GENETIC JOYCE STUDIES - Issue 14 (Spring 2014)


A Notion of Joyce’s Time:

Interpreting the Diacritics of Finnegans Wake 124.8–12

Tekla Mecsnóber[1]


This article contains many diacritical signs, which may not appear correctly in each browser. For a precise rendition of the diacritical signs, please read the pdf version.


Introduction: the crux

There is a passage in the description of the mysterious letter of book I chapter 5 of Finnegans Wake that seems to have been as much of a challenge for scholars as it was for printers:


The unmistaken identity of the persons in the Tiberiast duplex came to light in the most devious of ways. The original document was in what is known as Hanno O’Nonhanno’s unbrookable script, that is to say, it showed no signs of punctuation of any sort. Yet on holding the verso against a lit rush this new book of Morses responded most remarkably to the silent query of our world’s oldest light and its recto let out the piquant fact that it was but pierced butnot punctured (in the university sense of the term) by numerous stabs and foliated gashes made by a pronged instrument. These paper wounds, four in type, were gradually and correctly understood to mean stop, please stop, do please stop, and O do please stop respectively, and following up their one true clue, the circumflexuous wall of a singleminded men’s asylum, accentuated by bi tso fb rok engl a ssan dspl itch ina, — Yard inquiries pointed out → that they ad bîn “provoked” ay fork, of à grave Brofèsor; àth é’s Brèak — fast — table;   ; acùtely profèššionally piquéd, to=introdùce a notion of time [ùpon à plane (?) sù ’ ’ fàc’e’] by pùnct ! ingh oles (sic) in iSpace?! (FW 123.30–124.12)[2]


Although Joyce’s exploration of a range of punctuation symbols in this passage has been quite widely recognised,[3] little detailed commentary has been made on the use of diacritics. Among the exceptions is Bernard Benstock, who suggested that in this passage “even the most insignificant bit of type, minuscule accent marks that change the way a letter sounds but have no voices of their own, are given an opportunity to speak”, as if “to justify Leopold Bloom’s assumption that ‘Everything speaks in its own way’” (U 7.177).[4]

Michael Kaufmann saw an instance of his concept of the “literal” or “typographical enactment” of content in the passage, and also pointed out the phonetic potential of the diacritics: “Print breaks apart like the split china and glass, imitates voice phonetically and accentually, introducing “time” onto a plane/plain surface. The signs literally point—and name themselves (‘ fork,’ ‘à grave,’ ‘acùtely profèššionally’). They look like they mean.”[5] Although Kaufmann’s phonetic readings are not entirely incontestable on the basis of the text he quotes (the marking of “à grave” and “ùpon à plane” is rather counter-intuitive and does not follow any established phonetic notation, and “acùtely” has a grave, rather than acute accent in the text he quotes), his remarks remain among the most serious and useful explicit efforts to make sense of these diacritics.

Erik Bindervoet and Robbert-Jan Henkes added a further valuable interpretative layer by making use of earlier textual variants and providing a reading of Joyce’s original diacritics in the appendix to their Dutch version of Finnegans Wake.[6] (Similar restored readings are also given in the appendix of the 2012 Oxford World’s Classics Finnegans Wake edition by Robbert-Jan Henkes, Erik Bindervoet and Finn Fordham, and in the main text of Danis Rose and John O’Hanlon’s 2010 Houyhnhnm Press and 2012 Penguin edition.[7]) Henkes and Bindervoet’s “Dutch” text extends the typographic multiplicity of the passage by adding extra diacritics as well as mathematical and other symbols, italics and small caps, and thus clearly stresses the importance of these marks, although not in a – phonetically or otherwise – clearly motivated way (p. 124). In what follows, I shall attempt to add to the existing interpretative lore by providing a genetically informed contextualized reading of Joyce’s use of diacritics in Finnegans Wake 124.8–12.[8]


Punctuation, diacritics, women: the history and sociology of writing

On the surface, the text discusses the significance of certain holes in the sheet of the “mamafesta” (or motherly manifesto, FW 104.04) in helping us to identify the document’s writer. These “foliated gashes” are quite clearly associated with bodily aggression (“stabs,” “wounds”), (male) sexual activities (“the piquant fact that it was but pierced butnot punctured”) and possibly with transgression (“stop, please stop, do please stop, and O do please stop”). Equally clear is the link with punctuation. While the original text was written in an unbroken or “unbrookable” text (like the scriptio continua of early Greek and Latin manuscripts) and “showed no signs of punctuation of any sort,” the subsequent piercing introduced the four classic punctuation marks (the comma, the semicolon, the colon and the full stop) that indicate gradually increasing lengths of pauses.[9] Two of the perspectives suggested here – masculinity and punctuation – had been associated in Ulysses already, albeit in a negative way: Joyce banned punctuation marks from the closing episode of the book specifically to stress that this last interior monologue belonged not to an orderly male mind but the sort of formless primeval female matter that he described to Frank Budgen in August 1921 as a “perfectly sane full amoral fertilisable untrustworthy engaging shrewd limited prudent indifferent Weib [woman]” (SL 285).

As mentioned earlier, it has less often been noted that many of the Wakean “paper wounds” serve not as punctuation, but as diacritics. This is evident enough both from verbal allusions (like “circumflexuous,” “accentuated,” “grave,” or “acutely”) and from the profusion of diacritics at the end. In addition, as part of the general discussion of the development of various scripts and alphabets from their “early muddy terranean origin” (FW 120.29), the text also refers to the history of Hebrew diacritics. When it calls the “mamafesta” the “Tiberiast duplex” (FW 123.30–31), it evokes the scholars of Tiberias who worked out the characteristic system of diacritics that from the seventh century onwards served to indicate the canonical pronunciation and chanting of Hebrew biblical texts.[10] Moreover, the “unbrookable script” suggests not only the ancient scriptio continua, but also a not yet fully developed Gaelic typeface (like the one used in Charlotte Brooke’s seminal 1789 collection, Reliques of Irish Poetry, and therefore later called the Brooke type) and a not quite “broken” blackletter type (the German word Fraktur derives from Latin fractus, “broken”) – both of which also use dots (“punctuation”) as diacritics. It is also useful to remember that in July 1924, as if to stress the same link between punctuation and diacritics, Joyce had extended the Ulyssean association of punctuation with the male mind to diacritics: he insisted that Valery Larbaud’s French translation of the “Penelope” episode should be printed not only without punctuation marks, like the original English version, but also without the accents required by normal French orthography (Letters III, p. 99).[11]


Diacritics on page 124 of the 1939 Finnegans Wake

Although it appears to be clear enough that this passage is concerned with the phenomenon of diacritics at several levels, it is difficult to make much more of it on the basis of the text as it appeared in the 1939 first edition of Finnegans Wake and as it has been usually reprinted since then. The 1939 text quoted above contains 16 instances of only four diacritics: several grave accents (as in Brofèsor or, seemingly perversely, in acùtely), two acutes (in ath é’s and in the French-based piquéd), one circumflex (bîn), and two carons (both in profèššionally).[12] Although these diacritics add a strikingly alien(ating), non-English, international touch to the text,[13] there seems to be no language that would require exactly this set of diacritics and could serve as a frame of reference. Carons or háčeks ( ˇ ) are typical of Slavic and Baltic languages. The combination of grave, acute and circumflex accents is familiar from both classic polytonic Greek and French, but grave accents are also used in many other languages (including Italian and Scottish Gaelic), just like acutes (in Irish Gaelic, Hungarian, Czech and Slovak), and the circumflex (in Esperanto, modern Turkish and Romance languages like Italian and Romanian). Moreover, while the value of the sound indicated by the character š is roughly the same here as it would be in Slavic languages, the use of accents does not seem to follow any linguistic logic at all: the vowels represented by the grapheme ù in acùtely and ùpon are radically different.


Diacritics in the “first draft” version (1923/24) and the manuscript addition (c. 1925)

A comparison with the manuscripts, typescripts and proofs of this passage, however, suggests that originally Joyce appears to have intended a much more complex – and, I shall suggest, more consistent – use of diacritics. This complexity seems to have suffered greatly from several subsequent transcriptions and publications.

This is what the “first-draft” version that David Hayman reconstructed and dated to December 1923 or January 1924 looks like:


The original document was what is known as unbreakable script tracery, that is to say, it had no signs of punctuation of any kind. On holding it to the light it was seen to be pierced or punctuated (in the university sense of the word) by numerous dots cuts and gashes inflicted made by a pronged instrument. These paper wounds, four in type, were gradually understood to mean stop, please stop, do please stop, and O do please stop respectively and investigation showed that they were provoked by the fork of a professor at the breakfast table professionally trying piqued to introduce tempo into [a plane] surface by making holes in space.[14]


While clearly occupied with punctuation, this passage contains neither terminological references to diacritics nor any use of them. These seem to have been inserted into the text from a handwritten note dating probably from March 1925 or somewhat earlier when Joyce reworked this passage for publication in the July 1925 number of T. S. Eliot’s influential literary magazine The Criterion (1922–1939).[15] It is in this note that we first encounter an array of – unfortunately not always very legible – diacritics as well as references to grave, acute and circumflex accents. My reading of the additional passage inserted after “These paper wounds, four in type, were gradually understood to mean stop, please stop, do please stop, and O do please stop respectively and” to replace the words “investigation showed that they were provoked by the fork of a professor at the breakfast table professionally trying piqued to introduce tempo into [a plane] surface by making holes in space” is as follows:


, following up their one true clue, the circumflexuous wall of a singleminded men’s asylum, accentuated by bi tso fb rok engl a ssan dspl itch ina, — Yard enquiries pointed out → that the͡y ad̔ ad̓ bîn “provòked” by fork, ŏf ă grave Profès̄s̸or; ăth é’s Brèak – fast – table ; ; acùtely profèššionally piquéd, to = introdụ̀çe ă nòti͡ön ŏf ti̅m̅e̅ [ŭpòn ă plāñe (?) sụ̀ ’’rfáç’e’] by pụ̀nc͞t ! ingh o̓les (sic) in iSpåce ? ! (JJA 46: 355–6)


As compared to the 16 instances of diacritics in the 1939 version, this manuscript note has about 20 more. These include aspiration symbols of the Greek type (the smooth breathing or spiritus lenis,  ̓, as in a̓d and o̓les, and, possibly – but not in my reading – the rough breathing or spiritus asper,  ̔ ), the macron ( ˉ , as on the first s in Profès̄s̸or or in plāñe), the tilde (   ̃ , as in plāñe), the umlaut ( ¨ , as in noti͡ön), the ring above ( ˚ , as in iSpåce), the breve ( ˘ , as in ŏf ă grave and ŭpòn ă, only in my reading), the dot below (or underdot,   ̣ , as in introdụ̀çe and pụ̀nc͞t), the cedilla (introdụ̀çe, sụ̀ ’’rfáç’e’), the slash (  ̸ through the second s in Profès̄s̸or) as well as the type of tie that is shaped like a double inverted breve (  ͡  , linking the last two letters in the͡y and two or three letters in noti͡ön) and the continuous line (overbar) covering (at least) the last three letters of the word ti̅m̅e̅ and, possibly, the last two letters of pụ̀nc͞t.[16]


Interpreting the diacritics

Not surprisingly, there was no orthography in 1925 that would have required exactly this set of diacritics either.[17] The aspiration signs are typical of ancient Greek, the macron is and was used, for instance, in Latin dictionaries and textbooks, Latvian, and the Romanized transcription of Sanskrit, the breve in (textbook) Latin, Romanian, Esperanto and several Cyrillic alphabets, the tilde in Romance languages like Spanish and Portuguese (as well as alphabets influenced by them), the umlaut in German, Scandinavian languages, Hungarian and Slovak, the ring above in Scandinavian languages (å) and Czech (ů), the dot above in Irish Gaelic (especially when written in the Gaelic script), Polish (ż) and Hebrew (when written with Hebrew characters), the dot below most famously in Hebrew (and in Romanized alphabets of several languages outside Europe), the cedilla in Romance languages (ç), and the slash or stroke through most obviously in the Scandinavian ø and in a number of currency symbols (like £, $ or ¢).[18] The tie is and was most often seen in phonetic transcriptions and respellings to indicate diphthongs, triphthongs, affricates or double articulation. A macron or an overline also occurs in Latin prosody to mark long syllables and in certain phonetic notations to mark long vowels and special pronunciations of consonant letters. An overline was also used in the medieval scribal tradition to indicate abbreviations, at first especially those of sacred names.[19] (From this perspective, the Joycean overbar above time is a fine touch, possibly suggesting both the long pronunciation of the vowel letter i and a scribal abbreviation of “time” as a nomen sacrum in the age of Bergson and Einstein.)

These last few examples may in fact hold the most likely clue to what in the 1939 version looks like an entirely gratuitous spattering of diacritics. Joyce’s original manuscript version strongly suggests that Michael Kaufmann’s insight about the phonetic nature of this passage was fundamentally correct: the grave “Profès̄s̸or” is indeed trying to “introduce a notion of time upon a plane surface” by making phonetic marks which indicate the precise pronunciation – in time – of the letters written on the surface – in space – of the sheet of paper. Thus, the diacritics of this passage are indeed not only sexual, but also “sexophonologistic” (FW 123.18): they mark both the gender of their maker and the pronunciation intended by him.

In particular, the text appears to use grave accents in order to indicate primary stress, the acute accent perhaps for a partial reduction to [ɪ],[20] the breve – I suggest – for vowels reduced to the schwa [ə], the macron (as well as the ring and circumflex) for long vowels, the aspiration sign (although the wrong one) for the [h] sound, carons for the palatalized ([ʃ]) pronunciation of s, the ö for a reduced schwa-like vowel, the cedilla for the fricative ([s]) pronunciation of c, the tie for diphthongs and contractions, the overbar perhaps for double articulation or etymological reconstruction (in pụ̀nc͞t for ‘punch’). Also, when Joyce replaced the initial P of (what by then had become) Profès̄or by a doubly underlined B around June 1927 (JJA 46: 427–8 and 419), he may also have evoked the phonetic convention in which (small) capitals stand for the devoiced (and thus also “German”) pronunciation of voiced consonants.[21]

Predictably, Joyce’s use of diacritics is by no means completely consistent (length, for instance, is indicated by three different diacritics), but that was also the case, as we shall see, with various phonetic notations that he knew. While the function of some symbols remains unmotivated, it seems clear that with many of the diacritics Joyce used in this passage, he at least partially meant to give the effect of phonetic marking. Although some of his markings are idiosyncratic (gratuitous, tongue-in-cheek, or “faulty”), the international nature of these diacritics inevitably evokes other internationally inspired phonetic marking systems of the late nineteenth and early twentieth centuries.


Phonetic notations in Joyce’s reading

Joyce would have known several such marking systems. He was a user of the first edition of A New English Dictionary on Historical Principles (NED, later known as The Oxford English Dictionary) as it was published in fascicles making up 10 volumes between 1884 and 1928. In harmony with my reading of Joyce’s “mamafesta” passage, the New English Dictionary used the breve to mark reduced (“obscure”) unstressed vowels and no diacritic at all for short stressed vowels. The macron indicated long vowels, with values based on “European” (e.g. Italian) vowels, so that been was respelled as bīn (cf. Joyce’s bîn).[22] Although adapted from the NED, the Concise Oxford Dictionary of Current English (1911) used the same diacritics rather differently: the breve marked short stressed vowels (rather than unstressed reduced ones), while the macron marked the long pronunciation of vowel letters based on the English readings (so that no respelling, only a macron was needed in tīme – as was the case in Joyce’s text as well).[23] Elocution manuals like Bell’s Standard Elocutionist by David Charles Bell and Alexander Melville Bell, of which Joyce seems to have had the 1892 edition, [24] had their own systems: while the breve indicates short vowels and the macron long ones for the Bells as well, the non-identical vowels of pŭll and ŭp get the same symbol, while the a with macron is used for the uncharacteristic value (both in English and continental terms) represented by the letter in āll.[25]

The case of American dictionaries was still more confusing, as Joyce must have known if he did indeed read one of the books that he asked Sylvia Beach to order in July 1923, only four months after he began working on the Wake, namely Frank H. Vizetelly’s Essentials of English Speech and Literature (1915) (Letters III, 77-8). In the section rather misleadingly called “Phonetics and Pronunciation”, Joyce could here read a lengthy exposition of the ills of various phonetic notation systems using “chaotic aggregations of dots and dashes above and below the letters, together with curves and curlicues”,[26] including those of the Oxford English Dictionary and Webster’s American dictionaries. Some of Vizetelly’s examples – as well as perhaps his repeated derisive remarks on diacritic “dots and dashes” – could be directly relevant to Joyce’s “dots cuts and gashes” passage. Thus, in addition to some rather regular uses of the macron, breve and umlaut, Vizetelly cites uses of the i with circumflex (î) to denote the sound of i in marine (as in Joyce’s bîn for ‘been’), the u with one dot under to denote the sound of u in wulf, with two dots under for the sound of u in rude (cf. Joyce’s introdụ̀çe), the c with cedilla for the sound of the first c in civic (as in Joyce’s introdụ̀çe and sụ̀ ’’rfáç’e’), the double inverted breve for diphthongs as in oil (cf. Joyce’s the͡y).[27] If he followed up on Vizetelly’s lead, Joyce may have found out that at least certain editions of Webster’s American Dictionary employed an s with a slash through (1845) or an s with a bar under (1865) to denote the sound of s in environs and has (cf. Joyce’s Profès̄s̸or).[28] This last example may also owe something to the confusion of voiced and voiceless consonants that Joyce’s Professor, who became at least partly German(ic) in later versions as a Brofès̄or and a Brotfressor (FW 124.25, cf. German Brotfresser, “bread-devourer”), could have been prone to.[29]

While Joyce may have been inspired by Vizetelly’s Essentials of English Speech and Literature to pay extra attention to the “disorderly Babel of sound-symbols,”[30] the Wakean passage under discussion also clearly relies on other systems of phonetic notation. Another  potential candidate is the notation of the International Phonetic Association (IPA). Joyce certainly knew this from sight at least partially, since this was the system used, “to bring out clearly the insufficiency of the ordinary spelling”, in Otto Jespersen’s Language: Its Nature, Development, and Origin (1922).[31] (This was a book that Joyce read in the late summer of 1923, roughly at the same time when he also showed interest in Vizetelly’s work.[32]) Joyce may even have taken the symbol in “by fork” from Jespersen’s book, which employs a similarly sans serif version ([ʌ]) of the inverted v symbol of the IPA and defines its phonetic value as similar to what the unstressed article a would have in by a fork.[33]

While the IPA alphabet won Vizetelly’s approval by endeavouring to reduce the use of “dots and dashes” through the introduction of separate symbols (often inverted letters) for the most important sound values, other systems relied rather heavily on diacritics. Such a notation is used, among other contemporary books,[34] in Antoine Meillet and Marcel Cohen’s Les Langues du Monde (1924), a work Joyce certainly knew and used in 1937.[35] In fact, Joyce’s use of diacritics in the 1925 addition bears a remarkable similarity to the notation of this volume. In particular, besides the conventional use of the macron for long vowels and the breve for reduced ones, this book employs the spiritus asper for aspiration (cf. Joyce’s “smooth” a̓d and o̓les), (small) capitals for devoiced consonants (cf. Joyce’s Brofès̄or), the s with caron for the sound corresponding to the English sh (as in Joyce’s profèššionally), the acute accent for primary, the grave accent for secondary stress (cf. Joyce’s loosely inverted use in profèššionally and é’s) and the o with umlaut for a reduced schwa-like vowel (“e muet” as in the French “prenez,” cf. Joyce’s noti͡ön). Meillet and Cohen’s work also uses three further diacritics that are present in Joyce’s passage, although with rather less similar values: the underdot for a particularly “closed” articulation (or “fermeture”, cf. Joyce’s gratuitous – or Frenchified – sụ̀ ’’rfáç’e’), the n with tilde for the palatalised nasal sound corresponding to the French gn (cf. Joyce’s plāñe), the a with a ring (å) for a “vowel between a and o” (cf. Joyce’s iSpåce).[36]



Phonetic rationales

To recover with certainty the exact sources of Joyce’s use of diacritics in the 1925 addition to the “mamafesta” chapter may not be possible, but it might be also not quite necessary. What seems more important is that Joyce appears to have been aware of several systems of phonetic notation, and made a point of incorporating the phenomenon in his Work in Progress. Moreover, as the question of internationally applicable phonetic notations was very much of a current issue, and drew plenty of commentary in the kinds of linguistic works that Joyce seems to have explored in the early 1920s, it is most likely that he was also to some extent familiar with the arguments regarding the aims and best forms of these notations.

Essentials of English Speech and Literature would, for instance, have informed Joyce about the uses of the American “National Education Association Alphabet” that Vizetelly advocated. These uses included the teaching of reading to children, a possible spelling reform, the correction of non-standard pronunciation, and the codification of a good American pronunciation – presumably also for the benefit of the masses of non-English speaking “international” immigrants reaching the country at the time. A similar case could be made for the New English Dictionary, whose obvious orientation to native speakers of “English” is paired with a colonial and international orientation, betrayed by the publisher’s rather proud acknowledgement in the last volume of the first edition that Oxford University Press is present in cities like Leipzig, Copenhagen, New York, Melbourne, Capetown, Bombay, Calcutta, Madras and Shanghai.[37]

More explicitly internationally oriented systems, like the International Phonetic Alphabet, would also aim at facilitating the teaching of foreign languages, “the recording of languages hitherto unwritten”, linguistic research – and, with a gesture also shared by several other linguistic projects of the pre-World War I years and the interwar era – at serving “the advancement of peace between nations.”[38] To illustrate the practical potential of the IPA alphabet for the writing of diverse languages, the “official organ” of the International Phonetic Association, entitled Le Maître Phonétique, used this system to print all its contents (except for proper names and numbers) throughout several decades – including the 1920s – from its inception in the late 1880s.[39] Although the pacifist aspirations of the IPA were clearly not achieved, the possibility of creating an alphabet capable of being used by all nations for all languages – and thus facilitating international communication – was very much part of the phonetic discourse of the age and was, for instance, explicitly discussed (and, on practical grounds, rejected) in Ferdinand de Saussure’s epochal Cours de linguistique générale (1916).[40]

The internationalist and pacifist perspective of the IPA can itself be seen as a revision of the rather explicitly imperialist goals that its most important predecessor, Karl (or Carl) Richard Lepsius’ very influential Standard Alphabet for Reducing Unwritten Languages and Foreign Graphic Systems to a Uniform Orthography in European Letters (1863) was meant to achieve. Lepsius’s system, which was propagated at an 1853 “Alphabetical Conference” in London, was adopted as the Church Missionary Society’s standard and could well have been the ultimate source of much of the phonetic notation in Meillet and Cohen’s Les langues du monde, professed two aims.[41] The “scientific aim” was to “bring these [mainly non-European] languages with their literature within our reach, and to increase our knowledge of the nations to which they belong”, while the “practical” aim was the rather explicitly colonialist agenda “to facilitate the propagation of the Christian faith and the introduction of Christian civilisation among heathen nations”.[42]  

Although Joyce was not necessarily aware of all details of the rationales behind these phonetic systems, the passage he intended to add to the “mamafesta” text in 1925 suggests that he did know several such systems well enough to rely on them. As he had also studied numerous European languages and exploited the national associations of several diacritics in Ulysses,[43] Joyce must have been also aware that these – more or less explicitly international – systems of phonetic notation typically drew on diacritics used for languages of specific nations.[44] In effect, then, Joyce seems here to be adding a multi- and international system of diacritics to what was becoming the increasingly multi- or international language – or “anythongue athall” (FW 117.15–16) – of the Wake.[45] Moreover, by employing these diacritics, Joyce may also evoke some fundamental linguistic and language political discourses surrounding the notation of pronunciation. Some of these, like those concerned with the potential confusion caused by too many diacritics (as was the case with Vizetelly and Saussure) and those trying to advance the cause of mutual understanding and peace between the nations (as was the case with the IPA), clearly resonate with the key Wakean motifs such as Babelic confusion and brotherly war.


Joyce’s diacritics between 1925 and 1939: reductio ad absurdum

The further career of the diacritic complexity in Joyce’s 1925 manuscript addition suggests, however, that – pace Vizetelly – fewer diacritics are not necessarily better than more. Joyce’s breves (or possibly rough aspirations) were quickly misinterpreted as grave accents by the typist, who also overlooked the cedilla in introdụ̀çe. In turn, the smooth aspirations that were added very legibly to the typescripts were printed on the pages of The Criterion either as breves (ăd) or as acute accents (óles) (JJA 46: 393, 398, 415, 428). Thus, the most “realistic” facets of Joyce’s quasi-phonetic notation were blurred, and the original rhythm of stressed and unstressed vowel letters (as in ŭpòn ă plāñe) was reduced to a meaningless uniformity (ùpòn à plāñe). Still, although the printers of The Criterion could apparently not cope with the underdots and the slashed s̸, they could manage the ties and overbars, so that the final text still displays more than 30 diacritics and manages to create an impression of phonetic notation (JJA 46: 427–8).

Two years later, the galley proof (dated 2 July 1927) of the same passage for its reprinting in transition 5 (August 1927) omits all macrons, the ring above the a, the tilde above the n, all markings above noti͡ön, and half a dozen grave accents. It also turns the acute accent of the typescripts into a grave one above the a in sù ’ ’ fáç’e’, so that what remains appears phonetically even less motivated: ùpon à plane sù ’ ’ fàç’e’ (JJA 46: 438; dating: JJA 46: 419). Finally, the galley proofs of the 1939 first edition of Finnegans Wake, dating from 1937–1938, complete the process of diacritic erosion. Although Joyce seems to have made a point here of saving the grave accent of profèššionally (indicating the primary stress) from corrupting into an acute one, other diacritics – such as the phonetic tie on the͡y and the overbar on pùnc͞t – were lost, thus effacing further crucial markers of Joyce’s original textual choices (JJA 49: 167; dating: JJA 49: 1). In consequence, page 124 of most editions of Finnegans Wake has diacritics that display much less “method” and much more “madness” than Joyce originally intended, and the passage seems to culminate in a thorough diacritical “confusioning of human races” (FW 35.5), in which many of the implications of the original manuscript version appear totally lost.

Why Joyce allowed this “corruption” to happen is a relevant question.[46] One probable factor is the fact that at the time when the passage first went into print, the writer – who could be rather cavalier in his use of, for instance, French diacritics[47] – had severe eye problems resulting in repeated operations and other treatments, and he was often obliged to rely on the help of friends and family as he tried to correct the typescripts and proofs.[48] Another factor may be that he was very anxious to see his new work in print and felt that he did not have much time or indeed chance to ask for too many corrections. A further likely reason is that the printers of The Criterion, transition and Faber and Faber – just like the printer of the first edition of Ulysses – probably experienced genuine typographic limitations in trying to provide type combinations with the often arcane diacritics required by Joyce’s text, however corrupted.[49]


Conclusion: recovering ă nòti͡ön ŏf Jo͡yçé’s t̅i̅m̅e̅

It seems that Joyce, who had explored several diacritics in Ulysses, actively suppressed the accents in the 1924 French “Penelope,” and devised a mock-scientific phonetic notation for the description of the mamafesta in 1925, lost this diacritic enthusiasm by the time of the 1939 edition. Thus, the “nòti͡ön ŏf ti̅m̅e̅” that Joyce – emulating his textual scholar and linguist Profès̄s̸or/Brofèsor – hoped to introduce “ŭpòn ă plāñe (?) sụ̀ ’ ’rfáç’e’” in his 1925 text came to be ironically encoded into the 1939 version as a corruption and reduction of the very diacritics that were meant to convey it. However, probing into the genetics of Joyce’s text gives one the chance to recover Joyce’s lost notations, and with them, some of the interpretative clues that can also help provide a “notion of the time” in which Joyce was, clearly, acùtely profèššionally provòked to pụ̀nc͞t culturally and politically relevant international o̓les in the iSpåce of the Wake.

[1] I am profoundly grateful to Fritz Senn of the Zurich James Joyce Foundation for giving me not only the opportunity to do research at the Foundation, but also full access to all materials held by it. I also thank Sabrina Alonso for looking up and sharing textual data from the James Joyce Archives. My research has been greatly assisted by the wealth of texts that have been made accessible online via the Open Library and Google Books. Finally, I must thank my family for giving me the time to research and write this study.

[2] This is the version given in the 1939 Faber and Faber edition and in most later editions.

[3] A recent example is Mikio Fuse’s theologically informed genetic study of this chapter, “The Letter and the Groaning: Chapter I.5” (How Joyce Wrote Finnegans Wake: A Chapter-by-chapter Genetic Guide, ed. Luca Crispi and Sam Slote, Madison and London: University of Wisconsin Press, 2007, pp. 98-123), in which, following the lead of the first draft version of this passage, he expands on punctuation but not on diacritics. Roland McHugh’s Annotations to Finnegans Wake (revised edition, 1991) offers conspicuously little insight here, but it does mention punctuation: “The Book of Kells has 4 types of full stop. Sullivan dates it on basis of punctuation” (Baltimore and London: Johns Hopkins University Press; p. 124). McHugh also provides the Latin meaning of circumflexus as “bent; a vault” and paraphrases like “bits of broken glass & split china”, “Scotland Yard” and “by a fork”.

[4] Bernard Benstock, “Bedevilling the Typographer’s Ass:  Ulysses and Finnegans Wake”, Journal of Modern Literature, vol. 12, no. 1 (March 1985), p. 32.

[5] Michael Kaufmann, Textual Bodies: Modernism, Postmodernism, and Print (London and Toronto: Associated University Press, 1994), pp. 75, 77, 75.

[6]  James Joyce, Finnegans Wake (Amsterdam: Atheneum–Polak & Van Gennep, 2002) p. 636.

[7] James Joyce, Finnegans Wake, ed. Danis Rose and John O’Hanlon (London: The Houyhnhnm Press, 2010, and, as The Restored Finnegans Wake, London: Penguin, 2012), p. 98, and James Joyce, Finnegans Wake, ed. Robbert-Jan Henkes, Erik Bindervoet and Finn Fordham (Oxford World’s Classics series, Oxford: Oxford University Press, 2012), p. 634. The appendix of the latter edition is the work of Henkes and Bindervoet and is based on their appendix to the Dutch Finnegans Wake

[8]  I have presented much of the textual argument of the present paper in Joyce, the Universe and Everything: A Reading of Finnegans Wake 123.30-124.12”, in Whack fol the dah: Writings for Ferenc Takács on his 65th birthday, ed. Ákos Farkas, Zsuzsanna Simonkay and Janina Vesztergomi (ELTE Papers in English Studies, Department of English Studies, Eötvös Loránd University, Budapest, 2013), pp. 141-158. In an earlier essay entitled “‘Inbursts of Maggyer’: Joyce, the Fall, and the Magyar Language”, I gave a rather different, non-genetic reading of the diacritics of this passage, arguing that Joyce exploited them here as markers of the obscurity and carnality of postlapsarian languages and their written forms (Hungarian Studies 26 (2012)1, 93–106, DOI: 10.1556/HStud.26.2012.1.8).

[9] This tradition of assigning pauses of gradually increasing length to the comma, semi-colon, colon and the full stop was very much present in a book whose contents were widely discussed in the British press and appear to have informed the “Eumaeus” episode of Ulysses, The King’s English by H. W. and F. G. Fowler, first published in 1906 by Oxford University Press. In chapter IV (“Punctuation”), the Fowlers begin their discussion of modern lapses of punctuation by reminding the reader of the basics that are clearly too often forgotten: “we observe that the four stops in the strictest acceptance of the word (,) (;) (:) (.) … form a series (it might be expressed also by 1, 2, 3 and 4), each member of which directs us to pause for so many units of time before proceeding” (220). For a detailed assessment of the significance of the Fowlerian discourse for Joyce’s works, see Andrew Gibson, “Joyce through the Fowlers: ‘Eumaeus’, The King’s English and Modern English Usage,” in Joycean Unions: Post-Millennial Essays from East to West, European Joyce Studies 22, ed. Brandon R. Kershner and Tekla Mecsnóber (Amsterdam and New York: Rodopi, 2013), 225–244.

[10] Peter T. Daniels and William Bright, The World’s Writing Systems (New York and Oxford: Oxford University Press, 1996), p. 491.

[11] See also Adrienne Monnier, The Very Rich Hours of Adrienne Monnier, trans. Richard McDougall, intr. Brenda Wineapple (Lincoln, NA: University of Nebraska Press, 1996), pp. 131–2. Joyce’s association of the correct use of diacritics with education is also evident from a 1925 remark on his concierge writing with orthographic mistakes and “without accents” in French (15 August 1925, SL 308).   

[12] Throughout this paper, boldface will be used to highlight characters and combinations of characters quoted as (strings of) graphic symbols. The only exception to this is the passage quoted from A First-Draft Version of Finnegans Wake, where David Hayman’s original notation is preserved.   

[13] The general lack of diacritics in English appears to have created a corresponding diacritical purism in many English and American type designers. Eric Gill, who was to design the cover of the 1936 Bodley Head edition of Ulysses, made it clear in a magisterial, if rather cavalier, passage in his 1931 An Essay on Typography that Roman letters with diacritics (such as those in the Czech alphabet) are as foreign to “the English language” and its impeccable typographic pedigree as Greek, Fraktur or Cyrillic: “Lettering for us is the Roman alphabet and the Roman alphabet is lettering. Whatever the Greeks or the Germans or the Russians or the Czecho-Slovaks or other people may do, the English language is done in Roman letters, and these letters may be said to have reached a permanent type about the  first century A.D.” (reissue of 2nd ed., London: Dent, 1941; p. 32). Gill’s last statement chooses to overlook the entire later development of minuscule letters, as well as the patterns of local variants, gradual changes and sudden revivals in the history of the Roman alphabet.  

[14] James Joyce, A First-Draft Version of Finnegans Wake, ed. David Hayman (Austin: University of Texas Press, 1963), text from p. 89, dating p. 294. In Hayman’s notation (which is preserved here), Joyce’s additions are printed in italics, additions to additions are in square brackets, cancellations are crossed out, and substitutions are in bold face (cf. p. 44). Hayman notes that Joyce wrote “understand” for “understood” and by error repeated the article before “plane.”

[15] Joyce worked on the revision of this text for The Criterion between late February and late March 1925; see Letters III, pp. 114–7. The writer may at this time, however, have also incorporated passages written somewhat earlier. This was certainly the case with some other parts of the Wake Joyce was working on three months later: “I am working hard at Shem and then I will give Anna Livia to the Calendar. … I have got out my sacksful [sic] of notes but can scarcely read them, the pencillings are so faint. They were written before the thunder stroke” (letter to Harriet Shaw Weaver, 13 June 1925, SL, p. 307). As Joyce had been experiencing recurrent problems with his eyes for months, it is difficult to say which “thunder” or eye attack he had in mind.

[16] Rose and O’Hanlon as well as Henkes and Bindervoet read a rough breathing or spiritus asper̔ ) wherever I read a breve ( ˘ ). These marks are easy to confuse in handwritten forms (the Criterion printed breves or acute accents where the typescripts seem to have rather clear instances of the spiritus lenis), but the use of the spiritus asper would be linguistically entirely unmotivated where Joyce would appear to use them. In contrast, breves were widely used, as we shall see, in phonetic notations of English (and other languages) to indicate unstressed or short stressed vowels. In the deleted first version of a̓d, Joyce may have used a breve or a rough breathing, and both the ăd and a̔d readings could be motivated. Idiosyncratically, Joyce seems in this passage to use the smooth breathing ( ̓) to indicate the [h] sound, as if it was a rough breathing. The reading of “surface” is particularly unclear: the presence of the r is uncertain and the reading of the vowel(s) after the f could be ȧı (as in Henkes and Bindervoet’s 2002 Finnegans Wake, p. 636 and in Henkes, Bindervoet and Fordham’s 2012 text, p. 634), ai (as in Rose and O’Hanlon’s 2010 Wake, p. 98), á (as in subsequent typescript versions and The Criterion) or áı. Among these readings, ai and á seem to conform to the rest of the MS addition, which does not contain a distinctive overdot or an undotted i, but does display both acute accents (ăth é’s) and a limited number of respellings (bîn, pụ̀nc͞t ! ingh o̓les, iSpåce). The tie bar linking the c and t in pụ̀nc͞t could be read as either a straight or a curved one; is given and is left uncorrected as a straight one on subsequent typescripts. The circumflex in bîn and the overbar in time are placed above the dot of the i in Joyce’s manuscript.

[17] The dating is important: the late 1910s, the 1920s and the 1930s saw heavy orthographic and typographic changes in many nations.

[18] Several of these diacritics (the cedilla, the umlaut, the circumflex and the breve) are also used in the romanized alphabets of Azerbaijani, Turkish and other Turkic languages; these were introduced around or after the time that Joyce added this note; see Lenore A. Grenoble, Language Policy in the Soviet Union (Dordrecht: Kluwer Academic Publishers, 2003), pp. 49–50.

[19] For scribal abbreviations, see, for instance, Zsigmond Jakó and Radu Manolescu, A latin írás története [A History of the Latin Script] (Budapest: Európa, 1987), pp. 200–2.

[20] This is true of àth é’s (‘at his’) and sụ̀ ’’rfáç’e’ (if the latter does have an acute accent). The accent in piquéd is borrowed from French and indicates a different pronunciation, IPA [eɪ].

[21] Joyce made this change on a page of the Criterion (July 1925) revised for the printer of transition 5. It would seem that Joyce used double underlining to indicate regular (rather than small) capitals, and Brofès̄or has been printed with a (regular) capital in practically all editions from the 1927 transition version onwards.

[22] New English Dictionary, ed. James A. H. Murray et al., vol. 1 (A–B) (Oxford: Clarendon, 1888), p. xxv.

[23] Concise Oxford Dictionary of Current English, adapted by H. W. Fowler and F. G. Fowler (Oxford: Clarendon, 1911), p. xii.

[24] James S. Atherton, The Books at the Wake (Carbondale: Southern Illinois University Press and London and Amsterdam: Feffer and Simons, 1974), p. 87, citing Thomas E. Connolly, The Personal Library of James Joyce (Buffalo, N.Y.: University of Buffalo Press, 1955) p. 8.

[25] The examples I quote are from p. 15 of the 1878 edition of Bell’s Standard Elocutionist (London and Belfast: William Mullan and Son).

[26] Frank H. Vizetelly, Essentials of English Speech and Literature (New York and London: Funk and Wagnalls, 1915), p. 274.

[27] Vizetelly, pp. 274, 281, 286.

[28] Noah Webster, American Dictionary of the English Language, rev. ed. (New York: Harper and Brothers, 1845; and Springfield, Mass.: Merriam, 1865).    

[29] Both of these clearly “Germanised” forms first occur in the pages of Criterion (July 1925) revised for the printer of transition 5 (see JJA 46: 427–8).

[30] Vizetelly, p. 286.

[31] The quotation is from Otto Jespersen, Language: Its Nature, Development and Origin (London: George Allen and Unwin Ltd., 1922), p. 10.

[32] For Joyce’s reading and use of works by Otto Jespersen, see Erika Rosiers and Wim Van Mierlo, “Neutral Auxiliaries and Universal Idioms: Otto Jespersen in Work in Progress,” in James Joyce: The Study of Languages, New Comparative Poetics 6, ed. Dirk Van Hulle (Bruxelles, etc.: PIE Lang, 2002), 55–70, esp. pp. 56–8.

[33] Jespersen, Language, p. 16. Jespersen’s book uses serif IPA symbols for most other sounds and identifies the value of the IPA symbol [ʌ] as that of the vowel in the English word cut, which in many dialects, including Hiberno-English ones, is quite close to the schwa of the unstressed a. Joyce’s symbol in “by fork” can of course be linked also to his siglum for Shaun (the carrier of this letter) and to the logical and mathematical symbol meaning “and”, but neither of these interpretations fits into the verbal context as smoothly as the IPA-based one.   

[34] Antoine Meillet’s earlier Introduction á l’étude comparative des langues indo-européennes (1903) and his former professor Ferdinand de Saussure’s Cours de linguistique générale (1916) both use a system of transcription similar to that of Les langues du monde; both use, for instance, the š symbol for the sound commonly spelled in English as sh. This specific phonetic sign had been introduced in the second edition of C. R. (Richard) Lepsius’ influential Standard Alphabet for Reducing Unwritten Languages and Foreign Graphic Systems to a Uniform Orthography in European Letters (London: Williams and Norgate and Berlin: W. Hertz, 1863; pp. 12, 64), which also recommended the systematic use of various other diacritics, and assigned some of these – the breve, the macron, and the point below – the same value as in Les langues du monde.

[35] Antoine Meillet and Marcel Cohen, eds, Les langues du monde (Paris: Libraire Ancienne Éduard Champion, 1924). Vincent Deane gave a detailed description of Joyce’s notes on Meillet and Cohen’s book in “Les langues du monde in VI. B. 45,” A Finnegans Wake Circular 3.4 (1988) 61–74, remarking that all materials from it that were used by Joyce appeared on a set of Finnegans Wake galleys dated 12 March 1937, but received by Harriet Shaw Weaver on 16 May 1938 (p. 62). As only a few of these notes (as transcribed by Deane) and virtually none of their Wakean forms (in I.5, I.6 and I.7) display the diacritics of Meillet and Cohen, it would seem that by this time Joyce’s interest in this feature had faded. In view of the use of diacritics in the passage added to I.5 in 1925, it is not unlikely, however, that Joyce had known and explored this book already in 1924 or 1925 from a different perspective. This would not be alien to Joyce’s habits: as Erika Rosiers and Wim Van Mierlo establish, Joyce read and used Otto Jespersen’s works at several times over a period of about 15 years; see “Neutral Auxiliaries and Universal Idioms.”

[36] Meillet and Cohen, Les langues du monde, pp. xi–xiii. Joyce’s use of the ringed å instead of an ā may in fact be motivated. The writer, whose “Dano-Norwegian” was good enough to be able to write a letter in it to Ibsen in 1901 (JJII 85-7), may have known that å was a rather recent (1917) addition to Norwegian (following a longer Swedish tradition) to mark a sound earlier spelled with a double aa and going back to a long [a:] sound, thus being in fact a historic alternative to the ā notation.

[37] See the front matter of Volume X, Part II of the New English Dictionary (1928).  

[38] Paul Passy and Daniel Jones, eds, Principles of the International Phonetic Association (supplement to the Maître Phonétique, Sept–Oct. 1912), pp. 2–4; the “tendency” of the activity of the Association to advance peace is mentioned on p. 4. The Statutes of the International Phonetic Association (1995) still mention this: “The Association considers that in pursuing its aim it makes a contribution to friendly relations between peoples of different countries,” accessed 1 October 2012 <>. The advancement of peace was also a more or less explicit goal of the many international auxiliary languages that were proposed between the late nineteenth century and the late 1930s (Volapük, Esperanto, Idiom Neutral, Ido, Novial, Basic English), many of which made it into Finnegans Wake (cf. e.g. FW 116.26, 116.31, 117.13–4).

[39] The International Phonetic Association was founded in 1886 as the “fonètik tîtcerz asóciécon and its first periodical publication was called dhi fonètik tîtcer, which used a gradually evolving phonetic notation. The first version of the IPA alphabet was introduced in 1888. See dhi fonètik tîtcer, no. 1 (May 1886) and Passy and Jones, Principles of the International Phonetic Association, p. 1.

[40] Copies of the January and February 1896 number of the Le Maître Phonétique, scanned by Google Books, are available through the Open Library. Saussure’s Cours de linguistique générale (1916) explicitly states the question, “Is there a case for replacing conventional orthography by a phonetic alphabet?”, and then lists a few counterarguments, one of them being the confusingly large number of diacritics needed; see Ferdinand de Saussure, Course in General Linguistics, trans. Roy Harris (London: Duckworth, 1983), p. 33 (Introduction, chapter VII, section 3).

[41]Lepsius’s system was first published in German as Das allgemeine linguistische Alphabet (Berlin, 1854), but the really influential second (revised) edition came out in English in 1863. Meillet and Cohen’s notation overlaps with that of Lepsius’s Standard Alphabet (1863) in several instances, including the use of the caron and ring with consonants and that of the breve, the macron, and the point below with vowels, while their use of the spiritus asper for aspiration coincide with the first version of Lepsius’s system; see C. R. Lepsius, Standard Alphabet (1863), especially pp. 11-12 and 46ff. For a brief summary of Lepsius’s significance, see Michael K. C. MacMahon’s chapter entitled “Phonetic Notation” in Daniels and Bright, The World’s Writing Systems, p. 835.     

[42] Lepsius, Standard Alphabet, p. 23.

[43] For an account of Joyce’s exploitation of the national implications of diacritics in Ulysses, see my essay entitled ‘Diacritic Aspirations and Servile Letters: Alphabets and National Identities in Joyce’s Europe’ (forthcoming in Doubtful Points: Joyce and Punctuation, ed. Elizabeth M. Bonapfel and Tim Conley, European Joyce Studies series, Amsterdam and New York: Rodopi).

[44] Lepsius’s adoption of the š symbol was, for instance, explicitly motivated by the fact that it was in use in “several Slavonic languages”; see Standard Alphabet, p. 12.

[45] While Joyce’s early explications of Wakean passages tended to involve languages historically connected with the Dublin setting of the book (Irish, Latin, Danish and English), a letter dated 27 January 1925 also gives Greek, German, Japanese and Italian as sources for “words expressing nightmares”; see Selected Letters, p. 306. A  passage in book III, chapter 4 (now FW 565.25–8) contained a few Esperanto sentences as early as October–November 1925, according to David Hayman (pp. 255, 324). By August 1928, Joyce could say that he used 29 languages and language variants for a particular passage; see Selected Letters, p. 333.

[46] Wim van Mierlo discusses some possible reasons for Joyce’s apparently inconsistent attitude to the accents in another phrase (in FW 4.2–3); see “Editing the Wake,” review of Finnegans Wake, ed. Danis Rose and John O’Hanlon, James Joyce Literary Supplement 25.2 (Fall 2011) 6–9.

[47] For instance, Joyce dropped several diacritics and punctuation marks when he copied passages from Léon Metchnikoff’s Le civilisation en les grandes fleuves historiques into Notebook VI.B.1 in 1924; see James Joyce, The Finnegans Wake Notebooks at Buffalo VI.B.1, edited by Vincent Deane, Daniel Ferrer and Geert Lernout, (Turnout, Belgium: Brepols, 2003), p. 8.

[48] When between late February and late March 1925 Joyce worked on the revision of this piece, for most of the time he could not write normally or read print – let alone, one would think, diacritics; see Letters, vol. 3, pp. 114–7. When the typescript came back from Lily Bollach in early April 1925, there were plenty of major corrections needed. Some problems may easily have been overlooked in the rush, especially diacritics, which Joyce may not have seen at all: “The piece for the Criterion nearly drove me crazy. It came back from the typist (to whom I was too blind to explain its labyrinths) in a dreadful muddle. Yesterday with 3 magnifying glasses and the help of my son we chopped it up and today Mr Morel will come and sew it up on his sewing machine [typewriter]. I want to send it off before I go to the clinic” (letter to Harriet Shaw Weaver, “written in black pencil in large letters,” 11 April 1925, Letters III, p. 119). In late April, Joyce also tried to correct the proofs of another section (FW 30–34) he was to publish in Robert McAlmon’s Contact Collection of Contemporary Writers (Paris, 1925), but his eyesight made this temporarily impossible; see The Letters of James Joyce, ed. Stuart Gilbert, vol. 1 (New York: Viking Press, 1957; reissued with corrections 1966), p. 227; Letters III, p. 119. As the JJA does not reprint the Criterion proofs of this passage, it is not clear whether Joyce made any corrections to the diacritics then. When he went through the pages of Criterion two years later in July 1927, marking them up for the printer of transition 5, the only change he made here was to replace the initial letter of Profès̄or with a B (written in large letters in pencil); see JJA 46: 427–8.

[49] Although Maurice Darantiere took great pains to set and correct the text according to Joyce’s instructions, his first Ulysses (Paris: Shakespeare and Company, 1922) – and many subsequent editions – omitted carons in the mock-Czech “Přhklštř” (in all copies), and did not appear to have Joyce’s macrons in the mock-Sanskrit words “tālāfānā, ālāvātār, hātākāldā, wātāklāsāt” in at least one copy, no.  785; see the facsimile of this copy in the Oxford World’s Classics series (Oxford: Oxford University Press, 1993), pp. 289, 294, and the proofs in The James Joyce Archive, vol. 25, pp. 54, 65, 81, 97, 102. Other copies (no. 925 at the Zurich James Joyce Foundation and no. 257, published in facsimile by the First Edition Library, Shelton, Connecticut) have macrons placed at uneven heights, which would suggest that the printer had to resort to makeshift solutions.. In contrast to the 1934 first American edition by Random House (and to Samuel Roth’s previous serialization in Two Worlds Monthly, vol. 2, no. 3, February 1927, p. 345), most European editions of the novel in the 1920s and 1930s do contain these diacritics (including the 1922 edition by the Egoist Press, the 1929 French translation and the 1936 Bodley Head edition), although some (such as the 1926 reset second Shakespeare and Co. edition, the 1927 German translation or the 1932 Odyssey Press edition) clearly involved typographic difficulties and use a different font or very thin or uneven diacritics. The 1984 Gabler edition restores both the macrons and the carons (12.353–4, 12.565).