The widespread notion that the fundamental transition in the human trajectory from tiny, mobile populations of hunter-gatherers to sedentary, politically complex ‘civilisation’ took place through a Neolithic Revolution, foisted upon us by dramatic climate change towards the end of the Pleistocene, has long invited scepticism within archaeology. Critiques are various and diverse, but two are perhaps particularly telling. The first arises from the biological perspective that agriculture is but the visible end-point of a protracted co-evolutionary process between humans and parts of their environment, meaning that understanding this defining transformation in human ecology lies not in the Neolithic itself but far earlier, during its long prelude.

A second objection, meanwhile, follows from the recognition that institutionalised political complexity can emerge in societies without agriculture. First invoked by ethnologies of complex hunter-gatherers along the Pacific coast of North America, these have often entailed the exploitation of rich, sustainable, easily seen and harvested marine and aquatic resources. Archaeological embodiments of this are now envisaged worldwide, but an early, oft-cited exemplar was Michael Moseley's (Reference Moseley1975) Maritime foundations of Andean civilization, which promulgated the idea that on the coast of Peru complex society had first been sustained not by food crops, but by marine resources. Here monumental civilisation flowered during the third millennium BC, epitomised by the monumental site of Caral, Norte Chico, today reconstructed and marketed as the ‘first city in the Americas’. Caral was neither a city, nor the first, but regardless, it represents only the Late Preceramic culmination of a long process of transformation of human society and environments in the wider Andean region, the understanding of which must therefore be sought far earlier, in its Middle Preceramic (aka Middle Archaic) prelude between around 8000 and 5000 BP. Yet while that prelude is known through diverse sites along the Peruvian coast and in the Andean highlands (many studied decades ago, only reported in synthesis and now destroyed by development), few have been studied in the hinterlands between the littoral and high Andes.

Into this relative lacuna then comes Pernil Alto, the latest output of the German Archaeological Institute's two-decade Nasca-Palpa Archaeological Project directed by Marcus Reindel and Johny Isla, about a Middle Archaic site in the Andean foothills of the Río Grande Valley, southern Peru. Such sites in such locations are extremely rare, as its author Hermann Gorbahn notes in his introduction, because they have been obscured by geomorphological changes and later agriculture. Pernil Alto was only fortuitously discovered through investigation of an overlying Initial Period site. What is almost as rare, moreover, is the publication of any such site in such meticulous and beautiful detail.

Pernil Alto comprises 21 chapters falling into three sections. Five introductory chapters outline the methods used to derive different lines of evidence, review the Archaic Period state of the art and describe the site and its putative palaeoenvironmental context. Thereafter six chapters set out the results of excavations and analyses. These are quite exemplary presentations. Detailed stratigraphies recorded according to a Harris matrix and then verified and rectified through the calibration and modelling of 70 radiocarbon dates as a sequence of phases provide a coherent chronostratigraphic framework: basic archaeological information without which findings have, at best, to be taken on faith, or, at worst, have no meaning, and yet which are all too rarely presented with the clarity and precision shown here. Detailed descriptions of those findings follow in the forms of architecture, burials, artefacts and plant and animal remains, including wooden and fabric artefacts preserved over five millennia in the region's arid climate. Illustrations are beautiful, often in colour. Indeed, Pernil Alto includes some remarkable findings: a carved camelid figurine, surely among the oldest unequivocally dated art objects in South America; the oldest securely dated and identified sweet potato remains anywhere; and extraordinary evidence for the early use of red and yellow mineral pigments.

Through eight chapters of the final section, Gorbahn, leveraging the resources and scope of the Nasca-Palpa Archaeological Project, marshals an extraordinary breadth of evidence, including, inter alia, using artefacts as proxies for different activities, quantifications of plant micro and macrofossil remains, strontium isotopes in human remains and palaeodemography. This evidence offers a compelling and detailed picture of settlement, economy and society in one of the earliest permanently occupied villages in the Andes towards the end of the fifth millennium BP. One can cavil at some attempted quantifications of small samples, but data throughout are set within precise and transparent chronostratigraphic contexts. Gorbahn shows how the occupants of Pernil Alto built and organised their dwellings, buried their dead within them after abandonment, and subsisted by floodplain farming of beans, sweet potato, pumpkin, achira and guava, supplemented by gathered tree legume pods and some hunting of birds, guanaco, deer, guinea pig and freshwater shrimp. Here, some 60km from the sea, marine resources could only ever be incidental to subsistence.

Permanent villages with similar material cultures and horticultural assemblages had flourished somewhat earlier around 6000 BP along the Pacific coast, including nearby at Santa Ana and La Yerba III on the estuaries of the Río Grande and Río Ica, respectively, and at Paloma and Chilca I in the Chilca Valley. Yet those villages’ diets, as directly reflected in their inhabitants’ isotopic signatures, were dominated by marine resources. One of Gorbhan's many contributions in Pernil Alto therefore is to offer rare insight into how, during the fifth millennium BP, settlement began to move inland, alongside increasing farming of food crops. On the south coast these changes did not culminate in the precocious developments of the Late Preceramic farther north: a stark contrast that had nothing to do with marine resources, but rather with differences in how exploitation of those marine resources could be intensified through an agriculture of cotton for fishing nets in close proximity to the Pacific Ocean (just as Moseley originally envisaged). Pernil Alto is indispensable reading for scholars of the Andean Archaic and indeed anyone interested in the wider story of transformation from human cultures dependent on environments, to environments dependent on human cultures.