The study of literary art raises two sets of difficulties. Firstly, those pertaining to its material, to what is usually called words. Secondly, those pertaining to the principles of construction of the art.

In the first case, our object of study is found to be closely linked to our practical consciousness, and frequently has meaning only by reason of this connection. The existence of such a relation is too easily forgotten, and literary study is pursued by means of other relations from practical life which are purely arbitrarily introduced.footnote¹ The result is that the heterogeneous, polysemic character of the material, dependent on its role and destination, is ignored. In the same way, no account is taken of the fact that there are in a word inequivalent elements depending on its concrete function; one element may be promoted at the expense of the others, which are, in consequence, deformed or occasionally degraded to the status of mere neutral accessories. Potebnia’s grandiose attempt to construct a theory of literature moving from the word as

The second difficulty arises because the principle of construction or formation is habitually treated as a static principle. For example, the mode of criticism which consisted of questioning and judging the characters of a novel as if they were living people has been abandoned, but no-one can claim that biographies of characters and attempts to establish historical reality from these biographies have definitively disappeared. All this rests on the postulate of the static hero. Here we should remember Goethe’s remarks on artistic fiction, on the dual sources of light in some Rubens landscapes, and on dual facts in Shakespeare: ‘In the higher regions of art, where a picture becomes a picture properly speaking, the artist has freer play, and may even resort to fictions . . . The artist speaks to the world by means of a whole . . . but if this is against nature, I would claim that it is higher than nature.’ At one time Lady Macbeth tells us: ‘I have given suck’, and later we are told that she has no children; she is justified as a character because, for Shakespeare, ‘what matters is the power of each of these apostrophes. It is imperative always to avoid analysing the painter’s brush-strokes and the poet’s words with such minuteness and exactitude . . . the poet at any moment makes his characters say what is suitable, just and good at that moment, without overloading his mind by worrying whether these words may not contradict some others.’ And Goethe explains this from the standpoint of the constructive principle of Shakespearian drama: ‘Further, Shakespeare never expected that his dramas would some day be read in printed books, that the letters would be counted, compared and inventoried; as he wrote, he had the stage before his eyes; he saw in his dramas something living and mobile which had to pass rapidly from the stage to the eyes and ears of the spectators, something which could not be expected to be pinned down and criticized in detail since it was merely a matter of acting and producing an impression.’footnote²

Thus the static unity of the character (like all static unities in literary works) proves to be extremely unstable; it depends entirely on the principle of construction and may oscillate in the course of a work in a way prescribed in each particular case by the general dynamism of the work. It is enough that there be a sign designating the unity: its category legitimizes the most striking effective violations, obliging us to consider them as equivalences of the unity.

But this unity is not, as might naïvely be imagined, the static unity of the hero; the sign of a static entity is replaced by the sign of dynamic integration, of integrity. There is no static hero, there are only dynamic heroes. And the sign of a hero, the name of a hero are enough; we need not detail the hero in every given situation.

The stability and solidity of static habits of consciousness appear in the case of character. We find the same state of affairs in the problem of the ‘form’ of a literary work. The famous analogy: form/content = glass/wine has been abandoned. But all the spatial analogies applied to the notion of form have importance only in their lack of truth. In fact, the stasis implied by the notion of space constantly slips into the concept of form (whereas spatial forms should be seen as dynamic forms sui generis). It is the same with terminology. I daresay that in nine cases out of ten the word ‘composition’ denotes for the researcher some idea of static form. The notions of ‘line’ and ‘strophe’ are imperceptibly drawn out of the dynamic series; repetition is no longer seen as a matter of different intensities according to different frequencies and qualities; the dubious notion of ‘the symmetry of compositional facts’ is produced, dubious because there can be no question of symmetry if it is a matter of reinforcement.

The unity of a work is not a symmetrical, closed entity, but a dynamic integrity with its own movement; its elements are not united by a sign of equality and addition, but by a dynamic sign of correlation and integration.