Sunday, 16 February 2020

Did Roman engineers stand under bridges?

Did Roman engineers or architects have to stand underneath their bridges, to prove that they were properly built? This story sounds weird, and it’s totally implausible. (Bear in mind that the Romans built bridges to go over water.)
Hmmm, thinks: if I were a Roman engineer, where would be the best place for me to stand under this bridge so it can be tested? (The Ponte di Tiberio, Rimini, dating to the principates of Augustus and Tiberius, early 1st cent. CE)
If you haven’t heard the story before, I’ll grant that it is niche. But like so many myths about antiquity, it does pop up all over the place. Here’s Nassim Taleb in a 2012 book:
First, never get on a plane if the pilot is not on board. ...

The first heuristic addresses the asymmetry in rewards and punishment, or transfer of fragility between individuals. Ralph Nader has a simple rule: people voting for war need to have at least one descendant (child or grandchild) exposed to combat. For the Romans, engineers needed to spend some time under the bridge they built -- something that should be required of financial engineers today. The English went further and had the families of the engineers spend time with them under the bridge after it was built.
-- Taleb, Antifragile (2012), chap. 23
We shouldn’t expect Mr Taleb to be very accurate about the Romans, mind. In 2017 he infamously had an online shouting match with Professor Mary Beard, the eminent Roman historian, in which he insisted tenaciously (and falsely) that there was racial purity within each province in the Roman empire.
I’d better grant that later in the same chapter Taleb adds a couple more snippets about the Romans which are at least partially accurate. (1) Roman soldiers had to swear a military oath on joining the army (apparently Taleb believes other armies don’t do that); (2) there existed an extremely rare military punishment called decimatio, the random execution of 1 in every 10 or every 100 soldiers (Taleb comments that ‘putting more than 10 per cent to death would lead to weakening of the army’ -- apparently a 10% casualty rate wouldn’t do that). It’s clear he gets his ancient history mainly from popular culture.
Taleb didn’t invent this story, but I do wonder if it’s because of his book that the notion entered popular culture.
Predictable aswer alert! (QI, ‘Keys’, 2013)
Stephen Fry. In Roman times, they’d get the constructor of the arch to stand right under the arch when the support scaffolding was taken away, just to show that he had faith enough in his own, er ...

Tim Minchin. Well, it’s natural selection of arch-builders, isn’t it. Is that guy any good? Well he’s still here!

Isy Suttie. I like that idea of getting people to test things. It’s like going to a barbecue and getting someone to try the sausage.
-- QI, series 11 episode 8 ‘Keys’ (first broadcast 25 October 2013)
Notice how Stephen Fry spots the problem with having this as a story about bridges: he makes it about arches instead. There’s no basis for that either, just some imaginative rewriting, to try to get the story to make some kind of sense.

It wasn’t QI or Taleb that invented the story. The oldest version I’ve found is a signature line used in a USENET post back in 2004:
"When Roman engineers built a bridge, they had to stand under it while the first legion marched across. If programmers today worked under similar ground rules, they might well find themselves getting much more interested in Ada!"
-- Robert Dewar
-- Preben Randhol, post to comp.lang.ada, 11 Feb. 2004 (alternate link)
Robert Dewar was a computer scientist who ran a company involved with the Ada programming language. Is he the ultimate origin of the myth? Who knows.

But I will say this: to me, it sounds awfully like the kind of thing you might hear from a tour guide.

I have a sneaking suspicion that the story really originates with the Pons Fabricius, in central Rome. It’s about 10 minutes’ walk from the Forum and from the Circus Maximus, and just around the corner from the Theatre of Marcellus. It was built in 62 BCE, and it’s still in use for pedestrians and cyclists to cross between the east bank and the Isola Tiberina.
The Pons Fabricius, a.k.a. Ponte Fabricio
Not that Fabricius had to stand underneath his bridge while the legions marched across! No no, I have in mind something much more mundane. It’s a simple misinterpretation. You see, there are inscriptions on the side of the bridge recording who built it and who restored it. Things like

L(ucius) Fabricius G(ai) f(ilius), cur(ator) viar(um) faci<e>ndum, c<u>ravit

Lucius Fabricius, son of Gaius, curator in charge of making roads, supervised (the building of the bridge)
But in separate places, the following gets tacked on:

idemque probavit

and the same man (i.e. Fabricius) approved it
(Don’t mind the spellings, that’s just what Latin looked like when Caesar was in his 30s.)
The thing is, the word probavit is ambiguous. Probare can mean ‘approve’, but it can also mean ‘test, demonstrate’. The same ambiguity can be seen in two English words derived from probare: ‘approve’ and ‘prove’, with ‘prove’ in the sense of test (as in, ‘the exception proves the rule’).

The modern Italian derivative, ha provato, is more specific. It’s almost always going to mean ‘he tried, he demonstrated, he tested’. So if someone like a tour guide were explaining or describing the inscriptions, I’m imagining they might well give the the meaning as ‘test’.

And just to show how plausible this is, here’s an ancient history website set up by an Italian family that reports the inscription exactly like that.
A latin inscription above the arch, on both sides of the bridge reminds us that it was built by Fabricius curator viarum (warden of roads) and that "idemque probavit" - he personally tested it.
-- ‘Ancient Roman Bridges’,, May 2006
I suggest we’ve got three stages in the development of the myth:
  1. The original: ‘Fabricius supervised the bridge, and the same man (idem) approved it (probavit).’
  2. An intermediate version, like on the MariaMilani site, with the mistranslations: ‘Fabricius built the bridge and personally (mistranslation of idem) tested it (mistranslation of probavit).’
  3. The misinterpretation of the mistranslation: ‘Fabricius built the bridge and personally tested it by standing underneath it.’
When tourists in Rome want to walk across a real, ancient, Roman bridge, they’re going to be crossing the Pons Fabricius. It’s nice and central, ten minutes’ walk from the Forum, as I said. So this is a misinterpretation that stands a good chance of going viral.

Of such things are myths made. Even ones as small as this.

Monday, 10 February 2020

The Epic Cycle wasn't as popular as you think

The Epic Cycle is perhaps the most famous group of lost texts of all time. They haven’t existed for at least 1500 years. Yet, when people studying Greek myth nowadays learn that they once existed, they’re often inexorably drawn to the lure of what has been lost -- what might have been.

What is the Epic Cycle? For those lucky people who are about to learn this for the first time, the Cycle was a group of eight early epic poems about the Trojan War -- the legendary war over Helen, fought between the city of Ilion or Troy, and an alliance of Greek heroes. Together the eight epics formed a complete poetic account of the war.
The wooden horse in the film Troy (2004). The horse used in the film is now by the waterfront in the nearby city of Çanakkale.
The two surviving Homeric epics, the Iliad and Odyssey, were reckoned among the eight. For the other six we have titles, summaries, author names. Of the poems themselves, we have only a few isolated snippets of text.
  • Kypria. This epic covered everything from the wedding of Thetis up to the start of the Iliad, in the ninth year of the war. (Some scholars like Jonathan Burgess think it originally covered the whole war. That’s more than just speculation, but we don’t have time to talk about it today.)
  • Iliad. This one survives.
  • Aithiopis. This covered two major episodes: the arrival of Penthesileia and her Amazons, and her death at the hands of Achilles, then the arrival of Memnon and his Aithiopes, and his death at the hands of Achilles. (Memnon’s Aithiopes are kind of linked to the real Ethiopia, but only kind of.) And then Achilles dies too.
  • Little Iliad. This covered various prophecies that had to be fulfilled before Troy could be defeated, like the theft of the Palladion and the story of Philoctetes. Also, the wooden horse gets built.
  • Sack of Ilion (or Iliou persis). The wooden horse goes into action, and Troy is razed to the ground. (Incidentally, the historical Troy was inhabited continuously through the end of the Bronze Age until about 950 BCE. The traditional date for its destruction is 1184 BCE. The real Troy survived after that date for about as long as the USA has existed.)
  • Returns. The homecomings and/or deaths of the major Greek heroes ...
  • Odyssey. ... except Odysseus, who gets a whole epic to himself.
  • Telegony. Another one with two episodes, synthesising inconsistent traditions about Odysseus’ later career and death, in northwest Greece and in central Italy.
See the ‘further reading’ list below for the surviving summaries and other material: they can be found in West 2003.
By the way the Brad Pitt movie, Troy (2004), uses material from the Iliad, but none of the others. The non-Iliadic bits of the film -- some of the Iliadic bits too -- are based on original material, combined with some other ancient sources.

What if we had even one of these epics? How great would it be to have the story of Achilles’ death? What literary glories are we missing out on?

These are the questions that tantalise Cycle fans. Let’s boil the questions down into slightly more academic terms:
  1. How different were the Iliad and Odyssey from the rest of the Cycle?
  2. What about the Theban cycle?
  3. Why did only the Iliad and Odyssey survive?
  4. When was the Epic Cycle lost?
The fall of Troy: the oldest surviving visual depiction, a Cycladic vase from Mykonos, ca. 670 BCE, roughly contemporary with the Iliad. Centre: the wooden horse, with peepholes for the Greek soldiers inside. Right: the death of Astyanax, son of Hector, perhaps being thrown from a tower by Odysseus, while the child’s mother Andromache reaches out her hands to plead for his life.

1. How different were the Iliad and Odyssey from the rest of the Cycle?

It’s best to withhold judgement on this, because it’s just too speculative. We have Aristotle’s opinion that the Kypria and the Little Iliad weren’t as good as Homer. But it’s a bit tendentious. The Little Iliad seems to have had much more unity of plot than he lets on. Around Aristotle’s time ‘cyclic’ became a generic word for tiresome, rambling storytelling. But we don’t know exactly how it came to have that sense. The Greek word kyklikos literally means ‘circular’, but it has other metaphorical meanings too; and there’s testimony linking kyklikos as a literary term to Antimachus, an epic poet who lived a few decades before Aristotle.

We can assume the lost epics weren’t as good as Homer. Anything more than that is speculative. There’s a famous article condemning the literary qualities of the Cycle, mostly because of its fantastic elements (Griffin 1977) -- but bear in mind that we’d be raising eyebrows at the Iliad, too, if only a summary survived. Just imagine: ‘Achilles’ horses talk to him, then a river chases him across the battlefield.’ You can’t judge literary quality from a summary.

2. What about the Theban cycle?

There’s no such thing as a Theban cycle. It never existed.

Poems about Thebes did exist! But no cycle. Modern scholarship has often grouped together four lost epics, the Oidipodeia, Thebaid, Epigonoi, and Alkmaionis, but there’s no reason to imagine they were grouped together in antiquity. No source, anywhere, ever mentions a ‘Theban cycle’. The idea was invented in the 19th century by the scholar Friedrich Welcker.

Some ancient sources do refer to a ‘cyclic Thebaid’. Others assign stories that may have belonged to Theban poems to ‘the cyclic (ones)’ -- poets? summarisers? mythographers? Who knows. The most robust interpretation is that ‘cyclic’ could be used as a catch-all term for any early epic that wasn’t the Iliad or Odyssey. Or maybe they’re references to Antimachus’ Thebaid. Either way, there’s no suggestion of a group of four epics.

One of the Tabulae Iliacae -- miniature carvings depicting Greek heroic legends, made in the early Roman principate -- lists the Oidipodeia and the Thebaid together, and mentions a ‘cycle’ shortly afterwards. But it still isn’t a Theban cycle. The tablet also lists two other epics, the Danais and the Titanomachy, which are totally unrelated. Plus, the word ‘cycle’ seems to be the next item in the list, not an umbrella term for the other titles.

Two poems do get grouped, but only sometimes, and never as a ‘cycle’. The Thebaid and Epigoni both get assigned to Homer by two early sources, Herodotus and (probably) Alcidamas. Also, the first line of the Epigoni survives, and its wording suggests the existence of a previous story. So these two may have gone together as a pair -- sometimes. But no group of four.
  • ‘The cyclic Thebaid’: Thebaid fragments 2, 3, 6 ed. West. Theban-related stories ‘in the cyclic (writers)’: Thebaid frs. 9 and 11 West; Epigoni fr. 3 West.
  • Titanomachy, Danais, Oidipodeia, and Thebaid listed together in conjunction with a ‘cycle’: Tabula Iliaca 10K ed. Sadurska, the ‘Borgia tablet’ = Cyclus Epicus fr. 2 Bernabé. Note that ‘Titanomachy’ is a supplement for ]μαχίας.
  • Thebaid and Epigoni assigned to Homer: Epigoni fr. 1 (from the Contest of Homer and Hesiod, quoting the first line of the Epigoni, probably based on Alcidamas) and fr. 5 (= Herodotus 4.32) ed. West.
In particular, no ancient or mediaeval source ever mentions a ‘Theban cycle’, contrary to what some people claim.
Sure, it’d be great to know more about the Theban epics. I’d love to have the Thebaid in particular! The Iliad has a few odd features that seem to be inspired by Thebaid material. For example, the fact that Agamemnon sometimes lives in Argos instead of Mycenae. Also, the Homeric formula anax andron Agamemnon ‘Agamemnon lord of men’ sounds like it was designed with Adrestus, anax of Argos, in mind: Agamemnon is a basileus, not an anax.

That doesn’t mean I have to assume the epics were ever a tetralogy. It’s high time to abandon that invention. There never was a Theban cycle.
The wooden horse imagined in Lego by ‘Brickman’, Ryan McNaught (‘Let’s go build’ exhibition, Te Papa, Wellington, Dec. 2017. Photo by T. Schaefer.)

3. Why did only the Iliad and Odyssey survive?

We’re damned lucky they did survive. It wasn’t a foregone conclusion. The Homeric poems didn’t hit the big time until the late 500s BCE, maybe a century and a half after the Iliad was composed. Until that moment, they might easily have gone the same way as the Thebaid and the Cycle.

We have only a couple of mentions of Homer in settings earlier than 500 BCE, and there it’s pretty clear that the name referred to epic poetry in general -- a bit like using ‘Hollywood’ to refer to all films regardless of where they’re made. In one story, set in the early 500s, it’s clear that ‘Homer’ means the Thebaid, not the Iliad or Odyssey:
For when Cleisthenes (tyrant of Sicyon) made war against the Argives, firstly he banned rhapsodes in Sicyon from competing in (performing) Homeric epic, because Argos and the Argives get praised so much all the way through. And second, there is a hero shrine to Adrestus, son of Talaos in the marketplace of Sicyon, and Cleisthenes wanted to cast him out of the country because he was Argive. ...

(Unable to ban the cult of Adrestus directly,) he introduced (a shrine to) Melanippos, on the grounds that he was Adrestus’ archenemy, since he had killed Adrestus’ brother Mekisteus and his son-in-law Tydeus.
-- Herodotus 5.67
The Iliad does have lots of references to all the Greeks as Argeioi, ‘Argives’. But this story is entirely about things from the Thebaid -- Adrestus, king of Argos, which made war on Thebes; Melanippus, one of Thebes’ defenders.

The fame of the Iliad and Odyssey suddenly skyrocketed with the advent of performances at the Panathenaia festival in Athens. After that they never lost their popularity. From about the 520s BCE onwards, their survival was guaranteed. The Cycle just didn’t get as lucky.

4. When was the Epic Cycle lost?

Look, the Cycle was never popular. It never enjoyed any prestige. It never had a wide readership. We have plenty of citations of it, sure, but only in antiquarian material -- scholars citing obscure words from an early text, abstruse mythological details, that kind of thing.

But even with scholars, hardly any of them knew the Cycle firsthand. They repeat odd facts and words without any context. Often it’s obvious that they've only encountered the material in earlier scholarly works. We can literally count on one hand the ancient writers who claim to have read any of the actual poems: Herodotus, Aristotle, and Pausanias. That’s it.

Pausanias is the latest. He’s a travel writer, living in the 2nd century CE. He states explicitly that he has read the Kypria and the Little Iliad. (Some scholars doubt even that this is true.) In one passage (10.29-10.30) he cites the Returns, so he may have known that poem as well. It isn’t impossible that some other late authors might have known some of the poems -- maybe Athenaios, maybe Porphyry -- but they don't say outright that they knew them, so we can't be sure.

Does this sound overly sceptical? Let’s put it in the context of which early poems people were actually reading. We have thousands of fragmentary literary papyri from Egypt, mostly Roman-era. The best represented author is Homer, unsurprisingly: there are hundreds of copies of the Odyssey, well over a thousand of the Iliad. If we look at lost authors, we see some of the big lyric poets -- Archilochus; Simonides; Sappho’s poetry was still a school text in the 7th century CE. For lost epics, the big name is Hesiod: we’ve got around sixty fragmentary copies of the Hesiodic Catalogue of Women (more than either of the surviving Hesiodic poems).

Here’s the question you should be immediately be thinking of. How many papyri of the Epic Cycle do we have?

If you guessed ‘none at all’, then congratulations, you are an excellent guesser of papyrus quantities. This doesn’t necessarily mean the Cycle had already disappeared completely. But it does show that it was way less popular than any other early poetry we know of.
Note. For completeness, I’d better note that one papyrus does appear in Bernabé’s edition as Little Iliad fr. 32. But no one believes it’s genuinely from the Cycle: Bernabé himself catalogues it as a ‘doubtful fragment’.
So if no one was reading the Cycle, how did the stories survive? In the Hellenistic and Roman eras there was a fashion for mythological manuals, encyclopaedias of myth, and prose summaries of myth. That’s how we know about the Cycle: the summaries that have survived were apparently copied from one of those manuals. The summariser makes it clear that the poems weren't popular in his own time:
... the poems of the Epic Cycle are preserved and have many people interested in them, no so much because of their merit, but because of the continuity of the material in it.
-- Proclus, Chrestomathy §20 ed. Severyns
The material also had a vogue in the visual arts. There are some Megarian ‘Homer cups’ from the 3rd-1st centuries BCE. And I’ve already mentioned the Tabulae Iliacae miniatures, from around the time of Augustus. These adapt many scenes that we know of in the Cycle summaries, but without trying to copy the poems or their summaries slavishly.
The most famous of the Tabulae Iliacae: tablet 1A, the Capitoline tablet (Rome, Musei Capitolini, Sala delle Colombe inv. 316). The left side of the tablet is missing. The panels down the right side illustrate books 13 to 24 of the Iliad, summarised in tiny writing on the pillar to their left. The central panel shows the destruction of Troy. At bottom centre are scenes relating to lost Cyclic epics. Perhaps the most striking thing about this tablet is its size: it’s tiny. It’s just 28 cm wide.
Several tablets mention Cyclic epics: tablet 1A has captions mentioning the Aithiopis, Little Iliad, and Sack of Ilion; 2NY and 6B mention the Sack of Ilion alongside the Homeric epics; 9D mentions the Iliad, Aithiopis, and Sack of Ilion; 7Ti mentions the Little Iliad and Sack of Ilion, and refers to events from the Aithiopis.

But these weren’t working directly from the poems either. They’re using summarised forms, the kind of thing you get in an encyclopaedia. One giveaway is that though the artists clearly spoke Greek perfectly well, they don’t use the spelling that you’d find in an early epic: their spelling is phonetic. They write Αἰνήας for Αἰνείας, Ποσιδῶν for Ποσειδῶν, Ἰλίας μεικρά for Ἰλίας μικρά, that kind of thing. If they’re not familiar with the spellings used in early epic, that means they weren’t reading early epics.

Another giveaway is the phrase used for the wooden horse. Tablet 1A calls it the δούρηος ἵππος (again phonetic, for δούρειος ἵππος). But that phrase could never have appeared in an epic poem. It doesn’t scan. Whatever the Little Iliad called the wooden horse, it wasn’t that. When Homer mentions the wooden horse in Odyssey 8, he calls it the δουράτεος ἵππος.

But guess what we find when we look at the summary of the Little Iliad? Yup: δούρειος ἵππος, just like in the tablet.

No one was reading the Epic Cycle. People lapped up Cyclic material in secondhand accounts instead.

It’s possible Pausanias is telling the truth, and that he found intact copies of the Kypria and Little Iliad in a library in Athens. But even if he is, they must have been among the last copies still in existence. We don’t know if the poems ever even got to Alexandria. And no ancient writer ever claims to have seen a copy of the Aithiopis or the Telegony. I’d bet those poems were lost even before the Roman conquest. (Which is a pity -- those are the most interesting ones!)

Even if the poems did survive, they were very obscure. When Roman poets like Vergil and Ovid went looking for Cyclic material, it’s most likely that they got hold of summaries, in Rome, rather than making a research trip to Athens like Pausanias did.
Note. The last part of this post is based on a paper I gave at the ASCS 41 conference in Dunedin in January 2020, titled ‘The Aeneid and the Epic Cycle’. Abstracts can be found here, and the slides I used here.

References and further reading

  • Bernabé, A. 1996. Poetarum epicorum graecorum testimonia et fragmenta vol. 1, 2nd ed. (1st ed. 1987). Teubner.
  • Burgess, J. S. 2001. The tradition of the Trojan War in Homer & the Epic Cycle. Johns Hopkins University Press.
  • Davies, M. 2001. The Epic Cycle, 2nd ed. (1st ed. 1989). Bristol Classical Press.
  • Davies, M. 2014. The Theban epics. Harvard University Press.
  • Fantuzzi, M.; Tsagalis, Ch. (eds.) 2015. The Greek Epic Cycle and its ancient reception. Cambridge University Press.
  • Gainsford, P. 2015. Early Greek hexameter poetry. Cambridge University Press.
  • Griffin, J. 1977. ‘The Epic Cycle and the uniqueness of Homer.’ Journal of Hellenic Studies 97: 39-53.
  • Huxley, G. L. 1969. Greek epic poetry from Eumelos to Panyassis. Faber and Faber (London).
  • Sadurska, A. 1964. Les tables iliaques. Państwowe Wydawnictwo Naukowe (Warsaw).
  • Sammons, B. 2017. Device and composition in the Greek Epic Cycle. Oxford University Press.
  • West, M. L. 2003. Greek epic fragments. Harvard University Press (Loeb 497).
  • West, M. L. 2013. The Epic Cycle: a commentary on the lost Troy epics. Oxford University Press.