Welcome to a Text Technologies online workshop on recovering text in manuscript images with Professor Bill Endres this Friday!
The workshop will present some of the techniques Bill wrote about in his new book, Digitizing Medieval Manuscripts, which can help recover hard to read text in manuscripts. These techniques can be used on images of different resolutions and quality, even those taken with camera phones.
Bill will do a basic introduction to ImageJ and show us how to divide images, apply alternative LUTs to increase contrast, split an RGB image and how to use alternative color spaces for recoveries. The plan is to have a 60-90-minute hands-on session with Bill showing techniques and the participants trying them out on their own computers.
To participate please register here or contact me on twitter, @ffrwelin.
When: Friday April 3 at 10am PDT/ 12pm CDT/ 1pm EDT (in your timezone, click here)
Bill Endres is an Assistant Professor at the University of Oklahoma, who has worked extensively with the digitization of the St Chad Gospel in the Lichfield Cathedral.
Last week I was at LODLAM for the first time. LODLAM is an unconference for Linked Open Data – Libraries, Archives, Museums and this year it took place at the Getty Museum in LA. Most of what I know about LOD I know from the Digital Humanities at Oxford Summer School workshop on Linked Data I had the privilege to attend a few years ago, through funding from Riksbankens Jubileumsfond. And it was especially great to meet Terhi Nurmikko-Fuller, the convenor of the workshop and my main teacher there, again.
I have two main takeaways from LODLAM. The first was that while I thought it was a very good and interesting conference and I got quite a few useful references out of it, I don’t think LOD is going to be very central for my current project. (And now that I’ve written that, I’m scared that those might have been some famous last words.) Which is great! It helps me further narrow down what this project is about. Rob Sanderson also gave a really interesting talk touching on the intersection and differences between LOD and IIIF, which I also enjoyed and found very helpful (slides here!).
The other main thing this conference reiterated to me was the importance of community and community building for these kinds of enterprises. I know that the idea of community as infrastructure isn’t a new one but at least to me there were a few things I had not necessarily considered. Deciding on community standards takes a lot of time and effort, and needs to be allowed to do so. On the way there will be dead ends and things that don’t quite fit – and maybe what I’m realizing is that that is okay. Not only for the lessons that can be incorporated into new guidelines and best practices but for the experience in itself. (For the record, I’m notoriously anti all “the journey was the destination” talk but I GUESS sometimes the journey actually is the destination.) Sometime, we need to talk about who can put their time into community development and outreach.
Another thing I wanted to mention is that I do like the unconference format, that is, the participants propose and decided together what the sessions are going to be about and everyone has a chance to contribute in the discussion or by adding things to the shared document with notes. I think it works really well for meetings that focus on community, standards and development. At Linked Pasts IV in 2018 I was also especially impressed with how well the format worked with getting people together to form working groups.
— Agnieszka trying to drink out of a fire hose (@ffrwelin) February 3, 2020
A few links:
Science Stories – the prize winner from this year’s LODLAM challenge, telling the stories of women in science, using IIIF and LOD to aggregate and present the information. I also thought that the Dutch project Reconstructions and Observations in Archival Resources – Golden Agents was really cool with how they dealt with uncertainty in using archival materials for people and places.
Project Passage – a report on creating linked data without the users having the technical knowledge using Wikibase.
Read the Docs – a free hosting solution for technical documentation which also offers version control and multiple formats.
For my PhD (2017) I worked on a Swedish late-medieval, multitext manuscript, Cod. Holm. D 3 also called Fru Elins bok, ‘Lady Elin’s Book’. One of the things that made me choose to work on this manuscript is that it has often described as “useless” and “full of errors”. As my focus wasn’t on textual criticism and finding the textual archetype (the authors original intent) but more on the materiality of the manuscript in its context (cf. Material Philology) this manuscript was perfect to show how manuscripts that had been deemed worthless still hold valuable information.
D 3 is a paper manuscript dated to around 1488 with watermarks (Åström 2007) and the name comes from the colophon at the end of the first text of the manuscript calling upon God to protect the excellent Lady Elin who owns the book. The owner has been identified as Elin Gustavsdotter Sture, who died around 1495 (Wiktorsson 1997). She was married to a half-brother of King Karl Knutsson of Sweden and then to Erik Axelsson Tott, who was regent of the Swedish realm during two periods in the 1450s and 1460s. She was also the daughter of Märta Ulfsdotter Sparre, who also owned a manuscript, Cod. Holm. D 4a, written around 1450.
Lady Elin, “ffrw elin”, named in Cod. Holm. D 3. Photo: National Library of Sweden.
The two manuscripts are very similar – from sharing many of the same texts, some even in the same order, down to the word level. Because of the similarities many previous scholars have seen D 4a as the actual exemplar of D 3. This narrative has been argued for in (most of) the latest publications (e.g. Wollin 1997, Kornhall 1959, Olson 1921)* but there has always been information that didn’t fit.
I made a close study of D 3’s Flores och Blanzeflor, the Swedish translation of the French romance Floire et Blancheflor, which Olson (1921) was convinced was a copy of D 4a. In my dissertation I argued that that is not the case. While there are very striking similarities, D 3 also contains clear evidence of traces of another textual witness, which seems related to another branch of the stemma, a now lost manuscript. The differences are such that I found it clear that D 3’s Flores could not be a copy of D 4a’s. This conclusion is also in line with the discussions of exemplars for the other texts in D 3, but that is a whole, complicated post for another time.
*Yes, recent in this case does go back to the 1920s.
Literature
Backman, 2017. Handskriftens materialitet. Kornhall, 1959. Den fornsvenska sagan om Karl Magnus : handskrifter och texthistoria. Olson, 1921 [1956]. Flores och Blanzeflor. Wiktorsson, 1997. “On the Scribal Hands in the Manuscripts of Skemptan” in Master Golyas and Sweden : the transformation of a clerical satire : a collection of essays, eds. Ferm & Morris. Wollin, 1997. “The Lord-Abbot and his Texts” in Master Golyas and Sweden : the transformation of a clerical satire : a collection of essays, eds. Ferm & Morris.
I (Agnieszka Backman) am a Wallenberg postdoc at Stanford University working of the materialities of medieval manuscripts on digital platforms from a multimodal, socio semiotic perspective.
I finished my PhD on the materiality of a Swedish medieval manuscript, D 3 called Fru Elin’s bok ‘Lady Elin’s book’ in 2017 at Uppsala University (full text, in Swedish but abstract and summary in English). For the last two years I have worked in The Norse perception of the world project (www.uu.se/norseworld), which is mapping foreign place names in medieval Swedish and Danish vernacular texts.
Manuscript Materialities 