For this week’s musing I wanted to talk a bit about citation systems. In particular, you all have no doubt noticed that I generally cite modern works by the author’s name, their title and date of publication (e.g. G. Parker, The Army of Flanders and the Spanish Road (1972)), but ancient works get these strange almost code-like citations (Xen. Lac. 5.3; Hdt. 7.234.2; Thuc. 5.68; etc.). And you may ask, “What gives? Why two systems?” So let’s talk about that.
The first thing that needs to be noted here is that systems of citation are for the most part a modern invention. Pre-modern authors will, of course, allude to or reference other works (although ancient Greek and Roman writers have a tendency to flex on the reader by omitting the name of the author, often just alluding to a quote of “the poet” where “the poet” is usually, but not always, Homer), but they did not generally have systems of citation as we do.
Instead most modern citation systems in use for modern books go back at most to the 1800s, though these are often standardizations of systems which might go back a bit further still. Still, the Chicago Manual of Style – the standard style guide and citation system for historians working in the United States – was first published only in 1906. Consequently its citation system is built for the facts of how modern publishing works. In particular, we publish books in codices (that is, books with pages) with numbered pages which are typically kept constant in multiple printings (including being kept constant between soft-cover and hardback versions). Consequently if you can give the book, the edition (where necessary), the publisher and a page number, any reader seeing your citation can notionally go get that edition of the book and open to the very page you were looking at and see exactly what you saw.
Of course this breaks down a little with mass-market fiction books that are often printed in multiple editions with inconsistent pagination (thus the endless frustration with trying to cite anything in A Song of Ice and Fire; the fan-made chapter-based citation system for a work without numbered or uniquely named chapters is, I must say, painfully inadequate.) but in a scholarly rather than wiki-context, one can just pick a specific edition, specify it with the facts of publication and use those page numbers.
However the systems for citing ancient works or medieval manuscripts are actually older than consistent page numbers, though they do not reach back into antiquity or even really much into the Middle Ages. As originally published, ancient works couldn’t have static page numbers – had they existed yet, which they didn’t – for a multitude of reasons: for one, being copied by hand, the pagination was likely to always be inconsistent. But for ancient works the broader problem was that while they were written in books (libri) they were not written in books (codices). The book as a physical object – pages, bound together at a spine – is more technically called a codex. After all, that’s not the only way to organize a book. Think of a modern ebook for instance: it is a book, but it isn’t a codex! Well, prior to codex becoming truly common in third and fourth centuries AD, books were typically written on scrolls (the literal meaning of libri, which later came to mean any sort of book), which notably lack pages – it is one continuous scroll of text.
Of course those scrolls do not survive. Rather, ancient works were copied onto codices during Late Antiquity or the Middle Ages and those survive. When we are lucky, several different “families” of manuscripts for a given work survive (this is useful because it means we can compare those manuscripts to detect transcription errors; alas in many cases we have only one manuscript or one clearly related family of manuscripts which all share the same errors, though such errors are generally rare and small).
With the emergence of the printing press, it became possible to print lots of copies of these works, but that combined with the manuscript tradition created its own problems: which manuscript should be the authoritative text and how ought it be divided? On the first point, the response was the slow and painstaking work of creating critical editions that incorporate the different manuscript traditions: a main text on the page meant to represent the scholar’s best guess at the correct original text with notes (called an apparatus criticus) marking where other manuscripts differ. On the second point it became necessary to impose some kind of organizing structure on these works.
The good news is that most longer classical works already had a system of larger divisions: books (libri). A long work would be too long for a single scroll and so would need to be broken into several; its quite clear from an early point that authors were aware of this and took advantage of that system of divisions to divide their works into “books” that had thematic or chronological significance. Where such a standard division didn’t exist, ancient libraries, particularly in Alexandria, had imposed them and the influence of those libraries as the standard sources for originals from which to make subsequent copies made those divisions “canon”. Because those book divisions were thus structurally important, they were preserved through the transition from scrolls to codices (as generally clearly marked chapter breaks), so that the various “books” served as “super-chapters”.
But sub-divisions were clearly necessary – a single librum is pretty long! The earliest system I am aware of for this was the addition of chapter divisions into the Vulgate – the Latin-language version of the Bible – in the 13th century. Versification – breaking the chapters down into verses – in the New Testament followed in the early 16th century (though it seems necessary to note that there were much older systems of text divisions for the Tanakh though these were not always standardized).
The same work of dividing up ancient texts began around the same time as versification for the Bible. One started by preserving the divisions already present – book divisions, but also for poetry line divisions (which could be detected metrically even if they were not actually written out in individual lines). For most poetic works, that was actually sufficient, though for collections of shorter poems it became necessary to put them in a standard order and then number them. For prose works, chapter and section divisions were imposed by modern editors. Because these divisions needed to be understandable to everyone, over time each work developed its standard set of divisions that everyone uses, codified by critical texts like the Oxford Classical Texts or the Bibliotheca Teubneriana (or “Teubners”).
Thus one cited these works not by the page numbers in modern editions, but rather by these early-modern systems of divisions. In particular a citation moves from the larger divisions to the smaller ones, separating each with a period. Thus Hdt. 7.234.2 is Herodotus, Book 7, chapter 234, section 2. In an odd quirk, it is worth noting classical citations are separated by periods, but Biblical citations are separated by colons. Thus John 3:16 but Liv. 3.16. I will note that for readers who cannot access these texts in the original language, these divisions can be a bit frustrating because they are often not reproduced in modern translations for the public (and sometimes don’t translate well, where they may split the meaning of a sentence), but I’d argue that this is just a reason for publishers to be sure to include the citation divisions in their translations.
That leaves the names of authors and their works. The classical corpus is a “closed” corpus – there is a limited number of works and new ones don’t enter very often (occasionally we find something on a papyrus or lost manuscript, but by “occasionally” I mean “about once in a lifetime”) so the full details of an author’s name are rarely necessary. I don’t need to say “Titus Livius of Patavium” because if I say Livy you know I mean Livy. And in citation as in all publishing, there is a desire for maximum brevity, so given a relatively small number of known authors it was perhaps inevitable that we’d end up abbreviating all of their names. Standard abbreviations are helpful here too, because the languages we use today grew up with these author’s names and so many of them have different forms in different languages. For instance, in English we call Titus Livius “Livy” but in French they say Tite-Live, Spanish says Tito Livio (as does Italian) and the Germans say Livius. These days the most common standard abbreviation set used in English are those settled on by the Oxford Classical Dictionary; I am dreadfully inconsistent on here but I try to stick to those. The OCD says “Livy”, by the by, but “Liv.” is also a very common short-form of his name you’ll see in citations, particularly because it abbreviates all of the linguistic variations on his name.
And then there is one final complication: titles. Ancient written works rarely include big obvious titles on the front of them and often were known by informal rather than formal titles. Consequently when standardized titles for these works formed (often being systematized during the printing-press era just like the section divisions) they tended to be in Latin, even when the works were in Greek. Thus most works have common abbreviations for titles too (again the OCD is the standard list) which typically abbreviate their Latin titles, even for works not originally in Latin.
And now you know! And you can use the link above to the OCD to decode classical citations you see.
One final note here: manuscripts. Manuscripts themselves are cited by an entirely different system because providence made every part of paleography to punish paleographers for their sins. A manuscript codex consists of folia – individual leaves of parchment (so two “pages” in modern numbering on either side of the same physical page) – which are numbered. Then each folium is divided into recto and verso – front and back. Thus a manuscript is going to be cited by its catalog entry wherever it is kept (each one will have its own system, they are not standardized) followed by the folium (‘f.’) and either recto (r) or verso (v). Typically the abbreviation “MS” leads the catalog entry to indicate a manuscript. Thus this picture of two men fighting is MS Thott.290.2º f.87r (it’s in Det Kongelige Bibliotek in Copenhagen):
MS Thott.290.2º f.87r which can also be found on the inexplicably well maintained Wiktenauer; seriously every type of history should have as dedicated an enthusiast community as arms and armor history.
And there you go.
Bret Devereaux, “Fireside Friday, June 10, 2022”, A Collection of Unmitigated Pedantry, 2022-06-10.