Iñaki Intxaurbe, researcher at the EHU-University of the Basque Country, analysed the distribution of images in nine caves in the Basque Country, using network-based analysis and statistical techniques
By studying the recurrent associations between the images, Intxaurbe identified the hierarchical structural patterns followed by prehistoric drawings. Bison, horses and goats appear in the centre of the compositions, and the rest of the animals are distributed around them. The horses are depicted facing the right, and the other animals, in general, facing the left. Although the reasons for this are unknown, the study shows that images correspond to a specific internal organisation.
Postdoctoral researcher Iñaki Intxaurbe conducts his work at the Department of Graphic Expression and Engineering Projects of the EHU-University of the Basque Country: “I work with rock art, with figures drawn on the walls of caves. I make digital copies of these figures so that I can later carry out simulations and statistical analysis”, he explains. Specifically, he analysed 500 graphic motifs found in nine caves in the Basque Country, which were made approximately 16,000 to 14,000 years ago. “I wanted to know whether these images were created at random or whether they followed some kind of pattern”, he added.
Graphic expression can be defined as the visual representation of ideas, concepts or information. If Palaeolithic art shows indications of collective thought or symbolism, it can be understood as a structured communication system. “Given that cave art is a form of visual and graphic expression, it would be reasonable to think that there might be some kind of syntactic order behind it. Syntax determines the organisation of words and sentences, and studying it allows us to understand the structures used in a language”, he says. To carry out the study, Intxaurbe used statistical methods and a network analysis of the entire set of figures to decipher the relationship patterns between the different elements.
The images are not organised at random
The results show that “cave painting was done in an organised manner, not at random.” In general, the bison, horses and goats occupy central positions, while the other animals perform other significant functions around them. “The bison are usually orientated towards the left and the horses towards the right, apart from a few isolated cases in which the reverse occurs: the horses are facing left and the other animals are facing right. This demonstrates that the system is not random and that behind these images is a general idea, an important element within its symbolism”, says Intxaurbe. The research demonstrates the existence of a structural pattern associated with these three predominant animals: “The statistical analyses suggest that all images in the caves were organised according to the presence and orientation of these animals, within some levels of significance.”
The study does not offer explanations as to the reason for these practices. What it does show, however, is that behind the drawings is an organised structure, though it is not yet possible to determine the intended message. The research also confirms some ideas put forward in the 1950s: “At that time, it was already thought that rock art was not random. It was known that most of the animals were orientated to the left. Some authors suggested that this was because most of the artists were right-handed and therefore painted the animals facing left. However, our work shows that the horse is one of the most depicted animals and that, statistically, its predominant orientation is the opposite: towards the right”, explains the EHU researcher.
He adds: “The cave paintings were not created for leisure purposes. We know they were important to their creators, as they were only made in some caves, not in all of them. We now know that cave art at the end of the Palaeolithic had a syntactic structure and that behind it was an idea or graphic code that we don’t understand but which was important to them.” They chose specific caves for them. We now know that cave art at the end of the Palaeolithic had a syntactic structure and that behind it was an idea or graphic code that we don’t understand but which was important to them.”
Open source
Intxaurbe combined different methodologies in his research: “I used multivariate statistical analysis and network analysis, and the results are highly promising for conducting comparative analysis of Palaeolithic graphic systems.” The EHU researcher offers a useful tool for studying the internal organisation of symbolic systems. In fact, he published both the methodology and the database in open source so that it could be used by other researchers. “In the future, they could also be applied to caves studied by other teams. This will allow us to continue to expand the database and to compare the results with caves in France, the Iberian Peninsula or any other part of the world.”
The methodology could also be useful for studying historical inscriptions, palaeography or historical linguistics, since it is based on co-occurrence analysis and combinatorial structures. “For example, it could be applied to the Iberian system or the Basque system to study how sequences we don’t yet understand are organised, and to see whether they follow some kind of network structure. This could help, for example, to determine the syntax of other languages that have not yet been deciphered”, he concluded.