Mythic chronology in Pausanias: preliminary data analysis
written by Greta Hawes
Later this week I will be presenting at the Celtic Classics Conference some research that I’ve been working on with Xinyi Xu, a PhD candidate in the Faculty of Engineering at the University of Sydney.
The BIG IDEA is to try to understand whether Pausanias has a consistent and coherent understanding of what happened when within the mythic period. So, for example, does he understand that one event happened before another, or that two mythic heroes belonged to the same generation even if they came from distant cities?
It’s difficult to work this out just by reading Pausanias since the work is so long (10 books!) and the mythic material so scattered and disjointed. Of course, Pausanias does make some explicit points about mythic chronologies being wrong (e.g. Daidalos could not have joined in a particular expedition because he wasn’t alive at the time… 10.17.4). But what if we could go beyond these obvious statements and look for underlying patterns in his mythic content that might reveal what he was doing with myth, and not just saying about it?
This is where having a good dataset comes in handy!
The BIG IDEA is essentially this: there are two ways in which time is established in Greek myth: by generations (i.e. heroic genealogies) and by events (i.e. essentially the interactions between these heroes).
Most genealogies and events in Greek myth are highly local: they are pertinent to a single city or region. But some are not. So, events like the Trojan War, the expedition of the Argonauts, and the return of the Heraclids affect lots of heroes from across Greece; conceptually, all the heroes who fight at Troy, or take part in any other such expedition, should belong to roughly the same generation. A few heroes also have an outsized impact on genealogies. Heracles is a great example of this since he fathers children across the Mediterranean; again, conceptually at least, all of Heracles’ children should belong to the same generation.
So, I have come up with a plan to work with Pausanias’ data in two different ways and see where the chronological inconsistencies lie. Firstly, I have worked with family connections that we have captured to establish a kind of relative genealogical chronology for all these heroes. Usually this involves counting back from the “anchor generation” of those that fought at Troy. So, a hero who fights at Troy will be labelled “TWG” [= “Trojan War Generation”]; his parents will be labelled “TWG-1”, their parents “TWG-2”, and their parents "TWG-3”, etc.
The next step will be to work with Pausanias’ data from the perspective of mythic events. So, I will take each of these entities, now labelled with a chronological marker determined by genealogy, and see whether they interact only with other heroes of roughly the same generation.
To do all this, I needed someone with excellent network analysis skills, and so was delighted to meet Xinyi Xu, who took on the task of modelling this data and visualising the results. We are only part way through the analyses currently, but we do have some interesting results to share at the conference, so I wanted to add them here to allow others to see them as well.
We have been focussing at this stage on identifying which mythic people are relevant to this project, and what their family relationships look like.
Pausanias mentions 1 805 mythic “people” (individuals and collectives) in his work. (You can see all the entities he mentions here.) Of these, only 1 691 are relevant to our project in that they conceptually exist in a particular point in mythic time. (We need to exclude immortal beings like gods and personifications, and multi-generational collectives like the Amazons.) Xinyi’s preliminary analyses now show that of these, 1 192 have family connections that are described in Pausanias.
This is where things get truely interesting. Pausanias often just mentions genealogies in passing. And yet, much of this scattered information turns out to fit together into a remarkable whole. Xinyi has shown that 72% (858/1192) of the mythic people with genealogical connections in Pausanias belong to a single family — we call this the “big family”.
Network graph showing “the big family” in Pausanias, using preliminary data from MANTO 9 July 2024, created by Xinyi Xu.
Nodes show gender (pink: female; blue: male; green: undefined). Edges show types of relationships (blue: child; purple: spouse; green: descendant; orange: sibling). Arrows point from parents to children.
For high resolution version click here.
Beyond the heroes that cluster into this “big family”, we get a large number of much smaller families.
Network graph showing the “small families” in Pausanias, using preliminary data from MANTO 9 July 2024, created by Xinyi Xu.
Nodes show gender (pink: female; blue: male; green: undefined). Edges show types of relationships (blue: child; purple: spouse; green: descendant; orange: sibling). Arrows point from parents to children.
For high resolution version click here.
You can see here that many of our heroes form simple two-node networks (usually parent-and-child) or quite small clusters. In the centre we get larger families extending several generations. Those who know a bit about Greek myth will notice that some of these smaller families are actually connected to other ones, and so there is some artificiality in here.
In some cases, we may still need to go back and re-work our data to capture implicit genealogies that we may have missed. But in most cases we are finding that really crucial information (like the fact that Hecabe was Priam’s wife, and Polyxene was their daughter) is simply not given in this text. Pausanias would simply expect his readers to already know such basic “facts”!
For this reason, Xinyi ran a second analysis, which pulled in all the genealogical information about this same group of mythic people from all sources in MANTO. This analysis shows that the “big family” would expand with such implicit connections counted: in this scenario 83% (1040/1250) of Pausanias’ people are genealogically related to each other.
With thanks to the team that collected data from Pausanias: Kennis Barker, Audrey Coleman, Glen Goodwin, Greta Hawes, James O’Maley, Rosie Selth, Scott Smith, Ari Toumpas.
And thanks also to my Macquarie University interns and research assistants who have worked on the generational data analysis: Gabriele Dolphin, Rowan McIntyre, Rosemary Selth, Elle Platt.