Evaluating the effects of parallax in archaeological geometric morphometric analyses

Abstract

Geometric morphometrics is a powerful set of techniques that can be used to visualize and analyze the shape of artifacts. With the growing use of geometric morphometrics in archaeology, it is important to understand and identify limitations in the method. One such limitation is the accumulation of measurement error. Here, we investigate the impact of parallax or the effect of the position of an object in relation to the camera. We designed an experiment to assess the effect of parallax on measurements of artifact morphology by photographing a sample of artifacts at close range (50 cm) and systematically shifting the fixed angle of the camera relative to the artifact in five steps: 90°, 95°, 100°, 105°, and 110°. We took digital images of geometric microliths from three Jordanian Epipalaeolithic sites at each of the camera angles. We then digitized the outline of each artifact using 24 sliding landmarks. Our subsequent analyses of microlith shapes grouped by camera angle show that they are statistically indistinguishable from each other, which suggests that within these parameters, parallax has little effect on geometric morphometric measurements. While taking digital images directly above artifacts is ideal, the angle at which previously published photographs of artifacts is sometimes unknown. Our findings suggest that small deviations of the camera angle (up to 20° from horizontal) will not significantly impact geometric morphometric analyses.