Two substantial scholarly treatises on the history of the villanelle are in existence: Ronald E. McFarland’s The Villanelle: The Evolution of a Poetic Form (Moscow, ID: University of Idaho Press, 1987) and Julie Ellen Kane’s unpublished dissertation, “How the Villanelle’s Form Got Fixed” (Louisiana State University, 1999).[i] The latter dramatically revises the former in some important respects. Together they provide a detailed history of the villanelle that contradicts almost all the briefer accounts.
McFarland’s book, as it happens, is the cited source on the villanelle in Poetic Designs, the reference I consulted at the commencement of my own interest in the form. The Villanelle attempts three things: to locate instances of the villanelle, to rank the intrinsic worth of these instances, and to narrate a consistent history of the villanelle. The first goal is fulfilled by one appendix listing sixty-seven villanelles in Spanish, Italian, French, and English dating as far back as the fifteenth century and a second appendix reprinting the full text of thirty-three additional contemporary villanelles in English.[ii] Throughout, evaluation is predicated on the opposition of terms such as “sophisticated” and “trivial”; “serious” and “light”; and “experimental” and “conventional”; with the former terms positive and the second terms negative. It is primarily a poem’s content, or subject matter, that is subjected to this evaluation, but also such formal attributes as enjambment and alteration of the refrain. For instance, of Edmund Gosse’s “Wouldst thou not be content to die” (1878), McFarland opines that
for the most part the poem is conventional in sentiment and essentially descriptive or pictorial in mode. Despite the implied presence of another person, no genuine dramatic moment occurs in the poem, nor is there any inner tension on the part of the speaker. Certain elements of Gosse’s art, however, should be credited. Unlike some of his followers, he does allow some variation in the refrain lines, and he moves fairly naturally from the second line of each stanza (without excessively pronounced end punctuation) to the refrain line; the penultimate tercet moves easily into the concluding stanza. (64-5)
Evaluations such as these constitute the majority of the book.
McFarland’s literary-historical narrative is also essentially an evaluation, but an evaluation of the villanelle as a form. He argues that the villanelle has become a versatile form in the hands of good modern poets, one that is capable of expressing the deepest philosophies and emotions:
The villanelle has [...] escaped the nearly automatic triviality of its early application in English and in many French poems. Trivial poets will always exist, though, and they will turn to the villanelle on occasion. The form does lend itself to sometimes delightful comic uses. If nothing else, I hope this study proves its flexibility, even when the poems are composed according to the strictest rules.
McFarland observes that what he calls the “loose-lined” villanelle (characterized by enjambment and varied refrains) has come to be increasingly common, and this seems coextensive with the introduction of more “profound” themes than those explored by the sixteenth-century and nineteenth-century practitioners of the form:
By the end of the 1950′s [...], the villanelle had been proven, at least in the hands of able poets, capable of the profoundest themes, and of tones ranging from the comic to the apocalyptic, from the meditative to the enraged. Most important, however, were the achievements with the form by poets of major stature, from Pound and Auden to Thomas and Roethke. Among these poets some rather cautious experimentation with the form set the groundwork for poets of the next generation, who were to examine rigorously its potential and its limits. (96)
By this account, modern poets surpassed their Victorian (and Renaissance) predecessors in thematic exploitation of the villanelle, while contemporary poets have surpassed their modern predecessors in formal experimentation with the villanelle.
McFarland is chiefly interested in this aesthetic “evolution” of the villanelle as it has played out in the twentieth century, but he does include two chapters on the villanelle’s origins in sixteenth century France and fifteenth century Italy, and he in fact sees this as one of the distinctive contributions of his work:
Very few writers of the villanelle in English are aware of its complete history any more than I was when I began research into the subject. Many poets have been influenced only by direct predecessors who have written villanelles in English. Some have been aware of French antecedents, usually limited to Jean Passerat. Few, however, have known of its Italian origins. While the primary intent of this book is to examine the development of achievements in the form by English and American poets, I have devoted almost half of it to the history of the villanelle in Italian and French poetry. (x)
It is this early history of the villanelle that Kane’s dissertation focuses on, and this early history it radically alters. While Kane does correct McFarland on some factual assertions, it is primarily his smooth narrative of gradual development (“evolution”) that her work interrogates and disrupts.
Kane reports that she can find only eighteen examples of poems designated as “Villanelle” or “Villanesque” written in France between 1553 and 1627, and reveals that
no two of the eighteen are identical in rhyme scheme, length, and syllable count. They do not resemble each other but, in most cases, each resembles an actual musical villanella or villancico. It cannot possibly be said that there was anything resembling a poetic “form,” let alone a fixed poetic form, for the villanelle in the sixteenth century (155).
The villanella and the villancico were at that time Italian and Spanish dance-song forms, musically simpler than the polyphonic madrigal. French poets adopting the title “Villanelle” or “Villanesque” thus probably meant to indicate that their poems, if set to music (as was commonly done), should have simple rather than complex settings; the simplicity of musical setting usually suggested a “simplicity” of character or theme, too, best evoked by terms such as “pastoral” or “rustic.” Kane explains that sixteenth-century poets in England used the term “Neapolitan” or “Napolitane” (and, later, the term “canzonetta”) rather than the term “villanella,” but that although the terms were different, “In both countries it was strongly associated with music and with an ‘old-fashioned’ oral poetic tradition of semi-improvising lyrics to a preexisting tune” (202). In other words, to title a poem “Villanelle” would have been something akin to titling a poem “Blues” today. Structurally, the villanella had no rule other than that it usually had a refrain, which was–as in the popular song forms of any era–a single refrain, not an alternating one. The terms “villanella” and “villanelle” referred to musical distinctions, not verbal distinctions; they were by no means set poetic forms, as the sonnet then was, and even as the triolet and the rondeau were.
McFarland’s first two chapters, then, on the early history of the villanella and the villancico, are shown to be more irrelevant than inaccurate–though also inaccurate, as for instance on the date of composition of Passerat’s “J’ay perdu ma Tourterelle,” which Kane corrects from “ca. 1590″ to 1574. McFarland’s first two chapters, in fact, continually mention the fact that the villanella was associated with music and that neither it nor any sixteenth-century French villanelle had a fixed form, but without seeming to recognize that this makes nonsense of his narrative of steady development from that age to this. Being in fact more interested in evaluation than in evolution, despite its title, McFarland’s work misses the point that in modern usage, the term “villanelle” means a poem modeled after the scheme A’bA” abA’ abA” abA’ abA” abA’A”. His work The Villanelle is based on an equivocation, since the first two chapters are accounts of “the villanelle” in a different sense–roughly and briefly, the sense of “peasant song.” Even my dictionary defines the word “villanella” as meaning “a rustic Italian part song,” whereas the separate word “villanelle” is in prosody “a short poem of fixed form, written in tercets” etcetera.[iii]
The only poem of the Renaissance on the now-familiar scheme A’bA” abA’ abA” abA’ abA” abA’A” was Jean Passerat’s “J’ay perdu ma Tourterelle,” first published in 1606. Kane’s research, methodologically committed to the systematic examination of primary sources, shows that the scheme of this poem was only “fixed” as a schematic type in the nineteenth century. She reports that in 1845, Théodore de Banville published a parodic poem on the model of Passerat’s “Villanelle,” titled “Villanelle de Buloz,” in the Parisian periodical Silhouette. “Villanelle de Buloz” was reprinted in the Journal d’Alençon in February of 1857 and also appeared later that year in the book for which Banville is best known, his Odes Funambulesques. When Théodore de Banville’s Petit traité de poésie française came out in 1872, it defined the villanelle as a type of fixed form:
La Villanelle est divisée en tercets. Elle commence par une vers féminin.
Il ne paraît pas qu’elle comporte un nombre fixe de tercets.
Elle est écrite sur deux rimes: l’une, masculine, qui régit le second vers de tous les tercets; l’autre, féminine, qui régit les autres vers. [...] Et rien n’est plus chatoyant que ce petit poëme. On dirait une tresse formée de fils argent et d’or, que traverse un troisième fil, couleur de rose! (214-5)
The Villanelle is divided into tercets. It begins with a feminine line.
It does not appear to consist of a fixed number of tercets.
It is written on two rhymes: one masculine, which governs the second line of all the tercets; the other feminine, which governs the other lines. [...] And nothing is more sparkling than this little form. One might say that it is a plait woven from strands of silver and gold, through which winds a third strand the color of a rose![iv]
The 1751 edition of Richelet’s Dictionnaire de rimes also contributed to later confusion,[v] but in Kane’s work it is primarily Banville who is credited with creating the legend of the schematic villanelle out of whole cloth–though even his scheme differs in several respects from that usually cited by handbooks today, since Banville declared that the entire poem could be of any length and that the refrains should have feminine endings, stipulations that have disappeared from English definitions. The nineteen-line model was solidified in the public imagination by one Joseph Boulmier, who published in Paris in 1878 an entire book of schematic nineteen-line villanelles explicitly modeled upon Jean Passerat’s “J’ay perdu ma Tourterelle.” Banville’s Petit traité was read with interest and admiration by Edmund Gosse and Austin Dobson, who went on to write schematic villanelles of their own and to champion the form in England as one that ought to be “revived.” It is not too much to say that instead they manufactured the form as a type.
From that point on, Kane and McFarland and other scholars are in essential agreement on the basic history of the villanelle, from its popularity at the end of the nineteenth century with a coterie of vers de société poets (including Gosse, Dobson, W. E. Henley, Andrew Lang, Ernest Dowson, and Oscar Wilde) to its featuring prominently in Joyce’s Portrait of the Artist as a Young Man; from its successful remaking by Dylan Thomas and Theodore Roethke at mid-century to its current firm standing in contemporary English poetry. But contemporary handbooks, guides, manuals, primers, introductions to poetry in English have frequently reproduced each others’ inaccuracies and assumptions about the early history of the villanelle.
These works often make two broad claims that Kane’s work definitively disproves: that the villanelle is essentially French, and that it was schematized in Italy or France at some vague period during or before the Renaissance, by Passerat or by one of his predecessors or contemporaries. Mary Oliver’s Rules for the Dance (1998) calls the villanelle “originally and still a French form [...] usually, but not always, in iambic pentameter” (53). Stephen Adams’s Poetic Designs (1997), as we’ve seen, reports that “The meter can be iambic trimeter, tetrameter, or pentameter (or, I suppose, any meter a poet can sustain). Like the triolet, this form [...] was associated with pastoral in French Renaissance poetry” (93). Dacey and Jauss’s New Formalist anthology Strong Measures (1986) states in an appendix with definitions of forms that the villanelle is “A French syllabic form. [...] The lines may be of any single length” (449). Judson Jerome’s The Poet’s Handbook (1980) explains that “French and Italian poetry produced many fixed forms, often with complex rhyme schemes which are difficult to adapt to English because rhymes in our language are so much more scarce. [...] Many villanelles are in tetrameter” (129). Both the 1974 and 1965 editions of the Encyclopedia of Poetry and Poetics published by Princeton inform the reader that the villanelle is “A Fr. verse form, derived from an It. folk song of the late 15th- early 17th c.” and that it “has since Passerat retained the following pattern [...]” (893, 893).
The confusion about the villanelle’s meter exhibited in these examples is only one indicator that the form’s history has never been well and thoroughly understood by Anglophone poets and scholars. The confusion can also be attributed to the fact that the villanelle, like the majority of French poetry, has never been associated with a particular meter. In fact, it is almost contradictory to speak of a form as both metrical and French, since almost all French poetry is essentially syllabic, making the distinction irrelevant.[vi] Dacey and Jauss get closest to the truth, albeit a bit tautologically, when they label the villanelle “A French syllabic form.” Villanelles in English, of course, are free to be accentual-syllabic if they like, though equally of course they can also be accentual or non-metrical.[vii] Judson Jerome might have been thinking of nineteenth-century villanelles such as W. E. Henley’s “A dainty thing’s the villanelle” and Oscar Wilde’s “Theocritus” when he made the assertion that “Many villanelles are in tetrameter” while Mary Oliver might have been thinking of twentieth-century villanelles such as “Do not go gentle into that good night” and “The Waking” when she maintained that villanelles are “usually [...] in iambic pentameter.” Meter is a major part of the definition of most English poetic forms, and these forms tend to be accentual-syllabic–indeed, the term “meter” is sometimes informally used as a synonym for accentual-syllabism. It is clear, at any rate, that the habit of associating poetic form with meter is widespread among English-speaking scholars, and that this habit has in the case of the villanelle led to error and inconsistency.
There are handbooks and guides that go into more accurate detail, however, and there are some that refrain from going into inaccurate detail, preferring to give the alphabetic rhyme scheme of villanelle almost without embellishment. The entry in the 1965 and 1974 editions of the Princeton Encyclopedia of Poetry and Poetic Forms wisely makes no mention of meter, for instance. The 1993 New Princeton Encyclopedia of Poetry and Poetics quietly points out that the villanelle “first had as its only distinguishing features a pastoral subject and use of refrain; in other respects it was without rule.” The entry goes on to assign primary responsibility for the fixed nineteen-line villanelle to the English, not the French:
While the French poets who revived the villanelle in the later 19th century treated it as a stanza type, their English counterparts, however, invested it with the status of a fixed form. Although Austin Dobson tried to present the villanelle to the English as he found it in Banville, declaring “there is no restriction as to the number of stanzas,” his compatriots stuck rigidly to the 19-line Passerat model popularized by Joseph Boulmier. (1358; abbreviations expanded)
This entry was written by Clive Scott, then lecturer in French and now Professor of European Literature at the University of East Anglia; it is telling that a more accurate account of the villanelle comes from a lecturer in French than from a scholar of Anglophone poetry.[viii]
French versification is to be distinguished from English versification (syllable-stress metre) by the following broad principles: the integrity of the French line depends on the number of its syllables rather than on the number and nature of its rhythmic segments (French: mesures, English, ‘feet’); the position of French ‘accents’ (equivalent of English stresses) is determined by the syntactic structure of the line rather than by the inherent stress patterns on individual words; the French accent falls on the last accentuable syllable of each syntactic unit in the line, and since these units naturally vary in length, French rhythmic measures obey no law of recurrence and no principle of regularity, and thus have no connection with the notion of beat; because French accents are linked also with pitch, and because the French line always ends with an accentuated syllable, there is a natural tendency in French verse for the end of the line to coincide with a syntactical break, that is, to be endstopped; it is for this reason that enjambement is potentially a greater transgression in French verse than in English. Individual lines of verse in French thus have a peculiar rhythmic autonomy and the rhythms of one line in no way predict the rhythms of the line following. (198)
It should also be mentioned that there are experiments with non-syllabic meters in French, not least of which is the vers mesurés movement in which Jean Passerat took part in the sixteenth century. Vers mesurés are essentially the equivalent of quantitative verse in English; both are an attempt to reproduce the meter of classical Latin poetry through an emphasis on the duration of syllables rather than on the number of syllables.