The chaotic character, the polyvalence, the multiinterpretability of this polylingual chaosmos, its ambition to reflect the whole of history (Quinet, Michelet) in terms of Vico’s cycles («jambebatiste»), the linguistic eclecticism of its primitive glossary («polygluttural»), the smug reference to Bruno’s torture by fire («brulobrulo»), the two obscene allusions that join sin and illness in one single root, these are just some of the things this sentence manages to suggest—in a first, cursory reading—thanks to the ambiguity of different semantic roots and the disorder of its syntactic construction.
Semantic plurality is not enough to determine the aesthetic value of a work. And yet it is precisely the multiplicity of the roots that gives daring and suggestive power to the phonemes. In fact, a new semantic root is often suggested by the juncture of two sounds, so that, in the end, auditory material and referential repertory are indissolubly fused.
On one side, the desire to produce an open, ambiguous communication affects the total organization of the discourse and determines both the density of its resonance and material and the proportional calibration of the relationship bet ween its sounds and its rhythm reverberate against a backdrop of references and suggestions, thereby increasing their echoes. The result is an organic balance such that nothing can be extracted from the ensemble, not even the slightest etymological root.
Theoretically, both Dante’s tercet and Joyce’s sentence result from an analogous structural procedure: an ensemble of denotative and connotative meanings fuses with an ensemble of physical linguistic properties to produce an organic form. From an aesthetic standpoint, both forms are «open» in that they provoke an ever newer, ever richer enjoyment.
But in Dante’s case, the source of this pleasure is a univocal message, whereas in Joyce’s it is a plurivocal message (not just in what it communicates but also in how it communicates it). Here, aesthetic pleasure is augmented by another value that the modern author is trying to attain—the same one that serial music is after when it attempts to free music from the compulsory tracks of tonality by multiplying the parameters along which sound may be organized and tasted; the same one that «informal painting.’ is after when it proposes different angles of approach for each and every painting; and the same one that the novel aims at when it no longer offers us one story and one plot per book but tries, rather, to alert us to the presence of more stories and more plots in the same book.
Theoretically, this value should not be confused with aesthetic value: to succeed aesthetically, the project of plurivocal communication must be incorporated into the right form, since this alone can endow it with the fundamental openness proper to all successful artistic forms. On the other hand, plurivocality is so much a characteristic of the forms that give it substance that their aesthetic value can no longer be appreciated and explained apart from it. In other words, it is impossible to appreciate an atonal composition without taking into consideration the fact that it wants to provide an alternative, an openness, to the fixed grammar, the closure, of tonal music and that its validity depends on the degree of its success in doing so.
This value, this second degree of openness to which contemporary art aspires, could also be defined as the growth and multiplication of the possible meanings of a given message. But few people are willing to speak of meaning in relation to the kind of commu nication provided by a nonfigurative pictorial sign or a constellation of sounds. This kind of openness is therefore best defined as an increase in information. Such a definition, however, forces us to move our investigation onto a different level and to demonstrate the validity of information theory in the field of aesthetics.
III. Openness, Information, Communication
In its advocacy of artistic structures that demand a particular involvement on the part of the audience, contemporary poetics merely reflects our culture’s attraction for the «indeterminate,» for all those processes which, instead of relying on a univocal, necessary sequence of events, prefer to disclose a field of possibilities, to create «ambiguous» situations open to all sorts of operative choices andinterpretations.
To describe this singular aesthetic situation and properly define the kind of openness» to which so much of contemporary poetics aspires, we are now going to make a detour into science, and more precisely into information theory, hoping it will provide us with a few indications that might prove useful to our research. There are two main reasons for this detour. In the first place, I believe that poetics in certain cases reflects, in its own way, the same cultural situation that has prompted numerous investigations in the field of information theory.
Second, I believe that some of the methodological tools employed in these investigations, duly transposed, might also be profitably used in the field of aesthetics (as we shall see, others have already done this). Some people will object that there can be no effective connections between aesthetics and information theory, and that to draw parallels between the two fields can only be a gratuitous, futile exercise. Possibly so. Before engaging in any kind of transposition, let us therefore examine the general principles of information theory with no reference to aesthetics, and only then decide whether there are any connections between the two fields and, if so, of what sort, and whether it might be profitable to apply to one the methodological instruments used in
1
Information Theory
Information theory tries to calculate the quantity of information contained in a particular message. If, for instance, on August 4 the weather forecaster says, «Tomorrow, no snow,» the amount of information I get is very limited; my own experience would have easily allowed me to reach that conclusion. On the other hand, if on August 4 the forecaster says, «Tomorrow, snow,» then the amount of information I get is considerable, given the improbability of the event.
The quantity of information contained in a particular message is also generally conditioned by the confidence I have in my sources. If I ask a real estate broker whether the apartment he has just shown me is damp or not and he tells me that it is riot, he gives me very little information, and I remain as uncertain as I was before I asked him the question. On the other hand, if he tells me that the apartment is damp, against my own expectation and his own interest, then he gives me a great deal of information and I feel I have learned something relevant about a subject that matters tome.
Information is, therefore, an additive quantity, something that is added to what one already knows as if it were an original acquisition. All the examples I have just given, however, involved a vast and complex amount of information whose novelty greatly depended on the expectations of the receiver. In fact, information should be first defined with the help of much simpler situations that would allow it to be quantified mathematically and expressed in numbers, without any reference to the knowledge of a possible receiver. This is the task of information theory. Its calculations can suit messages of all sorts: numerical symbols, linguistic symbols, soundsequences,andsoon.
To calculate the amount of information contained in a particular message, one must keep in mind that the highest probability an event will take place is 1, and the lowest is o. The mathematical probability of an event therefore varies between I and o. A coin thrown into the air has an equal chance of landing on either heads or tails; thus, the probability of getting heads is 1/2. In contrast, the chance of getting a 3 when rolling a die is 1/6. And the probability that two independent events will occur at the same time is the product of their individual probabilities; thus, when rolling a pair of dice, the probability of getting a t and a 6 is 1/36.
The relationship between the number of possible events in a series and the series of probabilities connected to each of them is the same as that between an arithmetic progression and a geometric progression, and can be expressed by a logarithm, since the second series is the logarithm of the first. The simplest expression for a given quantity of information is the following:
odds that addressee will know content of message after receiving itInformation = log———————————————————————
odds that addressee will know content of message before receiving it
In the case of the coin, if I am told that the coin will show heads. the expression will read:
1 /
log2 1 / 2 = 1.
Information theory proceeds by binary choices, uses base 2 logarithms, and calls the unit of information a «bit,» a contraction of «binary» and «digit.» The use of a base 2 logarithm has one advantage: since log,2 = 1, one bit of information is enough to tell us which of two probabilities has been realized. For a more concrete example, let’s take a common 64square chessboard with a single pawn on it. If somebody tells me that the pawn is on square number 48,
