Tirra-Lirrical Ballads: Source Hunting with the Lady of Shalott

Early in 1868, two critics were speculating about the origin of Alfred Tennyson’s already classic “The Lady of Shalott.” Here is Frederick James Furnivall writing to William Michael Rossetti with an answer from the horse’s mouth: “As you kindly took trouble about The Lady of Shalott for me, you are entitled to a copy of Tennyson’s own account:—‘I met the story first in some Italian novelle: but the web, mirror, island, etc., were my own.’”¹ A notebook from Tennyson’s days at Trinity College records, “Legends. / The Lady of Scalot. Novelle Antiche,” apparently confirming that he had found inspiration in a thirteenth-century collection of tales called Cento novelle antiche (One hundred ancient tales).² The collection was known in the nineteenth century for having inspired many of the stories in Boccaccio’s 1353 Decameron. Among the hundred ancient tales is, indeed, a brief novella about a “damigella di Scalot” who died for love of Lancelot.

But no amount of explanation or evidence has satisfied scholars, who have doggedly sought additional sources for Tennyson’s 1832 ballad. Without discounting connections to the Scalot novella, they have argued for a range of other influences, from the Malory romance Tennyson claimed not to have used to Sappho, Spenser, and Shelley.³ Even the avowed Italian source has yielded more questions than answers: Was Tennyson working from an Italian-language edition of Cento novelle antiche available to nineteenth-century readers of Italian, or was he working from an English translation of the novella in Thomas Roscoe’s 1825 The Italian Novelists? What use did he make of two contemporaneous poems by Louisa Stuart Costello and Letitia Elizabeth Landon also derived from the Scalot novella?⁴ And why is the Italian story missing so many of Tennyson’s crucial details: the “web, mirror, island, etc.” that he claimed as “my own”?⁵ These discrepancies led Isobel Armstrong to the agnostic conclusion that the poem “has no source, and is in fact the conflation of a number of mythic structures”—but the same mismatches have tantalized many other source-hungry critics with the prospect of further discoveries.⁶

Why did Victorians care so much about the origin of this poem, and why do we still care? Perhaps it is the vertigo a reader experiences on first learning that this consummately English ballad about England’s mythic past is so embroiled in Italian literary history. I myself have spent countless hours tracking the source of Tennyson’s source. On a tip from the Victorian translator Thomas Roscoe, who suggested that the novella was first recorded by Dante’s mentor Brunetto Latini, I have searched Latini’s work for references to the Lady of Shalott—or to the variants “Scalot” and “Scalotta”—and I have searched for information about the Arthurian romance that Latini was thought to own “a beautiful copy of.”⁷ I was hoping to find in the poem’s shadowy prehistory some clue to the form of “The Lady of Shalott,” to learn something more about Tennyson’s particular variety of lyrical ballad.

But the poem’s source problems are only replicated in the literary history of the tale. In choosing the story of the Lady of Shalott—by whatever means he found it—Tennyson was engaging with a source question as much as a source. In the 1814 History of Fiction, a work that Roscoe drew on for his introduction to The Italian Novelists, John Colin Dunlop characterizes Cento novelle antiche as both an origin point and a collection of enigmatic histories. It constitutes “the first regular work of the class [of short fiction] in Europe,” and it “laid the foundation of the most splendid efforts of human genius.” At the same time “it was not a new and original production, but a compilation of stories already current in the world.” The diverse tales, from diverse literary cultures, may have been recorded by one author or many, “but who these authors were,” Dunlop announces, “is still a prob lem in the literary annals of Italy” (my emphasis). Rejecting authorship claims for both...