On the Purported Evidence for Feasting at Pueblo Alto

Trash Mound from Pueblo Alto

Many recent interpretations of Chaco Canyon see it as a site of pilgrimage, and this is often specifically seen as taking the form of regular region-wide ritual events involving communal feasting, construction work on the massive buildings in the canyon, trade involving various mundane and exotic items, and ritual breakage of pottery and deposition of it in the mounds accompanying most great houses.  This idea, which has been incorporated into a wide range of models of Chacoan society both hierarchical and egalitarian (although it is especially important to egalitarian models), is heavily dependent on data gathered in the excavation of Pueblo Alto by the Chaco Project in the 1970s.  In addition to excavating about 10% of the great house itself, Project personnel excavated a trench and several stratigraphic columns in the large trash mound to the southeast.  What they found was a series of well-defined layers.  Some of these, toward the bottom, seemed to consist mostly of construction debris, and others, toward the top, consisted mostly of windblown sand and redeposited artifacts, but the ones in between seemed to show a pattern of large, well-defined deposits.  This was interpreted as being quite different from the expected pattern from the regular deposition of domestic trash from a residential site, and the theory developed, particularly by Wolky Toll, explained it as the result of occasional massive depositional events in which large amounts of pottery and other artifacts were deposited all at once.  Toll estimated that the number of layers approximately matched the number of years during which Pueblo Alto was occupied, and that they therefore accumulated as the result of annual events in which numerous people came from throughout the region to attend events at Pueblo Alto (and presumably at other great houses too).  As part of these events, pilgrims would probably have brought offerings of items from their home areas, thus explaining the huge amount of imported goods at Chaco as well as the lack of apparent exports.  These items would have included mundane items like wood, corn, and pottery, as well as more exotic things like turquoise and Narbona Pass chert.  People may have also worked on constructing the great houses as part of some sort of ritual offering of labor, which would explain the massive scale of these buildings despite the small permanent population of the canyon itself.  While there is a certain amount of evidence for residential use at Pueblo Alto and other great houses, it indicates a pretty small population relative to the size of the buildings, and Toll’s model interprets this as a small “caretaker” population of what were primarily non-residential, public structures with large plazas that could serve as the sites of ritual feasting and other activities during these festivals.

Furthermore, the composition of the artifact assemblage found during the excavations of the Pueblo Alto mound seemed to offer an interesting possibility for another ritual activity.  Basically, there was a huge amount of pottery in it, especially gray utility ware, much of it imported from the Chuska Mountains to the west.  Based on the number of rim sherds in the excavated portion and an estimate of the size of the whole mound, Toll calculated that 150,000 vessels were used during the 60-year period (AD 1040 to 1100) during which Gallup Black-on-white was the predominant decorated type, a period that roughly corresponds to the height of the Chaco system.  This works out to 2500 vessels a year, or 125 for each of the 20 households estimated to have lived at Pueblo Alto at any one time.  This is a huge number compared to ethnographically documented rates of pottery usage and breakage or ratios seen at small sites, and to Toll it suggested that the pottery deposited in the mound was probably not broken in the course of everyday life at Pueblo Alto but was instead broken deliberately in rituals associated with the annual pilgrimage fairs.  Ritual breakage and deposition of pottery is a known Pueblo practice, but this would be on a scale not seen at any other known site.  Nevertheless, this is the model of the formation of the Pueblo Alto mound that has been widely accepted and incorporated into a wide variety of interpretations of the Chaco system that differ wildly in many respects but all have some sort of pilgrimage function for the canyon as part of its regional role.  It’s important to note that this is the only direct evidence for a pilgrimage function known from excavations at Chaco.

Niche at Pueblo Alto

I think it’s pretty plausible that pilgrimage and communal feasting took place at Chaco, but I’ve increasingly come to think that the evidence from the Pueblo Alto mound is extremely weak.  There are a few different reasons I don’t buy it, and a couple of the most important ones are well illustrated by two articles published in American Antiquity in 2001.

One of these, by Toll himself, is part of the same special issue on the organization of production at Chaco that included Colin Renfrew‘s model of Chaco as a “Location of High Devotional Expression” or pilgrimage center.  Toll likes this idea, obviously, and his article, in addition to summarizing the known data on the production and use of pottery at Chaco, attempts to take a closer look at the Alto data to evaluate Renfrew’s model.  This basically involves looking at each of the “event layers” (as distinguished from the construction and post-occupational layers) and calculating the proportions of local and non-local ceramics and lithics (as well as the different types of pottery forms and wares) in each.  These were then compared to the proportions in the mound as a whole, other Chaco Project excavations dating from the same period, and excavations dating earlier and later.  If Renfrew’s theory works, the event layers should show higher proportions of imported ceramics and lithics, as well as possibly higher proportions of ceramic types likely to have been used in feasting, compared to these other data sets.

Flake of Narbona Pass Chert with Ant at Pueblo Alto

Toll does his best to spin the results to be consistent with Renfrew’s model, but looking at the actual numbers in his tables, there’s just nothing there.  The event layers are virtually identical to the whole mound, which isn’t really surprising given that they comprise a large portion of it, and both are very similar to contemporary non-mound contexts in most ways.  Earlier and later contexts are different in interesting ways, but that’s neither here nor there in terms of evaluating Renfrew’s model.  Chuska ceramics are a case in point: they comprise 33.4% of the event layers, 30.8% of the whole mound, and 33.1% of the contemporary non-mound contexts.  That doesn’t look like a meaningful difference to me.  In the cases where the mound layers do differ from other contemporary contexts, they generally have fewer exotic materials.  For example ceramics from the Red Mesa Valley comprise 3.3% of the event layers, 3.8% of the whole mound, and 4.6% of the contemporary non-mound contexts.  The presence of Narbona Pass chert is something of an exception, with the proportions for the event layers, the whole mound, and other contemporary contexts being 26.1%, 26.4%, and 19.5% respectively, but stone from the Zuni Mountains has proportions of 2.2%, 2.4%, and 10.9% (!) respectively, which suggests that there’s just no pattern here in which the event layers or the mound as a whole contain higher proportions of imported material.  Basically, Chaco was awash in all sorts of imported stuff during this period, and it was not particularly concentrated in the mound more than anywhere else.  The mound has lots of imports because there were lots of imports all over the place, not because it was formed as the result of annual pilgrimage feasts.

The biggest difference between the mound and other contemporary contexts comes with the forms of pottery.  Generally, the forms of pottery found at sites in this area at this time are whiteware bowls, whiteware jars, grayware jars, and redware and brownware bowls.  Red and brown bowls were long-distance imports and are found in small numbers at most sites.  The other wares were local, at least in a general sense, and while there was surely some variation, the standard idea about functions is that gray jars were used for cooking food, white bowls were used for serving food, and white jars were used for storage.  Thus, for feasting contexts an unusually high number of white bowls, and possibly gray jars, would be expected.  Since red and brown bowls are likely to have had symbolic or ritual importance, given the distances from which they were imported, they may occur in higher frequencies in feasting or ritual contexts too.

Shell Bead at Pueblo Alto

The pattern in the Pueblo Alto mound, while distinct from other contemporary sites, didn’t really match these expectations.  The most obvious difference was a much lower proportion of white bowls: 27.0% for the event layers versus 32.7% for the whole mound and 33.4% for other contexts.  This was balanced by a higher proportion of gray jars, which Toll interprets as still giving evidence for feasting, but this is mighty weak evidence for pilgrimage and feasting, even though the high proportion of grayware that came from the Chuskas during this period means that the high proportion of gray jars in the mound contributed to a higher level of Chuskan imports.  Red and brown bowls were also much rarer in the mound (both in the event layers and in the whole thing) than in other sites.

So, despite Toll’s efforts to show the data from the Pueblo Alto mound supporting his and Renfrew’s pilgrimage theories, I don’t buy it.  That’s not to say that there was no pilgrimage or feasting at Pueblo Alto, of course, just that this evidence doesn’t show that there was any.  And, remember, this is the only evidence out there for feasting and pilgrimage at Chaco.

Plaza at Pueblo Alto

But what about those unusually large, distinct layers in the mound?  Don’t those indicate unusual depositional events consistent with annual feasting and the deliberate breaking of huge numbers of pots?

Well, no.  To understand why not, we turn to the second paper published in 2001 on this topic, written by Chip Wills.  Wills is a Chaco Project alum who worked on the Pueblo Alto mound excavations, so he knows what he’s talking about here, and what he says is that the layers in the mound aren’t at all necessarily evidence for annual feasts.  Basically, what he says is that there’s nothing special about the layers in the mound.  They’re not really bigger or richer in artifacts than deposits found elsewhere.  He has a lot of specific criticisms of Toll’s interpretations and methodology, but that’s the gist of it.  He says that the unusually well-defined nature of the layers could well be the result of natural processes on layers deposited in various ways, so that it doesn’t necessarily indicate occasional large deposits rather than steady trash accumulation.  Most importantly, he finds that there isn’t actually a clear distinction between the “construction” and “event” layers, and that it’s quite plausible that the whole mound, or at least the vast majority of it, resulted from the deposition of construction debris from the various stages of construction and remodeling at Pueblo Alto.  There would have been other “depositional streams” as well, including the dismantling of earlier architecture below the present site (which is known to have been present from the excavations).  Wills doesn’t deny that ritual may have played a role in the creation of the mound, since the construction of the great house itself may well have been a ritual act, but he does deny that there is any sign that the actual contents of the mound indicate that it resulted from occasional ritual deposits rather than a combination of construction debris and regular trash dumping.

Rim Sherd at Pueblo Alto

Okay, but what about all those smashed vessels?  Basically, Wills doesn’t find Toll’s calculations convincing.  He says that Toll calculated the number of vessels based on the number of rim sherds found, then extrapolated that number to the whole mound based on the excavated portion.  The assumptions here are that each pot is represented in the mound by a single rim sherd, and that sherd density throughout the mound is constant.  Neither of these is really reasonable, although the “one rim sherd per vessel” one is particularly problematic.  It was apparently based on the fact that few rim sherds from the same vessel were found, but what are the odds that only one rim sherd from each pot made it into the mound?  Extrapolating the number of vessels is tricky, of course, and obviously the raw sherd counts can’t be a reliable way to do it (since vessels varied in size), but this rim sherd idea is questionable at best.  The idea of uniform density is really just an example of a reasonable assumption given an unknowable reality, but it’s still not necessarily right.  Wills mentions another estimate of 30,000 vessels for the whole mound, which he also attributes to Toll, and this produces much more reasonable per year and per household numbers suggesting that the observed sherd density could easily reflect regular domestic trash.  He also notes that it was the number of households, based on architectural data, that was held constant when this seemed to conflict with the number of vessels deposited, but he doesn’t elaborate on the implications of this beyond noting the privileged place architecture tends to hold in population estimates at Chacoan sites.

I originally read both of these articles a couple years ago when I was first starting to work at Chaco, and at the time I found Toll’s more convincing.  Rereading them now, though, I find Wills more convincing, and his arguments have never really been squarely addressed by Toll or anyone else associated with the pilgrimage/feasting theory (although they are occasionally mentioned in passing).  Chaco may well have been a pilgrimage site and the location of communal feasts, but it’s important to note that the Pueblo Alto trash mound doesn’t provide evidence for this idea, and neither does anything else.

Toll, H. (2001). Making and Breaking Pots in the Chaco World American Antiquity, 66 (1) DOI: 10.2307/2694318

Wills, W. (2001). Ritual and Mound Formation during the Bonito Phase in Chaco Canyon American Antiquity, 66 (3) DOI: 10.2307/2694243

Advertisement