We take, as a straightforward illustration of this economy of witnessing and verification, an example from Roland McHugh's The "Finnegans Wake" Experience (McHugh, 85-6). The trail McHugh follows starts with a letter in which Joyce claims that he is including a reference to Emanuel Swedenborg, among others, in the current draft of II.1 (LettersI, 302). The final text bears no apparent reference to Swedenborg, but through the intervention of this note from B.33--"new teeth grow / at 81 in ES" (VI.B. 33: 187)--and its source in Swedenborg's biography which explains that he did indeed grow a new set of teeth at the age of eighty-one, the following line duly appears as the sought-after reference: "got a daarlingt babyboy bucktooth, the thick of a gobstick, coming on ever so nerses nursely, gracies to goodess, at 81" (FW 242.08-10). The notebook entry and its source thus corroborate the testimony of the letter that there is indeed a Swedenborg reference in the final text which would otherwise be missed by all but the most Swedenborgian of readers. Indeed, McHugh notes that he initially thought this passage to refer to Kierkegaard's William Afham; but once sprung, Swedenborg's newly grown teeth have circumscribed McHugh's reading. Subsequent to McHugh, Danis Rose and John O'Hanlon have catalogued further references to Swedenborg from the biography via the commodious B.33 index (Rose and O'Hanlon, 8-11).
The general economy of the testimony of genetic fragments is not necessarily as straightforward as implied by Swedenborg's bite. Each textual witness participates within the project of writing whereof it testifies. No witness in a textual hierarchy can provide a definitive testimony since each witness itself rests within the testimonies of other witnesses. This admission of a nonbounded intertextual circulation shifts us to what Schork pejoratively calls the "expansive view" of genetic criticism (Schork, 108-9). We will try to demonstrate that any operation of textual verifiability is expansive. Jean-Michel Rabaté provides a precise elaboration of the implications of such a claim: "The examination of the pretext uncovers the same mechanism as the final text, but it also reveals a method and a final philosophy... an infinitization of language results equally from the protocols of decipherment and genetic exegesis" (Rabaté, 88).
The circumscribed view also contends with this potential infinitization of referential play. Already in the Swedenborg example there is a codependency between the textual and the pretextual matter: Joyce's letter leads to the notebook which depends upon a prior source. Furthermore one could include the draft stage in which the reference first appeared in order to corroborate the link between the notebook entry and the final text. In other words, although this methodology tends towards, as Schork states, circumscribing the possibilities of reading, its potential for interpretive adjudication rests within a mobile and expansive circulation of textual witnessing. Witnessing moves up and down, pellmell through the genetic chain.
We would now like to turn to Jacques Aubert's treatment of the problems of reading and articulating the ostensive first word of Finnegans Wake because he also raises, perhaps inadvertently, some of these same problems concerning the circulation of testimony. The problems Aubert identifies in reading riverrun are also problems of geneticism and witnessing.
Aubert begins by arguing that the articulation of "riverrun" (FW 003.01) depends not upon any meaning of the word itself, but rather upon economies of differentiation registered by both the word and its context. Alone, the word suggests several incongruous possibilities. The word has an excess of possible signification because it belongs to multiple grammatical genres: noun and verb combined (river run with an elided pause); solitary substantive (with an elided article ( ) riverrun); literary allusion; &c (Aubert, 69-70). Already this solitary word is excessively expansive.
Of course, many critics over the years have proposed their own readings of the possible meanings of this one infamous word. Aubert's initial and provisional solution to this inscrutable effluvia of possible signification lies in reading the entire sentence in order to determine the syntactic rôle of "riverrun." Paring away the subordinate clauses, Aubert arrives at the syntagm's verb: "brings." The verb thus partially circumscribes "riverrun" as a substantive. This delimitation can also be justified by a downstream examination of this sentence's archive: in the first draft there is the germ of the first fragment, albeit without the infamous substantive: "brings us to Howth Castle & Environs!" (FDV: 46; JJA 44:3; LettersI 247). In essence Aubert has proposed a reading that begins with the syntagm's being pared down to its initial formation. Aubert has gone downstream within the syntagm's syntax to arrive at the verb, and through the archive we have rearrived at the same conclusion.
Through such a reading, riverrun appears in the place of a subject (Aubert, 70-1). A noun is actualized through its articulation by a verb. However Aubert points out that the recourse to syntactic norms yields an anacoluthic sentence in that the substantive "riverrun" lacks some kind of article. The word still exceeds its circumscription by the verb. A definite article is available by the aposiopesis on the last page of the book: "A way a lone a last a loved a long the" (FW 628.15-6). The break--the unenjoined article--is supposedly to be read as a jointure between the first and last sententious fragments. But this fracture between fragments might not yield a simple jointure. We will return to this problem of fractured articulation in our conclusion. Aubert does claim a continuity predicated upon this circular discontinuity:
In the final analysis, the point is this: the interplay is out of place, that is, it does not occur between 'river' and run,' but in front of the first of these two words. In the beginning, there was play on play, and there lies the real impertinence of the text. And within this initial play the potential for articulation and nominalization alike is contained: they will spring out of the space between the two poles represented by 'the' and 'riverrun' (Aubert, 72).Aubert thus demonstrates how a grammatical indetermination (of the silent pause between the first and last words) surrenders to determination. "Riverrun" is determined by neither its expansive semantic overtones nor its delimiting syntactic rôle, but rather its articulation arises from a circulation or interplay between the expansive modalities and the circumscribing contexts. Aubert has thus provided us with a paradigm for genetic criticism by insisting upon a mobile adjudication between contested possibilities of reading which could be opened along the axes of a text's archive.
An individual witness can circumscribe a reading, but a cluster of witnesses necessarily tends towards an expansion of the modalities of reading. The interplay between circumscribing testimonies is expansive; and no one reads just one word in the Wake. Indeed, when "riverrun" was first added, it appeared as two words: "river" and "run" (JJA 44: 105). [I have recently learned from Luca Crispi that the word "river" was added to this phrase in the hitherto-unknown previous draft-stage: Joyce's corrections to the carbon of the typescript he had sent to Harriet Shaw Weaver, the correction was made between the15th and the 29th of November, 1926. This manuscript is part of the Sylvia Beach archives at Buffalo, TS 6J-1A. "River" came first, then it ran.] The words remain separate in the next draft level, the first typescript (JJA 44: 145), although there is an indication that the words should be joined. In the duplicate typescript Joyce indicated that this word is to be capitalized, although this orthography was not retained (JJA 44: 174). The proofs for transition 1 contain the word as we know it today (JJA 44: 204).
This excursion through the genetic archive testifies on behalf of the first apparent reading of "riverrun" as a subject and a verb: a reading noted by Aubert and countless others. But different archival tributaries testify to some of the other possibilities registered by Aubert. Looking for the genetic derivation (which means, from the Latin, to change the course of a river) of "riverrun" seems to be not unlike searching for the origin of the Nile.
The confluence between Aubert's syntactic and our genetic readings is lost if we wade further down one tributary of the archival stream--back to B.15, the principle notebook for I.1--in order to locate a potential first element of the sentence. In B.15 we find "Howth Castle and Environs" (VI.B. 15: 33). The apparent genetic source of ALP's rearriving riverrun is the destination naming HCE. This recalls John Bishop's reading that the first paragraph "teems with forms of 'the zeroic' couplet HCE and ALP, which Joyce modulates in the course of the book" (Bishop, 367).
Along another tributary, in B.9--which was composed in the first half of 1925, slightly over a year before Joyce began work on the first chapter--one finds two likely antecedents to the word "riverrun": "riverrain" (VI.B. 9: 45) and "riverend a" (VI.B. 9: 148). The latter made its way into I.8 (FW 203.18); it was added to the manuscript in September 1925 (it was added to the page proofs for Le Navire d'argent [JJA 48: 164]; the context "the riverend name" of Michael Arlow, a gentleman associated with ALP, seems to bely the notebook association of ALP riverend). Although B.9 was not primarily used for the composition of I.1, there are at least two entries near "riverrain" which were added into the drafts of I.1 during the same period as the insertion of "riverrun," but into different draft stages (such as "rill" [VI.B. 9: 50] which appears in the first draft of the second half of the chapter [JJA 44: 99; FW 023.17]; and "wurrum," an addition to the same draft stage [JJA 44: 85; FW 019.12]; this draft stage was also late November 1926. In a letter to Weaver [May 13, 1927], annotating a brief passage from I.1, Joyce explicitly associated rill with a [Selected Letters 321], thereby corroborating a conceptual association between the I.1 entries in B.9 and ALP).
This is still not enough evidence to definitively convict riverrain into service as a direct ancestor for "riverrun." Indeed, the bulk of the material surrounding the entry "riverrain" appears in I.8. These include a diagram of a river detour (VI.B. 9: 53) and a diagram of a river branching off into tributaries adjacent to the ALP siglum (VI.B. 9: 57). "Riverrain" is an indifferent witness to the Wake's first word.
No one of these antecedents to "riverrun" seems to be a definitive singular source, indeed they all serve as a plurality of witnesses. Without knowing of the corroborating B.9 evidence Aubert signals riveraine as one possible overtone to "riverrun." This lexical overtone of riveraine connects the opening word to the washerwomen of I.8 because "the 'riveraines' in the book are the washerwomen positioned on the banks of the river, a symmetrical pair of witnesses and commentators on events, on the flow of events. If this value remains scarcely perceptible of a first reading, by a second one this fact is known. ... This reactivation is based on the interplay between 'river' and 'riveraine,' the interplay of stable and fluid, of witness and event; there are several options: witness viewing event, and event viewing the witness, and so on" (Aubert, 76).
Aubert thus claims that the specific interplay between "river" and "riveraine" is a modulation of the general interplay between expansiveness and circumscription. Returning to our genetic inquiries, we will now claim that notebook entries--in the sense that they are draft witnesses--are riveraines: witnesses standing on an unstable border between stability and flux or ground and water. And like the washerwomen, these textual riveraines are unstable and labile witnesses from swerving shores to bendings abeyed.
In conclusion, we will now return to the question concerning the articulation of "riverrun." The final the articulates riverrun but only indefinitely; for indeed such articulation renders the end of the book indefinite, if only to fin again. Aubert does not radicalize his argument concerning the "true genealogy of the noun" (Aubert, 72) by admitting that the interplay of the diverse effects flowing from "riverrun" are themselves pronominative. These divagations between presence and absence take the place of a name. In other words, the Wake is perhaps not unlike the reverend letter in that it is a "prepronominal funferal" (FW 120.09-10). The disarticulation performed by the word itself falls within a disarticulation of syntactic placement.
The game of enjoining article to word is not a definite (re)articulation of a sentence cleaved in twain at opposite ends of a book. On the one hand there is an additional interruption, a witness to the time and place of writing, found at the bottom of the last page:
PARIS,The book Finnegans Wake thus folds around a witness of the completion of the work in progress: its dating. This signature interrupts the rejoining of the first and final fragments. Indeed the date, unlike the last sentence, ends with a period. The date thus imposes a finality and a closure which remain unproffered by the last sentence. The possibility of the noun's articulation out of a differentiation proffered by disenjoined articles is itself differentiated by a testimony to the process of writing, a testimony which is apparently rendered terminal by virtue of the period. The "commodius vicus of recirculation" cannot be neatly folded upon itself in a definitely articulated or perfectly circular act of bookish rejoining and completion because of a witness to this task of writing.
1922-1939.
One could read a complete, unfractured sentence here, "A way a lone a last a loved a long the riverrun past Eve and Adam's..."; but there is the interruption of the book itself, or rather an interruption of the book's exteriority, an interruption suspended by the covers and binding of the volume itself. The "commodius vicus of recirculation" is pre-dicated upon an interrupted interruption of the book's very pre-position: the pre-positioning of riverrun found in the last word and the pre-position of the Wake, its imbricated pretexts. The commdiousness of the Vichian recirculation remains "usylessly unreadable" (FW 179.26). "Riverrun" is thus a disjointed resumption of a perpetual fragmentary redeenunciation. The notebooks, like the Wake itself are expansive witnesses to the in progress working of this project of fragmentary redeenunciation.
Jacques Aubert, "Riverrun," trans. Patrick O'Donovan, Post-Structuralist Joyce, eds. Derek Attridge and Daniel Ferrer (Cambridge: Cambridge University Press, 1984) 69-77.
John Bishop, Joyce's Book of the Dark (Madison: University of Wisconsin Press, 1986).
Jack P. Dalton, "Re 'Kiswahili words in Finnegans Wake' by Philipp Wolff," A Wake Newslitter (old series), 12 (1963): 6-10.
James Joyce, A First-Draft Version of "Finnegans Wake," ed. David Hayman (Austin: University of Texas Press, 1963); abbreviated as FDV.
Roland McHugh, The "Finnegans Wake" Experience (Berkeley: University of California Press, 1981).
Laurent Milesi, "L'idiome babelien de Finnegans Wake," Genèse de Babel, ed. Claude Jacquet (Paris: CNRS, 1985), 155-213. Our translation.
Jean-Michel Rabaté, "Pour une cryptogénétique de l'idiolecte joycien," Genèse de Babel, ed. Claude Jacquet (Paris: CNRS, 1985), 49-89. Our translation.
Danis Rose and John O'Hanlon, "Constructing Finnegans Wake: Three Indexes," A Wake Newslitter (new series), XVIII.1 (1980): 3-15.
R.J. Schork, "By Jingo: Genetic Criticism of Finnegans Wake," Joyce Studies Annual 1994, 104-27.
George Tobridge, A Life of Emanuel Swedenborg, second edition (London: Warne, 1912).