This volume presents the results of the substantial development-related excavation (covering 1.58 ha) of a Romano-British villa in the northern outskirts of Bristol. While the site’s quality and complexity were not anticipated initially, nevertheless ‘the project circumstances allowed for carefully staged, research-driven fieldwork; and the post-excavation analysis could be placed within modern investigative frameworks’ (1). However, the site’s complexity (and, presumably, related financial constraints on the project) prompted removal of some of the middle parts of the sequence (below the late Roman buildings) by machine in order to access earlier features (6); the extent of such removals is not shown or quantified, so the implications for the site record and its analysis can only be guessed at. Curiously, the principal project aims set out in the preface to the report (xv) are not explicitly referenced thereafter.

The report structure is straightforward, set out in five chapters: Introduction and background (1–8), Excavation Results (9–64), The Artefacts (65–163), Biological Evidence (164–94), Discussion (195–220). The concise text is supplemented with ‘an interactive web-based GIS “Storymap”’ (8), but this is aimed at a general audience and adds nothing of substance to the main account. Illustration of the site sequence, apart from the main period plans (none of which shows the location of excavated interventions), is based almost entirely on photographs. Period shading superimposed on orthomosaic plans of buildings enhances their definition superficially, but obscures detail; complementary conventional plans would have been useful. There are only five section illustrations.

The site sequence follows a well-established pattern, developing from a series of oval enclosures probably not established before the Roman conquest (88, notwithstanding the notional period 1 date range of a.d. 1–100), through major more rectilinear reconfiguration and the first appearance (perhaps alongside a rebuilt roundhouse) of a rectangular timber building (BG, period 2.1, c. 100–250). Further development (period 2.2, c. 250–350) saw rectilinear stone-founded buildings established on the north and east sides of the central space; the latter, main ‘villa’ house a simple west-facing structure with four rooms and a cross passage. Only in period 2.3 (c. 350–410) did the site attain characteristics which in the authors’ view mean that it ‘might be termed a villa’ (37). The much-expanded house (T3, now with a portico, and wings with heated rooms and a small bath suite) was fronted to the west by two courtyards, the outer with a further agricultural building (R and W) to each side. The latest additions, on the east side of building T3, dated after c. 395; activity into the fifth century is therefore certain, but its duration is not. The relationships of post-Roman rubble layers with robber trenches and other features (fig. 2.42) were complex, so there is no simple narrative and chronology of contrasting or complementary processes of decay or active recycling of building material. There are hints of significant early medieval activity in the vicinity.

The primary discussion is broadly chronological, dealing with the character of the site in successive phases and then putting each into its local/regional context, before moving to other topics. A key issue, both here and in Chapter 2, concerns dating evidence: the aim (xv) ‘to obtain a high level of chronological precision for the structural sequence’ has in truth not been met, and chronological uncertainties imposed limitations on the contributions of the finds and environmental evidence when these could not be assigned to a specific period (e.g. 167 for the animal bone), to the detriment of better understanding of the site. What emerges beyond this? The broad sequence of site development is clear, but the detailed picture of the critical middle stages of its evolution is obscure.

There are questions about some aspects of the structures. Some (such as structure BG) are simply incomplete in plan. Period 2.3 structure R, ostensibly a straightforward aisled building, is shown in the reconstruction illustration (fig. 5.3) with aisles wrapped round all four sides of the building, but neither this nor certain other issues concerning internal elements are discussed. Late developments in the north wing of building T3 may relate to a small bath suite, but the relevant elements are only illustrated by a single oblique figure (2.31) and details of its working remain obscure. A suggestion (99) that roofing tile was brought to the site as hardcore is of interest, but perhaps rightly treated with scepticism later (207).

The broad character of the site’s economy is also clear enough. The period 2.1 animal bone assemblage was ‘dominated’ (168) by sheep/goat remains, but cattle became more important thereafter, paralleling an increase in cereal production (171), principally of spelt wheat. These are now well-established patterns, but since phased sample sizes are small the general trends, while plausible, are not very securely based. How presumed arable surplus was treated is less certain. Connections with the relatively local nucleated site of Sea Mills (inter alia perhaps as the site of a mill, though there were surely others closer at hand) are seen as particularly important. There are potential indications of such a connection in certain imported objects, notably a small copper alloy lamp (132–3 — incidentally, finds distributions (figs 3.14 and 3.15) are a welcome feature), but beyond that the links are less clear. While Sea Mills perhaps declined in importance in the later fourth century (140–2, 214), Dings took its full (‘modest’, 205) villa form at this time (caution over use of the label is commendable, but prevarication is distracting). Dings probably sat at the top of an immediately local settlement hierarchy, but relationships with the numerous neighbouring settlements (on fig. 5.5, but why not also on fig. 5.1, with relevant Margary road numbers, and the southward extension of RR541a from Wickwar to RR54?) remain to be clarified.

Overall, however, despite the difficulties with which the authors have had to contend, this report makes a significant contribution to knowledge of the diversity of the regional settlement pattern, aspects of the later Roman rural economy, and broad trends in these through time.