one letter for conspiratorial reasons has established a second, internal rhythm, and all remaining words have been equally reduced. It would be strange and monstrous to suggest that Pushkin had structured this note, one that he would have found tragic, with conscious care for its rhythmic or phonological organization; the point, rather, is that the immanent and unconsciously operative laws of autocommunication display certain structural features that we commonly observe in the example of a poetic text.
These peculiarities are even more noticeable in the following example, stripped of both the mnemonic and the conspiratorial functions and presenting auto-messaging in its purest form. We are speaking here of the unconscious notes that Pushkin made in the process of reflection, quite possibly without realizing he was doing so.
On May 9, 1828, Pushkin wrote the poem “Alas! The Language of Garrulous Love” [Uvy! iazyk liubvi boltlivoi], dedicated to Anna Alexeevna Olenina, whom he was then courting. There we find the following note:
ettenna eninelo
eninelo ettenna.12
Beside the note: “Olenina Annette.” Over “Annette,” Pushkin had jotted “Pouchkine.” It is not difficult to reconstruct the line of thought: Pushkin was thinking about Annette Olenina as a fiancée and wife (the note “Pouchkine”). The text presents anagrams (one reads them right-to-left) of A. A. Olenina’s first and last name as he was thinking of her in French.
The note’s mechanics are interesting. Initially, the name is transformed through its reverse reading into a conventional index, and then the repetition establishes a certain rhythm, while the transposition rhythmically disturbs that rhythm. The poem-like character of such a construction is obvious.
* * *
One can represent the mechanism for information transfer along the “I—I” channel as follows: a certain message is introduced in a natural language, then a certain supplemental code is introduced that constitutes a purely formal organization, one that is constructed syntagmatically in a specific way and is simultaneously either liberated completely from semantic meanings or strives toward such liberation. A tension arises between the initial message and the secondary code, fostering a tendency to interpret the text’s semantic elements as having already been included within the supplemental syntagmatic construction and now receiving new—relational—meanings from their mutual correlation. However, while the secondary code strives to liberate the primarily signifying elements from the general semantic embedded in the primary code, this does not occur. The shared semantics remains, but they are overlaid with a secondary semantics formed on account of those shifts that arise out of using signifying units to construct a language of different kinds of rhythmic series. But the text’s semantic transformation does not end there. The proliferation of semantic connections within the message muffles the primary semantic connections, and the text can behave, at a given level of perception, as a complexly organized, asemantic message. But highly organized, asemantic texts tend to become the organizers of our associations. We ascribe them associative meanings. Thus, in scrutinizing the pattern on the wallpaper or listening to non-program music, we ascribe specific meanings to the elements of these texts. The starker the syntagmatic organization, the more associative and free the semantic connections will become. Accordingly, the text along the “I—I” channel tends to become overgrown with individual meanings, and it begins to serve as an organizer of the scattered associations that accumulate in a given person’s consciousness. It restructures the personality that has been involved in the autocommunicative process.
In this way, the text carries a threefold significance: the primary level is in the language itself; the secondary comes about on account of the text’s syntagmatic reorganization and the tension among its primary units; and the third level arises from extratextual associations of varying degrees—from the most general to the extremely individual—being drawn into the message.
There is no need to prove that the mechanism we have described can simultaneously be presented as typical of the processes that form the basis of poetic creation.
The poetic principle, however, is one thing, but actual poetic texts are another. It would be an oversimplification to identify the latter with the messages being broadcast along the “I—I” channel. An actual poetic text is broadcast along two channels simultaneously (the exceptions being experimental texts, glossolalia, texts like asemantic children’s school rhymes and zaum,viii as well as texts in languages their audiences do not understand). It oscillates between the meanings transmitted across the “I—HE” channel and those formed in the process of autocommunication. Depending on its movement toward one axis or the other, and on the text’s orientation toward one kind of transmission or the other, it is taken to be a “poem” or “prose.”
Of course, the text’s orientation toward the primary linguistic message or a complex restructuring of meanings and the proliferation of information in itself does not mean that it will function as poetry or as prose: what comes into play here is the correlation with these concepts’ general cultural models within a given era.
And so we can conclude that the system of human communication can be constructed in two ways. In one instance, we are dealing with certain information given in advance and traversing from one person to another, and with a code that is constant within the limits of the entire communicative act. In the other, we have an increase in information, its transformation, its reformulation, during which it is not new messages but new codes that are introduced, and the receiver and the transmitter are combined in one person. In the process of such autocommunication the individual personality is itself reformulated, and a rather broad range of cultural functions is tied to this, from the sense of your own separate being that a person needs to have in certain kinds of culture, to self-consciousness and autopsychotherapy.
The role of such codes can be played by various kinds of formal structure—the more asemantic their organization, the more successfully they serve the function of reorganizing meanings. Such are spatial objects that, like patterns or architectural assemblages, are destined to be contemplated, or temporal ones, like music.
Verbal texts present a more complex issue. Insofar as the autocommunicative nature of a transmission can be masked by its assuming the forms of other aspects of communication (for example, a prayer can be perceived as a communication not with oneself, but with an external, almighty power; a repeat reading, the reading of a text that is already familiar—by analogy with the first reading—as a communication with the author, and so on), the addressee who is receiving the verbal text must decide what it is that has been transmitted to him—a code or a message. Here, to a significant degree, it will be a question of the receiver’s frame of mind, insofar as one and the same text can serve as message or code or, oscillating between these poles, one and the other simultaneously.
Here one ought to distinguish between two facets—the properties of the text that allow it to be interpreted as a code, and how the text functions, which allows it to be used in this way.
In the first case, the need to receive the text not as a usual message but as some coded model is marked by the formation of rhythmic series, of repetitions, by the appearance of supplemental patterns that are completely superfluous from the point of view of communication within the “I—HE” system. Rhythm is not a structural level in the construction of natural languages. It is no accident that while the poetic functions of phonology, grammar, and syntax find their bases and analogues in corresponding, non-artistic levels of the text, one can point to no such parallel for meter.
The rhythmic-metrical systems are transferred not from the “I—HE” communicative structure, but from the “I—I.” The projection of the principle of repetition into the phonological and other levels of natural language constitutes autocommunication’s aggression toward any linguistic sphere that is alien to it.ix
Functionally, whenever it adds no new information to what we already have and transforms the self-awareness of the individual who generates texts and transfers messages already in hand to a new system of meanings, the text is used not as a message but as a code. If Reader N is informed that a certain woman named Anna Karenina has thrown herself in front of a train because of a tragic love affair, and instead of attaching that message to those