"Cyberinfrastructure for the Humanities and Social Sciences"
John M. Unsworth
Chair, Commission on Cyberinfrastructure for the Humanities and Social Sciences, American Council of Learned Societies
Dean and Professor, Graduate School of Library & Information Science
University of Illinois at Urbana-Champaign
Research Libraries Group Annual Meeting
National Gallery of Art
2:15-3:15, April 26th
In January, 2003, a blue ribbon panel appointed by the National Science Foundation and led by Dan Atkins, of the University of Michigan, completed a report called "Revolutionizing Science and Engineering through Cyberinfrastructure" (http://www.communitytechnology.org/nsf_ci_report/). This report is a kind of provocation for the ACLS Commission, and there is a lot of other activity of this sort, going on right now—for example:
Digital Archiving and the National Archives and Records Administration (2004)
NSF "Post Digital Library Futures" report (2003)
NRC "Beyond Productivity" report (2003)
The United Nations World Summit on the Information Society (2003, 2005)
In a press release that followed the publication of the Atkins report, Peter Freeman, the assistant director of Computer and Information Sciences and Engineering (CISE), the NSF directorate that commissioned the report, said: "The path forward that this report envisions. . . truly has the potential to revolutionize all fields of research and education." Certainly, the report has had a significant impact on the rhetoric, and perhaps also on the priorities, not only of CISE, but also of other parts of NSF, and on other funding agencies concerned with information technology as it supports research.
So, what is cyberinfrastructure? Here's how the Atkins report addresses that question:
"The term infrastructure has been used since the 1920s to refer collectively to the roads, power grids, telephone systems, bridges, rail lines, and similar public works that are required for an industrial economy to function. Although good infrastructure is often taken for granted and noticed only when it stops functioning, it is among the most complex and expensive thing that society creates. The newer term cyberinfrastructure refers to infrastructure based upon distributed computer, information and communication technology. If infrastructure is required for an industrial economy, then we could say that cyberinfrastructure is required for a knowledge economy. " (5)
So, cyberinfrastructure is the infrastructure for a knowledge economy. And why should we care about it? Well, we all live, and will continue to live, in that knowledge economy, so we all have at least the same interest we would have in good roads and bridges, good telephone systems and power grids. And why should the humanities and social sciences care about it? Because we can make it a better infrastructure, if our perspectives, our training, and our expertise are included in its design and deployment. After all, science—whose goal is predictive certainty—only has half the picture. Uncertainty (or ambiguity, if you prefer) is the other half, and the humanities and social sciences celebrate that, explore it, tolerate it, and understand it better than the sciences do. Or, at another level, if science and engineering are about what we can do, the humanities and social sciences are about what we should do. If we don't know what we can do, we don't know what choices to consider, but if we don't know what we should do, we don't know which choices to make. Cyberinfrastructure is no different, in that respect, from atomic energy, biotechnology, or any other challenge: it is not only a scientific challenge, with scientific outcomes: it is also a social and human challenge, with outcomes that the humanities and social sciences are best equipped to understand.
The "overarching finding" of the Atkins report
. . .is that a new age has dawned in scientific and engineering research, pushed by continuing progress in computing, information, and communication technology, and pulled by the expanding complexity, scope, and scale of today's challenges. The capacity of this technology has crossed thresholds that now make possible a comprehensive "cyberinfrastructure" on which to build new types of scientific and engineering knowledge environments and organizations and to pursue research in new ways and with increased efficacy. Such environments and organizations, enabled by cyberinfrastructure, are increasingly required to address national and global priorities, such as understanding global climate change, protecting our natural environment, applying genomics-proteomics to human health, maintaining national security, mastering the world of nanotechnology, and predicting and protecting against natural and human disasters, as well as to address some of our most fundamental intellectual questions such as the formation of the universe and the fundamental character of matter."
I agree with all of this, and I'm certain that Dan Atkins, and many other scientists, would agree that along with all this new knowledge, these new certainties, will come new uncertainty, and new quandaries, that science itself, by itself, will not be able to resolve. But if the humanities and social sciences want to have some influence in the process now underway to design our information technology environment over the next decade, then we need to articulate our needs and our potential contributions—and even more than that, we need to articulate the importance of the humanities and the social sciences, for the amelioration of the human condition. That's something we haven't done very well since progress displaced enlightenment as our culture's highest value, and it's something that we still don't do very well.
For example, in the "Summary of 2003 Fiscal Year Budget Request," the NEH argues for the humanities as follows:
In the 1965 legislation that established the National Endowment for the Humanities, the Congress of the United States declared that "democracy demands wisdom and vision in its citizens" and posited that "promoting progress in the humanities" was the surest route to such wisdom. . . . The National Endowment for the Humanities helps Americans develop wisdom and vision through the study and contemplation of the record of human thought. The study of history, literature, languages, philosophy and other humanities subjects help us not only to better understand our own nation, but other cultures as well."
All well and good, and I believe also true, but unfortunately we in this country seem to believe that wisdom is a great deal less expensive than knowledge, and knowledge—especially knowledge with practical consequences—is what we're willing to spend money on.
But the humanities and the social sciences have access to knowledge that does have enormous practical consequence, and the future will be better, or worse, depending on whether that knowledge is part of our "knowledge society." In order to be a meaningful part of that society, I would argue, the humanities and social sciences will need computational methods and they will need access to the kind of vast datasets that make computational methods both necessary and useful. Computational methods already have a place in the social sciences, and they have a foothold now in the study of literature, history, art, and other humanities disciplines. The humanities, in particular, has been without a galvanizing methodology for a generation now, and it is being, and will be, revolutionized by information technology as profoundly as any of the sciences.
That revolution, though, has some preconditions: it requires a motivating factor, to move the disciplines toward new methods, and it requires the means—both intellectual and financial—to adopt, refine, and disseminate those methods for the rising generation of scholars.
The motivation, if it comes, will come in the form of very large datasets that can only be manipulated and interpreted with the aid of computers. We're getting there, already, with digital libraries, and if projects like DLF's distributed digital library come about, we'll seem suddenly to have arrived. But even that scale is not quite what I have in mind. To arrive at terabytes or petabytes of humanities or social science data, we will have to effectively address two issues—intellectual property (the primary data-resource-constraint in the humanities) and privacy rights (the primary data-resource-constraint in the social sciences). In a sense, as John King pointed out to me, this is no different from the struggle, in computational science, with resource constraints on memory, bandwidth, or processing speed: unless we radically increase these resources, we radically limit the kinds of questions we can ask and answer. We generally think of intellectual property and privacy rights as legal issues, which of course they are, but short of sweeping legal remedies, which I don't expect and don't actually desire in either case, I think the solutions to these problems will be, to some significant extent, technical. That makes them, in effect, the primary "cyberinfrastructure" research agenda for the humanities and social sciences. If we tackle that agenda successfully, the humanities will have access to the full recorded history of the 20th century—music, film, text and image, all in digital form; the social sciences will have access to the full record of societies, populations, individuals. There is tremendous danger of abuse, here—as there is with any other research that has profound practical consequences—but there is also a tremendous opportunity to learn, to understand, even to achieve some wisdom. Lest that emphasis on big datasets sound too much like "rugged informationalism," let me emphasize that I agree with David Weinberger that "what the world needs [is] people who know how to manage metadata, navigating the twisty darkness of the ambiguous world while preserving the value of the unspoken." I would just add that if we only have access to metadata, and not to the data itself, we won't be able to really plumb ambiguity's "twisty darkness"—but I would also agree that "preserving the value of the unspoken" is one of the major challenges for the humanities, especially, as it grapples more deeply with computation and information science.
If you look at where the funding for cyberinfrastructure and the research programs that support it is coming from, you might conclude that not all communities, not all classes, and not all categories are going to be equally well served in its design and deployment. To begin with, and lest we lose sight of this, the largest investment will be from the commercial sector, and it will take the form of developing products, not doing basic research, much less doing education. A knowledge society implies an information economy, and we have already seen that the owners of large caches of information—television and film studios, the recording industries, publishers—are not eager to achieve technical solutions to the problem of restricted access.
In the area of basic research, most of which is now done in universities and colleges, the big dog is health research, which accounts for more than half of federal spending on research: the NIH's budget in 2004 was around $28B. By comparison, the NSF's budget was about $5.5B (an increase of $171M over 2003), within which CISE represents about ten percent, or $584M. By comparison, again, $584M is about what the largest private foundations give away in a year, in this country—with the exception of the Gates Foundation, which gives away about twice that. Descending the scale of funding and influence, the 2004 budget for IMLS was about $262M (roughly half of the budget for CISE); the budget for NEH was $162M (less than the increase for CISE, over 2003); bringing up the rear, the budget for NEA was $139M—just under half a percent of the budget for NIH. In fact, add the budgets of NEA, NEH, and IMLS together, and you won't quite equal the budget for CISE, which is one of the mid-range budgets in NSF.
"Cyberinfrastructure" is more than just hardware and software, more than bigger computer boxes and faster wires connecting them. The term describes new "research environments" in which disciplinary experts, in interdisciplinary teams, supported by specialized computational support staff, have global, instantaneous access to enormous computing resources. And although the redaction of the Atkins report in NSF presentations subsequent to its publication has tended, in my view, to emphasize only that last point—enormous computing resources—if you read the report itself, you may be struck, as I was, by its emphasis on human resources, on organizations, and on education and training, for example in passages like this one:
"This vision of science and engineering research involves significant educational dimensions. The research community needs more broadly trained personnel with blended expertise in disciplinary science or engineering, mathematical and computational modeling, numerical methods, visualization, and the sociotechnical understanding about working in new grid or collaboratory organizations."
Here's another such passage:
"Human resources are critical to making cyberinfrastructure and applications work, keeping them working, and providing user support. In the interest of funding more grants, NSF has arguably under-supported the recurring costs of permanent staff, preferring to focus resources on direct research costs and 'hard' or 'tangible' assets. In the ACP, human resources are the primary requirement in both development and operations, and success is clearly dependent on adequate funding both in centers and in the end-user research groups."
What's being said, in these passages, of the importance of discipline-specific computational support in the sciences could also be said of computational humanities or social science. And with respect to the nature of research projects themselves, it would be as true, these days, of the humanities as of science to say that "many contemporary projects require effective federation of both distributed resources (data and facilities) and distributed, multidisciplinary expertise, and that cyberinfrastructure is a key to making this possible."
What the Atkins report says—that one could not say of the humanities or perhaps even of social science—is that "prior investments provide a sound foundation for the ACP." In fact, prior investment in cyberinfrastructure for the humanities and social sciences is tiny, by comparison to what it has been in the natural sciences, and that puts us in a rather different position, with rather different needs. In addition to the two "resource constraints" I discussed earlier—intellectual property and privacy rights—we have another resource constraint, and it is a constraint on a human resource, namely those disciplinary computational specialists. Education and training, therefore, must be even higher on the agenda for the humanities and social sciences than they are for computational science or computer science itself. Schools like the one I recently moved to, schools of library and information science, are our best bet for producing those specialists, I believe, and I firmly believe (with Margaret Hedstrom and John King) that libraries are one of the principal places that they will do their work. I also think, though, that we will have to reassert, more generally, the importance of mathematics in general education and in the liberal arts curriculum, beginning as early as middle school and high school. We will need English majors who have a background in logic, who can handle statistics, who do math, if we are going to turn out a generation of disciplinary specialists who can bring the accumulated wisdom of the humanities to bear in computational contexts—perhaps in helping build ontologies for scholarly projects in disciplinary contexts, or building tools for data-mining in the context of humanities research. I've met a few of them at the University of Virginia, and even graduated a couple with Ph.Ds in English who have gone on to tenure-track appointments as humanities computing specialists in English departments; I'm meeting and graduating a lot more of them now, at the University of Illinois' Graduate School of Library and Information Science. These newly minted scholars, some of whom are specialists in disciplines of the humanities, or social sciences, and some of whom are specialists in information science, have arrived at that expertise without abandoning mathematics and logic. Consequently, they have absorbed and naturalized computational methods, and they hunger for more data. Given the necessary resources, they will—I am convinced—find novel ways to bring their disciplines to bear on the uncertainties, the quandaries, the moral and aesthetic challenges, as well as the practical problems, of "the knowledge society."
Well, those are some of my starting hypotheses, as chair of the ACLS commission on cyberinfrastructure for the humanities and social sciences. I expect that the work of the commission, over the next year, will test those hypotheses in various ways, and will alter them as a result. I present them here, not as predictions of the commission's outcomes, but to mark my own starting point, and I would welcome your response to any of the points just raised. The Commission itself will meet tomorrow morning for the first time, with each other, and tomorrow afternoon, with you. During the coming year, it is charged to:
It will attempt to accomplish this by a combination of public meetings, with invited testimony as well as open discussion, presentations at conferences and professional meetings, a web-based survey of the broadest possible cross-section of humanities and social science scholars and students, and through the solicitation and consideration of written input from experts unable to testify at public meetings. We will also present the draft of our report for public comment, and we expect to revise that report as a result of the public response to the draft. The dates and locations of the public meetings are outlined in a handout in your registration packet, and that handout also names the members and advisors of the Commission. You are invited not only to respond with questions and comments now, and tomorrow afternoon, but also to address written comments, especially, to any member or advisor of the Commission, or by email to firstname.lastname@example.org, which comes to me.
I'll close by observing that there is a kind of ten-year cycle to the sort of thing we're doing. Ten years ago, it was the National Information Infrastructure, and various commissions and committees around that term, that sparked a good deal of the priority-setting and decision-making that set the research agenda for the next decade. The humanities and the arts were a small part of that conversation, and there were some outcomes from that, but I hope this time around the engagement is a more profound one, and I hope the outcomes are more lasting. I also hope that the bridges we build in this process are bidirectional, and encourage collaborations and provocations that finally unite CP Snow's two cultures, and deconstruct that binary opposition once and for all.
ACO*HUM (Advanced Computing in the Humanities, sponsored by the European Commission, published in 1999)
The National Information Infrastructure: Agenda for Action (1993)
Ron Brown, Secretary of Commerce and Chair, Information Infrastructure Task Force
Humbul Humanities Hub
Voice of the Shuttle
H-Net: Humanities and Social Sciences Online
NEH-funded online projects
Association for Computers and The Humanities
Association for Literary and Linguistic Computing
National Initiative for Networked Cultural Heritage
Consortium for Computers in the Humanities/Consortium pur ordinateurs en sciences humaines (coch-cosh)
Association for Computational Linguistics
The American Association for History and Computing
Conferences and Publications:
Culture, Creativity and Information Technology
Social Sciences Research Council
History and Geography: Assessing the Role of Geographical Information in Historical Scholarship (2004)
Digital Resources in the Humanities
Inaugural Conference on Computational Social Science (2003)
[and see The GIS History Project (1996), http://www.geog.buffalo.edu/ncgia/gishist/, also
Past Time, Past Place: GIS for History (ESRI: 2002)]
ACLS Occasional Paper No. 41, "Computing and the Humanities: Summary of a Roundtable Meeting" (1998)
ACLS Occasional Papers no. 36, "New Connections for Scholars: The Changing Missions of a Learned Society in an Era of Digital Networks" (1997)
Institutional Models for Humanities Computing
Jahrbuch fˆºr Computerphilologie 4 (2002)
Humanities and Arts on the Information Highways
Malhotra, Yogesh; Abdullah Al-Shehri & Jeff J. Jones (1995). National Information Infrastructure: Myths, Metaphors And Realities [WWW document]. http://www.brint.com/papers/nii/