Main »

Center Net 2010

centerNet summit

Introductory Session

centerNet had their first international summit at King's College London. The summit was supported by the NEH and organized by Neil Fraistat and Kay Walter. The summit was a chance for directors of centres and funders to talk to each other, to to develop collaborations, and to develop regional groups.

Neil Fraistat talked about the history and need for centerNet. centerNet held a American meeting in 2007. It now has 220 members worldwide. Out of that meeting came 3 mandates:

• Make the network international
• Make the network broad and inclusive
• Make the network free

A centre should be larger than a single project and have some hope for persistent. centerNet believes that centres can achieve more together than apart.

centreNet has two strategic alliances. One with CHAIN that brings together large infrastructure projects that promotes humanities cyberinfrastructure. The second is with CHCI to promote intersections between computing centres and interdisciplinary institutes in the humanities. We now have an official affiliation with CHCI. Jointly we have identified two directions

• Digital Disciplines
• Digital Publics

Asia Pacific Region

Jieh Hsiang from National Taiwan University presented a summary of the Asia Pacific region activities. Their main activities in the area of DH include:

• Building large-scale digital archives of cultural heritage
• Developing information infrastructure for digital contents
• Integration of IT and digital contents into teaching and research

They have had two meetings and formed a council. Dr. Hsiang summarized activities happening in Australia, China, Japan, Taiwan and Taiwan.

The region celebrates diversity while trying to find common ground including a common cyberinfrastructure. They are looking to integrate and share cross-cultural archives. They are developing regional workshops and conferences.

European Region

Jan Christoph Meister and Lorenza Saracco presented about the European region. The major stakeholders include:

• Traditional disciplines (both the empirical and hermeneutical disciplines)
• New disciplines like game studies and new media
• Computer science and informatics
• Libraries and infrastructure providers

In Europe there are national and regional centres (regions within nations) but less pan-European coordination. There are larger EU initiatives supporting infrastructure, archives, policy development, and virtual centres. In 2002 they developed the European Strategy Forum on Research Infrastructure which has developed a roadmap (2006). It is launching 44 projects.

There is a strong focus on eScience and infrastructure:

"eScience/eHumanities only to become reality when we are successful in building infrastructure to enable it."

They have a model that has three layers:

• Common data services
• Community support services
• Users and data generators

There are 5 projects in humanities and social sciences. DARIAH is one example, CLARIN another. Also important is CESSDA for archives.

UK and Ireland

David Robey talked about how there is a Network of Expert Centres that has been going for a while and is affiliating with centreNet (which explains why they aren't part of Europe.) Their funders are JISC and the Arts and Humanities Reserach Council. The AHRC used to support the AHDS and ICT Methods Network, but those were cut. The NEC is trying to rebuild a virtual institutional infrastructure http://www.arts-humanities.net . This is to provide a knowledge base and community support. The NEC is actively involved in DARIAH and CLARIN, but continued participation depends on the AHRC providing its national share of funding to DARIAH and CLARIN.

Canada and the USA

Michael Eberle-Sinatra talked about SSHRC and CFI in Canada. CFI funds infrastructure and SSRHC funds research. SSHRC is about to unveil a new architecture which should include a new international partnership program.

Brett Bobley talked about the NEH and the Digging into Data program in particular. He talked about how the humanities are international by definition and that therefore we need to build international infrastructure.

Other Funder's Lightning Round

• Bridging Cultures is a new NEH initiative to facilitate bringing people together to understand different cultures. There will be soon some new grant programs in this area.
• Alastair Dunning talked about JISC and their funding of digitization projects.
• DFG is a German research funding agency. Most of their funding goes into individual projects and centres. They have a joint program for digitization with NEH and an international training program.
• In France there is the CNRS. The CNRS supports ADONIS which is building infrastructure for French digital humanities.
• The ESF is not a funding organization, but an association of member organizations. They provide a bridge between policy and research. They do supervise workshops and ask their member organizations to fund them. They also help with communication and they have have committees to promote strategic directions.
• The Dutch Research Council supports national research institutes (none in the humanities, for some reason), research projects, and now a program on research infrastructures (of which only 5% is for the humanities.) There are also funds through other agencies.
• The Japan Society for the Promotion of Science covers not just science but also the humanities and social sciences. It supports international collaborations. They support international fellowships.
• The National Science Council of Taiwan supports research including international collaborations around digitization of cultural heritage.
• CIFAR (Canadian Institute for Advanced Research) is a private/public consortium that has funded for the most part the hard sciences. They have run a consultation to figure out how to fund the humanities. There are potentially some new initiatives comings.

One thing that stands out is the move by funding agencies to find ways to support international projects. There are nation-to-nation initiatives and programs in individual countries that allow others to participate, but we don't have a real international program unless one thinks of the EC/ESF work in Europe.

Breakout Groups

I was in a group of centres. We talked about common research areas including:

• Visualization and image processing
• Text encoding, archives, and digital editions
• Game studies, interactives, and new media
• High-performance computing
• Digital humanities skills and pedagogy
• Empirical studies of books
• Annotation tools
• Analytics
• Ontologies for cross-disciplinary database integration

We then talked about opportunities for collaboration:

• Graduate student internship
• Graduate student mentoring, review, and supervision
• Doctoral (or graduate student) consortium
• Site visits and program review
• Scholar exchange
• Expertise bank
• Discovery registries
• Data access
• Sponsored bogs, podcasts, tweets, and other born-digital types of communication and collaboration

Problems

• Money
• IP issues
• Cultural differences and language differences
• Differences in graduate programs
• Administrative barriers
• Lack of truly international funding opportunities or ways to hookup coordinated projects across boundaries

How could centerNet benefit our centres:

• It can validate and provide visibility to digital humanities
• It can be an advocacy body
• It can foster ideas in early seed stages
• It can matchmake between centres

Gathering

We then gathered and shared what our breakout groups discussed. Here are some of the points brought out

• Visibility - valuing digital humanities research and explaining it to a larger community
• Optimization - doing things in a more informed way and better coordinated way
• Sustainability and stability - finding ways to sustain projects so that they are stable.

There was a discussion in more than one group about top-down vs bottom-up models. Twitter can do more than dedicated efforts to broaden knowledge.

The general overlaps in research were organized into three groups:

• Methodological - types of analysis, crowdsourcing as a method, use of markup or databases, and visualization
• Topical - history, english, languages, theatre, art,
• Outcomes - software development, infrastructure, resources and datasets, and e-learning

A suggestion for collaboration was to learn from open projects that can bring in individuals at other nodes. Centres can help each other with advocacy, sharing case studies, strategies for promoting research, documenting policies and procedures, advice, access to resources, standards, and training.

Ideas for funders included using centres and libraries as resources for curation of digital resources. Funders could encourage good practices in preservation. Centres could be used as a resource for proposal reviewing - centreNet could develop a registry of expertise for funders. Funders could promote international collaboration, student exchange and training.

The funders felt that digital humanities are important but they don't want to fund them as centres so much as fund projects and infrastructure. They discussed the lack of esteem for digital outcomes from research (the website vs the book.) They believe centerNet should be involved in advocacy and policy. centerNet could paint a picture according to digital humanities so that individual countries could see what they are missing.

Regional Groups

We then broke into regional groups.

There is a lots of enthusiasm for regional events (regional in the sense of North-East USA). One idea is for centres to write into grants funding for consultants from other centres.

Housing can be a problem for visiting scholars. Centres could try to make it easier for people who want to visit by documenting basic things like how to get housing. How can we make it easy for visitors, even without funding?

We talked more generally about serving a discovery function such that people could find out what expertise is where. Realistically this would take a lot of work to maintain - how do we now develop such things so that they can be funded. Can discovery functions be divided into smaller parts such that individual centres could take some on?

We should think of a model where centerNet calls for initiatives or proposals in areas where we would value coordination. centerNet could call for proposals to develop a directory and authorize a center to apply to develop a registry.

Centers nearby have developed activities together like forums, training, and grants. We don't however have much collaboration over larger distance, for example we don't have many partnerships with Mexico.

We should think about ways to develop satellite events that are accessible to graduate students. We should think about THATcamps and other unconference types of events that are more inclusive.

centerNet should think about taken a stronger advocacy role. Funders of all kinds need a community to speak up and represent the needs of researchers. centerNet could also advocate with elected officials.

Funders could think about grants for finishing projects and think about as many collaborative research programs as possible.

Day 2: Talk by Jon Orwant

Day 2 started with a talk by Jon Orwant from Google Books. He talked about how Google is new to supporting the humanities. They are convinced there is a new style of research based on large-scale which applies not just to data but to text.

At this point they have scanned 12 million books, 5 billion pages. More than half is non-English. Now that we have all this data, what can we do with it. Google is interested in hypotheses that can be tested on large amounts of data as they have with Google books. Jon described a linguistics project testing regularization of verbs. Now they are trying to figure out scale out these types of projects.

One approach is to identify subcorpora that Google can extract so that researchers can crowdsource the correction and analyze it. The next phase would be to let people automatically extract subcorpora to work with and the last step is to figure out how to let people run their computations directly on the Google books corpus.

Jon talked about the importance of rewarding the developers of tools. Google wants to not only support research, but also support the developers of tools. They don't think you should support top-down work so much as bottom-up work. Standards are not worth effort compared experimentation. John doesn't want to see any particular standard in a proposal, but that there be a structure so that others can learn from the project and solve related proposals.

Jon talked about an exemplary proposal from Dan Cohen looking at changes in religion in the Victorian age. Cohen's project was an example of a hypothesis that could be tested on lots of data that Google has. Some problems they say in proposals were:

• Some researchers wanted to do things that were naive (as in easily using natural language processing)
• Some proposals wanted to do things where the details were a problem
• Some researchers wanted access to parts of the corpus that Google couldn't share for legal reasons

Google will soon announce the results of the first round. They are not committing to another, but may have one a year from now.

There was a good question about the diversity of data formats - has the experience of the digital humanities been that we need some top-down data format standards to ensure interoperation.

Regional Summaries

We then had reports from the regional reporters. I reported on the conversation among the North American region (see above.)

The Europeans are dealing with fragmentation. They believe that there should be regular meetings. The meetings should deal with teaching issues, summer schools, project development, and thematic clusters. Bringing regular meetings would help weave them together. They also felt that these meetings should be face-to-face, not virtual. They imagine a rotating meeting that goes from country to country.

Language can also be a barrier as in some countries even using an English term like "digital humanities" raises hackles. They need to solve the practical problem of language.

There is a strong need for database of projects in Europe, but this will have to wait. They need project classification schemes for this and for funding agencies.

In terms of cooperation they think it is best to start with bi-national projects rather than multi-national cooperation.

Suggestions of how to work with funders included the need for a common classification scheme. They also think it is important to talk early and often - and involve funders in design process of projects.

There is a dominant focus among funders on infrastructure that gives the false impression that digital humanities will emerge naturally if the infrastructure is perfect.

CenterNet centers could get involved in policy and strategy documents.

The Asia/Pacific region agreed that they all have the same problem that is to explain to funders what the digital humanities is as it is such a new field in their region. They need to also convince their colleagues about the importance and excitement of the field. The first and most important thing for this region is promote the awareness of this field. CenterNet can help with this and can provide independent reports to funders about the field. It was also suggested by some of the funders that they should advertise the interdisciplinary nature of the field and the social impacts.

They also discussed the importance of scholarly impact and the need for more journals in the field.

They see the need for mutual visits and internships for graduate students. They also talked about resource sharing as there is a fair amount of similarity between, for example, archives. They should also think about multi-lingual archives. They could also share citation tools.

Finally they talked about an annual digital humanities conference in the region that could include a doctoral consortium.

The UK Ireland meeting noticed that they do have a network of centres. One barrier to collaboration is that centres are often in competition with each other. They need financial incentives to encourage collaboration. Funders should encourage collaboration.

They believe that funders should insist on sustainability through using standards and existing infrastructure. They did note that many institutions don't have, or can't afford to build, big repositories or digital humanities centres. How can funders and centres help those at places without big support.

They tried to think of some cost-effective ways to help:

• Sharing expertise - provide experienced help to starting centres.
• Funding for networks or workshops would be useful

What could centers do to help funders? They felt that evidence of value would help funders. They could help projects show their impact.

Finally, they feel funders should do more to check conformance to data-management plans.

Breakout Groups

We then had a breakout group to talk about outcomes. We talked about the short term opportunities:

• centerNet is now real
• Dialogue with the funders
• centerNet is useful for exchanging views across regions
• Meeting with funders is a real outcome - the opportunity to show funders the global scope of this field
• Meeting with people within the region was helpful
• Could centerNet use the manifesto that came out of France? Can centerNet develop and promote strategic documents?
• It would be helpful to build a map or registry of projects and the field - not top-down or bottom-down, but middle-up and down
• We could develop a graduate consortium
• It would be useful to develop a dialogue around core skills and curriculum
• We could develop a series of strategic documents
• We can discover a lot through the narratives from other regions
• Evidence of impact could be documented
• Digital humanities needs to work with library - centerNet need to include more libraries
• It is useful to have national networks in addition to regional and international networks
• Increasing the visibility of the field
• centerNet should think about affiliating with ADHO

We talked about longer term outcomes:

• Encouraging interdisciplinarity and multi-linguality
• Involving other types of funders
• Developing a culture where centerNet doesn't do things, but supports centres to do things
• Developing opportunities for shared and hybrid instruction and supervision - overcoming the administrative barriers
• We need to develop a democratic structure for centerNet
• We need to encourage interoperation and openness
• What would centerNet look like if digital methods become general in the humanities?
• centerNet could facilitate dialogue for and around strategic policies.
• centerNet could assist with universities developing
• centerNet can help think critically about the university and scholarship
• We need to think through what centerNet should do that other organizations like SDH or ACH don't.

What steps are most important next? Our group identified the following:

• Regional meetings, workshops and conferences
• Form a subgroup to identify opportunities and call for initiatives
• Translation - make the web site multilingual by December
• We need to adopt a strategic principle that we don't publish strategic documents in only one language

Next Steps

There was some general discussion about what centerNet can be and how it is different about other ADHO groups. At the end we identified the following workgroups for which we hope people will volunteer. These workgroups are open to all. At the meeting we identified individuals to start discussions in these areas.

• Business Plan and Governance
• Advocacy
• Multilingualism
• Consultancy "SWAT" Teams
• Evidence of Value and Impact of DH Projects
• Internationalization
• Education, Graduate Training and Curriculum
• Social Media and Web Presence

After the meeting we heard from Lorna Hughes and Arianna Ciulla that the ESF will be funding a European group called NADIMA around methods.