8

New Perspectives in Archaeology: Process, Adaptation and Behaviour, 1962–1980

Potassium-Argon Dating and the Oxygen Isotope Record

In making cross-continental climatic correlations to build their global culture histories, archaeologists were making several assumptions. First, they assumed that the glacial phases in Europe were accompanied by concomitant climatic responses in more southerly latitudes, principally in terms of precipitation, and that the different climatic curves could therefore be ‘wiggle-matched’ to find the best fit. The validity of these matches was tested and calibrated using other proxies, vegetation histories, faunal complexes, astronomical calculations and archaeology, each of which brought additional layers of assumption. The last explicitly assumed that the succession of stone tool industries in Africa, Europe, India and the Levant ran more-or-less in parallel. If Africa was the cradle of humanity and the crucible of culture then some time lag between regions was acceptable, and local cultural facies were only to be expected, but the fundamental phases of cultural development seen in the Earlier Palaeolithic record were, geologically speaking, contemporaneous across the Old World, with the arrival of handaxe cultures marked by the Chellean/Abbevillian of the second (Kamasian) pluvial at Olduvai, of the Mindel-Günz in Europe and of the second (Mindel) glacial in India (Terra et al. 1938). In calendrical terms derived from astronomical calculations, this was believed to equate to a period ~400,000–500,000 years ago (cf. Zeuner 1946). Little had really changed since de Mortillet.

These long-laid foundation stones were totally and finally undermined by the application of radiometric and geo-chemical dating methods to the question of Pleistocene chronology, techniques developed in other fields which gave absolute calendrical age-estimates (albeit with large margins of error). When writing the revised and expanded fourth edition of Dating the Past, half a century after Ernest Rutherford had first used the decay of radioactive elements to estimate the age of the Earth, Zeuner (1958) could list only two commonly used radioactive techniques, radiocarbon and radium dating, the former not much use beyond 35,000 years ago, the latter having a minimum range of ~150 million and thus totally irrelevant to the Palaeolithic. By the late 1980s, Aitken (1990) was able to discuss the pros and cons of over a dozen radioactive and chemical techniques applicable to a range of materials of Pleistocene age, including potassium-argon (volcanics), uranium-series (speleothems, travertine, bone, teeth, shells), fission track (volcanics), luminescence and optically stimulated luminescence (burnt materials, certain sediments), electron-spin resonance (speleothems, travertine, teeth, bones, sediments), amino-acid racemisation (shells), magnetic dating (sedimentary and igneous rocks) and obsidian hydration (volcanic glass). Few of these scientific dating revolutions would have such a dramatic impact on the Earlier Palaeolithic as the first.

In 1957, Jack Evernden and colleagues at Berkeley announced the potassium-argon method for dating volcanic and other igneous rocks older than ~300,000 years old, which, given the right materials, was applicable to the Lower and Middle Pleistocene (Evernden et al. 1957). The same year Evernden accompanied Louis Leakey to Olduvai, where they began to collect samples from the gorge. By the time he announced Zinjanthropus, Leakey (1959) had already determined that the fauna and Oldowan industry of Bed I was not Middle Pleistocene but Lower Pleistocene in age, but this could hardly have prepared him for the results. Samples from Olduvai sites FLK 1, FLK NN and MK1 produced an average K-Ar age estimation for the ‘Zinj floor’ in Bed I of 1.75 ma BP, while the top of Bed I, in which the Chellean appeared, had an average age of 1.23 ma BP. Everything was much older than expected and as shown by subsequent analysis, much older than the supposed equivalents in Europe. K-Ar dating of volcanic materials in the Rhine Valley, for example, established that the Günz glaciation was a maximum of 500,000 years old (Frechen and Lippolt 1965, Table 8.1), while the Abbevillian-Acheulean industry at Torre in Pietra in Italy was dated to 430,000 years old, again younger than the >500,000 year dates for the later Chellean II–III at Olduvai (Evernden and Curtis 1965). The cross-continental climatic and archaeological correlations had greatly compressed the African record and were simply invalid.

Table 8.1 Frequency of various uses (counting edges) of stone tools from Clacton and Hoxne Lower Industry (modified after Keeley 1980, 1993).

Inferred Use

Clacton Count

%

Hoxne Count

%

Swanscombe Count


WoodWhittling

10

20

2

5.9

1

Wood Chopping

4

8

1

2.9

0

Wood Sawing

2

4

0

0

0

Wood Scraping

5

10

2

5.9

0

Wood Wedging

1

2

5

14.7

0

Hide Scraping

3

6

4

11.8

0

Hide Cutting

1

6

2

5.9

0

Hide Unknown Action 0

0

1

2.9

0

Meat Cutting

10

20

10

29.4

1

Bone Chopping

0

2

5.9

0

Boring

4

8

2

5.9

0

Plant Cutting

0

3

8.8

0

Other

10

20

0

0

2

Radiometric dating techniques also provided critical tie-points for other established and developing relative chronologies, such as the climatic record preserved in deep-sea cores. Differences in the abundance of species of planktonic foraminifera preserved in deep sea sediments had been linked with changes in past climates since the 1930s (Ericson et al. 1965). The invention of the piston corer in the 1940s had made it possible to recover sediment cores up to 15m long, each containing detailed and continuous records of cyclical climate change, but until the advent of radiocarbon dating the age of the sediments remained floating and essentially unknown. Radiocarbon allowed the sediments from the last 35,000 years to be directly dated, from which an average accumulation rate and therefore the age of the entire core could be estimated.

During the 1960s two basic approaches were applied to the deep sea-core records. David Ericson and his team at Columbia University (e.g. Ericson et al. 1965) used the frequency and distribution of 18 common species of foraminifera to calculate palaeotemperatures, finding evidence for four major cold stages over the past one and a half million years, with the great interglacial (Mindel-Riss) lasting over 0.6 million years. In Chicago, Cesar Emil-iani (1955) had employed many of the same common species of foraminifera, but instead of faunal zones based on species tolerances and distribution, used variations in the isotopic composition (Oxygen-16 vs Oxygen-18) of their calcareous tests to calculate palaeo-temperatures. This showed a much greater number of climatic oscillations, with some 14 cold-warm events since the (supposed) beginning of the Pleistocene (Günz glaciation), in keeping with the solar radiation curve but contrary to the conclusions of Ericson’s team and that derived from terrestrial pollen records (Zagwijn 1957, 1960, 1961, 1963; West 1963). Emiliani (1955) originally calculated the duration of the Pleistocene to have been ~300,000 years (based on radiocarbon), but later revised this to 600,000 (1966).

Direct potassium-argon dates on deep-sea cores were unhelpful, as they produced dates with error ranges greater than the duration of the climatic events being measured. They did, however, allow geologists to establish an absolute timescale for the known polarity reversals in the earth’s magnetic field (as recorded in its rocks) which could then be used to provide further tie-points in the deep-sea cores and allow better estimates of deposition rates and deformation (e.g. Cox et al. 1968). The results favoured a longer chronology for the Pleistocene (Glass et al. 1967; Ericson and Wollin 1968), but were not incompatible with the oxygen isotopic curve, as Emiliani had not yet examined the earliest parts of the cores in any detail (Shackleton and Turner 1967).

Comparing only the records for the past 500,000 years, Shackleton and Turner (1967) concluded that the two methods were in good agreement, but that Emiliani’s method was far more sensitive to temperature change and provided a more detailed picture. The evidence from the lacustrine sediments at Marks Tey in Essex, which showed annual laminations that could be counted and contained a pollen profile spanning most of the ‘Great Interglacial’ (Hoxnian, Mindel-Riss, Holsteinian), also supported Emiliani’s shorter chronology, as did oxygen isotope measurements on Pleistocene ice preserved in the Greenland ice sheet (Dansgaard and Tauber 1969). These again indicated seven or more independent glaciations during the past 425,000 years, not just the four usually recognised.

Shackleton (1967) reasoned that Emiliani’s oxygen-isotope curve not only measured the temperature of seawater between cold and warm phases (which he estimated varied by up to 6°C), but must also reflect the extraction of large amounts of water from the oceans during periods of extensive continental glaciation, with concomitant melting and recirculation of this water as glaciers receded to present levels during warm periods. The relationship between the two was complex, with the more recent and higher-resolution record from the last interglacial (MIS 5), showing a significant time-lag between temperature change and ice volume (Shackleton 1969). This had a direct bearing on the nature and pattern of human occupation in northern latitudes, which might have experienced drops in temperature without full extension of the glacial ice or concomitant reduction in sea level. Indeed, it seemed unlikely that all cold events were of sufficient duration for major ice sheets to accumulate and advance.

There was clearly much potential for correlating the continuous oxygen isotope sequence with the fragmentary terrestrial record. Shackleton (1969) considered the regional pollen sequences (e.g. West 1956, 1957, 1963; Vlerk 1955, 1957; Zagwijn 1960, 1961) to provide the best available record of climate change on land, but noted that the classic last interglacial pollen spectrum (Ipswichian, Riss-Würm, Eemian) could only be correlated with the first warm peak of MIS 5 (5e) after which temperatures declined rapidly, with bad winters and cool summers.

In fact, the pollen record signalled only three glacials and three interglacials during the past ~500,000 years (West 1963), not the 12–14 events seen in the isotopic record. The additional warm periods (MIS 7 and 9) either had minimal expression on land, in other words there was no direct way to correlate oceanic and terrestrial records, or they had vegetational sequences that could not be differentiated from other Interglacials, either because they were in close succession and very similar or because correlation and seriation of fragmentary evidence had led to the creation of false successions (Sutcliffe 1976). Reviewing the evidence from central Europe, Frenzel (1973) reached a similar conclusion. There were far more cold and warm episodes than could be accommodated in the traditional model, and there were botanical reasons why a number of interglacial sites formerly believed to represent a single warm interval, could not have been contemporaneous.

A classic case was found in the series of deposits situated between 5 and 15m on the north bank of the Lower Thames at Ilford, Aveley and Thurrock, the contents and relationship of which had been a cause of friction since the 1860s (Dawkins 1867; Wood 1867, 1868). Pollen analysis suggested to West (1969) that these sites all belonged to the Last Interglacial (Ipswichian/Riss-Würm/Weichselian), although Conway (1970) objected that such a correlation ‘did violence’ to other lines of evidence. Little Thurrock, for example, had a Hoxnian mammalian fauna (cf. Sutcliffe 1964), Rhenish molluscan elements that filled a gap in the Swanscombe sequence and a Clactonian flint assemblage (Wymer 1957). The site was in his opinion Hoxnian in age. Sutcliffe (1975, 1976; Sutcliffe and Kowalski 1976) also emphasised the difference in the mammalian faunas at Aveley and Ilford, which were dominated by a small Mammuthus trogontherii (the Ilford-type ‘dwarf’ mammoth), horse and Stephanorhinus kirchbergensis (Merck’s or steppe rhinoceros), and were thus entirely different to the hippopotamus, straight-tusked elephant, fallow deer and Stephanorhinus hemitoechus (narrow-nosed or forest rhinoceros) fauna of the classic Ipswichian site at Trafalgar Square (West 1957; cf. Franks 1860), which was tellingly situated at a much lower elevation. He reasoned that they must belong to different parts of an interglacial or different interglacials, the weight of evidence leading Sutcliffe to conclude that they represented a pre-Ipswichian, post-Hoxnian interglacial, such as were signalled in the isotopic record.

Using central European loëss and terrace sequences as a link to the MIS curve, Kukla (1977) also concluded that sites had been mis-correlated and that the classic four-glacial framework should be abandoned. Nevertheless, pollen-based chronologies continued to dominate until the 1990s and beyond (e.g. Mitchell et al. 1973; Roe 1981; Zagwijn 1985; Gibbard 1985, 1994).

Tooling Up to Face the Environment

Whatever their chronological shortcomings, palynological studies still provided an increasingly high-resolution picture of local environments and short-term climate change, greatly enhancing the information gained from mammals, molluscs and plant macro-fossils. This finally made it possible for archaeologists to examine the Lower Palaeolithic using the ecological approaches long advocated by prehistorians of later periods such as Gordon Childe and Grahame Clark, although much of the existing Palaeolithic data was quite unsuitable for the task.

Despite their historical importance and celebrity, the questions now being asked meant that most of the classic sites of Europe had really been discovered “too early for their own archaeological good” (Roe 1981, 172). The assemblages they had produced tended to be poorly contextualised, widely dispersed and biased by the individual peccadilloes of the original collectors. Many lacked flakes, cores and small flake tools, either because they were not looked for, not spotted, or not kept. Sure, more representative samples could be found, particularly those from the famous floors and workshops (e.g. Atelier Commont, see Bordes and Fitte 1953), but these still lacked fine spatial information that might help better understand the variations in stone tools and their relationship to the environment. Meticulous large-scale excavation was required to collect new representative samples, to record in detail their vertical and horizontal distribution, and to establish a firm environmental context, but opportunities were limited. The rate of discovery of significant sites had greatly diminished in much of Lower Palaeolithic Europe, as finite resources were exhausted and extraction methods changed, while many pits, including classic localities in the Thames and Somme, had been abandoned, refilled and built over (Wymer 1956; Agache 1958). The best that could be expected were witness sections or small excavations, aimed at providing confirmation of old reports, firmer contexts and new artefacts from which to make some sense of the old collections (Wymer 1956, 1957, 1961, 1964; Agache 1958; Howell 1966; Bourdier 1974).

The death of Henri Breuil in 1961 also marked the passing of numerical subdivisions based on type fossils and connoisseurship. In South Africa, where only four Acheulean sites had been properly excavated (Sterkfontein, Klipplattdrift, Three Rivers and the Cave of Hearths), the deaths of Riet Lowe in 1956 and Goodwin in 1958 were likewise followed by reappraisals of the eight-fold succession in handaxe development, the distinct status of the Fauresmith and the anthropogenic origin for the Kafuan (Mason 1961).

There were far greater opportunities in East Africa, which was entering a golden age of discovery led by a generation born exclusively in the twentieth century, including Mary Leakey (1913–1996), J. Desmond Clark (1916–2002), F. Clark Howell (1925–2007), Maxine Klein-dienst (b. 1936) and Glynn Isaac (1937–1984), who were actively conducting meticulous multidisciplinary excavations on primary context ‘occupation floors’, retaining all cultural and non-cultural materials and spatially plotting all significant finds in three-dimensions.

To use the information from these occupation floors to understand the behaviour (‘life-ways’) of Lower Palaeolithic humans, Maxine Kleindienst at the University of Chicago advocated using qualitative and quantitative methods with standardised measurements and typologies, like those used by Bordes in Europe but tailored to meet the specific requirements of the African material (Kleindienst 1961, 1962). Kleindienst’s system was hierarchical, recognising four major classes (shaped tools, modified tools, utilised tools and waste), each broken down into a series of types and sub-types (Figures 8.1 and 8.2). It was based on overall morphology but took no account of the degree of finish or workmanship shown in each class of object. This broke with previous classifications of Acheulean industries, which relied on subjective assessments of the quality and sophistication of handaxe manufacture, and while Kleindienst was explicitly seeking behavioural explanations, artefacts were still treated as the product of culture, “of habit, pattern and conformity to the traditions of a given time and place” (1962, 83). Kleindienst was also more acutely aware that conclusions drawn from statistical studies were only ever as good as the samples used, and that greater attention should be paid to site formation process (taphonomy or the science of burial), for example the winnowing and concentration of small objects by freshets and the differential erosion of different rock types.

Line drawings of different types of stone tool

Figure 8.1 Kleindienst’s Late Acheulean tool classes (after Kleindienst 1961).Line drawing of different types of cleaver and handaxe

Figure 8.2 Sub-types of handaxes and cleavers in the East African Late Acheulean (after Kleindienst 1962).

Analysis of the assemblages from the multiple occupational floors at Olorgesailie and Isimila (Howell et al. 1962), all assigned to Leakey’s Acheulean Group 9 or 10, showed that while all African assemblages were similar in general content, the variation in the relative proportions of each component was greater than previously recognised (Figure 8.3).

series of ten graphs showing different frequencies of stone tools in different levels at Olergesailie.

Figure 8.3 Component analysis of ten occupation surfaces at Olorgesailie (after Kleindienst 1962).

Kleindienst identified four main concentration types:

1. Occupation areas with a high percentage of handaxes/cleavers/knives, low percentage of other large tools, and low percentage of small tools

2. Occupation areas with a low percentage of handaxes/cleavers/knives, low percentage of other large tool, and high percentage of small tools

3. Occupation areas with equal proportions of handaxes/cleavers/knives and small tools, but low percentage of other tools

4. Occupation areas with higher proportions of other large tools compared to handaxes and flake tools

Observing patterned variation was one thing, interpreting its meaning was quite another (Kleindienst 1961, 51). None of the variables involved–time, space, function, style–could be fully controlled, a problem exacerbated by the fact that the makers of the tools were not modern Homo sapiens, rendering actualistic ethnographical studies of uncertain relevance. Time was probably not the main factor controlling variation at Isimila and Olorgesailie, although it was important to keep in mind the fact the occupation sites did not all represent the same amount of time. Some might capture a few years or decades worth of accumulation, others centuries. Each of these scales had a different story to tell from the gravel accumulations spanning tens of thousands of years.

Few African sites contained enough waste to warrant classification as workshops. Instead there was an inverse correlation between shaped tools and flakes, suggesting that the handaxes, cleavers and knives had been made elsewhere and imported onto sites in a finished form. Small tools and flake tools were more often accompanied by manufacturing debitage and were thus likely made on the spot for immediate use. This was also reflected in the patterns of raw material use: small tools tended to be made from the most homogenous local material available (often quartz), while large tools tended to be on exotic materials, silicified limestone, quartzite and metamorphic cataclastites. Stone had been carefully selected, matching the functional requirements of the tool to the physical properties (shape, size, hardness, durability) of the material.

Kleindienst concluded that the spatial variation in the major classes was probably related to site function or the range of activities undertaken at a site, hominins “doing different things to different degrees at given times and places” (1961, 44). High proportions of small tools might reflect, for example, a greater woodworking component. There might equally be some seasonal (wet vs dry season) or social (age/sex) segregation of different activities. The former, at least, could potentially be tested using other proxies. Variation within each class of tools, the different types of handaxes and cleavers, however, was probably related to style (Kleindienst 1961, 1962). She also observed that the differences between concentration types A and B described the contrasts between the Late Acheulean and Hope Fountain industries. They were probably functional variants.

Using Kleindienst’s method for the sake of clarity and comparability, J. Desmond Clark (1964) examined the relationship between climate and culture change at Kalambo Falls, although he unwisely accepted a set of early and very young radiocarbon estimations at face value.1 Pollen from the site showed that the four stages of Late Acheulean occupation took place when the environment was like that of today, the valley bottoms dominated by reed-beds and grassland, with woodland on the slopes. During the final Acheulean and Acheu-lean/Sangoan transition conditions became progressively wetter and cooler, and by the end of the Sangoan the vegetation had become more open, with a marked decline in wet forest habitats. During the same period, East Africa had seen the extinction of a high proportion of surviving Middle Pleistocene faunas (although no bone was preserved at Kalambo Falls). The MSA was associated with a dry climate and essentially modern fauna.

The new methods confirmed that major differences existed in the industrial groups at Kalambo Falls. The artefacts on the Acheulean occupation floors were spread over a much larger area than those on the Sangoan floors, suggesting to Clark that Acheulean populations were bigger. In terms of artefact types, Acheulean floors were dominated by handaxes, cleavers, knives and regular scraping tools, whereas the Sangoan was characterised by small informal flake tools and heavy choppers. Clark suggested that these differences reflected the dominant activities at the site, tentatively inferring from the morphology of the working edges that the Acheulean toolkit had been used for cutting and dismembering large carcasses, whereas the Sangoan was more concerned with chopping and gouging hard vegetable materials, such as wood. Clark (1964, 97) acknowledged that these functional inferences might be incorrect, but nonetheless felt it was “obvious that a change in the economy (in the dietary habits and the functional emphasis of the stone equipment)” had taken place and “that large animals generally formed a higher proportion of the faunal remains associated with Acheulian industries than they did with the Middle Stone Age”. Clark (ibid.) saw climate as the key driver underlying all these changes, the wetter, cooler conditions after the Acheulean “determining the direction this change would take; directions which differed in the woodland savannah from those in the open grasslands… [making] environment… the most important single factor influencing cultural and biological evolution” (ibid.).

Cooke (1962) likewise interpreted the ‘cruder’ Sangoan industry of Zimbabwe (then Southern Rhodesia) as a response to drier climates, which would have forced people to adopt a fully rather than seasonally nomadic lifestyle, stick close to rivers and invest little effort in tools, which thus were made on the spot and rarely transported. The density of artefacts at a site and the range of tools present might thus simply reflect the duration of occupation, sites used for longer having a larger number and variety, related in turn to the abundance of local resources (i.e. the Hope Fountain variant).

Ethnographer Richard Lee was not so convinced by simple climatic arguments. Lee (1963) divided southern Africa into five ecological zones (desert, grassland, savannah, forests and macchia, a Mediterranean-type vegetation), each of which was ranked on the basis of game, plant resources and water into habitats that were favourable, marginal and unfavourable for human occupation. He concluded that even at its driest, southern Africa would still have had 1,200,000 square miles of favourable habitats, with over a million square miles suitable for permanent occupation, regardless of climate. It was true that the Acheulean was rarely found in the moist forests, possibly because their subsistence was tied to large game hunting on the open savannah, but Lee could find no evidence that desiccation would produce stress, dislocation and culture change everywhere. The favourable zones were large and continuous, giving no reason to suspect major changes in economy, social isolation or the ‘speeding up’ of culture change during dry periods, as imagined by Desmond Clark (1960). The overall picture was one of large-scale ecological stability under different rainfall regimes. For Lee, archaeologists urgently needed to rethink their ideas on the role of climate and environment in driving Palaeolithic culture change.

Tracing Artefact Function

A major sticking point in ecological and functional interpretations was the fact that the actual use of most artefacts was still unknown, despite the functionally loaded names they had been given by the pioneers (Bordes 1969). Most archaeologists relied on form, edge damage, guesswork, ethnographical analogy and, at best, use-experiments, to infer general patterns of use for different objects (e.g. Clark 1964; Keller 1966). This allowed ideas that had long been rejected–and rightfully forgotten–to be resurrected. Jeffreys (1965) for example, opined in the pages of Man that handaxes were projectiles that would have been particularly useful for hunting birds, that pointed forms were held at the tapered end and hurled like a stick-grenade, and that the tool should be renamed the ‘hand-bolt’. This was suggested without a single reference to the nineteenth-century dismissal of the idea.

The English translation of Russian archaeologist Sergei Semenov’s Prehistoric Technology (1964) introduced ‘traceology’ or use-wear analysis to countries outside of the USSR. This technique employed the study of macroscopic and low-powered microscopic analysis of damage, scratches and striations on the edges of stone tools to determine the direction and motion of use, which, when combined with extensive experimental replication, allowed the probable function of stone tools to be determined in a more scientific manner. The method certainly broke new ground (Lacaille 1965, although note that similar methods had been simultaneously developed in the United States, e.g. Sonnenfeld 1962), but it was extremely specialised and time consuming to master and Lower Palaeolithic applications were slow in coming (Howell 1966; Keller 1966).

Working at Oxford under the supervision of Derek Roe in the 1970s, Keeley (1980) developed new techniques of microscopic analysis which he applied to a range of Lower Palaeolithic materials from Britain (Table 8.1). Keeley’s method used both low-powered and high-powered microscopy to study scratches and polishes formed on the edges of stone tools during use, from which he inferred the motion of the tool and the hardness of the material with which it had been in contact. Most stone tools had been used to process wood or animal products (hide and flesh), with very little evidence of use on plant foods, possibly because the plants used by Lower Palaeolithic humans required little processing, or because plants played only a small role in hominin diets at Hoxne, Swanscombe and Clacton where they might have been seasonally unavailable. Most plant processing seemed to be aimed at making other tools from wood (Figure 8.4).

Line drawing of stone tools, showing location of use-wear and inferred function, ranging from woodworking to hide-skinning.

Figure 8.4 Function of stone tools from Hoxne and Clacton, as inferred from high power micro-wear analysis (after Wymer 1985).

Keeley found no evidence that the Clactonian choppers were regularly used as tools, the damage identified by Warren (1923d) and Wymer (Singer et al. 1973) probably being the result of post-depositional spalling. Of the 22 choppers he examined from Clacton, only two showed any evidence of having been used, in both cases to chop bone. Most of the evidence Keeley found for chopping was on large flakes. Hoxne and Clacton also differed in the number of flakes that had been used for tools, 10% compared to 25%. According to Keeley (1980, 1993) this reflected the fact that handaxes were being made and used at Hoxne, providing a bespoke cutting tool and generating an astonishing number of flakes and chips too thin and brittle to be of use. In general, the flakes produced during handaxe manufacture seem to have been rarely used, most utilised flakes were detached from cores. Many tasks, moreover, appeared to have been accomplished with unretouched edges, although even these showed areas that had been modified to remove a spur or to blunt opposing edges to facilitate comfortable handling. The weakly developed wear patterns further suggested that these were rather ad hoc and rapidly discarded implements. Only pieces requiring a regular edge (for scraping) or a steep strong edge (for chopping) were retouched, and it was possible that the retouch found on hide-scrapers may be product of resharpening rather than initial shaping, as scraping quickly dulls an edge (Keeley 1993, 131).

As Keeley’s method demanded pristine artefacts, even the smallest amount of post-depositional abrasion could alter or destroy use-polish, he was only able to determine the use of three handaxes. All showed evidence of meat polish, but due to the small sample size he cautiously stuck to the received wisdom that they were multipurpose tools, although he did emphasise their usefulness in a range of carcass processing tasks away from the home base, the flake tools being used in ‘central locations’ (tacitly implying that the site of High Lodge operated as a home base). A twisted ovate from Hitchin, despite appearing in mint condition, failed to provide any indication as to the functional purpose, if any, of this particularly intriguing edge configuration. This is still a problem in interpreting the form today (White et al. 2019).

Despite its enormous early potential, microwear never really solved the problems of stone tool function, nor has it revealed clear activity differences between sites. Most Palaeolithic sites show the same range of basic tasks regardless of artefact or culture. Furthermore, when others tried and failed to replicate Keeley’s results in blind tests (Newcomer et al. 1986; Bamforth 1988), faith in microwear studies faltered, particularly in Britain. Too much, it seemed, depended on experience and subjective classification. It was hardly a sure fire or even scientific route to understanding functional variation in the Palaeolithic. It was quicker, and perhaps no less accurate, to take an informed guess. This was not true on the continent, however, where microwear continued to develop and is now a fairly routine part of lithic analysis; although the range of operations performed at most sites is still fairly monotonous.

Zinjanthropus the Toolmaker?

Leakey’s (1959) pronouncement that Zinj was the maker of the Oldowan was not intrinsically unreasonable and was quickly endorsed by others (Howell 1960; Washburn 1960), but it was in hindsight somewhat rash. Robinson and Mason (1957) had been far more circumspect about the association of stone tools and australopithecine fossils in the Middle Breccia at Sterkfontein. They thought it false logic to assume that the most advanced primate found in the fossil record at a site was the maker of the stone tools: on that basis one would have to conclude that the handaxes at Olorgesailie had been made by giant baboons (Robinson and Mason 1962). Moreover, since there was a general negative correlation between australopith fossils and stone tools, where one was common the other was usually absent, it was equally reasonable to conclude that their association at Sterkfontein was co-incidental. The high level of technological competence detected in the Sterkfontein stone tools, which were similar in age to the Oldowan from Upper Bed I at Olduvai (regardless of whether one believed Leakey’s 1.75 million-year-old age, Koenigswald et al.’s (1961) more conservative date of 1.3 million years or took all the dates cum grano salis, Straus and Hunt 1962), provided further reason to be suspicious. It seemed improbable that Australopithecus could have developed from a non-cultural stage to an advanced lithic culture in the short span of time separating the Pink Breccia at Sterkfontein from the Middle Breccia, when there were more advanced hominins living contemporaneously in the region, for example Telanthropus capensis (now subsumed within Homo erectus) at Swartkrans (Robinson 1953). Lithics were also very few and far between, and certainly not in situ; assuming the Middle Breccia represented several millennia, the whole collection could be accounted for by just two to three discards per year finding their way into the sediment. Robinson (1961) suggested instead that Telanthropus had entered South Africa as the Middle Breccia began to accumulate, where they found an existing population of non-cultural australopiths. He also perceived a certain illogicality in those who uncritically accepted that australopiths had made stone tools (Washburn 1960), but rejected Dart’s (1957) ‘osteo-donto-keratic industry’2 as carnivore food refuse rather than humanly fashioned tools (Washburn 1957).

Zinjanthropus soon joined the menu. New excavations in Bed I and Bed II at Olduvai (Leakey 1961; Leakey et al. 1964) revealed that another more advanced hominin had existed contemporaneously with Zinj. Olduvai hominid 7 (OH7) was more gracile, more evolved and had a larger brain than Zinj (675–680cc to 530cc). Leakey decided that this hominin was much more likely to have made the Oldowan and he assigned it to a new taxon, Homo habilis (handy man), the oldest member of our genus3. A rough circle of stones on the living floor at DK I was interpreted as a windbreak, the earliest evidence of artificial structures. Higher up in Bed II at LLK II, an even more advanced skull (OH9), similar to those found at Steinheim, Broken Hill and Saldanha, was found associated with a Chellean industry (Leakey 1961). This provided Leakey with the only fossil evidence for the makers of the Chellean, the Atlanthropus specimens from North Africa occurring in his opinion with Acheulean tools (contra Arambourg 1955). The European and Asian fossils were also now known to be much younger, Middle Pleistocene or later (Howell 1960, 1966). Leakey may have done a complete u-turn over Zinj, who was now regarded as a victim or intruder, but he did so on his own terms. Leakey had by now also determined that in order to properly understand the meaning of patterns in the Early Palaeolithic record, archaeologists would need much better information on living apes, setting in motion the primatological careers of Jane Goodall, Diane Fossey and Biruté Galdikas.

The ‘New’ Archaeology: An American Revolution

Trigger (1989) traced the origin of the term ‘New Archaeology’, as well as many of its main tenets, to a 1959 Science paper by Chicago-based archaeologist, Joseph Caldwell (1916–1973). It was here that Caldwell (1959) examined trends in North American archaeology during the 1950s, observing that while its fortunes had been greatly improved by radiocarbon, a less celebrated but more important shift was occurring in the way American archaeologists thought about the past. He noted that up until World War II their chief concern had been the description of sites and the description (or definition) of prehistoric cultures, a particularistic and classically culture historical approach. In the optimistic and materialistic decades since (Trigger 1989), American archaeologists had developed a greater concern with cultural processes and more general explanations, similar to what Childe and Clark had been doing in Europe, but because North America contained evidence for both living and prehistoric ‘ethnographic’ societies, this was more directly informed by anthropology, particularly the cultural ‘neo-evolutionism’ of Julian Steward, Leslie White and Walter Taylor. This tended to regard human societies as resistant to change unless change was forced upon them and sought answers to more functionally orientated questions concerning the relationship between culture and environment, how culture change might occur without outside influences or individual acts of genius. Culture was not a collection of independent and equally important artefacts, but a functionally integrated system; Caldwell (1959, 129) supposed that “behind the infinite variability of cultural facts and behind the infinite and largely unknown detail of historical situations we shall discover the workings of a finite number of general cultural processes”. From these it should eventually prove possible to construct archaeological postulates, or hypotheses, which could be tested using appropriate cultural facts. The aim of archaeology, then, should be the explanation of culture change in terms of cultural processes. Culture history had been the necessary prelude to this new phase of research (Trigger 1989).

Photograph of two hominin skulls from Olduvai, of different species.

Figure 8.5 Handy and Tasty? Left: Olduvai hominin 5 (OH5), Zinjanthropus boisei; right: OH24, Homo habilis (this specimen was discovered in 1968) (Photos © Smithsonian Institution).

The next generation of archaeologists transformed this nascent New Archaeology into a decade-defining counterculture movement, shedding the shackles of tradition between the end of the Chatterley ban and the Beatles’ first LP.3 They were informally led, with almost “messianic fervour” and a “self-assured, domineering brilliance” (Meltzer 2011, 3; Straus et al. 2011, 328), by newly appointed Chicago professor, Lewis R. Binford (1931–2011). Binford had been a student of Leslie White, Alfred Spaulding and James Griffin at Michigan, the last, his first doctoral advisor, coming to symbolise to Binford everything that was wrong with traditional archaeology (Meltzer 2011). Over the next several decades, Binford, “pushed, pulled, or otherwise cajoled archaeology into becoming more anthropological, evolutionary and scientific” (Meltzer 2011, 3). He adopted a deliberately polemical and provocative style aimed at showing the advantages that New Archaeology brought and maximising the distance between the new and the old, frequently caricaturing the arguments of others in order to drive home a point. New Archaeology was thus delivered as a dramatic and (in their eyes) necessary rupture with the past, a “brand new epistemological perspective” (Binford 1968a, 17). It was the first real paradigm change since 1859. It even had a new vocabulary: jargon.

Trigger (1989) and Meltzer (2011) both provide excellent summaries of Binford’s work, the full scope and reach of which lies well outside this book. It is important here only to examine the impact of New Archaeology on the Palaeolithic, a period shrouded in uncertainty but long enough for evolutionary processes to operate, which involved species of human who were not fully ‘modern’, and where behavioural or ‘Processual’ approaches (as the New Archaeology would later become known) found a natural playground.

Binford’s starting point was that “American archeology is anthropology, or it is nothing” (Binford 1962, 217, citing Willey and Phillips 1958, 2). To be relevant to anthropological aims, however, archaeology needed to move beyond ‘explication’, meaning detailed classification and ordering, and into explanation (ibid.). In other words, Binford wanted to know the ‘hows and whys of culture change, not just the what, where and when’ (Binford 1964, 425). He was highly critical of the ‘normative’ view, which saw culture as a historically transmitted and homogenous whole, a mental construct of ideas shared by all members or by specific subsets of society and which formed the ideational basis of life (Binford 1965, 203). For Binford, even the normative studies that attempted explanation tended to be ‘particularistic’, seeking to understand specific events (Binford 1962, 217).

Binford advocated a different approach, one focussed on systems and processes, how the different parts (‘traits’, ‘variables’) of the archaeological record fitted together, and how change in one part was related to changes in other parts, and in the whole system. Like Caldwell, Binford (1962) argued that not all artefacts should be treated as equal and comparable, and one could certainly not assume that cultural similarities/differences reflected the blending and influences of different historical traditions. Historical explanations, where they could be justified, simply explicated mechanisms of cultural process such as migration, but if migration explained culture change, what explained the migration?

Following Leslie White (1959, 8), Binford regarded culture as humanity’s “extra-somatic means of adaptation to the physical and social environment, which was participated in, not shared” (Binford 1965). Technology formed the interface between people and their environment, what Steward had called cultural ecology, but it needed to be explored in a systematic rather than deterministic fashion, based on adaptive requirements and coping strategies. One should not be surprised, thought Binford, to find that different groups at similar levels of complexity in similar environments had developed similar solutions. Culture change could thus best be understood as the adaptive response in cultural systems to natural environmental change. Systems would tend to maintain equilibrium unless forced to change by external forces, although those changes would be internal to the culture, a social response within the system. There was little room in Binford’s world for spontaneous inventive genius.

Material culture was not, furthermore, simply technology. Artefacts needed to be understood in functional context and within the social, technological and ideological sub-systems of the total cultural system. True, archaeologists could not dig up a social system, but they did “excavate the material items which functioned together with these more behavioural elements” and “combined to produce a systematic and understandable picture of the total extinct cultural system” (Binford 1962, 218–219). Binford regarded some objects as technomic artefacts, the primary functional context of which was coping directly with the physical environment and in which variation was most easily explained in ecological terms, such as the nature, distribution and density of resources, the energetic cost (economy) of extraction, exploitation, loss, breakage and so on. Other artefacts were socio-technic, having their primary purpose in social sub-systems, while ideo-technic artefacts had their primary function in ideological systems, signifying and symbolising ideological rationalisations of the system, providing the symbolic milieu for enculturated individuals. Culture was also participated in differentially. Different social units (based on age, sex, status, etc.) performed different tasks, often at different locations, and people’s ideas about their culture would vary depending on how they personally functioned within it. More complex societies would have a greater number of cultural arenas. Cutting across all of Binford’s ‘technics’ were formal characteristics that one might call ‘style’, not directly explicable in terms of material, technological or functional requirements but serving as a way of symbolically promoting group solidarity, the basis of group awareness and identity.

In order to capture the full cultural system, Binford (1964, 1965) advocated a regional rather than site-based approach to analysis, looking at how mobile peoples moved across their landscape, the types of resources they encountered and the things they did at different times in different locations. He emphasised the importance of sampling and data quality, and the need for new analytical methods that went beyond style and chronology. He promoted a multivariate approach to the archaeological record, seeking hitherto unseen relationships between types of artefact, and he called for a new approach to ethnography. Instead of drawing direct analogies, archaeologists should be looking for general patterns of behaviour from which to build hypotheses about the past that could be tested using independent archaeological data (Binford 1967, 1968a, 1968b, 1968c). Archaeology should be a science that helped anthropologists to understand patterns in the past in terms of the ‘laws of human cultural behaviour’ (Binford 1968a).

The Bordes-Binford Debate

Binford’s early worked examples focussed on the many issues he had with the practice and interpretation of prehistoric archaeology in North America, but he soon found an equally rich source of irritation in the Mousterian of the Old World (Bordes 1961b; Binford and Binford 1966). A decade earlier, François Bordes (Bordes 1951, 1953d, 1961a) had introduced a standard typology and new standards of analysis to Lower and Middle Palaeolithic archaeology, which had allowed him to make objective, statistical comparisons between assemblages (Chapter 7). His methods identified six main variants and sub-variants in the Mousterian of south-west France: MTA (divided into A and B), Typical, Denticulate and Charentian (divided into Quina and Ferrassie: see Chapter 7). Bordes could detect no chronological pattern in the occurrence of the variants, no linear evolution over time; rather variants alternated seemingly at random, as exemplified at Combe Grenal (although, see Mellars 1965 who identified a definite chronological patterning in the occurrence of Ferrassie, Quina, MTA A and MTA B). Bordes considered three main hypotheses to explain the variation and alternation of industries: 1) that it represented seasonal variation in the toolkits used by a single tribe; 2) that it represented environmental variation (through time) in the toolkit used by a single tribe; or 3) that the variants represented the products of different tribes, each with its own tradition of stone tool manufacture.

It seemed unlikely to Bordes that Neanderthals had switched their toolkits four times a year. The thickness of sediment in each cultural layer alone suggested much longer time periods than a single season, and it was unrealistic to suppose that Neanderthals had rigidly used certain caves for certain things in certain seasons and that they continued this practice for such extended periods. In any case, the evidence from reindeer antlers and teeth suggested many caves had been occupied year-round (Bordes 1961b, 806). Environmental variation was also difficult to demonstrate–assemblages resembling those from south-west France were found in entirely different environments in Tunisia, while single geological layers could contain two different Mousterian variants produced under the same conditions (Bordes 1961b, 806–807). Bordes (ibid.) therefore concluded that the presence of different cultures within the Mousterian complex was an ‘established fact’.

From an ethnographical perspective, Binford and Binford (1966) argued that the type of stylistic variation posited by Bordes would be expected to show spatial clustering, reflecting geographical and social distance of the makers. In the Dordogne, however, they overlapped and interdigitated in time and space, suggesting that the tribes that made them had lived side by side, coming into continuous contact without apparently influencing each other in any way. This was simply not what one would expect among Homo sapiens populations and, if correct, would suggest that the social behaviour of Neanderthals was vastly different.

Binford and Binford therefore set out to test a fourth theory, that the patterning in the Mousterian was functional, appealing to the already well-worn ‘four-differences’ argument–that the different toolkits reflected different activities undertaken in different caves at different times. This might relate to the local environment, season of use, composition of the group and so on, but ultimately it was functional variation interrelated with the physical and social environment. The Binfords used early computers to perform multivariate factor analysis of 40 variables on three sites, two from the Levant (Yabrud I and Mugharet es-Shubbabiq) and one from France (Hoopeville), intuitively assigning tools specific primary and secondary functions, such as light and heavy-duty cutting, scraping, perforating, incising, sawing, planing, gut-stroping or hunting. From this they identified five main factors, reflecting four types of inferred activity:

· Factor I: manufacture of tools from non-flint materials

· Factor II: killing and butchering

· Factor III: cutting and incising (food processing)

· Factor IV: shredding and cutting (of plant materials)

· Factor V: killing and butchering

These roughly corresponded to Bordes’s MTA (Factor III), Typical (I), Denticulate (IV) and Ferrassie (II and V) variants, although there was no exact correlation. The underlying assumption made by the Binfords was that the objects found at a site reflected the activities undertaken there. The internal variability of outputs within Neanderthal groups (the percentage of Levallois, cores vs flakes, the frequency of finished tools) was therefore conditioned by context, the location and nature of raw materials, the location of intended use and the distance between them. Bordes’s so-called Upper Palaeolithic types (types 30–37 and 40) were just indicators of so-called Upper Palaeolithic activities, not of any direct evolutionary relationship between the Mousterian and EUP (Aurignacian or Perigordian).

For the Binfords, the excitement lay not in functional interpretations per se, but in what functional variation allowed them to infer about Neanderthal social systems. The fact that not every activity was conducted at a site suggested that Neanderthals had organised themselves into task-specific sub-groups to exploit different parts of their landscape, probably based on age and sex, and that at other times they had come together in different social units to do other things (family time, food sharing). These might have varied seasonally, but they could generally be broken down into extraction tasks (Factors II, IV, V), carried out at work camps and conditioned by the distribution and size of the targetted resources (rock, prey), and maintenance tasks (Factors I, III) carried out at base camps selected for other properties (protection, centrality). This would suggest that Neanderthals were socially little different to modern ‘fission-fusion’ groups of hunter-gatherers (Binford and Binford 1966). They later confirmed their preliminary analysis using 41 levels from Combe Grenal, where they identified a larger number of factors, combinations of which were related to climate change sufficient to alter the local ecology (Binford and Binford 1969). Thus, the alternation of industries related to cyclical changes in the use of the cave over time.

Bordes (e.g. 1969; Bordes and de Sonneville-Bordes 1970), by now the undisputed doyen of French Palaeolithic archaeology, remained largely unmoved by this Binfordian rhetoric. He saw no need for the rigid dichotomy between culture and function drawn by the Binfords. All cultures were a product of their environment. People would not develop implements for killing mammoth, techniques for hunting mammoth, and objects fashioned from mammoth ivory if there were no mammoth (Bordes 1969, 2). Merely by using culture, humans changed their environment–a stone tool changed the environment, as did a hut, or clothes, or fire. Palaeolithic humans were not the playthings of nature, but increasingly effective partners (ibid.). Stone tools told archaeologists about both the needs of people in the past and the traditions they developed to service those needs, not simply one or the other. Bordes thought it regrettable that many typological terms carried functional connotations–an accident of history–but this did not justify the use of functional typologies and he doubted microwear would provide definitive answers. Ultimately, Bordes thought that the actual function of a stone tool was irrelevant, if one’s main aim was to write (pre-)history.

Philosophy aside, there were significant analytical difficulties with the Binfords’s model (see Bordes and de Sonneville-Bordes 1970; Mellars 1996), not least functional inferences that had no firmer basis than those of Lubbock or Mortillet and the frankly unintelligible use of statistics. Nevertheless, by applying processual methods to a such a fundamental Old World problem, they brought New Archaeology to global attention and attracted a highly receptive new generation of postgraduate students willing to follow Binford down an increasingly anti-historical path. Not for them “migrations and invasions, man’s innate desire to improve himself, the relation of leisure time to fine arts and philosophy”, awful cliches that continued to appear in the literature (Binford and Binford 1969, 87). Instead, they set themselves the task of studying humans as just another component of an ecosystem, a unique culture-bearing component maybe but nonetheless one whose behaviour could be rationally determined. Humans were just ‘another unique species’ (Foley 1987a) and could be understood in the same evolutionary terms as any other.

Glynn Isaac and the Archaeology of Human Origins

Glynn Isaac (1937–1985) earned his first degree in Natural Sciences (geology, zoology, archaeology) from the University of Cape Town, before going up to Cambridge to study for an MA and PhD in Archaeology (Gowlett 1990). Cambridge in the late 1950s was anything but a theoretically stagnant pool full of floating cultural historians (Gowlett 1990). Among the teaching staff were pioneers of ecological approaches (Grahame Clark), statistical and geographical methods (Charles McBurney) and economic archaeology (Eric Higgs), while Isaac’s friends and classmates included those grappling with theoretical concerns (David Clarke), as well as new analytical techniques (Derek Roe). In 1966, he joined J. Desmond Clark on the teaching staff at Berkeley, forming another mutually beneficial collaboration.

Isaac’s doctoral research was based on his own excavations at the Acheulean ‘occupation’ site of Olorgesailie (Isaac 1968a, 1977a), an opportunity gifted by Mary and Louis Leakey who were too busy at Olduvai Gorge to give proper attention to other sites. Similar circumstances led to Isaac joining Richard Leakey’s team at Lake Turkana, Kenya, as project co-leader. The Turkana excavations, which thrust sites such as Koobi Fora into the international spotlight, were clinically executed and recorded, exposed primary-context accumulations of bones and stones over large areas, and provided Isaac and the 1970s generation of research students4 with a new window into the deepest past. Thanks to erroneous dating, they were also, for almost a decade, believed to predate the Olduvai sequence by three-quarters of a million years–an episode known as the ‘KBS Tuff controversy’ (Lewin 1987).

In 1969, Isaac’s postgraduate student Kay Behrensmeyer found stone tools in a volcanic tuff forming part of the Omo Formation of Northern Lake Turkana (southern Ethiopia and northern Kenya). Argon-Argon dating at Cambridge (Fitch and Miller 1970) provided an age estimate of 2.61 million years for the deposit, which although incongruous with the palaeontological evidence from the region (Maglio and Cooke 1978) and subsequently unrepeatable (Lewin 1987), was accepted over a biostratigraphical technique with a known poor track-record. The age made very little difference to technological or behavioural interpretations, other than giving an exaggerated sense of stasis, but the stratigraphical association of the KMN-ER 1470 Homo skull proved a major bone of contention for reconstructions of human biological evolution (see Lewin 1987 for a full discussion of controversy). By the mid 1970s, mounting problems led Isaac to seek a second opinion from an independent laboratory, producing a revised date of 1.6–1.8 million years (Curtis and Cerling 1975; Curtis et al. 1978), within the range of dates known from Olduvai. While an unfortunate and misleading interlude, it did not detract from the behavioural information contained in the Turkana record.

Isaac’s background, influences and attitude gave him a developing perspective on the past which, in terms of aims, ambitions and analytical rigour, shared much in common with the New Archaeology, although this was far from true of his theoretical and philosophical outlook, the way in which he practiced and wrote archaeology. Isaac (1971b, 124) regretted the “unnecessary sound and fury” that had “accompanied the numerous declarations by angry young men”, some of whom appeared to be engaged in a deliberate iconoclastic revolt against the older generation.5 This did not mean, however, that their ideas signified nothing nor that traditional archaeologists had anything to fear (cf. Hawkes 1968). The amount of data and number of professional archaeologists had grown exponentially since World War II and there was clearly a need for greater systematisation and a more explicit framework. Archaeology could not afford to rely on models of the past based entirely on the authority, opaque personal assumptions and nomenclature understood by only a few well-travelled connoisseurs. The record was now providing more than just an outline of major chronological divisions and was proving itself capable of supporting much richer exegesis, but this required the development of explicit new concepts. Isaac shared Hawkes’ skepticism for the growing cult of numbers among ‘statniks’. Data needed to be useful, relevant and meaningful. He looked forward to seeing where the generation game might lead, but whatever the outcome, it was an important stage in the growth of the discipline, showing that it had become more self-reflexive, more mature and more explicit in its assumptions (Isaac 1971b).

Isaac’s brand of archaeology was not a science, a social-science or a branch of the humanities–it was a unique pursuit in which all three met (Isaac 1971b). He was of course concerned about the chronological ordering of artefacts, but his key interests lay in questions about past human behaviour: hunting practices, diet, the length of occupation, group size and composition, mobility patterns, tool use and so on, all encapsulated within an evolutionary framework (Isaac 1967, 1972b). Any attempts at answering such questions, however, depended entirely on the quality of the data, and before drawing behavioural inferences it was vital to determine to what extent their sites had been rearranged by geological processes, especially on the so-called occupation floors where the frequency and distribution of artefacts was assumed to reflect human behaviour. In other words, archaeologists needed to understand site formation processes (taphonomy) in order to understand which parts of the patterning were humanly significant and which were not.

Isaac’s (1967) early experiments showed that in fluvial settings flakes could be preferentially removed, while elongated objects such as handaxes were susceptible to alignment in the direction of flow and frequently developed an upstream tilt. Even where flakes could be conjoined and bones re-articulated, some movement was the rule. Bones were also highly susceptible to dispersal by scavengers and destruction by natural weathering and erosion. It was not safe to assume that the species and anatomical composition of a faunal assemblage was a product of human hunting, butchery or transport practices, as certain bones were selectively destroyed depending on their structure and density, and on the size of the animal (Isaac 1967, 1971a, cf. Lubbock 1865, 183–184). Isaac (1969) thus began to wonder whether the differential distribution and density of cultural materials observed at Olorgesailie (Isaac 1968a) and Olduvai (Leakey 1967), where dense central clusters of bones and artefacts over an area of ~5m × 6m had been interpreted as indicating that a group of 10–20 people had been present, might reflect clast behaviour in fluvial environments not human aggregations at lakesides.

The main challenges were thus to understand the various processes involved while recognising that our archaeological windows were vanishingly small fragments of past systems, biased subsamples of already tiny samples. Indeed, Palaeolithic archaeologists had few useful sites to work with, which were widely and sparsely distributed across space, and spanned vast time periods during which we might expect to see profound changes not only in culture but also in human neurophysiological systems, particularly the brain (Isaac 1972a). Isaac also felt that the direct analogies offered by ethnography were dangerous when applied to the Lower Palaeolithic, because they made little accommodation for extinct modes of behaviour that exerted their own selective pressures and feedback mechanisms on biological, cultural and social systems (ibid.).

Subsistence patterns provided a key departure point for Isaac’s developing model of early (i.e. Oldowan) hominin behaviour. Having been a participant in the almost legendary Man the Hunter conference (Lee and Devore 1968), where it was noted that meat rarely made up more than 50% of the diet of modern African and Australian hunter-gatherers, Isaac (1971a) suspected that the meat component of the Lower Palaeolithic menu had been similarly overestimated. Some level of carnivory could be assumed–it was an ancient trait shared with other apes and monkeys–but the questions of when, why and how it had become intensified into the modern hunter-gatherer mode of life, and whether hunting created social pressures that promoted brain expansion and linguistic communication (Washburn and Lancaster 1968; Krantz 1968), remained unanswered. The data were certainly not yet well-enough understood to decide whether humans or climate had been responsible for the extinction of ~40% of large mammalian species (including Primates, carnivores, elephants, ungulates, camels, giraffes, bovids) claimed to have occurred at the end of the Acheulean (Martin 1966, 1967; Leakey 1966). That said, Isaac saw rare but strong evidence for specialised hunting by Acheulean groups, for example the dismembered Pelorovis (wild cattle) herd that had been trapped in mud at Olduvai BK II (Leakey 1957), the 50 giant gelada baboons at Olorgesailie DE/89B (Isaac 1977a) and the elephant remains at Torralba and Ambrona (Howell 1966).

The last two sites sat on the 40–45m terrace of the Rio Ambrona in central Spain, and showed numerous Acheulean horizons in fluviatile, lacustrine and colluvial sediments of Mindel-Riss age6 (Butzer 1965; Freeman and Butzer 1966; Howell 1966). Excavations at Torralba (Howell 1962) had revealed possibly intact occupation surfaces in Unit IIC, a grey colluvium, with ~700 artefacts, including hundreds of waste or manufacturing flakes (30%), bifaces (5%, predominantly cleavers) and flake tools (50%), plus possible bone and wooden artefacts. The associated mammalian assemblage contained equids, bovids and other large animals but was dominated by the remains of straight-tusked elephant, with several dense clusters of bones and stones in what seemed to represent two different types of subsistence behaviour (see also Freeman 1978).

One cluster contained the left side of an elephant in a semi-articulated state but with the skull and pelvis completely missing; none of the bones present had been broken to extract marrow (Freeman and Butzer 1966). The skeleton was associated with a lithic industry comprising flakes, three cores, two flake tools and a cleaver all made on raw materials obtained over 3km distant. Some tool manufacture had probably taken place at the site (ibid., 16), but most seemed to have been subsequently removed for use elsewhere. Three further clusters were dominated by broken elephant bones (some belonging to the same individual as the main cluster) and a sparse lithic toolkit consisting of cleavers, scrapers, cores and flakes.

Freeman and Butzer (1966) concluded that the main cluster was a kill-site where primary processing took place, stripping the flesh but not breaking the bones, and other areas were where the secondary processing occurred, where parts of the animal were divided and broken for marrow. Fire may have been used for protection on these subsidiary sites, allowing people to tarry for longer in relative safety. As side scrapers were present, these locations might also have been used for processing non-food resources. They could detect no evidence for culture change through the sequence, the variation over short distances probably relating to localised functional differences; it served as a warning against reconstructing culture history from such small windows into the past. The importance of Torralba and Ambrona to prehistoric populations probably lay in their location, on a N-S route that funnelled migrating game through a landscape dominated by pine parkland.

Despite the many shortcomings of the available data, Isaac felt sure that some conclusions could be drawn about the social organisation and economy of Lower and Middle Pleistocene humans. The African Pleistocene record contained a low-level background scatter of artefactual material, estimated to be about one object per square kilometre, presumably the result of random discards by peripatetic hominins (Isaac 1968b, 1975a). Within this scatter were dense concentrations of artefacts and other traces of occupations, such as animal bones. To Isaac these clusters represented well-defined home bases. The bone accumulations further confirmed that hunting (and/or scavenging) was of prime importance, and hinted at the fact that food was shared among the group. The size of some sites might also suggest that such groupings were larger than nuclear families, and perhaps constituted local bands. Isaac saw the combination of hunting, home bases and food sharing as a uniquely human behavioural package, one not seen in normal patterns of primate feeding but which had clearly developed very early in the evolution of Homo (Isaac 1967, 1969, 1971a, 1972a, 1972b). This further implied that early humans must have developed some form of containers, even though direct evidence was lacking (Isaac 1969).

A Kaleidoscope of Chronological and Geographical Patterning

In Europe, the use of elaborate numerical sequences of stone tool development died with the Abbé Breuil, perhaps the only truly qualified practitioner, in 1961. Five years later, F. Clark

Howell’s (1966) assessment of the European Lower Palaeolithic admitted only three or four stages of the Acheulean (Lower, Middle, Upper and Final) plus the Clactonian, all of which were highly variable with later industries quite likely to contain types from one or more earlier stage. John Wymer (1928–2006) felt that Breuil’s scheme had never had any success in Britain and attempts to assign individual handaxes to his stages had only resulted in confusion (Wymer 1968, 46). He was quite certain that several stages of the Acheulean were present in Britain–Early, Middle, Late-Middle and Late–but only the Middle (as found at Furze Platt and the Swanscombe Middle Gravel) and Late-Middle phases (as at Wansunt and the Swanscombe Upper Loams) could be distinguished with any confidence. He found no grounds for any serial subdivision of the Clactonian, which in his reading of the geology represented the earliest concrete evidence of human occupation in Britain, first landing immediately after the Anglian (Mindel, Elster) glaciation (Wymer 1961).

A similar trend towards simplification (or lumping) was happening in East Africa. By the end of 1963, the Leakeys had fully excavated four sites in Bed I and nine sites in Bed II at Olduvai, amassing a wealth of highly detailed spatial and chronological data, and identifying new geological marker horizons (formed of aeolian tuff) that could be traced over long distances (M. Leakey 1967, 1971). Based on this new evidence, Mary Leakey (1967) found no reason to amend the classification of the Bed I industries, they were all still Oldowan, although she did expand the toolkit from choppers, flakes and flake tools to include polyhedrons, discoids, spheroids (bolas stones?), proto-bifaces, heavy-duty (>50mm) scrapers, light-duty (<50mm) scrapers, burins and seven types of chopper, plus utilised-flakes, waste flakes and manuports (stones that were not natural to the location and which had presumably been introduced, but then not worked, by hominins).

In contrast, Louis’s scheme for Bed II, which had proposed a gradual transition from the Oldowan to the Chellean followed by a linear evolution through 11 stages of the Chelles-Acheul culture, was radically altered (M. Leakey 1967, 417). The new excavations had revealed several problems with the earlier stratigraphical sequencing, and it was now clear that sites once thought to belong to different periods and contain different cultural stages were the same age. Mary Leakey also suggested that the Oldowan Culture (not just a few Oldowan forms) continued in a developed state into Lower Bed II, with the objects once classified as crude Chelles I handaxes now regarded as ‘proto-bifaces’ of Oldowan manufacture. In this new model, true handaxes emerged in Middle Bed II, but there was no gradual evolution from Oldowan to Chellean to Acheulean. Rather, the Lower Acheulean appeared fully formed with simple but delicate handaxes and cleavers. Leakey suggested it was probably an intrusive culture from elsewhere. It retained Chellean types but these never occurred to the exclusion of Acheulean forms. She further observed that the Acheulean did not occur in all levels at all sites in Middle and Upper Bed II, but interdigitated with the Developed Oldowan, characterised by a few “ill-made small bifaces”.7

Ignoring the Binfords’s recent but well-known functional diatribe in France, Mary Leakey interpreted the pattern at Olduvai as representing the products of two distinct culture groups, who had occasionally come into contact with each other, the degree of that cultural contact measured by the variable frequency of handaxes in the Oldowan. Where they occurred in extremely low numbers, they might be little more than objets trouvés (Leakey 1967, 432). She further speculated that the Oldowan had been made by Homo habilis, and the Acheulean by Homo erectus, an intrusive species to Olduvai which had brought with it a new type of culture (1967, 1971). Later analysis (Leakey 1971) revealed two facies of the Developed Oldowan: Developed Oldowan A (DOA) from the Sandy Conglomerate in lower Middle Bed II which showed an increase in the frequency of sub/spheroids and in the number of light-duty tools; and Developed Oldowan B (DOB) in Upper Bed II, which included all the latter types plus some handaxes. There was thus an uninterrupted local evolution from the classic Oldowan in Bed I and Lower Bed II, to DOA in lower Middle Bed II and DOB in upper Bed II. This left Leakey with a new problem, that of differentiating the DOB from the Acheulean when both contained handaxes and both could occur in the same horizon. To resolve this, she rather arbitrarily decided that to qualify as Acheulean an assemblage must contain more than 40–50% bifaces (as a percentage of the tools), while the DOB always had less than 40%. Additional criteria were also used when necessary, such as the diminutive size and crude manufacture of many DOB bifaces. Indeed, in later publications Mary Leakey (1975) wondered whether, despite being dominated by typical Oldowan forms, the character of the handaxes at some DOB sites (such as the large and boldly flaked bifaces from MNK Main and TK Lower Floor) meant that they should really be classified as Acheulean.

Further excavation between 1968–1971 concentrated on the top of the sequence at Olduvai, Beds III, IV and the Masek Beds. Contrary to her husband’s earlier assessment, Mary Leakey (1975) could find no progressive trend in the manufacture of handaxes through the 1-million-year time span of the Acheulean at Olduvai, although there was a marked increase in the frequency of small flake tools and a concomitant decrease in choppers during this period. Each assemblage was consistent within itself, but distinct from the rest, which suggested to Leakey (1975, 491) that the

artefacts from each site or level were made by a group of people who had their own tradition in toolmaking and standard tool kit, from which they did not deviate to any appreciable extent. Such groups may, perhaps, have consisted of clans or tribes or even family units.

These Acheulean tribes had not been alone but had continued to share the landscape with the makers of the Developed Oldowan, which by Upper Bed VI times had reached a new stage (DOC) dominated by light-duty and ‘microlithic’ tools (48%), and with other new features such as pitted anvils and punches. The main factor distinguishing the two, however, was the use of manufacture and use of large flakes in the Acheulean and the lack of them in the DOB (Leakey 1975, 485).

Table 8.2 Distinguishing features of the Oldowan sub-stages and Acheulean (after Leakey 1975).

Industry

Key Characteristics

Acheulean

Relatively high proportions of bifaces, such as handaxes, picks and cleavers

Developed Oldowan C

Relatively high proportions of light-duty scrapers, with awls, outil écailles, laterally trimmed flakes, pitted anvils and punches

Developed Oldowan B

Relatively high proportions of spheroids, sub-spheroids and battered blocks, awls and outil ecailles

Developed Oldowan A

Relatively high proportions of choppers, spheroids and battered blocks, with new forms such as proto-bifaces and awls

Oldowan

Relatively high proportions of choppers, discoids and polyhedrons

Glynn Isaac’s work in East Africa had left him equally critical of the serial subdivision of stone tool industries. In his experience, the difference between industries of the same age was often greater than between the supposed evolutionary stages. He recognised only four divisions: Oldowan, Developed Oldowan, Lower and Upper Acheulean, the second and third of which might represent facies of the same industry. The differences between the Lower and Upper Acheulean were technological rather than morphological, the Lower Acheulean being characterised by simple workmanship with minimal, bold knapping indicative of hard hammers, the Upper Acheulean by more refined forms with shallow invasive scars typical of soft hammer percussion. There were no differences in elongation or length over time. The pattern of culture change seemed to Isaac (1969) to be one of thresholds or major ruptures, rather than gradual change; long periods of stasis followed by great leaps forwards. Features were either present or absent with no intermediate expressions. A similar sequence was seen in North Africa, where the quality of the evidence had dramatically improved since Breuil’s wartime visit and where Biberson (1961, 1967) and Balout (1967) now detected the local development of handaxes from an early pebble-tool culture similar to the Oldowan into Early, Middle and Evolved Acheulean.

Isaac (1972a) also returned to the often-overlooked fact that Early Palaeolithic stone tool technology was marked by two seemingly opposite trends–static conservatism and dynamic variation. Conservatism was often taken as an indicator of cognitive limitations and the resistance of primitive culture to change, although another way of viewing it was that the tools kits were, in selective terms, in equilibrium with the level of behavioural adaptation and ecological setting (Isaac 1969). They were generalised enough to cope with multiple situations. The most complex artefacts, thought Isaac, might provide an index of the mental capabilities of the makers, while the presence of standardised forms might provide a means of assessing the linguistic capacity of Oldowan and Acheulean humans, assuming that the rules behind toolmaking were comparable to those employed in language (Chapter 9).

Handaxes and other ‘large cutting tools’ showed greater standardisation than other forms and were clearly made to a set of arbitrary design rules, although Isaac (1977a) remained equivocal whether the various forms represented ‘real or users’ types’ or whether they were modern constructs created for the convenience of archaeological classification and discourse (see also Callow 1976). Isaac’s statistical analysis of the tools from Olorgesailie indicated that different classes of biface–handaxes, cleavers, picks, trihedrals knives and discoids–were “recurrent improbable combinations of attribute states” (ibid., 120), but he did not consider them to be real modalities, rather “arbitrary zones within a structured continuum” (Figure 8.6). Cleavers did appear to form a “modality that is weakly separate from handaxes” (ibid.), but even these could be seen to blend into classic handaxes via chisel-ended forms (ibid., 123). Small tools on the other hand were unstandardised and appeared to have been made on an ad hoc basis for functional purposes, either to modify an edge or the shape; core and flake working was simply widespread and generalised (Isaac 1977a). Neither had much taxonomic value, their importance lay in what they could reveal of human behaviour.

Diagram showing a central modal form, with other forms extending off it like tendrils plus a 3-D graph showing the frequency of the different types.

Figure 8.6 A field of morphological gradients. A) Diagram illustrating Isaac’s concept that categories of large cutting tools formed zones within a field of morphological gradients, this particular scheme constructed using the forms found at Olorgesailie; B) The frequency of major categories presented as a 3D histogram (redrawn after Isaac 1967).

Variation within and between stone tool assemblages had been interpreted in various ways, but it seemed to Isaac (1972a) that it was controlled by three factors–the raw material used, the design concept and skill of the worker. The design concept was related to two further factors, the function of the tool and the socially transmitted ways of making that tool. There was no need to regard these, or the third possibility that form and technology was controlled by raw materials, as competing hypotheses.

There was for Isaac, moreover, a compelling alternative, one that owed more to David Clarke’s (1968) systematic formulation of cultural processes than functionalist ways of thinking. Isaac (1969) argued that the pattern of variation in time and space was stochastic; craft norms had randomly ‘walked’ or ‘drifted’ over time, with no direction or apparent sense. He drew the analogy of a ball bearing dropped into a bowl, its path unpredictable and marked by numerous short-lived states. This, he argued, should be the null hypothesis, that variation in culture was largely the outcome of unpredictable micro-cultural processes, conceivably influenced by idiosyncratic styles of individual craftsman or isolated hunting bands (Isaac 1969, 19, 1975a, 1975b). Where innovation had occurred, it was rarely transmitted: low population density and poor communication networks would hinder its spread, and equilibrium would soon be restored through stochastic processes.

With such a “kaleidoscopic view of chronological and geographic patterning… classifiable palaeocultural entities might be apparent”, but it was never likely to provide the high-resolution cultural histories of later prehistory (Isaac 1972b, 387); sites were just too few and far between in both space and time. The propensity of the record to throw up new variants, such as the scraper-core or core-scrapers from Karari escarpment (Harris and Isaac 1976), adequately showed the dangers of assuming our knowledge of variation was complete. Before any historical narrative was even attempted some clarity was still urgently required over basic terminology and correlations (Isaac 1975a). Sites once believed to be Middle Pleistocene in Africa had been shown by new dating techniques to be of a much older, Lower Pleistocene age. Africa had been occupied for a million years before humans first stepped foot into Europe, making the use of such terms as Early Acheulean in Europe totally inappropriate.

It was also wise to remember, Isaac reminded, that even though 90% of the literature involved stone tools, archaeologists still had no idea what they were used for or how they otherwise functioned in the economy and ecology of the makers (Isaac 1969, 1971a). He thought it was necessary to re-orientate stone tool research, away from imponderables and towards more directly observable phenomena. Isaac (1972b) noted that the number of standardised tool types increased over time, from eight in the Oldowan, 12 in the Lower Acheulean, 17 in the Middle Acheulean and more than 40 in the Late Acheulean. Design complexity and levels of standardisation also increased over time, within wide tolerance limits (Isaac 1972b, 1975a), while variation became more localised and geographically distinct. Together these probably showed the evolution of greater cognitive power and greater social standardisation, with more elaborate rule systems and more rigid boundaries between groups.

Isaac was developing and testing models just as much as Binford (cf. Toth and Schick 1986), but his philosophy was to keep all possibilities as a series of working hypotheses to be discussed and elaborated, knowing that the data just wasn’t yet good enough to support dogmatic assertions of any theoretical flavour. The research programme he designed and implemented at Koobi Fora from the mid-1970s was thus directly aimed at providing new types of information to answer a new set of questions, far removed from the old obsession with ordering stone tools. Placing modern hunter-gatherers at one end of the scale and chimpanzees at the other, Isaac (1975a, 1975b, 1976b, 1977b) set himself the task of understanding where on this spectrum Oldowan and Acheulean humans had sat, in terms of speech; tool use and manufacture; structures; meat eating, food sharing and collective provisioning; the division of labour by age and sex; the organisation of daily life around temporary fixed locations; family units; the use of symbols; and the observance of detailed rule systems that regulated the mode of execution of most individual and social functions. Isaac and his team were more than up to this daunting task.

The Exaggerated Death of Culture History

Despite the processual onslaught, Culture Historical approaches did not simply shrivel and die. Traditional archaeologists often agreed with the key aims of the New Archaeology: to make archaeology more scientific, more ecological, more anthropological and to explore more fully the meanings of similarities and differences in the archaeological record (Isaac 1971b; Trigger 1989). Indeed, some had been actively pursuing this agenda for years (Desmond Clark, for example); yet they now felt alienated by the combative rhetoric, the jargon, the poorly explained statistics and the meaningless facts, which threatened to overwhelm the subject like a ‘tide of mud’ (Hawkes 1968). They also resented the implication that historical approaches were outmoded and theoretically naive. However much New Archaeology wanted to share in the prestige of the natural sciences and whatever methods it used, its underlying aims were still historical (ibid.). The “scientific facade was far too grandiose for the modest historical building behind it”, and it was dehumanising (Hawkes 1968, 257).

The well-publicised Bordes-Binford debate (Binford and Binford 1969; Bordes and de Sonneville-Bordes 1970) did little to sway French archaeological opinion during the 1970s. Many French Lower Palaeolithic publications of this period, whether introducing new finds or discussing old materials, were content to apply Bordes’s typological system, adopt or dispute his preferred dating (cf. Bordes 1968) and follow a cultural interpretation (Lejards 1965; Huguenin and Rigolot 1971; Huguenin and Meunier 1972; Guillaume 1974; Lumley 1975. Meignen 1975; Tuffreau 1976, 1979 a, 1979b; Tuffreau et al. 1975; Monnier 1976). It was equally irrelevant to the emerging technological school led by Jacques Tixier (see Chapter 9), which shared with processual approaches (see Collins 1975 and other papers in Swanson (ed) 1975) an interest in experimental archaeology to understand technology and allow more precise analyses of how prehistoric knapping events had unfolded, but which took a far more humanistic view based on the anthropological writings of André Leroi-Gourhan, putting the knapper, rather than external factors, at the centre of events (Tixier 1967; Lenoir 1975; see Chapter 9). Indeed, for Lenoir (1975), all hominin knappers consciously or unconsciously imposed their cachet or ‘stamp’ on the objects they made. Debitage (waste) had individual style, retouch had individual style and shape had individual style, but these were all formed within a specific tradition of manufacture from which individual technical and aesthetic habits had been acquired. The two were not divisible–the group influenced the behaviour of the individual, and the individual likewise affected the behaviour of the group. The function of any object was a social construct.

If France was resistant to brash American exports, Britain was caught entirely unprepared. When Derek Roe (1937–2014) began his PhD in 1961 there had been no new synthesis of the British Palaeolithic since the second edition of John Evans’s Ancient Stone Implements in 1897. Roe therefore set out to survey all existing collections of Lower and Middle Palaeolithic artefacts (Roe 1968a) and use the largest and best contextualised assemblages to examine variation in British handaxes. To do this Roe (1964, 1968b) developed (in collaboration with another of Charles McBurney’s students, Glynn Isaac) a new morphometric method of recording handaxe shape that relied on a series of linear measurements and indices, and a unique system of graphical representation, the tripartite diagram. This could be used to examine whole assemblages statistically and without recourse to formal typology (Figure 8.7).

Diagram showing Derek Rtoes methods of measurement and representation.

Figure 8.7 Roe’s (1968b) Morphometric system. Top: system of measurements and indices; middle: tripartite graphical representation; bottom: traditions and groups after secondary sorting.

When applied to the 38 most intact and stratigraphically secure handaxe assemblages available, Roe found that while there was considerable variation within each site, the assemblages naturally divided into two main groups: one consisting of assemblages dominated by pointed handaxes, the other by ovate handaxes. Metrical cleavers made up just 2% of the British sample, and by Tixier’s (1956) definition most were not cleavers at all, just handaxes with tranchet edges (see White 2006a). Using the finer levels of shape variation within each category, combined with a series of technological characteristics (tranchet tips, twisted tips, twisted edges), Roe further divided the data into seven groups. Pointed group I contained most of the pseudo-cleavers, often in association with concave-edged ficrons. Pointed group III contained only one site, Wolvercote, and was defined on the basis of its characteristic ‘slipper-shaped’ plano-convex handaxes. Ovate group V contained crude hard-hammer handaxes, while group VI had high frequencies of twisted ovate handaxes. Groups II (pointed tradition with ovates) and VII (ovate tradition with less pointed-ovate forms) had fewer diagnostic features and seem largely to have consisted of those assemblages that did not belong anywhere else. Group IV was a dustbin group that contained the most poorly provenanced assemblages.

Roe (1968b, 1981) interpreted these groups as reflecting different traditions of manufacture, encompassed within a general evolutionary trend in technique and refinement.8 Unfortunately for Roe, he could find little chronological or geographical patterning in his data, most sites could still only be tentatively assigned to one of the traditional Alpine phases or one of the pollen-based British interglacial periods. He could, on purely typological grounds, detect Early (group V), Middle (group I and II), late Middle (groups VI and VII) and Micoqian (group III) phases, but both the traditions and the groups seemed to overlap in space and time. Forms were certainly not exclusive to the traditions or groups they defined, there was no question of one tradition or group wholly succeeding the other, and there was no linear sequence of replacement. Group V did appear, on geological grounds, to be the earliest (Roe 1975), and at Swanscombe, handaxes evolved from typical Middle Acheulean points to Late Acheulean ovates, but the most recent excavations at Hoxne (Wymer 1974) had revealed the opposite succession, from ovates to points. As a concession to New Archaeology, Roe (1968b, 75) acknowledged that “variation of a purely functional kind” was “likely to be superimposed” upon any cultural pattern, the same bands of hominins perhaps leaving industries of different types at different places to service immediate local needs, but stressed that the data and the dating was simply not up to the task. The sequences were too coarsely understood and it was not even possible to state that two industries had been made in the same millennium. To overcome these difficulties, Roe later recruited Lawrence Keeley to examine function directly through high-powered microscopy, but his results on handaxes were inconclusive (see preceding).

In a contemporary study, Collins (1969) used assemblage-level methods akin to those devised by Kleindienst and Bordes to describe and seriate the British Lower Palaeolithic, identifying eight stages of the Acheulean (oldest to youngest: Barnfield I and II; Cuxton I and II; Northfleet I-III and Elveden) and four of the Clactonian (Ricksons, Jaywick I and II and Mildenhall). He rejected raw material and functional explanations, thinking the dichotomy between function and culture to be a false one, and saw no evidence to refute the traditional hypothesis that this variation reflected communities of shared cultural tradition. Using distribution maps and environmental data from pollen and fauna, Collins proposed a recurrent association between the Acheulean, the hunting of big game and open environments, and between the Clactonian, forested habitats and a diet based on fish, small mammals, birds and vegetable resources. Collins believed that the Clactonian was culturally linked to the pebble and chopper industries of eastern Asia and eastern Europe, who had been the first groups to colonise Britain after the retreat of the Elsterian (Mindel, Anglian) glaciation because they were pre-adapted to the rapidly expanding forested environments. He associated this non-handaxe lineage with Homo erectus and early Neanderthals. Handaxes had been made by another group of humans, early Homo sapiens. The paper, published with comments in the US journal Current Anthropology, was not well-received, with Lewis Binford commenting that Collins’s hypotheses were not hypotheses, in the scientific sense of the word, and Sally Binford coldly stating that it captured almost every fallacy possible in the writing of prehistory. Others were kinder but equally unconvinced, Isaac doubting whether the absence of handaxes was a good criterion for inferring a continuous cultural tradition although he agreed that the “handaxe design range” was not

simply a natural outcome of the mechanical regularities of stone fracture, and there is at least the basis for defending the view of Collins and others that handaxes are sufficiently improbable to represent a specific thread of cultural continuity which links dispersed series of occurrences.

(Isaac in Collins 1969, 306)

Based on his own new and well-targeted excavations in the Thames Valley and East Anglia, Wymer reached the firm conclusion that the Clactonian represented the earliest occupation of Britain, and that no convincing evidence existed for a human presence, either handaxe- or non-handaxe-making, before the Anglian (Wymer 1961, 1968, 1974). From the environmental profiles from Clacton (Singer et al. 1973) and Swanscombe (Waechter 1973) Wymer determined that the Clactonian had first arrived during the dying stages of the Anglian glaciation, survived throughout the pre-temperate and early temperate phases of the Hoxnian interglacial (Ho biozones I and II), before disappearing sometime before Ho IIc. It is at this point that the Acheulean made its earliest appearance, as represented by the lake beds at Hoxne and the (somewhat later) Middle Gravel at Swanscombe. This suggested to Wymer that the Acheulean infringed on long-held Clactonian territory, presumably driving them out or absorbing them (Figure 8.9).

A graph showing Wymer’s 1970s Palaeolithic framework for Britain, with an initial Clactonian followed by a confusion of Acheulean.

Figure 8.9 John Wymer’s chronological scheme for the Palaeolithic of Britain. Based mostly on the evidence from Swanscombe, Hoxne and Clacton, Wymer was already beginning to see the difficulties of correlating the wider climatic record with the palynological record. Note that although he has the traditional glacial-interglacial names, he indicates nine not five climatic fluctuations (after Wymer 1974).

The presence of contemporaneous point-dominated and ovate-dominated handaxe assemblages at Swanscombe and Hoxne further showed that different local traditions had existed within the Acheulean (Wymer 1974). There was still no obvious chronological pattern as far as Wymer could detect, and only a Middle and late Middle Acheulean could be identified with any confidence, although there were strong hints of an Early Acheulean. Wymer saw no distinct line between the Acheulean and Levallois, which sometimes occurred together at Late Acheulean sites in the Thames, and little evidence for serial subdivisions within Levallois, just Early and Late phases each typified by the industries found beneath and above the coombe rock at Northfleet. Wymer (1968) also developed his own typological system (Figure 8.8), which he presented graphically as a matrix of variation that could easily accommodate, even highlight, sub-type levels of variation. This method was almost never used outside Britain, but like Roe’s method was widely adopted by generations of British scholars keen to communicate their findings in a regional context. It is still not uncommon to see British workers using both methods, whatever their theoretical leanings, although three-dimensional scanning and computer assisted analysis are possibly in the process of making these schemes redundant.

A grid showing the types and type labels in Wymer’s typology

Figure 8.8 Wymer’s typology, not entirely dissimilar to Isaacs’s field of morphological variation (after Wymer 1968).

The terrible truth faced by workers in the 1970s and 1980s was that, despite more than 100 years of study, the European Palaeolithic collections now contained little of value for answering the sort of behavioural questions in vogue in African archaeology. What Collins did achieve, though, was to put the European Acheulean firmly within Binford’s crosshairs (Binford in Collins 1969; Binford 1972).

In other regions of the world, local archaeologists were largely playing catch-up with the more traditional developments in the classical British and French schools of thought, by abandoning Breuil and adopting both a simpler framework and more statistical methods of analysis. But their data was far from suitable for tackling the processual agenda. In Israel, Isaac Gilead’s (1970, 1977) doctoral thesis used typology, technology and a metrical system based on Roe’s method to explore the growing number of handaxe sites in the Levant, most of which were poorly age-constrained and which included surface scatters. The earliest records from the region were undoubtedly the Oldowan and Early Acheulean industries from Ubeidiya, which were initially dated to 620±120 ka BP (Horowitz et al. 1973) and since to over 1.2–1.4 million years old. The Middle Acheulean was found at Latamne (Clark 1966) and Gesher Benet Ya’aqov (Stekelis 1960), where it was marked by an increase in true cleavers and use of flake blanks from prepared cores for both cleaver and handaxe manufacture. Gilead’s Late Acheulean comprised the earlier cordiform-handaxe dominated sites at Maya’an Barukh (Stekelis and Gilead 1966) and Qatafa D2, and the later more pointed handaxe assemblages in which Levallois became more prominent, as found at Qatafa D1, Evron and Kissufim. His final expression of the handaxe was found in the Acheulo-Yabrudian facies at Tabun and Yabrud. Cleavers were found throughout the Acheulean time range, although they were not typologically consistent with either Tixier’s or Kleindienst’s schemes (Gilead 1973). Metrically, Gilead was able to detect some trends, but these were masked by raw material quality and form. It was also not possible to map industries onto vegetation or climate; the distribution of sites mirrored modern areas suitable for human occupation and were close to the basalt and limestone regions for raw materials.

Archaeology in post-partition India was controlled by the government, whose research agenda was heavily focussed on later periods (Khatri 1962a; Allchin 1963; Ghosh 1971). Writing in 1962, Khatri (1962a) was able to name only two projects operating under the auspices of the Indian Council of Scientific and Industrial Research: Explorations for the Remains of Early Man in India, awarded in 1957 to Dr M.R. Sahni, president of the Palaeontological Society of India, and Investigation of the Pleistocene Deposits in the Valleys of Chambal, Narmada and Godavari, recently awarded to Khatri himself.

Despite the long history of colonial research, the Pleistocene of India was not well understood (Malik 1965) and the concerns of these projects remained largely culture historical. They saw the addition of a new variant, the pebble-tool Mahadevian (Khatri 1962b) which “anticipated the Chellean” and suggested that the long-mooted parallel cultures (Soanian and Acheulean) were part of an evolving lineage that mirrored developments seen in Africa, but which perhaps occurred independently. Ghosh (1971) later suggested that they formed part of the same cultural complex, bound together by Levallois technology and reflecting local raw material conditions. As more excavated assemblages and associated environmental data became available, such as those from the Bhimbetka rock shelter (Misra 1978) and the open-air ‘occupation floors’ at Chirki-On Pravara (Corvinus 1968, 1970) it also became possible to use statistical methods and examine finer levels of assemblage variation. Using metrical and statistical methods drawn from Bordes and Roe, Joshi and Marathe (1977) dismissed Zeuner’s idea that cleavers were associated with forested environments, noting that handaxes exceeded cleavers by three to six times in moist areas, moist-dry areas and forested areas. Cleavers did, however, grow in importance from the early to Late Acheulean, during which time handaxes also became more finely made.

In eastern Europe, the discovery of the ‘Buda’ Industry, evidence of fire-use and human remains from Vérteszöllös, Hungary (Kretzoi and Vértes 1965), reaffirmed the expected absence of handaxes and the existence of an independent pebble-tool industry (of Clactonian aspect) in this region (Valoch 1968). Further evidence for the absence of handaxes from eastern Asia did the same for the AC-TC industries east of the Movius Line (e.g. Soejono 1961; Walker and Sieveking 1962; Ikawa-Smith 1978).

The vast lacuna in Central Asia, however, made it difficult to establish any direct links between these regions. Critically reviewing the evidence from the USSR (including Azerbaijan, Armenia, Georgia, Russia, Ukraine and Soviet Asia), Klein (1965, 1966; Bahder and Klein 1965, Figure 8.10) could find no evidence of Chellean occupation, and only three convincing Late Acheulean sites (Azych Cave, Azerbaijan; Satani-Dar, Armenia; Kudaro I, Georgia, see also Bahder 1968). Klein noted several difficulties in working in this area–the language barrier and methodological differences (specifically the failure to adopt Bordes’s typology) being chief amongst them. While occurrences had been claimed, it was not clear to Klein whether the terms Acheulean and Chellean were being used in a temporal rather than cultural sense, and if the latter whether this followed Breuil or Commont. Rogachev (1964) similarly concluded that humans had not settled the Russian Plain until the Upper Pleistocene or Late Acheulean/Early Middle Palaeolithic, while from a typological perspective, the earliest occupation of Mongolia was Levallois-Mousterian or Late Acheulean (Gabori and Merbs 1963). But did this mean that humans had been absent until relatively late, or that the earliest occupation had been a non-Acheulean culture like the Soan or AC-TC, hints of which had been found in Kazakhstan (Bahder 1968)? None of these uncertainties were helped by the lack of a decent Quaternary framework for the region, which meant that even those rare finds not made on the surface were barely datable (Rogachev 1964; Ravskii et al. 1968).

Line drawings of bifacial stone tools from sites in the old Soviet Union.

Figure 8.10 Back in the USSR. Top: Chellean handaxes (surface finds) from Satani-Dar, Armenia. Middle: Late Acheulean handaxes and points from the Lashe-Balta, South Ossetia. Bottom: Late Acheulean cores and a point from Satani-Dar (after Panichkina 1950). Scale = 3cm.

The once-pivotal South African record was in no better state (Mason 1962). The first controlled excavations and quantitative analysis of the Stellenbosch type site at Bosman’s Crossing (Seddon 1966) produced samples too small (n=134, with seven handaxes/cleavers) for valid comparisons with other key sites, such as Hangsklip and Montagu Cave (Sampson 1974). Butzer’s (1974) geo-archaeological analysis of the sites of Doornlaagte (Mason 1966) and Rooidam (Fock 1968) confirmed the impression that these sites preserved intact Acheulean living floors, but details of the finds were (and remained) unpublished and both lacked the bones or botanical remains needed to infer the types of activities in which humans had engaged. Furthermore, apart from the site of Amanzi Spring (Inskeep 1965), which had been dated using an inappropriate method (radiocarbon, Vogel and Waterbolk 1967), the South African Acheulean and Fauresmith remained ‘irremediably’ undated (Deacon 1966; Isaac 1975a).

More robust statistical analysis and mapping ultimately led Humphreys (1970) to the conclusion that the supposedly characteristic Fauresmith types of South Africa were also present in the Acheulean, and were almost exclusively made on lydianite, in contrast to the lavas and quartzites used for Acheulean types. Furthermore, the distribution of Fauresmith assemblages was perfectly correlated with outcrops of lydianite, with one exception–Riverview Estate–which, somewhat alarmingly, had been noted as different to other Fauresmith occur-rences anyway (Humphreys 1970). He concluded that the Fauresmith should be abandoned, and the assemblages subsumed into the Late Acheulean. Only well-dated and properly excavated sites would truly solve the problem of lithic variation (ibid.).

Large excavated assemblages were also at a premium in North Africa, although this region had the advantage that the rhythms of marine transgressions and regressions provided a useful chronological framework that allowed correlation with Europe and East Africa (Tixier 1956, 1967; Biberson 1961, 1964, 1967; Balout 1967; Balout et al. 1967). They were good enough, though, to allow French archaeologists Balout and Tixier to pioneer a new and explicitly technological approach to lithics, and provide a bespoke regional typology built on a Bordesian chassis. This record was not, however, used to study the potential links between environment, technology and cognition, but for culture historical reconstruction, the validity of which outside the coastal zone was unknown.

Given the parlous state of the global Palaeolithic archive, which simply had not been collected with the type of questions now being asked in mind, it is not surprising that discoveries and syntheses still involved detailed descriptions and culture historical approaches, which remained the basic prerequisite for any interpretation. Where sites were few and assemblages small, as in south-east, north-east and north-central Europe, type fossils also continued to form the basis of culture historical reconstruction (Gamble 1986, 178). If measured by the number of practitioners rather than the amount of noise generated, culture history has always been alive and kicking.9

Transitions in Early Prehistory: Putting an End to the Lower Palaeolithic

The terms Lower, Middle and Upper Palaeolithic had been in use since at least the beginning of the twentieth century, although it is difficult and probably unwise to attempt to pinpoint a precise origin.10 In their earliest form they were merely summary devices. Déchelette (1908a) used Palaeolithic, Quaternary and Pleistocene synonymously and interchangeably, his chapters on the Lower Palaeolithic (Acheulean and Chellean epochs), Middle Quaternary (Mousterian epoch) and Upper Quaternary (Reindeer epoch, Aurignacian and Solutrean phases), blending geological and archaeological nomenclature. Obermaier (1909) also recognised an Upper Palaeolithic starting with the Aurignacian, but assigned the Mousterian and all earlier industries to the Lower Palaeolithic. Breuil (1913, 1932a, 1936) frequently made use of Lower and Upper divisions, but while he occasionally described materials as Middle Palaeolithic, it was largely irrelevant to his idea of gradually evolving phyla. Other authors of the 1920s–1940s likewise used them as convenient terms that were apparently so well known that no definition was necessary, even when they were being replaced (e.g. Goodwin 1926). They also provided some level of safety in a discipline that was beginning to realise that naming global entities after individual European type-sites was not entirely appropriate or useful.

By the 1950s (e.g. Bordes 1950a, 1950b) the Lower Palaeolithic had become shorthand for the various chopper-core and handaxe industries of the Lower and Middle Pleistocene, Middle Palaeolithic had become synonymous with the classic Mousterian industries of the Upper Pleistocene (and the Neanderthals), while the Upper Palaeolithic was marked by the appearance of blade-based industries and Homo sapiens during the later Würm. By adopting a chronological division between the Lower and Middle Palaeolithic, Bordes conveniently elided the issue that the key typological markers, handaxes, Levallois and well-made scrapers, were found on both sides of the line. He was aware of this tension, which he tried to resolve by invoking a Mousterian stage of evolution that had occasionally been attained by pre-Würmian Acheulean and flake-industry populations (Bordes 1977). This still meant, though, that sites such as Baker’s Hole, England (Wymer 1968), and Biache St Vaast, northern France (Tuffreau 1979a, 1979b), which were technologically and typologically indistinguishable from numerous Middle Palaeolithic assemblages yet dated to before the penultimate glaciation, were Lower Palaeolithic. Undated sites with both handaxes and Levallois could be Upper Acheulean, Early Mousterian or MTA, depending on the frequencies of each type.

Such a permeable boundary was something Tuffreau (1979a, 1982) and others at the 1980 Haifa Conference celebrating the 50th anniversary of the Mount Carmel excavations could no longer accept (Ronen 1982 and papers therein). In reviews of the Lower and Middle Palaeolithic in Britain, Southern and Northern France, Italy, the Levant, Germany, Czechoslovakia, the Balkans, India, Africa and South-East Asia, there was near consensus that a more appropriate division between the Lower and Middle Palaeolithic was the emergence of Levallois and a range of sophisticated flake-tools, after which handaxes, where they had been present, disappeared at varying rates. There was now an early Middle Palaeolithic of late Middle Pleistocene age, and a later Middle Palaeolithic of Upper Pleistocene age. Although primarily a technological and typological definition devised largely for the purposes of culture historical narratives, which some would have preferred to discard as an outmoded and unnecessary level of classification (Bar-Yosef 1982), the appearance of Levallois at the new Lower-Middle Palaeolithic transition did have implications for wider behavioural and social changes, if such could be found (ibid.).

The Haifa definition was quickly and widely adopted by archaeologists of most theoretical viewpoints. Regardless of how one wished to explain the change or how long the transition lasted, the permanent appearance of Levallois tolled the death knell for the Lower Palaeolithic way of life.

Calm Before the Storm

Binford’s early forays into the realm of Palaeolithic archaeology were limited; indeed had it not been for the popular impact of the functional debate with Bordes, traditional culture historians of the period might barely have noticed the New Archaeology at all. Binford’s crusade was fought on the battlegrounds of method and theory, with well-aimed targets in the fields of North American archaeology (see papers in Binford 1968d). With two exceptions (Sally Binford’s paper on Levantine Middle Palaeolithic variability and James Sackett’s on the Upper Palaeolithic of south-west France), the New World similarly dominated the seminal New Perspectives in Archaeology (Binford and Binford 1968), a volume that “showcased the new archaeology” (Meltzer 2011, 10) and “set the agenda for a whole generation of research” (Shennan 1989, 832), particularly in America but increasingly abroad.

At this point, however, Lewis Binford’s research took a “right-turn” (Binford 1983a, 106). Further work with Sally Binford on the fauna and lithics from 55 excavated Mousterian levels at Coombe Grenal left the pair with an “embarrassment of recognisable patterning, and more correlations between things than anyone had ever imagined” (Binford 1983b, 66). And that was the problem. Archaeologists tended to rely on their imaginations to provide the behaviour organisation and conditions behind the archaeological record, but it was abundantly clear to Binford (ibid., 67) that “knowledge of the dynamics of technological organisation and the organisation of tools and work was inadequate for the development of a theory of site formation”. We could not rely on our imaginations to understand the behavioural patterns behind the archaeological record, because we were totally ignorant of what most of those patterns might mean. We had no idea of what other things might have to be equal in order for our empirical propositions11 about the past to be properly tested. We needed better insights into the organisational strategies of living people, how material objects were generated and used within cultural systems, and how they varied over time and space, and under different conditions. We needed a way of translating the static archaeo-logical record into a dynamic behavioural system. Thus, in the spring of 1969 Binford went to the American Arctic to begin studying site formation processes among caribou-hunting Nunamiut people of Alaska. Fully aware of the differences between modern human hunters and Neanderthals, and that Alaska was not a good analogy for Pleistocene France, Binford focussed on how people organised themselves in their landscape, the ways in which technology and material culture slotted into the system, and on the different patterns of debris they left behind them. This new line of research would eventually provide Binford with a powerful set of observations from which to understand the causal processes and conditions behind archaeological patterns and variation, a methodological bridge between the unanimated past and living present. Binford (1977b, 1978a) called this ‘Middle Range Theory’ (see Chapter 9). He had also begun to suspect that many of the so-called living floors were nothing more than fluvial jumbles (1977b).

While these ideas fomented in Binford’s mind, he turned to the patterning in the Early Palaeolithic record only occasionally. Examining the use of models in Palaeolithic research, Binford (1972) re-iterated his issues with the traditional assumption that because material culture was patterned, the pattern must be cultural. He pointed to the variable presence of Levallois in the Acheulean as a classic problem for such interpretations, because patterning in one characteristic varied independently of the other. Which, he wondered, was the true reflection of the culture–the shape of the tools, or the technology used to make them? And if it could be shown that raw material had caused the patterning, was it in any way representing the shared ideas of the group? At best it showed a reservoir of shared behaviour used according to circumstances, with any chronological or spatial patterns merely reflecting the distribution of the conditioning contingencies (ibid., 126).

The statistical approaches to the East African Acheulean pioneered by Kleindienst and Clark (see preceding) provided Binford with a large data set with which to explore variation in the East African Acheulean, once the figures had been suitably modified to meet his requirements.12 Binford observed, as both these authors had done before, that scrapers and other small tools tended to be associated with evidence for the butchering of animals, while handaxes, bifacial knives and cleavers were found at sites where the butchery and consumption of meat was not important. This, Binford hypothesised, might also explain the rarity of handaxes from areas north of the Alps, Caucausus and Himalayas, because hunter-gatherers in these northern latitudes tended to have an essentially meat-based diet.

Binford (1972) had further realised that the type of patterning he was seeing among the Nunamiut and other modern hunter-gatherers, was different from the patterning he saw in the Lower and Middle Palaeolithic record. Given the vast amount of time involved and the number of small, isolated populations that must have existed across the globe, he would expect to see hundreds of well-defined cultures if the traditional assumptions that cultural transmission, drift and replacement were correct. But there were three major traditions at most, the Oldowan, Acheulean and Mousterian. In the past, Binford argued, archaeologists had tended to assume that Lower and Middle Palaeolithic humans shared the same capacity for culture as modern humans, but this was in fact a major research question and Binford now doubted that it was true. Much modern hunter-gatherer culture was related to the conscious articulation of distinctiveness, but there was no evidence of ethnic expression in the Lower or Middle Palaeolithic, where the variation was largely functional and therefore more suited to behavioural interpretation. That Bordes’s Middle Palaeolithic typology was applicable everywhere, whereas Madame Bordes’s Upper Palaeolithic typology was applicable only in southwest France was evidence that the variation was caused by different factors, specifically that style in lithic assemblages–meaning arbitrary form imposed in an active and consciously expressed modern-human way–was absent before the emergence of modern humans and the Upper Palaeolithic (Binford 1973).

Table 8.3 The eight key tenets of the culture history model (after Binford 1972).

Culture Is …

Rationale

1) Localised in Individual Human Beings

Derives from the fact that only humans are considered to have the biological capacities for cultural cognition

2) Transmitted Among Human Beings

Through learning individuals assimilate culture from other persons

3) Shared

Sharing results from the cumulative effects of common learning experiences in short the degree of association of interaction occurring among human beings

4) Cumulative

It may be generated only by acts of invention

5) Derived From Humans

Inventions, once made, add to the alternatives among which choices may be made and serve as the basis for recombination into new inventions

6) A Continuum

It is a continuum because the succession of individuals in a generational succession is a continuum. Since culture is transmitted across generations, it also is a continuum

7) Continuously Changing

This derives from the fact that culture bearers are continuously being replaced, thereby ensuring that the cumulative effects of individual choices at any two points in time vary as the composition of individuals in the popultion varies. This also derives from the expectation that individuals are continuously being presented with new alternatives through acts of invention

8) Gradual to Change

This is expected because individuals are replaced gradually in human populations resulting in minute shifts in relative popularity among varying alternatives

Turning Bordes’s and Hawkes’s words against them, Binford (1972) argued that their notion that culture was transmitted passively from parents to offspring turned humans into uninventive automatons who would behave exactly like their ancestors unless social channels were interrupted or changed. He dismissed Collins’s cultural interpretations as fallacy–an archaeologist’s ability to synthesise their results into different taxonomic units that they called traditions was in no way a test of the idea that these differences represented degrees of ethnic affinity. Time and space were not explanations, they were simply reference dimensions. Artefact clusters or groups only showed that some systematic process was at work. It told nothing of what those processes were. More precise definitions of cultural taxonomy (e.g. Clarke 1968) did not bring us closer to those controlling factors either.

Binford now began to realise that functional explanations were not as straightforward as he had once presumed. In the Acheulean of East Africa and the Levant, Clark had identified four major cultural zones (the North African Zone, the Levantine Zone, the African open savannah zone and the African closed/forested savannah zone) defined by ecology, food resources, technology, typology and raw materials (Clark 1975). Within each zone, humans tended to leave four generalised artefact facies, characterised by qualitative and quantitative differences in the composition of the lithic assemblages (Clark 1966; Clark 1975; see Kleindienst’s Variants A–D). Clark (1966, 1975) felt these similarities could hardly be explained by cultural links, and most likely related to the fact that Lower Palaeolithic humans had simple, generalised needs that only required a simple, generalised toolkit. Variation was therefore largely functional and more-or-less independent of climate. Advocates of functional models also assumed that the objects found on Palaeolithic occupation floors directly reflected the types of activities that humans had performed there (Binford 1973): they were, after all, short-term camps that preserved evidence of short-term activities and should therefore provide the best evidence for the function of different toolkits. Such reasoning had led Clark (Clark and Haynes 1970) to conclude that carcass butchery was associated with small flake tools, not handaxes, although this did not mean that the handaxe was not an essential piece of butchery equipment at other locations (Clark 1975). Redating of the East African Acheulean to a much older period (Cole and Kleindienst 1974) did little to unseat the conviction that variation was functional and not a ‘function of time’.

Binford’s Nunamiut research quickly disabused him of these armchair assumptions (1973, 1977a, 1978b, 1979, 1980). Through his primary fieldwork he began to see the real complexity behind the archaeological record for the first time, introducing a series of organisational principles and a new vocabulary that would, in the coming decades, completely shape Lower Palaeolithic research, particularly in Anglophone countries (e.g. Gamble 1986). What had started out as an assumption–that Early Palaeolithic technology was an expendable technology (a term Binford regrettably replaced with expedient in later writing) that responded to and reflected local circumstances–was now a major behavioural variable for Palaeolithic archaeology.

In general, the artefacts the Nunamiut tundra-dwellers left at a site did not reflect activities undertaken there because their technology was largely curated. Valuable tools were retained and maintained for some time, they were ‘personal gear’ that passed through sites without being left behind. Other tools had different life-expectancies and were discarded at different rates. The degree to which an assemblage reflected the activities performed at a place therefore varied with the complexity of the ‘economic techno-logistics’ used by a group. The more tools were curated, maintained, transported and reused, the weaker the correlation between function and industry at any given activity site. The degree to which style was imposed on objects also depended on its life-expectancy. Such investment was only given to highly curated objects, whereas the form of expedient tools was again driven by short-term functional requirements. Binford (1973) was not denying that handaxes contained stylistic information, just pointing out that this had never been satisfactorily demonstrated on his terms–patterning may speak of isolation, but it did not imply that these were consciously maintained ethnic units.

The Nunamiut had three categories of tool, which in Binford’s (1979) developing organ-isational approach were termed active (in use), passive (stored) and insurance (cached). The latter were often left in potentially useful spots while people were engaged in other activities and may have been incidental to the main purpose of the site. Well-used locations often accumulated furniture or facilities, large pots, large rocks and so on. What was discarded at a site, the start of the journey into the archaeological record, was largely food waste plus lost and abandoned artefacts. Part-processed food stuffs, tools and broken tools usually returned to the village (Binford 1977a). Therefore debris and activity were only loosely connected: only ‘situational’ gear, things made on the spot to service an immediate and perhaps unforeseen need, was really correlated with function, and such tools tended to be informal, highly varied and contingent on the properties of the local raw material, though they were excellent examples of behavioural flexibility (Binford 1983b).

Personal gear, on the other hand, saw greater investment in the procurement of good raw materials, and manufacture was often staged, with different parts of the sequence occurring at different points in time and space. This did not need to involve special purpose trips–resource procurement could be either ‘direct’ (usually within a short foraging radius) or ‘embedded’ (things brought back to camp while out doing other things, working on the principle that only a fool comes home empty handed). This had implications for the presence of exotic rocks, often taken as evidence for direct procurement and subject to principles of optimal foraging theory, in as much as far travelled materials were expected to have been pre-shaped in order to reduce bulk and carry greater value. In fact, they could just reflect pieces dropping out at various points in a seasonal or annual round. The type of personal gear carried similarly depended on season, and also on sex, women having a different toolkit to men. There would thus be major differences in artefactual content and social information potential at residential versus special purpose sites.

Binford’s arctic hunters were very different from tropical hunters (Binford 1979, 1980). Whereas the Nunamiut procured 70% of their food during the two seasonal caribou migrations (each lasting about 15 days), which was then processed and stored, tropical hunter gatherers such as the San tended to procure food daily, a system of immediate rather than delayed returns (Binford 1979). This formed the basis of Binford’s two main types of hunter-gatherer settlement system, foragers and collectors.

Collectors were logistically organised. They tended to set up camp in one place and provisioned themselves (collectively) by sending out special-purpose task groups to key points in the landscape, where they would hunt and gather resources that were then returned to the camp. Daily mobility was high, but residential mobility was low, perhaps two seasonal camps and two residential moves in an annual round. They practiced anticipated hunting, going to places where herds were known to congregate or had been spotted from a hunting stand, rather than wandering the landscape looking for random opportunities. Hunters would have numerous hunting stands and caches scattered around the landscape, where they would engage in several other unrelated craft activities, often simply as boredom reducers (Binford 1978b). The nature and number of tasks at a hunting stand further depended on the number of people present, the more people, the higher the amount of time spent socialising at the expense of other things. On average only 24% of the hunter’s time was spent looking for game. Collectors existed on a spectrum of simple to complex, with mobility highly correlated with latitude, groups tended to be more residential and less mobile in more northerly, colder regions.

Foragers on the other hand were less logistically organised. Instead, they mapped onto their environment, making seasonal or more frequent moves to exploit resources as they become available, which were immediately consumed rather than stored, a form of ‘tethered nomadism’ (Binford 1980). Foragers also tended to practice a more opportunistic or encounter hunting strategy, taking anything that happened to cross their path or that they otherwise stumbled upon.

By examining the archaeological residues left behind by hunters at a hunting stand in relation to the actions that had generated them, Binford noted that when sat around a hearth, men tended to drop, toss, place, dump and brush material aside in relation to their peri-personal space, forming a roughly u-shaped accumulation of debris with two defined concentrations, the ‘drop and toss zones’. These were open down-wind and looked suspiciously like the concentrations of debris described as huts or windbreaks in Africa.

Binford was clear that these were not cultural differences, just contrasting behavioural strategies practiced by cognitively identical modern humans, but it was through these concepts that he began to detect organisational differences between the Earlier and Upper Palaeolithic. If Binford spent the 1970s being relatively quiet on the Old World front, it was simply because he was reloading (Meltzer 2011).

Notes

· 1 These placed the Late Acheulean at a mere ~58,000 years ago. The Sangoan was dated to 39–41,000 years old and the onset of the MSA at 36,000 years. This had a major impact on the length of time over which Clark believed these cultural changes to have occurred, and on his correlations with other sites and chronological frameworks.

· 2 Dart (1957) argued that the dominance of certain animal body parts on australopithecine sites was because they had been selected for their intrinsic functional properties, mandibles with rows of teeth as saws, humeri as maces. Later taphonomic studies (Brain 1983) showed that Washburn was correct to assume that carnivores had been the main agent of bone accumulation, and that the australopithecines had been the hunted not the hunters.

· 3 With apologies to Philip Larkin.

· 4 Among the students to work with Isaac at Koobi Fora were R.J. Blumenschine, H.T. Bunn, D.P. Gifford (later Gifford-Gonzalez), J.W.K. Harris, E.Kroll, K. Schick, J. Sept, C. Sussmann and N. Toth. Apologies to those not mentioned. See Debating the Oldowan, Chapter 8.

· 5 Isaac’s (1972b) essay was a prize-winning submission to the Whither Archaeology competition run by the journal Antiquity. This had sought replies on all sides to Jacquetta Hawkes (1968) infamously pessimistic and reactionary response to New Archaeology.

· 6 The archaeology at Torralba occurred in nine or ten discrete occupation floors. At Ambrona, no floors were identified. The Ambrona material was divided into Lower and Upper Occupation.

· 7 The Lower Acheulean was found in Bed II at sites EF-HR, FLK S II, HWK II, VEK II and CK/EK. The Developed Oldowan was found in Middle Bed II at MNK II, FC II, FC II West and in Upper Bed II at TK II.

· 8 Roe’s groups were initially constructed with the aid of univariate statistical analyses. The general patterns were later confirmed by multivariate analysis of the same and different data (see Graham and Roe 1970; Doran and Hodson 1975; Callow 1976, 1986a).

· 9 As a postgraduate in the 1990s I asked Clive Gamble to read a draft of one of my early papers. In a marginal comment to a statement I made about the ‘demise of culture-history’, Clive wrote: “When did it demise? I see it around me every day”

· 10 As far as I can ascertain de Mortillet (1873) did not use the terms (contra Jelinek 1977), but did divide the Palaeolithic into assemblages without bone implements and assemblages with bone implements, which would later form one of the defining features of the Upper Palaeolithic.

· 11 The italics signify the fact that Binford (1983a, 66–67) explicitly sought to distance himself from the rigid hypothetico-deductivist methods adopted by some of the younger second-generation New Archaeologists. He also deplored their tendency to create new sub-disciplines (Behavioural Archaeology, Social Archaeology, etc.), which he saw as sociological posturing (Binford 1983a, 108).

· 12 To be fair to Binford, Isaac (1977a) was forced to make similar adjustments to make his data fit with published figures.

If you find an error or have any questions, please email us at admin@erenow.org. Thank you!