The recent publication by Brepols of the first three notebooks at Buffalo is to be widely welcomed. There is one area however that was not perhaps sufficiently covered and that is which notebook entries were transferred to which drafts. There is an Appendix to each volume which tells you which notebook entries appear on which pages of Finnegans Wake, and there is an Abstract of Draft Usage which lists the draft stages to which material was added. However there is no table that arranges the transferred notebook entries according to the draft stages in which they were used. To illustrate this rather technical matter, I am supplying two tables of the information here, compiled from the Brepols edition. The tables list the draft pages of a particular draft stage and, for each page, the location of the notebook entries transferred to it are indicated.

In the first table I have included a column of brief notes about the relevant draft pages in the Archive. After both is a brief analysis. The notes give a rough idea of how the notebook entries were used (whether as a small self-same unit or to help make up a new unit) and to give a sense of proportion, i.e. how many revisions on that page there were. The analysis that follows shows what issues are opened up by this table – and what further work for geneticists it invites.

The research carries with it the suggestion that the editors of the Buffalo notebooks might incorporate elements of it as an appendix in their edition. Now for the context behind the tables: in the early months of 1930 Joyce used VI.B.29 (and other notebooks) to supplement HCE's citybuilder's crescendo speech that had appeared in transition 15 and would eventually form the ending of chapter III.3. This was the twelfth draft level for the section: there would be three more before the piece was published in a special edition of Haveth Childers Everywhere by Jack Kahane. We can see how Joyce flicked through his notebook from the following table:

Use of VI.B.29 on III§3B.12. February 1930

So can we uncover a system for Joyce's use of the notebooks for revision? It varies from draft page to draft page. Sometimes the notes used are close to one another in the notebook (as on page 86), possibly from a cluster of notes with a shared source. At other times they are disparate (as on page 104). For some pages the notebook provides nothing; elsewhere five or six entries may be used (p. 106, 104). As we shall see, Joyce did not necessarily move forward through the notebook in order for each draft page. One things is clear: he was ranging widely over the notebook, and other notebooks – as if to ensure an aesthetic of continuous relocated dislocation, as if a portion of text could connected anywhere, no matter how disparate.

While we can get an idea of how Joyce moved in and out of this notebook as his revisions progressed, for a fuller idea we have to recognize – again – that Joyce was in fact using several notebooks simultaneously in revising a passage. This fact intensifies our sense of the bizarre complexity of Joyce's method of composition, especially in the way he was combining scattered elements into one 'revision unit' – whether consisting of two words or longer phrases. This cannot be properly assessed, however, until the elements used from other notebooks on a given draft page are collated with this one.

To indicate how VI.B.29 was used on a sample page, we can transcribe JJA59:104 as follows. The words in plain type are the material Joyce was revising. Those in italic are the revisions, and those in bold italic are revisions that draw on VI.B.29. The notebook location is indicated in square brackets after each unit (hyperlinks would obviously make this textually clearer).

From this we can see the order in which Joyce used the notebook. He picked out a word from page 2, moved forward to 61 and found another, then moved forward again to 78. Then he flicked back to 28, then on again to 51. In addition we can get a sense of proportion: the typescript Joyce was revising consisted – on this page – of 228 words. He added another 52, increasing the text by about 23%, just 9 of which had a source in VI.B.29.

Now, to get a further sense of proportion and a comparative perspective, we can table the next but one draft level, that is III§3B.14. This was a set of revisions made for the printer of Haveth Childers Everywhere:

Use of VI.B.29 on III§ 3B.14, Spring 1930

Compared with III§3B.12, the notebook at this level is being far more heavily ransacked – not just because the section has become longer but there is greater use per page. Only on the first page of the typescript is VI.B.29 not used. Every other page has something from VI.B.29, and page 198 has sixteen elements. Many of these make up a list (a list within a list) of architects who helped design and build civic Dublin. The notebook is proving more valuable, or Joyce is wishing to have more of the intertextual colour it can provide within his textual mosaic. Continuing this comparative use of the tables, we could eventually have a sense of how the use of a notebook rose and fell during composition, and its relative importance compared to other notebooks.

The immediate stage of research following this one will be to integrate other notebook usage and arrange it according to the draft pages – then we will have an idea of how Joyce pulled out his material from what kind of variety of sources.