Skip to main content

Paul Pettitt

Professor Paul Pettitt from our Department of Archaeology and Professor Bob Kentridge from our Department of Psychology explores the origins of human art through Visual Palaeopsychology.

Archaeology is the only science that can provide tangible evidence of our biological, cognitive and behavioural evolution. The oldest formal period, the Palaeolithic (‘Old Stone Age’) saw the origins of Homo sapiens, art, burial, and long-distance exchange, all in the context of repeated dispersals out of our African homeland increasingly across the Old World and into the new. By the end of the Pleistocene (‘Ice Age’) around 11,500 years ago, only distant islands and the poles remained to explore, and already, many groups had begun complex interactions with plants and animals that would soon result in domestication, agriculture, writing, and urban settlement. 

Visual culture – whether we call it ‘art’, ‘marking’, ‘symbolism’ or ‘making’ arose in the Palaeolithic. Palaeolithic hunter-gatherers have left traces of their visual culture in deep caves, on cliff walls, and in the form of carved, engraved and painted portable objects of bone, stone, antler and mammoth ivory. These form a rich database for reconstructing the origins and long-term evolution of art – something that seems effectively unique among humans. But while archaeologists have a suite of sophisticated tools to analyse these and the sediments in which they are found, interpretations of the ‘earliest art’ tend to be informal and non-testable. A number of specialists have forwarded attractive hypotheses as to why characteristics of our visual system might have been influential in the way art first arose, yet the few attempts so far to link archaeology and psychology tended to draw on theories that modern scientific psychology rejected decades ago. Hence, important theories about such a truly unique and universal human characteristic have remains conjectural. 

Durham’s Centre for Visual Arts and Culture (CVAC) formed a fertile ground for the two of us to begin a collaboration around five years ago. One of us (PP) researches the origins of art in the Middle (Neanderthals) and Upper Palaeolithic (Homo sapiens), the other (RK) conducts experimental studies of visual perception and attention and their brain mechanisms. These might seem distant topics, and aside from a shared enjoyment of Hawkwind’s music, we might never have rubbed along were it not for the warm encouragement of CVAC’s director, Ludmilla Jordanova. It didn’t take many beers to discover that we shared many common interests, focussed on a desire to place research into the psychological factors that might have influenced the origins, production, and uses of art in our deep evolutionary past on a firm scientific footing. We were lucky enough to attract a brilliant archaeology graduate, Lisa-Elen Meyering, for one of CVAC’s doctoral studentships, to develop the fledgling discipline we dubbed ‘Visual Paleopsychology’. Our interdisciplinary and cross-faculty collaboration has been such a success that we now have a dozen or so graduate students and academics working together in the Durham Paleopsychology Group, and a joint AHRC/German DFG-funded research project. Here is a flavour of our work to date. 

A little after 40,000 years ago our ancestors began to create figurative art, drawing, engraving and painting the large herbivores on which they were dependent for survival, such as wild horses, deer, bison and mammoths, onto cave walls, tools, and other portable objects. This was not the earliest manifestation of visual culture: Eurasian Neanderthals and their African and Near Eastern Homo sapiens contemporaries had been producing non-figurative markings using wet pigments since at least 100,000 years ago. So, what brought about the profound and abrupt shift to figurative work? People might simply have imagined that they saw animals on the irregular surfaces of cave walls. This is not as outlandish as it may sound. We have all seen faces in clouds or in a slice of toast. Imagine that you are in a dim cave, with only a flickering tallow lamp to light your way. It’s quite likely that you’ll imagine seeing glimpses of things that are important to you as the light dances around. This is called ‘pareidolia’. It is not limited to seeing faces, poodles or planes: people’s pareidolic imagery take the form of important or familiar things. Bison, horses, mammoths, reindeer and other herbivores loomed large in the lives of our hunter-gatherer ancestors. It’s a product of the way the brain has evolved, to make quick sense of suggestions of threatening or valuable things: better to over-interpret and be wrong, than to be eaten or to starve. 

But how to avoid this being simply a ‘just so story’? We thought this proposition was testable. If cave depictions were stimulated by pareidolia then the cave walls should affect modern viewers in the same way as our ancestors, who after all shared our ‘suggestible’ visual system. If a natural contour evoked the shape of a horse’s back and a horse had been painted at that point (as if often the case) then, we thought, all we need to do is to erase the art and track the eyes of modern participants and see if they pareidolically see the same. We couldn’t of course head to Spain and start scrubbing away some of the world’s oldest art away, but we could do this with modern technology.  Our PhD student Izzy Wisher took hundreds of photographs of Spanish cave walls with Palaeolithic art on them (and some cave walls without), creating digital 3D models of those cave walls using high-resolution computer photogrammetry. She then inserted these models into a virtual reality (VR) cave system. A participant in our experiments could explore the caves in VR, armed with a virtual flickering torch for illumination. We could ask them questions about what they saw, or imagined they saw, and track their eye-movements from inside the VR headset, so we could build up a detailed picture of exactly what was attracting their attention in the simulated environment. The results were very encouraging and, even more satisfyingly, they fitted in beautifully with the results of our next study.   

Palaeolithic artists could be minimalists when depicting animals – a few lines to bring out the curve of a back or a striking muzzle. Lisa-Elen researched the salience of such art pecked and engraved onto cliff walls in Portugal’s Coa Valley. The art’s locations probably related to places where it was possible to monitor prey animals at a distance. We hypothesised that the parts of an animal shown in an incomplete depiction corresponded to the parts of the animal that were most helpful identifying it in the far distance. We adapted a method used to study the perception of faces in experimental psychology to test our hypothesis. The ‘Bubbles’ program presents participants with a view of a tiny part of an animal and asks them to identify it. They attempt this many times, the program switching between animals and moving the viewing window to a new location every time, continually adjusting its size to keep the task both from being too easy or difficult. After hundreds of trials the computer program adds together the views through all of the windows where an animal was correctly identified, subtracting those that caused incorrect identifications. Running this experiment with many observers and combining the results shows us the parts of an animal that were most likely to contribute to a correct identification, such as what bit of a horse is best at defining a horse at a distance. There was a pleasing correspondence between the results of our bubbles experiments and the parts of animals that were preferentially featured in rock art. Somewhat to our surprise (but pleasingly so) we also saw that participants in our VR experiment had often directed their gaze to these same locations, even when they were just viewing a bare wall where they imagined seeing one of these animals. As we predicted, the bits of animals of most use are those that tend to be stylistically exaggerated in Palaeolithic art. 

Our final study is very high-tech and is a work in progress. We are researching how people used flat pieces of slate bearing faint engravings on a 16,000-year-old campsite on the banks of the Rhine. These are hard to see, and to make matters worse, many different engravings of different animals are sometimes made on top of one another on the same piece of rock (known as plaquettes). We wanted to know whether, when the engravings were fresh it was easier to see the newest picture or whether, by turning the plaquette (these are generally hand-sized) one could discern the separate pictures. For some, there is even an impression that the shape of the rock might have inspired the choice of animal to engrave (pareidolia again). The plaquettes are far too delicate and precious to use themselves in experiments, so again we turned to VR, using a dedicated high resolution 3D scanner and digital microscopes to capture 3D digital models of them. We can then simulate the effect of fresh engraving (pale rock dust sitting in the newly inscribed lines), or even remove the engravings in VR. To explore how we can use natural lighting from different angles to adjust the visibility of engravings our participants manipulate 3D printed models of the plaquettes and see the VR plaquettes turn with them. Our combination of 3D digital modelling tracked 3D prints, VR and eye-tracking is cutting edge and gives an overwhelming sense of reality to the plaquette observers see in VR. In VR we have exquisite control over the art, its context and lighting, and a continuous record of the way a person’s gaze passes over the plaquette as they explore it. We can use a multitude of methods from psychology to analyse this eye-tracking data and to compare it to computational models of visual processing in the brain as we slowly unravel how our evolutionary past guided the nature of our earliest visual culture. 

Find out more: