the case of the aztec temple square - cumincad

2001: ACADIA34

SimulationSection 1

From Aztec Pictograms to Digital Media - TheCase of the Aztec Temple Square

Antonio Serrato-CombeThe University of Utah, USA

AbstractVirtual archaeology or re-creating ancient worlds digitally is not a new item.While the virtual reconstructions of Cahokia on the Mississippi or the RoyalCemetery at Ur in Iraq have provided us with glimpses of how those ancientsites might have looked, the ‘quality’ of the digital end product has been lack-ing. This is because virtual archaeology is in its infant stages. This paper makesthe point that in order to truly develop the bases of a new cognitive science,virtual archaeology has to incorporate a willingness to achieve higher digitalmodeling and rendering qualities. In other words, our ability to explore, tointerpret and to appropriately use digital tools needs to aspire to greater andmore penetrating abilities to reconstruct the past.

This paper presents the case of the digital reconstruction of the Aztec TempleSquare. This is a unique project because, unlike other sites in antiquity wherethere is a substantial amount of archaeological evidence, the Aztec site containslittle or almost no evidence. Most of what we know of the case comes fromIndian manuscripts and Spanish chronicles.

KeywordsVirtual Archaeology, Reconstruction, History

2001: ACADIA 35

Antonio Serrato-CombeFrom Aztec Pictograms to Digital Media -

The Case of the Aztec Temple Square

1 IntroductionOn or around November 1, 1519 Hernán Cortéssent ten scouts to look ahead for what was to be-come the Spanish conqueror’s last fifty miles intoTenochtitlán, the capital of the Aztecs, the larg-est Pre-Columbian empire in the Americas. Thoseexplorers came back to Cortés with stories of ‘greatcities and towers and a sea, and in the middle of ita very great city’. A week later, the Spanish armyfinally reached Iztapalapa, a town of about 15,000people just five miles from the center of the city.It is most likely that is was from this point thatEuropean eyes first saw the Aztec Temple as wellas other remarkable structures in the center of acity ranked as one of the most populated in theworld at the time. The military conquest ofTenochtitlán in 1521, was followed by the almostcomplete destruction of not only the entire Azteccity, but its cultural artifacts as well. Two hun-dred and sixty nine years went by with almost noconcern or interest for the rich cultures that ex-isted before the arrival of Cortés. By chance, in1790, while updating the layout of the centralplaza in Mexico City, several large monoliths ofPre-Columbian origin saw again the light. This

single event marked the beginning of a long pro-cess of attempting to understand and reconstructwhat the Castillian expedition so feverishly de-stroyed. These efforts culminated with theTemplo Mayor Project undertaken by the Mexi-can National Institute of Anthropology and His-tory. This project brought to light vast amountsof data on the life, customs and achievements ofthe Aztecs in the sixteenth century.

Figure 1. Aerial view of the central area in Tenochtitlán, the capital of the Aztecs

Figure 2. The Aztec Temple as illustrated in a codex drawing

2001: ACADIA36

Despite these efforts, the question of how did the TemploMayor Square really looked like is still shrouded inmystery. Unlike other great sites of antiquity inMesopotamia, Greece or Rome where today onecan still find remnants and cultural artifacts lead-ing to superb digital theoretical reconstructionsof these sites as has been documented by Forteand Siliotti in Virtual Archaeology, the case of theAztec Temple Square is unique. Cortés and hisfollowers tore down every single structure on sightand scorched every Aztec painting and manuscriptthey could find.

This paper documents and explains some of thesteps taken to reconstruct the area known as theAztec Temple Square or Precinct, and more spe-cifically, the architecture of the Aztecs in the1500’s. Included in the presentation is the use ofa variety of digital tools and techniques that as-sisted in a) analyzing historic interpretations, b)debunking myths, c) providing accurate visual-ization of structures and d) highlighting archi-tectural qualities.

2 The ProblemIn the search for solutions to the difficult prob-lem of attaining believable visualizations, it quicklybecame evident that previous attempts at recon-structing Aztec architecture suffered from:

· lack of sufficient evidence· incorrect interpretation of historic data· unsound perspective drawing techniques· cultural bias

Perhaps one of the best examples of the problemsenumerated above is the illustration of the AztecTemple Square by Cortés himself shown in Fig-ure 3. In this drawing, it is clear that what theSpanish conqueror drew was actually a picture ofa European fortified town complete with towers,minarets and draw bridges. Certainly not an ac-curate representation of what his own eyes sawwhen he actually climbed the steps of the Aztectemple in the center of the main square. LikeCortés, historians over the last one hundred yearshave also attempted to illustrate the location andform of major structures in the core of the city.These works have sometimes followed particularidiosyncrasies resulting in formal solutions witha low level of credibility. A case in point are theo-retical reconstructions done in the 1960’s wherea strong French Beaux Arts look was given toAztec architecture. In other instances, delinea-tions of the Temple Square area have suffered fromserious perspective drawing problems. In this par-ticular case the drawings have depicted Aztec ar-chitecture with building proportions that couldnot have been built due to the lack of construc-tion materials and expertise available at the time.Finally, another significant problem over the yearshas been the cultural bias. For some, Aztec de-sign was of no relevance compared to otherMesoamerican civilizations such as the Maya orInca. For others, Aztec architecture was not onlyof huge size and importance, but surpassed allother previous Pre-Columbian civilizations interms of quality and variety of execution.

3 The Role of Digital ToolsBefore any attempt to use digital tools was un-dertaken, it was deemed of the utmost importanceto plan the work in terms of understanding theunique characteristics of Pre-Columbian design.This paramount point of departure was criticalbecause the central idea behind the project wasnot simply to achieve the reconstruction and pro-duce some ‘cool’ views of Aztec architecture withthe aid of modeling and rendering software as hasbeen done by many, but to intensely study theprinciples of design behind the creation of monu-mental pyramids and shrines. At this point, it be-came clear that the Aztec Temple Square was theconcrete manifestation of a unique system of be-

Figure 3. Plan of the Aztec Temple Square by Hernan Cortés.

SimulationSection 1

2001: ACADIA 37

lief and view of the universe. The square was thecenter stage for Aztec drama. It served as the set-ting for colorful visual displays of highly ener-gized relationships between temples and moun-tains, between social groups, and between humanbeings and their gods. Also, the Aztecs understoodvery well the principles of synesthesia. The AztecTemple was a perfect example where visual atten-tion was of primary importance to the operationof drives, instincts, emotions, feelings and bodilyprocesses of human beings and thence to the sur-vival and well-being of the Aztec culture. Theconcept of seeing divine images in dramatic ap-pearances under the auspices of a powerful, au-thority was central in the orchestration of archi-tectural spaces within the precinct.

So, the questions and therefore the goals of theproject at this point focused on:

· how to best use digital tools in translating the avail-able historic sources, especially sixteenth centurychronicles, into meaningful architectural form

· how to re-interpret and de-codify ancient Aztec manu-scripts

· how to add special meaning to digital renderings be-yond the simple geometric reconstruction

Based on the above, the project was organized inthe form of illustrated case studies to create sceneswith as much meaning as possible on particular

architectural issues. The continuity in the minicase studies was not based on historic chronol-ogy, but on related architectural considerations.And, the presentation proceeded from macro scaletopics to end with building details, from largerplanning and urban design issues, to small scaledecorative motifs. The ultimate goal was to linkhistoric data with digital technologies in order tocommunicate how the Aztecs saw their world.

4 Analyzing historic interpretations - translat-ing historic sources

One of the most common pitfalls of virtual ar-chaeology comes from the fact that the person orteam developing computer models and render-ings excels in these endeavors but knows very littleabout the history he or she is trying to visualize. Acase in point are the several computer models ofthe Aztec Square developed in the 1990’s.

Figure 4 shows four models of the same pyramid-shrine to illustrate the issue. The top left image isa photograph from an Aztec clay model uncov-ered by the archaeological record. The top rightimage is a digital model based on the originalAztec clay model. The center image is a digitalmodel developed in the early 1990’s showing arather exaggerated height of the structure, with avery small shrine at the top. Here, it should benoted that while the digital model is well as-sembled, it is not conceivable that the Aztecs atthe time had the technical know-how to buildmasonry structures with the proportions shownon this model. The bottom illustration depicts amore authentic shrine, based on both the archaeo-logical record and building proportions and de-tailing derived from similar shrines in surround-ing Aztec communities. The conclusion in thiscase is clear: in the rush for digitally re-creatingthe past obvious mistakes were made. The basicgeometric configuration was incorrect. Buildingdetails and architectural styles were wrong too.Because digital reconstructions give the ‘feeling’of being real, the general public is unaware of theproblems associated with them, especially in re-lation to their veracity. A superficial inspection ofthe digital models on the Aztec Temple shown inVirtual Archaeology reveals that the ‘Aztec capital’sancient splendor has been restored by means of

Figure 4. Different versions of the same Aztec shrine, rangingfrom a clay model at the top left, to a digital model based on thearchaeological record.



2001: ACADIA38

computer processing, which has recreated theoriginal in all its clarity and color’. However, it istoo bad that every single element shown in thepublication is incorrect, from the general place-ment of buildings, to the choice of building de-tails and colors. To make the situation worse, theaccompanying text did not even mention the terms‘theoretical reconstruction’ or ‘probable massing’.It simply put forward a series of images with ex-aggerated terminology. What this means is thatdigital reconstructions of the past require highethical standards. They also necessitate a multi-disciplinary approach. Geometry, site analysis,mapping, lighting, environmental conditions, allhave to be part of a larger effort that includes thesearch for all evidence, including both the his-toric and archaeological record. In the case of theAztec Temple Square digital project, more thanfive hundred sources were consulted, rangingfrom the accounts in Latin of the fifteenth cen-tury, to the most recent archaeological excavationsby the Mexican Institute of Archaeology and His-tory. In this regard, it should be noted too thatone of the most reliable sources for the projectwas the digital reconstruction of other sites wherethe archaeological record was more profuse. Ap-

plying what was learned from those sites to theAztec capital proved to be invaluable.

5 From Aztec manuscripts to digital mediaBecause Cortés pulverized most of the physicalevidence, only two reliable historic sources wereleft to decipher the Aztec Temple Square archi-tectural puzzle: chronicles written in the sixteenthcentury and several codices dating before andimmediately after the Spanish conquest. In thelast category only very few extant manuscriptswere considered authentic and suitable for serv-ing the theoretical reconstruction efforts. Whileit is possible that codex drawings were only in-tended to be seen as pictograms with no architec-tural significance, there were enough elements inthe pictograms to begin to derive and sketch pos-sible architectural delineation. A significant prob-lem was, obviously, that Aztecs did not use thesame architectural notation we use today. So, us-ing what appeared to be a combination facade andsection representation in the codices shown inFigure 6 (Codex Cospi in particular), the explo-ration began by digitally removing all the elementsthat were not considered architectural. Then thesection part of the codex drawing was convertedto an elevation as shown in Figure 7 (typical Az-

Figure 5. Digital reconstruction of Tula, the capital of the Toltecs. This site contains a substantial amount of archaeological evidence,and was used for comparison purposes.

SimulationSection 1

2001: ACADIA 39

tec drawings represented buildings in a combina-tion elevation-section manner).

In other words, a mirror elevation was then cre-ated and placed side by side. The final result wasstill a composite drawing in that it showed ele-ments apparently drawn in section — like whitecoatings and what appeared to be structural rein-forcing rods – and elements that were clearlydrawn as a facade or elevation. This particularreconstruction revealed three volumetric ele-ments: a base, the structure itself, and a decora-tive building termination sitting on a flat roof.Structurally, it showed a post-and-beam construc-tion system. The outside of the building was prob-

ably a lime-based stucco applied on a frame. It isvery interesting to note that the interior seemedto be covered with a straw-mat type of finishplaced diagonally. As the post-and-beam struc-ture did not show any type of cross-bracing, thediagonally placed mats may have added a rein-forcing quality to the whole, preventing poten-tial collapses. It was also evident that it was notpossible to arrive at conclusions on dimensionsand proportions. However it is very likely thatthe platform was raised about 1.5 meters from theground as was done in other historic sites. In sum,this process was used throughout the project, en-listing a variety of historic sources in order to ar-rive not only at the most correct geometric rep-resentation of buildings, but to their full charac-terization with colors, textures and special details.A different approach was pursued in regards tothe placement of major structures as shown inFigure 8 illustrating the area around the mainpyramid.

Figure 6. Section of the Cospi Codex used as the basis for recon-structing prototypical elevations.

Figure 7. Composite elevation drawing derived from the Aztecpictogram in the Cospi Codex.

Figure 8. Site plan of the Aztec Temple



2001: ACADIA40

Archaeological and two-dimensional data fromthe Templo Mayor research and related projectswas first scanned with the highest level of preci-sion possible. The scans were then embossed digi-tally using a variety of contrast levels and lightingangle effects. The idea was to arrive at the opti-mal geometric description upon which a founda-tion plan could be correctly traced as shown infigure 8. Upon completion of the optimized map,the process continued by outlining the major axesand intersecting volumes at ground level with theoutline of the foundation shown in darker lines.Because archaeological data only contained tracesat the foundation level, with no information onheight, it was necessary to complete the geomet-ric construction using the accounts of chroniclersof the time. This process proved to be extremelycomplicated because unlike the Gizeh pyramids,the Aztec Temple was built following a layer pro-cess. Aztec architects did not have an establisheddefinitive plan when construction began as evi-denced in the numerous transformations of thestructure. The historical record revealed that overtime, the only constants were the orientation, ac-cess, the double staircases and the twin shrines atthe top of the temple. Based on this argument,several study digital models were developed inorder to test if what they showed matched theaccounts of the sixteenth century Spanish chroni-clers. One by one these models were discardedbecause their general configuration did not cor-respond to the accounts. At the end only twopropositions became viable. These were modeledand mapped accordingly for full rendering andvisualization. What came up from this process isa vision of the Aztec Temple that greatly differedfrom previous visualization efforts. Not only thetesting of the models produced a more accuratearchitectural representation, but the drawingswere devoid of perspective errors that had plaguedprevious visualization efforts. In the end, the vi-sualization resulted in the debunking of severalmyths including the idea that the Aztec Templesquare was surrounded by a crenellated fortifiedwall and corresponding moat. In sum, what waslearned about this particular phase of the project,is that digital modeling of historic sites cannot betaken lightly. Very serious mistakes can be maderesulting in depictions that are actually more

problematic than the ones they were supposed toimprove.

6 Virtual archaeology and quality of represen-tation

The third major component of this paper is thequestion of quality of representation. At the mostbasic level, virtual archaeology requires research-ers to make decisions on geometric construction.In most cases many gaps have to be filled withassumptions and interpolations between knowndata. Then, at a higher level, the researcher hasto make decisions on original colors and texturesin order to correctly map the geometry. In thisinstance, more suppositions have to be formulatedin an environment where the degree of specula-tion is much higher because for the most part thisinformation has long vanished. At a third level,there is the question of experiential quality. Inother words, how would special lighting and otherenvironmental factors affect the quality of visual-ization. And finally, there is the issue of correctlyrepresenting the more subtle cultural qualities thatconformed historic sites. In other words, whileone can today produce accurate geometric repre-sentations of the Bakenrenef tomb in SaqqaraEgypt or the tomb of Pakal in Palenque Mexicoit is more likely that these two sites had a clearlydifferent ‘tone’ and ‘feeling’. Or, how to illustratethe unbelievable drama that occurred at the topof the Main temple, where tenths of thousandswere believed to have had their hearts ripped outof their chests.

In order to solve these issues and the extremelydifficult challenge of making the visualizations

Figure 9. View of the platform atop the Main Aztec Temple. Itwas on this platform where countless individuals lost their lives inthe last days of the Aztec Empire.

SimulationSection 1

2001: ACADIA 41

believable, which is what virtual archaeology isall about, the process involved the breaking of thetask into several components that were consid-ered essential in defining the architectural quali-ties of Aztec architecture. Briefly, these compo-nents were: a) site planning principles, b) use ofopen spaces, c) architectural delineation of spe-cial structures, d) architectural devices, e) build-ing materials and, f) the temples at night. For eachcategory a number of modeling and renderingstudies were conducted attempting to visualizewith the highest fidelity possible the embeddedqualities. Because the length of this paper doesnot allow for the full presentation of each onethese aspects, only two items will be presented inan abbreviated form: delineation of special build-ings, the Skull Rack, and the shrines at night.

7 Skull Rack or Tzompantli AreaIn the ruins at Persepolis, the visitor can see inthe reliefs numerous accounts of human sacrifice.The royal tombs at Thebes do not leave doubtthat sacrifices actually took place. In India, hu-man skulls adorn many deities such as Bhavani.Tenochtitlán also saw human sacrifice. However,what made human sacrifice somewhat differentfrom those other civilizations, is that the Aztecsconnected human sacrifices with religious prac-tices by a ruling class desirous to claim its hege-mony and power through fear.

From a computer modeling and rendering pointof view, the challenge consisted of illustrating amost dramatic event within the context of a uniquebuilding type - a structure with thousands ofbloody skulls strung along horizontal poles.

Rather than literally illustrating such an overpow-ering edifice, it was decided to express the con-cept architecturally as a building that acted ascounterbalance or gateway to the main temple.In other words, if processions toward the templetook place, which is very likely, they must havehad to pass either through or along the sides ofthe skull rack and its bloodstained content. Be-fore worshiping at the temple, individuals werethus forced to experience the view of thousandsof skulls as proof of the tremendous power of theirrulers and accompanying system of belief. Accord-ing to the ideas admitted by the Aztecs, they werenot committing any injustice. Actually, Aztecsbelieved that once individuals were sacrificed, theybecame consecrated, that is, they belonged to thegods. Individuals were thus transformed into amystic substance. Therefore, Tzompantlis werenot just skull depositories, they were sacredshrines or temples. For the final execution of themodeling and rendering of this section of theTemple site, photographs were taken from hu-man skulls found during recent archaeologicalexcavations in Mexico City. These images werethen placed within a series of models that showedin the background several shrines west of the maintemple all with a most stunning visual result.

8 At NightDid all activity die at dusk? The Aztecs did nothave candles - these were brought to the Ameri-cas by the Spanish. However, this does not implythat the capital of the Aztec empire was a quietplace at night. Religious leaders wanderedthroughout the night lighting torches, chantingand whispering prayers. Aztec youth were sentby their preceptors to bathe in the city’s water-ways. Privileged classes probably entertainedguests throughout the night, while merchantsbrought goods from surrounding communities ontheir swift canoes. The city at night probably wasnot a grim or gloomy place; it was punctuated bythe brightness of numerous wood resin torches.Also, it was at night that the most sacred ritualstook place as the Spanish chronicled ... “In suchway flutes were played: when it was late at night,when it was close to midnight, then flutes wereplayed, then the temple ritual of bloodletting wasperformed....when it was midnight, the flutesFigure 10. View of one edge of the Aztec Skull Rack, with the

shrines on the West side of the Major Temple in the background.



2001: ACADIA42

Figure 11. The Aztec Temple Square at dusk.

Figure 12. The Aztec Temple Precinct as seen from the West Gate

SimulationSection 1

2001: ACADIA 43

played again. With this, the priesthood. awakenedand thus began what was called the Guard ofThose Who Play the Big Drum’.

Finally, it is probable that when the various de-signs and reconstructions of the Aztec templeprecinct were contemplated, its architects saw anopportunity to express and make tangible the dia-lectic that set in motion the Aztec universe. Inreality, the Aztec Temple square was like a largetemple where pyramids, dance platforms, schools,areas for the priests, and shrines were only piecesof a larger vision of the cosmos. Its form and pres-ence matched the complexity of a system of beliefdependent on gods who were symbols of mys-tery. The basic difference between the Acropolisin Athens and the Aztec Templo Mayor precinctis that in the former, Greek deities imposed upontheir following a permanent, static cult while inthe latter the religious experience was more dy-namic and ritualistic. Huitzilopochtli, Tláloc,Tlahuizcalpantehcutli and many other Aztec dei-ties only became ‘alive’ through the giving ofoneself to them. In other words, Aztecs lived anddied for their gods. In the Aztec temple project,understanding these extremely complex param-eters was the key that opened the door to higherquality virtual archaeology visualization. It wasonly after the realization such profound debt tothe gods demanded, that the final ingredient ofhigh quality visualization became a reality. Thesecomponents were tied to architectural spaces thata) were ample enough to permit the elaborate ritu-als and festivals, b) integrated sacred physical ele-ments to the design, c) facilitated magical prac-tices and rituals and d)followed religious andphilosophical constructs.

9 ConclusionsFirst, the field of virtual archaeology has entereda new phase. With current modeling and render-ing tools we are now able to generate reasonablevisualizations of ancient sites such as the AztecTemple in Mexico. Because these visualizationsare relatively easy to produce and contain a muchhigher level of information compared to olderrepresentational aids, the risk of producing im-agery that is not historically correct and there-fore misleading is much higher than ever. There-fore it is essential that we re-invent the discourse on

virtual archaeology. We need to be as knowledgeableas possible not only in the use of digital tools but in thehistoric and cultural context of the sites that are beinginvestigated.

Second, poor quality virtual archaeology visual-ization is now commonplace. We must strive notonly for more historically accurate representa-tions, but for higher rendering quality, especiallyin the area of cultural interpretation. We must re-alize that the most difficult thing to do in computermodeling and rendering is to visualize what cannot beseen, but we know exists.

Notes1) Forte Maurizio, and Alberto Siliotti, eds. (1997). Virtual

Archaeology. New York: Harry N. Abrams.

2) The Codex Cospi is at the Biblioteca Universitaria, Bo-logna. It is a skin screenfold type of manuscript consist-ing of 20 leaves (24 painted pages, 13 on obverse, 11 onreverse). It was first published in Kingsborough,Antigüedades de Mexico, Volume II.

Additional SourcesRivera, Antonieta and Wojtowicz, Jerzy, Aspects of

Tenochtitlan: Nature of CD-ROM Production in the Con-struction of Content (ACADIA 97 Conference Proceed-ings) Cincinatti, Ohio (USA) 3-5 October 1997, pp. 319-327.

Lewin, Jenniffer S. and Gross, Mark D. Resolving Archaeo-logical Site Data With 3D Computer Modeling: The Case ofCeren (ACADIA 96 Conference Proceedings) Tucson(USA) October 31- November 2, 1996, pp. 255-266.

Martini, Kirk, Ancient Structures and Modern Analysis: Inves-tigating Damage and Reconstruction at Pompeii (ACADIA97 Conference Proceedings) Cincinatti, Ohio (USA) 3-5October 1997, pp. 283-293.



the case of the aztec temple square - cumincad

Documents