The college is proud to announce that the1999 recipient of the Collen Award is Dr. Joshua Lederberg, biologist, geneticist, computerscientist, and schola. Dr. Lederberg's lifelong contributions to biomedicine and science,as well as to medical informatics make him highly deserving of the recognition embodiedin the Collen Award. Joshua Lederberg was born in Montclair, NewJersey, on May 23, 1925, the son of a rabbi. His interest in a scientific career was mentionedin an essay that he wrote at the age of seven. His family moved to New York City when Lederbergwas a child, and he was able to attend Stuyvesant High School, which concentrated in the sciences.Upon graduating from high school at 16, Lederberg took advantage of a local scholarship to attendColumbia University. And subsequently began medical school there. After his experimentswith Edward L.Tatum, that demonstrated sexual recombination in bacteria, Lederberg decidedto leave medical school to pursue a Ph.D., which he received from Yale in 1948 at theage of 23. He then joined the genetics department at the University of Wisconsin. In 1958 whenhe was but 33 years old, Joshua Lederberg received a telegram of the sort that onlya handful of scientists every receive. He was awarded the Nobel Prize for his discoveriesconcerning genetic recombination and the organization of the genetic material of bacteria. He sharedthe prize with George W. Beadle and Edward L. Tatum. The work that lead to Lederberg'sNobel began in 1945 when he was 20, and still a medical student. Lederberg's work, whichformed the basis for his Ph.D. dissertation demonstrated that bacteria can, in fact, reproducethrough sexual recombination, and opened up the genetics of micro-organisms to the traditionalmethods of the field. Lederberg is known not only for his demonstrationof bacterial conjugation, but also for his discovery of transduction and for coiningthe term plasmid. The Nobel Prize gave him a reputation among the general public thatallowed him to pursue his interests at the intersection of science, policy and society.Between 1966 and 1971, he wrote a series of over 200 articles for the Washington Post,which converted into lay terms many of the pressing scientific issues of the day. Typicalarticles included themes such as "Russian computers are having delusions or are they?"and the premonitory "Congress should examine biological warfare tests", a topic which continuesto draw his interest and involvement at the national level to this day.Josh has had a long interest in formalizing the intellectual processes of scientific investigationand discovery. But on coming to Stanford in 1959 to found the Stanford Department of Geneticshis study of informatics took on new dimensions. He took classes in programming and began aseries of systematic investigations of using computers in biomedicine. By the mid-1960s,Josh was working on ways to systematize the enumeration of chemical structures. This earlywork would find expression later on in the DENDRAL project.[Donald A. B. Lindberg:] It's hard to remember when I really first met Josh, but I can remembervividly the first serious conversation I ever had with him. It was 1964. It was my firstmeeting as a Markel Scholar. The group was in the Ahwanee Lodge at Yosemite, and thespeakers had been Jacob Bronowski, Lee Dubridge, and Josh Lederberg. And I remember seeingJosh in the sun at a coffee break, sitting alone at a table, sort of lost in thought,and I decided that he probably wouldn't mind being interrupted, and so I asked him whatwas he thinking about -- seemed like not too impertinent a question -- although it probablywas -- and he said he was thinking about this: If he had a chemical molecule in his mind,and he could see it clearly, what would be the next one that logically should be considered?That struck me as kind of an amazing thing to be thinking about.The post-Sputnik start of the US space program in earnest -- one of Lederberg's new areasof attention was the exploration of outer space in search of life forms. He coined theterm "exobiology" to describe this new field dealing with the search for the origins oflife. He invented new ways to represent information about molecular structures, such as proteins,in the computer, so they could be analyzed and compared. The 1964 printout shown heremay be the first time that a protein sequence was stored in a computer for analysis.[Edward Feigenbaum:] In 1964 when I first met Josh, he was writing his first large-scalecomputer program in the dialect of ALGOL. The program was called DENDRAL for dendridicalgorithm but the reference to dendrites was metaphoric. Josh was creating a complete systemfor the enumeration of acyclic chemical structures. When I arrived at Stanford in January of 1965,I discussed with Josh my goal of creating models of the thinking processes of scientists,especially the processes of empirical induction with which hypotheses and theories were inferredfrom data. Josh, suggested a task of inferring organic chemical structures from mass spectraldata and enthusiastically entered into a collaboration. My sketch of Lederberg and DENDRAL is necessarilybrief and leaves out so much of his tangible contributions to our work in artificial intelligenceand its applications. And it leaves out almost all of his intangible contributions, the textureif you will of the day to day interactions with Josh. These interactions were full ofwisdom and guidance on how to run substantial interdisciplinary projects of the kind wewere doing in expert systems, insights on the nature of knowledge and cognitive processesin science, and creative ideas for getting us over today's hurdle and tomorrow's hill.In truth, a flood of creative ideas. If we had had the energy and resources we couldhave probably done ten other projects with these ideas. Working with Josh on our commonproblems in modeling expert thought in science and biomedicine gave me the privileged opportunityto witness a great intellect at work, applying superb analytic thinking skills with the enthusiasmof a young hacker for how much fun it was to program a computer to do these things.And it was not only fun, but it was important in the history of computer science.The DENDRAL collaboration lead Josh to devote a great part of his energies to computer science,including the underlying languages and tools necessary to express his ideas about molecularbiology and genetics. [Lindberg:] Walking across the lovely campusat Stanford with him following a morning in which I don't remember the real circumstancesbut a whole lot of very nice research was reported at Stanford -- I must confess whileI liked it all, I was sick and tired of hearing about LISP and list-processing languages.So as we hiked off to the commissary or wherever the faculty club or wherever we ate, I saidto Josh, what would happen if some malignant deity all of a sudden destroyed, eliminatedLISP there isn't any more LISP. What would you do? He said, I guess, Don, I would takeFortran and write LISP. Lederberg insisted on participating first-handin the development, testing, and use of the computing systems that fascinated him, nomatter how cumbersome the early human computer interfaces were.[Carl Djerassi:] So we got together with Ed Feigenbaum, who had been working on artificialintelligence in general, to see what we could apply to chemistry, in particular the structuralelucidation of organic compounds, which was precisely my area. And then they came to meand said, look, can we just simply try and simulate in a computer how Carl Djerassi thinks?And that was really the beginning of a very fruitful collaboration for well over a dozenyears in which a lot of other people were involved Tom Rindfleisch, Bruce Buchanan,and we published a awful lot of papers in which we did try to simulate a chemist's thinkingin computer-intelligible language. What was most interesting, it stimulated us to do alot of chemical research to fill gaps, intuitive gaps, in our own thinking, which the computerpicked up because the computer program wasn't that intuitive yet. And that was really quiteinteresting, because chemists don't really wonder how they are thinking, they just doit. That really taught us an element of intellectual rigor. And it was a wonderful collaborationof people in medicine and computer science and chemistry.As the DENDRAL programs evolved and made progress toward modeling the thought processes of chemistsworking to unravel molecular structures, Josh's interests returned to the broader possibilitiesof intelligent computer applications. The SUMEX-AIM resource that he created in theearly 1970s produced many systems that were the seeds of artificial intelligence researchfor years to come. [Thomas Rindfleisch:] About a year after Icame to Stanford, a small group of us, led by Josh Lederberg, synthesized the ideas fromACME and DENDRAL in the form of a proposal to the NIH called SUMEX-AIM. SUMEX-AIM containedtwo fairly outlandish ideas at the time. The first was to develop a national communityof AI in medicine application groups and the second was to interconnect them using networkingtechnology that was just then under development by the ARPA community in the form of the ARPANET.SUMEX-AIM was the first non-DOD research resource on the ARPANET. In fact, we were node number56 among a total of 63 on the entire network in that current implementation. SUMEX lastedfor about 19 years and was the nursery for over 20 AI-in-medicine projects that werethe best known in that era. Josh wrote a very important essay in 1977 that described theseideas. It was entitled Digital Communications and the Conduct of Science: The New Literacy.I'd like to quote a brief passage from this essay. In it he discussed how the convergenceof economical digital communications with computer-aided tools would facilitate theinterconnection of users separated both in time and space, and how this medium will increasethe thoughtfulness of communication, return literacy in the efficient and precise useof language, and enhance scientific discourse in many ways. As I scan the e-mail in my inboxtoday, I'm not sure how much progress we have made in the literate use of language, butit is absolutely clear with the current form of the Internet that digital communicationshave profoundly changed the way science is conducted.Josh left Stanford in 1978 to become president of The Rockefeller University in New York.Despite the heavy burdens of administering one of the nation's leading research universities,he continued to explore the synthesis of computing and communications tools with scientific researchon many fronts. [Edward Shortliffe:] I would say that Joshcreated an environment in which I was taught to understand computer science as science,and not simply the construction of artifacts. The scientific approach to computing, wherebyone seeks to be sure that the results can generalize, where the experimentation is itselfsubjected to proof, and where there is an emphasis on methodologic development thesewere key elements in the environment and the way in which I learned computer science becauseof the scientific influence of people like Josh at that time. I saw a brilliant mindin action, and I saw in him an interest in basically all things intellectual and an abilityto innovate in very diverse fields. There was much to emulate in Josh Lederberg andhis mentoring. To this day, Josh Lederberg continues to bringa sensitivity and appreciation of computing and communications to his advisory work andto his influence at the national and international level. He became president emeritus of Rockefellerin 1990 and returned to his roots as a citizen scientist. He has been particularly activein applying scientific understanding to the direction of research, to public health, andto policy. As was mentioned, he has worked hard at educating the public and governmentabout the risks of biological terrorism and warfare and advising on detection and countermeasures. Josh has continued his active research laboratoryat Rockefeller, working on the genetics, biochemistry, and evolution of DNA; on the computer modelingof scientific reasoning; and on exobiology and the search for the origins of life. Hehas been honored by the National Library of Medicine in their Profiles in Science program.The NLM is collecting and organizing Josh's voluminous papers as a contribution to thehistory of science. Most befitting Josh's career in medical informatics, this collectionis available on the NLM's Web site. Josh's work in informatics is a lasting legacy, andhe continues to pose challenges for the field. [Lederberg:] One principle I'd like to havemy name attached to, Lederberg's Principle, is that machines will become really smartonly when (1) they can directly read the literature and (2) spend some time living in the realworld, where the survival of the fittest is what will determine who's out there. As longas we have to spoon feed them, datum by datum, they're going to be moving in a very cumbersomeand costly way indeed. It is with pride, admiration, and affectionthat we salute Joshua Lederberg as he receives the 1999 Morris F. Collen Award.