It is becoming impossible to maintain a full record of the progress of science and technology. A large fraction of the scientific and engineering knowledge that humanity now possesses was discovered in our own lifetime (some nine-tenths of all scientific papers ever written were written by scientists who are alive today). That calls for a corresponding exponential increase in historical documentation--the preservation of source materials that we and our posterity require in order to understand the processes of modern discovery. Historians and archivists are making vigorous documentation efforts, but funding for their work has not increased over the past generation; thus decade by decade an ever smaller fraction of scientific work has been documented in even the most superficial manner. In entire fields there is little but the scientific papers themselves, which give a grossly incomplete and often distorted view of how the research was really organized and pursued; still less is preserved documenting the history of technology. One way to meet this challenge is to develop innovative methods using the technology scientists themselves have adopted.
We propose to experiment with using sites on the World Wide Web to gather historical source materials. The Internet is becoming the method of first choice for researchers to store and exchange information. Web sites offer the most flexible and appealing way to exploit the Internet. A Web site dedicated to a particular scientific development of the past could become a magnet attracting scientists themselves to submit unpublished documentation-- reports, and correspondence, autobiographical reminiscences, historical narratives, photographs, commentary on materials previously submitted, indeed information and analysis in any form. To explore this novel approach to documentation we propose to launch a cohort of independent sites that address a single highly important and explosively growing field: the earth sciences with their associated research technologies.
The sites will be constructed under the aegis of leading scientific organizations, namely the American Meteorological Society (AMS), the American Geophysical Union (AGU), and the Center for History of Physics of the American Institute of Physics (AIP). These organizations arewell placed to overcome the chief problems of using the World Wide Web. Web sites for historical documentation will require not only technical expertise, but a commitment to permanent accessibility of the structured information (transported to whatever future formats evolve). Each of the three organizations already has a Web server containing scientific publications which they will maintain for posterity. New Web sites are easily lost to view amid the chaotic growth of the medium. All three organizations have mechanisms for reaching the relevant populations of researchers. They can offer guarantees of value through well- established mechanisms, developed for their print publications, to certify the validity of their sites as reliable repositories for particular types of information.
Each of the three organizations will continue to develop the informational content of its Web sites independently. All will experiment with different structures in some respects. However, we will share the most difficult and expensive task, development of a common Web site "look and feel" and database software. Throughout this work, the staff will keep in close contact to learn from each others' experience.
The project will conclude with reports evaluating the experiment. The chief outcome will be a body of experience on how to construct successful sites (or, how not to construct them). Will this new technology make a continuing and widespread historical documentation effort feasible at reasonable cost? Each site will be judged successful if it attracts a substantial body of historical source materials, helps ensure the preservation of documentation that might otherwise be lost, and otherwise shows promise of providing future scientists, historians, and other scholars with the wherewithal to produce usable historical accounts of a particular topic.
It will be particularly important to determine whether further sites can be launched with little difficulty to document other topics. If our experiment succeeds there is every reason to expect that other research communities will be stimulated to document their own histories in a similar manner. Such volunteer effort could lead to a permanent habit, underwritten by the disciplinary societies and other scientific organizations, of documenting the history of their fields.
Evaluation will be overseen principally by committees of scientists and historians of science already established for other purposes: the History of the Atmospheric Sciences Committee of the AMS, the History of Geophysics Committee of the AGU and the Advisory Committee on History of Physics of the AIP. Each will be asked to review all the Web sites. We will thus have three final reports, from different perspectives, frankly describing the problems encountered and the success or failure of the particular approaches used.