established works of art, t hen an infinite regress of prior-established works
of art is generated so that t here could never have been a first work of art and hence
no present works of art. Since t here are works of art, t he similarity conception as
t he whole story is wrong, and t here woul d have to be some non-similarity ur-work
or ur-works of art t hat have priority over "similarity" art. Thus, some kind of non-
similarity foundation woul d be required by t he similarity view.
The atomists, t he Platonists, and t he similarity theorists all begin by focusing
on t he order t hat they not e in things: t he Platonists infer Forms to explain iden-
tities in experience and to be t he intensions of all kinds of words, which they take
to be manifested in genus-species relationships; t he atomists infer conclusions
about micro-structures to explain identities of physical phenomena in experience
but have not hi ng to say about intensions; and t he similarity theorists forego
inference and focus solely on experienced similarities, out of which they try to
construct intensions. I shall not discuss t he similarity theories further.
2. 2
Recently, some philosophers of language have tried to work out a way to adapt
t he insight of t he t ongue-t i ed Greek atomists about t he essences of things to t he
probl em of t he application of words to things. One of t he t hi ngs t hat has enabled
these philosophers to try to do what t he atomists could not try to do is t hat t here
are now well-worked-out and accepted theories of t he mi cro-st ruct ure of things
t hat were not available to t he atomists. In general, this new t echni que, according
to t hose who advance it, is said to approach t he question of t he application of
words to things t hr ough their extensions rather t han t hr ough i nt ensi ons.
3
This new
approach supposedly contrasts sharply with t he traditional, Platonic-tinged
approach to meani ng. One begins, using t he new approach, with descriptions of
features t hat function mor e or less like intensions which serve to focus on some
gr oup of things (an extension), and t hen in t he cases of natural kinds one dis-
covers or proceeds with t he assurances t hat an essential, underlying propert y of
t he members of t he gr oup of things (the extension) can be discovered t hat
uniquely picks out t he gr oup of things, or some significant subset of t he gr oup of
things. This underlying property, if discovered or discoverable, is what identifies
this kind of t hi ng in all possible worlds in which t hat kind of t hi ng exists. Some
philosophers of art have subsequently tried to apply this new t echni que to t he
philosophy of art.
The approach of Platonists is a top-down approach; for t hem, t he Forms, which
function as intensions, are given and complete, so that, for t hem, intensions come
George Dickie
first as ready made and det ermi ne extensions, t he members of which are mere
appearances. The new philosophers of language, on t he ot her hand, use a bottom-
up approach; for t hem, t he discovered or discoverable essential propert y of t he
members of an extension constitutes its nat ure.
These philosophers of language begin, not with a discussion of natural kinds,
but with a discussion of proper names as rigid designators. According to this view,
a proper name such as "Aristotle" is a rigid designator, a rigid designator being
somet hi ng t hat picks out t he same object in all possible worlds where it exists.
Also, according to this view, proper names are i nt roduced by a baptism or dubbi ng,
and we track their referents t hr ough t he world by means of causal historical chains.
Proper names, on this view, function by means of reference or extension rather
t han by intension. We in later generations have come to believe various proposi-
tions about Aristotle, and earlier philosophers of language tried to use these beliefs
in one way or anot her as intensions to pick out Aristotle in all possible worlds.
Almost all of our beliefs about Aristotle, however, could be false, and t he rigid-
designator approach avoids using t hem in any way.
These philosophers of language t hen go on to apply t he rigid-designator
approach to words for natural kinds. In t he case of an element such as gold,
various properties such as being yellow, very malleable, and so on served to focus
on a gr oup of objects (an extension); it was later discovered t hat all or many (a
significant subset) of t he objects have a particular atomic number which t hen
henceforth serves as t he essential, underlying property t hat picks out t he gr oup of
gold things. Gol d' s essential property is being t he element with t he atomic number
79, and this means t hat gold is identical with t he element with t he atomic number
79 and that gold is necessarily t he element with t he atomic number 79. In t he
case of a compound such as water, t he essential, underlying propert y t urns out to
be a particular molecular combi nat i on of elements, namely, H
2
0 ; thus, water is
necessarily H
2
0 . In t he case of a species of plants or animals, t he essential, under-
lying propert y woul d perhaps t urn out to be somet hi ng like a particular DNA
profile or whatever t he correct underlying propert y is for species. These underly-
ing properties serve to identify, for example, gold and water in all possible worlds
in which t here are such substances. In t he cases of elements, compounds and
species, t he essential properties are underlying because they are micro-structures.
The discovery of t he essences of natural kinds - "gold, " "water," and so on - is
approached t hr ough extensions. These philosophers of language have advanced
t he insight of t he Greek atomists to a remarkable degree.
2. 3
James Carney (1975, 1982) at t empt ed to take t he insight of t he philosophers of
language even further by applying t he rigid-designator approach to t he probl em
of what has been characterized in t he past as t he defining of "art." For this appli-
48
Defining Art: Int ensi on and Extension
cation to be possible, according to Carney, it must be t he case t hat a paradigm set
of objects had been dubbed "art" and t hat t he dubbers believed t hat t he objects
share a universal property t hat is a nature, j ust as all pieces of gold share t he nature
of having t he atomic number 79. And, for Carney, j ust as gold is necessarily t he
element with t he atomic number 79, art woul d necessarily be whatever had t he
universal property. Carney suggests t hat it is counterintuitive not to have t he belief
t hat works of art have such an extension-determining nat ure. Carney t hen says:
It is not unreasonable to suppose that what Danto, Dickie, and others have called
the "artworld" is the subclass in a society that determines the universal property
and that they rely on theories of art to do this. The artworld would be analogous to
the metallurgists for "gold," and art theories would play a role similar to scientific
theories and "gold" in that they would be taken as hypothesizing the extension-
determining property of art. (1975: 200)
A few lines later Carney writes, "Theories such as t he imitation t heory or expres-
sion t heory woul d be adequate, since they hypothesize a universal property for
paradigms" (1975: 201) .
What Carney is saying is t hat any art t heory t hat claims t hat artworks share an
essential property, which of course is every traditional art t heory except t he simi-
larity view, is a candidate for being fitted to t he rigid-designator approach. Wi t h
all t he historical theories of art t hat have been put forth, Carney' s final say on
t he mat t er has to be hypothetical: "If t he paradigms [of art] have a universal
property, t hen t here is a way to det ermi ne with certainty whet her x is art: x is
art if x has t he universal propert y" (1975: 200) .
This hypothetical resolution leaves open t he possibility t hat different universal
properties mi ght be det ermi ned by different members of t he artworld, assuming
for t he moment an underst andi ng of t he artworld as Carney is envisaging it. On
Carney' s view, this apparent difficulty is resolved because of t he nat ure of t he only
t wo possibilities. If a disagreement arises among t he members of t he artworld over
t he common nature of artworks, t hen t he members may decide t hat all t he old
paradigms of art do not share a single underlying nat ure and it will t urn out t hat
t he old paradigms separate into t wo or mor e extensions, each with its underlying
nature, and t here will be t wo or more kinds of art. Or, on t he ot her hand, if a
disagreement arises within t he artworld over t he common nature of artworks, say,
over whet her a new kind of t hi ng is art, with one side citing one nat ure and anot her
side citing anot her nature as art -det ermi ni ng, t he t wo sides either accept t he t wo-
or-more-kinds-of-art solution j ust discussed or can agree on one nature and t here
will be only one kind of art. So, t he members of t he artworld will cither disagree
and t here will be more t han one kind of art, or they will agree and t here will be
one kind of art.
Carney' s view raises t hree questions. First, can art theories play t he kind of role
t hat scientific theories play in connect i on with t he rigid-designator approach to
natural kinds? Carney' s answer is "Yes, if they assert a claim of a universal prop-
George Dickie
erty for t he paradigms. " This answer raises a second question: "Which art t heory
is analogous to t he atomic t heory t hat yields t he atomic number 79 for gold, t he
molecular t heory t hat describes t he molecular st ruct ure of water, or t he biologi-
cal t heory t hat specifies DNA profiles?" Hi s answer is that it is t he art t heory (or
theories) with t he universal propert y (or properties) det ermi ned by t he members
of t he artworld. This answer raises a third question: "Does t he artworld function
as Carney envisions?" He does not answer t he third question ot her than to say,
"it is not unreasonable to suppose" t hat it does (1975: 200) .
Peter Kivy was t he first to attack Carney' s suggested approach to t heori zi ng
about art (1979). Kivy does not comment on Carney' s claim about actions of t he
members of t he artworld, focusing solely on his claim of close analogy between
art theories and scientific theories. Where Carney sees analogy, Kivy sees disanal-
ogy We are not prepared, Kivy says, to accept an art t heory in t he way t hat we
"are prepared to accept a scientific account of t he internal st ruct ure of a natural
kind" (1979: 430). He agrees t hat in t he scientific domai n t here is a history of
discovering t hat earlier theories are false and of their being replaced by new
theories, and t hat this bears some resemblance to art theories being replaced by
later art theories. But, he says, t he succession of scientific theories is different in
t hat it reveals an increasing scope and ability to deal with t he data. Such success
in t he scientific domai n inspires a confidence t hat is not found in t heori zi ng about
art by philosophers.
Thomas Leddy was t he next to attack Carney' s view (1987). He appears to
accept Kivy's poi nt about t he disanalogy between art theories and scientific t heo-
ries, but he focuses on a logically prior analogy alleged in Carney' s view - his con-
t ent i on t hat t he artworld determines t he universal property of art analogously to
t he way metallurgists det ermi ne t he nat ure of gold. Carney begins by talking about
artworld members "hypothesizing" about t he universal property of art. Leddy
notes, however, t hat it appears to t urn out t hat on Carney' s view t he members of
t he artworld supposedly determine t he nature of art by deciding on a universal
property. This contrasts sharply with how metallurgists determine t he nat ure of
gold: they discoverthe universal propert y of gold (1987: 264) . Thus, Leddy uncov-
ers anot her difference where Carney' s view requires similarity.
What of Carney' s claim, which neither Kivy nor Leddy addresses, t hat t he
members of t he artworld function to det ermi ne (even if it is only by deciding) t he
nature of art? Carney says t hat t he members of what Dant o, I, and others have
called t he artworld det ermi ne t he universal propert y of art. This makes t he art-
world into somet hi ng like a legislative body t hat deliberates and issues directives
t hat are bi ndi ng on t he ot her members of society. First, it is not indicated who
t he "ot hers" are, but t he accounts t hat Dant o and I have given of t he artworld
are very different, al t hough this was perhaps not so clear in t he mi d-1970s, when
Carney published his view, as it is now. Carney writes t hat Dant o' s st at ement t hat
"It is t he t heory that takes it [Brillo Box] up into t he artworld" means that we can
take his st at ement to imply "that t he extension of t he t erm ' art' is det ermi ned by
t he theories of art held by t he artworld" (1975: 201) . Perhaps Dant o' s account
50
Defining Art: Int ensi on and Extension
in "The Artworld" can be const rued to fit t he rigid-designator approach as Carney
says, but I do not think this is Dant o' s view, and I, for one, do not think t hat t he
artworld functions as a legislative body.
Richard Wollheim has attributed to me t he view of t he artworld as legislative
body and t hen has gone on to ridicule t he view (1987: 14- 15) . This underst andi ng
of t he institutional t heory of art deserves to be ridiculed because t here is no reason
at all to think t hat t he artworld or any aspect of it acts like a legislative body -
with meetings and decisions, and with declarations and proclamations. Fort u-
nately, t he view of t he artworld t hat Wollheim attributes to me is not one t hat I
have ever held (Dickie 1993: 69- 71) , al t hough he and a number of ot her people
seem to have t hought I did. I have always underst ood t he artworld to be a back-
gr ound for t he practice of creating and experiencing art - a background t hat is an
essential part of t he practice.
So, t here seem to be t wo strikes against Carney' s view - Kivy's and Leddy' s -
and perhaps a third against his underst andi ng of t he nat ure of t he artworld.
I do think t he rigid-designator approach can be fitted to t he imitation t heory
of art and to some versions of t he expression t heory of art, but not because these
theories mi ght be held by members of t he artworld. The t wo theories can be fitted
to t he approach because they are what I have elsewhere called "natural-kind t he-
ories" (1997a: 25- 8) . The philosophers who were t he proponent s of these t wo
theories were at t empt i ng to identify art with one, single, particular kind of human
activity - imitating or t he expression of emot i on - t hat can quite reasonably be
regarded as natural or what today woul d be t hought of as hard-wired. Not e t he
parallel here with gold, water, and species. In t he case of gold and t he ot her ele-
ment s, it t urned out that t here is a fairly small number (something over a hundred)
elements, and physicists discovered t hat each one has one distinct atomic number
t hat uniquely picks it out . (Apparently, isotopes can be ignored. ) In t he case of
water and t he ot her compounds, it t urned out t hat t here is a very large number
of compounds, but even so physicists and chemists discovered t hat each one has
one distinct molecular configuration t hat uniquely picks it out . The case is perhaps
similar for species. It is quite reasonable to suppose that some day t he now-hi dden,
underlying nat ure of t he behavior of imitating or expressing of emot i on will be
discovered by a scientist - some breed of psychologist/biologist. These t wo ways
of acting woul d be natural-kind behaviors of natural-kind beings. I am speaking
here of imitating and expressing as such; how such imitating and expressing are
st ruct ured and directed and t oward what may vary from culture to culture. By t he
way, these behaviors are not limited to human beings.
Unfort unat el y for t he imitation and expression theories, t here is no good reason
to think t hat either of these behaviors is identical with art - some works of art are
not imitations or expressions of emot i on and some imitations and expressions of
emot i on are not art - which is why t he theories have been almost universally
rejected. So, al t hough t he imitation t heory and t he expression t heory are t he sort
of theories t hat could be fitted to t he rigid-designator approach, because t here is
reason to think t hat imitation and t he expression of emot i on have underlying
51
George Dickie
essences, it woul d be a mistake to try to do so because t hose behaviors j ust do
not match up with all our artworks. They are t he wr ong theories to use.
2. 4
There is perhaps a way to fit certain aspects of t he rigid-designator approach to
art, namely, to approach t hr ough an extension, looking for an underlying prop-
erty, al t hough not one t hat functions in all t he ways t hat a property such as being
H
2
0 does. Consider Carney' s procedure. His account of t he application of t he
rigid-designator approach to natural kinds can be summarized by picturing t he
following pairs: gold./physicists, water/physicists and chemists, and species/molecular
biologists. Carney t hen tries to use t he rigid-designator approach with t he specifi-
cation of t he essence of art, pairing art and members of the artworld. In effect, Kivy
and Leddy in different ways poi nt out t hat Carney' s pairing is not analogous to
t he earlier pairs which all involve scientists. In order to extend t hose aspects of t he
rigid-designator approach t hat I wish to use, t he second place in t he art/ pair
should be filled with t he name of some kind of scientist. In t he above discussion
of t he fitting of t he imitation and expression theories to t he rigid-designator
approach, t he second place in t he pair was filled by psychologists/biologists, who
presumably woul d focus on t he behavior specified by t he t wo old theories of art.
But what I am envisaging here is scientists who woul d focus directly on t he art of
our culture or ot her cultures.
Before trying to apply t he aspects of t he rigid-designator approach I wish to
use to t he complicated not i on of art, consider how it mi ght go with a simpler cul-
tural concept. Assume an ant hropol ogi st goes to work, say, in t he 1920s on a par-
ticular sout h Pacific island culture. In landing on t he island, t he ant hropol ogi st ' s
native translator is drowned, so she must carry out her studies wi t hout t he benefit
of access to t he islanders' language. One of her observations is t hat many people
are referred to as pukas but not all. She t hen observes t hat only males are pukas,
al t hough some of t hem are fat and some skinny, some are short and some tall, and
so on. So what is a puka? In her further observing of t he social st ruct ure of t he
islanders, our ant hropol ogi st discovers t hat t he teenage boys and girls regularly
engage in promiscuous sexual behavior wi t hout anyone disapproving, but t hat
about age 16 on t he day of t he summer solstice they are compelled to st op this
behavior by everyone in t he society. The 16-year-olds t hen have t he opt i on of par-
ticipating in an elaborate ceremony in which a male and a female are paired and
thereafter must maintain a monogamous relation. Those who choose not to par-
ticipate in t he ceremony are thereafter not permi t t ed to engage in sexual activity
with a partner wi t hout social disapproval. It is t he males who do not participate
in t he ceremony t hat are thereafter pukas. So, al t hough it was not at first evident
to our anthropologist, t he "underlying" feature in t he case of pukas is t he prac-
tice of t he members of t he culture, treating a male 16 years old or older who has
52
Defining Art: Int ensi on and Extension
refused to participate in t he ceremony t hat regularizes sexual activity among
persons over 16 in a certain fashion. The practice t hat our ant hropol ogi st takes
not e of is not underlying or hi dden in t he way t hat t he universal properties of
gold, water, and species are, but it is not as obvious as t he colors of t he islanders'
clothes either. It takes some observing, inferring, and t heori zi ng to arrive at an
underst andi ng of t he cultural practice, but t he practice in a way is transparent. The
cultural practice is underlying but t here to be taken not e of.
When writing down her notes on t he island culture, our ant hropol ogi st trans-
lates "puka" as "bachelor." When at t he end of her first year on t he island anot her
native translator arrives, he says t hat t he translation is close but incorrect and t hat
t here is no exact equivalent for "puka" in English. Our ant hropol ogi st woul d have
discovered t he underlying nat ure of pukas, but she woul d t hen have gone on to
translate t he word "puka" wrongly. American and European societies do not have
pukas because we do not compel t he ceremony and behavior described above. Any
society t hat did compel t he ceremony and behavior described woul d have pukas,
even if only this island in fact had this practice. Pukas are individuals compelled
and regulated as described, and they woul d be such in all possible worlds in which
they exist. Our ant hropol ogi st woul d have const ruct ed an essentially correct
t heory of one aspect of t he island culture wi t hout t he help of t he intensional
cont ent of t he island language, al t hough her use of "bachelor" in her account is
not quite right. Pukas and bachelors are similar in a certain central respect, but
they also differ in some i mport ant central aspects, so we cannot say that they have
t he same underlying nat ure.
The underlying nat ure t hat bachelors have and t he underlying nat ure t hat pukas
could have, unlike t he physical reality t hat gold, water, and species have, are cul-
tural realities (cf. Searle 1997). Such cultural natures are or woul d be a small
part of a larger reality t hat is constituted by webs of relations t hat are or could be
instituted by a society of persons.
2. 5
Theori zi ng about art began in and has been carried out t hr oughout almost all of
its history in t he Platonic mode, which focuses on discovering t he intensions of
our words. Even when philosophers have given up on Forms, they have contin-
ued t he top-down approach of seeking intensions - in t he analysis of concepts, in
ordinary language, or in j ust seeking definitions wi t hout saying how they are doi ng
so. The "intensional" approach, which focuses on language, will not work for
such t erms as "gold, " "water," and their like, but it appears to work for t erms
like "bachelor." Carney himself ment i ons that, unlike "gold, " "bachelor" has a
specifiable intension. He writes, "A term like ' bachelor' in its accurate adult use
is i nt roduced as a synonym for ' adult male not previously mar r i ed' " (1975: 199).
Carney' s saying t hat "bachelor" is "i nt roduced as a synonym" makes it sound as
53
George Dickie
if t he t erm was put i nt o t he language in t he way a technical t erm is i nt roduced by
a philosopher or a logician, but of course he does not mean t hat it was. "Bache-
lor" came i nt o t he language as a co-relative t erm to "marriage, " and bot h t erms
(and many others) ride on practices t hat we have instituted as a cultural gr oup.
So, "bachelor," unlike "gold, " has an intension, but on t he ot her hand, "bache-
lor," like "gol d, " can be approached t hr ough its extension; al t hough in t he case
of an English word like "bachelor," we never bot her to do this because, as native
speakers of English, we have intensional access to its meani ng. In t he imagined
case of pukas, t he language-deprived ant hropol ogi st is forced to approach t hr ough
an extension because she lacks intensional access to t he islander' s language.
The concepts bachelor and puka, unlike t he concepts gold, water and tiger, are
transparent, t hat is, a person who knows t he culture in which t he concept func-
tions, knows t he concept. The natures of bachelors and pukas are underlying in
t he sense that one cannot j ust look at an individual and see t hat he is a bachelor
or a puka; one must know whet her an individual is enmeshed in t he relevant
cultural relations.
The nature of bachelors and t he nature of gold are similar in t hat they are dis-
coverable by empirical inquiry. The natures differ in t hat t he nature of bachelors
depends on cultural developments - decisions and t he like - whereas t hat of gold
does not . So, one is t empt ed to say t hat t he nat ure of gold cannot change, which
is t rue, but t hat t he nat ure of bachelors can, which is misleading. The use of t he
word "bachelor" can change in several different ways. Assume that at a given time
t he word has one meani ng. The word could change its meani ng entirely while still
having only one meani ng. It could change entirely and have t wo completely new
meanings. The word could acquire a second meani ng with t he original meani ng
remaining unchanged, and so on. But even if t he word "bachelor" changed in
one of these ways or even if t he word ceased to exist as a word in t he language,
t he condi t i on of being an unmarri ed adult male and t he cultural practice of
categorizing such an individual need not change. The condi t i on and practice
remain a logical possibility even if "bachelor" ceases to exist in t he language and
every person is in fact married or previously married.
I believe t he concept art is like t he concept bachelor (and t he concept puka),
and t he word "art" is like t he word "bachelor" (and t he word "puka"). Of course,
no one has ever felt t he need to put forth a t heory of bachelorness. Plato seems
never to have felt t he need to write a dialogue about it as he did about piety,
friendship, and justice, or even to attack it in passing as metaphysically inferior and
psychologically dangerous as he did art. Why has no one ever felt t he need for a
t heory of bachelorness? Perhaps because "bachelor" does not serve as an evalua-
tive weapon-word in t he way t hat "art" does, and, no doubt , for a variety of ot her
reasons; but probably no one has ever felt t he need for a t heory of bachelorness
because intensional access to its meani ng is so easy and uncontroversial. In any
event, Plato did theorize about art and did attack art on metaphysical and
psychological grounds, and philosophers have been trying to theorize about art
ever since. Intensional access to "art" is obviously much mor e difficult t han to
Defining Art: Int ensi on and Extension
"bachelor," and it is clearly much more controversial. We do not seem to need
to apply what may be called "t he extensional approach" to "bachelor," but perhaps
t he difficulty and controversy involved with "art" can be avoided by using t he
extensional approach with it.
Earlier I not ed t hat in order to use t he extensional approach to "art" t he second
place in t he art/ pair should be filled with t he name of some kind of scien-
tist. Following t he pattern established with t he case of bachelors and pukas, I think
t he pair should be art/cultural anthropologists. Art is, I have l ong believed, a cul-
tural not i on, and cultural anthropologists are t he scientists that deal with cultural
phenomena. My own belief in art as a cultural phenomenon is demonst rat ed by
t he fact that t he institutional t heory of art, which I have been defending in one
form or anot her for a l ong time, is clearly a cultural theory. Perhaps it is wor t h
not i ng that Leddy twice alludes to art as a cultural concept by contrasting it with
what he calls natural science concepts in his 1987 article t hat I discussed above,
but t hat is all he says on t he topic.
When I say t hat art is a cultural not i on, I mean t hat it is a phenomenon t hat
has been invented by a cultural gr oup and t hat it is not a genetically det ermi ned
behavior like mating, eating, and t he like.
4
In saying this, I do not mean to suggest
t hat only we or some small gr oup of societies have art or our concept of art. I
think that Denni s Dut t on is probably right that all human societies have art
(2000). Of course, t here could be and mi ght be a human society t hat does not
yet in fact have art. I only wish to say of each society t hat has art t hat it was
invented at some poi nt in t he past. It is, of course, possible t hat in t he case of a
given society art was i mport ed from anot her culture before it could be invented
indigenously, so in this case t he poi nt in t he past referred to woul d be in t he past
of anot her society.
What woul d be t he general features of t he application of t he extensional
approach to t he not i on of art using t he art/cultural anthropologist pairing? We
woul d be looking for a cultural phenomenon t hat is shared by all cultures or at
least many cultures, since, as not ed, a given culture mi ght not yet have art. The
phenomenon woul d have to be of a cultural nat ure t hat is t he same in all cultures
t hat have it. That nat ure woul d have to be of a rather abstract kind, given t he
widely differing kinds of art t hat t here are. And it mi ght be t hat t here are things
t hat very closely resemble a kind of art t hat are not art; t hat is, t here could be an
arbitrariness about one t hi ng' s being art and a very closely resembling t hi ng not
being art. Such arbitrariness is perhaps inevitable where cultural matters are con-
cerned, because cultural matters are how they are as t he result of how a culture
has "set things up" at some time or duri ng some period in t he past.
It woul d perhaps be best to begin our looking within our own culture. We
woul d approach our own culture as t he imagined ant hropol ogi st did t he island
culture, except t hat we have t he advantage t hat we are native speakers of our own
language. Of course, various aspects of language could be handicaps that lead
native speakers astray. Words in their dictionary senses typically have a number of
different meanings, which could be confusing. Further, we (and any native speaker
55
George Dickie
of any natural language) can make a word mean almost anything by t he use of
inflection, irony, gestures, juxtaposition, and t he like, and this t oo may confuse
us in t he cases of t he meanings of some words. All of this linguistic flexibility
may obscure our view of t he practices t hat underlie t he specific meani ngs of
some words. We mi ght be better able to isolate these specific meani ngs of these
practice-dependent words if we were language-deprived observers, like t he
imagined anthropologist. Of course, language-deprived observers can make t he
kind of mistranslation imagined.
The first lesson to be derived from t he imagined case of t he language-deprived
ant hropol ogi st ' s mistranslation is t hat in t he case of t he meani ng of culturally
det ermi ned words, access to t he intensional cont ent of t he language of t he
members of t he culture involved can be useful. The second lesson to be derived
from t he case of t he language-deprived ant hropol ogi st (one i gnored by t he purely
intensional approaches of historical philosophers of art) is t hat in t he cases of
certain concepts and t he words t hat go with t hem, knowl edge of t he practices t hat
underlie t he concepts and their words is crucial to their meani ng. Applying these
t wo lessons, it may be possible to make some progress in t heori zi ng about art.
To apply t he extensional approach to t he not i on of art using t he art/cultural
anthropologist pairing, one needs to look for linguistic usage that is integrated with
cultural practice analogous to t he islanders' use of "puka" and their practice
involved in organizing t he cultural activities of pukas, and our use of "bachelor"
and t he practice involved in our organizing t he cultural practice of bachelors.
If we start, as cultural anthropologists, with our own linguistic usage, t he usages
of "art" woul d have to relate to a provisional description in t he way t hat t he
usages of "gold" relate to t he properties of being yellow and malleable as a pro-
visional description. In t he way t hat physicists t hen focus on yellow and malleable
objects on their way to discovering t he atomic number of gold, cultural ant hro-
pologists will have to start by focusing on objects t hat satisfy t he provisional
description. The usages we will find with "art," however, because of our wide
linguistic flexibility, will be all over t he place. But not e t hat t he usages of "gold"
will be t oo; consider, "You struck gold!" said to someone who bought Xerox
stock early on or to t he discoverer of an i mport ant scientific t rut h. Despite all of
its many and varied usages, with "gold" we somehow wi nnowed our way down
to a preliminary extension of yellow and malleable objects and t hen proceeded
to atomic theorizing. Wi t h "art," we will have to wi nnow our way down to a
reasonable preliminary extension and t hen look for a cultural practice t hat
underlies this preliminary extension or some significant subset of it.
We can think of t he history of t he philosophy of art - from Plato to Dant o -
as a kind of wi nnowi ng of t he preliminary description of "work of art," al t hough,
as we shall see, t he process does not always eliminate items from t he extension of
works of art - sometimes it adds t hem.
Not even t he earliest imitation theorist woul d have been moved to think of
driftwood as art if some ot her ancient Greek, foreshadowing Morris Weitz and
some ot her t went i et h-cent ury philosophers, had ut t ered t he Greek equivalent of
Defining Art: Int ensi on and Extension
"That driftwood is a lovely piece of sculpture. " He woul d not have considered t he
driftwood to be within t he extension of art, not only because it is not an imita-
tion but also because it is not an artifact, t hat is, a human artifact. Someone has
to be in t he grip of a philosophical movement to think that a piece of driftwood
all by itself is an artwork because it has been referred to as a lovely piece of sculp-
t ure. Thus, non-artifactuality was wi nnowed out of (or perhaps it should be said
never got into) t he preliminary description of "work of art" by virtue of some-
t hi ng like common sense.
Imitation theorists were, however, moved - eventually anyway - by anot her
kind of case: things t hat were obviously not imitations but which seemed like art-
works nevertheless. Being bot hered by this kind of case forced an expansion in
their concept i on of t he intension of artworks and caused t hem to cease being imi-
tation theorists. The expansion of t he intension of artworks by counterexamples
advanced by others or j ust noticed has been a standard feature of t heori zi ng about
art. Not i ce t hat to work as a counterexample to someone' s theory, t he alleged
counterexample must plausibly fit into an extension despite t he fact t hat it lacks all
or some part of t he theorist' s underst andi ng of t he intension of a t erm under
consideration. This kind of philosophical move has been responsible, not for a
wi nnowi ng out of characteristics from t he preliminary description of "work of
art," but for an addi ng of characteristics. This kind of move eliminates theories -
t he imitation t heory of art, for example, and, I think, t he expression t heory as
well.
I think t hat virtually all philosophers of art - present and past - are and always
have been agreed t hat it is poems, painting, plays, sonatas, sonnets, sculptures, and
such familiar items t hat are works of art and t hat this is t he extension that they
are and always have been t heori zi ng about - trying to state t he intension t hat fits
it. There has of course been some disagreement as to whet her to count Dada
objects and similar things as works of art, but this is a skirmish of little signifi-
cance. I have maintained t hat Dada objects are theoretically useful because they
have helped us gain insight i nt o t he art-making cont ext in which works of art are
embedded, either because Dada objects are works of art or because they are not
works of art but have been mistaken for works of art by some people. In any event,
let us set aside this dispute and focus at t ent i on on t he huge gr oup of works of art
about which t here is compl et e agreement.
The philosophical probl em with t he large gr oup of items that constitutes t he
extension of works of art has always been t he great diversity of its members. This
het erogeneousness has been t he great barrier to t he traditional at t empt s to extract
t he intension from t he exhibited features of these works.
5
By "exhibited features"
I mean characteristics t hat can be noticed by directly experiencing works of art -
for example, t hat they are representations, are expressive, are delicate, and t he like.
What exhibited features could be found to be exemplified in all these many and
diverse works of art? And what a ready source of counterexamples t hat same diver-
sity has been against all t hose imitationists, expressionists, and t he like who have
tried to specify partial or complete conditions from among exhibited features.
57
George Dickie
Focus now on present-day theories of art. I think Dant o' s at t empt to charac-
terize art in t erms of aboutness is an example of t he traditional search for t he inten-
sional meani ng of "work of art" among exhibited characteristics, and I believe t hat
his t heory is vulnerable to t he traditional kind of counterexample attack (cf. Dickie
1993: 76- 7; Carroll 1993). Beardsley' s at t empt to characterize art in t erms of aes-
thetic character is also an example of t he same traditional search among exhibited
characteristics and is subject to t he same kind of attack (1979: 729, 1983: 299) .
On t he ot her hand, I think Jerrold Levinson' s historical t heory and my institu-
tional t heory are different from t he traditional theories and can be underst ood as
attempts to discover t he underlying nature of t he extension of works of art - t he
underlying nat ure being t he non-exhibited, feature of works of art t hat ties t hem
together. I will not discuss Levinson' s t heory here because I have discussed its
difficulties elsewhere (1997a: 22- 4) .
What woul d a search by a cultural ant hropol ogi st uncover in a study of "how
we deal wi t h" works of art? I believe such a search woul d reveal an underlying
cultural st ruct ure like t hat envisaged by t he institutional t heory of art. The insti-
tutional t heory is an account of t he cultural st ruct ure within which works of art
are produced and function, and t he st ruct ure itself is specified in t erms of a variety
of cultural roles.
In any event, even if cultural anthropologists could not find a cultural struc-
t ure identical with t hat described by t he institutional theory, I believe they woul d
find a st ruct ure very much like it. That is, they woul d find a st ruct ure of t he general
sort t hat t he five declarations I gave as definitions in The Art Circle can serve as
a summary account of. These five are as follows:
1 An artist is a person who participates with underst andi ng in t he making of a
work of art.
2 A work of art is an artifact of a kind created to be presented to an artworld
public.
3 A public is a set of persons t he members of which are prepared in some degree
to underst and an object which is presented to t hem.
4 The artworld is t he totality of all artworld systems.
5 An artworld system is a framework for t he presentation of a work of art by an
artist to an artworld public. (1997b: 80- 2)
The st ruct ure presented by these five woul d be t he sort of t hi ng t hat constitutes
t he cultural essence of t he institution within which art has its being. And t he
second declaration (definition) is t he sort of t hi ng t hat can be taken as a state-
ment of t he cultural essence of a work of art, namely, a st at ement that identifies
art with t he complicated property of being an artifact of a kind created to be pre-
sented to an artworld public.
58
Defining Art: Int ensi on and Extension
2. 6
Assume for argument ' s sake that "A work of art is an artifact of a kind created to
be presented to an artworld public" captures t he cultural essence of art because
t he st at ement has been arrived at by means of a search by anthropologists into
cultural structure. Would t hat mean t hat it states t he intension of "work of art"?
When Carney applies t he rigid-designator approach to "work of art," he concludes
t hat "work of art" has no intension. Remember, however, what Carney says
about "bachelor," namely, t hat it has an intension. So, Carney treats words like
"bachelor" differently from t he way in which he treats words like "gold. " He
does not explain why "bachelor" is different from "gold, " and t here is not hi ng in
his article t hat requires him to do so. I, however, have claimed that what I am
calling "t he extensional approach" can be applied to "work of art" and have also
tried to apply it to "bachelor." So, I need at least to show how "bachelor" has an
intension if t he extensional approach can be applied to it.
In effect, I have agreed with Carney t hat "bachelor" means "an adult male not
previously married" and added that being an adult male not previously married is
t he underlying cultural nature of bachelors. How does t he case of t he nature of
bachelors, which is derived by means of t he extensional approach, differ from t hat
of t he nature of gold, which is derived by means of t he rigid-designator approach?
And how does t he case of "bachelor" differ from "gold" so t hat "bachelor" can
have an intension and "gol d" does not?
Consider t he question of how t he nat ure of bachelors differs from t he nat ure
of gold. First, t he nat ure of bachelors is a cultural nature; it is a cultural status.
Being a bachelor is of course not a legal status, but it is related to and derived
from marriage, which is a mat t er of law - law being an officially enacted cultural
phenomenon. Marriage is a cultural (legal) way of organizing various central
aspects of t he lives of human beings. Being a bachelor is a cultural (non-legal) way
of organizing aspects of t he lives of some human males - how and under what cir-
cumstances, for example, certain human males are invited to dinner. The nat ure
of gold, on t he ot her hand, is not cultural, but physical. There woul d be gold if
t here were no cultures anywhere. Ther e could not be bachelors wi t hout a culture
and in fact not wi t hout a culture t hat has marriage. We have no control over t he
nat ure of gold, but we do have control over t he nature of cultural things, al t hough
it is a complicated matter.
There is anot her i mport ant difference between t he nat ure of gold and t he
nat ure of bachelors. An individual sample of gold in t he actual world (the element
with t he atomic number 79) woul d be gold in all possible worlds, t hat is, t he indi-
vidual sample of gold is necessarily t he element with t he atomic number 79. In
contrast, an individual bachelor Adam in t he actual world mi ght be a married man
in some possible worlds and t hus not a bachelor in all possible worlds, t hat is, t he
individual bachelor Adam is not necessarily a bachelor. Being gold is a nat ure t hat
mi ght be called an intrinsic nature, while being a bachelor is a nat ure t hat an indi-
59
George Dickie
vidual acquires by fitting into a cultural context. It is wor t h not i ng in passing t hat
particular works of art are like bachelors in this respect. For example, a particular
physical object t hat is a work of representational art in t he actual world mi ght not
be a work of art in some possible world, because that possible world lacks t he cul-
tural institution of art and t hat particular physical object woul d t hen merely be a
representation; t hus, a particular work of art in t he actual world is not necessarily
a work of art.
Second, in t he case of discovering (or j ust knowing) t he nature of bachelors,
t here is no intellectual division of labor t hat requires experts to discover t hat nat ure
as is required in t he case of gold. Gold has a hi dden nature t hat requires highly
specialized individuals - physicists, chemists, and t he like - to discover its nat ure.
The nat ure of bachelors, al t hough it is underlying in being a culturally founded
phenomenon, is known to virtually everyone; its nature is transparent rather than
hidden.
Consider now how "bachelor" differs from "gold. " First, "bachelor" is a cul-
tural t erm in t hat it derives in part from t he cultural t erm "married, " al t hough of
course it also involves t he biological t erms "adult" (which may be given some cul-
tural cont ent by way of a roughly specified age designation) and "male. " "Gol d"
is not a cultural t erm but a physical t erm, t hat is, it refers to objects with a
physical nat ure.
Second, "bachelor" has some similarity to technical words t hat are stipulated
to have a certain meani ng. In t he case of technical terms, it is typically individual
persons who do t he stipulating, but in t he case of words like "bachelor," it is t he
culture t hat does t he "stipulating" or somet hi ng like stipulating t hat may be called
cultural det ermi nat i on. The exact nature of this det ermi nat i on is vague, and fur-
t her mor e t he courses of t he various det ermi nat i ons probably differ in t he cases of
different cultural t erms. The cultural det ermi nat i on of t he meani ng of cultural
words is of course closely related to t he fact that we have some control over t he
natures of cultural matters.
There is an intimate connect i on between "bachelor" and t he nat ure of
bachelors - a sort of congruency in which t he intension of "bachelor" and t he
nature of t he members of t he extension of "bachelor" are det ermi ned by our
culture in a logically concomi t ant way. The word "bachelor" (and its definition)
and t he nature of bachelors are bot h t he same kind of thing, namely, they are
bot h cultural product s. Bot h are co-ordinated and mi rrori ng product s of how we
organize t he intertwined pathways of our linguistic and behavioral lives. "Gol d"
and t he nat ure of gold lack this intimate, mi rrori ng relationship. We are partici-
pants in t he initiating and maintaining of t he cultural natures of t he kinds of things
of which being a bachelor is a prime example; these are cultural kinds. In contrast,
we have no such "insider" relation to t he natures of natural kinds.
The inarticulate Greek atomists were on t he right track about gold. In a way,
t he Platonists woul d have been on t he right track if they had t urned their atten-
tion to transparent t erms like "bachelor"; they woul d have had, however, to focus
attention, not on rational intuition of Forms in which supposedly every sort of
Defining Art: Int ensi on and Extension
Notes
I wish to thank Suzanne Cunningham and Ruth Marcus for reading and commenting on
earlier versions of this chapter.
1 In ancient Greek times perhaps this would have been called efficient causality.
2 In ancient Greek times this might have been called formal causality.
3 This approach is derived from the well-known work of Saul Kripke and Hilary Putnam.
4 For a discussion of cultural-kind and natural-kind theories of art, see Dickie (1997a:
25-8).
5 The distinction between the exhibited and non-exhibited features of works of art was
first made and used by Maurice Mandelbaum (1965).
References
Beardsley, M. (1979). "In Defense of Aesthetic Value." In Proceedings and Addresses of the
American Philosophical Association. Newark, DE: American Philosophical Association.
(1983). "Redefining Art." In M. J. Wreen and D. M. Callen (eds), The Aesthetic Point
of View: Selected Essays. Cornell University Press.
Carney, J. (1975). "Defining Art." British Journal of Aesthetics, 15: 191-206.
(1982). "A Kripkean Approach to Aesthetic Theories." British Journal ofAesthetics 22:
150-7.
Carroll, N. (1993). "Essence, Expression, and History." In Rollins (1993): 79-106.
(ed.) (2000). Theories of Art Today. Madison: University of Wisconsin Press.
Dickie, G. (1993). "An Artistic Misunderstanding." Journal of Aesthetics and Art
Criticism, 51: 69-71.
(1993). "Tale of Two Artworlds." In Rollins (1993): 76-7.
(1997a). "Art: Function or Procedure - Nature or Culture?" Journal of Aesthetics and
Art Criticism, 55: 25-8.
(1997b). The Art Circle. Evanston: Chicago Spectrum Press. Originally pub. 1984.
Dutton, D. (2000). "But They Don' t Have Our Concept of Art." In Carroll (2000):
217-38.
Kivy, P. (1979). "Aesthetic Concepts: Some Fresh Considerations." Journal of Aesthetics
and Art Criticism, 37: 423-32.
Leddy, T. (1987). "Rigid Designation in Defining Art." Journal of Aesthetics and Art
Criticism, 45: 263-72.
61
t hi ng participates, but on t he underst andi ng of our participation in cultural prac-
tices that are involved with transparent t erms.
The cultural t erm "work of art" is, I believe, like t he cultural t erm "bachelor"
and not like t he physical t erm "gold. " There is an intimate cultural connect i on
between "work of art" and t he nat ure of a work of art t hat our cultural ant hro-
pologists can discover such t hat that nat ure can be converted into a definition of
"work of art." I hope it will be discovered that t he definition is my institutional
one.