1. Introduction

1.1 The Pleistocene

The Pleistocene is the geological period during which multiple ice ages, or glacials, occurred. The last glacial ended about 11,700 years ago, at the beginning of the Holocene. The Pleistocene and Holocene Epochs together are termed the Quaternary Period. The Pleistocene began 2.58 million years ago (2.58 Ma), and we now know that there were numerous ice ages during this epoch. Those of the Middle and Late Pleistocene (together accounting for the last c. 0.8 million years) were more severe than those occurring earlier.

The Pleistocene was not continuously cold; instead, there were periodic warmer episodes, termed interglacials, during which conditions were similar to those of the Holocene, which is generally regarded as merely the latest of numerous interglacials. This glacial–interglacial oscillation is a principal characteristic of the Pleistocene and has been used as a framework for dividing the Quaternary into different climatic phases (Shotton 1973; Imbrie and Imbrie 1979; Bowen 1999).

The sequence of alternating warm and cold Pleistocene climatic episodes is best understood from long sedimentary sequences in the deep oceans and from the deepest ice cores from Antarctica. Both of these data sources yield their climatic signal as fluctuations in the proportion of the oxygen isotopes ¹⁸O and ¹⁶O (Shackleton and Opdyke 1973; Lisiecki and Raymo 2005; Text Box 1). The greater resolution now available, especially from ice cores, has revealed shorter-timescale climatic fluctuation overprinting the glacial–interglacial cycles.

High-resolution records of late Pleistocene climate — gleaned from palaeoenvironmental studies from the last glacial — suggest that this cold stage was punctuated by several oscillations of warmer climate, although these were not as warm as full interglacials. Such fluctuations are called interstadials, and the term stadial is used for the particularly cold parts of glacial stages during which ice sheets extended beyond the present Arctic and Antarctic regions.

The distinction between interglacials and interstadials is essentially one of length and intensity. The formal definition of an interglacial in north-west Europe requires the presence of deciduous woodland (Turner and West 1969).

The glacials and interglacials recognised in Marine Oxygen Isotope curves are classified as numbered stages. These are counted downwards through the oceanic sedimentary sequence: the Holocene is Marine Oxygen Isotope stage 1 (MIS 1); the last glacial maximum is MIS 2 (Figure 1). The MIS curve is not a simple fluctuation between interglacial and glacial maxima and minima. It shows considerable complexity, with substages recognised during the various interglacial stages. Thus MIS 5 is subdivided into MIS 5e to 5a: 5e is the Ipswichian interglacial; 5d a cold stadial; 5c a warm interstadial; 5b another stadial; and 5a another interstadial.

The changes in climate through the Quaternary have been driven by the effects of variations in the eccentricity, axial tilt and wobble of the spinning Earth and its orbit around the Sun, known as Croll–Milankovitch cycles. In the last million years or so the dominant influence has been the shape (eccentricity) of the Earth’s orbit around the Sun, which gives rise to the 100,000 years (100 ka) climate cycles that have dominated during this period (for example Imbrie et al. 1993; Figure 2).

1.2 The Quaternary stratigraphic framework

Figure 3 summarises our current understanding of the chronology of the British Palaeolithic record, compared to the MIS record and the timing of the connection of Britain to mainland Europe.

This classification of Pleistocene strata is based on the recognition of temperate and (less commonly) cold-climate proxies in certain deposits, together with evidence for the deposition of some sediments under warm (temperate) conditions and others under intensely cold or even glacial conditions. For many years in Britain, this division was based on palynological distinctions between interglacials (summarised by Mitchell et al. 1973).

The glacial episodes were characterised by major continental ice sheets. During two of these, land-based ice extended across large parts of Britain (Bowen et al. 1986; Clark et al. 2012). These were the Anglian (c. 450,000 years ago; equivalent to MIS 12) and the Devensian (c. 110,000–11,700 years ago; equivalent to MIS 5d–MIS 2). The Anglian was Britain’s most extensive ice covering (Figure 4).

Together these two glaciations were responsible for almost all glaciogenic diamicton (formerly called ‘boulder clay’) surface deposits in Britain. Between the Anglian and Devensian there may have been more than one glacial ice advance southward across England, although the evidence only remains where it was not destroyed by the later Devensian ice advances (Lee et al. 2011). It is thought that people did not live in Britain during glaciations.

Mitchell et al. (1973) recognised just two interglacials between the Anglian Stage and the Holocene. These are the Hoxnian Stage (equivalent to MIS 11) and the Ipswichian Stage (equivalent to MIS 5e). The Ipswichian and the Holocene are separated by a final glaciation in the latter part of the Devensian. The time interval between the Hoxnian and Ipswichian interglacials, however, appears to represent more than a single interglacial–glacial cycle (Bowen et al. 1986; Bridgland 1994; 2006).

Mitchell et al. (1973) also identified by palynology an interglacial immediately before the Anglian glaciation, the Cromerian. This is also recognised to be an oversimplification, and the Cromerian is now divided into at least four interstadials. Data from vertebrates and non-marine Mollusca, however, suggest at least five distinct warm episodes within what would once have been called ‘Cromerian’. These probably represent isotopic substages within MIS 21–13.

The term ‘Cromerian Complex’ is now generally used for this sequence of interglacials and the cold periods that bridge the late Early Pleistocene and the early Middle Pleistocene. The oldest of these interglacials has a reversed magnetic polarity, indicating that it pre-dates the Matuyama–Brunhes palaeomagnetic reversal (c. 780 ka) when the Earth’s magnetic north and south poles changed to their present polarity (see Figure 1).

Artefacts have been recovered from Cromerian Complex interglacial deposits. Of particular importance in distinguishing between these interglacials is the change, during MIS 15, in water-vole molar tooth morphology (see The ’Vole Clock’).

MIS 22, immediately before the Cromerian Complex, coincides with the first of the intensely cold glacials that have occurred only since the 100 ka climate cycles began (see above). One British archaeological site could be older than this — Happisburgh 3 — where artefacts occur in reverse-magnetised sediments. These have been attributed to MIS 25 or 21, late in the (reversed polarity) Matuyama chron (Parfitt et al. 2010). This dating would make Happisburgh 3 the earliest known occupation of Britain during the Lower Palaeolithic (but see Happisburgh 3 for discussion of the complexity of dating this site).

The MIS 11 Hoxnian interglacial is well represented in lacustrine basins formed during the preceding MIS 12 Anglian glaciation. The Hoxnian type locality is a kettle-hole lake overlain by fluvial deposits in Suffolk (Ashton et al. 2008); and the para-stratotype is a complete lake sequence at Marks Tey in Essex.

In the Lower Thames, the Hoxnian interglacial is well represented in the north Kent sites of Dartford Heath and Swanscombe (Figure 5). At Swanscombe a hominin skull fossil was found, along with many flint artefacts, animal vertebrae and molluscan fossils. There are three superimposed Lower Palaeolithic industries at this site: a basal Clactonian, an assemblage with pointed handaxes, and an upper handaxe assemblage with distinctive twisted edges, the latter thought to represent MIS 11a (Bridgland and White 2015).

The MIS 9 ‘Purfleet’ interglacial is securely established in the British terrestrial record in the Corbets Tey Terrace, east of London.

The Lower Thames sequence is of considerable importance because it forms a staircase of four terraces, within which all four of the post-Anglian interglacials are represented (Figure 5). Investigations at Purfleet, Essex confirm the correlation of the sediments there with the relatively short but strikingly warm MIS 9e interglacial optimum (Bridgland et al. 2013). This site contains three major Palaeolithic tool industries in superposition: Clactonian, overlain by Acheulian, overlain by Levallois. The Acheulian represents the Lower to Middle Palaeolithic transition (Wymer 1999; White and Ashton 2003; White et al. 2011; but see White et al. 2024).

The next-youngest interglacial, equivalent to MIS 7, is known as the Aveley interglacial. It is complex, with perhaps three temperate peaks, although none was as warm as the Ipswichian or earlier Purfleet interglacials. Human occupation of Britain during MIS 7, within the early Middle Palaeolithic, shows consolidation of Levallois knapping and a decline in handaxe use. An exception is Pontnewydd Cave, Clwyd, a rare North Wales interglacial context where numerous handaxes of MIS 7 age were found (Green 1984).

There is no conclusive evidence for hominin presence between MIS 6 and 4, which includes the Ipswichian MIS 5e interglacial. No archaeological material or butchery damage to any of the large vertebrate bone collections from that stage has been found. This is probably because Britain was an island at this time. The return of hominins to Britain in the later Middle Palaeolithic, during MIS 4/MIS 3, has been observed at a few open-air (e.g. Lynford Quarry) and cave sites (e.g. Pin Hole, Creswell Crags). Humans do not appear to have been present in Britain during the Last Glacial Maximum (MIS 2).

Terrestrial records are typically discontinuous, patchy and confined to single glacial–interglacial periods. Therefore a robust chronological framework is required to enable correlation with the universally applicable and continuous framework provided by the oceanic oxygen isotope signal and with the ice core records. River terrace and raised beach sequences, however, can provide terrestrial frameworks in uplifting areas (e.g. Bridgland 2000; 2006; Bridgland et al. 2004), as these contain a sequence of distinct interglacial deposits in many parts of Britain (e.g. Bridgland 2010; Bridgland and Allen 2014).

1.3 Palaeogeography

The landscape and environment that the early occupants of Britain inhabited was, for much of the time, very different to today’s. During the predominantly colder Pleistocene, sea level was generally much lower because global water was locked up in larger polar ice caps. Before MIS 12 there was a ‘British Peninsula’ at the north-west extremity of the European continent, rather than an island Britain (Preece 1995; Figure 6a). The timing of and mechanism for the formation of the Strait of Dover is controversial, but it seems likely that this took place during the Anglian (MIS 12) as a result of the overflow of a glacially dammed lake in the southern North Sea basin (Figure 6b). This drained into the English Channel and cut the earliest Dover Strait.

At the same time, the route of the Thames moved farther south into its modern valley through London (Bridgland 1994) and the Bytham river was obliterated by the Anglian ice sheet, which engulfed its valley. Parts of the former valley provided post-Anglian drainage routes (Figure 6c), but the huge river system was not restored. It was replaced by a proto-Trent system that required two further climate cycles and another glaciation before it reached anything like its modern configuration. Its drainage into the Humber did not occur until deglaciation at the end of the Devensian (Bridgland et al. 2014; 2015; Figure 6d). The Solent river was unaffected by glaciation; its eventual demise was caused by the widening of the English Channel, probably during MIS 6, which drowned its lower reaches and separated the Isle of Wight from the English mainland (Westaway et al. 2006).

1.4 Fitting the archaeological record into this dynamic landscape

The Ancient Human Occupation of Britain (AHOB) project has revealed human occupation in the Early Pleistocene. Sites are known at Happisburgh 3, Norfolk, and at Pakefield, Suffolk, which produced Lower Palaeolithic artefacts dated to MIS 25/MIS 21 and to MIS 17, respectively. The British archaeological record also covers much of the Middle Pleistocene. Lower Palaeolithic human occupation is known from sites such as MIS 13 Happisburgh 1 and the Boxgrove raised beach, West Sussex.

During the MIS 6 glacial, the final stage of the Middle Pleistocene, hominins disappeared from Britain and were absent during the last (Ipswichian) interglacial. Hominins probably did not return until MIS 4/MIS 3, when Late Pleistocene Neanderthals, using Middle Palaeolithic Mousterian stone tools, have been found at a number of sites (e.g. Lynford Quarry). Modern humans appeared slightly later, using a more sophisticated Upper Palaeolithic technology.

Two separate lithic technologies coexisted in Britain during the latter part of the early Middle and most of the late Middle Pleistocene. These are Clactonian assemblages (Mode 1, characterised by flakes) and Acheulian assemblages (Mode 2, characterised by handaxes). The distinction between these two knapping technologies is far from straightforward, however, as both industries produced identical flakes and cores. Clactonian assemblages cannot be recognised definitively unless handaxe making is not represented at all (McNabb 2007).

A further potential advance is the matching of handaxe typology to particular Pleistocene stages (Bridgland and White 2015), which has developed out of a more reliable understanding of the climatostratigraphy and dating of Quaternary deposits across Britain. Handaxe making dwindled in importance once the Levallois technique using prepared cores (Mode 3) appeared. The Levallois industry heralded the transition into the Middle Palaeolithic, and occurred over a wide area around the MIS 9–8 transition.

The Upper Palaeolithic probably first appeared during MIS 3 (by c. 43 ka) with the arrival of Homo sapiens (e.g. Kent’s Caverns; Proctor et al. 2017). Upper Palaeolithic technologies are characterised by blades from prepared cores (Mode 4).

During MIS 2, however, there was probably another complete depopulation of Britain. People returned only as the climate ameliorated at the beginning of the Lateglacial and into the Holocene.

1.5 Shorter-timescale divisions of the Late Pleistocene

The Late Pleistocene began with the warming transition that led to the MIS 5e Ipswichian interglacial (see Figure 3). The preceding glacial produced the most extensive glaciation of the neighbouring part of the European mainland, although the equivalent British ice sheet was smaller than at least two earlier ones (White et al. 2016). This episode was clearly one of severe cold, which probably explains the lack of compelling evidence for human occupation of Britain during MIS 6 and MIS 5e. There is no good evidence that humans returned before MIS 3.

The level of resolution available for the various palaeoclimatic records of the Late Pleistocene is significantly greater than that for the Early and Middle Pleistocene thanks to evidence from ice cores, especially for fluctuations during the last climate cycle (Text Box 2; Figure 7). The MIS 3 interstadial was relatively cold and unstable compared to the previous warm stages of the last million years.

The ice-core record shows that much of the last climate cycle (since MIS 5e) has been characterised by high-frequency, high-amplitude climate oscillations of c. 500 to 2000 years duration (Figure 7), known as ‘Dansgaard–Oeschger’ cycles. These cycles show abrupt warming by 5–8°C within 50 years, perhaps within as little as a decade, followed by more protracted cooling.

The high-resolution temperature record derived from the ice cores can be used to define Late Devensian chronostratigraphy. The record is divided into a series of alternating Greenland Stadial (GS) and Greenland Interstadial (GI) stages. GS-1 represents the pre-Holocene Younger Dryas (in Britain called the Loch Lomond) Stadial and GI-1 represents the Bølling–Allerød (in Britain called the Windermere) Interstadial (Figure 8).

Back to contents | Scientific dating methods for the Pleistocene