It has been decades since I last saw reference to Muller-Beck's 1960s notion that the first Americans had a ‘non-blade Mousteroid technology’, and later an ‘Aurignacoid blade technology’. When those terms leapt out at me from the pages of the Dry Creek volume, it was startling. But then there is something of an ‘insect preserved in amber’ quality to Dry Creek. Originally drafted in the mid 1980s, the manuscript on this important Alaskan Late Pleistocene site went unpublished. In an all too familiar story, the site's investigator, Roger Powers, got distracted by other projects (let someone without sin cast the first stone: I cannot). He fully intended to see the manuscript published when freed by retirement in 2003; that plan ended with his death just months later. But in a nice twist to the usual ill-fate of such projects: two of his one-time students, Ted Goebel and John Hoffecker, now prominent Arctic prehistorians, took it upon themselves to see the volume to publication. The first two-thirds of Dry Creek present the manuscript as it was written in the 1980s, with Goebel applying only a light editorial touch and occasional notes. That is the ‘insect in amber’ part. The last third is a chapter by Kelly Graf and colleagues based on follow-up excavations at Dry Creek in 2011, previously published in shorter form (Graf et al. Reference Graf, DiPietro, Krasinski, Gore, Smith, Culleton, Kennett and Rhode2015), and a final chapter by Goebel and Hoffecker, ‘A Dry Creek retrospective’, which delivers more than its title implies.

Dry Creek was the first site in Alaska to yield what appeared to be two distinct Late Pleistocene cultural components: Component I dated to just over 11000 radiocarbon years, with a bifacial technology that subtly hinted it might serve as a Clovis progenitor; and Component II, dated to ~10600 radiocarbon years with a predominantly microblade/microcore technology that firmly linked Dry Creek to sites in north-east Asia (there is a Late Holocene component as well, but the archaeological spotlight was fixed firmly on the earlier ones). The 1980s chapters provide a detailed description of the work, the site stratigraphy and geology, its assemblages and spatial patterning, what was learned and some of the questions that lingered: most notably, whether the difference in the technology between Components I and II was historically meaningful (different cultures) or instead marked differences in activities carried out by the “same stock of people” (p. 201). Critics wondered about that difference too, doubted it was genuine and argued instead that it might be a result of post-depositional movement or even sampling error. These are the challenges that the 2011 fieldwork sought to resolve—and did (spoiler alert: the technological differences between the components are real, and now better chronologically anchored and stratigraphically sorted, although there is still uncertainty as to what the differences mean, as Graf and colleagues note in Chapter 8).

Not all the original chapters have aged gracefully, but that is to be expected. Much has been learned of Late Pleistocene Alaska since, partly owing to later work by Powers and his students, and, of course, as methods evolve. Were the spatial analysis of the occupation floors to be done today, for example, it is likely that activity ‘clusters’ would be defined quantitatively, rather than on visual inspection supplemented by differences in raw materials (as in Chapter 5). Nonetheless, the original chapters provide important data on the formal tools (although the flakes that comprise the bulk of the nearly 35000 artefacts from the site received only limited attention, a shortcoming not addressed until decades later), and yield insights into activities that took place in each of the early components (Chapters 4 and 5). Both show evidence of tool refurbishment and manufacture, with the larger Component II assemblage—which, curiously, has clusters with bifacial points yet no microblades— marking either a wider range of activities or maybe repeated visits. Given the site's artefacts and topographic position, it was inferred that the site was a temporary hunting lookout (‘spike camp’) used on different occasions by hunters who filled their downtime retooling their weaponry and snacking on meat from recent kills (Chapters 5 and 7).

The archaeological inferences are buttressed by Dale Guthrie's analysis (Chapter 6) of the meagre faunal remains from Dry Creek—tooth scraps of Dall sheep, bison and wapiti (elk), the latter two en route to Arctic extirpation. From those teeth, but mostly drawing on his deep knowledge of Arctic animals, ecology and phenology (modern and Pleistocene), Guthrie makes a compelling case for the season of occupation (autumn and winter), its environmental context, why it was an attractive overlook for hunters, how they may have moved about the landscape, their hunting strategy (opportunistic) and even what Dry Creek might imply for the development of big game hunting on the North American Great Plains. I am not altogether willing to follow the last point, mostly because he was relying (understandably) on 1980s views of Paleoindian adaptations. No matter: this ideas-rich chapter alone justifies the price of the book.

So what are we to make of Dry Creek more than three decades on? As Goebel and Hoffecker (Chapter 9) show, the core inferences drawn in the original work about the activities that took place here have largely stood the test of time. The effort to tie the Component I bifacial points (now known as Chindadn) to Clovis came to naught, although subsequent work has shown that they may link to sites in north-east Asia, and possibly represent a population that lingered in regional refugia during the Last Glacial Maximum. Component II, although clearly distinct and now demonstrably at least 2000 years younger than Component I (with the Younger Dryas chronozone in between; Chapter 8), still remains a puzzle, at least in regard to the relationship between its microblade and biface technologies, whether these were from the same occupation on site, and how these relate to complexes of comparable age now known on both sides of the Bering land bridge (Beringia).

Along with their retrospective, Goebel and Hoffecker summarise what has been learned since of Alaskan Late Pleistocene prehistory and where Dry Creek fits in (it is still among the oldest sites), and what its components and the complexes they represent—as well as emerging genetic and genomic evidence—may suggest of the initial peopling of eastern Beringia and the Americas. Such notions will surely change with new evidence, as the authors note, but altogether it is a thoughtful and useful synthesis.

Dry Creek emerges from the amber as a well-reported, well-illustrated summary of a key Late Pleistocene Alaskan site and what it tells us of Beringian prehistory. It was worth the wait.