In archaeology, context is everything. Or, so I’ve been told.
Archaeological “context” means at least two different things. One is depositional: what was found where, with what? That is, “context” is association. The other meaning comes from CRM: a “historic context,” according to the Secretary of the Interior, is a succinct summary of historical information informing evaluations of significance. That is, putting a site “in context.” Neither associational context or historic context are easy or straight-forward. Both are critically important. I’ll address “associational context” first, and then return to “historic contexts” later in this essay.
I have been taken to task, more than a few times, for ignoring “context” – particularly in A History of the Ancient Southwest(2009). “Context” in these critiques typically refers to associations and details that can support interpretations differing from mine, about a particular area or problem. People accuse me of overlooking their favorite details – details which, to be sure, might actually be important. You could see this coming. In A History of the Ancient Southwest, I tell a two-thousand year story in half of 250 pages (the other half is an intellectual history of Southwestern archaeology). The effort was necessarily stream-lined. I had to steer a storyline through eddies and shoals of scholarship, and I passed by many branches and bayous. Contexts (with a few exceptions in extended footnotes) were referenced by citations of reports and syntheses. I don’t present, directly, all the details – but, thus, I invite dissection in detail. Everybody has their own take on their site, or their area, or their period, which they know best. It’s the classic Kidder quandary: everyone said Kidder had a great job of synthesis his A Study of Southwestern Archaeology, except for their particular areas, which he got wrong. I’ve heard much the same about A History of the Ancient Southwest.
Recently, my friend Chip Wills, writing about his work at the key Basketmaker III site of Shabik’eschee at Chaco, critiques my use/misuse of context (W.H. Wills, F. Scott Worman, Wetherbee Dorshow and Heather Richards-Rissetto, “Shabik’eschee Village in Chaco Canyon: Beyond the Archetype,” American Antiquity vol. 77, no. 2, 20112). My principal sin, it seems, is an “…alternative interpretation that is not based in direct analysis of Shabik’eschee and relies on an archetypical mode of inference that ignores intra-site temporal and spatial variation (see especially Lekson 2009)” (Wills et alia p. 343). And, again: “… based on synchronic perspectives that ignore basic contextual evidence for occupational fluidity (e.g., Lekson 2009:67)” (Wills et alia p. 346)
In my discussion of Shabik’eschee, I was relying on Wills’s contextual evidence (presented in Wills and Windes 1989). It’s true: I use other people’s data, “pre-crunched.”* Surely that’s why data and syntheses are published. The presenters of those data have certain rights and privileges, but those do not include immunity from re-interpretation and second-guessing. Wills’s interpreted Shabik’eschee as a sort of Navajo “outfit,” repeatedly and seasonally re-occupied by a few families, possibly with a haitus when the place wasn’t occupied at all. “The most direct analog for our hypothetical settlement system is historic Navajo residential strategies in the same region.” (Wills et alia p. 343 – a point to which I will return). In A History of the Ancient Southwest, I disagreed with that interpretation. I called Shabik’eschee a village – of which, more below. I do not deny Wills’s contextual evidence (how could I?) but I interpret it differently than did he, and I am not alone: Wills lists a number of archaeologists who question the Navajo model, including several working directly with Basketmaker III – a period I know mainly from site visits, reports, and Will’s and others’ hard work. My primary reference was Paul Reed’s 2000 edited volume, Foundations of Anasazi Culture, which brought together a data and interpretations on Basketmaker III across the northern Southwest.
The key issue, as I read Wills and his colleagues, is that not all structures at Shabik’eschee were contemporary; and of course I agree (and I said so, without equivocation, on the offending page 67). I thought Shabik’eschee was a very interesting early village – small, dynamic as short-lived pit-houses must be, not a Pueblo, but still a village. By using the term “village” I lay myself open, apparently, to charges of an “archetypical mode of inference.” I would hope that I’m free of that sin; my distrust of anthropological classifications is second only to my distaste for French philosophical window-dressing. Indeed, I can claim an early interest in “occupational fluidity.” A couple of decades ago I simulated Mimbres Valley pit-house sites assuming short-term house use and high mobility – much as Wills suggests for Shabik’eschee – in contrast to the then-prevailing notion that Mimbres Valley pit-house sites were permanent “villages” (in a 1989 “gray lit” synthesis, later published as Archaeology of the Mimbres Region). After adjusting and re-adjusting my several constants, the model worked: Mimbres pit house sites could be made to seem much like Navajo outfits, fulfilling the prophecy. But Mimbres, I came to realize, was more complicated than that: (Late) Pithouse Period Mimbres sites were almost certainly permanent villages tied to and fueled by canal irrigation systems. Why then, were Mimbres villagers living in short-term pit-houses? The easy answer is that, in the Southwest, pit-houses do not necessarily indicate mobility: consider Hohokam’s centuries-long towns and villages, with pit-house domestic architecture (“houses in pits”). That answer is easy, but it ignores context – and we should never ignore context. In this case: history. Historical contexts, I will argue, are every bit as important as deposition and association – the latter two well represented by Wills, the former perhaps less well.
An example of historical contexts, building on the Mimbres case: in A History of the Ancient Southwest and in Archaeology of the Mimbres Region, I argued (i.e., interpreted) Mimbres as a player in much larger regional historical contexts, first deeply engaged with Hohokam when Hohokam was going strong (700-950 or so) and then swinging to Chaco when Chaco had its day in the sun (900-1150 or so). (Many – but not all – of my Mimbres colleagues are committed to Mimbres free of external entanglements; they are wrong.) Mimbres’s historical context might explain the misfit between ephemeral house forms and permanent settlements. Technology (for example, irrigation acquired from Hohokam) could change rapidly and irrigation entails sedentism (or vice versa); but I submit that house-form would change much less rapidly (I commend to you Amos Rapoport’s House Form and Culture, a neglected classic from an earlier age). Mimbres of the 8th – 10th centuries formed Hohokam-style permanent village, with short-lived Mogollon pit-houses. It took a while for architecture to catch-up: somewhere around 1000 (and the shift of interest to Chaco), Mimbres switched from pit-houses to masonry pueblos. Hohokam never really made that switch: pit-houses worked well through their millennium-long run, and only very late gave way to adobe compounds (but that’s another story, with other relevant contexts).
What contexts should we consult for Shabik’eschee? Let’s start with Basketmaker III first; and then think bigger both in time and space. Everyone knows Shabik’eschee is a big Basketmaker III site, maybe the biggest. “It is absolutely true that Shabik’eschee is enormous compared to most other BMIII sites in Chaco” (Wills et alia p. 343). Indeed, in Chaco there’s only one (known) contender: 29SJ423 (which Wills and his colleagues discuss), at the other end of the canyon. Taken as individual sites, Shabik’eschee and 423 have no known rivals beyond Chaco, too: the vast majority of Basketmaker III sites are one or two pit-houses (an educated guess: over 90%). A Basketmaker III site with ten pit-houses is a prodigy.
How big was Shabik’eschee? Wills says at least 60 pit-houses, plus quite a few more outside the conventional site boundaries, with many buried in canyon bottom alluvium (Wills et alia p. 332). And 423 was of comparable size. Of course the pit-houses were not all contemporary, but the size of Shabik’schee and 423 as sites commands our attention. Viewed in the context of Basketmaker III archaeology, they are phenomenal: village-stable or occupationally fluid, they are big sites. What was it about Chaco the produced sites of that size?
But wait, it gets better! Wills makes an excellent point: “…Shabik’eschee and 29SJ423 were not two separate Basketmaker sites or ‘villages,’ but rather the east and west extremes of a single settlement or community that was stretched out along the entire canyon floor between…” (Wills et alia p. 342). I like his interpretation very much, because the idea of a 15-km-long Chaco Basketmaker III “community” has featured in one of my regular AIA and public lectures for some years – indeed, since the last time Payton Manning won a Super Bowl.
To which I now turn: not the Super Bowl, but the 15-km-long Basketmaker III site, and what it might mean. Insofar as we know, there’s nothing in the northern Southwest that approaches the scale and density of Basketmaker III in Chaco. That could change: ongoing work along the flanks of the Chuska may tell another tale. But as currently known, Chaco during Basketmaker III times was by far – BY FAR! – the biggest, baddest, oddest, most interesting “site” in its time and place.
And, if we look at 15-km-long Chaco Basketmaker III in historical context, it gets really interesting. My AIA lecture was (and is) “A Millennium on the Meridian.” I revisit the Chaco Meridian and update it with new information (most of which appears in a series of footnotes to A History of the Ancient Southwest). To make a long story short, much as Basketmaker III Chaco was the biggest, baddest, etc site of its time and place, each of the succeeding Pecos System stages had one conspicuously biggest, baddest, etc sites: Basketmaker III, Shabik’eschee/423 at Chaco; Pueblo I, Ridges Basin/Blue Mesa just south of Durango; Pueblo II, Chaco Canyon (again); Pueblo III, Aztec Ruins; and Pueblo IV, Paquimé. Insofar as we know, each of those sites were uniquely large and complicated, without peers in their times and places; they clearly are the key sites for understanding those times and places; and they are all on the Chaco Meridian. As displayed below: the vertical lines represent longitudes; the site names and horizontal lines and symbols represent the sites’ east-west dimensions – for Shabik’eschee/423 about 10’ longitude or, as noted, 15 km. The Chaco Meridian began as a four point problem: north, Chaco, Aztec, and Paquimé. The first three points are now widely accepted: Chaco moved north to found Aztec Ruins. There was abundant evidence at the turn of the millennium to support that interpretation, and much new evidence from Paul Reed and his colleagues working at Salmon and Aztec makes this even more certain. The fourth point, Paquimé, is still a matter of debate – another story, told elsewhere. The realization that Shabik’eschee/423 and Ridges Basin/Blue Mesa preceded Chaco, Aztec and (perhaps) Paquimé adds momentum to the Meridian going forward – but also casts Shabik’eschee/423 in a new light, looking backward.
In my discussion of Shabik’eschee and 423 in A History of the Ancient Southwest: “I say this because we know what happened earlier in the Tucson Basin [another story: read the book]…And because we know what happened next!” Ridges Basin/Blue Mesa, and then Chaco. Historical contexts are ascritical as stratigraphy and association; indeed historical contexts provide the frameworks in which to interpret associational contexts.
But what sort of history? A History of the Ancient Southwest focused on political history, for several reasons: first, political systems – particularly flashy hierarchical systems – are easier to see than, for example, ideological systems; second, ancient North America was awash with polities and political systems; and third, because political systems in the Southwest are overgrown by a jungle of environmental data or swept under the prayer rug of ritual – it’s hard to even see ‘em through all that tangle and matting. On this blog/book, I recently despaired of Southwestern archaeology’s fixation on ritual. Elsewhere, I’ve questioned our relentless promotion of environment and climate (read the book). This is not to say that ritual and environment were not important. Of course they were! But they have been elevated to Prime Movers eclipsing all other modes of power: politics, economy, and so forth.
In the case of Shabik’eschee, for example, Wills and his colleagues propose a web of local environmental conditions to explain Chaco Basketmaker’s unseemly immensity, by suggesting that specific conditions at Chaco created bigger, denser settlements than Basketmaker III elsewhere:
“Given the episodic occupation patterns at excavated sites, we also suggest that individual households, the probable units of production, were able to shift locations in response to locally altered conditions for small-scale agricultural plots determined by the dynamically changing configuration of the floodplain associated with aggradation and channel movement.” (Wills et alia p. 342)
The usual suspect: the environment did it. To be sure, the Southwest is mostly desert and not easy to farm. But that does not make its human history a secondary derivative of rainfall regimes. Hohokam, in the 8th century, got around its environment by investing in extensive canal systems, and raised cities. Chaco, in the 11th century, organized a regional economy, administered by a city-state capital. Conversely, Mesa Verde, in the 13th century, deliberately chose NOT to develop canal irrigation, which might have avoided or alleviated the Abandonment of the Four Corners. Choices, both positive and negative, shaped history in the ancient Southwest. But these are the kinds of choices – political, economic – which we have been taught to believe were impossible for “middle-range” or “intermediate” Southwestern societies. Those societies had no political or economic histories – or, so we were told.
Where to look for appropriate models for Chaco Basketmaker III? Models from intermediate societies? Shop at home, think locally: Chaco Navajo – Wills’s Navajo model of Shabik’eschee. Again, I am not unsympathetic to the application of modern Southwestern Native groups to distant Southwestern pasts. They have the considerable advantage of being real societies operating on the same lands. In the Mimbres studies discussed above, I applied the economic geography of the eastern Chiricahua Apache to the earlier Mimbres Pithouse Period. Apache yearly rounds were intriguingly complicated (and impressively long-distance), and mapped reasonably well on the known distribution of Mimbres Pithouse Period sites. Voila! But, much like my cherished simulation of mobile pit-houses, I ultimately discarded Apache-Pithouse parallels because Mimbres Pithouse societies were clearly more complicated – economically, socially, politically – than the Apache people who roamed the same lands, five centuries later. Context was key: the two had very different historical contexts, historical trajectories.
Basketmaker-as-Navajo. Wills notes that Navajo mobility produces a cluttered landscape, in which a few families create impressive numbers of sites. (Surely sheep have something to do with this?) And the data (Archaeological Surveys of Chaco Canyon, 1981 – dated but handy) supports his view, I think. Al Hayes’s survey found 135 Basketmaker residential sites spanning three or four centuries, and 377 Navajo residential sites of a comparable span of time. Twice as many Navajo sites, but given the difficulty of seeing Basketmaker sites and the ease of seeing Navajo hogans, I think the numbers are comparable. Hayes may well have seen half or less than half of Chaco’s Basketmaker sites – a problem Hayes fully recognized, and a situation well-documented by Wills and his colleagues. Both hogans and pit-houses had short use-lives. Thus, Basketmaker III archaeology resembles Navajo archaeology, at first blush. There may be a problem: no Chaco Navajo site approached Shabik’eschee’s or 423’s size – and that’s what all the fuss is about. A few extraordinary Navajo sites outside Chaco, such as Big Bead Mesa, reach Shabik’eschee numbers of houses. Those extraordinary sites demand specific, historical explanations: Big Bead Mesa was the product of war between Navajos and Spanish. So far, no one has suggested Shabik’eschee as fortress/refuge – different contexts, indeed!
Not environmental contexts (sheep do matter!), but historical context. What is the proper context for interpreting a Southwestern site, such as Shabik’eschee? Until fairly recently, we cast our nets close. The site catchment area was popular, not so very long ago. Wills expands Shabik’eschee’s context farther, to the whole of Basketmaker III in Chaco Canyon – and that’s great! And, going even farther, he offers a useful survey of Chaco’s place in Basketmaker III in northwestern New Mexico, prefiguring new CRM in the Chuska Valley. A good start, a great start…but perhaps not quite far enough. In A History of the Ancient Southwest, I proposed that Southwestern sites could be contextualized at even bigger scales, both geographically and historically. The geographic context for southwestern sites should encompass Mesoamerica, indeed all of temperate and tropical North America: they thought big, so should we. And temporal contexts should be expanded beyond site-formation, “intra-site variation” and “occupational fluidity” to actual historical events and developments before and after the site in question.
Since the publication of that book, I’ve tried to follow my own advice, with some success. Expanding the geographic context brought a degree of resolution to the “mystery of Chaco Canyon” – the altepetl, a Mesoamerican socio-political formation from Chaco’s time, better fits Chaco’s facts than any other interpretations currently on offer. Now I’m stumbling through new methods to shift historical context from humanistic narrative (of my last book) to tools that can be applied systematically, perhaps even scientifically – with appreciative nods to path dependency and contingency and counter-factuals. I’ve come think of this method as “triangulation:” fixing the entity in question (“????” below) in historical context by viewing it from what happened before (prior), what happened after (post), and what happened in its contemporary times (peer). (My “triangulation” differs from Patrick Kirsch’s; he uses the term to describe multiple lines of evidence, an admirable strategy.)
We don’t always know priors, posts, and peers, but with Chaco we certainly know the posts and peers: after Shabik’eschee/423 came Pueblo I, Pueblo Bonito, and Chaco’s golden age. Historical contexts are as important as stratigraphy and association for understanding southwestern sites and social trajectories, particularly those which refuse to conform to our expectations for “intermediate societies.” We know what happened next! For Basketmaker III at Chaco, the historical transformation into the closest thing to a city ever seen in the Ancestral Pueblo region.
I can conjure no plausible counter-factual history that turns 18th and 19th century Chaco Navajo into their region’s center and capital – absent colonial intervention. Big Bead Mesa did not last. The Navajo resisted Spain, Mexico and the United States; Bosque Redondo forged a bitter, artificial unity; and Federal policies produced Window Rock. Left to their own devices, Navajo people would be living traditional lives which don’t include constitutions and presidents and Window Rock. Or so they tell me.
Chaco’s world was Hohokam and Mesoamerica. And Mesoamerica, at least, included governments and rulers – if not constitutions and presidents. Spain, Mexico, and the United States were imperial intrusions, but Mesoamerica was kindred civilization – of which Chaco clearly aspired to be a part (read the book). The fundamentally different historical contexts of Basketmaker III and Navajo makes one an unlikely model for the other (either way: models must be reversible), for anything beyond simple subsistence activities tied to specific conditions in the canyon. A superficial equifinality, I think.
Which brings me, in a roundabout way, to my final point: the Southwestern urge to simplify – an urge stronger than sex, sin, or belief. “Simplify” not the logic of our work, nor the structure of our data, but the very societies we purport to study. We insist – a bedrock belief – that the Southwest was a simple place, and that interpretations making ancient societies simple, therefore, are preferable to interpretations suggesting otherwise. I have (too often) noted with amusement and chagrin the righteous satisfaction going ‘round the audience, when a solid conservative speaker lays low the suggestion that Chaco, or Hohokam, or Paquimé might have something more than a Pueblo. We know what we like, and we like our ancient peoples simple.
Anthropology teaches us that Native societies north of Mexico were “intermediate” or “mid-level” — essentially people without history (or the mechanisms for history), which is why anthropology scooped ‘em up. Anthropology wants a simple Southwest, so we low-ball the past. At risk of offense, I say that’s not good scholarship or sound science; that – as I argue in A History of the Ancient Southwest – is prejudice, inherited from our anthropological forefathers. The fiats declaring the Southwest simple are so hoary and so basic that we today do not even see the dead hand of our discipline’s past. They come forward from the very beginnings of American Anthropology (read the book).
Shake ‘em off! Shake off those cold, dead hands! I submit that interpretations straying into realms – social, political, ritual – heretofore reserved for Mesoamerica are not outrageous or extreme. They are realistic assessments of Chaco: they are its historic context. And in that context, a Navajo-like Basketmaker III seems to me the extraordinary claim – not impossible, but far less likely than a lively, complicated Basketmaker III.
Wills is one of our best; his archaeology is smart, sophisticated, and perceptive. His thinking on Chaco consistently favors interpretations of a simpler ancient society than other models on offer, such as mine. His views are consistent with the data, but so are mine. I’m sure that many (most?) Southwesternists will prefer Wills’s Chaco to mine. Why? I am not asking for arguments of data, because the data support either interpretation. (Trust me on this.) I ask, why is a simple Southwest a priori preferable to a complicated, dynamic, historically contextualized Southwest?
Parsimony has no place here. “Parsimony” is a rule of thumb in logic: in explanations, all things being equal, a simpler logical structure is preferable to a more convoluted logical structure. Parsimony applies to the logic of our interpretation, NOT to the phenomenon we hope to understand. Just as nuance is the refuge of the scoundrel, no people can be great who have been subjected to parsimony. Nuance is not inherently evil of course, but too often nuance complicates for complication’s sake, making our work more complicated than it needs to be. Parsimony has its place, but not in the definition of the entity in question, the object of study, the explanandum. Mis-applied, parsimony makes things simpler than they really were. That’s no service to the ancient people we serve.
Triangulation – historical contexts – might help us avoid the procrustean bed of consensus, move us out of our comfort zone. A half-century ago, Gordon Vivian looked at Kin Kletso this way: priors, posts, and peers (within the canyon). He gave us a Chaco that overarched conventional contexts, far beyond individual sites; and he asked for historical contexts beyond the Southwest – at least, beyond Pueblo ethnography. He was right on both counts: Chaco Canyon itself is the “site” (as Wills rightly notes); and we must escape the heavy gravity of Southwestern ethnology – with its “simple” societies – to understand it. In archaeology, context is everything.
* For example, in the “pre-crunched” syntheses of Paul Reed’s Foundations of Anasazi Culture: The Basketmaker-Pueblo Transition; and two remarkable, recent books: Lisa Young and Sara Herr’s (2012) Southwestern Pithouse Communities, AD 200-900; and Richard Wilshusen, Gregson Schachner and James Allison’s (2012) Crucible of Pueblos: The Early Pueblo Period in the Northern Southwest.