Toward an Interactive History of Science and Technology: Reflections on the Dibner/Sloan Web Project

Paper presented at the LXV plenum of the Russian National Committee of the History and Philosophy of Science and Technology, Moscow, May 2002

by Slava Gerovitch

There is nothing outside the Web.

- virtually Derrida

History Ex Machina

In May 2001, I gave a mock talk at the MOCKolloquium at the Dibner Institute for the History of Science and Technology, entitled "The Automatic Historian.". In that talk, I proclaimed the end of cut-and-paste history and the dawn of a new era of point-and-click history. Every joke has a grain of truth, and so did this one. A year earlier, in May 2000, the Dibner Fund and the Alfred P. Sloan Foundation launched a three-year project, History of Recent Science and Technology on the Web (HRST), hosted by the Dibner Institute and led by Jed Z. Buchwald. The main purpose of the project is to explore ways to cope with the phenomenal scale and complexity of modern science, which makes the traditional mode of doing history virtually obsolete. It becomes increasingly difficult, if not impossible, for individual historians to navigate in the sea of information produced by scientific and technological activity in recent decades, to select relevant pieces, to analyze them in a timely manner, and to produce comprehensive interpretations.

The proposed solution has two aspects: automation and collaboration. The project attempts to raise the productivity of historical work by using the latest Web technology and by enlisting scientists and engineers as collaborators. Instead of an individual historian, who used to go about collecting archival material, interviewing specialists, and writing conference papers, the center of historical research activity now shifts to a website. The HRST website serves as a digital archive of texts, images, audio, and video information; it is also equipped with sophisticated tools for searching, expanding, editing, and commenting on this diverse pool of historical data, including special software modules for the development of online multi-thread timelines, multi-subject bibliographies, and multi-party interviews. The website contains a complex hierarchy of permission levels; it provides different degrees of access to various parts of the site for different users and makes it possible to organize small working groups that can share data and comments among themselves.

A great hope of the HRST founders was that this website would be used primarily by scientists and engineers, experts in their own fields, who would select most important primary sources and deposit them on the site, compile timelines and bibliographies, add comments, and organize online discussions. Historians working on the project would largely serve as facilitators, helping the primary users to cope with the intricacies of the Web technology. The website itself would serve as the true engine of historical research. 

The main HRST site consists of five projects devoted to specific historical topics: Bioinformatics, Materials Research, Molecular Evolution, Physics of Scales, and the Apollo Guidance Computer. All five projects share the same software, but their approaches are somewhat different. Each group of historians develops its own strategy in dealing with the target audience of scientists and engineers whom they invite to interact with the site. Bioinformatics, for example, is a relatively narrow research field located at the intersection of academic and business circles. The Bioinformatics website has advanced graphics, easy tools for the quick upload of digital archives, and game-like controls for manipulation with timelines. Materials research, on the other hand, involves a very large and diverse group of scientists, engineers, and industrialists scattered around the globe. Historians from the Materials Research group often go abroad to conduct interviews, and they have published several analytical essays on their site, trying to engage their collaborators in critical reflection over the history and status of their discipline. The Molecular Evolution group works with a small number of leading researchers living in different countries. They have posted crucial primary sources on their site and organized an online discussion around them. The Physics of Scales group also deals with a small number of leading researchers scattered around the globe. They have posted proceedings of important conferences in their field and conducted a number of live interviews. The Apollo Guidance Computer group works with a small number of engineers, now mostly retired, who worked together on a big project in the 1960s. This group organizes off-line activities, such as workshops and live interviews, then posts their transcripts and digital copies of archival materials on the site, and attempts to generate discussion around these materials. As a member of this group, I am responsible for the section devoted to Computing in the Soviet Space Program. All five groups have carried out important research, obtained unique materials, and raised significant historical questions for discussion. 

As the HRST project progressed, however, historians began to face significant difficulties in their attempts to engage scientists and engineers in direct online participation. Some scientists found the interface too complex, some did not have Internet connection, some were too busy, and some simply were not motivated enough to devote their time to historical reflection. Arne Hessenbruch, a member of the Materials Research group, has summed up typical problems in his article Compromises and Opportunities: History of Recent Science and Technology on the Web. Describing his efforts to attract attention of the participants of the Fall 2001 Materials Research Society (MRS) meeting, he writes: 

Individual attendants at the conference showed no interest whatsoever. I had booked a presence with a poster describing our project and inviting people to contribute to the website. With the help of Babak Ashrafi I also had a laptop hooked up to the internet, so that conference visitors could browse the site and respond to an interview set up for the purposes on the spot - we also handed out flyers containing help on how to do the interview from home. The MRS advertised our poster session very generously in the conference program, the website, and on posters throughout the conference building. But out of the approximately 4,000 attendees, a big fat zero responded. 

Hessenbruch concludes that "in order to involve living materials researchers we have to present them with something that is geared more to their interests. Our own historical questions will not do." He proposes to work closely with professional societies, such as MRS, asking them to stimulate collaboration between scientists and historians and even delegating them the choice of interviewees and the selection of questions. 

Although this strategy may lead to some local advances in this particular complex project, it seems that in general this strategy fosters marginalization of historians in a Web-based historical study. If professional historians no longer select, organize, and present material, pose questions for study, or add critical reflection to the flow of reminiscences, the academic standards of such a study would be hard to maintain. Furthermore, I do not believe that there exists a fundamental division between "our historical questions" and "questions that interest scientists." Scientists usually show great interest in historical interpretations, especially if they are challenged with alternatives to their own vision. 

In my view, the problem of collaboration in a Web-based historical study is tightly connected with the problem of automation, and they must be solved together. If a research project requires collaboration, it must be human collaboration. Facing the HRST website with its entire arsenal of sophisticated online tools, scientists often still prefer email communication with historians. It is not the website that mediates between the scientist and the historian, but rather the historian who mediates between the scientist and the website. 

As I argued in my article on automation in The Encyclopedia of Computer Science, automation "does not simply transfer human functions to machines, but involves a deep reorganization of the work process, during which both the human and the machine functions are redefined." The HRST website did not so much simplify as it transformed the work of historians. Their job was hardly made easier by the use of the Web; instead, it became further complicated: now they not only had to collect historical material, as before, but also digitize and post it on the site. It was historians, rather than scientists and engineers, who uploaded most of the digitized sources on the sites and posed most of the questions for discussion. Historians had to organize and display their findings not in the process of writing a book, as usual, but "in real time", in the course of the project. Their role proved crucial for organizing, stimulating, and maintaining scientists' and engineers' interaction with the site. For example, an online discussion organized by the Molecular Evolution group received a crucial impulse only as the result of a provocative intervention by a historian (the very same Arne Hessenbruch). 

Historians had to learn complex technical skills and new forms of collaboration. The most difficult transition they had to make was the shift to a new medium for knowledge transmission, the transition from the book to the website. 

The Book vs. the Website

In his article, Hessenbruch raises an important issue of fixity and trustworthiness of Web publications. He compares the website to the book, and concludes that "the management of digital data is not fundamentally different from that of print media: both require an infrastructure and resources for routine maintenance." The knowledge that the first printed publications in early modern England faced the same problems that now afflict the Web is reassuring, but it does not help us position the website vis-à-vis the book in today's social world. The academic book industry already has a well-developed system of scholarly standards, a set of conventions for indexing, footnoting, and referencing, a system of manuscript peer-review, and a certain amount of trust placed in the names of well-known university presses. The book has already achieved the dominant status in academic circles, and the website has a very long way to go before it can compare to the book on any of these criteria. 

In my view, however, there is no need to try to beat the book in its own game. The website has its own, unique role to play, and the Web provides means for a new kind of writing well suited for this role. Instead of trying to eliminate some of the obvious differences with the book, website creators might try to make the most of these differences to generate new types of texts and new modes of interaction with the text. 

For example, website content is not completely fixed, and one might see this as a drawback compared to the fixity of the book, printed in thousands of identical copies. On the other hand, this dynamic quality of Web publications opens endless opportunities for correcting errors and adding new material to the existing text, making it a growing, living entity. Through the mechanisms of indexing and hyperlinking, the addition of every new page to a website implicitly or explicitly effects other pages on that website, producing a cumulative effect. Perhaps a couple of generations down the line people will be staring at a regular book in disbelief, wondering what is the use of such a fixed object which cannot be easily modified and transformed. 

Web publishing also promises to liberate writing from its unidirectional, "linear" characteristic, much reviled by the warriors against logocentrism. Instead of prescribing a particular order of reading, as in a book, a website potentially allows for different paths through the same content. Graphic Web design provides opportunities for an effective use of images, which are inherently non-linear. By shifting visual accents on a page and offering various options for navigation, one can avoid telling predestined stories or imposing explicit hierarchies. One can give a greater navigation initiative and interpretative power to the reader. For example, instead of using a list, one can organize entries in a circle or a hexagon, as on the Materials Research site. There is no single "standard" reading of a website. One can present different interpretations without privileging one of them. 

Most importantly, web publishing is interactive. The HRST software makes it possible to put "add a comment" button at the bottom of any page and then to display the original text along with the entire chain of responses, additional hyperlinks, images, audio, or video. This feature is not currently used to its full potential. For example, Hessenbruch's article, which raises important issues for discussion, is presently posted on a webpage that does not allow for comments. It is truly difficult to organize collaboration when a historian is giving a monological speech. The Web as a whole represents multiple voices, and this medium works best if this quality is maintained even in its smallest parts. [After reading this remark, Hessenbruch added a discussion forum, which made it possible to post comments. - S.G.]

New technical tools, narrative devices, and social interactions facilitated by the Web open for historians new exciting opportunities. Taking full advantage of such opportunities suggests a completely new mode of operation for the historian. 

From Automation to Interaction: Some Suggestions

In my view, a collaborative history site could be most productive as a place of interaction not only between historians and scientists, but also - most significantly - among historians, especially if they are separated by institutional and national borders. Ordinarily historians of recent science collect their primary materials themselves and keep their findings to themselves, using only small excerpts in their publications. After they are finished with a topic, these materials are discarded, and the next generation of historians has to start all over. If primary materials are placed on-line and preserved indefinitely, a new type of relationship could be established among historians - not competitive, but collaborative.

A collaborative history site might play a role similar to that of a shared genome database. Specialists can use it freely and receive credit for their contributions. Gradually it could become an organizing nexus of professional activity. Individual projects would flow through it, adding to it and benefiting from it. Sharing primary sources could free up time for really interesting work - analysis and interpretation. When the focus shifts to interpretation, interaction among historians becomes crucial. Instead of scattering their efforts in time and space over many separate publications, they could correct and comment on primary sources and on each other's work "in real time."

The next phase of the HRST project could involve not just collecting more material for the site, but also working on historical hypertext essays, both individual and collaborative. The combination of a digital archive and a digital journal/discussion forum would enrich both sides.

A historical website opens up a unique possibility of doing cross-national research projects, which are usually impeded by linguistic and logistic barriers. My own work on the comparative history of onboard computers in the US and Soviet space programs, though still in its initial phase, shows significant promise of benefits from bringing together primary sources and interpretative perspectives from two different national contexts. I am trying to build an international community of scholars from the US, Russia, France, and Germany working on this comparative topic and using a common pool of primary sources in English and Russian (the latter backed up with translations into English). These days historians are paying more and more attention to comparative, cross-national studies of science. For example, the collection Science and Ideology: A Comparative History, edited by Mark Walker, is published this year by Routledge (I confess, I have an article in it). The HRST software team has recently added a Russian-language capability to the site; this work has to be expanded to achieve full functionality of foreign-language submissions and usability for foreign users.

Undergraduate students increasingly move their research from libraries to the Internet, and it is very important to establish examples of quality historical sites on the Web and to set standards for this type of discourse. The next phase of the HRST project could serve precisely this purpose.