Osteological Age & Sex Estimation

Children's skeletons change through tightly scheduled developmental milestones — tooth formation and eruption, bone growth, and growth-plate fusion — that track chronological age closely, so subadults can often be aged to within a year or two. Adults have finished growing, so age must instead be read from degenerative changes such as joint-surface remodeling, and the rate of degeneration varies with activity, health, and individual biology. This loose correlation means adult estimates are inherently broad, and accuracy declines further with age, so the oldest adults can only be placed in wide, open-ended categories.

The pelvis is the most reliable, because female pelvic anatomy is shaped by the demands of childbirth: traits such as the ventral arc, the subpubic concavity, and a wide greater sciatic notch strongly indicate female sex, while their absence indicates male. The skull provides secondary evidence through the relative robustness of the mastoid process, supraorbital ridge, mental eminence, and nuchal crest, scored on a gracile-to-robust scale. Buikstra and Ubelaker standardize the scoring of these traits. Sex is estimated only for adults, since these features are not reliably dimorphic before puberty, and is reported as a probabilistic assessment.

Because each skeletal indicator correlates only imperfectly with true age, any single feature yields a wide and potentially biased estimate, and indicators can disagree. Combining several — multifactorially or by seriation — produces a more robust estimate and exposes conflicts that would otherwise be hidden. Reporting the result as an interval with a confidence statement, and as a standardized age class, conveys the genuine uncertainty rather than a false point value. This honesty is essential downstream, since paleodemographic analyses depend on per-individual age estimates whose uncertainty is properly characterized.