«Item type text; Dissertation-Reproduction (electronic) Authors Munro, Natalie Dawn Publisher The University of Arizona. Rights Copyright © is held ...»
Table 5.21: Fragmentation indices for hare and gazelle long bone ends from the Natufian layer at Hayonim Cave.
Cooking In the Paleolithic periods animals were most often cooked by roasting. Roasting is defined here as cooking without containers, and includes heating meat directly above an open flame or on hot coals, either by direct or indirect exposure (i.e., through the
Upper Paleolithic, yet no firm evidence for this exists in southwest Asia until the Neolithic ceramic period. The most obvious archaeological indicator of cooking is burning, though bone often does not bum even when roasted or boiled for prolonged periods (e.g., Kent 1993).
Bones burned as a result of cooking may be confounded by the effects of burning by natural occurrences, secondary human activities, or disposal in hearths. These processes must be separated as much as possible before interpretations of human cooking can be made. Indirect exposure to fire results fi-om the protection of bone by an intervening substance such as sediment or meat, the first being intentional and the second not. Burning experiments by Stiner et al. (1995) indicate that bone buried within 6 cm of a hot coal bed by a bonfire on the ground surface can be charred. The intensity with which bone is burned, and whether or not it was in direct contact with the flame can often be partly evaluated from its color in the absence of diageneic processes that may cause discoloration (Shipman et al. 1984; Stiner et al. 1995). When burned, bone first turns black as the organic content carbonizes, then gray to bluish-gray to white with progressively more heating as oxidization of the carbonized organics occurs (Brain 1981).
Calcination occurs only if bone comes in direct contact with the heat source. Bone that is protected from the heat source by a few centimeters of insulation may turn a grayish color, though it never reaches the full calcination stage. Unfortunately, several other taphonomic processes can discolor bones and frequently mimic burning. At Hayonim
common. Burning was recorded in this study only when it was certain, thus burning counts are likely underrepresented, particularly in the Hilazon Tachtit assemblage.
Much information on cooking and burning is anecdotal. To be more useful this information needs to be integrated for systematic analysis. The first step is to compare fi'equencies of biuning at the taxon, element, and portion levels to determine if burning is randomly distributed across the prey anatomy (more likely natural or secondary processes) or biased toward certain parts (more likely intentional). Non-random patterns can then be examined more closely for regularities in the coloration and intensity of burning to identify cooking techniques by prey species.
Results for Cooking At Hayonim Cave and Hilazon Tachtit there is significant variation in the proportion of burned specimens by taxon (Table 5.22), though at least 14.0% of all prey taxa from Hayonim Cave are burned. These proportions are high in comparison to ethnographically derived figures fi*om cooked bone, regardless of whether the animals were roasted or boiled (Kent 1993). They are thus interpreted as partly, if not largely the result of secondary burning following deposition. This point is underscored by data in Table 5.23, which lists the spatial distribution of burning in Hayonim Cave. There is much variation in the frequency of burned bone among Loci, Areas, and Graves, regardless of the relative abundance of taxa present, confirming that spatial context is a primary determinant of burning frequency. Burning frequencies vary across space due to differentiation in the use histories of each area. For example, a large space along the
calcined bone and ash. The fill is a light gray color, a stark contrast to the surrounding brown sediments. This is an extreme example and, with the exception of hearth features, differential burning in sediments is rarely so obvious. However, this example does illustrate the complicated stratigraphic and variable spatial history of the Hayonim Cave deposits. Repeated use of fire, as well as small scale displacement of sediments undoubtedly exerted secondary damaging effects on bones (including intrusive species) and other materials previously deposited in the fill, influencing the firequency of burning on bones in the immediate vicinity.
High rates of secondary burning likely reflect the intensity of site and/or feature use, since the longer a site is occupied, the more likely bones will be burned through secondary processes. The high rate of burning thus suggests regular reuse of Hayonim Cave in comparison to Hilazon Tachtit, the latter of which has much lower rates of burning across the board (5.9%). This is most likely a combined effect of the conservative approach used to identify buming at Hilazon, but also an overall lower firequency of buming.
Table 5.23: Proportion of burned bone specimens by spatial region (Loci, Areas and Graves) from the Natufian layer at Hayonim Cave.
Despite the strong influence of localized secondary burning events in Hayonim Cave, it is clear that the taxonomic differences in burning frequencies shown in Table
5.22 crosscut spatial boundaries, and are not determined only by localized depositional histories or changes in the relative abundance of species from one area to another. This point is clarified in Table 5.24, which compares variation in the frequency of burned hares and tortoises across space at Hayonim Cave. Although the rates of burning for the two species vary with spatial location, differences in burning frequencies between the two species are consistent throughout, with hares burned more often than tortoises regardless of spatial location.
Other differences in burning frequency between taxa shown in Table 5.22 hold
(23.1%) are burned in significantly higher frequencies than are tortoises (14.6%) and gazelles (18.7%). Distinctive treatment of prey species during processing, cooking, or disposal therefore clearly had a secondary yet significant impact on the taxonomic distribution of burning at Hayonim Cave.
Table 5.24; Proportion of burned hare and tortoise elements by spatial region within Hayonim Cave.
Note that though the frequency of burning for the two species rise and fall in tandem with changes in space. In all cases, hares are burned in higher frequencies that tortoises.
Separating out differences in the treatment of prey proves problematic. Little can be said of cooking techniques for gazelle, since burning is evenly distributed across gazelle elements and bone portions. The high frequency of burned carnivore bones is provocative, but closer investigation of their spatial and elemental distribution reveals that the inflated frequency of burning is largely determined by the spatial location of fox
area along the east side of the cave wall. All bones from this area were intensely burned and covered in a light gray matrix of ash.
The burning damage on hare bones is unusual not only for its frequency, but also its coloration. A substantial sample of hare bones (23.3% of burned NISP) are burned a faint grayish blue color in localized patches. Comparisons with bones burned in experimental conditions by Stiner et al. (1995), indicate that the color does not match the blue shade a bone passes through just prior to calcination, but is close to the faint greyishblue color that occurs only as a result of indirect contact with a heat source. A significant portion of the hare bones were most likely burned while protected from the heat source either by a thin layer of dirt or perhaps by the flesh of the animal itself during cooking.
Post-depositional buming is a less likely cause since hares are the only species that exhibit this unusual coloration, even though their bones were interspersed in the fill with all other species.
Table 5.25: Frequency of burned hare long bone shafts and long bone articular ends from the Natufian layer at Hayonim Cave.
Prop = proportion.
Buming is evenly distributed across hare elements, but, at the portion level of comparison, long bone shafts are burned in consistently greater frequencies than the
sample size is small (NISP = 9). On average, hare long bone shaft fragments are burned more than 50.0% of the time, much higher than the average rate of 27.2% for the complete hare sample. Shafts are also highly fragmented, and burning may have occurred as part of a more general human activity that also involved bone breakage (see below).
Gazelles The representation of gazelle body parts in Hayonim Cave was determined predominantly by human-mediated processes. The Hayonim gazelle assemblage provides some evidence for skinning, disarticulation, and defleshing, and, more vividly, intensive marrow extraction. The question of grease rendering is more problematic due to the potentially confounding influences of other mechanical processes, namely trampling. A few other lines of evidence (i.e., attrition of the hare assemblage is not density-mediated and fragile partridge long bone shafts are much less fragmented than those of gazelles) however, suggest that, though trampling probably played a role in assemblage formation, processing activities such as grease extraction were more influential. Trampling is expected to preferentially destroy low-density bone following deposition via periodic, repetitive mechanical loading on bones lying on or just below the living surface. Bones with low mineral density are less resistant to mechanical loading and thus more subject to destruction than denser bone. Unlike human processing effects, however, trampling is expected to differentially affect bones according to variation in bone density independent
activities is obviously mediated by variation in bulk bone grease yields and is expected only in animals of suitable body size (e.g., ungulates).
In the Hayonim Cave assemblage, density-mediated destruction occurs in the gazelle assemblage only. Smaller prey species (e.g., partridges and hares) show less pronounced attrition by density-mediated sources, as indicated by high completeness ratios for partridge long bone shafts (see Table 5.16) and the insignificant correlation between hare bone mineral density and survivorship (see Table 4.10). The fact that gazelle bones are interspersed with the bones of other taxa throughout the Hayonim Cave deposits means that density-mediated processes were not spatially localized but are specific to gazelle.
Though the relationship between bone survivorship and mineral density was shown to be significant for gazelle, the relationship is stronger when dense long bone shafts are removed from the equation. Bone survivorship was thus also influenced by other processes that crosscut bone tissue density boundaries. This pattern is exactly that expected from the combined effects of grease and marrow extraction. Grease rendering preferentially destroys soft, low-density long bone ends and vertebrae, and cold marrow extraction tends to fragment long bone shafts, though dense articular long bone ends are expected to survive. It has also been shown that, regardless of bone density, gazelle bone portions with poor survivorship are highly fragmented in the Hayonim Cave assemblage.
Cancellous bone and shaft fragments are therefore not entirely missing from the
identifiability as a result. Though this pattern is precisely that expected to result from marrow and grease extraction, it does not preclude trampling from playing a role.
Figure 5.11: Average fragment size for all taxa at Hayonim Cave.
Ta.xa in descending order of average fragment size.
In combination, these results point strongly to human processing activities as the single most important determinant of skeletal part frequencies in the Hayonim Cave gazelle assemblage. Trampling likely played an unspecified secondary role, as indicated by the small average fragment size for all taxa (Figure 5.11). On average, the Hayonim Cave assemblage is fragmented, but some bone portions are far more fragmented than others. The influence of in-bone nutrients likely began well before assemblage formation, playing a major role in prey transport decisions, particularly after the advent and proliferation of groundstone technology during the Upper and Epipaleolithic periods.
Increasing site permanence during the Epipaleolithic associates with increasing
Natufian required much work to manufacture but also offered significant bone processing advantages, undoubtedly increasing the value of what were previously considered lowutility prey bones.
Overall the evidence for processing activities from Hayonim points to very intensive use of gazelle carcasses. Natufians made thorough use of all components of the gazelle carcass, stopping only at very small marrow sources in compact elements, which were still extracted close to 50% of the time. Gazelle were transported to the site whole or nearly whole, stripped of meat, processed for marrow and grease and even used for raw materials for the manufacture of bone tools and ornaments. The faunal assemblage is highly fragmented overall, with an average fragment size of only 2.1 cm, attesting to the tendency of Natufian foragers to bash things up.