The change which constitutes a problem is an altered relationship between two dependent variables.[21] Thus the automobile is a problem in the city not because there are so many automobiles but because there are too many for the width of the streets, too many for the number of competent drivers, because the too narrow streets are filled with too many cars driven too recklessly for the present ability of the police to control them. Because the automobile is manufactured faster than old city streets can be widened, because some persons acquire cars faster than they acquire prudence and good manners, because automobiles collect in cities faster than policemen can be recruited, trained or paid for by slow-yielding taxpayers, there is an automobile problem made evident by crowding, obnoxious fumes and collisions. But though these evils seem to arise from the automobile, the fault lies not in the automobile but in the relation between the automobile and the city. This may sound like splitting hairs, but unless we insist upon it we never define a problem accurately nor lay it open successfully to solution. The problem of national defense, for example, can never be stated by a general staff which draws upon its inner consciousness for an estimate of the necessary force. The necessary force can be estimated only in relation to the probable enemy, and the military problem whether of peace or of war lies always in the ratio of forces. Military force is a purely relative conception. The British Navy is helpless as a child against the unarmed mountaineers of Tibet. The French Army has no force as against fishing smacks in the Pacific Ocean. Force has to be measured against its objective: the tiger and the shark are incomparable one with the other. Now a settled and accepted ratio of forces that might collide is a state of military peace. A competitive and, therefore, constantly unbalanced ratio is a prelude to war. The Canadian border presents no military problem, not because Canada’s forces and our own are equal but because, happily, we do not compare them. They are independent variables, having no relation one with the other, and a change in the one does not affect the other. In capital ships we are confronted now with no naval problem in the Atlantic or in the Pacific, because with Britain and Japan, the only two comparable powers, we are agreed on a ratio by treaty.[22] But for all types of ships not subject to the ratio there is a naval problem in both oceans, and if the Washington Treaty should lapse the problem which it settled would recur. It would recur because the synchronized progress of the three navies would be replaced by a relatively uneven progress of each as compared with the others. 4