Keywords

1 Introduction

In this paper we explore the regionalisation of rock art styles in the terminal Pleistocene of Northern Australia (c. 18–12 ka) and the Late Pleistocene of Europe (c. 45–12 ka) to elucidate examples of human network building at different temporal and spatial scales (Johnston et al. 2017; Ouzman et al. 2017) (Fig. 5.1). The early development of regional rock art styles provides one important stream of evidence for how people have negotiated change and cross-cultural interaction, utilising rock and mobiliary art within social networks and for information exchange. In Northern Australia, human-focused rock art provides a rich repository of human sociality with a focus on body ornaments and perishable accoutrements not normally recovered from excavation contexts. Of particular interest in this respect is the Gwion (or Gwion Gwion) art style in the Kimberley (Northwest Australia). This sits at an important juncture of social and demographic reconfigurations, dated to the terminal Pleistocene/early Holocene transition (~12 ka) as outlined in several foundational papers characterising time brackets for earlier figurative animal art from 17 ka – 13 ka and for human Gwion Gwion art at c. 12 ka (Finch et al. 2020, 2021). We will discuss these findings in the light of rock art from the European Upper Palaeolithic and the transformative environmental changes that occurred during and after the Last Glacial Maximum (LGM). We explore a range of interpretations between the two hemispheres. For Northern Australia, we build upon recent and emerging research across the wider region that includes the suggestion of the existence of a tropical Pleistocene era culture bloc with possible regional cultural exchanges (Florin et al. 2020; Veth et al. 2021). In the context of European Palaeolithic art, we consider the intersection of past social and cultural meanings at different scales. For both the northern and southern hemispheres, we contextualize the respective phenomena in relation to recent archaeological evidence, new radiometric dates, emerging palaeoclimatic frameworks, and modelled changes in social organization. Through our analysis, we want to explore late and terminal Pleistocene networks through rock art and mobiliary figurative expressions. Our explorations will draw attention to the existence and persistence of past regional relationships that emerged at opposite ends of the world and perhaps in response to similar global glacio-eustatic drivers (see discussion in Aubert et al. 2018). While these networks should not be regarded as historical precursors to the age of globalization that characterises the modern world, they allow insights into the sometimes vast, spatial dimensions of human flexibility, adaptability, and fundamental sociality through deep time. We aim to explore the tensions between global narratives – here profiling the expanding and contracting worlds and human responses due to glacio-eustatic and climatic changes – and local cultural expressions that see forager-gatherer-hunters engage in intensive production of anthropomorphic form(s) in both the southern and northern hemispheres, yet unquestionably as unique local expressions with clear emblematic differences.

Fig. 5.1
A map of Kimberley region showing the current landmass, Drysdale river, and King George River.

Map of the Kimberley region and Arnhem Land showing extent of the coastal plain towards the Timor Sea at peak LGM c. 20 ka. (Map: Emma Beckett).

Full size image

2 Art, Regionalisation, Globalization, and Networks

One of the most persistent topics in the study of past hunter-gatherer societies is the attempt to understand and model their spatial behaviours. Before the widespread adoption of farming and animal husbandry practices after ca. 10 ka ago, virtually all human populations relied on different forms of mobility and territorial marking to fulfil their social and economic needs. Both within hunter-gatherer studies and in relation to the need to develop a systematic basis for archaeological inferences, studies into mobility patterns and related causalities have increased rapidly over the last several decades (Kelly 1992; Perreault and Jeffrey Brantingham 2011; Whallon 2006). From the 1960s onwards, an era which can be described as the modern phase of hunter-gatherer studies, the crucial role of the environment was thoroughly recognised and explored in increasingly sophisticated ways. Probably the most influential early paper in this respect was Binford’s (1980) Willow smoke and dogs’ tails, in which he proposed a basic distinction between a forager and a collector mobility pattern. This distinction is fundamentally connected to the spatial and temporal distribution of resources in an environment and the respective human responses to these challenges. From these insights, a range of conclusions can be drawn about a wide range of human behaviours at different temporal scales and with reference to the impact on different material cultural expressions ranging from the pattern of sites across landscapes, the structure of camp sites and down to preferred hunting equipment (see e.g., Bleed 1986). However, while such a human behavioural ecology approach has its clear merits, human societies and behaviours do not exist in isolation and thus issues of demography, the creation and maintenance of social networks, and the mechanisms for the exchange of information within and between populations need to be equally considered. In this paper, we aim to explore some of these wider themes in relation to the possible roles of rock art and mobiliary art objects (after Conkey 1989).

Explicit attempts have been made in Australia to theoretically engage rock art styles with archaeology and palaeoenvironmental studies and many of the polarities inherent in previous approaches (for example, informed versus formal; symbolic versus functional; ritual versus mundane; gendered versus ungendered) are now being unpacked in more nuanced ways (chapters in McDonald and Veth 2012; Veth et al. 2016, 2021). The fact that rock art can signal information at many levels and has agency between culture groups and across time and space as well as inter-generationally is a recurrent theme as is its organising role in ideational, sensory, social organizational, religious, hierarchical, territorial, and economic domains. The information content of rock art, when viewed within its larger archaeological and environmental contexts, can inform on multiple facets of past behavioural systems (Porr and Veth 2017). And as such, we can use different theoretical approaches for different scales of analysis.

At a broad scale, an evolutionary approach has the significant advantage in that it does not assume (or require) ‘uniform’ human intentionality for the continuities and changes in the styles used. Put another way, proliferation events in successive rock art style phases and gradual changes between these (Travers 2015) can be viewed as patterned symbolic behaviours that are the human evolutionary outcomes of new ecological states. The same pressures will influence other non-symbolic social and economic behaviours (such as group mobility and stone artefact technological systems) and, therefore, these archaeological records are coupled with changing art production modes. An example would be the greater degree of homogeneity in rock art styles across arid regions due to the low density and high mobility of groups, and with a high degree of language/dialect permeability (McDonald and Veth 2013a, b). Innovation, learning and adoption of rock art operate as humanly mediated outcomes. However, the success of its transmission through space and time will be influenced by the direction and intensity of environmental, climatic, and hydrological changes.

In evaluating rock art style phases covering many millennia, we argue that the temptation to use single prime movers to explain change (such as climate, demography, or artistic convention) should be avoided. Changes are likely to be multi-factorial and coupled rather than deterministic, with longer-term processes favoured. We suggest that these explanatory frameworks can encompass evolutionary social biology approaches and Group Boundary Formation (GBF) theory (Foley and Lahr 2011); climate drivers and biome changes as these influence hunter-gatherer mobility patterns (Kuhn et al. 2016); and changing information-exchange networks (McDonald and Veth 2013a). Additionally, the repeated and recursive use of places and existing art (McDonald and Veth 2013b) and cladal, non-progressive models, for changing art styles (Johnston et al. 2017; Travers 2015; Veth et al. 2016; Welch 2016) should also be incorporated as these are well-documented human behaviours over time. Settlement and aggregation behaviours can be understood according to GBF and Information Exchange Theory (IET; and see Veth et al. 2021). Using GBF and IET, different settlement and mobility patterns are modelled to be in phase with – and not determined by – different environmental settings (Veth et al. 2000; Whallon 2006). That social groups would have experienced various stressors with these environmental, territorial and consequent demographic changes is not in question. We argue that changes in a group’s boundary maintenance and information exchange behaviours are often expressed and managed most visibly via rock art styles, a theme developed in detail by Conkey (1978, 1980 and 1984, also see Wobst 1977). Style phases are a specific coding of a group’s self-image and subsistence as groups engage with environmental and cultural drivers. In the Australian case studies discussed below, during higher precipitation regimes, more ‘closed’ information systems tend to occur with greater emblematic group-identifying behaviours and higher stylistic heterogeneity. In contrast, arid phases generally tend to result in more ‘open’ information systems where more stylistically homogeneous schemes occur across permeable group boundaries and over larger landscapes (McDonald and Veth 2013a). While some of these latter aspects can also be discerned in European Palaeolithic art, it also needs to be recognised that the dimensions discussed so far generally do not explain all aspects of figurative imagery. Material culture is always polysemic and meaning is established relationally (e.g., Motta and Veth 2021). This applies both to past meanings and those created through modern research questions and analyses (Porr 2018). Material culture items can generally be related to different causalities and different rhythms that operate at different temporal scales. At smaller scales, social meanings and particular contexts of use and consumption can often be discerned while at lager scales regional and interregional social networks become relevant as they relate to particular environmental conditions.

3 Europe

Europe comprises the western portion of the Eurasian sub-continent which is the largest land mass on the planet (Fig. 5.2), stretching from the Atlantic Coast to the Eastern shores of Siberia. This enormous region has a human history that covers at least two million years as evidenced by early hominin sites in Georgia and China (Muttoni et al. 2018). However, here, we will focus on a much later period and ‘constrained’ area, the Upper Palaeolithic of Europe, which is broadly dated to between ca. 45 and 12 ka ago. During this time, Neanderthals were replaced by Homo sapiens populations and from ca. 40 ka ago, Europe was exclusively inhabited by the latter (Higham et al. 2014). The ability to create regional or interregional networks might have significantly contributed to the survival rates and expansion of Homo sapiens populations during times of environmental stress. As French (2021) has argued, the Neanderthal-to-Homo sapiens transition marks an important threshold in the demographic history of Eurasia. It is the time that saw a crucial expansion of social lives and an increase in the connectedness of human societies on increasingly larger scales. Populations, “while still small, were no longer ‘small-scale’” (French 2021, 173).

Fig. 5.2
A map of the Eurasian sub continent showing the current land mass, major rivers and oceans.

Map of the Eurasian subcontinent showing key regions discussed in this paper. (Map: Emma Beckett)

Full size image

3.1 Environment and Palaeo-Climate

The time period that is relevant for the European Upper Palaeolithic falls within Marine Isotope Stages (MIS) 3 and 2. Most importantly, it encompasses the Last Glacial Maximum (LGM), which is currently dated to between ca. 27 and 20 ka. While significant climatic fluctuations still occurred, the LGM was characterised by generally very cold conditions and a considerable extension of Alpine and Scandinavian glaciers, which made occupation of the northern latitudes of the continent challenging for human populations. It is clear that large parts of the current temperate zone of Europe were unoccupied during the LGM (Maier and Zimmermann 2017). The climatic downturn was followed by a complex but generally rapid phase of amelioration. The retreat of the glaciers was followed by the establishment of open steppe and tundra habitats, which were superseded by both open and increasingly closed coniferous and deciduous forest types. The first modern humans entering Eurasia sometime between 50 and 40 ka would have encountered deciduous woodlands in the southern parts of the region, for example, around the current Mediterranean and Black Sea, and coniferous forests that stretched across the whole northern sections of the continent. Overall, during MIS 3, habitats exhibited much variability, offering a range of adaptive opportunities, including mountainous and coastal environments. With the cascading down of temperatures, these opportunities partly narrowed but others were enhanced or emerged. The period between ca. 27 and 20 ka saw the expansion of boral and tundra conditions further towards the south, impacting the productivity of plant and animal resources. However, during this time, Europe also became part of one of the most significant and rich glacial environments connected to one of the most iconic Ice Age animals. The concept of the so-called ‘mammoth steppe’ was first proposed by Guthrie (1990), who argued for the existence of this specifically Middle and Late Pleistocene habitat that does not have a clear current ecological equivalent. This steppe was the product of a unique combination of relatively low temperatures and marked aridity during the Pleistocene, which both created conditions for a rich soil and abundant growth of grasses. This environment supported an extraordinarily rich biomass of grazing animals (mammoths, bison, horses, woolly rhinoceros, reindeer etc.) together with communities of large predators (lions, leopards, wolves, and hyenas) (Gamble 2013, 232–236). When climatic conditions deteriorated further around the LGM, however, human populations retreated towards refugia in the south of the continent, for example in Southwestern Europe. Accordingly, the amelioration of climate after ca. 18 ka allowed populations to expand again northwards and resettle regions that were either previously covered by glaciers or had endured arctic conditions during the LGM.

3.2 Archaeology

The climatic and environmental changes described above equally presented challenges and opportunities for human occupation. After the initial occupation of Europe by Homo sapiens groups, these populations followed complex patterns of expansion and contraction interacting with geographical, environmental, and climatic variables. The first fully established Upper Palaeolithic technocomplex is the so-called Aurignacian (ca. 45 to 28 ka). It is generally accepted that the Aurignacian was exclusively associated with Homo sapiens. Compared to the preceding Middle Palaeolithic, it is characterised by a much higher standardisation in lithic and organic technologies, which are recognisable from the Iberian Peninsula to the Caucasus. These similarities seem to reflect more integrated and stable social transmission mechanisms that operated over vast distances and for several 1000 years (White et al. 2015). The Aurignacian is also the time during which personal ornaments become much more common in the archaeological record. They were made from a range of organic materials and were often highly standardised in their shapes and decorations. It has been argued that the distribution of these implements across Europe reflects the existence of ethno-linguistic groupings during the Aurignacian (Vanhaeren and d’Errico 2006). At a smaller scale, it appears that differences in ornamentation’ allowed for the differentiation of local social groups inhabiting neighbouring valleys (Dutkiewicz et al. 2018). Despite these undeniable complexities, the settlement patterns across Eurasia were rather diffuse and population densities were probably quite low (Schmidt and Zimmermann 2019). At a local level, however, differentiations between habitation sites and locales for other and probably restricted purposes can be observed, for example, for some cave sites in the Swabian Jura, Southwest Germany (Porr 2015).

During the subsequent Gravettian phase (ca. 28 to 20 ka), the increasingly cooler and dryer conditions allowed human populations to make use of the opportunities presented by the mammoth steppe environment (Bicho et al. 2017). While the Aurignacian was mostly restricted to Europe and parts of Anatolia, the Gravettian became a true Eurasian technocomplex and typical tool types from this period can be found over vast distances across Europe and the Eurasian Plain. The presence of large and predictable herds of herbivores allowed the development of much more focused settlement patterns and demographic expansion. Large numbers of mammoths and other grazing herbivores were not only utilised for food. Their bones were also used to construct massive open-air habitation structures that were possibly occupied on a semi-permanent basis enabled by food storage practices. Most of these mega-sites occur along the Desna/Dnieper river system and are integrated into large settlements of up to 10,000 square metres (Soffer et al. 2000). From this period, there are also a number of complex multiple burial sites containing rich grave goods that point towards the emergence of social stratification (Pettitt 2011).

These complex social and economic structures were disrupted during the height of the LGM, which severed lifeways and communication pathways operating across the Eurasian plain (Maier and Zimmermann 2017). During the coldest periods of the LGM, large parts of Central and Eastern Europe were uninhabited. Subsequently, Europe was resettled from Southern European refugia while Eastern Europe was recolonised from refugia along the Caucasus and other Central Asian Mountain ranges. In Western Europe, the main post-LGM technocomplex was the Magdalenian that can be found as far as present-day Poland. In Eastern Europe, so-called Epi-Gravettian technocomplexes show more continuity with pre-LGM traditions but in both cases, the trajectories were leading towards various Late Palaeolithic technocomplexes of highly mobile foragers in a world of increasingly forested environments (Wygal and Heidenreich 2014; Gavrilov 2021; Miller 2012; Lengyel et al. 2021).

3.3 Rock Art and Mobiliary Art

It has long been noted that the emergence of Upper Palaeolithic technocomplexes across Eurasia appear to be connected with the regular occurrences of unequivocal figurative imagery in the archaeological record (Lorblanchet and Bahn 2017). While different archaeological artefact categories do not map neatly onto art/stylistic categories in the European record, already during the Aurignacian the imagery shows a high level of diversity and occurs both in the form of rock art and mobiliary objects. The latter include the famous Grotte Chauvet and the ivory statuettes of the Swabian Jura Mountains (Floss 2015; Clottes 2003). During this early period, figurative representations, which overwhelmingly depict animals, are not as standardised as organic or lithic tools. The spatial patterning seems rather to follow the evidence of personal ornaments, outlined above, with broadly shared conventions and regional variations. The sample size for figurative representations is, however, very small and thus no quantitative analyses are possible. A qualitative comparison between the Grotte Chauvet and the mobiliary art of the Swabian Jura caves shows a broad similarity in the choice of motifs (with a preference for lions) and some stylistic conventions but also a lot of variability (a trend first observed by Clottes 1995 and elsewhere). Cave art appears to be absent from Central and Eastern Europe, instead being restricted to Southwest Europe. Overall, the available evidence appears to mirror the structure of Aurignacian demography with relatively low population densities and a high degree of social and cultural variability. The tension between shared conventions and expressions of individuality can also be seen in some of the collections of figurative art themselves. In the Swabian Jura statuettes, the depictions of mammoths and lions show coherent references to animal behaviours (along a carnivore/herbivore opposition) but little stylistic standardisation (Porr 2010a). Figurative imagery also exhibits a complex dialectic between change and continuity through time, which is evidenced by the initial occurrence of a female statuette at Hohle Fels cave (Germany) and a comparable depiction at Chauvet Cave (France) during the Aurignacian, long before the proliferation of this motif during the Gravettian period (Porr 2010b). Figurative imagery does not simply change with environmental conditions and adaptive strategies; it also follows its own rhythms.

Some of the most well-known figurative forms from the European Palaeolithic are the female statuettes from the Gravettian period. The most famous of these statuettes is the so-called Venus of Willendorf. These representations occur mostly as mobiliary or portable items, however other expressions are also known such as the famous engraving at Laussel, France (Gaudzinski-Windheuser and Jöris 2015). One particularly striking feature of these female statuettes is their enormous geographical range. They occur from Iberia to the Baikal region of Siberia. Across this enormous area and despite some significant variations, the statuettes share several common features: they are usually carved in stone or ivory, they often depict unclothed and, less commonly, clothed women. In Western Eurasia, they seem often to depict mature women with sometimes exaggerated sexual attributes. Contextually, the statuettes almost always occur in habitation sites, and they appear, therefore, to be connected to a domestic and public sphere. They are a regular feature in the extensive habitation structures mentioned above. In a seminal paper, Gamble (1982) argued that the statuettes reflect extensive networks of interactions and alliances which operated to mitigate risk during the harsh climatic conditions of the LGM. The statuettes were consequently interpreted as “visual mechanisms of information exchange designed to establish and maintain alliance networks between groups living up to thousands of kilometres apart” (French 2021, 242). The Willendorf-style motifs are, therefore, interpreted as evidence for the existence of open social systems, which is consistent with the ecological structure of the mammoth steppe, as outlined above (e.g., Conkey et al. 1997; Soffer et al. 2000). This inference is further supported by demographic reconstructions that seem to suggest that Gravettian interaction networks allowed for the survival of viable populations, linked as nodes, during the LGM despite local extinction events (Maier and Zimmermann 2017).

The role of rock art and mobiliary art during the European Upper Palaeolithic has also been explored for post-LGM societies. The proliferation of mobiliary art objects and painted cave sites in the Franco-Cantabrian region in Southwest Europe after ca. 18 ka has been explained as a reflection of demographic refugia effects and the necessity to negotiate increased population densities (Conkey 1987; Rivero and Sauvet 2014; Gravel-Miguel 2016; Fuentes et al. 2019). In terms of the interrelatedness between figurative imagery and social networks, the re-colonisation of Central and North-western Europe was also accompanied by the emergence of new and widely distributed motifs (Maier 2012). During the Late Magdalenian period between ca. 16 and 13 ka, the so-called Gönnersdorf-style female figurines emerge (Bosinski 1987). They are characterised by highly abstract and standardised depictions of the female body. Most famous in this context are the numerous engravings on slate plaques at the site of Gönnersdorf, Germany. Here, they are arranged in groups and seemingly engaged in communal activities or rituals (Bosinski et al. 2001). They also occur as mobiliary objects and were apparently personal items. These objects were, for example, found at the Magdalenian sites of Oelknitz and Nebra (Fig. 5.3) (see Braun 2018 for an overview). Finally, in some exceptional cases, they were also manufactured from lithic materials as evidenced by some extraordinary findings in Poland (Fiedorczuk et al. 2007). It has been argued that these representations are connected with the rapid (re)colonisation of post-LGM habitats (Maier 2015, 2017). They possibly reflect the establishment of networks between pioneering groups with their level of abstraction acting to suppress individual expression. Gaudzinski-Windheuser and Jöris (2015, 312) have argued that the respective long-distance communication networks “focused on the mandatory functioning of individuals within a group to ensure survival. Thus, the individual sphere must have been subordinate for the group and was presumably reflected in the absence of depictions of individuals”. While the details of this interpretation need to be addressed in future contextual analyses, during the Late Palaeolithic period, mobiliary art clearly played a crucial role in the establishment and maintenance of large-scale cultural entities and reflects open systems of social interaction.

Fig. 5.3
A photograph and caricature of three female figurines.

So-called Gönnersdorf-style female figurines characterised by highly abstract and standardised depictions of the female body from the Magdalenian site of Nebra, Germany. (Landesamt für Denkmalpflege und Archäologie Sachsen-Anhalt; photo: Juraj Lipták)

Full size image

4 Northern Australia and Sahul

4.1 Environment and Paleoclimate

Our modelling for northern Australia considers changes in the intensity of the Australian–Indonesian summer monsoon, rainfall, temperature, and sea levels. Sea level modelling shows that between 14 and 12 ka vast areas of the coastal plain (~200 km of coastline of the Bonaparte shelf), representing significant estates for NW Kimberley peoples, was drowned due to sometimes rapidly rising sea levels (Ishiwa et al. 2016; Williams et al. 2018), and especially between Meltwater Phase 1A and 1B (Finch et al. 2021). This is precisely the time which sees a switch between large naturalistic animal depictions dating to times of lower sea level (and possibly to the peak of the Last Glacial Maximum at 19 ka) towards the efflorescence of anthropomorphic images reflecting the demographic packing of populations slowly relocating to interior estates. In the case of the Kimberley region (some 424,000 km2), the effects of the possible loss of the Indo-Australian Monsoon before the Last Glacial Maximum (LGM); the drowning of the Bonaparte Basin after the LGM; ENSO intensification from 4 ka and then relaxation of ENSO during the last 1.5 millennia, resulted in significant changes in precipitation, available territory, and seasonality (Veth et al. 2016). While the LGM reconfigurations of northern Australia did not act to isolate groups in glacial refugia as occurred in Eurasia, parts of some bioregions may have been used less frequently or intensively (Law et al. 2021; Veth et al. 2022). During the LGM a coextensive land bridge emerged between the Kimberley and Arnhem Land to the east, accompanied by the emergence of style regions in each culture bloc, between which there could have been some degree of information exchange, as expected in an open and permeable exchange mode Lewis 1988). Major falls in sea level up to -130 m at the peak of the LGM (c. 19 ka), created contiguous cultural landscapes over the broad and shallow continental shelves of the north, which were then subsequently drowned and lost entirely by 6 ka.

4.2 Archaeology

The Kimberley was a vast and continuously occupied hunter-gatherer landscape as deduced from both the archaeological record (Vannieuwennhuyse et al. 2017; Veth et al. 2019, 2021) and recent genomic studies (Bird et al. 2018; Malaspinas et al. 2016). The known occupational history of the Kimberley extends back to at least 50 ka (Norman et al. 2022; Veth et al. 2019; Wood et al. 2016), with strong evidence that it has likely acted as a large-scale refugium in the larger Australian context through all of this time (Veth 1993; Williams et al. 2013; Wood et al. 2016). Because the region extends from the interior to the sea including plateau, riverine and plains hosting desert, riparian, woodland and (sub)tropical vegetation we might expect very different human adaptive strategies through time. And that is precisely the patterning emerging (Fig. 5.1), with detailed site-catchment archaeologies being described from the edge of the Great Sandy Desert (Balme et al. 2019; Veth et al. 2009), along the Devonian reef and sandstone plateau (O’Connor and Fankhauser 2001; O’Connor et al. 2014; Veth et al. 2019) and from the western and northern maritime zones and subcoastal riverine catchments (Dortch 1977; O’Connor 1999).

There are marked introductions of technologies and practices through time, such as the edge-ground axe and ground bone points at 46 ka, application of ochreous pigments by 40 ka, and a sequence of very different rock art styles dated from at least c. 20 ka. There is also the long-distance transport of shells for ornaments and implements at 30 ka, and bifacial and backed point production by 5 ka (Hiscock et al. 2016; Langley et al. 2021; Maloney et al. 2014). Through the lenses of Information Exchange and Group Boundary Formation Theory, different scenarios of cultural/economic and symbolic sharing can be expected across the Kimberley depending on (a) sea level and position on the continental shelf, (b) latitude, (c) catchment types and (d) resource patchiness in addition to predominantly social trajectories (Motta et al. 2020; see summary in Veth et al. 2021). Distinct regional material cultures are in place by the terminal Pleistocene and unquestionably by the mid-Holocene. Against the backdrop of changes in sea levels, monsoonal intensity, and vegetation structure, is the emergence of the material signatures of distinct polities, likely also of related language and dialects, which show both persistence and flexibility in long-term residence patterns across the Kimberley. The case for continuity is overwhelming as is the complexity of technological and symbolic practices associated with these non-Pama-Nyungan speakers. The central role of the Kimberley region in the settling of Australia (Bird et al. 2018), emergence of complex symbolic practices (Veth et al. 2021), and regionally distinct land-use and resource extractive practices (Hiscock et al. 2016; Ouzman 2021) makes it the ideal landscape to examine issues of regionalism and long-distance exchange and connection. Unsurprisingly, in many respects it shares some of the long-term occupational patterns and regional symbolic variabilities seen in Arnhem Land to the east (Lewis 1988).

4.3 Rock Art

Across northern Australia, advances in scientific dating techniques are confirming terminal Pleistocene ages for human figures in rock art; specifically, age constraints of 12 ka for Gwion Gwion figures and an earlier date and brackets for Irregular Infill Animal Art of 17–13 ka (Finch et al. 2020, 2021; Jones et al. 2017, 2020). This research provides opportunities to develop a more robust chronology and to explore how we understand the role of style, and how linear the progressions may be, in these regions. Superimposition sequences in both areas are indicative of early figurative traditions that change in significant ways through time (Chaloupka 1993; Walsh 1994; Welch 2016). In both the Kimberley and Arnhem Land, two of the richest bodies of figurative rock art globally, earliest art styles are dominated by naturalistic animals which currently have absolute age brackets from AMS as early as 17 ka (Finch et al. 2021; Jones et al. 2020) with very few figurative human depictions (Fig. 5.4). Humans are more obviously present via hand-prints and stencils. However, by around 12,000 years ago, a significant shift happens in northern Australia to focus on humans, in scenes, rich with dress and material culture displays (Fig. 5.5). This is happening at a time of significant territorial loss with sea level rise and inferred social responses, which include the proliferation of Gwion and Dynamic figures. That this pattern is observed in both regions suggests dynamic cultural exchanges are being reflected in shared rock art traditions (Lewis 1988; May et al. 2018; Taçon et al. 1999; Veth et al. 2011).

Fig. 5.4
A photograph of a rock face with paintings of three kangaroos.

Panel of early Irregular Infill Animals, here three life-size kangaroos in red ochre with, selective battering of anatomical features, King George River. (KV_BAC_KGR_020: BAC/UWA, and Ambrose Chalarimeri)

Full size image
Fig. 5.5
A photograph of a rock face with paintings of a giant creature with shoulder arm and wrist ornamentations and waist band.

Large Gwion figures with details of headdress, shoulder arm and wrist ornamentations, tassel waist band and carrying dilly bags, King George River. (KV_BAC_KGR_003: BAC/UWA, and Ambrose Chalarimeri)

Full size image

What concerns us here is why two contiguous areas of the northern Australia, the Kimberley and Arnhem Land (Fig. 5.1), should have both witnessed a proliferation of rock art focusing on grouped (read socialized) humans. Both regions show prolific details in headdress, hair belts, necklaces and bangles, dilly bags and pubic aprons, as well as wooden artefacts including spears, digging sticks and ceremonial ‘staffs’. While the schemata are not identical, with Kimberley Gwion figures generally shown in more erect and formal stances (in contrast to subsequent dynamic Elegant Action Figures – Walsh 2000), in comparison to the Dynamic Figures of Arnhem Land, there are many common attributes which allows us to think about these as ‘cousin-styles’.

The most recent phase or substyle amongst Gwion art is the shift to dynamic figures (Fig. 5.6), labelled ‘Elegant Action Figures’ by Walsh (1994, 2000). This shift sees a move away from the codified, large Gwion scenes of parallel, richly garbed humans suggestive of ceremony, to scenes of life, such as hunting, sex, and camping. With this shift, there are increasingly gendered figures which are smaller, with a simplification of form and reduction in dress, and the inclusion of animals, largely macropods, appearing as key figures within scenes. Additionally, the highly visible placement of earlier Gwion is no longer dominant, and in contrast, discrete panels that engage with the less obvious features of the rock (e.g., an erosion stain; a rippled under-ceiling surface) are incorporated into placement choice.

Fig. 5.6
A photograph of a rock face with paintings of humans, kangaroos and rivers.

Transitional Elegant Action Figures showing animated and juxtaposed humans, kangaroos and therianthrope on shelter roof, King George River. (KV_BAC_KGR_028A: BAC/UWA, and Ambrose Chalarimeri)

Full size image

5 Discussion

Both study areas from the northern and southern hemispheres have experienced significant environmental changes during the Late Pleistocene. The impacts of climatic fluctuations around the LGM on Eurasia and Northern Australia (Kimberley and Arnhem Land) have been substantial with significant impacts on the mobility of human populations. In each example, we can track human adaptive and social responses. Although these are not seen as simple reactions to changing environmental conditions and resource distributions, we can discern coherent patterns emerging in both regions in both mobility configurations and forms of visual communication. Across the Kimberley and Arnhem Land, there are comparable visual repertoires between approximately 17 ka and after 12 ka (see Jones et al. 2017, 2020), which follow a largely similar trajectory over time. In both regions, and dating from the tail end of the LGM, the earliest phase of figurative rock art is dominated by naturalistic depictions of large animals. This imagery appears to have a mostly public character for the following criteria (a) the motifs are generally on single panels most of which have wide viewscapes, (b) animal types are aggregated in related scenes (c) they are portrayed as life-size motifs with peripheral infilling, and (d) they exhibit communal and complex relational ontologies (see Motta and Veth 2021). This latter characteristic is shared with the following phase of art where it switches to a range of elaborately decorated human figures, the Gwion Gwion style phase, often engaged in communal and seemingly ritual activities (Fig. 5.5). This phase is followed by a diversification of the art, which becomes more stylistically encoded and restricted, whilst less elaborate (Fig. 5.6). Finally, during the early Holocene and from the mid-Holocene onwards, publicly displayed art forms begin to dominate again, and these seem to articulate with a stronger local focus and the formation of estates within larger interconnected cultural blocs. Across Europe, the changes in the art over time seem to relate to an initial settlement of the continent by modern human groups, who were connected with each other but still displaying a considerable degree of variability, and especially in their figurative visual repertoires. This phase was followed by a long phase of vast open networks between hunter-gatherer populations, who sometimes followed strategies of reduced mobility accompanied by an increase in social complexity. These open networks continued throughout and after the LGM in different forms and at different scales. Only with the onset of the Post-Pleistocene reforestation, did populations increasingly form locally focused social and economic strategies.

Art forms reflect and mediate these changes in complex ways. Across Europe, mobile art and rock art seem to reflect the nature and reach of communication networks that articulate with broad environmental changes and the availability of resources. Only in Southwest Europe and under demographic refugia conditions, do local focus areas appear, being created around communal and highly elaborate rock art sites. These are now some of the best-known Palaeolithic painted caves. They reflect a highly structured and complex social landscape, which most likely relates to social group formations and respective restrictions, rules, and interconnections. In terms of motif choices, no dramatic switches in the overall repertoire can be discerned. Human and animal depictions are continuously created in parallel. However, animal imagery continues to dominate throughout, and human depictions remain under-represented (Rivero and Ruiz 2019, 6). This is a slightly different pattern than the mobiliary art of the Gravettian and Epigravettian, which shows a greater proliferation of human depictions in the form of female statuettes. However, these findings need to be considered with regard to taphonomic processes, which negatively impact parietal art in caves more than rock art. Throughout, animal and human depictions are highly contextual and the female statuettes may have had a role in negotiating social identities across domestic and ritual spheres. While they reflect large-scale patterns, they probably played more active roles in the negotiations of social life than in the establishment of communication networks themselves.

Across the two Northern Australian regions we argue that rock art has played a crucial role in the negotiation of social formations, places, and boundaries. We believe two key drivers for change over time were the loss of significant habitable land areas and the associated changes in available resources. Consequently, the dynamics of GBF (Veth et al. 2021) are reflected in the appearance of stylistic heterogeneity and boundary marking and differentiation; whilst the paired transition across these regions with inferred encoding of information readable across the northwest, indicates continued exchange of visual strategies, as understood through IET. The initial period of naturalistic depictions of animals, the ‘external gaze’, appears to reflect a focus that is arguably similar to some European Palaeolithic art. As recent research in Indonesia has established, this focus also has a similar antiquity in Borneo and Sulawesi (Aubert et al. 2019; Brumm et al. 2021). In Southeast Asia and Northern Australia, they do not appear to have been spatially focused to the same degree. They seem to reflect a more flexible spatial and social organisation of interconnected hunter-gatherer groups. During the subsequent period of rapid inundation of the coastal plain (c. 12–8 ka), visual systems in the Kimberley and Arnhem Land begin to exhibit a less diffuse spatial pattern. Emergent regionalism occurs from the terminal Pleistocene with loss of territory and the reconfiguration of smaller and bounded groups. Interestingly these two adjacent regions fall within the northerly and more ancient non-Pama Nyungan language family, independent of the Pama-Nyungan family spreading across the rest of the continent during the Holocene (Bouckaert et al. 2018).

The visual repertoire from c. 12 ka onwards shows a high degree of standardisation together with a strong focus on human depictions, as outlined above. The land becomes marked with representations of ritual activities, with people invested in rituals themselves. It is difficult to avoid the impression that the art played a role in negotiating or renegotiating the relationships between human groups in often expressive and publicly visible forms. The general switch from a focus on animal to human depictions represents the greatest difference between the Northern Australian and the European art repertoires examined in this paper.

In Southeast Asia, a similar switch has been recently described from rock art imagery in Borneo. The so-called datu saman figures exhibit a close similarity with some Gwion art in the Kimberley (after Aubert et al. 2018). Interestingly they are dated to only 1.6 ka earlier at 13.6 ka than Gwion. It would be tempting to infer the existence of a system of interrelated cultural areas between Northern Australia and Southeast Asia during the Late Pleistocene in the form of a ‘trans-Wallacean tropical cultural interaction shere’. The current radiometric dates seem to suggest that the situation cannot easily be resolved, and it rather appears that we are dealing with parallel developments of the negotiation of social identities through visual expressions of ritual activities. Nevertheless, these may be rooted in similar deep cultural and ideological foundations and perhaps shared common drivers associated with the LGM.

During the latest stages of the Pleistocene and in the Holocene, rock art becomes less and less central across Europe for the establishment of social relationships. The reforestation across the temperate zone of the continent pushed communities towards less large-scale mobility and economic patterns. Visual systems of communication become more difficult to discern with more material expressions made from perishable materials (see Rivero and Ruiz 2019 for an overview). In contrast, across the Kimberley and Arnhem Land, hunting and gathering groups started to organise themselves in local estates and created a complex social and cultural landscape. This configuration articulates well with the recent Monsoonal weather pattern and the respective environments and resource availability patterns.

Based on ethnographic records, these estates were connected to spiritual ancestors such as the famous Wandjina of the Kimberley. These areas were, nevertheless, connected with each other and it remains an open question if human groups and individuals actually stayed within these estates or expanded their ceremonial ranges, as recorded ethnographically (e.g., the Kunapippi ceremony). Within this system emerges the Wunan network, a Kimberley-wide social network and trade system with reciprocal obligations, that sees the movement of valuable objects, such as pearl shell from the coast into the arid zone (Akerman and Stanton 1994; McCarthy 1939). These trade routes, driven by individuals responsible for links within the chain, endure into the recent past (Bradshaw et al. 2021). Archaeological and ethnographic evidence supports the deep interconnectedness of people at different scales. Exchange networks not only encompassed the Kimberley or Arnhem Land, but also adjacent desert groups. In both case studies, the available evidence from the rock art, mobiliary art, archaeological and environmental contextual information allows us to gain important insights into the development of local, regional, and interregional patterns of connections that sometimes reach almost global dimensions over many thousands of kilometres.

6 Conclusion

Our comparison of largely figurative art rock art schema between Europe and northern Australia uncovered some parallel developments in art repertoires, with a common foundation of dominant figurative animal iconographies followed by subsequent shifts in style and theme due to environmental challenges associated with the Last Glacial Maximum. These environmental vectors included glacial refugia in the north and sea level changes in the south. In tracking such global responses, this paper speaks to global drivers with local responses. While the post-LGM response in northern Australia, and possibly in the broader trans-Wallacean region, was a distinctive switch to predominantly socialised human figures, European rock art shows a more complex pattern, although human representations do also become more common in some regions. Whilst we argue for open information exchange systems in both regions, interactions in Europe were facilitated through different media with a distinct focus on mobiliary art such as the Willendorf-style and Gönnersdorf-style figurines. While environmental drivers associated with the LGM create globally shared phenomena, specific cultural responses are localised and regionally constituted, as evidenced in the unique inter-regional patterns demonstrated in this paper.