Genetic Joyce Studies - Issue 7 (Spring 2007)
The Mystery of the Fuga per Canonem Solved
Susan Brown

Waiting for us in a closet in Paris was a trove of hundreds of pages of previously believed lost Joyce manuscripts and notes, only one page of which has inspired this inquiry. It is a crucial page, however. It contains Joyce's fugal intentions for "Sirens."

Joyce's claim of a fugal structure for "Sirens" has seduced but ultimately defeated critics for over eighty years. There is no need to rehearse the dozens of conflicting attempts to explicate the fugal structure of "Sirens" which have kept critics busy. The theories range from the argument that a strict canonical structure applies to the episode to the claim that "Sirens" intentionally mirrors a Bach fugue to the more recent position that Joyce's fugal reference was a "hoax" (Sebastian Knowles lists "fugal form in 'Sirens'" under "bogus statements" in the index of Dublin Helix 167). More importantly, no thesis has emerged as sufficiently definitive to become the standard reading in Joyce studies or in the classroom.

But then, we've had very little to work, and what we do have is contradictory and inaccurate. In his schemata, Joyce listed the technique for "Sirens" as "fuga per canonem," expanding this claim to "all the eight regular parts of a fuga per canonem," in a July 1919 letter to Harriet Shaw Weaver (Letters I: 129). In June of 1919, according to Georges Borach, Joyce enlarged on this plan in a conversation but now described "Sirens" as a "fugue with all musical notations" (459). Any definitive reading of "Sirens" is immediately frustrated by the mutually exclusive terms fuga per canonem and fugue. Not only did Joyce create the confusion by using both terms, but Stuart Gilbert's commentary on "Sirens" --- which Gilbert claims "reproduces" "word for word" information given to him by Joyce --- also interchanges the terms (ix). However these terms are not interchangeable. In fact, the contrary. The fuga per canonem is a strict form also known as a round (as in "Three Blind Mice" or "Frere Jacques"; see Honton 41) which, according to Grove's Dictionary of Music and Musicians, is "descended from the contrapuntal experiments of mediaeval monks" in the sixteenth century. The second is a radically complex contrapuntal form developed in the seventeenth and eighteenth centuries, most innovatively by Bach, and --- Grove's explains --- is "contrary to the rules" (Williams 114-15).

In addition, critics have struggled with Joyce's mysterious and vague claim of "eight regular parts" of the fuga per canonem. The vagueness of the term "parts" and the claim of "eight regular parts" has engendered a critical free for all, and, over the past eight decades "all the eight parts" have been interpreted as voices, themes, or physical divisions. This reference has remained unresolved, however, and rightly so. Neither the fuga per canonem nor the fugue have a prescribed number of parts, divisions, voices, or themes, much less "eight."

A final critical conundrum concerns Joyce's competence with musical theory; did he comprehend the intricacies of the fugal form to have applied it in a way that is discernible, consistent, and/or meaningful? Even though Joyce could sing well and play the piano, accepted wisdom has it that Joyce couldn't read music or comprehend complex music theory. As Jurgen Grandt points out in his article "Might Be What You Like until You Hear the Words," "Joyce had limitations that would preclude a comprehensive understanding of the abstract elements of the fugue." Grandt adds that Otto Luening, who spoke at length with Joyce about contrapuntal and polyphonic music, "does not believe that the fugue serves a governing framework in 'Sirens'" (76).

Whether by accident or design, Joyce has left us clues in his own handwriting --- not all of which are legible --- which resolve these puzzles and reveal the process by which Joyce attempted, as Daniel Ferrer comments, to give "Sirens" "the intricate structure of one of the most forbidding musical forms" (63). Two previously "missing" drafts for "Sirens", found at the turn of the century in an apartment in Paris and now housed at the National Library of Ireland and specifically Joyce's notes on the cover of one draft, offer us critical clues to solving an 80-year-old mystery: what did Joyce mean by the fuga per canonem technic, and how was he, if at all, attempting to apply it in "Sirens" ?

On the inside cover of a copybook, catalogued as II.ii.3 in the new Joyce manuscripts and notes in the National Library of Ireland and which represents the missing first half of the "Sirens" draft, Buffalo MS V.A.5. (JJA 13: 32-56), Joyce has written the following numbered list of eight terms or phrases in Italian underneath which I have included his previously unpublished parenthetical notes:

Fuga per canonem



(reale in altro tono: in raccorciamento)

soggetto + contrasoggetto in contrapunto


(proposto --- codetta)


(nuovi rapporti fra detti: parecchio)


tela contrappuntistica


stretto maestrale

(blocchi d'armonia)

(mystery word)


(NLI MS 36,639/09, p. FCV)

Although those of us that have viewed the "list" as published in abbreviated form by Michael Groden and Daniel Ferrer in the 2001 JJQ, agree on a transcription of most of the terms, work remains to be done on one obstinate phrase --- labeled above as "mystery word" --- for which no one has yet claimed a definitive transcription or translation. Ferrer, whose advice and translations were invaluable in my project, suggests the following alternative transcription:

5) contra esposizione

(nuovi rapporti fra i detti: parecchi) (divertimenti)

stretto maestrale

(blocchi d'armonia)

(rovescii antesi)

On a broader scale, the new "Sirens" manuscripts, the fuga per canonem notes, and Joyce's source for these are critical keys as we connect the dots about why and how Ulysses underwent a radical change at this moment in its composition. As a comparison of the two newly discovered drafts reveal, the original plan for the episode did not include Bloom; he is missing from the first half of the earlier draft. Ferrer asserts in his article, "What Song the Sirens Sang . . . Is No Longer Beyond Conjecture," Joyce suddenly embarked on "a new departure in the development of Ulysses," the use of "counterpointed voices" with the insertion of Bloom and Bloom's interior monologue into the narrative (59-60).

It is after Joyce has investigated the fugue and cribbed these notes onto the cover of the II.ii.3 copybook that Joyce suddenly leaves behind the security of the initial style and commits to creating a new type of literature. Ferrer concludes, "The presence of these notes on the threshold of the second draft and the juxtaposition with the earlier version. . . . gives us the impression of observing at close range a crucial turning point in the history of Ulysses --- one could say in the history of literature" (63).

The discovery of these exciting new materials, however, is only the beginning of the scholarly odyssey. Yes, the fugal structure was more than a metaphor or bogus claim for Joyce. As Groden points out, "Here is his (Joyce's) indication to himself of a fugue's structure, which he apparently planned to superimpose onto an episode that was already partially drafted" (44). However, what Joyce meant by these eight terms and how he applied them is far from immediately clear. One stumbling block is that these eight parts are possible or potential parts of a fugue, not a fuga per canonem. Nor is it possible to simply, as many of us tried, to take these terms as a jumping off place for analyzing the radical new rewriting of "Sirens."

As my initial translation below indicates, several terms and phrases in the notes could not be comfortably and logically translated; nor do they represent known musical concepts:

fugue according to rule

1) subject

2) countersubject

(real [answer] in other key: in diminution)

3) subject and countersubject in counterpoint

4) exposition

(proposed or leading part --- coda or digression)

5) counter-exposition

(new relations among several designated)

(musical digressions)

6) (mystery word) contrapuntal

7) masterly stretto

(blocks of harmony)

(mystery word)

8) pedal

Although I can't pretend to have exhausted all possibilities, no contemporaneous standard musical dictionary or encyclopedia in English or Italian offered a definition or entry for "fuga per canonem," "tela contrappuntistica," or "stretto maestrale."

If we are to discover what Joyce thought he was doing when he altered his game plan for "Sirens" and changed literature forever, the solution, as those who work with Joyce's genetic materials know, is not to seek meaning from the notes themselves. We need to investigate the correspondence between the notes and their source. Although, as Grandt points out, Joyce may have discussed music extensively with Philip Jarnach and Otto Luening (76-77), historically, Joyce turned to a lay text or encyclopedia for his notes, not conversation. And, as I will describe, it was from the most popular musical reference text that Joyce cribbed his fugal notes.

Hopefully my findings proffered here will initiate a new critical debate about how to apply these materials and perhaps bring us closer to a critical consensus before another eighty years pass.

The entry for the fugue written by Ralph Vaughn Williams in the 1906 edition of Grove's Dictionary of Music and Musicians is the source for Joyce's fuga per canonem notes which appear in Italian on the cover of the new discovered "Sirens" copybook.

The identification of the source also clears up the primary obstacles which have prevented Joyce critics from arriving at a standard reading: the confusion between the fuga per canonem and the fugue (Joyce's error while skimming); Joyce's claim of "eight" parts (his own musically inaccurate but impressionistic invention abstracted from the source); and Joyce's mastery of complex fugal theory (bogus to none).

A correlation between the fuga per canonem notes and Williams's entry conform to Joyce's cribbing habits and explain Joyce's use of Italian, the specific terms he chose to crib, the obscure or untranslatable terminology, and his parenthetical notations.

The notes and source don't represent the intertexuality typical in Ulysses. Joyce's relationship with Grove's represents pure cribbing --- not the literary allusions that mark his textual exchange with existing philosophical, religious, aesthetic, and theosophical texts and theories.

Neither the notes nor the source can be fruitfully treated as prescriptive. Joyce, as was his pattern cribbing from esoteric sources, is often inaccurate, sloppy, incomplete, illogical, and impressionistic. Furthermore, even if he understands his own notes (which he often does not), his application of his notes is symbolic and metaphoric --- not according to rule. Thus, readers who expect to find in Joyce's texts exact fidelity to a source or concept often miss what is there. Because Joyce applied concepts as needed --- impressionistically, not accurately or rigidly --- the notes and the source give us new information to explore what Joyce thought he was doing.

Locating the second edition of Grove's the source of Joyce's fuga per canonem notes was an odyssey in itself. None of the books on music or musicians in either Joyce's Trieste or Paris libraries contain the terms in Joyce's list or conform to his note-taking habits. The one music text in Italian from Joyce's Trieste library, Manuale di Musica by Gustavo Magrini (1916), devotes only a page to the topic of the fugue, and fails to correspond to the notes. Based on Joyce's typical selection of sources for his notes on such subjects as math, physics, science, and theosophy, logic suggests it was a contemporaneous encyclopedia, reference book, or standard popular text, one he may have owned or found in the Zurich library. However, the standard dictionaries and encyclopedias in Italian and English to which Joyce would have had access in Trieste and Zurich yielded no pertinent entries. The fugue entry in the first edition of Grove's Dictionary of Music and Musicians, first published in 1878 (reprinted in 1890 and 1900) and written by William.S. Rockstro, was not Joyce's source. Nor were any of the late Nineteenth Century popular handbooks on the fugue such as those by Rockstro, Ebenezer Prout, and James Higgs.

However, a comparison between Joyce's notes on the fugue and the entry for "fugue" in the second edition of Grove's Encyclopedia of Music and Musicians published from 1904 to 1910, confirms that this edition of Grove's was Joyce's model for the fugal experiment in "Sirens." The relevant fugue entry, signed by Rostro's successor, R.V.W., or Ralph Vaughn Williams, a British composer popular since the turn of the century, appears in the second edition, Vol. II (1906), which was published by Macmillan Company, edited by J.A. Fuller Maitland, and reprinted in 1925. See the attached pages. In Williams's entry, Joyce's eight terms appear in sequential order, but, except for the first term, the originals are in English, not Italian.

A quick perusal of the source explains this. When describing the subject (italics Williams) of a fugue --- term one in Joyce's list --- Williams records in Italian one of the three types of subjects as a "'soggetti.' (See Soggetto)" (116). Attracted to the one term in Italian in the entry, Joyce then copied that term and translated the rest of his notes into Italian. This also explains the mystery phrase, "tela contrappuntistica." It is not the translation of a standard musical term at all, but Joyce's idiosyncratic translation into Italian of "contrapuntal web" --- a descriptive phrase used by Williams to explain the "middle section" of a fugue. In Italian tela de ragno means spiderweb or web of spider; however, the word "tela" alone does not mean web. This is Joyce's fanciful creation of a phrase in Italian.

That he often thought in Italian --- having lived in Trieste --- is reinforced by the fact that he translated English into Italian when taking notes long after he vacated an Italian speaking environment. As Aida Yared uncovered, when Joyce was taking notes for Finnegans Wake from Richard Burton's Arabian Nights he, on one occasion, read something in English and wrote the note in Italian, "suggesting," says Yared, "he was 'thinking' in Italian" (Yared). Stuart Gilbert expands on this Irish habit, saying, "To the Dubliners, music was essentially an Italian art, and they always liked to allude to songs by their Italian names even though the opera whence they came was by a non-Italian composer and usually sung in English" (248). Thus, the fugue entry in Grove's was "sung" by Williams, a non-Italian writer in English, but translated into Italian by Joyce.

The fact that the original eight terms are typographically set off by italics or quotations is another important indication that Williams is Joyce's source. As Phillip Herring observes in his work with Joyce's note-taking, "when in unfamiliar waters, Joyce skimmed." Yet, astonishingly, the notes Joyce "skimmed" from esoteric and complicated texts usually represent essential concepts --- not random jottings. That is also the case here. What we may call the Stephen Dedalus School of Speed Reading explains Joyce's method.

As I have explained in other studies, Joyce identified the key concepts from esoteric sources on math, physics, linguistics, and theosophy by jotting down words and phrases which the author has set off typographically: usually, italics, quotation marks, headings, bullets or lists, upper case, and/or bold font. Thus, Joyce's notes from complicated texts --- as is the case with his notes from Williams --- were pre-selected for him by his source. We're reminded of Stephen's self-effacing revelation, "Two pages apiece of seven books every night, eh? . . . Hurray for the Goddamned idiot" (40-41)!

Also relevant when verifying a source, such as the Williams's entry, are three other habits typical of Joycean note-taking: Joyce cribbed idiosyncratic turns of phrase, proper names, references to humans, and/or analogies using humans; the ordering of Joyce's notes closely parallels the sequential order in the source, but usually a few phrases or terms are not in sequence; and a small cluster of notes may come from one page, several consecutive pages, or various pages throughout a source.

See Figures One through Seven where these note-taking patterns are easily discernible in a correlation between Joyce's sources and Phillip Herring's transcriptions in Joyce's Notes and Early Drafts for Ulysses and Joyce's Ulysses Notesheets in the British Museum (see also Brown "the Geometric" 398-420). Figures two and three [Brown Figure Two; Brown Figure Three] illustrates Joyce extrapolating notes randomly and out of order from Betrand Russell's Introduction to Mathematical Philosophy from which he chooses words and phrases which either reference humans and/or are in italics or quotations; he would have found the father and brother metaphors --- particularly "the paternal grandfather is the square of the father" --- appealing. Figure Five [Brown Figure Five] illustrates Joyce cribbing headings, phrases in italics, and proper names from a three-page section of Poincaré's description of non-Euclidean Geometry in Science and Hypothesis. Figures One, Six and Seven [Brown Figure One; Brown Figure Six; Brown Figure Seven] records Joyce taking notes from Bloom's geometry text, The Short, but yet Plain Elements of Geometry by Gaston Pardies from which Joyce jotted down unusual wording or spelling, analogies and metaphors, references to humans, and words set off typographically by italics or upper case. Figure Four [Brown Figure Four] demonstrates Joyce's following the same pattern from Sir Thomas Heath's Euclid, vol. I; the impressionistically gathered notes reveal no real understanding on Joyce's part of the original material from a 12-page section of Euclid concerning the Fifth Postulate; he was cribbing for key terms pre-selected for him by Sir Thomas. In Figure Eight [Brown Figure Eight] Joyce follows the same technique; the references in Draft V.A. 8, an early draft of Cyclops (scene 2) transcribed by Herring, replicate the seven principles of "Esoteric Buddhism" --- the only text typographically segregated in A. P. Sinnett's The Growth of the Soul --- as Joyce attempts to place Cornelius Kelleher on the astral plane; the chances of Joyce having read and digested Sinnet's entire book seem unlikely when all he had to do to allude to the major concepts was copy them from the author's own list. Joyce's notes from Field's The Traveller's Hand-book for Gibraltar (Figure Nine) [Brown Figure Nine] follow this pattern as proper names and/or words in italics are lifted out of order from several consecutive pages. As Figure Ten [Brown Figure Ten] confirms, Joyce took one set of late notes for "Penelope" by pulling from a typographically isolated table in Sir William Robertson's From Private to Field Marshal.

In a word, Joyce the polymath was Joyce the fraud.

With Joyce's note-taking habits as a guide, a correlation between the fuga per canonem notes and Williams's entry resolves the long-standing critical controversies over the terms fuga per canonem and fugue and over Joyce's claim of "eight regular parts" for the fugal structure of "Sirens." First, a comparison of the notes with the source makes clear why and how Joyce confused the fuga per canonem and the classical fugue, thus confusing Joyce scholars for eighty years. In a word, he fell victim to his own speed reading methods and failed to realize that the entry from which he was cribbing describes the possible parts of the fugue, not the fuga per canonem. On the first page of his entry, Williams explains that, in the Sixteenth Century, the term fugue "meant a movement in canonic form; indeed the name 'canon' is merely short for 'fuga per canonem,' a fugue according to rule" (114). [114groves] Joyce lifted the term fuga per canonem, which is set off by quotation marks; copied it as his heading; and then skimmed ahead to the first paragraph containing an italicized word, missing the critical fact that the entry addresses the fugue, not the fuga per canonem. By skipping forward looking for words set off typographically, Joyce skimmed over Williams's crucial comment: "From the time of Bach the word 'fugue' has connoted a very definite musical form which will now be described." Joyce's confusion is even more understandable when we see that Williams opens the paragraph in which the italicized word "subject" appears saying: "We can now proceed to a detailed description of the fugal form" (115). [115groves]

Moreover, although it is now clear that "the eight regular parts of the fuga per canonem" refers to consecutive movements or sections, not eight voices or themes, Joyce draws an inaccurate conclusion from his skimming of the source, one which has confused critics for decades. Williams makes no such claim of "eight parts" or any "regular parts." In fact, Williams consistently qualifies the existence of any of these movements --- particularly the counter-subject, counter-exposition, and stretto maestrale --- as choices which "sometimes" are employed by the composers of fugues. "All the eight regular parts" of the fugal structure is Joyce's idiosyncratic invention, a structure which he abstracted subjectively from the source.

A comparison of Joyce's notes with his source also resolves some of the puzzling terms on his list. In the section of Williams's entry which introduces the second term in Joyce's list, the countersubject, two parenthetical phrases from Joyce's cover notes appear and are thus explained. "Reale in altro tono" is a translation in Italian of the passage from Grove's: "The answer may be a repetition of the subject in a different key. This is called a real answer" (116). [116groves] Joyce's note on the "answer" is elliptical, however; with Williams as a guide, we see that "reale in altro tono" means "real [answer] in another key." On the next page, Williams tells us that the "answer" is occasionally in "diminution" which Joyce translated into Italian as "in raccorciamento" (117). [117groves] Following this passage, the term "counter-subject" appears in italics and Joyce translates it as "contrasoggetti." We next observe that Joyce abstracts part three of his eight parts, "soggetto and contrasoggetto in contrapunto," from a single sentence in Williams: "The countersubject is usually in double counterpoint with the subject, designed, that is, to appear either above or below it as occasion requires" (117). [117groves]

Joyce lists part four as the "esposizione" (exposition) with the terms "(proposto --- codetta)" in parenthesis. On page 117, "codetta" appears in italics and in the next paragraph the term "exposition" is also printed in italics. The source for proposto or propose may be the passage: "When the subject and answer have been thus propounded, the other voices enter in turn with subject or answer alternately" (118). [118groves] In music terminology, proposto may also mean "leading part" or subject. In the discussion of the exposition, Williams comments fancifully in a way that would have caught Joyce's attention and perhaps inspired his note: "Now it is necessary, before the subject as the hero of the plot, sets out on its career of adven­ture, that its nature and characteristics should be thoroughly impressed on the attention" (117). [117groves]

In the paragraph introducing the "counter-exposition," Williams describes the "contraesposizione" as "a whole series of extra entries" which may be the source of Joyce's parenthetical note "divertimenti" under "contraesposizione." Likewise, Williams says the counter-exposition "presents the subject in a new aspect with regard to the counter-subject" (118) [118groves], a passage which arguably is the origin of Joyce's parenthetical note "(nuovi rapporti fra detti: parecchio)," which can be loosely translated, "new relations among several designated. . . "

As discussed earlier, the next "part" in Joyce's list of "eight regular parts," "tela contrappuntistica, has been creatively translated from the descriptive phrase, "contrapuntal web." Joyce in essence invents an Italian term for "a series of episodes . . . interspersed with entries of the subject in various new situations and guises," a second phrase from the source which may have instead been the origin of the note, "nuovi rapporti fra detti: parecchio."

On the next page we encounter the seventh "part" on Joyce's list, "stretto maestrale," the term italicized and defined thus in Grove's: "A stretto in which all the voices take part, and in which each voice takes up subject or answer in turn in their entirety and without any modification is called a masterly stretto or 'stretto maestrale'" (119). [119groves] In the sentence following the italicized word "pedal," the eighth term on Joyce's list, Williams claims, "in many cases, right at the close, the contrapuntal texture gives way to massive blocks of harmony" (120), the phrase which Joyce translates into "blocchi d'armonia" and places out of order beneath the preceding notation, "stretto maestrale," rather than under "pedale."

These final notes on the list signal a needed caution. Similar to the lack of sequential order of Joyce's notes from Russell or Heath, Joyce's parenthetical notes from Grove's do not always correspond to their order in the source. In a second example, Joyce places the parenthetical note "codetta" under "esposizione" but the term actually appears in the Grove's entry before the term "exposition" is introduced (117-18).

A second caution concerns Joyce's lack of accuracy and depth in his use of this source. The fact that he missed the distinction between the fugue and fuga per canonem illustrates the need for this concern. Just as he did with his inaccurate and incomplete cribbing and application of terms and concepts from math texts, he jotted down terms and phrases from the Williams entry that --- by virtue of typographical distinction, not because he understood and abstracted the essence of the text --- are the key concepts in the composition of a fugue. Terms also caught his eye because they were in Italian, referenced humans, or represented interesting turns of phrase, not because he understood their complex musical meanings. This replicates Joyce's notes from science and other "less familiar waters" where the allusion serves a stylistic or thematic purpose, but Joyce's understanding of the original text is minimal. For example, we have his use of Euclid's Fifth Postulate as a metaphor in "Ithaca" where his geometry is inaccurate but his reference to non-Euclidian Geometry is thematically crucial (Brown "The Geometry" 174-92; Wilcox 643-49). As he did with many complicated sources, Joyce was using his source as an inspiration --- not a template --- for the structure, rhythms, metaphors, and themes in his musical experiment, "Sirens."

Many provocative passages, like snapshots, also appear throughout the pages of Grove's, further tying this source to Joyce's notes and suggesting fodder for critical exploration. For example, Joyce's Wagnerian goals for "Sirens" correspond to a reference in Williams. In the conversation quoted by Georges Borach, Joyce associates the writing of "Sirens," the fugue form, and Wagner's 'Die Meistersinger' when he says, "A quintet occurs in it ["Sirens"], too, as in the Meistersinger, my favorite Wagner opera" (327). In the Grove's entry, Williams comments, "the prelude to Wagner's 'Die Meistersinger' is a well-known instance of a movement where three subjects are at first presented separately and harmonically as in a sonata, and afterwards combined as if in the stretto of a triple fugue" (120). The mention of this particular opera in connection to fugal form by both Williams and Joyce seems far from coincidental.

Back to the big picture. How, after all these decades, can the sudden appearance of Joyce's notes on a previously unknown draft of "Sirens" and the correlation of those notes with the source bring us to a better understanding of the thematic content of the "Sirens" episode and Joyce's aesthetic goals for this chapter and for the whole novel? Unfortunately, the new notes and their source bring us no closer to definitive answers.

For example, the new materials make it clear Joyce intended a "subject," an "answer," and a "countersubject" as parts in the fugal structure of "Sirens." But the usefulness of the new materials stops there. Is the subject love and the countersubject betrayal? Is the subject Bronze and the answer Gold? Or is the subject a multiple subject with several voices as described by Grove's, so that the sirens are the subject and both Bloom and Blazes represent the answer? Are we to see Bloom as the subject and Blazes as the answer? Or women as the subject and men as the answer? And where does Molly fit in? Since Joyce jotted the note "(reale [answer] in altro tono: in raccorciamento)" under "contrasoggetto", has he confused the answer with the countersubject? Easily done in this complicated entry about a complicated musical form. If so, is the answer/countersubject a "repetition of the subject in a different key"?

While a correlation of Joyce's notes and his source fails to illuminate Joyce's intentions, Joyce may have passed on to his future readers some guidance for interpreting his fugal notes for "Sirens" through Stuart Gilbert's 1930's study, which Joyce oversaw. Immediately arresting is that Gilbert italicizes many of the same terms which are italicized in Williams, the very terms Joyce cribbed. For example, Gilbert tells us: "The Subject is obviously the Sirens Song; the Answer, Bloom's entry and monologue; Boylan is the Counter-Subject" (253; italics Gilbert's). He also tells us: "The Episodes or Divertimenti are the songs by Mr Dedalus and Ben Dollard. The Episodes, Subject, Answer and Counter-Subject are often bound together con­trapuntally in the narrative or in the texture of Mr. Bloom's monologue" (253). If we follow Gilbert, Joyce intended the subject as female seduction; the answer as Bloom in the despondent tone of the betrayed; the countersubject as Boylan, the cocky male version of the seducer; and the episodes and divertimenti as the songs actually sung at the Ormond.

We have good reason to trust Gilbert's explication of "Sirens." Gilbert claims to reproduce "word for word," information given him "by Joyce," specifically for this chapter (viii-ix). In fact, in light of Gilbert's claim that his study was "endorsed by Joyce" and contains nothing "that did not receive his full approbation," it seems possible that Joyce gave Gilbert access to the fuga per canonem notes (as he gave him the schemata for the whole book).

The fact that there are many parallels in wording between Gilbert and Williams further hints that Joyce --- as Gilbert claims Joyce often did --- shared with Gilbert his source (see "Preface"). For example, Williams heralds the "contrapuntal web" marking the middle section of the fugue as the first example of "serious modulation" in the fugue where "the composer is usually said to be 'free' to proceed as he likes" (120). Gilbert reflects the wording and content of Williams's entry as he describes "the process of modulation" which he then defines with examples from Ulysses as something akin to "free association" (254). Likewise, Gilbert's discussion of Joyce's stretto mirrors Williams: "a stretto in which all the voices take part" is often reserved for the climax of the fugue (Williams, 119-120). Gilbert tells us, "The bringing together of all the persons at the bar, under abridged names, and of fragments of thematic material from preceding pages, in a swift simultaneous clink of libation is clearly the stretto of the fugue" and he quotes the famous passage, "Near bronze . . . Dollard" which appears just before the final paragraphs of "Sirens" (Gilbert, 256; U 278). In addition, in speaking of the close of "Sirens," Gilbert reprises Williams' comments about the close of a fugue. Says Williams: "the natural desire for home, but home in a new aspect" is the "climax of the fugue" and "is usually heralded by a return to the original key" with the "pedal on the domi­nant and sometimes also on the tonic. In many cases, right at the close, the contrapuntal texture gives way to massive blocks of harmony" (120). Gilbert echoes Joyce's source when he says about the end of "Sirens," "After the dominant comes inevitably the return," and he quotes as his example from "Sirens," the "climax of Martha" which is "rendered in almost technical terms: "dominant to love to return with deepening yet rising chords of harmony" (257). Finally, Williams, in a second reference drawing on the "web" metaphor, cites the "leit-motif of Wagner's musical dramas" and describes "a 'leit-motif' imposed on the polyphonic web of Wagner's music" (121). This parallels Stuart Gilbert's quoting well-known scholar and critic Professor Curtius, who claims, "The literary technic" in "Sirens" "is an exact transposition of the musical treatment of the leitmotif, the Wagnerian method" (243).

I believe that Gilbert's study of "Sirens," because he claims to have taken it directly from Joyce while he completed the French translation of Ulysses, is a logical and useful reference for connecting the new drafts, the fuga per canonem notes, the fugue entry in Grove's, and the final version of "Sirens."

What I have attempted to provide is a set of findings and suggestions upon which scholars and lovers of Joyce will go forth and explicate. Hopefully, with these new keys --- Joyce's own notes and the source of these notes --- we will better understand Joyce's goals and observe the process by which he attempted to work them out in "Sirens." Moreover, these new genetic and source materials may also inspire critics to realize why and how Ulysses underwent an unexpected shift in content and style at this moment of its composition and to explore this radicalization of literature that changed the writing of fiction.

Joyce would have warmed to a comment by Williams which he also understood well from his conversations with Otto Luening about the complexity of contrapuntal techniques; says Williams: "But some subjects are so difficult to manage that nothing but the insight of genius can make the connection between the two sufficiently obvious to ensure recognition" (123). Joyce, who surely felt he was in competition with Shakespeare and Dante for immortality, would be inspired by this prompt.

Yes, Mr. Joyce is smiling somewhere. If only one page of hundreds among the new manuscripts inspired this critical inquiry, he will be keeping scholars busy for an extra hundred years.


Bowen, Zack. Musical Allusions in the Works of James Joyce. Albany: U of NY Press, 1974.

Borach, Georges. "Conversations with James Joyce." Trans. Joseph Prescott. College English 15.6( 1954): 325-27).

Brown, Susan Sutliff. The Geometry of James Joyce's Ulysses: From Pythagoras to Poincare. Diss. U of South Florida, 1987. Ann Arbor: UMI, 1987.

____. "The Joyce Brothers in Drag: Fraternal Incest in Ulysses." Gender in Joyce. Eds. Marlena Cocoran and Jolanta Wawczkya.

Cole, David W. "Fugal Structure in the Sirens Episode of Ulysses." Modern Fiction Studies 19 ( ): 221-226.

Ferrer, Daniel. "What Sons the Sirens Sang. . . Is No Longer Beyond All Conjecture." JJQ 39.1 (2001): 53-68.

Gilbert, Stuart. James Joyce's Ulysses. New York: Vintage Books, 1955.

Groden, Michael. "The National Library of Ireland's New Joyce Manuscripts: A Statement and Document Descriptions." JJQ 39.1 (2001): 29-52.

Herring, Phillip F., ed. Joyce's Notes and Early Drafts for Ulysses. Charlottesville: UP of Virginia, 1977.

____, ed. Joyce's Ulysses Notesheets in the British Museum. Charlottesville: UP of Virginia, 1972.

Honton, Margaret. "Thou Lost One: All Songs on that Theme in 'Sirens.'" JJQ 17 (1979): 41-48.

Kidd, John. "Arguments for a Facsimile of the First Printing of Ulysses, Revised to Be a Corrected Reprint of 1922 Ulysses." Quoted in The Panizzi Lectures. Ed. McKenzie. London: The British Library, 1986.

Knowles, Sebastian D. G. Dublin Helix. Gainesville: U of Florida P, 2001.

Levin, Lawrence. "The Sirens Episode as Music: Joyce's Experiment in Prose

Polyphony." JJQ 3 (1965): 12-24.

Williams, Ralph Vaughn. "Fugue." Groves Dictionary of Music and Musicians.

Vol. II. Ed. J.A. Fuller Maitland. Boston: The Macmillan Company, 1906.