Quantifying Levallois: a 3D geometric morphometric approach to Nubian technology

Archaeological and Anthropological Sciences (2025) 17:88 https://doi.org/10.1007/s12520-025-02199-2 RESEARCH Quantifying Levallois: a 3D geometric morphometric approach to Nubian technology Emily Hallinan1 · João Cascalheira1 Received: 15 October 2024 / Accepted: 27 February 2025 / Published online: 24 March 2025 © The Author(s) 2025 Abstract Levallois technology, a hallmark of Middle Palaeolithic stone tool manufacture, involves sophisticated core reduction strategies that have major implications for understanding human cognitive and technological evolution. However, traditional methods of analysing Levallois cores often fail to capture the nuanced variability in their morphology. This study introduces a novel application of three-dimensional geometric morphometrics (GM) to quantify the shape variability of Nubian Levallois cores from the Nile Valley and Dhofar regions. By employing this technique, we analysed core surfaces and preferential scar shapes, identifying distinct regional and technological patterns. Our results reveal significant inter-regional differences in core elongation and surface convexity, highlighting the importance of shape-oriented, rather than metric-based, analysis of prepared cores. This new GM approach offers a robust and replicable tool for investigating lithic variability and holds potential for broader applications in Palaeolithic research, enhancing our understanding of human technological adaptations. Keywords 3D geometric morphometrics · Levallois · Lithic technology · Nubian Levallois cores Introduction Levallois technology is regarded as a hallmark of stone tool manufacture in the Middle Stone Age in Africa and the Middle Palaeolithic of Europe and western Asia, associated with at least three hominin species – early humans, anatomically modern humans and Neanderthals (Foley and Lahr 1997; Tryon et al. 2005; Hublin 2009). Named after the area in France, Levallois-Perret, where it was first described (Boucher de Perthes 1857; de Mortillet 1883; Commont 1909), Levallois is a type of prepared core technology that involves preparation of the convexities of the core surface and faceting of the striking platform to control the shape and size of the final end-product. The degree of planning involved in the pretermination of the Levallois end-product has been viewed as an indicator of cognitive complexity owing to the conceptualisation of the final core and product form, the sequential method required to achieve it, and social learning to transmit the process (Schlanger 1996; Lahr and Foley 2001; Wynn and Coolidge 2004; Lycett et al. 2016). * Emily Hallinan 1 Interdisciplinary Center for Archaeology and Evolution of Human Behaviour, University of Algarve, Faro, Portugal The implications of Levallois for technical and economic behaviour has also seen varied interpretations, related to its efficiency (Brantingham and Kuhn 2001; Sandgathe 2004; Lycett and Eren 2013) and the role of its end-products (Sisk and Shea 2009; Eren and Lycett 2016; Shimelmitz and Kuhn 2018). Despite the central place that Levallois occupies in human evolution, its definition (Copeland 1983; Van Peer 1992; Dibble and Bar-Yosef 1995) and identification of both the technology (Hu et al. 2019a, 2019b; Li et al. 2019; Pallo 2022) and specific Levallois methods in assemblages (Rose et al. 2011; Goder-Goldberger et al. 2016; Blinkhorn et al. 2021a, 2021b; Hallinan et al. 2022a) remain highly debated topics. Definitions of Levallois have shifted over the past decades, yet shape remains a key concept in all of them. In an early typological approach, Bordes (1950, 1961, 1980) emphasised that the predetermined shape of the Levallois flake was achieved through special preparation of the core prior to its removal, whereby different configurations of preparatory flaking generated end-products with different morphologies – flakes, blades or points. Accordingly, the morphology of the Levallois core was strongly tied to the end-product shape: an oval core produced an oval flake, a triangular core produced a triangular flake. Subsequent technological approaches shifted attention away from the final Vol.:(0123456789) 88 Page 2 of 22 flake form towards understanding the Levallois reduction process (Bar-Yosef and Dibble 1995). Boëda’s (1988, 1994, 1995) widely applied Levallois concept focuses on the geometric structure of the core as a volume, worked through a series of steps and fulfilling certain criteria. The volume of the Levallois core must consist of two asymmetric convex hemispheres that possess a plane of intersection at the core’s margin. These hemispheres or surfaces are hierarchically related, each serving a specific, fixed role in reduction: a preparation surface for the production of striking platforms, and a flaking surface for the removal of the Levallois product. The flaking surface possesses both lateral and distal convexities that must be maintained to control the direction of force for the end-product, which is removed parallel to the plane of intersection. Viewed within a technological framework, it is the pattern and orientation of the preparatory flaking – not the core shape itself – that determines the morphology of the end-product (Boëda 1995; Van Peer 1992). However, whether lithic analyses are conducted from a typological or technological perspective, they encounter the same limitations – that shape is a qualitative, rather than quantitative variable. Geometric morphometrics and Levallois morphology Geometric morphometric (GM) approaches provide a statistical framework for studying shape variation – where shape is the specific geometric configuration of a specimen – independently of size (Slice 2007). Originally developed for applications in biological sciences (Rohlf and Marcus 1993; Adams et al. 2004), GM has gained traction as a powerful tool for quantitative analysis of shape across different aspects of lithic studies, though most commonly applied to bifaces, flakes and points since these can be consistently orientated and aligned according to geometrically correspondent points (e.g. Lycett et al. 2006; Archer and Braun 2010; Iovita 2011; Archer et al. 2016, 2018; Herzlinger et al. 2017; Herzlinger and Grosman 2018; Archer and Presnyakova 2019; Okumura and Araujo 2019; Timbrell et al. 2022a). As a consequence, most current GM studies that relate to Levallois technology have focused on debitage, using two-dimensional (Eren and Lycett 2012; Picin et al. 2014; Buchanan et al. 2023) and, occasionally, three-dimensional techniques (Chaćon et al. 2016; González-Molina et al. 2020; Delpiano et al. 2021; Bustos-Perez et al. 2024). The problem of identifying homologous points between specimens for comparison has meant that the application of GM to cores has been limited, since this artefact type displays marked variability in form through continuous, non-uniform reduction trajectories. However, the highly structured geometry of Levallois cores means that c (...truncated)