Monday, September 15, 2008

What do we mean by “urban”? Comments on definitions and standards of scholarship

What do we mean by “urban”? Can archaeologists define terms like this however we please? Should journals go along with idiosyncratic definitions?

In a recent article in the journal Science, Michael Heckenberger et al (2008) use the term “urban” to describe large village settlements in the Xingu region of the Amazon from the late Pre-Columbian period. These sites conform to no published or recognized definition of “urban” that I am aware of. They are far too small to fit the demographic definition of urbanism and they do not have concentrated urban functions required by the functional definition of urbanism (for discussions of these approaches to urban definition, see Smith 1989, 2008b). I sent off a letter to the editor of Science pointing this out, but as usual, the letter was not accepted. You can see it here.

My concern here is with the issue of idiosyncratic definitions and their role in scholarship. If these sites can be considered “urban,” then that term loses much of its analytical strength for comparative analysis (their definition of urban is “multicentric networked settlement patterns, including smaller centers or towns,” p.1214). There is already a fair amount of confusion and debate about the nature of urban settlements and how to identify and study them archaeologically (Smith 1989). If we accept any interesting-looking ancient settlement system as an urban system, then that concept becomes so enlarged that is ceases to be useful. Many societies in the American Southwest, or Neolithic societies in the Old World, could be described as multi centric networked settlement patterns. These are fascinating settlement systems, but there is no reason to call them "urban." Societies can be interesting and important without being urban.

After writing my letter to Science, I came across a brief essay by V. Gordon Childe (1957), who made the same argument I made when he complained about the use of the term “urban” in the journal Antiquity to describe the Neolithic site of Jericho (Kenyon 1956).

“The use of the terms ‘civilization’, ‘city’, ‘town’, by Kenyon (Antiquity 30, 192), Wheeler (ibid,. 132-34) and the Editor himself (ibid., 129) in reference to neolithic Jericho, in no wise enhances the transcendent significance of the site. It just deprives prehistorians of convenient terms for giving expression to economic and sociological distinctions that can be recognized in the gross material data provide by dirt archaeology.” (Childe 1957:36).

Evidently we are having the same discussion fifty years later. Perhaps few archaeologists today recall the debates over the urban status of Jericho. But even so, one would hope that the reviewers and editors of Science would think twice about promoting an idiosyncratic definition of a rather common concept.

The vague and expanded concept of “urban” promoted by Heckenberger et al. does not advance scholarship, and it could even be harmful to comparative urban analysis. I am trying to promote the idea among non-archaeological urban scholars that ancient cities and urbanism are interesting and relevant to their concerns. I think archaeological scholarship on ancient cities has something to contribute to scholarly understanding of general processes of urbanization and to our understanding of modern cities (Smith 2008a). But if it looks like archaeologists are willing to call all sorts of non-urban societies “urban,” I fear that this will impede communication across disciplines and potentially restrict modern urbanists from taking ancient cities seriously.

I don’t think archaeologists should go around proposing idiosyncratic definitions of standard concepts, and I don’t think journals should promote this practice. It may be fine in postmodern scholarship, but not in empirical scientific scholarship.


Childe, V. Gordon
1957 Civilization, Cities, and Towns. Antiquity 31:36-38.

Heckenberger, Michael J., J. Christian Russell, Carlos Fausto, Joshua R. Toney, Morgan J. Schmidt, Edithe Pereira, Bruna Franchetto, and Afukaka Kuikuro
2008 Pre-Columbian Urbanism, Anthropogenic Landscapes, and the Future of the Amazon. Science 321:1214-1217.

Kenyon, Kathleen M.
1956 Jericho and its Setting in Near Eastern History. Antiquity 30:184-197.

Smith, Michael E.
1989 Cities, Towns, and Urbanism: Comment on Sanders and Webster. American Anthropologist 91:454-461.

2008a Ancient Cities: Do They Hold Lessons For the Modern World? Paper presented at the 73rd Annual Meeting, Society for American Archaeology, Vancouver, British Columbia.

2008b Aztec City-State Capitals. University Press of Florida, Gainesville.

4 comments:

JetJill said...

I'm quite glad you tried to write to Science, but I am curious as to why they rejected your letter. Any ideas? Frankly, I suspect the editors may look on archaeology as a much less rigourous science than many of the other contributing fields - despite the seemingly high incidence of errata and misconduct in, for instance, bio fields. I wonder if a general audience would consider a term like "urban" to be self-evident or mutable, such that there is no acceptable usage.

Michael E. Smith said...

A cynical explanation for their failure to publish my letter is that it makes the journal look bad - they screwed up and published an article with some very unscientific aspects (the usage of the term "urban" as well as the online supplementary materials, which are quite laughable in comparing the Xingu sites to the major cities of the ancient world).

A more likely explanation is that they didn't have space. As you suggest, I would guess they look down on archaeology as less rigorous, or at least less important, then other fields, so they simply did not want to waste space on it in their letters section.

Marv said...

First, I applaud your effort on the problem of generalizing terminology.
Second, a question: Early sites in Peru such as Caral are touted as urban. How do these fit into your definitions?

Michael E. Smith said...

From a functional perspective (which I tend to favor), Caral and other early coastal sites in Peru are clearly urban in nature. See some of my publications for more discussion of approaches to urban definitions.