close this bookVolume 3: No. 21
View the documentResearch software announcements
View the documentManagement culture
View the documentResearch culture
View the documentInternet news
View the documentJob opportunities
View the documentDiscussion lists
View the documentAI software
View the documentComputists' news

Ernest Henley, past president of the American Physical Society, has warned the House that the US research community is growing too fast. Universities should "consolidate and cooperate," and NSF should be more selective so that funds are not spread too thinly. Increased investment alone is not the answer. [Robert L. Park, WHAT'S NEW, 5/21/93.] (This may be a battle between constituencies. Those that "have" want bigger, longer grants for large projects at established centers and labs. Those that "have not" want easy small grants for exploratory work by new researchers.)

D.R. Forsdyke's "On Giraffes and Peer Review" (FASEB Journal, 4/93) claims that peer review is failing. Giraffes have a nerve bundle running down to the chest, around a major blood vessel, and back up to the larynx -- clearly a mistake due to evolution. Sometimes design by revolution is preferable. Forsdyke sees the North American peer-review system (for biomedicine grants) as evolved for grantsmanship rather than scientific ability. "The less obvious an idea is, the more difficult it is to communicate." Grantsmen review each other's proposals and also review (and approve!) the grant-review process. Problems are blamed on lack of money or lack of alternatives. Meanwhile there has been competition at the expense of collaboration, and an increase in fraud and plagiarism. Forsdyke advocates bicameral review, with a PI's track record judged by peers and the proposed work judged by agency staff. A sliding funding scale would then apply, with track record getting the most weight. Details can be found in his papers in FASB J. 5, 2312-2314 (1991) and Accountability in Research 3, 1-5 (1992). [forsdyke, sci.research, 5/17/93.]

(I am not convinced that creative ideas need be obscure, or that projects without explainable benefit should win public funding. Grantsmanship and scientific merit can be complimentary rather than inimical. Besides, learning to use the system works better in the short term than bitching about the system. Funding by track record perpetuates an "old boy" system that would stifle progress. What is really needed is a statement of why the work needs doing, what evidence -- if any -- supports the approach, what resources are needed, and why we should believe that the PI can do the work. This last element may be based on track record, but track record alone is not sufficient. Agency staff should act as venture capitalists, drawing on the scientific community for guidance. That is how peer review currently works, except that the process has become too mechanical. Bureaucrats are playing the role of Lucy in the cake factory, watching the assembly line accelerate into chaos. At NSF, most proposals are never read by program directors -- there's no time! ARPA makes better use of program managers. Hire good people and give them a chance to be professional.)

National labs have always "employed the best scientists and left them alone," according to William Burnett. With current funding, it is necessary that needs be identified, goals set, and progress measured. Industry must be involved in lab R&D from the beginning, and labs must not be "shelves of technologies waiting to be commercialized," according to William Spencer. The House science committee under George Brown (D-California) is reevaluating the US national lab system. 30 labs employ 56K people for $6.5B/year. Brown's H.R. 1432 would require the Secretary of Energy to prepare a plan to consolidate the nuclear weapons programs, add committees to advise the labs on tech transfer, and streamline the Cooperative Research and Development Agreement (CRADA) process (but require more accountability on CRADAs over $500K). DOE lab mission statements would be made explicit, including a goal of improving science, math and engineering education. Congressional witnesses were not entirely happy about having more advisory committees, and there has been debate over whether competition in ideas is more beneficial than lab consolidation. [Audrey T. Leath (, FYI #67, sci.research, 5/20/93.] (Competition in ideas. You don't hear about that from people who want easy grant money for past productivity. Past success is a good predictor of future effort, but so is a good research proposal.)

"An Insider's Guide to Choosing a Graduate Adviser and Research Projects in Laboratory Sciences" can be FTP'd as jce.send in pub/Advice.For.Grad.Students on Only slightly different from the version in J. of Chemical Education, 1993, 70, pp. 303-306. Comments to Marshall Dermer ( [sci.engr, 5/15/93.]

"Real-World Engineering: A Guide to Achieving Career Success," (IEEE Press, 1991) by L. Kamm includes tips on time management and company politics. "What they didn't teach you in engineering school." [Daniel Romanchik (, sci.engr, 5/23/93.]

Jorn Barger is circulating a memoir of life in Roger Schank's Institute for the Learning Sciences (ILS) at Northwestern University. Jorn is no longer working there, and vents a little steam as he describes his adventures in cognitive psychology. Jorn writes, for instance, of a not-invented-by-Roger syndrome that forces each new project to have a theoretical tie to one of Schank's publications. For catching up on Roger's work, Jorn recommends "Tell Me a Story" or perhaps "The Connoisseur's Guide to the Mind." These and other ILS reports are available, free to academics, from Elizabeth Brasher-Brown ( [,, 2/7/93.]