A Newly-discovered Medieval Legal Manuscript
Guest blog by Ross Macdonald
Readers with an interest in undiscovered antique legal texts will be fascinated by a recent report by scholars from Northwestern University, Illinois into a medieval law text discovered in the binding of a 16th century volume. (A summary article is at https://www.livescience.com/59925-imaging-reveals-medieval-text.html and the full article, which appears in (2017) Analytica Chimica Acta, at http://www.sciencedirect.com/science/article/pii/S0003267017307201?via=ihub ; henceforth “Analytica”). The finding brings to mind the techniques used at Vindolanda, on tablets from London’s Walbrook, and carbonised scrolls from Herculaneum. As such it has received press coverage, some of it – in the way of second-hand press coverage – going breathlessly beyond what was actually said (“a sixth century manuscript”). This blog briefly considers the legal significance of the text deciphered.
First, though, the document and the technology applied to it (with apologies for any mis-statements arising from my effort to summarise). This involves a printed (1537) copy of Hesiod’s Works and Days which Northwestern’s Library acquired in 1870. It has now been found that the volume incorporated an earlier manuscript legal text (which turns out to be from Justinian’s Institutes): not, as in some cases, as strips of padding in the spine, but re-using two parchment pages to cover the endboards. In the process the pages were treated to remove most of the existing writing, but some survived partly because the ink had “burnt” into the material; and the pages had been trimmed heavily at the sides and corners to fit the boards. The research team set out to identify the text if possible, by imaging it in different spectra. Their achievement, which is remarkable, has been to make most of of the writing – apart from some of the smallest script – readable. (It should be said that many passages only make sense once their legal source is identified and compared with other editions; information which was not available to the research team. I found the editions from Venice (1618) and Lyon (1627) – the first available online at https://babel.hathitrust.org/cgi/pt?id=ucm.5324323553;view=1up;seq=171 , the second at https://droitromain.univ-grenoble-alpes.fr/ – particularly useful in that respect: both retain the Vulgate text but print it in extended form – the “Analytica” like many medieval versions contracts words very heavily – and both add extra commentary to the standard medieval Gloss.)
So, the point of the project was partly to test the potential of the team’s methods for future use, and in that they have been brilliantly successful.
The legal significance of the texts is, therefore, very much secondary. How were the passages to be identified? They consist of central texts surrounded by marginal glosses – one margin having been cropped away from each page – and various interlineations by later users. Fortunately a (relatively) easily-readable passage in the gloss on one page discussed Adam’s marriage status (or lack of it under canon law), and the text could be identified as Book 1.10.4-5 De Nuptiis of the Institutes. This led the team to postulate that the manuscript had been compiled specifically for educating canon lawyers; that it represented a relatively early stage in textual development – the glosses being fragmentary; and, going further, that it may have been discarded when changes in canon law made the old text obsolete.
While possible, that is not the most obvious explanation. There are two crucial points. First, the other page covers Institutes 1.12.4-5 Quibus Modis Ius Potestatis Solvitur (ie loss of paternal authority, particularly by military service or capture). This has nothing to do with canon law; indeed this page has more notes by users than the other; and the two pages are not contiguous. So the manuscript need not have been prepared in a specifically canonist teaching context. Secondly, the gloss (so far as I can make out) is not fragmentary, except insofar as all legal glosses were – though by looking only at “Analytica” one admittedly can hardly tell, words being so highly abbreviated and lacking most of the helpful markers (indexing by letter, line breaks,etc) seen in many editions. At any rate, comparison with print editions suggests that all the expected gloss words are present (though, as I note later, not all necessarily in the right order). In short, the manuscript looks like a standard edition of the Institutes with the usual Glossa Ordinaria.
So it seems likely that the source manuscript was a standard “pecia” style copy of part or all of Book 1, perhaps commissioned by a student (not being a palaeographer, I make no comment on the likely date). With the era of print editions (often including more extensive critical textual apparatus) there would be ample reason to dispose of a worn old manuscript.
One might add that the gloss layout seems technically rather inexpert: the gloss does not match up with the adjacent Institutes text (some appears to relate to text which would have been on a previous page); there is not the “bespoke” page-by-page adjustment of text layout which is common at least in printed editions; and there is a significant error in the copying, which I now turn to.
The page on Institutes 1.12.4 contains a long interlineation near the foot, in the gap between the central text and the marginal gloss. (While the image embedded in the authors’ livescience.com blog is just readable – enough at least to identify the text – the version embedded in the “Analytica” article itself has slightly poorer resolution.) It is, compared with the very small and now-illegible other notes on that page, fairly firmly written. The passage puzzled me, but it turns out – when “Analytica” is compared with printed editions – to be a misplaced marginal gloss: one which should have belonged in the (now-lost) left-hand margin but had been missed out by the copyist and added in at some later date. That sort of error is common in legal manuscripts: thanks to the abbreviations, and the frequent repetition of important words, it was easy for a copyist to jump a couple of lines by mistake. The “Analytica” gloss actually differs slightly from that in printed editions: they read “bestiis. Deducendis forte, vel etiam cum comederet cum eis, vel alia bestialia faceret”; “Analytica” seems to have “comederent… facerent” but again that sort of variation was common, as the “n” was often skipped in abbreviation and the associated “macron” (ē) might be mistakenly omitted or, conversely, extended in full form thus altering the tense.
As for the other brief notes in smaller script, only a few letters can be made out. They seem to include additional citations – with the occasional “FF” sign which denotes passages from the Digest – but I could not be sure.
At least two other unusual variants in the text are worth noting. First, in Institutes 1.12.4, at (in the Birks/Macleod translation) “to release a son from the bonds of his authority”, Krueger has “relaxare”. The more common late 15th century and 16th century reading was “liberare”. But Cujas (as the Lyon 1627 edition notes) pointed out that “omnes veteres codices” had “relaxare”. This, too, is the “Analytica” reading, which may perhaps suggest a relatively early manuscript date.
The other relates to Institutes 1.10.4, in Krueger “Duorum autem… liberi… iungi possunt” (ie can marry). The common 15th/16th century version was similar, “coniungi”. But there was another possible reading, discussed at length by Cujas (again excerpted in the Lyon 1627 edition): “non possunt”; this went back to a commentary by (or attributed to) the Theophilus who participated in Justinian’s work of compilation. And we see “non” interlineated in the “Analytica” manuscript. Since it was re-used as binding in 1537, we seem to see here a Humanist influence predating Cujas.
So, arguably, a relatively “ordinary” legal text. But its very ordinariness can make it in its way as valuable as egregious nuggets. Any addition to the corpus of medieval legal manuscripts is useful, and the potential of the Northwestern University team’s method is immensely welcome for future research.
Ross Macdonald is a doctoral candidate at Edinburgh Law School