Are Works of Art in the Mind?

 A Response to Collingwood’s Theory of Art

 

By Robin Phillips March, 2003

 

 

In this essay I want to evaluate the claim that works of art exist in the mind. I will begin by considering Collingwood’s idea about what art is and how this leads him to the position that works of art exist only in the mind. I will explain why I think this idea is problematic and then go on to consider an idea that provides an alternative to Collingwood’s position.

Collingwood offers a positive definition of art in terms of the expression of emotion. Before the artist expresses his emotions, the emotions exist in the form of inchoate feelings. That is to say, the artists feels something but is not quite sure what it is he feels. As the artist engages in artistic creation his feelings gradually become clarified. So, writes Collingwood, “When creating for ourselves an imaginary experience or activity, we express our emotions and this is called art.”

Although this definition could easily be criticized on the grounds of necessary and sufficient conditions, it is sufficient for our purposes simply to consider the three central aspects to a work of art that are given here: (A) expression; (B) creation; (C) the imaginariness of the artistic object. (I have listed them in this order since, as we shall shortly see, this seems to be the order of logical priority in the outworking of Collingwood’s argument.)

Since, on Collingwood’s view, “Expression is an activity of which there can be no technique”[1], he is compelled to discuss artistic creation in terms that distinguish it sharply from craft. He sets forth a number of differentiations between art and craft, the primary ones being that art does not involve a distinction between means and end, nor does it operate according to a preconceived plan, nor does it build out of pre-existing material. He thus draws a distinction between ‘making’ and ‘creating’, the former being relevant to craft, and the later referring to something that is non-technical and arises ex nihilo. It is this sense of creation that Collingwood believes to be involved when an artist produces something. As he writes,

 

Art has something to do with making things, but these things are not material things, made by imposing form on matter, and they are not made by skill.[2]

 

So, positively speaking, of what does artistic creation consist? According to Collingwood it consists of an activity that occurs in the mind of the artist by which his emotions are expressed and transmitted to the minds of others. This brings us to the third facet of Collingwood’s definition of art, namely the imaginariness of the artistic object. By imaginary Collingwood means that works of art do not belong to the category of what is real[3], but “‘exist in a person’s head’ and nowhere else…”[4] Though the artwork may be externalised, this is secondary and it is not a necessity for the work to be fully finished and complete.

Collingwood attempts to explain this with a curious parallel between works of art and bridges. Collingwood points out that in the case of a bridge, the bridge may be said to exist in the engineer’s mind before the plan is committed to paper and before the bridge is built. At this stage, the bridge is imaginary, according to Collingwood. When the bridge is built, however, “we say that it exists not only in his head but in the real world.”[5] It comes as a surprise when Collingwood says that “The same distinction applies to such things as music”, given that he is trying to establish that works of art do not exist in the real world. As Collingwood’s explanation (for we had best avoid the term ‘argument’) unfolds, however, we discover that the main function of his bridge analogy is actually to contrast it with works of art. Collingwood lays the ground for this contrast by discussing the two stages involved in making a work of art, using the musical arts as his paradigm.

The first stage is when the artist/composer creates the work in his head. As Collingwood writes, “The actual making of the tune is something that goes on in his head, and nowhere else.”[6] The composer may then go on to a second stage in which the piece will either be performed, hummed, written down, etc.. Collingwood acknowledges that this second stage may at first seem to resemble the construction of a bridge. However, he thinks that this second stage should actually be seen as an accessory to the first stage (admittedly a very useful accessory!), since

 

The noises made by the performers, and heard by the audience, are not the music at all; they are only the means by which the audience, if they listen intelligently (not otherwise), can reconstruct for themselves the imaginary tune that existed in the composer’s head.[7]

 

Collingwood goes on to deny that creating a tune and then notating it is comparable to making a plan for a bridge and then building it, and he also denies that it is comparable to the relation between plan and bridge. (By ‘plan’ he means “a kind of thing that can exist in a person’s mind.”[8]) Instead he suggests that creating a tune and then notating the tune is comparable to the plan for the bridge and the specifications and drawings. Although this seems straight forward enough, one wonders why Collingwood stops here and does not refer to the physical acts of humming, singing or performing (which he had earlier acknowledged as belonging to the second stage of composing), seeing as these things seem to provide the real parallel with physical bridges. By relocating the parallel at this point, the musician composing a tune in his head could be said to be comparable to the engineer thinking up a bridge; the composer notating the tune could then be comparable to the engineer drawing up the specifications; and when one stands to behold the finished bridge, that could be comparable to one listening to someone perform the finished composition.

On the other hand, even if we accept Collingwood’s idea that works of art exist only in the mind, it remains uncertain what this involves beyond a mere verbal adjustment. I showed earlier that when one accepts his definitions of expression, craft, creation, etc., his conclusions about the imaginariness of art fall into place, yet there seems no logical reason to prefer his definitions in the first place. Further, there seems to be good grounds for rejecting each of his definitions, though I would like to limit myself to the problems with his definition of expression as “an activity of which there can be no technique.”

If the absence of technique is a necessary condition for expression, then much of what passes for art must be excluded. This was part of Collingwood’s motivation, as he wanted to be able to dismiss such things as technical or formulaic art, yet it is not simply these genres that rely on technique. Within any artistic genre there is a confluence between creativity, on the one hand, and forms and procedures governing the medium, on the other; the later being classifiable under the rubric of technique. These forms (which can be learned through technique) provide the structure through which creativity (which cannot be learned through technique) can flow. Such things as how to hold a paint brush, how to use perspective, how to compose a piece that conforms to the structure of a sonata, how to achieve the appearance of weightlessness in ballet, etc., are among a myriad of techniques used by artists. Such techniques assist with creative expression. [9] For example, it would be difficult to express one’s emotions through sculpture if one has not first learned some basic techniques governing the medium. It would be hard to get up on stage and express one’s emotions through the art of ballet if one had not first learned the techniques of ballet dancing. It is only in contemporary art, where one can fling paint at the canvas, pound on the piano with one’s feet, or run around impromptu on stage, that there is a complete absence of technique.

Consistent with the consequences of rejecting technique, and consistent with the consequences of creation ex nihilo, is Collingwood’s idea that an artist has no idea of the general character of a work prior to creating it. He may sit down to write a play, but he has no idea whether it will be a tragedy or comedy. Exactly what is involved practically in the process of artistic creation remains vague for Collingwood. This is perhaps because the only process that would be consistent with all his conditions would be art that springs spontaneously from the muses – that is, pure inspiration and no skill, as Plato says of poetry.

I have suggested that technique can play an important role in helping us to express our emotions. In a similar way I would like to suggest that pre-existing material may also help the artist in this regard. (The assumption that the only purpose in art is to express one’s emotions may easily be challenged, yet my purpose is to show that Collingwood’s ideas are faulty even without rejecting this initial assumption.) The idea that works of art arise ex nihilo comes from the following suppressed premise: that when (or if) one’s inchoate feelings take form in matter as an artwork that this is always subsequent to the mental design of that work; being subsequent, therefore, the artwork can be said to be complete prior to the working of any material. Yet the anecdotal evidence of many artists (including my own experience as an amateur composer) does seem to suggest that the working of material is often the means, and not merely the effect, by which many artists express their emotions. Geniuses like Mozart and Beethoven could compose complete symphonies in their minds before mechanically transcribing them on paper, yet this seem to be the exception (and indeed, that is why such people were considered geniuses). Most children and amateur musicians, when composing music, usually always do so by sitting down at their instrument and experimenting with this or that pattern or combination of notes until they find something that expresses their emotions.[10]  Before the physical reality of the material sounds, the particular expression of emotions yielded by those sounds was unavailable. Apart from very simple pieces, it is only those who have a special ability that can compose something in the mind and then sit down and “reconstruct for themselves the imaginary tune that existed in {their} head.”[11]. Even the outstanding composer Haydn would often sit at the keyboard to compose his symphonies.

Other art forms must also be considered in this regard. How many painters are able to conceptualise the detail involved in a finished work prior to the externalisation process? Even if such conceptualisation were commonplace, is the actual skill that it takes to accurately represent the idea on canvas to be considered irrelevant in assessing one’s skill as an artist? I might be able to visualize a whole corpus of sculptures that are of equal quality to those of Michelangelo, but surely my skill as an artist would be inferior to Michelangelo if I am unable to actually create those objects. Here lies a problem with Collingwood’s theory which, if correct, leads to the counter-intuitive position that one’s greatness as an artist has nothing to do with actually making the sculpture or actually painting the work.[12]

Even if our discussion were to be confined to geniuses such as Mozart and Leonardo who, it is true, could conceptualise works prior to externalisation, one could certainly argue that the potentiality of actualisation is presupposed in the conceptual act. When Beethoven went deaf and had to compose his symphonies purely in his mind, he was able to ‘hear’ the music in his head only because he first knew what the various instruments sounding like. A man born blind could not create works of visual art in his head since such creation presupposes a prior knowledge of colour and form. If a blind person can imagine such things, it is only because he has learnt about it from someone who can see. So creating works of art in one’s mind, even if they are never physically actualised, presupposes knowledge of the content of the physical medium. So even when the mental creation is historically prior to the physical actualisation, there is no corresponding logical priority.

So far I have spoken about this from the composer’s point of view. But what about the audience? Collingwood suggests that when an audience listens to a piece of music, they reconstruct the imaginary tune that existed in the composer’s head. But how is this possible unless the piece means the same thing to the listener as it did to the performer, and means the same thing to the performer that it meant to the composer? If the composer is himself the performer, then we must assume that as he performs the piece he is expressing exactly those emotions which he felt when he originally composed the piece in his mind.

It hardly needs pointing out that such assumptions are exceedingly problematic. In his essay ‘The work of music and the problem of its identity’, Ingarden discussed the indeterminate aspects of a score. He pointed out that although a score contains instructions to the performer, the individual performer is left with a great deal of latitude which he must use if is to avoid making the piece sound mechanical and lifeless. The factors governing how the performer uses this latitude may certainly have something to do with the performer’s own emotions, and it is not reasonable to assume that this will resemble the intentions of the composer. Neither is it reasonable to assume that the emotions evoked in a person who listens to a piece will resemble those of the person performing the piece. This is true across all the arts. When one views a painting, watches a ballet, reads a poem, etc., can the emotions that one experiences ever be said to exactly resemble the emotions of anyone else? Even when the same person returns to the view, watch or listen to the same artwork again, is it reasonable to assume that it will mean the same thing the second time?

Despite the senses in which Collingwood’s theory runs counter to experience, it nevertheless does appeal to certain aspects of our experience of art works. For example, it appeals to our sense that artworks are imaginative, expressive and that they are more than just craft. Further, by postulating that works of art exist in the mind, Collingwood answers a problem concerning the identity of art - a problem that is particularly relevant when considering the ontological status of musical and literary works. It is simple enough to say that Michelangelo’s La Pieta exists at the Tate, but where do musical works exist? If they do not exist in the mind, should they be identified with an original manuscript, a specific performance or every single performance of a given piece? Each of these answers would seem to lead to problems. We know that a poem or a Sonata would not be lost even if the manuscript were destroyed, as long as there were copies or memories. Should the music or the poem then be identified with all the copies? In that case, how would we account for orally transmitted poetry like the Iliad? Identifying such things with a performance or the sum total of all the performances is equally problematic, for we feel that the work does not suddenly collapse into non-existence every time the performance ends or that it is created anew at the next performance.

To resolve difficulties such as the above, it is helpful to utilize the nomenclature of types and tokens, of which space permits only a very brief discussion. A type is the generic entity under which a class of particulars of an invention may be correlated and is identifiable by the particulars/tokens of that type. Thus, Beethoven’s ‘moonlight sonata’ or Keats’ ‘Ode to a Nightingale’ or Tschaikowsky Swan Lake may all be said to fall under the category of types. On the other hand, the individual performances of the ‘moonlight sonata’, the individual copies or recitations of ‘Ode to a Nightingale’ or the specific performances of the Swan Lake ballet, may all be said to be tokens of the types. The type shares those properties with the tokens that the tokens posses intrinsically (that is, by virtue of being a token of that type) and not those properties of a given token that are extrinsic, such as that it was performed last night by the European ballet or that it exists on my bedroom shelf. This helps to explain what we mean when we say that we have heard the ‘same’ piece on a number of different occasions. Though there is a numerical difference between the different performances, recording or reproductions[13], they can be seen to be qualitatively identical by virtue of being tokens of the same types.[14]

Using this nomenclature we are able to transcend the problems the arise in identifying a work with either a concrete externalisation or a mental construct, for, as Hanfling has pointed out, “a type cannot…be classified within the traditional mental/physical dichotomy.”[15] It is not hard to see how this provides an objection to Collingwood’s idea concerning the mental ontology of art. Since an art type cannot be said to be located anywhere – because it is not a ‘thing’ as such but a logical construct or concept – the work cannot be said to be located in the artist’s head. If it were located in the artist’s head, we would have no way to affirm that when the audience hears the piece that it is the same piece. Using type and tokens, however, gives us a way to establish identity and individuation. We can also establish that because knowledge of the type is only accessible through the tokens, that once a piece is performed, the token (and, by extension, the type) becomes publicly available and is not the private property of the artist. That is why the piece can continue to exist even after the artist is dead (something it would be difficult for Collingwood to affirm!), and why we can speak of great works of art as living on and being eternal.

So we see that the type/token nomenclature, though complex, is not counter-intuitive for it captures the sense we have of artworks existing independent of ourselves and enduring through time. It captures the sense that as soon as a work is created, it takes on an existence independent of the composer. That is, after all, why art is so attractive to human beings, for in creating, listening or viewing something that is enduring and can outlast ourselves, it enables us to feel that we are participating in something bigger than ourselves. It is bigger than ourselves because it does not simply exist in our own minds.

 

 

Bibliography

 

 

Oswald Hanfling, Philosophical Aesthetics: An Introduction (Blackwell Publishers in Association with the Open University, 1992).

 

R. G. Collingwood, ‘The Principles of Art’ in Theories of Art & Beauty (Milton Keynes, The Open University, 1991).

 

Richard Wollheim, ‘Art and its objects’ in Theories of Art & Beauty (Milton Keynes, The Open University, 1991).

 

Roman Ingarden, ‘The work of music and the problem of its identity’ in Theories of Art & Beauty (Milton Keynes, The Open University, 1991).

 

 



[1] Collingwood, p. 357.

 

[2]  Ibid, p. 355

 

[3] As he writes, “A work of art need not be what we should call a real thing. It may be what we call an imaginary thing.” Ibid, p. 362.

 

[4] Ibid, p. 364.

 

[5] Ibid, p. 362.

 

[6] Ibid, p. 364.

 

[7] Ibid, p. 366.

 

[8] Ibid, p. 363.

 

[9] Just as technique plays an important part in art, so expression can play an important part in craft. For example, an engineer may certainly express his emotions in preferring one design of a bridge over another (which, incidentally, would be a sufficient condition for the bridge to be art, according to Collingwood’s definition).

 

[10]  I am working on the basis of Collingwood’s assumption that all art expresses one’s emotions, though such an assumption might well be challenged.

 

[11] I have even found that some composers of special ability could only compose in the physical way I have mentioned, where the piece evolves through an experimentory process of trial and selection.

 

[12] Poetry is somewhat different since it operates in words, and there is no skill involved in transferring a word from thought to paper or voice, and hence one might argue that a poem could exist fully completed in one’s mind in a way not possible in other genres – but that is a point I will not pursue.

 

[13] I mention reproductions here to show that the idea of tokens can (if only in theory) be extended to the visual arts, though this creates certain complexities that I wish not to pursue.

 

[14]  It is not just with works of art that the type/token distinctions apply, for our entire language is permeated with the types and tokens, though unfortunately there is not space to develop on this. See Wollheim’s ‘Art and its Objects.’

 

[15]   Hanfling, p. 81.

 

 

 

 

Return to Robin Phillips HOMEPAGE

 

 

You are invited to join my mailing list!

 

As a member of my mailing list, you will receive automatic notification about additional material and features on this site, as well as occasional newsletters. To join, send a blank email to

 

largerhope  @  tiscali.co.uk

 

with “Join” in the subject heading. To unjoin, send a blank email with “Unjoin” in the subject heading.

 

(Note: for anti-spam purposes, the above email address has had spaced inserted before and after the @ sign. The address will only work after deleting these spaces.)