Ammianus and the difference between chapter headings and text

Texts of Ammianus' history are usually printed divided into chapters - between six and sixteen per book of 30 or so pages- each of which has a heading. Editors never tell you the status of these chapters -- but the divisions were made and the headings written in 1681 by Adrien de Valois (Hadrianus Valesius) for his revised version of his brother Henri de Valois' edition of 1636. Adrien intended them as epitomes at the beginning of each book, though more often than not they are printed at the start of each chapter. A couple of years back I published an article on these headings (Classical Philology 104 (2009), 233-242). One effect of them is to make Ammianus look more late antique and less classical, more like Eusebius and less like Tacitus, because many late antique texts have chapter headings, whether authentic or editorial, and classical texts don't. Another issue, since the texts were written by a learned scholar with a deep knowledge of the period, is that sometimes the headings reflect Adrien's reading or inference rather than Ammianus' text.

To give two examples of many: (1) the chapter heading at 30.10 says that the emperor Valentinian II was acclaimed at the town of Brigetio (Szöny in Hungary): this is an inference from Ammianus, who is unspecific, but in fact a contemporary text states that he was acclaimed emperor in Aquincum (Budapest). Plenty of scholars wrongly place Valentinian's elevation in Brigetio, following Adrien's chapter heading. (2) Or take the heading of 25.7, which calls the peace treaty of 363 by which the emperor Jovian ceded the city of Nisibis to Persia "very shameful but necessary." Adrien took his wording from another historian, Eutropius; but it is misleading since Ammianus thought the treaty shameful and unnecessary.

I have now found another case which I missed in my article, thanks to discussion with Dr Kyle Smith. When Nisibis (Nusaybin, on the modern Turkish/Syrian border) was surrendered to the Persians, the inhabitants were allowed to leave under the treaty. The heading of chapter 25. 9 includes the following: oppidani inviti patria excedere et Amidam migrare compulsi, "the townsfolk were compelled against their will to leave the homeland and move to Amida." But though Ammianus describes compulsion, the townsmen are not described as moving to Amida, (modern Diyarbakır) (25.9.6): exin variae complentur viae qua quisque poterat dilabentium, "then the various roads were filled with people slipping away wherever each was able to." It would make sense for people to move to Amida, which had been sacked by the Persians in 359 but which became a much more important centre in the following centuries; and Zosimus, whose account is close to Ammianus and who is presumably Adrien's source, tells us (3.34.1) that "most, indeed nearly all emigrated to Amida, a few settled in other cities."

Adrien's chapter heading here is not badly misleading; but if I had been asked I would have said that the citizens of Nisibis were resettled in Amida, and a closer look shows that the situation was not as simple or as orderly as that. Ammianus reinforces his pathetic picture of the refugees by not naming their destinations - indeed implying (see above) that they are going on various roads in different directions.

So men were appointed to drive them out, who threatened death if anyone postponed departure, and the walls were filled with wailing and laments, and through all parts of the city there was a single sound of everyone groaning, since the matron tore her hair on being driven out an exile from the home in which she had been born and brought up, and the mother bereft of her children or widowed from her husband was driven far from their graves, and a tearful throng embraced the doorposts of their houses or the thresholds and wept (25.9.5).