The evolution of human manipulative abilities may be clearly linked to the evolution of speech motor control Both creativity and complexity in vocal and manipulative gestures may be closely linked to a single dimension of brain evolution — the evolution of absolute brain size. Inferring the linguistic capabilities of earlier hominids from their lithic artefacts, however, required us to take account of domain-specific constraints on manipulative skill In this article we report on a pilot flint-knapping experiment designed to identify such constraints ‘in action’.

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.

The Lowed/Middle Palaeolithic site at Red Barns, Portchester, on the outskirts of Portsmouth (SU 608063), was re-investigated in summer 1999. Three test-pits succeeded in relocating and exposing the artefact-bearing horizon first discovered in 1973 by J.C. Draper of Fareham and last seen in 1975 when a rescue excavation took place. This excavation produced a massive (in every sense) lithic collection including seven sediment samples, 8678 flint artefacts, 2058 flint nodules and a staggering 18,423 thermally fractured flint pieces (Gamble & ApSimon 1986). Recent study of the 1975 material (Wenban Smith et al. forthcoming) has demonstrated that the site is older than previously thought, dating to at least 200,000 BP and probably to nearer 400,000 BP, and that lithic technology at the site was dominated by the production of pointed plano-convex handaxes. Study of the organizational structure of the lithic production gave an insight into the patterning of Archaic hominid behaviour, with the site serving as a locale were handaxes were regularly made, but from which they were normally removed before being used and abandoned elsewhere.

The site at Red Barns was excavated in 1975, but the large lithic collection remained unstudied following preliminary examination. This paper reports on further analysis of the lithic material from the site, together with a reappraisal of the faunal remains and original mineralogical analyses, and the results of processing sediment samples from the 1975 excavation. An abundant molluscan assemblage was recovered from the deposits covering the main archaeological horizon, allowing climatic/environmental reconstruction and amino acid dating. The synthesis of these data indicates the site to be older than previously thought, dating to between 425,000 and 200,000 BP.

Analysis of the lithic material has suggested that the site is an undisturbed palimpsest of flint tools and debitage. The poor, severely frost-fractured nature of the raw material used for knapping, together with the location of the site on a Chalk outcrop, have enabled investigation of some assumptions about the influences upon knapping technology of i) poor quality raw material and ii) local availability of flint fresh from Chalk bedrock. The persistent manufacture of finely worked plano-convex handaxes suggests that, even in an area where fresh Chalk flint must have been abundant, the immediately available poor quality flint source was not a bar to formation of an assemblage dominated by handaxe production. Secondly, the emphasis on carefully shaped pointed plano-convex handaxes suggests that this shape was both deliberately imposed and not dictated by a lack of local availability of flint fresh from the Chalk. Behaviour at the site was investigated by analysis of the organisation of the lithic production and it was demonstrated that, while some handaxes and flake-tools were abandoned at the site and some flake core reduction also took place there, the dominant pattern was for handaxes to be made at the site and then removed and abandoned elsewhere.

A map made by F. N Hayward showing the location of the Lower Palaeolithic site of Baker's Hole, found in the Hayward Archive stored by the British Museum at Franks House, shows that the site was not at the location suggested by modern authors. The new location, near some remnant Pleistocene deposits, opens the possibility of using modern techniques to date the archaeological material recovered from the site in 1909.