Inverted Radiocarbon Dates and a Crystal-Filled Pit
At a limestone rockshelter in southwest Asia, a 2 by 3 m excavation exposed six stratigraphic layers. Nine AMS radiocarbon dates on charcoal cluster between 24,000 and 19,000 cal BP, but three dates invert the stratigraphic order by 400–700 years, and a later pit cuts Layers 3 and 4. The lithic assemblage of about 1,200 pieces includes refits that join pieces from Layers 2 and 3. In a small sub-floor pit capped by a limestone slab, excavators found 14 ochre-stained quartz crystals and a pierced marine shell. The faunal analyst reports mostly gazelle with cut marks, and micromorphology has not yet been done. A colleague wants to submit a paper calling the pit 'the earliest evidence of ritual behaviour in the region'; a second colleague wants to redraw the layer boundaries and discard the inverted dates.
1. How should the inverted radiocarbon dates and the intrusive pit be handled analytically?
Read the worked answer
Treat the dates as a set, not individually. Charcoal gives a terminus post quem only if the wood was already dead before burning; old-wood offsets and the intrusive pit can produce exactly this pattern. Steps: place every date in its stratigraphic and feature context, including the pit, and flag dates from contexts of doubtful integrity; run a Bayesian phase model using the stratigraphic ordering as prior information, which will downweight or identify outliers; report calibrated ranges with the model output and the outlier analysis rather than a simple average; and submit additional samples from secure contexts, preferably short-lived taxa such as seeds or bone rather than charcoal. Discarding the inverted dates outright is also wrong - they may date the pit or a later use of the shelter, which is information. The pit fill should be dated separately from the layers it cuts.
2. Evaluate the 'ritual' claim: what alternative explanations and middle-range reasoning apply?
Read the worked answer
The pit is suggestive, not demonstrative. Ritual is a behavioural inference, not a material category, so it requires middle-range argument: what would a non-ritual deposition of ochre-stained crystals under a slab look like, and can those alternatives be excluded? Candidates include caching of raw material, a child's collection, a construction deposit, or post-depositional movement. Ochre has many uses - tanning, adhesive, body decoration, medicine - and the crystals may be manuports. The pierced shell is a stronger ornament signal but still not ritual per se. What would strengthen the case: micromorphology of the fill and slab to show deliberate placement and sealing, use-wear and residue analysis on the crystals, spatial comparison with other pits at the site, and evidence that the deposit is not a cache. 'Earliest evidence' claims also require systematic regional comparison, not a single site.
3. The refits cross Layers 2 and 3. What does that imply for the occupation history, and what should be done next?
Read the worked answer
Refits joining Layers 2 and 3 mean the layer boundary is not a depositional boundary - at least part of the assemblage is a palimpsest, or the layers are archaeologically isochronous in that area. Implications: minimum occupation counts and per-layer assemblage comparisons are unreliable, and the knapping event may be a single episode, which changes estimates of discard rate and intensity of site use. Next steps: plot the refits in three dimensions, examine the sediment boundary with micromorphology, and determine whether the join crosses a real disconformity or a gradual transition. If the layers are geogenic rather than cultural, the sequence should be described as one deposit with internal variation. This also bears on the ritual claim, because the pit's stratigraphic priority depends on the integrity of those same layers.