Text into data: Apollodoros 2.1.1-2 and the early Peloponnese

written by Greta Hawes

The MANTO methodology is simple in concept: take a bit of text, identify its mythic entities and then draw ties between them. But as always the devil is in the details.

As we have worked to collect data, we have had to confront and resolve innumerable problems that we could not have even imagined before diving into this project. And from these particular problems we create a set of principles to guide future work.

As an example of the basic difficulties of identifying what exactly comprises a ‘mythic entity’, in this post I’m going to talk through a passage from Apollodoros’ Library which was critical in pushing us towards the principle that place names (sometimes) matter more than physical locations.

Here is the passage:

Phoroneus ruled over all the territory that would later be called the Peloponnesos. With the nymph Teledice he had Apis and Niobe. Apis turned his power into tyranny, was a violent dictator, and named the Peloponnesos Apia after himself … Argos received the kingdom and called the Peloponnesos Argos after himself

(2.1.1, 2 trans Smith & Trzaskoma)

We can quickly pick out the ‘easy’ bits, i.e. the genealogical connections. To capture those we need just a single tie (as always, entities are in CAPS):

APIS, NIOBE is child of PHORONEUS, TELEDICE

(Of course, there is much more possible information that we could give about this little family unit (Apis and Niobe are siblings... ) and much more accurate information (Niobe is daughter of Phoroneus and Teledice, Apis is son…), but this is not necessary since it is done globally throughout the dataset using reversals. Thus, two agents who have the same parents will be listed as siblings. Gender is an attribute of each entity, so the gender-neutral ties are transformed into gendered ones when the data is visualised, as in Yaya Lu’s project.)

We would then want to capture the fact that we have listed here three successive rulers of the same kingdom:

PHORONEUS rules THE PELOPONNESE (570577)

APIS rules THE PELOPONNESE (570577)

ARGOS rules THE PELOPONNESE (570577)

So far, this seems quite straightforward. Of course, we have had to use a name for this kingdom, ‘the Peloponnese’ that it will not receive until later, when it is named after Pelops (described in Epitome 2.9). Nonetheless, we can accurately locate it as a place using the Pleiades dataset: the number ‘570577’ identifies it as a particular region in cartographical space.

Next we would want to capture the idea that Apis and Argos both named their kingdoms after themselves, i.e. that they were both eponyms of this region of Greece. But ‘APIS is eponym of THE PELOPONNESE’ makes no sense — Pelops will be eponym of the Peloponnese!

Here we need to recognise that sometimes what we call something is more important than its fundamental ontological identity. Phoroneus, Apis and Argos all rule the same territory — which we call by convention the Peloponnese — but their names connect to different place names. So, to capture these eponymic relationships we will need:

APIS is eponym of APIA (570577)

ARGOS is eponym of ARGOS (570577)

ARGOS changes name of APIA (570577) to ARGOS (570577)

Notice that this last tie introduces some chronological specificity: we now have a datum that captures the fact that Apollodoros asserted that the name Apia was used before the name Argos was. Notice also that these new entities Apia and Argos share their Pleiades URN with the existing entity, ‘the Peloponnese’. This is because they should all have the same longitudinal and latitudinal data as attributes. As we gather data, we need to keep in mind that Apia and Argos exist really only as place names and that if we want to connect entities to this kingdom that have nothing to do with the name of it, we should use the entity ‘the Peloponnese’ as we did for capturing its various rulers. So we give Apia and Argos the attributes of ‘alternative name for the Peloponnese’ to show that these entities are subsidiary to the main entity, which we by convention give the label ‘the Peloponnese’.

Names are crucially important in Greek myth: so many stories tell of why a place has the name it has, and so often we suspect that mythical characters exist only because someone of that name must exist to give his or her name to something. Much of the development of MANTO has been about getting the balance right between thinking of entities as ontological ‘things’ with distinctive, definable qualities (‘the large peninsula protruding into the Mediterranean Sea with representative points Lat. 37.2539576404, Long. 22.3130920268’) and recognising that the labels given to these things are nonetheless significant. This is where entity attributes become crucial. We must give our entities labels (‘Apia’, ‘Argos’, ‘the Peloponnese’), but the data that sits under these labels as attributes of the entity are where the true power of discrimination resides. So, in this passage we have two entities called ‘Argos’; we can identify these are different from one another because they belong to different categories (one is a place, the other an agent), and because they have different identifiers:

[place] Argos - alternative name for the Peloponnese

[agent; male] Argos - son of Zeus and Niobe

And as the dataset has grown, so too has our list of Argoi - we currently have:

[place] Argos - alternative name for the Peloponnese (570577)

[place] Argos - alternative name for the Argolid (570104)

[place] Argos - city in the Argolid (570106)

[agent; male] Argos - son of Zeus and Niobe

[agent; male] Argos - Panoptes ('the all-seeing'), guards Io

[agent; male] Argos - Argonaut, builder of Argo

And this is without counting ‘Pelasgian Argos’ or ‘Amphilochian Argos’ or ‘the Argos contingent’ …

All this looks tidy and well-structured once more, but if we want to turn text into data, we need to keep room for a bit of messiness. After all, our sources are frequently less-than-clear about which ‘Argos’ they mean. Apollodoros - sometimes perhaps deliberately - does not distinguish between Argos the city and Argos/Argolid the region; and then certain tragedians take to calling Mycenae by the name ‘Argos’, and so, when Euripides mentions that the Cyclopes built the walls of Argos (Trojan Women 1081-99) what should we do with that factoid?

The devil is in the details.