Alba Regia. Annales Musei Stephani Regis. – Alba Regia. Az István Király Múzeum Évkönyve. 23. 1984-1985 – Szent István Király Múzeum közleményei: C sorozat (1987)

Tanulmányok – Abhandlungen - Choyke, A. M. – Bartosiewicz László: Animal exploitation and its relationship the bone deposition at Lovasberény-Mihályvár. p. 7–18.

Bone Wo г к i n g When the bone material from Lovasberény-Mihályvár was originally sorted emphasis was not placed on bone working. Information was only gathered on the proportion of worked bone per species. Nothing is said about bone part here. Bone and tool types not recognized by the archaeologist and separated out at the beginning were not recorded in detail. There are thus 27 worked cattle bone, most of which are likely to be what we currently term large mammal ribs. At this level of fragmentation it is not always possible to distinguish cattle rib from that of horse or red deer. Five tools are made from the bone of sheep or goat and 5 from pig, 2 of which include first phalanges which were polished. There are 24 worked remains from red deer. Most of these are probably cut off antler tine, very few of which were actually used as tools. The same holds true for the 3 worked pieces of roe deer bone. There are also 21 pieces of worked bone on unidentifiable fragments. This makes a total of 85 pieces which are worked or at least manipulated some way. Later we returned to study the 43 tools which had been recognized by the excavator. There are no identified cattle, although most probably the 14 large rib fragments, worked as scrapers-smoothers on the corpus costae, belonged to this species. There are 5 worked sheep/goat bones of which 4 are metapodial fragments. One of these is a double pointed tool found in an oven with stone weights and fish bone. It has been suggested that this tool form is that of a fishing gorge passed between the gills of a small bait fish. The second sheep metacarpal is some kind of a small, broad perforator pointed at the tip, broadening sharply and flat in cross section. It was twisted in a hard material. The latter two metapodial tools are not made from casually broken diaphysis splinters, but from carefully split bone. As noted previously, bone from domestic or wild swine was rarely worked. There are 2 pieces at this site, a tibia diaphysis Class 1 scraper and a fibula carefully worked into a long needle form with flint and sandstone. There are 14 pieces of deer bone which show signs of working. Of these 6 are unused tines which were cut off at the beam, probably during the making of implements such as the 5 hafted burr and beam tools. These are all characterized by flat and hollowed butt ends, a hole drilled through the side and a smoothly cut off eye tine by the burr. It seems likely that the hollowed out end was fitted with a stone or metal blade of some sort. There is one tine which has scoring at one end grooving lower down and a hollowed out base. There are two finely carved antler barbs or projectile points. Only one tool is made from deer bone. This type, a mandible burnisher with high wear and polish on both sides of the diastema, is more commonly made on sheep/goat mandibles and found more often at eastern sites. There is 1 roe deer antler prong which may or may not have artificial sharpening at the tip. The single horse metapodial spur has been worked into a blunt ended needle form with a hole drilled dorso-ventrally through the distal end. Another blunt needle was made on a small rib. There are 3 splinters from small mammals worked into dullbulbous tip tools. Aside from the 14 large rib scrapers, there is a scraper from a roughly worked and barely used large splinter which is unidentifiable as to bone or species. Numerical Seriation As was demonstrated previously, the extensive excavations uncovered a relatively great number of features with rich faunal material in them ensuring a desirable proportion between variables and observation entities (cases), which provided sufficient data for multivariate evaluation. The statistical procedure chosen was not particularly developed for archaeological seriation. Instead, results of a factor analysis, as available in the BMD package (Frane — Jennrich

—Sampson 1981), were adapted to provide differing aspects for potential seriation. Univariate statistical calculations introducing this method have been successfully used in the identification of bone deposition patterns for the purposes of a taphonomic study (Bartosiewicz 1984) as well as associated coefficients of single correlation. In a wider sense, factor analysis itself, in combination with other multivariate techniques offered by the BMD software was also useful in outlining horizontal and vertical distribution of a variety of artifacts (Akazawa

—Hanihara 1983). Remaining within the context of zooarchaeological applications another example is worthy of mention. When animal remains are studied in terms of the representation of skeletal elements this method of data matrix reduction was efficient in outlining butchering patterns (BlNFORD 1981). While this last approach is more functional, the species composition of the material was chosen as input data for the purposes of this study, because it was thought to be indicative of changes in consumption activities and possibly general changes in subsistence practices over time. In this work, even the rudimentary distinction between skeletal regions would have resulted in a subdivision of the raw data, increasing the number of variables thus introducing a great proportion of zeroes into the data matrix, in addition to the distorting effect of correlated skeletal elements within individual animals. These two influences would have increased the incidence of combinations of ambiguous interpretations unnecesarily. Table 2 shows the univariate statistics obtained for the greater part of the material recovered from 82 features. While the mean values reflect the proportions of the faunal list of the total sample, coefficients of variation suggest a marked diversity of distribution between features. This is particularly characteristic of the bones from large game, red deer and wild pig. Small domestic ungulates seem to be most evenly spread out over the site. Although one part of the deviations is undoubtedly caused by random deposition, a basic patterned structure emerges in the single correlation matrix in Table 3. The high positive correlations between caprines and domestic pig as well as cattle and horse are of particular interest. Negative correlations do not occur in the matrix. This shows the lack of extreme complementarity between species in terms of meat consumption (Bartosiewicz—

Sáfár 1983). I