The Inclusion of Paralipomena in Genetic Editions


To study a work’s creative process and reconstruct the dynamics of the composition history, ›genetic criticism‹ is particularly interested in so-called paralipomena (short jottings, notes, etc.). Because these paralipomena – by definition – do not belong to any specific version, they are often left out of critical editions. In scholarly editing, the notion of ›version‹ is usually defined as a text that already possesses a minimum of syntactic and structural characteristics of the published text. But in a work like James Joyce’s Finnegans Wake, characterized by numerous linguistic distortions, a single word can undergo several stages of transformation even before it is incorporated into a draft or ›version‹ of the text. This necessitates a smaller concept of the ›version‹ (›word version‹ or ›Wortfassung‹ as opposed to ›Textfassung‹). By using a separate XML tag to indicate paralipomena and external reference tags to indicate their relation to previous and subsequent word versions, it is possible to chart the development of verbal transformations and include paralipomena as an integral part of a work’s genesis.

To investigate how digital philology may be instrumental in visualizing the genesis of twentieth-century writings, this essay focuses on the last work by James Joyce, Finnegans Wake, or »Work in Progress« as it was called during the seventeen years of its composition history. The methodology employed is based on the writings of genetic critics such as Louis Hay, Almuth Grésillon, Daniel Ferrer, Jean-Louis Lebrave and Pierre-Marc de Biasi. Genetic criticism (›critique génétique‹) is a relatively young discipline, developed mainly in France since a few decades. Traditionally the analysis of manuscripts was regarded almost exclusively as preparatory work in view of an edition. One of the great merits of critique génétique is that it has changed this relation: the research objective is the examination of the writing process; a genetic edition is only one of the many possible research intstruments to find the crucial moments in the composition process. It is also a way to share the result of this genetic analysis with the whole research community. Genetic criticism distinguishes itself from ›textual criticism‹ and ›scholarly editing‹. Nonetheless, genetic critics also recognize that they need to make their findings public. This may also imply editing, but a form of editing in which the main focus is not so much the reading text but the composition process that preceded it, the so-called avant-texte.

Because the genesis of a literary work is usually a complex process and often has a hypertext-structure (with blocks of text that are being moved and shuffled around), electronic scholarly editing can be of help in mapping out the writing process. In order to make the writing history surveyable the James Joyce Center at the University of Antwerp tries to give shape to this hypertextual composition process in a digital environment, using the markup language XML (eXtensible Markup Language) to chart the text structure and encode manuscript features such as deletions, additions and other peculiarities in the documents. The encoding is based on crucial notions such as document, text, version, and work as defined by Peter Shillingsburg in his book Scholarly Editing in the Computer Age, notably the chapter entitled »Ontology«. The <text> tag is used to indicate »the actual order of words and punctuation as contained in any one physical form«[1]. With reference to Joyce’s works, the physical form is in most cases paper. As a physical vessel the document contains only one text, but it may contain more than one version of more than one work – a »work« being »the message or experience implied by the authoritative versions of a literary writing«[2].

The <div> tag is used to indicate a version, that is »one specific form of the work – the one the author intended at some particular moment in time«.[3] The German scholarly editor Siegfried Scheibe gives the following definition of ›version‹: »Textual versions are achieved or unachieved elaborations of the text that diverge from one another. They are related through textual identity and distinct through variation.«[4] This implies that one single correction is enough to create a new version. For pragmatic reasons, however, the definition of a version as employed hereafter is broader than a writing layer, that is one version may consist of several writing layers.

Apart from versions, a document may also contain fragments of text (jottings, notes, reflections, try-out sentences, etc.) which strictly speaking do not belong to a version of a work. These so-called ›paralipomena‹ are indicated by means of the tag <div0> (to distinguish them from the versions, marked by the <div> tag). Since the Greek etymology of ›paralipomena‹ means ›what is left out‹, they have traditionally often been left out of scholarly editions. Nonetheless, it is important to include them in a genetic edition, especially in a case like Joyce’s Finnegans Wake.

Finnegans Wake was published in 1939. The book is about a Dublin family. HCE and ALP have three children: two brothers (Shem and Shaun) and their sister Issy. The father, HCE, is said to have committed a crime in Phoenix Park, but nobody actually knows what happened; the whole book can be seen as a web of rumours. The two brothers are each other’s opposites. Shaun the Postman is the respectable citizen, Shem the Penman is the artist. The writing process of this work is preserved on various kinds of documents. The focus in this article will be on the holograph material: the so-called B-notebooks, C-notebooks, Notesheets, and Copybooks or drafts.

1. Textual versions (›Textfassungen‹)

Finnegans Wake consists of four Books, subdivided into seventeen chapters; each chapter consists of several sections. The unit of the section is usually (in its first draft version) not longer than a few pages. This was Joyce’s favourite working unit. Of each section multiple versions are extant. So, to edit Finnegans Wake, one could start from the final narrative structure and present the multiple versions of all the sections of all the chapters. But this final narrative structure was not established yet in the early stages of the writing process. At the very beginning, Joyce wrote short vignettes, which he called »active elements«. In a letter he described them as follows: »these are not fragments but active elements and when they are more and a little older they will begin to fuse of themselves«[5].

In this process of »fusion« the materiality of the documents played a significant role. One of these documents is the Guiltless copybook (called »Guiltless« because that is the first word of the copybook). This 90-page document contains different versions of different sections of 6 chapters, written in less than 6 months’ time. Eventually, it would take Joyce seventeen years to complete the whole book. Joyce started writing in his Guiltless Copybook with the intention of writing on the right-hand side and using the left-hand pages only for additions. But after about nine pages, Joyce started filling the blank verso pages. Sometimes, he also wrote in retrograde direction, so that the sections got more and more entangled. Toward the end of the copybook, he seems to have proceeded in a more systematic way again, but the documentary context of the entanglement, especially in the first half of the copybook, is a crucial element in the process of »fusion«.

This writing method necessitates an editorial approach that allows for different approaches to view the manuscript material from different perspectives. Whereas a (1) teleological approach presents all versions of all sections in their disentangled state, a (2) documentary approach offers the possibility to study precisely the entanglement. It is also possible to rearrange the sections in the order of their composition, that is a (3) chronological approach. These three approaches work with the section as the smallest unit. The final structure of the book determines the shape of this part of the edition.

2. Word versions (Wortfassungen)

But a considerable part of the writing process took place at a much smaller level. Editing multiple versions implies a definition of the version as a relatively large unit of text, consisting of more or less well-formed sentences. For instance, Siegfried Scheibe distinguishes the »first version« from what precedes, by defining it as a text that »possesses the textual structure of the work […] contain[ing] articulated, connected sentences«[6]. By including the notion of ›Text-‹ the German notion of ›Textfassung‹ expresses these syntactic presuppositions quite explicitly.

In the case of Finnegans Wake, however, it seems necessary to additionally work with a smaller concept of the version. Most words in Finnegans Wake are distortions or combinations of existing words, and it usually took Joyce several »versions« to arrive at the final stage of this process of distortion. So, we need a concept like »Wortfassungen« or »word versions« to denote different versions of one single word or word-unit. The first »word version« is in most cases a note or an excerpt derived from a source text. Whenever Joyce used a note, he crossed it out with a colour crayon, a system he devised in order not to use any note twice.

At this level, Joyce’s writing method consists of two movements, which are described in Finnegans Wake as »decomposition« and »recombination«[7]: a source text is »decomposed«, because Joyce only picks out a few words and jots them down in his B-notebooks. Each B-notebook is based on several source texts and since Joyce almost never indicated the reference of the source text, he creates an amalgamation in which a note from a newspaper can end up next to a note from for example the Encyclopaedia Britannica, without any distinction. This obliteration of the original context creates opportunities for new associations. After a while he used several B-notebooks to write his drafts by means of a »recombination« of the decomposed elements.

In 1933, after ten years of working in this way, Joyce asked a French lady, Mme Raphael, to transcribe all the undeleted notes from his old B-notebooks into new notebooks, the so-called C-notebooks, in a very clear handwriting. For instance, notebook VI.C.01 consists of the leftovers of three notebooks (VI.B.16, VI.B.11, and VI.B.34).

From the C-notebooks, some notes were directly incorporated into a draft, but some had to pass through yet another intermediary stage, the so-called Notesheets. Since some of these notesheets are marked by one of the sigla »/\a«, »/\b«, »/\c«, and »/\d«, the hypothesis is that Joyce arranged them in four piles corresponding to the four chapters of Book III of Finnegans Wake. Consistent with the principle of teleological arrangement (following the narrative structure of the final version of Finnegans Wake), the James Joyce Archive (the Garland facsimile edition) organizes these notesheets according to the chapter of destination, using the numbering of the British Library. For instance, all the notesheets for the first chapter are grouped together, but the order of the notesheets within this group is not chronological. This is what the French genetic critic Pierre-Marc de Biasi refers to as the »rangement« or »static classification«.[8]

By retracing the prehistory of each word unit, it is now possible to rearrange each of the four piles of notesheets according to the writing sequence. This is what de Biasi refers to as the »classement« – a crucial phase in the reconstruction of the dynamics of a work’s genesis. This »classement« or »genetic classification« [9] allows us to follow the genesis not just of textual units or sections of the text, but also of a single word in Finnegans Wake. The most conspicuous stylistic device in this work is what Lewis Carroll called a ›portmanteau‹, that is two or more words packed into one.

From the perspective of scholarly editing, this lexical distortion of the portmanteau technique has some consequences for the presentation of the genesis, because – as argued before – it necessitates a more microscopic concept of ›version‹. From the perspective of genetic criticism, the main question is what this distortion-process may tell us about the dynamics of the composition and about Joyce’s poetics. Especially the indirect usage can cause major distortions, comparable to an increasingly intricate web of rumours.

For instance, from B-Notebook VI.B.11 (p. 131) the word-unit »all the world was June« was incorporated without distortion in the drafts of the Guiltless Copybook. Once it was incorporated in the draft, Joyce crossed it out to avoid double usage later on. As opposed to this example of direct usage, the cases of indirect usage often involve more distortion. Some of the notes on the same page of the same B-notebook (VI.B.11, p. 131) remained undeleted. These undeleted notes were copied into the C-notebooks by Mme Raphael in 1933, but not without mistakes. For instance, the word-unit »vice supreme regent« became »with supreme regard«. Joyce selected this note and added it to the second pile of the notesheets (»/\b«). It ended up in Book III, chapter 2: »With supreme regards«.

If we translate this into XML, starting for instance from the level of the Notesheets, the evolution of these multiple »word versions« can be presented as follows:

<div0 id="47486a-31v" rend="strikethrough">
<p id="47486a-31vh" jja="61:155" date="19360000">with supreme regard</p>
<xref bnb="B11:131f">vice supreme regent</xref>
<xref cnb="C01:183n" rend="strikethrough" color="orange">With supreme regard</xref>
<xref ms="47486a-92" jja="61:045" date="332036" section="III.2§2A.13/2B.11/2C.13">with supreme regards,</xref>
<xref fw="461.31">with supreme regards,</xref>

The <div0> indicates that this is part of the paralipomena of Finnegans Wake, i.e. the jottings and notes that do not belong to any version, although there is a link. This link is indicated in the last two <xref> tags. The tag

<xref ms="47486a-92" jja="61:045">with supreme regards,</xref>

denotes the first instance where the note is integrated in a draft of the text, indicating the British library number (47486a-31v) and the volume and page number of the James Joyce Archive (61:155). The last <xref> tag indicates the page and line number in the published version of Finnegans Wake where this note eventually ended up.

The XML transcription follows the structure of the physical documents, but the tagging of every word version also holds a record of its path through the genesis of »Work in Progress.« In a comprehensible web interface the <xref> tags which hold this information are translated on the fly into hyperlinks, using the XSL transformation language in a Java server environment, developed by Vincent Neyt. These hyperlinks allow users to pick out any entry from a notebook, get an overview of its genesis, and follow links to the previous and following versions within their documentary context.

The process of notetaking and the dissemination of these notes into drafts and ultimately into the final text, is made tangible in both directions: chronologically or counterclockwise, i.e. starting from a note’s first occurrence in a notebook (in this case bnb="B11:131f") among various other jottings derived from the most divergent source texts (such as a newspaper); or starting from the other end (from a draft or ultimately the published version), where all the passages based on notebook entries can be highlighted in different colours depending on the notebook of origin, revealing the subsequent process of »recombination.«

As this case shows, the distortion process often takes place in the so-called precomposition phase, i.e. before the incorporation in the version, when the notes are still paralipomena:

B-Notebooks: vice supreme regent

C-Notebooks: With supreme regard

Notesheets: with supreme regard

If a genetic edition only included the ›Textfassungen‹ of this »Work in Progress«, this case (»with supreme regards,« > »with supreme regards,«) would appear as a rather uninteresting invariant. Only by including all the preceding ›Wortfassungen‹ (the xrefs to the B-notebooks, the C-notebooks, and the Notesheets) it is possible to chart the development of the verbal distortion that is so typical of Finnegans Wake.

Dirk Van Hulle (Antwerpen)

Dirk Van Hulle
University of Antwerp

(23. Dezember 2005)

Works Cited:

Biasi, Pierre-Marc de

2004 Toward a Science of Literature: Manuscript Analysis and the Genesis of the Work. In (Jed Deppman/Daniel Ferrer/Michael Groden (eds.): Genetic Criticism: Texts and Avant-textes. Philadelphia: University of Pennsylvania Press, p. 36-68.

Joyce, James

1939 Finnegans Wake. London: Faber and Faber.

Joyce, James

1957 Letters of James Joyce. Ed. Stuart Gilbert. London: Faber and Faber.

James Joyce Archive

1978 63 vols. Eds. Michael Groden et al. New York: Garland.

Scheibe, Siegfried

1995 On the Editorial Problem of the Text. In: H. W. Gabler et al. (eds.): Contemporary German Editorial Theory. Ann Arbor: University of Michigan Press, p. 193-208.

Shillingsburg, Peter

1996 Scholarly Editing in the Computer Age: Theory and Practice. 3rd Edition. Ann Arbor: University of Michigan Press.


Shillingsburg (1996: 46).


Shillingsburg (1996: 176).


Shillingsburg (1996: 44).


Scheibe (1995: 207).


Letter to Harriet Shaw Weaver, 9 October 1923; Joyce (1957: 205).


Scheibe (1995: 196-97).


Joyce (1939: 614).


For a more detailed analysis of this composition process, see van Hulle (2006).


de Biasi (2004: 51).


de Biasi (2004: 52).