What is involved in editing an ancient text?
Stefan Sienkiewicz interviews Tobias Reinhardt
Following the recent publication of his new critical edition of Cicero’s Academica, and of his Translation and Commentary, Tobias Reinhardt talks to Stefan Sienkiewicz about what is involved in editing an ancient text.
What is involved in editing a text, in general and in particular?
Since we have no autographs of major texts written in antiquity, we rely for our knowledge of them on medieval manuscripts written on parchment (and later paper). The first step in producing an edition of a text is usually to establish what manuscripts there are. If you are fortunate enough to have access to facilities at Oxford, then work begins in the Weston Library, where there is an extensive collection of catalogues of libraries with manuscript holdings (shelfmark: R.Cat.). By working through that from beginning to end one can identify most of the manuscripts of a given Latin text. This information can then be supplemented, including by searching online catalogues of particular collections, which have become available more recently. In the case of Acad., where Cicero created three different editions, and where we have the latter half of the first edition and the first twenty percent or so of the third edition, both of which have a different transmission, one will be looking for the manuscripts of two separate texts. In addition, there are fragments of the final edition which are transmitted indirectly, through quotations in other authors; these require a different approach altogether. The next step, and it should be said clearly that one never starts from scratch in this task but stands on shoulders of other scholars, is to impose order on these manuscripts, by establishing their relationships to one another. One does this by looking for minimally one uncorrectable error, of a kind which no two scribes would make independently of one another: if two manuscripts share this error, and in addition have errors peculiar to themselves, one can assume that they descend from a third manuscript which exhibited the error in question. Using this principle, one can try and impose order on the manuscript tradition and will in many cases arrive at a tree-like structure (the stemma), which allocates a place to all extant manuscripts.
How do you decide which manuscripts to consult?
Assuming a stemma can be drawn, one normally excludes those manuscripts from the constitution of the text which derive from other extant manuscripts, the reasoning being that the manuscripts further down in the stemma cannot contain anything new that has been vertically transmitted. However, such later manuscripts may still be of value in places, e.g. when they carry corrections reflecting a lost part of the tradition which descends from higher up in the stemma, or when the corrections are emendations by a reader who had the understanding and the skills to correct the text without drawing on manuscript support. The Lucullus is an unusual case in that for two thirds of the text we exclusively rely on a grouping of Carolingian manuscripts which were created in relative geographical proximity to one another. Three of these manuscripts exhibit detailed corrections, and one of the decisions an editor has to take is to evaluate the nature of these corrections, and to decide then how best to use and present the evidence they provide. This in turn requires an understanding of the intellectual environment in which medieval manuscripts were created and used. An engagement with the reception history of a text is an integral part of editing.
Is there a lot of travelling?
One can look at lots of manuscripts in the original, and many editors see this as a benefit. At the same time, it is true that the prices for reproductions have come down significantly, and that many larger collections are gradually making their holdings available online as colour facsimiles. These will be good enough for editorial purposes in many cases. However, the digital image cannot completely replace the consultation of the original. For instance, whether a discolouration of the parchment under part of a word is just a feature of the parchment or due to an erasure which was written on secondarily is something which even a high-resolution colour image may not reveal. Here one needs to appraise the texture of the manuscript surface in this area to take a view.
When you decide between readings, to what extent is the decision driven by philosophical considerations, and to what extent by other considerations?
A good number of the textual problems of Acad., which is on the whole a comparatively well transmitted text, have a philosophical dimension. Some of these problems have to do with negations: being small words, they can get left out easily, especially if they are abbreviated, as is often the case in medieval manuscripts. In a text where the discourse is often couched in clipped, Stoicising style, negations can also give rise to the saut du même au même, where the scribe omits words between two negations. Greek terms (as opposed to transliterations), which Cicero used on occasion, tend to be garbled in manuscripts because knowledge of Greek became rare in the Latin west for several centuries until it was regained in the Renaissance. In some cases we would not necessarily suspect the text on formal grounds, but we have comparative material from Greek texts which is at variance with what Cicero says (doxographical passages, which report particular views of less well-known figures, are a case in point); this is cause of scrutiny and weighing of the evidence. Some passages are simply very hard to understand and thus prompted readers presumably already in antiquity to provide explanations in the margin of their text. The next person using the same manuscript as an exemplar would then mistake the marginal explanation for omitted material and mistakenly import it into the text.
There are, however, also many cases where stylistic, linguistic, or other non-philosophical considerations determine the editor's choice.