Tuesday, 24 June 2014

What Have the Romans Ever Mapped for Us? Results from the Latin Geographic Tradition

Having recently completed our first content workpackage (CWP1), dedicated to Early Geospatial Documents from the Latin Tradition, we'd like to take this opportunity to share the annotation data that we've compiled so far. Overall we have completed annotating place references in 33 documents (41 if we include additional language versions of the same document). Within these documents, we've identified 19,880 toponyms, and were able to establish mappings to Pleiades in 15,721 cases (79%).

Spatial distribution of toponyms annotated in CWP1 - Latin Tradition

You can find the complete list of our documents, along with a download link for the data, below. The annotations are stored in CSV format - i.e. they can be opened in a spreadsheet application, or imported into a database or GIS.

An additional part of our work in CWP1 (which will be important too for all our content work packages) has been to identify additional relevant documents as we go along. Our list of "geospatial documents" has therefore grown quite substantially. We have included these documents in our annotation tool Recogito. You can follow their status directly on Recogito's Latin Tradition landing page.

Now that we have finished work on the first of our six traditions, we are keen to get your feedback. In particular:

  1. We look forward to seeing what and how you make use of these data. We're sure that you'll use them in ways that we can't anticipate, and we'd love you to share that with us!
  2. The large number of documents, the ambiguity of the evidence and the comparatively short space of time mean that some of our identifications will inevitably be wrong or open to debate. We are planning a mechanism to allow people to suggest alternative suggestions or to indicate agreement and disagreement. In the meantime, feel free to contact us if you have proposals for corrections. We're also happy to hear suggestions for other early latin geographic documents which we may have missed.
  3. Since, as we expected, we could not fully annotate or geo-resolve all of our documents, we're interested in hearing from people who might be willing to join us in the challenge. If you would like to join in and help find places in the incomplete documents, feel free to get in touch and we may be able to provide you with a Recogito account. We'll have to roll this out slowly since Recogito is not a 'community tool' as such (with features such as full moderation, user profiles, etc) so please be aware that there may be a wait if we get a lot of volunteers!

Thursday, 22 May 2014

Future Footnotes, Reverse References and Bottomless Maps

Pelagios and the Graph of Historical Data 
are continually evolving in a number of directions that can sometimes make it hard to answer the simple question: ‘What are the benefits?’ Regular readers will know that these are many and various (and by no means all accounted for) but in this blogpost we’d like to outline two important ones that we've been thinking about for a while. We call these ‘Future Footnotes’ and ‘Reverse References’ (more on Bottomless Maps at the...well...bottom). 

Footnotes and references are pretty much the defining feature of textual academic discourse. It isn’t enough to have a bright idea or discover something remarkable – we have to relate our ideas and discovery to a wider body of research in order to locate them in the scholarly debate. But both suffer from severe limitations: footnotes can only point backwards and references can only point outwards.

Footnotes allow us to provide additional information that provides context or authority for an idea in a text (or even an image in the form of captions). Sometimes they are descriptive, but more often than not they are cross-references to previously written material – as an author we obviously can’t know about future material at the time of writing.  Because the graph of historic data is continually growing, Pelagios links act as a Future Footnotes. Annotation allows us (and anyone else in fact) to create hooks that connect our texts to both old and new material as it becomes available, without having to manually update those links ourselves. So, for example, when we annotate a reference to Londinium, for example, we don’t just say ‘here are the other references to Londinium that I am aware of at the time I wrote this’, but ‘here are other references to Londinium that the community is aware of, at the time you are reading it’.

References, on the other hand, allow us to provide evidence that backs up a point that we are making. They unilaterally point outwards because we only have the opportunity to refer to other work. In contrast, by annotating content we are simultaneously contributing our information to a wider cloud and so we create a Reverse Reference – i.e. it becomes directly available to other people through their own annotations. And to flip the logic of Future Footnotes, we don’t merely make it available to works that will be annotated in the future, but we make it available to works that have already been annotated as well. Thus we immediately make our work more accessible to precisely the people who might find it interesting.

So the benefits of Pelagios, and the Graph of Historical Data in general, are that they both future-proof and mutualise the cross-referencing that underpins academia in a way that has never been possible before. The analogies to footnotes and references aren’t quite perfect because they don’t account for the authorial stance – i.e. the desire for an author to selectively identify content, but they do indicate how radical this development in one of the most fundamental practices of academia can be. 

There is one additional benefit that we also think is essential: because the graph is open and distributed, it’s possible to create services that allow for seamless interlinking between online resources. In other words, you don’t have to go to a centralised portal or search service to discover relevant material. You simply discover it naturally, through hyperlinked references and footnotes in online books or articles (or webpages, or pictures, or maps, or songs or videos...and so forth). 

Of course curated portals and search services are valuable too, which leads us to our Bottomless Maps. There has been much discussion of the Deep Map in recent years – interactive digital maps that contain content that extends beyond the visual surface. Bottomless Maps, like the Pelagios heatmap for example, link an ever-growing (and thus to all intents and purposes infinite) quantity of content to the places they depict. While the scope of the Pelagios project is restricted to historic geographic concepts, the model which we have been collaboratively developing is applicable to any other kind of reference (people, periods, classifications, canonical text citations, and so forth). An ecology of other projects is now springing up to support this, so the combined (and evolving) graph of humanities data will ultimately become much more significant than Pelagios itself. We look forward to a future in which such cross-referencing is just as commonplace as footnotes and references are today.

*We here expand the notion of the Graph of Ancient World Data to include any content of a historical, classical or archaeological nature.

Friday, 7 March 2014

Greeking Out

This week marks a new and exciting milestone in the Pelagios 3 project - the start of work on the ancient Greek geographic tradition. There's more Latin to do of course: our work packages run on a staggered, overlapping 6-month basis, and, while we already have 19 documents in the system (some in both Latin and their modern language translation), future additions will include some major itinerary lists—including the Antonine Itineraries and Ravenna Cosmography—as well as a number of smaller but fascinating geographic sources such as the Haidra mosaic, some more inscribed vessels, and the Piazzale delle Corporazione at Ostia.

But from today we'll start introducing Greek documents into the system. Ancient Greek traditions of knowledge about geography extend far beyond Plato's "frogs around a pond" metaphor for Greek settlements around the Aegean Sea. From Homer's Odyssey, Greek texts push the boundaries of travel, exploration and knowledge, and Odysseus, the man who 'saw the cities of many men and knew their minds', stands as the archetypal explorer for Greeks who settled in places as far off as the Black Sea, Massalia (Marseille) and Libya. Later Greek authors like Hecataeus, HerodotusAristotle, PytheasEratosthenes, Hipparchus, Posidonius, Artemidorus and Ptolemy are largely responsible for the way we conceptualise geography today (indeed, Eratosthenes invents the discipline), and we still use the terms that they came up with—terms such as equator, meridian, parallel, latitude and longitude. At the same time, much Greek geography is almost cosmological in nature—an attempt to understand the form of the earth and its place in the universe.

Remarkably, however, given the number and detail of these ancient witnesses, almost no Greek maps survive, and it is debate whether maps were even a feature of Greek traditions of geographical knowledge. (A map documented in Herodotus's Histories, carried by a certain Aristagoras of Mytilene, becomes the site of contestation and debate, while Herodotus himself 'laughs at' the schematic representations of his contemporaries.) Instead Greek conceptualizations of the world were almost exclusively in a narrative form, from numerous periploi (sailing itineraries) to Strabo, whose Geografica remains central to our understanding of global geography in the transition to Empire. 

Working with Ancient Greek texts will introduce some new challenges for us to tackle. To begin with change in alphabet will take a little getting used to for some of the team! Fortunately recent work by Bruce Robertson, Greg Crane and others on OCRing ancient Greek means that we should be able to include a range of previously inaccessible texts. We can also draw on experience form the Hestia project and a promising new approach developed by Thomas Efer at the University of Leipzig that can identify toponyms in a Greek text by comparing it to a previously marked up English text. We don't yet know what will be the most efficient combination of methodologies but at least we have plenty to choose from.

We have enormously enjoyed working with the Latin texts and will continue doing so, but the possibilities for analysis opened up by annotating documents from these two strongly related yet radically divergent traditions are incredibly exciting.

Jerusalem depicted in the Madaba Mosaic (6th C. AD). Image from Wikimedia Commons.