The expression of sexual dimorphism in human skeletal remains from ancient Egypt: regional and temporal differences and the impact on modern and population-specific metric sex estimation methods

UoM administered thesis: Phd

  • Authors:
  • Emily Marlow

Abstract

When considered in the context of other sources of information, the analysis of human remains can provide important insights into the population of ancient Egypt. Sex is an important component both of the biological profile of an individual, and the demographic profile of a population. Metric sex estimation methods are usually preferred in instances where skeletal remains are fragmentary or damaged; however, metric techniques are prone to error as a result of several biases, notably population differences in body size and skeletal proportions. In addition, many commonly used metric equations have not been validated for use with ancient Egyptians, and few population-specific equations exist. The study sample consists of 318 adult individuals, represented by a complete skeleton (n=162) or isolated cranium (n=156). The majority of individuals date to Old Kingdom (n=106) or Late Period (n=154) Giza. In addition, 43 individuals date to Pre-dynastic Period Keneh, 13 individuals to Middle Kingdom Sheikh Farag, and two individuals to New Kingdom Thebes. The sex of each individual was estimated using standard morphological methods. A total of 63 dimensions of the skeleton, or as many as it was possible to obtain, were measured for each individual in the study sample. Tests of intra- and inter-observer error revealed that the majority of measurements used in the study can be reliably measured. Testing of 12 modern methods of metric sex estimation revealed total weighted accuracy rates as low as 30-40%; many of the methods were additionally exceptionally poor at correctly estimating the sex of males. Previously created population-specific metric equations produced total accuracy rates ranging from 79.2-100% when tested on the study sample. New population-specific equations created using data collected from the study sample are associated with high rates of correct sex classification, often in excess of 95%. Three discriminant functions and three logistic regression equations created as part of this research were tested on a sample of 119 adult skeletons from late Old Kingdom and Ptolemaic Period Saqqara. Two functions and one equation produced acceptable accuracy rates in males and females separately in the Old Kingdom subsample. Results of principal components analysis and adjusting for body size suggest that some skeletal dimensions are sexually dimorphic with regards to both size and shape. All results arising from this research are discussed in the context of known factors affecting skeletal size and proportion differences between different populations and the expression of sexual dimorphism.

Details

Original languageEnglish
Awarding Institution
Supervisors/Advisors
  • Ryan Metcalfe (Supervisor)
  • Andrew Chamberlain (Supervisor)
Award date1 Aug 2015