Morphometric evolution of the domestic triad, in western Gallia Narbonensis (southern France, Languedoc), between the 2nd c.BC and the 4th c.AD: Preliminary and critical use of log size index for diachronic analysis

Does the choice of a reference matter for log size index analysis? A case study from roman times in southern France

“…The data correspond to those analysed in Jeanjean et al (2022) and include 5533 measurements, from which 1950 correspond to cattle, 2197 to suids (wild boar and domestic pigs undistinguished), 273 to goat, and 1113 to sheep. The measurements come from almost 30 years of preventive archaeology and were all acquired by one of us (VF) during the faunal analysis of 64 archaeological sites.…”

“…The measurements come from almost 30 years of preventive archaeology and were all acquired by one of us (VF) during the faunal analysis of 64 archaeological sites. These sites are located in the South of France and cover the period of time ranging from 200 BC to 400 AD (Jeanjean et al, 2022). The measurements were acquired following the Von Den Driesch guideline (von den Driesch, 1976) to which were added few extra measurements (see Jeanjean et al (2022)).…”

“…These sites are located in the South of France and cover the period of time ranging from 200 BC to 400 AD (Jeanjean et al, 2022). The measurements were acquired following the Von Den Driesch guideline (von den Driesch, 1976) to which were added few extra measurements (see Jeanjean et al (2022)). To each measurement was assigned a dating corresponding to the mean of the terminus post quem (TPQ) and terminus ante quem (TAQ) of the stratigraphic unit where the bone was recovered (similarly as in Jeanjean et al (2022)).…”

“…The measurements were acquired following the Von Den Driesch guideline (von den Driesch, 1976) to which were added few extra measurements (see Jeanjean et al (2022)). To each measurement was assigned a dating corresponding to the mean of the terminus post quem (TPQ) and terminus ante quem (TAQ) of the stratigraphic unit where the bone was recovered (similarly as in Jeanjean et al (2022)). The formulation used for LSI computation was: LSI = Log (x/y) = Log (x) -Log (y), with x the measure of the studied specimen, y that of the reference, and Log being the natural logarithm.…”

“…Some others, like us (e.g. (Jeanjean et al, 2022)), chose to use as a reference dataset the average measurements of their own assemblage. To which extent the references differ from each other, and if and how the choice of the reference influences LSI analyses and interpretations, is not yet well known.…”