Assessing the reliability of silicone replicas in the analysis of sickle gloss at the early Neolithic site of Cheikh Hassan (Northern Levant, 11th millennium BP)

Over the past decade, confocal microscopy combined with quantitative texture analysis has opened new possibilities for studying prehistoric tool use. However, a key challenge remains in identifying the extent to which different acquisition methods—direct surface measurements versus silicone moulds—influence the textural properties of the original surfaces. This study evaluates whether high-precision polyvinylsiloxane can be reliably used for the identification of plant types harvested with lithic tools during the initial phases of agriculture in Southwest Asia. We compared texture parameters from original surfaces and moulds on eight experimental tools used to cut various plants (domestic, wild and wild cultivated cereals and reeds). Our results show that textures from direct measurements and those from moulds form clearly distinct groups, indicating that the two methods are not interchangeable for harvesting tools. The differences introduced by the moulding process follow a consistent pattern, rather than resulting from random variation. Based on this, we developed a silicone-based experimental model capable of distinguishing between tools used for harvesting different plant types. Applying this model to Pre-Pottery Neolithic glossy tools from Cheikh Hassan confirms a gradual shift from green to ripe cereal harvesting in the Middle Euphrates, in line with findings from neighbouring sites based on direct surface measurements. • Acquisition method—direct vs. silicone casts —significantly affects texture data. • Direct and cast-based measurements produce distinct, non-interchangeable patterns. • Casting introduces consistent, predictable changes in surface texture. • Cast-based experiments can identify plant types cut with Neolithic tools. • Cheikh Hassan shows a shift from green to ripe cereal harvesting in the PPN.