Pro-tip: if a manuscript is catalogued as containing ‘extracts from Isidore of Seville’s Etymologies’, check if it really is that.

Years ago, in a couple of related manuscripts in the Biblioteca Laurenziana in Florence, I found that a text described as ‘mostly from Isidore’ was actually a Lombard revision of Irish computistical and cosmological material from 747 (= Computus Amiatinus). Very little of it was actually Isidorian.[1] Around the same time, Cinzia Grifoni discovered that a text in St Gall catalogued as ‘excerpts from Isidore’ was actually a copy of the rare Recension 3 of Pseudo-Methodius’s Revelations. See: it is worth checking.

And so, in lockdown and my interest in watching The Phantom Menace with 5 year-old waning fast, I idly thought I would check some more ‘extracts from Isidore’s Etymologies’ in the famous computistical manuscript Cologne Dombibliothek 83-II. It is a great manuscript. It was written in or shortly after 805 for the library of Archbishop Hildebold of Cologne (d. 819), Charlemagne’s one-time palace chaplain. It may have been started earlier as there is a unique addition to Isidore’s chronicle that identifies the year as 798 – intriguingly as the year ‘when missi came from Greece to hand over imperium to [Charles]’.[2] The manuscript has been studied thoroughly by Arno Borst, Brigitte Englisch, Immo Warntjes, Donald Bullough, and Anton van Euw, and it has been online since 2005, so there should be little new to find.

So, those ‘extracts from Isidore’. In 1874 Jaffé and Wattenbach published their catalogue of the manuscripts of Cologne’s Dombibliothek. They noted that, at 15r to 36v, there was a text that began ‘Prologus sancti Esidori de numero’, which they said was extracts from Isidore’s Etymologiae, book III.[3] If you don’t know book III off by heart, it is the bit when Isidore outlines the mathematical disciplines of the quadrivium. Jaffé and Wattenbach gave no more details about the excerpts. In 1910 my great St Andrews predecessor Wallace Lindsay, editing the Etymologiae, suggested that it was not the Etymologiae but actually a witness to Isidore’s De numeris.[4] Sadly he had not actually seen the manuscript, as August Eduard Anspach showed in his lengthy 1912 illustration that the Cologne text was indeed mostly assembled out of the Etymologiae.[5] But even Anspach was only really concerned to establish the Isidorian content, and not the nature of the material with which it was mixed, so he didn’t give much information on the non-Isidorian passages. Only in 1939 were further parallels with insular computi, especially the Divisions of Time, noted by Charles Jones, although not extensively.[6]

Renewed attention in the 1990s started to make better sense of the text – at least after a new cataloguing effort in 1993 had only repeated Jaffé and Wattenbach for contents.[7] Fresh analysis by Diane Anderson in 1995[8] and for the exhibition Glaube und Wissen im Mittelalter in 1998 provided a new breakdown:

(1) 15r-20r = excerpts from Isidore, Etymologies III

(2) 20r-26v = a version of the popular Divisions of Time text, partly based on Isidore, and presented as a coherent extension of our (1)

(3) 26v-28v = extracts on astronomy from Isidore, also presented as a coherent extension

(4) 29r-36v = a computus focusing on lunar calculations, also excerpted from Isidore, which was probably considered a separate work from (1)-(3) as it had its own list of chapters.

The most recent reference to this material, treating (1)-(3) as a coherent text, is in Jacopo Bisagni’s excellent 2019 Kathleen Hughes Memorial Lecture. Bisagni is working on a new thorough analysis of the Divisions of Time, because there are many many versions of it, and no one really has any sense of how they relate to each other. He designates the Cologne text as containing an epitome recension of the Divisions of Time (Enlarged Version 1) – where, confusingly, the only other witness to this version of the text is mostly completely different (Laon BM 422 which was also long designated ‘extracts from Isidore’).[9] Bisagni’s principal concern thus far, however, has been that other witness, so he has not had much to say about the Cologne text.

One thing he does note about the Cologne text is that it contains passages on the largest units of time that were not included in the copy in Laon. What he doesn’t mention is that they contain a reference to the year it was composed. On f. 26v it states ‘from the nativity up to this year [there are] 797 [years and] from the beginning of the world [there are] 5,997’ (a nativitate Christi usque hunc annum dccxcvii, ab inicio mundi Vdccccxcvii). This only imperfectly fits the 798 date of the Isidore Addition earlier in the manuscript, which is written in a different hand with different orthography anyway. But more importantly, no one has ever mentioned that this dating clause exists. Because virtually no one has ever read it.[10] Because people thought it was just ‘extracts from Isidore’ and not a coherent and deliberate text written (possibly) in the circles of Hildebold in 797.

Why is this interesting? Because we start to move from seeing that part of the manuscript as a boring old partial witness to Isidore’s work, to seeing it as a more deliberate attempt to appropriate Isidore and the lost Divisions of Time (Enlarged Version 1) to create something new at the end of the eighth century. We could, for instance, read it alongside that Isidore’s Chronicle Addition. That is famously concerned with how Scripture says that the timing of the End of the World is unknowable – in part a challenge to anyone who thought that the world might end in the year 800 or so, calculated to be a round 6000 years after Creation. The Etymologies–Divisions hybrid looks in that context to an understanding of the world that is mathematical and rational, not boundlessly arbitrary and mystical. Learning is key – and indeed the Addition famously ends with a call to read and count better to avoid falling for false beliefs… and the Etymologies-Divisions hybrid is the very next text in the manuscript.

If you’ve ever read about proto-sciences and apocalypse under Charlemagne, you know that it was like that, and this 797 text is therefore no surprise. It is, however, always worth stressing that one of the reasons that early medieval science has a bad reputation is because much of the evidence has been and remains largely ignored or unanalysed. There is much work to be done.

[1] The article discussing this is nearly, nearly done – promise.

[2] This is now edited in vol 2 of Arno Borst’s Schriften zur Komputistik (Hanover, 2006).

[3] P. Jaffé and W. Wattenbach, Ecclesiae Metropolitanae Coloniensis Codices Manuscripti (Berlin, 1874), p. 29.

[4] W. Lindsay, ‘Isidorus “De numeris”’, Berliner Philologische Wochenschrift 30 (1910), 1144.

[5] A. E. Anspach, ‘Isidori Hispalensis “Institutionum disciplinae”’, Rheinisches Museum für Philologie 67 (1912), 556-68 at 563-68.

[6] See C. W. Jones, Bedae Pseudepigrapha (Cambridge, MA, 1939), 48-51.

[7] Handschriftenzensus Rheinland, no. 1046, ed. G. Gattermann with H. Finger & M. Riethmuller (Düsseldorf, 1993), 1. 620.

[8] Diane Warne Anderson: The Medieval Manuscripts of the Cologne Cathedral Library. Vol I, Ms. 1-100 (Collegeville, 1995).

[9] J. Bisagni, ‘From atoms to the cosmos: the Irish tradition of the division of time in the Middle Ages’, Kathleen Hughes Memorial Lecture 18 (2020), 28 n. 72.

[10] I am sure someone has read it and I have missed the reference, being in lockdown without a library.