Michael Mann has recently published a large book on the history of power which few historians or sociologists can afford to ignore.footnote¹ Its 549 pages are only the first volume out of three; Mann’s aim will at the end be a massive retheorization of the sociology of power, on the basis of historical examples taken from the span of human history from the Neolithic Revolution to the present day. Vol. II will complete the history in the capitalist/industrial world (it will include a discussion of gender relations too); Vol. III will be the theoretical volume. However, there is plenty of theory in Vol. I, integrated into the historical analysis with a stylish verve and (usually) a satisfying neatness. Mann is ambitious: he remarks in his first sentence that he initially wanted to refute Marx and ‘reorganize’ Weber and, although he indicates that he has stepped back a bit from this, he still does, really. Whether he has ultimately managed or not will presumably have to await Vol. III: not least because Marx (although not, of course, Weber) was not particularly concerned with the pre-capitalist world. But Mann’s models are already pretty clear in Vol. I, and we can learn a good deal about how far he has reached. Enough, indeed, for me: I am a practising historian, and prefer to discuss my theory as intermingled as possible with historical example.

Mann’s arguments are rich and complex, and one cannot possibly hope to engage critically with them all. He is also very skilled at covering his back with qualificatory paragraphs and throwaway lines that reintroduce the subtleties of analysis that the historian likes, and that Mann’s main lines of argument inevitably, and properly, exclude. His control over his material and over his models impresses the reader, almost throughout. He has not reorganized Weber (a hard task, given Weber’s complexity and heterogeneity), but it is with Weber that one finds oneself comparing the book, which is already no mean achievement. The other pole of the comparison would, however, not yet be Marx, but perhaps rather Perry Anderson, whose two books Passages from Antiquity to Feudalism and Lineages of the Absolutist State represented, a decade or so earlier, a similar sweep across history, expressed with a similar joie de vivre. That Anderson’s standpoint is Marxist and Mann’s what could be called post-Weberian is not a result of this time difference (Mann started his book in 1972), but the dates of publication of the two works, considering the difference in their intellectual positions, are emblematic: I will return to the point at the end. Mann is often tighter than Anderson, who, ironically, can easily get sidetracked into issues of strictly political history with insufficient regard for the socio-economic structures developing below; Anderson, by contrast, has a wider perspective than Mann, that in part derives simply from his use of books in more languages than Mann (whose bibliography is restricted to works in English, with a few in French), but is largely the result of a much greater interest in comparison. Mann’s (un-Weberian) caution about comparison is in fact one of his weak points, as I shall argue later.footnote²

My intention in this review will not be to take Mann on in order to rescue Marxist history from his critique. This would be premature, but also, more important, pointless: for Mann has a great deal to offer Marxists, and (as Weber did, though less covertly) employs a fair amount of explicitly Marxian analysis. Where he differs from Marxism, it is often because he wants to analyse different historical processes from those that form part of the Marxist project. So, instead, I will begin by following through some of his arguments in a constructive spirit, commenting on what I see as his strengths and weaknesses. I will look at his analyses of the origins of the state, of the development of empires, and of Christianity and the rise of Europe. To conclude, I will discuss some of the more general issues that arise from Mann’s discussions of the relationships he shows us between his four sources of social power, ideological, economic, military and political; for these four sources and their interrelationships form the theoretical core and the narrative dynamic of his book.

Mann alarms us on p. 2: ‘it may seem an odd position for a sociologist to adopt; but if I could, I would abolish the concept of “society” altogether.’ The alarm may not be Mann’s fault, for doubtless he wrote the sentence before Margaret Thatcher pronounced that society did not exist; in fact, however, he is not as radical as appears at first sight. He means, above all, two things: that societies are not unitary, for there are many intersecting networks of social relationships in any environment, which cannot be tightly articulated into a single social ‘structure’ or ‘system’; and that they are not bounded, for these networks overlap those in other ‘societies’, rendering any neat territorialization impossible. He defends this perspective against both Marxist and Weberian theory (pp. 11–18): a little unnecessarily, perhaps, for the Marxist conception of the social formation, even if it regards the economy as a whole as primary, by no means entails a ‘unitary society’—indeed, in many formulations, it explicitly excludes it; and I have encountered both Marxists and Weberians who have little trouble with the concept of unbounded social interaction, which seems to me, too, to be a normal part of any sophisticated historical analysis. Mann argues, further, that the non-coincidence of state, culture and the economy makes any simple causal models a priori invalid. He implies (p.4) that ‘ultimate primacy’ is something we can still look for (presumably he will unveil how in Vol. III); but ‘the economy’, or ‘ideology’, cannot do the job on its own simply because the networks of power are not superimposed. Instead, they float freely, and interrelate causally in different ways at different historical times. Be this as it may, Marxists will perhaps be relieved to find that Mann’s explanations are usually in reality, although always complex, largely materialist (more often, as I shall argue, than he says)—just as most people will be relieved to find that he uses the word ‘society’ throughout as a heuristic term.

How Mann proceeds can be clearly seen in his discussions of the origins of stratification, civilization and the state (distinct concepts, but in effect, in history, appearing together), which form the core of his first three historical chapters (pp. 34–129). He argues that human society evolves almost naturally towards ranking, but not beyond ranks to firm stratification, the creation of irreversible hierarchies that involve coercive power over people. Here, evolution stops and more complex social processes take over; change henceforth is more usually cyclic (the endless repetition of similar patterns) than structural (the irreversible move from one pattern to another), and the development of stratification is a structural change. Mann does not like the concept of social evolution, partly because it does not explain global human development (he spends many pages on this issue, although the fact that human societies have not all developed in the same direction or at the same rate does not need much demonstration); but also, more important for his model, because evolution is essentially an endogenous explanation, and Mann’s theory of unbounded networks leads him to regard endogenous structural change as virtually impossible. Mann has a reasonable point here; in particular, it does indeed make little sense to see any clearly bounded social groups before the development of the state. Instead, he explains his first state/civilized society, Sumer, as the product of a social dialectic, a ‘caging’ process in an irrigated landscape, where a network of small political groupings (the ancestors of the Sumerian city-states) interacted with nearby non-irrigating frontier societies. The inhabitants of the irrigated world were discouraged from abandoning rising rulers by the simple (‘caging’) fact that the non-irrigated world was poorer; the rulers began, perhaps, as local ‘big men’, but soon acquired real and permanent political power through their control of the exchange interstices between irrigated core and frontier periphery (pp. 73–89).

At first sight, this explanation does not differ greatly from a materialist evolutionist one, for Mann argues that this process happened most characteristically where an irrigated landscape interacted with a non-irrigated one (Sumer, China, the Indus, etc.). That is to say, when ranked societies got the chance, thanks to geography, to become stratified, they did so. Furthermore, when he confesses that Egypt, the major Eurasian exception to the model, managed to develop the state without any internal or external interactions (for Egypt had no independent cities; and no frontier either, for the Nile runs through desert—pp. 108–15), i.e. necessarily endogenously, one wonders whether Mann has not been too strict in his catastrophist anti-evolutionism.footnote³ But he is certainly right to stress that the rarity of the process and the length of time (many centuries) it took, even in successful areas like Sumer, mean that the break represented by stratification/civilization has to be explained as rigorously as possible. (It is noteworthy that he is more rigorous here, where there is of course no direct historical evidence at all, than he is in his discussions of the next major break, the origins of capitalism, where there is no shortage of material.) Mann’s network model leads him to stress the fruitful historical role of dual social systems, small city states in a wider cultural continuum (‘multi-power-actor civilizations’), which he finds in all his early civilizations except Egypt, as well as in Greece, (in part) Rome, and (with the city enlarged to the region or nation) medieval and modern Europe. He favours a double dialectic, that between the two levels of this continuum (city-state and wider culture) and that between the core cultural system and its frontier neighbours, in all his major historical cruxes, in fact. I like this model very much; it seems to me subtle, powerful, and capable of a great deal of interesting variation. Whether or not it is as anti-evolutionist as Mann says I think simply depends on one’s use of language. (The link between evolution and the dialectic goes back to Engels.) It is certainly not anti-materialist, however. Marxists are not likely to criticize it because of its rejection of purely endogenous explanation. They are likely to be critical, on the other hand, of Mann’s relative inattention to the relations of production, whose increasing exploitativeness he skips over rather fast (e.g. pp. 84, 101, 110, 113); the landlord–peasant relationship is mentioned, but never theorized, here or later, as we will see. But Mann is not so interested in the development of the economy; his conceptual focus, here as later, is really the state.

The next stage that Mann confronts is the development of empire, which he treats as a rising cyclical trend lasting nearly three millennia, starting with Sargon of Akkad and ending with late Rome (pp. 130–78, 231–300; the interval break is devoted to the discovery of iron, the Phoenicians, and in particular the Greeks). These chapters are highly satisfying and complex analyses of the effects of military power on overall socio-economic development. Here, in Mann’s analyses, three of his four power sources—economic, political and military—spin free of each other (ideology not, perhaps, quite yet, as we will see); and military power is in this historical epoch the causal element in the dynamic. Like other commentators, I think Mann’s distinction between political and military power is artificial, and insufficiently theorized (he will have to talk over doubters in Vol. III with more care). But this point is less important than his articulated and flowing description of the possibilities and limits of military control. He shows how hard it is, logistically, to maintain intensive military control over anything more than about a hundred kilometres’ radius (pp. 137–42); ancient empires were ‘extensive’ dominations, which all had to rule through local élites. Nonetheless, they were not, or not wholly, parasitic; they increased production by extracting surplus for the army, by maintaining peace, by protecting exchange, by building roads, by establishing economic value, by intensifying exploitation, by favouring literacy—economic development and repression (viz. economic power and political/ military power) hand in hand (p. 153). Mann works out the argument first for Akkad; but it still works, after at least eight cycles of imperial rise and decline, for Rome, where the organizational coherence of the legion allowed effective political control to extend far beyond its original radius (p. 272–80). Indeed, the argument for Rome (economic development related to repression) is a widely known one, for of course it involves not just the legionary economy but the greatest development hitherto known of slavery as a mode of production; the relationship would, indeed, have been even clearer if Mann had discussed Roman commercial activity more, for Roman commerce flourished very largely on the back of the state (see pp. 264, 271–2 for brief references; but the literature is largely in Italianfootnote⁴).