Works of Art

Generalising across the arts is a tricky business. Can we really expect to find anything in common between, say, Ulysses, Der Rosenkavalier, the ‘Donna Velata’ and Donatello’s St George in virtue of which they are all works of art? As if that were not hard enough, try adding prints, films, dances and buildings and the problem becomes intractable. Yet traditionally the aim of aesthetics has been to undertake an abstraction from the properties of particular works of art, and of different forms of art, with precisely the hope of isolating just those general features which are supposed to characterise or define the nature of art itself. To discover defining, or even characteristic, properties of art, if such there be, presupposes an answer to even more basic, ontological, questions concerning what sorts of things or entities works of art are. Are they physical objects, ideas, universals, classes, or what?

There is perhaps no better guide to the dauntingly complex issues involved in these questions than Richard Wollheim’s Art and Its Objects. First published over twelve years ago, this concise, elegant and wide-ranging book has established itself as an indispensable text for undergraduate courses in aesthetics. The second edition leaves the original text unchanged but adds a helpful analytical contents, summarising the argument, as well as an extended and up-to-date bibliography and six Supplementary Essays on related topics.

The content of the original Art and Its Objects falls roughly into two parts. The first investigates ontological questions, examining as its starting-point the hypothesis that at least some works of art are physical objects; the second explores the implications of the ‘amorphousness’ of the concept of art, from which the central thesis emerges that art is a ‘form of life’. The sections on ontology are unsurpassed in their clarity and incisiveness and many of Wollheim’s conclusions have become the orthodoxy. He carefully takes us through the arguments against identifying, say, the novel Ulysses either with any one copy of it or with the class of all copies or even with Joyce’s manuscript. Literary and musical works of art are not physical objects, or even classes of such objects, but, on Wollheim’s thesis, are ‘types’. Each copy or each performance is a ‘token’ of the respective ‘type’. These are technical terms, easier to illustrate than to understand. So, for example, the sentence ‘The cat sat on the mat’ contains six word-tokens but only five word-types; there are two tokens of the type ‘the’. Types and tokens have some but not all properties in common. My copy of Ulysses might weigh 13 oz. but that is not a property of Ulysses itself; however, both token and type share the property of having been read by me.

In contrast to those works of art that are types, like novels, poems, ballets, operas and symphonies, there are others which Wollheim characterises as ‘individuals’, such as paintings and sculptures. But even with individuals it is an open question whether the works of art themselves, the ‘Donna Velata’ or the St George, are identical in all respects with some physical object, paint and canvass, marble or stone. In the text of Art and Its Objects Wollheim considers several objections that might be raised to making a complete identification between an individual work of art and a physical object. Are there not some properties, representational or expressive, for example, which are possessed by some individual works of art but could not be possessed by any physical object? And what about more radical theories, such as those of the Idealist school associated with Benedetto Croce and R.G. Collingwood, which identify all works of art with some inner state of mind of the artist? Wollheim argues that there is nothing in these objections that forces us to conclude that no works of art can be physical objects. But likewise he is careful to stress that there is nothing that forces us the other way, making us identify even some works of art with physical objects.

Wollheim returns to these issues in the Supplementary Essays. For example, in Essay III he looks again at the ‘physical object hypothesis’, but, perhaps with excessive caution, still cannot commit himself either way. The nagging worry about physical objects is that they have properties which are unwelcome as properties of works of art. For example, they age, fade and decay. This provides some motivation for positing an ‘aesthetic object’ perhaps possessing only aesthetic properties, or identified only with some optimal state of the physical object. Wollheim is sceptical of the distinction between aesthetic and non-aesthetic properties, but he is tempted by the latter suggestion. Yet it, too, is not quite right: don’t we think of works of art as corruptible? Essay V further clears the way for the physical object hypothesis, with regard to representational properties. In the main text, Wollheim argues that looking at (representational) pictures involves seeing X, the medium or representation, as Y, the object or what is represented. He now prefers the notion of seeing Y in X, partly to allow for the fact that we can see X and Y at the same time. This is surely right: we do not cease to see the paint when we perceive the figure painted.

In Essay II, Wollheim puts pressure on the ontological distinctions, type / token / individual. Suppose a poet in the 20th century quite unknowingly produced an identical set of words to that of a 16th-century poet. Do we have two poems or just one? How many types and tokens are there? Wollheim sidesteps this one, but insists that a constraint on any correct answer be that the identity of all works of art depends essentially on their history of production. This identity condition is a theme of several of the Essays. Then again, suppose that technological innovations made it possible to reproduce paintings with almost total fidelity to the original. Would not pictures then become types rather than individuals? And would there be any aesthetic significance in this change? Wollheim agrees that pictures could in these circumstances be reclassified as types. But he argues in detail that because of the consequent changes in how an artist would conceive of his work the change of ontological category would have an aesthetic, as well as merely taxonomic, relevance.

The second part of the original Art and Its Objects develops the thesis that ‘art is a form of life.’ The parallel here is with Wittgenstein’s claim that language is a ‘form of life’, a claim intended to draw a contrast with atomistic theories of language in which meaning is viewed either as a relation between individual words and items in the world or as a mental experience accompanying the use of a word. For Wittgenstein, words have meaning only in virtue of having a function in human practices. We understand the meanings of words only by understanding the relevant practices. Wollheim draws instructive parallels between art and the Wittgensteinian conception of language, both from the standpoint of the artist and from that of the spectator of art. The artist’s creativity and the spectator’s understanding of a work of art can only take place against the background of a ‘form of life’ and cannot be explained in terms of any simple atomistic relation between a person and an object. Art comprises a complex institution or practice which makes certain activities possible. As Wollheim observes, we can reflect on the ‘pervasiveness’ of art by speculating on what would be different if art were to disappear from the world: we should lose not just works of art and artists but such amorphous items as artistic creativity and artistic appreciation.

The idea that art is a ‘form of life’, on the Wittgensteinian conception, is now widely accepted. But things have moved on since the programmatic suggestions of Art and Its Objects in 1968. The notion that certain human actions and concepts can be explained through their role in institutions or practices has played a major part in philosophical theories of action and of language over the last decade. It would have been instructive to have seen Wollheim developing the ‘form of life’ conception in the light of this recent work. Instead, he offers only critical remarks (Essay I) about one, as he sees it abortive, attempt to explain art in ‘institutional’ terms. This is the proposal, asssociated with George Dickie, that art can be defined in terms of the conferring of status on objects by appropriate members of a social institution, called the ‘art-world’. Wollheim takes a poor view of the theory and poses for it a dilemma: either it is not an institutional theory of art, if it requires that there be independent (good) reasons for conferring the status of art on an object, or it is not an institutional theory of art, if it allows the conferment to take place for no good reason at all. The points are telling, though the notion of ‘good reason’ might well be explained from within the theory. The lack of any follow-up to Wollheim’s own ‘form of life’ view is all the more disappointing since it looks like an embryonic, though different, institutional view itself.

Of the remaining essays, one (Essay IV) argues that ‘the task of criticism is the reconstruction of the creative process.’ There is much of interest here – though, as with many of the essays, this has more to do with the negative points in anticipation of objections than with positive descriptions of how this reconstruction might operate. The essay on ‘Art and Evaluation’ (Essay VI) is curiously disappointing, never advancing beyond the most basic preliminaries. An opportunity was missed to draw together the suggestions of the main text and the new ideas from the essays on the identity of works of art and on criticism: for together they almost palpably yield a conception of artistic value. Nevertheless, the essays continue the high standard set by the original Art and Its Objects, and this second edition must further endorse Wollheim’s reputation as a major contributor to aesthetics.

Nicholas Wolterstorff, in Works and Worlds of Art, begins where Wollheim leaves off. His central topic is that of representation in the arts and he draws his theoretical framework from the philosophy of action and from ‘speech act’ theories of language. Perhaps as a consequence, the book is written in a more austere, and technical, idiom than Art and Its Objects. However, the main theses are readily comprehensible and many of the details are fascinating and provocative. For example, we are offered finely detailed analyses of what it is to compose a musical work and to perform one; also, of what is and is not part of a fictional world. Are Huck and Jim really having a homosexual relationship, as Leslie Fiedler suggests, in the world of Huckleberry Finn? Wolterstorff has a keen eye for the logical conundrum. Suppose Henry Kissinger were to play the lead role in a play about Henry Kissinger. Would we have the same relation there between actor and character as we have between, say, an actor and Willie Loman in Death of a Salesman? And what is the relation anyway? Or how can impossible objects, like the Escher buildings, be represented in pictures?

The pivotal idea in the book is that of using an art artefact to perform the action of world projection. The ‘action of world projection’, for Wolterstorff, ‘constitutes the heart of representation’, which in turn he equates, without further explanation, with ‘mimesis’. The novelty of the book is the analysis of representation as a human action, rather than as a relation between a symbol and what it symbolises. So, for example, the action of Rembrandt’s creating the painting ‘Bathsheba’ counts as the action of Rembrandt’s projecting a world which includes a woman bathing. In its most general form, to project a world is to ‘introduce’ states of affairs (or propositions, which for Wolterstorff are the same), either with words or with a picture or in other ways, at the same time adopting some ‘mood stance’ towards these states of affairs – for example, an assertive or fictive or interrogative stance.

It follows that in writing this review I am ‘projecting a world’ and that world projection is as commonplace among historians, philosophers and gossips as among artists and novelists. Wolterstorff strives always for generality in his definitions and I think in the end this weakens the specific contribution he hopes to make to aesthetics. For it turns out that world projection is neither a necessary nor a sufficient characteristic for identifying works of art. Not all world projection is artistic and not all works of art involve world projection. Furthermore, while the notion seems to fit novels, dramas, films and pictures, Wolterstorff leaves us guessing as to how it might be applied to the more difficult cases such as music or sculpture or dance or architecture: yet he claims ‘it is to be found in all the arts, though not, indeed, in every work of art.’

Like Wollheim, Wolterstorff has a great deal to say about the ontology of works of art, though again the difficulty of generalising across the arts crops up. For example, he distinguishes between ‘occurrence-works’, those that can be performed, and ‘object-works’, those which yield impressions, castings, or examples. But from the start the distinction is uncomfortable. Wolterstorff soon concedes that literary works and films can belong to both categories and even drama and music, for which the ‘occurrence’ category seems to be tailor-made, have to be legislated out of the other category, on the stipulation that ‘the drama has no copies’ and that music is composed of sounds, not notes. Paintings belong to neither category but are considered, without much argument, to be physical objects. Somewhat gratuitously, only occurrence-works and object-works are named as ‘art works’, as distinct from ‘works of art’, reserved as a more general term which includes paintings. All this is unnecessarily ragged and no improvement on Wollheim’s distinction between individuals and types.

A more fruitful contribution to ontological issues comes in the thesis that art works are ‘norm-kinds’ and the notion of a kind plays a central part in the book. A kind is identified by the properties essential within it and a kind exists only in as much as the associated properties exist. In turn, a property, say ‘F’, exists only in as much as either something is F or something is not F. Needless to say, kinds are abundant. In composing a musical work, a composer ‘selects a set of properties required for correctness’. These are properties of sounds: the work thus composed is the norm-kind characterised by those properties. A curious consequence of these definitions is that ‘to compose is not to bring into existence what one composes. It is to bring it about that something becomes a work.’ The kind that characterises Beethoven’s Fifth Symphony existed while dinosaurs roamed the earth. Beethoven’s genius was selection, not creation. I wonder if this paradox could not be removed by distinguishing the existence of properties individually from the existence of complex combinations of properties. After all, not all kinds need to be ‘natural kinds’.

The longest section in the book is entitled ‘Worlds’ and deals with fictions and literary works. On this topic, Wolterstorff is at his best, not shying away from any of those perennially elusive problems about the nature of fictional characters and fictional states of affairs. Nearly everything here is illuminating and thought-provoking. He resists the fashionable equation of fictional worlds with ‘possible worlds’, pointing out that fictional worlds are by their nature indeterminate in ways that the logician’s ‘possible worlds’ cannot be. There are no answers to Keats’s questions:

Who are these coming to the sacrifice?
To what green altar, O mysterious priest,
Leads’t thou that heifer lowing at the skies ...?

Also, by careful application of the distinction between what is included in a fictional world and what is true in the actual world, he explains, for example, what it means to say that I (in the actual world) can see Hedda Gabler (in the fictional world) or how Napoleon can make an appearance in War and Peace.

Perhaps most interesting of all is the thesis that fictional characters are ‘person-kinds’ – a theory for which Wolterstorff claims Aristotle as an ancestor. We have a pervasive habit in our ordinary speech of talking of fictional characters as if they were real people. Isn’t Sherlock Holmes the most famous detective that most of us have heard of? But it takes a logician to sort out the Pickwickian tangles of what we really mean when we talk so freely of fictions. Can we refer to Sherlock Holmes in the way that Watson can? No, because we are referring to a character, he is referring to a man. Conan Doyle does not refer to either of them, he ‘fictionally projects’ them. A character, say, Chichikov, is a ‘person-kind’ – in this case, the Chichikov-in-Dead Souls kind; it is a kind having, essentially, all those properties indicated of it by Gogol in the world of Dead Souls. What about characters, like Faust, who seem to appear in different fictions with slightly different properties? Wolterstorff suggests that we distinguish the Faust character simpliciter from the Faust-in-Goethe’s-Faust character or the Faust-in-Marlowe’s-Dr Faustus character. All three are kinds, but the former has fewer essential properties than the latter two. Calling characters kinds is not to undermine their uniqueness or individuality. There is no implication that characters are stereotypes or that the kind must be exemplified elsewhere. The main problem with this view is in identifying what the kinds are. What are the essential properties of Anna Karenina or Dorothea? Won’t that be endlessly debated?