Conclusion

In all, it seems that violence by lords and armies (off the battlefield) can be linked tentatively to some typical settings. First, it was shared space that was often the theatre of violent clashes.

Second, on a number of occasions, the taxpayers came into direct contact with the ultimate recipients of the tax payments, the military. That means that on such occasions payments for the army or any given group of warriors had to be collected by these warriors themselves. This could take the form of tax cheques (barat); the ‘cashing' of such cheques often involved violence and could be an altogether chaotic process. The provisioning of the army on the march is another occasion; armies lived off the land regularly and did not even try to avoid that. Extra levies meant for the army on the march, and in particular the na^cl-baha, but also other forms of ‘awaridat, appear in the Seljuq period; they became very common indeed in the Mongol and post­Mongol periods. Both the tax cheque and the extra levy for the provisioning of the army therefore have two sides: they can be seen as a regular instrument of state finance, and they can appear as a form of looting and marauding. Even if they work without open violence, they cannot function without threat; immediate coercion is behind both these forms.

Third, beginning in the second quarter of the eleventh century, Turkmens and later other Turkish nomads migrated into Iran and other parts of the Middle East. Numbers are hard to come by and are open to discussion, but it is clear that this was a major migration. Conflicts arose from their need for pasture and their way of paying the army through shares in plunder. Some of the most violent clashes occurred where the Turkmens competed with other pastoralists and transhumant stockbreeders over pasture. Central Asian nomads continued to migrate into Iran during the Mongol period. The particular forms of violence linked to nomad migrations into Iran are one of the most salient features in the history of Iran in the period under study.

Fourth, local lords could turn into violent lords particularly in situations when sultanic power had broken down. This happened repeatedly during the period under study. Successions tended to be contested, and thus sultanic power could vanish even if the dynasty itself was well established. Dynastic change was also not infrequent; most dynasties did not outlive a century. However, there are also instances of local lords acting violently in regions and times where sultanic power was strong; in such cases, the sultan appar­ently had an interest in keeping them as allies (or vassals).

Beyond these settings, a constant level of violence was implied in processes of recruitment and of making people work for the army: forced service in the army and forced labour were features of everyday life.