THE MATERIAL ON THIS PAGE IS COPYRIGHTED AND SHOULD BE APPROPRIATELY CITED (C) 2007-2008, Mark Harlan
Dr Mark Harlan is known mostly for his groundbreaking work in bringing the personal computer to the contract work place, devising programs for
Human Systems Research to work on the first available PC (TRS-80). He is also widely acclaimed for his efforts while president of New Mexico
Archaeological Council (NMAC) in bringing NMAC and the Forest Service to the table for revised procedures for cultural resource management on land
under Forest Service jurisdiction.
His dissertation research focused on the analysis of the Chalcatzingo figurines. More recently he has returned to this work with new tools, technology,
and new concepts to arrive at an all-together-different understanding of what this assemblage represents.
However, my original data capture design did not contemplate addressing gender and the figurines give us clues to sex other than the presence of
clearly depicted breasts or pregnancies, although the clues are ambiguous. The first three specimens in the top row of the slide are unequivocal females,
having clearly depicted breasts, pregnancies or both. The next three figurines are not pregnant and appear to have less pronounced breasts, but they
would have classed as female when querying the database. The first two abdomen-leg combinations in the middle row give the impression of female
gender, with narrow waists and wider hips, while the next figurine fragment is more equivocal. None of these would class as female since they do not
show pregnancy. The final figure in the middle row seems heavily muscled but may have pendulous breasts and certainly has wide hips. Given life in a
farming community, there is no reason to discount female attributes because of the large arm muscles and my subjective impression is that this figurine
portrays an older woman. The final row is fully ambiguous, with the most difficult example presented last. Does this torso represent a particularly strong
female, with a narrow waist that may (or may not) be tapering outward to show us wide hips or is this the ideal male with powerful arms, a broad chest
and six pack abs?
Identity and Diversity in the Anthropomorphic Figurines from Chalcatzingo, Morelos, Mexico
Mark Harlan, presented at the 2008 Society for American Archaeology meeting in Vancouver, BC, Canada
When Mesoamericanists hear “Chalcatzingo,” they most often think of the famous bas reliefs. These figures invoke the supernatural more than they represent the
natural world, but the people who lived at Chalcatzingo in the Early and Middle Formative have offered us another insight into their quotidian lives. They created
numerous figurines almost certainly intended to represent themselves in miniature. My presentation today concerns the potential of these figurines to inform us
about identity within the prehistoric community.
The Chalcatzingo figurines have already received ample attention, including study specifically focused on identity in gendered and non-gendered forms. In the
early 1970s I conducted both an attribute-based analysis and a typological study. My own early efforts were severely limited by mainframe computer technology
and the level of abstraction inherent in multivariate statistical analysis, but did appear to indicate that figurine iconography changed from a uniform to a restricted
distribution over time, perhaps reflecting organizational changes in the community. That work left the content of iconography essentially unexamined. Working
alone and with Susan Gillespie, David Grove studied a subset of figurine heads, with far more attention to iconic content. Grove and Gillespie concluded that at
least some of Chalcatzingo’s figurines represent portraits of the site’s ruling elite. Ann Cyphers de Guillén later returned to the complete collection, using an
attribute-based approach closely similar to my own. Her work focused on gender and its social roles, concluding that the figurines preserve evidence of the
operation of female sodalities in the prehistoric community. Despite all this earlier attention, I remain convinced, that we have not exhausted the potential of this
important collection to inform on the nature of the prehistoric community and I have undertaken a complete reanalysis of my original data.
I have returned to my data and not the figurines themselves, since I have not been able to determine where the figurine collection resides or whether it even exists.
These data were created by observing variability on 6,149 figurine fragments. The observations characterize the manner in which the figurines depict objects such
as head coverings, body parts and – rarely – other clothing or accoutrement. This primary observation set was supplemented by recording which body parts were
present on each fragment, fragment size, and by taking a Munsell reading of the primary color. The dataset was constructed with the aim of discovering changing
patterns of interaction within the prehistoric community but contains much information that can be used to examine issues surrounding identity. For the current
effort, I have gone back to the original laboratory sheets to recapture the information and to set out in new analytical directions.
See Harlan 1979 for a clearer image
of this figure.
New directions require new approaches to the old data, thinking about the
organization of variability in the figurines themselves and using data mining
methods to work from the bottom up rather than multivariate statistical analysis
imposed from the top down. One useful approach is to envision the figurines as
partitioned loci for portraying objects. In this approach, the headdress area, head,
chest, abdomen and limbs can be seen as containers where objects are depicted.
A virtue of this conception is that it corresponds to pieces that were usually
individually modeled and appliquéd or tenoned together to create a complete
figure (although the figurines are almost never found unfragmented).
The objects depicted within the containers are mainly
body parts, although most figurines were originally
modeled with some kind of head covering, usually a
turban. Turbans can display various decorations, with
buttons most common. Many figurines also wear ear
spools of several designs and a few have some other
kind of garment. The most common garment is a simple
covering on the waist or pubic area, followed by various
types of sandals. Limbs may be shown in a range of
poses, but are seldom found attached to the trunk. It is
even rarer to see a figurine that depicts anyone doing
anything. The eighteen examples of women carrying
and sometimes nursing babies constitute the most
notable exception. Figurines sporting equipment for the
ball game also imply an activity, but there are only six
The variants on objects that may be depicted in each container are not uniformly distributed and, as shown, decline from the headgear area to the
feet. This is apparent whether one considers the variants on each object or the number of depictions of each variant.
The first point to emphasize is that figurines at Chalcatzingo are almost always found in fragmentary condition. Only 22 of the 6,149 cases recorded had a
complete set of body parts. Breakage tends to occur along the lines of attachment, so the head covering may detach from the head, or the head from the
body, the body from the trunk and so on. There are 465 theoretically possible combinations of parts, 50 of which actually occur. A reason why more of the
possible combinations do not occur is readily apparent when one considers differential survival rates. It is clear that fragments which include a head are
preserved far more frequently than fragments which do not. Actually, 2,228 of the fragments found at Chalcatzingo consist of solo heads, usually with
headgear still attached.
The fragments with heads do not survive more often because they are the most massive. Isolated chests and abdomens are each more massive than solo
heads, but far less likely to be encountered in archaeological context. Thinking in terms of iconography and considering the survival rate of each of these
“container areas” individually reinforces the picture. Arms are rare because they are easily broken off and the resulting fragments are small, but not as
small as solo turbans, which survive at a much greater rate. Heads are by far the most frequent solo piece but fourth in terms of the mass that would aid
The most parsimonious explanation for the differential survival of solo heads and fragments that include heads is that these pieces were curated before
entering archaeological context, greatly improving their survival rates. This would seem to indicate that the identity information encoded on the heads and
headgear was somehow important enough to preserve and it would be interesting to see how it co-varies with other information, such as gender. Such
efforts are hampered by the rarity of fragments that include the head and trunk of the figurine, and the even rarer preservation of pieces with limbs
attached. I have addressed this limitation by probabilistically re-associating fragments using data mining techniques, as summarized in the next slide.
The actual methodology employed a straightforward application of automated querying but is still too complex for today’s presentation. Focusing instead on the
results, we see a satisfying increase in the number of cases available to address questions concerning association of iconography coded in the Headgear and Head
containers with gender and other variability, encoded in the Chest, Abdomen and Limb containers. At this point, it is useful to consider the work of Grove and
Gillespie, since their studies have also addressed questions of identity, although they looked at only a subset of figurine heads.
Grove and Gillespie looked at the heads I had assigned to George Vaillant’s type C-8 and found a series of strikingly consistent depictions. They concluded
that these heads are miniature busts representing members of Chalcatzingo’s ruling elite. While the consistencies are indeed striking, several problems
remain. Clearly, portraiture can be no more than a component of figurine function, since C-8 figurines, while the dominant type at Chalcatzingo and more
common there than elsewhere, still represent much less than half of the total figurine heads from the site. Further other figurine types from Chalcatzingo
display much of the same iconography seen on the C-8s. These include the less finely modeled types Ch1 and Ch2 which I defined at Chalcatzingo and which
are found only there and in the surrounding region.
Besides sharing iconography with non-portrait figurines, the portraits seem to be rare. I queried my database for heads that have headgear, eyes, mouths and
ear ornaments that match examples of Grove and Gillespie’s Persons A through M. This produced only 142 hits, about two thirds of them for Persons A, B and
C. It is further interesting to note that, in my recombined data, 20 of the portrait heads are associated with chests that have breasts or abdomens that show a
pregnancy. For the present, it may be better to question the nature of personhood in the prehistoric community and tentatively identify these heads as
belonging to a unified representational style depicting a group identity. This leaves the question of whether we are dealing with groups of men, of women, or
groups that include both.
In her gender study, Ann Cyphers de Guillén asserted that 97% of the Chalcatzingo figurines represent females, although she has not detailed her basis for that
conclusion. Leaving aside for the moment questions concerning the nature of gender categorization in the prehistoric community and whether or not it was
tied to sexual dimorphism, assigning a given figurine fragment as “male” or “female is still more complex than it might seem. The first limiting factor is that
genitalia are never shown on figurines at Chalcatzingo. In my data, there are then only two objects that unequivocally inform on sex, breasts and
pregnancies. If the appropriate body parts are present it is only possible to assign a given fragment to a “female” category or to determine that sex could have
been indicated but is still unknown. Using the database of reconstructed figurines, the ratio is 42.7% female and 57.3% unknown.
Mark Harlan carried out the original study of the Formative Period Mesoamerican site of Chalcatzingo figurines as his
dissertation research, conducting field and laboratory work in 1973 and returning to the laboratory in 1976 complete his
database. Besides providing the basis for his 1975 dissertation at the University of Arizona, the data formed the basis for an
article on the development of complexity at the site, published in American Antiquity in 1979. Typological studies of the
figurine collection done in 1976 were published in 1987 in the volume Ancient Chalcatzingo edited by David C. Grove.
Subsequent to Harlan’s work, David Grove and Susan Gillespie performed limited restudies of the collection itself and of the
typological data Harlan had generated. Susan Gillespie examined the distribution of figurine variability between major areas
of the site on the basis of Harlan’s typological data as published in Ancient Chalcatzingo. Her work was intended as a
counterpoint to Harlan’s 1979 American Antiquity article, but failed to recognize figurine variability at the level of iconic
attributes, on which Harlan had based his study, cross-cuts figurine variability as captured by the traditional, whole piece
typology. Further, the patterns Harlan’s study appeared to indicate emerge through time and Gillespie, working with the
whole piece typology, had no way to achieve temporal control.
Also published in Ancient Chalcatzingo and in the journal Archaeology in 1984 Grove and Gillespie studied a subset of
figurine heads from the site, those Harlan had assigned to type C-8. They found striking similarities in headgear, eye
treatment and overall shape of the faces in some of these figurines and concluded that they are miniature busts that portray
individual rulers of ancient Chalcatzingo.
Working later in time, Ann Cyphers de Guillén restudied the entire Chalcatzingo figurine collection, using a methodology that
is closely similar to Harlan’s analysis of iconic attributes. With close attention to the figurine’s bodies and essentially ignoring
the iconography of their heads, she has posited that Chalcatzingo’s prehistoric inhabitants had female sodalities and that the
figurines functioned in ritual associated with those social groupings. Looking at both the figurines and the site’s monumental
art, she has also speculated that women may have played key political roles in the ancient community.
Mark Harlan’s 2008 presentation to the Society for American Archaeology in Vancouver, B.C., Canada is the first in a series of
reports he plans to provide on his reanalysis of his original Chalcatzingo figurine data. In it, he indicates that none of the
previous studies, including his own, have exhausted the potential of the figurines to inform on the structure of the prehistoric
community and its development from a simple farming village to a regional center.
For those familiar with the work of H.M. Wobst (1977), the
distribution of depiction density is no surprise. [Return to
Slide 7] It seems likely that the Chalcatzingo
anthropomorphic figurines depict the inhabitants of the
prehistoric community as they differentiated themselves
into groups whose identities could be readily discerned at
a distance. Before further exploring this concept, it is
necessary to consider how we encounter the figurines in
archaeological context where their fragmentary state
bedevils attempts to discern regularities. This
presentation only briefly glosses the complex area of
fragmentation and totally ignores the even more involved
topic of partibility.
This gradation of depiction density from top down is not an inescapable process of depicting the human form, as seen in these examples from the Old World.
It would seem best to discard Western concepts such as portraiture and clear gender assignment and return to the figurines themselves to seek the
information they encode. Based solely on the intensity and complexity of their use as iconic containers, the figurines’ makers intended that those who
viewed their creations instantly recognize individuals or social actors whose identities were signaled first in their headgear, next in the appearance of
their eyes and secondarily in the features depicted on their bodies. I suspect that a coherent narrative can be constructed around the identification and
distribution of social groups in the prehistoric community, and adaptation as their society morphed to complex form. It may even be that the narrative
will shape itself along the lines I first detected more than 30 years ago, but first the figurines themselves must reveal their individual stories and that will
require an enormous amount of additional data mining.
Cyphers de Guillén, Ann M
1983 Implications of Dated Monumental Art from Chalcatzingo, Morelos, Mexico, World Archaeology 13:382-393
1984 The Possible Role of Women in Formative Exchange in K Hirth (ed) Trade and Exchange in Early Mesoamerica, University of New Mexico Press, Albuquerque
1987 Las Figuras de Chalcatzingo, Morelos: Estudio de Arte y Antropología, Unpublished Ph.D. Dissertation, Universidad Nacional Autónoma de México
1993 Women, Rituals and Social Dynamics at Ancient Chalcatzingo, Latin American Antiquity 4:209-224
1998 Women, Rituals and Social Dynamics at Ancient Chalcatzingo in Kelley Hays-Gilpin and David S. Whitley (eds) Reader in Gender Archaeology, Routledge,
London and New York
Grove, David C
1987 Comments on the Site and Its Organization in David C. Grove (ed) Ancient Chalcatzingo, 420-433, University of Texas Press, Austin
Grove, David C. and Susan Gillespie
1984 Chalcatzingo's Portrait Figurines and the Cult of the Ruler, Archaeology, July-August: 27-33
Harlan, Mark E
1975 Prehistoric Exchange at Chalcatzingo, Morelos, Mexico. Ph.D. Dissertation, The University of Arizona, Tucson, University Microfilms, Ann Arbor, Michigan
1979 An inquiry into the development of complex society at Chalcatzingo, Morelos, Mexico: Methods and Results, American Antiquity 44:471-493
1987 Chalcatzingo's Formative Figurines, in David C. Grove (ed) Ancient Chalcatzingo, 252-263, University of Texas Press, Austin
1987 Description of Chalcatzingo Figurine Attributes, in David C. Grove (ed) Ancient Chalcatzingo, 491-497, University of Texas Press, Austin
Summary Statement and Background Information