Francesc Marginedas defends his Doctoral Thesis "Processing of human bodies in archaeological contexts of cannibalism"
This morning, in the Abric Romaní Hall at IPHES-CERCA and within the framework of the Doctoral Programme in Quaternary and Prehistory of the Universitat Rovira i Virgili, the public defence of the doctoral thesis titled “Processing of human bodies in archaeological contexts of cannibalism”, conducted by Francesc Marginedas, took place.
The thesis was supervised by Dr. Palmira Saladié, researcher at IPHES-CERCA, and Dr. Antonio Rodríguez-Hidalgo, researcher at the Institute of Archaeology-Mérida (CSIC–Junta de Extremadura) and associate researcher at IPHES-CERCA.The panel was composed of Dr. Silvia Bello, from the Natural History Museum of London, Dr. Noemí Salas, from the National Center for Research on Human Evolution (CENIEH), and Dr. Rosa Huguet, a researcher at IPHES-CERCA.
The IPHES-CERCA team congratulates the new doctor. Congratulations!
Summary of the Doctoral Thesis
This doctoral thesis presents a taphonomic analysis of three main archaeological assemblages, an isolated skull, a subrecent burial set, and comparative data previously published from four other sites. The three main assemblages show clear evidence of anthropic intervention and come from different chronological and geographical contexts: the Magdalenian of Maszycka Cave (Poland), the Neolithic of El Mirador Cave (Spain), and the Bronze Age in the same site (level MIR4). The isolated skull comes from the Bronze Age of Txispiri Cave (Spain), while the subrecent assemblage includes 19th-century burials found at Abric Romaní (Spain). This corpus is complemented by four previously published assemblages, from which only cranial remains showing evidence of cannibalism were considered, covering different periods and regions: the Lower Palaeolithic of Gran Dolina (Spain), the Magdalenian of Gough’s Cave (United Kingdom), the Neolithic of Fontbrégoua (France), and the Neolithic of Herxheim (Germany).
Based on these materials, the thesis is structured around three main objectives: (1) to analyse evidence of human activity related to ritual and symbolic contexts, with special attention to the manufacture of skull-cups; (2) to develop a new method for identifying the use of metal knives in the processing of human bodies during the Bronze Age, through the detection of microscopic residues; and (3) to study two archaeological contexts with evidence of cannibalism—one of them unpublished (the Neolithic of El Mirador Cave)—in order to contextualise them within a broader European chronological and cultural framework.
To achieve these objectives, various analytical methodologies were applied. The study of the cranial remains included the analysis of the distribution and frequency of cut marks using Geographic Information Systems (GIS) in the assemblages of Gran Dolina, Gough’s Cave, Herxheim, Fontbrégoua and El Mirador Cave (MIR4 level). In addition, machine learning techniques were employed to analyse fracture planes on skulls from MIR4, Txispiri Cave and Abric Romaní. A controlled experiment was also designed in which a deer was processed using copper and bronze knives, to establish reference parameters. Based on these, the chemical composition of metallic microparticles found in archaeological cut marks from MIR4 was characterised.
The study is completed with an anthropological and taphonomic analysis of the complete human assemblages from Maszycka Cave and El Mirador Cave (Neolithic levels), aiming to determine the minimum number of individuals (MNI), estimate their ages, and document evidence of processing and consumption. In the case of El Mirador Cave, radiocarbon dating and strontium isotope analyses were also carried out to determine the chronology and geographical origin of the individuals.
The results obtained make it possible to establish differentiated patterns depending on the archaeological context. The GIS analysis of cut marks on skulls reveals that, in contexts with a possible ritual significance (such as Gough’s Cave, Herxheim, Fontbrégoua or MIR4), marks are concentrated in specific areas associated with intensive cleaning of the skull. In contrast, in contexts without ritual indicators (such as Gran Dolina), the distribution is more dispersed and corresponds to functional processing. Additionally, machine learning has proven to be an effective tool for distinguishing between perimortem fracturing (as in MIR4) and postmortem fracturing (as in Abric Romaní), as well as for classifying fractures of uncertain diagnosis, such as the calotte from Txispiri, ultimately attributed to post-depositional taphonomic processes.
The new method proposed for identifying metallic residues on cut marks has shown great potential for application in Metal Age contexts. At MIR4, copper residues were detected on cranial fragments and on a rib, suggesting the use of copper in possible ritual manipulations. However, the scarcity of such residues and the difficulty in distinguishing between ancient and modern contamination present methodological challenges still to be resolved. The Magdalenian assemblage from Maszycka Cave allowed the identification of at least ten individuals (six adults and four subadults). The taphonomic analysis shows clear cut marks and anthropic fracturing associated with consumption, ruling out the previously proposed ritual or funerary interpretation. The identification of new postcranial remains with traces of skinning, disarticulation, and marrow extraction reinforces the hypothesis of cannibalism. The assemblage has been dated to around 18,000 cal BP.
Finally, in the Neolithic levels of El Mirador Cave, remains of at least nine individuals (three adults and six subadults) were identified, dated to the Late Neolithic (5720–5600 cal BP). The remains were found both in situ and mixed with materials belonging to a Chalcolithic burial. The taphonomic analysis reveals clear evidence of butchery, cooking, and consumption. Radiocarbon dating suggests a single event of cannibalism followed by the abandonment of the cave, while strontium isotope data indicate a likely local origin for the individuals consumed.