Category Archives: Medieval libraries

Computing through the Ages

It may seem unlikely now, but I was once a pioneer. Thirty years ago (gulp) I was part of a team creating and operating a unique computer program which would register around 600 first year A level students together with their subject choices, allocate the students efficiently to the timetabled classes and print out an individual timetable for each new student. We had two days to input the data, a day to perform the allocations and a weekend to print them out before the students turned up to start their courses. Privately, the weekend was there to allow us to do the job manually if it all failed.

The program worked on a BBC-B which looked like this:


Figure 1 BBC Computer 1980’s


It worked. And a process that had been done manually with lots of bits of paper with crossings out and erasures changed for ever and the college began to realise that there was a role for computers in educational administration. Job done.

2018 is a different world but I’m still keen on using digital methodologies to save work. And I’m still in awe of what these things can do.

So, when I started the PhD, an early decision was to see if I could use a database as part of my research methodology. Earlier this year I pulled this off and became the proud creator of a database of texts that were in the ninth century library catalogues in the medieval monasteries of St Gall, Reichenau and Murbach.

The originals did not look all that promising:


Figure 2 Page from St. Gall catalogue Cod Sang 271


But these are precious resources. Only five institutions still retain their ninth century library lists, the other two being Lorsch and the cathedral of Cologne and and I think my three are the only ones to be digitised so far. My plan was to present quantitative data on the types and origins of the texts that existed in the centres and use this to get a picture of the reading available to the monks.

At the beginning I was interested in the technology: whether to use Excel or Access, how to streamline input and how to create visuals from the data. But, unlike the 1988 project, it was not the technology that provided the challenge, it was the data itself. We did have some data related problems back then – I recall a fuss about a student with a long double-barrelled name which would not fit into the space allowed for it and had to be abbreviated – but in essence we were taking a functioning manual system and speeding it up. It was the computing itself that challenged us.

Nowadays, the tables have turned. The computing side is well established. There is an app, one that other people use and for which there are training courses. The work lay in structuring the information I had. What kind of text was an ‘Enchiridion’? Is a ‘commentary’ different from an ‘exegesis’? And so on. I resolved the dilemma as best I could by recording decisions and establishing my own definitions as I went.

I did OK and at the PhD Continuation Review earlier this year, the panel was pleased at the use of quantitative data and suggested this might be exploited further as the work progressed. At this point I decided that I needed to look into the whole issue more carefully and that was when I discovered Mark Merry’s handbook ‘Designing Databases for Historical Research’, which istoryhiis generously available online for free. You can find it here:

Because it is tailored to the needs of historians, it is remarkably useful in addressing the kind of problems we did not have with the 1988 project, but which do crop up a lot in my current research, for example, accounting for changes in the names of geographical entities or designing a database before you know what exactly your research is going to turn up.

What struck me though is that as well as tailoring advice to the needs of historians, there might be sense in tailoring a version for medievalists. The difference is not merely that a medieval information needs more of what Merry calls ‘standardisation’ – enabling the system to know that two ways of spelling a name actually refer to the same chap, for instance. Nor is it to do with all that Latin and the wobbly writing. It is that medievalists are dealing with a very different mindset and, inevitably, using sources which may be patchy and/or unreliable. And crucially, that we are trying to adapt what was often a very organic and unsystematic storage system to suit a totally alien technology.




Leave a comment

Filed under Learning Latin Late, Medieval libraries, Monastic life, PhD thoughts

The tale of three panthers and their teeth

st gall pantherFor long and weary I have kept a postcard on my wall which I acquired on a visit to St Gall several years ago. It comes from a ninth century manuscript of a well-known sixth century educational work by Cassiodorus called ‘The Institutiones’, which set out the plan he had for the education of the monks at the monastery he founded at Vivarium and it is only one of several diagrams in the work.[1] I love it and I have always called it ‘the panther’ although the official description is ‘a panther-like animal’. His function is to show the four divisions of the subject of Mathematics in the traditional ‘Seven Liberal Arts’, a kind of medieval syllabus for monks. It is part of the Quadrivium, the higher level of learning to be undertaken once the basics of Grammar had been mastered. The four divisions are here represented by the four feet of the panther, each labelled appropriately.

But it’s much more than that. Time and effort have been put into the drawing. The panther has a collar and lead and what seems to be a coat and if you are so inclined, it would be very easy to imagine that he is smiling at us. It would fit very happily into a school textbook, even a story book for young children. Did they do picture books in the ninth century?   And what do we know about the presence of panthers in the Carolingian Empire? Not a lot it seems. So I bought the postcard and kept it.

Some years later I started my PhD, which uses a group of ninth century manuscripts from the Abbey of St Gall and its close neighbour, the abbey of Reichenau. Cassiodorus cropped up again and, to my surprise, so did the panther.[2]

panther aug perg 241

A slightly posher version, with red detailing and the various titles of the four divisions in boxes hanging from cords attached to his feet, but still with a collar (albeit plain not spotted) and lead and a differently patterned back. His tail is also different: at Reichenau it forms a jaunty ‘S’ shape, at St Gall it is a passive downward slope, and he has two ears, one pointed and one rounded like the St Gall one. What is really striking however, is that this one is showing his teeth. If you look closely, both beasts have three teeth, but the way they are arranged is different – cheery grin for St Gall, threatening growl for Reichenau.  So I went looking for enlightenment and found Michael Gorman.

Michael Gorman does not apparently share my love of the panther but he does know about diagrams in Cassiodorus manuscripts. 3] Sadly, he hasn’t managed  to track back to the original but he has tracked the diagrams back to the eighth century, to what he think is the archetypal manuscript, namely Bamberg Patr.61, the text of which was used for Mynors’ printed edition in 1937.[4] The gap between the printed versions and the manuscripts becomes clear at this point. The manuscript has a total of 37 diagrams but Mynors only reproduced eight of them in his book. The text thus quickly became disassociated from the diagrams. Gorman points out quite rightly that this omission is actually quite a serious one: my panther is not an optional extra, he carries the names of the four sub-divisions and if he is removed, they are too.  But it is only now, when we can all look at lots of original manuscripts without leaving our desks, that we are rediscovering the diagrams. [5]


And I rediscovered another panther in Harley 2637, another ninth century manuscript of the Institutiones, which is readily available online from the British Library. It originates from either Western Germany or Eastern France and was first recorded at Cues on the Moselle. Whilst it cannot be traced back to a Carolingian centre, its panther clearly relates to the two examples from Lake Constance.

harley 2637 panther

The main difference between this panther and the other two is the increased use of colour, which seems to take the beast into the realms of fantasy and magic. The collar and lead here are important enough to split the heading into ‘Divisioma’ and ‘thematicae’, whereas in the other two diagrams the heading stands aloof, with a conventional gap between ‘divisio’ and ‘mathematicae’. The tail has the Reichenau curve, but with a tuft at the end which is missing both at Reichenau and St Gall, as if the model for both the Carolingian versions was somehow cut off at the end. But other features are the same. The three teeth are Reichenau style, but somehow look even meaner because the mouth is smaller, the boxes around the four headings are also Reichenau style. It has one ear, like St Gall, but it is pointed like Reichenau. The coat, if such it is, is divided from the lower body by a straight horizontal line, as is the case at Reichenau, whereas at St Gall, it curves down between the back legs. The Harley tail is plain, the Reichenau tail is coloured red and the St Gall tail has a zigzag pattern. There is clearly a relationship between them all and each has at least one unique element, but the St Gall exemplar is clearly distinctive.

These diagrams matter because they imply a different kind of reading from what we expect of ninth century abbeys.  You cannot read a panther aloud,  although you could perhaps read the manuscript alongside a student.  But this is not for use with a group.  These texts were intended for private reading.

That conclusion alone is worth bothering with.  But there is more.  The latest bit of the jigsaw comes from Sven Meeder’s contribution to the recent book of studies in honour of Rosamund Mckitterick.[6] Meeder points out that Bamberg Patr.61 is a ‘sister manuscript’ to St Gall 855, the original object of my interest, and uses this connection to support the view that the two manuscripts had a common ancestor, probably from either Monte Cassino or somewhere in northern Italy. Further, he demonstrates that a relationship exists between the St Gall manuscripts containing Mallius Theodorus’ ‘De metris’ and other multi-text manuscripts containing the same text which have their origin in centres south and north of the Alps. Checking out these manuscripts for evidence of panthers may well be worthwhile.

For now, the lesson for me is that although much energy and ink has been devoted to the investigation of textual links between manuscripts and this is right and proper, we might still, from time to time, spare a thought for the guys who drew the panthers.

[1] Cod Sang 855, p.276

[2] Aug Perg 241 p53

[3] M. Gorman ‘The Diagrams in the Oldest Manuscripts of Cassiodorus’ Insitutione’ in Revue Benedictine vol 110 Issue 1-2 January 2000 pp27 – 41

[4] Ibid. p28

[5] Ibid. p29 Sadly we don’t always manage to find the right places to read them. Mynors apparently pointed out that the diagrams existed in a total of six manuscripts, the ones under discussion here and  Bamberg Patr. 61, Paris Mazarine 660 and Berlin Phillips 1737. These are the ones I have not yet got access to.

[6] ‘Monte Cassino’s Network of Knowledge: The Earliest Manuscript Evidence’ in ‘Writing the Medieval West’ ed. Elina Screen, Charles West. Cambridge 2018 pp 141 – 142



Leave a comment

Filed under Learning Matters in the Medieval World, Medieval libraries, PhD thoughts

A dummie at the court of King Access


Dummies cover

I have blogged before on the joy of ninth century library catalogues, and in particular on the joy of reading ninth century librarians’ comments. Now for the hard stuff. In between chasing missing books, these guys did a fantastic job of listing what they had. For St Gall, there is a lot of fun to be had in matching the lists to the extant stock, but for Reichenau, just a boat ride away, not only do we not have the manuscript versions of the catalogues, we also don’t have any Reichenau manuscripts at Reichenau. Happily, and thanks to a diligent German scholar in 1918, called Paul Lehmann, we do have a printed copy of several ninth century catalogues, based on copies made of contemporary manuscripts in the eighteenth century. And Lehmann did track quite a few of the manuscripts mentioned in it amongst the stock transferred to Karlsruhe after the dissolution of the abbey so they can’t all be imaginary. But for the time being, it’s the catalogues that are exercising my mind. And what to do with them that might contribute to an understanding of the textual culture that the monks were living and working in.

My mind drifted to the idea of visualisations, thence to digitisation and – and this is a long leap – to the idea of a database of texts. I used to use databases for my work, but they came ready-made. Not only had someone else done the key to disc bit, but someone else had dreamt up the structure required. And the database programs available to me now are not the ones we created in 1999.

And so I became a dummie. And bought a book that showed me how.  In its own way.


And I have to admit I’m still a dummie, but I am a wiser dummie because I have a better idea of why it’s hard.

I thought it was hard because of the technology, but it wasn’t. I thought it was hard because of the circularity of the process of deciding which fields to include before you have entered all 700 records because it is only after you have entered those 700 records that you know which fields are needed and by then changing the fields is problematic. But it wasn’t really that either.

I did the right things. I started with apparently uncontroversial stuff like ‘Author’, ‘Title’, ‘approximate date’ and ‘Title of catalogue’. I spent rather more time and experimentation on the fields which were to categorise the type and content of the text concerned: ‘Letters’, ‘Sermons’, ‘Biography’ for example, as well as ‘Patristic’, ‘Grammar’ and ‘Astronomy’. I did quite a bit of sampling and refining. I came up with some sort of system to ensure consistency and found the kind of fixes that are required when ninth century monastic librarians meet 21st century technology. I learned quite quickly, for example, that ‘Author?’ is not a simple question if the title is ‘the Book of Genesis’ – I presume people have spent whole PhDs on less. Given that that is not an option, the choice is between leaving the field blank or simply entering ‘BIBLE’. For a couple more days I wrestled with this kind of thing, finding out along the way how to add new fields and build in a checking system at the end so that entries were consistent. And then I took a day or so off. And then I tried to enter 700 records.


And when it came to using the fields, I realised that I had the thing upside down. Being a dummie at setting up a database is easily cured by buying a yellow book. Being a dummie at describing ninth century texts will take a lot longer to fix. It took the plan of setting up a database to bring it home to me that I need to know much more about these texts and about the authors before I analyse them. I need to decide for example, which authors are ‘patristic’ because the database requires me to say something more definite than ‘This catalogue includes several patristic authors’. I need to find field labels that fit the historical context. Trying to split ‘Theology’ from ‘Philosophy’ is not going to work so I will put them together but can I lump Augustine’s biographical texts in with hagiography? Is Rhetoric part of Grammar? What is the difference between a commentary and an exposition if you only have the catalogue title? Is it possible to make sensible use of a record that just says ‘Alexandri libri’? If you have a title such as ‘de libero arbitrio’ does that mean it’s the one by Augustine that you already have or might someone else have used the same title?

I once heard someone who works a lot on digitising historical records say that the people creating the digital resource ended up learning more than the users of the resource ever could. That seems to be true here.  I’m learning a lot about what these monks read. But I still hope to have some visualisations by September.

Leave a comment

Filed under Learning Matters in the Medieval World, Medieval libraries, PhD thoughts