insists on the fact that the ambiguity (he uses the term «vagueness») of a poetic message is not limited to the semantic level (as is often the case with ethical terms), but rather extends to its syntactic construction, and, consequently, to the pragmatic level of psychological reaction. Similarly, Jakobson asserts that «ambiguity is an intrinsic, inalienable character of any selffocused message, briefly a corollary feature of poetry. Let us repeat with Empson:The machinations of ambiguity are among the very roots of poetry’ . . . The supremacy of poetic function over referential function does not obliterate the reference but makes it ambiguous» (Selected Writings, p. 238). As for the poetic word, cloaked with every possible meaning, see Roland Barthes, «Y atil use ecriture pot»tique?» in Le degri zero de l’ecriture (Paris: Seuil, 1953). These are essentially the same issues raised by the Russian Formalists: «The aim of poetry is to render the texture of words perceptible in all its aspects» (Boris Eikhenbaum, Lermontov Leningrad, 1924). In other words, for the Formalists the essence of the poetic discourse lay not in the absence of meaning but rather in its multiplicity.

The Divine Comedy: Paradiso, tr. Charles S. Singleton, Bollingen Series, vol. 8o (Princeton: Princeton University Press, 1975), canto 33, 11.
124126.

Openness. Information, Communication

Sec Stanford Goldman’s exhaustive study, Information Theory (New York: PrenticeHall, 1953), as well as A. A. Moles, Information Theory and Esthetic Perception. tr. Joel E. Cohen (Urbana: University of Illinois Press, 1966).

This definition can be traced back to a principle adopted by linguists, namely that, in phonology, every distinctive feature implies a choice between the two terms of an opposition. See N. S. Troubetskoy, Principes de phonologic (Pans: Klincksicck, 1949). esp. pp. 15, 33; Roman Jakobson. Essais de linguistique ,qinerale (Paris: Editions de Minuit, 1959), p. 104; and G. T. Guilbaud, La Cybernitique (Paris: Presses Universitaires de France. 1954). p. 103. As F. Boas has very clearly shown, the choice of a grammatical form by the speaker presents the listener with a definite number of bits of information. To give a precise meaning to a message such as «The man killed the bull.» the addressee must choose among a number of possible alternatives. In information theory, linguists have found a privileged tool for their investigation. Thus, the dialectics between redundancy and improbability in information theory (of which more later) has been measured against the dialectics between basis of comparison and variants, between distinctive features and redundant features. Jakobson speaks of a granulary structure of language that lends itself to quantification.

Max Planck, Wege zur physikalischen Erkenntnis (Leipzig: S. Hirzel Verlag), ch. 1.

Ibid.

Hans Reichenbach, The Direction of Time (Berkeley: University of California Press. 1956), pp. 5455. Unlike Reichenbach, Planck considers entropy a natural reality that excludes a priori all those facts that would seem empirically impossible.

Ibid., p. 151. 7. Ibid..p.167.

Norbert Wiener, The Human Use of Human Beings (Boston: Houghton Mifflin. 195o; rpt. New York: Avon Books, 1967), p. 31. In short, there is an equiprobability of disorder in relation to which order is an improbable event because it is the choice of only one chain of probability. Once a particular is realized, it becomes a system of probabilities in relation to which all deviation appears improbable.

For instance, a sequence of letters randomly drawn from the most probable trigrams in Livy’s language will yield a certain number of pseudowords with an unmistakable Latin sound: ibus, cent, ipitia, yetis, ipse, cum, vivius, se, acetiti, dedentur. See Guilbaud, La Cybernitique, p. 82. to. Wiener, The Human Use of Human Beings, p. 163.

Penguin Book of Italian Verse, ed. George Kay (Harmondsworth: Penguin,1958).
12.R. Shannon and W. Weaver, The Mathematical Theory of Communication (Urbana: Illinois University Press, 1949)
13.See Goldman, Information Theory, pp. 330331; and Guilbaud, La Cybemitique, p. 65.
14.Shannon and Weaver, The Mathematical Theory of Communication, pp. 99100, 104, to6.

Ibid., pp. 101102.
16.Giuseppe Ungaretti, «L’Isola,» in Life of a Man, tr. Allen Mandelbaum (New York: New Directions, 1958), pp. 5455.

The Russian Formalists had been dealing with the same question, though not in terms of information, when they came up with the theory of «estrangement,» or «defamiliarization» (priem ostrannenija). Extraordinarily enough, Shklovsky’s article «Iskusstvo kak priem» Art as devicewhich he wrote in 1917already anticipated all the possible aesthetic applications of an information theory that did not yet exist. «Estrangement,» for him, was a deviation from the norm, a way of confronting the reader with a device that would frustrate his systems of expectations and thereby draw his attention to a new, different poetic element. In this essay, Shklovsky is mostly concerned with illustrations of some of Tolstoy’s stylistic techniques, in which the author describes familiar objects as if he had never seen them before. A similar concern with deviations, and violations, of the narrative norm is also present in Shklovsky’s analysis of Tristram Shandy.
18.As did the Dadaists, and also Hugo Ball, who, in 1916, at the «Cabaret Voltaire» in Zurich, used to recite poetry in a strange, fantastic jargon. Similarly, certain contemporary musicians like to abandon themselves entirely to the whims of chance. All these, however, are marginal examples whose main experimental value is that they help set certain limits.
19.In other words, the fact that a work of art provides its audience with a certain kind of information certainly helps determine its aesthetic valuethat is, the way in which we «read» and appreciate it. This information plays a role in the total system and affects the form of the work. On the other hand, to believe that an analysis dealing exclusively with the informative value of a work might provide a satisfactory evaluation of that work would be somewhat naive. For an example of such naivete, see the symposium on «Information Theory and the Arts,» inJoumal of Aesthetics and Art Criticism, June 1959.
20.See Briefe an H. Jone und J. Humplick (Vienna, 1959). 21.Weaver, The Mathematical Theory of Communication, p. 117.

Moles, Information Theory and Esthetic Perception. Articles on the same subject have appeared in various issues of the Cahiers d’etudts de radiotelioisiots.

See Inccntri rnitsicali, vol. 3 (1959). the exchange between Nicolas Ruwet and Henri Pousseur.

Moles, IntOrmation Theory and Esthetic Perception, pp. 7879. «There is no absolute structural difference between noise and signal. They arc of the same nature. The only difference which can be logically established between them is based exclusively on the concept of intent on the part of the transmitter: A noise is a signal that the sender does not want to transmit.» «If the sonic material of white noise is formless, what is the minimum ‘personality’ it must have to assume an identity? What is the minimum of spectral form it must have to attain individuality? This is the problem of coloring white noises» (p. 82). This is also the problem that confronts the composer of electronic music.

This essay was originally written in 1960 for the fourth issue of lncontri musicali. The postscript was written six years later. Garroni’s critique was entitled La crisi semantica delle arti (Rome: Officina Edizioni, 1964), of which ch. 3 dealt with Opera aperta.

Goldman, Information Theory. p. 69.

If information theory corresponds to a statistical study of physical phenomena (seen as «messages»), this step will take us toward a communication theory that will deal specifically with human messages. The notion of «message» can operate on both levels, though we should not forget Jakobson’s objection to much theoretical work in communication: «Attempts to construct a model of language without any relation either to the speaker or to the hearer, and thus to hypostasize a code detached from actual communication, threaten to make a scholastic fiction out of language.» Roman Jakobson, Selected Writings (The Hague: Mouton, 1981), vol. 2, p. 576.
28.»Knowledge does not create the organization of its message; it imitates it to the extent that it is true and effective. Reason does not dictate its laws to the universe; rather, there is a natural harmony between reason and universe, since both obey the same general laws of organization.» P. Guillaume, La psychologie de la forme (Paris: Flammarion, 1937), p. 204.

«Several facts show that the perceptual interpretations of primary sensorial data are remarkably plastic, and that the same material, under given circumstances, may elicit very different reactions.» H. Pieron, in La Perception, a symposium volume (Paris: Presses Universitaires de France, 1955), P. / I
N. R. S. Lillie, «Randomness and Directiveness in Evolution and Activity in Living Organisms,» American Naturalist 82 (JanuaryFebruary, 1948): 17.

J. P. Kilpatrick, «The Nature of Perception,» in Explorations in Transactional Psychology (New York: New York University Press, 1961), pp. 4149.

«In perception, as well as in intelligence, nothing can be explained in terms of experience alone, and yet nothing can be explained without recourse, more or less substantial depending on the situation, to current or prior experience.» Jean Piaget, in La Perception, p. 21. See also Piaget, Les micanismes perceptifs (Paris: Presses Universitaires de France, 1961), p. 45,D: «The reason for the interactions between subject and object seems to be quite different from the one the founders of Gestalt theory have borrowed from phenomenology. The notion of a perceptual equilibrium suggested by facts is not the same as that of a physical field with a precise, automatic balance of the forces involved; rather, it entails active compensation on the part of the subject who is trying to moderate exterior disturbances.»

Jean Piaget, La psychologie de l’intelligence (Paris: A. Colin, 1947), chs. 1and 3.

Piaget, in La Perception, p. 28.

Piaget, La psychologie de l’intelligence, ch. 3. For a probabilistic study of perception, see Les micanismes perceptifi, where, after distinguishing between the operative processes of intelligence and those of perception, Piaget maintains that between the two «there is in effect an uninterrupted series of intermediaries» (p. 13). Experience itself