Verse structure and literary tradition: the interaction between rhyme and stress in the Onegin Stanza.
Rhythm and Rhyme
Across the world's folklore and literary traditions, the principles of verse organization take a bewildering variety of forms, ranging from parallelism and alliteration to regulation of syllabic length and complex stanzaic structures. (1) This formal diversity notwithstanding, it is typologically common across cultures to draw the distinction between maximally organized and less-organized varieties of marked ("literary") language, which roughly maps on the opposition between "verse" and "prose" in Western literatures. This distinction appears to be fundamental to the artistic use of human language, even though it can be complicated by ties of common descent, such as those uniting Slavic folk systems of versification, or historical influences that connect national literary cultures, as in the case of the borrowing of Classical Greek meters by the Romans. (2) Cross-cultural ubiquity of verse might lead us to downplay the evidence of literary history, foregrounding instead universal aspects of cognition; it also poses the problem of fit between languages and metrical systems, that is, of the extent to which the prosody of a particular language favors or rules out particular forms of verse organization. (3) Indeed, over the past decades, theory of verse has largely become a province of phonologists, while linguistics at large has moved closer to cognitive science and away from the study of literary forms and their evolution.
The very nature of the data, nevertheless, demands that the most challenging problems in verse theory be confronted with the combined ammunition of literary-historical and structural-linguistic methodologies. In this article, we address one such problem--the interaction between rhythm (stress patterns internal to the poetic line) and rhyme (repetition of line endings)--on the basis of an in-depth statistical study of the Onegin stanza, a poetic form that developed within Russian syllabo-accentual system of versification in the nineteenth and early twentieth centuries. In particular, we explore how different principles of metrical organization interact (a) within a particular literary work, (b) in a stanzaic form that evolves as part of a literary tradition, and (c) in a literary system viewed as a whole. Our approach to the use of scientifically verifiable methods in the humanities is in line with the Russian tradition of poetics, which has been open to statistical methods since the 1920s. (4)
In spite of substantial advances in the study of stress-based meters in traditions such as English and Russian and a well-established consensus on the typology of rhyme across European traditions, very little is known about how rhythm and rhyme are related. (5) As we endeavor to show, etymology, in this case, is less deceptive than it might appear. Both rhyme and rhythm are related to Latin rhythmus, itself a borrowing from Greek (Brogan 1053). As Emile Benveniste shows, the Greek term [TEXT NOT REPRODUCIBLE IN ASCII] was the principal means of conceptualizing form in Archaic lyric poetry and pre-Socratic philosophy and, only beginning with Plato, assumed the meaning "form of movement," particularly of recurrent movement. (6) Sprouting from this fundamental concept of form, the terms rhyme and rhythm, across European languages, came to denote aspects of formal organization that were associated with new vernacular systems of versification, as opposed to Classical prosody (Gasparov, A History 97-98). (In Greek and Roman literary cultures, stress patterns and parallel line endings were not an essential part of verse; the latter was, however, relevant for rhetorical prose.)
The fine-grained analysis of the inner dynamic of the Onegin stanza promises more than an insight into the interaction between basic constituent formal features of European postclassical verse. It offers a glimpse into the workings of what Mikhail Bakhtin describes as "genre memory," referring to the life of literary forms as symbolic structures that persist over centuries and millennia, accumulating meanings of which the author employing these forms is not fully aware (109-37). As we show in this study, later poets who took over the complicated machinery of Pushkin's Onegin stanza in a changing historical and poetic milieu, did not reproduce it exactly; their often unwitting adaptations reflect the immanent imperatives of verse structure and the general cultural-historical tendencies of the poetic cultures of late Romanticism and Modernism. The analysis of the evolution of the Onegin stanza provides, as it were, a microscopic view of a literary form undergoing processes of dissolution and regeneration. In this way, metrical analysis can contribute to the ongoing conversations in the field of Historical Poetics on persistence and innovation of forms, the limits of authorial control, genre as sedimented historical content, and the nature of literary tradition in the modern age (Kliger, "Dostoevsky and the Novel-Tragedy"; Kliger, "Resurgent Forms"; Kunichika; Kliger and Maslov).
Rhyme and Stanza
In a conversation with William Wordsworth, the ageing Friedrich Gottlieb Klopstock inquired "whether the rhymes of Pope were not more exact [than Dryden's]." The record of that exchange, as preserved by Wordsworth, continues: "This question I understood as applying to the final terminations; and I observed to him that I believed it was the case, but that I thought it was easy to excuse some inaccuracy in the final sounds if the general sweep of the verse was superior. I told him that we were not so exact with regard to the final endings of lines as the French. He did not seem to know that we made no distinction between masculine & feminine rimes; at least he put inquiries to me upon this subject." (7)
Russian versification, by and large, inherits the French attitude toward rhyme, with its more stringent expectations and a regular alternation between "masculine" (stress on the ultimate syllable) and "feminine" (stress on the penultimate syllable) line endings. (8) Moreover, in Russian poetry, rhyme has proven remarkably resilient as a major principle of verse organization. It is retained even by those twentieth-century poets who, like Vladimir Mayakovsky and Joseph Brodsky, experimented with rhythm in radical ways, to the point of relinquishing syllabo-accentual meters. While the evolution of Russian rhyme is relatively well-researched, little is known about the role of rhyme in the structure of verse, and about its relation to other principles of formal organization.
Rhyming is inextricably linked with stanzaic structure. In general, when there is a rhyme, we also find a stanza, ranging from a couplet, or pair of lines (rimes plates in the alexandrine), to longer and highly unorthodox sequences, which can be as long as twenty-nine lines, as in one of Voltaire's odes "To de Richelieu" (AbAbAAbbCCbC ddCeCeCeCeeCfCCfC). (9) The most familiar among the longer kinds of stanza are composed of fourteen lines; these include the different varieties of the sonnet and the Onegin stanza. Both the Shakespearean sonnet and Pushkin's Onegin stanza include a concluding rhyming couplet. This superficial resemblance, however, is outweighed by the systematic alternation, in the Onegin stanza, between two kinds of ending--"feminine" and "masculine"--generating a distinctive stanzaic architecture.
The structure of the Onegin stanza (capital letters mark feminine ending):
AbAbCCddEffEgg James Falen's translation of Pushkin's Eugene Onegin, the poem for which the stanza was originally devised, preserves the alternation of clausulae: But when the age of restless turnings A Became in time our young man's fate, b The age of hopes and tender yearnings, A Monsieur l'Abbe was shown the gate. b And here's Onegin--liberated, C To fad and fashion newly mated: C A London dandy, hair all curled, d At last he's ready for the world! d In French he could and did acutely E Express himself and even write; f In dancing too his step was light, f And bows he'd mastered absolutely. E Who'd ask for more? The world could tell g That he had wit and charm as well g (Falen 6; ch. 1, 4).
Another legacy of French verse in Russian is the use of "rich" rhymes (rimes riches). A rhyme is defined as rich if the same consonant precedes the rhyming vowels (e.g., strain/rain); on this definition, the category of rich rhyme encompasses "identical" (or tautological) rhymes in which the rhyming syllables have the same onset (write/right). (10) In the history of Russian verse, as shown by Gasparov, the preference for rich rhymes is very strong in the eighteenth century due to French influence; it declines in the nineteenth century, but increases again in modernist poetry. Gasparov's data is reproduced table 1. (11)
The poetic practice of Alexander Pushkin (1799-1837), while in general falling at a low point in the history of rich rhymes in Russian, displays greater attention to the onset syllable than that of his contemporaries. Whereas Gasparov's coefficient for Eugene Onegin (1823-1830) is 17 and Pushkin's later lyric (1824-1836) scores at 16, Iazykov and Baratynskii have coefficients at around 10. Indeed, as we show below, in Pushkin's crafting of the Onegin stanza, rich rhymes served as an essential structural element.
This study discusses three long poems that employ the Onegin stanza--Pushkins Eugene Onegin, mostly written in 1823-1830; Mikhail Lermontov's Tambov Treasurer's Wife, written in 1838, and Viacheslav Ivanov's Infancy, written in 1913-1918. (12) The poems represent three moments in the evolution of the form: crystallization in an originating work, destabilization by a younger contemporary, and renovation by a much later poet. Lermontov's and Ivanov's attitude toward Pushkin's legacy, which reflect their different positions in the history of literature, are in turn reflected in the subtle variation of the Onegin stanza. In Lermontov, these slight modifications betoken a continuous tradition that is being unsettled, whereas in Ivanov's case, they index a resurgent older form that is, in some ways, radically refashioned. While this variation can be recovered by the use of statistical methods, the poets themselves were only vaguely, if at all, aware of it Yet the value of the testimony that such evidence gives on the evolution of literary form is not thereby diminished; on the contrary, one might say that in this case poetic structures speak to us about historical change directly, without the added layers of the poets' intention and design.
Each line of the three poems was annotated for a variety of factors, such as the type of rhyme, the position of missing stress (if any), the type of the onset consonant, and the length of the rhyming word. In what follows, we only discuss factors related to rhyme and stress. (13)
The article is organized as follows. In the sections "The Onegin Stanza: The Quality of Rhyme" and "The Onegin Stanza: Deviations from the Metrical Template," we discuss the general properties of rhyme and meter in the Onegin stanza. In the section "Three Forms of Interaction Between Rhythm and Rhyme," we turn to various aspects of their interaction. Our conclusions are summarized in the section "Rhyme-Meter Isomorphism and the History of Poetic Forms," where we also present general observations on the evolution of poetic forms within a literary tradition.
THE ONEGIN STANZA: THE QUALITY OF RHYME
The Onegin stanza contains seven rhyming line pairs (A, b, C, d, f, E, g), which can be shown to differ in their preference for rich vs. poor rhymes. Table 2 presents the proportions of rich rhymes in these line pairs, separately for the three poems.
The major factor in the distribution of rich and poor rhymes is the type of ending: masculine (monosyllabic) rhymes are considerably more often rich than feminine (disyllabic) rhymes. The effect is very robust, and constant across the authors. Figure 1 juxtaposes the patterns from the three poems: all three show the same general outline. Since Pushkins Eugene Onegin is considerably longer than the two other poems, the contrast between masculine and feminine rhymes is more noticeable.
The correlation between quality of rhyme and type of ending is compatible with two explanations, which are not mutually exclusive, and are probably related. On the one hand, feminine endings involve more phonological material after the rhyming vowel, and hence allow for fewer rhyming options to choose from. On the other hand, the additional identical syllable following the stressed vowel compensates for a difference in the stressed syllable's onset consonant, so that even poor rhymes, in the case of feminine endings, may be perceived as strong enough. (15)
Table 3 lists the proportions of rich and poor rhymes divided by the type of ending. With minor variation, the general structure is stable across the three authors, and the difference is highly significant (p-value < 0.0001 (16)). In Eugene Onegin, 40 percent of masculine rhymes, and only 13 percent of feminine ones, are rich (a 27 percent difference). Lermontov's poem has the lowest proportion of rich rhymes, and it has the weakest masculine-feminine asymmetry: the difference in proportion is only 9 percent. Ivanov's Infancy, on the other hand, has the highest overall proportion of rich rhymes, and its masculine-feminine asymmetry is even stronger than in Eugene Onegin; we observe a 29 percent difference of rich rhymes between masculine and feminine lines.
These results suggest that the two later texts subject Pushkin's Onegin stanza to different kinds of transformation. Lermontov operates with a considerably looser formal structure, marked by the lowest proportion of rich rhymes, in accordance with the gradual decline in the use of rich rhymes in late Romantic Russian poetry. For Ivanov, by contrast, the Onegin stanza epitomizes the structural elegance of the "golden age" of Russian poetry, the period dominated by what Lydia Ginzburg describes as the poetic "school of harmonic precision" (shkola garmonicheskoi tochnosti) (13-43). Accordingly, Ivanov recreates Pushkin's model "hyper-correctly, making it more rigid than it originally was.
The difference between Lermontov's and Ivanov's adaptations of the Onegin stanza is also reflected at other levels of literary form. Lermontov repeatedly fails to complete the stanza, filling in only some lines: ten out of the total of fifty-four stanzas in Tambov Treasurer's Wife are incomplete, while Pushkin, in his 366-stanza poem, takes advantage of that license four times. Ivanov's Infancy, by contrast, only includes perfectly formed stanzas. Thematically, Lermontov's poem, in which an unwilling heroine foregoes her marital pledges at the instigation of her own miserly husband, radicalizes the satiric element of Pushkin's poem. In contrast to Lermontov's playful reversal of Eugene Onegin's plot, Ivanov casts an idealized figure of his mother as a new Tatiana who, in spite of the perturbations of the materialist age, preserves the pristine religiosity of a "Russian maiden".
Furthermore, it is significant that Ivanov borrows the Pushkinian stanza, uniquely in his oeuvre, for the autobiographical narrative of the first years of his life. Due to its literary-historical associations with the Golden age of Russian poetry, this form contains a promise of the original, Edemic harmony that Ivanov's poem seeks to evoke. Indeed, in the spirit of Symbolist mythopoesis (mifotvorchestvo), Ivanov, in Infancy, does not shun hagiographical motifs in creating a myth of the Poet who will recapture the lost powers of art. As this aim is formulated in one of his essays, "[t]he crisis in the art of the word is the widowhood of the poet who has lost the Psyche who was once alive in the word, who has lost the inner form of the word; it is a yearning for its renewal, vivification, transfiguration." (17) It is this project of the radical renewal of poetic language--indeed of the word as such--that informs Ivanov's self-consciously faithful, "hyper-correct" restoration of the Onegin stanza.
THE ONEGIN STANZA: DEVIATIONS FROM THE METRICAL TEMPLATE
Pushkin's Eugene Onegin, as also other poems employing the Onegin stanza, has a consistent metrical structure: the line consists of four iambic feet, with masculine or feminine endings depending on the position of the line in the stanza (hence each line is either eight or nine syllables long). Less than 40 percent of all lines, however, realize the abstract metrical template of iambic tetrameter faithfully in having four stresses, and deviations from the metrical template are unevenly distributed across the lines within the stanza. (18) The numbers in table 4 demonstrate that the proportion varies widely depending on the lines position: the first line has the highest proportion of folly stressed lines, and line 7--which is located roughly in the middle of the stanza--has the highest proportion of missing stress. The same data is represented in figure 2.
The proportion of folly stressed lines in figure 2 reveals the underlying structure of the stanza, which consists of three quatrains and a couplet (AbAb+CCdd+EffE+gg). As Barry Scherr notes, the peaks in the proportion of folly stressed lines mark the beginning of independent structural units. (20) The first quatrain is marked by a gradual change in the proportion of folly stressed lines. The second quatrain begins with an increase and shows a surprising sharp drop in folly stressed patterns at line 7. The third quatrain, again, starts with a peak. The same principle is at work in the concluding couplet, where we see an increased number of folly stressed lines at line 13.
A similar result on the inner divisions in the Onegin stanza was obtained by Tomashevskii and later corroborated by Postoutenko based on evidence from punctuation: assuming that punctuation marks tend to reflect major divisions in syntactic structure, the Onegin stanza, in Pushkin, can be shown to consist of four units--three quatrains and a couplet. Moreover, as in the case of stress, the syntactic units within the stanza become increasingly less autonomous (Tomashevskii 367-70; Postoutenko 151).
There is one remarkable feature of Pushkin's Onegin stanza that this hypothetical structure fails to capture: the rhythmic weakness of line 7. The high proportion of missing stress points to the poet's effort to contravene the reader's desire to break the long stanza into two parts (6+8 lines) by treating line 6, which forms a couplet with line 5, as the end of a unit. By under-stressing line 7, the poet avoids the impression that it may constitute a new beginning, since opening lines (cf. line 1) tend to be heavily stressed in Pushkin. This interpretation is buttressed by Postoutenko's data, according to which the second quatrain with parallel rhymes--CCdd--does not show the expected preference for a syntactic break between the pairs of rhyming lines. We are more likely to find a period after line 7 (36 percent) than after line 6 (31.1 percent). (21)
Notably, while the underlying quatrain structure of the stanza is clearly evident in Eugene Onegin, it is no longer detectable in the later texts. Figure 3 shows the distribution of folly stressed lines in Lermontov's Tambov Treasurer's Wife and in Ivanov's Infancy. The numbers, although quite small, reveal tendencies that are markedly different. In Lermontov, only the couplet is rhythmically distinct, whereas the quatrains are no longer detectable; on the other hand, Lermontov does maintain a sharp drop in fully stressed lines at line 7, in the middle of the stanza. Ivanov relinquishes all these compositional tendencies, instead slightly privileging the first members of rhyming pairs of lines: the first members of rhyming pairs are consistently more likely to be fully stressed than the second ones.
These differences point to the progressive obfuscation of the underlying structure of the stanza. In Lermontov, the structure is still detectable, but the boundaries between the underlying units are obscured by a low proportion of fully stressed lines. Remarkably, in Ivanov, who is otherwise following Pushkin's model quite closely, the underlying structure of the stanza no longer informs the rhythm of particular lines. For Ivanov, the Onegin stanza is already established as an unanalyzable unit in its own right, showing no effects of the original divisions. Similar to other Russian poets who turned to the Onegin stanza after Lermontov, Ivanov is no longer concerned with the danger of the stanza breaking into two halves (hence the lack of the drop in fully stressed lines at line 7). (22)
In fact, the gradual conventionalization of the stanzaic structure can be observed even within Pushkin's Eugene Onegin itself. If we divide the text of the poem into two portions, by chapter, it turns out that the underlying structure is obvious in the first portion, comprising Chapters 1-4, but becomes obscured in the later one (Chapters 5-8). The numbers are given in table 5,' figures 4 and 5 show the difference between the clearly structured stanza of the 41-4 shape (for the early chapters) and the obscured structure that aims, first and foremost, at avoiding the perception of line 7 as a new beginning (for the later chapters).
The observation that in Eugene Onegin Pushkin is handling a progressively more stable prosodic medium is in part consistent with the findings by Dean Worth, who studied the use of "grammatical" rhymes (where the identity of clausulae is morphological in nature) in Eugene Onegin. In particular, Worth shows that the earlier chapters of Eugene Onegin are more experimental with respect to rhyme than the later ones (with the notable exception of Chapter 8): "during the first half of Onegin, Pushkin gradually uses more and more innovative rhymes built on contrasting parts of speech, whereas the opposite is the case during the novel's second half, as if the poet gradually wearied of innovation and moved back toward the more traditional rhyme pairs containing grammatically identical classes, only to change again toward innovation in Chapter 8" ("Grammatical Rhyme Types" 46).
As Pushkin went on composing in the newly invented stanza, its inner structure, originally revealed in the distribution of fully stressed lines, became increasingly less apparent. In a sense, we observe here a process of crystallization of a new poetic form, which obeys its own principles such as the drop in stressing in the middle of the stanza. In this one aspect, Lermontov, in fact, simply continues a tendency that is already present in Pushkin; it is the immanent dynamic of a poetic form, rather than the poetics of a particular author, that is at issue. In other aspects, Lermontov moves toward an altogether un-Pushkinian stanzaic arrangement, marked by leaps in stressing between lines and, in the later segments of the stanza (the last "quatrain" and the concluding couplet), by progressive intensification of rhythmic structure. (23) Pushkins harmonic precision" is replaced by sudden shifts in poetic intonation typical of Lermontov's lyric. Ivanov's poetics, by contrast, accords with Yuri Tynianov's literary-historical principle of "the struggle with the ++fathers, in which the grandson turns out to be similar to the grandfather." This similarity, however, as Tynianov saw, is illusory, for "we should not forget that Pushkin was never a Pushkin scholar" (Pushkin nikogda ne byl pushkinistom) (182).
THREE FORMS OF INTERACTION BETWEEN RHYTHM AND RHYME
That the interaction between meter and rhyme has remained largely unstudied is not surprising: the two elements of formal organization of verse show fundamentally different properties. While manifestations of meter (rhythm) can be assessed with reference to individual lines, rhymes cannot be defined in absolute terms, as they depend on the relationship between two or more lines. On the other hand, as both meter and rhyme impose formal constraints on the composition of verse, one might expect them to show a compensatory relation: stricter adherence to the metrical template could be expected to counterpoise a looser rhyme, and vice versa. In what follows, we argue that while the interaction between rhythm and rhyme is indeed generally informed by the logic of compensation, some rather startling complications reveal that they can also work in tandem, perhaps pointing to a shared basis in what Roman Jakobson called "the poetic function" of language.
First vs. Second Rhyming Line
One factor that has an effect on the stress pattern is the line's position within the rhyming pair. The Onegin stanza consists of seven pairs of rhyming lines. The members of each pair differ in the degree of their integration into the stanza: the rhyming relation is realized on the second member of the pair, imposing additional constraints on the second line's ending, while the first member is only expected to adhere to the metrical template. In this sense, the two positions in the rhyming pair are inherently asymmetric. If meter and rhyme have a similar function, we should expect the asymmetry in the rhyming relationship to produce an asymmetry in the metrical structure. This is indeed what our data demonstrates.
In table 6, we present the proportions of fully stressed lines for the two positions in the rhyming pairs in Eugene Onegin. The first lines in a rhyming pair, which have no rhyming expectations to fulfill, are more likely to be fully stressed than the second ones (28 percent of fully stressed lines against only 25 percent). Even though the difference in proportions is not immediately striking, it is highly significant statistically (p-value < 0.003). The correlation between stress pattern and rhyme suggests that deviations from the metrical template are "licensed" by a strong rhyming relation.
The same correlation appears to hold in Ivanov's Infancy, in spite of the small numbers that make it more difficult to observe (p-value < 0.05, implying marginal significance). The metrical difference between the first and the second member in a rhyming pair is in fact more pronounced in Ivanov than in Pushkin: the 8 percent difference in the proportion of fully stressed lines between first and second positions corresponds to a 3 percent difference in Pushkin. (24) As in the case of richness of rhyme, Ivanov's version of the Onegin stanza is more rigid than Pushkin's original.
As we saw in the case of quality of rhyme, the structural organization of Lermontov's Onegin stanza is relatively loose. Predictably, Lermontov also differs from Pushkin and Ivanov in not displaying a correlation between conformance to the metrical template and position in the rhyming pair. The difference between the first and the second members of rhyming pairs, in fact, seems to point in a direction opposite of the expected pattern: first members of a rhyming pair are less likely to be fully stressed than its second members (this difference is not statistically significant, and is likely due to chance). In any case, the failure of Lermontov's verse to adhere to the logic of mutual compensation between meter and rhyme is noteworthy, suggesting a poetics that works against the givens of literary tradition.
Figure 6 sums up the distributions, illustrating the difference between first and second rhyming lines, and the difference between the three authors.
The overarching pattern observed in Pushkin and Ivanov suggests that greater rhythmic license is granted to the second line in a rhyming pair. This difference is correlated with a more general feature of stanzaic composition: the distribution of rhythmically strong lines in quatrains (particularly of the "abab" type), which tend to have fewer lexical stresses in the last line. This phenomenon was initially described with reference to Russian poetry by Georgii Shengeli (111-21), then further explored by Mikhail Gasparov ("Stroficheskii ritm"). This result was corroborated by Marina Tarlinskaja based on the analysis of iambic and dolnik stanzas in English and German. (25) More work is needed to establish the relative significance of the compensatory logic of lightening the second rhyming line and other tendencies in stanzaic composition, such as the short-last effect. (26)
Rich vs. Poor Rhymes
The hypothesis of functional relatedness of meter and rhyme makes another prediction: we would expect the quality of rhyme to be correlated with the line's degree of faithfulness to the metrical template. In particular, we would expect rich rhymes to compensate for deviations from the metrical template, and vice versa. To test for this effect, we looked at the second members of rhyming pairs and divided them into two groups according to the rhyme's quality (poor vs. rich). As predicted by our hypothesis, the proportion of fully stressed lines is higher in the case of poor rhyme than in the case of rich rhyme. In Pushkin's Eugene Onegin, the difference is highly significant (a 7 percent difference, p-value < 0.001). The same difference is present--and even more pronounced--in Ivanov's Infancy (a 10 percent difference, p-value < 0.05). In Lermontov, it is the least noticeable (a 6 percent difference), and the distribution is not statistically significant, that is, it could be due to chance.
Notably, the correlation is more pronounced in lines with masculine endings than in lines with feminine ones, where the same tendency is not strong enough to be statistically significant (cf. the counts in tables 10 and 11; data from Lermontov is not included in these counts, since it does not display a significant correlation between stress pattern and type of rhyme).
This difference between masculine and feminine rhymes can be attributed to the fact that rich feminine rhymes are relatively scarce, and for that reason lines that include them, from the point of view of production of verse, allow for fewer choices between stress profiles than lines ending with rich masculine rhymes. This tentative explanation needs further testing.
The Echo-Effect Between Rhyme and Rhythm
The data discussed in the preceding section confirms that a compensatory relation exists between quality of rhyme and faithfulness to the metrical template. We now turn to our main finding, which pertains to a different dimension of interaction between rhyme and stress: the relationship between the rhythm of lines that are related by rhyme. We use the term "stress configuration to refer to particular patterns of realized and non-realized metrical stresses within a line of iambic tetrameter. The term makes it possible to specify the rhythm of the line, as opposed to its underlying meter: a line that omits a stress in the second foot (e.g., "Theocritus and Homer bored him") has a different stress configuration as compared to a line that omits a stress in its third foot (e.g., "Of free and easy conversation") or a line with two omitted stresses in the second and third position (e.g., "Cite Juvenal in conversation"); all these lines nevertheless conform to the metrical template of the iambic tetrameter.
As already mentioned, stress configurations and rhyme cannot be directly compared, as they characterize different types of entity: rhythm is a property of individual lines, while rhymes define a relationship between two or more lines. To address the interaction between rhyme and rhythm, we explored relational--rather than absolute--properties of stress configurations. Specifically, we looked at pairs of rhyming lines and compared their stress configurations. The corpus was annotated not simply in terms of whether or not the line is fully stressed; we also marked whether or not the two lines from a rhyming pair show omission of stress in identical positions. (For example, two lines with an unstressed third foot were classified as identical in their stress configuration; but if one line lacked stress in the third foot, and the other lacked stress in the first and third feet, the two were classified as having different stress configurations.)
The counts in table 12 show a correlation between the type of rhyme and the (non)identity of stress configuration for Eugene Onegin. Rather surprisingly, lines related by rich rhymes are more likely to coincide in their stress configuration than lines related by poor rhymes (the effect is statistically significant, p-value < 0.002). The difference suggests that rhyming is not restricted to phonological identity of the line's endings; it is also reflected in the identity of the lines' rhythm.
The correlation between quality of rhyme and (non)identity of stress configuration is relatively weak, and it is only statistically significant in Pushkin. The other two poems show the same difference in proportions, but due to more limited data, the difference does not pass the statistical significance test (cf. table 13). While suggestive, the evidence of Lermontov and Ivanov should therefore be treated with caution.
A potential caveat is related to the possibility of dealing with highly correlated factors. There are indications that adjacent lines sometimes show parallel structure and may therefore be more likely to have "grammatically homogenous" rich rhymes. (27) One could hypothesize that the echo-effect in stress configuration could be due to syntactic parallelism between adjacent lines, since in the Onegin stanza four pairs of rhyming lines are indeed adjacent.
To explore this effect further, we isolated nonadjacent rhyming pairs: these are lines 3,4, and 12, which rhyme with lines 1, 2, and 9, respectively. If the echo-effect is due to adjacency, we should expect it to disappear or at least become weaker when the sample is restricted to nonadjacent pairs of lines. Quite surprisingly, the effect not only persists, but also becomes stronger when adjacent pairs of lines are excluded: the difference in the proportion of rich rhymes in table 14--for nonadjacent rhyming lines--is more pronounced than the nonsignificant difference in table 15, for adjacent ones. The distribution suggests that stress configurations tend to recur independent of syntactic parallelism; the "rhyming" effect in table 14 is significant with p-value < 0.004.
The additional counts in table 16 demonstrate that the quality of rhyme also shows no significant effect on the identity of stress configurations between adjacent lines in cases where the adjacent line does not rhyme. In this case, looking at the configuration of the preceding lines, we observe that rich rhymes do not produce an effect that could be compared to the one we observed with non-preceding, but rhyming lines.
The effect of echoing, or "rhyming," stress configurations is independent of syntactic parallelism or linear adjacency. We discuss the significance of this finding in the next section, relating it to the rest of our findings.
RHYME-METER ISOMORPHISM AND THE HISTORY OF POETIC FORMS
Our results reveal important functional similarities between meter and rhyme, suggesting that these two principles of organization of verse are intimately related, at two levels. At the level of absolute value, we looked at deviations from the abstract metrical template and discovered a compensatory relation between rhyme and rhythm. On the one hand, second members of rhyming pairs are more likely to deviate from the metrical template than their first members, due to being more strongly integrated into the stanzaic structure through rhyming. On the other hand, lines with rich rhyme are more likely to deviate from the metrical template than lines with poor rhyme, for the same reason. This result bears out Wordsworth's intuitive description of rhyme as just one aspect of structural organization of verse, no more important than others: "... it was easy to excuse some inaccuracy in the final sounds if the general sweep of the verse was superior" (1.94).
At the level of relative value, we showed that rhyme and rhythm can be isomorphic: lines related by rich rhyme are more likely to have identical stress configurations than lines related by poor rhymes. In Pushkin's Eugene Onegin, rhyme and rhythm both reflect the strength of formal association between the lines. This finding is in line with research on isomorphism between different levels of formal organization of verse initiated by Roman Jakobsons discussion of the poetic function. One of Jakobson's examples was Russian nineteenth century iambic tetrameter, which displays isomorphism between the structure of the syllable (alternation of nuclei and margins), metrical template (alternation of strong and weak positions), and the actual rhythmic realization (a stress configuration in which the second and the fourth positions are more likely to be stressed than the first and the third; cf. "Upon the crossroads of the sea") ("Linguistics and Poetics" 362-63).
Continuing this line of research, Mikhail Gasparov analyzed the "heaviness" of rhythm (the number of actual lexical stresses per line, ranging from 2 to 4) in lines comprising the quatrain of iambic (and trochaic) tetrameter with cross rhyme (abab) in Russian poetry. He observed that in the nineteenth century such stanzas develop a distinctive wave-like quality, in that the second and the fourth lines are least likely to be stressed. The alternation of heavy and light lines within the quatrain suggests that the Jakobsonian principle of isomorphism can be extended to the level of the stanza. Tarlinskaja supports Gasparov's case for such structural isomorphism based on the analysis of dolnik quatrains in English and German. (29)
Our findings in part vindicate this structuralist insight, while also foregrounding the primacy of the functionalist perspective on verse organization. The Onegin stanza is based on two related, but partially independent principles. One is faithfulness to the metrical template that helps hold verse together; in this respect, it functions in the same way as the rhyme. Here the two levels of structure are found to be in a compensatory relationship. The other, isomorphic principle is reflected in the echoing pattern of deviations from the metrical template: lines associated by rich rhyme tend to share the same stress configuration.
More generally, the holistic approach to the rhyme-and-stress system in the Onegin stanza is borne out by evidence from its later evolution. Lermontov and Ivanov, in their works that employ the same stanza, introduce changes at all levels of the system simultaneously. The loosening of structural requirements in Lermontov is simultaneously reflected in a smaller proportion of rich rhymes, less pronounced asymmetries between masculine and feminine rhymes, and a lack of perceivable interaction between rhyme and stress. In Ivanov, by contrast, the structure is organized in a more rigid way, as seen in the higher proportions of rich rhymes, the striking asymmetry between masculine and feminine rhymes, and in correlations between rhyme and stress. The two trajectories of development imply a spectrum of relative rigidity in the rhyme-and-stress system. The fact that restrictions on rhyme and stress co-vary supports the hypothesis of their functional relatedness.
The history of the Onegin stanza also suggests far-reaching conclusions on the evolution of poetic form in relation to changes in culturally prevalent attitudes to literary discourse. The "natural" life of literary forms is reflected in the progressive crystallization of a new stanza within the creative history of Pushkin's Eugene Onegin, as well as in its destabilization in Lermontov's Tambov Treasurer's Wife, a poem that, while uncomfortable with the established conventions of literary language, continues the process of obfuscation of the stanza's original composition that was initiated within Eugene Onegin. In this process, the refashioning of the form occurs spontaneously in the quest for a new expressive potential.
The "artificial" resurrection of older forms, characteristic of European Modernism in general, is represented by Ivanov's masterly evocation of the Pushkinian stanza in Infancy. Paradoxically, it is as part of such creative restoration that novelty often arises: while successful in bringing back some aspects of the Pushkinian stanza, and even making them more prominent than in the original, Ivanov also introduces a new principle of privileging the first members of rhyming pairs of lines. (30) Literary scholars should not forget that beyond the level of immediate aesthetic appreciation, unbeknownst both to the author and to the reader, poetic forms undergo immanent processes of modification, oblivion, and innovation. Statistical analysis of meter and rhyme offers a tool for restoring due significance to that under-explored aspect of literary history.
Although in his well-known discussion of genre memory Bakhtin does not address topics in Russian prosody, due to an interest in the longue duree, his observations provide us with a sound foundation for understanding how a form evolves within a briefer term. Foregoing the approach that favors the poets' power to design the text, we should instead assign the central role to what enables linguistic structures to reveal what history hides, yet poets feel.
(1.) For an introduction to comparative metrics rich in examples from different traditions, see Fabb.
(2.) See Jakobson, "Slavic Epic Verse." For a survey of historical ties between European traditions of verse see Gasparov, A History of European Versification.
(3.) Gasparov, A History, passim, and, more recently, Kazartcev, who demonstrates that influences linking different national traditions often trump natural language propensities. Hanson and Kiparsky postulate the principle of "fit" that requires languages to "select meters in which their entire vocabularies are usable in the greatest variety of ways" (294).
(4.) For important applications of the "Russian method," see Bailey; Gasparov, "A Probability Model"; and the works of Marina Tarlinskaja.
(5.) Important work on English metrics, focusing on iambic pentameter, includes Tarlinskaja, English Verse; Kiparsky; Hanson and Kiparsky. For introductions to Russian verse see Zhirmunskii, Introduction to Metrics; and Scherr, Russian Poetry. On rhyme see Eekman (general typology of rhymes), Samoilov (literary history of rhyme in Russia), Scherr, Russian Poetry 193-224, as well as the overview in Brogan who stresses the overall poverty of scholarship in the field. See also the English translation of the introduction to Zhirmunsky's classic monograph on rhyme.
(6.) Benveniste defines the original meaning of the word as "the form in the instant that it is assumed by what is moving, mobile, and fluid" (285-86).
(7.) Wordsworth 1.94. On the commitment to rich rhymes among the French Romantics cf. Zhirmunskii, Rifma 104-5, 318 and de Cornulier, with earlier literature cited therein.
(8.) In fact, the principle of masculine--feminine alternation was first borrowed from the French by German poets, whose influence was paramount in molding Mikhail Lomonosov's syllabo-accentual verse (Brogan 737; Scherr, Russian Poetry 211). The feminine rhymes in French are due to the retention of e muet, no longer pronounced in the spoken language.
(9.) Cf. Gasparov, "Strofika" 350. Gasparov also discusses forms of rhymed poetry with highly irregular stanza-like organization, especially favored in Russia by the Romantics. On the problem of two-verse units: Tomashevskii 297-98.
(10.) This distinction is important for English verse, in which identical rhymes (but not rich rhymes in general) are avoided. Wagner and McCurdy present a linguistic explanation of this restriction based on general avoidance of repetition at the end of phrases, observed in English but not in French.
(11.) In calculating the coefficients, Gasparov, "Evoliutsiia, rewards by extra points the use of several identical sounds before the rhyming vowel (ograda-otrada counts as 1; ograda-konokrada as 2; ograda-mimo grada as 3) (295). The increased use of rich rhyme in twentieth century Russian poetry compensates for the rise of inexact rhyme (the rhyme in which the consonants following the rhyming vowel do not coincide); see note 15 for further discussion.
(12.) Eugene Onegin consists of 5,100 complete lines, excluding two short portions that do not conform to the Onegin stanza: Tatiana's letter to Onegin in Chapter 3, and Onegin's letter to Tatiana in Chapter 8. Lermontov's Tambov Treasurer's Wife consists of 729 complete lines, Ivanov's Infancy, of 686 lines. The Onegin stanza was adopted for extended narrative in English by Vikram Seth in his novel in verse The Golden Gate (1986) and in Hebrew by Maya Arad in Another Place, a Foreign City (2003).
(13.) Earlier work on the structure of the Onegin stanza includes Shengeli; Tomashevskii; Postoutenko; Scherr, "Structural dynamics."
(14.) The order of rhymes in the table reflects the order in which they appear in the poem: f precedes E, even though the first E-line precedes the first f-line. For Lermontov, the number of rhymes is smaller than predicted from the overall number of lines, due to a large number of incomplete and missing lines (in such cases, no rhyme was considered to be present).
(15.) The latter explanation is indirectly supported by Gasparov's observation, based on large-scale statistical analysis, that inexact rhymes are more likely to be rich than inexact ones, and moreover, masculine rhymes with a different final consonant are more likely to have identical sounds "to the left" of the rhyming vowel than feminine inexact rhymes ("Evoliusiia" 298). Historically, compensatory "enrichment" of rhyme in twentieth century Russian poetry is due to the rise of inexact rhyme (cf. Scherr, Russian Poetry 200-201; Gasparov, "Evoliusiia" 338-39). Similarly, identical onset consonants compensate for differences in the following consonants.
(16.) To test whether a result is statistically significant, we use chi-square test (for large sample sizes) and Fisher's exact test (for smaller sample sizes); both tests calculate the probability of obtaining the given (or even more skewed) distribution on the assumption that the chosen factors play no role; e.g., the p-value < 0.0001 suggests that the same (or a more skewed) distribution of rich rhymes could be obtained by chance, without any influence from the type of ending, with the probability of less than 0.01 percent (which is an extremely significant value by all measures). P-values between 0.01 and 0.05 are considered marginally significant; p-values larger than 0.05 suggest that the distribution is likely to be due to chance.
(17.) Ivanov 165-66. Further on liturgical and Pushkinian allusions in Infancy, see Bird 94-102.
(18.) Our results replicate an earlier count by Shengeli (111), as recalculated by Gasparov, "Stroficheskii ritm" 189, based not on the proportion of fully stressed verse, but on a slightly different measure-a mean number of stresses in each particular line. The two measures largely coincide, with only slight differences related to the treatment of verses with two missing stresses.
(19.) The differences in the total number of lines per position are explained by the fact that not all stanzas are complete and some omit lines.
(20.) Scherr, Structural dynamics" 272-73. The original observation that the first line in any stanza is most likely to be fully stressed was made by Shengeli (109-21). This topic is explored further in Gasparov, Stroficheskii ritm," where the proportion of stressed lines is shown to decrease within AbAb stanzas in Pushkin's period. Accordingly, Scherr tentatively relates the decline in the prominence of the first line of the three quatrains to the difference in the rhyming scheme (Russian Poetry 273). As for the slight increase in the number of stresses on the last line of the second and the third quatrain, it is explained by Gasparov by a need to avoid the sense of closure that a steady decrease would convey ("Stroficheskii ritm" 190-91).
(21.) Scherr interprets this pattern differently, suggesting that line 7 could be marking the end of a unit (Structural dynamics" 278). It does not seem plausible, however, that a structural unit would end with a verse that introduces a new rhyme.
(22.) The distribution of stress profiles across the Onegin stanza in other poets writing after Lermontov (Stakhovich, Minaev, Severianin) is discussed by Mikhail Lotman.
(23.) As discussed in the following section, it is much more common for stanzaic constituents to end in a rhythmically "light" verse.
(24.) This effect could be due in part to the obfuscation of the internal underlying boundaries that are still transparent in the original stanza: in Pushkin, where each quatrain and the couplet show properties of independent units, the effect of the position in a rhyming pair might be made less obvious by additional factors. The two forces act together in shaping the dynamics of the Onegin stanza, and we have no way of assessing them separately.
(25.) Tarlinskaja, Shakespeare's Verse 266-78, Strict Stress-Meter 106-15. Notably, no such distribution is observed in Shakespeare's Sonnets, which tend to use heavily stressed maxim-like concluding verses and thus also do not show the lightening of the second member of rhyming pairs (Shakespeare's Verse 269-70). In general, comparison between the extent of faithfulness to the metrical template in the Russian Onegin stanza and in the English sonnet is complicated by the prominence of nonmetrical stressing in English poetry.
(26.) In quatrain composition, the last line of a stanzaic constituent is likely to be shorter (Hayes and MacEachern). Tarlinskaja shows that the last lines of quatrains in dolnik tend to become shorter in that they prefer monosyllabic rather than disyllabic non-ictic intervals (Strict Stress-Meter 110-15). Lev Blumenfeld offers an elegant explanation of the short-last effect in verse. The effect described in this section for Pushkin's Eugene Onegin cannot be due to the quatrain architecture, as it is more pronounced in the "a"-rhyme than in the b -rhyme pair. This effect is even stronger in Ivanov's Infancy, which does not retain the underlying stanzaic structure.
(27.) On the higher frequency of rich rhymes in adjacent positions in Russian verse, due most likely to them being grammatical (and syntactically parallel) the work of Lilly. On grammatical rhymes in Eugene Onegin cf. Worth; Postoutenko 152-54.
(28.) Reproduced from Table l in "Linguistics and Poetics" 363.
(29.) Tarlinskaja, Strict Stress-Meter 115. Gasparov, "Stroficheskii ritm," traces three phases in the evolution of quatrain architecture, which roughly correspond--in an isomorphic fashion--to the three phases in the evolution of dominant stress configurations from the eighteenth to the mid-nineteenth century observed by Taranovsky.
(30.) In Ivanov's Infancy, in the rhyming pairs of lines, the first member of the pair is consistently more likely to be fully stressed than the second one. An important topic for future research is the extent to which Ivanov's commitment to the sonnet, which is rather unusual in the Russian poetic tradition, might have influenced his modification of the Onegin stanza.
Bailey, James. Toward a Statistical Analysis of English Verse. The Iambic Tetrameter of Ten Poets. Lisse, Netherlands: Peter de Ridder P, 1975. Print.
Bakhtin, Mikhail Mikhailovich. Problems of Dostoevsky's Poetics (1963). Ed. Caryl Emerson. Minneapolis: U of Minnesota P, 1984. Print.
Benveniste, Emile. "The Notion of 'Rhythm' in Its Linguistic Expression." Problems in General Linguistics (1951). Trans. Mary Elizabeth Meek. Coral Gables, FL: U of Miami P, 1971. 281-88. Print.
Bird, Robert. The Russian Prospero: The Creative Universe of Viacheslav Ivanov. Madison: U of Wisconsin P, 2006. Print.
Blumenfeld, Lev. "End-Weight Effects in Verse and Language. A talk at the conference "Prosody Today: Comparative Perspectives on the Study of Verse," U of Chicago, March 2015.
Brogan, Terry V. F. "Rhyme." The New Princeton Encyclopedia of Poetry and Poetics. Princeton, NJ: Princeton UP, 1993. 1052-64. Print.
de Comulier, Benoit. "Rime 'riche' et fonction de la rime: le developpment de la rime 'riche' chez les romantiques." ['Rich' rhyme and the function of rhyme: the development of 'rich' rhyme among the Romantics] Litterature 59 (1985) 1115 25. Print.
Eekman, Thomas. The Realm of Rime: A Study of Rime in the Poetry of the Slavs. Amsterdam: Adolf M. Hakkert, 1974. Print.
Fabb, Nigel. Linguistics and Literature: Language in the Verbal Arts of the World. Malden, MA: Blackwell, 1997. Print.
Falen, James E. Eugene Onegin: A Novel in Verse, by Alexander Pushkin. Oxford: Oxford UP, 1995. Print.
Gasparov, Mikhail Leonovich. "A Probability Model of Verse (English, Latin, French, Italian, Spanish, Portuguese)." Trans. Marina Tarlinskaja. Style 21 (1987). 322-58 Print.
--. A History of European Versification. Trans. G. S. Smith and M. Tarlinskaja. Oxford: Clarendon P, 1996. Print.
--. "Strofika nestroficheskogo iamba v russkoi poezii XIX veka" [Stanzaic properties of non-stanzaic iamb in Russian 19th century poetry.] (1976). Izbrannye trudy, Vol. 3: O stikhe. Moscow: Iazyki russkoi kul'tury, 1997. 340-65. Print.
--. Evoliutsiia russkoi rifmy" [The evolution of Russian rhyme] (1984). Izbrannye trudy, Vol. 3: O stikhe. Moscow: Iazyki russkoi kul'tury, 1997. 290-339. Print.
--. Stroficheskii ritm v russkom 4-st. iambe i khoree" [Stanzaic rhythm in Russian iambic and trochaic tetrameter] (1989). Izbrannye trudy, Vol. 3: O stikhe. Moscow: Iazyki russkoi kul'tury, 1997. 181-95. Print.
Ginzburg, Lidiia. O lirike. [On lyric.] 2nd ed. Leningrad: Sovetskii pisateP, 1974. Print.
Hanson, Kristin, and Paul Kiparsky. A Parametric Theory of Poetic Meter." Language 72 (1996): 189-247. Print.
Hayes, Bruce, and Margaret MacEachern. "Quatrain Form in English Folk Verse." Language 74 (1998): 473-507- Print.
Ivanov, Viacheslav. "The Crisis of Humanism" (1919). Selected Essays. Trans. Robert Bird. Evanston, IL: Northwestern UP, 2001. Print.
Jakobson, Roman. "Slavic Epic Verse: Studies in Comparative Metrics" (1952), Selected Writings, Vol. 4. Gravenhage, Netherlands: Mouton, 1962. 414-63. Print.
--. "Closing Statement: Linguistics and Poetics" (1958). Style in Language. Ed. Thomas Sebeok. Cambridge, MA: MIT P, 1960. 350-77. Print.
Kazartcev, Evgenii. "Comparative Study of Verse: Language Probability Models." Style 48.2 (2014): 119-39. Print.
Kiparsky, Paul. The Rhythmic Structure of English Verse." Linguistic Inquiry 8.2 (1977): 189-247. Print.
Kliger, Ilya. Dostoevsky and the Novel-Tragedy: Genre and Modernity in Ivanov, Pumpyansky, and Bakhtin." PMLA 126.1 (2011): 73-87. Print.
--. "Resurgent Forms in Ivan Goncharov and Alexander Veselovsky: Towards a Historical Poetics of Tragic Realism." The Russian Review 71.4 (2012): 655-72. Print.
Kliger, Ilya, and Boris Maslov, eds. Persistent Forms: Explorations in Historical Poetics. New York: Fordham UP, 2015. Print.
Kunichika, Michael. The Scythians Were Here ...' On Nomadic Archeology, Modernist Form, and Early Soviet Modernity." Ab Imperio 2012 (2): 229-57. Print.
Lilly, Ian K. "On Adjacent and Nonadjacent Russian Rhyme Pairs." The Slavic and East European Journal 24.3 (1980): 245-55. Print.
--. "On Adjacent and Nonadjacent Russian Rhyme Pairs II." The Slavic and East European Journal 29.2 (1985): 188-97. Print.
Lotman, Mikhail Yur'evich. "Ritmicheskaia struktura oneginskoi strofy i problemy stroficheskogo ritma. [Rhythmic structure of the Onegin stanza and the problem of stanzaic rhythm.] Poetry and Poetics: A Centennial Tribute to Kiril Taranovsky. Ed. Barry P. Scherr, James Bailey and Vida T. Johnson. Bloomington, IN: Slavica, 2014. 61-89. Print.
Postoutenko, Kirill. Oneginskii tekst v russkoi literature. [The Onegin text in Russian literature.] Pisa: ECIG, 1998. Print.
Samoilov, David Samoilovich. Kniga o russkoi rifme. [A book on Russian rhyme.] Moscow: Khudozhestvennaia Literature, 1982. Print.
Scherr, Barry P. Russian Poetry: Meter, Rhythm, and Rhyme. Berkeley: U of California P, 1986. Print.
--. "Structural Dynamics in the Onegin Stanza." Formal Approaches to Poetry: Recent Developments in Metrics. Ed. N. Friedberg and B. Dresher. Berlin: de Gruyter, 2006. 267-86. Print.
Shengeli, Georgii Arkad'evich. Traktat O russkom stikhe. [A treatise on Russian verse.] Moscow: Gos. iz-vo, 1923. Print.
Taranovskii, Kiril. Ruski dvodelni ritmovi. [Russian binary rhythms.] Belgrade: Naucna knjiga, 1953. Print.
Tarlinskaja, Marina. English Verse: Theory and History. The Hague, Netherlands: Mouton, 1976. Print.
--. Shakespeare's Verse: Iambic Pentameter and the Poet's Idiosyncrasies. New York: Peter Lang, 1987. Print
--. Strict Stress-Meter in English Poetry Compared with German and Russian. Calgary, Canada: U of Calgary P, 1993. Print
--. "What is 'Metricality? English Iambic Pentameter." Formal Approaches to Poetry: Recent Developments in Metrics. Ed. N. Friedberg and B. Dresher. Berlin: de Gruyter, 2006. 53-74. Print.
Tomashevskii, Boris Viktorovich. "Strofika Pushkina" [Pushkin's stanzas.] (1958). Pushkin: raboty raznykh let. Moscow: Kniga, 1990- 288-483. Print.
Tynianov, Yuri. "Promezhutok." [An Interval.] Poetika, istoriia literatury, kino. Moscow: Nauka, 1977. 168-95. Print.
Wagner, Michael, and Katherine McCurdy. "Poetic Rhyme Reflects Cross-linguistic Differences in Information Structure." Cognition 117 (2010): 166-75. Print.
Wordsworth, William. The Prose Works of William Wordsworth. Ed. W. J. Owen and Jane Worthington Smyser. 3 vols. Oxford: Clarendon Press, 1974. Print.
Worth, Dean S. "Grammatical Rhyme Types in Evgenij Onegin." Alexander Pushkin: Symposium II. Ed. A. Kodjak, K. Pomorska, K. Taranovsky. Columbus, OH. Slavica, 1980. 39-48. Print.
Zhirmunskii, Viktor Maksimovich. Rifma, ee istoriia i teoriia. [Rhyme: its History and Theory.] St. Petersburg: Academia, 1923. Print.
--. Introduction to Metrics (1925). Trans. C. E. Brown. The Hague, Netherlands: Mouton, 1966. Print.
--. "Introduction to Rhyme: Its History and Theory" (1923). Trans. John Hoffman. Chicago Review 57.3/4 (2013): 119-27. Print.
CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE, LLACAN (LANGAGE, LANGUES ET CULTURES D'AFRIQUE NOIRE)
UNIVERSITY OF CHICAGO
TATIANA NIKITINA is a Researcher at the French Centre national de la recherche scientifique. She has published numerous articles on the grammar of West African languages and cross-linguistic variation in the encoding of space and person. Her current work focuses on the applications of statistical and experimental methods to the study of grammar and stylistic variation.
BORIS MASLOV is Assistant Professor of Comparative Literature at the University of Chicago. He is the author of Pindar and the Emergence of Literature (Cambridge UP, 2015), as well as of articles on the history of literary theory, the reception of Ancient Greek lyric, and comparative historical semantics.
Table 1: The coefficient of "richness" of rhyme, according to M. L. Gasparov ("Evoliutsiia" 321) 1745-1780 37,2 1780-1800 28,2 1800-1815 20,0 1815-1830 15,1 1830-1845 14,2 1845-1860 15,4 1860-1890 18,4 1890-1905 20,7 1905-1913 27,6 1913-1920 38,4 1920-1930 49,8 1930-1935 56,5 Table 2: Distribution of rich and poor rhymes within the Onegin stanza (14) Pushkin's Eugene Onegin A B C d f E g Rich 55 152 45 139 152 38 146 (15%) -41.50% (13%) (38%) (42%) (10%) (40%) Poor 311 214 320 226 211 325 216 Total 366 366 365 365 363 363 362 Lermontov's Tambov Treasurer's Wife Rich 5 (10%) 10 5 (90%) 17 19 8 17 (19%) (32%) (38%) (17%) (34%) Poor 47 42 48 36 31 39 33 Total 52 52 53 53 50 47 50 Ivanov's Infancy Rich 15 29 17 27 29 14 32 (31%) (59%) (35%) (55%) (59%) (29%) (65%) Poor 34 20 32 22 20 35 17 Total 49 49 49 49 49 49 49 Table 3: The difference in richness between masculine and feminine rhymes Rich Poor Total Pushkin's Eugene Onegin Masculine 589 (40%) 867 1456 (100%) Feminine 138 (13%) 956 1094 (100%) Lermontov's TTW Masculine 63 (31%) 142 205 (100%) Feminine 18 (12%) 134 152 (100%) Ivanov's Infancy Masculine 117 (60%) 79 196 (100%) Feminine 46 (31%) 101 147 (100%) Table 4: Fully stressed lines vs. lines with at least one missing stress in Onegin, by line (19) 1 2 3 4 5 6 7 Fully 152 94 90 81 117 95 66 stressed Missing 214 272 276 285 248 270 299 stress Total 366 366 366 366 365 365 365 8 9 10 11 12 13 14 Fully 96 104 85 88 89 106 90 stressed Missing 269 259 278 275 274 256 272 stress Total 365 363 363 363 363 362 362 Table 5: Fully stressed lines by position in the stanza, for the two parts of Eugene Onegin Chapters 1-4 1 2 3 4 5 6 7 Fully stressed 80 47 47 34 62 49 36 Missing stress 98 131 131 144 1116 129 142 Total 178 178 178 178 178 178 178 Chapters 5-8 1 2 3 4 5 6 7 Fully stressed 72 47 43 47 55 46 30 Missing stress 116 141 145 141 132 141 157 Total 188 188 188 188 187 187 187 8 9 10 11 12 13 14 Fully stressed 44 53 39 40 40 53 38 Missing stress 134 124 138 137 137 123 138 Total 178 177 177 177 177 176 176 Chapters 8 9 10 11 12 13 14 Fully stressed 52 51 46 48 49 53 52 Missing stress 135 135 140 138 137 133 134 Total 187 186 186 186 186 186 186 Table 6: Realization of metrical template in the rhyming pairs of lines, Eugene Onegin Fully At least one stressed line stress missing Total First member 724 (28%) 1826 (72%) 2550 Second member 629 (25%) 1921 (75%) 2550 Table 7: Realization of metrical template in the rhyming pairs of lines, Ivanov's Infancy Fully At least one stressed line stress missing Total First member 109 (32%) 234 (68%) 343 Second member 84 (24%) 259 (76%) 343 Table 8: Realization of metrical template in the rhyming pairs of lines, Lermontov's TTW Fully stressed At least one line stress missing Total First member 92 (26%) 265 (75%) 357 Second member 114 (32%) 243 (70%) 357 Table 9: Stress omission by type of rhyme (second members of a rhyming pair only) Pushkin's Eugene Onegin Fully stressed At least one line stress missing Total Lines with poor rhyme 487 (27%) 1336 1823 Lines with rich rhyme 142 (20%) 585 727 Ivanov's Infancy Lines with poor rhyme 53 (29%) 127 180 Lines with rich rhyme 31 (19%) 132 163 Lermontov's TTW Lines with poor rhyme 92 (33%) 184 276 Lines with rich rhyme 22 (27%) 59 81 Table 10: Stress omission by type of rhyme, masculine rhymes (second members only) Pushkin's Eugene Onegin Fully stressed At least one line stress missing Total Lines with poor rhyme 241 (28%) 626 867 Lines with rich rhyme 114 (19%) 475 589 Ivanov's Infancy Lines with poor rhyme 26 (33%) 53 79 Lines with rich rhyme 23 (20%) 94 117 Table 11: Stress omission by type of rhyme, feminine rhymes (second members only) Pushkin's Eugene Onegin Fully stressed At least one line stress missing Total lines with poor rhyme 246 (26%) 710 956 lines with rich rhyme 28 (20%) 110 138 Ivanov's Infancy lines with poor rhyme 27 (27%) 74 101 lines with rich rhyme 8 (17%) 38 46 Table 12: The "rhyming" of stress configuration, by type of rhyme, Eugene Onegin Identical stress Different stress configuration configuration Total Lines with poor rhyme 605 (33%) 1218 1823 Lines with rich rhyme 291 (40%) 436 727 Table 13: The "rhyming" of stress configuration, by type of rhyme Lermontov's TTW Identical stress Different stress configuration configuration Total Lines with poor rhyme 93 (34%) 183 276 Lines with rich rhyme 32 (40%) 49 81 Ivanov's Infancy Lines with poor rhyme 56 (31%) 124 180 Lines with rich rhyme 56 (34%) 107 163 Table 14: Identical stress configurations in nonadjacent pairs of rhyming lines (lines 3, 4, 12), Eugene Onegin Identical stress Different stress Total configuration configuration Lines with poor rhyme 280 (33%) 570 850 Lines with rich rhyme 106 (43%) 139 245 Table 15: Identical stress configurations in adjacent pairs of rhyming lines (lines 6, 8, 11, 14), Eugene Onegin Identical stress Different stress Total configuration configuration Lines with poor rhyme 325 (33%) 648 973 Lines with rich rhyme 185 (38%) 297 482 Table 16: Identical stress configurations in adjacent pairs of non-rhyming lines (lines 3, 4, 12), Eugene Onegin Identical stress Different stress Total configuration configuration Lines with poor rhyme 305 (36%) 545 850 Lines with rich rhyme 81 (33%) 164 245 Figure 1. Distribution of rich rhymes in three texts Onegin TTW Infancy A 55 5 15 b 152 10 29 C 45 5 17 d 139 17 27 f 152 19 29 E 38 8 14 g 146 17 32 Note: Table made from line graph. Figure 2. Fully stressed lines by position in the stanza, Onegin. 1 152 2 94 3 90 4 81 5 117 6 95 7 66 8 96 9 104 10 85 11 88 12 89 13 106 14 90 Note: Table made from line graph. Figure 4. Fully stressed lines by position in the stanza, Chapters 1-4 of Eugene Onegin 1 80 2 47 3 47 4 34 5 62 6 49 7 36 8 44 9 53 10 39 11 40 12 40 13 53 14 38 Note: Table made from line graph. Figure 5. Fully stressed lines by position in the stanza, Chapters 5-8 of Eugene Onegin 1 72 2 47 3 43 4 47 5 55 6 46 7 30 8 52 9 51 10 46 11 48 12 49 13 53 14 52 Note: Table made from line graph. Figure 6. Fully stressed lines by position in a rhyming pair first member second member Pushkin 724 629 Ivanov 109 84 Lermontov 92 114 Note: Table made from bar graph.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Eugene Onegin|
|Author:||Nikitina, Tatiana; Maslov, Boris|
|Article Type:||Critical essay|
|Date:||Dec 22, 2015|
|Previous Article:||Dual textual dynamics and dual readerly dynamics: double narrative movements in Mansfield's "Psychology".|
|Next Article:||Bruegel and John Berryman.|