Mapping the Symbolic Structure of Palaeolithic Rock Art Using Co-occurrence Network Analysis
摘要
This study examines the internal organisation of Magdalenian parietal art along the Bay of Biscay axis through the application of network-based and multivariate statistical analyses to a large, systematically documented graphic corpus. The dataset comprises more than 500 figures from nine decorated cave sites located in the Basque region, which constitutes a key transitional zone between the Iberian Peninsula and the rest of continental Europe. Palaeolithic rock art is approached as a structured semiotic system, in which thematic associations and formal conventions can be investigated independently of predefined symbolic interpretations. Theme co-occurrence networks were constructed at the panel scale using both frequency-based and Jaccard-normalised weighting, and were complemented by filtered networks, minimum spanning trees, and panel–theme bipartite models. The results reveal a non-random patterning of thematic associations, characterised by the recurrent dominance of large herbivores—particularly bison, horse, and ibex—and by the presence of hierarchical and modular structures operating at multiple analytical scales. Formal variables such as figure orientation and inclination also display structured distributions, supporting the existence of shared graphic conventions. Rather than advancing specific symbolic interpretations, this study identifies recurrent organisational patterns within Magdalenian parietal art and situates them within a broader structural framework. These findings highlight the potential of network-based approaches for comparative analyses of Palaeolithic graphic systems, particularly when integrated with detailed contextual and taphonomic assessment.