If poetry-as Columbia's Mark Van Doren used to tell students-was done in by steam heat, it may be that science will perish from hot air.

This is because of the ruling councils of science, prodded by political demands of goose feed, are being driven to compose plans for the unplannable. And the results are reams of head-spinning policy prose that would be funny but not for the fact that in this business, today's foolishness can become tomorrow's strategy.

Take, for example, testimony recently given to a House subcommittee by W. N. Hubbard Jr., the president of the Upjohn Company, in his capacity as chairman of the Planning and Policy Committee of the National Science Board, which presides over the Science Board, which presides over the government's billion-dollar a year National Science Foundation.

Asked to explain how the foundation figures out the best way to spend its reserach money, Hubbard answered that his committee "has, for over two years now, been trying to arrive at some statement that would allow our prioritization of programmatic efforts in light of less than infinite support, to be made in a systematic way. And so, at the last meeting . . ." he explained, "a good deal of wordsmithing occured, trying to set forth in more operant terms. . ." certain aspects of the foundations legislative charter.

Asked at the same hearing to explain how the foundation divides its budget among basic and applied research among basic and applied research and other activities, NSF Director Richard C. Atkinson departed from him usually lucid prose to state, "Obviously we are not going to come to some fixed formula on a optimum mix. That would be a mistake and any mix we thought of as constant into the future would be mistake. But, I do think we have the issue of worrying about an optimal mix. In considering issues of optimality, we have a look at what we are trying to optimize and the key issue here is that the foundation's role should be optimizing for the long term rather than attempting to optimize on a very short-term basis."

Meanwhile, politically generated demands for medical research to explain and predict itself have caused the National Institutes of Health to set in motion an enormous planning exercise involving thousands of scientists. Their efforts are aimed at providing HEW Secretary Califano with a set of "health research principles" that he proposes to use for writing biomedical research budgets in five-years chunks. Some of the prose from this stupendous undertaking suggest the need for professional cryptanlysis. For example:

"While there was no clear consensus within the Panel concerning the best set of categories to use or even concerning the optimal metaphor to employ (continuum, circle, foundation and superstructure, or feedback loops), there was general agreement that the concept of science base (or basic or fundamental research) should be construed broadly, without reference to any particular type or level of aggregation."

The reason that scientists and their adminstrators are being put through these ho*ops is that research is both expensive and unpredictable, and Califano and other politicians yearn for assurances that it's being well run. Their desires are understandable, especially since the space-age, blank-check era for science spawned a lot of sloppy practices and unfulfiled extravagant promises of big, utilitarian payoffs.

But the effect is that the formerly open line to the Treasury has been succeeded by unrealistic demands for planning and explaining work that can wither under such requirements. One need not have romantic views of the roles of serendipity and spontaneity in research to conclude that a bit of maneuvering room, some relief from the bookkeepers, and a moratorium on grand designs wouldn't cost all that much and might turn out to be both cheap and practical.