Originalveröffentlichung in: Bulletin of the Museum of Far Eastern Antiquities 22, 1950, S. 83-126 THE KARASUK CULTURE AND ITS SOUTH-EASTERN AFFINITIES BY KARL JET TMAR In 1945, Karlgren established that some types of weapons and tools from An-yang, dating back to the Yin Dynasty, re-appear in Suei-yiian and Siberia.1) The similarities are so frequent and so striking that a connection between them cannot be denied. If a series of important metal types are related, then the bearers of these objects must also be in some way historically connected. The only question is: what kind of relation existed between »Yin-time China and the Nomad culture of Suei-yiian and Siberia*?2) We must remember, moreover, that Siberia3) and Suei-yiian differ from one another to such a degree4) that the existence of affinities between these regions seems to be a problem in itself. Presuming that there could not be any relation between South Siberia and An-yang without contact with the regions of the Great Wall — which lies just in the middle — we are confronted with two special problems: 1. What are the relations between the high-culture of An-yang and the northern confines of China? 2. What are the relations between these »Nomads» of the northern confines of China and the inhabitants of South Siberia? Karlgren thinks it is possible to establish a typological and chronological series originating in An-yang, passing on to Suei-yiian and then to the Minusinsk region. Certain types entered even into the Pontic Scythian culture. By means of this typological series Karlgren concludes that China was the stronger partner, that the general »cultural current)) flowed towards the North and Northwest and that the objects in the Ordos region and still more so those in Siberia are younger than those in An-yang. Karlgren's wording, however, is very cautious. He says that he only wanted »to show that certain features of the Nomad art can best be explained as being due to an influence from Yin-time China». But !) BMFEA 17, 1945, pp. 101—144. 2) It must be remembered that it is an open question whether there already existed »Nomad cultures* in the modern sense. 3) In this case Altai and Minusinsk. 4) Not only geographically — about 1000 km. of desert and mountains — but also prehistorically and probably in racial respects. %) BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES it is impossible to speak of single elements apart from the entire culture (therefore he must deal with the Hien-yiin and Hiin-yu). Loehr's treatise and Karlgren's answer demonstrate the problem in its whole extension.1) Karlgren's conclusion, however, is in opposition to the old tradition of European scholars who used to look upon the West, and consequently the Nomads of the Northwest, as the giver and upon China as the receiver. Karlgren has amply emphasized their carelessness in presuming what ought to be demonstrated.2) Nevertheless, it will still be a long time before these scholars are convinced that the »current of culture» may go in the very opposite direction. This fact has been shown by Loehr's somewhat temperamental answer to Karlgren's interpretation. But there already exist a number of more modern reports which were not available either to Karlgren or to his opponent.3) These treatises deal with the Karasuk culture of Minusinsk, the northern pillar of the far-reaching cultural relations. I now propose to give a short summary of these publications. Their strongest point is that they add the results of paleoanthropology to the archaeological material. Of course, in the first place they contribute to the solution of the second partial problem, namely, the relations between Minusinsk and Suei-yiian. (I shall therefore set aside the first one). But they are also essential for the whole problem, as they put a limit to speculation, especially in chronology. I. ARCHAEOLOGICAL MATERIAL. First we shall look at the treatises published in western languages. Up to the middle of the twenties Central and North European scholars were leading in the archaeology of South Siberia. Accordingly, the whole world was well informed. Gero von Merhart's book sums up all independent Western research, but unfortunately it also marks the end of it.4) In his publication Merhart already takes into consideration that complex which is to-day called Karasuk, but he is unable to fix any date.5) But soon afterwards the collaboration of the rest of the world came to an end. Subsequently Western research was entirely confined to compilations and informa­ tion.6) Notably two groups of scholars took note of Russian results; on the one 1) Loehr 1949 a and 1949 b. Karlgren 1949, p. 23 »Postscript». 2) Similar discussions by Andersson 1943, pp. 282—291, and Creel 1937, p. 220. 3) Loehr knew of them, but he is unfortunately mistaken in hoping to. find a confirmation of his theories there. 4) Merhart 1926. Bibliography up to that date. 5) Merhart 1926, pp. 37—39. »Kurgane mit Platteneinzelgrab». He is speaking of a special tribe which has no part in the general development. In other words, he anticipates modern result3. 6) I do not except my own article. 84 KARL JETTMAR: THE KARASUK CULTURE ETC. hand Field, Prostov, Golomshtok and others1) who wrote in American journals, and on the other hand Tallgren and his collaborators in the ESA.2) It was just at this time that Russian explorers succeeded in attaining decisive results, but as they were principally interested in local development and so-called »stadiality», their conclusions could not settle all questions for us. A change occurred in the last few years before World War II. Then, at last, the Russian scholars could not help taking migration and diffusion into considera­ tion. During this time, Tallgren's work was already approaching its end; only the American reports continued. But these are mostly lacking in illustration and criticism. Though they are precise in detail3), they do not give a lively, well-rounded picture as do the works of Tallgren. So we are placed in the position of knowing names and sites without being able to operate with them or to form any idea of them. Gaul's treatise is simply the exception to the rule. Unfortunately it is based only upon articles of the »antimigrationistic stage» of Russian research.4) After this short account, we shall now review the history of the exploration of the Karasuk culture as presented in eastern publications. Teploukhov laid the foundation, distinguished the Karasuk culture from the rest of the finds and marked its place, in relative chronology. It ranges between Andronovo and that complex which he called »Minusinsk Kurgan culture». This order has not met with any serious opposition since;5) it is, moreover, supported by numerous transitional forms, at the beginning and at the end.6) Teploukhov accordingly regarded it as a part of an inner development, being nevertheless aware of its south-eastern relationship. By comparing especially the forms of knives with the knife-coins of China of Chou time he confirmed the correctness of his relatively late dating(10th—8th c. B. C.) During the following years, intensive excavations were carried out in the Minusinsk Basin7), and it became known that in the Altai and in certain regions of Kazakhstan similar types were spreading8) almost at the same epoch. Griaznov, *) Golomsthok, 1933. H. Field and E. Prostov, treatises in Am. Anthropologist, Am. Journ. of Arch., Am. Review of the Sov. Union, Southwestern Journ. of Anthr. Besides these, American material by Ward 1948. English material is scanty but excellent (Childe, Russians publishing in »Antiquity»). 2) Cf. Tallgren 1937 a, 1937 b, 1938. Of the others, in particular Salmony must be mentioned here. 3) Sometimes not even that. Cf. Henry Field and Kathleen Price: »Review of Soviet Archaeology 1919—1945 in Historic Perspective* p. 221, where they deal with our subject. But who would guess that »Din-Din» is a mistake for »Din-lin = Ting-ling»? That »In» means »Yin»? Why should we need two authors, when neither of them recognizes well-known Chinese names which have been phonetically translated into Cyrillian letters? 4) Gaul 1943. I am indebted to Prof, von Heine-Geldern for this treatise, and also for others which were not accessible in Vienna. s) Tallgren was sceptical, but later on he accepted name and dating. Cf. 1928 and 1938, pp. 121—128. 8) Teploukhov 1926, 1927, 1929 a, and 1929 b. 7) Kiselev 1949, p. 64. Survey. 8) Griaznov 1927, 1930 a, 1930 b, Kartsov 1929, Rykov 1935. Levashova 1939. Survey by Kiselev 1949, pp. 87—92. 85 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES a brilliant excavator, consequently spoke of a »Karasuk culture of the Altai». A long series of forms transitional to the groups of western civilization could be estab­ lished, but burial sites with typical Karasuk inventory were found as far west as Karaganda. Besides this, it was found that various Karasuk objects had spread still farther into the forests of the northwest. Griaznov likewise tried to solve the problem by typological research1). An analysis of his work shows that in 1941 he still believed in the chronology as well as in the derivations of Teploukhov. In this article he gives still further reasons in support of the assertions of Teploukhov. On the other hand, he is convinced, for example, that no foreign influences can be established in the whole history of socketed celts. Kiselev followed another path2). He, too, made extensive excavations. He may be called the official successor of Teploukhov. For a long time he made no mention of any foreign influences, but he began to bring a completely new order into the »Kurgan culture». This means a fundamental difference between the conception of Teploukhov and his own, for his basis is not the development of the rite of interment but the typology of tools and weapons. He asserts neither more nor less than that collective tombs and single graves might be coeval. The whole difference lies in the social order3). By his new typology that curious empty interval between the Animal Style elements of Karasuk and those of the »Kurgan II»4) is eliminated. This new grouping found its expression also in a new name. Kiselev called the period of Minusinsk Kurgans »Tagar»5) (or »Tagarsk», after Lake Tagar and the isle of the same name). Besides this, he points out that, in the West, Late Andronovo forms with Seima decor continue almost to the beginning of the Scythian time. That shows that the presence of such forms in the Minusinsk Basin does not mean that they were already spreading in Karasuk time. They might have been accepted at the beginning of Tagar time, when the contact with the West was reinforced. This is of vital importance for deciding the question of how strong the local element in Karasuk was. By this new sequence many of those types which were thought to be local in Karasuk must now be looked upon as having come from the West at a later period.6) But if the characteristic features of Karasuk were neither local nor related to the West, where else did they come from? Then it was that Kiselev finally began to take into consideration eastern affinities. The correctness of Kiselev's opinion has been generally acknowledged in the 1) Griaznov 1941. Cf. Engl, resume pp. 270—271. 2) Kiselev 1929 a, 1929 b, 1933 a, 1937 a, 1937 b, 1938, 1949, pp. 62—108. 3) E. g. Kiselev 1949, pp. 152—154. 4) In Teploukhov's scheme »IV», reproduced by Golomshtok 1930, p. 320, Salmony 1935, pi. Ill, Gaul 1943, p. 172, cf. Gaul 1943, pp. 173—174. His deliberations on this point have lost their importance. 6) Gaul (1943 p. 166 foot-note) was mistaken in writing about an »(ill-defined) later 'Tagar phase' of the Kara Suk complex*. He had not read Kiselev 1929 a and 1933 a. e) Kiselev 1949, pp. 110—128. 86 KARL JETTMAR: THE KARASUK CULTURE ETC. meantime.1) It seems justified not only because it simplifies the Tagar complex, but because it is confirmed on two later occasions. Firstly, there was the Chinese exhibition in Moscow in 1940, where the Russian scholars could observe finds from An-yang.2) This extended the horizon in an unexpected direction. Secondly, Kiselev undertook an oarchaeological journey» to Northern Mongolia sponsored by the Academy of Sciences.3) On this occasion he discovered so many identical forms in the museums that he could not but revise the old standpoints. Post-war excavations in Minusinsk confirmed that animal sculptures already belonged to Karasuk and hinted once more at foreign relations.4) Even now a good deal is not yet clear, but these reports have led up to a new idea, which I should like to present here. I wish to set aside all regions of obviously secondary character, e. g. the Altai, which do not give us any information about the age, derivation and eastern affinities of the Karasuk complex. Our chief attention will be drawn to the Minusinsk Basin, where we have plenty of well-examined material, and where we find the types so concentrated that it seems to be the very heart of this culture in Siberia. A. Burial Customs. Dwelling-sites are very rarely found intact, so that the whole work of exploration must begin with the graves. Stray finds have to be classified accordingly. That is why our description starts with the Karasuk cemeteries, of which nearly two dozen are known in the Minusinsk Basin. Kiselev's survey is based on the following burial sites:5) Askyz, excavated by I. P. Kuznetsov-Krasnoiarskii8) Mochov on the Uibat, excavated 1900 by Levashova Okunev, excavated 1926 and 1927 by Teploukhov near the Ninia river, excavated 1908(?) by Kuznetsov-Krasnoiarskii One near the Beia river, excavated 1936 by Kiselev7) Lugavskoie, excavated 1924 by Tugarinov8) Krivinskoie, excavated 1928 by Kiselev9) One near Lake Tagar, excavated 1894 by Adrianov10) !) Chernetsov, 1947. 2) Kiselev 1949, p. 85. 3) Kiselev 1947. 4) Lipskii 1949. 5) Kiselev 1949, pp. 62—67. 6) publ. 1889. ') publ. 1937 b. 8) publ. by Teploukhov 1927. ") publ. 1929 b. 10) publ. 1920, 1924, and OAK for 1894. 87 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES Bystraia, excavated 1929, 1930, 1931 by Levashova and Kiselev1) Podkuninskaia, excavated by Levashova 3 burial sites near Tes, among these Ilinskaia gora and excavated by Kiselev2) Georgievskaia gora 1929, 1931, and 1932 Ust Tes Dzhesoss, excavated by Adrianov3) Syda Ust Syda excavated by Kiselev4) Ust IErba Batenf^ ) excavate(^ Teploukhov5) Orak, excavated by Sosnovskii, 1927. There is one cemetery to be added: Near the town of Abakan, excavated 1946 by Lipskii.6) Gaul has only reported on the finds of Kara-Suk and Bateni, Askyz, Lugavskoie and of the Lake Tagar. He cites a publication by Kiselev7), but it is not contained in his material. One burial site often consists of more than a hundred single graves, but only 290 graves altogether have been analysed in detail. The result, however, is still poorer as most graves were robbed at an early period. We can easily imagine how early robbing started, bearing in mind that there were practically no precious metals in the graves. The robbers were digging for the bronzes. But the inventory of bronze is small enough even in graves that have not been robbed, especially compared with that of the following (Tagar) period. Some centuries later it would not have been worth while digging for these objects. In the Tagar period, moreover, we see that the robbers knew very well where the richly equipped graves of men were situated, and they knew them, too, in Karasuk. We are thus led to believe that contemporaries robbed them. However, it seems rather improbable that all irregularities in the position of the skeletons should have been caused by robbers. Secondary interment is quite possible.8) In any case, comparative statistical examinations cannot be made. The graves of Karasuk are marked by a fence of rectangular slabs placed on the surface of the soil and put into the ground in a vertical position. In the x) publ. by Kiselev 1937 a. 2) publ. 1929 b and 1937 a. 3) publ. 1902 and 1924 and OAK for 1894. 4) publ. 1937 a. s) publ. 1927. «) publ. 1949. ') 1937 a. 8) Kiselev 1949, p. 64. 88 KARL JETTMAR: THE KARASUK CULTURE ETC. northwest of Minusinsk, however, these stone fences are often erected in a circle, which is reminiscent of older forms from Afanasievo and Andronovo. This might mean that the influence of new elements was stronger in the southeastern region. Smaller rectangles often occur round a larger one. In the centre of these fences, under a low mound, there lies a generally trapezoid pit clad and roofed with slabs of Devonian sandstone. Sometimes there is no other fence, but even then it is easy enough to recognize the graves. They lie so near to the surface that the plates of the stone cist stand out conspicuously. The skeleton lies extended on its back or slightly turned to the left side, with its head at the broader basis of the trapezoid. Knowing this for certain, it is possible for us to form some idea of the orienta­ tion of the dead, even if we cannot derive anything from the bones themselves. The head is usually directed towards the east or east-northeast. Besides extended skeletons we also find flexed burials, i. e. the known form of the older graves of the same region. As yet we cannot find anything revolutionary* in the rite of burial; neither is there anything surprising about the number of persons who were buried in one and the same grave.1) Already before that time, single graves were predominant. Only occasionally was a child buried in the same stone chest. Very seldom a man and a woman are buried together. We sometimes find the graves of a male and a female lying symmetrically and surrounded by the same fence.2) In any case we have no reason to think of a change in the social order, e. g. a change from matriarchy to patriarchy, as Russian scholars are inclined to conjecture. I do not intend to give here more than a short outline, since Merhart, Teploukhov and others3) have already sufficiently described the graves. I shall complete the picture on two points only: Firstly, an approximate inner chronology of the Karasuk cemeteries can be fixed, though we have to consider local differences, e. g., in the north the cultural events often occur somewhat later; but it is certain that Ulus Orak, the »IArki» (near Bateni), and especially Ust IErba are the oldest sites, like Krasnii IAr in the Altai. The latest are Bystraia, Ilinskaia gora and Askys. The rest range between these extremes.4) Secondly, during the latest excavations a very peculiar thing was noted. In four cases, outside of the fence, in the north or at the north-eastern angle, a small pit has been found in the ground. It is laid out with three stone slabs which form an equilateral or isosceles triangle when viewed from above. It measures 12—20 cm. in the length, 12—15 cm. in height, the stones being 2 cm. thick. These stone cases were empty. What were these puzzling chests meant for? In the course of Kuznetsov-Krasnoiarskii's excavations small cubical holes, 1) Nor does distribution of the burial sites differ greatly from that of Andronovo. 2) Kiselev 1949, pp. 65—67. 3) Merhart 1924 and 1925; Merhart 1926, pp. 37—39. 4) Kiselev 1949, pp. 75—86. 89 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES each of 50 cm. in length on each side, were found in the ground which did not contain anything but bronzes and pottery.1) Kiselev gives us a further hint. He states that the graves of children, which are, moreover, unusually numerous, are mostly situated near the north-eastern angle of the fence. In Tashtyk, a child's grave was found which was curbed in with timbers forming a triangle. A further, ethnological, affinity is reported of some small tribes of the Amur region, who bury miscarriages and afterbirths in small triangular stone cases like these even at the present day.2) The inventory of the graves, which we shall now deal with, can be understood to be as follows: the dead were provided with clothes and food which they would need »on the road». In no case was any equipment for military or civil purposes in the other world given to them. This is proved by one characteristic feature: the few knives found in the graves are not near the dead bodies, but they are all found near the pots and with bones of animals.3) They obviously served as »carving- knives», not as weapons. The dead were provided not only with meat, since we found bones of animals, but also with pulpy or liquid food in pots. B. Pottery. Vessels constitute the most reliable inventory. A grave may often contain a considerable number of them. It is obvious that the form of the vessels differs completely from the Andronovo types.4) (PI. I; 7—11). The flat bottom has disappeared. All rela­ tionship with the culture of the »timber graves» of the Pontus are suddenly severed. Wherever the typical form of the Karasuk grave is met with, we find vessels with rounded bottoms. Hardly any exceptions have been seen. The shape of the vessels is mostly globular, or, sometimes, irregularly napiform with straight medium- high collar. Sometimes the collar is slightly flaring, as we know it from the Andronovo vessels. Occasionally the vessel widens elliptically at the sides, the bottom is flattened, but there is no edge between side and bottom as in the vessels found in Andronovo. Other vessels are ovally elongated towards the bottom, recalling the pottery that is usually assigned to »ancient times», that is to say to Afanasievo. The few exceptions can be reduced to two groups. On the one hand, vessels with flat bottom are found in graves which must be assumed to be the oldest, as is evidenced by the decor and technique as well as by the further inventory of metals and the position of the skeleton. They are most closely related to Andronovo. On the other hand, the flat bottom is known to exist on vessels that are judged to be the latest, next to Tagar (for reasons mentioned above), The ') Cf. Gaul, 1943, p. 164. 2) Lipskii 1949, pp. 79—81. 3) Griaznov 1941, pp. 237—239. 4) Kiselev 1949, pp. 78—82. Illustr. in: Teploukhov 1927, Kiselev 1929 b, Kiselev 1937 a. 90 KARL JETTMAR: THE KARASUK CULTURE ETC. Tagar time, indeed, brings a complete revival of the flat-bottomed vessels. Thus the »exceptions» really confirm the observation that the round bottom is a specific feature of Karasuk. But it is confined to the Minusinsk Basin, for the whole region of the Altai keeps its local flat-bottomed pottery besides the metal inventory which indicates Karasuk types. The south-eastern region (Suei-yiian) does not yet show anything similar to this development. But there is another new feature in Karasuk which we can find also outside the Minusinsk Basin, namely in the contemporaneous Late Andronovo of the West. The collar-base is set off in a very distinct manner, sometimes sharply accentuated by ornaments. This probably means a parallel development in both regions, based on the common tradition. As to the ornamentation, we have, on the one hand, vessels which show cog-like impressions in a comparatively primitive manner. Sometimes the surface seems to have been treated simply with a tuft of grass. One of the usual archaic patterns is the »pine tree» or »herring-bone» design. All these particulars are very well known from the earlier history of the upper Yenisei. We know them especially from the Afanasievo graves. When we discover, moreover, that this kind of ornamentation appears together with finds reminding us of the past in form also, we are tempted to speak directly of an Afanasievo Renaissance in Karasuk. There are other vessels, besides these, with triangles, rhombi and transverse bands marked by a dentated stamp, proving clearly the continuation of Andronovo features. Not all designs belonging to the same group are already met with in Andronovo graves of Minusinsk. An entire series of these designs is found exclusively outside this region, in the Late Andronovo of the West,1) for example, the distinctly marked collar-base, or even a zigzag line circling the pot in the equatorial height. Sometimes two zigzag lines are arranged one above the other, forming a broad band, and in other cases all these patterns are combined to form an intricate carpet-pattern. Here we see in a characteristic way that we stand at the eastern frontier of this kind of geometrical ornamentation.2) Besides pottery, it is also to be found on bronze objects. Geometrical ornamentation seems to be a specific feature of the western group of the Siberian Bronze Age. There is one point more, really the most elementary one, the technical execution of the vessels. Here, the inheritance of the past is not so strong as in the ornamentation. We are rather reminded of the change in the form of pottery. The method of manufacture is quite a new one: the vessels are hand-made and formed of very gritty clay. The outside is grey-brown, but the inside is dark with a bluish colour. The vessels are decidedly thin-walled, and their quality is superior to all preceding forms. Perhaps the sides were flattened with a hammer. At any rate they were diligently smoothened when still wet. It has repeatedly been !) Griaznov 1927. 2) In the Baikal cultures the ornamentation is entirely different. 91 BULLETIN OF THE MUSEUM OF FAR EASTERN ANTIQUITIES suggested that the outer layer was a glazing made of some special compound. This would at least explain that some pots show a tendency to exfoliate on the outside. In his »Mongolian Journey», Kiselev gives a hint that similar pottery can be found in the collections of Northern Mongolia. Nevertheless he explains in his later great treatise that he could not yet speak of any south-eastern affinities.1) As for the technique, there are two exceptional groups. Some vessels with their thick sides remind us of the past, some others found in late graves show some affinity with the Tagar period. Consequently, we may say that in form and technique we meet characteristics which indicate a breach with the past. What may have been the reason? It would be tempting to deduce from the return to the round-bottomed vessels a growing mobility of the population, i. e. nomad tendency. We are well aware, on the other hand, that the appearance of flat-bottomed pottery in Andronovo after the pointed pottery of Afanasievo had meant a more settled mode of living (with fixed fire-places?). This is now proved by the discovery of the remains of dwelling-sites belonging to Late Andronovo in the West. At any rate, Kiselev speaks of a fundamental change in the mode of living. In spite of this, we are not allowed to overestimate the force of this argument. But we shall see that it is at least an indication — the first indication of this kind. Besides this, in one characteristic element of the form, namely in the marked collar-base, and in most of the ornamentation, we see the steadily growing develop­ ment of the traditional Andronovo style. The most surprising feature is what I called the »Afanasievo Renaissance*). In the Minusinsk Basin the end of Afanasievo is usually dated 1700 B. C. That means that Afanasievo had already passed several centuries before. What are the reasons for this revival? I should like in this connection to refer to an article by N. M. Komarova.2) The author points out that a group of Afanasievo elements can also be isolated in Andronovo. She believes she can thus give them a chronological order. But these elements can be found in the next period as well. Consequently, we have to consider an inner ethnical order. Certain tribes of the Minusinsk region, possibly the descen­ dants of the Afanasievo people, have always kept to these forms of ornamentation, while the immigrated Andronovo groups preferred other designs. Komarova's only argument against this is that the Minusinsk Basin would have been too small for such a subdivision. Surely this is not correct. Probably the Karasuk time witnessed a strengthening of the former population of the region. Or maybe this part of the population had lived outside the Minusinsk Basin in Northern Mongolia, where we also find Afanasievo features, and came back in a larger immigration. But we shall have to deal with this problem later on. 1) Cf. Kiselev 1947, p. 359 and Kiselev 1949, p. 82. 2) Komarova 1947, pp. 57—59. 92