Identifications of studied bone specimens, and information from associated archaeological context, are fundamental to interpretation. It is often difficult to arrive at specific conclusions about bone assemblage formation because identifiable traces tend to be elusive and ambiguous, even when preserved. They are, however, crucial for establishing the reliability and validity of any inference about past subsistence and ecology.
Equifinality
The exact identification of clues surrounding assemblage accumulation and preservation often remains elusive because of equifinality. Cause and effect can be difficult to establish because dissimilar events can produce similar results, or because unvarying outcomes do not consistently result from the same event. The problem is compounded because an assemblage originally impacted by one variable can be subsequently modified by another. Bone assemblages usually bear multiple signatures that are often the result of disassociated events. It is our task to identify them as best we can, understand their origin, order their sequence of appearance, and evaluate their importance to the questions we ask about subsistence and ecology.
The ultimate cause of bone preservation lies in the variable structural density of its constituent materials. Harder portions of the vertebrate skeleton have a greater chance of surviving attritional forces. Evaluating the nature and degree to which the preserved assemblage was subjected to density-mediated attrition is of first-order importance for understanding how adequately the study assemblage reflects the original accumulation. Zooarchaeologists have established independent means to assess the variability of structural density in skeletons of different animal taxa. Through comparing skeletal part frequencies with a hypothetical assemblage that survives on the basis of structural density, we can assess whether or not assemblage preservation was influenced by any number of attritional forces whose effects are mediated by differential structural density of bone material. However, this can only demonstrate an ultimate cause because numerous different attri-tional events can be responsible for density-dependent survivorship. We must rely on observations of the recovered assemblage in order to suggest reasonable possibilities about proximate causation. Here, the likelihood that our assessments are correct is strengthened through corroboration from multiple lines of evidence (see Archaeozoology).
Subsistence Analysis
If it is reasonable to assume that the recovered assemblage was originally accumulated and deposited by humans for their consumption, then we can ask questions appropriate to human subsistence. However, we must be aware that quantitative and qualitative differences in the studied assemblage can result from factors that contributed to its modification during the time between original accumulation and eventual analysis. The basic analytical unit for all subsequent inferences is the NISP of animal taxa that are relevant to inferences about prehistoric consumption. However, different values of NISP can vary for reasons that are irrelevant to our inferences about subsistence. They may reflect simple variability between species and element, and are strongly affected by fragmentation, differential preservation, and the method of field recovery used.
Subsistence inferences based solely on NISP would be misleading, primarily due to specimen interdependence. Therefore, NISP is normally presented alongside an estimate of the minimum number of individuals (MNI) of a given taxon required to account for the NISP of that taxon in the sample. In its simplest form, MNI is equal to the highest number of either left or right paired element portions for each given taxon. The minimum number of elements (MNE) is similarly computed without distinguishing body side, and can be easily converted into minimum animal units (MAU), which relate MNE to individual bodies. Although not without their attendant problems, these derived permutations of NISP offer units of analysis that are more usable for questions about diet. Increasingly robust and reliable inferences about dietary contribution are possible when coupled with appropriate utility indexes. These are empirically determined estimates of total food value represented in different skeletal portions of different animal taxa.
Archaeological data can help us to identify what kinds of animals humans exploited, which sexes or ages were preferred, and when, where, and how they were procured, processed, and consumed. Nevertheless, the exact quantitative relationship between the recovered and deposited samples remains unknown. Moreover, we can never be certain how representative the deposited assemblage is of the original sample of animals that was procured and consumed by humans. Therefore, it is usually difficult to establish anything but a rough estimate of relative dietary contribution.
Palaeoecological Analysis
If we can establish how the recovered assemblage was originally accumulated and deposited, and understand how attributes of the study assemblage reflect biases introduced by the accumulating agent, then we can ask questions appropriate to palaeoecology. Again, we must be aware that quantitative and qualitative differences in the studied assemblage result from factors that contributed to its modification during the time between original accumulation and eventual analysis. However, we can never be certain that the quantitative structure of our recovered sample accurately reflects either the structure of the original accumulation or a hypothetical population of animals in the past. Reliable palaeoecological inferences are never based on treating taxa within the assemblage as variables, but as attributes that are either present or possibly absent.
Palaeoecological inference rests on analogical reasoning. It uses relevant attributes of analogous contemporary taxa whose archaeological counterparts are represented in archaeological context to inform about environmental conditions at the time and location of deposition. Animals that lived locally to the site of deposition are preferred for palaeoeco-logical inference, as are narrow-niched stenotopes or specialist species that tolerate a narrow range of conditions. Our inferences assume that the ecology of contemporary taxa has changed little from similar taxa in the past, and that they are stable in time and space. Their validity also rests on the assumption that archaeological resolution is clear and that the identified taxa used for palaeoecological inference were not transported to the site of deposition from far away. The likelihood that our palaeoecological inferences are correct, is strengthened through corroboration from multiple lines of evidence, particularly by using suites of taxa that point to similar conclusions.
World History
