Starch Grain Analysis

There is no single trait that proves antiquity, so confidence comes from a combination of optics and controls. Polarized-light birefringence and the Maltese-cross extinction confirm that a particle is genuine, well-ordered starch, and damage features can suggest processing such as grinding or cooking. But because modern starch is everywhere, the decisive safeguard is a control regime: sampling sediment away from the artifact, unused surfaces, and laboratory blanks, and comparing what those yield to the artifact residue. Pearsall stresses that without such controls a starch assemblage cannot be confidently attributed to past behavior, and contamination must be assumed until ruled out.

The two methods recover different parts of the plant record and are complementary. Flotation recovers charred macroremains — seeds, chaff, nutshell, wood — but systematically misses plants that are rarely burned or that lack durable hard parts, including many tubers and roots. Starch grain analysis recovers microscopic granules that lodge on tools, in pots, and in dental calculus, capturing exactly those starchy roots and tubers that flotation underrepresents and tying plants to specific artifacts and individuals. Used together, macroremains, starch, phytoliths, and pollen give a fuller, cross-checked picture than any single line of evidence.

No. Identification depends on how diagnostic a granule's morphology is, and many plants produce granules that overlap in size, shape, and hilum form, so they can only be assigned to a family or broader group. Some economically important taxa do produce distinctive granules — with characteristic size ranges, eccentric hila, or pressure facets — that allow genus- or species-level identification when compared against a good reference collection. Responsible analysis reports identifications at the rank the evidence supports, separating secure species-level calls from broad categories, rather than forcing every granule to a single plant.