Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!pasteur!ucbvax!hplabs!hplabsz!taylor From: taylor@hplabs.HP.COM (Dave Taylor) Newsgroups: comp.society Subject: Re: University Education and Industry Needs Message-ID: <1431@hplabsz.HPL.HP.COM> Date: 19 Jan 88 08:36:26 GMT Sender: taylor@hplabsz.HPL.HP.COM Organization: Hewlett-Packard Laboratories, Software Technology Lab Lines: 114 Approved: taylor@hplabs Kurt Guntheroth, a while back, asked an interesting question about the relationship of Universities and Industry when it came to funding projects. Dennis Hamilton and others have continued the discussion with various thoughts on the subject, to which I'd like to add my own. It seems to me that we've assumed a bit of an arbitrary distinction between universities and industry; where almost any level of interaction is questionable. Well, I'd like to propose instead that we consider them both areas in the same `research space'; Universities ===================+ --> Applied Research | + - - - - - - - - - - - - -|- - - - - - - - - - + | | +===========|=======================+ +=================|===========+ | | + - - - - - - -|- - - - - - - - - - - - - - - - + | / +================= Industry ========+ Areas of Research As we move to the right in this illustration we find more and more applied research, intended to result in products that the company can use to increase their bottom line; to make into products they can introduce to the market. Universities, on the other hand, are more inclined to do `pure' research (an interesting phase, actually; it implies that the research done by industries is somehow impure) which are intended less to become products to sell and more as vehicles for learning about basic, fundamental problems in the specific area of research. There is certainly some overlap as my illustration shows; I can think of many examples of pure research done in industry (for example, the continued research at Bell Labs and NASA) and examples of research which (somewhat ex post facto, I admit) seems to have been intended more as a product than research result (examples that pop up are UCSD Pascal and the University of British Columbia X.400 mail system - now sold by companies created, essentially, to sell the products). With this basic view espoused, let's now go and look at the question Kurt originally asked; "Why should industry give [money to universities] and then resign all control?" It is indeed an interesting question. I believe the answer is that the platform whereby most industry/applied research is done is firmly rooted in the fruit of pure research done five or more years in the past. It's a very symbiotic relationship between industry and univer- sities since neither can exist without the other (Where would the graduates go? Where would they get trained and qualified employees?). Let's leave that thought for a second... Based on my own experience, it seems that a couple of the basic tenets of research are that new and seemingly wild ideas mustn't be summarily discarded and that some reasonable percentage of the research done should ultimately fail. I believe that we're starting to hit on the problem that is typically encountered in industry-funded university research finally; that very few companies in industry are willing to give money to a `failed' project, so they instead impose sufficient constraints upon the project to ensure that it won't fail (but that, of course, it won't succeed either as pure research). In essence, I believe that a lot of industry funding is tied closely to the lack of willingness to `take chances' in research, to risk failure, or indeed to fail. "New Scientist" had a fascinating editorial a few months ago about the aura of failure that springs up about research that doesn't succeed in its goals, and the effect on the research staff. It is sufficiently alarming that the percentage of false results reported in technical and scientific journals has been increasing steadily for the past twenty years. One musn't fail. But for pure research, how can you *not* fail? It's like saying that you'll only bet on the pools if you're guaranteed of winning. And research is, fundamentally, a gamble. The researchers are betting that their theories will turn out to be true... As Universities have become more interested in expensive areas of research (like particle physics, IC wafer fabrication technologies, or parallel computation) they've found that it's very useful to have a big company or two along to help with the neccessary equipment and overhead. But nothing is really free, and they seem to be willing to pay the price of increased industry input in the traditional university role of research (the "ivory tower"). The case cited by Dennis Hamilton, where Eastman Kodak threatened to withdraw funding of a project at the University of Rochester if a request of theirs concerning admissions wasn't resolved as they choose, is an example of the frightening lengths that this can go to. I believe that Kodak should be reprimanded for trying to exert influence through blackmail, and that the board of regents of the University of Rochester should be reprimanded too; if they need to rely on the funding of companies that force them to adopt racist policies, then they're in much deeper than they think. I'd like to see a situation where industries help fund research in universities purely for humanitarian and philanthropic reasons, without any particular desire to influence or control either the research or the policies of the department/university itself. In fact, perhaps it should be legislated that a percentage of all corporate income should be shifted into University accounts earmarked for research purposes. Note that I've not even mentioned the other plague of the current spate of industry participation in university research; company confidential classifications extending to research done at public and private universities. In any case, this is indeed an interesting issue and I look forward to further discussion on it herein. --- Dave Taylor