A Day of DH at the Manuscript Division of the William Blake Archive

[Cross-posted with the Blake Archive’s submission to the official Day of DH blog!]

The Blake Archive has editors and assistants working at various campuses around the US, including a group at the University of Rochester. In residence at the University of Rochester, we have:

• Morris Eaves, co-editor of the William Blake Archive
• Esther Arnold, PhD student (English) and project assistant
• Laura Bell, PhD student (English) and project assistant
• Duncan Graham, undergraduate (Economics) and undergraduate intern/project assistant
• Sarah Jones, Managing Editor, Blake/An Illustrated Quarterly
• Gabi Kirilloff, MA student (English) and project assistant
• Rachel Lee, PhD student (English) and project coordinator
• Hardeep Sidhu, PhD student (English) and project assistant
• Lisa Vandenbossche, PhD student (English) and project assistant

Working off-site, we have:

• Andrea Everett, PhD student (English) and project assistant
• Ali McGhee, PhD student (English) and project assistant
• Nikolaus Wasmoen, PhD student (English) and project assistant

The Blake Archive team at the University of Rochester, affectionately known as BAND (Blake Archive, Northern Division) collaboratively authored this document; below you’ll find accounts from several people at different points throughout the day.

---

Esther Arnold (~ 9:00 AM on April 8, 2013):

Hello from the University of Rochester’s division of the William Blake Archive, where our focus has been on creating digital editions of Blake’s manuscripts and typographical editions. I’m transcribing Blake’s unfinished illuminated manuscript of the Book of Genesis (ca. 1827-27) at the moment. I’m on my second “pass” through the manuscript and am trying to make up my mind about transcription issues that came up during the first pass.

Many of the questions I’m asking relate to the underwriting that is legible throughout Blake’s manuscript–the text Blake wrote lightly in pencil before going back through the manuscript and writing it in darker pencil and/or ink. In many cases the underwriting (when legible) appears to match, in content and position, the clearer text written over it. To avoid creating a transcription that has large blocks of deleted text that are replaced by the same text (which would happen if we used a substitution tag for the underwriting and overwriting), I am trying to explain the presence of underwriting in a general way in object notes and then transcribe it only when it differs significantly in content or location from the darker, overwritten text. I am also supplementing with line notes.

Blake’s verse numbers, which are sometimes written twice, in a slightly different position, are forcing me to make some decisions about when to transcribe underwriting. I’m trying to decide whether to just describe these instances in textual notes, or to transcribe them as best I can. Below are a few examples from object 10 of Genesis:

Here is verse number 1, in which case there’s a clear gap between what seems to be the underwritten text (transcribed in gray) and the overwritten text (in black).  The change in position between these writings seems worth transcribing at first glance:

But the two writings of verse number 3 (below) overlap. Does it make sense to transcribe both writings, especially when the transcription doesn’t quite capture what we see in the manuscript (the overlap) anyway? The fact that neither “3” looks darker than the other also raises the question of which one is the underwritten “3”. Looking back at verse number 1, I wonder if I’m assuming too much in distinguishing underwriting from overwriting.

Verse number 7 is written twice too, but more or less in the same position. If I’m only transcribing underwriting that differs significantly in content or location, maybe the double 7 shouldn’t show up in the transcription at all:

It’s hard to determine where to draw the line when it comes to transcribing these underwritten verse numbers. Right now, I’m leaning towards relying on textual notes rather than transcribing in cases like this.

---

Rachel Lee (9:00 AM on April 8, 2013)

I’m the project coordinator for the Blake Archive, which means that I facilitate workflow between the University of Rochester’s wing of the Blake Archive and the folks at the University of Carolina, Chapel Hill (where the project managers Ashley Reed and Joe Fletcher live, as well as the technical expertise and servers).

Mondays are my busiest days on the Archive. Our weekly staff meeting–with all the BAND project assistants, editor Morris Eaves, and our undergraduate intern, Duncan Graham–is on Monday mornings. During the staff meeting, each person gives an update on their respective project. The meetings are also an opportunity to pose questions to the group–and with Blake’s manuscripts, there are plenty of questions for us to grapple with week after week.

This morning started with a Team Color Code meeting at about 9:15am (more on Team Color Code below). Our current project involves Blake’s manuscript The Four Zoas; we’re at the point where checking other editions of the manuscript–to see how other editors have tackled this intricate, complex work. I had two editions with me: a photographic facsimile by Cettina Tramontano Magno and David Erdman and a more speculative, dreamy edition by Landon Dowdey (which attempts to offer a clean and coherent reading text).

To prepare for the Team Color Code meeting, which takes place just before our staff meeting in the same room, I go to the new Blake Archive office to pick up some laptops, which we’ll need for both meetings. Our meeting room–which was difficult to secure, as our building is going through major renovations–has a perfect table and a walk-in safe (!!), but no technology to speak of (aside from a TV and VCR in the corner). We used to meet in a seminar with a lovely large, wall-mounted digital display, which made it really easy to examine thorny editorial cruxes as a group.

I grab the laptops, chat briefly with Esther, and head down to our meeting room. As I wait for the Team Color Code meeting, I brush up on the editorial history of the Four Zoas.

---

Team Color Code Meeting (Rachel Lee, Hardeep Sidhu, Gabi Kirilloff) (~9:15 AM-10:00 AM on April 8, 2013)

Team Color Code is a subset of BANDmembers working on devising a new “color code” for the transcriptions displays of our electronic editions. Present members include: Hardeep Sidhu, Gabi Kirilloff, Laura Bell, and myself).

Our present task is to devise a transcription display that adequately deals with the editorial conundrum known as the Four Zoas (FZ for short). We first implemented a color code a few years ago with the publication of the first Blake Archive manuscript, An Island in the Moon. Our thinking was that it would be great to use the affordances of HTML display to register particular sorts of changes within the manuscript. Here’s a quick example from Island in the Moon.

We group related changes, such as this deletion and addition) with a substitution tag <subst> to show that a set of revisions is related.

Once we started working on the FZ, however, it became readily apparent that our color code was wholly inadequate to the task. Rather than clarifying the text (which is the primary goal of our transcriptions), the color-coded display of the FZ created even more chaos.

Hence, the need for a new design. Although we still call ourselves Team Color Code, a better name would be Team Gray Scale, as our experimental system uses gray fonts, gray bars, and text placement to show the relationship between layers of revision.

Our pre-meeting meeting today was to work through some issues that have come up in the transcription to object 4 of the manuscript. Partway through our meeting, Morris Eaves arrives with the truly gigantic edition of FZ by Bentley, which we immediately consult. We only had a few minutes to compare our transcription to Bentley’s before it was time to select a few questions/problems to bring to the whole group–hopefully to get some feedback about how we’re deciding to solve some of the transcription and display challenges that have been coming up.

---

BAND staff meeting (10:00 AM -11:00 AM on April 8, 2013)

Here are the (lightly edited) minutes to today’s staff meeting, which we post each week to our (private) Google site, where we host some of the documentation and project pages (also google docs).

In attendance: Morris Eaves, Lisa Vandenbossche, Hardeep Sidhu, Esther Arnold, Duncan Graham, Gabi Kirilloff, Rachel Lee

Morris Eaves

Bentley’s GIGANTIC edition of the Four Zoas

• Morris just picked up Bentley’s edition of the Four Zoas
• The UR library had to purchase us a copy, which they kindly–and quickly–did
• This will be a great reference for Team Color Code
• Bentley’s edition and Keynes’ editions of Blake represent a new generation of thinking about the Four Zoas
• Keynes did his edition
• Bentley did major study of VALA alone (as his dissertation)
• Erdman did print edition of Blake
• Bentley did 2 vol. edition
• Erdman & Magno’s edition based on infrared photography focuses on the illustrations; there are no transcription

Some notable quotes

• “When it comes to scholarly tools, we got ’em!”
• “In the name of the Day of DH, we made a major print acquisition.”

Intern experiment

• This semester, the UR team is experimenting with having undergraduate interns
• Duncan Graham has been so exceptional that we want more!
• Morris is soliciting letters of interest from prospective interns
• We hope to have 1-2 undergraduate interns next semester

Team Color Code/Object 4 of Four Zoas

• We share our google document, which has notes/images of our main question and screen shots of our possible solution
• Line 9:

[Notes from our project google document, where we share proofreading notes, pose questions, and insert manuscript images to discuss.]

Object 4 line 09m: Hardeep: I like the way the new color-coded transcription handles this line. It makes sense and looks clean. Currently, that pencil line displays as a strikethrough (of “Like Sons & Daughters”). But does Blake ever underline? If so, is it possible that this isn’t a strikethrough of the ink line but that it’s an underline of the pencil? I think the current transcription is right, but I thought I’d mention it.

Gabi: In line 4.09 Justin’s transcription has the text in pencil as written over the text in pen:

In the previous transcription, the overwritten pencil text was treated separately as its own line. In our transcription at the moment, I mistakenly have it as underwritten text:

The same is true of the strikethrough in pencil. I’m not sure how to treat “Fairies of Albion” as the top layer – it seems wrong to have “Fairies of Albion afterwards Gods of the Heathen” as the top layer visible in the reading layer, and then to have “Like Sons & Daughters, Daughter of Beulah Sing” be visible in the dropdown as underwritten text. Should I treat the two as separate lines? Is there a way to treat them as one line, showing them both in the reading layer, but still indicate that “Fairies of Albion…” is an addition? This goes back to the addition at the level of a line versus at the level of the object issue.

If Justin’s transcription is correct, and the text and line in pencil are in fact revisions, then the following seems consistent with what we’ve been doing so far:

Here’s how we dealt with a similar issue from object 3:

If, upon looking at other transcriptions, we determine that the pencil is not overwriting, then perhaps we can just treat it the way we were or treat it as a separate line.

Duncan Graham

Proofreading Blake’s letters

• This week, he checked BAND’s transcription of letter 23 Aug 1799 with own his own transcription & the standard references we use (Bentley, Erdman, Keynes)
• He is continuing to proofread, and was pleased to find errors in the new letter
• In proofing, he came across an insertion (marked with a caret) that is displayed in the “old’’ color code, in which insertions display in blue text; he found this distracting. The caret makes it obvious that this is an insertion–why the blue?
• The old color code does a good job displaying more complex changes, such as substitutions. However, for consistency, we need to maintain color code in all instances. (Plus, Blake rarely uses carets.)
• As a reminder when proofing, be sure to check ALL the standard references and note in editors’ notes any differences in between them and our reading

Esther Arnold

• Started her post for DayofDH during her office hours before the meeting. It’s about verse numbers in the Genesis manuscript and the problem we call “overwriting,” which is when some text is written over other text. In Esther’s case, the overwriting is often identical between layers (that is, a “4” in ink is written over a “4” in pencil), but sometimes the two layers don’t line up. Sometimes there’s overlap, sometimes they’re far apart, etc.
• She’s having trouble coming up with guidelines to deal with overwrites–sometimes it’s hard to determine which layer is which
• She’s leaning towards discussing things in notes, unless something is obviously going on
• Overwrites are difficult to describe in useful way

Hardeep Sidhu

• Worked with Duncan on Friday
• Finished the transcription of a recent letter acquisition; still needs to fill in work info
• Started blog post; will post this week

Morris Eaves

• Andrew Jewell is editing the Willa Cather letters
• Cather wrote 3,000 letters (!!); all of a sudden, our letter project seems totally doable
• Cather letter project: first publishing plain text version, then digital versions
• General discussion: How do you collect/archive emails?

Lisa Vandenbossche

• Continuing to proofread letters that have been transcribed; started next batch
• First publication of letters
• Should be published today!
• Once published, Esther will crosspost an announcement to the UR Eng Dept website

---

 Lisa Vandenbossche (11:20 AM  on April 8, 2013)

I am proofreading some of the earlier Blake letters that were not a part of the set that were just published (will be published?) this week. These are letters that we are hoping to publish sometime this summer.

Someone has already gone through and transcribed the letters, so I am now checking that transcription against the original letter (to make sure I agree with the reading) and against the standard sources that the Archive uses for the letters. Our standard sources are: William Blake’s Writings, Ed. G.E. Bentley Jr.; The Complete Poetry and Prose of William Blake, Ed. David Erdman; The Letters of William Blake, Ed. Geoffrey Keynes. Once we have completed a transcription, we then look at how our transcription differs (if at all) from these standard readings. Any differences will be marked in an editor’s note in the Archive.

I am currently working on one of the few letters that we have that Blake did not write. This is instead a letter that was written to Blake on December 18, 1808 by his friend and sometimes patron George Cumberland. The letter concerns works that Cumberland had recently sent to Blake for his opinion on them. The letter itself is fairly straightforward, but included in the same packet (presumably to cut down on mailing costs) is a letter that Cumberland wrote to his son concerning a piece that he was to deliver to Blake and wait for feedback on. Blake was to then respond to Cumberland’s letter and send his response with George’s son while he waited. This part of the packet has quite a bit of overwriting and blotting, which we have captured in the transcription with color coded text. This is what the desktop looks like as I compare them:

---

Ali McGhee (1:04 PM on April 8, 2013)

I’m creating a transcription for Poetical Sketches, the first typographic edition that the Archive will publish. While it initially seemed that transcribing an already-typed work would be a simple task, there have been many unforeseen and unique challenges that have kept me busy! The work should be completed and available for perusal sometime this Spring.

The genesis and background of Poetical Sketches has proven fascinating. Although not among Blake’s best-known works, it provides readers with a look at the artist’s early influences and sheds light on the evolution of his poetic voice. It also contains a handful of frequently-anthologized poems, like the “Songs” “How Sweet I Roam’d from Field to Field” (the first Blake poem I remember reading) and “I Love the Jocund Dance” (objects 12 and 15). It includes playful, entertaining works, like the overwrought, high Gothic “Fair Elenor,” as well as complex, ironic musings on British history and nationalism, like King Edward the Third (object 31).

Poetical Sketches was printed in 1783, and 23 copies have currently been identified. The work, which was never formally published, exists only in the proof stage. Blake, who gave many of the copies as gifts, made minor alterations in several and left others untouched. The first copy the Archive will publish is Copy C. Copy C, which has minor changes, will provide the model for all other copies. Once the transcription is finished, making the others available will be a much faster process. Below is an example of a minor change in Copy C from 6.14, where “in” has been struck through:

While these small variations are pretty easy to deal with, the biggest challenges for this transcription have come with questions of formatting and spacing. Transcribing an already-printed work raises issues about how best to represent the page while adhering to Archive standards and practices. One issue I ran into early was representing initials, larger letters that come at the beginning of each new work in the Sketches. An example, below, is from “To The Evening Star” (object 7.4):

 

Currently, the Archive does not represent initials, so I had to fiddle with spacing and formatting to most accurately reflect the printed page while complying with the Archive’s goal of providing searchable, simple XML transcriptions. My final decision was to represent the increased spacing between the letters of the first word, while providing a <choice> tag to enable people who might search for “Thou.” I then decided on a standard indentation of 3 spaces for all lines following lines that begin with an initial:

The resulting transcription looks like this:

There are many little decisions like this that I’ve had to make while working on the transcription. Issues of formatting are always on my mind. Particularly challenging works have included the drama King Edward the Third, which requires some serious thought about spacing. Here’s the beginning of the work (object 31):

Transcription of pages like this requires a lot of estimation of space length, followed by multiple rounds of re-tweaking. Andrea Everett, another BAND project assistant, has been working with me on the transcription, and it’s really helpful to have two sets of eyes for pages like this one in order to ensure that our transcription is as accurate as possible. This page is still in progress, as I ran into a minor problem with the display that should be simple to resolve (famous last words!).

The advantages of encoding and publishing a typographic work include creating a searchable transcription for Archive users and providing easy access for students, researchers, and teachers of Blake’s works. Applying our rigorous editorial standards to the work has revealed new information that will prove illuminating for people interested in Blake’s artistic development. This will be an important addition to the Archive as we continue to evolve our goals and take on more of Blake’s works. As for me, Poetical Sketches has given me a chance to familiarize myself with this rich early work while getting into some of the nitty-gritty of XML transcription.

---

Nikolaus Wasmoen (2:10 PM on April 8, 2013)

My name is Nikolaus Wasmoen, and I am a PhD candidate who’s been with the Blake Archive group here at Rochester (“BAND” for “Blake Archive, Northern Division”) since 2010, when I began as a project assistant working mainly on our electronic edition of Blake’s letters. We are just about to publish the first group of these letters, with more coming soon thereafter (see below). As a new assistant in 2010, I joined during the early stages of transcription and metadata description of the letters, which coincided with the late-stage preparations for publication of the Blake Archive edition of An Island in the Moon, which was the first work to appear in the Manuscripts and Typographic Works section of the archive. While the letters have been in process, BAND and the other teams in the Blake Archive have published a handful of manuscript works, to which, as Ali explains, we are soon looking to add some typographical ones as well.

Building on the new manuscript transcription tag set that was developed for these letters, we have continued to adapt and, in some cases, expand our metadata descriptions, transcription tag sets, and various levels of editorial notes and apparatus to fit the “new” features (new to us, at least) that we encountered in the other manuscript and typographical works we’ve been preparing since Island. What we’ve followed tradition in grouping together as “letters,” for instance, is hardly a stable definition of these often hastily composed and transitory documents, which are not infrequently much less similar to each other as physical objects than other groups of objects brought together within Blake’s “works” as an artist and commercial engraver. This makes it a constant negotiation between efforts to maintain consistency across our editorial treatments in the archive, and the need to hone our approaches to various peculiarities in the ever-expanding archive of Blake’s works we are editing and publishing.

More on the letters project here at BAND:

As we are just about to announce the publication of the first group of letters in our edition, it seems like a good moment to take a look at the letter project as a whole, and the future installments we are looking forward to bringing out in this series alongside some of the other manuscript and typographical works we are preparing.

At the moment, we have photography for 62 of Blake’s letters (including one letter written shortly after his death by his friend George Cumberland, to another friend, Samuel Palmer, concerning Blake’s last days and funeral). For a variety of reasons, we are publishing these letters in groups in reverse chronological order, beginning with the letters Blake wrote toward the end of his life, 1825-1827, and proceeding to earlier periods. We are also continuing to add to our collection of letters as new objects are discovered or made available by the Blake Archive’s contributing collections.

The first group, to be announced any day now, will consist of 21 letters from 1825 to 1827. By the end of the semester, we are aiming to finish a further 5 letters, from 1808-1824, as well as an additional 3 letters in the 1825-1827 range that we acquired too late to include in the first group. Over the summer, we’ll then be looking to complete a group of 23 manuscript letters from 1791-1807, to which we will add a special group of 10 typographical letters that exist only through posthumous transcriptions of Blake’s original manuscripts (in a biography by Alexander Gilchrist, edited and expanded by his widow, the Rossettis and members of their circle in the later 1800s).

With so many other things in the pipeline for publication, these groups may be combined or expanded for final publication, but we’re excited to be able to finally share the first group of these objects in the Blake Archive, and look forward to continuing to expand this part of the collection.

---

 Laura Bell (5.11 PM on April 8, 2013)

This afternoon I’m working on two tasks for BAND, a transcription of A Descriptive Catalogue and a project to make our transcription standards available publicly.

I’ve been working on the transcription guidelines with our Project Coordinator, Rachel Lee. As we both have office hours at different times, we decided to create a google doc that we could both access as the main workspace for the project. In this way, we are always able to see what kind of progress we’ve both made, and can even use the document to ask each other questions and archive conversations about the project all in one place. These kind of collaborative tools are a really important part of our workflow at BAND. The transcription guidelines currently exist as a word document, and so our first task was to decide how to organize and format the information so that it could be accessible and easy to read online. Now we are converting it so that it will be ready for publication.

Our formatted google doc:

The .php file:

                                          

A Descriptive Catalogue is a typographical edition of a prospectus that Blake wrote for an exhibition of his own works. As well as a detailed description of the pieces that were exhibited (some of which are now lost), the Catalogue includes a heated discussion on European art and an analysis of Chaucer’s Canterbury Tales.

As a transcription project, A Descriptive Catalogue is especially interesting for the kinds of questions that it throws up as a typographical work, just as Ali Mcghee has discussed in her post about Poetical Sketches. For example, one problem I keep encountering is how to deal with words that the printer has split over two separate pages:

In this case, the word “himself” is divided between pages 24 and 25. The Blake Archive organizes works into objects; in this case one object is one single page, so what is the best way to encode the transcription, which looks like this:

I finally decided to add a <choice> tag to “him-” on page 25, thus allowing people to find the word, “himself” if they searched for it. I decided not to add an additional <choice> tag to ”self” as it suggests that the word “himself” appears in full on both objects and furthermore, the transcription and the image of the page make it clear that the word has been hyphenated. Now that I have decided on a solution to this issue, I’m working my way back through the transcription in order to make sure that I deal with all similar instances in the same way.