What Were Rock Art Sites Like in the Past? Reconstructing the Shapes of Sites as Cultural Settings

Abstract

Rock art research often focuses on the art, rather than on the site or its landscape. Yet what makes the art meaningful in culture is not just the paintings, stencils or engravings, but the individual and connected places where they are found and of which they are a part. Over time, those places can change, and sometimes dramatically so. To understand the art, attention thus needs to be given on these spatial contexts: the details of what was where in the past matter. In this chapter we argue that in archaeology—a discipline aimed at understanding the cultural past—the form of the landscape of rock art sites at the time the art was produced and engaged needs to be understood. We do so by investigating four dimensions of a site’s past landscapes: its past landforms; its palaeo-entrance; its palaeo-landmarks and pathways; and its past rock surfaces. Each brings new insights on the physical configuration of a rock art place, as context of its cultural significance and engagements.

1 Introduction

By definition, deep-time art in the twenty-first century begins with a conundrum made of history’s hermeneutics: the actions of people a very long time ago are seen and understood from present-day perspectives. In archaeology, the aim is to record what is there today at a given site and landscape, so that we can work out for the past what is not. But how a site is known today may not be the same as how it was known in the past, given that often time has changed both the physical properties and cultural contexts of a site. How, then, can the art, and the site, be more closely envisioned for what it was like at the time of its creation or use, if its images, rock surfaces and landscape settings, and their cultural positionings (the reasons why they were created or used in the first place), have seen the passage of time variably measured in the tens to tens of thousands of years? The problem is that often we do not even know what a site looked like when its art was made and when the site was occupied—here we use ‘occupation’ not just to mean ‘settlement’ or ‘lived in’, but to refer to any form of engagement, as in Ingold’s (2000) and Thomas’s (2008) ‘dwelling’ and ‘inhabitation’.

An important advance in global rock art research in recent years has been the ‘ontological turn’ (e.g. Moro Abadía and Porr 2021), the recognition that the world views (ontologies) of the people who made and used a particular rock art site define how the art was, and is, meaningful. But, again, to understand how the art was meaningful in the past requires its positioning in its past cultural landscape. Not all dimensions of this landscape can be reconstructed, for ‘landscapes’ are not just physical, relating to the world as it is known and experienced as much as to its physical shape and properties and the resources then available (Riemer et al. 2017). Yet the physical layout of a site and its environment can inform on that past cultural landscape nonetheless. How, for example, has a physical space changed to enhance or inhibit movement from one place to another? How has the shape and size of a wall surface transformed over time, so that at one time it was a canvas for painting, at another not? Knowing which transformations took place, how and when, may be important contexts to understand what was meaningful in the past, and what may not have been. Sometimes, especially in regions with First Nations Traditional Owners, the meaningfulness of the art or rock art site is explained through knowledge held in community histories and traditions. In the cave of Yalo, on the island of Malekula in Vanuatu, for example, Small Nambas community members know that a section of the spirit-cave is today devoid of rock art not because it wasn’t there, but because earthquakes collapsed sections of the wall, especially in 1965 (Wilson et al. 2000). The cave is a pathway and living landscape for the spirits of the dead, and the art a receptacle for the individuals who may have died but now reside within. In other areas again a powerful spirit-being, its presence visible at a low rock, guards the entry to the site and guides the newly arrived spirits to their residence in the cave. Other features of the cave walls are also knowing and prescient, such as a rock ledge in the cave, onto which community members throw pebbles to learn the gender of their next child. How people understand the features of the rock through culture gives shape to the local landscape inside the massive cave. The ontological dimensions of landscapes, particularly around landform highlights, are key to understanding sites invested in the past and the concentration of sites in geographical space.

Here we tackle the challenge of reconstructing the physical settings of past rock art sites through archaeomorphology (see below): seeing the art in its landscape setting at the time of its creation and use. While we do not focus on the ontological dimensions of any individual site or its landscape, our reconstructions are made in the understanding that they take us one step closer to the physical world of those who experienced that world. Too often in archaeology the art is recorded in all its minutiae, while the physical condition and landscape context of its creation and use are either ignored or given lip-service not so much because the researcher is not aware of the problem, but because reconstructing, and understanding, the place of the art in times past is beyond the immediate aims or abilities of the research program in question. The danger is then one of over-focusing on the image itself, at the expense of understanding, for example: (1) the broader landscape of the site and why its particular setting was chosen for occupation (e.g. Delannoy and Geneste 2020; Genuite et al. 2021a); (2) why certain zones of a site were chosen for activities that revolved around images and through which stories could be told; (3) what the accessibility to sites and decorated zones was like at the time the art was made; (4) which pathways artists and users of a site followed to get to the decorated zones; (5) how a site was architecturally altered by people at the time the art was made or used, and that would thus shed further information on the spatiality, materiality and reasons for the art (e.g. David et al. 2017; Delannoy et al. 2017); and (6) the taphonomic transformations that an art site witnessed subsequent to the art’s creation and the site’s visitation, and that would thus enable a better comprehension of the time of occupation and use. Together, an integrated archaeological and geomorphological (‘archaeomorphology’, see Delannoy et al. 2013, 2017) vision towards such questions allows for the revelation of critical spatial information that archaeology or geomorphology cannot address alone. Archaeomorphology enables the configuration of sites and landscapes to be reconstructed for the past, as context for visualising the environments in which people lived.

This chapter outlines how archaeomorphological approaches to rock art sites can shed important new light on broader landscapes and the internal settings of a site. Such a dual spatial approach brings in both the art site’s external environment and its internal site structure, at articulating spatial scales.

2 Reconstituting Site Landscapes at the Time of Occupation

Across the world, many major rock art sites and site complexes are found in settings with pronounced landscape features or relief (e.g. in Australia, the Arnhem Land plateau; in South Africa and Lesotho, the Maloti-Drakensberg Park; in the U.S.A., Chaco Canyon; in Colombia, Serranía de la Lindosa; in Baja California, Mexico, the Sierra de San Francisco). It is not unusual in such landscape settings for researchers to ask about possible connections between art sites and local features of the environment, e.g. as topographic landmarks that affected the choice of location for the art (e.g. David 2002; Gunn 1997; Monney 2012; Wilson and David 2002; for social anthropological approaches to “senses of place”, Feld and Basso 1996).

A major difficulty of such a palaeo-landscape perspective is that knowledge of a site and its surroundings is usually based on present-day landscapes. Was a site’s environment different in deep time when people engaged with the art? If so, what was the site like, what were its spatial characteristics, and what were its architectural features? Such questions are all the more important when dealing with old sites in fast-changing landscapes, for example where erosion is rapid (e.g. Foz Côa, Portugal: Aubry et al. 2012), where deposition builds up quickly (e.g. Roc-aux-Sorciers, France: Bozet and Miskovsky 2010; Genuite 2019), where speleothems accumulate on rock surfaces (e.g. Chauvet Cave: Delannoy et al. 2018; archaeological floors covered with flowstone at Bruniquel: Jaubert et al. 2016), where increasing or prolonged aridity renders previously permanent villages uninhabitable (e.g. Chaco Canyon, U.S.A.: Lekson 2006), where sea level change has altered the coastline (e.g. Murujuga, Australia: McDonald 2015), or where tectonic activity leaves fractured geologies or raised terraces above their palaeo-levels (e.g. some parts of Island Melanesia: Wilson et al. 2000). In some cases, such as among the hard quartzites of the Arnhem Land plateau and the Kimberley in northern Australia, landforms have changed little over tens of thousands of years (Delannoy et al. 2017; Genuite et al. 2021b; see also Pillans and Keith Fifield 2013), even since the beginnings of human presence around 50,000 to 65,000 years ago (Clarkson et al. 2017). But this is not the case for high to mid-latitude European and Asian sites for example—among many others across the world—that are more amenable to weathering and erosion (e.g. rock engravings, NW Spain: Pozo-Antonio et al. 2018; Daraki-Chattan rock art sites, India: Liritzis et al. 2019). In these latter cases, it is essential that robust geomorphological investigations are undertaken to reveal the physical layout of sites in their environments at the time of their use. This would also allow for the changing configuration of the landscape to be both characterised and dated. Yet while geomorphological studies can usually determine the origins and evolution of landscape features over time, a greater difficulty often lies in determining changes relating to the time of people’s presence at a site, because this usually entails finer-grained chronologies and a focus on discrete components of the landscape.

It is at the nexus of these temporal and landscape scales that the archaeomorphological study of Chauvet Cave, in the southeast Massif Central region of France, was undertaken. Chauvet Cave has been the subject of interdisciplinary research for over 20 years, since 1998. A major focus of the research has been to document the decorated panels, whose earliest rock art dates to c. 36,000 cal BP and is currently among the oldest known in Europe and across the world, as a step towards understanding why the art was made (Clottes 2001; Quiles et al. 2016; for earlier art in Africa, Europe and Southeast Asia, see e.g. Henshilwood et al. 2009; Aubert et al. 2018 respectively). But other aspects of the archaeology of Chauvet Cave have also featured prominently in the research, and are thought of as critical clues for the cave’s Upper Palaeolithic use and significance, including in relation to the art: archaeological and palaeontological remains, sometimes buried, sometimes on the floor, sometimes on the walls, along with the changing configuration of the cave itself, before, during and after human occupation, all need to be investigated so as to determine what people engaged with, and how they negotiated and created their living environments (Delannoy and Geneste 2020).

So far, most of this research has been concerned with the interior of the cave and its palaeo-entrance. More recently, archaeomorphological investigations have tackled the question of when the exterior environment attained its current form, and what it was like at the time of the cave’s occupation. The valley below the cave features the entrance of narrow gorges marked by a vast, and spectacular, natural arch (the Pont d’Arc) under which the Ardèche River flows (Fig. 10.1). If this network of gorges and landmark arch existed between 36,000 and 31,000 cal BP, when many of the paintings and stone arrangements in the cave were made, they must have to some degree, and in some ways, acted as reference points for the cave, and thus contributed to the cave’s signification for the communities in whose territories it lay. As waypoints and cultural markers the gorges and the arch fronting the cave would also have affected the choice of Chauvet Cave among the hundreds of other, often large cavities that open elsewhere in the deep valley of the Ardèche, although nowhere else, as far as we know, is any cave as extensively decorated or internally marked with artificial installations as Chauvet Cave. Addressing these notions requires researching key features of the contemporary landscape that relate to how the cave could be accessed, its spatial relations to other landmarks, and how the cave was used compared to other nearby sites. In this respect, four salient features of Chauvet Cave’s immediate surroundings are evident: the Pont d’Arc archway, the abandoned river meander at the base of the cliff (Fig. 10.1a–b), a natural ledge along the cliff that leads directly to the cave entrance (Fig. 10.1c), and the cave entrance itself.

Fig. 10.1

Chauvet Cave in its environmental setting. (a, b): The cave’s cliff-line overlooking an abandoned meander of the Ardèche River, and the Pont d’Arc archway through which the river flows today. Photo and artwork by Jean-Jacques Delannoy. (c): The long ledge in the rock that serves as a pathway at the base of the cliff and that leads to the Chauvet Cave entrance. Photos by Stéphane Jaillet

The question remains as to the morphological evolution of each feature, and how to date their origins and transformations. Although a prominent feature of the landscape, the Pont d’Arc archway cannot easily be dated, as its datable components eroded away as it evolved. Emphasis has therefore been placed on the terraces deposited by the Ardèche River, how and when they developed in articulation with the Combe d’Arc meander. The abandonment of the meander marks when the Ardèche River began to flow through the arch. Dating the last alluvial sediments to be deposited in the now-abandoned meander thus makes it possible to date the beginning of the landscape that we see today (Genuite et al. 2021a).

Three terraces of varied elevations relate to the Ardèche River’s alluvial history and to associated palaeoclimatic cycles. The highest (+30 m above the current riverbed) represents an ancient river level that would have risen above the top of the bridge of the Pont d’Arc archway. The intermediary level (+15 m) relates to a time when the Combe d’Arc meander was still an active channel of the Ardèche River; a layer of river pebbles has been identified through outcrops and subsurface Electrical Resistivity Tomography (ERT) in the now-abandoned meander. The lowest level (+8 m) is not observed inside the Combe d’Arc meander, indicating that the most recent perennial water flows along the Combe d’Arc meander are associated with the intermediary level (+15 m). That level dates to 124,000 ± 16,000 years ago (at 95% confidence), as determined by Electron Spin Resonance (ESR) ages on buried alluvial sediments: the cessation of flow along this part of the Ardèche River channel is associated with the opening of, and flow of the river through, the Pont d’Arc archway. This chronological time-line establishes that the arch has been a visually dominant feature of the Chauvet Cave landscape since c. 124,000 years ago, well before the first human entries into the cave (Fig. 10.2).

Fig. 10.2

Evolution of the Ardèche River’s gorges to the south of Chauvet Cave and relative to the opening of the Pont d’Arc archway. Upper left Upper alluvial level at +30 m. Lower left: Intermediate alluvial level at +15 m, representing the end of the severance of the Combe d’Arc meander and commencement of the opening of the Pont d’Arc archway. Right: Reconstruction of the landscape 124,000 years ago. At the time of Chauvet Cave’s first frequentations, around 36,000 cal BP, the landscape was similar, although by then the Ardèche River had stopped flowing along the Combe d’Arc meander, exclusively passing through the Pont d’Arc archway instead. Artwork by Kim Genuite

Archaeomorphology applies geomorphological methods, three-dimensional (3D) imagery and geochronological methods to archaeological questions, allowing researchers to determine that at the time of Chauvet Cave’s occupations, the broad topographic features of the surrounding landscape were similar to today’s, albeit with a different vegetation cover. This makes it possible to ask why Chauvet Cave was chosen to make its dense and extensive rock art and installations rather than the many other nearby cavities. Does the presence of nearby prominent landmarks feature as waypoints of extraordinary meaning towards a special cave in these choices?

3 Determining the Location, Shape and Size of Ancient Cave and Rock Shelter Entrances

Just as understanding the history of now-prominent landmarks is critical to understanding a site’s occupational context, so too is understanding the location and configuration of its palaeoentrance(s) at the time of occupation fundamental to understanding its visibility and accessibility. We cannot usually assume that a site’s entrance was the same in deep time as it is now, as has been shown for many sites around the world (e.g. Altamira, El Castillo, La Garma in Cantabrian Spain: Arias and Ontañón 2012; Sainz et al. 2000; Chauvet Cave, Cosquer Cave, Aldène Cave, Lascaux, Bruniquel in France: Ambert et al. 2005; Clottes 2001; Clottes et al. 2005; Jaubert et al. 2016; Rouzeau 1978; Cloggs Cave, Nawarla Gabarnmang in Australia: David et al. 2017, 2021; Delannoy et al. 2020). For some of these sites, their current entrances through sinkholes, narrow squeezes, unobstructed passages and so forth can significantly skew our perceptions of access and the contexts and spatial configurations of the archaeological features, thereby affecting also how a site is thought to have been used and socially engaged. It is therefore important to accurately determine the location and geometry of the entrances used by people in the past, as well as the condition and age of their closures or transformations. Recent research at Cloggs Cave (Australia), for example, has made it possible to reconstruct the location and configuration of the palaeo-entrance through which now-extinct Pleistocene megafauna entered the cave between 54,160 and 44,500 years ago. That entrance was significantly larger, more open and accessible, and less convoluted than the one used by people thousands of years later, the closure of the palaeo-entrance through accumulated floor sediments helping explain the hidden nature of the ritual activities that took place within the cave in those later times (David et al. 2021; Delannoy et al. 2020).

Similarly, Chauvet Cave’s palaeo-entrance was not the same during the time of frequentation as it is today. Three lines of enquiry relating to the cave’s palaeo-entrance are of particular interest: the visibility of the entry from distant vistas; the penetration of sunlight into the cave’s first chambers; and the age and cause(s) of its closure.

Chauvet Cave’s palaeo-entrance is no longer visible from the Combe d’Arc panorama (Fig. 10.1). Its obscurity was caused by the collapse of the cliff-face above the entry. Geomorphological studies coupled with ³⁶Cl cosmogenic dating of the escarpment, collapse and corresponding cliff scar have revealed three phases of rockfall, dated to 29,500, 25,000 and 21,500 years ago (Sadier et al. 2012). It is the third collapse, 21,500 years ago, that completely sealed the entrance of the cave from medium-sized (e.g. canids) to large fauna and human access, and that caused it to become invisible from the surrounding landscape.

Today, the Pleistocene entrance remains blocked by a scree cone (Fig. 10.3). The presence of this scree makes it difficult to imagine the Upper Palaeolithic entry-way: was it prominent, chaotically strewn with rock debris, large and open? Geomorphological clues on either side of the scree cone, coupled with 3D visualisation modelling of both the parts and of the whole—including traces of the ceiling’s outline, sections of visible wall, floor spaces etc.—have enabled an accurate reconstruction of the entrance (Fig. 10.3). Chauvet Cave’s Upper Palaeolithic entrance was a clear and defining element of the landscape when its artworks and installations were made and used, until 29,500 years ago (Delannoy et al. 2010). This timing corresponds to the cave’s second (Gravettian) and last phase of human frequentation (until its rediscovery by speleologists in 1994), signalling the cave entrance’s narrowing and subsequent closure and disappearance from the surrounding landscape.

Fig. 10.3

The Chauvet Cave entrance passage and reconstruction of its Upper Palaeolithic entry at the time the art was made. (a): Today, showing the scree cone from the cliff collapse that closed the entry-way. (b): The current landscape of the Combe d’Arc escarpment (the ‘Cirque d’Estre’) that led to the now-hidden cave entrance. (c): Reconstruction of the Chauvet Cave landscape 36,000 cal BP, showing the prominent cave entrance (dark area near the base of the cliff) towards the right of the image. Photo and artwork by Jean-Jacques Delannoy

The reconstruction of the Upper Palaeolithic entrance also raises further questions, particularly as to the penetration of direct and subdued sunlight into the cave. Could a sunlit cave entrance have affected the distribution of paintings in the proximal chambers and passages of the cave, or the pathways followed by people and animals? Cave Bear tracks first appear on the floor near the centre of the Salle des Bauges—the first large chamber after the entry—and, soon after, as scratch marks on the walls. It is these same walls that directed Cave Bears along passages that led to the deepest chambers.

In order to best perceive how daylight entered the cave, the solar path angles, the sun’s radiance at the winter and summer solstices, the strength of solar radiation, and the effects of albedo (reflection of incident light) from the limestone walls and floors (limestone and clay clasts) around 36,000 cal BP were all taken into account. Figure 10.4 shows the result. The pattern enables us to better understand the distribution of the rock art in the cave’s proximal chambers: all the art panels closest to the entrance are located in permanently dark areas not reached by direct or subdued sunlight. The implication is that even when the Upper Palaeolithic entrance was open and visible from a long way away, the art was not made to be seen in public view, remaining hidden from all but those who entered its darker and deeper recesses.

Fig. 10.4

Evolution of Chauvet Cave’s entrance and proximal cavities, and penetration of daylight at the start of the Salle des Bauges 36,000 cal BP, before the collapse of the overlying cliff. Plan and cross-sections by Jean-Jacques Delannoy; 3D light model by Kim Genuite

Knowledge of Chauvet Cave’s configuration at the time of its use by people is critical to understanding what the art was all about. The gathering of such knowledge has required dedicated geomorphological study supported by absolute chronometric ages and 3D modelling capable of integrating multiple spatial dimensions (e.g. the shape of the cavities, the entry of light), all informed by archaeological questions.

4 Reconstructing Past Passage-Ways in Underground Sites

So far, we have focused on the need to understand a site’s external environment and its entry-way and visibility from those surroundings at the time of its frequentation. The same logic can also be applied to another topic and scale of research: that of human passages and journeys within a site. While this question of pathways is rarely asked of most archaeological sites, it has been of considerable concern for larger cave sites with spatially variable archaeological signatures, and thus presumably variable functions (e.g. Cussac, France: Jouteau et al. 2019; Pillar Cave, Australia: Mardaga-Campbell 1986). This question is regularly asked especially in underground sites where travel is more restricted as a result of the absence of natural light and the presence of areas that are difficult to navigate (e.g. the narrowing of passages, presence of sinkholes and wells, seasonal inundations etc.) (Ambert et al. 2005; Rouzeau 1978). Such questions can relate to a range of factors, such as the distances covered underground; the by-passing, or crossing, of steep slopes or deep cavities such as sinkholes; the negotiation of obstacles such as slippery ground or subterranean bodies of water of a range of sizes; and the choice of taken paths. Was there a pattern to human movement in the past, or was it more random depending on the visitor? The answers to these questions are not trivial in understanding the archaeology and, with this, how and why people did what they did at a site in the past: a site can be visited opportunistically on a single occasion, or it can be organised to accommodate regular access to a particular zone. In all these concerns, there are many ways of walking, all social and cultural, and it is these social and cultural approaches that archaeology is interested in (see Ingold and Vergunst 2008). Again, research on such issues at Chauvet Cave is illustrative (Delannoy and Geneste 2020; Delannoy et al. 2012; Monney 2012).

Chauvet Cave is characterised by a succession of large, 30 to 50 m-wide underground spaces connected by narrow passages with, for the most part, relatively flat floors (Fig. 10.4). The more open, large and rather unrestrictive chambers allow for a relative freedom of movement within the cave. But this is a present-day perspective derived from modern electrical lighting and caving equipment. In Upper Palaeolithic times, crossing one of the large chambers with small artificial lamps or torches would not have produced as much light as when travelling along the narrow corridors (where light could reflect off the walls). Nor would it have been possible to fully illuminate underground spaces such as is possible today. At Chauvet Cave, research has thus paid particular attention to archaeological traces of the paths that people followed when in the cave. These are not abundant, but enough has been preserved to obtain a clear picture of how people travelled through the site: some human footprints in the Galerie des Croisillons, one of the terminal galleries; charcoal torch marks on the walls of low passages; upright Cave Bear bones marking the route along the Salle des Bauges. Other than these archaeological traces, further clues have been erased by subsequent water run-off, the passage of Cave Bears, or, more commonly, covered over by speleothems such as flowstone and stalagmites. The high-resolution mapping of the entire cave (at a scale of 1:50) has made it possible to reconstruct the floor as it was at the time of the cave’s Upper Palaeolithic use (Delannoy and Geneste 2020) (Fig. 10.5). Sediments on the floor were then more clayey and slippery than they are today. This is noticeable along some steep passages, where multiple Cave Bear slip-marks are evident (Fig. 10.6). This reconstruction of floor conditions at the time of human frequentation reveals that corridors and passage-ways that are now covered by speleothems were then very wet or otherwise impractical for human passage, causing detours not evident from today’s cave conditions. The study of such obstacles, coupled with ceiling heights, makes it possible to find ancient pathways whose use is confirmed by the presence of charcoal torch marks on walls and low ceilings.

Fig. 10.5

Floor conditions at Chauvet Cave today and 36,000 cal BP. Cartography by Jean-Jacques Delannoy. (a): The Galerie du Cierge today. (b): The Galerie du Cierge 36,000 years ago. Photo by Stéphane Jaillet, reconstruction by Yago Delannoy

Fig. 10.6

Anthropic installations in the Galerie du Cierge. Photo by Jean-Michel Geneste, plan and cross-section by Stéphane Jaillet, concept and graphic by Jean-Jacques Delannoy

The high-resolution archaeomorphological mapping of the cave has made it possible to identify a number of otherwise ambiguous anthropic structures on the floor, including some erected to navigate the pitch-black chambers and corridors. One example is a 50 cm-long × 35 cm-wide × 15 cm-thick rock slab artificially placed against the bank of a 60 cm-deep depression in the cave floor (Fig. 10.6). The slab was manually extracted from a suspended stalagmitic floor and transported over 30 m to its current location. Petrographic analysis and digital refitting of the slab onto its originating scar in the stalagmitic floor through the 3D model confirms the origin of this Upper Palaeolithic step as an artificial construction. It is of interest to note that the axis of the pathway that leads to the step is slightly depressed through compaction as a result of the repeated passage of both people and Cave Bears. This part of the cave also contains other human constructions: sets of slabs from the same suspended stalagmitic floor as the step noted above; an alignment of blocks whose joints were sealed by clay fill, beyond which water ponded as it flowed from one of the few sources of water in the cave (Fig. 10.5a). It is nevertheless difficult to infer with certainty the intentions of the people who built the damming wall, because a result of its construction, intentional or not, is the diversion of part of the waterflow to the NNW side of the chamber, rendering less humid the SSE side where people could then more easily walk.

The combined archaeological and geomorphological study of the floor of the cave around specific research questions—here concerning ancient pathways taken by the site’s occupants and the nature and layout of artificial installations—makes it possible to consider the art in the cave in relation to routes of travel and how people arranged space in ways that enabled them to negotiate constraints such as large spaces unable to be adequately lit, obstacles such as water ponds and slope failures, and cross-roads between chambers, and thereby to structure social activities within the cave. Such an approach to the study of the spatiality of rock art sites has also been employed elsewhere, and merits broader application globally (David et al. 2017; Delannoy et al. 2017; Jaillet et al. 2018; Monney and Jaillet 2019).

5 Reconstructing Rock Walls and Determining the Age of Their Paintings

Having begun with a wider vision, from the external environment increasingly focusing on the interior of Chauvet Cave, we now apply archaeomorphology more specifically to decorated walls. For this we shift to another site: JASRN-124 site 3 on the Arnhem Land plateau, Northern Territory, Australia (Barker et al. 2017). Unlike Chauvet Cave, JASRN-124 site 3 is an open-faced, mushroom-shaped rock shelter with its own analytical challenges: here there are numerous paintings on the vertical walls, the most prominent of which, and the reason as to why the site was studied, is a large painted bird thought by its recorders to be a representation of the Pleistocene giant bird, Genyornis newtoni, thought to have gone extinct more than c. 40,000 years ago (Gunn et al. 2011). The rock panel that houses the large bird painting was formed when the overhang collapsed, creating a large, vertical scar on the rock face. Determining precisely when the overhang collapsed, or how old the paintings are, through more standard approaches such as cosmogenic dating of the now-collapsed palaeo-overhang that once covered the now-painted wall, or excavating below the fallen boulders so as to retrieve buried fragments of use-worn ochre and other archaeological traces of occupation, was problematic as neither the boulders nor the rock face could be damaged due to their cultural significance, and excavation below the painted wall could not be undertaken due to the presence of large blocks carpeting the entire floor below the remnant overhang. The major question at this site was to determine when the overhang had collapsed, to work out whether the painting, which could only have been painted after the present rock face had formed, was compatible in age with the presence of Genyornis on the continent (for palaeontological details of Genyornis, see e.g. Murray and Vickers-Rich 2004). The three analytical obstacles were an inability to: (1) directly date the mineral pigments; (2) date their extremely thin underlying and overlying crusts (Hoffmann et al. 2018; Pike et al. 2012; Plagnes et al. 2003; Slimak et al. 2018); and (3) excavate below the boulders to find and stratigraphically date paint drops or other contextual archaeological deposits (David et al. 2017, 2019). However, another approach enabled the age of the overhang collapse, and with this the age of the present and now-painted surface, to be accurately determined.

The ability of archaeomorphology to resolve the archaeological questions posed of JASRN-124 site 3 revolves around its ability to provide information not so much on the archaeological deposits, as on the physical evolution of the site (e.g. Barker et al. 2017; David et al. 2017, 2019). Two archaeological excavations undertaken at JASRN-124 site 3 are particularly pertinent to this question: Squares B + E and Square D (Fig. 10.7). Both excavations were set below overhangs of the same geological strata whose collapse resulted in the formation of the rock scar that was subsequently painted with the large bird thought to be of Genyornis (Panel A) (Fig. 10.7). Petrographic and lithological characterisation (thickness, granulometry etc.) of the eight major quartzite strata (D0 at the base to D7 at the top) of the rock outcrop enabled each to be differentiated, and to thus determine from which part of the outcrop each fallen block came. Coupled with the buried soft sediment layers revealed by the archaeological excavations, individual rockfall events could be identified and dated by stratigraphic association. The accelerator mass spectrometry (AMS) radiocarbon dating of individual pieces of charcoal buried under and over the partially or completely buried collapsed blocks made it possible to fix in time a sequence of overhang collapses, and thus to determine when the extant walls were created. The paintings on those walls must necessarily be of the same age or younger than the rock collapses that created their rock surfaces. Two major rockfall events were thus dated. The first took place around 20,407–20,905 cal BP, the second around 13,739–13,976 cal BP. The second of these overhang collapses is the critical one, for it is then that the extant rock surfaces were created and subsequently painted (cf. Panels C and D of the western face of the site), including the rock wall with Panel A that contains the large bird painting (Fig. 10.7). This age determination allows us to conclude that rock art Panels C and D are younger than 13,739–13,976 cal BP and, therefore, not as old as what had been expressed by some researchers upon the discovery of what appeared to be a painting of the extinct bird Genyornis. Indeed, the large bird painting on Panel A is positioned adjacent to another bird and macropod paintings and to a particular kind of red hand stencil known as ‘3MF’ (3 middle fingers held together, the thumb and little finger splayed). 3MF hand stencils are typically associated with Dynamic Figures across much of the Arnhem Land plateau, a particular style of anthropomorph long thought to date to the terminal Pleistocene or very early Holocene. The post-13,739–13,976 cal BP age for the entire panel is consistent with this. Also consistent is a comparable age for zoomorphs on Panel C painted in similar style to the two bird paintings on Panel A. The determination through archaeomorphology of a maximum age of 13,739–13,976 cal BP for all these paintings and hand stencil, both on Panel A and Panel C, is coherent and consistent with the previously theorised approximate age for this painting style, estimated to have been extent in Arnhem Land c. 10,000 to 12,000 years ago (Taçon and Brockwell 1995) (for further details and images of the JASRN-124 site 3 paintings and 3MF hand stencil, see Barker et al. 2017: 491–92).

Fig. 10.7

JSARN-124 site 3 (Arnhem Land, Australia) and location of rock art Panels A and C. The six cells in the lower part of the figure show the geomorphological evolution of the southern edge of the rock shelter, as based from 3D mapping of the rock stack and the buried rockfall revealed by the archaeological excavations in Squares B + E. (After Barker et al. 2017)

The results of the excavation against the west wall of the site cannot be directly applied to the northern side where the large bird painting of Panel A occurs, because the presence of the large blocks at the foot of the northern wall do not allow the buried horizons to be spatially traced to the critical northern areas. To get around this problem, the morphology of the extant exposed walls and boulders, and that of the blocks revealed by the archaeological excavations, were incorporated in a 3D model of the site. On the digital model, special attention was paid to the collapsed strata that resulted in the creation of the extant wall subsequently painted with the large bird on the northern side of the site (Panel A), and how they interlocked with the corresponding collapsed strata on the western side of the site where both geomorphological and chronological details were available. This work of connecting the morphogenesis of one part of the site (the west and northwest) with another (the north) enabled the broader palaeo-morphology of the rock outcrop to be determined, and the age determination of Panel A to be constrained (Figs. 10.8 and 10.9). The paintings in Panel A cannot be older than 13,739–13,976 cal BP, and cannot therefore be of the long-extinct Genyornis newtoni.

Fig. 10.8

JASRN-124 site 3. Top left: the northern and western sides of the site as it is today. Middle Left: Reconstruction of the collapsed overhangs on the northern and western sides as they were prior to the paintings. Top right: Excavation Squares B + E (foreground), with Square D in the northwestern corner and closer to the large bird painting, hidden from view by surface boulders. Bottom right: rock art Panel C above Squares B + E. Photos and graphics by Jean-Jacques Delannoy

Fig. 10.9

Synoptic chart of archaeomorphological investigations at JASRN-124 site 3

Finally, we ask why on the northern side of the site, the collapsed overhang stands high above the present ground surface, whereas it is hardly visible above ground on the western side where it is largely buried under sand. This disparity has two causes: (1) major differences in the amount of rockfall, including the mass of individual boulders; and (2) unequal degrees of aeolian sedimentation on the two sides of the rock outcrop. On the western side, rock strata D3–D4 and D5a have fallen: part is buried (as revealed by the excavations in Squares B + E and D), and the rest is flush with the present ground surface. On the northern side, the entire overhang, consisting of rock strata D4–D5–D6–D7, totalling close to 3 m thick, collapsed. Aeolian sedimentation subsequently covered large parts of the floor of the site. Given the direction of prevailing winds, the sand settled against the leeward face: the western side of the rock outcrop.

The example of JASRN-124 site 3 is particularly instructive for the ability of archaeomorphology to resolve specific archaeological questions concerning the history of a site’s morphology, in this case concerning the age and, with this, the taxonomic attribution of a large bird painting originally thought by some researchers to be the extinct megafauna Genyornis newtoni. The articulation of archaeological with geomorphological methods and evidence to answer a united question enabled both a reconstruction of the site to be made for the terminal Pleistocene, and a better understanding of the progressive development of the rock outcrops’ painted surfaces.

6 Conclusion

The archaeology of rock art is in a strange position: the art itself is usually what captures the researcher’s imagination, so much so that research on the motifs’ age (e.g. Finch et al. 2021), manufacturing techniques and chaînes opératoires (e.g. Vergara and Troncoso 2015), paint formulas (e.g. Chalmin and Huntley 2018), style or design conventions (e.g. Taçon et al. 2020), spatial patterns within and between panels (e.g. Gunn 2018), symbolism (e.g. Solomon 2018) and other such concerns with the art often overwhelm the research, at the expense of critical details of the site and its broader landscape setting—aspects of central relevance for understanding how people engaged with the art. Those details relating to the art’s emplacement offer considerable information for better understanding the art’s social and environmental contexts (and therefore they help better understand the art itself). It is the place—the site and its surrounding landscape—that positions the art in a social and ontological geography. That is, place positions human activities and how people organise themselves socially and culturally in meaningful territories. An archaeology of place as it relates to rock art thus requires more than an ‘archaeology of art’ narrowly defined. Archaeomorphology is well placed to address this aspect of the past.

The examples presented in this chapter illustrate how rock art can be explicitly considered in relation to its contemporary landscape through evidence at multiple, articulating spatial scales. Once on the wall, the art establishes a new setting for future perceptions and social activities, including how a rock surface, site and broader setting will subsequently be engaged. It is thus important to determine not only the nature of artworks at specific points in time, but also of the rock surfaces, passage-ways, site entrances, and landmarks, for each falls into the visual and existential purview of ‘the rock art site’ and thus affected how and why people engaged with the site and its art. Let us not forget in this context that people do not, and did not, simply passively place art on pre-existing surfaces. Rather, sites and surfaces formed active components of the lived, social and ontologically and experientially meaningful world. In engaging with already-meaningful places, people both actively modified both rock surfaces (e.g. by preparing rock walls by scraping before painting (Delannoy and Geneste 2020)), site configurations (e.g. by removing rock pillars and ceiling rock layers (Delannoy et al. 2017)) and their environments (e.g. by positioning standing stones as place-markers visible over considerable distances (e.g. Gunn et al. 2012)). Archaeomorphology, along with other landscape disciplines such as palaeo-ecology, are well placed to contribute meaningfully to these aims, by understanding not just how things were, but how they became and continued to transform through time, so as to improve understandings of the archaeological (social) and environmental contexts of the art as positioned in places we wish to better understand.