A rather small minority geologists, meteorologists, biologists, archeologists, et cetera go on expeditions between stretches of office work or teaching. The rest work in laboratories, governmental and private. The private laboratories are those of manufacturing companies, universities, and other institutions like museums and medical centers, and a few well-fixed individuals. A little over a billion dollars is spent yearly on scientific research in the United States, half by the government and half by private agencies.
Among private organizations, colleges and universities tend to specialize more in pure science abstract problems about the nature of man and the universe while the other kinds favor applied science - engineering research on practical problems leading to improvements in ways of making and using things. However, just as there is no sharp distinction between these two kinds of research, so there is no sharp distinction between the organizations performing them. Many colleges and many industrial laboratories solve problems in both fields. Some of the laboratories of manufacturing companies do much basic research while others concentrate on specific problems.
Moreover these organizations do a lot of work for each other. A manufacturing company, for example, may rely upon its own engineers for ordinary design and improvement of its product, but resort to a university or government laboratory for advice on drastic new departures and difficult problems. The university laboratory likewise may at the same time perform engineering research for a manufacturer, research in sociology or economics for the government, and investigation of a problem in pure science for itself in order to advance the knowledge of mankind.
The "laboratory" using the word in its broadest sense will comprise a group of buildings or a single building or a part thereof, with equipment and personnel. This will be divided into parts, called "divisions" or "sections" or "laboratories" or whatever, each specializing in some field or branch of science. Then each of these subdivisions has two more or less distinct areas: the office space where the scientists have their desks separate cubicles for the higher ups; a general deskery for the rest and stenographers, and a "laboratory" proper where the physical work is done.
A "laboratory" in the last sense is seldom so neat and shiny as it appears in movies about noble scientists. Instead it is cluttered with odds and ends of electric cable, tubing, wire, glassware, stands, pencil stubs, mechanics' lunches, old technical magazines, ash trays made of discarded scientific paraphernalia, and so forth. Let some officious person "clean it up" and the scientists scream that they can no longer find anything: "I've been saving that roll of copper-nickel-iron wire for a year to use in Project 8663B, and now just when I have a use for it-"
There is no very serious problem in hiring a young scientist to work in a laboratory. That is one reason there are degrees: so that the young scientists come neatly ticketed. Hence if you have work to be done in biochemistry you hire a biochemist.
But then the trouble begins. If he turns out to be no good you fire him but how do you know when he is no good? Maybe the problem you gave him was inherently insoluble, so it is no discredit to him that he failed to solve it. On the other hand if he fails to solve several problems and then some other man succeeds, the first man doesn't look very good. I have known a third-rate scientist who would repeatedly turn in nice neat reports, carefully calculated and composed and reasoned and every time the report would be rendered useless by some glaring basic error in scientific assumptions.
Then, in most lines of work there is a rough correlation between a man's general work habits neatness, agreeableness, punctuality, et cetera and the merit of his work. But with scientists this does not apply. Some very good scientists are disorderly people who keep irregular hours and snarl at everybody. It takes a good administrator to weigh the worth of the men under him and to see through such superficialities as sloppiness of appearance or disagreeable personality.
Well, how do you get a good administrator?
Ah, there's one of the hardest nuts of all to crack. In most organizations, when you wish to choose a man to head the outfit or a section thereof, you pick the one in that group who has shown the most ability in his assigned work in the past. Now, if your outstanding scientist is also a good administrator you're in luck and there is no problem. But this doesn't happen any too often, because the qualities that make the best scientist are not at all those that make the best administrator. The scientist often cares little for people and doesn't get on well with them; the administrator must be an expert on people and must get along with them. The scientist is usually bored to tears by bookkeeping, finance, time-checks, organization charts, and maintenance problems; the administrator must seize all these with a sure and confident grasp.
Does it fallow that the man who knows the most about accounting, organization, and the like is the best lab administrator?
No, because here you run up against another paradox. A laboratory cannot be closely compared with a production; sales, teaching, or policing organization, all of whom can show results that can be measured in simple units: dollars, arrests, and so on. A production department produces things that can be simply evaluated; a laboratory produces knowledge, and nobody knows what the knowledge will be worth until a year or a decade or a century has elapsed.
But the administrator, playing nursemaid to a gang of idiosyncratic geniuses, must justify his fat salary. Unless watched he is likely to impose upon the laboratory a degree of regimentation, organization, and paper work that actually cuts down the productiveness of the group.
In large organizations there is a natural tendency for paper work to increase with time. Each executive thinks of some daily or weekly or monthly report on hours, progress, equipment, et cetera which, if he could only make everybody turn it in faithfully, would give him a much clearer picture of what is going on in his organization. So he institutes this report, but seldom thinks to abolish any of the reports established by his predecessors. As a result the paper work increases until the scientists are actually spending a fifth or a quarter of their time writing reports.
These contradictions and difficulties are least evident in laboratories that do cut-and-dried engineering tests. Such outfits run in a routine manner, with one engineer bossing a gang of undegreed technicians, in much the same manner as a production department.
But as soon as any originality or thought is allowed you have "research," and the more research you have, the "purer" the science, the greater are the administrative difficulties. The troubles also increase with the size of the organization, reaching a maximum in government labs and those of great private companies like du Pont and Bell Telephone. [footnote: I said great companies like them, not necessarily those two companies themselves. ] Though oceans of ink have been spilled on the respective merits and faults of capitalism and socialism, from the point of view of the ordinary working scientist or engineer, the difference between working for the government and for a large corporation could be put in your eye without discomfort.
Experience shows that for profound theoretical research in pure science, men work best when working alone or in a voluntary association of two or three brains, and when given the greatest possible freedom. On the other hand for solving practical engineering problems, or applying known principles, the best results are had by well-organized teams of specialists seeking well-defined objectives. University laboratories, which go in more for pure science, therefore tend towards the former type of organization if you can call it that - while industrial laboratories, devoting themselves more to practical engineering, incline towards the latter.