suspect on the basis that it might be trying to find a particular answer. Certainly we can find many examples of productive institutional interdisciplinary structures, and we cannot expect to find people eager to point out cases in which having the institutional structure has not been successful. Unfortunately, those less successful examples, the cases people are reluctant to discuss, are the cases from which we would learn the most.
If we examine how institutional structures affect the scientific enterprise, we must be prepared to go about it objectively, i.e., we must look for both the positive and negative consequences for science and education of institutionalized supporting structures with specific goals. We must look at not only those structures targeted at interdisciplinary work but also those designed for traditional disciplines. Are we better off with designated, targeted institutions, or with an informal, fluid enterprise? Do narrowly defined structures such as “centers,” whether funded or unfunded, contribute more or less to science and science education than larger, more broadly oriented, fluid organizations?
Two more of the most apparent (and related) changes in the U.S. chemistry graduate programs have been the increase in the length of time from entry to doctorate and the increase in the length of doctoral theses. Typical graduate careers, B.A. to Ph.D., lasted about 4 to 4 1/2 years in the 1950s and 1960s; now 5 to 6 years is more typical. The doctoral theses on my shelves were of the order of 80 to 130 pages into the 1970s but are typically about 200 pages or more now.
These are symptoms of change. What changes have been responsible for them? An obvious and rather trivial contributor is the use of the computer as a word processor, so that the cost of doing a long thesis is negligibly greater than the cost of doing a brief thesis. This contrasts with the situation of having to pay a typist, on a per-page basis, to put a thesis into acceptable form. Furthermore, it is vastly easier to write and revise a text with a word processor, so that we might expect the writing quality of the theses now to be higher than it was in the 1950s and 1960s. I am not convinced that has happened, but I hope someone does a doctoral dissertation on that subject so I can find out. One thing we can expect to improve with computerization of the scientific literature, something just now at its threshold, is the quality of the scholarship, specifically the accuracy of attributions and historical background. As the archives of journals are converted into searchable electronic forms, we can justifiably expect students to find and cite early works directly relevant to their research. The barriers to doing such searches have been significant, so much so that we have all heard outraged criticisms of how the literature has been overlooked, especially literature in languages other than English.
Another reason often given for the growing duration of the graduate career is the need to know much more than was needed 30 or 40 years ago. Certainly the body of scientific knowledge is vastly larger than it was in 1950 or 1960. We sometimes rationalize 6 years of graduate work on the basis that a student finishing a Ph.D. must know about the same fraction of the literature that was expected of a fresh Ph.D. in 1960. Hence, we argue that students must spend more time in school to absorb all that material. If this is so, should it be adequate justification for the increase in time to the Ph.D.? This takes us past the realm of identifying changes into the subject of the next section dealing with what graduate education should achieve.
The collaborations between universities and private firms have been growing rapidly and often involve graduate students. This has been a significant change in graduate education in chemistry and in the relationships among institutions and individuals. The history of this involvement is tied closely to the evolution of the research industries in America and to the perceptions in Congress of the role of government in assisting the evolution of science and technology. As a result of the apparently transient fad of industries divesting themselves of much of their longer-term research programs and seeking outside sources to do that research, the universities appeared to offer natural, low-cost alternatives. The most obvious symptom of the mood of Congress was the passage of the Bayh-Dole Act, which allows