Originally shared by Yonatan Zunger
Originally shared by Yonatan Zunger
On the subject of social networks, let me present to you Viking social networks. Or rather, the analysis of the social networks represented in sagas and epic poems, and the rather fascinating things they can tell us.
The one that caught my eye the most: the Táin Bó Cúailinge, the classic Irish epic, has long been thought to be entirely fictional. However, fiction leaves a trace: the mathematical structure of social networks which people make up (who talks to whom, etc) don't match the mathematical structure of "real" social networks. (This is sort of similar to Benford's Law, a method used to detect when people have faked data: it turns out that the distribution of digits found in real data and in numbers that people made up are different) Applying this analysis to the Táin, it was discovered that its graph is halfway between a real and fake one -- and that the fictional aspects of it were all clustered around a handful of characters who knew an unrealistic number of people. If those characters were amalgams of multiple people -- if, e.g., Queen Medb didn't actually speak to everyone directly, but sent messengers instead which the story just skipped -- this would actually be a much more realistic graph, suggesting a more factual origin to the tale.
Some of this can also tell us about modern social networks: for example, the life of Erik the Red (with his multiple deportations for manslaughter) lets us know that we've always had trolls, and the fact that Beowulf has realistic networks except for Beowulf himself suggests that those who hunt down trolls have always been foci of legend.
h/t Jennifer Ouellette
http://www.theverge.com/2014/7/15/5898653/mapping-mythical-social-networks
On the subject of social networks, let me present to you Viking social networks. Or rather, the analysis of the social networks represented in sagas and epic poems, and the rather fascinating things they can tell us.
The one that caught my eye the most: the Táin Bó Cúailinge, the classic Irish epic, has long been thought to be entirely fictional. However, fiction leaves a trace: the mathematical structure of social networks which people make up (who talks to whom, etc) don't match the mathematical structure of "real" social networks. (This is sort of similar to Benford's Law, a method used to detect when people have faked data: it turns out that the distribution of digits found in real data and in numbers that people made up are different) Applying this analysis to the Táin, it was discovered that its graph is halfway between a real and fake one -- and that the fictional aspects of it were all clustered around a handful of characters who knew an unrealistic number of people. If those characters were amalgams of multiple people -- if, e.g., Queen Medb didn't actually speak to everyone directly, but sent messengers instead which the story just skipped -- this would actually be a much more realistic graph, suggesting a more factual origin to the tale.
Some of this can also tell us about modern social networks: for example, the life of Erik the Red (with his multiple deportations for manslaughter) lets us know that we've always had trolls, and the fact that Beowulf has realistic networks except for Beowulf himself suggests that those who hunt down trolls have always been foci of legend.
h/t Jennifer Ouellette
http://www.theverge.com/2014/7/15/5898653/mapping-mythical-social-networks
Comments
Post a Comment