Maurice Mandelbaum—Family Resemblances and Generalization Concerning the Arts

[219]

AMERICAN PHILOSOPHICAL QUARTERLY
Volume 2, Number 3, July 1965

V. FAMILY RESEMBLANCES AND GENERALIZATION
CONCERNING THE ARTS

MAURICE MANDELBAUM

[SK's comments]

In 1954 William Elton collected and published a group of essays under the title Aesthetics and Language.

As his introduction made clear,
a common feature of these essays was

the application to aesthetic problems

of

some of the doctrines

characteristic of recent British linguistic philosophy.

... there have been a number of important articles which, in addition to those contained in the Elton volume, suggest the direction in which this influence runs.

...

"The Task of Defining a Work of Art"
by Paul Ziff,

"The Role of Theory in Aesthetics"
by Morris Weitz,

Charles L. Stevenson's
"On 'What is a Poem'"

and
W. E. Kennick's
"Does Traditional Aesthetics Rest on a Mistake?"

In each of them one finds a conviction which was also present in most of the essays in the Elton volume:

that it is a mistake

to offer generalizations concerning the arts,

...

to attempt to discuss

what art, or beauty, or the aesthetic, or a poem,

essentially is.

In partial support of this contention, some writers have

made explicit use

of Wittgenstein's doctrine of family resemblances;

Morris Weitz, for example, ...

However, ... [he] made no attempt to analyze, clarify, or defend

the doctrine itself.

...

I

The locus classicus for Wittgenstein's doctrine of family resemblances is in Part I of Philosophical Investigations, sections 65-77.⁷

---

⁷ ... A parallel passage is to be found in "The Blue Book": see ... The Blue and Brown Books (Oxford, Basil Blackwell, 1958), pp. 17-18.

---

In discussing what he refers to as language-games, Wittgenstein says:

Instead of producing something common to all that we call language, I am saying that these phenomena have no one thing in common which makes us use the same word for all—but they are related to one another in many different ways.

And it is because of this relationship, or these relationships, that we call them all "language."
(§65)

He then illustrates his contention by citing a variety of games, such as board games, card games, ball games, etc., and concludes:

We see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities of detail.
(§66)

I can think of no better expression to characterize these similarities than "family resemblances";

for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament, etc., etc. overlap and criss-cross in the same way.—

And I shall say: "games" form a family.
(§67)

...

[220]

...

Wittgenstein's concrete illustrations of the diversity among various types of games may at first make his doctrine of family resemblances extremely plausible.

...

Nonetheless, I do not believe that his doctrine of family resemblances, as it stands, provides

an adequate analysis

of why

a common name, such as "a game," is in all cases applied or withheld.

Is W saying

why

all of this is as it is?

Or just that it is?

He does say,

... it is because of this relationship ...

(§65)

Yet the next excerpt begins,

We see ...

(§66)

I read this latter as the post hoc "seeing" of the interested philosopher, who confines himself to saying only what can be said. The further developments which would enable the "analysis" to

apply

in

all cases

are not forthcoming.

This is in keeping with the 'stratified' or 'fractal' quality of W's thought: the concept game both contains language, and is contained by language. The "why" of this is inscrutable.

As for Carroll's contention which landed us here, it would indeed be a 'misuse' of the family resemblance thesis to take it as "an adequate analysis" when it purports to be merely an atomized true description at a particular level of observation. To assess for this, I would need to read Weitz much more closely than I have, and this seems not worth the trouble.

Consider first the following case.

Let us assume that you know how to play that form of solitaire called "Canfield";

suppose also that you are acquainted with a number of other varieties of solitaire ...

Were you to see me shuffling a pack of cards, arranging the cards in piles, [etc., etc.] ... you might say:

"I see you are playing cards. What game are you playing?"

However, to this I might answer:

"I am not playing a game; I am telling (or reading) fortunes."

Will the resemblances between

what you have seen me doing

and

the characteristics of card games

with which you are familiar

permit you to contradict me and say that I am indeed playing some sort of game?

Of course not, perhaps excepting some extraordinary contingencies which can safely be ignored.

Rules of

permission

are not rules of logic.
But W might say that this so-called

contradiction,

whether or not it is given voice, is a fait accompli in every respect which matters: everything you saw leads to the conclusion that I am playing a solitaire game; nothing you saw leads to any other conclusion. These are the facts of the case, upon which all concerned can agree.

The fact that you could be so easily and completely wrong here is hardly a refutation of the family resemblance thesis. Rather, the skeptic has to show that you are likely to be wrong more than occasionally; i.e., that such situations are endemic rather than exceptional.

The point of the example, I think, is to show in what type of situation such a mistake is bound to be endemic: if you literally cannot 'see' my cognition, then you cannot (you will not) take account of it in forming any concepts in which it is implicated; ergo, you may be surrounded by fortune tellers, yet you remain unable to form any concept of

telling (or reading) fortunes.

If that is the point, however, it is a point which cuts both ways, for in that case there simply would be no such concept to argue over, because it would not exist. If the concept does exist, it exists without endemic "contradiction".

Ordinary usage would not, I believe, sanction our describing fortune-telling as an example of playing a game, no matter how striking may be the resemblances between the ways in which cards are handled in playing solitaire and in telling fortunes.

On first pass, I wrote here:

We cannot normally have small talk with our intentional objects. If we could, who is to say what all they would tell us? No doubt it would be far more shocking than telling fortunes. But in all cases where we do not or cannot talk to them, we cannot correct our assessments in the above manner.

Or,
to choose another example,
we may say that while certain forms of wrestling contests are sometimes characterized as games
(Wittgenstein mentions "Kampfspiele")
an angry struggle between two boys, each trying to make the other give in, is not to be characterized as a game.

Yet

one can find

a great many resembling features between such a struggle and a wrestling match in a gymnasium.

But of course.

One can find

whatever one chooses.

What do we find??

Nothing W says above could be taken to deny that

a great many resembling features

can be found among objects called by different names, as well as they can be found among those called by the same name; nor does he identify any discrete threshold of similarity beyond which objects simply must be named the same.

Now, does the family resemblance theorist need to be able to identify this threshold? The point is, he cannot: if there is such a threshold, then all that lies afield of it has no name. Where is a horizon? What lies afield of it? We cannot say.

It is for this reason that any example Mandelbaum might 'name' here undermines his point at once, simply by having a name of its own. Evidently

Kampfspiele

and

struggle

already have (at least these) names. To point to resemblances among these differently-named things is a non sequitur vis-a-vis the family resemblance thesis. If these things are not (also) called games, or if they are only sometimes (also) called games, then any resemblances between them and games have not been sufficient to

form a family.

This much is perfectly encapsulated by Mandelbaum's own exposition above.

We do not need to litigate the threshold between chaotic resemblance and family resemblance if the names have already litigated it for us.

If this is not W's claim, and/or if W is making any broader claim than this, then I have more work to do trying to understand all of this.

What would seem to be crucial in our designation of an activity as a game is, therefore, not merely a matter of

noting a number of specific resemblances

between it and other activities

which we denote as games,

but

involves something further.

First pass note:

This is no counterexample, it is a restatement of W's argument.

Designation

per se is not a matter of

noting.

That is precisely W's point:
designation happens in absence of noting.

Now:

Based on the secondary literature I've perused, it sounds as if W is wide awake to the

involvement

of

something further.

If I've understood this much correctly, then obviously we are headed for what W would consider an ill-fated effort to give "an adequate analysis" of exactly what this "something" is. But really this issue marks precisely those 'limits of explanation' which philosophy cannot hope to breach.

To suggest what sort of characteristic this "something further" might possibly be, it will be helpful to pay closer attention to the notion of what constitutes a family resemblance.

Suppose that you are shown ten or a dozen photographs and you are then asked to decide which among them exhibit strong resemblances.⁹

---

⁹ ... Haig Khatchadourian has shown that Wittgenstein is less explicit than he should have been with respect to the levels of determinateness at which these resemblances are significant for our use of common names.

See "Common Names and 'Family Resemblances'," Philosophy and Phenomenological Research, vol. 18 (1957-58), pp. 341-358.

(For a related, but less closely relevant article by Professor Khatchadourian see "Art-Names and Aesthetic Judgments," Philosophy, vol. 36 [1961], pp. 30-48.)

---

Upon a quick read, the latter article strikes me as uninteresting and mistake-riddled. By far the most interesting thing about it is the resemblance (yuk!) of certain verbiage to Walton's canonical paper of a few years later.

The former paper, meanwhile, covers some of the same ground and seems just interesting enough to merit attention.

You might have no difficulty in selecting, say, three of the photographs in which the subjects were markedly round-headed, had a strongly prognathous profile, rather deep-set eyes, and dark curly hair.¹⁰

---

¹⁰ It is to be noted that

this constitutes a closer resemblance

than that involved in what Wittgenstein calls "family resemblances,"

since in my illustration the specific similarities all pertain to a single set of features, with respect to each one of which all three of the subjects directly resemble one another.

In Wittgenstein's use of the notion of family resemblances there is, however, no one set of resembling features common to each member of the "family"; ...

Thus, in order to conform to his usage, my illustration would have to be made more complicated, and the degree of resemblance would become more attenuated.

...

However, if what I say concerning family resemblances holds of the stronger similarities present in my illustration, it should hold a fortiori of the weaker form of family resemblances to which Wittgenstein draws our attention.

---

In some extended, metaphorical sense you might say that the similarities in their features constituted a family resemblance among them.

The sense, however, would be

metaphorical,

since

in the absence of a biological kinship

of a certain degree of proximity
we would be inclined

to speak only of resemblances,

and not of a family resemblance.

What marks the difference between a literal and a metaphorical sense of the

[221]

notion of "family resemblances" is, therefore, the existence of a genetic connection in the former case and not in the latter.

Wittgenstein, however, failed to make explicit the fact that

the literal, root notion of a family resemblance

includes this genetic connection

no less than it includes the existence of noticeable physiognomic resemblances.¹¹

---

¹¹ Although Wittgenstein failed to make explicit the fact that a genetic connection was involved in his notion of "family resemblances,"

I think that he did in fact presuppose such a connection.

If I am not mistaken, the original German makes this clearer ...

...

Modifying Miss Anscombe's translation in as few respects as possible, I suggest that a translation of this passage might read:

I can think of no better expression to characterize these similarities than "family resemblances," since various similarities which obtain among the members of a family—their build, features, color of eyes, gait, temperament, etc., etc.—overlap and criss-cross in the same way.

[§67]

This translation differs from Miss Anscombe's ... in that it makes more explicit the fact that

the similarities

are similarities among the members of a single family,
and

are not themselves definitive

of what constitutes a family resemblance.

---

I cannot imagine W disagreeing with the suggestion that

the similarities

are not themselves definitive

,

nor with the suggestion that he is speaking in an

extended, metaphorical sense

.

If there is any substance to this disagreement, it concerns the suggestion that

we would be inclined to speak only of resemblances,

and not of a family resemblance

so long as no

genetic connection

is concurrently in evidence.

This is precisely the point of the metaphor: we do speak of category members as if there were some "genetic connection" among them, yet it is rare that this connection actually exists.

When it comes to actual people, given only appearances to work with, it is not at all uncommon to suspect that two of them are related, nor to be wrong about this more often than one is right.

Now, as for the notion that our very concept of what it is for two objects to bear resemblance is a notion which grows up around tacit knowledge of myriad "genetic connections", this notion is quite in line with the Wittgensteinian gestalt, but it provides no ammunition for Mandelbaum here; for the fact remains that we readily form categories where there has been no such genetic descent.

Had the existence of such a twofold criterion been made explicit by him,

he would have noted that there is in fact an attribute common to all who bear a family resemblance to each other:

they are related through a common ancestry.

Such a relationship is not, of course, one among the specific features of those who share a family resemblance;

it nonetheless differentiates them from those who are not to be regarded as members of a single family.¹²

---

¹² Were this aspect of the twofold criterion to be abandoned,

and were our use of common names to be solely determined by the existence of overlapping and criss-crossing relations,

it is difficult to see

how a halt would ever be called

to the spread of such names.

Robert J. Richman
has called attention to the same problem
in
"'Something Common'," Journal of Philosophy, vol. 59 (1962), pp. 821-830.

He speaks of what he calls "the Problem of Wide-Open Texture," and says:

"the notion of family resemblances may account for our extending the application of a given general term, but

it does not seem to place any limit

on this process"

(p. 829).

In an article entitled
"The Problem of the Model-Language Game in Wittgenstein's Later Philosophy," Philosophy, vol. 36 (1961), pp. 333-351,

Helen Hervey also calls attention to the fact that "a family is so-called by virtue of its common ancestry"

(p. 334).

...

---

Here is Hervey on W's "notion of the model language game".

[337]

...

not without surprise perhaps, we find that Wittgenstein himself by no means confined his investigation to the enumeration of actual usages.
While this certainly is part of his task, it becomes quickly apparent that he did not believe that by this procedure alone he could achieve that 'complete clarity' at which he aimed
...
He states, in fact, that there 'is not a philosophical method', but 'different methods' which he likens to 'therapies'.
...
one does not need to look very far ... to discover that one of the most important of them is to be found in the notion of the model language-game.
...
it is one which [the philosopher] deliberately invents to assist him in his investigation.
And with this we come up against a problem, for has not Wittgenstein stated frequently that ordinary language is perfect as it is, that every sentence of our language, even the vaguest sentence, is 'in order as it is', and all that is to be done is to study what lies clearly before us?

Compare to the 'ideal-type' in sociology. At least sentences are not persons with purported rights and responsibilities.

---

As for talk of what

determines

our use of common names

and of all

halts

and

limits

on this process

flying the coop right along with necessary and sufficient conditions, this seems like a basic misunderstanding. W is not offering a positive account of any of this; his is a 'deflationary' account.

If, then, it is possible that the analogy of family resemblances could tell us something about

how games may be related to one another,

one should explore the possibility that,
in spite of their great dissimilarities,
games may possess

a common attribute which,

like biological connection,

is not itself one among their

directly exhibited characteristics.

Unfortunately, such a possibility was not explored by Wittgenstein.

Or does

the analogy

purport to

tell us

how games

are not

related

?

Is family resemblance itself a fact or an artifact?

To be sure,
Wittgenstein

does not explicitly state

that the resemblances which are correlated with our use of common names

must be of a sort

that are directly exhibited.

Nonetheless,
all of his illustrations in the relevant passages involve

aspects of games

which would be included

in a description of how a particular game is to be played;

that is,
when he commands us to "look and see" whether there is anything common to all games,
the "anything" is taken to represent

precisely the sort of manifest feature

that is described in rule-books,

such as Hoyle.

However,
as we have seen in the case of family resemblances,
what constitutes a family

is not defined

in terms of

the manifest features of a random group of people;

we

must first characterize

the family relationship in terms of genetic ties,

and then observe

to what extent those who are connected in this way resemble one another.¹⁴

How harsh to be with this passage?

It's clear enough that the admonition to

characterize   first

and only

then observe

is perfectly contrary to W's explicit argument.

Presume there's no case to be made that

we must

proceed this way. Still, it's plausible enough that this is how we proceed in the specific case of family: without prior (tacit) knowledge of

genetic ties

we simply could not possess the concept of a particular kind of resemblance which (1) prevails among members of "families", but (2) does not prevail (on the whole) among the aggregates of unrelated people that we typically encounter.

This much granted, it remains doubtful that W means to entangle us in this two-step.

---

¹⁴ Although I have only mentioned the existence of genetic connections among members of a family, I should of course not wish to exclude the effects of

habitual association

in giving rise to some of the resemblances which Wittgenstein mentions.

I have stressed genetic connection only because it is

the simplest and most obvious illustration

of the point I have wished to make.

---

A curious disclaimer. Is it really necessary?

When it comes to concept formation in general,

habitual association

is most of the story. But then, nonperson objects do not learn from their "associates"; rather, we learn (if we do) to associate the objects, in the peculiar sense of "associate" which attaches to the practice of using the same word to describe different objects.

A further curious aspect of this, then, is that it seems not to matter too much (if at all) what exactly the so-called "association" consists of: it matters only that there is such an association made, and not at all that it be in any sense 'real'.

e.g. My association of the song Goodie, Goodie with the clarinet may well be unique to me. At a minimum, certain more granular aspects of it are undoubtedly unique; we have only to zoom in until we land on them. Presumably, then, no one else will understand what I mean (and I will not be tempted to find out whether they do) so long as this association fails to reach critical mass. But who is to say that sharing in such an association (or not) is what makes it 'real' (or not)? What really changes with changes in sharedness? I suppose the answer is: the form of life is what changes this way. That is not nothing, but it is pretty close to nothing.

In the case of games,
the analogue to genetic ties might be

the purpose for the sake of which

various games were formulated by those who invented or modified them,

e.g.,
the potentiality of a game to be of absorbing non-practical interest to either participants or spectators.

If there were any such common feature

one would not expect it to be defined in a rule book,

such as Hoyle,
since rule books only attempt to tell us how to play a particular game:

our interest in playing a game, and our understanding of what constitutes a game, is already presupposed by the authors of such books.

Yikes! A genuine strawman traipses on stage!

It is Mandelbaum, not W, who has foisted

such books

upon us here.

It is true that

the sort of manifest feature

that we

see

when we

look

is of

the sort

that the

rule-books

traffic in. But the books are under no obligation to make philosophical sense! And we are under no obligation to take them as such!

Or . . . is this . . . (gasp!) . . . a mere

analogy

?

[222]

It is not my present concern to characterize any feature common to most or all of those activities which we call games,

nor would I wish to argue on the analogy of family resemblances that there must be any such feature.

If the question is to be decided,
it must be decided by an attempt to "look and see."

However, it is important that we

look in the right place

and in the right ways

if

we are looking

for a common feature;

'We won't know if we don't look.'

'Looking' at all of course does presume that we already know where and how. This point is well taken. Still, it verges on mauvaise foi to take W as having failed even to 'look'. W's well-known irreverence for background reading is immaterial here. He is plainly aware of the grand project of 'looking'. His remarks presume this project, and are are directed squarely at it.

we should not assume that any feature common to all games must be some manifest characteristic,

such as whether they are to be played with a ball or with cards,
or
how many players there must be in order for the game to be played.

If we were to rely exclusively on such features
we should,
as I have suggested,

be apt

to link

solitaire with fortune-telling,
and
wrestling matches with fights,

rather than (say) linking solitaire with cribbage and wrestling matches with weight-lifting.

This

aptness

is spurious and imagined. The reason why should be familiar by now: all positing of

features

is retrospective.

This is to say:

that we can identify a feature retrospectively does not prove or entail that this feature really was operative in the formation of the concept.

W makes no warrant that

we rely exclusively on such features

.

Rather, family resemblance is a metaphor which highlights the difference between (1) the actual process of concept formation, and (2) the retrospective view of any results.

It is, then, my contention that Wittgenstein's emphasis on directly exhibited resemblances, and his failure to consider other possible similarities, led to a failure on his part to provide an adequate clue as to what—in some cases at least—governs our use of common names.¹⁵

---

¹⁵ I do not deny that directly exhibited resemblances often play a part in our use of common names:

this is a fact explicitly noted at least as long ago as by Locke.

However,

similarities in origin,

similarities in use,

and

similarities in intention

may also play significant roles.

It is such factors that Wittgenstein overlooks in his specific discussions of family resemblances and of games.

---

How are these latter similarities to become known, if not by being exhibited ? He has parsed family very literally, and now he parses exhibited in the same vein. True, we would not say that genetic descent is "exhibited"; but (apropos of this recent British linguistic philosophy ) this is an artifact of grammar and usage and is minimally informative about the central question. Genetic descent CAN be exhibited: a trained geneticist could demostrate the fact for his colleagues, if not (truly) for those of us without such rigorous training; the rest of us would require a minimum of nine months time and unfettered access to some very private matters, but we too could witness a demonstration of descent. This is socially absurd; it is not philosophically absurd. Descent IS demonstrable empirically provided we accept certain background assumptions: the trained geneticist makes thousands of these, and this enables a highly streamlined "exhibit"; the layperson assumes only a few things about sexual intercourse and childbirth, and so their "exhibit" must be relatively quite tedious (for all concerned). If the question turns on the FACT of genetic descent then the fact either is or is not established (for us). And it IS a manifest fact; or at least it CAN be rendered as one. That it typically is not rendered this way is immaterial. -->

If the foregoing remarks are correct,
we are now in a position to see
that

the radical denigration of

generalization concerning the arts,

which has come to be almost a hallmark of the writings of those most influenced by recent British philosophy,

may involve serious errors, and may not constitute a notable advance.

II

In turning from Wittgenstein's statements concerning family resemblances to the use to which his doctrine has been put by writers on aesthetics,

we must first note what these writers are not attempting to do.

In the first place, they are not seeking to clarify the relationships which exist among the many different senses in which the word "art" is used.

Any dictionary offers a variety of such senses

(e.g., the art of navigation, art as guile, art as the craft of the artist, etc.),

and it is not difficult to find a pattern of family resemblances existing among many of them.

However, [this] ... has not ... been of interest to the writers of the articles with which we are here concerned.

In the second place, these writers have not been primarily interested in analyzing how words such as "work of art" or "artist" or "art" are ordinarily used by those who are neither aestheticians nor art critics;

their concern has been with the writings which make up the tradition of "aesthetic theory."

In the third place, we must note that the concern of these writers has not been to show that family resemblances do in fact exist among the various arts, or among various works of art;

on the contrary,

they have used the doctrine of family resemblances

in a negative fashion.

In this, they have of course followed Wittgenstein's own example.

Well, I gather that the scarcity of "positive"

doctrine

in W's work is hardly incidental. 'Deflationary' seems to do more justice to this issue than "negative" can.

...

However,
as the preceding discussion of Wittgenstein should have served to make clear,

one cannot assume that

if there is any one characteristic

common to all works of art
it must consist in

some specific, directly exhibited feature.

Like the biological connections among those who are connected by family resemblances,
or
like the intentions on the basis of which we distinguish between fortune-telling and card games,
such a characteristic

might be

a relational attribute,

rather than some characteristic at which one could directly point and say:

"It is this particular feature of the object which leads me to designate it as a work of art."

A relational attribute of the required sort might, for example, only be apprehended if one were to consider specific art objects as having been

created by someone

for some actual or possible audience.

On this, I keep coming back to the same issue:

If I have knowledge of a

relational attribute,

then for me this attribute is

manifest;

and if not, then not.

It's true that a work's complete catalog of "relational properties" typically is not compiled, and it probably cannot be compiled; and if not, then this catalog cannot be displayed alongside the work. It certainly cannot be read off the work's surface.

If this is what "manifest" and 'nonmanifest' mean here, then these are trivialities. If there is a dispute over art status which hinges upon the existence of such a property, the dispute is resolved precisely by "manifesting" the property for the benefit of any skeptics; then this property can no longer be called 'nonmanifest'.

The locution 'nonmanifest property' describes what an economist might call an 'information asymmetry'. This matter is entirely contingent and earthly; it is not in the least an ontological matter.

What is the remedy for an information asymmetry? It is for the party on the short end of it to gain parity of knowledge with his counterpart: he must learn what the other already knows. As always, anywhere the 'what' of knowledge becomes important, the 'why' cannot be far behind. We can observe the "relation" first-hand; we can be told of it by an expert, or by a rando on the internet; perish the thought that in unguarded moments we can even (think we can) read it off the surface of the work, though we ourselves would agree that it is not "manifested" there.

In art, the majority of our knowledge of "relational attributes" is so-called 'testimonial knowledge'. It doesn't take a Wittgenstein to be skeptical of any ontological conclusions which depend upon this kind of knowledge for their support. Reliability is the least of our problems in this regard. Asymmetry is the real bugbear. Asymmetry warrants questioning whether these are "properties" at all. Properties can be multiplied endlessly; anywhere we are not entirely overwhelmed by a profusion of properties, we can be sure that salience has already had its say.

The suggestion that the essential nature of art is to be found in such a relational attribute is surely not implausible

when one recalls some of the many traditional theories of art.

It's safe to assume that there is indeed a

relational

aspect of categories. Rosch, e.g., emphasizes the importance of dissimilarity as well similarity in category formation.

But again, to say that such "attributes" have been operative in concept formation just is to say that they have been "manifest" to those who come to possess these categories. Or, if you prefer it the other way around: to suggest that some given properties are not "manifest" just is to say that they have not been operative in concept formation.

The enfranchisement of nonmanifest properties as decisive of art status is therefore a pure stipulation. It is just another 'honorific' theory; only it is unclear just what basis for 'honor' simply must inhere in a work's "relations", whereas the theories namechecked by Weitz at least sought to honor something worth honoring, even if it turns out that not all artworks could possible share it.

I am coming around to the notion that this paper is 'impressive', at least in the sense of being ahead of its time.

For example, art has sometimes been characterized as being one special form of

communication

or of

expression,

or as being a special form of

wish-fulfillment,

or as being a presentation of

truth in sensuous form.

And it requires only one person (or at most two) to

communicate,

express,

or

wish

?

Or must there be more?

If a single act of "communication" between two people is all that is required, i.e. if this act need not be "manifest" to anyone beyond the dyad, then this is a severe deflation of Social and Institutional Theories of art, not their apotheosis.

Such theories do not

assume

that in each poem, painting,

[223]

play, and sonata there is

a specific ingredient

which identifies it as a work of art;

Hello?

Weitz has said that this is exactly what they

assume.

Have I missed something here?

rather,
that which is held to be common to these otherwise diverse objects is a relationship which

is assumed

to have existed,
or

is known

to have existed,
between

certain of their characteristics

and

the activities and the intentions

of those who made them.¹⁶

Taken literally, this can be actually squared with Rankian-Beckerian theory! For Rank has given a particular account of this

relationship

between

characteristics

and

activities

.

The first hangup is that Rank does not give any of the accounts that Art Theory has given.

The next hangup is that Rank's explanatory account of where art comes from psychically and socially in fact comes to share the fatal flaw of Weitz' 'honorific' theories as soon as it is recruited as a definition. The reason why is precisely that which set Rank on his path in the first place: some 'artists' (i.e. 'neurotics') never produce a 'work of art'. So even here we find something like the family resemblance thesis standing between us and a tantilizing candidate definition: the priest, politician and painter have almost everything in common, but it is their differences which loom largest, pragmatically and ontologically alike. This is family resemblance, avant la lettre, applied in precisely the same 'social' domain which so-called Institutional Theorists also seek to exalt, operationalized with a comprehensiveness they cannot achieve. And it too fails as a definition.

---

¹⁶ I know of no passage in which Wittgenstein takes such a possibility into account.

In fact,
if the passage from "The Blue Book" to which I have already alluded may be regarded as representative,

we may say that Wittgenstein's view of traditional aesthetic theories was quite without foundation.

In that passage he said:

The idea of a general concept being a common property of its particular instances connects up with other primitive, too simple, ideas of the structure of language.

It is comparable to the idea that properties are ingredients of the things which have the properties; e.g., that beauty is an ingredient of all beautiful things as alcohol is of beer and wine, and that we therefore could have pure beauty, unadulterated by anything that is beautiful

(p. 17).

I fail to be able to identify

any aesthetic theory

of which such a statement would be true.

It would not, for example, be true of Clive Bell's doctrine of "significant form,"

nor would it presumably be true of G. E. Moore's view of beauty,

since both Bell and Moore hold that beauty depends upon

the specific nature of the other qualities

which characterize that which is beautiful.

Curious argument here.

What is this

specific nature,

these

other qualities,

if not

properties

taken as

ingredients of things

?

This appeal to
"the other qualities"
(besides beauty itself)
which

characterize

the art-object
sounds suspiciously like the family resemblance thesis.

It would be strange indeed, then, to argue from this that Weitz et al have misapplied family resemblance. Rather, the charge would have to be that they have overlooked its prototype in precisely those works which they understand as taking the 'necessary and sufficient conditions' tack.

However, it may be objected that when I suggest that what is common to works of art involves reference to "intentions," I overlook "the intentional fallacy" ...

This is not the case.

The phrase "the intentional fallacy" originally referred to a particular method of criticism, that is, to a method of interpreting and evaluating given works of art;

it was not the aim of Wimsatt and Beardsley

to distinguish between art and non-art.

These two problems are, I believe, fundamentally different in character.

However, I do not feel sure that Professor Beardsley has noted this fact,

for in a recent article in which he set out to criticize those who have been influenced by the doctrine of family resemblances

he apparently felt himself

obliged

to define art
solely terms of

some characteristic in the object itself

(see "The Definition of the Arts," Journal of Aesthetics and Art Criticism, vol, 20 [1961], pp. 175-187).

Had he been willing to relate this characteristic to the activity and intention of those who make objects having such a characteristic, his discussion would not, I believe, have been susceptible to many of the criticisms leveled against it by Professor Douglas Morgan and Mary Mothersill (ibid., pp. 187-198).

---

While we may acknowledge that it is difficult to find any set of attributes—whether relational or not—which can serve to characterize the nature of a work of art

(and which will not be as vulnerable to criticism as many other such characterizations have been),

it is important to note that the difficulties inherent in this task are not really avoided by those who appeal to the notion of family resemblances.

As soon as one attempts to elucidate

how the term "art" is in fact used

in the context of

art criticism,

most of the same problems which have arisen in the history of aesthetic theory will again make their appearance.

In other words,
linguistic analysis does not provide a means of escape
from

the issues

which have been of major concern
in

traditional aesthetics.

This fact may be illustrated through examining a portion of one of the articles to which I have already alluded,

Paul Ziff's article entitled
"The Task of Defining a Work of Art."

To explain how the term "a work of art" is used, and to show the difficulties one encounters if one seeks to generalize concerning the arts,

Professor Ziff chooses as his starting point one clear-cut example of a work of art and sets out to describe it.

The work he chooses is a painting by Poussin,
and his description runs as follows:

Suppose we point to Poussin's "The Rape of the Sabine Women" as our clearest available case of a work of art.

We

could

describe it by saying,

first, that it is a painting.

Secondly, it was made, and what is more, made deliberately and self-consciously with obvious skill and care, by Nicolas Poussin.

Thirdly, the painter intended it to be displayed in a place where it could be looked at and appreciated, where it could be contemplated and admired. . . .

Fourthly, the painting is or was exhibited in a museum gallery where people do contemplate, study, observe, admire, criticize, and discuss it.

What I wish to refer to here by speaking of contemplating, studying, and observing a painting, is simply what we may do when we are concerned with a painting like this.

For example, when we look at this painting by Poussin, we may attend to its sensuous features, to its "look and feel."

Thus we attend to the play of light and color, to dissonances, contrasts, and harmonies of hues, values, and intensities.

We notice patterns and pigmentation, textures, decorations, and embellishments.

We may also attend to the structure, design, composition, and organization of the work.

Thus we look for unity, and we also look for variety, for balance and movement.

We attend to the formal interrelations and cross connexions in the work, to its underlying structure. . . .

Fifthly, this work is a representational painting with a definite subject matter;
it depicts a certain mythological scene.

Sixthly, the painting is an elaborate and certainly complex formal structure.

Finally, the painting is a good painting.

And this is to say simply that the

[224]

Poussin painting is worth contemplating, studying, and observing in the way I have ever so roughly described.

With reference to this description we must first note that it is clearly not meant to be anything like

a complete description

of the Poussin painting;

it is at most a description of those aspects of that painting which are

relevant

to its being called a work of art.

For example, neither the weight of the painting nor its insurable value is mentioned.

Thus,
whether because of his own preconceptions,
or because of our ordinary assumptions concerning how the term "work of art" is to be used,
Professor Ziff

focuses attention on some aspects

of the Poussin painting rather than

upon others.

In doing so,
he is

making an implicit appeal

to what is
at least

a minimal aesthetic theory,

that is, he is

supposing

that neither weight nor insurable value

need be mentioned

when we list

the characteristics which lead us to say

of a particular piece of painted canvas that it is a work of art.

Well,

we could describe it

this way. He does not say that we must, or that we should.

Sometimes a hedge like this is weaselly, other times it is perfectly appropriate and unremarkable.

I miss no opportunity to raise the issue of

relevance,

but I do so precisely to suggest

a means of escape from

those pesky

traditional  issues.

The solution could not be more obvious:

implicit appeals

to

theory

can be entirely

avoided

by "avoiding"

description

.

Now, W has indeed suggested that "description" take the place of theory. The suggestion here is: to describe just is to have a theory, if only an "implicit" one. Why? Because description is prospectively infinite; to give any description is, inherently, to cherrypick "relevant" features from among an infinite flux.

This is a serious and worthy issue, but I don't know that it much impacts W's assessment of either of "traditional aesthetic theories" or of the broader task of philosophy (including any Aesthetic philosophy). Mandelbaum himself raises the implicit-explicit distinction, and this too has to has to be taken seriously. Relevance is operative in every human perception and judgment, but we don't normally say that Phil has a "theory" of car repair, or that Joe has a "theory" of Instagram reels. In other words, the charge that any mere description of an artwork evokes at least an "implicit theory" is a charge that has any sense at all only on terrain whereupon "theory" has long prevailed, and from which it retreats only with protracted skirmishing. This is to say, essentially, that the relevance issue is fractal, or perhaps, as we say today, that it is 'meta': it's not relevant to talk about relevance, except when it is; the same for theory. Must I have a "theory" of philosophy if I only post annotations of certain articles that I've read? I want to say that I would post most of them, if I had the time and bandwidth; or maybe I only think this; maybe I cannot think otherwise. In fact there is some "theory" at work here, but I defy the reader to tell me exactly what it is, given only the end results to work with.

In other words, it is not Ziff but Mandelbaum, not the critic but the metacritic, who, by 'an act of relevance', transfigures ordinary acts into expressions of an "implicit theory". There is a time-honored way of dealing with this sort of situation, which is to speak of the purported theory as belonging not to the creator but to the work; that is, to the work itself, so to speak. In this case, we would then speak not of Ziff but of 'Ziff's article', 'the article', etc. This move sounds simple enough, but the tremendous revolt against it in the name of humanism shows just what a delicate and radical achievement it really would be. It is the appropriate hedge against the two unacceptable extremes of (a) saddling the creator of the work with full moral responsibility for the critic's interpretation of it, and (b) complete absolution of the creator from all responsibility for anything at all which may come of the work, directly or indirectly. The rhetoric of the work itself is a hedge against these two worst-cases, and a concession to the intractability of the issue of indirect responsibility.

Without such a hedge, every speech act is either socially meaningless or socially reprehensible. None can be anything in between these extremes, for these are the extremes to which any regime of full social responsibility for speech acts regresses.

What we need is a ritual scapegoat which can be ritually slaughtered without actually harming any living creatures. Similarly, we need 'heroes' whom we can be allowed to 'meet' and honor we can afford to bestow. These are the twin capacities in which the work itself serves.

In the second place, we must note that of the seven characteristics which he mentions, not all are treated by Professor Ziff as being independent of one another; nor are all related to one another in identical ways.

It will be instructive to note some of the differences among their relationships, since it is precisely here that many of the traditional problems of aesthetic theory once again take their rise.

For example,
we are bound to note that Professor Ziff related the seventh characteristic of the Poussin painting to its fourth characteristic:

the fact that it is a good painting is,
he holds,
related to the characteristics which we find that it possesses when we contemplate, observe, and study it.

Its goodness, however, is not claimed to be related to its first, third, or fifth characteristics:

in other words,
Professor Ziff is apparently not claiming that the goodness of this particular work of art depends upon its being a painting rather than being some other sort of work of art which is capable of being contemplated, studied, etc.;

nor is he claiming that its goodness is dependent upon the fact that it was intended to be hung in a place where it can be observed and studied;

nor upon the fact that it is a representational painting which depicts a mythological scene.

If we next turn to the question of how the goodness of this painting is related to the fact that it was "made deliberately and self-consciously, with obvious skill and care by Nicolas Poussin,"

Professor Ziff's position is somewhat less explicit,
but what he would say is probably quite clear.

Suppose that the phrase "obvious skill" were deleted from the description of this characteristic:

would the fact that this painting had been deliberately and self-consciously made,
and had been made with care
(but perhaps not with skill),
provide a sufficient basis for predicating goodness of it?

I should doubt that Professor Ziff would hold that it would, since many bad paintings may be supposed to have been made deliberately, self-consciously, and with care.

I have to defer full consideration of this argument to the occasion of my full consideration of Ziff.

For now:
the

goodness

of the Poussin is nothing more nor less than the final item in Ziff's

description

of it.

Again, it appears to be a description that we

could

give, not one that we must give.

This ascription of goodness indeed

depends upon

nothing named in the given passage. Rather, is is itself a "description" that one might give.

Yet, if this is so, how is the maker's skill related to the object's goodness?

Perhaps the fact that "obvious skill" is attributed to Poussin is meant to suggest that Poussin intended that "The Rape of the Sabine Women" should possess those qualities which Professor Ziff notes that we find in it when we contemplate, study, and observe it in the way in which he suggests that it should be contemplated.

If this is what is suggested by attributing skill to the artist, it is surely clear that Professor Ziff has without argument built an aesthetic theory into his description of the Poussin painting.

That theory is implicit both in the characteristics which he chooses as being aesthetically relevant, and in the relations which he holds as obtaining among these characteristics.

For now, it seems all that has been

suggested

is that the Poussin could be "described" as evidence of

the artist's

skill.

as well as it could be described as possessing

goodness.

Perhaps it also could be described as French, or as objectlike, or as irascible: these could be credits or debits, depending on source and context. It will be argued, perhaps, that "skill" is de facto laudatory, but this too has been shown to be contingent: Downtown Punks and Uptown Experimentalists alike have figured out how to turn this credit into a debit.

If it be doubted that Professor Ziff's description contains at least an implcit aesthetic theory,

consider the fact that in one of the passages in which he describes the Poussin painting
(but which I did not include ... ),

he speaks of the fact that in contemplating, studying, and observing this painting

"we are concerned with both two-dimensional and three-dimensional movements, the balance and opposition, thrust and recoil, of spaces and volumes."

Since the goodness of a painting has been said by him to depend upon the qualities which we find in it when we contemplate, study, and observe it, it follows that these features of the Poussin painting contribute to its goodness.

And I should suppose that they are also included in what Professor Ziff calls the sixth characteristic of the Poussin painting, namely its "complex formal structure."

Thus, presumably, the goodness of a painting does depend, in part at least, upon its formal structure.

On the other hand, Professor Ziff

[225]

never suggests that the goodness of the Poussin painting depends upon the fact that it is a representational painting, and that it has a mythological (or historical) subject matter, rather than some other sort of subject matter.

In fact, when he discusses critics such as Kenyon Cox and Royal Cortissoz, Professor Ziff would apparently—and quite properly—wish to separate himself from them,

rejecting the view that what makes a painting a good painting has any necessary relation to the fact that it is or is not a representational painting of a certain sort.

Thus, Professor Ziff's account of the aesthetically relevant features of the Poussin painting, and his statements concerning the interrelationships among the various features of that painting, define a particular aesthetic position.

The position which I have been attributing to him is one with which I happen to agree.

However, that fact is not of any importance in the present discussion.

What is important to note is that Professor Ziff's characterization of the Poussin painting contains an implicit theory of the nature of a work of art.

According to that theory, the goodness of a painting depends upon its possession of certain objective qualities, that these qualities are (in part at least) elements in its formal structure, and that the artist intended that we should perceive these qualities in contemplating and studying the painting.

(Had he

not

had this intention,

would we be able to say

that he had made the object self-consciously, deliberately, and with skill

?)

I gather that much of 'Wittgensteinianism' is concerned with precisely this kind of question, approached from the linguistic angle. Taking this tack to the extent of my present capabilities, I think that the first two

sayings

are basically tautologies of

intention.

The third,

skill,

is the really interesting case, because in fact it makes no more sense to say this in the presence of "intention" than in its absence. As Malcolm might have put it: artmaking and conoisseurship are 'occasions for intentions', but there is no 'inner process' of artistic intention.

Further, this implicit theory must be assumed to be a theory which is general in import, and not confined to how we should look at this one painting only.

Were this not so, the sort of description of the Poussin painting which was given by Professor Ziff would not have helped to establish a clear-cut case of what is to be designated as a work of art.

For example, were someone to describe the same painting in terms of its size, weight, and insurable value (as might be done were it to be moved from museum to museum), we would not thereby learn
how the term "work of art"

is to be

used.

Right . . . we would learn how it is used!

These are

uses!

And, qua uses, they are perfectly commensurable with Ziff's uses.

In failing to note that his description of the Poussin painting actually did involve a theory of the nature of art,

Professor Ziff proceeded to treat that description as if he had done nothing more than bring forward

a list of seven independent characteristics

of the painting he was examining.

In so doing, he turned the question of whether there are any features

common to all

works of art into a question of whether one or more of

these seven specific indices

could be found in all objects to which the term "work of art" is applied.

Inevitably, his conclusion was negative, and he therefore held that "no one of the characteristics listed is necessarily a characteristic of a work of art."

Sheesh!

How many

specific indices

must we generate and test for? Eight? Seventy? Or what?

However, as we have seen, Professor Ziff's description of the Poussin painting was not actually confined to noting the specific qualities which were characteristic of the pictorial surface of that painting;

it included

references to

the relations between

these qualities and the aim of Poussin,
and

references to

the ways in which a painting

having such qualities

is to be contemplated

by others.

Had he turned his attention to examining these relationships between object, artist, and contemplator,
it would assuredly have been more difficult for him to assert
that

"neither a poem, nor a novel, nor a musical composition can be said to be a work of art in the same sense of the phrase in which a painting or a statue or a vase can be said to be a work of art."²⁰

---

²⁰ ... For example,
Ziff denies that a poem can be said to be
"exhibited or displayed."

Yet
it is surely the case that
in printing a poem or
in presenting a reading of a poem,

the relation

between

the work

and

its audience,

and

the relation

between

artist,   work,   and audience,

is not wholly dissimilar to that which obtains when an artist
exhibits a painting.

If this be doubted,

consider whether there is not a closer affinity between these two cases than there is between a painter exhibiting a painting and a manufacturer exhibiting a new line of fountain pens.

---

In fact,

had he carefully traced the relationships which he assumed to exist among some of the characteristics of the Poussin painting,

he might have found that,
contrary to his inclinations,
he was well advanced toward putting forward explicit generalizations concerning the arts.

The tell here is that the

manufacturer

example is explicitly introduced as illustrating a difference in degree, as if all we needed to do to prove the similarity of painting with poetry is to locate some third term which shares a few (but not too many) similarities with both. But if Ziff's argument is not about the manufacture of fountain pens, then this is not the comparison implicated in his comparison of various arts. Rather, that comparison must be confronted as it is given by Ziff.

What of this?
Can a poem be

exhibited or displayed?

Well, . . . yes and no!

There are some differences. Mandelbaum even cannot deny this much. If we engage him sincerely, then what he says-without-saying here is that he considers those differences to be trivial; and perhaps by implication, Ziff takes them to be nontrivial.

If we arrogate to explain just why we find these differences to be nontrivial, we can expect Mandelbaum to accuse us of "having a theory" per which we have fixated upon certain implications ("generalizations"!) at the expense of others.

To shift the ground of debate to that of

relations

is to shift to a more general level. This is how the differences in "exhibition" of painting and poetry can be made to look trivial. But this also eliminates most (I might argue all) of the differences between art-objects and consumer goods.

III

While Professor Ziff's argument against generalization depends upon the fact that the various artistic media are significantly different from one another,

the possibility of generalizing concerning the arts has also been challenged on historical grounds.

It is to Morris Weitz's use of the latter argument that I shall now turn.

In "The Role of Theory in Aesthetics" Professor Weitz places his primary emphasis on the fact that art forms are not static.

From this fact he argues that it is futile to attempt to state the conditions which are necessary and sufficient for an

[226]

object to be a work of art.

What he claims is that the concept "art" must be treated as an open concept,
since new art forms have developed in the past,
and since any art form (such as the novel) may undergo radical transformations from generation to generation.

One brief statement from Professor Weitz's article can serve to summarize this view:

What I am arguing, then, is that the very expansive, adventurous character of art, its ever-present changes and novel creations, makes it logically impossible to ensure any set of defining properties.

We can, of course, choose to close the concept.

But to do this with "art" or "tragedy" or portraiture, etc. is ludicrous since it forecloses the very conditions of creativity in the arts.

Unfortunately, Professor Weitz fails to offer any cogent argument in substantiation of this claim.

The lacuna in his discussion is to be found in the fact
that
the question of

whether a particular concept is open or closed

(i.e., whether a set of necessary and sufficient conditions can be offered for its use)
is not identical with
the question of

whether future instances

to which the very same concept is applied

may or may not possess

genuinely novel properties.

Is this to say that

novel properties

can always be added to

necessary and sufficient

ones?

If so, then all well and good.
But why the easy presumption that

the very same concept is applied

in such cases?

The concern seems to be that closing a concept just does mean that it would not be applied beyond a certain (now arbitarily stipulated) degree of "novelty".

In other words,
Professor Weitz has not shown that

every novelty

in the instances to which we apply a term
involves

a stretching of the term's connotation.

Showing this NOT to be the case seems a much more difficult road to hoe. Either way, by this time we have made a hash of the whole question of apply ing a term. If a term IS applied, then there is no novelty to contend with. To challenge an application is not to challenge a stretching of connotation , it is to challenge the boundary of the concept itself.

By way of illustration,
consider the classificatory label
"representational painting."

One can assuredly define this particular form of art without defining it in such a way that it will include only those paintings which depict either a mythological event or a religious scene.

Historical paintings, interiors, fête-champêtres, and still life

can all count

as "representational" according to any adequate definition of this mode of painting,

and there is no reason why such a definition

could not have been formulated

prior to

the emergence of any of these novel species of the representational mode.

Well, now Our Man too has

turned the question of

whether there are any

features common to all works of art

into

something else,
namely into

a question of

whether one or more of these

above-given predicates

could

have been unwittingly pre-accommodated by some unwitting definer.

Indeed, this "could" happen. So far, so good.

Would it happen? That is the actually important "question", and it is the one we cannot possibly answer.

Thus,
to define a particular form of art
—and to define it truly and accurately—
is not necessarily to set one's self in opposition to whatever new creations may arise within that particular form.²²

This is circular and vacuous.

---

²² To be sure,
if no continuing characteristic is to be found,

the fact of change will demand that the concept be treated as having been an open one.

This was precisely the position taken by Max Black in a discussion of the concept "science."

(See "The Definition of Scientific Method," in Science and Civilization, edited by Robert C. Stauffer [Madison, Wisconsin, 1949].)

Paul Ziff refers to the influence of Professor Black's discussion upon his own views, and the views of Morris Weitz are assuredly similar.

However,
even if Professor Black's view of the changes in the concept "science" is a correct one
(as I should be prepared to think that it may be),
it does not follow that the same argument applies in the case of art.

Nor does the fact that the meaning of "science" has undergone profound changes

in the past

imply that further analogous changes will occur

in the future.

---

. . . , nor that they won't!

I have grabbed only a small chunk from Black:

[78]

...

Consider, if you will, the vast variety of activities of a scientific character which have occurred in the day now nearly past―the glass blowing and the dissecting, the manipulation of rulers, stop watches, test tubes, ... ; men fishing in swamps, solving differential equations, ... ; handling, manufacturing, observing, ... Is not the resulting impression one of the extreme diversity, not to say heterogeneity, of the activities which we are naturally inclined to regard as scientific? Yet there is something more than a mere aggregate here; we know that this apparent mesh of activities ...

[79]

... is unified by an extremely fine network of relationships. There is a pattern, but an extremely complex one.

The activities I have distinguished are not conducted independently and in isolation; ... calculation is performed for the sake of experimental and observational test; experiment is conducted in the service of generalization, which in turn uses theory, which provokes speculation, ... a maze of cross-connections and mutual dependencies. Science is an organic system of activities; and the pattern of its development is also organic.

... To do justice to our subject we must extend our survey in imagination to cover the history and development of scientific activity no less than its present condition. We shall then see that this vast symphony of activities displays superordinate rhythms of development and change; there will be brought vividly to our notice the striking variety of motives and circumstances which have fostered or hindered the progress of science, ... For sheer complexity of texture and incident, science is like life itself and as little to be reduced to formula.

... it may be [as "some scientists" hold] that the

[80]

very remote past consists largely of mistakes to be avoided. But it deserves to be remembered that the history of any scientific discipline intimately determines the current modes of investigation. ... To pretend otherwise is to claim for human reason, as manifested in scientific progress, a universality and fixity it has never manifested.   ...

...   None of the characters which we recognize in the scientific process are independently necessary or sufficient, but all supporting and jointly reinforcing one another give rise to the unique historical phenomenon.

...

[81]

... Astronomy makes no experiments, mathematics uses no observation, geography is mainly descriptive, [etc.] ... The characters mentioned ... may be present in higher or lower degree ... Their diminution removes from an activity the features we apprehend as scientific; their joint presence in high degree creates conditions recognized as pre-eminently scientific.

This line of thought will lead us to abandon the search for a timeless and immutable essence in favor of the identification of a system of overlapping and interacting criteria. I propose to call this "definition of concrescence." ... the simultaneous actualization of a number of mutually reinforcing characters, all of them capable of variations in degree. ... I am proposing ... that we treat "scientific method" as a historical expression meaning, among other things, "those procedures which, as a matter of historical fact, have proved most fruitful in the acquisition of systematic and comprehensive knowledge." On this approach,

[82]

the methodological problems involved in the definition of scientific method closely parallel those arising in an attempt to define Napoleon, the industrial revolution, slavery, or any other person or institution having historical actuality. In each such case, what we recognize as theidiosyncracy of the unique historical phenomenon is constituted by a growing together, a concrescence, of variable factors, interacting to produce the degree of unification and contrast with an environment which leads us to recognize a distinct entity.

...   To provide a satisfactory definition by concrescence we shall need (a) a description of the main factors engaged in the concrescence, (b) determination of their relative "weight" or importance, and (c) an account of their mode of interaction.

Great news!

There is a pattern,

and it is an

organic

one.

Yes, organic

like life itself

The connecting up of this 'organicism' with 'historicism' is, to me at least, equal parts novel and suspicious. Such notions as "men make their own history, but they do not make it as they please" begin to lose their luster once the ground of "history" is granted to be so "extremely complex" as to preclude being "reduced to formula". If the range of historically-conditioned outcomes turns out to be so complex that we cannot hope to take inventory of it, then calling these outcomes historically-conditioned is not really saying anything.

The notion of

mutual dependencies

among various scientific

activities

seems more worth taking seriously.

Calling this

a historical phenomenon

or a

network of relationships

seems diversionary and subordinate to everything that follows from the fact that these are "activities", doings, acts; and that they do, in a manner of speaking, "depend" upon each other.

In what manner is this D-word spoken?
It is the manner of an F-word, one which is conspicuous by its absence here.

A 'function', so-called, certainly has a 'story', in the sense that almost anything does. It has a past, but it does not make for much of a story.

Supporting and jointly reinforcing

is not much of a story.

But the

relational

tinge is baked right into the 'concepts' of support and reinforcement; and I would add (insist) that the elements of boundedness and necessity inherent in the concept of a function just are the elements which make the ascription of relationality sensible.

Without boundaries or necessities, 'relations' may be multiplied endlesslessly and whimsically. This has been called the

"That reminds me of . . ."

method, and of course we prefer to steer clear of it.

Whither art? Art too may have 'functions'. If it no longer has them, this would bail us out of having to consider analogies with the above; but this notion cannot be sustained. Ironically enough, the assumption that decadent art is not functional art betokens (what else?) the premature "closing" of two "historical" concepts at once; as if the functions of art were burdens from which freedom and abundance have released us. Hardly!

Really the victory over material necessity has intensified our hunger for an art which has what might be called an 'organic collective ritual function'. Recently I am forced to the conclusion that this is most of what's going on in what we call art. It has not been 'functioning' very well as of late; but that is hardly the same thing as not functioning at all.

I want to say that art cannot help but function this way, no matter how good or bad things get. But of course this is no kind of definition: it manages to be both over- and under-inclusive along various lines . . . just like always before. Black's descriptions above do seem, generally, to fit the case of art, too.

" I am proposing ... that we treat "scientific method" as a historical expression meaning, among other things, "those procedures which, as a matter of historical fact, have proved most fruitful in the acquisition of systematic and comprehensive knowledge." " (81) This seems to smuggle in a definition!

Consequently, it would be mistaken to suppose that all attempts to state the defining properties of various art forms are prescriptive in character and authoritarian in their effect.

This conclusion is not confined to cases in which an established form of art, such as representational painting, undergoes changes;

it can also be shown to be compatible with the fact that radically new art forms arise.

For example,
if the concept "a work of art" had been carefully defined prior to the invention of cameras,
is there any reason to suppose that such a definition would have proved an obstacle to viewing photography or the movies as constituting new art forms?

Hmm.

Yes,

there is

ample

reason to suppose

this!

He seems to be saying that, if only we had proceeded

carefully

enough, the resulting definition would have been able to accommodate

new art forms,

provided of course that these forms really were art. This just is what a good definition would do; if it does not do this, then it never was a satisfactory definition.

No art-properties better demonstrate the inadequacy of this view than do so-called "relational properties". Can we not find plenty of resistance to early Mechanical Photorealism which hinges precisely upon the argument that creators and viewers alike simply cannot have the same "relationship" with a photograph as they have with a portrait?

Do we not find here, too, that admitting consideration of "relevance" and "matters of degree" quickly makes the question intractible? If we look hard and close enough, we will find relational differences; and then, we will find some people who for whom these are nothing, along with others for whom they are everything.

Contrary to Black's view of science, we find Mechanical Photorealism at its height of "contrast with [the] environment" at the very beginning of its "history", after which this contrast is progressively attentuated.

To be sure, one can imagine definitions which might have done so.

However, it was not Professor Weitz's aim to show that one or another definition of art had been a poor definition;

he wished to establish the general thesis that there was a necessary incompatability, which he denoted as a logical impossibility, between allowing for novelty and creativity in the arts and stating the defining properties of a work of art.

He failed to establish this thesis since he offered no arguments to prove that new sorts of instantiation of a previously defined concept will necessarily involve us in changing the definition of that concept.

To be sure,
if neither photography nor the movies
had developed along lines which

satisfied the same sorts of interest

that the other arts satisfied,

and if

the kinds of standards

which were applied in the other arts
were not

seen to be relevant

when applied to photography and to the movies,
then
the antecedently formulated definition of art

would have

functioned as a closed concept,
and it

would have

been used
to exclude all photographers and all motion-picture makers from the class of those who were to be termed "artists."

However, what would the defender of the openness of concepts hold that one should have done under these circumstances?

Suppose, for example, that
all photographers had in fact been the equivalent of passport photographers,

and that they had been motivated by no other interests and controlled by no other standards than those which govern the

[227]

making of photographs for passports and licenses:

would the defender of open concepts be likely to have expanded the concept of what is to count as an art in order to have included photography?

The present inclusion of photography among the arts is justified,
I should hold,
precisely because photography arises out of the same sorts of interest,
and can satisfy the same sorts of interest,
and our criticism of it employs the same sorts of standards,
as is the case with respect to the other arts.

Bearing this in mind,
we are in a position to see that still another article which has sometimes been cited by those who argue for the openness of the concept "a work of art" does not justify the conclusions which have been drawn from it.

That article is Paul Oskar Kristeller's learned and informative study entitled "The Modern System of the Arts."

The way in which Professor Kristeller states the aim of his article suggests that he too would deny that traditional aesthetic theory is capable of formulating adequate generalizations concerning the arts.

He states his aim in saying:

The basic notion that the five "major arts" constitute an area all by themselves, clearly separated by common characteristics from the crafts, the sciences and other human activities has been taken for granted by most writers on aesthetics from Kant to the present day. . . .

It is my purpose to show
that this system of the five major arts,

which underlies all modern aesthetics and is so familiar to us all, is of comparatively recent origin and did not assume definite shape before the eighteenth century,

although it had many ingredients which go back to classical, mediaeval, and Renaissance thought.

However,
the fact that the classification of the arts has undoubtedly changed during the history of Western thought,
does not of itself suggest that aesthetic theory must undergo comparable changes.

Should this be doubted,
one may note that Professor Kristeller's article does not show in what specific ways

attempts to classify or systematize the arts

are integral to, or
are presupposed by, or
are consequences of,

the formulation of an aesthetic theory.

This is no minor cavil,
for if one examines the writers on aesthetics who are currently attacked for their attempts to generalize concerning the nature of art,
one finds that they are not (by and large) writers whose discussions are closely allied to the discussions of those with whom Kristeller's article was primarily concerned.

Furthermore, it is to be noted that Kristeller did not carry his discussion beyond Kant.

This terminal point was justified by him on the ground that the system of the arts has not substantially changed since Kant's time,

However, when one recalls that Kant's work is generally regarded as standing near the beginning of modern aesthetic theory
—and surely not near its end—
one has reason to suspect that questions concerning "the system of the arts" and questions concerning aesthetic theory constitute distinct, and probably separate sets of questions.

A survey of recent aesthetic theory bears this out.

Since the time of Hegel and of Schopenhauer there have been comparatively few influential aesthetic theories which have made the problem of the diversity of art forms, and the classification of these forms, central to their consideration of the nature of art.

For example, the aesthetic theories of Santayana, Croce, Alexander, Dewey, Prall, or Collingwood cannot be said to have been dependent upon any particular systematic classification of the arts.

In so far as these theories may be taken as representative of attempts to generalize concerning the arts, it is strange that current attacks on traditional aesthetics should have supposed that any special measure of support was to be derived from Kristeller's article.

There's hardly a word of this paragraph I wouldn't disagree with, but it has been so long since I first read Kristeller that I might as well never have read it at all, hence I defer to a closer consideration of that article.

Should one wish to understand why current discussions have overlooked the gap between an article such as Kristeller's and the lessons ostensibly derived from it,

an explanation might be found in the lack of concern evinced by contemporary analytic philosophers for the traditional problems of aesthetic theory.

For example, one looks in vain in the Elton volume for a careful appraisal of the relations between aesthetic theory and art criticism, and how the functions of each might differ from the functions of the other.

File this with Wolterstorff's claim that Adorno and Beardsley could not have theorized about a memorial concert.

A

[228]

striking example of the failure to consider this sort of problem is also to be found in John Wisdom's often cited dicta concerning "the dullness" of aesthetic theory.²⁷

---

²⁷ See "Things and Persons," Proceedings of the Aristotelian Society, Supplementary Volume XXII (1948), pp. 207-210.

---

In examining his views one finds that the books on art which Wisdom finds not to be dull are books such as Edmund Wilson's Axel's Castle, in which a critic "brings out features of the art he writes about, or better, brings home the character of what he writes about."

In short, it is not theory—it is not aesthetic theory at all—that Wisdom is seeking: he happens to be interested in criticism.

It's news to me that Wisdom rather than Passmore is the most

often cited

to this effect.

This paper turns out to be the third of three contributions to a symposium on utilitarianism. At the risk of failing to respect that context, the remarks on criticism are indeed interesting enough on their own to be worth a second look.

I do not wish to be taken as denying the importance of criticism,
nor as belittling the contribution which a thorough acquaintance with the practice of criticism in all of the arts may make to general aesthetic theory.

However,
it is important to note that

the work of any critic

presupposes

at least an implicit aesthetic theory,

which—as critic—it is not his aim to establish or, in general, to defend.

This fact can only be overlooked by those who confine themselves to a narrow range of criticism:

for example, to the criticism appearing in our own time in those journals which are read by those with whom we have intellectual, political, and social affinities.

When we do not so confine ourselves,
we rapidly discover that
there is, and has been,

an enormous variety in criticism,

and that
this variety represents
(in part at least)

the effect of differing aesthetic preconceptions.

To evaluate criticism itself we must,
then,
sometimes undertake to evaluate these preconceptions.

In short,

we must do aesthetics ourselves.

And then,

to evaluate

this evaluation, . . .

. . . , and so on . . .

I think we can see, by now, that criticism is not to be "evaluated" in this way, and not only because this threatens to set off an infinite loop of meta-evaluation. For consumers of criticism to

do aesthetics ourselves

is not to parse criticism but rather to deconstruct it. In other words, the entire conceit of the institution is for all

preconceptions

("aesthetic" and otherwise) to remain safely under the surface.

In one sense, this is how all formal academic discourse has to work, so as to avoid regressing to litigation of first principles. The specter of

self-confinement

to

intellectual, political, and social affinity

groups has of course turned out to be a far more serious problem than anyone of Mandelbaum's generation could have anticipated; nonetheless, it is a bit of a Red Herring in the present discussion. Relativizing "manifest" arguments to various background assumptions doesn't explain those arguments, it explains them away.

There is, fittingly, a pedagogical 'meta' level here to accompany and complement the discursive one: Mandelbaum's admonishment here is highly emblematic of the ascent of so-called 'critical thinking'. No longer is it enough to learn what a critic has to say; now we are charged with "evaluating" it; and it is made clear enough, i.e. by mention of "a narrow range", "affinities", "confine[ment]", etc., that this can only amount to a deflation by way of relativization.

It has been argued (convincingly, to my mind) that the dethroning of rote learning in favor of 'critical thinking' has proven to be a terrible mistake, purely from the standpoint of educational outcomes; but of course it's even worse than that: the very charge to enter imaginatively into others' arguments is neutered by this fixation on "evaluation" of these others' alleged "preconceptions".

However, for many of the critics of traditional aesthetics this is an option which does not appeal.

If I am not mistaken, it is not difficult to see why this should have come to be so.

In the first place,
it has come to be one of the marks of

contemporary analytic philosophy

to hold that philosophic problems are

problems which cannot be solved by

appeals to matters of fact.

Thus,
to choose but a single instance,
questions of the relations between aesthetic perception and other instances of perceiving

—for example, questions concerning psychical distance, or empathic perception, or the role of form in aesthetic perception—

are not considered to be questions with which a philosopher ought to try to deal.

In the second place, the task of the philosopher has come to be seen as consisting largely of the unsnarling of tangles into which others have gotten themselves.

As a consequence,
the attempt to find

a synoptic interpretation of some broad range of facts

—an attempt which has in the past been regarded as one of the major tasks of a philosopher—

has either been denigrated or totally overlooked.²⁹

---

²⁹ For example, W. B. Gallie's "The Function of Philosophical Aesthetics," in the Elton volume, argues for

"a journeyman's aesthetics,"

which will

take up individual problems,

one by one,

these problems being of the sort which arise

when a critic or poet gets into a muddle

about terms such as "abstraction" or "imagination."

For this purpose the tools of the philosopher are taken to be

the tools of logical analysis

(op. cit., p. 35);

a concern with

the history of the arts,

with

psychology,

or

a direct and wide-ranging experience of the arts

seems not to be presupposed.

A second example of the limitations imposed upon aesthetics by contemporary linguistic analysis is to be found in Professor Weitz's article.

He states that

"the root problem of philosophy itself is to explain the relation between the employment of certain kinds of concepts and the conditions under which they can be correctly applied" (op. cit., р. 30).

---

Therefore,
problems such as the claims of the arts to render a true account of human character and destiny,
or questions concerning the relations between aesthetic goodness and standards of greatness in art,
or an estimate of the significance of variability in aesthetic judgments,
are not presently fashionable.

And it must be admitted that if philosophers wish not to have to face either factual problems or synoptic tasks, these are indeed questions which are more comfortably avoided than pursued.

The Johns Hopkins University

---