Error is an occupational hazard of the literary life. As soon as a man sets pen to paper he sets himself open to two sorts of error: one, his own, which is bad enough; and, two, the misunderstanding of those who read him, which is infinite. (Allan 1952, ix)

This paper explores some of the ways in which the cultural and political history of the United Kingdom has affected writing about the Neolithic in what is now Scotland. It uses the terms ‘core’ and ‘periphery’ as they have been applied by historiographers writing about the operation of the cultural and political relations within the United Kingdom in the last three centuries. The paper is in two parts. The first part presents a summary of relevant literature generated by 20 years of debate about Scottish and ‘British’ identity, history, literature, and culture. The second part considers the ways in which the general issues have affected the writing of Scottish prehistory, concentrating on the Neolithic of lowland Scotland and examining in particular the writings and influence of Fox, Childe, Piggott, and Atkinson.

Sir Grahame Clark's interests in wetland archaeology were not restricted to his pioneering work at Star Carr and much of his writing was illuminated by his wide knowledge of the fruits of wetland research in many parts of the world. The paper presents the case for wetland archaeology, to show how it has expanded our knowledge of the past and has made prehistory more colourful and dynamic to both archaeologists and the public. In seeking the patterns of behaviour that existed in the past, six key elements contribute to our studies: environment and change; economy and subsistence; stratification and context; structures and activities; chronology and precision; and range of material culture. The evidence for all of these aspects is well-preserved in many wetland environments, and a number of key sites are identified and assessed for their contribution to prehistoric studies.

Koongine is a sizeable limestone cave set in a low ridge some 4 km from the sea in the lower South-east of South Australia. It was used for about 2000 years at the transition from the Pleistocene to Holocene, and then again during the last millennium. The sequence at this site exposes issues of the appropriate scale and form of explanation for changes in site use. The stratified deposits of stone artefacts provide an opportunity to define for the first time the nature of the ‘Gambieran’ Industry. This spatially and temporally restricted industry characterised by large convex scrapers made on large, often asymmetrical, flakes is otherwise known mainly from older surface collections. The formal definition of this local industry adds to the growing evidence of considerable variation in the earlier stone tools of Australia, and provides an additional basis for rejecting the concept of a widespread Core Tool and Scraper Tradition, and replacing it with a model which recognises a mosaic of different tool-making traditions embedded in local social, economic, and technological contexts.

This short paper describes the remains of prehistoric settlements, chambered tombs, a promontory fort, a prehistoric defensive wall, a rectilinear field system, and other field systems on marginal land at St David's Head. Antiquarians and archaeologists have known of these remains for over two centuries, but it is only through modern surveying techniques and aerial photography that their true nature can be appreciated. The defensive wall and associated rectilinear field system could have originated from the 2nd millennium BC through to the 1st millennium BC. Other field systems and settlements are likely to be of later prehistoric or Romano-British origin. Elements of the field systems have influenced and are preserved in the modern ‘Pembrokeshire’ landscape which borders the headland.

A long cropmark enclosure at Springfield, Essex, interpreted as a Neolithic cursus, was investigated betwee 1979–85 to confirm its date and establish a site sequence. The enclosure was c. 690 m long and 37–49 m wide the ditch being uninterrupted in all areas examined, features within the interior at the eastern end included a incomplete ring of substantial post-pits which it is suggested originally formed a complete circle. Peterborou pottery, predominately Mortlake style, Grooved Ware, a small amount of Beaker pottery, earlier Bronze Age urn sherds, and flint artefacts of the late 3rd-early 2nd millennium were recovered from the cursus ditch an other features. Collectively the evidence indicates a prolonged period of use. The results of the excavations a described, the site is discussed in its local and regional context and the implications of the excavation for our understanding of cursus monuments are considered.

A recurring theme of the Late Bronze Age is the apparent association between deliberate deposition of material and wet places. Recently, a human skull has been discovered within the basal sediments of a relict mire at Poulton-le-Fylde, Lancashire, dating to the later Bronze Age (c. 1250–840 cal BC). The find, which belonged to a c. 25–35 year old male, was located within a layer of silty wood peat elm deep, representing the ancient root system of a hazel copse and containing many hazelnuts and some charcoal. Palaeopathological investigation established the likelihood that the skull had decomposed before deposition and there are strong parallels between the find and its context and other prehistoric skulls recorded from British wetlands. The connection of the human remains with considerable amounts of hazel wood may also be of significance when viewed within the wider context of similar associations recorded from European bog-bodies. During the course of excavation and survey of the site worked wood fragments were recovered indicating both human and animal (beaver) activity, dating to the later Bronze Age and Early Iron Age respectively. The stratigraphic sequence indicated that organic sedimentation resulted from the rapid flooding of a formerly relatively dry landscape, perhaps as a result of the effects of beaver damming – a possibility which may hold wider implications for the archaeological interpretation of prehistoric pollen data.

The Azé Cave (Sâone-et-Loire, France) has yielded, among other archaeological and palaeontological remains, a Lower Palaeolithic industry within a layer dated through faunal remains to 400–350 kya. This industry is made up of local rocks, mostly poor quality flint, and also chert and crystalline rocks.

Only flint was definitely knapped; chert was only broken into pieces before being used. The process of knapping is opportunistic and the cores, seldom exhausted, usually bear two opposite reduction faces. The flakes often reveal patches of residual cortex. They are moderately thick and the striking platform angle varies within a wide range of values. Some of them seem to be the result of a tearing-off motion. All the flint pieces have been heavily retouched; half of them have been simply utilised, others are proper tools, mostly scrapers, usually with steep retouch. Fragments of chert, sometimes obtained by tearing-off (from the cave walls?) have been often retouched but with less intensity than the flint. A good number of more or less trimmed cobbles are also part of this collection. However, handaxes are completely missing.

The lithic industry of Azé 1–1 lacks standardisation for the core reduction process as well as for trimming or retouching. It fits well in the Lower Palaeolithic industry but the absence of handaxes is noteworthy. This industry offers one more contribution to the diversity of this cultural period.

This paper accepts the position that the European Aurignacian should be seen as a reflection of behaviour connected to a modern human dispersal. A two-phase dispersal model (‘Pioneer’ and ‘Developed’ fades) is proposed to explain the variations in artefactual diversity and spatio-temporal patterning, enacted by directional, rapid movement across the continent rather than by a ‘Wave-of-Advance’. Presumed behavioural signatures of this population dispersal, notably what is here termed ‘behavioural flexibility’, are also explored.

This paper reports on the recovery of Palaeolithic flint artefacts and faunal remains from fluvial gravels at the base of a sequence of Pleistocene sediments revealed during construction works at two sites to the south of Swanscombe village, Kent. Although outside the mapped extent of the Boyn Hill/Orsett Heath Formation, the newly discovered deposits can be firmly correlated with the Middle Gravels and Upper Loam from the Barnfield Pit sequence dating to c. 400,000–380,000 BP. This increases greatly the known extent of these deposits, one horizon of which produced the Swanscombe Skull, and has provided more information on their upper part.

Comparison of the lithic assemblages from volume-controlled sieving with those from general monitoring demonstrated that artefact collections formed without controlled methods of recovery, such as form the majority of the Lower and Middle Palaeolithic archaeological record, are likely to be disproportionately dominated by larger, more visible, and more collectable neatly-made handaxes to the detriment of more poorly made, asymmetrical handaxes and cores, flakes, and percussors. The lithic assemblage from the fluvial gravel was confirmed as dominated by pointed handaxes, supporting previous studies of artefacts front the equivalent Lower Middle Gravel at Barnfield Pit. The raw material characteristics of the assemblage were investigated, and it was concluded that there was no indication that the preference for pointed shapes could be related to either the shape or source of raw material.

This paper also reviews the significance of lithic assemblages from disturbed fluvial contexts, and concludes that, contrary to some current perspectives, they have a valuable role to play complementing less disturbed evidence in developing understanding of the Palaeolithic.

A new date for the assumed first humans in Mallorca disagrees strongly with the earliest dating of them. New light on the first settlement of Mallorca comes from the discussion of dates and the review of the empirical evidence for that settlement. We conclude that human arrival on Mallorca must be situated in the 3rd millennium cal BC.

Thirty-four years ago there appeared in the Proceedings of the Prehistoric Society the third and final paper in a series reporting the research of Eric Higgs and his colleagues into the climate, environment, and Stone Age record of north-west Greece (Higgs et al. 1967). This paper discussed two limestone rockshelters, Kastritsa and Asprochaliko, comparatively in what is now a classic example of site catchment analysis. Kastritsa, in the Pamvotis lake basin, and Asprochaliko, some 35 km away in the Louros river valley (Fig. 1), were destined to become famous, not only for being amongst the first sites in Greece to yield stratified Palaeolithic deposits, but also for the methodological and interpretative issues raised by their archaeological records. The ideas presented in the 1967 paper aroused considerable debate and controversy regarding when and during which seasons the sites were used and whether they were complementary parts of the region's Upper Palaeolithic settlement system (Higgs et al. 1967; Bailey et al. 1983; Bailey 1997a; Green 1997). The essential assumption underlying Higgs' original hypothesis was that some of the strata recovered from these sites were contemporaneous. In recent years the upper part of Kastritsa (strata 1 & 3) has also been considered to be partly coeval with two of the rockshelters in the Voidomatis river valley, namely Klithi and Megalakkos (Fig. 1).