The manuscript Modena, Archivio Capitolare, O.I.2, which transmits the Liber legum of Lupus, is one of the most well-known manuscripts containing capitularies. Its content and the remarkable pictorial program have already been the subject of numerous studies. Nevertheless, some questions remain up to debate, among them the significant issue of its date of origin. Hypotheses on this matter span a wide range, from the mid-9th century to end of the 10th century. With this contribution, we aim to present new arguments in favour of a late dating of the manuscript.

The palaeographical classification of the manuscript is complicated due to the fact that several scribes were involved in its production. Even within what is believed to be the work of a single scribe, one encounters significant variation in the letter forms (Pohl 2001, p. 127; Golinelli 2008a, p. 229). So far, is has not been possible to assign the manuscript to a specific scriptorium, although Modena is now considered the most likely place of origin (Pagano 2001, pp. 254–257; Golinelli 2008a, pp. 23f.).

Fundamental for the dating is a booklet preserved as the final quire, containing a calendar and computistical texts that can be confidently dated to around 991. This is because the calendar is followed by an appendix in the form of a shortened Easter table (fol. 211v), which begins precisely with said year (Borst, Reichskalender 1, p. 119). Based on this evidence, but also drawing from palaeographical arguments, earlier scholars generally agreed on the 10th or 11th century as the time of origin (see the summary of earlier dating proposals since the 18th century: Russo 1980, pp. 608f.; Pagano 2001, pp. 248–250; Golinelli 2008a, pp. 21f.).

More recent studies, however, tend to date the main part of the manuscript earlier, placing its origin in the second half of the 9th century. The work of Annalisa Bracciotti, who conducted a palaeographical analysis of the manuscript in the context of her study on the Origo gentis Langobardorum (Modena O.I.2, fols. 5v–7v), is of particular note. Her assessment refers not only to the sections of the manuscript containing the Origo gentis, which was written by a different scribe than the later portion (beginning at fol. 52r), but rather applies broadly to the entire codex (Bracciotti 1998, S. 59f.). The early dating proposed on the basis of her palaeographical judgment also fits well within the context of the episcopate of Bishop Leodoin of Modena (869/871–891), as during his tenure the library in Modena appears to have possessed a significant collection of legal manuscripts. As a result, this dating has since gained wide acceptance. Examples include Pohl 2001, pp. 125f. (referring to Fornasari, Collectio, p. 251, who as early as 1966 linked the script to that of a correcting hand found in the Modena codices O.I.4 and O.I.12 and dated it to the 9th century); Golinelli 2008a, p. 24; Nicolaj 2013a, pp. 287f., 291f., 302 (although both Pohl and Nicolaj do not rule out a later dating to the end of the 10th century); Heil M 2017, p. 6; Mischke 2024, p. 500; and more recently Thom Gobbitt (who also refers to personal communications with Colleen Curran, Anna Dorofeeva, and Evina Stein, all of whom support a dating to the late 9th century; Gobbitt 2025, p. 252).
The apparent contradiction between the proposed early dating of the main part of the codex and the unambiguous reference to the year 991 in the final quire was seemingly resolved by Pohl through a detailed codicological analysis. He confirmed an observation already made by Giuseppe Russo (Russo 1980, p. 610) that the fascicle containing the calendar was added at a later stage and thus, from a codicological standpoint, is not suitable for determining the date of the manuscript as a whole. This makes it plausible that the main section of the manuscript, containing the legal texts, was produced in the 9th century and only later, in the 10th century, the calendar was added.

However, even Pohl noted that there is no visible palaeographical difference between the final section containing the calendar and the preceding parts of the manuscript: „Die Schrift der letzten Lagen entspricht dann weitgehend derjenigen in den auf das Kalendarium folgenden komputistischen Texten und sieht nicht über 100 Jahre älter aus.“ (Pohl 2001, p. 127; see also Russo 1980, p. 610, who already observed “molte… affinità paleografiche” between the sections). Scholars are divided on the question of whether a change of scribe occurs in the final quire of the manuscript. Johannes Merkel (Merkel 1858, p. 596) believed that the entire codex was written by a single hand, a view initially shared by Alfred Boretius (Boretius 1864, pp. 32, 35). However, in the introduction to his edition of the Leges Langobardorum, Boretius revised his position, now arguing that the final part was written by later hands (Boretius 1868, p. XLI; followed by Krusch 1924, p. 110).
Rudolf Buchner disagreed, attributing the entire final section of the manuscript – from fol. 55 to 215 – to a single scribe (Buchner 1940, p. 64). Bracciotti, on the other hand, considered the calendar section to be younger than the rest of the codex (Bracciotti 1998, p. 58). Other scholars have noted the involvement of multiple hands in general terms but without assigning specific elements to specific scribes (e.g. Russo 1980, p. 610; Golinelli 2008a, p. 11), with the exception of Gobbitt 2025, who does distinguish between different hands, although he does not include the final quire in his analysis.
During our transcription and subsequent collation of the manuscript for the digital edition, we paid particular attention to the question of how many hands were involved in its production. In this, the palaeographical expertise of our colleague Grigorii Borisov proved especially valuable: he was able to identify the profiles of three main scribes. The first two alternated, each writing a few pages or shorter sections in the first part of the manuscript up to fol. 51v. Beginning at fol. 52r, however, a third main scribe takes over and continues alone through the rest of the codex (except for fols. 53v–54v, which were again written by one of the two earlier scribes). The diverse nature of this third hand can be seen, for instance, in the abbreviation for -orum, which appears in two different forms (e.g., fol. 97v, lines 19 and 20: once as a reverse et-ligature, and once as a simple cross). A striking feature is the ex-ligature, where the x is written as a vertical cross. This form appears alongside an alternative version in which the x resembles the standard Saint Andrew’s cross. A clear example of the parallel use of both ligature forms can be seen on fol. 55r, lines 15–18 and 25–28. Aside from such variable elements, this scribe’s hand is easily identifiable thanks to consistently and regularly used abbreviations, ligatures, and letterforms, as well as certain linguistic characteristics—for example, the frequent use of the diphthong ae at the beginning of words (aecclesia, aedictus, aeternum, etc.). Despite the graphical differences between the legal and the computistical parts of the manuscript, the features of this main scribe’s hand from the legal section are clearly visible in the final quire as well.

However, in the Easter table – with the only clear pointer for dating the manuscript to around 991 – shows a conspicuous peculiarity: the section covering the years 991 to 1006 appears on fol. 211v where it was squeezed into the lower right corner. The continuation, covering the years 1007 to 1158 (see Merkel 1858, p. 603), is found on fol. 212r–v, which is a leaf that was inserted into the final quire.

It appears that the scribe initially began with the Easter table on the final page of the calendar, using the remaining blank space, and then continued on a separate leaf that was inserted for this purpose into the quire. However, it is also conceivable that the leaf now numbered fol. 212 – with the Easter table starting from the year 1007 – was written separately. The scribe could have subsequently added the earlier years 991–1006 on the preceding page after having inserted fol. 212 into the quire. But if the Easter table is indeed an addition, and that it may even have been added only after the final quire was completed, it is all the more important to determine whether it was written by the same hand or by a different one. However, the material available for palaeographical comparison is limited: the Easter table contains mainly Roman numerals and capital letters used for the lunar letters. Nevertheless, the writing style and consistent forms of the capital letters (including two alternative forms for E and G each) provide a solid basis for concluding that this section too was written by the main scribe responsible for the entire final part of the manuscript. Antonio Ciaralli was able to confirm this assessment with further paleographical arguments, and we would like to thank him warmly for his expertise.

If the scribe of the final quire is indeed the same person who wrote the legal section of the manuscript, then a dating to the 9th century is hardly reasonable. While it is conceivable that the same scribe first copied the legal texts and then, years later, added the booklet with computistical content, it is implausible that nearly a hundred years could have passed between these two tasks.
The findings presented here are still provisional and will be further developed in a forthcoming study by Borisov on scribal hands with a high degree of graphical variation. For the time being, however, the often-contested dating of Modena O.I.2 to the “late 10th century, around 991”, as already supported by Mordek, appears to remain valid.

B. Mischke

(translated by Maximilian Lange)

Merkel 1858
Boretius 1864
Boretius 1868
Krusch 1924
Buchner 1940
Russo 1980
Bracciotti 1998
Borst, Arno, Der karolingische Reichskalender und seine Überlieferung bis ins 12. Jahrhundert, Teil 1 (MGH Libri memoriales 2, 1; Hannover 2001)
Pagano 2001
Pohl 2001
Golinelli 2008a
Nicolai 2013a
Heil M 2017
Mischke 2024
Gobbitt 2025