Divje babe I - an attempt to apply statistical analysis to the mass animal remains in the palaeolithic archaeology. II. Broken-open long marrow bones of cave bear

Abstract

On the basis of the standardized sample containing all determinable long marrow bones of cave bear (T. 1), and consisting of 16 archaeologically fertile and 7 archaeologically sterile arbitrarily-chosen staratified units (cf. I. Turk, J. Dirjec, Fig. 1), the hypothesis has been tested that the antropogenic breaking-open of the long marrow bones of cave bear had once occured, presumably for the consumption or extraction of marrow, on the newly discovered palaeolithic cave site at Divje babe I, near Cerkno (Slovenia, Yugoslavia). Previously, the marrow of the recent brown bear had been analysed for the purpose of determining the marrow's nutritiousness and, above all, the content values of some of the essential substances. These had been found to be present in substantially larger quantities than e.g. in the analysed marrow of red deer (Tt. 2, 3). The bones from the sample have been classified as shown in T. 1 and Fig. 2. The classification has followed major quantitative differences between the marrow in the bones of adult (group I) and juvenile specimens (group II) (R.J. Blumenschine, 1986, 37, 132; D.C. Thomas. E. Broughton, 1978). These differences are also found to be observed by the now living predators and scavengers in the consumption of marrow in terms of a definite preference for the bones of adult animals (R.J. Blumenschine, 1986, 58). Furthermore, the classification has been made in adherence to the significant differences in the amount of marrow between the subgroup humerus-femur-tibia (Aa) and radius-ulna-fibula (Bb). These differences have been expected to produce some global differences between the two subgroups if the extraction of marrow had been confirmed. The distribution of the two subgroups into bones integral (A,B) and bones broken open (a,b) is irrelevant in the present case. The only thing that is relevant here is the number of fragments which will increase in direct proportion to the degree of fragmentation of bones. Since an increase in fragmentation tends to affect the proportion of determinable to indeterminable fragments, the decision has been made to additionally keep account of the proportions between the archaeologically fertile and archaeologically sterile units. This latter division has been made to conform to any of the possible global differences between the units expected to be detected if the palaeolithic visitors to the cave had indeed taken to extracting marrow. Freshly accumulated long marrow bones of adult cave bear subjects could only have been crushed by hyenas, bears themselves, and especially by ancient men. Since no osteoodontological or other remains of cave hyena, which are only rarely traced in the palaeolithic sites of Slovenia m general, had been unearthed at the site treated here, the only two biogenic agents left for consideration are man and cave bear. Yet the high degree of fragmentation so typical of the Divje babe I site can never be the result of mere intensive decomposition of bones caused by natural agencies because the bones in general are in a very good state of preservation, with most fragments displaying rounded adges of fractures that had mostly originated in the time prior to the final integration of the fragments into the sediment. This was the reason why all of the tests had been prepared so as to make it possible for any global differences that could be associated with the activities of man to become visible, as well as to bring down to the lowest possible degree various disturbances caused by other living and non-living agencies that should be taken into consideration as well. In testing the hypothesis concerning the anthropogenic causes for the fragmentary state of the long marrow bones of cave bear from Divje babe I, only non-parametric statistics has been used, since it is known to impose fewer tasks on the person analysing the data, although it usually yields rather rough results. The results of Friedman's Fc test for dependent samples, p=0. 05 (H. Biining, G. Trenkler, 1978, 217 ff.), which had been used to test the global differences between the (a) and (b) subgroups in the units exhibiting cultural remains and features, have proved different taphonomical features for the humerus-femur-tibia subgroup in comparsion with those of the radius-ulna-fibula subgroup, both in adult and juvenile specimens (Fig. 2). The homogeneous character of the bones containing larger amounts of marrow (humerus-femur-tibia) is above all significant with the adult specimens, since it has been testified to both between the bones and between the units. The results of the test are in agreement with the hypothesis of the extraction of marrow, however they fail to yield any further details on the fact. This leaves us room enough to make different alternative explanations, which, however, have nothing to do with the hypothesis proposed above. The results of Wilcoxon's one-sided test for dependent samples and the exact distribution, p=0.05 (H. Biining, G. Trenkler, 1978, 186 ff.), which had been used to test the differences between the (a) and (b) subgroups in the units exhibiting cultural remains and features, have shown significantly lower values in adult specimens' bones that contained larger amounts of marrow, whereas the analougous bones of juvenile specimens have displayed just the opposite results (Fig. 3). One way of explaining this fact is to seek the cause in the increased fragmentary state of the bones from certain subgroups within the two groups, and the minimized determinabilty of osteological fragments related to it. For this reason the proportion of determinable to indeterminable fragments in the 16 archaeologically fertile and 7 archaeologically sterile units have been put to Wilcoxon's one-sided test for independent samples, p=0.05, with the result of establishing that the proportions in the archaeologically fertile units are significantly higher than those in the archaeologically sterile units (cf. p. 36), which is direct opposition to the conjecture of a higher amount of indeterminable fragments in the archaeologically fertile units owing to the more intensive process of fragmentation. Another disturbing point is the complementarity of the results in both subgroups of long marrow bones of both the adult and juvenile specimens. One of the possible alternative interpretation of the given situation is the most direct one, i.e. that the fragments represent the natural state of affaires, which, of course, is in direct opposition to the hypothesis as it has been proposed above. The results of Wilcoxon's one-sided test for independent samples and the exact distribution, p=0.05 (H. Biining, G. Trenkler, 1978, 145 ff.), between the archaeologically fertile and archaelogically sterile units, have shown that the former will invariably yield more remains (T. 4). In addition to the absolute ones, a measure of qualitative aberrations between the two groups and further subgroups of the long bones would naturally be expected to be found in the course of testing the above hypothesis. Some inexplicable qualitative aberrations that had already been established in the analysis of skeletal elements (I. Turk, J. Dirjec, 1989) had completely blurred major absolute quantitative differences between the two groups of units. That is why the proportion of the humerus-femur-tibia to radius-ulna-fibula subgroups have additionally been tested both in adults and cubs. Wilcoxon's one-sided test for independent samples, p=0.05, has shown these proportions to be significantly higher in the archaeologically fertile units both in the juvenile and adult specimens, which means that, very likely, there are no relevant differences between the two groups. Both the ponderings of the naive intuitive and the formal quantitative techniques would suggest the following conclusions: The analysed remains of cave bear had been accumulating at the entrance to the cave most probably because of the natural dying-off subjects and the autopodal accumulation of their skeletal remains. The agency of early men in this process must have been negligible, if at all present. The fact that ancient men would break open the bones of cave bear, and even some of the human long marrow bones which resembled them (in their morphological and structural features) (P. Villa et al., 1986), has indeed been proven by a number of finds from later archaeological periods. The analyses of recent brown bear's marrow have confirmed the edibility and nutritional value of the marrow, and even the presence of some essential substances. However, the results of the executed statistical analysis do not admit of the hypothesis of the anthropogenic breaking-open of the long marrow bones of cave bear from the Divje babe I site on the basis of the criteria proposed above. Therefore, the conclusion can be drawn that the broken-open limb bones of cave bear reveal no solid evidence of any activities of the palaeolithic visitors to the Divje babe I cave. There are several alternative explanations for the fragmentary state of the bones with predatory animals as the principal agencies, especially the cave bear himself. There is also a very simple, and also very probable alternative: All the analysed remains represent the natural state of affairs, in which the bones that have been shattered by whatever reason to a higher degree always outnumber those which have been less shattered. Logically, such an explanation can only be accepted within the two groups (adult-juvenile) but not between them.