Topper Site

What happened

1. 1998Goodyear's team digs up to a meter below the Clovis level and finds unusual stone pieces up to two meters below the surface.
2. 1999Outside geologists led by Michael Waters (Texas A&M) visit and propose a full geological study of the site.
3. 2001Geologists take plant remains deep in the Pleistocene terrace; OSL on soils above 'ice age strata' puts pre-Clovis at over 14,000 years.
4. 2004May: backhoe and hand excavation reaches ~4 m below the surface in the terrace; T. Stafford collects charcoal from a black stain where pieces lay.
5. 2004November: Goodyear announces dates of 50,300 and 51,700 14C BP (UC Irvine); a third sample is modern, attributed to an intrusion.
6. 2004Michael Collins, who had seen the pieces, calls them geofacts; David Anderson and others await formal publication.
7. 2005Goodyear reviews four eastern pre-Clovis sites, Topper included, at ~12,000-16,000 RCYBP; the abstract does not mention the 50k dates.
8. 2009Waters, Forman, Stafford & Foss publish a geoarchaeological study of Topper and Big Pine Tree; Waters' team had worked there since 1999.
9. 2018Goodyear & Sain maintain a pre-LGM pre-Clovis core-and-flake occupation (Clariant complex) in the Pleistocene terrace.
10. 2024Jennings et al. report a probable Clovis hearth and domestic structure, part of Topper's separate Clovis occupation.

Objections 0 of 4 proved wrong

• ?
  The stone pieces from the dated level are natural fractures (geofacts), not human-made tools.

  Natural, not human · Michael Collins, Univ. of Texas (2004); Waters et al. 2009 UNVERIFIED · 2004

  Unresolved — Proponents still describe a core-and-flake industry in the terrace (S2, 2018); no study settling human versus natural origin was found in the sources read.
• ?
  Earlier New World claims over 40,000 years had not survived professional examination, so the dates need careful verification and reporting in scientific journals.

  Contradicts the accepted model · David Anderson, Univ. of Tennessee (2004) · 2004

  Unresolved — No proponents' journal report of the ≥50,300 dates was found in the sources read; Goodyear's 2005 chapter (S8) covers younger pre-Clovis evidence only.
• ?
  One of three samples from the dated level gave a modern age, attributed to an intrusion, so the level was not sealed and the pieces' association with the dated carbon is uncertain.

  Layers or association · Intrusive date reported by T. Stafford (2004) · 2004

  Unresolved — ARCHAEOLOGY (S10) expected the context to face close scrutiny; no micromorphology or refit study of this level was found in the sources read.
• ✓
  The dates show only that the material is at least 50,300 radiocarbon years old; its absolute age is not known.

  Dating or samples · T. Stafford, the dater (2004 announcement) · 2004

  Held up — The dater's own caveat; ≥3 earlier dates on the lower soils were also beyond 50,000 years (S10). Not disputed in the sources read.

Evidence at the time within ~5 years of the claim

• Unambiguous evidenceNo — Stone pieces at a chert quarry. Michael Collins, who had seen them, called them geofacts; specialists awaited study of human versus natural origin.
• Documented contextPartly — Dug ~4 m into the Pleistocene terrace by backhoe and hand. Of three samples from the dated level, one was modern and attributed to an intrusion, so the level was not sealed.
• Independent dating methodsNo — Only radiocarbon dated the deep level; the OSL dates in S9's timeline are on soils above 'ice age strata' (pre-Clovis 'older than 14,000').
• More than one labNo — Both finite dates came from one lab (UC Irvine), on samples Stafford collected. The lab for the ≥3 earlier >50,000-year dates (S10) is not stated.
• Professional recoveryYes — Excavated by Goodyear's SCIAA team (Allendale PaleoIndian Expedition, with public volunteers), by backhoe and hand; Stafford collected the dated samples.
• Peer-reviewed reportNo — Announced by press release; Anderson awaited formal publication (S10). Goodyear's 2005 chapter covers only ~12,000-16,000 RCYBP evidence. S1 (2009) unread.
• Open to criticsYes — Outside geologists led by Waters worked at the site from 1999 (S9), and critic Michael Collins had seen the pieces by 2004 (S10).
• Independent teamsYes — Outside geologists led by Waters (Texas A&M) studied and dated the site's soils from 1999, alongside the project; they published separately in 2009 (S1).
• Outside stakeNo — No commercial, religious or nationalist motive documented. Goodyear was a USC research archaeologist at SCIAA (since 1974) and founder of the Allendale PaleoIndian Expedition.

What would settle it for a field partner

1. Date the sands of the artifact-bearing terrace level by luminescence (OSL/IRSL) for a finite burial age, after checking whether S1 already reports such ages.Medium cost
   Would show: A finite age for the level, which the radiocarbon minimum ages (≥50,300 14C BP) cannot give.
2. Blind test: lithic specialists sort the terrace pieces mixed with chert fragments collected from sterile terrace sediment, without knowing which is which.Low cost
   Would show: Whether the claimed tools can be told apart from natural breakage in the same deposit.
3. Micromorphology on any archived sediment from the 2004 black stain, or on a newly exposed equivalent stain in the same terrace level.Medium cost
   Would show: Whether the burnt layer is an in-place human combustion feature or natural burning or root staining, and whether intrusions cut it.
4. Use-wear and residue analysis at high magnification on pieces from the dated level, with naturally broken chert from the same terrace as controls.Low cost
   Would show: Edge wear or organic residues that natural fracture would not produce.

What would change this record

Refuted if blind analysis finds the terrace pieces indistinguishable from naturally broken chert in sterile terrace sediment. Supported if pieces sealed in the dated level show use-wear or residues natural breakage cannot make, and finite luminescence ages confirm its age.

Notes

• Status. Dates from the level are minimum ages only (≥50,300 14C BP). Collins (2004) judged the pieces geofacts; the proponents still present a pre-LGM occupation in the terrace (S2, 2018). Human origin remains unsettled.
• Accepted view then. Pre-Clovis Monte Verde (~14,500 cal BP) had been accepted in 1997 (S11); no New World claim over 40,000 years had survived professional scrutiny (S10).
• Dating basis. Two radiocarbon dates, 50,300 and 51,700 14C BP (uncalibrated), on burnt plant remains, UC Irvine lab; the dater said the absolute age is not known. bp_min is the proponents' 'at least 50,000 years'.
• Scope: only the 2004 finds deep in the Pleistocene terrace. Pre-Clovis finds of 1998-2003 lay above the terrace, in soils outside geologists dated to ~16,000 years (S9, S10); the Clovis occupation (S4, S7) is separate. Stafford both collected the 2004 samples (S9) and co-authored S1. S1 could not be read (paywalled; open copies blocked), so reports that it doubts the human origin of the sub-Clovis pieces come only from search summaries: UNVERIFIED. Sources differ on Stafford's affiliation: Stafford Laboratories, Boulder (S9) vs University of Wisconsin (S10). The expedition includes public volunteers under professional direction (S9).

People and evidence

• Proponents. Albert C. Goodyear, Douglas A. Sain
• Critics. Michael Collins, Michael R. Waters
• Evidence types. radiocarbon, stratigraphy, geology, lithics · 2 independent lines
• Record confidence. medium — Key facts rest on two 2004 reports read directly (S9, S10) and abstracts (S2, S8, S11). S1, the main external study, is unread (paywalled), so none of its findings are used.