Выбрать главу

The good administrator is the one who can tell what is the optimum degree of organization and control for his particular group doing his particular kind of work. He must be able to manage his geniuses, to spur them on, and to protect them from uninformed outside interference without pampering them to the point where they take advantage of his liberality as some scientists, like other people, will do if encouraged.

He must watch out, for instance, for the scientist who has been working for a long time on a problem and is hesitant to apply the crucial test for fear it will prove that he is on the wrong track, and who therefore goes on indefinitely refining his apparatus and devising subsidiary tests and generally fiddling around. The administrator must prowl around the laboratory enough to keep track of what goes on without driving his charges wild by constantly breathing dawn their necks. He must be able to ask for clearly defined results without imposing upon the scientist his ideas of how the results should be obtained and at the same time be ready to offer sound advice if asked, or to take corrective action if the scientist obviously has got off the track and is getting nowhere.

One authority recommends that a good laboratory administrator should be somebody like a patent attorney, who has had plenty of contact with science without himself being primarily a scientist. Another suggests: "All administrators should be women, of the aggressive but motherly type. I think they would possibly do a very good job, being willing to provide for their mad little creatures in return for a kind word now and then, and no ego buildup."

As a practical matter the administrator, even if not primarily a scientist, should have a pretty good grounding in science, so that he shall be able to judge his juniors' results and make intelligent suggestions when asked. Also, many technical people get huffy when asked to take orders from nontechnical people, whom they regard as virtual illiterates.

Well, then, suppose you pick your administrator from the ranks of nonscientists who still know enough science for the task. What do you do to reward the scientist who has worked long and well for you and who in a more conventional organization might reasonably expect to be promoted to command of the group? For while most scientists are not extravagant hedonists with violent power-complexes, they do - like other people like to live reasonably well, and their egos demand at least some satisfaction in the form of promotion and pay-raises.

Sometimes, as I said, a good scientist is also a good administrator, in which case there is no problem. But sometimes you have to cook up a scheme for raising a scientist's pay and rank without actually putting him in charge of a department. You call him a "consultant" or a "senior engineer" or something, and reward him by giving him more freedom and less paper work instead of the reverse, which is the normal lot of the executive.

As an example of the absurdities inherent in this paradox, there was a world-famous physicist who worked on the Manhattan Project at Los Alamos during World War II. To give him a rank commensurate with his standing in the field the United States Government had to hire him at one of their top rates, which meant that when the organization chart was drawn up this man was Number Three in command at Los Alamos. Ordinarily this meant nothing; Numbers One and Two ran the place and let Number Three sit in his chair and cerebrate. But the time came when Numbers One and Two were both away at once and our hero found himself acting director. A maintenance man came in to see the director and gave the physicist a long report on some perilous fire hazard that he had found, which should be attended to at once. When the maintenance man had finished, the physicist looked at him vaguely and said:

"But that has nothing to do with physics!"

And thereupon he rolled up his eyes and withdrew his soul once again into the Nirvana of the higher mathematical physics.

The scientific administrator's lot is a pretty exacting one. He has to go as far as he can to give his scientists optimum working conditions, but he can seldom go so far as they would like. They would like laboratories stocked with unlimited equipment which they could simply pick up in the stock room and walk off with, without even signing a receipt, and a minimum of supervision and reports. Then the administrator would not know what was going on, and when Congress or the Front Office or the Faculty Committee came around to ask what his geniuses were doing to justify their expense he would have a tough time answering which would not be good for next year's appropriation.

Or he might find that one of his bright boys had gone off into some line of work that happens to fascinate the scientist but that has nothing to do with the overall objectives of the organization.

On this last point, a tot of trouble would be saved if the administrator would make it quite clear to the scientist when he hires him, just what is expected of him. If the man is employed to do dull routine tests on hydraulic valves he should be made to understand that fact clearly, and if he doesn't wish to test hydraulic valves he does not have to take the job. But if the scope of the job is not made clear the scientist may get a bright idea for an improvement in rotameters and neglect the valves for his wonderful new discovery. And then trouble pops and the administrator is caught in the middle between the outrage of the scientist who is insulted by the lack of appreciation of his genius and the equal outrage of the man who is paying to have valves tested and expects just that.

Another thing the administrator has to watch out for is overorganization, a disease that flourishes in large organizations in general and governmental departments in particular. We all know something about the general principles of organization; authority must be congruent with responsibility, fields of authority must be sharply demarcated, objectives must be clearly stated, and so on. And we know something about line versus staff organization and organization charts.

Now, there are people who love those little charts with their lines and boxes as a pig loves mud. If given a free hand they will think up so marry interlocking relationships and lines of authority and committees, and ordain so many committee meetings and consultations and memoranda in octuplicate, that the organization will be paralyzed by sheer complexity. No organization will work very well if it is so complicated that even those running it can't understand it.

To give an example, during the late war I once had charge of the War Production Committee of the Naval Air Material Center in Philadelphia. This was the governmental equivalent of a labor-management committee in private industry, and was supposed to process suggestions for increasing production. Somebody in Washington set up the most elaborate scheme you ever saw, according to which several hundred people in the shops and laboratories were to be organized into five levels of committees. The shops elected the first set of committeemen, who in turn elected the second from among their number, and so on.

As this was too complicated to be workable we simplified it to a mere four levels and went ahead. We had three hundred sixty committeemen working about one hundred fifteen hours a week on WPC work, plus five full-time employees including Lt. Comdr. de Camp - and three part-time officials spending another two hundred fifty hours a week on this activity. And most of the "work" was sheer waste motion. The useful part of it could have been done by a mere handful of people spending a fraction of the time, while the rest actually produced instead of talking about production. [footnote: I am happy to say that some months before the end of the war, the admiral and I agreed to abolish the whole monstrosity, and I was allowed to go back to gadgeteering. ]

Over-organization is especially pernicious in dealing with scientists, who mostly like to work by themselves and who to some extent became scientists so that they could do so.

And speaking of line and staff organization, there are two ways of handling the minimum routine work and paper-shuffling that must be done. One is to divide it or rotate it among the scientists, so that each gets his fair share of drudgery. The other, which is better if the organization is big enough to afford it, is to setup service departments to handle such matters. A scientist is not normally expected to clean the floor under his desk; you hire a janitor for that. By the same token you can take much routine letter-writing; material ordering, and the like off his shoulders by providing expert help.