¹ It is very nearly a century if it be carried back to Danilevski's publication, in 1869, of his Russia and Europe. But that work remained wholly unknown in Western Europe until its translation into German by Nötzel in 1920. The time elapsed is two thirds of a century if it be carried back only to the publication of Brooks Adams’ Law of Civilization and Decay in 1898, which it must be since the book deals directly with the subject, however ineffectively. The last half century has, however, been the most active period, containing the work of both Spengler and Toynbee, as well as that of less speculative thinkers, such as Sorokin and, pre-eminently, A. L. Kroeber.

² The distinction does not obtrude very much, however, for the development of culture configurations in themselves is excluded.

³ Even in his first three volumes, Toynbee made an attempt to get further than this(Toynbee, A. J., A Study of History, I-III, London, O.U.P., 1934; classification of societies, I, 131–133Google Scholar) and recently, in his twelfth volume, he attempts (Ibid., XII, 546–561)a complicated functional-typological classification. Both these classifications are of some interest. Kroeber also began to formulate a typology and classification of civilized societies, which was based differently from Toynbee's (Kroeber, A. L., ed. LeGuin, Charles A., A Roster of Civilizations and Cultures, Viking Fund Publications in Anthropology, No. 33, New York, Wenner-Gren Foundation, 1–62Google Scholar). I do not think the time has yet come to attempt a detailed classification; far more knowledge is needed. Kroeber's partial classification, posthumously published, is valuable, but it is unfinished and experimental only. The short list given here (text, below) is the same, except for elimination of the “Russian Society” (I have changed my mind about that) as that which appears in my book, The Origin of Civilized Societies (Princeton, University Press; London, Oxford University Press, 1959Google Scholar); trans. E. Gaenschalz as Der Ursprung der Hochkulturen(Stuttgart, Kohlhammer, 1962, Urban Bücher), hereafter cited as The Origin.

⁴ Cf. The Origin, pp. 6–7.

⁵ Below, pp. 414–16.

⁶ None, at least, which offers more than guesses, or opinions given at second hand.

⁷ Albright, W. F. in The City Invincible, ed. Carl H. Kraeling and Robert M. Adams(Chicago, University Press, 1960), p. 72Google Scholar. There is alsoa body of controversial literature on the subject, notably by Kathleen M. Kenyon, Robert J. Braidwood and V. Gordon Childe in Antiquity and elsewhere. All these scholars and James Mellaart have contributed largely to the knowledge which now exists about the earliesttowns, of which the two best known are Jericho and çatal Hüyük.

⁸ The statement rests upon my argument that it was in the tropical forest that the society originated (The Origin, pp. 112–116). The argument remains fully valid and has been strengthened in a small way by confirmation that in the modern province of Tabasco, in tropical forest, the people of the Olmec culture lived similarly to the early Mayas; see Willey, Gordon R., “Mesoamerica” in Courses toward Urban Life, ed. Braidwood, R. J. and Willey, G. R. (Chicago, Aldine, 1962), pp. 84–101Google Scholar and references there given. It is perhaps a little more likely that very early beginnings were in Olmec country than in Maya country, but they must quite early have been diffused from the one to the other region.

⁹ John A. Wilson held, at least until December,1958, that Egypt began without towns, which would certainly imply that none existed in thevalley of the Nile before the riseof the civilized society. In the futile discussion ofthe subject (not necessarily futileon other matters about Egypt) published in City Invincible (pp. 124–164), Wilson finally(Ibid., pp. 163–164) gave a little feeble evidence for thinking that the Thebaid did not originally contain towns and that the same might be true too of the region of Memphis.On purely general grounds, I think Wilson may be right. Others in the discussion in Chicago all thought he was wrong, but the discussion was not of origins; it was vaguely of conditions at undefined, presumably mostly early, dates.

¹⁰ It arises, however, out of K. C. Chang's remarks (Courses toward Urban Life, pp.68–69) to the effect that earlier archaeological evidence of cities in North China suggests separate and self-contained units, whereas for later periods (Shang and later black pottery?) the culture seems to have become to some extent unified, etc. There could, of course, be other explanations of Chang's impressions of the evidence, but the one I suggest fits well with the detailed findings for the earlier period of Andersson, J. G. (articles in Ostasiatiska Samlingarna and Children of the Yellow Earth, trans. Classen, London,K. Paul, 1934)Google Scholar, passim.

¹¹ If this opinion is correct, it fits well with the absence of such people in or near Egypt and hence with the absence of fortified towns there: except in the Delta region and until about 5000 B.C., the land outside the Niledepression had long ago been dried out. These considerations could affect the cases of Middle America, China and also India, but too speculatively to be worthy of discussion yet.

¹² In fact, I reject categorically all arguments, from those of Childe onward, to that effect. Where, as is fashionable at the moment, there is no argument, but merely a tacit assumption that towns are synonymous with civilization, the issue is evaded even if unintentionally. This does not necessarily inhibit theamassing of new empirical evidence, but it gives the unfortunate impression that those whodo it have stopped thinking at the point the Greeks reached.

¹³ The Origin, pp. 94–97.

¹⁴ Ibid., pp. 120–128.

¹⁵ Butzer, Karl W., Environment and Archeology: An Introduction to Pleistocene Geography(Chicago, Aldine, 1964), pp. 449–453Google Scholar.

¹⁶ Ibid., pp. 461–464.

¹⁷ See the argument in The Origin, pp. 67–69. There is no need to repeat it here. It is merely necessary to drop out the factor of desiccation; the remainder stands.

¹⁸ We have been disposed to describe the doctrines so set up in the early stage - the Age of Faith - of civilized societies as myth. So it is in the literal Greek meaning of the word, but that must not mislead people into thinking that it was not regarded as fact- even as ascertained and fixed truth - by the charismatic leaders who propounded the doctrines and by their followers who - more or less - lived by the doctrines. Henri Frankfort's concept of the “mythopoeic mind” (See Kingship and the Gods: A Study of Ancient Near Eastern Religion as the Integration of Society and Nature, Chicago, University Press, 1948Google Scholar, and Frankfort, , Frankfort, , Wilson, , Jacobson, and Irwin, , The Intellectual Adventure of Ancient Man, Chicago, University Press, 1946Google Scholar, etc.) is, on this subject, both good and bad. It was certainly very useful and still is that he thought it worth enquiring into the subject and presenting a sophisticated understanding of it to specialists in late prehistory, who, nowadays, tend to be rather impatient of such things.On the other hand, I doubt very much that there is such a thing as a mythopoeic mind, distinct from other human minds, or operating very differently; the idea sounds likea survival from the imaginings of Lévy-Brühl. In any case, to think that such minds dominated all thought before the Greeks is antediluvian and will not be discussed here.My “Concept of the ‘Conglomerate Myth’ ” (Proceedings of the Tenth International Congress of Philosophy, ed. Beth, , Pos, and Hollak, , I, Amsterdam, 1948, 74–81Google Scholar) is an early attempt, and certainly an insufficient one, to grapple with such matters, but it may still be useful. -We are not in any difficulty in substantiating the mentality and the mode of leadership at the origin of the primarycivilized societies, as these matters are treated here. We need not be afraid of accepting them as fact in the absence of direct contemporary evidence in written documents or archaeological finds; of the latter, indeed, a few exist. Quite adequate authority exists in later written records, and extrapolation back from them gives the most direct evidence available even if it is inevitably full of gaps and unsure in detail. Moreover, the beginning of every new cycle in the later development of civilized societies confirms the generalities of the charismata of the first cycles of the primary societies and indeed serves to show what was general and positive to them all. (I relied on these two kinds of supportin writing Chap. V, “The New Religions”, of The Origin, pp. 129–171; I think now that I should have broadened out the doctrine, in the cases where the evidence warrants it, and brought agriculture in more fully, rather than insisting so narrowly on water-gods. I make no apology for a few flights of fancy; they are needed and can easily be thrown out if disproved.)

¹⁹ It is very hard to say how many and it certainly varied. In Mesopotamia, about which there is now some knowledge for the period, it was about 1000 years, but in Egypt, Crete and China it was probably less. For India there is a complete blank, a muddle for Middle America, less muddle but a lot of argument stillfor Peru. These figures (which really matter rather little) are based on very recent research, mostly archaeological, and they are highly tentative. - N.B. This essay has now passed beyond the period covered by The Origin.

²⁰ The last three paragraphs are constructed out of bits and pieces known for one or some societies in each instance and in a few cases for them all. The pieces are collected, naturally, from all sorts of sources, consisting principally of archaeological work, but already including some written materials, all of later date, also some modern theory.Two works of synthesis on the New World societies have proved very useful: A Reappraisal of Peruvian Archaeology, ed. Bennett, Wendell C. (Memoirs of the Society for American Archaeology, No. 4)(Menasha, Wisconsin, 1948Google Scholar); Willey, G. R. and Phillips, P.,Methods and Theory in American Archaeology (Chicago, University Press, 1958Google Scholar). The training both in archaeology and in ethnology of the American anthropologists writing in these two volumes is of special importance for interpretation of their archaeological findings in social or cultural terms. There is some recent good work on Mesopotamia, which is, however, difficult to find in published form. It is worth looking at City Invincible, pp. 20–46, 61–123, and 269–328, also at Kramer, S. N., History Begins at Sumer(London, Thames and Hudson, 1958), pp. 46–84, 261–284Google Scholar, and idem, The Sumerians:Their History, Culture and Character (Chicago, University Press, 1963)Google Scholar. Valuable ideas for China areto be found in Owen Lattimore, Inner Asian Frontiers of China(American Geographical Society Research Series, No. 21) (London and New York, 1940)Google Scholar. Two theoretical works have contributed to understanding the changeover from early priestly rule, both on the seven firstoccasions in history and on later occasions. The first is Sorokin, Pitirim A., Social and Cultural Dynamics (New York,American Book Co.,1941), IV, pp. 138–142, 428–432Google Scholar; in Sorokin's terminology the changeover would be described as a transition from dominance of an “Ideational Socio-Cultural Supersystem”to dominance of an “Idealistic Socio-Cultural Supersystem”; the “supersystems” are modes of thinking, essentially of gaining and verifying knowledge - epistemological systems, that is to say. Sorokin is at pains to argue that the different modes of the two (and a third) period do not exclude other modes, but merely predominate over them in the period in question. The same predomination is intended in the terminology I use for the (three) periods, Age of Faith, Age of Reason and Age of Fulfilment; cf. above, n. 4. Sorokin's interpretation is based on events in the history of the Graeco-Roman and Western societies, but is applicable, in my judgment, to similar changes in all civilized societies at all times. The second work is Bergson, Henri, Les Deux Sources de la Morale et de la Religion (Paris, 1932Google Scholar); trans. Ashley Audra, R. and Brereton, Cloudesley, The Two Sources of Morality and Religion (New York, 1938Google Scholar); his conceptions of “dynamic religion” and “static religion” are suggestive although not very directly applicableand they make use of his concept of “élan vital”, which I find groundless. Comparison of events in the very early history of the primary civilized societies with events in their history at similar crucial stages in their later history and with events at such stages in the history of later societies has been found readily feasible and most useful. Nations are authenticated for Mesopotamia, Peru and Middle America. They can be regarded as reasonable inferences for Egypt and Crete, but there are no data worth the name about nations in the very early history of India orChina.

²¹ Cf. discussion by Steward, Theory of Culture Change, pp. 203–206.

²² Greek culture was later extended to the Roman Society (Rome-Latium-Etruria), whereafter I apply the term “Graeco-Roman”to the joint societv

²³ See Christensen, A., L'lran sous les Sassanides (Copenhagen, Levin … Munksgaard,1936), pp. 30–38, 141–178, esp. 35, 153–159Google Scholar.

²⁴ In fact we know far more about the former than the latter. What we do know, however, includes, most appropriately, the doctrine of theT'ien Ming, the “Decree of Heaven”, which combines the two. I follow the interpretative work of Creel, H. G. on this subject; see Studies in Early Chinese Culture, First Series (Baltimore, TheJohns Hopkins University Press, 1937), pp. 89–93Google Scholar and passim.

²⁵ The differences led Toynbee, in his first classification of civilized societies, t o postulate a “Hittite Society” alongside his “Babylonic”. In his second the Hittite has, rightly, been degradedto “satellite” status and alongside it appear, with queries (?) “Elamite”and “Urartian” (See above n. 2). It is true that divisions in the Mesopotamian Society in its second cycle appear very early. I should say that they arepolitical rather than religious or generally cultural. There is a good account of them from a special political point of view by Brundage, B. C. in Feudalism in History, ed. Coulborn, R. (Princeton,University Press, 1956), pp. 97–108Google Scholar. But the difficultyof characterizing them in any of those three ways is to be expected at any cyclic transition.

²⁶ For my opinion on the place of the caste system in Indian history - summarily stated, however; I have not published it fully - see Comparative Studies in Society and History, I (1958–1959), pp. 48–52.

²⁷ See discussion in Sansom, George, A History of Japan to 1334 (Stanford, University Press, 1958), pp. 69–74.Google Scholar

²⁸ Op. cit., XII, 187–209. His chief interest in the phenomenon appears to be that such empires were a means whereby farming peasantrieswere induced to carry on their backs a ruling clique which, after a varying periodof oneor a few centuries, became corrupt, incompetent and excessively expensive, whereuponit was overthrown and a new one, not yet degenerate, took its place.

²⁹ In the Byzantine Society, however, the stateof Bulgaria, peripheral nevertheless at the time of its origin, was perilously close to the nuclear territory of the society.

³⁰ What is judged is style within the meaning of the concept given by Kroeber, A. L. in Style and Civilizations (The Messenger Lectures for 1956) (Ithaca, Cornell University Press, 1957)Google Scholar.

³¹ The word needs no apology even though there are so few such cases in history that we cannot be quite certain where the boundary between normal and abnormal should be drawn.