I have to see a thing a thousand times before I see it once.
(Thomas Wolfe, You Can’t Go Home Again, 1942)
The purpose of this chapter is to provide: (1) background and contextual information for each of the 30 sites fTom which our observations on their prehistoric faunal assemblages were derived. (2) To describe and tabulate the perimortem taphonomic conditions of these assemblages using the 26 variables defined in Chapter 2. The descriptions and comparisons are drawn largely from the quantitative information about each variable provided in Tables A1.1-A1.26. These are located in Appendix 1, with assemblages listed by site number. Our first conference presentation of this research was at an international field meeting held at the Kurtak archaeological station directed by N. I. Drozdov (Turner et al. 2000). A brief English version appeared in Turner et al. (2001a).
Some additional untabulated information about specific pieces of bone is drawn from our data-collecting forms. For instance, we think that in some cases it is useful to describe several variables for a given piece, whereas in most cases this additional description does not seem worthwhile.
We have chosen to arrange the order of presentation of our assemblages alphabetically for the following reasons: (1) We initially thought that we would easily develop a dichotomous classification containing two distinctive sets of sites - archaeological and paleontological. This soon proved not to be possible due to our discovery of major carnivore presence in several of our “archaeological” sites. Some sites that have been previously identified as archaeological have turned out to possess a highly significant carnivore presence, as inferred fTom the types and amount ofperimortem damage to their faunal content. In fact, we doubt if any archaeological site described in this book has not been influenced to some extent by carnivores that also occupied these sites, regardless of whether they were cave or open-air types. (2) We considered other forms of presentation such as regional, ecological, temporal, etc., but abandoned each because of the information that we felt the reader would have to already possess in order to make the ordering useful instead of being a burden of additional background learning. Moreover, the senior author is well aware of how very little geographic, archaeological, and natural history knowledge non-Russians scholars possess about Siberia. (3) On the basis of many faunal studies of large Pleistocene animal remains, it is surprising how often the same set of animals turns up in northern Eurasian assemblages - mammoth, horse, bison, reindeer,
Wooly rhinoceros, cave hyena, cave lion, wolf, and several smaller species. While predator and human hunting selection, as well as preservability, may influence this community, nevertheless, these species were present in nearly all of the Pleistocene sites. This substantial faunal homogeneity suggests that there were no marked environmental differences except in the very far north and the very far south of Siberia. Some species are more common in one assemblage and less common in others. This may not reflect environmental differences as much as human and carnivore hunting preferences affected by seasonality, population size, sex ratio, mean age of human and carnivore groups, and various unfixed environmental considerations such as local forest fires, epidemics, severe winter storms, etc. (4) The names of several of these sites will be strange to most readers of this book, so the simplest possible ordering, namely alphabetical, was thought to be the most useful form of organization.
A few remarks need to be made about statistical analysis. At the beginning of our study we had assumed that multivariate statistical analyses would be ideal. We soon abandoned this viewpoint as it became clear that there were a host of data acquisition problems that would make such sophisticated analyses meaningless. These included much missing data; limited and even impossible identifications (such as species, skeletal element, etc); hard-to-standardize types of variation (damage form); subtle-to-significant differences in the ways the assemblages were originally collected, saved, partitioned, cataloged, and curated; and other problems of creating a database utilizing non-standardized, between-institution procedures and practices. In other words, we suspect that there is an undefinable but real amount of non-comparability between some of our assemblages, and even non-comparability within a specific assemblage, that makes a general plan of complex statistical analysis nonsensical and meaninglessly expensive. We have developed a very respectable database, but because we are so familiar with it, we are certain that it is not up to the task of supporting complex statistical analysis, and cannot meet many of the assumptions of multivariate analysis. Therefore, we limit ourselves to a few univariate statistical comparisons, and these only when we feel the data sets are adequately matched in quality, or where we need a statistical test to help decide if a probable real difference exists between two sets of not especially similar information.
World History
