Tecaxic-Calixtlahuaca "Roman head"

What happened

1. 1933José García Payón's excavation at Calixtlahuaca reportedly finds the head in a burial offering under three intact floors of a pyramidal structure.
2. 1960Ignacio Bernal brings the head to the 34th International Congress of Americanists in Vienna; it is not generally accepted as evidence of contact.
3. 1961García Payón publishes a two-page note in Boletín INAH, his only publication on the head in his lifetime.
4. 1961Heine-Geldern publishes 'Ein römischer Fund aus dem vorkolumbischen Mexiko' ('A Roman find from pre-Columbian Mexico').
5. 1979García Payón's fuller excavation account appears posthumously; H&G call it detailed, while Smith finds little specific context in it.
6. 1995The head is tested by thermoluminescence at the Forschungsstelle Archäometrie, Heidelberg.
7. 1996Paul Schmidt tells Hristov that the head is known to be Colonial and was planted as a joke; he later names Hugo Moedano, citing John Paddock.
8. 1999Hristov and Genovés revive the claim in Ancient Mesoamerica, reporting a TL age of 1780 ± 400 b.p. and no sign of post-Columbian intrusion.
9. 2001Schaaf and Wagner, who ran the TL test, say the reported figure and its reference were wrong; the corrected limits are 870 BC-AD 1270.
10. 2001Hristov and Genovés reply, adopting the corrected TL limits and rebutting the post-1492 import and 'planted' objections.
11. 2003Michael E. Smith's web page lists six alternative explanations and calls the case for a Roman object 'weak'.
12. 2015Huster and Smith date Calixtlahuaca's last phase (Yata) to AD 1450-1530, later than the AD 1510 abandonment H&G assume.

Objections 1 of 7 proved wrong

• ?
  The thousand-year gap between the head's style and the rest of the offering cast doubt on the find's reliability; the 1960 congress did not generally accept it as evidence of contact.

  Layers or association · Participants, 34th International Congress of Americanists (Vienna) · 1960

  Unresolved — The association still rests on García Payón's account (S6); no re-examination of the deposit is reported in the sources checked.
• ✕
  The head is a Colonial-period object: the Museo Nacional de Antropología catalogued it as Colonial, and in 1996 Paul Schmidt told Hristov that 'everybody knows that the head is Colonial'.

  Natural, not human · Museo Nacional de Antropología catalogue; Paul Schmidt (UNAM) · 1996

  Proved wrong — For manufacture only: the corrected TL limits end about AD 1270, and the lab says they 'probably' rule out a Colonial manufacturing date (S2). Colonial deposition is a separate, open objection.
• ?
  The head was planted as a joke during the dig: Paul Schmidt said so in 1996 and, in a letter of 6 March 2000, named Hugo Moedano as the prankster, citing John Paddock.

  Fraud or provenance · John Paddock (oral), relayed by Paul Schmidt (UNAM) · 1996

  Unresolved — Hearsay: Paddock and Moedano had died. Senior scholars Hristov asked had not heard it (S6); Smith says many Mexican archaeologists have, and he could neither confirm nor reject it (S8).
• ✓
  The TL age Hristov and Genovés reported (1780 ± 400 b.p.) and its cited reference were wrong; a standard TL determination could not be made on the head.

  Dating or samples · Peter Schaaf and Günther A. Wagner (the Heidelberg TL laboratory) · 2001

  Held up — Stated by the lab that ran the test; H&G's 2001 reply adopts the corrected, much wider limits (S3). The lab says the result cannot decide the head's origin.
• ?
  Even if Roman, the head could have been brought by a Spaniard and placed in an offering made in the early Colonial period; the contents and context cannot show the offering is pre-Conquest.

  Layers or association · Critics answered in H&G (2001); Michael E. Smith (c. 2003) · 2001

  Unresolved — H&G cite intact floors, untouched gold and site abandonment in AD 1510 (S6), but the site's last phase runs to AD 1530 (S4). No direct date on the offering is reported.
• ?
  García Payón kept few field notes, so the head may have entered the Calixtlahuaca collections by error after excavation; the Toluca collection includes vessels donated from other sites.

  Fraud or provenance · Michael E. Smith · 2003

  Unresolved — H&G reply that García Payón never listed any donated item as excavated, and that his 1979 account describes the context with a photo, plan and plates (S7). No inventory audit is reported.
• ?
  The Roman attribution rests on experts' statements, with no published comparison to known Roman figurines; the head may be a post-Roman European Christian figure, the first professional reaction.

  Natural, not human · Michael E. Smith (citing the initial 1960s reaction) · 2003

  Unresolved — Boehringer, Heine-Geldern and Andreae (as quoted) called it Roman, but no side-by-side comparison was found; the TL range, reaching AD 1270, also allows medieval manufacture (S2).

Evidence at the time within ~5 years of the claim

• Unambiguous evidencePartly — Certainly a human-made clay head, but its Roman (Old World) origin rested on stylistic opinion alone; the 1960 congress did not generally accept it as evidence of contact.
• Documented contextPartly — The excavator's account puts it under three intact floors of a pyramidal structure (S6), but only a two-page note was in print by 1966; Smith judges the record below professional norms.
• Independent dating methodsNo — No physical date on the head until the 1995 TL test; in 1960-66 its age rested on style alone.
• More than one labNo — No laboratory analysed the head in 1960-66; the later TL work (1995) came from one Heidelberg lab.
• Professional recoveryPartly — Dug in 1933 under José García Payón, a professional archaeologist, but in the 1960s he was often asked whether he was present at the find; he said he was (his son, per S6).
• Peer-reviewed reportNo — By 1966 only short notes were in print: García Payón (1961, 2 pp.) and Heine-Geldern (1961, 3 pp.), per H&G's references. No full technical report.
• Open to criticsPartly — The head was brought to the 1960 Vienna congress for specialists to see, but the excavation record stayed largely unpublished until García Payón's posthumous 1979 account (S8).
• Independent teamsPartly — Outside scholars judged only its style (Heine-Geldern 1961; Boehringer, date not given); no independent group re-examined the deposit or analysed the object in the 1960s.
• Outside stakeUnknown — No documented commercial, religious, nationalist or fame-seeking motive found for the 1960-61 proponents or the 1999 revivers.

What would settle it for a field partner

1. With INAH permission, re-date the head by TL or OSL at a second laboratory and publish the raw data; check feasibility first, since the 1995 test gave no standard TL determination.Medium cost
   Would show: Whether firing can be narrowed enough to separate Roman, medieval European and Mesoamerican manufacture.
2. Petrographic and chemical sourcing (e.g. INAA or pXRF) of the clay against Mediterranean and central Mexican references, plus CT or X-radiography to show if the head was moulded or hand-modelled.Medium cost
   Would show: Whether clay and forming technique match Roman-period Mediterranean workshops, later European ones, or local Mexican ones.
3. A published side-by-side comparison of the head with dated Roman and post-Roman European terracotta heads, by two or more specialists working independently.Low cost
   Would show: Whether the Roman attribution survives comparison with real comparanda rather than resting on stated opinion.
4. Archive study: García Payón's 1933 notes and 1979 texts, the MNA catalogue history behind the 'Colonial' entry, and the Toluca offering inventory (Smith, Wharton & McCarron); trace the Paddock story.Low cost
   Would show: Whether the head is securely tied to the 1933 offering or could have entered the collection later, and whether the planting story has any documentary basis.

What would change this record

Toward refutation: clay or technique matching Mexican or post-Roman European work, archives showing the head joined the collection later, or documents confirming the planting story. Toward vindication: Roman-period Mediterranean clay plus a securely pre-Conquest offering.

Notes

• Status. Open since 1960. Its proponents maintain a Roman origin; the 1960 congress was not convinced, and the site's later excavator calls the case 'weak'. TL cannot decide origin, and no retest of the head is reported since 1995.
• Accepted view then. No Old World object from a pre-Columbian context in Mesoamerica had been accepted as evidence of contact before 1492.
• Now understood. ~870 BC-AD 1270 (TL, date of firing only)
• Dating basis. Stylistic attribution by classical archaeologist Ernst Boehringer, as reported by Hristov & Genovés. Their 1999 revival added a TL age of 1780 ± 400 b.p., which the lab later called wrong (S2).
• Correction. Published figure corrected: Hristov & Genovés (1999) gave a TL age of 1780 ± 400 b.p. The lab that ran the test says that figure and its reference were wrong; its limits are 870 BC-AD 1270 (S2), adopted in H&G's reply (S3).
• Correction. Smith (S8) says García Payón published the head in 1960. H&G's reference lists (S6, S7) date his note to 1961, with 1960 as its showing at the Vienna Congress of Americanists. The 1961 note itself was not seen.
• The whole case rests on one object; H&G note that no Old World find from Mesoamerica has been accepted as proof of pre-1492 contact (S6). H&G date the burial to AD 1476-1510 but the passages read do not say how. Andreae's Severan-period attribution (c. AD 193-235) is known only as H&G quote it from Domenici 2000: 29, not located (UNVERIFIED). The draft's 'fragment of a larger figurine' detail was dropped: no source outside Wikipedia was found. H&G's two web texts differ on small details (1997 vs 1998 for their poll of senior scholars; page ranges for García Payón 1979).

People and evidence

• Proponents. Robert Heine-Geldern, Ernst Boehringer, Romeo H. Hristov, Santiago Genovés T.
• Critics. Michael E. Smith, Peter Schaaf, Günther A. Wagner, Paul Schmidt (relaying John Paddock)
• Evidence types. art, artifacts, stratigraphy, tl · 2 independent lines
• Record confidence. medium — Provenance rests on both sides' own web texts (S6-S8), read directly but not peer-reviewed; S1-S3 read as abstracts only; the 1961 notes and Andreae's quote were not seen.