Sunday, July 10, 2011

Why build a city on a mountain?

The Aztec-period city of Calixtlahuaca covered this mountain

Here is a photo of Cerro Tenismo, the volcanic mountain that was covered by the Aztec-period city of Calixtlahuaca. I've directed an archaeological project at this site for several years, and I am still puzzling about why the founders of this city decided to build on a mountain. Some of this post is taken from an older post (of the same name) on the Calixtlahuaca Project blog.

Calixtlahuaca  covered most of the top and sides of this mountain, plus another hilly area to the southeast (to the left, in this photo). The several thousand inhabitants built their houses on stone terraces, which were also farmed with maize and maguey plants. The city was founded ca. AD 1100, and was occupied until the first couple of decades after the Spanish conquest of Mexico in 1521.

The first place to look for comparative insights into Calixtlahuaca's hilltop location is to other Mesoamerican cities built on mountains. Monte Alban and Xochicalco are two of the largest and best known examples, both powerful capitals during their day. It has long been clear to archaeologists that these cities were built where they were for reasons of defense. Images of mountaintop cities in Mesoamerican pictorial codices (see my entry on the Calixtlahuaca blog on these) tend to show battles and defensive walls. But for several reasons, we don't think that defense was a major factor in the layout of Calixtlahuaca:
  1. We did not find any defensive walls or ditches.
  2. The largest civic buildings were not built in a protected location.
view down the hill from excavation unit 323
The second factor is quite striking. The royal palace was at the base of the hill, completely unprotected, as was a large unexcavated platform (Sructure 16). The two largest temples, structure 3 (circular temple, dedicated to the wind god Ehecatl) and structure 4 (rectangular temple, dedicated to the rain god Tlaloc) were built part-way up the hill, but closer to the base. Again, these were relatively unprotected. When defense is an issue, the main civic buildings are almost always built at the top of the mountain or hill (again, think Monte Alban or Xochicalco, or any one of innumerable hilltop cities in the Old World).

Well, what is so surprising about building a city on a mountain if defense was NOT a major consideration? The answer is the effort required to build the site. Every house that was built had to be accompanied by the construction (and constant maintenance) of stone terraces. Temples 3 and 4 required massive platforms and large excavations into the hillside to build level areas for these temples and their groups.

There is much flat land in the surrounding Toluca Valley, so it would not have been hard to find a level location for the city. This was rich farmland, and we don't think populations were so high that people couldn't build their settlements on the plain. One factor that comes to mind about Calixtlahuaca's mountain location is show: the city would have looked very impressive to visitors approaching from the north, with its north flank covered with houses and large temples. But how can such a hypothesis be tested?
Ephesus

In order to gain additional perspective on Calixtlahuaca's location, I have been looking for other ancient cities around the world whose residential zones were built on mountainsides, with the civic architecture at the base of the hill. Ephesus (the Roman occupation) is one example (see photo), and I am looking for others. If you have suggestions—whether examples of similar premodern cities, or ideas about how to interpret Calixtlahuaca's location—let me know!

3 comments:

  1. Although the vertical scale is not nearly the same, this same question could be asked of every northern Mesopotamian tell settlement. As the site grows vertically, it becomes more of a challenge to supply water to households, which appears to have been done by wells to the water table rather than runoff capture that (I think) was common to Mesoamerican sites. In the Mesopotamian case, the answer (to me at least) appears to stem from the signficance of place: through time, these cities developed religious and political significance that made their continual settlement (and preferential re-settlement) likely, even in the face of logistical challenges like water supply.

    The "classic" Mesopotamian plan has the civic/ceremonial structures at center and on the highest (=oldest) part of the mound, but there are many instances were palaces were built in lower areas. My own take on this patterning assumes that the elites that constructed them were either unable or unwilling to displace the occupants of the high mounds, who we can assume had long established physical and social ties to their neighborhood, including the interment of their ancestors beneath the floors of their homes. Dislodging them would require a centralization of political power that I think would have been exceptional, rather than the rule for Mesopotamian elites (in the Bronze age, at least; the Iron Age Assyrians were a different story). Building in lower/outer areas avoided this confrontation. It also provided more abundant and flat land on which an idealized plan could be realized; building on an older and more established part of the mound would have necessitated accommodating the existing physical terrain, which had in some cases developed over millennia and was very uneven, and also the social terrain, including various constructions that could not be removed for various social, political, religious reasons etc.

    Not sure any of this is relevant to Calixtlahuaca, and I suspect political centralization under the Aztecs might have been closer to the Assyrian situation- they could move entire villages and towns for their demographic and urban planning purposes.

    Kudos on the new blog, Mike- this is my first comment but it's been in my reader since its inception.

    ReplyDelete
  2. Dr. Smith,

    Mesoamerican codices usually depict towns with the hill glyph, I know you know this, but I wonder if there isn't some practical reason for this. I know from my experience in Afghanistan that prominent terrain features are the most useful navigation tools in mountainous terrain. I have often thought about this while traveling in Mexico. For example, the site of Chalcatzingo was built at the base of a mountain that can be seen for miles. Easy to find. Also, the original civic-ceremonial center at the site I am working was established at the base of Cerro Dainzú, and the later occupation at the base of Cerro Danush. These are both prominent lone hills that are visible from long distances. There is also clearly some deeper meaning to mountains. For example, I believe that Cerro Danush played a similar role to Cerro Gordo or Tlaloc (mountain of sustenance) During the Late Classic period, a temple and elite residence was built into the summit of the mountain, and I believe that the elites were trying to assert more control over ceremonial activities. After the collapse of the Monte Albán state, however, offerings were still left on the summit of the mountain--showing its continued importance to the community. In the Late Postclassic, the settlement of the site was nucleated on the valley floor just west of the mountain, but a small shrine was constructed on the summit. A map from the 1580 Relación Geográfica shows the mountain in way overblown proportions to the towns in the area, clearly showing the community's identification with it.

    ReplyDelete
  3. @Ronald-

    I agree that mountains had symbolic significance in ancient Mesoamerica. But how does that observation help explain the setting of any individual site? First, most Mesoamerican cities/towns/villages were NOT located on mountains. So what factors determine location? In the case of Aztec-period cities, most were not on mountains. So why was Calixtlahuaca in a different setting? Second, how can one establish that the topographic placement any individual city was due to symbolic features, or, alternatively, that such symbolism was irrelevant to settlement setting? I don't know how to operationalize such concepts.

    ReplyDelete