The three metaphors

Peter Steiner

pp. 39-116

The Machine

To make two bald statements: There’s nothing sentimental about a machine, and: A poem is a small (or large) machine made of words. When I say there’s nothing sentimental about a poem I mean that there can be no part, as in any other machine, that is redundant. [...]

There is no poetry of distinction without formal invention, for it is in the intimate form that works of art achieve their exact meaning, in which they most resemble the machine, to give language its highest dignity, its illumination in the environment to which it is native.

—William Carlos Williams, Collected Later Poems

Probably the best known Formalist model was advanced by Viktor Šklovskij, the self-proclaimed “founder of the Russian school of Formal method” (Šklovskij 1923b, p. 317). His answer to the question “what is Formalism?” was very clear: “In its essence the Formal method is simple—a return to craftsmanship” (ibid., p. 327; Brik 1923, p. 214). Technology, that branch of knowledge pertaining to the art of human production, was the predominant metaphor applied by this model to the description and elucidation of artistic phenomena.3

Šklovskij’s obsession with the machine analogy was well known to his contemporaries. In a commemorative article about Jurij Tynjanov, Lidija Ginzburg recalls a random chat of 1925 in which Tynjanov had tried to differentiate his own approach to literature from Šklovskij’s. “Viktor is a fitter, a mechanic.—And a chauffeur, someone prompted.—Yes and a chauffeur too. He believes in construction. He thinks that he knows how the car is made...” (in Kaverin 1966, p. 90). Tynjanov did not have to explain his phrase because the hint was transparent to everyone. He was alluding to Šklovskij’s bon mot in a 1922 letter to Roman Jakobson: “We know how life is made and how Don Quixote and the car are made too” (Šklovskij 1922a, p. 24; Larin 1923, p. 89).

Šklovskij did not reserve his car/literature analogy for the inner Formalist circle. Quite the contrary: it recurs again and again as the central image in his scholarly, pedagogical, and creative texts as well. For example, in his booklet The Technique of the Writer’s Trade (1928), Šklovskij advises aspiring prose writers about how to read literature:

If you wish to become a writer you must examine a book as attentively as a watchmaker a clock or a chauffeur a car.

Cars are examined in the following ways: The most idiotic people come to the automobile and press the balloon of its horn. This is the first degree of stupidity. People who know a little more about cars but overestimate their knowledge come to the car and fiddle with its stick-shift. This is also stupid and even bad, because one should not touch a thing for which another worker is responsible.

The understanding man scrutinizes the car serenely and comprehends “what is for what”: why it has so many cylinders and why it has big wheels, where its transmission is situated, and why its rear is cut in an acute angle and its radiator unpolished.

This is the way one should read (Šklovskij 1928b, pp. 7-8).

What this technological metaphor meant for the study of literature is apparent in the introduction to On the Theory of Prose—the most scholarly of Šklovskij’s books: “In the theory of literature I am concerned with the study of the internal laws of literature. To draw a parallel with industry, I am interested neither in the situation in the world cotton market, nor in the policy of trusts, but only in the kinds of yarn and the methods of weaving” (Šklovskij 1925, p. 5). Because of the repeated use of the machine analogy, I shall term this trend in Formalism “mechanistic.”

The source of Šklovskij’s technological metaphor is rather complex. It betrays first the influence of Italian Futurism, with its cult of the machine as the most crucial factor in the birth of the modernist artistic sensibility. But in Russia it also indicated a certain political stance. It was related to the leftist intelligentsia’s yearning for a radical transformation of Russian society. The mastery of technology was often seen as the ultimate means to this end. Lenin’s famous equation—“socialism = the Soviet government + electrification”—was an expression of this belief, as were the unrealizable Constructivist projects of scientifically designed socialist cities, or Vladimir Majakovskij’s statement that a single Ford tractor is better than a collection of poems.

Šklovskij’s interest in literary know-how was conditioned by pragmatic concerns too. The Formalist leader did not enter the field of Russian letters as an academic observer or an armchair theoretician, but as an active participant—a creative writer. From this perspective, the problems of literary production were of paramount significance. Yet it was precisely in this area that previous Russian criticism exhibited a curious lacuna. Whereas for all the other arts technical knowledge was considered vital to both historical and practical study, in literature technique was consigned to schoolbooks on poetics that were mere catalogs of tropes, figures, and meters derived from Greek and Latin models. It was this gap that mechanistic Formalism, concerned with the literary technē, set out to close.

The selection of the machine as the controlling metaphor of his theoretical model served Šklovskij in yet another way. It furnished a frame of reference that enabled him to treat literature in a manner radically different from that of pre-Formalist critics. At the risk of oversimplification, one might claim that traditional literary scholars were concerned above all with what the work conveyed. To understand this “what,” students of Russian literature looked beyond the work: into its author’s life, the philosophy supposedly embodied in it, or the sociopolitical events that gave rise to it. This “what,” customarily called the content of the literary creation, was opposed to its how, its form. And even though the meaning of these two notions varied from critic to critic, the “what,” the message of the literary work, always seemed the decisive member of the pair. Form was relegated to a mere auxiliary mechanism necessary for expressing content, but completely dependent upon it.

By focusing on the nuts and bolts of poetic texts, the internal laws of literary production, mechanistic Formalism radically reversed the value of content. Mocking traditional critics, Šklovskij wrote: “The present-day theoretician, in studying a literary work and considering its so-called form as a shroud that must be penetrated, is mounting a horse while jumping over it” (Šklovskij 1925, p. 162). The “how” of literature gained decisive prominence in the mechanistic model, and the machine analogy furnished the conceptual viewpoint that enabled Šklovskij to redirect attention from the external conditions of the literary process to the internal organization of the work.

Disjunction was the key logical principle by which mechanistic Formalism organized its basic concepts. This principle split art decisively from non-art (or byt, “everyday life”4) and expressed their mutual exclusivity in the following set of polar oppositions:


The first concept in the table, de-familiarization (ostranenie), has today gained wide currency. The word was coined by Viktor Šklovskij to account for the special nature of artistic perception. In his 1914 manifesto, The Resurrection of the Word, Šklovskij presented the dialectics of de-familiarization and automatization in this way: “By now the old art has already died, but the new has not yet been born. Things have died too: we have lost the sensation of the world. We are like a violinist who has stopped feeling his bow and strings. We have ceased to be artists in our quotidian life: we do not like our houses and clothes and easily part with a life that we do not perceive. Only the creation of new forms of art can bring back to man his experience of the world, resurrect things and kill pessimism” (In Stempel 1972, p. 12).

In this early formulation, the principle of de-familiarization is closely linked to the poetics of Russian Futurism, a movement that sentenced past art to death and set out to create artistic forms more attuned to the iconoclastic tastes of radical youth. As his mechanistic model developed, Šklovskij began to replace the existential frame of reference with terminology that would better fit his machine metaphor. It was economy, or more precisely, energy-efficiency, that eventually became the criterion for differentiating between automatized and de-familiarizing modes of perception.

Šklovskij’s concept of artistic perception has its roots in the positivistic belief in art’s economizing of mental energy, in particular the principle of least effort that Herbert Spencer (1820-1903) had declared the universal law of style. In the Russian context, Spencer’s theory had found an echo in the writings of Aleksandr Veselovskij (1838-1906), one of the few critics of the past whom the Formalists did not completely disregard. In the third chapter of his unfinished Historical Poetics, Veselovskij had used Spencer’s principle of the economization of mental energy to support his differentiation of poetic from prose style. Poetry achieves its results with a paucity of means impossible in prose, as witnessed in its unfinished periods, elisions, and omissions. Veselovskij especially stressed the role of rhythm and rhyme, the predictability of which purportedly saves us from wasting energy in frustrated expectations (Veselovskij 1940, p. 336). It was this assertion that Šklovskij challenged. “The idea of the economy of energy as the law and goal of creativity might be correct when applied to a particular case of language, ‘practical’ language;” but “the language of poetry is a difficult language, language which is made difficult and hampered” (Šklovskij 1925, p. 115). According to Šklovskij, the perception of art manifests not the law of least effort but the law of maximal effort.

The explanation of this claim offered by the mechanistic Formalists is elegant in its simplicity: artistic form is difficult because it is made so. The teleology used in this argument is in perfect harmony with the technological metaphor. The work of art as a product of an intentional human activity is a functional object whose purpose is to change the mode of our perception from practical to artistic. This change can be effected in several ways, most simply by displacing an object from its customary context. “In order to render an object an artistic fact it must be extracted from among the facts of life... it must be torn out of its usual associations” (Šklovskij 1923a, p. 115).

The Formalists were not so much interested in ready-made objects or found art as in the artistic work as a complex artifact. For this reason the concept of “displacement” was always secondary to that of the “device,” which pertains specifically to the production of the work. “Every art,” argued Šklovskij, “has its own organization—that which transforms its material into something artistically experienced. This organization is expressed in various compositional devices, in rhythm, phonetics, syntax, the plot of the work. It is the device that transforms extra-aesthetic material into the work of art by providing it with form” (ibid., p. 138). The device changes extra-artistic material into art, forming it anew and in this way de-familiarizing it. The cardinal position of the concept of the device is apparent in Jakobson’s programmatic statement: “If literary history wishes to become a scholarly discipline it must recognize the artistic device as its sole hero” (Jakobson 1921a, p. 11).

It must be stressed, however, that despite their obvious similarity there is an important difference between Jakobson’s and Šklovskij’s notions of the device. For Jakobson, the material of verbal art was language and hence he conceived of poetic devices as linguistic by their very nature. Šklovskij did not deny that in poetry language itself is de-familiarized. “But,” he hastened to add, “there are works of art in which the aesthetic perception of divergence rests outside the word, where the word is disregarded, is not felt, or has ceased to be felt” (Šklovskij 1927, p. 8). These are, obviously, works of literary prose—the main field of Šklovskij’s expertise. In this literary form, the source of de-familiarization is the deformation not of language but of events and happenings in the process of their verbal representation. Accordingly, the devices that Šklovskij studied most closely were those pertaining to prose composition and narrative.

The difference between literary narrative and the events it narrates in Šklovskij’s understanding is that between the device and the material. A prose work is an intentional construction, whereas the events represented in it are merely the material for this construction. The corresponding terms in the sphere of narratology are “plot” and “story,” the two modes in which events “occur” in literature. Story was understood as the series of events ordered according to their temporal succession (as they would have occurred in reality) and as Tomaševskij stressed, according to causality (Tomaševskij 1925c, p. 136). Plot, on the other hand, was the liberation of events from temporal contiguity and causal dependency and their teleological redistribution in the literary text. The story, equated with material, served the artist as a mere pre-text for plot construction, a process governed not by external causes but by internal, formal laws. Here form, conceived “as the law of construction of the object” (Šklovskij 1929, p. 60), was opposed to “motivation” defined by Šklovskij as the “extra-literary [bytovoe] explanation of plot construction” (Šklovskij 1923c, p. 50). Motivation was seen as playing only a secondary role in the literary construction, for “the forms of art are explained by their artistic regularity and not by extra-literary motivation” (Šklovskij 1925, p. 161).

By relegating material to a mere ancillary position, the mechanistic Formalists ascribed value to it only insofar as it contributed to the technique of the work itself. Material was deprived of any emotional, cognitive, or social significance. Thus, a literary construction was nothing more than “pure form—a relation of materials” (ibid., p. 162). Or even more radically, “values became artistic material, good and evil became the numerator and the denominator of a fraction and the value of this fraction equaled zero” (ibid. p. 169).5

The position of the mechanistic model in the overall picture of Russian Formalism is rather peculiar. Perhaps the term “teaser” (probnik), which Šklovskij used to describe his own existential predicament, best characterizes the role this model played in the history of the movement (Šklovskij 1924, pp. 66-67). From the vantage point of hindsight, the mechanistic metaphor represents a transitory stage in Formalism. Šklovskij’s The Resurrection of the Word was, without any doubt, the first attempt at formulating some of the basic principles of literary study that later acquired the name of the Formal method. But in marking the beginnings of the Formalist enterprise, over the course of time this text inevitably became marginal in view of further developments. A historical marker, it seems, plays a double role. It is not only the boundary that separates two successive developmental stages, but also the point of their contact. Thus, while Šklovskij’s 1914 manifesto revolutionized literary studies by injecting into them principles of the avant-garde artistic practice of Russian Futurism, at the same time it carried over a large remnant of the older critical tradition.

As I shall illustrate later, mechanistic Formalism was in some respects a mirror image of Veselovskij’s poetics. We have already seen how its key term, “de-familiarization,” was derived from its predecessor by reversing Veselovskij’s criterion of poetic style. But Šklovskij was able to do so because he was brought up on Veselovskij’s system and shared some of its postulates. While subverting some of Veselovskij’s principles, Šklovskij covertly borrowed others from the nineteenth-century philologist. He was certainly aware of the perils that this inverse parallelism posed to his own theorizing. “I am afraid of the negative lack of freedom,” he complained. “The negation of what others are doing ties me to them” (Šklovskij 1926, p. 52). And it was this link to nineteenth-century philology that at least in part was responsible for the quick aging of the mechanistic model. In fact, most of the subsequent developments of Russian Formalism might be seen as a series of corrections of and departures from the original Šklovskian metaphor.

In his perceptive study of Veselovskij’s poetics, Boris Èngel’gardt described it as consisting of two integral components: the history of literature in the strict sense of the word, and the theory of the genesis of poetry from extra-aesthetic phenomena (Èngel’gardt 1924, pp. 90-91). The great Russian philologist conceived of literature, first of all, as part of the larger cultural context. According to his famous formula of 1893, the history of literature is the “history of social thought in imagistic-poetic experience and the forms that express it” (Veselovskij 1940, p. 53).

The role of the literary historian, then, is to recover the causal relations among successive elements of social thought. “When studying a series of facts,” Veselovskij argued, “we observe their successivity, the relation of what follows to what precedes it. If this relation recurs we begin to suspect a certain regularity. If it recurs often enough we cease to speak of preceding and following and substitute the terms cause and effect.” To establish the true regularity of the phenomena studied, however, historians must extend their research to the series contiguous to the one under investigation, to discern whether the cause of change does not lie outside it. They must also test knowledge gained from one series on other similar series to discover whether a causal relation obtains there as well. “The more such tested recurrences,” Veselovskij concludes, “the more probable it is that the resulting generalization will approximate the precision of a law” (ibid., p. 47).

The history of literature for Veselovskij is an incessant interaction between two factors: the passive artistic form and the active social content. What differentiates literature from other intellectual practices (philosophy, religion, and so forth), and hence what makes it possible to speak about the history of literature, is the repertoire of elementary poetic forms that express thought. Those forms—various types of imagery, parallelisms, or plot constructions—which Veselovskij outlined in his genetic studies of poetry, are passed from generation to generation in the same way as every national language and are recombined in every literary work.

From this perspective it might appear that literary history is simply the permutation of the same forms without any actual change, but Veselovskij claims that literature does evolve, that the constant poetic forms are continuously imbued with new content. This content does not come from literature itself but from developments in social life and corresponding transformations in the human spirit. Thus, the engine of literary history according to Veselovskij lies outside literature and the task of the historian “is to study how new life content, this element of freedom that rushes in with every new generation, fills the old molds, those forms of necessity in which the entire previous development has been cast (ibid., p. 52).

This short presentation of Veselovskij’s views on literary history should suffice to explain Šklovskij’s attitudes toward him. Šklovskij’s radical separation of literature from other spheres of social life, his rejection of the causal explanation in literary studies—all of this can be seen as resulting from his negative relation to Veselovskij. Yet it must be stressed that despite this, Šklovskij did not banish diachrony from literary studies, and in fact affirmed the historical dimension of verbal art. As Jurij Striedter has observed, de-familiarization, the key concept of mechanistic Formalism, as the juxtaposition of old and new, definitionally presupposes some form of temporality (Striedter 1978, p. 1). Nevertheless, Šklovskij’s notion of literary history deviated radically from Veselovskij’s.

At the outset it must be said that Šklovskij’s treatment of literary diachrony is not altogether consistent. The charismatic Formalist leader did not study this topic systematically, and in the course of time changed his mind about some important issues. The concept of de-familiarization is a case in point. In his Resurrection of the Word, Šklovskij argued that what art modifies above all is our habitual perception of the world. Art develops in order for us to regain a feeling for objects (and language) that have become automatized in our perception. This notion of de-familiarization is the direct reverse of Veselovskij’s idea of literary change. For him it was the evolution of life that revitalized petrified artistic forms, whereas for Šklovskij the evolution of art revitalized the automatized forms of life. Nevertheless, this reversal still proceeds from an inevitable relationship between literature and everyday life, which Šklovskij’s mechanistic model denied. The value of art is a function of its utility for byt, and hence cannot be separated from it.

For this reason Šklovskij subsequently modified his notion of de-familiarization. As early as 1919, in the OPOJAZ collective volume Poetics, he declared that the development of art is totally immanent. New works come about to change our perception not of byt but of the artistic form itself, which has become automatized through our acquaintance with older works. “The work of art is perceived against the background of and through association with other works of art. Its form is determined by its relation to other forms that existed prior to it.... A new form appears not to express a new content but to replace an art form that has lost its artistic quality” (Coll. 1919, p. 120).

The admission that the work of art is peculiar because it differs not only from everyday reality but from earlier works as well introduces an element of chaos into the two-term system of mechanistic Formalism. Though Šklovskij still upheld the original opposition of art and byt, he was forced to complicate the category of art with a secondary dyad, canonized/noncanonized art. He took this step in his short booklet on Vasilij Rozanov. “In every literary period,” Šklovskij wrote, “not one but several literary schools may be found. They coexist; one of them is the canonized apex and the others are a noncanonized [lower stratum].... While the forms of the older art become as little perceptible as grammatical forms in language from elements of artistic intention [ustanovka] turning into ancillary, non-perceptible phenomena—the new forms of art that substitute for the older ones are produced in the lower stratum. A younger school bursts into the place of an older one.... However, the defeated school is not destroyed, does not cease to exist. It is only displaced from the top to the bottom... and can rise again” (Šklovskij 1921, pp. 5-7).

This model of immanent literary history, however, begs certain questions. First of all, what is the ontological status of noncanonized literature? Within the framework of mechanistic Formalism this category is a conceptual bastard, in that it is composed of artworks whose form, paradoxically, is not perceptible. One might also inquire whether this model, which treats literary history as an “eternal return” of the same artistic forms, does not preclude the possibility of any actual developmental novelty. Earlier I argued that a similar problem had existed for Veselovskij when he insisted that every literary work is a recombination of the same elementary poetic forms. But because he did not conceive of literary history as an immanent process, formal repetition nevertheless implied for him novelty in content. This avenue was closed for Šklovskij, however, who programmatically refused to deal with the issue of literary content.

Locked in his mechanistic metaphor, Šklovskij could provide no viable answer to the ontological status of noncanonical art. It was only in another Formalist model, the one advanced by Jurij Tynjanov, that this issue was addressed. Tynjanov’s studies of the change that Russian literature underwent in the eighteenth and nineteenth centuries refuted the basic premise of mechanistic Formalism, the strict separation of art from byt. As he illustrated convincingly, the line separating literature from non-literature is flexible. What bursts into the place of canonized art may not be noncanonized art at all, but extra-artistic phenomena; moreover, the deposed canonized art may not only descend to lower strata in the artistic hierarchy but leave the sphere of art entirely and become extra-artistic.

Though Šklovskij admitted in a letter to Tynjanov that he was impressed by this argument, his overall reaction was ambivalent. On the one hand, Šklovskij seemed to reject the concept of immanent literary development to which he earlier subscribed: “We claim, it seems, that the literary work can be analyzed and evaluated without leaving the literary series... However, the notion of literature changes all the time. Literature extends and absorbs extra-aesthetic material. This material and those changes which it undergoes while in contact with the material already aesthetically transformed must be taken into account.” On the other hand, Šklovskij insisted that once this material becomes a part of art it loses its original ties with life and becomes a component of artistic form. “Literature lives while extending over non-literature. But the artistic form accomplishes a peculiar rape of the Sabines. The material ceases to recognize its master. It is deformed by the law of art and is perceived outside of its original context (Šklovskij 1926, p. 99).

Because of the rule of exclusion underlying the binary model of mechanistic Formalism, the approach was unable to provide a description of literary change that would adequately account for the interplay of the literary and nonliterary spheres. Šklovskij’s position was inevitably contradictory. He was aware of the historical relativity of the concept of literature, but could not take full advantage of his knowledge without destroying his conceptual frame. Caught in this paradox, he was unable to offer any solution. The conclusion of his letter is an example of what Richard Sheldon termed “the device of ostensible surrender,” that is, an overt capitulation hiding covert intransigence (Sheldon 1975, pp. 86-108). “Answer my letter but do not drag me into the history of literature,” pleaded Šklovskij. “I will study art, realizing that all its dimensions [veličiny] are historical” (Šklovskij 1926, p. 100).

While evading the problem of the interaction between literature and byt, Šklovskij’s immanent literary history did offer a solution to the second problem: artistic novelty. In a succinct history of the novel, Šklovskij depicted artistic change as follows. Like all narratives, the novel’s artfulness lies in the transformation of a lifelike story (fabula) into a literary plot (sjužet). This task is complicated by the composite nature of the novel, by the fact that it is a concatenation of several short stories. The history of the novel from this perspective is a succession of different motivations for the device of fusing short stories into larger wholes. In the most elementary novels (for example, Don Quixote), it was the protagonist who strung the pieces together. After this method became automatized, the psychology of the hero was used as the connecting thread. The works of Stendhal, Tolstoj, and Dostoevskij provide ample variations on this psychological motivirovka (motivation). Eventually even this mode of fusion wore out. The audience’s interest in connecting individual pieces waned and the segments themselves began to attract attention. At this moment, motivation itself turned into a device. The individual segments were brought together in a negative way to show the reader that they had nothing in common, that their connective tissue was simply a technical device enabling the writer to make them into a novel. This is the method of modern novels, Šklovskij claims, most notably of his own epistolary novel Zoo (Šklovskij 1923e, pp. 83-85).

It is instructive to compare this history of the novel with the earlier account of literary development found in Šklovskij’s booklet on Rozanov mentioned before. Both proceed from an immanent notion of literary history driven by the opposition, de-familiarization/automatization. But whereas the Rozanov booklet presents literary change as an infinite permutation of the same poetic forms, Šklovskij’s history of the novel adds something new to this scheme. The master device of this genre—the fusion of the constituent stories into a larger whole—remains the same, but different literary periods introduce different motivirovki. What the source of these new motivations is, Šklovskij does not say, and one might intuitively surmise that it is byt. This assumption does not contradict his two-term model, for as I showed earlier, the motivation of a device for him is merely an auxiliary component of the literary construction.

Šklovskij’s foray into the history of the novel is noteworthy for yet another reason: its conception of historical process. According to this conception, the development of a literary genre is not an uninterrupted continuum, a chain of works successively de-familiarizing each other, but instead a qualitative leap, an abrupt ascent to a higher level of literary consciousness. There seems to be a qualitative difference between the way elementary or psychological novels are produced and the way their modern counterparts are. The earlier works presuppose a “naive” attitude toward writing. The author portrays characters and their psychic lives without being aware that all of this is nothing but an excuse for fusing short pieces into a novel. The modern novel is based on a self-conscious attitude toward writing on the author’s part, deliberate debunking of “deceptive” artistic practices. The modern novelist says that the emperor is naked, and by eliminating “fictitious” motivations lays the devices of his trade bare.

This ironic attitude toward literary production stems in turn from the writer’s historical self-awareness, his or her reflexiveness about the logic of literary history. For example, the “naive” novelist creates characters and events without realizing that in fact he is complying with the historical demand for de-familiarizing artistic form. The “cunning” modernist, conscious of his historical role, proceeds differently. He analyzes the present state of literature and designs his writings in such a way as to achieve the maximal artistic effect, he does not merely deviate from previous conventions, but shows that they are mere conventions. By stripping bare the very process of literary creation, the modernist de-familiarizes artistic form anew, thus reaffirming the logic of literary history.

By merging literary theory and practice, its istoria and poeisis, Šklovskij also effectively subverted Veselovskij’s objectivist literary history. For Veselovskij, the literary historian’s task was to reconstruct the causal chain of the literary series. From Šklovskij’s point of view such an approach to history writing was a mirror image of the “naive” novelist’s attitude toward literary production. Not only were novelists unaware of their actual role in the historical process, but objectivist historians seemed equally ignorant of the aesthetic presuppositions involved in their practice. Because the literary series is virtually an infinite continuum, objectivist historians had to focus on only certain works, authors, or periods. And because they were dealing with literary phenomena, the ultimate criterion for this selection was their own literary sensibility. Thus, despite its claims, objectivist historiography never actually recaptured the literary past “as it was” but always provided varying, distorted pictures of it. The remedy Šklovskij proposed was the same one he put into practice in his creative writing. Literary history should turn in upon itself and lay bare the devices of its trade, instead of the pretended reconstruction of the literary past, literary history should become “the gay business of [its] destructtion,” a self-conscious “misreading” of history according to modern artistic principles (Šklovskij 1923b, p. 220).

Hence, the job of the literary historian in Šklovskij’s view was complementary to that of the artist. The artist revitalizes literature by creating new poetic forms that replace old, automatized ones: the literary historian does so by recycling these old forms through a de-familiarizing recreation of them. “We are losing the living perception of Puškin,” Šklovskij argued, “not because our byt and language are far removed from his, but because we did not change the standard (the criterion) to which we compare him.” Aiming at his own camp, Šklovskij continued, “the study of literary traditions, the Formal study of art in general, would be utter nonsense if it did not provide us with a new perception of the work.” Therefore, he concludes, “the task of the Formal method or at least one of its tasks is not to ‘explain’ the work but to impede its perception, to renew the ‘set toward the form’ that is characteristic of the work of art” (ibid., p. 205). He put this call into practice in the same article by presenting a new Puškin—a master parodist, a Russian follower of Laurence Sterne—whose Evgenij Onegin lays bare the devices that created its literary form.

This program for a new literary history, however, did not receive much of a welcome from the Formalists. The Muscovite Grigorij Vinokur, for example, in his review of the anthology on Puškin in which Šklovskij’s essay had appeared, declared that its author “lacks any—even the most elementary—sense of history” (Vinokur 1924, p. 264). This negative reaction was in part conditioned by the fact that most of the other members of this movement did not share Šklovskij’s passion for mingling scholarship with art. Even those who, like Tynjanov, applauded the artistic boldness of Zoo and its highly unusual blend of literary theory and creative writing (Tynjanov 1977, p. 166), refused to go the full route with Šklovskij and radically relativize their notion of literary history. They viewed Šklovskij’s approach as a manifestation of aesthetic egocentrism, an ahistorical “imposition upon the past of current modes of poetic production,” for which they had already blasted the older generation of literary scholars (Jakobson 1921a, p. 5).

The rejection of Šklovskij’s approach to literary history by his comrades-in-arms had a certain justification. His reading of Evgenij Onegin was arbitrary insofar as it was motivated by his idiosyncratic literary sensibility rooted in the iconoclastic poetics of Russian Futurism. Such an orientation was clearly unacceptable to the young theoreticians striving to establish an “objective” science of literature. Yet, at the same time, one might ask whether the Formalists in their campaign against historical relativism were not blind to the historical relativity of their own enterprise. As Jurij Striedter argues, most of the later Formalist reconstructions of the literary past “did not reflect on what was principally the historical character of their own school and its system, nor did they incorporate it in any way into their theory and analysis” (Striedter 1978, p. 11). Of the Formalists, only Èjchenbaum was willing to take Šklovskij’s challenge seriously and translate it into a more cogent scholarly program for a self-reflexive historiography. “In its essence,” he wrote, “history is a discipline of complex analogies, a discipline with a dual vision: the facts of the past are discerned as significant and enter the system invariably and inevitably under the aegis of contemporary problems... History, in this sense, is a particular method for studying the present through the facts of the past (Èjchenbaum 1929, p. 49). Ultimately, one may speculate that such a stance could have developed into what modern critical theory calls the history of literary reception. But before this happened, Russian Formalism itself was transformed into a historical phenomenon.

Šklovskij’s concept of literary history constitutes a programmatic rejection of Veselovskij’s poetics. It either reversed or subverted all the crucial notions of its nineteenth-century predecessor concerning the development of literature. This is not to say that Šklovskij’s relationship to Veselovskij was purely negative. According to Èngel’gardt, whose account was quoted earlier, Veselovskij’s system involved not only literary history but also a theory of the genesis of poetry from extra-aesthetic phenomena. To this latter domain, in my opinion, mechanistic Formalism is very closely linked indeed.

Let me briefly characterize this aspect of Veselovskij’s theory. In his genetic studies Veselovskij strove to establish which phenomena of primitive culture evolve into the simplest poetic forms. In order to do so, he dissected the literary work into its smallest elements—motifs, epithets and formulas—which he then pursued across the entire range of literatures of different nations and periods. Thus, aside from its historicity, Veselovskij’s poetics can be described as genetic, inductive, and comparative.

The main thrust of mechanistic Formalism is also decidedly genetic. It tries to establish how a literary work arises from extra-literary phenomena. Šklovskij revealed his bias toward a genetic explanation when he wrote, “Phenomena can be grasped best when we can understand the process of their origin” (Šklovskij 1923b, p. 74). Because the most basic premise of mechanistic Formalism was never to seek an explanation for the facts of art among the facts of byt, its adherents disregarded all general cultural preconditions. Works of art were seen as intentional artifacts, and to grasp them meant to explain how they were made. The titles of some essays, for example, Šklovskij’s “How Don Quixote Is Made,” or Èjchenbaum’s “How Gogol’s ‘Overcoat’ Is Made,” bear witness to this genetic approach.

The titles of these essays might, however, be misleading. They seem to suggest that by focusing attention on the genesis of particular literary texts, the mechanistic Formalists were studying their actual origins as individual and unique creative acts. Nothing would have been more alien to the Formalists and the tradition that they continued. The sober positivist Veselovskij had already waged a war against the Romantic myth of the literary work as a totally subjective expression of a strong individual. Assessing the state of his discipline in 1870, he wrote, “contemporary scholarship has taken the liberty of looking at the masses, which until now have stood behind [the heroes], deprived of any voice. It has discerned a life and movement in them which, like everything else that takes place on a grand spatiotemporal scale, is imperceptible to the naked eye. It is here that the hidden springs of the historical process ought to be sought.… The great individuals now appeal as reflections of this or that movement prepared for among the masses” (Veselovskij 1940, p. 41). The author, in Veselovskij’s view, is merely a crystallization of poetic traditions and social currents existing independently of the author, and it is precisely these general preconditions of literary creation rather than any unique creative act that form the true object of scholarship.

The Formalists followed in Veselovskij’s footsteps, though instead of attacking Carlyle and Emerson they turned against more recent psychological critics. Pointing a finger at the Freudian method, Šklovskij wrote: “Least of all should one become involved with psychoanalysis. It analyzes the mental trauma of only a single man. But the single man does not write; it is the time, the school-collective that writes” (Šklovskij 1929, p. 211). As Osip Brik put it: “OPOJAZ thinks that there are no poets and literati but poetry and literature. Everything written by a poet is significant only as a part of his work in the common enterprise and is absolutely worthless as an expression of his ‘I.’ … The devices of the poetic craft must be studied on a grand scale, along with their differences from contiguous spheres of human work and the laws of their development. Puškin did not create a school; he was merely its head.” And to make his point stick, Brik declared: “If there were no Puškin, Evgenij Oneginwould have been written anyway. America would have been discovered even without Columbus” (Brik 1923, p. 213).

Given such a strong Formalist aversion to the individual aspect of the literary process, it is obvious that Šklovskij and Èjchenbaum were aiming at something other than a simple description of two disparate creative acts. Replying to a self-imposed question “what is significant about the Formal method?” Šklovskij wrote in his characteristic staccato style: “What is significant is that we approached art as a production. Spoke of it alone. Viewed it not as a reflection. Found the specific features of the genus, began to establish the basic tendencies of form. Grasped that on a large scale there is a real homogeneity in the laws informing works. Hence, the science of literature is possible” (Šklovskij 1926, pp. 64-65). What the Formalists subscribing to the mechanistic model set out to investigate, therefore, was the general technology of literary production and the laws that govern it, rather than the genesis of some randomly chosen texts. Both Šklovskij and Èjchenbaum utilized Cervantes’s and Gogol’s works as case studies to outline the broader principles that generate prosaic works in two different genres: the novel, and the short story oriented toward oral delivery.

The genetic approach was not merely a heuristic device for the mechanistic Formalists: they believed that the process of making art is intimately connected to the process of its perception. As Šklovskij wrote, “art is the way to experience the making of a thing while what was made is not really important in art” (Šklovskij 1917, p. 12). The perception of the work is thus nothing but the re-presentation of the intentional creative process which gave birth to the perceived work. And because the device is the “main hero” of this process, it should be the focus of attention for the student of literature. It is here that the inductive and comparative methods enter the scene. The literary work is dissected into such elementary devices as repetition, parallelism, gradation, and retardation, and the existence of these devices is ascertained through a comparison of the most heterogeneous materials—folksongs, tales, high literature, even film stories. The results then serve as a verification of the original premise of mechanistic Formalism about the heteromorphism of art and byt.

Earlier I noted the unenthusiastic response that Šklovskij’s theory of literary history elicited among the Formalists. The same applied to his poetics. The first disagreement with the mechanistic model concerned the ontological status of the device. According to Šklovskij, the device was the smallest universal and virtually independent element of artistic form migrating from work to work. Viktor Žirmunskij objected that it does not exist independently but only as a part of the work and its actual value is always determined by the immediate whole in which it belongs: “The poetic device is not an independent, self-valuable, quasi-natural-historical fact. The device as such—the device for the sake of the device—is not an artistic element but a conjuring trick… The same device, from the formal point of view, very often acquires a different artistic meaning depending on its function, i.e., on the unity of the entire artistic work and on the general thrust of all the other devices” (Žirmunskij 1928, p. 52).

Žirmunskij’s comment implies a second objection to inductive poetics, namely, its disregard for the holistic nature of the literary work. The mechanistic model conceives of the work as a mechanical aggregation of its parts. This seems to be the gist of Šklovskij’s slogan that the “content (soul) of the literary work equals the sum total of its stylistic devices” (Šklovskij 1925, p. 8). The critics of this notion pointed out that the literary work is not a mere aggregate, but that it possesses a certain inward quality which belongs to it only as a whole and which is lost when it is mechanically dissected into its parts. “The search for the minimal atom of the text betrays a materialistic quasi science,” wrote Boris Larin condescendingly. “Every adolescent can dissect a frog believing that he is Harvey. In the same way it is easy for everyone to follow a little matrix and list on file cards the words in a pre-Petrian tale, the epithets of Puškin or ‘sound repetitions’ in verse, or to separate the speeches from conversations in Don Quixote. The results of such an analysis can, of course, be utilized in many ways, but what I am aiming at is the inadmissibility of these oversimplified methods in obtaining the material of study itself. In stylistics we must not for a moment lose sight of the interrelation of elements, the wholeness of the artistic text” (Larin 1923, p. 62).

Indeed, though at times Šklovskij appears to be aware of the Gestaltqualität in the work of art, he has difficulty in finding its locus. As Victor Erlich has pointed out, Šklovskij’s confusion over the word “form” has its roots in this problem. “The Russian Formalist leader seemed to fluctuate between two differing interpretations of the term: he could not make up his mind as to whether he meant by ‘form’ a quality inherent in an esthetic whole or an esthetic whole endowed with a certain quality” (Erlich 1981, p. 187). It is thus not surprising that Šklovskij’s work is riddled with contradictory statements concerning the holistic nature of the literary work. He insists upon its integral nature, stating that “nothing can be subtracted from a literary work” (Šklovskij 1923c, p. 16), but then declares that “the unity of the literary work [is]… a myth” (Šklovskij 1925, p. 215). Though most of the Formalists probably would have subscribed to the first statement, only a very few would have agreed to the second. To see the literary work not as a conglomerate of devices but as an intrinsically unified whole required another perspective—a metaphor quite unlike that offered by the mechanistic Formalists.

The Organism

The spirit of poetry, like all other living powers [...] must embody in order to reveal itself: but a living body is of necessity an organized one,—and what is organization, but the connection of parts to a whole, so that each part is at once end and means!

— Coleridge, Shakespearean Criticism

Zweck sein selbst ist jegliches Tier, vollkommen entspringt es
Aus dem Schoss der Natur und zeugt vollkommene Kinder.
Alle Glieder bilden sich aus nach ewgen Gesetzen,
Und die seltenste Form bewahrt im geheimen das Urbild.

— Goethe, “Metamorphose der Tiere”

A belief in the holistic nature of the literary work compelled other Formalists to seek a different conceptual frame for their study of literature. As the mechanistic Formalists, drawing their inspiration from the realm of technology, probed into the clockwork of devices in the literary work, another group of Formalists turned to biology and its subject matter—the organism—as their model. In a methodological article, “The Boundaries of Literary Theory as a Science,” Boris Jarcho, a member of the Moscow State Academy for the Study of Arts, noted three similarities between the literary work and the biological organism: (1) both are complex wholes composed of heterogeneous elements; (2) both are unified wholes; (3) in both the constitutive elements are hierarchically differentiated, in that some are essential to the unity of the whole and others are not (Jarcho 1925, p. 59).

The literary work may be compared to the biological organism in other respects as well. Just as each individual organism shares certain features with other organisms of its own type, and types that resemble each other belong to the same species, the individual work is similar to other works of its form (for example, the sonnet), and homologous literary forms belong to the same genre (for example, the lyric). As a result of this organization, the work and the organism can be conceived generatively. New configurations both similar and dissimilar to previous ones are constantly arising, so that individual structures appear not as discrete entities but as the momentary stages of an ongoing morphogenetic process of transformation. This generative character of the organism, along with the holistic one mentioned earlier, was exploited by these literary scholars in what I shall call the morphological trend in Russian Formalism.

The name for this trend, morphological Formalism, is drawn from the writings of the Formalists themselves. For this reason we must scrutinize the name closely. The Formalists used the term in a variety of ways. Even the arch-mechanist Šklovskij sometimes referred to the Formalist movement as a “morphological school” to avoid the pejorative connotations of the label “Formalism” (Cf. Šklovskij 1923c, p. 40, or 1922b, p. 59). However, this usage did not imply that Šklovskij had consciously explored the parallel between art and the organism. Of the Formalists who did use “morphology” in its biological sense, some did so in order to emphasize the holism of the literary work and others its generative nature.

The first group cannot be defined with any precision. It includes some Formalists who subscribed fully to the organic model (Žirmunskij and A. Skaftymov), others who resorted to this metaphor only occasionally (Èjchenbaum), and still others for whom the holistic study of the literary work was just a step to the generative model (M. Petrovskij). Consequently, the term “morphology” subsumed a wide range of meanings. For Žirmunskij it was equivalent to taxonomy, which “describes and systematizes poetic devices” prior to the study of their “stylistic functions in the typologically most essential poetic works” (Žirmunskij 1928, p. 55). For Èjchenbaum, morphology meant something akin to formal anatomy (Èjchenbaum 1922d, p. 8), whereas for Petrovskij it included both the anatomy of the work (static description) and its physiology (dynamic functioning) (Petrovskij 1925, p. 182). In general, the term “morphology” was not as crucial for these Formalists as other terms, such as “organism.” But for the Formalists emphasizing the generative nature of the work—above all Vladimir Propp and Michail Petrovskij in his later writings—“morphology” was a key term and they used it in a very restricted sense. To understand this as well as the concept of “organism” we must examine the biological theories underlying morphological Formalism.

Emanuel Rádl has stated that “in biology, from the eighteenth century onwards it has been believed that the quintessence of an organism is revealed by its form and structure” (Rádl 1930, p. 120). There were two opposing theoretical views explaining the actual forms of organic bodies. Georges Cuvier (1769-1832), the father of paleontology and comparative anatomy, described the organism by proceeding from the parts to the whole, the latter conceived as the “correlation of parts.” An organism was a functional system in which each element acquires a specific position according to its function. The holistic nature of the organism and the functionality of its parts were accepted as premises by Johann Wolfgang Goethe (1749-1832) the pioneer of morphology. Goethe, however, did not proceed from the individual organism but instead from the general whole—the a priori “ultimate phenomenon’’—to the individual organism, an actual transformation of this phenomenon. He envisioned morphology as a science concerned “with organic shapes… their formation and transformation” (Goethe 1887-1912, sec. 2, vol. 6, p. 293).

Because of their different points of departure, Cuvier and Goethe emphasized two aspects of the concept of the type. For Cuvier, organisms belonging to a specific type could vary from one another only in their peripheral parts. He believed, as William Coleman has observed, that “the functionally integrated animal, a specific type, could not significantly vary in any of its parts or operations without abruptly perishing” (Coleman 1964, p. 3). Goethe’s notion of nature as a continual transformation produced an opposite view of the type. He saw biological wholes as Dauer im Wechsel(continuity in change), as creative forms or processes rather than static correlations. Ernst Cassirer succinctly summarized the difference between these two great biologists, whom he called “morphological idealists.” According to Cassirer, “Cuvier advocated a static view of organic nature; Goethe a genetic or dynamic view. The former laid its stress upon the constancy, the latter on the modifiability of organic types” (Cassirer 1945, p. 106).

With these two notions of organism in mind we may return to the Formalists. Let us begin with those who shared Cuvier’s static notion of the organism. Some of their isolated criticisms of the mechanistic model have already been mentioned. The general disagreement between the mechanistic and morphological approaches, however, is determined by their opposing notions of teleology, which must be examined more fully. In his introduction to the Russian translation of Oskar Walzel’s book, The Problem of Form in Poetry, Viktor Žirmunskij pointed out the ambiguity inherent in Šklovskij’s programmatic slogan “art as device.” On the one hand, the device provides a purposive explanation of art—as a means of affecting the perceiver’s reception (the principle of de-familiarization). On the other, it provides a functional explanation of art—as a means of affecting the teleological organization of the work (the manipulation of extra-artistic material).

Žirmunskij unequivocally rejected the purposive explanation of art. He argued that the aesthetic effect of the work is a bundle of multifarious consequences which cannot be reduced to de-familiarization alone. The perception of the work is not limited to the pure enjoyment of self-centered devices but “implicitly it includes cognitive, ethical, or religious elements” (Žirmunskij 1928, p. 161). This is especially valid for literature, Žirmunskij continues, because its material—language—is not purely formal as is the material of music, but always carries meaning. In addition to this linguistic meaning, literature as a thematic art employs translinguistic meanings which it shares with other thematic arts such as painting. Thus, Žirmunskij concludes, the strictly formalistic approach to art practiced by the theoreticians of OPOJAZ is incapable of dealing with literature in its totality and must be augmented by thematic studies. “The study of poetry as art requires attention to be paid to its thematic aspect, to the very selectionof the theme as well as to its construction, compositional elaboration, and combination with other themes” (ibid., p. 169).

The inclusion of thematics into literary studies provided Žirmunskij with further ammunition against the definition of art as an effect upon the perceiver. Thematics links literature not only to other thematic arts but to the extra-artistic sphere as well, and hence to general culture. Because the cultural configuration and the place of literature within it are in constant flux, to seek the essence of literature in the reaction of readers would be futile. Their reactions change as the culture changes, and each new reading in a shifting cultural milieu will bring about a new perception of the work. Therefore, to study literature from the reader’s point of view would lead the student of literature to a relativism that would threaten the very identity of the work. Curiously, Žirmunskij stated his distaste for Rezeptionsästhetik most clearly in his refutation not of Šklovskij but of Tynjanov. “Further research in this direction leads to a theory according to which in different periods different elements can become the dominant of the same work, i.e. can acquire ‘constructive’ relevance... In other words: the work of art is not ‘formed’ by the author but by the reader and the history of criticism and readers’ taste replaces historical poetics as the study of the change in literary forms and styles” (ibid., p. 356).

This statement, however, should not lead us to conclude that Žirmunskij identified the teleology of the artistic device with the intention of the artist. Despite his disagreement with the mechanistic Formalists, Žirmunskij shared their rejection of psychologism. “Every work of art,” he wrote in the introduction to his Byron and Puškin, “has a special kind of ideal existence—fully autonomous and independent of the subjective processes in the creator’s and perceiver’s consciousness.” Following this precept, Žirmunskij concentrated “above all on the study of works themselves. The writer’s ‘personality’ and in particular his empirical, biographical personality, his human—all too human—psychology are thus excluded, as well as the study of the milieu that educated and formed him” (Žirmunskij 1924, pp. 197-98).

Nevertheless, Žirmunskij did not subscribe to the extreme social determinism advocated by Brik. Evgenij Onegin, in Žirmunskij’s opinion, demonstrates a certain degree of poetic individuality in relation to the works of other authors, which is undoubtedly related to the idiosyncrasies of Puškin’s personality. Yet this differential quality, he insisted, is the property of the literary text, and the critic must infer it from the work itself and not from circumstances that are external to it. Thus, without denying the importance of the author for the work, Žirmunskij was not interested in the artist as a concrete psychophysical entity. Rather, he conceived of the artist as a specific final cause who gave rise to the work as a unified whole. Instead of speaking of the writer’s intentions, Žirmunskij spoke of the “unity of the artistic task” or the “general form-giving principle,” which he even called “entelechy” (Žirmunskij 1923b, p. 6)—the Aristotelian term used by the neo-vitalists at the beginning of this century.

Though he rejected the purposive implications of the formula “art as device,” Žirmunskij did endorse its functional meaning. The teleology of the device rests in the function it performs within the work. “Poetics studies the literary work as an aesthetic system determined by the unity of the artistic goal, i.e., as a system of devices [my italics]. Thus, in the study of the artistic work we consider metrical construction, verbal style, plot composition, and the selection of a particular theme as devices, i.e., as aesthetically relevant facts determined by their artistic teleology” (Žirmunskij 1928, p. 158).

Although Žirmunskij believed that this interpretation of the device was inherent in Šklovskij’s slogan, there is a substantial difference between the two theorists’ views of the role of the device within the work. For Šklovskij the device simply transformed the non-artistic material into an artistic form; for Žirmunskij the device helped to meet certain requirements within the work in which it occurred. Šklovskij spoke of the work as a “sum of devices”; Žirmunskij called it a “system of devices.” Within an additive whole, it is the presence or absence of a device that matters; within a system, the presence of the device is taken for granted and it is its interrelatedness to other devices that counts.

The concept of the literary work as a system necessitated a redefinition of the concept of the device. It could no longer be seen as a purposive manipulation of material, but instead must be viewed as a functional exploitation of this manipulation. A poetic device, for example rhyme, was not to be described as a particular sound repetition but as a functional element within the literary whole. To define rhyme as “a sound identity occurring at the end of a line from the last stressed vowel on” would be to treat it non-functionally. This definition is inadequate, according to Žirmunskij, because it concerns the sound aspect of rhyme and virtually ignores its compositional role as a marker of the rhythmic series and the strophic organization. Rhyme, said Žirmunskij, is “every sound repetition which carries an organizing function in the metrical composition of a poem” (Žirmunskij 1923a, p. 9).

The redefinition of the device introduces a further complication into the binary model of mechanistic Formalism. Šklovskij’s opposition of material to device does not allow for the functionality of the device, because it does not posit any source of unity for the functional elements of the work. Therefore, Žirmunskij decided to augment this opposition with a third term, “the teleological concept of style as the unity of devices” (Žirmunskij 1928, p. 23). This notion would account for the interconnectedness of the devices of a text as well as the essential wholeness of every work of art. “Only if the concept of ‘style’ is introduced into poetics,” Žirmunskij argued, “can we consider the basic conceptual framework of this discipline (material, device, style) complete” (ibid., p. 51).

Žirmunskij’s critique of the mechanistic model and his emphasis on the functional interrelatedness of elements within the work inspired several contemporary literary scholars. Among them, the one closest to his position was Aleksandr Skaftymov—a professor of Russian literature at the Saratov University. Though actually quite remote from the mainstream of Formalism, Skaftymov’s writings during the early twenties bear the clear stamp of what I term the morphological metaphor. His embracing of this model was most likely the result of personal contact with Žirmunskij, who was conducting a course in theoretical poetics at the Saratov University at this time. Yet, despite their sharing of the organicist view, Žirmunskij and Skaftymov reacted against two different theoretical positions. Žirmunskij argued against the relativism of Rezeptionsästhetik, whereas Skaftymov’s target was the genetic method which attempted to explain the literary work through the extra-literary phenomena surrounding its origin. As I showed earlier, this method had already been criticized by the mechanistic Formalists, who argued that the composition of a work is not determined by the factors of byt present during its creation but instead by the general laws of literary production. For the organicist Skaftymov, however, the key to understanding the literary composition lay in its inner teleology.

Skaftymov treated the work as a totality unified from without by an artistic intention which within the work becomes a form- giving dominant. “A study whose aim is to reveal the nature of a teleologically formed object must inevitably conceive of this object as a unity. This concept is then expressed in the description of the relations between the constitutive elements and the general system of coordinations and subordinations which exist within the [artistic] whole” (Skaftymov 1924b, p. 135). All the components of the work are drawn into this system of relations, including those thematic components that in some respects may exceed the limits of the work. “Elements of psychology, history, sociology, and so forth, fragmentarily contained in the work, are not interesting in themselves but only in the teleological thrust they obtain in the general unity of the whole” (ibid.).

Skaftymov applied his teleological approach most successfully to the study of byliny—Russian oral heroic epics. In his monograph, The Poetics and Genesis of Byliny, he presented a theoretical alternative to the genetic study of folk poetry elaborated by Veselovskij and his followers (Skaftymov 1924a). This so-called ethnographic school had aimed at establishing a link between the elements of the bylina and the historical events that supposedly gave rise to individual compositions. The “original’’ bylinawas presumably diluted and transformed in later renditions, so that it was difficult to discover. Nevertheless, Veselovskij’s school believed that a thorough study of the variants would ultimately lead to the kernel of the bylinawhich would directly reflect an actual historical situation. Against this conception, Skaftymov posed his opposite view that byliny are literary compositions unified from within, their elements determined by their functions within these wholes and only secondarily by their extra-literary significance.

In opposing the “ethnographic” school, Skaftymov offered an all-encompassing critique of the inductive approach in literary studies. He rejected it on two grounds. Epistemologically, he claimed, a pure induction is a fiction: “It is no secret that even observation and classification of a multiplicity of varied facts is always performed according to some a priori principle.” From a practical standpoint, he argued that the inductive method is incapable of dealing with the organic wholeness of byliny. Poking fun at the inductivists, he quipped, “we walked around it, we discerned some of its features, and without grasping their internal significance or their essence, we began to explain their growth and development. Comparing random bits and pieces of the bylina, we fragmented it, and then we combined those pieces, believing that we had recreated the extinct forms of the past. Out of a living organism we made mechanics” (ibid., p. 49; p. 43).

According to Skaftymov, any analysis of byliny as functionally integrated organisms would reveal a single, dominant, compositional goal: the effect of surprise, which all the elements of the bylina help to establish. The basic binary structure of the bylina is conditioned by this goal, consisting of two parts and portraying two main protagonists. It begins with an introduction in which the hero and his adversary are contrasted, the hero as a rather inept figure (too young, in fragile health, and so on) and the villain as the possessor of a superhuman power. The second part of the bylina depicts the fight between the hero and his enemy. From the introductory description of the two it appears that the hero does not stand a chance. Indeed, the actual fight is very short and the vanquished party gives up with only a token resistance. The appeal of the bylina, however, rests in the fact that the loser is not the underdog of the introduction—the hero—but the villain, whose success seemed to be guaranteed. The hero’s victory ends the bylina, for “immediately after the decisive moment, the progression of the plot ends; the singer has nothing more to speak about” (ibid., p. 61). Only a brief formulaic conclusion expressing the gratitude of those saved by the hero or the general joy over his victory is attached to the finished story.

All the elements of the bylina, whether formal or thematic, are subordinated to the goal of creating an unexpected solution. For example, narration and description alternate in order to reinforce the bylina’s binary articulation. In the introduction description prevails; in the fight narration does. Moreover, the description focuses solely on the two main protagonists. All the secondary characters remain underdeveloped since they serve merely as the background against which the two main characters operate.

The total subordination of elements to a single structuring principle provides Skaftymov with a base from which to attack the genetic method, which had concentrated on thematic details—the names of characters and localities, the social organizations depicted, and so on—in order to reconstruct the origins of the bylina. However, as Skaftymov convincingly argued, these details are utterly secondary in the teleological structure of Russian heroic epics. For this reason, names are freely altered from one performance to another even by the same narrator, and the social interactions among the characters do not reflect the ideology of their time but the requirements of the plot structure. Skaftymov’s conclusion that “every genetic study of the bylina requires a preliminary description of the inner constitutive meaning of its parts” epitomizes not only this study but the works of “morphological” Formalism in general (ibid., p. 127).

Although both Žirmunskij and Skaftymov understood the literary work as a functionally integrated organism, there was a slight difference in the way they conceived of this organism. Žirmunskij saw it above all as a harmony of functional parts, whereas Skaftymov saw it as a hierarchically organized whole in which the function of some parts was determined by other, dominant ones. This divergence results from the different aspects of Cuvier’s zoological theories emphasized in each Formalist. To Žirmunskij, paleontology seemed the more valid metaphor. If the work was a system of parts whose functional correlations constituted a harmonious and unchangeable whole, it resembled more a dead fossil animal than a living, changing organism. Indeed, Žirmunskij compared the task of the student of style (the crucial concept of his art theory) to that of the paleontologist. “Just as a paleontologist can reconstruct from a few little bones of an unearthed animal—provided he knows their function—the entire structure of the animal, the student of artistic style… can reconstruct in general terms an organically integrated structure, ‘predict’ its presupposed forms” (Žirmunskij 1928, p. 51).

On the other hand, Skaftymov’s treatment of literature, especially the bylina, was more like Cuvier’s comparative anatomy and theory of the biological type. Skaftymov was keenly aware that there was considerable variability among the individual byliny, but he saw this variability as limited to secondary elements which were functionally insignificant. Cuvier’s notion of the variability of the individuals of a given species was quite similar. As William Coleman has observed, “Cuvier did not deny the existence of variation. His plan was to reduce variation to its proper limits, and the anatomical rules provided the initial key to the problem. From the primary fact of the integral harmony of the organism it was recognized that certain organs were more important to the animal than others: heart, lungs, nervous system were more important than hair, skin, color, or size. These circumstances demanded a certain stability or invariability of the central organs and permitted the almost unlimited variation of peripheral features” (Coleman, 1964, p. 143). In like manner, the essential elements of the bylina—its two main protagonists and two narrative sequences—would correspond to the indispensable parts of the organism and its secondary features—the names, social status of the heroes, societal mores present in it—would correspond to the hair, skin, coloring, and so on.

It is interesting that Skaftymov’s concern with the variability of the individual compositions belonging to a genre helped to prepare the way for another brand of morphological Formalism, which I shall discuss presently. This approach was inspired by Goethe’s transformational concept of organic form. Skaftymov’s characterization of the bylina, for example, has a distinctly Goethean ring: “Everything in the bylinais in flux. Its existence always was and will be in an uninterrupted creative process begun no one knows where, when, or by whom. The bylinais not something ready-made, but is always in a state of becoming” (Skaftymov 1924a, p. 36). Despite this assertion, Skaftymov’s study pursues not the process of becoming per se, but rather what was stable and unchangeable in it. As a search for the functional invariant in a genre, it is quite different from those Formalist genre studies searching for transformational rules.

Vladimir Propp and Michail Petrovskij were the two most prominent Formalists to transfer Goethe’s morphology from the organic to the literary form. Although the relationship between the “static morphologists” and Cuvier was only implicit (his name is absent from their writings) the “transformational morphologists” proclaimed their spiritual indebtedness to Goethe openly through the epigraphs in both Propp’s Morphology of the Folktale and Petrovskij’s “Morphology of the Short Story.” The importance of these epigraphs for a proper understanding of their conceptual framework cannot be overstated. Propp himself made this clear in his answer to Lévi-Strauss’s review of the 1958 English edition of his book, claiming that the omission of Goethe’s epigraphs in the English version had caused Lévi-Strauss to misunderstand his method.”6The historical context surrounding the publication of the two “morphologies” is also noteworthy, for they followed very closely on the publication of Goethes morphologische Schriften by Wilhelm Troll. According to Horst Oppel—the historian of the morphological method in German literary-studies—this publication “paved the way for the acceptance of morphology as method” (Oppel 1947, p. 13). Troll’s edition of Goethe appeared in 1926, Petrovskij’s “Morphology of the Short Story” in 1927, and Propp’s book in 1928.

Besides the external signs of kinship between Goethe’s method and those of Propp and Petrovskij, there is an essential similarity in their epistemological presuppositions. Goethe constructed morphology as a science on the assumption that despite all the heterogeneity of organic phenomena a single underlying principle unites them. This idea occurred to him during a trip to Italy in 1786, where he encountered new and exciting plants. “In this new manifold I came across here the following idea became more and more vivid to me: namely that all the forms of plants perhaps developed from a single form. This in itself would enable us to define species and genera correctly...” (Goethe 1887-1912, sec. 1, vol. 30, p. 89). His search for the archetypal plant or animal (Urpflanze or Urtier) of which all the actual forms of a given species were metamorphoses is paralleled in Propp’s and Petrovskij’s search for the archetypes underlying all the actual forms of the two genres that they dealt with—the fairy tale and the short story, respectively. And just as Goethe conceived of organic forms as processes rather than products, the two Formalists defined their genres in terms of transformations, not as sets of fixed features. Significantly, each quoted Goethe’s statement, ‘‘Gestaltenlehre ist Verwandlungslehre” (the theory of forms is the theory of transformations), Propp choosing it as the epigraph for his eighth chapter and Petrovskij as the motto for his entire study (ibid., sec. 2, vol. 6, p. 446).7 This Goethean principle was the basis for their literary inquiries.

Propp’s book is well known so I shall deal with it only briefly. It is noteworthy that his motives for studying fairy tales were similar to Skaftymov’s for oral heroic epics: dissatisfaction with the genetic approach previously used. We can even catch an echo of Skaftymov’s stress on descriptive over genetic investigation in Propp’s claim that the discussion of the morphology of fairy tale must precede the question of its historical roots. “Historical studies may appear more interesting than morphological investigations... but the general question of where a tale comes from remains, on the whole, unsolved, though even here laws of genesis and development undoubtedly exist that are still awaiting elaboration… However, we maintain that as long as there is no correct morphological study there can be no correct historical study. If we do not know how to dissolve the tale into its constituent parts we cannot carry out a comparative study... [And] if we cannot compare one tale with another how can we study the relation of the tale to religion or myth?” (Propp 1928a, p. 26) This quotation is indicative of the unique direction Propp took in his morphological study of the fairy tale. Unlike all the other Formalists employing the biological model, he accepted the challenge of inductive poetics, attempting to isolate the smallest constituent of the genre he studied. In fact, he accused Veselovskij—the main representative of inductive poetics—of not being analytic enough, pointing out that the motifs he advanced as the minimal elements of narrative were readily divisible into smaller units.

Propp was not merely more analytic than Veselovskij; the real difference between them was the manner in which they tackled the problem of the minimal unit. This difference resembles the contrast between the “mechanistic” and “morphological” concepts of literature. Veselovskij, a true inductivist, believed that the part is prior to the whole, not only for the sake of the descriptive procedure but in the genesis of the work as well. For this reason, in describing individual motifs he paid no attention to their relationship to the wholes they composed, since the latter were posterior combinations. Propp’s organicism prevented him from being an inductivist of this type. He agreed with Veselovskij that the “part is prior to the whole for descriptive purposes,” but he would not claim that it was prior in an absolute sense. His definition of the minimal unit of the fairy tale treated it teleologically in terms of its role within the whole. On the most abstract level, he conceived of the fairy tale as a narrative about actions performed by certain characters. And it is the actions and not the interchangeable characters, that count. Characters as carriers of these actions, are functionally indispensable, but what is important is not their individuality but their function, that is, their “action defined from the point of view of its relevance for the course of action.” Thus, Propp’s definition of the minimal unit of the fairy tale as the “function of acting characters” differs from Veselovskij’s notion of the motif as minimal unit, not so much in that the former is smaller than the latter, but that it is a part of a functionally integrated whole, whereas the motif is a part of a mechanical aggregate (ibid., p. 22; pp. 30-31; p. 29).

From what was just said, it might seem that Propp’s notion of the functionality of organic parts matches Cuvier’s, but in fact it is quite close to Goethe’s. Goethe insisted upon the functional definition of parts over the static description. “Function correctly grasped is the being conceived in activity.” Thus, when “we are concerned with the human arm, [we are in fact dealing] with the front legs of an animal” (Goethe 1887-1912, sec. 2, vol. 7, p. 200). The variety of forms these limbs can attain is almost unlimited, but by acknowledging their functional similarity a morphologist can study and compare them. It was through just such a functional reduction that Propp succeeded in establishing thirty-one elements as the basic units of every fairy tale. These elements do not exist in isolation but are interlocked in a configuration—the compositional scheme of the fairy tale. The final test of Propp’s method is not only to prove that all fairy tales are composed of the same elements but of the same elements in an identical sequence. By comparing the schemes of various tales Propp arrives at the invariant—the ultimate Ur-Typ of which all fairy tales are transformations.

After discovering the generic invariant of the fairy tale (what Propp called the composition), he might have been expected to outline the laws governing its transformations. This aspect of morphology is conspicuously missing from his book, however. Instead, he discussed transformation in an article in the fourth volume of Poètika published by the State Institute for the History of the Arts. “Transformations of the Fairy Tale” appeared separately from The Morphology of the Folktale, for reasons that I shall soon discuss. First let us consider the morphological theories of Michail Petrovskij.

Petrovskij’s attempt at a morphology of the short story differs from Propp’s, despite their common model. Petrovskij shows not the slightest interest in proceeding inductively from the minimal units of the short story. The elements of narrative with which he operates are defined functionally, but are certainly not the simplest possible. Moreover, the material the two morphologists consider differs. Propp analyzed a genre that was no longer a vital art form. Petrovskij’s object of study, on the other hand, was very much alive at the moment he attempted to describe it. Consequently, his definitions are much less formalized than Propp’s. The two genres also differ in their structures. In the short story there are two temporal sequences—that of the narrated event itself and that of its presentation. In the fairy tale both the number of elements and their sequence are fixed. The only thing that can vary is the appearance of the performers of the functions. Therefore, while Propp could present a single sequential formula for all fairy tales, Petrovskij had to account for two levels—the “disposition” or temporal sequence of events and the “composition” or narrative sequence of these events.

The pair, disposition-composition, does not coincide precisely with Šklovskij’s opposition of story and plot. Petrovskij, unlike Šklovskij, did not believe that the material of a prose work was life as such. Instead he emphasized that life as the material of literature “is always restructured life… it is always a selection” (Petrovskij 1927, p. 72). Literary material is a semantically unified configuration, a life endowed with specific meaning. For this reason Petrovskij often used the term “plot” to designate what Šklovskij meant by “story.” In general, the terms “plot” and “disposition” are interchangeable in Petrovskij’s system.

The place of the short story as a genre, according to Petrovskij, exists between the anecdote and the novel. What distinguishes it from the novel is that it contains only a single event. It differs from the anecdote in treating this single event not in isolation but as part of a larger context. Conceived in this way, the minimal scheme of a short story’s disposition contains three parts: the “kernel of the plot” (that is, the event itself), and the two connectors that link it to its larger context—the Vorgeschichte or as Petrovskij hesitantly translated it in a footnote, “the plot prologue,” and the Nachgeschichte, “the plot epilogue.” The composition of the short story, or the presentation of its plot, has a corresponding three-part scheme. First is an introductory “exposition” which leads toward the “climax” of the story (naprjaženie) and culminates in what Petrovskij calls the pointe, the moral of the story. The middle term of both the disposition and the composition of the short story, the “kernel of the plot” and the “climax,” can be further subdivided into the “complication” (zavjazka), the “climax proper” or “knot of the plot” (the moment of highest tension), and the “resolution” (razvjazka). This scheme can be visualized as shown in the diagram.


In the second part of his study Petrovskij illustrates the transformations of the basic scheme of the short story through specific examples. The fourth tale of the first day of the Decameron is the simplest story analyzed. Its content is rendered succinctly in the short synopsis that introduces the tale. “A monk, having fallen into a sin deserving a very grievous punishment adroitly reproaching the same fault to his abbot, quitteth himself of the penalty” (Boccaccio 1925, p. 30). The complication of this story arises when the abbot surprises a young monk with a girl in his cell in flagrante. To escape the punishment the monk pretends to leave his cell and go to the forest to collect some wood, hoping that the abbot himself will fall into sin with the girl. This indeed happens and is secretly witnessed by the young monk. The moment of the highest tension follows when the abbot calls the young monk and threatens him with prison for his deed. The crisis ends in a happy resolution for the monk, who reveals to the abbot that he knows as much about the abbot as the abbot knows about him and thus “quitteth himself of the penalty.”

Because this tale is a short story and not a simple anecdote, the event does not appear in isolation, but is introduced by the Vorgeschichte in which the situation of the event and its two main protagonists are described. In the composition of the tale, this description functions as the exposition of the event, preparing the way for the climax. Symmetrically, at the end of the tale the event is concluded by the Nachgeschichte consisting of a single sentence which describes the new relationship among the members of the triangle after the event: “Accordingly, [the abbot] pardoning him and charging him to keep silence of that which he had seen, they privily put the girl out of doors and it is believed that they caused her return thither more than once thereafterward” (ibid., p. 32). This sentence, Petrovskij argues, is not only the Nachgeschichte of the plot but also the pointe of the composition. The phrase “and it was believed...” goes beyond merely connecting the event with a larger context but involves the presentation of this event by the narrator. In general the morphology of this tale follows closely the basic scheme of the short story without any transformations. All the parts of the scheme are present and there is no discrepancy between the disposition and composition. For this reason, Petrovskij calls this tale, using Goethe’s terminology, the Urphenomenon of the short story (Petrovskij 1927, p. 76).

In contrast to the simplicity of Boccaccio’s tale, de Maupassant’s short story “Le Retour” represents a radical transformation of this basic genre scheme. Šklovskij once observed that this story is a variation of the famous plot, “a man at the wedding of his wife,” which differs from the others of its type by presenting this surprising plot in a rather low-key manner (Šklovskij 1919d, p. 120). Petrovskij’s analysis amplifies this impromptu observation. He characterizes the kernel of the plot as the “return of the husband who disappears without a trace after his grass widow marries someone else and starts a new family” (Petrovskij 1927, p. 81).

The striking feature of this story is that its composition differs from its disposition. The narration begins with an exposition describing the seaside, the cottage of the family Martin-Levesques, and its inhabitants. The complication starts when one of the girls notices the reappearance of a stranger three times in one day. After the exposition, comes the first part of the Vorgeschichte telling why the family has a hyphenated name. But immediately afterward, the narration returns to the present and describes a hostile dialogue between the stranger and Mrs. Martin-Levesques. In the evening when Mr. Levesques returns, the stranger disappears. The event recurs the second day but now the complication changes into the knot of the plot. Mr. Levesques, who that day remained home, speaks to the stranger and finds out that he is no other than Mr. Martin. The second part of the Vorgeschichte which follows explains that he did not die in a shipwreck as was believed, but was captured instead by savages who held him for twelve years. The two husbands decide to solve their problem by going to the priest. On the way they stop at a cafe for wine and the story ends with the following dialogue: “And the tavern-keeper, three glasses in one hand and a carafe in the other, approached, large of paunch, ruddy, fat, and asked with a quiet air: ‘What, you here, Martin?’ Martin replied: ‘I am here’” (Maupassant 1903, vol. 17, p. 137).8

This abrupt ending seems to leave out some basic parts of the short story scheme—the resolution, pointe, and Nachgeschichte. Petrovskij argues that Maupassant’s story represents a radical transformation of this scheme rather than a truncation of it, however. He compares the resolution of “Le Retour” to a draw in a game of chess. “A game of chess can end with the victory of the white or of the black side, but can also end in a draw. The meaning of the draw arises from the entire preceding game but it in turn provides the game with meaning. After a great dynamic tension everything results in zero” (Petrovskij 1927, p. 85).

The pointe of the story rests precisely in this “incomplete resolution.” It forces the reader to “shift retrospectively the semantic center of the story from the facts to the attitude toward them… The irony of the story consists in the fact that in this ordinary fishermen’s milieu an unusual conflict loses its unusualness, becoming colored by the gray, indifferent light of its heroes’ psyche” (ibid.). But in addition to this “incomplete resolution” Petrovskij argues that the story does contain the equivalent of a resolution which suggests the outcome of the event. This is the conversation of the two male protagonists before they go to the priest. There Martin proposes to keep the house and in return not to press any demands for his wife’s return. Although readers are left in suspense as to whether this proposal is the actual resolution, it presents them at least with a plausible possibility. This “equivalent of the resolution” then serves as a functional equivalent of the Nachgeschichte linking this single event to the larger context of life.

Though Petrovskij goes on to analyze two other short stories, the two examples discussed so far are sufficient for a general understanding of his method. As I pointed out earlier, the main difference between the static and transformational morphologists was the latter’s intention to go beyond a discovery of the invariant of a genre to outline the rules governing the transformations of the invariant in individual literary works. It is important to ask whether they were successful. Goethe had outlined a basic “double law” governing the formation and transformation of all organic wholes: “(1) the law of internal nature according to which plants are constituted, and (2) the law of external circumstances according to which plants are modified” (Goethe 1887-1912, sec. 2, vol. 6, p. 286). Petrovskij completely ignores the relation of the literary work to external circumstances. In this respect, he is even more radical than Šklovskij in purging extra-literary phenomena from literary studies. Though the relation of literature to byt in Šklovskij’s system was secondary, it was at least implicitly present, since life was considered the material of literature, but Petrovskij cut even this link to extra-literary phenomena by declaring literary material pre-poetic, that is, structured according to the requirements of literature. The spiritus movens of transformations must therefore be in the internal nature of the genre itself. What it is, however, we may only guess. It is not the tension between what Šklovskij termed canonized and new forms, a notion that would explain a particular transformation at a particular time, nor can it be an inner necessity stemming from the basic scheme of the short story that Petrovskij had outlined. Instead of a theory of transformations we are presented with a set of ad hoc rules which pertain to individual transformations within the stories analyzed but are far removed from constituting the Verwandlungslehre of the genre.

Propp’s attitude toward the transformational rules of the fairy tale is more complex than Petrovskij’s. As we have seen, he does not discuss it in The Morphology of the Folktale, though he mentions in the introduction that the original manuscript had contained a section on this issue that was dropped (together with some other parts of the manuscript) for stylistic reasons.9In the same year that the book appeared, Propp published the article mentioned earlier dealing with the topic he had omitted in the book.

Yet after reading this “spin-off” piece one begins to doubt that mere stylistic reasons had led Propp to omit it from the larger text. More likely it was his failure to elaborate any general transformational theory that prompted his decision. For a Goethean morphologist, the elaboration of transformational rules is as important as the isolation of the generic invariant. To eliminate this issue “for the sake of brevity and a more vivid presentation” seems a rather high price to pay, especially by someone who otherwise demonstrates little consideration for his reader. The omission casts considerable doubt on the legitimacy of the term “morphology” used in the title of the book. And from his remarks addressed to Lévi-Strauss (quoted earlier) it is obvious that the author himself was not unaware of this fact. “To be absolutely precise,” he wrote, “I should not have spoken of ‘morphology’ but used the much more restricted concept of ‘composition’ and called the book The Composition of the Folkloric Fairy Tale” (Propp 1976, p. 140).

Propp’s article “The Transformations of the Fairy Tale” in conjunction with his book shows that, unlike Petrovskij, he takes into account both aspects of Goethe’s “double law.” The book discusses the constitution of the genre as a particular configuration of functional elements, whereas the article deals with the external circumstances that modify this genetic invariant. As Propp argues in the latter, “the causes of transformations often lie outside the tale, and without taking into account comparative material from the environment of the tale, we shall not grasp its evolution.” Propp hastens to add that the external causes do not modify the whole fairy tale but only some of its parts: “There is a great difference between organic formations and the fairy tale. Whereas in the first, the change in one part or feature causes a change in another, in the fairy tale every part can change independently of the other parts” (Propp 1928b, pp. 72-73). Instead of offering general rules explaining the particular modifications of the basic scheme in different milieus, Propp provides four criteria for distinguishing the variants of a part of a fairy tale from the original one (a fantastic treatment is prior to a rational one, a heroic to a humorous one, and so on) and twenty modifications which a single element might undergo (reduction, amplification, corruption, and so on).

Propp’s search for the transformational rules of the genre led him into problems with the biological metaphor, because unlike other morphological Formalists, he overextended it. Despite many similarities, there is obviously an essential difference between a literary and a genuine organic whole: literary works are intentional objects endowed with an immaterial meaning but organisms are empirical objects whose proper existence is in the realm of material reality. The other morphologists were keenly aware of this difference. In fact one of their main arguments against the mechanists was that they reduced the literary work to a mere formal construction and paid little if any attention to literary semantics. Instead of such a monistic notion of the literary work, the morphological Formalists conceived of it in a dualistic manner—as a unity of the formal construction (we might say, of the material vehicle) and theme (semantics in the broadest sense of the word). Propp did not accept this dualistic vision. In pursuing the organic metaphor, he conceived of the fairy tale as an empirical object and analyzed it not as a semantic but a formal construction.

Whether this division of a work into formal and thematic components is justified is another matter. Nonetheless, this distinction is a handy way of discussing the category of the “function of an acting character,” which Propp found so crucial. The monist Šklovskij had treated characters as primarily a part of the formal construction. For example, he claimed that Don Quixote was a device for stringing disparate motifs together into a narrative whole. Don Quixote’s characteristics per se were irrelevant: Šklovskij shows that they actually change as the narrative unfolds. What remains constant is Cervantes’s use of that character in his manipulation of the material. On the other hand, the dualist Skaftymov argued that the formal aspects of the prose work are subordinate to its thematics. Therefore, he analyzed the way in which the characters of Dostoevskij’s Idiot function within the overall unity of its theme. He was especially interested in the traits of literary figures, examining their actions and interactions as contributions to their characterization. He considered the deep inner conflicts within Dostoevskij’s characters and the discord among them as supporting the general theme of the novel, the dialectic resolution of contradictions through forgiveness. It is obvious that Propp’s conception of the function of a character is closer to Šklovskij’s than to Skaftymov’s. Šklovskij and Propp do differ, of course: in Šklovskij’s opinion, the character of Don Quixote links disparate motifs; in Propp’s view it is the fairy tale characters’ actions that create linkage by necessitating the actions of other characters. But both theorists treat the character as merely a part of the formal construction.

I began this discussion by accusing Propp of overextending the biological metaphor, for in treating the literary work as a formal construction, he was reducing it to an empirical object. This assertion requires some clarification, because Propp selected only certain empirical characteristics of the fairy tale as crucial for his morphological analyses of the genre. The most important of these was its temporal extension. Propp conceived of the fairy tale as a narrative unfolding in time as a string of events. All the constitutive parts of the fairy tale that he considered relevant are related to the temporal flow of the narrative, whereas all static or atemporal features are dismissed as secondary.

In his morphological analyses Propp operates with two types of formal units—the simple and the complex—which he terms “functions” and “composition,” respectively. The functions, that is, the functions of an acting character, participate in the temporality of the narrative because when one appears, the other necessarily follows, until their entire sequence (the composition, the basic generic scheme) is complete. The other type of simple unit in the fairy tale does not contribute to the narrative flux; this is what Propp calls the static element. “A motif like ‘Baba-Jaga gives Ivan a horse’ consists of four elements of which only one represents a function [the verb]; the others are static” (ibid., p. 71). The static elements are the attributes of the acting characters which make up what we might call the thematic aspect of the tale. Because they do not influence the narrative flux, however, Propp treats them as accidental embodiments of the functions, irrelevant to the morphology of the tale. The static elements combine in actual fairy tales with functions, or better, provide the latter with flesh and blood, and in the predetermined sequence they create the “unique” plot of the variants of the fairy tale. Propp does not pay any attention to what might be called the overall theme of the tale. Whether it is Baba-Jaga who gives the horse to Ivan or Ivan who gives it to Baba-Jaga, the temporality remains the same. For Propp, the plot of the fairy tale is nothing but an actualization of the fairy tale’s composition—the narrative flux itself. As a result, Propp was indignant when Lévi-Strauss in the review referred to “plot” as “theme”: “for a folklorist and a literary scholar, the ‘plot’ is the center of attention. In Russian the word ‘plot’ as a literary-theoretical term has acquired a very specific meaning: the totality of the actions and events which are unfolded in the course of narration… However, for Professor Lévi-Strauss the plot is uninteresting, he translates it into French as ‘theme.’ He most likely prefers it because ‘plot’ is a category pertaining to time whereas ‘theme’ lacks this feature. There is, however, no student of literature who would accept such a substitution. We can understand these two terms in many ways but never can we identify them or substitute one of them for the other” (Propp 1976, pp. 145-46). Needless to say the term “theme” does not appear in Propp’s morphological investigations of the fairy tale.

Propp’s conception of the fairy tale as an empirical, temporally extended object led him to stress the formal units that constitute narrative flux and to disregard the fairy tale’s semantics. Admittedly, this radical reduction paid off in his search for a generic invariant, for the wealth of semantic nuances had blinded earlier students of folklore to the formal regularity of the fairy tale. As soon as transformation is the issue, however, all those features, all the semantic nuances that differentiate one tale from another, become crucial. Indeed, Propp’s genre definition disregarded these very features. As Lévi-Strauss jokingly observed, “Before Formalism we were certainly unaware of what these tales had in common. Since Formalism, we have been deprived of any means of understanding how they differ” (Lévi-Strauss 1976, p. 133). Propp cannot have his cake and eat it too. One cannot have transformational rules without treating semantic features as elements of a system. Because a fairy tale is not an empirical but an intentional object, the static elements and their relations must be taken into account if we are to grasp the unity of the fairy tale in its process of transformation.

Lévi-Strauss convincingly showed that the semantics of the fairy tale is crucial. The specific acting characters that fulfill a function are not altered arbitrarily. For example, three birds may fulfill a function in a certain tale: an eagle, an owl, and a crow. Though on the formal level it makes no difference which bird fulfills the function, from the semantic point of view these birds are opposed to each other in significant ways. The eagle is diurnal, whereas the owl is nocturnal; as predators both are opposed to the scavenger crow (ibid., p. 135). From this example it follows that the acting characters are not accidental embodiments of minimal functions but partial meanings whose dynamic interplay encompasses the overall structure of the meaning of a given fairy tale. They are interconnected, and a change in one leads to a change in all the others.

The Formalists who were inspired by Goethe’s concept of morphology actually failed to realize their goal—to isolate the transformational rules of a literary genre. Despite their claims, they were ultimately quite close to the “static morphologists.” Even though they attempted to grasp literature as a process, they succeeded only in pinning down the invariants of the genres they studied. By conceiving of these invariants as a functional correlation of parts they arrived at the same organic metaphor as the morphological Formalists who were proceeding from Cuvier.

The application of the biological metaphor to literature demonstrated that the literary work is not a conglomerate of devices but a functionally integrated whole whose elements are determined by the role they fulfill in the literary organism. When they attempted to cross the boundaries between literature and extra-literary phenomena, however, the morphologists did not fare well. Though they criticized the mechanistic Formalists for their radical separation of art and byt, they themselves were unable to bridge this gap. Instead they replaced the mechanistic opposition of art and byt with the regular versus the accidental. In other words, they admitted that the internal organization of a literary work is subject to influences from the nonliterary world, but they saw these influences as random and secondary to an understanding of the inherent regularities of literature.

The inability to bring literature and life together quite strongly affected the morphological Formalists’ attitude toward literary history. They rejected the mechanists’ immanent approach, but as long as they saw the extra-literary sphere as incidental to the internal constitution of literature they could not develop a systematic explanation of literary change. Concerned with the identity of literature in its internal regularity, they had no place in their theories for the vicissitudes of history. Thus they willingly traded the insecurity of change for the certitude of identity, diachrony for synchrony. For them the theory of literature was independent of and prior to its history.

In his 1922 inaugural lecture at Saratov University, Skaftymov separated the theoretical and historical aspects of literary studies, giving precedence to theory. “I contrast the theoretical to the historical view on the following points: (1) A theoretical knowledge grasps the object in its inner constitution; a historical study views the object in the process of its becoming, (2) A theoretical study takes into account the holistic correlation of the constitutive elements of the object; a historical knowledge is concerned with cause and effect relations (causality)” (Skaftymov 1923, pp. 55-56). The theoretical approach alone, Skaftymov believed, is adequate for the treatment of a literary work as an aesthetic object and all historical facts play a merely auxiliary role in it. Moreover, a history of any phenomenon can be studied fully only after its identity is established theoretically. Quoting Žirmunskij—another Formalist relying on the morphological metaphor—Skaftymov declared, “Only a ‘theoretical poetics’ can construct the system of scholarly concepts which the historian of literature needs for solving his concrete historical problems” (ibid., p. 67; Žirmunskij 1921, p. 51). This notion surfaces six years later in Propp’s book on the fairy tale in a passage quoted earlier: “Historical studies may appear more interesting than morphological investigations… However, we maintain that as long as there is no correct morphological study there can be no correct historical study” (See Propp 1928a, p. 26).

The rift between theory and history and the privileged position the morphologists accorded synchronic studies were not, however, shared by all Formalists. It was in response to these issues that a third Formalist model arose which sought to treat literature as a strictly historical phenomenon.

The System

No poet, no artist of any art, has his complete meaning alone. His significance, his appreciation is the appreciation of his relation to the dead poets and artists. You cannot value him alone; you must set him, for contrast and comparison, among the dead. I mean this as a principle of aesthetic, not merely historical criticism.

— T. S. ELIOT, “Tradition and the Individual Talent”

I call the third Formalist model “systemic” because it uses the metaphor of the system as its primal frame of reference. The role of systemic Formalism was to fill the gaps left by the other two metaphors; to describe the relationship between art and byt and provide an account of literary history capable of explaining the dynamic interplay between these two domains.

The name “systemo-functional” was chosen by Jurij Tynjanov, the main proponent of the model, to designate his approach to literary studies (Tynjanov 1977, p. 295).10It points aptly to the holistic and relational nature of the approach. These features, too, indicate the link between the systemic metaphor and new developments in such other disciplines as psychology, logic, and linguistics. I would like to outline briefly the way advances in these areas helped to shape the systemic metaphor of Russian Formalism.

One of the leading Gestalt psychologists, Kurt Koffka, devoted part of his Zur Analyse der Vorstellungen und ihrer Gesetze [On the Analysis of Representations and their Laws] (1912) to the distinction between “descriptive” and “functional” concepts in psychology.11 Descriptive concepts like “color” or “image” are those which involve direct experience and “derive from simple perception and the descriptions of experiences” (Koffka 1964, p. 238). Functional concepts, such as “distortion of memory” go beyond simple perception. They are used “to put experience into relation with other objects, either with other experiences or with stimuli” (ibid.). Koffka concludes that “all functional concepts have as their basis experiences that have somehow been made objective. This kind of concept formation is of the same type as the formation of concepts [in physics]” (ibid., p. 242).

The literary scholar must also distinguish between concepts relating to the direct experience of literary texts and concepts that bring these into categorical relation. Tynjanov expressed this distinction as the opposition between “literary fact” and “literature.” He noted that “whereas a hard definition of literature is more and more difficult to make, every contemporary can point his finger at what is a literary fact. He will tell you that this or that as a fact of byt or of the poet’s private life” is not a literary fact, “while something else certainly is” (Tynjanov 1929, p. 9). “Literature” is a notion of an order distinct from the literary fact. It is a functional concept relating notions of direct literary experience, Koffka’s descriptive concept. This distinction, which at first glance might appear obvious, was quite important to the systemic Formalists. Until it had been elaborated, literary critics frequently identified literary facts with literature in general, confusing a particular literary sensibility, for instance, with the theory of literature itself.

The rise of the relationalist outlook at the turn of the century was obviously fostered by new advances in the sciences. But it would have been unimaginable without the support of the philosophers and logicians who provided its epistemological justification. In his influential book Substanzbegriff und Funktionsbegriff (1910) Ernst Cassirer declared that the “two chief forms of logic which are especially opposed to each other in the modern scientific development, are distinguished… by the different value which is placed upon thing-concepts and relation-concepts” (Cassirer 1923, p. 9).12 The “thing-concept,” whose origin Cassirer traces back to Aristotle, is characteristic of traditional concept-formation based on the process of abstraction. A general concept, it was believed, was derived from particulars by abstracting their similar feature. The fallacy of this approach, according to Cassirer, rests on the presupposition that similarities are not merely a principle of logical ordering but real properties of objects. Thus, in the process of abstraction what is non-essential to objects is eliminated in order to discover their unchangeable substance. With this substantialist view Cassirer contrasts the relationalist one, in which similarity is not considered a property of objects but a categorical tool that enables us to unite disparate objects in a single concept. Thus, similarity is one of many possible principles of logical ordering that give rise to “relation-concepts.” As Cassirer explains the process, ‘‘all construction of concepts is connected with some definite form of construction of series. We say that a sensuous manifold is conceptually apprehended and ordered, when its members do not stand next to one another without relation but proceed from a definite beginning, according to a fundamental generating relation, in necessary sequence. It is the identity of this generating relation, maintained through changes in the particular contents, which constitutes the specific form of the concept” (Cassirer 1923, p. 15).

There are several clear points of contact between Cassirer’s and Tynjanov’s theories. Most important is their common use of the mathematical function as a model for concept-formation in general. Quoting the German logician Moritz Drobisch, Cassirer asserts, “Every mathematical function represents a universal law, which, in virtue of the successive values which the variable can assume, contains within itself all the particular cases for which it holds.” Moreover, this concept of function “is not confined to mathematics alone,” but “extends over into the field of the knowledge of nature” (ibid., p. 21). Tynjanov, taking Cassirer’s lead, crossed even this boundary and applied the notion of function to the study of cultural phenomena as well.

Concept-formation in literary studies is more complex than in psychology or the natural sciences. The task of a psychologist or a physicist is to match two sets of givens: Cassirer’s “objects of the first order” (or Koffka’s descriptive concepts) and “objects of the second order”—those concepts “determined by the form of the generating relation from which they proceed” (Koffka’s functional concepts) (Cassirer 1923, p. 23). Besides these two givens, however, the student of literature must also deal with “objects of a third order”—socially shared sets of conventions which determine the existence and identity of the objects of the first order.

An awareness of this problem most likely came to Tynjanov through his acquaintance with the theories of Ferdinand de Saussure. The Swiss linguist had pointed out that speech phenomena are implementations of the underlying linguistic system shared by speakers of each particular language, a system he termed la langue. His example of the knight in chess illustrates this notion well. The identity of this piece is purely relational. If the physical piece is lost during the game the knight can be replaced by any object, even something as different as a matchbox. The equation of two such objects is not the same as bringing them together in the concept-forming mode I have just described. There the knight and the matchbox would be subsumed under a single concept through a logical relation introduced from outside the game. In the game, however, their relationship is generated from within because the matchbox, like the knight, becomes liable to the same set of rules—the game itself. The substitution of one object for another depends on “an unchangeable convention, the set of rules that exists before a game begins and persists after each move” (Saussure 1959, p. 88). In the same way that a piece in a game of chess derives its identity from an underlying system of rules, the identity of a linguistic fact is a function of the underlying linguistic system—la langue. As Saussure argues, the socially shared linguistic code “is necessary if speaking is to be intelligible and produce all its effects” (ibid., p. 18).

The analogy between language and literature is obvious. The identity of every literary fact is determined by sets of norms we call genres, schools, or historical styles. Significantly, even the fact that an utterance is considered literary is determined by the existence of a social habit we call “literature.” Thus Tynjanov asks, “Is the so-called immanent study of a literary work… outside of its interrelation with the literary system possible?” The answer is negative: “Such an isolated study of a work is a mere abstraction similar to the abstracting of an individual element from the work” (Tynjanov 1929, p. 34). A literary work is inseparably linked to the literary system, and outside this context loses its identity.

Tynjanov’s distinction between “literary fact” and “literature” and between both these concepts and “literary system,” and his relational approach to concept-formation all show his affinity to the theories of Koffka, Cassirer and Saussure. Tynjanov, however, departed from all these thinkers in two significant respects: he approached his material dialectically and historically. Perhaps it was the tradition of Hegelianism in Russian intellectual life that led him to conceive of literature as a dynamic hierarchy, an ongoing struggle for domination among parts and wholes. It is this intrinsic dynamism of literary structures that Tynjanov identified as the distinctive feature of literature. “Literature is a speech construction perceived precisely qua construction, i.e.. literature is a dynamic speech construction” (Tynjanov 1929, p. 14).

Here the concept of the “dominant” enters Tynjanov’s system, which according to Jakobson’s later assessment “was one of the most crucial, elaborated, and productive concepts of Russian Formalist theory” (Jakobson 1978, p. 82). But because of its wide currency among the Formalists, we must differentiate among its various usages. The term itself was borrowed from Broder Christiansen’s Philosophie der Kunst. In discussing the perception of a work of art he wrote, “It happens only rarely that the emotive factors of an aesthetic object participate equally in the effect of the whole. On the contrary, normally a single factor or a configuration of them comes to the fore and assumes a leading role. All the others accompany the dominant, intensify it through their harmony, heighten it through contrast, and surround it with a play of variations. The dominant is the same as the structure of bones in an organic body: it contains the theme of the whole, supports this whole, enters into relation with it” (Christiansen 1912, pp. 241-42).

The notion of the dominant as a skeletal, form-giving element in the static hierarchy of holistic correlations caught the fancy of some of the Formalists. Boris Èjchenbaum, who was responsible for this borrowing from Christiansen, occasionally used the term in this sense. In his analysis of Anna Achmatova’s early poetry he tried to isolate the “essential dominant determining the major facts of a style,” in this case, her “striving for laconicism and energy of expression” (Èjchenbaum 1923, p. 63). This meaning of the dominant fit very well the conceptual frame of those morphological Formalists who discussed literature as an organism. Thus, according to Skaftymov, the role of the literary scholar was to “reveal the interrelations of the work’s compositional parts, to point out the emerging dominants and among them the final concluding and all-embracing point which in turn was the basic form-creating intention of the author” (Skaftymov 1924b, p. 24). On a more empirical level, Žirmunskij spoke of metaphor as the “capital device, the stylistic ‘dominant’” of Aleksandr Blok’s poetry (Žirmunskij 1928, p. 221).

For the systemic Formalists, however, it was Èjchenbaum’s reinterpretation of Christiansen’s term that was accepted instead. Èjchenbaum used “dominant” to refer to a specific element within a literary work which is brought into the foreground and “deforms” to its needs all the other elements. He saw the work not as a harmonious correlation of parts and wholes but as a dialectic tension among them. “The work of art,” Èjchenbaum argued, “is always the result of a complex struggle among various form-creating elements; it is always a kind of compromise. These elements do not simply coexist and ‘correlate.’ Depending on the general character of the style, this or that element acquires the role of the organizing dominant governing all the others and subordinating them to its needs” (Èjchenbaum 1922b, p. 9). In the type of lyric poetry Èjchenbaum analyzed, the dominant was intonation, because it deformed the other aspects of the poems—syntax, word order, and so on.

Keeping in mind this notion of the dominant we may consider Tynjanov’s definition of literature. The perception of a speech construction qua construction is based, according to Tynjanov, on our awareness of the hierarchical organization of such a construction caused by the tension between the dominant and the subordinated elements. “Art lives through this interplay, this struggle. Without the sensation of subordination, the deformation of all the factors by the factor fulfilling the constructive role, there would be no fact of art... If the sensation of the interplay of factors (necessarily presupposing the presence of two elements the dominating and the subordinated) vanishes, the fact of art is obliterated: it becomes automatized” (Tynjanov 1924, p. 10).

This is Tynjanov’s most basic definition of literature, but it omits one important point. For Tynjanov, the very identity of a literary fact rests in its relation to the underlying literary system: “Whether a fact is literary or not is a function of its differential quality (i.e. whether it is related either to the literary or the extra-literary series)” (Tynjanov 1929, p. 35). This means that not every strikingly organized speech construction will be perceived by us as literary. It is rather the other way around: the special perceptibility of a speech construction comes about only through its comparison to other speech constructions considered by us literary. Thus, a construction that appears merely “usual” can, at one moment, become a literary fact because of the unusual nature of the immediately preceding literary tradition against whose background it is perceived, and vice versa. “Transrational language [zaum’ ] always existed in the language of children and mystics, but only in our time did it become a literary fact. And, on the other hand… charades, logogriphs are children’s games for us, but in Karamzin’s period in which verbal trifles and the play of devices were foregrounded, they were a literary genre” (Tynjanov 1929, p. 15). Thus, not perceptibility per se but perceptibility vis-à-vis the literary system is, for Tynjanov, the opposite pole of automatization.

By applying the opposing values, perceptible/automatized, to the literary system, Tynjanov exposed the relativity of the notion of the literary fact. That the literary system is a social institution and as such liable to change means the literary facts of different periods might be quite dissimilar. How then is it possible to construct a literary series, to discover a relation that would encompass under the single category of literature all the disparate literary facts? Because the relativity of literary facts is historical. Tynjanov found the answer to this question in literary history. “Only in evolution,” he claimed, “can we analyze the ‘definition’ of literature” (ibid., p. 14). Literary facts of various periods, disparate in themselves, become related if they are placed within a concrete historical process and viewed according to the logic of this process.

Tynjanov conceived of the logic of literary history dialectically. To be meaningful, the perceptibility of a speech construction needed an opposite—the automatization of this perception. Literary change is triggered by the tension between these two. “Evolution is caused by the need for a ceaseless dynamics. Every dynamic system inevitably becomes automatized and an opposite constructive principle dialectically arises” (ibid., p. 15). The life of a literary fact is the vacillation of a linguistic construction between these two poles. It is lifted from the sphere of automatization to replace some older constructions, which in the course of time have become automatized; for a longer or shorter period of time it is perceptible, only to become automatized again and replaced by some newer constructions.

Thus, the literary series conceived historically is an ongoing struggle of dialectically opposed speech constitutions. It is a succession of literary facts which exhibit contrastive principles of construction. From this perspective only a negative definition of literature is possible. The identity of the literary series rests in a constant negation of its identity by its members.

Literature as a concept did not, however, occupy a central position in the theories of the systemic Formalists. The true crux of their thought was the notion of the literary system—the ultimate arbiter of what is and what is not a literary fact. As I pointed out above, Tynjanov derived this concept from Saussure’s langue— the linguistic system underlying the facts of speech. Certain critics of Tynjanov, such as Viktor Vinogradov, claimed that his theory was nothing but a “re-telling of Saussure in literary-historical terms” (Vinogradov, 1930, p. 24).13 In my opinion this judgment is a polemical exaggeration. It takes only a brief glance at Saussure’s concept of langue to see how different it is from Tynjanov’s literary system.

First of all, Saussure’s langue is static, devoid of any evolutionary dynamics. In fact, he declared it incompatible with history, as his famous division of linguistics into synchronic and diachronic studies attests. Further, Saussure saw changes in langue as catastrophic. They are brought about at random from the outside, and once they penetrate the system they destroy it and establish a new system different from the previous one. For this reason the system of langue is absolutely autonomous. As the concluding words of Saussure’s Course state: “the true and unique object of linguistics is language studied in and for itself” (Saussure 1959, p. 232).

Tynjanov’s “literary system” differs front langue in every one of these respects. The separation of synchrony from diachrony was utterly alien to his historical orientation. “The juxtaposition of synchrony and diachrony,” Tynjanov and Jakobson wrote in 1928, “was the juxtaposition of the notion of system to that of evolution; it becomes meaningless as soon as we recognize that every system exists in evolution and on the other hand that evolution is necessarily systemic” (Jakobson & Tynjanov 1928, pp. 36-37). Tynjanov took to heart Šklovskij’s notion that opposing literary schools—the canonized and non-canonized—coexist in every literary period. The literary system is not a balanced, harmonious structure like langue but is intrinsically unbalanced, torn by conflicting tendencies to preserve the status quo and to change it. Such a system simultaneously contains its past and points to the future. Its past is contained in constructions that have been literary facts: its future rests with the constructions negating this automatized past which are about to become literary facts. It is impossible to extract from this ongoing process an atemporal slice of the synchronic “present,” Tynjanov declared. “The literary epoch, the literary present, is not at all a static system opposed to the dynamic, evolving historical series. The same historical struggle of different layers and formations which exists in the diachronic historical series goes on in the present” (Tynjanov 1929, p. 11).

If the literary system at every moment contains developmental tendencies, Saussure’s belief in the asystemic and catastrophic nature of changes in langue does not apply to it. The developmental character of the system also makes untenable Saussure’s claim that the external impulses for change are totally random. They appear random only if viewed in separation from the literary system, from the perspective of the developmental regularity of the literary system there is no randomness. To underline this difference, the systemic Formalists divided Saussure’s diachrony into two categories: “the genesis of a literary phenomenon” on the one hand, and “its evolutionary significance, its place in the evolutionary series” on the other (ibid., pp. 12-13; see also p. 31). The specific origin of a literary phenomenon is a cross-section of many impulses—biological, psychological, social—and so in its full complexity might be random. But the fact that this configuration of extra-literary factors was incorporated into the literary series, that it crystallized into a literary fact—an element of literary history—can always be explained in reference to the evolution of the literary system. Thus, while “it is impossible to construct a genetic history of literature,” it is quite possible to write a history of the literary system (ibid., p. 386).

An especially clear illustration of Tynjanov’s claim is Èjchenbaum’s dispute with Lev Trockij. To discredit the Marxist approach to literary history Èjchenbaum shrewdly employed an example that Trockij himself had used against the psycho-biological interpretation of art: can J. M. W. Turner’s role in the evolution of European painting be deduced from the fact that he suffered from astigmatism? For the young Trockij, this was an inadmissible reduction of a social to a biological phenomenon—a stance subsequently applauded by Èjchenbaum. But when Trockij later attacked the Formalist concept of literary history, he denied art its specificity by conceiving of its evolution as an extension of class struggle. There is, of course, a difference between treating art through a biological and a sociological frame of reference, and one could argue that sociology is the more relevant concern. But this answer would not satisfy the Formalists, for whom both biology and sociology were capable of explaining only the genesis of the work and not its evolutionary significance. “Art has its specific ‘sociology’ and its laws of evolution,” Èjchenbaum argued in the tones of a literary historian. “If they tell us that a writer was psychologically a representative of a certain class, it is just as true as that Turner was an astigmatic, but ‘it does not concern me’ because these are facts of a different order [than artistic facts]” (Èjchenbaum 1927, p. 286).

In other words, not every Russian nobleman born in the 1820s turned out to be a Tolstoj, nor did every astigmatic painter born in the late eighteenth century produce work of Turner’s quality. The reason for the systemic Formalists’ rejection of the possibility of a genetic history of literature is that the number of extra-literary impulses instrumental in a literary change is limitless. Only those impulses that mesh with the developmental tendencies of the literary system have a chance of influencing the system. In Tynjanov’s words. “An ‘influence’ can be successful at a time when there are literary conditions for it and in the direction indicated by those conditions” (Tynjanov 1929, p. 46).

The belief in the immanent development of the literary system might seem to bring the systemic Formalists close to Saussure’s view of the absolute autonomy of langue. If everything literary is determined solely by the preconditions of the literary system, this system would indeed be “in and of itself the unique object of literary studies.” This similarity to Saussure is hard to deny and has a historical justification. Saussure and Tynjanov’s emphasis on the autonomous character of their systems was meant to establish their respective fields of study as independent disciplines. It would be wrong, however, to see Tynjanov’s position on this issue as absolutely set and inflexible. He effected a gradual relativezation of the original Formalist position on the autonomy of the literary system.

Only at the very end of the movement, though, did the systemic Formalists succeed in advancing a coherent theory of the relative autonomy of the literary system. I refer here to Tynjanov’s and Jakobson’s nine-point thesis written in 1928. In this scheme, Tynjanov rejected his strictly deterministic conception of the literary system according to which the domination of one principle of construction necessarily and unequivocally causes the rise of a single contrastive principle, which in time becomes the new dominant. Instead he proposed a more pluralistic view according to which several new principles of construction different from the dominant emerge and struggle for control. Moreover, Tynjanov recast his entire concept of the relation between literature and extra-literary phenomena. He conceived of the entire culture as a complex “system of systems” composed of various subsystems such as literature, science, and technology (Jakobson & Tynjanov 1928, p. 37). Within this general system, extra-literary phenomena relate to literature not in a piecemeal fashion but as an interplay among systems determined by the logic of the culture to which they belong. Thus, among all the pretenders to dominance in the literary system, the one that converges with the developmental tendencies of the overall cultural system becomes the victor.

This, of course, is a highly abstract scheme which—because time ran out for the Formalists—they never put into action. Nonetheless, it indicates the road the systemic metaphor was taking to release literature from the social vacuum into which it had been forced by the Formalists’ belief in the autonomy of the literary system. By the same token the theses demonstrate the deep-seated difference between Saussure’s and Tynjanov’s thought, making a simple equation of their theories impossible.

So far, I have discussed systemic Formalism only in relation to other fields of knowledge, but it is also useful to compare it with the other two Formalist metaphors. Systemic Formalism is the most advanced stage of the movement. That it was qualitatively different from the other models was obvious to its contemporaries. In 1927, Viktor Žirmunskij, for example, felt compelled to add a footnote to his 1919 review article of the OPOJAZ anthology Poetics, in which he termed Tynjanov a “neo-Formalist” in order to distinguish his approach from the “original” Formalism (Žirmunskij 1928, p. 356).

It is worth specifying precisely what that difference is. For example, at first glance it might appear that systemic Formalism resembled the morphological model with its notions of system and function. In fact, the coincidence of vocabulary is a matter of homonymity and not a sign of any conceptual affinity between the two Formalisms. The morphological approach used these terms in a biological sense, whereas the systemic one used them in a mathematical-logical sense. For the former, “function” denotes the role an element performs within a whole; this whole is a system because it is an interplay of functional elements held together by what Žirmunskij once called the “unity of artistic goal.” For the systemic Formalists, function was the relation of the interdependent variables, and system a hierarchical set of interdependent variables.

In general, these two Formalisms were mutually antipathetic. The systemic Formalists perceived the morphologists as mere fellow travelers, whereas for the morphologists those who considered literature a system were extremists and radicals, Žirmunskij, after his split with OPOJAZ in 1922, was the most hostile of the morphologists, and the systemic Formalists never tired of accusing him of academic eclecticism. Žirmunskij quite properly objected to some of the extreme postulates of the systemic metaphor, such as the immanent development of the literary system and its strict determinism, which as I noted earlier was later abandoned by the systemic Formalists themselves. Žirmunskij’s critique lacked effect because he was unable to offer a viable alternative hypothesis as to how literature is connected with the overall development of culture or what brings together all the disparate human activities of a particular historical moment. Instead of elaborating these problems Žirmunskij hid behind a smoke screen of vague terms such as “the uniform perception of life,” “the psychological background of an era,” or the “uniform life tendency,” which he had borrowed from contemporary German aesthetics (Cf. for example Žirmunskij 1928, pp. 55-58).

The relation of systemic to mechanistic Formalism was quite different from its relation to morphological Formalism. Members of the two groups were personal friends and their theories tended to overlap. Tynjanov accepted many of Šklovskij’s key concepts and freely acknowledged his debt. Such surface similarity should not obscure the important differences between the two Formalisms, which transcend the metaphoric divergence to involve the mode of concept-formation underlying each model. Tynjanov did not passively borrow Šklovskij’s terms but always reformulated what he borrowed, fitting it into a different conceptual frame.

In Cassirer’s terms, one might say that the “thing-concept” dominated Šklovskij’s thinking and the “functional concept” was crucial to Tynjanov’s. Šklovskij proceeded from the assumption that an unchangeable literary essence—“literariness”—was intrinsic to every literary phenomenon. On the infra-literary level, that is, the level of elements composing the work, he isolated the device as a monad of literary form endowed with “literariness” regardless of its context. On the highest extra-literary level, the level of all human activities, Šklovskij drew a sharp line between phenomena with a literary essence and those without it. For him, the facts of literature were incompatible with the facts of byt.

The middle infra-literary level, the level composed of literary works themselves, played havoc with the mechanistic metaphor. If Šklovskij’s theory had been ahistorical, this level would have posed no problem to him, for he considered all literary works to be essentially the same, differing only in the way they were made. Because, as I argued earlier, such was not the case, he encountered difficulties. To maintain the separation of literature and life he had to locate the source of this change within literature itself. For this reason, he introduced the new opposition of “canonized” and “non-canonized” literature. But this opposition was incompatible with the substantivist nature of mechanistic Formalism. If all literary works were literary, but some at a given moment were more literary than others, it is not an unchangeable essence but a changeable relationship among works that constitutes literariness.

This was the point of departure for systemic Formalism. Unlike Šklovskij, Tynjanov did not locate the differential quality of literary phenomena in the phenomena themselves. Instead he found them literary by virtue of the relation in which they participated at the level of a single work, literature in its totality, and the whole national literature of a given time.”14 Each of these was a system for Tynjanov, a set of interdependent variables, no element enjoying a privileged status prior to its incorporation into the appropriate system. Moreover, Tynjanov considered systemic not only the organization of each level but the interrelations among the levels as well. Thus, the minimal system—the literary work was a variable in the higher literary system, and in turn this system was a variable in the ultimate cultural system.

At the infra-literary level, Tynjanov warns against the futility of any inquiry into the constitutive parts of a literary work that separates them from their context. “Analyses of the isolated elements of a work—plot and style, rhythm and syntax in prose, rhythm and semantics in verse—were enough to convince us that the abstraction of these elements is permissible to some extent as a working hypothesis, but that all these elements are correlated and interacting. The study of rhythm in verse and in prose revealed that the very same element performs a different role in a different system” (Tynjanov 1929, p. 33). Thus, the literary purport of a device is derived solely from the context into which it is incorporated and the material of a literary work is not determined by its extra-literary substance but only by its place in the literary construction. Clearly aiming at Šklovskij, Tynjanov wrote: “It is self-evident that ‘material’ is not at all the opposite of ‘form’; it is also ‘formal’ because there is no material which would be external to a construction… Material is that element of the form that is subordinated for the benefit of the foregrounded constructive elements” (ibid., p. 15).

Tynjanov follows the same pattern with the narrative aspect of the literary work, which Šklovskij had split into the literary “plot” and the lifelike “story.” I have pointed out that some Formalists expressed misgivings about the manner in which Šklovskij differentiated between these two notions. They argued that the story is not merely a sequence of events but a semantic structure—a sequence extracted from its context and endowed with meaning. Tynjanov agreed with this qualification. As he wrote, “story is the entire semantic scheme of the action” represented in the literary work (Tynjanov 1977, p. 341). But his conception of plot and how it is related to “story” was different from the other Formalists’. He did not see plot merely as a literary redistribution (composition) of the sequence of events but as something more intimately related to the overall structure of the work. “The plot of a work, is defined as its dynamism comprised of the interplay among all the correlations of material… stylistic, story-related, and so on” (ibid.). Story—the configuration of events depicted in the work—is only one among many variables in this process.

Story is thus related to plot as a partial configuration to the complex configuration encompassing the work as an overall system. However, this part/whole relation must not be viewed as static. As Tynjanov stressed several times, “the unity of a work is not a closed symmetrical whole but an unfolding dynamic integrity; among its elements stands not the static sign of equation and addition, but always the dynamic sign of correlation and integration” (Tynjanov 1924, p. 10). The relation of the story and plot was no exception to this rule. In every literary work (lyrical poetry included) a struggle goes on between the two. In some works, for example the traditional novel, the semantics of events clearly dominates the overall structure of the work, whereas in others the plot unfolds outside the story. In both cases it is the relationship between them that exerts a decisive influence on the overall meaning of the work.

It is important to notice that as the systemic metaphor developed, its treatment of the infra-literary level underwent a gradual expansion. In the beginning Tynjanov was primarily interested in the relations among the textual elements themselves, but in the course of time he began to focus more and more on the vertical connections between this and higher levels and their impact upon the relations among the infra-literary elements. In the earliest stage of his career, “deformation” was the term he used to describe the makeup of a literary work.15It was a set of hierarchically related elements in which the dominant (or as Tynjanov often calls it, the “constructive factor”) deforms to its needs the “material,” that is, all the other subordinate elements. While the constructive factor and material are variables in the sense that any linguistic element can become the dominant of a work, the subordination/superordination relation is constant: it is precisely this hierarchical tension among the elements of a speech construction that renders it a literary fact.

It became obvious to Tynjanov that there was a constancy in the constructive factor and material of different literary works. Genre and any other systems larger than the work determine the hierarchical arrangement of elements within it. Thus, the simple notion of deformation was subsequently replaced by a more comprehensive concept, the “principle of construction,” which denotes the deformation of a specific material by a specific constructive factor. Tynjanov’s probes into the difference between prose and poetry, for instance, revealed that the “principle of construction in prose is the deformation of sound by meaning,” whereas the “principle of construction in poetry is the deformation of meaning by sound” (Tynjanov 1977, p. 55). As long as poetry is perceived as different from prose, the internal organization of every poetic work will be based upon the deformation of meaning by sound regardless of the specific form this deformation takes. In this way the principle of construction vertically integrates the system of a single work into the overall literary system and renders the relations of the infra-literary elements a function of the next higher level.

In the last stage of his theoretical career, Tynjanov attempted to link the infra-literary textual elements to the extra-literary level as well. He introduced the notion of the “constructive function” of an element that consists of two simultaneous relations: infrarelations proper, which he called the “syn-function” or the relations of an element “to the other elements of a given [work-] system:” and intra-literary and extra-literary relations, which he termed the “auto-function” or the relations of an element “to the similar elements of other work-systems and even of other series” (Tynjanov 1929, p. 33). This distinction resembles to some extent the Saussurean opposition between syntagmatic and associative relations in language, the first being in praesentia vis-à-vis the other elements of the syntagm in which they occur, and the second in absentia, “present” only in the linguistic system. The different modalities of these relations are reflected in the fashion in which Tynjanov links the syn- and auto-functions. The auto-function is potentially the precondition of the constructive function of an element within the work, but the syn-function determines its actual constructive function. Tynjanov offers the following illustration. An archaism appears in a literary work. Its existence there is determined by its auto-function, the relation of this word to the lexical system of a given language. But its syn-function—its incorporation into the work—determines whether the archaism serves as a lexical signal of high style (Michail Lomonosov’s usage) or of an ironic standpoint (some of Fëdor Tjučev’s archaisms).

Tynjanov’s treatment of the intra-literary level was equally relational. Šklovskij had set aside his substantivism in treating it, so it is no surprise to find here a confluence of the mechanistic and systemic metaphors. In particular, the concept of parody used in Šklovskij’s studies of Sterne and Puškin is echoed in some of the earliest of Tynjanov’s work.

For Šklovskij, parody was above all a means of de-familiarizing automatized literary forms through the laying bare of automatized devices and the displacement and violation of customary literary norms, and its aim was to provide us with a new perception of literary form. “The appearance of Tristram Shandy,” Šklovskij argued, “was motivated by the petrification of the devices of the traditional roman d’aventure. All of its techniques had become totally automatized. Parody was the only way to rejuvenate them. Evgenij Onegin was written… on the eve of the rise of a new prose. The molds of poetry were cooling off. Puškin dreamt of writing a prosaic novel; rhyme bored him” (Šklovskij 1923b, p. 219).

In his earliest studies, Tynjanov exhibited a keen interest in works oriented toward other works, especially parodies and stylizations. The similarity between the two lies in the fact that “both are leading a double life: behind the plane of the work stands the second plane, the stylized or parodied one” (Tynjanov 1929, p. 416). Apart from this kinship there is an important dissimilarity between them. In a stylization the plane of the work is congruent with the background: in a parody there is an incongruity between the two planes. This notion of parody approaches Šklovskij’s. It is this incongruity of the new and the old, the parodying and parodied, that shakes out perception and renders the literary form de-familiarized. Tynjanov differs from Šklovskij in his use of the concept of parody, however. As Jurij Striedter observes, “while for Šklovskij parody serves first and foremost as the testing and verification of his previously formulated thesis of art as estrangement, for Tynjanov the literary-historical analysis of parodistic texts and the subsequent ‘theory of parody’ are the starting point for a... theory of literary evolution” (Striedter 1977, p. 459).

For literary evolution, conceived as a struggle for domination of different elements, the “dialectical play of devices” in parody becomes an important vehicle of change (Tynjanov 1929, p. 455). Nikolaj Nekrasov’s parodies of Lermontov’s poems are a case in point. Nekrasov arrived on the Russian literary scene in the 1840s after the long domination of Romantic poetry which, in the works of Puškin (1799-1837) and Lermontov (1814-1841), established the canon of Russian verse. The clumsiness and prosaic quality of Nekrasov’s poems contrasted sharply with this smooth and elegant tradition, although his role in the development of Russian poetry proved considerable. As Tynjanov put it, “The ‘impossible,’ unacceptable form of Nekrasov, his ‘bad’ verses, were good because they displaced automatized verse, because they were new” (ibid., pp. 11-12; see also Èjchenbaum 1927, pp. 77-115). Thus, Nekrasov’s early parodies of Lermontov’s poems were an important element in the process of literary change toward post-Romanticism. “The essence of his parodies does not rest,” according to Tynjanov, “in the mocking of the parodied but in the very sensation of the displacement of the old form through the introduction of a prosaic theme and vocabulary’’ into poetry (Tynjanov 1929, p. 401). And although the mechanism of Nekrasov’s parodies was quite simple, “the combination of elevated rhythmical-syntactic figures with ‘low’ themes and vocabulary,” they marked a departure from the Romantic canon (ibid.).

As Tynjanov further elaborated the systemic metaphor, his view of the infra-literary level broadened and he eventually transcended the mechanistic model. He realized that not only parodies and stylizations but all literary texts are directed toward other works. The identity of a work in respect to genre, style, or school, indeed its very identity as literature, is based on its relations to other literary works through the underlying literary system. The principle of construction—a special relation between the dominant constructive factor and the subordinate material—was the means Tynjanov used to link the internal organization of a work to the appurtenant literary system. Tynjanov even went so far as to identify the principle of construction with the literary system itself. Every speech construction exhibiting a particular hierarchical organization of linguistic elements perceptible to us becomes by virtue of this a literary fact.

The connection between the principle of construction and the literary system is especially apparent from a developmental perspective. As soon as an automatized “principle” is negated by a new principle, its systemic existence becomes clear. For only if we conceive of the new principle as a dialectic negation of the old literary system can we perceive its implementation as a literary fact and not merely a mistake.16 On the other hand, the new principle must be system-creating, must be implemented in more than a single “accidental” speech construction. Tynjanov’s model of literary change thus contains four stages: “(1) the contrastive principle of construction dialectically rises in respect to the automatized principle of construction; (2) it is applied—the constructive principle seeks the easiest application; (3) it spreads over the maximal number of phenomena; (4) it is automatized and gives rise to a contrastive principle of construction” (Tynjanov 1929, p. 17).

In 1928 Tynjanov replaced the principle of construction with the “literary function” in a wholesale revision of his terminology. He conceptualized the three literary levels I have discussed as three sets of functions: the constructive function corresponding to the infra-literary level, the literary function to the intra-literary level, and the social function to the extra-literary level. This shift was not a question of mere nomenclature: there were important conceptual differences between, for example, the principle of construction and the new “function.” Tynjanov characterized the principle of construction as a “concept which changes and evolves constantly,” (ibid., p. 16) whereas the literary function undergoes a much more gradual change, evolving “from epoch to epoch” (ibid., p. 41).

It is probably wrong to find in this terminological shift the hint of a more static and restrained view of literary change. In fact, if we scrutinize the meaning of “system” in regard to each pair of terms, we discover that they are neither contradictory nor incompatible. Tynjanov uses “system” for entities as different as the works of one author, literary styles and schools, genres, and even prose and poetry in general. Naturally, each of these subsystems evolves at a different speed. To draw a parallel from social history—a favorite ploy of the Formalists—we must distinguish between recurring coups d’état, which simply recycle the ruling elite, and genuine social revolutions which establish new economic-political formations. Literary evolution has both frequent coups and rare genuine revolutions. Though Tynjanov fails to provide us with any clear picture of the hierarchy of literary subsystems or a timetable of their evolution, the principle of construction does seem to apply to more limited subsystems which change rapidly, whereas the literary function applies to more general and hence more stable subsystems.

The functional concept formation of systemic Formalism was also apparent on the extra-literary level. This level was inaccessible to the mechanists because they programmatically separated art from byt, literature from life. From very early in his theoretical career, Tynjanov questioned the rationale behind this artificial distinction. “I do not object,” he wrote polemically, “to the ‘relation of literature and life.’ I only doubt whether this question is properly posed. Can we say ‘life and art’ when art is ‘life’ as well? Do we have to seek some additional utility of ‘art’ if we do not seek the utility of ‘life’?” (Tynjanov 1924, p. 123).

This assertion was not meant to deny literature an identity of its own. In fact, it was just the other way around. Byt is an amorphous conglomerate of the most disparate phenomena. Against the background of this nebulous domain the various specialized human activities stand out—the arts, science, technology—which in themselves are systems. These systems introduce specific functions among the heterogeneous phenomena by either incorporating their forms into a system in the course of its development or by rejecting them. Thus, when the fact of byt is rendered a function of a particular series it becomes a fact of that series (for example, a literary fact), or, on the other hand, after losing its affiliation with that series it turns into a fact of byt. As Tynjanov wrote, “byt is teeming with the rudiments of various intellectual activities. It is made up of a rudimentary science, rudimentary art and technology: it differs from a full-fledged science, art, and technology in the way that it deals with [phenomena]. The ‘artistic byt ’ is thus different from art in the role art plays within it, but they touch upon each other in the form of the phenomena [they both deal with]” (Tynjanov 1929, p. 19).

Defining literature as a “dynamic speech construction,” Tynjanov saw “byt [as] correlated with literature primarily through its speech aspect,” since speech phenomena exist in both byt and literature. For this reason he termed the “most immediate social function” of literature its speech function (ibid., p. 42).17Our language behavior is a complex structure of various forms, patterns, and modes of discourse—some of them well defined, others more fluid—which evolve alongside the entire structure of human communication. In every historical period these forms of discourse are differentiated according to which series they belong to: some are considered literary; others belong to byt. But from the developmental perspective, the boundaries between these two domains are far from being fixed and the forms of discourse vacillate between them. According to Tynjanov, “Every linguistic fact of byt has multifarious and complex functions which are interlocked in a struggle. Under certain conditions one of these functions—the literary—becomes foregrounded”; at this moment a linguistic fact turns into a literary fact (Èjchenbaum & Tynjanov, 1926, p. 10). The process works the other way round as well: a literary fact becomes automatized, its literary function recedes, and it turns into a neutral linguistic fact—a fact of byt .

Tynjanov called this intricate interplay of literary and extra-literary discourse ustanovka. The term is very resistant to translation or explanation. It has two common meanings in Russian, as Jurij Striedter has pointed out: “intention” on the one hand, and “orientation,” on the other, “the idea of positioning oneself in relation to some given data” (Striedter 1978, p. 2). From Tynjanov’s point of view, these meanings have a serious drawback—they are teleologically founded. In both cases what is implied is a psychological subject of action who either projects his intentions into the object he creates or whose orientation (mental attitude) is instrumental in the act of perception. Neither the “intentional” not the “affective” fallacy stemming from these meanings of ustanovka accords with the objectivist thrust of systemic Formalism. This model strove to replace the psychological subject of the literary process with transpersonal, self-regulating systems. Tynjanov emphasized several times that in his usage ustanovka is devoid of all its teleological, intentional connotations (Tynjanov 1929, p. 43 or p. 49).

Through this usage of the term Tynjanov tried to express an important feature of the literary system. In adjusting itself to extra-literary modes of discourse, the literary system exhibits a self-regulating quality characteristic of all teleological processes. This quality is not introduced from without through a psychological subject; it is an intrinsic properly of the literary system.

The point Tynjanov intended to make in this rather clumsy way is grasped today in the distinction between “goal-intended” and “goal-directed” behavior, or between teleology and “telenomy.” In discussing Jakobson’s concept of linguistic change, Elmar Holenstein has provided a succinct summary of this distinction. “Goal-intended behavior is based on conscious ideas, convictions, wishes, and intentions. These act as the cause of a particular behavior.” In contrast, “a process is designated as goal-directed when it evokes the appearance of goal-intended behavior but no consciously acting subject is discernible.” The essential feature of a goal-directed or “telenomic’’ process, according to Holenstein, is its “directive correlation”: “a process is regarded as telenomic when it is bound to another process in such a way that it not only causes it, but is in turn steered in its own course by the other process” (Holenstein 1976, pp. 119-20).

In deciding whether Tynjanov’s ustanovka is a directive correlation, let us look first at an application of this concept to some literary material. Perhaps the best illustration is Tynjanov’s discussion of the transition from Classicism to Sentimentalism in Russian literature. There, the ustanovka of the dominant genre of Russian Classicism, the ode, was toward the rhetorical speeches delivered before large audiences. The techniques of this type of discourse are clearly echoed in the “oratorical” odes of Michail Lomonosov (1711-1765), the best-known poet of the period. Lomonosov’s odes were persuasive in thrust, trying to sway the listener’s opinion. His rhetorical stratagem was not to appeal to the listener’s reason, but to his or her emotions. To achieve this goal, he structured his odes by combining distant and heterogeneous elements: the unusual nature of such combinations was calculated to have a maximal emotive impact.

Moreover, the ustanovka toward an oral delivery highlighted several other feature’s of the ode. The intonational tone aimed at the richest possible changes in vocal height, and the stanzaic structure was subordinated to this aim. Copious sound repetitions in euphonic and onomatopoeic constructions also forced the phonic aspect into prominence. And the “oratorical” ode achieved a semanticization of sound. Not only phonemes but meters were linked to particular meanings. In addition to accentuating the sound level, the ustanovka toward an oral delivery made possible the use of gestures as rhetorical means. These became semanticized through a secondary code of what Tynjanov called the “gestural illustrations” of odes (Tynjanov 1929, p. 61). Sometimes these “illustrations” actually became the main vehicle of meaning: words turned into stimuli for specific gestures. Finally, the imagery of Lomonosov’s odes was also geared toward emotive persuasion. Here combinations of semantically distant words (motivated often by sound) resulted in a change in the habitual meaning of these words, in a semantic shift capable of affecting the listener’s emotions.

By the end of the eighteenth century the ode had become automatized. It began to be used in non-literary ways, as a salutory speech or supplicatory verse, for example, so that it gradually became a fact of byt. This transformation, Tynjanov points out, did not affect the genre of the ode alone: the entire canon of high Classicism was becoming automatized. A new principle of construction arose, and a “small emotion, small form came to the fore” (ibid., p. 21). In fact, the entire system of social intercourse changed. The new environment of salons cultivated the art of conversation, a discourse light and personal, playful and sociable. The ustanovka of the literary system rendered many of the forms of social intercourse literary. Especially important among them, Tynjanov claims, was the epistolary form in which the new Sentimentalist principle of construction found its optimal implementation. “Implicit meaning, fragmentariness, the small ‘household’ form of the letter, all of this motivated the introduction of trifling subject matter and devices in contrast to the ‘grandiose’ devices of the eighteenth century.” From a fact of byt, the letter became an important literary genre. Karamzin’s Letters of a Russian Traveler (1791) marked a new stage in the history of Russian prose and even the subsequent generation of Romanticists paid close attention to the epistolary form. Only in the course of further development did the letter revert to what it is today—a fact of byt (ibid., p. 21-23).

This example, I believe, provides some basis for terming literary ustanovka a directive correlation. It locates the cause of literary change not in a teleological subject but in the dynamic interaction among systems. “It is clear,” Tynjanov wrote of the vicissitudes of literary history, “that what matters here are not individual psychic conditions, but objective ones, the evolution of the functions of the literary series in relation to the most immediate social series” (ibid., p. 45). It is also possible to argue that in some of its aspects ustanovka is not, properly speaking, a directive correlation. The interaction between literature and the “most immediate social series” that it describes is somewhat one-sided. Tynjanov was spiritually still too close to Formalism to be able to abandon a belief in the autonomy of literature. He saw literary development as determined mainly by the internal conditions of the literary system and regarded the extra-literary context as secondary, merely complementing the internal developmental causes by providing literature with speech constructions fitting the needs of the de-automatizing principle of construction.

The concept of ustanovka appears to be perched somewhere between the theoretical frame of Formalist poetics and post- Formalist tenets. By rescuing literature from the social vacuum into which it was placed by Formalism, ustanovka clearly points beyond this literary-theoretical school of thought, by not providing any avenue for the active involvement of social systems in literary development, however, ustanovka remains rooted in the Formalist postulate of autonomous literature.

That the concept of ustanovka was pointing toward the future is obvious from the nine-point thesis that Tynjanov wrote with Jakobson in 1928, quoted earlier. Only here was the literary system fully incorporated into the overall cultural system of systems and literary evolution conceived in terms of a directive correlation between those two systems. As the penultimate point of this interesting document states:

The discovery of the immanent laws of literature (language) permits us to characterize every concrete change in literary (linguistic) systems but does not permit us either to explain the tempo of evolution or to determine the actual selection among several theoretically possible evolutionary paths. This is because the immanent laws of literary (linguistic) evolution are indefinite equations which, while limiting the number of solutions, do not necessarily leave only a single one. Which pathway or at least which dominant is chosen can be determined only through an analysis of the correlation among the literary and other historical series. This correlation (the system of systems) has its own structural laws which should be studied (Jakobson & Tynjanov 1928, p. 37).

The Tynjanov-Jakobson theses occupy a crucial position in the history of Slavic poetics. The fruit of a collaboration between a leading Formalist, who had earlier lectured before the Prague Linguistic Circle, and the Circle’s vice-chairman, it represents a definite point of contact between Formalism and what later became known as Structuralism. The theses’ boldly charted design transcends the Formalist mode of inquiry, yet there was no opportunity for the Formalists to apply them to concrete literary material. They served as a springboard for the earliest Structuralist literary-historical studies, which aimed at demonstrating that literary development cannot be studied in isolation from the overall development of society (Cf. for example, Mukařovský 1934).

Tynjanov’s effort to eliminate the psychological subject from literary studies, to describe literary process in terms of objective, inter-systemic mutations, is to a large extent a child of its time. As I indicated earlier, all the Formalists, regardless of the theoretical model to which they subscribed, argued vehemently against psychologism and subjectivism in literary study. The systemic model in general followed this pattern. It is true that Tynjanov included the subject (especially the author) among his theoretical topics, but by “de-psychologizing” and “de-subjectivizing” him Tynjanov ended by fusing the subject with the literary system. The subject served the system in two capacities: first, as an unconscious generator of the varied principles of construction needed by the system for its constant rejuvenation; and second, as the system’s vehicle of literary sensibility signaling the automatization of the dominant principle of construction and thus triggering its replacement by a contrasting principle. In each case the subject is completely subordinate to the system. What matters in literary process are not a subject’s volitions, feelings, or actions but the internal conditions of the impersonal, self-regulating system.

Tynjanov’s concept of the author was influenced by Tomaševskij, probably the first among the Formalists to succeed in separating the authorial subject—for Tomaševskij a legitimate object of literary study—from the author as a concrete psychophysical being, whose locus is outside of literature. Tomaševskij treats the concept of the author from a dual perspective: the production and the reception of the literary text. The reader’s “struggle to comprehend the creative unity in a poet’s works naturally entails an interest in the writer as a kind of concrete unity. Thus, the reader is not satisfied with comprehending the abstract unity of poetic works. This unity must be embodied, named, recognized. The life of the poet is the frame which conveniently and simply fits his creation.” Such a conflation of Wahrheit and Dichtung, of an individual and a style, is, in Tomaševskij’s opinion, one source of the conceptual confusion in which “poetic individuality is comprehended as personal individuality” and the “key to artistic unity is sought in the unity of a personality” (Tomaševskij 1925b, pp. 56-57).

This confusion, however, can be introduced into the perception of the text deliberately by its author, who in one way or another establishes a link between his or her work and life. From the standpoint of literary production, Tomaševskij distinguished between two types of authors: those with a biography and those without one. It is the former who contribute to the confusion, for their texts acquire specific meanings and significance in connection with their author’s biography. Tomaševskij sees Voltaire as the first author with a biography in the history of modern literature. “Voltaire’s works were inseparably linked to his life. He was not only read; he was sought by pilgrims. The admirers of his oeuvre were also worshipers of his personality; the opponents of his works, his personal enemies. Voltaire’s personality unified his oeuvre. His works are not the first thing that comes to mind when his name is mentioned. Even today when most of his tragedies and poems are completely forgotten, the image of Voltaire is still alive and these forgotten works still shine by the reflected light of his unforgettable biography” (Tomaševskij 1923a, p. 6).

This distinction between the two types of authors does not mean that the literary critic should study an author as a concrete psychophysical individual. On the contrary, Tomaševskij’s article specifically argues against this approach. For him the term “biography” has two senses. In one sense, it is a documentary narrative produced by traditional literary criticism, a collection of facts from the poet’s private and public life. Tomaševskij claims that this kind of biography has very little to do with literary studies. “As far as ‘documentary’ biographies are concerned, they all fall into the sphere of the history of culture on a par with the biographies of generals and inventors. For literature and its history they are mere external, though necessary, sources of reference and auxiliary material” (ibid., p. 9).

What is integral to literary studies is what Tomaševskij calls the artistic, ‘‘legendary” biography. It differs from the “documentary” biography in that it is not generated by literary critics but by writers themselves conscious of the fact that “their lives will be the permanent screen against which their works will be projected.” Such reflexiveness affects authors in two ways. It “forces them, on the one hand, to stage the epic motifs in their lives, and on the other hand, to create for themselves an artistic biography—a legend with a well-calculated selection of real and fabricated events” (ibid., pp. 6-7). Despite the fact that in this type of biography it is impossible to demarcate with any precision where Dichtung ends and Wahrheit begins, or perhaps because of it, Tomaševskij finds the artistic biography proper to the sphere of literary studies. “‘This is because these biographic legends are literary interpretations of the poet’s life, interpretations which are essential as the perceptible background of a literary work, as a premise taken into account by the author when he created his works” (ibid., p. 8). In other words, the artistic biography presents us not with the author as a concrete psychophysical subject but with the authorial subject, that is, the author’s image refracted through the literary medium.

Furthermore, the artistic biography is a literary matter because both the existence of “authors with biographies” and the contents of these biographies are conditioned by the literary conventions of a period. In Russian literature of the mid-nineteenth century, for instance, after a proliferation of authors with biographies in the Romantic period, “authors without biographies” became the norm. As Tomaševskij put it, “the poet-hero was replaced by the poet-professional, the entrepreneur. The writer wrote, sent his manuscripts to the printer, and did not permit any views into his private life.” By the same logic, if there is a period of “authors with biographies” these biographies exhibit the characteristics demanded by the period’s literary conventions. Describing a collection of biographies of fashionable belletrists of the turn of the century, Tomaševskij wrote: “They all scream over one another that they have not studied anything because they were kicked out of high schools and technical schools, that they have nothing but a pair of torn pants and a couple of buttons, and that all this is because they do not give a bloody damn” (ibid.). Such examples illustrate how tenuously the artistic biography is linked with the “real” author. What emerges from it instead is the authorial subject—a figure whose birthplace and domicile are purely literary.

Tynjanov found Tomaševskij’s concept of biography congenial for a simple reason. It softened the rigid Formalist opposition of literature to author, while at the same time, by separating the literary from the nonliterary side of the author, it preserved the Formalist belief in the autonomy of literature. Despite the fact that several of Tomaševskij’s points turn up in Tynjanov, the latter approached the problem of the literary author from a different perspective, through the category of the proper name. It is the author’s proper name, Tynjanov believed, that is primarily responsible for the confusion between literature and byt. The name simultaneously denotes an individual tangential to the literary system and entities as essentially connected to this system as texts, literary schools, genres, or periods.

Tynjanov’s interest in the proper name was probably motivated by its importance in philosophical discussions of the late nineteenth and early twentieth centuries, outlined in the first chapter. In the argument between those who maintained that the proper name is a senseless mark and those who insisted that it is a shorthand description, Tynjanov sided with the descriptivist camp. This position fit well with his overall relational outlook, in which the identity of phenomena is a function of their context. The descriptivist view also conformed to the peculiar status of proper names in literature, which of course was Tynjanov’s central concern. As soon as we designate someone a literary author, we place that individual in a special context. He or she becomes a component of a literary process and the person’s identity becomes circumscribed through the texts, genres, or periods with which he or she is associated. For example, the names of Puškin, Dostoevskij, or Tolstoj denote only indirectly the real-life individuals Messrs. Puškin, Tolstoj, or Dostoevskij. Instead they refer to purely literary entities, particular stylistic features, sets of texts, and so on.

The peculiar status of the author’s name is especially palpable in the reception of literary works. Here the name serves as a kind of bridge where impulses coming from the text meet with extra-textual information. On the one hand, the name might arise from the text itself as the label of its stylistic individuality: “There are stylistic phenomena which lead to the person of the author… the particular vocabulary, syntax, and especially the intonational outline of the phrase; all of this more or less alludes to the ungraspable and yet at the same time concrete features of the narrator… the name is the last limit of this stylistic person’s literary concreteness (Tynjanov 1929, pp. 26-27).

The name, on the other hand, is also attached to the text from the outside, and through its connotations it introduces specific information and expectations into the reading of the text. The case of the pseudonym is especially telling. “Taken in its extra-literary aspect,” argued Tynjanov, “a pseudonym is a phenomenon of the same order as an anonym” (ibid., p. 28). Its purpose in literature, however, is completely different. “When a nineteenth-century writer signed an article ‘An Inhabitant of New Village’ instead of using his name, he obviously did not wish to convey to the reader that he lived in New Village, because the reader does not have to know this at all. But precisely as a result of this ‘purposelessness’ the name acquires different features—the reader selects from the concepts [in the pseudonym] only what is characteristic, only what in some way suggests a character for the author, and applies these to the features that arise from the style, the peculiarities of the narrative [skaz ] or the preexisting stock of similar names. Thus, New Village is for him the ‘frontier,’ and the author of the article a ‘recluse’” (ibid., p. 27).

According to Tynjanov, real proper names affect our reading of a literary text in the same manner as pseudonyms. The only difference is that the connotations of proper names are not derived directly from the words that make up the name but from the literary reputation of their bearers. Tynjanov uses Tomaševskij’s concept of biography to clarify the notion of literary reputation in terms of Tomaševskij’s “author’s artistic biography”—the blend of real events, hearsay, and outright fabrication that constitutes the image of an author. This image carries the same proper name as the psychophysical individual existing behind it, but this is merely a case of homonymity and should not suggest that the two entities are identical. Tynjanov differentiates between the “author’s individuality”—a set of personal characteristics irrelevant from a purely literary viewpoint—and “literary individuality”—a set of features representing the author in the reader’s mind. Though there is always some partial overlap between the two, students of literature should keep them distinct in their minds. The structure of “literary individuality” is ultimately a function of the literary system, whereas the author’s individuality is accidental from the standpoint of this system.

A careful differentiation between the author’s individuality and literary individuality is necessary in the study not only of the reception but of the production of literary works. As Tynjanov wrote, “it is very common today to substitute the problem of the ‘author’s individuality’ for the problem of ‘literary individuality.’ The problem of the psychological genesis of every phenomenon is thus substituted for the problem of evolution and literary change, with the suggestion that instead of literature we should study the ‘creator’s personality.’” Tynjanov points out the fallacy in this view using a parallel from social history. “To speak of the creator’s personal psychology and to see in it the source of the originality of a phenomenon and its significance for literary evolution is like claiming in an interpretation of the origin and significance of the Russian Revolution that it happened because of the personal idiosyncrasies of the leaders of the fighting parties” (ibid., pp. 12-13).

The regularity of literary production can be studied only within its actual context, which is provided, Tynjanov believed, by the state of the literary system. Within this context, an author’s individuality figures only as an accident. It is a conglomerate of haphazard activities in which some might become relevant for literature but only if required by the developmental needs of the system. All the author’s intentions, originality, and so on play no role in literary change. The new “principle of construction” always “arises on the basis of ‘fortuitous’ results and ‘fortuitous’ deviations, mistakes,” not because it was planned (ibid., p. 18). From the systemic point of view the authorial subject’s role in literary production can be studied only within the framework of “literary individuality.” This individuality, however, is a transformation of the “author’s individuality.” It is a configuration, a selection of certain of the subject’s actions which became enmeshed in the history of the literary system. Here too the same name stands for both “individualities” but again this is not a sufficient reason to conflate them.

In addition to the author, of course, another subject participates in the literary process—the reader. Although the systemic Formalists at least paid lip service to the authorial subject, the perceiving subject was virtually ignored. Tynjanov discussed the reader in two contexts. In his discussion of verse language he employed several basic categories pertaining to the reader’s consciousness, such as retention and protention, successivity and simultaneity, or mental attitude (Tynjanov 1924, pp. 28-45). Tynjanov’s goal was not the “phenomenology of reading” but the nature of poetic rhythm. Therefore, he did not treat these categories in a systemic fashion: they served him rather as heuristic devices to demarcate verse language from prose. Tynjanov also includes the reader in his studies of literary change, as an accessory to the literary system, or more precisely, as the very self-consciousness of this system that prompts it to seek a new principle of construction.

In harmony with the overall thrust of systemic Formalism in both of these instances, the reader is purged of all possible subjectivity and accidentality. Readers are first reduced to the inter-subjective basis of human consciousness. In the service of the system, moreover, they are as much present at the birth of a literary work as are the authors, and the readers’ acceptance or rejection of the work as literary is an externalization of the current state of the literary system. At the time the work is produced there seems to be no doubt as to its literariness—“every contemporary can point his finger at what is a literary fact” (Tynjanov 1929, p. 9). Yet at the moment the readers cease to be a part of the context from which the work arose, Tynjanov loses interest in them. Now chance prevails and the reading turns into a “misreading.” “It is not true,” Tynjanov argued, “that works cannot live ‘through the centuries.’ Automatized objects can be used. Each epoch focuses on certain phenomena of the past which are akin to it and forgets the others. But these cases are, of course, secondary phenomena, new work on old material. The historical Puškin differs from the Puškin of the Symbolists, but the latter is incompatible with the evolutionary significance of Puškin in Russian literature” (ibid., p. 12).

How incompatible the study of literary reception was with systemic Formalism can best be illustrated by Tynjanov’s 1929 article “On Parody.” There he attempts to rebut Žirmunskij’s charge that Tynjanov’s neo-Formalism aims at replacing “historical poetics with the history of criticism and readers’ tastes” (Žirmunskij 1928, p. 356). “It is utterly impossible to separate the author of literature from the reader because they are essentially the same. The writer is a reader too, and the reader carries on the writer’s job of constructing the literary work. This contrasting of reader and writer is furthermore incorrect because there are different readers and writers. The writer of one cultural and social system is closer to the reader of the same system than to the writer of a different system. The issue of ‘reader reception’ arises only if it is approached from a subjectivist-psychologistic standpoint, and not if it is studied systemo-functionally” (Tynjanov 1977, pp. 294-95). In other words, for a systemic Formalist, the only legitimate object of literary studies is the self-regulating literary system. The perceiving subject is either treated as an appendix of this impersonal system or ignored.


3 The Formalist S. Baluchatyj characterized his method as a “techno­logical literary discipline” (Baluchatyj 1927, p. 7). G. Vinokur des­cribed stylistics as “a kind of ‘linguistic technology’” (Vinokur 1929, p. 9). B. Èjchenbaum summed up the early phase of Formalism as fol­lows: “In recent years, students of liter­ature and critics have paid atten­tion above all to questions of literary ‘techno­logy’” (Èjchenbaum 1929, p. 50).

4 My translation of byt as “everyday life” is a rather inadequate ren­dition of a highly evocative Russian term. According to Roman Jakob­son, byt is “the tendency toward stabilizing the immutable present and the gradual accretion of the stagnant slime to it, the stifling of life by tight and petrified molds,” the antithesis of “the creative impulse toward the transformed future... It is cu­rious,” Jakobson continues, “that while in the Russian language and literature this word and its derivatives play quite a significant role... European languages lack any corresponding nomenclature” (Jakobson 1931, p. 13). For this reason, I have retained byt in all quotations from Formalist texts. In my own prose I alternate byt with “life.” If, however, the word “life” appears in quotation marks it is a translation of the Russian žizn‘. The adjective bytovoj is rendered as “extra-artistic” or “extra-literary” depending on the context.

5 Šklovskij was far from consistent in his arguments, and though his position in general was that form determines material, sometimes he was willing to argue precisely the opposite. It is interesting for this study that his concessions to material were also couched in a simile from the realm of technology: “If a mechanic wished to substitute a steel part of a machine for a bronze or an aluminium one, this new part cannot be a copy of the old one. A new material requires a new form” (Šklovskij 1923b, p. 18).

6 Propp wrote, “Professor Lévi-Strauss knows my book only in the En­glish translation. But its translator allowed himself an impermissible liberty. Not understanding the function of the epigraphs which at first glance do not seem to be explicitly connected with the text, he con­sidered them useless ornaments and barbarously omitted them… all these epigraphs… had the purpose of ex­pressing what was left unsaid in the text of my book...” (Propp 1976, p. 135). Lévi-Strauss’s “L’Analyse morphologique des contes russes,” (1960) reviews the first English edition of Propp’s Morphology of the Folktale(1958).

7 The epigraphs of other chapters of Propp’s book are from the following writings of Goethe: “Introduction” (ibid., sec. 2, vol. 6, pp. 298-9); “1stchapter (ibid., vol. 8, pp. 221-22): “2ndchapter” (ibid., sec. 1, vol. 35, p. 16); “9thchapter” (ibid., sec. 4, vol. 8, pp. 232-33). It is noteworthy that Skaftymov too uses a quotation from Goethe as the epi­graph for his essay on Dostoevskij’s Idiot (see Skaftymov 1924b, p. 135). However, in contrast to Petrovskij and Propp he does not quote from Goethe’s scientific works or diaries but from Faust and the two lines used, “Willst du dich am Ganzen erquicken,/ So musst du das Ganze im Kleinsten erblicken” (“If you want to enjoy the whole / You must learn to see the whole in the smallest part”), do not pertain to transformation but to the relationship of the parts and wholes.

8 “Le Retour” is published under the title “A French Enoch Arden” (Mau­passant 1903, p. 137).

9 Propp explained, “For the sake of brevity and a lively presentation we were forced to omit many things that a specialist would like to keep. In addition to those parts appearing below, the original draft of the work contained a study of the rich sphere of the acting character’s attributes… it deals in detail with the questions of metamorphosis, i.e., of the transformations of the tale” (Propp 1928a, pp. 6-7).

10 Unfortunately, the adjective “systemic” that I use for this Formalist model carries certain biological connotations (relating to the body as a system) that I do not intend. Its only possible replacement, “syste­matic,” is even less felici­tous, however, because of its primary meaning of “methodical” or “thorough.” I have chosen “systemic” therefore, in its sense of “relating to a system,” and hope that the reader will not be distracted by the specifically medical or biolo­gical usage of the term.

11 Koffka’s work seems to have been well known in the teens in Russia: the Formalists certainly were aware of it. It was the topic of Professor Georgij Čelpanov’s seminar held at the Moscow University in 1915/1916. Roman Jakob­son participated in this seminar (See Holenstein 1975, p. 62).

12 The Formalists made a few references to Cassirer’s book. Èjchen­baum cites it in his diary in January 1919 as one of the books to be consulted on issues of methodology (see M. O. Čudakova’s com­mentary in Tynjanov 1977, p. 455). A passage from Cassirer’s work is quoted by Sergej Karcevskij (1927, pp. 13-14). This passage is sub­sequently quoted by V. Vinogradov in his critique of Tynjanov’s method (1930, p. 59).

13 More recently, Frederic Jameson has asserted that “Tynjanov retains Saus­sure’s basic model of change, in which the essential mechanisms at work are the ulti­mate abstractions of Identity and Difference” (Jameson 1972, p. 96).

14 The three-level scheme that I outline here simplifies Tynjanov’s actual thought somewhat. The middle, infra-literary level in particular comprises several subsystems—genres, literary schools and styles. Tynjanov did not pro­vide any clear-cut picture of this level of system, however.

15 Apparently Tynjanov was not very happy about this term. He complained to Grigorij Vinokur in a letter of November 7, 1924: “My term ‘deformation’ is infelicitous; it should have been ‘transformation’—then everything would be in its place.” (Tynjanov 1977, p. 517).

16 That is, it is a mistake from the point of view of the system, not from that of the creating subject. As I argue later in this chapter, the systemic Formalists considered the author’s intentions irrelevant to literary change and claimed that it is an author’s unconscious slips rather than conscious efforts that give birth to a new principle of construction.

17 The concept of the “auto-function” discussed earlier, as a language link between literature and extra-literary phenomena, thus can be seen as one aspect of the overall “speech function” of literature.

This document is unfortunately not available for download at the moment.

Not implemented yet !
First Edition
(1984) "The three metaphors", in: Steiner Peter, Russian formalism, Ithaca, Cornell University Press, pp.44-137.