Institute for Advanced Technology in the Humanities had a 10th anniversay symposium to which I was invited. The celebration both looked back at the birth of IATH and looked forward. Some excellent talks. I think I rather like such events.
Some thoughts on the conference.
1. Problem of celebrating the start of an institute and the turns it can take. In the celebration there is buried an assessment as to where it can turn. The question is where will IATH be 10 years from now?
2. Problem of scaling collaboration beyond the (academical) village. What works where there is geographic collocation is difficult to scale to a national or international level. As the provost said, IATH worked partly because of the collocation of groups and people in the Library. How can the success of IATH be scaled up to a national or internation level if collocation is important.
One solution is to scale in – bring people in rather than try to duplicate IATH on a larger scale. Create a centre that can bring others in creating a national institute on the campus of UVA.
3. In some ways the issues now in humanities computing are not techical but social or political. (Were they ever anything else?) Humanities computing because of the challenge of collaborative large (and expensive) projects, has introduced academic administration back into the humanities. We have to learn and solve the problems about big humanities.
We don’t really know how to talk about the administration of knowledge in the humanities in a disciplined way because we don’t think it is a worthy subject. We did, however, in this symposium do what we can to tell stories about the start of IATH and the difficult decisions made back then. These stories are not just war stories we tell to celebrate success – they are a way of thinking about a key moment in the administration of any initiative – namely its birth.
4. Outreach vs. Inreach. Outreach reaches out – IATH has succeeded at the “shock and awe” of outreach – the unit has an international reputation. Now, can we think about how to let others reach into IATH. How many guests can a unit handle?
5. One way to deal with scale and preservation problems is to end things and restart new things. What if IATH was closed – what would be left? What has it spawned? Or, what can IATH become and what will it stop doing? In a way, Jerry in his talk alerted us to things that have been spawned by IATH – NINES, ARP(?) and the Speclab.
6. We want computers to make it so. We expect that if we put all sorts of stuff in that we may get thought out. We believe in the technical version of the hidden hand – good things will happen without our planning them. As Jerry pointed out in his talk, this is not so. You don’t get intelligent digital media out of a collection, no matter how big.
7. Tools problem – how can we do anything useful? Matt nailed the problem of tools in his talk – a problem having to do with our amateur character. We don’t really do UI, we don’t market? We don’t document. We just demo and give papers. The software dies.
8. What do we not know? How has information technology shown us what we don’t know? Jerry points out how he has learned how little he understands about how the book works and this is one of the benefits of digital humanities. The question is what practices can one try in the face of ignorance. How can we imagine the unkown.
9. The history of tools. 1. Learn Snobol. 2. Build Garden Tool for all. 3. Build Web Services. Will we return to learning Snobol or some text processing higher order language?
10. Dr. Wulf (?) of the American Academy of Engineers talked about significant changes in science and engineering that are connected to humanities computing. The big problem is now complexity and emergent properties. He thinks science will not solve these, but engineering will. He suggested that the humanities have been dealing with complexity and emergent properties, and therefore might be able to contribute to the discussion. He challenged us to develop models that can predict what we care about (not in the sense of predict the future, but predict cause and effect within the model.) This sounds like Willard’s point about models.