Norse Greenland has been seen as a classic case of maladaptation by an inflexible temperate zone society extending into the arctic and collapse driven by climate change. This paper, however, recognizes the successful arctic adaptation achieved in Norse Greenland and argues that, although climate change had impacts, the end of Norse settlement can only be truly understood as a complex socioenvironmental system that includes local and interregional interactions operating at different geographic and temporal scales and recognizes the cultural limits to adaptation of traditional ecological knowledge. This paper is not focused on a single discovery and its implications, an approach that can encourage monocausal and environmentally deterministic emphasis to explanation, but it is the product of sustained international interdisciplinary investigations in Greenland and the rest of the North Atlantic. It is based on data acquisitions, reinterpretation of established knowledge, and a somewhat different philosophical approach to the question of collapse. We argue that the Norse Greenlanders created a flexible and successful subsistence system that responded effectively to major environmental challenges but probably fell victim to a combination of conjunctures of large-scale historic processes and vulnerabilities created by their successful prior response to climate change. Their failure was an inability to anticipate an unknowable future, an inability to broaden their traditional ecological knowledge base, and a case of being too specialized, too small, and too isolated to be able to capitalize on and compete in the new protoworld system extending into the North Atlantic in the early 15th century.
We explore developments in tephra science that consider more than chronology, using case studies of morphological transformations of tephra deposits. Volcanic processes and prevailing weather conditions determine the distribution of tephra deposits immediately after an eruption, but as these freshly fallen tephra become part of the stratigraphic record, the thickness, morphology and definition of the layers they form changes, reflecting the interplay of the tephra, climate, Earth surface processes, topography and vegetation structure, plus direct or indirect modification caused by people and animals. Once part of the stratigraphic record, there can be further diagnostic changes to the morphology of tephra layers, such as the creation of over folds by cryoturbation. Thus, tephra layers may contain proxy evidence of both past surface environments and subsurface processes. Transformations of tephra deposits can complicate the reconstruction of past volcanic processes and make the application of classical tephrochronology as pioneered by Thorarinsson (Sigurður Þórarinsson in Icelandic) challenging. However, as Thorarinsson also noted, novel sources of environmental data can exist within transformed tephra sequences that include the spread or removal of tephra, variations in layer thickness and internal structures, the nature of contact surfaces and the orientation of layers.
Recent research in ecology suggests that generic indicators, referred to as early warning signals (EWS), may occur before significant transformations, both critical and non-critical, in complex systems. Up to this point, research on EWS has largely focused on simple models and controlled experiments in ecology and climate science. When humans are considered in these arenas they are invariably seen as external sources of disturbance or management. In this article we explore ways to include societal components of socio-ecological systems directly in EWS analysis. Given the growing archaeological literature on ‘collapses,’ or transformations, in social systems, we investigate whether any early warning signals are apparent in the archaeological records of the build-up to two contemporaneous cases of social transformation in the prehistoric US Southwest, Mesa Verde and Zuni. The social transformations in these two cases differ in scope and severity, thus allowing us to explore the contexts under which warning signals may (or may not) emerge. In both cases our results show increasing variance in settlement size before the transformation, but increasing variance in social institutions only before the critical transformation in Mesa Verde. In the Zuni case, social institutions appear to have managed the process of significant social change. We conclude that variance is of broad relevance in anticipating social change, and the capacity of social institutions to mitigate transformation is critical to consider in EWS research on socio-ecological systems.