It must not be simply assumed, as is often done by the Chinese and some of the European historians, that the Turkish and Mongolian tribes were so savage or so pugnacious that they continually waged war just for the love of it. The problem is much deeper, and to fail to recognize this is to fail to understand Chinese history down to the Middle Ages. The conquering Chou established their garrisons everywhere, and these garrisons were surrounded by the quarters of artisans and by the villages of peasants, a process that ate into the pasturage of the Turkish and Mongolian nomads. These nomads, as already mentioned, pursued agriculture themselves on a small scale, but it occurred to them that they could get farm produce much more easily by barter or by raiding. Accordingly they gradually gave up cultivation and became pure nomads, procuring the needed farm produce from their neighbours. This abandonment of agriculture brought them into a precarious situation: if for any reason the Chinese stopped supplying or demanded excessive barter payment, the nomads had to go hungry. They were then virtually driven to get what they needed by raiding. Thus there developed a mutual reaction that lasted for centuries. Some of the nomadic tribes living between garrisons withdrew, to escape from the growing pressure, mainly into the province of Shansi, where the influence of the Chou was weak and they were not numerous; some of the nomad chiefs lost their lives in battle, and some learned from the Chou lords and turned themselves into petty rulers. A number of “marginal” states began to develop; some of them even built their own cities. This process of transformation of agro-nomadic tribes into “warrior-nomadic” tribes continued over many centuries and came to an end in the third or second century B.C.