Q You have championed the idea of reference-citation indexing for more than 4 decades. Its commercial success and diffusion throughout the scientific community must be personally very gratifying to you. But, those early years of trying to realize your vision must have been difficult, both personally and professionally. Was it difficult to persevere?
A It’s all relative isn’t it? At first, the SCI [Science Citation Index] was not a universally accepted idea. It required a lot of steady convincing over 2 decades of continuing education. The SCI became a “standard” bibliographic tool within 10 years. Another key question was affordability, which also depends upon perceptions of value. SCI was often perceived as expensive in part because so much else is free or subsidized, like MEDLINE. And many people could not recognize the ability to retrieve information using cited reference searching. Even natural language searching, which was incorporated into SCI quite early, was not traditional.
SCI has been a hybrid system for most of its life. Only during the first few years was it limited to the Citation Index proper. Today, natural language—that is, title searching—is widely accepted. MEDLINE picked it up about 10 years ago. But ISI introduced title searching early on. We added the Permuterm Index to SCI over 30 years ago to aid librarians in finding key references.
Q In our Web era, the notion of linking seems to be an easier concept for people to grasp. In my mind, some of the most successful products have been those that enable connections of some kind. In reference citations, we’re talking about connections to other scholars’ works. Do you feel that this era makes it easier to accept a product like SCI?
A Yes, we always described the SCI and the scientific literature as networks and often created topographic maps demonstrating document links. Derek Price published his classic paper, “Networks of Scientific Papers,” 35 years ago. Based on citation linking, ISI pioneered a number of projects, like tracing the history of science. And I am still working on that.
Q Were you involved in the development of the ISI Web of Science (WOS)?
A Somewhat, but the folks at ISI were fully capable of implementing it without me. I was in on some early discussions, but in fact the development of WOS was defined by the CD-ROM edition. It had to be modified to deal with the limitations of the Web. There are still search capabilities in the CD-ROM version that need to be included in the Web version.
Q But on the Web you can access 25 years or more?
A The full WOS covers 55 years of source literature, beginning with 1945. The CD-ROM version begins with 1980, and you can only search 1 year at a time. On the WOS, a related-records search covers 50-plus years simultaneously. Regardless of the age of a source paper, you can find related, earlier—as well as subsequent—papers. Many users don’t even understand how bibliographic coupling works, but it produces a highly useful result ranked by citation strength or relevance. That’s why I call it “hypersearch.” I’ve said that the SCI is the ultimate hypersearch product. I think Ted Nelson is credited with the notion of hypertext, but I doubt that he knew the SCI even existed.
Q You were ahead of your time, I think.
A Yes, insofar as few people recognized the retrieval potential of citation links. Bibliometrics, however, had been studied much earlier, but the SCI made the field of scientometrics a practical reality.
Q Were you influenced by Shepard’s Citations?
A When I saw Shepard’s for the first time, I recognized immediately that it provided the structure I was looking for. The SCI is a variant of the legal citator. But Shepard’s was limited to case law, not the literature of law. The SCI is an index to the literature, and it is at least an order of magnitude larger. Shepard’s for the legal literature did not exist until ISI started the Social Sciences Citation Index [SSCI]. We had been indexing law reviews in SSCI for a decade before Shepard’s began.
Q You’ve done many different things; your background is in science. You’ve been a chemist, information scientist, an editor, publisher, database producer. What do you consider yourself to be, foremost?
A A science communicator, but I have defined myself as an information scientist for 40 years. When I started my consulting business I called myself an information engineer. You don’t see that term very much. But someone in the Pennsylvania state government wrote me saying I couldn’t call myself an engineer because I wasn’t a graduate of an engineering school. Ironically, it wasn’t long after that I was teaching information science at [the University of Pennsylvania’s] Moore School of Electrical Engineering.
We have a parallel situation today with information architects. I have been told that the American Institute of Architects tried to prevent information architects from using that term. But the term was adopted too quickly for them to prevent it.
Q The information architects are an active special-interest group (SIG) within ASIS&T (American Society for Information Science and Technology) now. I also think information architecture was the topic of the ASIS annual conference last year and is continuing to this year, is it not?
A ASIS sponsored a midyear Summit Conference on IA [information architecture] in Boston. It was so successful that we immediately formed a new SIG group. We have a similar meeting scheduled early next year for the West Coast.
Q Let’s talk about ASIS&T, of which you are now president. Back in 1968, you successfully proposed the change of name from the American Documentation Institute to American Society for Information Science. Then, to energize the organization and reflect the changes in the industry, you just successfully pushed the change from ASIS to ASIS&T—adding “and technology.” What are your priorities for ASIS&T now, besides recruiting? [Editor’s Note: For more on the name change, see Paula J. Hane’s NewsBreak on page 24 of the October 2000 issue and at http://www.infotoday.com/newsbreaks/nb000828-1.htm.]
A A lot has changed since the last vote on this issue over a decade ago. Tom Hogan (president of Information Today, Inc.) was president of ASIS then. He was not in a position to wage an educational effort that would convince 75 percent of the voting members. Even so, a 57-percent majority voted in favor. In February, “only” 70 percent voted for the change and 85 percent voted to change the constitutional requirement to 67 percent.
The new name better reflects what is happening in society. Since information science was vague or ambiguous to many people, we not only failed to recruit new members, but also lost many who were mainly applied-information scientists. There were many new information professionals who had just never heard of us. The new SIG for IA is tangible proof that a large group of applied-information scientists did not recognize that ASIS was the right professional affiliation for them. The broader name will help us recruit, but a name change alone is not enough. We also have to provide the right programs. And much else in ASIS&T needs to change, including the SIGs, which need to re-evaluate their definitions.
Q It sounds like leading ASIS&T could be a full-time job.
A ASIS&T officers are all volunteers. We don’t have a paid CEO in the traditional sense. Our executive director in Washington [DC], Richard Hill, is the full-time person supported by a small staff. Most members cannot afford the time that the presidency really requires. Institutional backing or resources are important. Some “retired” members like myself can devote more time than others who have to earn a living.
Q You don’t sound very retired to me. Let’s talk about your work on The Scientist, of which you are president and editor in chief. It is available entirely for free on the Web (http://www.thescientist.com). Is it advertiser supported?
A Advertising is the primary source of income. There’s a controlled distribution of 52,000 print copies mainly to life scientists. On the Web, our appeal is more broadly based, often drawing in educated consumers. Over 75,000 researchers are signed up for our new e-mail distribution, which is growing, without promotion, by about 500 members every day. Almost half of that growth is overseas.
Q How is free Web availability of journals affecting publishers and database producers, like ISI, which has linking agreements with publishers?
A The Scientist was the first full-text, continuously published journal on the Internet—even before there was a Web. We made a deal with the National Science Foundation to put us on the AT&T server, and people used the old file-transfer protocol. My idea at the time was that we could eventually charge a tiny fee for each hit—say 50 cents an article—that with enough volume could work. That model has not yet been implemented. It is not easy to collect such small fees, so now we have the advertising-supported model. But there will be many models of Web use for the same content. Even consumers will pay for relevant content if it is affordable.
Q What about the high subscription costs for some of the commercial, scholarly, scientific journals versus some of the alternative initiatives like PubMed? There have been some published data about the profit margins for some commercial publishers being excessively high and libraries struggling with the high costs. What are your thoughts?
A Private enterprise has been involved in publishing scientific periodicals for centuries. At the same time, there were relatively low-cost, professional-society and government-supported publications. Recently, the government has once again become highly competitive. It is politically correct to offer free MEDLINE or PubSCIENCE service. In the marketplace, if someone has a monopoly, they will ordinarily behave like a monopoly. If nobody is willing or able to challenge them, they’ll charge what the market will bear. But, I think the tide’s already beginning to turn. Publishers and societies are adapting to the changing market place. Added value is the name of the game, and that is where firms like ISI can excel.
Q Do you think increasing numbers of scholarly journals will be published directly to the Web?
A Many electronic publications already exist, even more are inevitable. As with print journals, some will depend upon volunteer editors. Some of those folks will become entrepreneurs. Some won’t. I’ve known journal editors who started as volunteers, but when the journal evolved into a valuable property they became capitalists. Some professional societies make a lot of money from journals that often support other society activities. They have learned to expect institutional libraries to finance those activities. To offset this imbalance, government agencies will have to increase their level of “overhead” support.
But the vast majority of electronic journals are now also published in print. Everyone’s moving ahead with electronic licensing deals. Both librarians and scientists have to recognize that publication is part of the research process. We have to support larger library budgets just as we support greater funding of research. Why are we paying more for a journal? Is it because it’s growing? If so, then we can better understand price increases. Some journal price increases are clearly outrageous. Some publishers were clearly taking advantage of a tacit monopoly in certain areas. It’s very hard to compete with established journals, but publishing has always attracted new entrepreneurs. Now publishers are bundling print and electronic licenses, including access to the archives.
In the long run, if publishers deliver added value, they will survive. That’s the role publishers should play or they become superfluous.
As a scientist, if I am going to publish in a journal, I want to retain the right of self-archiving. Many publishers have already recognized that, as, for example, JASIS [Journal of the American Society of Information Science]. So as long as I maintain my own personal Web site, colleagues can consult my work without jumping through hoops. The individual university may have to think about whether it’s going to be the archivist for material published by its scientists. Neither publishers nor governments can guarantee that things will be archived forever. There is considerable anxiety about the lack of permanence on the Web.
ISI has managed, in spite of government and other competitors, to be innovative and provide a product that people are willing to pay for. ISI always had and continues to have competition in many forms. Collectively, the full-text journals present another piecemeal way to obtain information that previously they could only get from ISI. But the way this information is packaged is critical.
Q Yet, ISI is pursuing full-text linking agreements with publishers.
A It’s symbiosis. The readers of the full-text journals benefit from having the citation links both ways. ISI users benefit by having the links to the full text. ISI is trying to increase those links. ISI also has to be more and more clever with what it does with the indexed information. There are still many interesting ways to manage information and to make searching easier.
Q What would you choose to implement if you were at the reins of ISI right now?
A I am personally interested in pursuing the use of ISI and other databases for historical research. When you search a database, there are ways of presenting information that would make the result more meaningful historically. I also think that citation-in-context will be a big step forward. Henry Small (ISI’s head of contract research) did research on it. Now there are full-text systems for doing this that get you closer to the author’s thinking. I also think there is a need for systematic reviews of the literature.
Q What about some of the data-visualization software that’s out there?
A Visualization is clearly a very exciting field, but whether anything is practical yet is another question. I’m only aware of experiments. We’ve done co-citation mapping for 30 years. The system, called the Atlas of Science, later changed to Research Reviews, was abandoned by ISI for a number of reasons. To me it is still a valid concept. Someone else may pick up on it. You need a powerful system if you’re going to do visualization in real-time on the Internet. Incidentally, there’s a SIG on visualization in ASIS&T.
Q Are there any trends in information retrieval right now that impress you?
A I started out working on machine translation of languages. This is now called computational linguistics. I think they’ve made considerable strides. I use computer translations quite often. I’m rusty on some of the foreign languages that I once knew. I find SYSTRAN convenient to get the gist of a foreign article [http://www.systransoft.com].
I’m also looking forward to the day when all new, published books will be covered in the citation index. That isn’t happening yet. It isn’t a trivial project to implement and will require the cooperation of hundreds of publishers.
Q Let’s end by talking about some of the activities supported by the Eugene Garfield Foundation.
A It’s a “plain vanilla” foundation, which means it can only support tax-exempt institutions—everything from minority scholarships at Drexel [University], Beta Phi Mu graduate stipends, fellowships, lectureships, travel, history of information science, etc. We are also involved in supporting the homeless, civil liberties, and Research!America. I’ve just joined the board of Research!America. Encouraging the public to support NIH [National Institutes of Health] and NSF [National Science Foundation] is the best way to benefit biomedical research and the information world. Supporting research is supporting information. If there’s growth in research funding, then the information industry, including libraries, will benefit.
Q I’m sure that your many fans in
the library and information fields are grateful for your many contributions
and now for your continuing support. Thank you for your time today and
best of luck in your ongoing endeavors.
For more information about The Web of Knowledge, A Festschrift in Honor of Eugene Garfield, visit http://www.infotoday.com/catalog/asis.htm. The book will be officially released this month at the ASIS&T Annual Meeting in Chicago. Many of Garfield’s writings can be found at his Web site at http://www.garfield.library.upenn.edu.
For more information about ISI, visit http://www.isinet.com.
Paula J. Hane, co-editor with Barbara Quint for NewsBreaks,
is contributing editor of Information Today, a former reference
librarian, and a longtime online searcher. Her e-mail address is firstname.lastname@example.org.
|Table of Contents||Information Today Home Page|