The upgrading of a 19 km length of the A1 trunk road in North Yorkshire generated the opportunity for an extensive programme of archaeological work between 2013 and 2017 which produced exceptional results of a kind that could not reasonably have been predicted by a research design written before the start of work. The excavations provided a transect through the walled town of Cataractonium (Catterick) and its suburbs, the roadside settlement at Bainesse 2 km further south on Dere Street, and a Late Iron Age and early Roman settlement at Scotch Corner at the junction of Dere Street and the trans-Pennine route over the Stainmore Pass towards the north-west and Carlisle. The scale of the investigations is staggering, without doubt the largest single fieldwork campaign to date in northern Britain, and one that has produced large, rich and varied artefact assemblages and better preserved human skeletal remains than are found at many sites in the North.

The sheer quantity of evidence evidently posed considerable challenges to the project team from Northern Archaeological Associates and a variety of other consultants. The decision was taken to present the findings in four online pdf volumes which in total amount to around 3,000 pages. The rationale for an exclusively online solution is said to have been ‘to reduce costs and minimise any environmental impact associated with production’ – surely more the former than the latter. Was this the correct decision? The reporting is very much in the grey literature tradition, a blow-by-blow description of virtually every feature and lengthy detailed catalogues. The format does not encourage succinctness and there is copious repetition and reiteration. The volumes lack the strong editorial hand that would have made for a tighter, more easily readable text. Very few people will have the time or motivation to read all that is presented here. Rather it is a source to be mined, and given the detail of the reporting, almost everyone will be able to find what they want in these pages. An alternative approach would have been to produce a single hard copy volume that sets out and contextualises the principal findings and provides detailed linkages with the digital archive for those that want to delve deeper. This matters, because there is a risk that the exciting results will not receive the attention or dissemination they deserve, despite their evident importance. The Scotch Corner site, in particular, should be mentioned in all future general accounts of the province. A permanent public display in the village would also have provided local legacy (little is said about public archaeology outcomes). This is not to criticise the authors or their employer. Commercial archaeology can sometimes be disparagingly characterised as digging by rote. But that is manifestly not the case here, as the enthusiasm of the authors shines through. They evidently recognised the significance of what they were finding – likely once-in-a-career highlights – and have done their utmost both in the field and in post-excavation to do justice to the evidence and make the most of it. Their willingness to engage with a wide constituency of experts and specialists should be applauded, and this has made for considered interpretation and discussion.

The most unexpected discoveries were made at Scotch Corner, where a relatively narrow excavation transect across an evidently complex site was provided with greater contextualisation by geophysical survey of the surroundings. Haselgrove provides a summation of the findings and their implications in an excellent preface. Scotch Corner was a node of a polyfocal oppidum centred on the Stanwix dykes. Occupation commenced in the pre-Roman period, probably in the period c. 55 cal. b.c.–cal. a.d. 15, and thus before the dykes, which date from c. a.d. 30/40–65/75. The presence of high-grade deposits of copper was surely instrumental in the choice of the site and the metal was mined from the very beginning of the occupation sequence. High-quality imported Roman artefacts reached Scotch Corner from c. a.d. 15, a couple of decades after their first appearance at Stanwix. Despite the exceptional finds, the plan of the undefended settlement is unremarkable, with little evidence of systematic planning. The excavation produced the fourth-largest assemblage of pellet moulds from Britain. These remarkable finds pose considerable problems of interpretation. Do they denote coin manufacture, as is commonly supposed, or might they have functioned in the production of micro bullion? If coins, then presumably some of the issues of the North-East Group (usually associated with the Corieltauvi) were in fact manufactured in Brigantia. Worthy of further consideration is a pers. comm. suggestion by Haselgrove that a certain Volisios named on some coins of the North East Group might in fact be the individual known from Tacitus as Venutius, the sometime consort of Queen Cartimandua.

A second significant discovery from Scotch Corner is the demonstration that the engineered roads in the vicinity of the site originated in the pre-Conquest period, an occurrence that has sometimes been asserted in other parts of Roman Britain but rarely demonstrated as clearly as it is here. Formalisation of the road system belongs to the period after a.d. 55, a time when occupation within the excavated part of the settlement diminished in intensity. Contrary to how we now perceive the road system on the map of Roman Britain, the original course of the Roman road from the south was across the Stainmore Pass towards the north-west, with the road towards Piercebridge and the north-east a later addition. A revival in activity at Scotch Corner c. 70–85/90 is associated by the authors with a military or official presence, and they term the settlement at this time a ‘proto town’. There is no indication as yet of a fort at Scotch Corner, but given the significance of the road junction one may yet be found, perhaps a precursor to that at Catterick established c. 85. Occupation had to all intents ceased at Scotch Corner by 135/50.

Investigation at Catterick in 28 separate trenches and other narrow strips of investigation sampled the vicus outside the Flavian fort, the subsequent defended settlement on the same site and the Brompton suburb on the north bank of the river Swale. There was no examination of the forts. The excavations have clarified the chronology of the main structural sequence at Catterick and amplified our understanding of the character of the occupation. The earliest activity was at Brompton; it dates to the early Flavian period and thus precedes the first fort of c. 85. Outside the fort, a ditch strengthens the argument for a Flavian annexe housing the baths. A substantial Trajanic ditch suggests a defended vicus that contained at least one roundhouse, a notable association, while a bridgehead defence dates to the same time. The fort is conventionally regarded as having been abandoned c. 120, although a continuing army presence in some form is likely, given that the site continued to play a role associated with military supply as demonstrated by timber granaries associated with spelt wheat and barley. Rebuilding of the mansio and bath house is dated to c. 160, and as the mansio shares a common alignment with the stone fort, the authors argue that the latter should also be dated to this time rather than the early third century, as previously argued. The settlement, which can reasonably be classed as a small town, was equipped with defences at this time that enclosed 10.5 ha on either side of the river.

The free-standing town wall dates to the early third century, not the latter part of that century as was thought. There was a fire in the early decades of the third century, and burnt deposits preserved some valuable structural insights, such as a building roofed with turves set on birch boughs. Masonry buildings were widespread from the middle of the century, although few complete or near-complete plans were recovered. A collapsed masonry wall of one building indicates an original height of 3.45 m. The north gate was partially demolished in the fourth century, although the defences continued to be maintained in some form. Coins of the House of Theodosius were rare and there is no conclusive evidence for fifth-century occupation, although the late recutting of the town ditch could suggest that parts of the circuit at least were maintained in some form. Four sunken-featured buildings are the first evidence for late fifth- or early sixth-century Anglo-Saxon occupation within the walled area, although occupation had shifted to the site of the present village by the mid sixth century.

The roadside settlement at Bainesse was only 2 km from the fort and town, but was presumably under a different system of administration and jurisdiction from Cataractonium. Established in the Flavian/Trajanic period, Bainesse lay at the navigable limit of the river Swale and perhaps served as a transshipment centre for agricultural produce brought by road or destined for it. The structural evidence suggests that parts of the settlement had gone out of use by the early to mid fourth century. But the burial evidence is instructive here. A cemetery set back from the road frontage contained 232 inhumations and 17 cremations within the area examined (its full extent is unknown). The artefacts associated with the burials suggest use from the later second to the early fourth century, but a programme of radiocarbon dating of 83 inhumations and 19 cremations coupled with Bayesian modelling was applied on a scale not hitherto reported from a cemetery in northern Britain. The scientific dating programme (‘marine corrected model’) demonstrates at 95 per cent probability that burial commenced in the period 140–210 cal. a.d. and ended between 435 and 480 cal. a.d. (six burials date fully within the fifth century on the model). Cremation occurred throughout the life of the cemetery, while stone cist graves were not used before the mid fourth century. A ring ditch perhaps denotes an intrusive, non-local, burial rite. The absolute dating programme has important implications for the traditional chronologies of many Romano-British cemeteries: burial could continue longer than the artefacts otherwise suggest. Bone preservation was good by northern British standards and Bainesse is the largest modern sample of Roman-period burials in the region. The osteological analyses are presented in detail. The authors of the cemetery volume formulated various optimistic research questions to guide their post-excavation analysis. To varying degrees, most remain unanswered.

This review has touched on only some of the findings from the project, but the wealth of data should be evident. National infrastructure projects provide unique opportunities to examine sites that would otherwise be unlikely to be developed. They also bring with them levels of funding which enable substantive campaigns of excavation and post-excavation analysis. The English Government is currently committed to further major investment in public infrastructure, and in the context of these volumes the proposed upgrade of the A66 should reveal more of the fascinating Scotch Corner site. Further surprises can realistically be expected.