X-Ray Fluorescence Sourcing

Because volcanic glass is chemically homogeneous within a flow and chemically distinct between flows. A single obsidian source has a consistent trace-element signature throughout, so any fragment of it gives essentially the same composition, and different volcanoes differ enough in elements like rubidium, strontium, zirconium, and yttrium to be told apart. This combination makes assignment clean and reliable, which is why obsidian sourcing — beginning with Renfrew, Cann, and Dixon's pioneering studies — became the model application of compositional provenance and why portable XRF is so effective for it.

It can be, but only with care. Portable instruments are convenient and non-destructive, yet analyzing small, irregular, or unprepared surfaces introduces geometry and matrix effects, and pXRF generally has poorer detection limits and precision than benchtop wavelength-dispersive XRF, NAA, or ICP-MS. Shackley stresses the need for calibration against matched reference standards and for treating uncorrected pXRF data as semi-quantitative. For robust, homogeneous materials like obsidian, well-calibrated pXRF gives excellent source assignments; for heterogeneous or weathered materials, results must be validated more cautiously.

It cannot be correctly identified, and the danger is misassignment. Source assignment works by matching an artifact to the closest known source signature, so if the real source was never sampled, the artifact may be forced onto the nearest available group or, with a strict distance threshold, flagged as unassigned. Shackley emphasizes that the comprehensiveness of the source library limits the whole analysis. Good practice is to use a quantitative distance cutoff so that genuinely unmatched artifacts are recognized as such rather than wrongly attributed, prompting a search for additional sources.