partments of all the armies in the world. But the experts will remain human beings. They will enjoy power, and their temptation will be to appoint themselves censors, and so absorb the real function of decision. Unless their function is correctly defined they will tend to pass on the facts they think appropriate, and to pass down the decisions they approve. They will tend, in short, to become a bureaucracy.
The only institutional safeguard is to separate as absolutely as it is possible to do so the staff which executes from the staff which investigates. The two should be parallel but quite distinct bodies of men, recruited differently, paid if possible from separate funds, responsible to different heads, intrinsically uninterested in each other's personal success. In industry, the auditors, accountants, and inspectors should be independent of the manager, the superintendents, foremen, and in time, I believe, we shall come to see that in order to bring industry under social control the machinery of record will have to be independent of the boards of directors and the shareholders.
3
But in building the intelligence sections of industry and politics, we do not start on cleared ground. And, apart from insisting on this basic separation of function, it would be cumbersome to insist too precisely on the form which in any particular instance the principle shall take. There are men who believe in intelligence work, and will adopt it;
