[To return to the demonstration page: FW: 48.
To jump to the basics of the notation system or its details: authorized changes between levels; changes within levels; TDs and IMs: transmissional departures and instructions missing; further symbols.]
The representation of the revision history of Finnegans Wake shows additions, deletions and changes which Joyce made to his documents from his first draft (level 0) through to the final corrections he made in 1940 on unbound pages of the published book (level 13). 1 The list of these revision levels may be seen here, and also in the dialogue box which will appear when one runs the “reduction program”, fw.exe, in the electronic version. This program allows the user to display , for any Wake page in chapters 2-4 (FW: 30-103):The representation is not a diplomatic transcription of the documents but a telling of the genetic story. Therefore:
- the full genetics of the text on the Wake page selected, from first draft to level 13, the published and corrected book;
- the differences which are introduced at any level from the text as it had been established at the immediately preceding level;
- the changes made on any continuous set of levels to what preceded them--for instance, one might call for “genetics from first draft through level 2” or “genetics of the galley proofs, levels 9 to 11, as they revise the typescript, level 8+T, which was their model before the galley revisions began”;
- the “plain text” output of any revision level, i.e., what the text looked like after revisions on that level were complete (and therefore also what the text was which the next revision level was to work on).
- The printed representation, in our published book, will show for each Wake page the basic genetics of that page (first draft to level 13), much like the demonstration page here (though the print version will for ease of use omit some details, like typographical errors and double punctuation). The CD which will accompany the printed book contains an application (the “reduction program”) permitting the user to display for any Wake page its full genetics or to display any revision level or set of levels, with any combination of the less significant details desired. There will also be, similarly to this demonstration, copies of the full genetics of each page with links to representations of each revision level for that page, for quick reference.
- Immediately corrected errors on a level, such as when a typist overstrikes a wrong letter with the right one, have not been recorded. Typographical errors (errors of transcription from one level to another, on which they are corrected) are more significant, since they often prove Joyce's intention to preserve unusual words, and they may be displayed or suppressed in the electronic version at the user's choice, but are not shown in the print version.
- The story is told word by word, not letter by letter. If Joyce changes a word by lining out and replacing a single letter in it, the representation shows the whole word as deleted and then replaced with the word as changed.
- Purely typographical matters are treated by conventions. When Joyce indicates italics by underlining or marginal instruction, the italics are shown from the start rather than showing the passage first underlined and then changed to italics.
- No effort is made to reproduce the oddities of spacing typists introduce (some, for instance, insert a space before a colon, others not) which there is neither reason to think Joyce intended nor sign of in the printed book. Ligatures are, however, marked as they arrive and depart from the sequence, as saying something about Joyce's spelling preferences.
- “Double punctuation”: if Joyce inserts a phrase, for instance, he will often place a comma at the end of his overlay even though one is present already in the substrate, presumably so that his copyists will be in no doubt. Double punctuation is not shown in print, and may be displayed or not in the electronic version.
- The printer's final hyphenation in the published book is usually uninteresting, and no such hyphens are marked . The very few cases of interest receive textual notes.
Changes which appear in transcriptions made by typists and typesetters and which are not authorized by Joyce's holograph or other surviving instruction are marked using two systems between which the user of the electronic version may choose. (It is not yet decided whether to print IMs or TDs only.):.
- If what is wanted is the record of the surviving documents in the revision chain, all unauthorized transcription changes may be displayed as “TD” (“transmissional departure”), thus: “ [x TD text x]”, which says that the word “text” appears without Joycean warrant on the document on which revision level x is being made.
- But since it is sometimes certain that documents in the revision chain have not survived, the user may choose to view all changes assignable to those documents as “IM” (instruction missing). Also, sometimes (though rarely) changes in substantives appear which are unimaginable as copyist's errors and which must be the result of some form of extradraft instruction: for instance, no typist or typesetter added on his/her own whim “straw braces” after “fight shirt” (FW 85:34 at level 9) or substituted “send treats in their times” for “send them Treats” (FW: 92.22 at level 8+T). These changes too are marked and may be viewed as “IM”. E.g., “ [x IM text x]”.2
- One exception: changes on the final page proofs, which are missing, are exempted from the system and are all by convention assumed authoritative (with qualifying notes where needed), with the level marked “12*”, the star meaning “document missing”.
Matters of textual interest and important facts about the documents—ink colors of overlay, unusual locations of drafts on supplementary pages, etc.—are discussed in copious “TN”s (textual notes), displayed as endnotes to the printed representation of each Wake page and accessible by hypertext links in the electronic version.The notation system:
Boldface designates first-draft text (level 0, in the system of the “draft tables” of The James Joyce Archive, used with a few modifications throughout). Display in red designates text deleted at any revision level (so bolded red designates text inscribed and deleted in the first draft). Normal type designates text which survives at the highest level being displayed. The first draft may thus be “read off” from the bold; the document at the highest level being displayed (including the last, the corrected book) may be read off from whatever is not in red. And the red shows what was present before that highest level, but which was dropped or changed within the selected range.
Superscripted bracketed numbers enclose changes at a given level after the first draft: “a [1 <long> high 1] perch” (FW: 31.02) shows “high” replacing the first draft's “long” at level 1. Overlay additions to a document on which revisions are being made are indicated by a pair of inward-pointing superscripted carets, each pair with a superscripted capital letter (A>. . . <A) which designates the degree of overlay. For example, “[3 A>, faced B>feets<B to the east,<A 3]” (FW: 76.11-12) shows “, faced to the east,” added as first-degree overlay to a typescript, and then “faced” (but not the comma) scratched out and the second-degree overlay “feets” replacing it.3
The specifics are perhaps best listed, with examples:
1) Authorized changes between levels. Note well: Where the first element after the level-indicating bracketed number is in red, that element was present on the level immediately below the level indicated (with one rare exception, for which see “initial inscription and deletion” under “Changes within levels”).
. . . [x newtext x] . . .
: currente calamo addition: “the [4 relics of the 4] bones” ( FW: 91.05). The phrase “relics of the” is added at level 4, during Joyce's fair copying of the level 3 typescript.[x A>newtext<A x]: overlay addition: “It was the Lord's [3 A>own<A 3] day [3 A>for damp<A 3]” (FW: 51.24). Joyce overlays “own” and “for damp” on the level 3 typescript, which is the substrate of the level 3 revisions.[x <text> x]: omission: “had not been [1 [4 <very> 4] 1] many” (FW: 41.10-11). Joyce adds “very" in his level 1 copy of the first draft, but drops it at level 4.[x <oldtext> newtext x]:substitution: “to ask [1 <if he could> could he 1] [2 <say> tell 2] [1 <what it was> how much 1] o'clock [1 it was 1]” (FW: 35.18-19). Oldtext is omitted during Joyce's copying and is “silently” replaced by newtext (therefore, no overlay carets); the symbols “< . . .>” mean that what they contain is not struck out, but rather that it has not been copied forward. Here, while copying at level 1 Joyce changes “to ask if he could say what it was o'clock” to “to ask could he say how much o'clock it was”, and while copying at level 2 replaces “say” with “tell”.[x oldtext newtext x]: replacement currente calamo: “any luvial mud [4 on to o'er 4] his face” (FW: 86.19). Oldtext on the substrate is copied to the new document, but then struck out and newtext inscribed as Joyce's copying continues. (This is quite rare between levels, but more common within a level.)[x oldtext A> newtext <A x]:replacement by overlay: “to [3 repeat A>pianissime<A 3] a slightly varied version” (FW: 38.30-1); or “with [3 deadly accuracy A>Anny Oakley deadliness<A 3]”(FW: 52.01); or “such a [8 wonderful A>wanderful<A 8] noyth untirely”. (FW: 59.13). This is more common than replacement currente calamo and is, in fact, Joyce's main revision technique. Note that he almost never (I cannot find an unproblematic example) simply strikes text without replacing it: if he did, it would look like these examples without the overlay.. . . [x <text as of previous level> x] . . . [x text copied at this level x]:movement between levels: “petties and shouting [4 <A>whether he was sure of his> 4] name [4 <B>names<B . . .That would do,<A> 4] . There were cries . . .” (FW: 87.29); “. . . altar [4 . And a new complexion was put on the matter when to the A>The mixer, accordingly, was bluntly broached as to whether he was one of those B>lucky cocks<B for whom the audible - visible gnosible - edible world existed. That he was only too sure of it.<A Whether he was sure A>too<A of his names . . .” (FW: 88. 04-12). Joyce drops the level 3 passage which, much extended, becomes FW: 88.11-90.33; he then copies it at level 4 where he precedes it with fresh overlay. (Note: often movement between levels is caused by a typist or typesetter misinterpreting Joyce's indexing marks: marked the same, but as TD in both locations.)2)Authorized changes within levels:
[x {newtext} x]
:initial inscription and deletion: “Was six pounds fifteen [1 {A>in all}B>in round figures <B<A 1] taken [1from A>off<A 1] you” (FW82.14-15). This is the rare false start, in which Joyce writes something new as the first element of change at a level and then deletes or replaces it. The special symbols “{. . .}” are needed to show that the element was not present on the previous document, unlike all other first elements which are in red. In the example, note the difference: “from” was present in the first draft, “in all” was not.[x . . . text . . . x]:deletion of text new at a level: “[8 A>. . . the refaced unmansionables of brown gingerdied hue . . .<A 8]” (FW: 82.26: “gingerdied” added); or “[8 A>and to league his pagan lot, palm and patte, with a papishee : B>For mine qvinnie I thee rape C>giftake<C and bind my hosenband I thee haltar<B<A 8]” (FW: 62.06-9: note “giftake” added as third-level overlay). Note that Joyce strikes out “brown” and “rape” but continues with “gingerdied” currente calamo and “giftake” in overlay.[x A>text<A x]
:overlay, first degree: “in fealty sworn [9 A>(my bravor best! my fraur! )<A 9]” (FW: 38.33). Note that all overlay carets “travel with” the text they mark. Consider: “the [3 overture [9 A>overthrew<A A>overthrewer<A 9] 3] to the [3 A>third<A 3] last [3 day A>days of [5 TD <pompery<A> Pompery 5] 3]”. Here the carets around “overthrew” show that it was overlay at level 3, where it replaced the first draft's “overture”; the caret after “pompery” travels back to show that “days of pompery” was the phrase replacing “day”, before the TD which capitalized “Pompery” at level 5.
[x A>. . . B>text<B . . . <A x]:overlay, second degree: “[3 A>in a burst B>loudburst<B of poesy<A [4 , 4] 3]” (FW: 91.03). The “burst” of the first overlay addition is struck out in revision and replaced by “loudburst”.[x A>. . . B>. . . C>text<C . . .<B . . .<A x]:overlay, third degree: “after which [3 A>stag luncheon B>and a few ones more C>, flushed with their firestufffostered friendship,<C to celebrate yesterday<B<A 3]” (FW: 42.06-7). After adding “stag luncheon” to follow “after which”, Joyce adds to it “and a few ones more to celebrate yesterday”, and then inserts “, flushed with their firestufffostered friendship,” within that.
[x A>. . . B>. . . C>. . . D>text<D . . .<C . . .<B . . .<A x]:overlay, fourth degree: “not forgetting [3 A>a B>particularist<B prebendary pondering on the Roman question B>easter, C>the tonsure question & Greek uniates D>, plunk em,<D<C<B<A 3]” (FW: 43.12-13). To “Roman easter”, replacing “Roman question”, Joyce adds first “the tonsure question & Greek uniates” and then “plunk em,”. The embeddings occasionally reach to six degrees.[x . . . A>text first placed<A . . . x] . . . [x B>text next placed<B x]:movement within a level: “Sordid Sam [8 B>, the unwashed<B 8], haunted always by his ham, [8 B>the unwished<B A>, at a word from Israfel the summoner,<A 8] passed away . . . propelled from [8 {A>the unwashed<A} 8] Behind into the [8 great A>the unwished<A greatTN1 8] Beyond” (FW: 49.21-5). Joyce overlays “the unwashed” and “the unwished” to precede “Behind” and “Beyond”, and moves them (with pencil lines invisible in the Archive) to follow “Sam” and “ham”. Note that the second placement is marked as second degree overlay since it is within the same level as and replaces deleted first degree overlay. (This form of revision is very rare.)3)Unauthorized differences between levels:
TN1 Note that when level 8 first indexes “the unwished” to replace “the great”, Joyce fails to delete “the”, so that the instruction would yield “the the unwished”. Since he reindexes for placement after “his ham” (with pencil lines which, while very heavy and plain, are for some reason invisible in the Archive reproduction), “the great” is left as output of level 8.
[x TD <oldtext> newtext x]
:transmissional departure: “wore a raffles ticket [5 TD <A>in<A> on 5] his hat” (FW50.26). The level 5 typist mistakes Joyce's level 4 “in” for “on”. At level 4 Joyce first wrote “on” but overwrote the “o” to “i” the typist made the wrong choice.[x IM <oldtext> newtext x]:instruction missing: “proclaim him [5' IM <a descendant> offsprout 5'] of vikings” (FW30.08-9). Proofs of the Contact Collection (level 4*) are missing, on which surely Joyce surely instructed for the change we see in the printed Contact Collection.[12* A>. . .<A 12*]:the page proofs: The one exception to the marking of IMs is that all changes which must have been made on the missing page proofs are shown as overlay revisions, for user convenience: “[4 . . . [9 A>, In Custody of the Polis,<A 9] [9T TD : 9T] 4] [12* A>Boawwll's [13 alocutionist A>Alocutionist<A 13] , [13 deposed A>Deposed<A 13] ,<A 12*]” (FW: 72.16). Here the epithet which brings Joyce's total to 111, “Boawwll's alocutionist, deposed,” (modified a bit in Joyce's post-publication corrections, level 13), must be an addition made on the page proofs.4) Further symbols:
[x . . . - x] and [x - . . . x]: revisions which cross page boundaries: that a revision is not fully contained on a single Wake page is shown by one hyphen preceding the level indicator on the page beyond which it is continued and another after the level indicator on the page on which it is continued: “. . . [1 . . . [2 [3 which A>what<A 3] all did - 2] - 1] [page break] [1 - [2 - so [3 that A>as<A 3] he was able to add 2] . . . 1]” (FW:54.36-55.01). The level 2 overlay “which all did so that he was able to add” extends across what become pages 54 and 55. Note that this is contained within a level 1 revision which has also begun on page 54 (or previously) and continues on page 55 (or beyond).. . . ¤ . . .:word illegible: “ [3 . . . Sycomore Lane [4 <A>Arrah,<A Nick, ses she, you've the nock, ses she, A>with your ¤ , ses she, ye ¤, B>says she<B<A> . Fine feelplay . . . 4] . . . 3] “ (FW: 95.21). Two words, in a sentence not copied from its initial draft, are enticing but illegible.. . . ¤ . . .:letter, letter sequence or mark of punctuation illegible. Unlike the preceding, this mark may cover more than one letter, because it is far harder than with words to tell how many letters are involved when you can't make them out!: “[3 A>. . .B>. . . at Multifarnham. Ba¤----------------. That would do,<B<A 3]”: After the two or three illegible letters which follow “Ba” in the initial draft of what will become the thunderword at FW: 90.31, Joyce puts as placemarkers a line and two part-lines of dashes.. . . ?text? . . .:reading not certain: “[8 . . . Bigamy Bob and [8+ <and ?sould? Sin Shanock> his old Shanvocht 8+] . . . . 8]” (FW: 48.03). What is bounded by superscripted question marks is not so illegible as to warrant the refusal of a reading, but is far too doubtful to propose with confidence.[x •A>text . . .<A x]:insert mark missing: “[5 A>(commonplace!)<A 5] [8 •A>, since known as Whiddington Wild,<A 8] his simple . . . .” (FW: 52.10). The “Whiddington Wild” overlay is drafted on level 8 in the top margin with no index mark and no mark in the text. I show it where placed at 8+. A textual note usually gives details.. . .×. . .:insert mark present but insert missing: “[3 A>. . . the Roman× Godhelic faix . . .<A 3]” (FW: 91.35-6). There is an “R” index mark following “Roman” at level 3, but no overlay matches it. As above, this situation will usually be discussed further in a textual note.
Notes
1 The symbols used are modelled as closely as utility and user-friendliness permit on those used in the 1986 Ulysses of which Hans Walter Gabler was main editor, especially those symbols which designate documents with numbered superscripted brackets and those which indicate the degree of overlay addition to the text by a system of caret marks (modified from Gabler's system). The most notable departures are that 1) first draft is in boldface, thus, rather than bracketed as zero level, “[0 . . . 0]” , and 2) deletions are shown in red, thus rather than with pointed brackets, “<. . .>”. (Two specific varieties of deletion, one of which does involve pointed brackets, are discussed below).
2 All differences assignable to a known missing document are marked IM, rather than some IM and some TD, although transcription errors can as easily have been made on missing documents as on surviving ones. On the other hand, there are some documents on which so many differences appear, especially involving punctuation and word distortion, that one suspects either a missing lightly revised document or some other form of extradraft instruction. While such differences are all marked “TD” because authorization is too uncertain, all instances where instruction seems likely are pointed out in textual notes.
3 “Overlay” means that text is inscribed interlinearly or somewhere else--a margin, a verso of a preceding page, a supplementary sheet. Textual notes tell where, if it is not nearby on the document being revised. Joyce marks his insertion points with a variety of symbols such as caret marks, connecting lines, letters and numbers. If these are missing that fact is marked, as is the presence of such indices without matching overlay (see below).