Ballin-Smith, B., Carter, S. P., Haigh, D. and Neil, N. R J. (1994). Howe: Four Millennia of Orkney Prehistory Excavations 1978-1982. Edinburgh: Society of Antiquaries of Scotland. https://doi.org/10.5284/1000184.

The excavation at Howe, Stromness, Orkney, was one of the largest and most complex undertaken in Scotland at the time. It provided a hitherto unparalleled opportunity to excavate fully the remains of a broch settlement. The discovery of earlier Neolithic structures and enclosed Early Iron Age settlements added importance to the site. The earliest levels, a sequence of fragmentary structures, were dated by incorporation of material of Neolithic type and the nature of the structures, including a stalled tomb and a Maeshowe chambered tomb, comparable to others of the period on Orkney in the fourth and third millennia BC. Sherds of Beaker pottery suggest some activity probably around the turn of the second millennium, but no Bronze Age phase as such could be identified. Occupation resumed in the Early Iron Age (eighth---fourth centuries BC). The first of three major structural complexes, the roundhouse, has brackets of occupation somewhere between the fourth and third centuries. There was no direct radiocarbon dating evidence for the first broch, which overlay the roundhouse; extrapolation from the earliest dates for the second broch put its construction and occupation in the second and first centuries cal BC. This Middle Iron Age broch and settlement was the most massive and best preserved of the structural complexes, overlying and partly destroying the earlier evidence; first to fourth centuries cal AD cover a number of periods of collapse and rebuilding. The later Iron Age farmstead occupied a period from fourth to seventh or eighth centuries cal AD.

The earliest Iron Age phases on the site (Phases 3-6) follow a chronological break after the Neolithic (Phase 2), bridged only by the deposition of fragments of two Beaker vessels. Activities on the site over this period, from the 6th century BC to the start of the first century cal AD, cover the two main areas of the original tomb mound and the outlying ditch. Three main events took place in the area of the tomb mound: stone was robbed from the central cairn and entrance passage of the Phase 2 tomb; a roundhouse was constructed over the remains of this tomb and, finally, the collapse and demolition of the roundhouse was followed by the construction of a broch, Broch 1. Within the area of the Phase 2 tomb ditch, four main events could be recognised: an enclosure gully was dug into the upper fills of the ditch which were Iron Age in date; this gully was then replaced by another similar gully backed by a stone wall; partial replacement of the second gully and wall by a clay-cored rampart and rock-cut ditch was followed by numerous modifications to this rampart and ditch. Between these two main groups of contexts lay the scattered and fragmentary remains of successive contemporary settlements. Most of the Iron Age structures occurred to the S of the old tomb mound. There is no evidence for major breaks in the occupation sequence of the settlement between Phases 3-6 and it has therefore been assumed that these phases represent stages in the development of a single Iron Age settlement whose plan remained consistent through to Phase 7.

A large collapse of the broch walls, resulting in the abandonment of the tower, marks the division between Phases 7 and 8, sometime during the 4th century cal AD. The tower was used as a dump and subsequently an iron-working shed was built into the rubble of the broch collapse. The focus of the settlement shifted to the NE building which formed the nucleus of a domestic settlement with a cluster of yards, interpreted as the farmstead. The focus then switched to a new stalled building in the W of the site and in Later Phase 8 back to the modified E building. These continuous additions, modifications and periods rebuilding are covered by 12 stages, 1-4 in Early Phase 8 and 5-12 in Later Phase 8. The main environmental changes which distinguished Phase 8 was the decline in the use of wood as a fuel and the substitution of heathy turves. There is a diversification in arable agriculture with the presence of hulled six-row barley, the appearance of flax and possibly cultivated oats in the farmstead. There is also a marked decline in red deer with greater reliance on sheep in particular in this mixed farming economy. Working in iron continued in this phase, with a late surge before its general decline. Antler working was still practiced and composite hair combs, previously found on other sites in later Iron Age/early Norse levels, were produced. Copper alloy jewellery continued to be important, distinctive objects from the earlier stages representing a period of the mid-4th to 6th centuries AD and from the later stages material ranging from 2nd century BC to 7th-9th AD. A gaming board usually associated with Norse levels was also found. The occupation of the settlement could have ended during the 7th century AD, but may have continued on to as late as the 9th century.

This chapter presents the environmental data from the site in seven sections, beginning with an introduction and followed by the specialist reports on plant remains, animal bone, bird remains, fish, marine mollusc and non-marine mollusca. Analysis of the data has been made by phase, with limited data deriving from Phases 1-4, some useful information from Phases 5/6 but the major part of the evidence offering detailed discussion is limited to the middle and later Iron Age, Phases 7 and 8, perhaps covering a timespan from the 1st to the 8th century cal AD. This represents the first comprehensive collection of information on the Iron Age environment from an excavation in the Northern Isles. Analysis of the plan and animal remains has shown that the Phase 7 settlement practiced a mixed arable and pastoral agriculture, growing naked six-row barley and keeping cattle, sheep and some pigs. This was supported by the hunting of wild animals, birds and fish, and by the gathering of wild plants and marine shellfish. There is no reason to believe that this was simply a consumer settlement, as there was evidence for cereal crop processing, the primary butchering of animal carcasses, and animal dung. The farmstead of Phase 8 presents new additions to the crop range of the Phase 7 settlement such as flax and hulled six-row barley; in addition, there are further economic changes such as the appearance of rabbits, the increase in domestic animals at the same time as the decline in deer and the use of healthy turves for fuel compensating for the decline in the use of wood.

The human bones were very fragmented and showed signs of considerable disturbance. Many individual bones were excavated from rubble contexts and wall fills and although these were, wherever possible, fitted on to an individual skeleton, in most cases this could not be done. Using the left femur as an indicator, a minimum of eight bodies were identified, mainly from the Iron Age phases. These were represented by five adults, one juvenile and two foetal or full-term skeletons. A single vertebra was found in the fill of a Neolithic ditch.

This chapter draws together all the evidence and the site is discussed under the following headings: the sequence of occupation, Neolithic building and rebuilding, the advent of the Bronze Age, Beaker traces and a Bronze Age hiatus, early Iron Age enclosure and building development, middle Iron Age broch tower and settlement, the post-broch late Iron Age, material culture and discernible changes, economic base, ritual and belief and the site in wider context. There is a growing list of Orcadian Neolithic sites which, after apparent long-term abandonment were restructured in the Iron Age.