Linking the Ancient World (or Laying Down the LAW)
I just returned from LAWDI, the Linked Ancient World Data Institute at ISAW, the Institute for the Study of the Ancient World in New York City. It was a three day conference in which tech and non-tech people came together to discuss progress in the ideal of sharing archaeological, philological, and geographical data across the web to allow for far-reaching research. We also compared the size of our problems.
[some photos by Daniel Pett of the Portable Antiquities Scheme / British Museum]
I've always been interested in ancient networks--the hows, whos, wheres and whys of interconnections thousands of years ago. That's led me to economic research, trade networks and the like, and eventually led me to studying 'money' as a connecting concept. That's not to say that all interaction is monetary, and I don't completely mean money in the modern sense. 'Currency' might be a better word as it indicates anything circulating as a norm among people, representing their interconnectedness. Today our 'currency' is information.
In ancient studies, we're trying to form knowledge strands and networks that feed off of and increase themselves. Essentially that's what the 'web' is, as the name suggests, but not all networks feed well into each other. We need information to fit, to be recognizable as reliable data of a particular kind. So when I say we're linking the ancient world, I mean we are trying to make sure that the disparate resources available on the web concerning ancient history, archaeology, epigraphy, and the like, can be joined into a much greater whole, something that will expand our research capability and our understanding overall. Furthermore, it will allow for any new additions to the network, both born-digital and made-digital, to be connected as well.
Google can already search web pages, but it can't always recognize specific data and provide it for researchers. It comes up with millions of pages that might mention something vaguely similar to what the researcher wants at varying degrees of trustworthiness. Rankings and narrow search terms try to improve hit relevance, but it isn't all that conducive to structured research on a specific topic. And there might not be a lot of reliable data out there on that topic at any rate. We may think that the web contains the sum of human knowledge, but it has a lot more on Sasquatch than it does on Sumerian standards of measurement -- even less on the data that can prove or disprove those standards (or Sasquatch, for that matter).
Hard data, reliable data, is growing on the web, but it still needs to be identifiable. It can never be perfect, nor can the system that harvests it, but we can improve on both. And as we think long and hard about how to make a good system, we are led deeper and deeper into the concept of linking and the concept of the web. In short, we get tangled up in strands, spending our time thinking about what the spider at the center might look like. Indeed, a good portion of linked data discussion has for a long time been caught up in philosophical issues. What does a Resource Locator (or Identifier) really point to? Does it really exist? or is it only a representation, or a representation of a representation?
It sounds silly at first. Anything in the virtual world can't be _real_. But a base stand-in for the real thing can be useful. If I have a reference to me, that reference might be to a picture of me, a picture from a different angle or at a different time, a drawing of me, a drawing I did, a paper I wrote, a mention of me in a paper... all of whose creation dates are much different than the 'creation date' of me. Any link certainly can't return me in real space, but does it have to return anything? Maybe it is only an acknowledgement that I exist and anything attached to that reference concerns me or a representation of some aspect of me, my work, etc.
As interesting as the philosophical discussion is, it takes up time that might be better used to actually _do_ something -- like put data out there. And that's where the debate has largely turned. Most people are saying 'what's the point?' We can't agree, but we can make some sort of workable compromise that gets our networks to function, so let's do that.
The first step is making any data presence on the web stable. It needs to be in the same place, accessible and cite-able, for the future. It must be open to all, with no restrictions. Copyright or the concept of intellectual property applied to raw data just doesn't work. We have to build new theories on solid information. Attribution to the person/s who gathered it is important, and that is why the data needs to be stable and cite-able, but restricted use makes no sense.
As one of my colleagues noted, all research is linking -- making connections between concepts, people, places, things. Thanks to LAWDI I'm now more firmly linked with a community that is both studying the ancient world and making those studies and the data that forms the core of them more accessible online, preparing the future of research. LAWDI was a resounding success and I thank all of the organizers for inviting me.
[some photos by Daniel Pett of the Portable Antiquities Scheme / British Museum]
I've always been interested in ancient networks--the hows, whos, wheres and whys of interconnections thousands of years ago. That's led me to economic research, trade networks and the like, and eventually led me to studying 'money' as a connecting concept. That's not to say that all interaction is monetary, and I don't completely mean money in the modern sense. 'Currency' might be a better word as it indicates anything circulating as a norm among people, representing their interconnectedness. Today our 'currency' is information.
In ancient studies, we're trying to form knowledge strands and networks that feed off of and increase themselves. Essentially that's what the 'web' is, as the name suggests, but not all networks feed well into each other. We need information to fit, to be recognizable as reliable data of a particular kind. So when I say we're linking the ancient world, I mean we are trying to make sure that the disparate resources available on the web concerning ancient history, archaeology, epigraphy, and the like, can be joined into a much greater whole, something that will expand our research capability and our understanding overall. Furthermore, it will allow for any new additions to the network, both born-digital and made-digital, to be connected as well.
Google can already search web pages, but it can't always recognize specific data and provide it for researchers. It comes up with millions of pages that might mention something vaguely similar to what the researcher wants at varying degrees of trustworthiness. Rankings and narrow search terms try to improve hit relevance, but it isn't all that conducive to structured research on a specific topic. And there might not be a lot of reliable data out there on that topic at any rate. We may think that the web contains the sum of human knowledge, but it has a lot more on Sasquatch than it does on Sumerian standards of measurement -- even less on the data that can prove or disprove those standards (or Sasquatch, for that matter).
Hard data, reliable data, is growing on the web, but it still needs to be identifiable. It can never be perfect, nor can the system that harvests it, but we can improve on both. And as we think long and hard about how to make a good system, we are led deeper and deeper into the concept of linking and the concept of the web. In short, we get tangled up in strands, spending our time thinking about what the spider at the center might look like. Indeed, a good portion of linked data discussion has for a long time been caught up in philosophical issues. What does a Resource Locator (or Identifier) really point to? Does it really exist? or is it only a representation, or a representation of a representation?
It sounds silly at first. Anything in the virtual world can't be _real_. But a base stand-in for the real thing can be useful. If I have a reference to me, that reference might be to a picture of me, a picture from a different angle or at a different time, a drawing of me, a drawing I did, a paper I wrote, a mention of me in a paper... all of whose creation dates are much different than the 'creation date' of me. Any link certainly can't return me in real space, but does it have to return anything? Maybe it is only an acknowledgement that I exist and anything attached to that reference concerns me or a representation of some aspect of me, my work, etc.
As interesting as the philosophical discussion is, it takes up time that might be better used to actually _do_ something -- like put data out there. And that's where the debate has largely turned. Most people are saying 'what's the point?' We can't agree, but we can make some sort of workable compromise that gets our networks to function, so let's do that.
The first step is making any data presence on the web stable. It needs to be in the same place, accessible and cite-able, for the future. It must be open to all, with no restrictions. Copyright or the concept of intellectual property applied to raw data just doesn't work. We have to build new theories on solid information. Attribution to the person/s who gathered it is important, and that is why the data needs to be stable and cite-able, but restricted use makes no sense.
As one of my colleagues noted, all research is linking -- making connections between concepts, people, places, things. Thanks to LAWDI I'm now more firmly linked with a community that is both studying the ancient world and making those studies and the data that forms the core of them more accessible online, preparing the future of research. LAWDI was a resounding success and I thank all of the organizers for inviting me.