A new mythography for our digital world

Written by Greta Hawes and Scott Smith

MANTO was born out of frustration.

Scott was annoyed that he couldn’t find a list of the homelands of the individual Argonauts, let alone a map showing them. And why had no-one cataloged all of the tombs of heroes mentioned by ancient geographers?

Half a world away, Greta was tired of resources for myth that relegated Imperial authors to the status of belated documentors of fact, and that assumed the tradition must be organised by people and families, rather than the places, objects and cultural phenomena that increasingly captured her attention.

And so they - slowly but surely - found themselves embarking on a task that many had taken on before.

The desire to bring order to Greek myth shines out from the earliest Greek prose writing. The “mythographic impulse” to systematize conflicting, constantly shifting oral oral tales is itself quite an odd thing: Marcel Detienne described these writers acted like anthropologists, describing and analyzing Greek culture from the outside. But there is nothing “un-Greek” about mythography. Researchers in the past couple of decades have shown that the impact of mythographers only appears modest. In reality, they deserve attention alongside the actors in the Theatre of Dionysos or generations of bards reciting epic poetry, or the festival choruses as a crucial force that shaped the mythic tradition we recognize today.

This is not to say, of course, that Greek mythography is a homogeneous genre. The form and functions of mythographic texts varied widely: Acusilaos’ genealogical work (late 6th/early 5th century BCE) would have looked quite different from the Tragodoumena of Asclepiades of Tragilos (4th c. BCE), which summarized tragic plots, or Herodoros’ massive biography of Heracles (c. 400 BCE). And the fragments of all three can be contrasted with Apollodoros’ (2nd c. CE?) surviving Library, which gives a linear ‘history’ from the creation of the world to the return of Heracles’ descendants several generations after the Trojan War.

The end of antiquity did not put an end to such labors. Late antique, Byzantine and medieval scholars collected and continued the work of summarizing and interpreting Greek myth. Renaissance humanists took up the challenge of large-scale systematization:  Boccaccio’s compendious De genealogia deorum (14th c. CE) offers stories and allegorical interpretations for over 720 mythical characters, organized genealogically, using material drawn from over 200 ancient authors. Its immense popularity inspired further massive tomes, including Giraldi’s De deis gentium varia et multiplex historia and the Mythologiae of Natale Conti (16th c. CE). In our own time, we may point to Edith Hamilton’s extremely popular Mythology or to Timothy Gantz’s magisterial Early Greek Myth (1993) as examples of this impulse still in operation.

If this long and continuing history demonstrates anything, it is that achieving a coherent and comprehensive account of Greek myth is impossible. The fundamental obstacles are not just a lack of resources or intellectual ability; there are also practical ones of having to shape material to fit the constraints of writing. Take, for example, Apollodoros’ Library, so often consulted for its clear treatment of myth. In fact, its coherence comes at the cost of completeness: Apollodoros suppresses variants o as not to interrupt his narrative flow. He includes only minor points of dispute which merely nod towards the controversies behind the curtain without distorting the singular narrative that is his organizing framework. In other words, Apollodoros has scrubbed away the messiness to provide his system of Greek myth. Most crucially, his decision to start at the beginning and unfurl the Greek past in a long linear accounts which go, generation by generation, down family lines is perfectly tailored to the technology he works with: long rolls of papyrus.

Developments in writing technologies have of course made variants easier to deal with. Modern source books, like Gantz’ Early Greek myth or Fowler’s Early Greek mythographers retain the genealogical and episodic structure familiar from Apollodoros, but because the reader has also detailed indexes, she can skip from section to section, honing in on particular fragments or characters and consulting just the relevant pages, which might now be scattered through the work.

MANTO is an experiment in entirely rethinking how to organize the data of Greek myth outside of these analog contexts. Computing excels in the storage and retrieval of information: digital methods give us unprecedented opportunities to aggregate and sort through complex data, to identify relationships within it and to observe trends operating at scale. These structural advantages allow us to conceive of Greek myth not as a series of characters described in a series of texts, but as a full network of entities, in which objects and places might be just as important as people. In other words, MANTO will find room for Mycenae, the tomb of Cassandra, and the Scepter of Agamemnon alongside Agamemnon, and give them biographies which — despite intersecting significantly with his — are all their own. Such non-linear data can then be explored from any chosen angle: we could investigate how the myths of smaller cites connected into networks controlled by more powerful neighbors since in MANTO every place is an autonomous node and not consigned to a footnote or dealt with in a cross-reference. We might observe how these networks change over time, as certain texts find and lose favor, new places are colonized, and old cities abandoned.

As a digital platform, MANTO is is content neutral: any text can be added to MANTO and the model itself imposes no hierarchy of value. In effect, MANTO does not know (as we scholars undoubtedly do) that Homer is a more important source than Apollodoros, or that Nonnus is irredeemably late and strange. Such agnosticism fits with the democratizing sensibility of we scholars of mythography, who have shifted the emphasis from the prestigious contexts and canonical (poetic) texts of the archaic and classical periods and towards “pedestrian” prose authors. But more than this, because new texts can always be added, MANTO can keep pace with emerging interests in, for example, Imperial literature, and it can quickly integrate new discoveries.

The power of a digital dataset is that it can make relatively trivial work that has traditionally done “by hand.” Once the dataset exists, it can be reused infinitely to produce authoritative answers to questions not yet conceived when it was built. So, we could easily set it the task of quantifying all the mythic figures named by Diodoros, or displaying all of the places mentioned by Apollodoros, or the overlap between Latin and Greek authors. We could calculate family trees found in tragic texts, or patterns of colonization recounted in Latin geography. Then there are the more complex questions almost impossible to resolve using traditional methods: “Are Pausanias’ accounts coherent or are there contradictions in the material? How many?” “How coherent is chronological organization across all known texts?” “How similar are Acusilaos and Hesiod in terms of their data?” “In which period are Homer's mythic variants most frequently accepted as authoritative? In which period Aeschylus’?” “What happens to Greek myth with the rise of Rome?” “What happens to myths when cities are abandoned?”

In all this excitement over digital potentials, we must remember that this technology brings with it its own particular constraints. Digital models are only as insightful as we make them. Some of this ignorance can be useful: a computer simply accepts that that hero has four mothers recorded, three tombs, and two different stories of being killed. Nonetheless, points of ambiguity, uncertainty, and dispute are difficult to deal with in a binary system. The rules of this universe must all be specified in advance, but that is the subject of the next post.