Beginning with Alexander Marshack's interpretation of engraved lines as lunar calendrical notations, a number of highly controversial claims have been made concerning the possible astronomical significance of Upper Palaeolithic images. These claims range from lunar notations, to solstice observances in caves, to constellation representations. Given the rare nature of artefacts and images that lend themselves to such interpretations, these claims are generally difficult to evaluate on the basis of archaeological data alone. However, comparative ethnology can provide at least a way of assessing the plausibility of such astronomical claims. If the premise is accepted that at least some of the Upper Palaeolithic groups were complex hunter-gatherers, then astronomical observances, or the lack of them, among ethnographic complex hunter-gatherers can help indicate whether astronomical observations were likely to have taken place among Upper Palaeolithic complex hunter-gatherers. A survey of the literature shows that detailed solstice observances were common among complex hunter-gatherers, often associated with the keeping of calendars and the scheduling of major ceremonies. Moreover, aggrandizers in complex hunter-gatherer societies often form ‘secret societies’ in which esoteric astronomical knowledge is developed. The existence of calendrical notations and secluded meeting places for secret-society members are suggested to be at least plausible interpretations for a number of Upper Palaeolithic caves and images.

Figurines—miniature human representations modelled in clay or stone — are one of those key categories of prehistoric material which no archaeologist who finds one can ignore. Whether working in central America or southeast Europe, or in any of the many other contexts in which figurines abound, they form a central class of material which generates a heightened level of interest and attention. But however numerous and how-ever intriguing, prehistoric figurines have another crucial quality — that of ambiguity. Without the help of textual evidence, can prehistoric figurines be confidently interpreted or understood? Can we ever hope to know what an individual figurine was meant to represent, or why it was modelled in the way it was? Yet the challenge of interpretation can hardly be refiised. For figurines illustrate self-awareness, which is a unique human characteristic. It is this dilemma — the impulse to interpret, but the difficulty of doing so convincingly — which is the focus of the present Viewpoint.

Figurines are found in many (though not all) regions and periods of prehistory. The earliest — the female forms once referred to as ‘Venus figurines’ — date back to the Upper Palaeolithic. At the other end of the scale, figurines are still in active production today, in the form of dolls, models and statues. In a prehistoric context, figurines have multiple dimensions of interest and meaning. In first place, there is the issue of sex and gender. Many figurines are clearly female, yet their gender significance, in both social and cognitive terms (rather than in simplistic notions of Mother Goddess or sex object), has only recently begun to be considered in a serious and critical way. Then there is the aspect of human self-awareness which the figurines so vibrantly express. Figurines also encode important cognitive elements in the modelling and representation of the human form, their makers frequently exaggerating some features or concealing others. Nor, ultimately, can we avoid the question of belief, and the ritual context in which so many figurines were made or used.

The contributors to this Viewpoint feature all believe that figurines can indeed be interpreted. But they also lay stress on the vital importance of context and definition. Prehistoric figurines cannot be understood as isolated artefacts, but must be seen as products of particular societies. How far we can penetrate into their meanings — and into the minds of their prehistoric makers — is the fundamental question which underpins this discussion. Can we interpret figurines? And if so, how should we go about it?

Judith Butler's proposal that embodiment is a process of repeated citation of precedents leads us to consider the experiential effects of Mesoamerican practices of ornamenting space with images of the human body. At Late Classic Maya Copáan, life-size human sculptures were attached to residences, intimate settings in which body knowledge was produced and body practices institutionalized. Moving through the space of these house compounds, persons would have been insistently presented with measures of their bodily decorum. These insights are used to consider the possible effects on people of movement around Formative period Olmec human sculptures, which are not routinely recovered in such well-defined contexts as those of the much later Maya sites.

Recent archaeological research now views the northwest European Neolithic and Early Bronze Age as a period of separation from a resilient complex of traditions of Mesolithic and even Palaeolithic origin. Extending this insight to recent findings in archaeoastronomy, this article treats the sarsen monument at Stonehenge as one among a number of monuments with lunar–solar alignments which privileged night over day, winter over summer, dark moon over full. The aim of the monument builders was to juxtapose, replicate and reverse certain key horizon properties of the sun and the moon, apparently with the intention of investing the sun with the moon's former religious significance. This model is consistent with both current archaeological interpretations of burial practices associated with the monument, and with recent anthropological modelling of hunter-gatherer cultural origins.

Archaeologists have been naive to expect that symbolic behaviour can be inferred more or less directly from the characteristics of an archaeological record. This article examines various construals of symbolism and its relationship to signs, indexes and style that are derived from more fully axiomatized theories of cognition in psychology, anthropology, linguistics and primatology. The authors' intent is to develop a more adequate operational definition of symbolism, and to determine whether our more stringent criteria for inferring symbolic behaviour can be met in ‘pre-modern’ contexts. A review of the archaeopalaeontological record suggests that the roots of symbolism can be traced to the Lower and Middle Palaeolithic, but that evidence for a robust pattern documenting its presence is not found until the Upper Palaeolithic. It is argued that symbolism had adaptive significance, but that it probably developed in contexts distinct from those in which it is manifest today.

To build a theoretical and empirical foundation for interpretation of the absence, segregation or simplicity of infant burials in archaeological contexts, we review social theories of emotion, inter-disciplinary views on the relationship between mortality rates and emotional investment, and archaeological interpretations of infant burial patterns. The results indicate a lack of explicit theory in most archaeological accounts and a general lack of consideration for individual variation and the process of change in mortuary practice. We outline the tenets of Bowlby's attachment theory and Stroebe and Schut's dual process model of bereavement to account theoretically for pattern, variation and change in modes of infant burial. We illustrate the value of this psychology-based perspective in an analysis of Victorian gravestone commemorations of infant burials in 35 villages in rural south Cambridgeshire, England, where individual and class-based variation, relative to falling mortality rates, is best explained as a function of coping strategies and contextually based social constraint on the overt representation of grief and loss.

The manufacture of stone tools is an integral part of the human evolutionary trajectory. However, very little research is directed towards the social and cognitive context of the process of manufacture. This article aims to redress this balance by using insights from contemporary neuroscience. Addressing successively more inclusive levels of analysis, we will argue that the relevant unit of analysis when examining the interface between archaeology and neuroscience is not the individual neuron, nor even necessarily the individual brain, but instead the socio-cognitive context in which brains develop and tools are manufactured and used. This context is inextricably linked to the development of unique ontogenetic scheduling, as evidenced by the fossil record of evolving hominin lineages.

Evidence for a hunter-gatherer range-expansion is indicated by the site of Station One in the northern Sudan, a surface scatter of chipped stone debris systematically collected almost 40 years ago, though not studied until present. Based on technological and typological correlates in East Africa, the predominant use of quartz pebbles for raw material, and the production of small bifacial tools, the site can be classified as Middle Stone Age. While often appearing in East African assemblages, quartz was rarely used in Nubia, where ferrocrete sandstone and Nile pebble were predominantly used by all other Middle Palaeolithic/Middle Stone Age populations. Additionally, façonnage reduction is characteristic of lithic technology in East Africa in the late Middle Stone Age, while Middle Palaeolithic industries in the Nile Valley display only core reduction. It is proposed this assemblage represents a range-expansion of Middle Stone Age hunter-gatherers from East Africa during an Upper Pleistocene pluvial.

Generation after generation, life after life, human existence flows through time. Person to person, community to community, the relationships of social existence spread out in space. Archaeology has come up with many different approaches to the central questions of temporality and sociality, but it has not been very successful with either.

At issue in this article is the soundness of archaeological inferences which proceed stepwise from data about the material culture of Middle Stone Age humans, via assumptions about their symbolic behaviour, to the conclusion that they had modern language. Taking as paradigmatic the inference that the humans who inhabited Blombos Cave in South Africa some 75,000 years ago had fully syntactical language, the article argues that the inferential step from symbolic behaviour to modern language lacks the required warrant. This step, it is shown, is not underpinned by an adequate bridge theory of the putative links between symbolic behaviour and modern language. The bridge theories invoked to date to shore up the Blombos inference are flawed, for instance, in that they incorporate untenable assumptions about language, including an incorrect view of the expressive power of relatively simple linguistic means.

Taking the social agency of dead bodies as its main theme, this introduction discusses the articles in this special section on the bioarchaeology of post-mortem agency and discusses theoretical concerns relevant to the (bio)archaeological analysis of agency. In particular, the argument that only biological living persons have social agency or impact the direction of social processes, or the decisions of others is challenged. The challenges of defining and archaeologically detecting such agency of past humans as well as socially alive entities such as ghosts, corpses, relics and totem heads are discussed. A review of the ways investigations of the agency of the dead might unify bioarchaeologists and further integrate social archaeology into future research is also presented. It is argued that a holistic anthropological approach to humans and their ability to impact their surroundings is not complete if the symbolic and material effects of the dead are not considered.

Defensive and fortified settlements are often places of relatively dense nucleation by people with few viable alternatives, resulting in the imperative need to establish consensual rules for living together. What behaviours and attitudes were adopted in defensively nucleated places? Recent investigations at Ayawiri (Machu Llaqta), a densely settled hillfort town of the western Titicaca Basin of Peru inhabited c. ad 1300–1450, shed light on the nature of social life within the defensive community and the workings of social distinction, priority in settlement, conformity and publicity. I draw on the evidence from Ayawiri to discuss how defensive settlements create and reinforce certain ways of living by reformulating the basic logics of sociality and the material and spatial realms through which they work.

The numerical notation associated with texts and other representational media used in ancient societies is an important means by which past cognitive processes may be reconstructed. No satisfactory typology exists, however, to help understand the relationship between numerical symbols and cognitive processes. As a result, theories concerning the development of numeration remain mired in a unilinear and ethnocentric framework in which our own (Hindu-Arabic or Western) numerals are seen as the ultimate stage of evolution. It is suggested herein that there are two separate dimensions that need to be considered when classifying and evaluating numerical notation systems, and that these dimensions are structured in highly constrained ways. A new typology is presented in which systems are classified into five major types on the basis of these dimensions. Using this typology, a multilinear model is presented for the patterned diachronic change in numerical notation systems, which refutes both unilinear evolutionary theories and radically relativistic propositions regarding how individuals in pre-modern societies represented numbers.

Archaeological excavations in 2002–3 at Hattab II Cave in northwestern Morocco revealed an undisturbed Late Palaeolithic Iberomaurusian human burial. This is the first Iberomaurusian inhumation discovered in the region. The skeleton is probably that of a male aged between 25 and 30 years. The individual shows a characteristic absence of the central upper incisors reported in other Iberomaurusian burials. Accompanying the burial are a stone core and a number of grave goods including bone points, a marine gastropod and a gazelle horn core. Thermoluminescence dating of a burnt stone artefact in association with the burial has provided an age of 8900?1100 bp. This is one of the youngest ages reported for the Iberomaurusian and raises questions about persistence of hunter-gatherer societies in the Maghreb and the potential for continuity in burial practices with the earliest Neolithic.

The debate regarding the status of the Blombos ochre engravings and shell beads for gauging the timeline of when cognitive abilities and symbolic intent appeared has been controversial. This is mainly due to the fact that what is referred to as symbolic is often too loosely defined and is therefore attributed to artefacts in an indiscriminate way. Recent evidence from various concurrent sites in southern Africa, including Blombos, provide the opportunity for a more nuanced analysis of the probable level of symbolic intent and how this relates to neuro-cognitive precursors. In what follows, it will be shown that, although some of the southern African artefacts do indeed demonstrate particular kinds of ‘symbolic’ intent, others need to be approached with caution. Data from the visual brain is presented that provides crucial evidence as to the appropriate level of intent suggested by the engravings and shell beads from the southern Africa context.

Campfire rituals of focused attention created Baldwinian selection for enhanced working memory among our Homo sapiens ancestors. This model is grounded in five propositions: the emergence of symbolism occurred late in the archaeological record; this emergence was caused by a fortuitous genetic mutation that enhanced working memory capacity; a Baldwinian process where genetic adaptation follows somatic adaptation was the mechanism for this emergence; meditation directly affects brain areas critical to attention and working memory; and shamanistic healing rituals were fitness-enhancing in our ancestral past. Each proposition is discussed and defended. Supporting evidence and potential future tests are presented.

Ochre is found at many Middle Stone Age sites and its use is often associated with enhanced mental abilities and symbolism, but the links between the visible uses of ochre and cognition have not been clearly defined. By establishing the technology and processes involved in using ochre, one can determine the skill, knowledge and cognitive abilities required to execute those activities. This is done here by constructing thought-and-action and inferential cognitive sequences for the various ochre activities performed at Sibudu, KwaZulu-Natal, South Africa. Powder production alone is not an indicator of complex cognitive processes, although some planning, foresight and knowledge of materials is required. Some ochre powder was used in the creation of hafting adhesives which is a cognitively demanding process requiring attention-switching, response inhibition, analogical reasoning and abstract thought. The direct transfer of ochre powder from an ochre piece to a soft material through grinding and rubbing requires some complex thought and action procedures — analogical reasoning and the ability to multi-task and switch attention between activities. Scoring a piece of ochre with a sharp tool does not necessitate enhanced executive functioning. However, some engravings demonstrate intentionality and an awareness of space and symmetry that may imply abstract thought.

Within the recent popularity of Actor-Network Theory (ANT) in material culture studies, scholars tend to lose sight of its origin in ethnography of laboratory work. In particular, ANT studies how scientific facts are constructed and stabilized in laboratories so that they become universally accepted, seemingly platonic, categories. This article returns to this initial insight and links it to the long-standing issue of archaeological types. Analysis of the practices of production, consumption and distribution of terra sigillata — Roman archaeology's most salient pottery type — shows how it became a category, how it was stabilized as such, and how this process imbued sigillata with specific agentic properties that allowed it to shape the range of possible actions in the past. By reframing platonic types as constructed categories, they can become active elements in our historical narratives.

This article reviews the evidence for planning behaviour in Middle Pleistocene hominids. It documents the way in which raw material procurement and tool production were structured during the Middle Pleistocene occupations of the Spanish sites of Sierra de Atapuerca, Torralba, Ambrona and Aridos. Differences in the use of raw materials for different kinds of tool or end-product allow inferences to be drawn about pre-Neanderthal intentionality and cognitive ability. The overall pattern of technological behaviour demonstrated by this study is far removed from the purely ‘opportunistic’ and can reasonably be described as involving both forethought and planning. The work is presented from a techno-economic perspective based on the differential use of raw material types present in the lithic assemblages of these sites, and the proximity of sources of these raw materials in the surrounding landscape.

Neuropsychological and developmental studies suggest human musical ability has a deep evolutionary history; but we do not find evidence of the manufacture and use of instruments, with which musical behaviours have often been assumed to be equated, until 70,000 years after the advent of Homo sapiens. This anomaly is addressed by examining the evidence from the fossil record for the evolution of the physiology and neurology required for musical behaviours, with the aim of identifying the development of the physiological and neurological capacity to produce and process melody and/or rhythm. Aural and vocal sophistication appear to have developed in tandem, beginning with full bipedalism around 1.75 million years ago, until a vocal apparatus similar to the modern was present in Homo heidelbergensis 400,000–300,000 years ago. Prosodic and structural aspects of both speech and music production and processing are lateralized in the brain in similar ways suggesting evolutionarily-shared foundations for these mechanisms.