The Collapse of Complex Societies 
by Joseph Tainter.
Cambridge, 250 pp., £27.50, June 1988, 0 521 34092 6
Show More
Show More

The main text of this book takes up only 215 pages. It tends to be repetitive and includes a number of not very well designed diagrams and maps. To that is added a list of about 630 references and an index which does not include all the references. Furthermore the references are so constructed that the reader is left unsure about whether the author has really consulted his source or why. For example, Paul Valéry died in 1945 at the age of 74. He had been elected a member of the Académie Française in 1925 and is primarily renowned as a poet. But Tainter brands him as the ‘noted French social philosopher’, with the suggestion that he was still alive in 1962. From time to time he asks his readers to consider the implications of his story for the world that we now live in, but he would certainly argue that his story is an archaeological story. And that, too, raises problems.

In the course of history many thousands of human settlements have left their archaeological traces in the ground. In some cases, excavation shows that the original site was simply abandoned and then plundered; in others, new ‘cities’ were built on top of and from the residues of old ones; in the vast majority, archaeologists have not so far got around to doing anything at all. As Tainter makes clear, the circumstances in which the societies which are now represented by plundered ruins met their end are very varied. In most cases, what is presented by the archaeologist as an historical reconstruction of events is simply guesswork. And the further back you go in time the more guesswork is involved. Tainter seldom writes about anything later than 1100 AD.

Although archaeologists and social anthropologists frequently find themselves in the same academic faculty and can even, on occasion, share a subject-matter, they are, in a case such as that of Joseph Tainter and myself, poles apart. This is because social anthropologists are primarily interested in a rapidly disappearing living present; they suffer from an illusion that if they do not record everything within the next ten years there will be nothing left to write about. By contrast, archaeologists of the Tainter sort are only concerned with the remote past. They are aware that what has actually been excavated is only a tiny fraction of what might have been excavated. They are also aware that since excavation techniques are improving all the time there can be no real urgency; the longer the delay, the greater the reward. But careers are at stake; men like Joseph Tainter are under tremendous pressure to make sense of their diggings long before they have the data which would permit such elucidation. Tainter’s solution is to claim that the various rules of history claimed by his predecessors, if suitably modified, will fit all possible cases, past, present and future. The magic that will fit everything is ‘collapse and the declining productivity of complexity’.

Here is an example. Between 1950 and 1968, Stuart Piggott and Mortimer Wheeler both wrote with extreme confidence about the overthrow of the Harappan (Indus) civilisation by invading Aryans at the beginning of the second millennium BC. I have long found this particular style of reconstructed history totally unconvincing, but it seems evident that for most of the time that he was working on the present book Tainter, like many other scholars, followed the Piggott/Wheeler line. Then, just as he was coming to the end, a mass of material was published (in 1982 and 1984) which is quite incompatible with the Piggott thesis. Only a very careful reader will appreciate that Tainter realises this inconsistency. His argument about Harappa is split up: at page 6 half a page; at page 48 12 lines; at page 62 five lines; at page 71 six lines; at page 204 four lines. At page 6 we are discussing ‘a highly centralised society in which the state controlled many facets of daily living’. At page 204 we are told that ‘current research suggests that there were indeed several independent Harappan states.’ This is no way to marshal an argument on one side or the other. Tainter would have been well advised simply to forget about Harappa just as he has managed to forget about a vast number of other archaeological sites which do not fit his theory.

Tainter is reticent about his past career, but the present book is published in the high-status Cambridge series ‘New Studies in Archaeology’ edited by Colin Renfrew and Jeremy Sabloff and is a development from a paper presented to the Society for American Archaeology in 1982. Tainter lists five earlier publications which have appeared since 1977. They all relate to American archaeology in New Mexico and Arizona. Tainter evidently belives that the data from such sites, about which he is quite well-informed, can best be interpreted by reference to sites in various parts of Asia and the Mediterranean, about which he would seem to know very little.

The third section of Tainter’s first chapter is entitled ‘Collapse in History’. It takes up 12 pages. We are given the names of eight ancient ‘civilisations’ followed by eight pre-Colombian Mesoamerican societies then two seemingly wholly inappropriate ‘simple’ societies, the Kachin and the Ik. We do not return to the Kachin or the Ik until the final chapter, but in other respects most of Tainter’s ‘examples’ are drawn from this list. But what are they examples of? The answer to that question is ‘variations in complexity’, a topic which is discussed at length in Chapter Two. Tainter takes his cue from R.H. McGuire, a specialist in the prehistory of South-Western Arizona. Complexity is here contrasted with ‘simplicity’. The experts on the latter theme are cited as Morton Fried, Max Gluckman, Marshall Sahlins – all social/cultural anthropologists. The distinction is important. Tainter would have us believe that prehistorians are concerned with the dead, and social and cultural anthropologists with the living.

Outside the Americas, Tainter’s list is: the Western Chou Empire, the Harappan civilisation, Mesopotamia, the Egyptian Old Kingdom, the Hittite Empire, the Minoan civilisation, the Mycenean civilisation, the Western Roman Empire. The parallel American list is the Olmec, the Lowland Classic Maya, the Mesoamerican Highlands (Teotihuacan, Tula and Monte Alban), Casas Grande in the far North of Mexico (it lasted from 1060 to 1340 AD), the Chacoans of the San Juan Basin in north-western New Mexico, the Hohokam of the southern Arizona desert who survived in diminished form into the 15th century AD, the Huari and Tiahuanaco of Peru. The Aztec in Mexico and the Inca in Peru about whom much more is known are virtually ignored. Tainter doesn’t really have any information about Chou China other than that it was a feudal state. There is no systematic archaeology. What he says about these people is 95 per cent fictional. But he cites the standard English language ‘authorities’. For the Harappan civilisation there has been a spate of recent literature based on detailed excavation, but Tainter is unfamiliar with most of this. Most of his references insist that the Harappan civilisation was a centralised regime, but excavations by M.R. Mughal supported by commentary from Gregory Possehl, dated 1982, lead to the opposite conclusion.

A confident one-page summary makes statements about the reduction of the Mesopotamian population between 600 and 900 AD for which there seems to be no evidence. Tainter ignores the fact that this was precisely the period that corresponds to the rise of Islam. The Prophet Muhammad was born about 570 AD.

What Tainter says about the Egyptian Old Kingdom seems archaeologically up-to-date, but when he writes of the fifty-year period subsequent to 2181 BC that ‘national centralisation collapsed ... Tombs were plundered, royal women were clothed in rags, and officials were insulted,’ he is letting his imagination run away with him.

And so on. The Hittites, the Minoans, the Mycenaeans are all represented by Tainter as civilisations of the same kind. A ‘great ruler’ builds a palace and enters on treaty relations with other ‘great rulers’. Then ‘disaster strikes’. Cities are destroyed; the population declines. History repeats itself in a predictable cycle. And perhaps if you are not too fussy about the smaller details, this is true. Every state has a beginning, a middle and an end. And if you are patient and wait long enough it will (eventually) start again.

Elsewhere Tainter discusses the decline of the Roman Empire at some length; he also mentions the Maurian Empire of Asoka; Byzantium, Spain, the Netherlands and Easter Island are disposed of in less than one page taken all together, but the later history of China, Korea, Japan, the whole of South-East Asia, Indonesia, Iran, the Mongolian Empires, Turkey, the Macedonians, Classical Greece is ignored. More surprising is the fact that, apart from the Hittites, the succession of ‘fertile crescent’ civilisations, from Ebla on-wards, which have been dug over very enthusiastically by Middle-East archaeologists are not discussed at all. Nevertheless Tainter seems to believe that his very non-random sample of evidence outlines a general case and tells us something we did not know before.

The heart of the matter is to be found in Chapter Five. It contains condensed, but relatively detailed, information about three of Tainter’s alleged ‘collapses’: the Western Roman Empire, the Lowland Classic Maya, and the society which flourished in the Chacoan Canyon between about 725 and 1225 AD. Just why these three cases should be compared is not clear, but if you start out with the circular assumption that, despite the contrast of time and space and ecology, they must be comparable because all three examples are cases of ‘the collapse of complex organisation’, you will find that it is so. Tainter produces the formula: ‘the collapses of the Western Roman Empire, the Southern Lowland Maya and Chacoan society can be understood as responses to declining marginal returns on investment in complexity.’ Perhaps, perhaps not. Complexity is in itself very complex!

Most European readers of the London Review will probably never have heard of Chacoan society. The name refers to a Pueblo-type cluster of ruins lying somewhat to the east of the present Navaho Reservation in New Mexico. It has been abandoned for the past 800 years and now rates as a National Monument. Tainter presumably uses it as an example because he has worked as an archaeologist on this particular dig. As it happens, I have been there myself. Certainly it is an unusual sort of place, but not a good site on which to base a law of historical decay. And that is the trouble. Tainter freely admits that he is trying to do the same sort of thing as Arnold Toynbee attempted in his 12-volume Study of History or Oswald Spengler in The Decline of the West, but on a very reduced scale – just 215 pages. I am puzzled as to why the practice of archaeology should lead its practitioners into fantasies of this kind.

At page 205 we come back to the Kachin and the Ik. Since I myself am responsible for any delusions that are current about the Kachin, I would like to deny that the distinction between gumsa (hierarchical/autocratic) and gumlao (egalitarian) political organisation is of the kind described by Tainter when he alleges that I have ‘argued that in simpler societies investment in political expansion, with insufficient return to the local level, engenders disaffection and collapse.’ The gumlao are not a ‘collapsed’ form of the gumsa (or vice-versa). Most gumlao communities are larger than most gumsa communities. But why should this matter?

The final chapter informs us that ‘collapse is recurrent in human history; it is global in its occurrence; and it effects the spectrum of societies from simple foragers to great empires.’ ‘Collapse is fundamentally a sudden, pronounced loss of an established level of sociopolitical complexity.’ The ‘collapse’ and ‘complexity’ are so defined that if you have the one you lose the other, and since, by Tainter’s own showing, there is no single cause of ‘collapse’, complexity must be that combination of factors which in any particular case results in collapse. Q.E.D.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.


Vol. 11 No. 5 · 2 March 1989

It is an honour to have my book, The Collapse of Complex Societies, reviewed by so eminent an anthropologist as Edmund Leach (LRB, 10 November 1988), and I am sorry to hear of his recent death. His review, however, contains many misunderstandings and deficiencies which I feel obliged to address. For readers of this reply, the book sets out to explain why societies collapse – that is, drop suddenly to a lower level of complexity (as happened, for example, to the Western Roman Empire, the Maya, and many other societies). The thesis is that societies become vulnerable to collapse when, as costly social features are added, the investment in social complexity reaches the point of diminishing returns.

I will not deal with those passages in which Leach impugns my character. Since Leach acknowledges that he had never heard of me, readers may draw their own conclusions about his competence in this regard. Readers may also, if they wish, join me in speculating why Leach found it necessary to write an ad hominem review. Perhaps the answer can be found in Leach’s own words: when scientists disagree philosophically, they often argue by ‘[throwing] doubt on the credentials of the investigator’ (‘Concluding Address’, in The Explanation of Culture Change).

I will deal first with the misrepresentations and fabrications in Leach’s review. 1. Even the most hurried reader will discern that Leach is completely wrong when he asserts that I am ‘only concerned with the remote past’. The book deals with regularities in social evolution, past, present and possibly future. 2. I never claimed, as Leach states, ‘that the various rules of history claimed by [my] predecessors, if suitably modified, will fit all possible cases.’ To the contrary, while I respect the work of my predecessors in this area, it was the deficiencies in existing theories that prompted me to undertake the study. 3. I most certainly do not follow ‘the Piggott/Wheeler line’ regarding the putative Aryan invasion of Harappa. In fact, my remarks on that thesis are quite critical. 4. Leach claims that the five references to my own work ‘all relate to American archaeology in New Mexico and Arizona’ (thus implying that I am unqualified to write about other places). In fact, only one of the five is about the American South-West. 5. Leach tells his readers that I ‘evidently [believe] that the data from [South-Western] sites … can best be interpreted by reference to sites in … Asia and the Mediterranean.’ Not only do I not believe that, I nowhere asserted any such thing. 6. The statement that ‘Tainter would have us believe that prehistorians are concerned with the dead, and social and cultural anthropologists with the living’ is a complete fabrication. I have never written anything of the sort. 7. I did not, as Leach claims, represent the Hittites, Minoans and Mycenaeans ‘as civilisations of the same kind’. 8. Finally, I nowhere stated that I am ‘trying to do the same thing as Arnold Toynbee … or Oswald Spengler’. Leach has employed a peevish and undisciplined imagination in place of a careful reading of the book.

Leach claims that what I wrote about Chou China ‘is 95 per cent fictional’, that in describing the collapse of the Egyptian Old Kingdom I was ‘letting my imagination run away’, and that there is ‘no evidence for Mesopotamian population decline between 600 and 900 AD. Here Leach lets his standards of scholarship sink even lower. Each of my descriptions of collapsing societies is taken from standard, competent authorities, whose works I have painstakingly cited. Leach could easily have checked any of these sources, but preferred facile disparagement to honest debate. In regard to the Western Chou, I wrote 104 words about the collapse of this state. I wish Leach had told us which five words he considers accurate.

Leach complains that I treated the Aztecs, the Inca, Byzantium, Spain, the Netherlands and Easter Island briefly, and ignored altogether ‘the later history of China, Korea, Japan … South-East Asia, Indonesia, Iran, the Mongolian Empires, Turkey, the Macedonians, Classical Greece’. He is partly right: I didn’t discuss any of these to any significant degree. Why? Because most of these societies never collapsed, and collapse is the subject of the book. This is a fatally revealing point: Leach did not understand the topic of the book he criticises so vehemently.

On the basis of one visit – during which he apparently learned little – Leach dismisses ‘Chacoan Canyon’ [sic] as ‘not a good site on which to base a law of historical decay’. Aside from the fact that I advanced no such ‘law’ (nor any notion of ‘historical decay’), Leach would do better to restrict his pronouncements to Burmese ethnography and French poets. Chaco Canyon itself is but a small part of what archaeologists call Chacoan Society, which is represented today by a network of ruins and interconnecting roads spanning more than 85,000 square kilometers. Chacoan Society developed over several centuries to a level of complexity without parallel in the prehistoric northern South-West. Then, within a few decades, it collapsed. This is precisely the sort of occurrence that students of collapse want to explain, and Chaco is one of our best-known archaeological cases.

Joseph Tainter
Corrales, New Mexico

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN

Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences