One of the most significant factors in the historical movement from tribal society to capitalism is the question of what forms of community take shape that at the same time determine and designate the sphere of action of the individual, and his relation to the larger social unit. Actually, after tribal society, only three forms of societal cohabitation have emerged: those based on politics, the military, and economics: the polis, the Greek city-state (which lived on after a fashion in the Italian commune and the German Hansa towns of the late Middle Ages and early modern period, which were mainly organised for commercial purposes), the empire (which was the basic societal form from the beginnings in the ancient Orient up to the twentieth century and gave rise to an extraordinary abundance of varieties of form), and finally, the nation-states of capitalism, which have gone through a variety of phases. Within these societal formations, various things determine the position of the individual. As is known (Marx formulated it precisely), capitalism is the only "economic societal formation!"2, and it is not easy to identify the "preponderant factors" of the others. The community of the polis consisted of citizens with full civic rights, was bound together by shared religious acts and was organised politically; in its classical form it was based on direct, therefore not representative, democracy. Its functioning was, naturally, governed by a number of great aristocratic clans, which directly controlled military matters. There law-givers, magistrates and executive were the repositories of power, and the full citizen was related to the community by right (therefore not by virtue of high birth, wealth or any other consideration), but his subsistence was based on the work of a vast number of slaves. We know that this form presumed a small community; in the case of Sparta, a few thousand full citizens controlled tens of thousands of helots3, but in the case of Athens, the ratio of politai to slaves was twenty to one4, and the corruptible democracy – a hotbed of permanent stasis and contention – took catastrophic decisions on many occasions. It is not surprising that Athenian theories of the state were without exception undemocratic.5 From ancient Egypt and Mesopotamia, through Achaemenid Persia, the hellenistic and Roman empires, and classical Islam continued in the Ottoman Empire, to the industrialised empires of modern Europe, Tsarist Russia and the Soviet Union, empires have constructed a range of models6 among which perhaps the most significant have been the type dubbed by Weber "patrimonial".7 Since we shall be returning to this when we characterise Islamic society in what follows, it will suffice to note at this point that with regard to its essential nature, the empire is a formation based on taxation, presumes the ruler's autocracy in military terms, is typically created by conquest, and embraces numerous ethnicities, cultures and languages. It is essential that the military-political exercise of power is indifferent to the economy, and so is based not on a definite economic pattern – determined by ownership, division of labour and means of production – but on one in which the most heterogenous economic circumstances can be found, that is to say, in this case it is not the economic factor that defines power-sharing but on the contrary, wealth follows from rank (this explains the powerlessness of those engaged in economic life, principally traders). The ruler of the patrimonial empire, in Weber's artful description, regarded his country as a huge princely estate to be taxed at pleasure8, the structure of which is shown by the trinity of ruler, servants and subjects. This was therefore in essence indifferent, or in a worse case hostile, to the community.9 It is therefore clear that the umma cannot be assessed from the aspect of practical rule. In today's terms, the former might be regarded as a sort of civil society that appears as a counterweight to the ruler's limitless power. In itself, the umma is hard to grasp – it is a virtual construct which defines itself inwardly against the abuses of power/authority, outwardly against the world of the unbelievers, and the paradox that to this day marks the society created and shaped by Islam is the nature of the relationship between Islam-community that has no formal institutions, and the ruler (khalifa, amīr al-mu'minīn, malik, sultan) on whom there are no legal restraints: which of them is the repository of societal hegemony, and what tools does the umma have against wielders of power? Furthermore, the problem with the term is that compared with earlier concepts, it only acquired its current meanings ("the nation of Islam"10, "the religious-political community of Islam"11, "the Sunnite community based on sharica, divinely guided, which has gained it continuity from ijmac"12) only in recent centuries, and indeed only in the most recent times, and both Muslim popular opinion and much of the specialist literature tends to take these back to early days, even though at that time these meanings cannot be deduced either from Mohamed's practices or from the Koran. The latter links a divergent (ethnic, religious, ethical) meaning to a variety of (not only Muslim) communities, and its meaning resembles rather the related South Arabic term lummiyya 'tribal alliance'.13 We may therefore interpret it as initially as "a body of a tribal nature created for a political purpose" which as yet has nothing to do with the other organisation (defined in what follows), that is, the religious community established by the Prophet and directed by his successor.14 A most important and indisputably contemporary document for its first meaning – that which may be regarded as fundamental in the Islam of Mohamed's time – is the so-called "Constitution of Medina", which, along with the Koran, is perhaps our most important contemporary evidence for the second half of the Prophet's activity and for the state-formation of early Islam. In pursuance of this, a constantly renewed debate began in the twentieth century, starting with J. Wellhausen; this document is not easy to interpret and agreement has been reached on only a few questions.15 Here, of course, only hypotheses concerning the umma are important to us. In Watt's opinion (obviously as an acceptance of later interpretations), it is a theocratic community directed by Allah, which replaces the traditional tribal organisation based on bonds of blood. Later, he changes his opinion fundamentally: the Koran does not support him; umma is not distinguishable from qawm; there is no question of a religious community; it seems to be essentially a mixed alliance of Arab and Jewish clans. Wellhausen emphasises the military nature of the alliance (Schutzgemeinschaft Allahs). In Serjeant's view, it is a matter of a political alliance of independent tribes without any religious connotation. Finally, Gil strips the concept of its functions; according to him, it simply means a "group". Our opinion is that Mohamed did not use the term in any kind of clearly crystallised sense (in the Koran, we tend to find milla used with religious connotation16). The concept designates a tribal alliance of a political-military nature, consisting of a mixture of elements, and it may be said that of the concepts this far analysed this is the only one that forms an authentic part of the beginnings of statehood under Mohamed, convincingly demonstrating that inchoate Islam had in many respects not radically parted from tribal society. In its essential characteristics, the umma of Mohamed's day displays elements of the latter. The later umma did not continue these beginnings.