Again, both of these scholars emphasized that frontier society allowed more room for heterodoxy, heterogeneity, and mobility.
As to their dissimilarities, K
pr
l
and Wittek held divergent opinions on the issue of the “tribal factor” in early Ottoman state building. Neither of them understood tribal formations in the sense utilized by modern anthropologists, however. For both historians (and all the Ottomanists of the time) tribalism entailed consanguinity; that is, a tribe would in essence have to be composed of blood relations whose ancestry ought to be traceable to a common origin, at least in principle. Given that, Köprülü was ready to accept that the Ottomans hailed from a tribe belonging to the Kayi branch of the O
uz Turks, as most of the sources maintained and as eventually became official Ottoman dogma. However, Wittek pointed out, the earliest reports about the ancestry of Osman and his tribe are from the fifteenth century, and more significantly, there are a good many divergences in the genealogies different sources provide for the Ottomans. On the basis of these discrepancies, Wittek concluded that the early Ottomans cannot have been tribally associated; otherwise, they would have a consistent genealogy to show for it. In the same vein, Mehmed II would have been unable to toy around with the idea of propagating a Comnenian lineage for his family. Even after Wittels objections, Köprülü insisted on the validity of the Ottomans' Kay1 identity, while at the same time maintaining that this was a secondary issue since, to him, the Kayi origins did not contribute anything specific to the rise of the state.
There was also a major difference of approach between the two scholars. K
pr
l
looked on the frontier society as a broad canvas composed of a variety of social forces (tribesfolk, warriors, dervishes, a
s, emigré scholar-bureaucrats), all of whom made their own significant contribution to the state-building potential of the Turco-Muslim principalities. All this eventually came under the domination of some descendants of the Kayi tribe because the latter happened to be located in a region that circumstances favored. Wittek, on the other hand, focused his attention on one specific element within the
uc (term for
frontier in medieval sources, pi.
uct) society, the gaza milieu and its ethos, as being instrumental in the emergence of the principalities and ultimately of the Ottoman state, which overran the others. To him, the political history of the frontiers was made by bands of gazis, warriors of the faith, who spread across the frontier areas as Seljuk power diminished and formed aspiring emirates, among which the band led by Osman Gazi carried the day because of its fortunate position. The earliest Ottoman sources, an inscription from 1337 and AhmedFs chronicle, completed ca. 1410, both full of references to the House of Osman as gazis, confirmed in Wittek's opinion the significance of the gaza ethos for the early Ottoman thrust.
These gazi bands may have drawn members from some tribes but were not composed of tribal groups as such; rather, they consisted of warrior-adventurers from various backgrounds. In relation to the Mente
e emirate, for instance, he had argued that the “gazi pirates” who founded this stateling were “originally a mixture of Turks and indigenous elements from the neighborhood of Byzantine territory” who were soon joined by “a large number of Byzantine mariners…owing to their unemployment.”
21 To borrow more recent terminology, gazi bands were “inclusive” entities for Wittek, and tribes were not. Since he held that tribalism required consanguinity (which, he argued, later Ottoman genealogies were unable to establish anyway), and since the warrior bands whom he deemed responsible for the creation of the principalities were anything but consanguineous, he rejected the notion that a tribe could have been instrumental in the foundation of the Ottoman state. The cohesiveness of the political-military cadres of the emirates came from shared goals and faith, not blood.
The differences between K
pr
pr
pr
l
's account of the Ottoman foundations, where he insisted on the presence of a lineage-based tribe as well as an ethnic stock and spoke against emphasizing the conversions, was very easy to read as nationalistic propaganda. And indeed his book is not free of blatant excesses.
On the other hand, K
pr
pr
pr
l
explicitly stated that he intended to view the rise of the Ottoman state on the basis of the morphology of Anatolian Turkish society “and the evolution of its religious, legal, economic, and artistic institutions more than its political and military events.”
24 Even though his aim was partially to depict early Ottoman history as a continuation of late-Seljuk
