The Old Question
- The Sources of Social Power. Vol I: A History of Power from the Beginning to AD 1760 by Michael Mann
Cambridge, 549 pp, £37.50, July 1986, ISBN 0 521 30851 8
Books on the theme of society-down-the-ages generally fall into one of two kinds. Either they are a narrative synthesis organised according to some preconceived criterion of historical significance, or they are an attempt to test against the historical evidence some would-be general theory to the effect that demography, class struggle, national psychology or whatever it may be is the master key to the explanation of the whole long story. Michael Mann, however, deliberately places himself mid-way in-between. As a result, he risks being simultaneously attacked by one set of readers for writing potted history at a safe distance from the sources, and by another for theorising at an insufficiently high and abstract level. But this is, in a sense, the point of the exercise. ‘Historical sociology’, so called, needs to vivify the higher theory with the right infusion of historical detail and vice versa, and it is by this that its success or failure requires to be judged.
The full text of this book review is only available to subscribers of the London Review of Books.
Vol. 9 No. 6 · 19 March 1987
SIR: I am puzzled by W.G. Runciman’s review of my book, The Sources of Social Power. Volume I: A History of Power from the Beginning to AD 1760 (LRB, 19 February). If it is as bad as the substance of his comments would indicate – lengthy enumeration of five ‘serious and even crippling defects’, plus solecisms and spelling errors – why is my ‘show still on the road’ for Volumes II and III, as he briefly suggests when opening and closing his account? This may have doubly puzzled your readers since he gives no indication of what the show is or what are its main arguments. Perhaps in rebutting his five defects I can briefly indicate the nature of my show.
1. I identify four fundamental sources of social power: economic, ideological, military and political. As an orthodox Weberian, Runciman wants to reduce this to three. He accuses me of confusing ‘forms’ with ‘means’ of power, and regards political power as not analogous with the others (in his own work he combines it with military power). Frankly, I do not care much for analogical neatness when dealing with the mess that is society and history. I notice only that four especially powerful types of organisation, each with its distinctive form (political is centralised and territorial, military is concentrated and coercive), appear ubiquitously in the history of societies, and that to study their interrelations affords us a decent level of overall understanding. My history then attempts to prove this, with a distinctive methodology which concentrates on the logistics of power – how messages, materials and manpower are communicated and controlled by power organisations. Perry Anderson and John Hall have labelled my entire approach ‘organisational materialism’. Your reviewer gives only the barest hint of this.
2. Clearly, however, if I have ‘a tendency to lapse into oversimplification’, my history cannot prove anything. Runciman gives three examples. One, a trivial metaphoric aside on Byzantium which is given too precise a context, I concede. The second example is my unwillingness to concede that Rome at its height was simply ‘more powerful’ than Classical Greece. Who is oversimplifying, I ask? The context is an argument that only rarely can societies be ranked simply above or below each other in power resources. They often differ qualitatively as well as quantitatively in their power achievements. The example still seems apposite: many of the Romans who conquered Greece and found themselves adopting Greek ways asked themselves the same question: who had really ‘won’? But again, nothing important hangs on the example. The book contains more central ones – for instance, the distinction between intensive and extensive power that is used throughout. Hence the argument that Medieval Europe developed extraordinary intensive powers, such as small groups pioneering local agricultural innovations, compared to the more extensive powers, the pacification of large territories, mobilisation of enormous armies, etc, of China in the same period. Runciman alludes briefly to this in saying I ‘condescendingly dismiss’ Chinese achievements as merely extensive. He gives no hint either of the meaning of ‘extensive’, or that the whole point is not to rank Medieval Europe above China, but to assert the qualitatively different nature of their achievements.
Only the third ‘oversimplification’ would matter, since it is part of an important theme, the developmental role of salvation religions. I assert, as he says, that early Christianity and Islam both possessed a ‘particularly dynamic and contradictory power form’. He finds this unacceptable. As he gives no clue of the nature of that power form, or why he disagrees, the reader cannot judge. My claim is that Christianity and Islam both differed from the other world religions in incorporating into their core doctrines both an egalitarian doctrine of individual salvation and an authoritarian justification of secular authority (Islam incorporated the latter rather less). I argue that this was not only contradictory, but also led to a dynamic interaction between church-state hierarchies and radical reform movements. The argument is not original to me, I discuss it at length, and I stand by it.
But only the third example matters anyway. Is this all my oversimplifications amount to? What about my core arguments detailed below in response to Runciman’s fourth ‘defect’? Are these all acceptable?
3. I do not discuss Africa or East Asia, complains Runciman. True. No, it is even worse than he says. I do not discuss anywhere outside the ancient Near East, the Mediterranean and Western Europe – apart from two comparative forays into the origins of civilisation and salvation religions and a few asides. But then my book does not claim to be a world-history. It is a history of power, and the Introduction makes quite clear that my history is of the ‘leading edge’ and most dynamic power region in any one period within a single civilisational area. The ‘inadequacy’ of my treatment of Byzantine or Chinese history is surely an understatement. They are not ‘treated’ at all!
4. This is a very similar defect: I discuss Medieval European development without much reference to China! What an outflanking movement Runciman has mounted! We may ask why he mounts it. His review is titled ‘The Old Question’, meaning, he says, ‘why did “modernity” arise in the European West?’ That may be his question, even Max Weber’s question, but it is not mine. I am concerned with power development, to be sure. But ‘the European miracle’ is only one of half a dozen I discuss, from the breakthrough to civilisation onward. My predominant method is historical rather than comparative because they occurred in a just-about continuous and cumulative sequence of development. I have attempted to reconstruct the sequence. This enables me to make some points concerning development itself: to construct a list of the techniques that enabled ‘power jumps’ to occur (from irrigation through the invention of the state, diffuse upper-class literacy and Marius’s pole, to commodity production) which demonstrate the importance of all four power sources; to analyse the geographical migrations of the leading edge of power; to mount forays against a too-static comparative sociology. Has he noticed any of this?
5. Most crushing of all, he fails to perceive any overall argument. Some of my cases are interesting if speculative (ancient Mesopotamia), or persuasive though eclectic (Rome). But they stand quite independently of my general arguments at the beginning and end of the volume. He then enumerates his four solecisms-misspellings-typos. I am slain!
Except that there is a general model which does inform the individual cases. Societies are viewed as multiple, overlapping and intersecting socio-spatial networks of power – principally mixed or occasionally ‘pure’ networks of economic, ideological, military and political power. Each network has a logistical infrastructure whose capacities are carefully assessed, and whose development is charted historically. No single one can be assigned historical primacy. To ascertain their exact and varying powers we must analyse long-run processes of historical development and interaction. Hence there follows a theory of the origins of the state and social stratification based on the ‘social and territorial caging’ provided successively by alluvial agriculture and military conquest; a dialectic over two millennia between Empires of Domination and ‘multi-power-actor civilisations’; a phase in which normative pacification is provided by salvation religions; an explanation of European development in terms of localised and non-coinciding economic, military and political power networks within the broader normative pacification of Christendom; the emergence of the modern world through the interactions of capitalism and a multi-state system; plus persistent themes such as the limited but developing role of classes and nations in the history of agrarian societies. This is to state my themes in 150 words. The book is 200,000 words. Of all this Runciman gives not a word.
Why has he so consistently missed the point? Is it because he is preparing a work whose model is the Weberian class-status-power, whose method is comparative, and whose cases include Byzantium and China. Is this the only legitimate sociological enterprise?
W.G. Runciman writes: I am sorry that Michael Mann should be so upset by a review which sought to convey that despite its failings his book deserves the serious attention of sociologists and historians alike. But failings they are. I see no reason to modify my view that his treatment of ‘power’ rests on a conceptual mistake, that he sometimes (but only sometimes) lapses into misleading oversimplification, that a so-called ‘history of power’ needs to cover more than its ‘leading edge’, and that selective narrative is no substitute for comparative analysis. Nor does his 150-word synopsis disclose any more of an ‘overall argument’ than I had perceived already: there is, rather, a general approach whose value my review endorsed and an interesting set of, to use his own word, ‘themes’.