Stone Age symbolic behaviours: questions and prospects

Andrea Vianello
(Graduate School of Archaeology, University of Sheffield)

Introduction

In the evolutionary perspective typical of biology, the more complex something is the more recent it is. Thus, symbolic behaviours and the complex symbolic systems they generate should be very recent because they appear late in the archaeological record and they are not very different from contemporary behaviours. Biological studies demonstrate that different species of a certain family of animals may co-exist without the need for one to be superior in absolute terms to the others. However, in the case of humans, anatomically modern Homo sapiens (AMHS or Homo sapiens sapiens), the absence of any human species other than ours in contemporary times has always been interpreted considering our ancestors less evolved, which translates in less intelligent considering the primary characteristic of AMHS. What happens then if an archaeologist unearths evidence of symbolic behaviours in the Middle Palaeolithic or evidence of complex symbolic systems among the Neandertals? Was the mind of our ancestors closer to ours than we would like to admit or do we have a problem in defining symbolic behaviour? Perhaps both, as we shall see.

In the following sections we shall review some recent developments on early symbolic behaviours and the search for the origins of them. The behaviours recognised are undoubtedly advanced and find no parallels in any known animal, suggesting that they are indeed a form of human symbolic behaviour that needs to be explored. However, the recognition of differences between the early and mature symbolic behaviours will prompt a concise review of the state of research on both biological and archaeological grounds of the advent of mature ("modern") symbolic behaviours. This review of evidence will span nearly the whole length of the Stone Age and include two parallel perspectives, the biological and archaeological ones. Although there is no attempt to provide a detailed and comprehensive summary of these studies, the resulting picture should provide some of the least controversial elements that may help in assessing any theoretical criteria on the determination of symbolicity. In the conclusions this work will try to provide a contribution on the formation and testing of theoretical criteria of symbolicity founded on the evidence as a whole. In addition, some methodological and interpretive issues will be explored as they come out throughout the sections.

Searching the first symbol: red ochre

In a recent paper (Hovers et al 2003), archaeologists have reported about the use of ochre at Qafzeh Cave. The authors suggest that human beings in the region demonstrate evidence of symbolic behaviours because complex symbolic systems can be inferred from the archaeological record as early as 92,000 years ago, in the Middle Palaeolithic. To date, this may be the earliest case of symbolic behaviours known to science. Red ochre was skilfully produced at Qafzeh Cave and probably used in connection with burials, without having any apparent practical function. There are a number of other depositions where red ochre has been detected and in every case red ochre is apparently used in non-utilitarian contexts. The hypothesis of red ochre being used as material expression of symbolic behaviours is simply tested verifying that non-utilitarian purposes were the likeliest. The criterion according to which non-utilitarian contexts are the product of symbolic behaviours may seem weak from a theoretical and methodological point of view, but is solid in practice, as we shall see. In the case of Qafzeh Cave, a few ochre-stained lithic artefacts have been recovered and suggest that the red hue had a symbolic meaning. Hafting, the use of ochre as painting substance, medicine and aid for hide tanning have all be considered (Hovers et al. 2003: 505), but they remain unproven and unlikely primarily because red ochre had been produced from cores of different mineral compositions that required different processing to obtain always the same colour. If practical reasons were paramount in the production of ochre, then variability in the colour may have been resulted as part of experiments finalised to improve both techniques and uses. We should also keep in mind that the minerals processed looked differently at the natural state and no selection by colour is apparent for the stone employed in tools-making, which are rarely stained.

The ochre is present in several layers of Qafzeh cave dating from 100,000 to 90,000 years ago (Hovers et al. 2003: 501). In later layers ochre has not been found, but there is still evidence of human presence. This suggests that ochre, for some reason unknown to us, was used only for a period, a fact that reinforces the interpretations of red ochre as symbolic. In the period of use, the excavators argue that an association between human remains, fires, and inedible molluscs can be identified. They (Hovers et al. 2003: 508) note that "the presence of human burials exclusively in the same stratigraphic block as the ochre is more telling". The recognised burials of an adult female together with an infant (Homo 9 and 10) and of an adult (Homo 11) with deer antlers, which possibly were a grave gift are used by the authors to support that the individuals occupying the cave were capable of symbolic behaviour already at that early time. Evidently, burials and grave goods, if proven, are not only proof of complex behaviours, but also proof of meditation, abstract thought and therefore cognition. Red ochre would then add to the picture, but perhaps it would not be determinant to recognise symbolic behaviours as early as the Middle Palaeolithic. In the Levant, ochre is used again in the Natufian period, dated to 12,700 - 10,500 years ago and remarkably again associated clearly with burials (Hovers et al. 2003: 510). Not only there is a long break between the two periods of use, but also two human species are involved: Neandertals about 100,000 years ago and Homo sapiens sapiens in the Natufian period.

Early symbolic behaviours

The clearest evidence of Palaeolithic symbolic behaviours in the archaeological record is provided by funerary contexts. Only a few burials of Neandertals in the Middle Palaeolithic are known. Middle Palaeolithic burials are variable and range from the simple caching of bodies to excavated graves, where grave goods were deposed. Pettitt (2002: 8-9) proposes six categories in which these early burials should be divided according to the complexity of the inhumation, but the first category is reserved for naturally occurred caching. In trying to provide some examples for each category, Pettitt has shown that a categorisation of burials based on complexity often groups burials according to their chronology. The earliest burials of his second category, the simple caching of dead, may be dated about 250,000 years BP (e.g. burials at Pontnewydd Cave dated 225,000 years BP), while the earliest burials of the sixth and last category, named "ritualised burial", date from 27,000 years BP (Pettitt 2002: 9). Qafzeh is placed on the third category, named "simple inhumation". The funerary practice and the use of colour as symbol are found together only from about 100,000 years ago, within a coherent framework of complex symbolic behaviours. Separate antecedents for both can be traced in the use of ochre and the caching of the dead, but earlier symbols do not seem to have been part of a complex symbolic system or to constitute evidence of symbolic behaviour.

At Kebara Cave (Schwarcz et al 1989; Valladas et al 1998), Israel, there is one of the earliest and most convincing cases of excavated grave and evidence of secondary activity as well. The grave has been cut through two hearths in the underlying layer and according to the excavators the posture of the body was such that the head intentionally remained exposed. At a certain stage, probably after the decay of the flesh, the head was removed by other humans. The intervention of a carnivore animal has been ruled out on the ground that the mandible, hyoid and a right upper molar were recovered from their correct anatomical positions. This deposition dates about 60,000 years ago and may be seen as the antecedent of later burials when secondary activity becomes common. At Shanidar Cave, the few depositions found represent over 10,000 years of activity as they date from 60,000 to 45,000-50,000 years ago.

Pettitt (2002: 18) concludes "that the focus of Neandertal social life was the body" and remains cautious on the significance of any burial before 27,000 years BP. Despite some evidence of symbolic behaviour, it seems that indeed the human body was central to any symbolic activity, whether it was a burial or the processing of soft tissues. Even the red ochre, because of its possible association with blood, appears linked to the body. Moreover, all these symbolic activities may simply replace natural ones. For instance, the use of red ochre may have been seen as a replacement of flowing blood when the death was not associated to blood loss, which may have been seen as the natural way to die. For instance, the symbolic meaning of red ochre in Qafzeh Cave has been interpreted as "death" (Hovers et al 2003). In the same way, defleshing may be seen as the natural occurrence for bodies, as the consumption of the flesh as meat if the natural event observed was the activity of a predator or scavenger. In that case, both proven defleshing and presumed cannibalism may be seen as symbolic re-enactments of natural occurrences. Regarding symbolisms connected to the body, it should be noted that both "life" and "death" cannot be expressed without the use of symbols. The red hue may have meant on occasion, blood, fire, flesh and meat consumed at the time of death as well many other concepts. Thus, most if not all of these concepts may have used to express the two terms, while understanding that both are part of the ineluctable natural cycle of fertility.

The positioning of the body, the deliberate removal of the head and the increasingly evident need to bury the dead in a constructed space support the possibility that complex symbolic systems existed. The use of ochre and the practice of burying the dead are not homogenous in the form both in space and time during the Palaeolithic. In particular, studies on red ochre (Knight, Power, and Watts 1995; Power and Aiello 1997; Hovers et al. 2003) demonstrate the variability of meanings. The handful of burials at Shanidar Cave spanning over 10,000 years are also evidence suggesting the absence of standardised rituals in this period, though the several symbols of later rituals can often be recognised.

Burying the dead is a deliberate action that implies extended social ties, strong enough to survive the occurrence of death. Unlike red ochre, there is no possibility that such behaviour was initiated to imitate something existing in the natural environment. In the case of red ochre, the colour has been associated to various natural elements of the same colour, such as blood. The use of colour is therefore understood as a word in a language, a symbol meaning something which can be found in nature. In the case of burials, accidental sepultures were unlikely to be discovered easily and to have provoked imitative behaviours. Instead, it is probable that burials were seen as a form of protection from scavengers, but this care implies strong social ties with the buried individuals and the thinking a relationship beyond death with those individuals, which is evidence of elaborated cognition.

Ethological observations have proved that mammals such as the elephant or the polar bear recognise defleshed members of the same species previously met and this triggers memories and behaviours that can be described as grief or mourning, but this behaviour is always strictly linked to the memory of specific individuals. The mammalian behaviour stops short of burying the dead, which would be physically possible, and therefore the action of burying the dead may be used as proof of advanced cognition, well over self-consciousness. Symbolic behaviours such as the presumed ochre-staining of the body or the proved ochre-staining of tools and later of skeletons (e.g. Paviland Cave) seem to be a by-product of advanced cognition. Symbolic behaviours stemmed from the need of expressing abstract thoughts originated by a self-conscious being relating with other individuals perceived as similar self-conscious entities. Individuals burying the dead must have enough cognitive capacity to separate themselves from the rest of the natural world, i.e. they must be self-conscious. They also must have the cognitive capacity of distinguishing the surrounding natural environment and recognising similar members. Furthermore, they must have the cognitive capacity of understanding that the other members of a social group are similarly self-conscious. This means that the world was perceived as the self, the others and the natural world. In burying the dead, a relatively simple and obvious action for contemporary humans, Palaeolithic hominins demonstrated that they understood and related to the others differently from the natural world because in burying a dead they interpreted the will of that individual and acted independently of a direct interaction or natural stimulus. The hominins burying other individuals thought the dead would fear scavengers or would like for any other reason (e.g. hiding, return to earth after life-cycle, etc.) to be buried and this happened because they imagined themselves in that situation. They then acted because in this mirroring of the self into the other they probably hoped the same care would be reserved to them and this may have helped them in overcome the fear of death. However, the tripartite division of the world in self, others and natural world is very complex and far from anything observed in any animal. Accepting a mature cognitive capacity of hominins well beyond self-consciousness, symbolic behaviours becomes therefore the language through which unphysical ideas could be expressed. Thus, self-consciousness must have appeared before graves, as early as evidence of caching the dead, about 250,000 years ago and perhaps even earlier. Complex cognitive capacity appeared probably after 250,000 years ago but certainly before 92,000 years ago, when at Qafzeh Cave there is evidence of symbolic behaviour, which originates from complex cognition.

Have AMHS individuals got their own symbolic behaviour?

Language probably offers the best insight of the complexity of human behaviour. However, direct evidence of language is very late, though human language capable of transmit abstract and real meanings seems to have existed as early as 60,000 years ago (Davidson and Noble 1993: 363). In seeking the very first evidence of language, burials, symbols and several other traits the evidence points to a past when the anatomically modern Homo sapiens (AMHS) had yet to be culturally distinguishable and spread all over the globe. This causes a split between the modern anatomy and behaviour.

It is evident considering red ochre and burials that Neandertals were capable of abstraction and complex symbolic systems. Although it is still unclear whether the Neandertals merged with or were replaced by modern humans, it is evident that the behavioural difference between the two species was not very sharp at the time of co-existence. The earliest evidence of modern anatomy is ca. 160,000 years ago (White et al. 2003), but the earliest evidence of the symbolic use of red ochre, one of the traits of behavioural modernity in a long list recently criticised (Henshilwood and Marean 2003), dates not later than 230,000 years ago and may be as old as 285,000 years ago (McBrearty and Brooks 2000, Barham 1998, Barham and Smart 1996). None of the traits (Henshilwood and Marean 2003) identified and perhaps not even a particular combination of these may be able to define modern behaviour. This happens because the environmental situation in which hominins lived was variable and therefore the traits displayed do not seem to follow any repetitive pattern. A hypothesis could be built on the basis that a pattern in the traits displayed is the fingerprint of modern behaviour, but with no pattern available the hypothesis is unsustainable. Furthermore, the context of deposition is often badly preserved, leaving archaeologists to deal with partial and unreliable information that would be difficult to use for testing purposes.

To sum up, the archaeological evidence proves that hominins have had symbolic behaviours, but these varied for long before some fixed patterns could be recognised. Thus, there are two different periods in human history in relation to symbols. In the first period, which predates AMHS, self-consciousness, the capacity of abstraction and symbolic systems as well as a few of the traits used to describe modern behaviour appeared. In the second period, patterns in symbolic behaviours can be identified and the traits increase exponentially. On one hand, there is no evidence that this change has been directly fuelled by biological changes because the path bringing humans to use symbolic systems began at the time of Homo heidelbergensis or perhaps already the late Homo erectus and progressed in different species such as Neandertals and Homo sapiens. On the other hand, it is evident that AMHS progressed further than any other hominin, resulting in their definitive evolutionary success.

As a result, archaeologists can focus on either the appearance of the first symbolic systems and behavioural traits or their late organisation in ritualistic patterns. This division seems important and evident on the archaeological record but it cannot be justified on material culture alone because it would be just a change in culture. In Europe, perhaps the best studied case, the arrival of Homo sapiens with a more sophisticated culture in which the sophisticated art work was included has prompted Henshilwood and Marean (2003: 635-636) to argue that art work may be the decisive trait in distinguishing modern behaviour, of which modern symbolic behaviour is a subset. However, they point out that art work needs to be considered all over the globe and when this is done, there is no evidence supporting a sudden arrival of the behaviour associated with a specific species outside Europe. They also suggest that rapid changes in the production of distinctive and consistent artifact styles may provide the marker for behavioural modernity. This brings us to the starting point where changes in material culture would represent changes in behaviour, though in Henshilwood and Marean's hypothesis only specific types of cultural changes could be considered for this. Although this limitative definition helps in considering human behaviour overall, when considering exclusively symbolic behaviours we need to find a single change capable of accounting for the emergence of modern symbolic behaviour in an archaeological record punctuated with small changes.

New prospects are offered by the human genome, which may be used to find genetic changes dated about 40,000-30,000 years ago according to Henshilwood and Marean (2003), or in the incertitude of the date one specific gene that may be responsible for it should be searched (Klein in Henshilwood and Marean 2003). Although not yet feasible, this approach seems already doomed by methodological problems. In the former hypothesis the link between genetic change and behavioural change would be forcefully synchronised to a date obtained from the archaeological record. In the latter hypothesis a single change at chromosomal level would have had enormous effects abruptly.

Even considering this approach from a purely biological point of view, its weaknesses remain. Changes in brain capacities and different human species cannot be detected in the archaeological record very easily because changes in material culture are frequent and small whereas changes in anatomy from an evolutionary perspective are far in between. To this regard, the biological theory of punctuated equilibrium proposed by Gould and Eldredge (1977) is born from the observation that rapid periods of modification are distanced by long periods of slow progress. This does not happen regularly in the archaeological record.

The modern human anatomy first appeared at least 160,000 years ago, but until 40,000 to 30,000 years ago there is little evidence of any difference in behaviour among the co-existing human species, and the size of brain was very similar across species, particularly AMHS and Neandertals, with the latter having the largest brains. This means that it took at least 120,000 years for a change in the general anatomy to become apparent in the archaeological record. Thus, finding a chromosome that may have changed 40,000-30,000 years ago does not ensure that the change it triggered has manifested itself immediately. Moreover, there has not been a single important change 160,000 years or so ago that simply manifested itself later, but a series of rapid changes which then continue more slowly. In 120,000 years, the time from the anatomical change to the appearance of behavioural modernity, many genetic changes happened. The biological evolutionary system seems to work like a cycle. Repeated environmental changes, new needs or any other stable condition which would benefit from anatomical changes stimulate modifications that may involve several alterations of the genetic code at once, so that eventually a new species is originated. This is the period of rapid change, which is followed by one of slow stabilisation and progressive enhancement of the anatomy. This hypothesis would account for both rapid biological transformations and progressive cultural variations. During the rapid changes period the anatomy would adapt to suit the new situation that prompted the change in first place, easing the burden of everyday life without causing major behavioural changes. Instead, the slow changes period would exploit the new possibilities offered by the major anatomical transformations effectively causing shifts in culture, style and behaviour, slowly and progressively.

Modern symbolic behaviour may be defined as the symbolic behaviour exclusive of Homo sapiens sapiens from about 40,000 years ago and unparalleled for complexity in any other human species. Accepting the hypothesis of evolutionary cycles, modern symbolic behaviour would be the product of slow refinements within a framework of continuous advancement, but it makes sense to separate it from a biological point of view because it has been very stable, it has provoked no further speciation and has reached a level of complexity unparalleled among other human species. From an archaeological perspective it makes sense as well because rituality appears coherently for the first time, while art and symbolic systems become increasingly important. As we have seen, both biological and archaeological considerations support the separation of the modern symbolic behaviour from previous behaviours, but there is no reason to claim a clear superiority of this behaviour from the previous ones. From the first symbolic behaviour detected about 250,000-230,000 years ago to the earliest modern symbolic behaviour detected about 40,000-30,000 years ago there are 200,000 years in between of steady cultural and behavioural progress. For instance, the Neandertals were probably behaviourally very similar to the contemporary AMHS, and we may be fooled by the European culture of the time that was not state-of-the-art until new people arrived in the correctness of summing biological and cultural differences. On one hand, it is likely that the Neandertals coped well for a while with new cultural changes, perhaps as well as Barbarians coming into contact of the Roman culture. On the other hand, the AMHS individuals bringing new cultural elements had just developed them from a culture similar to that still practised by the Neandertals, so that there was just a small progress in between the two cultures. The large cultural gap we see is partly due to the 30,000 or more years that separate ourselves from both Neandertal and Palaeolithic AMHS individuals and partly to the exceptional cultural and behavioural developments that were caused by the combination of increased sociality and the exploitation of new anatomical capacities. We shall now explore in more detail the cultural and biological developments that made possible modern behaviour, and particularly modern symbolic behaviour.

Mature symbolic behaviours

During the period from 40,000 to 30,000 years ago modern human behaviours appear. A few examples will help in identifying some of the commonest modern symbolic behaviours. Particular attention will be given to burials in order to provide comparative material with the earlier sepultures we have already examined.

Three small animal figurines, made of mammoth ivory, have been recently found at the cave site of Hohle Fels, in southwestern Germany (Conard 2003, Sinclair 2003). One is shaped like a bird, another like the head of a horse and the third may be half animal and half human. Twenty other ivory figurines had been reported from the region previously. Charcoal in animals painted in the Salle du Fond and the "horse" panel of the Hillaire chamber in the Chauvet cave, France, have been dated between 32,400 and 29,700 years ago (Valladas et al 2001). At Dolni Vestonice one of the earliest "Venus figurines" known is dated 26,000 years ago. Venus figurines, as they are called, become increasingly frequent and peak at about 15,000 to 10,000 years ago. They represent women, often pregnant, with exaggerated sexual attributes. To the same period dates a male burial found at Paviland Cave, Wales, which is known in the archaeological literature as "Red Lady of Paviland" (Aldhouse-Green 2000). The skeleton as well as the grave goods had been stained with red ochre. The grave goods consist of an ivory rod, bracelet fragments and perforated periwinkle shells. Furthermore, it is possible that two blocks of limestone marked the grave itself. The cave was used by humans for long, perhaps sporadically, and great quantities of Palaeolithic tools spanning from 35,000 to 11,000 years ago have been recovered since the discovery of the burial, back in 1823. However, the early date of discovery means that the context was inadequately recorded, so that it remains unclear what the grave assemblage comprised.

Since most of the Upper Palaeolithic complex symbolic systems are attached to funerary activities, it seems useful to consider the burial practice in the Mesolithic Levantine Natufian culture, which is the earliest deliberate and systematic funerary practice known. The Early Natufian period is significantly more recent than any other period considered previously, dating from 12,500 to 11,000 years ago. In addition, the population was largely sedentary or at least moving within a restricted territory. This suggests that social ties played an enormous role at this moment of history. Because the Natufian populations settled within small territories, the range of social connections increased exponentially. The original extended groups, typical of hunters-gatherers societies, are still recognisable in death because the Early Natufian sepultures were primarily group graves.

The funerary practice seems initially motivated by the attempt to preserve and show the unity of the original groups, a need that implies a large society composed by several groups. The stable settling in a territory produced a repetition of encounters between the same individuals, creating a social network of interconnections. This network produced a population in which each member was aware of the others and the territory was the unifying element rather than blood or the appurtenance to a group. The burials themselves ended up in reinforcing the link between humans and territory, because the space in which the people of Natufian culture was not only shared with other living humans, but also with dead humans that not necessarily were connected to each member. As a result an anthropocentric dimension of space replaced the natural dimension in the minds of humans. The people of Natufian culture lived in a human-centric space, where nature could be controlled to some extent. The creation of a space by and for humans is not new however. Early burials, both simple caching of corpses or deliberate sepultures, are found in contexts extensively used by humans for very long times. It is possible therefore that the sepultures were an attempt to mark the territory as "human", to separate it from the natural world. This hypothesis would also explain what linked sepultures thousands of years apart: the location. Caves, providing shelter to the primordial communities, were evidently the most obvious locations for social gathering and probably the cradle of the concept of anthropocentric space.

Grave goods, such as dentalium shells or bone beads and pendants or fox teeth, occur relatively frequently among Early Natufian burials (Belfer-Cohen 1995). The Early Natufian skull cache at Erq el Ahmar (Neuville 1951), dated about 11,000 years ago, is the only evidence of skull removal, which is common in the following Late Natufian period. Kuijt (1996: 329) has produced an interesting table that summarises some variables related to the burials across four consecutive periods in the Natufian area, evidencing how each variable is dependent upon the set of beliefs current at each time and location. Weinstein-Evron and Ilani (1993) have reported the use of ochre at the site of el-Wad Cave in Early Natufian layers dated 13,000 to 11,000 years ago. About eighty earths of red, orange, yellow and brown local minerals containing ferric oxide, silica or alumina as well as prepared ochre had been recovered. Heating experiments on similar minerals (Weinstein-Evron and Ilani 1993: 466) demonstrated that in all cases the end-product would have been red ochre.

Grave goods, skull removal and use of red ochre are just some of the variables that were part of the funerary rites practised by the people of Early Natufian culture. Remarkably, the grave at Paviland Cave, 26,000 years ago or the typical burial at el-Wad Cave 13,000 years ago appear extremely similar in the presence of shells and bone or ivory items as grave goods and use of red ochre. This similarity is even more remarkable because Late Natufian burials are evidently different from Early Natufian burials. The broad view that allows to recognise the similarity in the symbolic systems used at Paviland Cave and el-Wad Cave, which are far apart both geographically and chronologically, should be expanded to include the whole Stone Age and Africa at least. We would then notice that 96,000 years ago at Qafzeh Cave red ochre was skilfully produced and human remains were possibly associated to inedible molluscs (shells) and animal bones. In Middle Stone Age Africa red ochre was also used (McBrearty and Brooks 2000). We may also notice how the practice of skull removal appears and disappears throughout long periods and vast distances. The simple fact that each variable appears and disappears may be a meaningful pattern as well. For example, red ochre is used at Qafzeh Cave 96,000 years ago when it had great importance, particularly the specific red colour, but then it disappeared in later layers (Hovers et al 2003: 494) and at Shanidar, on sedimentary grounds, seven adult Neandertals "were deposited on separate occasions over at least 15 ka, i.e. from c 45-50 ka BP (Shanidar 1 and 5), to perhaps considerably before 60 ka BP (Shanidar 4, 6-9)" (Pettitt 2002: 8).

What is, if any, the difference between Palaeolithic and Natufian burials? In the Middle Palaeolithic, several symbolic systems were occasionally manifested and scarcely integrated, but there is no evidence of a structured ritual. It is very likely that different groups knew one or more of these symbols and handed them down generation after generation, but they were apparently unable to combine them and hand down the particular combination as meaningful per se. This happened in spite of likely intercultural contacts between different human species, suggesting that the problem was not the exchange, or perhaps sharing, of complex symbolic systems among groups, but the capacity of expressing beliefs was either missing (non-AMHS humans) or still immature (AMHS). There was the capacity to express single meanings through symbols though.

From about 40,000 years ago the frequency of contacts among groups increased, perhaps because severe environmental conditions may have reduced the range of mobility. The culture and behaviours of neighbouring groups probably merged. There is no evidence of standardised burials at first, which happens only very late, with the Natufian culture. However, fine arts appear in their stunning beauty and art may suggest that a biological change happened to make this possible. With the appearance of art and then systematic burials, there is one element that is as new as overlooked: repetition. For example, in cave art certain subjects are repeated several times, in different places and at different times. This repetition causes standardisation. Because several motifs-symbols are "fixed", they are together and new ones seem to be born as variant of those known. The capacity of creating new symbols from other symbols appearing for the first time. The process occurring may be described as "ritualisation". Any criteria we may use to recognise modern symbolicity needs therefore to look for rituality. This is a key discriminant between archaic basic symbols and mature complex symbolic systems.

Evolution and brain: the difficult relationship

The burials of non-modern hominins already show that many symbolic activities were practised, indeed all the main symbolic activities in the Early Neolithic can be traced back to the Palaeolithic, when these were apparently occasional. Complexity increases progressively (Pettitt 2002) in the archaeological record because of the number and combination of symbolic activities practised each time, not because single symbolic activities appear more sophisticated.

In short, it seems that non-modern burials tried to recreate a natural scene adding some elements that were felt as missing by the mourners. For example, the interment and use of red-ochre may have been two symbols that recreated what were considered ordinary natural occurrences, such as the natural covering and the loss of blood. The physicality of the natural world was paramount, with the body being at the centre of the scene and all the elements directly pointing to natural occurrences in death. Palaeolithic minds materialised their abstract thoughts into the physical world. People of modern symbolic behaviour instead do not seem to have perceived the world directly through their biological senses. They mediated the physical perception through the I/self-consciousness and expressed their understanding of reality through symbols. Modern symbolic behaviour in practice is the use of symbols as words in a language; full expressivity was reached through coherently structured symbols because the perceived reality was an abstract belief.

The archaeological record suggests therefore that a significant change happened with the advent of modern symbolic behaviour and this was driven by increased social contacts and produced complexity only in the structuring of symbols, not their construction. The increased sociality may have acted as a constant environmental factor that produced changes at evolutionary level, but given the origin any change must have interested especially the brain. However, the brain itself never survived, so cranial capacity calculated from the skull is the only data available.

Remarkably, Neandertals reached a superior cranial capacity than the contemporary AMHS but AMHS only managed to achieve a full modern symbolic behaviour. Although minimal variations in cranial capacity do not affect intelligence, a reversion of any consolidated evolutionary trend may be explained only by a change in the relationship of the factors that in the first place started the trend. When the australopithecines employed primarily bipedal locomotion, the arms were increasingly free of their original burden of supporting the body and the secondary function, handling, became primary. Handling stimulated the brain and because early hominids were physically too small to support a brain capable of more intelligence, growth started naturally. The process peaked between 600,000 to 100,000 years ago. On this regard, it is interesting to notice how apparently the earliest Neandertals developed after the earliest AMHS, suggesting that the evolutionary trend of cerebral growth (table 1) continued in a progressive way, but for some reason the most "capable" brain in size was unsuccessful in the struggle for survival.

Table 1: comparison of cranial capacities

The brain of AMHS decreased in size during Mesolithic in Europe and Middle Stone Age in Africa (Henneberg 1998) and then continued the trend of growth. Intelligence does not depend only on the volume of brain, which does not seem to have been an important limiting factor since the late Homo erectus. The increase in size of brain, called encephalisation, targeted especially the cerebrum, the area of the brain dedicated to sophisticated functions. Thus, encephalisation made possible complex symbolic systems, but it is important to notice that encephalisation reached a sufficient volume to support complex symbolic systems about 500,000 years ago with Homo heidelbergensis but the earliest complex symbolic systems detected date about 230,000 years ago. >From that moment AMHS appears 160,000 years ago in Africa and Neandertals appear in Europe and Levant about 130,000 years ago, two species originated perhaps by the same environmental factor, i.e. increased social relationships in different regions. The Neandertals seem however to have continued the path of expanding the size of the brain, perhaps hitting a physical limit in doing so. On the contrary, AMHS keeps expanding its cerebral capabilities as well, but the volume of the brain shrinks during this process exactly at the time the symbolic behaviour becomes apparent. What happened then to AMHS?

In short, it may be that the cerebrum may have fully developed. The cerebrum is the corrugated superficial area of the brain that hosts the most complex cerebral functions and that today forms the largest part of the brain. The corrugated structure of the cerebrum means that it is a naturally compact tissue, which would have allowed the expansion of the mass while density increased. Following this hypothesis, the expansion of the mass should have been compensated for a while by the increase in density. This process kept the brain size stable for long, but eventually the cranial capacity may have been reduced temporarily only to continue a slow expansion later. After all the evolutionary trend that expanded the brain may have not stopped: challenged by the dimensions necessary to support a "plain" brain capable of the next level of intelligence, nature may have just adopted a more functional form, expanding the cerebral faculties and reducing the physical space necessary. Remarkably, the brain is today a dynamic organ capable of small expansions and contractions within as little as three months because of its management of density (Draganski et al 2004). This capability, named neuroplasticity, may be the true innovation of AMHS and the one responsible, after 120,000-150,000 years from the biological change, of symbolic behaviour.

To conclude, two possible hypotheses linking biological and archaeological data have been presented. Both require much more than a single chromosomal change. Moreover, to manifest their effect on the archaeological record and behaviour the biological transformations needed about 250,000 years for complex symbolic systems to appear and about 150,000 years for modern human behaviour. Both biological changes were extreme enough to cause speciation and the effect on the archaeological record is sudden and extreme as well. The distance in time between biological events and archaeological effects is such that more research and many debates will certainly continue, but at least we have seen that it is possible to link the two set of data.

Conclusions

This short and incomplete discussion of Stone Age archaeological evidence finds its reason to be in the search of criteria capable of correctly recognise symbolisms at the dawn of humanity. We have seen that the key problems are in fact intertwined and not entirely within the dominion of archaeology. Biology is as important as archaeology, but both data sets are still fragmentary. On one side, the archaeological record points clearly towards an early introduction of complex symbolic systems, but these were not clearly associated to AMHS and modern symbolic behaviour. The difficulty may have been surpassed if archaeological data would not become apparently contradictory suggesting that symbolic behaviour, as we know it, appeared not earlier than 40,000 years and was associated specifically to AMHS. As a result, criteria of symbolicity have been built from various perspectives, such as biological (e.g. relationship AMHS - symbolism), archaeological (e.g. non-utilitarian tools are symbolic), ethno-anthropological and philosophical. Integration of datasets has always been the major weakness in this field of research.

The methodological issues of identifying symbols cannot be separated from the origins of symbols as instead Nowell (symposium papers: 12) suggests referring particularly to the theoretical debates. If we cannot agree on what complex symbolic systems and symbolic behaviour are, then we cannot recognise them securely. In trying to recognise some phases into the development of modern symbolic behaviour, we may be able to recognise what symbolic behaviour is and what symbols are and begin to formulate criteria capable to recognise fully modern symbolic behaviour as well as any meaningful degrees of it. The archaeological evidence, though fragmentary, demonstrates that throughout the Palaeolithic a few specific symbols were part of the life of every individual, or it would be impossible to have their repetition enormously distanced in both geographical and spatial terms. Bouissac (symposium papers: 19) points out that replication is indeed "the surest indicator" of symbolic value.

It would be pretentious to affirm that this brief review of evidence has contributed to our understanding of biological and genetic processes in the development of intelligence, cognition and symboling. However, it has put forward the need of integrating various types of evidence as ordinary methodological procedure. As we have seen, one piece of evidence may explain another, easily crossing the boundaries of different fields. In this work the archaeological perspective has been used as primary source of evidence, at least because it is better known to the writer, but biological evidence has been used to form a balanced picture of the past research and to explain the archaeological record.

It is likely that consciousness, the ability of abstract thought and symbolism have originated early in human evolution and have been refined ever since. Ultimately, if a change in human anatomy freed two limbs and stimulated new advanced neurological functions that lead to cognition, than it is possible that the creation of a social environment in which cognition was stimulated produced a change in the brain that allowed members of one species to proceed further on the evolutionary process and supersede any other species.

Questions about the relationship of Neandertals to our ancestors and the origins of consciousness have received a timid answer here, in spite of not having been asked or especially targeted. In brief, Neandertals were probably biologically different from our ancestors mainly in the structure of the brain. There is no physical proof of this as only cranial capacity is available, but this finding seems to fit with the overall archaic features of early Neandertals. Neandertals were adapted to the European environment and appear to have very much and it is unclear as yet what archaic feature may have been determinant in their demise. Yet, their culture and behaviour do not seem to have been archaic in anything up to the arrival of AMHS in Europe and their almost immediate development of modern behaviour. For what concerns consciousness, this feature should have appeared before any symbolic system as it may have been the direct cause for their emergence. These statements only call for further research and they are presented here to provide a perspective as broad as possible of both questions and prospects in this field of research.

We will never be sure of which thoughts moved the individuals at Qafzeh Cave to use specifically red ochre, but we may gain an understanding of why they used red ochre (e.g. it was an item connected to abstraction) and what, in general terms, they wanted to do with it (e.g. they wanted to express certain meanings in a symbolic way). As soon modern symbolic behaviour appears the complex symbolic systems change dramatically. We may find difficult to argue about the similarities of the burial at Paviland Cave and the Early Natufian burials, which are not and cannot be product of the same culture, but they both originated from a common cultural background unimaginably long (not less than 100,000 years). The symbols used may been developed and employed for extremely long periods, but it took as little as the transition from Early to Late Natufian to see previous symbols disappear and new ones appear originating both from the Palaeolithic culture (e.g. skull removal) or being formulated within that culture (e.g. reburial of multiple crania as secondary activity and symbolic absence of grave goods).

Art is perhaps the first evidence of a cultural dynamism unknown previously, but such dynamism can be explained in the same way for both art and funerary contexts. The increase in social relationships meant that symbols (and culture) could be passed to other individuals more easily as well as new variants could be produced and spread in shorter times. The relationship between social dynamics and symbolic meanings is evident in the Natufian period, when the merging of different groups has explosive effects on symbolic expression and can be guessed and perhaps proven in the case of European Palaeolithic art, which spread relatively quickly with constant motifs. The final result is that symbolic expressivity mutated from material representation of abstract ideas in a physical world to a proper language, understood by most and dynamic, where symbolic materials are the words describing the metaphysical (i.e. beyond the physical; abstract) world of the I/self-consciousness. Continuing the metaphor of language, the structuring of words into sentences equals in our case to the structuring of basic material symbols into rites.

Thus, the recognition of basic words-symbols and their minimal structuring give us an introspective into the Palaeolithic world, where simple criteria such as the non-utilitarian one inform us of people capable of metaphysical thinking. The recognition of "replication" or ritual activity is determinant for modern symbolic behaviour to be recognised. Indeed, the term "ritual" may substitute the overused and unclear term "modern". Humans showing ritual (modern) symbolic behaviour may have been psychologically more complex than other contemporary individuals not showing this behaviour, no matter their species. However, we should not forget that they were all humans capable of thinking beyond the physical world and the immediate.

In conclusion, not a single criterion used to identify symbolicity should be abandoned, no matter how weak its theoretical basis is. The five criteria suggested by Bouissac (symposium papers), the non-utilitarian criterion and any other are all appreciated. However, criteria of symbolicity need to be used as a set and never individually. Rituality is the watershed between archaic and modern symbolic behaviour, but rituality evidently did not appear overnight (especially if genetic changes had a role) and therefore the overall complexity must be assessed taking in consideration various criteria. A methodology capable of consider the whole range of evidence is important and therefore archaeology, cognitive sciences and biology should be used together. Also, very long periods and large areas should be considered unitarily because the recognition of symbols throughout Stone Age offers the possibility of finding repetitive patterns back and forth. Thus, not generic repetition but rituality should be placed as watershed criterion between archaic symbolic behaviours involving a few "complex symbolic systems" on their own and modern symbolic behaviours using symbols as words and rites as structured sentences. In this view, rituality may be defined as a form of language by means of symbols, i.e. a powerful communication tool capable to express abstract ideas in the physicality of the material world. In this way, because the materiality can be perceived through the senses by any individual in a very similar manner, rituality as a language can always be understood to some degree, capable as it is of circumventing any cultural, geographical and chronological barrier which limits verbal communication.

Andrea Vianello
University of Sheffield
Department of Archaeology and Prehistory
West Street, Sheffield
S1 4ET
United Kingdom

Bibliography

Aldhouse-Green, S. 2000. Paviland Cave and the 'Red Lady': a definitive report, Western Academic & Specialist Press, Bristol.

Barham, L. S. and Smart, P. 1996. 'Early date for the Middle Stone Age of central Zambia', Journal of Human Evolution, no. 30, pp. 287-290.

Barham, L. S. 1998. 'Possible early pigment use in South-Central Africa', Current Anthropology, no. 39, pp. 703-710.

Belfer-Cohen, A. 1995. 'Rethinking Social Stratification in the Natufian Culture: The Evidence from Burials', Oxbow Monograph, pp. 9-16.

Conard, N. J. 2003. 'Palaeolithic ivory sculptures from southwestern Germany and the origins of figurative art', Nature, no. 426, pp. 830-832.

Davidson, I. and Noble, W. 1993. 'Tools and language in human evolution', in Tools, language, and cognition in human evolution, eds. Gibson, K. and Ingold, T., Cambridge University Press, Cambridge, pp. 363-388.

Draganski, B., Gaser, C., Busch, V., Schuierer, G., Bogdahn, U. and May, A. 2004. 'Changes in grey matter induced by training', Nature, no. 427, pp. 311-312.

Gould, S. J. and Eldredge, N. 1977. 'Punctuated equilibria: the tempo and mode of evolution reconsidered', Paleobiology, no. 3, pp. 115-151.

Henneberg, M. 1998. 'Evolution of the human brain: Is bigger better?' Clinical and Experimental Pharmacology and Physiology, vol. 25, no. 9, pp. 745-749.

Henshilwood, C. S. and Marean, C. W. 2003. 'The Origin of Modern Human Behavior: Critique of the Models and Their Test Implications', Current Anthropology, vol. 44, no. 5, pp. 627-652.

Hovers, E., Ilani, S., Bar-Yosef, O. and Vandermeersch, B. 2003. 'An Early Case of Color Symbolism: Ochre Use by Modern Humans in Qafzeh Cave', Current Anthropology, vol. 44, no. 4, pp. 491-522.

Knight, C., Power, C. and Watts, I. 1995. 'The human symbolic revolution: A Darwinian account', Cambridge Archaeological Journal, no. 5, pp. 75-114.

Kuijt, I. 1996. 'Negotiating Equality through Ritual: A Consideration of Late Natufian and Prepottery Neolithic A Period Mortuary Practices', Journal of Anthropological Archaeology, vol. 15, no. 4, pp. 313-336.

McBrearty, S. and Brooks, A. S. 2000. 'The revolution that wasn't: a new interpretation of the origin of modern human behavior', Journal of Human Evolution, vol. 39, no. 5, pp. 453-563.

Neuville, R. 1951. Le Paléolithique et le mésolithique du désert de Judée, Paris.

Pettitt, P. 2002. 'The Neanderthal dead: exploring mortuary variability in Middle Palaeolithic Eurasia', Before Farming, no. 1 (4), pp. 1-19.

Power, C. and Aiello, L. C. 1997. 'Female proto-symbolic strategies', in Women in human evolution, ed. Hager, L. D., Routledge, London and New York, pp. 153-171.

Schwarcz, H., Buhay, W. M., Grün, R., Valladas, H., Tchernov, E., Bar-Yosef, O. and Vandermeersch, B. 1989. 'ESR dating of the Neanderthal site, Kebara Cave, Israel', Journal of Archaeological Science, no. 16, pp. 653-661.

Sinclair, A. 2003. 'Art of the ancients', Nature, no. 426, pp. 774-775.

Valladas, H., Mercier, N., Joron, J.-L. and Reyss, J.-L. 1998. 'GIF laboratory dates for Middle Palaeolithic Levant', in Neanderthals and Modern Humans in Western Asia, eds. Akazawa, T., Aoki, K. and Bar-Yosef, O., Plenum, New York, pp. 69-75.

Valladas, H., Clottes, J., Geneste, J. M., Garcia, M. A., Arnold, M., Cachier, H. and Tisnerat-Laborde, N. 2001. 'Palaeolithic paintings: Evolution of prehistoric cave art', Nature, no. 6855, p. 479.

Weinstein-Evron, M. 1993. 'The Natufian Use of el-Wad Cave, Mount Carmel, Israel', in The human use of caves, eds. Bonsall, C. and Tolan-Smith, C., Archaeopress, Newcastle upon Tyne, pp. 155-166.

White, T. D., Clark, J. D., Beyene, Y., Woldegabriel, G., Hart, W. K., Renne, P. R., Gilbert, H., Defleur, A., Suwa, G., Katoh, S., Ludwig, K. R., Boisserie, J.-R. & Asfaw, B. 2003. 'Stratigraphic, chronological, and behavioural contexts of a Pleistocene Homo sapiens from Middle Awash, Ethiopia', Nature, no. 423, p. 742-747.