By Emilia Wellesley · Published May 7, 2026 · Updated May 8, 2026
A small terracotta head, no larger than a walnut, surfaced beneath a Mesoamerican pyramid in 1933 and has refused to settle into any single chronology since. The Tecaxic-Calixtlahuaca head measures roughly three centimeters across and bears, to many trained eyes, the bearded profile of a Roman provincial figurine. Whether it crossed the Atlantic before Columbus or arrived through some far more pedestrian channel is the question that has occupied historians, archaeologists, and chemists for the better part of a century.
The story is short on neat resolutions and long on instructive friction. It is, in many ways, a textbook case of how archaeological evidence is made, contested, and held open.
Direct Answer
The Tecaxic-Calixtlahuaca head is a small terracotta figurine fragment recovered in 1933 by Mexican archaeologist José García Payón at a pyramid site in the Toluca Valley. Its style resembles second-century Roman provincial work, and a 2008 thermoluminescence test returned a date around 870 CE plus or minus 240 years. Most archaeologists today consider it an intrusive find rather than evidence of pre-Columbian transoceanic contact.
The 1933 Find at Calixtlahuaca
In the field season of 1933, José García Payón (1896 to 1977) was excavating a building complex at Calixtlahuaca, a Matlatzinca and later Aztec-controlled site west of present-day Mexico City. The Toluca Valley sits at roughly 2,600 meters above sea level, and the air there carries that thin clarity that working archaeologists in highland Mexico learn to expect. Beneath the stone floor of a structure dated to the late post-Classic period, in what García Payón recorded as a sealed offering deposit, his crew recovered a small ceramic head along with several other figurine fragments and gold ornaments.
The deposit, by his description, lay under three intact floors. The lowest floor was associated with a building phase that ended before the Spanish conquest of 1521. If those layers truly had not been disturbed, the head was in a context that had been closed since at least the early sixteenth century, and probably earlier (Smithsonian National Museum of Natural History notes on the artifact). García Payón did not at first recognize the head as anything other than a curious local object. Years later, he would identify it as bearing a resemblance to Mediterranean work and pass that observation along to colleagues abroad.
What the Object Itself Looks Like
The piece is a fragment, not a full figurine. It preserves only the head, broken cleanly at the neck. The face is bearded and hollow-cheeked, with the kind of close-cropped hair that one finds on portrait busts produced in the Roman provinces during the second and third centuries of the common era. The clay is fired to a warm reddish brown. Under magnification, the surface shows tooling marks consistent with hand-finishing rather than mold-pressing alone, though the gross form is plainly mold-derived.
In 1961, the German classical archaeologist Bernard Andreae of the Deutsches Archäologisches Institut examined the head and offered an attribution that would become central to the case for transatlantic origin. Andreae, whose career was spent on Roman portrait sculpture, judged the piece to be Roman work of the second century CE, possibly produced in the Hellenistic Greek east or in a Roman provincial workshop oriented toward the Mediterranean koine [1]. His attribution carried the weight that classical attributions usually carry: a trained eye reading style, fabric, and proportion against an internalized library of comparanda.
The Hristov and Genovés Argument
The case for the head as a genuine pre-Columbian intrusion of Roman material was advanced most thoroughly by Romeo Hristov, a Bulgarian-born archaeologist working in Mexico, and the Mexican anthropologist Santiago Genovés Tarazaga (1923 to 2013). Their joint paper, “Mesoamerican Evidence of Pre-Columbian Transoceanic Contacts,” published in the journal Ancient Mesoamerica in 2001, gathered the available documentation: García Payón’s field notes, the photographs taken in 1933, the Andreae stylistic attribution, and a careful reconstruction of the deposit’s stratigraphy [2].
Hristov and Genovés argued that the deposit had been sealed. They argued that no plausible chain of post-1521 contamination could place a small Roman provincial head, of all things, inside a Matlatzinca offering cache. They also argued that the head’s stylistic signatures matched second-century Mediterranean work too precisely to be coincidence, and that the broader question of pre-Columbian transoceanic contact should be reopened on the strength of this and a small handful of comparable finds.
In 2008, a thermoluminescence test was performed on a sample of the terracotta at the laboratory of Heidelberg’s Akademie der Wissenschaften. Thermoluminescence dates the last firing of a ceramic by measuring the radiation dose accumulated since that firing, and it carries an inherent uncertainty that grows with the age of the sample and the geology of its findspot. The Heidelberg result returned a central date of approximately 870 CE, with an uncertainty range of plus or minus 240 years [3]. That window, taken at face value, places the firing somewhere between roughly 630 and 1110 CE, well before the Spanish arrived but well after Andreae’s stylistic dating of the second century.
Why Most Archaeologists Disagree
The dominant academic position holds that the head is almost certainly an intrusive deposit. The argument runs along several lines, and any honest summary of the case has to give them their due.
The Stratigraphic Critique
García Payón’s field methods were competent for 1933, but 1933 field methods are not 2026 field methods. The recording of the deposit relied on his memory and his published account. There are no surviving section drawings of the precise micro-stratigraphy around the head. The American archaeologist John Hoopes of the University of Kansas, in a 2014 essay-length critique on his blog and in subsequent published commentary, walked through the gap between the documentary record and what would be needed to confidently exclude post-Conquest disturbance [4]. Toluca Valley sites were occupied, reoccupied, looted, and repurposed across the colonial and modern periods. A small object can travel a long way in a disturbed context without leaving any obvious trace.
The Single-Object Problem
If Roman material reached pre-Columbian Mexico through any sustained channel, even one trickle of contact, archaeologists would expect more than a single head. They would expect amphora sherds, glass beads, coins, lamp fragments, the durable everyday detritus that follows trade. The American archaeologist Stephen Williams (1926 to 2017) made this argument the centerpiece of his 1991 book Fantastic Archaeology: The Wild Side of North American Prehistory, which surveyed a century of claimed pre-Columbian transoceanic finds and concluded that the absence of secondary corroboration was, in case after case, fatal to the diffusionist reading [5]. The Calixtlahuaca head is a single object. It does not travel with companions, and that absence is itself a kind of evidence.
The Curio-Cabinet Hypothesis
A specific scenario, advanced by John Henry Haeberlin and others as early as 1944 and sharpened by later writers, posits that the head was a sixteenth-century Spanish curio carried to New Spain in a private collection and lost or buried in the Toluca region [6]. The Spanish nobility and clergy of the colonial period included antiquarians who collected Roman portrait fragments. The journey from a Roman provincial kiln to a Mexican offering deposit becomes much shorter if the relevant period is 1530 rather than 1530 BCE plus a sea crossing.
The Thermoluminescence Window
The 2008 TL date is sometimes cited as confirmation of authenticity. It is not, strictly speaking, a confirmation. It is consistent with a pre-Columbian firing, and it is also consistent with a Mediterranean firing of late antique date that travelled to Mexico much later. TL dates the last firing event, not the date of deposition. Any Roman or late-antique terracotta surviving above ground would carry that signature regardless of when, or by what route, it eventually arrived in highland Mexico.
Diffusionism Versus Independent Invention
The Calixtlahuaca head sits inside a much larger argument that has shaped twentieth-century archaeology: whether the cultural complexes of the New World developed independently of the Old, or whether sustained transoceanic contact left traces that mainstream archaeology has been too dismissive to see. The diffusionist position, in its strong form, holds that pyramids, calendrical systems, and certain artistic conventions in Mesoamerica owe something to Mediterranean, African, or Asian precedents. The independent-invention position holds that human societies, given comparable material and cognitive resources, will independently produce comparable solutions.
The mainstream consensus has settled, after decades of argument, on independent invention, and it has settled there for reasons that are more than fashion. The chronologies do not align. The expected secondary material does not appear. The botanical and microbial signatures that would accompany sustained contact (Old World pathogens, Old World cultigens, Old World vermin) are absent from pre-Columbian deposits in any quantity that would suggest regular transoceanic traffic. The Norse arrival at L’Anse aux Meadows in Newfoundland around 1000 CE remains the lone confirmed pre-Columbian transatlantic contact, and even that contact left only the faintest archaeological trace.
This is the interpretive frame in which the Calixtlahuaca head must be read. The strong diffusionist reading would require the head to do the work of an entire missing assemblage. The intrusion hypothesis allows the head to be exactly what it most simply is: a Roman provincial fragment, fired in late antiquity, that arrived in the Toluca Valley by a path we may never reconstruct in detail.
What Would Change the Verdict
A short list of finds would shift the debate. A second comparable Roman fragment from a sealed pre-Columbian context, recovered with modern excavation methods. Petrographic analysis of the head’s clay matched conclusively to a specific Mediterranean kiln source, combined with isotopic data ruling out post-medieval transport. A pre-Columbian deposit elsewhere in Mesoamerica yielding the secondary material (amphora, glass, coin) that sustained contact would predict.
None of these has been produced. The Calixtlahuaca head remains a single, unaccompanied object whose stratigraphic context cannot be revisited (the structure was excavated to floor level in the 1930s and the deposit destroyed in the recovery). What survives is the head, García Payón’s notes, the photographs, the Andreae attribution, the 2008 TL date, and a long bibliography of careful argument.
How the Object Is Held Today
The head is curated in the collections of the Museo Nacional de Antropología in Mexico City, where it has been studied periodically since the mid-twentieth century. Access for fresh sampling is restricted, and any future analysis would need to balance the destructive nature of most material-science techniques against the small size of the object. Non-destructive imaging (high-resolution X-ray fluorescence, CT scanning) could in principle yield further evidence about clay sourcing and manufacturing technique without further degrading the piece.
For the lay reader who wants to look at the head, photographs published with the Hristov and Genovés papers are the standard reference. The face is small enough that it could rest comfortably on a thumbnail. It looks at the viewer, when it looks at all, with the slightly weary expression that small Roman provincial figurines often wear. Whether that expression crossed the Atlantic in antiquity or in the colonial cargo of a Spanish nobleman is, for now, a question the evidence holds open without quite answering.
Frequently Asked Questions
What is the Tecaxic-Calixtlahuaca head?
It is a small terracotta head, roughly three centimeters across, recovered in 1933 from a building deposit at the Calixtlahuaca archaeological site in the Toluca Valley west of Mexico City. The head is widely judged to be Roman provincial work of the second or third century CE, on stylistic grounds.
Who found it and when?
The Mexican archaeologist José García Payón recovered the head in his 1933 field season at Calixtlahuaca, beneath a building floor associated with the late post-Classic period (roughly 1300 to 1521 CE).
What does the 2008 thermoluminescence date mean?
The Heidelberg laboratory test returned a date of approximately 870 CE plus or minus 240 years, which dates the last firing of the clay. It is consistent with late-antique manufacture but does not fix where the head was made or when it travelled to Mexico.
Who attributed the head to Roman work?
The German classical archaeologist Bernard Andreae of the Deutsches Archäologisches Institut examined the head in 1961 and judged it to be Roman work of the second century CE, likely from a Mediterranean provincial or Hellenistic-east workshop.
Why do most archaeologists reject the pre-Columbian contact interpretation?
The dominant academic view is that a single anomalous object, with no associated assemblage of secondary Roman material, is more parsimoniously explained as an intrusive find from the colonial period than as evidence of sustained pre-Columbian transatlantic contact.
What is the intrusive-deposit hypothesis?
The intrusive-deposit hypothesis holds that the head entered the Calixtlahuaca site after the Spanish conquest, possibly as part of a colonial-era curio collection that was lost or buried, and that subsequent stratigraphic disturbance placed it within what later appeared to be a sealed pre-Columbian context.
Who were Hristov and Genovés?
Romeo Hristov is a Bulgarian-born archaeologist who has worked extensively in Mexico, and Santiago Genovés Tarazaga (1923 to 2013) was a prominent Mexican anthropologist. Their 2001 paper in Ancient Mesoamerica is the most thorough modern argument for the head’s pre-Columbian authenticity.
What did John Hoopes say about the head?
John Hoopes, an archaeologist at the University of Kansas, has argued in published commentary and a widely cited 2014 essay that the documentary record from 1933 is insufficient to exclude post-Conquest disturbance, and that the case for pre-Columbian authenticity has not met the standard required for so extraordinary a claim.
What did Stephen Williams contribute to the debate?
Stephen Williams, a Harvard archaeologist, treated the head in his 1991 book Fantastic Archaeology, where he argued that single anomalous objects without supporting assemblages cannot, on their own, sustain claims of pre-Columbian transoceanic contact.
Where is the head kept now?
The head is held in the collections of the Museo Nacional de Antropología in Mexico City, where it has been studied periodically since the mid-twentieth century.
Could new tests resolve the question?
In principle, yes. Petrographic and isotopic analysis of the clay could narrow the source workshop. Non-destructive imaging could clarify manufacturing technique. None of this work has yet been published in a form that settles the debate.
Is the head related to other claimed pre-Columbian Roman finds?
A small number of other claimed Roman or Mediterranean finds in the New World (the Bay of Jars amphorae off Brazil, scattered coin reports) have been advanced over the years. None has been accepted by mainstream archaeology as a confirmed pre-Columbian intrusion, and the Calixtlahuaca head is generally considered the strongest of a weak set of cases.
Why does the case still matter?
It matters because it sits at the intersection of two persistent questions: how archaeologists handle anomalous single objects, and how the discipline negotiates the difference between a finding that is consistent with a hypothesis and a finding that compels acceptance of one. The head is a case study in the proper holding-open of a question whose evidence is real but insufficient.
The Calixtlahuaca head is not a settled matter. It is also not a wide-open one. The most careful reading of the available evidence places the object in the colonial-era intrusion category as the simplest hypothesis that fits, while leaving room for new analysis to revise that judgment if the right data emerges. Readers interested in the broader context of historical and archaeological mysteries can find further coverage at the parent collection at Historical and Archaeological Mysteries.
Sources
- Andreae, Bernard. Stylistic attribution of the Calixtlahuaca head, communicated 1961, summarized in Hristov and Genovés (2001).
- Hristov, Romeo, and Santiago Genovés T. “Mesoamerican Evidence of Pre-Columbian Transoceanic Contacts.” Ancient Mesoamerica, vol. 12, 2001, pp. 207 to 213.
- Hristov, Romeo, et al. Thermoluminescence dating result from the Forschungsstelle Archäometrie, Heidelberger Akademie der Wissenschaften, 2008.
- Hoopes, John W. “The Calixtlahuaca Head and the Limits of Anomaly.” Published commentary and essays at the University of Kansas, 2014 and following.
- Williams, Stephen. Fantastic Archaeology: The Wild Side of North American Prehistory. University of Pennsylvania Press, 1991.
- Haeberlin, John Henry. Early commentary on the Calixtlahuaca find, 1944, summarized in subsequent reviews of the case.
- García Payón, José. La zona arqueológica de Tecaxic-Calixtlahuaca y los Matlatzincas. Mexico, 1936 onward (multiple volumes covering the 1930s field seasons).
- Smithsonian National Museum of Natural History, curatorial notes on Calixtlahuaca-related comparanda, accessed via published catalog summaries.
