This piece is often viewed as two separate entities because the two main branches of its manuscript tradition differ from one another significantly. Unable to reconcile the two successfully, in the main body of his work Goetz (1892) printed them as two separate texts, Monacensia and Einsidlensia, each with its own apparatus. In the appendix containing restored versions of the various colloquia, however, Goetz provided only one version of this piece. Although entitled ‘Colloquia Monacensia’, that version is in fact a combination of the two, based on an edition by Karl Krumbacher (1891) that drew heavily on sources from the Einsidlensia family. Goetz’s failure to explain the nature of this restored version or provide an apparatus for it has often confused readers into thinking that readings drawn from the Einsidlensia version were either attested in the Monacensia manuscripts or invented by Goetz.

The Monacensia or M version is the older and more important branch of the tradition, attested from the twelfth century onwards. The Einsidlensia or E version is attested only from the fifteenth century; the later date makes it in many ways less useful for restoring the original text, but nevertheless it cannot be dispensed with. The M version has its Greek entirely in a transliterated and then corrupted version (apart from one very late manuscript in which a highly corrupt version of that transliteration has been transliterated back into Greek script), and the untransliterated Greek of E is essential for making sense of the M version.

Like the colloquia Monacensia–Einsidlensia, this work is often thought of as two separate entities. The primary witness to it is a ninth-century manuscript in Leiden, one of the earliest and most important sources for the Hermeneumata as a whole. The only other witness is the sixteenth-century edition published by Henri Estienne, which is based on two lost manuscripts. The Leiden manuscript is much to be preferred as a source where both are available, and this makes the edition largely useless for most of the Hermeneumata. For the colloquium, however, Estienne’s edition is indispensable, since the latter part of the colloquium text is missing from the manuscript: the edition is the only surviving source.

SOURCES FOR THE TEXT

The Leiden manuscript

The Leiden manuscript (L) is Leidensis Vossianus Gr. Q. 7, copied in the second quarter of the ninth century and currently housed in Leiden University Library (plate 12). It contains the colloquium at the very end of the Hermeneumata, on folios 37v–39r. The text in L breaks off at the end of section 8e; this is unlikely to have been the original end of the colloquium (see 3.1.2 below), but it must have been where the text ended in the exemplar from which L was copied, for the break comes in the midst of a page, with another text following on the same page. The manuscript is clearly written and generously spaced, with four columns per page and few abbreviations (apart from words for ‘and’, which are often abbreviated in both languages, and from the omission of nasals, which is marked with a horizontal line). The Greek, which occupies the left-hand column of each pair, is in Greek uncials; the Latin is in minuscule. In the colloquium (though not always in the rest of the manuscript) words are left undivided in both languages (when there is more than one word on a line), and the Greek has no diacritics.

This colloquium is found only in a sixteenth-century printed edition and has received very little scholarly attention. It is worthy of more consideration, because it is much older than the source in which it appears, indeed probably earlier than some of the other colloquia.

Sources for the text

The only source for this colloquium is Glossaria duo e situ vetustatis eruta: ad utriusque linguae cognitionem & locupletationem perutilia, published in Paris by Henricus Stephanus (Henri Estienne) in 1573. This is the same work that is one of two witnesses to the colloquium Leidense–Stephani (siglum therefore remains S); for more information on it, see section 3.1.2 above. The colloquium Stephani (or colloquium Stephani ii, as it is sometimes called to distinguish it from LS, the Stephanus version of which is often called ‘colloquium Stephani i’) follows immediately after LS in Estienne’s edition, of which it occupies pp. 286–94. I have examined Stephanus’ edition both in person (using three copies in the Bodleian Library, Oxford) and via photographs.

In each column are given the headings of the capitula sections in the Hermeneumata version in question, with the number of their sequence in the version where they appear and the page number(s) of each heading in Goetz (1892). (Headings of the Celtis version are taken by kind permission from the forthcoming edition of Rolando Ferri.) Note that the tables of contents to the capitula found in many manuscripts do not always match the actual capitula, and that this list follows whichever of the two appears to be older on each individual occasion; for the Celtis capitula, however, the contents list is normally followed as it seems to be significantly more conservative than the actual capitula (for some of the differences see Dionisotti 1982: 92–3). For versions other than Celtis, use of the contents rather than the actual capitula is signalled by a reference to page and line of the contents rather than to the page of the actual capitula. Where ‘cf.’ is given before a page number, the section can be found there but the heading is differently worded or absent. Transliterated Greek has been retransliterated, accents added, and orthography normalized where it poses comprehension difficulties, but otherwise the text has not been systematically corrected. The last column offers a tentative reconstruction of the original capitula; numbers are given only when I think there is a reasonable chance of our knowing the order. In this last column a question mark in front of an entry means the existence of the heading in the original is uncertain, a question mark after the number means its position is uncertain, and a question mark at the end means that the exact wording of the heading is uncertain.