The three articles that I have selected to annotate all deal with various aspects of paleopathological trauma analysis. The first is a seminal article that represents the first of such analyses to consider, acknowledge, and attempt to address the potential confounding effects of fragmentation and differential preservation on quantitative analyses of traumatic lesions in archaeological skeletal samples. The second is more focused on methodology, providing a comparison of various quantitative methods and evaluating how method choice can affect quantitative and qualitative results. The third is an example of an analysis of traumatic lesions in a fragmentary, commingled skeletal sample similar in condition to the material comprising the Smith’s Knoll assemblage. These articles illustrate different issues, all of which are important considerations in the analysis of paleotrauma and all of which are explored in my own work.

Lovejoy, C. Owen, and Kingsbury G. Heiple

1981 The Analysis of Fractures in Skeletal Populations With an Example From the Libben Site, Ottowa County, Ohio. American Journal of Physical Anthropology. 55: 529-541.

This article is often said to represent a pivotal point in the recording of trauma in archaeological collections; it is commonly cited as the first example of a paleopathological study to provide an epidemiological, population-based evaluation of paleotrauma that considered skeletal completeness as a factor that may affect quantitative analyses in archaeological collections. Lovejoy and Heiple’s (1981) article is therefore an extremely important publication to address in a paper that discusses quantitative analyses of archaeological traumatic lesions. Up to the point at which this article was written, the majority of fracture analyses had been primarily descriptive and had focused mainly on case studies. Lovejoy and Heiple (1981) instead aimed to evaluate the overall frequency of fracture lesions in the individuals of the Libben population, a sample from a Late Woodland site in Ohio. They heavily emphasize the fact that in order to extract meaningful information that can be used for population analyses, the standardization of methods is required; as such, they are extremely clear regarding the methods that they have used.

For the individuals in the Libben sample, the authors evaluate the frequency of antemortem or healed fracture lesions in only complete long bones, both by element and by individual. They then examine potential trends in the frequency data by age at death, sex, and element side, and go on to evaluate fracture frequency by years at risk, fracture etiology, and indications of patient care. The fracture data for this population are compared with data from non-human primates, and Lovejoy and Heiple (1981) assert that while comparisons with other aboriginal populations, both ethnographic and archaeological, would be ideal for the contextualization of the patterns observed in the Libben sample, suitable studies with well-defined methods had not yet been completed at that point. The authors provide minimal contextual information regarding the sample under study, and therefore the article is more osteological than bioarchaeological in nature. The authors provide raw data on the number of fractures by element at the beginning and the results of various rate calculations and statistical analyses following this and interspersed with discussion. This more concept-structured paper is therefore composed almost entirely of results and discussion, with little separation between the two. Methods are well described, and are presented along with the corresponding results and discussion rather than in a separate section. This paper is an extremely useful and important article for paleopathologists studying trauma in archaeological populations, but may have limited utility for other researchers. The most important contribution of this study is the authors’ demonstration of the fact that “the systematic observation and recording of fracture data can provide valuable information in the analysis of both human and non-human primate populations when properly adjusted for demographic variables” (Lovejoy and Heiple 1981, 541).

This paper was the first to engage with a question that is central to the paper I am writing, namely the issue of how differential completeness and preservation of archaeological skeletal material impacts the quantitative evaluation of traumatic lesions. Lovejoy and Heiple (1981) chose to deal with this issue by including only complete skeletal elements in their analysis. This is an effective solution, however it is not a viable option for the analysis of collections that are heavily fragmented and in which very few elements remain complete and undamaged, as is the case for the Smith’s Knoll collection. This article is representative of the direction in which paleopathological analyses of archaeological trauma have progressed since moving beyond the focus on individual case studies. The majority of more recent paleopathological analyses of trauma reference this study as a seminal article and important turning point, and many paleopathologists have utilized the methodology implemented by Lovejoy and Heiple (1981). While the approach taken by these authors has been extremely influential and has succeeded in beginning an anthropological discussion around the effect of differential preservation on the visibility of injuries in the skeleton, there are significant limitations in the application of these methods to fragmentary material. The analysis of fragmented, disarticulated, and commingled skeletal remains is the area of study for which this discussion proves most relevant, but it is also an area in which full discussion of how to approach methodological standardization has stalled. This paper therefore presents questions that can and should be reframed in terms of the analysis of fragmentary remains, and this is an issue I hope to explore in my own work. Lovejoy and Heiple (1981) serves as an excellent starting point for opening the discussion of how methods can be standardized to account for differences in preservation and completeness of the skeletal samples under investigation.

Judd, Margaret A.

2002 Comparison of Long Bone Trauma Recording Methods. Journal of Archaeological Science. 29: 1255-1265.

Judd’s (2002) article discusses various methods of quantitative analysis applied to the paleopathological investigation of traumatic lesions, and evaluates to what extent their results may be altered by including more fragmentary material in the calculations, using the example of an ancient Nubian skeletal sample. In this methodological paper, the author evaluates the preservation and location of lesions on long bones using a segmentation method that incorporates recommendations made by Buikstra and Ubelaker (1994) with a clinically-developed method of determining articular segments. She then applies two variations of each of five different methods of recording long bone trauma to her archaeological sample, and evaluates differences in the quantitative and qualitative results. Judd (2002) emphasizes the importance of standardization not only in the methods used to evaluate the frequency of trauma, but also in the methods used to assess completeness and to determine inclusion criteria. This standardization is most important to facilitate comparisons between the results of analyses completed on various archaeological samples as well as between archaeological and clinical studies. Judd’s (2002) article corresponds organizationally to a more typical “scientific” paper, with clearly differentiated sections for background information, methods, results, and discussion. She provides a comprehensive background of quantitative paleopathological analyses of traumatic lesions, highlighting the approaches that have previously been applied and how they have or have not taken issues of preservation and completeness into account. For the Nubian sample, Judd (2002) determines that quantitatively, few meaningful or statistically significant differences existed between methods at the population level. However, the exclusion of fragmentary elements that contain fractures may affect the patterns of fractures observed in the collection, and may therefore have an impact on the predominance of various lesion etiologies such as accidental or violence-related injuries. Judd (2002) therefore concludes that the interpretation of traumatic injuries in a population based on qualitative data may be affected by method choice; she stresses that significant quantitative differences may also exist if the collection under study is severely damaged, and therefore methods of recording and analysis should be selected primarily based upon the integrity and preservation of the skeletal material in a given sample.

This paper is very methods-focused rather than bioarchaeological, and the archaeological sample discussed is used primarily to illustrate the effects of altering methodologies rather than to evaluate and interpret the frequency and pattern of trauma in the population. It is therefore primarily useful for paleopathologists who will have occasion to use the methods discussed, as it has the potential to impact on research design but doesn’t provide much information that could be used in other capacities. The major contribution of this paper is the author’s evaluation of whether there are any significant and meaningful differences between various methods that are used to quantitatively analyze trauma in archaeological populations, and her conclusion that for the sample she has chosen to analyze there are no significant quantitative differences if more fragmentary skeletal material is included in the analysis. It is important to note, however, that the exclusion of injuries present in fragmented and incomplete elements can affect more than just the quantitative results, but could result in differential evaluation of comparisons between sex or age groups, the analysis of injury recidivism or multiple injuries present in the same individual, and the prevalence of injuries that can be attributed to accidental or violence-related etiologies.

This article provides a very important evaluation of how quantitative results can vary based on the inclusion of skeletal material of differing preservation and completeness. Previous researchers working with more fragmentary collections did acknowledge the potential confounding effect of the degree of completeness of skeletal remains, but there was little consistency in the manner of estimating and accounting for completeness in quantitative analyses. Judd’s (2002) analysis represents the first attempt to actually analyze how the inclusion of more fragmentary material might affect quantitative and qualitative representations of trauma frequency as well as how one might go about standardizing the estimation of element completeness.

Judd (2002) discusses several important issues related to the quantitative analysis of paleotrauma, namely the importance of methodological standardization to enable the comparison of results. Comparability is an extremely relevant issue, as the establishment of the context of a given sample in relation to other archaeological samples as well as relevant clinical studies is essential to the classification of an analysis as bioarchaeological. This article continues and elaborates on the discussion initiated by Lovejoy and Heiple (1981), a discussion that I hope to contribute to. Despite Judd’s (2002) contribution to this dialogue, however, quantitative analyses of trauma published since 2002 have not necessarily incorporated her recommendations for the standardization of not only quantitative methods but also the estimation of element completeness. I think that the concerns raised by Judd (2002) can be reframed in light of more recent attempts at quantifying trauma in fragmentary material, as well as of newer methods of recording fragmentary skeletal material that have come to light since this article was published, such as the zonation method that was applied in recording the Smith’s Knoll collection.

Cunha, Eugénia, and Ana Maria Silva

1997 War Lesions from the Famous Portuguese Medieval Battle of Aljubarrota. International Journal of Osteoarchaeology. 7: 595-599.

This article represents an example of a paleopathological analysis of traumatic lesions in a heavily damaged and severely fragmented skeletal collection. Similar to the individuals comprising the Smith’s Knoll assemblage, warriors killed during the medieval battle of Aljubarrota in Portugal were buried in an ossuary following the battle. Due to years of surface exposure prior to burial the remains display significant damage and fragmentation; the majority of the sample consists of long bone fragments. The paper is presented as a preliminary analysis of the traumatic lesions in the sample; Cunha and Silva (1997) specify that frequency data are not presented in this article due to the preliminary nature of the results. However, as a more detailed summary of the results was published only in a site report volume in Portuguese rather than a refereed journal, any quantitative results that were published remain inaccessible to many English-speaking researchers. The authors provide an overview of the types of injuries present in the collection, in conjunction with a detailed description of the types of weapons that were used during the battle based on historical documentary evidence. The violent injuries present at Aljubarrota are also considered in the context of other medieval European battle collections, of which there are relatively few, making this osteological war series particularly important evidence for battle conditions during the medieval period. This paper is clearly written as a preliminary report on injuries that the authors plan to review in more detail in further publications, and as such very little is said about the methodology used to assess skeletal lesions. The paper goes through historical background and a description of archaeological analyses and the sample itself, and then jumps into the results of their initial analyses, which are presented in conjunction with their discussion. Cunha and Silva (1997) use this article to explore the types of traumatic lesions they have observed in the sample from Aljubarrota, and how these injuries may be related back to the circumstances of the battle with which they are associated; this information is most relevant to paleopathologists but also to historians or archaeologists studying medieval warfare.

While this paper does not provide a quantitative analysis of the traumatic lesions observed by the authors, it does an excellent job of demonstrating some of the other issues associated with the bioarchaeological analysis of very fragmentary collections. Cunha and Silva (1997) illustrate some of the difficulties they encountered in terms of even basic skeletal recording procedures like age determination; basic recording must be undertaken before any more detailed analyses of traumatic lesions can proceed and therefore difficulties in the former have the potential to significantly affect the latter. This paper also demonstrates related limitations in processes more directly tied to trauma analysis, such as the difficulty of looking at the age distributions of injuries and the impossibility of viewing multiple injuries in a single individual. Cunha and Silva’s (1997) article therefore demonstrates very clearly many of the difficulties associated with the paleopathological analysis of skeletal collections that contain fragmentary and commingled remains. Their reluctance to present any quantitative results at this relatively early stage in their research also illustrates what I would consider to be a widely held view within paleopathology toward quantitative analyses in these types of collections. The calculation of trauma frequency in fragmentary skeletal material is seen by many paleopathologists to be a difficult and involved process, requiring serious consideration, which may or may not turn out to be valuable or even possible. Therefore, while this report is published far enough along in the research process to present results and interpretative conclusions regarding injury patterns and warrior behaviors in terms of weapon use, armor, and even some aspects of participant identity, it is considered too preliminary to discuss any quantitative conclusions.

Cunha and Silva’s (1997) article addresses many of the limitations and difficulties associated with the paleopathological analysis of poorly preserved skeletal collections. This study also illustrates the paucity of accessible quantitative data from these types of assemblages, and indirectly demonstrates how this may be related to underlying attitudes held by paleopathologists toward these types of analyses. It is these attitudes that I am hoping to address with my paper, in order to begin a productive dialogue about the limitations but also the potential value of quantitative analyses in these “difficult” skeletal assemblages.