The contents of the library of Alexandria

Q: 'What is the greatest unsolved mystery of all time?'
A: 'The contents of the Library at Alexandria. Unfortunately that's unsolvable.'

-- opinion discussion on a social forum, 1 Dec. 2016

The 'library of Alexandria' has a very strong brand nowadays. Its image, especially in the English-speaking world, has been shaped by three key moments in modern culture:

• Edward Gibbon's Decline and fall of the Roman Empire, vol. 5 (1781)
• Carl Sagan's TV series Cosmos, episodes 1 and 13 (1980)
• Sid Meier's Civilization video games (1991 to present)

Am I just picking on popular culture? How do I justify picking these events? Well, take a look at how common the phrase 'library of Alexandria' has been in the relevant periods. Here's the Google Ngrams graph for its frequency in the fifty years surrounding Gibbon's work:

The upswing is even more accentuated if you look at a longer time-span, like a century.

The phrase 'Alexandrian library' shows a similar trend, but it also enjoyed more of a vogue before Gibbon came along. In fact it was nearly always the more popular form of the phrase up until recently. References to the 'Alexandrian library' prior to the 1780s, when Decline and fall came out, rarely focus on its destruction: even a 1753 book about Hypatia only mentions the library in passing once. Pre-Gibbon occurrences of 'library of Alexandria' are nearly all caused by a modern publisher called 'Library of Alexandria' which has recently reprinted a number of books from that period.

Here's the graph for the half-century surrounding Cosmos and Civilization:

(The 'combined' line is added by me.) The phrase's increasing popularity is easy to see, but also notice how 'library' gradually acquires a capital L. By the late 1980s, in the wake of Cosmos, capital-L actually becomes more popular than small-l. It's around the same time that 'Library/library of Alexandria' overtakes 'Alexandrian library'. Capital-L drastically increases its lead over small-l, and 'library of Alexandria' over 'Alexandrian library', after the release of Civilization III in 2001 (2 million copies sold by 2003; in-game info sheet) and Civilization IV in 2005 (3 million copies sold by 2008; in-game info sheet). Google Ngrams doesn't present data after 2008. But Civilization V (2010) has sold around 10 million copies on just one sales platform, so don't go expecting the curves to flatten out or decline.

Longer-term graphs show Gibbon's and Sagan's influence in English very clearly, and perhaps surprisingly, also in French. (For the French one you'll have to click the 'search' button after loading the webpage, because of a bug: Ngrams doesn't handle the apostrophe in bibliothèque d'Alexandrie well.) In some other European countries, though, Gibbon had no impact at all: if you look at graphs for German and Russian you can see that there was no interest at all in the Bibliothek von Alexandria or the Александрийская библиотека until the 1980s and 1990s. Gibbon wasn't translated into Russian until 1883, and not into German until 2003, so in those countries the effect is down to Cosmos, Civilization, and increased contamination from the English-speaking world in the internet age.

All three -- Gibbon, Sagan, and Meier -- are responsible for the idea of the library as something uniquely irreplaceable. (And as we've seen in a previous post, that perception is untrue.) Sagan is especially responsible for the modern obsession over the 'destruction' of the library, and for the library's mystique in popular culture. Gibbon is responsible for making the library emblematic of a supposed conflict between Christian anti-intellectualism and pagan scientism, claiming that Christian-pagan violence in 389 CE was the occasion for its destruction. (A lot of sources, like this mostly fictional Wikipedia page, report Theophilus' campaign against pagan temples as happening in 391; Gibbon actually opts for 389, following Marcellinus Comes [see here, under the consulship of Timasius and Promotus.)

In the Civilization games, all the myths pop up simultaneously. Info sheets in Civilization I, II, III, IV, and V all claim:

Religious fanatics destroyed the library in 391 AD, after nearly 700 years of operation. Today, only a portion of the catalog survives, providing us with a mere hint of what treasures the library contained.

'The Great Library' as seen
in Civilization VI (2016)

(Civilization I and V have minor variations.) Not a word of this is likely to true, but it must surely have helped inspire sentiments like the one I quoted at the top. The latest game in the franchise, Civilization VI (2016), has re-written this passage:

Sources differ widely on responsibility for the fiery destruction of the Great Library’s collection of texts; usual suspects listed include Julius Caesar’s troops in 48 BC, Roman Emperor Aurelian c. 270 AD, and others. But it does appear that the last vestiges were burned in 391 at the orders of the Patriarch Theophilus to eradicate pagan influences in Egypt -- not the last time Christians would burn books.

Well ... it's an improvement, of sorts. The idea that the temple of Serapis still housed a library in 391 is doubtful; evidently we're still taking Gibbon at face value and blaming Christianity for the loss of information from antiquity; and here we have even more emphasis on the library's destruction as if it had any impact. However, there are some factual things creeping in too, and the words 'it does appear' are a small concession to the imaginary nature of the story.

Let's move on from the 'destruction' of the library. I want to address the concern raised at the top: the idea that the contents of the libraries are some great mystery.

It certainly would be nice to have a detailed catalogue, not to mention more texts. However, the fact that a lot of information is missing doesn't stop us from forming some very good ideas about it, based on the information we do have. Ancient writers regularly cite each other, and from those citations it's possible to infer an awful lot about what doesn't survive.

Editions of lost texts are a niche industry, but very important to our understanding of antiquity. Here are a few highlights in the modern scholarship --

• Over 1000 lost authors in the fields of history and geography alone -- the monumental Fragmente der griechischen Historiker, compiled initially by Felix Jacoby
• 216 lost tragic playwrights, about 170 of them earlier than the fire in 47 BCE -- Snell's 1971 edition of Tragicorum graecorum fragmenta (for the well-known playwrights, their hundreds of lost plays have large editions of their own)
• Around 80 lost philosophers prior to 400 BCE -- Diels and Kranz' Fragmente der Vorsokratiker
• About 200 lost epic poems prior to 400 BCE -- spread across several editions
• Around 90 lost lyric and elegiac poets prior to 300 BCE -- Page's Poetae melici graeci, Gentili and Prato's Poetarum elegiacorum testimonia et fragmenta

Alberto Bernabé's edition
of lost Orphic poems
(1612 pages)

Then there's the Kassel and Austin edition of lost comic poets, in eight large volumes. And I'm not even going to try to count up the later philosophers and poets whose books would have been in Alexandria, as well as in ancient libraries at Athens, Pergamon, Antioch, and elsewhere. And then there are other genres -- mathematics, natural sciences, prose fiction, mythography, technical manuals, and so on. Compared to the vast amount of books that were written in the Hellenistic period, the early ones are small fry. Wehrli's edition of lost Hellenistic philosophers has twelve volumes, and that's just for one school of thought.

The fragments of these lost books are often small. Some authors are just a name. But we do know an awful lot about some of them. Take for example Timaeus of Tauromenion, a tremendously important historian of early Italy. His work is lost, but in the New Jacoby edition, the fragmentary Greek text still adds up to something like 30,000 words. That's more than we've got of some intact historians! One of the most important foundations for the Greek mythological canon, second only to Homer, was a poem called the Catalogue of Women: it's lost, but between 1100 and 2000 lines of it survive, depending on how you count. However you look at it, it's bigger than either of the intact Hesiodic poems!

We don't have the actual texts of most books held in ancient libraries. We don't have much in the way of catalogues (nothing at all, in the case of Alexandria). And no doubt there are many, many more lost authors where no trace survives. But we have so many reports, about thousands of lost books, that we have a very, very good idea about the kinds of things that we're missing out on.

The last remains of the library were destroyed within a year of Hypatia's death. It's as if an entire civilization had undergone a sort of self-inflicted radical brain surgery so that most of its memories, discoveries, ideas, and passions were irrevocably wiped out. The loss was incalculable. In some cases, we know only the tantalizing titles of books that had been destroyed. In most cases, we know neither the titles nor the authors.

-- Carl Sagan, Cosmos, episode 13 'Who speaks for Earth?' (1980)

This is pure romance: invented by Gibbon, popularised by Sagan, and finally embedded by Civilization into a ruthlessly teleological view of the history of knowledge.

'Self-inflicted radical brain surgery'

There's no great tragedy, and not much mystery either. The claim 'The last remains of the library were destroyed within a year of Hypatia's death' is not true. Even if it were true, there's no good reason to suppose it would be important. It makes no sense to weep over Alexandria, nor any other library for that matter, because that's just not how texts got lost. No ancient library ever had the potential to be a repository that could last 2000 years. No modern library either, for that matter: libraries disappearing is just what happens if you wait that long.

In the same way, there's no point wondering or agonising over what arcane secrets might have been in ancient libraries. If you're wondering about the kind of stuff that we've lost, just open an edition of fragments. It's all there. It's there in truckloads.

(Do learn Greek first, though.)