19 Sabine Schmidt: Conflict and violence at the local level ing beads, guns and cloth. Cattle were used to provision the trading caravans. Wars between the Chagga chiefdoms were fought in order to procure the *items" needed for a continuing engagement in the caravan trade: captives, ivory, and cattle were taken in these fights (Moore 1986: 36). The discussion about the role of trade and its link with warfare leads me to conclude that indeed integration within a world-system can have different effects on warfare patterns and is thus one explanation for the non-corre- lation between the world-system position measure and the warfare variables.? (2) The second explanation for the obtained results I want to discuss points to- wards a conceptual weakness of world-system theory. It has been addressed among others by Smith (1985), who criticizes anthropologists for accepting the world-system paradigm too readily: “The result is an interpretation of local history in which the masses play only a pas- sive role. Anthropologists may not see the masses as passive, but they do see them as responding to a world made largely by others. They assume, in other words, that local communities adapt to external pressures (or die out), rarely considering the possibility that local-level processes actively shape the larger picture.” (84) Again I will use an example to show how local circumstances have interacted with ex- ternal impacts to shape patterns of conflict management. After the establishment of German colonial rule in Fast Africa in 1885 subsequently taxation was introduced (1897) and the cultivation of cash-crops like coffee and cotton was initiated. The reac- tions of the local populations in response to these measures were not uniform and de- pended on the degree of interference with the demands of the local peasant agriculture (Hyden 1980: 43). In the Chagga region around Mt. Kilimanjaro, coffee cultivation was introduced. The Chagga grew coffee alongside bananas and its cultivation did not require initial changes in the existing system of cultivation (Moore 1986: 117). Thus it did not interfere with the traditional subsistence agriculture and, as would be ex- pected, no resistance against the introduction of cash crop cultivation resulted. In the area of the Rufiji Valley a cotton cultivation scheme was initiated, here pea- sants were forced to work the communal plots. The rewards for the labor were low and in combination with the amount of time spent in these activities had a deleterious effect on subsistence agriculture. Hyden (1980: 43) attributes the outbreak of a rebellion in the Rufiji Valley in 1905 to these circumstances. The example serves to demonstrate how similar external impacts, the introduction of cash crop production, had very different effects on the local populations. The diffe- rential outcome, rebellion in one instance, no resistance in the case of the Chagga, can only be explained in terms of the interaction of local circumstances and external im- pact, but not by referring to the external impact alone. E - This claim is further supported by the different correlations calculated for two of the five variables measur- Ing world-system integration, trade and markets, and cropping systems, but again see footnote 6.