The question of how and why change happens in the past is central to all historical sciences, and archaeology is no exception. As this book amply demonstrates, however, the way in which change is conceptualised has itself changed over time, and these shifts have major implications for our interpretations of the past, and our ethical and political commitments in the present. Our current theoretical paradigm tacitly assumes that stasis, a lack of change, is the norm, punctuated by shorter periods of dynamic transition. Such thinking underlies what Crellin calls our block-time view of the past, in which, for example, the shift to farming signals the beginning of the Neolithic, and a long period of broadly similar existence is then transformed again by the use of metal in the Bronze Age. For Crellin, this gets things the wrong way around—it is change rather than stasis that is the norm, and the archaeologist's role is to understand the temporal and spatial scales at which this change is occurring.

The book is divided into three parts. Part one introduces the problems of approaches to change, outlining seven weaknesses of current archaeological thinking. In addition to the mixing up of change and stasis outlined above, we focus too much on progressive and teleological narratives, on revolutions and origins at the expense of messy and complex processes, and rely on deterministic and monocausal explanations, all within an anthropocentric framework that places humans at the centre of the world. Despite these weaknesses, Crellin retains an admirable belief in the political power of archaeology as the only discipline capable of providing grand narratives of the entire human past. This privileged position enables us to intervene in modern debates on issues such as migration, climate change, religion, gender and power, and in fact we have a moral responsibility to do so.

Part two examines how change has been conceptualised in the history of archaeological thought. Here the block-time approach is also in evidence in the narratives we create, with culture-history giving way to processual and then post-processual approaches, each conceptualised as a discrete grouping of scholars and ideas. Although the first chapter in this section largely retains this grouping, the next three use themes (time, scale and biography) to dissemble the traditional narrative. This is largely effective, allowing for more nuanced discussion and interesting insights. Case studies, mainly drawn from the UK and Ireland, including the author's own research on the Isle of Man, are employed judiciously in standalone sections to support the arguments made. The breadth of ideas and disciplines invoked and interwoven is impressive—for example, Chapter 4 opens with a discussion of the political implications of scales of analysis and moves on to cover the Annales school, structure and agency, new implications arising from the Bayesian analysis of radiocarbon dates and the emergence of modern humans. If this sounds like a hodgepodge, it does not feel like one to the reader.

Part three offers the solution to the problems outlined in the first two parts of the book: relational approaches to archaeology. Relational approaches have emerged over the last decade as thinkers such as Latour, Deleuze and Guattari have replaced Bourdieu, Foucault and Giddens as post-processualist lodestars. Broadly speaking, they emphasise the role of non-human agents (plants and animals, but also objects and landscapes), and hold that the properties of individuals and things are not fixed, but emerge from the relations between them. Crellin provides a nuanced discussion of the major theoretical positions and concepts grouped under this banner, including actor-network theory, symmetrical archaeology and entanglement, before outlining her own position, which draws on the Deleuzian idea of assemblage and Jane Bennett's concept of vibrant matter. Throughout this section, complex ideas are explained in clear and cogent prose, drawing out significant implications from what may appear to be subtle philosophical differences. Chapter 8 provides a case study of the approach, using a detailed examination of flat axes to comment on the emergence of the Bronze Age in Britain and Ireland, while the final chapter provides a summary and reflexive critique.

Writing about theory is a difficult task, particularly for an author committed to capturing messiness and complexity. For reasons of space, different and nuanced positions must be collapsed or elided in order to form a coherent narrative. One of the great strengths of this book is the clarity of the discussion of theoretical ideas and concepts, especially those most closely aligned to the author's own. There are inevitably gaps, especially in relation to processualism, where Binford and Schiffer do much of the heavy lifting and there is little engagement with more recent scholarship. A deeper engagement with processualist thought might also have led to more of an interest in the sorts of information that can be gleaned from the archaeological record. The book deals with the theoretical, political and ethical constraints on what can or should be said by archaeologists, but, with the exception of Bayesian approaches, there is almost no discussion of the interpretive affordance offered by the material record itself (see Perreault Reference Perreault2019).

Writing about theory is also a call to change, to follow the author(s) into their version of archaeology. So am I a convert to the new materialism? Not quite. For me the central issue, raised by Crellin herself, lies in the distinction between description and explanation. In a relational world where everything is connected, how can we establish which linkages are causal and which are not? If we cannot do this, our ability to understand why things change collapses, and we are left with a series of descriptions of what happened without any underlying explanations. I suspect, however, that this book will convert a few, with its promise of a political and ethical archaeology with room for both detail and nuance and grand narratives. And that would be no bad thing.