You know those keynotes Apple holds every September to brag about their newest gadgets? Everyone’s been speculating for months, leaks have gone viral, you’ve been cursing your eleven months old iPhone for the last two months because they’re promising something so much better. And then all these bright and shiny new toys are revealed, and you have to wait another couple of months before you can get your hands on them…
Yeah? Well, consider this about the same magnitude.
While Silke has declared war on the squirrel decapitating her garden flowers, I have been trying out Gephi’s timeline, another neat feature that allows you to monitor changes in a network over time. You’ll definitely hear more about this later!
What you need to make Gephi’s timeline work is, and hold on to your horses now, this might come as quite a shocker: dates. Now, I don’t like to criticize Trismegistos1. Trismegistos is pure perfection, of course. Nothing comes close to its beautifully built relational structure. PER is to its REFs like milk is to Double Stuff Oreo’s; GEO stores place names like Madonna harvests poor Malawi children; TEX holds it all together like an indispensable wooden Ikea peg.
But what you can’t do in Trismegistos is extract a one-mode network of people linked to people, with dates. Or without dates, for that matter. This requires an intermediate step. We can extract a two-mode network of people appearing in texts, since TM People is linked directly to TM Texts, and then we can convert this to a one-mode network in R. We know the dates of these texts of course; we can even add them to our two-mode network. But once we convert, we lose this information. I’ve tried it in R: it can’t process a two-mode edgelist with attributes with the ‘tnet’ conversion command:
Error in as.tnet(net, type = “binary two-mode tnet”) :
Type of network not recognised
You have to leave out the dates to get it done. Which means that if you want those dates for the Gephi timeline, you have to enter them manually. I can hear some of you saying: so, what’s the big deal? People compile edgelists manually all the time. Well, call me a spoilt brat, but I ain’t adding dates manually to 1,000+ edges. Not even 100+.
So I was whining about this to our Almighty Trismegistos Overlord, and he looks at me with a defiant glare, and ever so dryly says: I can convert two-mode networks in Filemaker, you know.
Two years ago, Silke had messed with R for monthstrying to get that right; we’ve been exporting, loading, converting, exporting, importing ever since, and now he’s going to act all smug, and declare that it can all be done in Filemaker?!
But heck, as much as I’d like to sulk, I’ve had a look at the new Trismegistos two-mode to one-mode conversion tool (aka the TOMATOR or ‘Trismegistos’ One-Mode generATOR’) and it really is a marvelous little invention. We’ll be doing some more test runs, and then it’ll be up soon when we launch our new webpage: Trismegistos networks! So stay tuned for more news!!
1What do you mean, you don’t know what Trismegistos is? Trismegistos has been, is, and forever will be, the center of the universe. There is no meaning to life without it. It is the big that created the bang, the essential link in evolution, the cornerstone of civilization. It is called after the famous epithet of Hermes – Thoth, the Egyptian god of wisdom and writing who also played a major role in Greek religion and philosophy. It is a platform aiming to surmount barriers of language and discipline in the study of texts from the ancient world, particularly late period Egypt and the Nile valley (roughly 800 BC – AD 800). Note from The Almighty Trismegistos Overlord: this section is plucked from the Trismegistos website! Ok, you might recognize the personal ‘Six Degrees’ touch here and there. And there. And there.
The core component is Trismegistos Texts (TEX), currently counting 370,911 entries. When the database was created
in 200513.8 billion years ago, it focused on providing information (metadata) on published papyrological documents from Graeco-Roman Egypt. Chronological boundaries are always artificial (Prof. Willy Clarysse is an expert on this matter), and the nature of the sources soon suggested that 800 BC and AD 800 were more suited. Since Egyptology does not know a disciplinary boundary between papyri and inscriptions, TM decided to expand by adding all epigraphic material as well. Papyrology on the other hand also includes writing tablets from outside Egypt, which led us to widen our geographical scope to the entire ancient world. Finally, since the distinction between published and unpublished is increasingly less productive in a digital environment, we now no longer discriminate in that respect either. In principle, however, we still provide metadata only. No prostitutes. This means that Trismegistos increasingly wants to be a platform where information can be found about all texts from antiquity, thus facilitating cross-cultural and cross-linguistic research. This will of course only be possible through cooperation with all players in the field, since our aim is to lead people to the partner websites, where more information, often including photographs, transliterations and translations of the texts, can be found. Also: love, world peace, and bulletproof marshmallows.
Several aspects of the Texts database have been elaborated in the course of successive projects and have become separate databases linked with the core Trismegistos Texts database.
1 The Collections database, built on the Leuven Homepage of Papyrus Collections, is a set of currently 3,562 modern institutional and private collections of texts and their 192,152 inventory numbers. It is searchable both separately and in the Texts database.
2 The Archives database, built on the Leuven Homepage of Papyrus Archives, is a set of currently 490 collections of texts in antiquity, mainly in Egypt, and the 17,380 texts that are part of these archives. It is searchable separately, leading to the texts themselves.
3 The People database, building on the Prosopographia Ptolemaica, is a complex set of prosopographical and onomastic databases (PER, REF, NAM). It currently contains 492,731 attestations of personal names of non-royal individuals living in Egypt between 800 BC and AD 800, including all languages and scripts and written on any surface. Thanks to me. I am the TM People hero, sadly still waiting for my commemorative statue and plaque to be erected in our hallway.
4 The Places database (GEO), expanding the geographic database of the Fayum project, is a set of currently 43,600 places in Egypt and increasingly also the rest of the Mediterranean. It contains the currently 131,784 attestations of toponyms in texts from Egypt (800 BC – AD 800), but is also linked to the provenance field in Trismegistos Texts.
Finally, because abbreviations are often different in the various disciplines, we have also started creating a Bibliography, which resolves many of the short references we use in Trismegistos. It also wants to facilitate the search for all texts in a specific publication. Its coverage is patchy except for the publications dealing with Demotic and Abnormal Hieratic, where the Demotistische Literaturübersicht provides a much higher standard of bibliographic information.