Dating petroglyphs and pictographs follow the same approach (finding the best area to test under the art, plus scraping and AMS dating of arbitrary levels). But their fallen material requires different analyses. We look for fallen art debris instead of pigment using sieves rather than cameras. We increase level thickness from 5 to 10 mm to capture petroglyph tools and their fragments exceeding 5 mm in thickness. Glue sheets are not needed. For more information, case studies of petroglyph dating are found in this website for Coso and Gabriola Island.
When a petroglyph is pecked out of a rock face with a hammerstone, fragments from both hammerstone and parent rock fall to the feet of the artist. Harder hammerstone fragments are reduced to a fine white flour, while softer rock art fragments are larger, like naturally fallen rock fragments. Our method of finding petroglyph age entails dating tiny amounts of natural or cultural organic matter in close proximity to art debris, including complete hammerstones and grinders. We do this by scraping 12 mm arbitrary soil levels under the art and differentiating cultural from natural fragments. Hammerstone pock marks and percussion theory suggest their fallen fragments have sharp edges, while soil fragments are rounded from weathering. Flour is also made when grinders are used to smooth the pocked petroglyph. Our secondary proof is to see if flour particles are also flat and sharp-edged; hence, distinguishable from natural soil particles.
Field sieving permits differentiation of larger hammerstone chips from rock art fragments by their hardness, mineral type and shape. Sieves with mesh sizes above 1.5 mm can aid differentiation, but microscopic examination is needed for samples smaller than 250µ (microns). Fallen hammerstone debris varies in size from complete tools to microscopic particles, but are confined to levels when the art was made. Tests dug away from the art indicate their absence, making it easier not to confuse them with soil chips and grains. Within the cultural levels, rock art tools are often of quartzite, basalt and quartz, minerals readily identified under a dissecting or low power microscope. Parent rock art material is often softer sandstone, limestone and schist, but may be harder like Coso rhyolite where naturally fallen flakes are much smaller and indistinguishable from petroglyph flour and native soil.
To test to see if petroglyph flour particles follow the same outline as larger, flat, sharp edged fragments, we compared debris below natural and laboratory-made sandstone petroglyphs uncontaminated with soil. Debris under petroglyphs that were artificially ground using a sand slurry between a line abrader and parent rock showed a mix of flat sharp particles with rounded particles, the latter likely from fallen un-used slurry. For particles under 250µ we found the trend continued with readily visible silica or sand-like material down through the silt sizes of <50µ to clay sizes of <2µ. As examination and comparison between natural soil and artificial petroglyph samples revealed no clear distinguishing traits, we decided not to pursue it with SEM studies.
In attempting to differentiate flatter petroglyph flour from rounder soil grains we used differential sedimentation. After mixing filtered samples with water into a slurry, we let them settle though a 1.5 m long plastic pipe full of water. We expected the sharper particles would settle last, but examination of the contents of different layers within a cc or so of sediment failed to distinguish natural from manmade particles by shape only. Better results might be obtained by omitting sand as an abrasive while making the petroglyph.