Art as actual object and virtual object

In The Archaeology of Art (2018), the archaeologists Andrew Jones and Andrew Cochrane put forward a new ontology of art for prehistoric archaeology. They argue against a representational treatment of art because it reduces archaeological artefacts to a carrier of meaning that can be extracted from it through the act of interpretation (Jones and Cochrane Reference Jones and Cochrane2018, 15; Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 3; Jones Reference Jones2017, 170–2). When we treat prehistoric art as a symbol, we neglect its material dimension, the fact that it has been shaped through the working with materials and that this process of making forms an integral aspect of it (Jones and Cochrane Reference Jones and Cochrane2018, 19–43; Alberti et al. Reference Alberti, Jones and Pollard2013, 22–23). Jones and Cochrane propose an ontology of art as process (Jones and Cochrane Reference Jones and Cochrane2018, 137–72; Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 2; Jones Reference Jones2017, 173), which means that the process of making, the engagement of the work by a viewer, etc., are all integral aspects of the art object. The reduction of prehistoric art to the motif or image as such therefore misconceives that the meaning of prehistoric art might have been situated in its process-character rather than in any representational function it might have fulfilled (Jones and Cochrane Reference Jones and Cochrane2018, 186–7; Alberti et al. Reference Alberti, Jones and Pollard2013; Back Danielsson et al. Reference Back Danielsson, Fahlander and Sjöstrand2012). This leads to a new conception of rock art in which rock art is not a symbol of a certain religious conception of the world or the visualization of a certain narrative, but in which rock art was meaningful as a practice in which different actors came together (Jones Reference Jones2006, 215; Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 4, 6, 13; Jones Reference Jones2017, 172). A good example is how the cup-and-ring art of Neolithic Britain is situated on rock surfaces that show a deliberate concern with geological features (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 19–35; Bradley Reference Bradley2023, 14). This indicates that it was not the cup-and-ring motifs as such that were important, but the relation between the motifs and the geology of the rock surface (Jones Reference Jones2006, 217–18). One could also see this as an indication that it was the establishment of a relation with the rock or place that must have been significant. The meaning of cup-and-ring art is then situated in the act of making, in the material interaction between carver and rock as the establishment of a relation with matter or place. The ontology of cup-and-ring art then becomes transformed from the condition of motif/symbol to the condition of process (Jones Reference Jones2006, 222–3; Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 177–80).

I want to argue that this new ontological conception of the art object, as proposed by Jones and Cochrane, in fact misconceives the ontological dimension of rock art. What defines a work of art is that it is not only an actual object (meaning that it has an empirical dimension, for example that it is made out of stone), but is also a virtual object (Deleuze 1968/Reference Deleuze2014, 273; Adorno 1970/Reference Adorno2013, 119–20, 132–3). By virtual object I mean that a work of art makes visible something that is invisible (Deleuze 1968/Reference Deleuze2014, 71; Deleuze and Guattari 1991/Reference Deleuze and Guattari1994, 193–4). A work of art makes something appear that is distinct from its empirical dimension even though it is through this empirical dimension that art makes visible something that is non-empirical. This ‘something’ can be called the image-character of art or the virtuality of art. The main mistake in the theorization of art is that one often does not distinguish between the ontology of the image and the ontology of the art object. An image is defined by immateriality, it is not actual, it does not have any actual shape of its own (although it has a form, which is immaterial) and therefore has no appearance in the actual world. What happens in art is that the immateriality of the image becomes situated within an actual object so that the image can appear (Husserl Reference Husserl2006, 186–91). The appearance of the image is not equivalent to representation. Representation reduces the ontology of the image to actual appearance, in the sense that art makes, for example, an animal appear even though no actual animal is present. The image-appearance is then actual appearance without the actual object to which it refers being present. This neglects that the ontological dimension of the image is non-actual; its ontological dimension is that of virtuality. The concept of the virtual (and the distinction between virtual and actual) is based on the philosophy of Gilles Deleuze (Deleuze 1968/Reference Deleuze2014, 272).

Deleuze’s philosophy has become inspirational for prehistoric archaeologists, and it has been adopted in the interpretation of prehistoric art.Footnote ³ Jones and Cochrane for example, rely on Deleuzian philosophy in their proposition of the prehistoric image as process (Jones and Cochrane Reference Jones and Cochrane2018, 25–26). However, this interpretation of Deleuzian philosophy takes insufficient account of the distinction between virtual and actual. The concept of the virtual defines a condition in which no actual form exists, in which there is no distinction between determined forms (Deleuze and Guattari 1987/Reference Deleuze and Guattari2013, 15). The virtual is the realm of Ideas, which do not have a determined form, but are differentiated from one another through speed and slowness, movement and rest (Deleuze 1968/Reference Deleuze2014, 269–70). It means that the virtual consists of attachments/relations that differentiate Ideas from one another without any clear boundary or actual form being established. The virtual realm of Ideas is the realm of unconscious aspirations and desires (Deleuze 1968/Reference Deleuze2014, 130–1, 136–9). The virtual side of an entity is what defines its unconscious desires, its phantasies or dreams, which by definition are not actual. This explains why the virtual is equivalent to the condition of the image: it is immaterial without being shapeless (as an image is not formless even though it has no actual form: Deleuze 1968/Reference Deleuze2014, 240–1, 272). In order to understand the operation and appearance of art, one needs to acknowledge that art consists of two sides: an actual side (the materials that were used to make the artwork) and a virtual side (which is the image that appears). The ontology of a work of art is based on a movement of tension between its virtual side and its actual side. Virtual and actual are heterogeneous, which means that they can never be reconciled with one another (Deleuze 1968/Reference Deleuze2014, 366). Every artwork therefore has an inner dynamic and is defined by tension, even though we might be looking at a static image (Adorno 1970/Reference Adorno2013, 110–11).

This ontological conception of the artwork, based on Deleuzian philosophy, is different from the ontology of the artwork put forward by Jones and Cochrane. One cannot reduce rock art to a process of working with materials because this only takes into account the actual dimension of rock art, neglecting its virtual dimension. It is precisely because Jones and Cochrane are concerned with a critique on art as symbol or representation that they put their emphasis on materials and the engagement with them (Jones and Cochrane Reference Jones and Cochrane2018, 19–43; Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 10; Jones Reference Jones2017, 171). The concept of matter/material that Jones and Cochrane adopt is indeed not a static conception of matter in which matter has a clearly defined actual form, but matter as a process of becoming (Jones and Cochrane Reference Jones and Cochrane2018, 22–30; Alberti et al. Reference Alberti, Jones and Pollard2013, 22–23; Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 5). However, the ascription of a process-character to actual reality/matter neglects that there is an ontological distinction between the realm of virtuality and the realm of actuality. When matter operates as a continuous process of becoming, it simply takes up the shapeless condition of the virtual, which means that the distinction between virtual and actual has been eliminated (Govaerts Reference Govaerts2022). This makes it impossible to understand that the ontology of art is based on the tension between virtual and actual and the complex reversal between the two. Although Jones and Cochrane are correct that one cannot reduce art to representation/meaning, one also cannot reduce art to its material dimension.

If the ontology of the art object cannot be reduced to its material dimension, one also cannot reduce the experience of art to empirical sensation.Footnote ⁴ The experience of the senses remains bound to the actual world, what we can actually see, feel, hear, touch, etc. One cannot see the virtual object or image, because it is immaterial. Although the realm of the virtual is defined by the condition of the imaginary/phantasm, and the immaterial/invisible, one nevertheless cannot reduce it to a mere phantasy experience because it has an ontology of its own, which is part of the real. The virtual and actual are the two halves of which the real consists, and their relation is non-hierarchical (Deleuze 1969/Reference Deleuze1990, 22–28, 39–41), which means that the actual world is not more real than the virtual world. It is through art that the virtual can appear in actual reality (Deleuze 1968/Reference Deleuze2014, 71). However, the experience of art can easily be reduced to a sensible experience, as after all, one is not looking at something invisible, but at something that is made out of very concrete materials (paint, stone, etc.). If art is based on appearance, why is the appearance of art not a sensible appearance? One must make a distinction in this case between an aesthetic experience and sensible experience. A sensible experience is limited to actuality, whereas an aesthetic experience is based on the experience of the invisible within actual form. Within aesthetic experience, one becomes aware of something that goes beyond the sensible dimension of nature (Kant 1790/Reference Kant2007, 175, 212). Traditionally, this has been understood as the feeling of the purposiveness of nature, the feeling that the world is designed by a superhuman being (God) (this is aesthetic experience as defined by Kant 1790/Reference Kant2007, 330). Within the context of prehistory, this Kantian concept of aesthetic experience is of course not immediately applicable. Nevertheless, this does not mean that one should replace aesthetic experience by sensible experience. Rock art is based on a dynamic relation between virtual and actual, which requires a conception of experience that is not limited to actual reality. The virtual does not appear here as a metaphysical purposiveness, as a supernatural dimension of the real, but as the other half of the real, as the realm of the imaginary.

Modernist abstraction versus prehistoric abstraction

The concept of experience that Deleuze puts forward takes the form of the concept of affect (Deleuze and Guattari 1991/Reference Deleuze and Guattari1994, 164). This concept has been adopted within prehistoric archaeology as a non-representational conception of experience in which a subject does not interpret the real, but stands in relational continuity with the world (Fowler Reference Fowler2013, 35; Hamilakis Reference Hamilakis2013, 124–25; Jones and Cochrane Reference Jones and Cochrane2018, 25–27; Harris Reference Harris2021, 59). Affect is a type of experience that overcomes the distinction between subject and object, subject and world, because it is a creative type of experience that operates as a form of becoming/process. This means that experience is not an experience of a pregiven world, but is constitutive of the subject and the world of the subject. By defining the experience of prehistoric art as affect, Jones and Cochrane draw attention to the fact that prehistoric art was a relational practice in which human, matter, motif and landscape stood in a dynamic relation to one another (Jones and Cochrane Reference Jones and Cochrane2018, 184–5). However, one has to take into account that Deleuze’s concept of affect cannot be limited to the encounter with sensibility. Experience as affect means that one encounters something that one cannot take into account by one’s senses, that violates common sense, that poses a challenge to one’s experience and faculties (Deleuze 1968/Reference Deleuze2014, 183–4). What is encountered is not the sensibility of the actual world, but the forces and formlessness of the virtual (Deleuze 1968/Reference Deleuze2014, 71). This type of experience is what defines the experience of modern/contemporary art (Deleuze 1968/Reference Deleuze2014, 71, 310; Deleuze 1981/Reference Deleuze2005, 25–26; Deleuze and Guattari 1991/Reference Deleuze and Guattari1994, 193–4). Modern art poses a challenge to the faculties of the viewer by confronting them with something that is not part of actual reality, that they cannot recognize. Modernist abstraction is a good example because it resists recognition (as it has no figurative forms) and also resists the sensibilities of the actual world because it puts in front of the viewer shapes and colours that never occur together in such a combination in actuality. What is important for my argument is that modern art violates the condition of image-appearance that we see in prehistoric rock art. Modernist abstraction and prehistoric abstraction do not have the same ontology (compare Figs. 1 and 2). In this sense it is important to acknowledge that Deleuze developed his concept of affect in relation to the encounter with modern art (Deleuze and Guattari 1991/Reference Deleuze and Guattari1994, 164).

Figure 1. Yves Klein, Blue Monochrome, IKB 191 (1962). Source: Wikimedia Commons, file IKB 191.jpg. The work is ineligible for copyright and in the public domain.

Figure 2. Detail of cup-and-ring art from the rock art panel at Achnabreck, Kilmartin, Scotland. Source: iStock.com/milehightraveler, photo ID 1176427811.

What happens in Modernist abstraction is that the ontology of image-appearance is violated (Deleuze 1968/Reference Deleuze2014, 10, 28, 183–4; Deleuze 1969/Reference Deleuze1990, 63). We are no longer dealing with a relation of tension between virtual and actual (which is what occurs in rock art), but with a conflation between virtuality and actuality itself (Govaerts Reference Govaertsaccepted a). Modernist art is an actual form that has itself taken up the condition of the virtual. Deleuze speaks about the material itself passing into sensation: ‘(…) it is no longer sensation that is realized in the material but the material that passes into sensation.’ (Deleuze and Guattari 1991/Reference Deleuze and Guattari1994, 193).

This means that actual/material form becomes itself virtualized, which is not the same as the appearance of the virtual (as image) in actual form. Within modern art, the abstraction that appears is not an image-form like a motif, but a shapeless form (like in monochrome painting: see Fig. 1) or the act of painting itself (as in Pollock’s paintings). We are dealing with two very different types of abstractions. In the case of rock art, abstraction operates as a virtual–actual relation in which the virtual appears (as image) in actual form (see Fig. 2), while in the case of Modernist abstraction, the distinction between virtual and actual has disappeared. The correct term to describe this condition is the simulacrum. Experience as affect does not operate as the establishment of a relation with the world, but as a complete shattering of the actual world (Deleuze 1981/Reference Deleuze2005, 25–26; Deleuze and Guattari 1991/Reference Deleuze and Guattari1994, 193–4). The shattering/destruction of actual reality is what occurs in modernity. Modernist art wanted to reveal the dimension of the virtual as such (completely leaving actual forms and common sense behind), which is why Modernist abstraction is no longer an image-appearance of the virtual within actual form (Govaerts Reference Govaertsaccepted a). This indicates why Deleuze’s concept of affect cannot be understood as a sensible experience. It also cannot be understood as an aesthetic experience because aesthetic experience is based on a relation between virtual and actual in which the virtual appears (as image). This means that, whereas we commonly associate aesthetic experience and art with Western modernity, it is in fact modern art that has made aesthetic experience and image-appearance impossible (transforming it into a simulacrum) (Baudrillard Reference Baudrillard1994; Reference Baudrillard2005; Govaerts accepted Reference Govaertsa). If the true ontological condition of art is based on this dynamic and non-hierarchical relation between virtual and actual, then it is to prehistoric societies and not to the modern West that we need to look in order to understand the ontology of art and aesthetic experience (Govaerts accepted Reference Govaertsb).

Cup-and-ring rock art as actual object and virtual object

In order to define the experience of rock art as an aesthetic experience based on the non-hierarchical relation between virtual and actual, I will now look at the cup-and-ring rock art tradition of Neolithic Britain. Cup-and-ring rock art is a type of rock art that was executed in the open landscape on flat horizontal rock surfaces, which originated in the Middle/Late Neolithic period and continued to be used into the Early Bronze Age (Bradley Reference Bradley2023, 1). The work of Andrew Jones and colleagues argued that the cup-and-ring art of the Kilmartin landscape in Scotland was not made by a nomadic population that moved through the landscape, but by a sedentary community of farmers who lived in the Kilmartin landscape (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 245). This is why the rock art was not meant to be viewed from a distance by people who travelled through the landscape, but was executed at places that were meaningful and known by the local community (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 246). I want to argue that the ontology of cup-and-ring art must be understood as consisting of both actual object and virtual object. Cup-and-ring art as actual object refers to the actual properties of the rocks and motifs: the form they have, their material properties, etc. Cup-and-ring art as virtual object means that it overcomes the boundaries, oppositions, etc., of its empirical form and makes possible what for its actual side is impossible. This is what the virtual object does: it has the condition of a dream-state, something that empirical reality could aspire to, but can never achieve. One could interpret this in the sense that cup-and-ring art makes it possible to bring together disparate elements of the empirical world that are opposites. This is visible, for example, in the relation between rock and water during the British Neolithic.

In a recent paper, Richard Bradley analysed how cup-and-ring art is very often situated at rivers, landing places for boats and at viewpoints over the sea (Bradley Reference Bradley2023, 10–11, 17). Also, the work of Aaron Watson found evidence that certain rocks at the estate of Ben Lawers in the southern Highlands had unusual properties in the sense that the act of carving must have produced remarkable quantities of mica dust that sparkled in the sunlight and which would have given a silvery lustre to the rock surface (Was and Watson Reference Was, Watson, Chittock and Valdez-Tullett2016, 158–60). This same sparkling quality is also manifested in quartz. Concentrations of pieces of quartz have been found both around decorated rock surfaces and as deposits within the cracks/fissures of the rock surface at Kilmartin (Was and Watson Reference Was, Watson, Chittock and Valdez-Tullett2016, 157; Bradley Reference Bradley2023, 13). The sparkling quality of the rock made it possible for a parallel to be drawn between the reflective surface of water and the reflective surface of the rock, which was observed by Aaron Watson at the site of Allt Coire Phadairlidh, where a rock art panel commands a direct view of Loch Tay below (Was and Watson Reference Was, Watson, Chittock and Valdez-Tullett2016, 158–61; Bradley Reference Bradley2023, 14). At Achnabreck (see Fig. 3), the rock art is situated in a position which commands a view of the island of Arran 45 km away (Bradley Reference Bradley2023, 14). The transformation of an actual object into a simultaneous virtual object overcomes the empirical oppositions between materials. They become connected in a rhizomatic way in which no oppositions exist. However, because we are not dealing with an equation of virtual and actual, the oppositions between stone and water are not overcome completely but only exist as constellation on the same plateau. This is also why we cannot reduce the experience of rock art to empirical sensation. The idea that light, sound, touch played a role in the experience and execution of cup-and-ring art (Was and Watson Reference Was, Watson, Chittock and Valdez-Tullett2016, 169–70) should be understood as a reversal of the senses, in which the virtual takes up actual form and actual form becomes rhizomatic. This means that the experience of light is never just about light but acts as a reversal in which oppositions are reconciled. This could be related to the example given above in which the reflective surface of both the water and the rock through sunlight draws a continuity between the two. We are here not dealing with the senses, but with an unhinging of the senses and faculties (Deleuze 1968/Reference Deleuze2014, 185) so that perceptions and affections do not experience actual forms in empirical reality but perceive them in terms of rhizomatic connections in which oppositions can be reconciled. This interpretation of the experience of cup-and-ring art as being an aesthetic experience (rather than sensible experience), and of cup-and-ring art being simultaneously an actual object and virtual object, means that we are dealing with the autonomous logic and ontology of art and the aesthetic.

Figure 3. Cup-and-ring rock art panel at Achnabreck, Kilmartin, Scotland. Source: iStock.com/milehightraveler, photo ID 1176427937.

Defining rock art as assemblage based upon the relation between virtual and actual

One of the main concerns in any study of rock art is the relation between rock art and the place/landscape in which it was executed (Jones Reference Jones2006; Reference Jones2017, 175–6). Rock art cannot be reduced to the painted or engraved motifs because the surface on which it was executed and the place in which it was situated form an integral aspect of rock art itself. If rock art must be understood in relation to its place, this opens up the question of whether place operated as a context for rock art or whether it was more than a context and was included within the logic of the rock art itself (being part of the picture rather than being a context for the picture). In recent archaeological theory, we see that the concept of landscape as context has been critiqued (Jones Reference Jones2006, 211–12; Alberti et al. Reference Alberti, Jones and Pollard2013, 27–30). Landscape as context neutralizes the landscape, transforming it into a background for cultural execution (the execution of rock art) rather than being an active actor in its own right (Jones Reference Jones2006, 214). This is why landscape as context has been replaced by the concept of assemblage (Alberti et al. Reference Alberti, Jones and Pollard2013, 30). The concept of assemblage comes from Deleuzian philosophy and means a non-hierarchical form of attachment in which entities do not have a pre-given form or identity but shape themselves through the attachments in which they are involved, attachments that change continuously (Deleuze and Guattari 1987/Reference Deleuze and Guattari2013, 24). By defining the landscape as assemblage, it no longer operates as a background for cultural action but operates as a constellation of actors in movement (Jones Reference Jones2017, 87). There is then no such thing as a background or context anymore, only a dynamic interaction between various actors that do not form a homogeneous ground and whose interaction is ongoing and therefore cannot be stabilized. A good example of rock art as assemblage (Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 17) is put forward in Richard Bradley’s discussion of Bronze Age Scandinavian rock art. He argues that it was through touch by running water, or the appearance of the rock art under certain light conditions, that rock art started to operate and became meaningful:

We are then not dealing with a landscape context for rock art, but with the inclusion of landscape features as actors that formed an integral part of the operation and meaning of the rock art itself (Jones and Diaz-Guardamino Reference Jones, Diaz-Guardamino, David and McNiven2017, 9, 11). Because the running water and light conditions only manifest themselves at certain times of the day or year, they cannot be understood as landscape context, but as elements within a changing assemblage of actors. This also indicates that rock art might be less about the rock art motifs depicted than about acting as a principle of connection between different actors or places (Jones Reference Jones2006, 220; Jones Reference Jones2017, 92). In my discussion of cup-and-ring art, I explained how it operated as a principle of synthesis between heterogeneous materials (stone and water) that included the disparate actors that make up the landscape into one plateau.

However, what needs to be understood is that there is a difference between rock art as one actor among others that makes up an assemblage and rock art as principle of connector that holds the various actors of the assemblage together. Although there might at first not seem to be a distinction between the two, they define two different ontological conceptions of world. In the former case (rock art as one actor within an assemblage), the world operates as an assemblage of actors that exist in constant interaction with one another without the possibility of sharing a plateau or world. In the latter case, rock art operated as a principle of connection that created a shared plateau in which the different actors of assemblage could be situated. It is then through rock art that all entities become included within a shared world, a shared space-time. This plateau-constitution must not be understood as a homogeneous space, or as the elimination of difference between the actors of assemblage, but as the ‘plane’ of a picture in which the virtual and actual side of the real are integrated and in which the invisible can appear. We are dealing with a relation in which the virtual becomes situated on the side of actual reality, which allows the empirical differences between actors to be included within a shared world, a shared logic. The landscape is then neither a background nor an assemblage of actors, but a plane/plateau in which actors are integrated within a shared world through the autonomous logic of the aesthetic/picture.

The cup-and-ring rock art at Kilmartin, Scotland must not be understood as one element within a wider landscape assemblage, but as a principle of connection/relation that operated as a virtualization of the actual landscape in which disparate actual places became integrated into one plateau. This means that we are dealing with a simultaneous affirmation and overcoming of site-specificity. Within the landscape at Kilmartin, we see that cup-and-ring art was executed alongside the routes and entrances into the Kilmartin landscape, of which the Achnabreck panel is a very good example (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 227–8). The panels at entrances tend to be the most complex because it is here that different communities could encounter one another and where the rights of movement within different territories had to be established (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 243). Apart from entrances into the Kilmartin region, cup-and-ring art is situated at significant places in the landscape on lower hill slopes, on upper hill slopes and in upland areas (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 236). What defines these places as significant is that they often overlook the coast/sea or are positioned near rivers or landing places for boats (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 228, 232; Bradley Reference Bradley2023, 10–11, 17). However, although cup-and-ring art was clearly connected to the site-specificity of place, I argue that it also works against the site-specificity of its location. This simultaneity of site-specificity and non-site-specificity is necessary in order to virtualize the landscape and to transform it from a constellation of disparities into one plateau. It is only through the unification of disparate elements of the landscape into one plateau that it could take up the condition of an ancestral homeland for the local community (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 247).

Jones and colleagues made in this sense the important observation that cup-and-ring art was not a rock art tradition based on the accumulation and erasure of motifs through superimposition, but a rock art tradition in which the motifs endured (Jones and Diaz-Guardamino Reference Jones and Diaz-Guardamino2019, 179). When new motifs were executed, the older composition was respected and new motifs had to integrate within the composition that was already present. This indicates a concern with continuity between past, present and future, in which motifs were considered to be part of an ancestral tradition that had to be respected (Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 245–7). Interesting in this regard is the parallel between the cup-and-ring motifs and the natural features of the rock surface. This parallel indicates that the natural features of the rock could have been regarded as an older rock art tradition (that in fact never existed) to which cup-and-ring art related (Bradley Reference Bradley1997, 153; Jones et al. Reference Jones, Freedman, O’Connor, Lamdin-Whymark, Tipping and Watson2011, 33). The concern with ancestry goes beyond site-specificity because it invents something that did not exist. The ancestral tradition was an ‘imagined tradition’ and the ancestral landscape of Kilmartin an ‘imagined landscape’ because there was no ancestral tradition on which the Neolithic communities could have relied (Bradley Reference Bradley2007, 87). This idea of ‘imagination’ is important because it is a characteristic of the virtual. The visualization of the virtual on the side of the actual (which occurs within art) makes it possible to make rhizomatic connections that transcend the limitations of the empirical landscape and matter. One way in which this could be interpreted is through the idea that cup-and-ring art makes it possible to bring different places of the landscape together that are empirically distinct (Jones and Diaz-Guardamino Reference Jones and Diaz-Guardamino2019, 178). Another way in which the visualization of the virtual on the side of the actual could be interpreted is through the idea of creating an ‘imagined community’ (Bradley Reference Bradley2023, 17) and ‘imagined ancestry’ that had no correspondence with actual, empirical reality.Footnote ⁵ This discussion of cup-and-ring art indicates that it was less concerned with the actual features of the landscape, and more with a movement of reversal between virtual and actual. Rock art includes the landscape within its own logic, the non-hierarchical relation between virtual and actual, and in this way virtualizes the landscape. It is this inclusion of the landscape within the logic of connection between virtual and actual that transforms it into an ancestral world.

Distinguishing between transcendental empiricism, simulacrum and prehistory

On the basis of this analysis of rock art, I want to argue that archaeology needs a new conception of assemblage based on the autonomous logic of the aesthetic. I already discussed above why archaeology has replaced the concept of landscape (as context) with the concept of assemblage. This same shift towards assemblage is visible in the theorization of the social. The social as assemblage requires us to understand that the social is never merely given but needs to be produced (Alberti et al. Reference Alberti, Jones and Pollard2013, 173–6; Latour Reference Latour2005; Jones Reference Jones2012, 22–28). This is a performative conception of the social in which there is not something like an underlying social structure of which cultural artefacts/artworks are a symbolic reflection. The social as assemblage means that the social consists of the continuous shifting attachments/relations between various entities: artefacts, humans, non-humans, environment, concepts, beliefs, etc. (Fowler Reference Fowler2013, 48). Concepts and belief are in this case only one element in the assemblage and no longer operate as condition/background or representation. These attachments are heterogeneous because there can be no homogeneous space in which all entities are included when the social has no a priori background condition. This idea of an a priori condition is based on Kant’s concept of the transcendental (Kant (1781/Reference Kant2004, 93–100). The transcendental was invented by Kant in order to develop the idea that one never has an experience of the real as it is in itself, but that experience has an a priori structure that defines in advance what can and cannot be seen and how it can be seen. Kant’s philosophy provided in this sense the basis for cultural relativism (Kohn Reference Kohn2015, 314), the idea of different cultural perceptions. The reason why the concept of the transcendental has been challenged is because it relies upon a distinction between nature and culture that many societies do not have. This also means that a conception of the social based upon an a priori/transcendental structure is an invention of modernity and therefore not universally applicable to other societies.

The challenging of the concept of the transcendental, however, is not without its problems. When we want to get rid of the concept of the transcendental, experience and the social become without condition, which means that experience and the social itself are no longer possible. The replacement of the concept of the transcendental by assemblage therefore requires a new principle that can explain how the social is held together. How does assemblage make experience and the social possible? We need to understand that assemblage is based upon a relation between the two elements of the real: virtual and actual (Deleuze and Guattari 1987/Reference Deleuze and Guattari2013, 50–51, 67; Govaerts Reference Govaerts2022, 13–17). The relation between the two is only possible through art. It is within art that the virtual can be situated on the side of the actual and can appear. It is therefore art that provides a new principle of connection that holds the two dimensions of the real together (virtual and actual) and which makes the social and a shared world possible. Instead of the transcendental (as a priori condition), it is now the autonomous logic of the aesthetic that holds the social together.Footnote ⁶ This conception of assemblage is distinct from the way assemblage is usually defined: as an assemblage of human and non-human entities (Latour Reference Latour2005). This latter conception of assemblage, I argue, defines the ontological condition of the contemporary world in which the distinction between virtual and actual has disappeared (Govaerts Reference Govaerts2022).Footnote ⁷

This means that actual reality, instead of being characterized by distinct determined forms/entities, has taken up the rhizomatic/formless condition of the virtual. This is why an assemblage of human and non-human entities does not consist of actual entities that interact with one another, but of entities that lack any determined form of their own. Entities (both human and non-human) have taken up the ontological condition of process, which means that they are defined through the attachments in which they are involved with other entities; attachments that change continuously. These other entities in turn consist themselves of a cluster of attachments that are continuously modified. An entity is then not an actual entity, but a cluster of attachments or relations, which are not static, but are involved in a continuous process of change (Deleuze 1969/Reference Deleuze1990, 18, 132, 164, 257; Latour Reference Latour1993, 87; Ingold Reference Ingold2006, 13; Reference Ingold2007: 35). This is what we see in archaeological theory today: a conception of the real as being relational in which entities lack a determined actual form and solely consist of continuously shifting attachments with other entities, that also do not have an actual form of their own, which is more commonly labelled as relational ontology (e.g. Ingold 2008; Jones Reference Jones2012; Alberti et al. Reference Alberti, Jones and Pollard2013, 37–114; Fowler Reference Fowler2013, 20–67; Hamilakis Reference Hamilakis2013, 115–16; Harris and Cipolla Reference Harris and Chipolla2017, 140–3; Crellin et al. Reference Crellin, Cipolla, Montgomery, Harris and Moore2021; Harris Reference Harris2021, 27–30). Such a condition, however, does not define the ontological condition of prehistoric worlds, but the conflation of virtual and actual that occurs in contemporary capitalism. Deleuze theorizes this condition through the concept of the simulacrum (Deleuze 1969/Reference Deleuze1990, 2, 187, 256).

What needs to be understood is that Deleuze’s definition of the real as transcendental empiricism is still based on a conception of the virtual as transcendental (Deleuze 1968/Reference Deleuze2014, 178, 316). The virtual operates here for Deleuze as a condition of experience, that makes experience as genesis possible, which gives rise to the real. Deleuze considers actuality to be a mere effect, an outcome of experience as individuation, rather than having any value itself (Deleuze 1968/Reference Deleuze2014, 212, 270, 308). This means that Deleuze reduces actual reality to a mere appearance (Deleuze 1969/Reference Deleuze1990, 9, 130, 133). When the virtual as condition of experience becomes more real than actual reality, one could say that Deleuze initiates a process of reversal in which the virtual takes the place of actual reality. This is what occurs in his conception of a plane of immanence or metaphysical surface (Deleuze 1969/Reference Deleuze1990, 244), which means that the virtual has risen to the surface. However, when the virtual takes up the place of the actual world, it has lost its virtual condition (which is by definition invisible). This is why Deleuze’s transcendental empiricism, even though it sets out to be a relational conception of the real based on experience as determination, is in fact a loss of the real, a conflation of virtual and actual in which virtual reality takes the place of actual reality itself (Govaerts Reference Govaerts2024, 127; accepted a). This is an ‘upside down’ world or wonderland in which nothing makes sense anymore (Deleuze 1969/Reference Deleuze1990, 1–3, 78). This loss of the distinction between virtual and actual is a result of Deleuze’s conceptualization of the virtual as a transcendentalism. Deleuze both maintains the virtual as condition (of experience) and wants to overcome it by making the virtual more real than the actual. This can only lead to a situation of paradox in which the virtual as condition has itself become the realm of actuality. This is distinct from the relation between virtual and actual that I proposed. In order to clarify this, I would like to put forward the following three distinctions:

1. 1. Transcendental empiricism: This is based on a virtual–actual relation in which actual/empirical reality is the product of a virtual field that conditions it. The relation between virtual and actual is asymmetrical (Deleuze 1968/Reference Deleuze2014, 270, 308), in this case because the actual is dependent upon the virtual, is conditioned by the virtual. Deleuze therefore fails to uncouple the virtual from a transcendentalism.

2. 2. Simulacrum: Experience operates as simulacrum when the realm of virtuality and actuality are conflated with one another. This conflation means that actual reality takes up the rhizomatic/formless condition of the virtual, which transforms actual entities into a constellation of attachments/relations in continuous movement. The conflation between virtual and actual leads to a loss of the real because the relation between virtual and actual has become impossible. The simulacrum is a logical outcome of a transcendental empiricism because it occurs when the virtual (as condition that is responsible for the determination of empirical reality) takes the place of empirical reality (Deleuze 1969/Reference Deleuze1990, 91).

3. 3. Prehistory: I argue that prehistory requires a non-hierarchical relation between virtual and actual in which the virtual no longer operates as transcendental condition of the actual world but is only another side of the real. The virtual operates in this case as the realm of the imaginary or Ideas, which is invisible and immaterial, and not as a transcendental condition for actual reality. A non-hierarchical relation between virtual and actual means that there is a simultaneous movement from virtual to actual and from actual to virtual. This is what occurs in pictures/art because art is an empirical object in which an image appears.

This distinction between transcendental empiricism, simulacrum and prehistory gives a more complex idea of what Deleuze’s philosophy can contribute towards prehistoric archaeology. It is through rock art that one can theorize a different virtual–actual relation that is distinct from Deleuze’s conception of the real as transcendental empiricism.

The alterity of autonomous abstraction

My statement that the ontology of prehistoric rock art is based upon a new relation between virtual and actual in which virtual and actual stand in a non-hierarchical relationship needs to be situated within the contemporary debate on ontology within archaeology and anthropology (Alberti et al. Reference Alberti, Jones and Pollard2013). Benjamin Alberti argues that an ontological understanding means both a taking seriously of the world of other people and a radical questioning of our own theoretical (modern Western) assumptions:

The philosophy of Deleuze, and particularly his conception of transcendental empiricism as a ‘relational ontology’, has contributed to the ontological turn in archaeology and anthropology because it requires us radically to rethink our Western distinction between subject and object, nature and culture, and puts the emphasis on relations (Holbraad and Pedersen Reference Holbraad and Pedersen2017, 182–3). Alberti is critical about this idea of relational ontology because, even though it helps us to rethink Western categories, it simply puts forward a new conception of the real based on relations (Alberti and Marshall Reference Alberti and Marshall2009, 349). This does not allow us to understand how worlds can be radically different from one another, how ‘multiple ontologies’ are possible. I critiqued the idea of relational ontology as well, but for a different reason. I showed above that relational ontology is the condition of the simulacrum, which is the condition of contemporary capitalism. The idea that relational ontology questions our Western modernity is therefore incorrect, because it is what defines our contemporary condition. I also showed how it is a logical result of a continued form of transcendentalism (in Deleuze’s transcendental empiricism), which means that ‘relational ontology’ is in fact a logical result of the discourse of Western philosophy, rather than a radical rethinking of this tradition. Instead of relying on a new general conception of the real, Alberti argues that it is through the encounter with otherness, the strangeness of archaeological material, that our thinking has to be questioned, without having an anchor point in a new theory of reality (Alberti Reference Alberti2016, 173–4). A new philosophical theory of the real does not allow otherness to occur because it neutralizes it by incorporating it within a general theory. Alberti argues for a ‘critical ontology’ (Alberti Reference Alberti2016, 174), which must be distinguished from a typology of different ‘ontologies’. Such a typology is not unsettling enough because the ‘other’ then becomes simply a category within a typological exercise without the activity of (Western) thinking itself being questioned. Critical ontology is more radical than simply understanding the world of past or indigenous societies. It requires a new form of thinking and therefore a new form of doing archaeology.

My argument for a non-hierarchical relation between virtual and actual is more radical than simply an application of Deleuzian philosophy to the prehistoric record because it questions the concept of ontology itself. The concept of ontology can be translated as ‘being’.Footnote ⁸ It is meant to counter the reduction of prehistoric artefacts to a representation or symbol, which neutralizes them, and ignores what cannot be captured by a concept, what resists interpretation (Alberti et al. Reference Alberti, Jones and Pollard2013, 16). The concept of ontology is opposed to the idea that the world of prehistoric societies was simply based upon a different cultural perception of the real (Alberti and Marshall Reference Alberti and Marshall2009, 346). Ontology therefore challenges our common categories and perceptions because it requires us to conceive that what we consider to be the ‘real’ can operate in various different ways. However, even though the concept of ontology indeed challenges aspects of our thought, one cannot neglect that ontology as a concept is a product of Western philosophy (metaphysics). In this sense, one could say that the concept of ontology has an inherent limitation to it that prevents an engagement with radical otherness. Radical alterity, I argue, can only be achieved if the concept of ontology (and therefore of metaphysics) is itself challenged. The non-hierarchical relation between virtual and actual that I proposed is post-ontological and post-metaphysical because it is opposed to the idea of the determination or being of actual entities. Instead, it puts forward a conception of world/entity that is simultaneously actual and virtual (without their distinction being violated), which is what occurs in art. This does not mean that a conception of world based on the non-hierarchical relation between virtual and actual transforms the world/entities into fiction. Instead, it should be considered as a radical form of alterity that forces us to conceive that there is something more real than ‘being’, which is the image-appearance of art. It means that a world is not (necessarily) based on being (as the general condition underlying all actual entities), but upon the constitution of a plane/plateau in which the virtual and actual side of the real are integrated. It is then the picture/art that anchors the world rather than ‘being’.

It is through a questioning of ontology itself as a (Western) concept that a new form of thinking becomes visible. The non-hierarchical relation between virtual and actual requires us to conceive of a form of abstraction that does not operate as epistemology, but as abstract/virtual entity. It is through the autonomous logic of the aesthetic/art that one can start to understand the existence of virtual or invisible entities. A virtual entity is neither a mere product of the human imagination or phantasy (a hallucination, imaginary projection) nor a product of human epistemology (a theoretical concept), but part of a dimension of the real that has an autonomy of its own (Deleuze 1968/Reference Deleuze2014, 376). A virtual entity must be understood as the autonomous existence of an Idea. I argued before that the virtual is the realm of the imaginary and Ideas. The fact that Ideas are a dimension of the virtual and not the actual is why the world has meaning before we know what it means, before we try to understand it by human understanding (Deleuze 1969/Reference Deleuze1990, 48). The autonomous existence of meaning/Ideas indicates that meaning means something different in this case than human knowledge or theoretical/representative concepts. Also, it cannot be reduced to an Idea in the Platonic sense, as a supernatural/metaphysical Idea that defines the essence or model of actual entities. Neither idealism nor realism makes any sense in the context of prehistory because both imply a conception of reality as given, as complete in itself, either awaiting human projection of meaning (idealism) or the discovery of its essential meaning (realism). The autonomous existence of Ideas, on the other hand, points to a conception of Idea as abstract autonomous entity, as an invisible/virtual entity, which is not an abstract epistemological concept or metaphysical essence. An example of an abstract autonomous entity is the ancestors. Such an abstract entity is invisible precisely because it is an abstraction or generalization. It is through art that one can relate to these ‘autonomous’ abstractions, as it is in art that the virtual appears. This is why it is within the experience and engagement with artworks that prehistoric societies could establish a relationship with the ancestors.

One could relate this to a statement made by the anthropologist Howard Morphy for the Yolngu of Aboriginal Australia:

For the Yolngu the paintings were not representations of the ancestors, but manifestations of them. This manifestation, however, required activation. It required the repainting of the designs in ceremonies, the enactment of them, in order for the ancestral beings to manifest themselves. The ancestral beings can be defined as an autonomous abstract entity that was invisible, but could appear/manifest itself in art. This is why the image-appearance of prehistoric art is not representational, but a becoming present of an abstract virtual entity within actual reality. However, the reason why the manifestation of the ancestors requires experience by a viewer is because it is through a viewer that the autonomous logic of the aesthetic, which is based upon appearance, can establish itself completely. Because art operates through appearance, it requires a viewer in order for its logic to be complete. This gives rise to a paradoxical situation in which the manifestation of the ancestors is indeed dependent upon human perception without being a mere projection of the human mind or a representation.

Distinguishing between Neolithic rock art and Upper Palaeolithic cave art

The idea that the ontology of rock art can be defined as a simultaneous movement from virtual to actual, and from actual to virtual, however, loses its relevance when it cannot be specified for different types of rock art through space and time. There are many different types of rock art, which did not all operate within the same way. A general ontology of rock art therefore contributes to the loss of difference between different worlds rather than doing justice to alterity. It is for this reason that it is necessary to theorize how the non-hierarchical relation between virtual and actual can operate in different ways and give rise to different worlds. I will therefore consider how the cave art of Upper Palaeolithic Europe is concerned with a different operation of the non-hierarchical relation between virtual and actual than is the case in the cup-and-ring rock art of Neolithic Britain.

The non-hierarchical relation between virtual and actual in the case of Upper Palaeolithic cave art must be related to the social organization of Upper Palaeolithic hunter-gatherer bands. The Upper Palaeolithic world was a paradoxical world because it required hunter-gatherers to establish relationships with other bands to secure their form of life (Montelle Reference Montelle2009, 41–42). This was necessary because most of northern and central Europe was covered with ice sheets, which forced hunter-gatherers into a relative small area (Montelle Reference Montelle2009, 27–29, 31, 34). This gave rise to a paradoxical situation in which a hunter-gathering way of life (which is by definition a way of life based on movement/encounter) had to be aligned with a shared conception of space and time between different bands (which is normally associated with sedentism) (Montelle Reference Montelle2009, 35–39; Govaerts accepted b). The paradox is situated in the fact that a shared conception of space-time did not have to replace a hunter-gatherer way of life, but had to ensure that such a way of life could be continued. The non-hierarchical movement between virtual and actual must have operated in this case as a way to overcome the distinctions between different bands and create a plateau in which information could be shared or distributed. This is what the virtual does: it overcomes empirical differences and creates a space in which things can be brought together that are empirically distinct. However, we are not dealing with a virtualization of the landscape in which actual disparities were integrated into one plateau (as is the case in the cup-and-ring rock art of Neolithic Britain), but with a paradoxical social logic in which the world simultaneously operated as a shared world and as a constellation of parts without these parts having any integration.

The difference between Upper Palaeolithic rock art (see Fig. 4) and Neolithic rock art (see Figs. 2 and 3) is that the virtualization to which cave art gave rise must have been a theatricalization of the world through which actuality could be seen as other than it is (Montelle Reference Montelle2009, 9). The archaeologists Paul Pettitt and Derek Hodgson have argued that the experience of Upper Palaeolithic people must have operated as a form of projection of animal contours on ambiguous surfaces, which is related to the sensitivity of the hunter-gatherer to detect animal shapes in the environment (Hodgson and Pettitt Reference Hodgson and Pettitt2018, 593–7). It is through virtualization as theatricalization that the world can be seen as other than it actually is, that an animal can become more than prey (taking up a position within a shared system of information), without that this virtualization operates as an integration of the actual world. It is not a real integration, but an ‘as if’ integration, a theatrical integration, which leaves the actual (hunter-gatherer) world untouched. This is what must have defined the paradoxical social logic of the Upper Palaeolithic world: a continuous (non-hierarchical) reversal between a theatrical ‘as if’ world that allowed a shared space-time between different bands, and a world based on a cluster of parts that lacked integration (which is the world of the hunter-gatherer that is based on movement and encounter with the animal, but is not concerned with a shared conception of space-time) (Govaerts accepted Reference Govaertsb). This is the only way to explain how the Upper Palaeolithic world was simultaneously based on hunting and gathering and on a shared ‘plateau’ between different bands (that was necessary to secure a hunter-gatherer form of life rather than transforming it).

Figure 4. Panneau des Lions, Grotte Chauvet-Pont-d’Arc (Ardèche). Source: photograph by A. Frich. Copyright holder: Centre national de Préhistoire/Ministère de la Culture (France).

The archaeologist Steven Mithen has argued that Upper Palaeolithic cave art was an art of information (Mithen Reference Mithen1996, 196; Reference Mithen and Mithen1998, 182). This does not mean that cave art was representational. The simultaneous reversal between virtuality and actuality that defined the paradoxical social logic of Upper Palaeolithic bands must have also occurred in cave art. This could explain why there was both a concern with the actual realistic features of animals (evidence of close observation of anatomy and depiction of how animals move), but also with theatricalization (which we see in the way the animal figures are included within the features of the cave wall to make them part of a theatrical situation (Pfeiffer Reference Pfeiffer1982, 136, 142–3; Aujoulat Reference Aujoulat2004, 228–32; Hodgson and Pettitt Reference Hodgson and Pettitt2018, 591). Upper Palaeolithic cave art as an art of information requires a radically different concept of information; information not as representation or theoretical knowledge, but as virtual entity that had an autonomy of its own. Whereas the virtual took the form of the ‘ancestors’ in the case of cup-and-ring rock art, in the case of Upper Palaeolithic cave art we are dealing with an autonomous form of ‘information’. It requires us to conceive of a form of information that was neither a mere epistemological product of the human mind nor a metaphysical essence, but that had an existence of its own, which could only manifest itself in the encounter with cave art. Western art could be considered as having inherited something of this autonomous form of information/meaning, even though in a far less sophisticated way. This is because the meaning of a work of art resists reduction to both sensibility and conceptualization, and therefore exists regardless of whether it is understood (or precisely because it cannot be straightforwardly understood). Information in the Upper Palaeolithic therefore must have been less produced, stored or memorized by Upper Palaeolithic communities, than activated within the experience of cave art. It is the taking seriously of ‘autonomous abstraction’ as a new type of entity that forces the prehistoric archaeologist to radically question his own way of thinking. This is how an ontological investigation can become ‘critical ontological’ (Alberti Reference Alberti2016) and even challenge the concept of ontology itself (which is what I think prehistoric archaeology requires).