How ancient writers marked the start of each new book

I’ve been translating a section of Theodor Birt’s Die antike Buchwesen for the last couple of days.  The book may be elderly, but is still the standard reference.  Pages 14-146 are concerned with prefaces, introductions, tables of contents and the like.  Here is what he has to say.  It is somewhat discursive, unfortunately.  I have renumbered the footnotes.  Note that he says that narrative works, such as Eusebius HE and PE, always avoid “proemia”, favouring linking words.  Both in fact have tables of contents — prographai — which suggests that these must be later additions, if this thinking were sound. But Polybius is a counter-example.  Here is what he says.

[p.140] So much for the individual book in its relationship to the whole work.

The diligent reader who picked up one [book] after another, could ensure the correct order of the rolls from the labels or σίλλυβοι. Such frequent book exchange, with the necessarily distracting break in reading, which was unavoidable, was injurious to continuous reading; the end of the roll that had been read had to be rolled up tightly, its cover put on and [p.141] the roll put carefully to one side before the next could be opened and unrolled, for which a slave may have been helpful. Moreover, it was desirable that a book picked up occasionally, out of context, could be read independently and remain understandable. So it was appropriate for the writer, to have something about the book at the front, in order to inform each reader.

A work with a purely scientific purpose such as Strabo’s certainly could not be based on the pre-Alexandrian approach of texts only being split into books as an after-thought (1), laconically indicating where each had left off.  In others we find, however, the custom of the proemium already formed, at first quite clearly merely for the practical purpose of orientation: the contents of the previous roll were in summary once again more or less briefly recalled, and the next task to be treated stated. At an early period Polybius does this scrupulously sometimes; one of his προγραφαί is summarised eightfold in no less than seventeen lines: …. To which is added in two short lines:  … The purpose of these προγραφαί, as Polybius (XI proemium) gives it in a remarkable manner, is: ‘to allow the reader to orientate himself, but also to encourage one who happens to pick up the book to read it, and finally to allow those who just want to look up something quickly to do so”. The proemia of Diodorus probably come closest to that pedantic type. In Varro it is simplified to even such a simple phrase such as in De lingua lat. VI init.: “origines verborum quae sint locorum et ea quae in his, in priore libro scripsi, in hoc dicam de vocabulis temporum” with the [p.142] addition: “atqui si qua erunt ex diverso genere adiuncta, potius cognationi verborum quam auditori calumnianti geremus morem.” Lucretius also summarises at length at the beginning of his books, usually with a “Et quoniam docui” (III, 31 f., 26 f. IV, VI, 43 f.), once in a full 25 verses (V, 56-81). How much more elegantly did his successor in the didactic poem, Virgil, briefly introduce wine-making with a “Hactenus arvorum cultus et sidera caeli, nunc te Bacche canam eqs.” and beekeeping with a “Protinus aerei mellis Caelestia dona exsequar eqs.” But we see that length also at later times, such as recur in the dream-books of Artemidoros. But it should be noted that this advice to the reader was sometimes unconnected with the actual text of this book, i.e. apparently only attached to the outside of the roll; this we see first reliably in the προγραφαί of the first six books of Polybius (2), the praefationes of Jerome (3), and it also seems to apply to those of Martial (4).

 A writer of taste, especially a poet, could not be content with such sober notes, however.  He reshaped the proemium, which was merely an index, into a proemium-shaped [p.143] excursus.  The beginning of each new roll is a fresh invitation to the reader to an important discussion and a common work: it was a reminder, so as to put forward a more general idea; the author, by speaking in a disinterested way, masked his zeal in a pleasant manner. The threshold of each roll is like that of a hospitable house: the meal may not be available immediately behind the covered vestibule.  Lucretius himself diverts the attention of Memmius, before he sets off and speaks, in a well-planned manner in what is possibly a preparative commonplace, which can fill a quarter or half of a hundred verses; similarly also Virgil to Maecenas, after he briefly has stated the contents of the book (Georg. I, 5-42, II, 4-8, III, 3-48). Diodorus also usually writes in the manner of Lucretius, either in technical discussions, or generally considered thoughts … — once almost with an ἀθιπενοι προλέγειν (XIX, 1, 9) — making a summary and disposition (so IV, V, XIV, XV, XVIII, XIX). Oppian and Manilius proceed in the same way. The latter in his astronomy wanders even more extravagantly than Lucretius (II v. 1—149, III 1—95, IV 1—121, V 1—31).  And Cicero acknowledges this as his own principle: in singulis libris utor prooemiis (ad Att. IV 6, 2).  It is easy to see how these pieces in Cicero’s books are added to the front externally; they betray their purpose very clearly. He himself writes about this once a very instructive note to Atticus (XVI 6, 4): “I have sent you my single book De gloria; but it has got the wrong proemium, which is already used for the third book of the Academica. I had  not remembered, when I was at Tusculanum, that I had used it before; afterwards on the ship, I was reading my Academica and discovered the problem. I immediately got a new one on paper for De gloria; it is enclosed; you must cut off the old one and stick this one on  (illud dissecabis, hoc agglutinabis)”. But Cicero explains the cause of this error to Atticus: “It’s because I have compiled a collection of proemia in their own roll; from which I choose one when I need a σύγγραμμα.  But Pliny proceeded quite otherwise. The [p.144] “medicinae ex animalibus” could not, unlike most chapters of his encyclopidic treatise, be contained in a single book; although Pliny elsewhere avoids real proemia, he interrupts the three books in the “medicinae ex animalibus” at the beginning of the second (XXIX) by giving a short history of medicine, at the beginning of the third (XXX) by giving a similar piece on magic.  The appearance of these introductions thus should relieve the monotony of these series of books for the reader. Nor was it accidental, but intended by Pliny, when he gave the book of the “medicinae ex aquatilibus” (XXXII), both front and rear tractates (5).

Narrative works always avoid proemia, however.  The Metamorphoses, the Aeneid, Lucian’s True history, the novel of Longus [=Daphnis and Chloe] and the like would have been greatly affected. The invocation of the muse in Apollonius Rhodius init III. and IV init. is not intended as an introduction, but as the start of special new material; accordingly II. init lacks such a passage (6).  The proektheses of Diodorus on the other hand seem intended almost to isolate the individual books, and so are different in principle to the προγραφαί, in the terminology of Polybius (7), which, as previously mentioned, were only attached to the outside of the roll, and therefore did not break up the text itself; the first hexad of Polybius only had these προγραφαί, and therefore there was in it, according to Polybius VI init., the [Greek] from book to book. Certainly the proemia of Xenophon’s Anabasis are simply such προγραφαί; based on the [Greek] Polybius. XI. init.

And only in such works, with  no proemium, was a textual uncertainty possible of the kind  that we encounter in VergiI. The sixth book of his Aeneid points to the dominant tradition of the verse “Sic Fatur lacrimam classique immittit habenas / Et tandem Euboicis Cumarum adlabitur oris.” The tradition in [p.145] Servius (on V fin. and VI init.) teaches that this is as arranged by the editors Tucca and Varius, while Virgil himself instead allocated the verses to the fifth book’s end, and that then “Probus and others” restored the arrangement of Virgil. Similarly it could be argued that even in Homer ὧς εφατο κλαίουσα once stood at the end of the book (II. 23. fin).

There was a practical, though inconspicuous way, in works without proemia, to help the memory of a reader who went from one roll to another. Care was taken that the final sentence of a book was repeated at the beginning of the next one, either exactly or slightly modified form; usually a little sentence expressing a transition; i.e. according to the testimony of the best manuscripts of Strabo (8) III fin.  …[several lines of Greek] Pliny VII Nat. hist fin. says: “nunc revertemur ad reliqua animalia primumque terrestria“, and then VIII init.: “ad reliqua transeamus animalia et primum terrestria“; first IX fin. has: “hinc volucrum naturae dicentur“, and then X init.: “sequitur natura avium“. Eusebius places at the end of the second book of his Ecclesiastical History “καὶ τὰ με καψὰ Ἰουδαιουσ ἐν τούτοις ἦν”, and book three begins with the same words without καὶ; the same happens in the transition from IV to V; the second book of his praeparatio [p.146] evangelica begins with the final words of the first: “τὰ με το προειρημένον περιέχει τρόπον” etc. (9).  See also the repetition of the διὸ in Porphyry “de Abstinentia” II init. from I fin. This also is present in pre-Alexandrian texts, as we see in Theophrastus’s History of Plants (VII fin. and VIII. Init.). We have here a trick, perhaps less the author’s own, so much as his publisher’s.

Instructive as the beginning of the book is, more still is the behaviour of authors at the conclusion of each book, to which we now briefly turn. …  (Birt continues with a discussion of colophons)

(1) For this subject see chapter 9.
(2) Polyb. XI. proemium says that only his first six books had prographai, the following instead had proektheseis, before each new Olympiad; both types of preface have similar purposes for his work, and the difference is that the prographai deteriorate easily through accidents of the copyist’s art (δυὶ πολλὰς αἰτιυς καὶ τὰς τυχούσας ὀ λεγωρούμενον καὶ φθειρόμενον), while the proekthesis have a safer place, as more closely linked to the work (…). Book V has not come down to us with a προγραφή. — That the opening words of this excerpted passage are incorrect has already been observed; they can be completed thus ….
(3) Hieronymus comm. Ezech. V praef.: Ne librorum numerus confundatur et per longa temporum spatia divisorum inter se voluminum ordo vitietur,praefatiunculas singulis libris praeposui: ut ex fronte tituli statim lector agnoscat quotus sibi liber legendus et quae nobis prophetia ezplananda sit.
(4) Martial’s (and Statius’) prefaces stand outside of the actual book: IX init. stands extra ordinem paginarum, VIII init.. in ipso libelli huius limine; also II init. stands before the pagina prima.
(5) This is done rigorously in De Halieuticis S. 159 ff.   [= Th. Birt, De Halieuticis Ovidio Poetae falso adscriptis (Berlin, 1878) — RP]
(6) Quintilian’s remark IV proemium 4  applies here also. 
(7) See above P. 142 note 1. 
(8) This is witnessed by ms. Mediceus Β for all four books, Parisinus C for III. fin. And IV fin., and Mediceus k forV. and VI fin; although in the latter case Β is missing some of the words in k giving τρόπον, but only as far as Ἰταλικῶ, the material omitted in Β is probably an expansion in k. These details of the ends of the books I owe to Prof. Niese.
(9) On this feature in Eusebius see Heinichen’s “Excursus XV” to Eusebius’ Kirchengeschichte, Bd. III, S.445.