Dating and context
Halligan et al. date Unit 3c tools to approximately 14,550 cal BP using associated wood, not the stone itself. [S27 · Halligan et al. (2016)]
| Find | Published relationship |
|---|---|
| Biface 12080-1 | Main paper reports seven adjacent dates; register distinguishes above/below/beside/level relationships, including older material |
| Flake 12242-1 | Two associated wood dates; screen-recovered find with inferred collection context |
| Modified tusk, recovered 1993 | Attributed to Unit 3c; not one of the newly excavated tools |
The paper’s 71 wood dates concern the broader stratigraphy, not 71 independent dates on one human action. All 42 excavation measurements in Tables S3/S4 are now image-checked and indexed; the 29 core dates are outside this register. [S27 · Halligan et al. (2016)]
A resource-use episode
Tusk grooves are interpreted as working during extraction. Butchery or scavenging is supported; a kill, typical menu and permanent settlement are not demonstrated. The pond-edge setting differs from today’s submerged site. [S27 · Halligan et al. (2016)]
Next checks and atlas coverage
Next obtain the original OxCal commands and output to resolve the exclusion-count discrepancy and test sensitivity to the retained discordant pair. The atlas uses a rounded contextual point, not zero uncertainty. Later Unit 4 artifacts remain distinct. Compare daily life.
Original sample relationships
The original excavation register preserves Tables S3/S4, including older wood and bark. Its Figure S12 audit joins all 37 samples from 50N/23E: eight are marked excluded and 29 appear as retained, individually or combined. Section 2.4 says nine were removed. No ninth laboratory ID is inferred. The extra Table S3 row, UCIAMS-153343, is from 50N/22E and is not evidence of an additional outlier. [S195 · Halligan et al. (2016)]
Among seven biface-related results, 127305 (bark) and 141950 (bald-cypress wood) are marked excluded. The other five appear as 141949, combined 127302/127304 and combined 127303/141951. Several cypress twigs remain in the model; the exclusions are not a blanket rejection of that species. Older plant material can predate deposition and does not independently demonstrate earlier tool use. [S195 · Halligan et al. (2016)]
UCIAMS-141934 and 141935 remain combined despite the authors’ reported failed chi-square test. Published model agreement is not a probability that the chronology is correct. We have inspected the publication, not rerun the model or inspected original certificates. The approximately 14,550 cal BP assessment remains attributed to the authors; this audit supplies no replacement date. [S195 · Halligan et al. (2016)]
Field-context precision
The biface was recovered in situ, but §1.3 gives its northing as 50.23 while Table S10 gives 50.32; both give E23.22 and 10.84 m below datum. Preserve both until the original records resolve the difference. [S195 · Halligan et al. (2016)]
Flake 12242-1 was found in the screen while a roughly 10 × 15 cm corner was excavated. The authors describe cleaning, unchanged diver/hose positions and a restricted excavation area as reasons to trust its context. Its tabulated coordinates represent that inferred context, not a directly observed in-situ artifact point. Samples were taken at the top/bottom of the excavated level, but Table S4 lists both near-flake measurements at depth 10.78; a lab-ID-to-upper/lower assignment remains unlinked. [S195 · Halligan et al. (2016)]
Atlas/globe coverage checked: AT-PL-3C retains its contextual estimate with these limitations. Excluded older wood and later Unit 4 samples do not receive the Unit 3c human-activity date. No new traditions evidence arose in this check.