Twenty years ago Barbara Voorhies visited Paso de la Amada, Mexico, where I was working on my first Mesoamerican field project. It was the winter of 1995 and Voorhies led me and the other student excavation supervisors on an excursion into the Soconusco estuary of southern Mexico to visit the Tlacuachero shellmound, where she had been working since 1973. It was then—and still is—the best documented Late Archaic-period (5500–4000 BP) site from Mesoamerica's tropical lowlands. Scotty MacNeish in Tehuacán and Kent Flannery in Oaxaca had each established similar baseline evidence of the Archaic period from the dry highlands of Mesoamerica in the 1950s and 1960s. Subsequent to the pioneering work of these three scholars, however, our understanding of Mesoamerica's Archaic period has developed surprisingly slowly despite its tremendous comparative significance for those who study the Neolithic worldwide (see Rosenswig Reference Rosenswig2015 for an extended review).

MacNeish and Flannery each assembled a team of specialists and produced what remain canonical publications on the Mesoamerican Archaic. The work of Voorhies differs in that she has continued to collect further data with a new field season at Tlacuachero in each of the past four decades. The monograph reviewed here reports her most recent work in 2009. With the help of nine collaborators, the floral, faunal and lithic data and radiocarbon dates are presented, as well as detailed spatial patterns and a reconsideration of formation processes. Eleven chapters are presented on high-quality paper in the UCLA Cotsen Institute of Archaeology's monograph series. Large photographs and drawings effectively illustrate the excavations at Tlacuachero, and I particularly appreciate the 29 plan maps of the floors indicating a variety of artefact and ecofact distributions. This is an easy-to-read monograph with the basic archaeological data clearly framed in terms of their anthropological significance.

Tlacuachero is a 125m-long and 7m-high shellmound within the mangrove estuary of the Soconusco region on the Pacific coast of Chiapas, Mexico. It consists of an upper ceramic-bearing component of chronologically mixed deposits (Chapter 3). A lower component of unconsolidated bedded shells is documented down to the water-level, where it is dated to 5000 cal BP. The key elements of this volume are occupation floors located 4.6m below the mound summit and newly dated to between 4935 and 4350 cal BP (Chapter 2). The 2009 investigations uncovered 33m² of these floors, bringing the total exposed to 120m². Lest the significance of this site be unclear, it contains the only Archaic-period living surface from which horizontal patterning is published from anywhere in lowland Mesoamerica. Originally interpreted as a single floor, after four seasons of work, three successive floors are now recognised at the central and highest part of the mound. The floors were laid down sequentially but offset so that their edges do not correspond.

One might reasonably ask whether a single season of fieldwork merits publication, especially as Tlacuachero has already been excavated three times and the results extensively published. The answer to this question is yes, and there are four specific reasons why this is so. First, chemical analysis of the floors themselves (Chapter 5) document that they were not built of clay transported to the site (as previously believed), but are actually carbonate hydroxyapatite that developed from wood ash and ground-up shell from activities at the mound. Second, formation and use of the three floors occurred over 600 years (Chapter 2), and each floor was used for much longer than previously thought (p. 179). Third, the spatial arrangement of features on floors 1 and 2 (Chapter 4), as well as their association with micro-refuse (Chapter 6) and phytoliths (Chapter 7), demonstrates that the architecture was not from a residence, as previously posited (p. 57). Instead, postholes are now interpreted to represent a wall-less sun shelter, drying racks, work tables and other temporary features (Chapter 10).

The fourth contribution of this monograph is interpretive and emphasises the motivation for constructing shellmounds (Chapter 11). The final chapter is a response to a yet-to-be-published paper by John Clark and John Hodgson as to whether Archaic-period shellmounds in the Soconusco estuary were intentionally constructed platforms rather than refuse heaps, and Voorhies has long-maintained the latter. She dubs these positions “constructionalist” vs “accretionalist”, and observes that a similar disagreement has arisen wherever large shellmounds are documented (p. 195). Voorhies reviews four New World instances of this disagreement from north-west Mexico, south-east USA, central California and south-east Brazil. In each case, archaeologists had first assumed accretionalist interpretations of shellmounds to be refuse piles; more recent investigators have subsequently proposed constructionalist interpretations: purposely built temples, burial mounds or other sorts of ceremonial constructions. This shift in interpretation parallels the earlier culture-historical and processual proclivity for functional explanation, followed by a more recent penchant for symbolic and spiritual understandings.

Voorhies effectively demonstrates that such opposed interpretations are similar in each region. But they represent a false dichotomy. Practical and ceremonial functions often exist simultaneously. For example, water-filled trenches can serve as both defensive moats and irrigation canals, and city walls generally function defensively to ward off attack as well as materialising social and political distinctions based on habitation within or beyond their circuits. Shellmounds likewise often accrete or develop as food-processing middens, but their placement, orientation and shape can also serve ceremonial purposes. Further, their meaning will change if burials or other ritual features are later added. The crux of the disagreement is whether the meat from the shells was eaten as they were piled up or whether they were gathered solely as construction material and served no nutritional purpose. The careful excavation, detailed reporting of stratigraphy and consideration of formation processes presented in this monograph demonstrate that at Tlacuachero the clam meat was eaten. For other regions, this is an empirical question that can be evaluated in the manner clearly laid out by Voorhies and her colleagues in this exemplary monograph.