a web of myth
Written by Greta Hawes and Scott Smith
A Byzantine scholar reports that his copy of Apollodoros’ Library began with the boast that no one would now need to read Homer or the tragedies because this mythography has ‘all that the world (kosmos) contains’. MANTO is a modern mythography; although our goal is not to replace the ancient poets (quite the opposite!), we too aim to replicate this mythic cosmos.
Greek myth is immersive because it creates a world of its own. This storyworld seems almost limitless: heroes have vast families, they visit each other and give gifts of special objects. Feuds continue down through the generations, as do curses, and marriage alliances can prove surprisingly important.
Each story can be continued in almost any direction: take a basic narrative like Heracles killing the Hydra. This is one of the 12 labours that the hero performs for Eurystheus, so already there is a broader context at play. He is helped by his nephew Iolaos, whom he later marries off to his first wife Megara (Apollodoros 2.6.1). The killing takes place near Lerna, at the springs of Amymone: these were named for one of the Danaids who, after she arrived from Egypt, was shown the springs by Poseidon (Apollod. 2.1.4). She later gave birth to Nauplios, the eponym of the harbour-town nearby. Pausanias tells us that in the 2nd c AD locals would point out a plane tree near the springs where they said the Hydra was born (2.37.4). And the two weapons that Heracles first tries to use against the Hydra - his club and his bow - also have their own stories. Heracles’ club was an olive branch cut either at Nemea (Apollod. 2.4.11) or somewhere along the Saronic Gulf (Pausanias 2.31.10). A generation later, the bow ends up in the hands of Philoctetes; Sophocles’ play tells of how Odysseus must travel to Lemnos to retrieve it from him since there is a prophecy that without it the Greeks will not be able to take Troy.
Each time we encounter a name in Greek myth, we can follow a new story off in a different direction. The mythic storyworld is thus a kind of ever expanding network:
A mythic network featuring Heracles’ killing of the Hydra. Graphic: Glen Goodwin
Most crucially, modelling the mythic storyworld as a network shows clearly that it is not only heroes and gods who spur ever more stories; weapons, trees, places, and landmarks also do this. In effect, these inanimate objects are just as active in holding the web together.
MANTO is in essence a large, complex dataset that captures relationships between all the various things that make up the mythic storyworld. Whereas the Greeks understood their storyworld as simply existing in an intuitive (which is not to say always coherent!) way, we are working in a digital space where each element needs to be defined and described as precisely as possible. The “nodes” in this network are entities of several different types:
Agents are person-like entities: gods, heroes, monsters etc
Collectives are groups of agents with a shared identity: so, the Argonauts, the Centaurs, the Danaids
Places are locations in mythic storyworld. Some of these have analogues on the historical map (Athens, Thebes, the River Acheloos), others do not (the Underworld)
Landmarks are immovable built or natural features like springs, tombs, or buildings
Objects are moveable entities without agency but with a specific identity, like the Cattle of Geryon, or the Scepter of Agamemnon
Events are significant occasions like the Return of the Heraclids or the Funeral Games of Patroclos.
When we create entities, we attribute to them a kind of ontological existence. This is actually trickier than it sounds, since we suddenly have to confront some thorny problems: Is ‘Tartaros’ always distinct from ‘the Underworld’? Was the gods’ Olympian home conceptually the same as the Mount Olympos in northern Greece? Do collectives that form for a particular reason and then disband (like the Argonauts) really belong alongside collectives that exist because all of the members are sisters (like the Gorgons or the Danaids)? Who counts as an Olympian? What defines a nymph?
Finding our way through these weeds has been one of the challenges of the past year’s work on the data structure. But of course these questions don’t require yes/no answers. Greek myth was tolerant of plurality and ambiguity, so we must allow for a lot of messiness.
Our data treats inconsistencies not as bugs in the system but as features of it. We do this by creating '“ties” which are assertions made about mythic entities in ancient literature. So, to capture Heracles’ labor described above, we would create a tie that says:
‘Heracles kills The Hydra at The Spring of Amymone with the aid of Iolaos using The Club of Heracles and The Bow of Heracles at the command of Eurystheus’ (the words in bold font are entities).
We would also have one that captures Apollodoros’ assertion that:
‘Heracles creates The Club of Heracles at Nemea’
and one that captures Pausanias’ assertion that:
‘Heracles creates The Club of Heracles at The Saronic Gulf.’
Both of these factoids exist in our dataset even though they cannot both be true at the same time. Greek myth is full of these variants and versions; one thing MANTO will do is reveal them at unprecedented scale.
Our aim is not to decide whether Heracles’ club was made in Nemea or nearby along the Saronic Gulf - that would misconstrue how storytelling works. Rather, we want MANTO to help us to navigate the tradition and understand it in all its complexity by revealing what the source material contains. Another example: our dataset includes both the factoid that the Aegyptids were buried at Argos, and the factoid that they were buried at Lerna. A curious user would be directed to the passage in Apollodoros where he describes the Danaids burying the heads of the Aegyptids in Lerna but their bodies at Argos.
No database will ever replicate the fullness of ancient Greek storytelling: if you want to understand Greek myth, you need to spend hours and hours with Homer and Sophocles, Pausanias and Apollodoros, Pliny and Pomponeius Mela. What MANTO does is different: it deals with data and relationships. By breaking up these stories into small, constituent parts and then using those parts to build up a storyworld, we are creating something that couldn’t exist in antiquity (because it requires non-linear digital methods) but which nonetheless approximates what we have observed through years of reading: that every ancient storyteller held a conceptual network of myth in his or her head.
To see how we go about creating this data, read on.
