4.2 State of preservation, MNE, MNI, element representation index and biological profile of the individuals
Due to the high fragmentation of the materials and their splitting among different institutions, the MNE, MNI and element representation index (Table 3) may be in all likelihood underestimated, being it difficult to confirm or exclude all the possible associations among the fragments of each skeletal district.
More in detail, as regards the overall element representation index of the different skeletal districts, it may appear rather low. This may be more evident if we compare the results from the Farneto rock shelter with other element representation indexes from other archaeological contexts for which the type of deposition is known and/or the excavation documentation is clear (cf. Robb et al. 2015; Knüsel et al. 2016). The dispersion of so many elements can be attributable to different causes, such as inaccuracies during the recovery, that surely occurred for the Farneto rock shelter, as well as the impossibility to count and study many fragments still embedded in the sediment blocks (Fig. 2c, d). However, the most fragile bones (e.g., sternum, vertebrae) easily undergo natural mechanical destruction because of their internal structure and composition, also taking into account the instable karstic environment of the area. Moreover, as gnawing traces were found, it is possible that also carnivore activity may have caused the dispersal of similar bone elements. On the other hand, the dispersal of the smallest bones (e.g., phalanges and other hand and foot bones) can be attributable to one or more intentional displacements of the remains for ritual or practical reasons, during which only the biggest bones are collected and the smallest ones are left or forgotten (Knüsel et al. 2016).
In the Farneto rock shelter, despite the rather low overall element representation index, all the bones are more or less attested and their number is quite consistent. The most robust elements, such as humeri and femurs, are the better-attested, while the most fragile elements, such as sternum and vertebrae, are underrepresented. The smallest hand and foot bones are clearly underrepresented but still present in a coherent number. Therefore, we may assume that any bone seems deliberately selected or stored. Moreover, the degree of fragmentation reached by the different bones seems coherent with the original fragility/robusticity of each one, as shown by the high number of little (0–20 mm; Fig. 4a) cranial fragments. As regards the representation of left vs. right sides, adult left elements are slightly more attested than the right ones (Table 3). Nevertheless, this difference seems too small to suppose an intentional selection of left elements.
Regarding sex and age at death, both male and female individuals and all age classes are attested, so that a horizontal intentional selection of the individuals to be buried can be excluded. The paucity of perinatal or very young infant remains, testified by one single specimen (the complete left humerus C4-61), can be explained by several reasons that are not mutually exclusive: their natural fragility (Gordon and Buikstra 1981; Walker et al. 1988; Guy et al. 1997), the inaccuracy and misidentification during the recovery (Schaefer et al. 2009), possible underlying ideologies and/or funerary practices that originally led to their exclusion (Tocheri et al. 2005; Halcrow et al. 2008; Cveček and Schwall 2022).
4.3 Bones colour, burning and gnawing traces
Colour During the study of the assemblage, in the lack of any information regarding the original position of the human remains, the annotation of bone colour and staining should have been functional to the reconstruction of the original disposition of the skeletal materials, but only a few differences were noticed. As already mentioned, the vast majority of the bones have a very light white colour, thus in all likelihood, they were lying on a soil that contained calcium carbonates (Dupras and Schultz 2013), such as a gypsum cave environment, typical of the area. In fact, among the radiometric dated samples, the 5 brown elements, presenting a colour compatible with a soil rich in humic acids (Dupras and Schultz 2013), were clearly recognized as extraneous to the prehistoric context (Table 1). The interpretation of the light reddish specimens is difficult, but we can propose that subadult bones react differently to the microenvironment of deposition, perhaps with a different chemical exchange with soil substances (Guy et al. 1997). Anyway, this colour is more similar to the light colour of the majority of other bones than to the brown colour of the more recent specimens excluded from the study.
Burning The reddish brown and grey colour of the very few burning traces detected on the skeletal elements is compatible with an estimated fire temperature between 285 and 645°C (degree 2–3 of Shipman’s scale; Shipman et al. 1984; Fig. 5a). These few traces seem related to accidental contact with fire. In fact, in the Eneolithic Italian sites (in Tuscany, Grifoni Cremonesi 2001; Lombardy, Barfield 2007; Emilia Romagna, Cavazzuti 2018) where cremation is considered part of the funerary ritual, the percentage of burned skeletal remains is much higher.
Gnawing The total number of bone elements affected by gnawing activity may be underestimated due to the state of preservation of some skeletal remains (e.g., surface covered by incrustations). Nevertheless, some rodents (Fig. 5b) and carnivore (pits and punctures; Fig. 5c, d, e) gnawing traces were found. In the case of the adolescent right tibia (C7-AMH-223; Fig. 5d) the two punctures distant 38 mm from each other are compatible with the dentition (distance between maxillary and mandibular canines) of Canis familiaris, Canis lupus or Ursus arctos (Murmann et al. 2006; Pokines 2013), animals that inhabited the Farneto area in prehistoric times. In the surrounding areas, such as in the nearby territory of Monterenzio Vecchia (Bologna), the presence of the three species is documented from the Eneolithic to the Bronze Age (Sala 1980; Maini 2012; Ciucani et al. 2019). The most ancient specimen of Canis lupus was found in the Neolithic site of Razza di Campegine (Reggio Emilia; Cazzella et al. 1976), while several canid remains were found in Bronze Age sites all over the region (Farello 1995; De Grossi Mazzorin 1996; Koupadi et al. 2020), occasionally accompanied by Ursus arctos remains, such as in the sites of Crocetta di Sant’Agata Bolognese (Bologna; Maini 2012) and Monte Castellaccio (Bologna; De Grossi Mazzorin 1996). Moreover, skeletal remains of Ursus arctos were recognized among the human remains from the Farneto rock shelter in the 90s by Gianni Giusberti (Facchini et al. 1999). All these results are coherent with an original deposition inside a cave environment, where human bones were likely exposed or buried in shallow pits, easily reached by wild animals (Pokines 2013).
Within the overall sample, several lesions that mimic puncture marks may be detected, in particular on cranial vault surfaces (e.g., frontal bone C1-9), but they are more likely the result of marks left by falling blocks or produced during the rolling of the bones, which probably originated from a higher place (Fernández-Jalvo and Andrews 2011).
4.4 Peri mortem lesions
The presence of several peri mortem lesions on both cranial and postcranial remains suggests that the Farneto prehistoric group performed practices of treatment of the cadaver on at least some individuals of both sexes and all age classes. These practices encompassed various activities such as disarticulation (lesions related to joint structures, i.e., epiphyses, entheses of ligaments, capsules or tendons stabilizing some joints) or dismemberment (bone fractures, deep chop marks on epiphyseal regions or on diaphysis) and scarnification/defleshing (cut marks or chop marks in correspondence of large muscular masses, as well as fine cuts aimed at cleaning the bones even from thin residues of soft tissues or just the periosteum; Table 4; S1). In some cases, episodes of interpersonal violence or of violent actions on cadavers (e.g., ritual sacrifices, anthropophagic practices) can be hypothesised (cf. Duncan 2005; Belcastro et al. 2010; Cavazzuti et al. 2020; Janković et al. 2021).
Due to the fragmentation of the bones and the state of preservation of their surfaces and fracture margins (e.g., incrustations, decortication, but also use of resins and consolidants during previous studies; Fig. 6), many lesions could have not been preserved or their characteristics be no longer discernible and interpretable. The fact that peri mortem lesions were found especially on femurs could be also the result of accidental sampling due to the major robusticity (dimension and cortical thickness) ensuring a better preservation of this bone. For these reasons, we did not calculate the frequency of lesions and the most affected skeletal districts. It is thus possible that their presence has been underestimated.
The morphology of many cut marks (e.g., flaking of borders, shoulder effect, parallel microstriae on the bottom of the sulcus; Fig. 9) is typical of stone tools, more or less retouched (Greenfield 1999; Moretti et al. 2015). Indeed, copper blades have not been found in the site (Nobili et al. 2017). The chop marks, instead, indicate powerful actions during cleaning of fresh cadavers with thick muscular masses or during dismembering practices (Fig. 10b, c, d), that in some cases could have caused bone breakage (e.g., humerus C4-58; S1).
As regards crania, our results allow hypothesizing disarticulation/dismembering of both the mandible from the cranium and the cranium from the trunk, as well as defleshing (Table 4). Lesions such as the chop mark on the right mastoid process (C1-7; Fig. 8c), as well as the lesions detected on mandibles (V-AMH-200 and C2-13), may be the result of the attempt to disarticulate the mandible from the cranium (Table 4). In the first case, the lesion is also compatible with disarticulation of the cranium from the column, as are the lesions found on two cervical vertebrae (Fig. 9a, b; S1).
The presence of cut marks of various dimensions (from very thin, barely perceptible lesions, to long and deep ones, e.g., on frontal bone V-4; Fig. 7) on cranial vaults suggests the presence of cleaning practices performed with different degrees of strength. Some lesions (especially, but not only, patterned fractures; Kimmerle and Baraybar 2008; Table 4) seem to indicate violent actions whose aim cannot be ascertained (interpersonal vs. ritual forms of violence, on the living person vs. on the cadaver), even if accidental trauma (e.g., falls, falling rocks) cannot be excluded (fractures in V-2B, frontal and occipital fractures in C1-6).
In two cases, the particular features associated with the lesions (the cut marks in correspondence of the perforating lesion in V-3; Fig. 6c; and the morphology of the margins of the lacuna in V-2; Table 4) could suggest some forms of therapeutic interventions on previous wounds. The absence of traces of bone reaction attests that the individuals did not survive to both the wounds and the operations. Medical surgery has been hypothesised for the coeval cranium of the Marcel Loubens Cave, located not far from the Farneto rock shelter (Belcastro et al. 2021), and for other Italian crania dating back to Neolithic times (Germanà and Fornaciari 1992; Formicola et al. 2012).
The lesions detected on the postcranial bones suggest that corpses were disarticulated and/or dismembered and cleaned from soft tissues (S1). Lesions in correspondence of the epiphyses are likely related to disarticulation. More in general, the lesions seem to have been produced by different tools and in different manners. In particular, where sharp force lesions are associated with crushing (some vertebrae, ribs; Fig. 10a; scapula and some long bone epiphyses), it is possible that a blunted blade acted where cancellous bone is covered by a thin layer of cortical bone. Another possibility is that the presence of soft tissues protected in some way the bone from the blade, causing crushing instead of clear cuts during corpse treatment (cf. Shipman and Rose 1983). As regards thin cut marks, these suggest the use of sharp blades during cleaning practices (Fig. 9g, h).
A few elements present typical fresh bone fractures (Fig. 11; S1). In one case (the complete right tibia C7-AMH-223; Fig. 11a, b; S1), a butterfly fracture has been observed, which is considered a typical peri mortem fracture (Cappella et al. 2014; Reber and Simmons 2015). In addition, the femoral diaphysis fragment C5-BL1-SN (Fig. 11c, d) and the bone flake C5-AMH-79 (Fig. 11e) present a pattern of features (spiral outline, smooth fracture surface and acute fracture angle) considered indicative of fresh bone breakage. Moreover, the presence of peeling (Fig. 11f) or spalling (Fig. 11c, d) of the cortical surface in correspondence of bone breakages is reported in literature as indicative of peri mortem fractures (Andrews and Fernández-Jalvo 2012; Pickering et al. 2013; Knüsel and Robb 2016), therefore possibly occurred during corpse treatment, even if also other taphonomic factors may produce similar traces, as it has been noticed as a result of Ursus arctos gnawing and scavenging activity, leaving on the bones fracturing, peeling, crenulation, tooth pitting and scoring (Arilla et al. 2014).
However, some other fractures present typical fresh bone features together with dry or mineralised bone breakage characteristics, making their interpretation doubtful (S1). In the frame of mortuary corpse treatment, intentional fresh bone breakages are more likely related to dismembering practices. Thus, the contextual presence of sharp force lesions interpreted as such reinforces the hypothesis that interventions on cadavers were at least partially responsible for those bone fractures as well. Considering the instability of the karstic environment of the site, post mortem accidental events could account for further damage.
4.5 The Farneto rock shelter funerary practices in the framework of Late Neolithic/Early Eneolithic Central and Northern Italy
Due to the circumstances of the discovery and the retrieval of the human skeletal remains from the Farneto rock shelter (Fantini 1959, 1969; Busi 2018), it is not possible to establish which type of burial or deposition (e.g., primary vs. secondary, collective) the individuals originally had. The most part of the skeletal remains are fragmented and they appear commingled and chaotically disposed also in those sediment blocks that keep together some fragments still today (Fig. 2c, d). The archaeological objects, out of the original context, cannot be interpreted as grave goods because of their uncertain origin and collocation. In any case, given the chronology of the skeletal assemblage attributable to a final phase of the Neolithic and an early phase of the Eneolithic period, grave goods were more probably represented by the ornamental elements found in the site (Fig. 2a), rather than the pottery objects which are in all likelihood mostly datable to subsequent periods of frequentation (Nenzioni and Lenzi 2022).
Given the conformation and nature of the deposit (Fig. 1a, b), it is possible, as L. Fantini originally thought (Fantini 1959, 1969), that the materials were probably originating from a higher place, a natural cavity likely used for funerary purposes, slipped down during ancient times and in the most recent years due to mudslides and rockfalls. Due to these events, which are typical of the karstic environment characterizing the ‘Parco dei Gessi Bolognesi e Calanchi dell’Abbadessa’ (Grandi and Pisani 2017; Pisani et al. 2019; Pisani 2022), the actual front of the Farneto hill facing the Zena stream has retreated some tens of meters compared to prehistoric times (Nenzioni and Lenzi 2022). The occurrence of this kind of soil movements in subsequent times can easily explain the main characteristics of the most part of the fracture margins, that is compatible with dry or mineralised bone fractures.
The analysis of peri mortem lesions suggests that the individuals from the Farneto rock shelter were subjected to disarticulation practices and intentional cleaning from soft tissues. Although the original type of deposition cannot be ascertained, corpses manipulation may be consistent with the original presence of secondary burials. This evidence can be related to our knowledge about Eneolithic funerary practices in Central and Northern Italy (Cocchi Genick 2009, 2014), and in Emilia Romagna in particular (Miari 2013, 2018; Belcastro et al. 2021; Miari et al. 2022).
Eneolithic funerary contexts in Emilia Romagna are represented by both necropolises/isolated pit graves and burials inside natural cavities (Cardarelli 1992; Miari 2013, 2018; Cocchi Genick 2014). Thanks to the results of the radiometric analyses, the obtained absolute dates from the Farneto rock shelter materials demonstrate an early use of natural cavities for funerary purposes in the central area of the region, starting from a final phase of the Neolithic period. This early frequentation is also confirmed by other absolute dates recently obtained at the MPLD in San Lazzaro di Savena, regarding some materials from the surrounding areas of the ‘Gessi della Croara’ (Nenzioni and Lenzi 2022). They are two human specimens from ‘Cava I.E.C.M.E.’ and ‘Grotta dell’Ossobuco’, respectively dating back to 3800 − 3642 cal BC and 3475 − 3370 cal BC. For this period, the available comparisons in Italy are really few. An example is represented by the site of ‘Poggio di Spaccasasso’ (Grosseto, Tuscany, Central Italy), where the funerary context was used from the second quarter of the IV millennium BC throughout the entire Eneolithic period (Volante 2018; Volante and Pizziolo 2019). Another comparison may be found in the site of ‘Grotta Bella’ (Terni, Umbria, Central Italy), where the use of the cave for funerary purposes during the Late Neolithic was recently hypothesised in the so-called ‘Sala dello Scheletro’ (Larocca 2022).
With regard to the funerary rituals performed in the Eneolithic contexts in Emilia Romagna, both inside natural cavities and pit graves necropolises, the presence of dislocated, commingled or even isolated human remains is attested and this evidence is commonly interpreted as the result of intentional manipulation practices (Miari 2013, 2018; Cocchi Genick 2014; Cavazzuti 2018; Cavazzuti et al. 2020).
As concerns natural cavities, the use of these contexts as secondary burials places during the Eneolithic period in Emilia Romagna is already well known. In the XIX century, Don Gaetano Chierici first discovered the presence of human remains in the ‘Tana della Mussina’ (Reggio Emilia), where most skeletons appeared disarticulated and some of them were partially burnt. The context was interpreted as a collective burial that was frequented at various times (Tirabassi 2013; Tirabassi and Valzolgher 2018). In the Re Tiberio Cave (Ravenna) collective burials were present as well, hosting disarticulated and displaced bones. Here, the almost complete absence of crania was noteworthy, indicating the existence of an intentional selection of the bones (Miari 2013; Cavazzuti 2018). Moving to pit graves necropolises or contexts, manipulation practices of the crania and of the upper part of the skeleton are attested in the necropolis of ‘Celletta dei Passeri’ (Forlì-Cesena; Miari et al. 2017) and in the context of ‘Fornace Cappuccini’ (Ravenna; Antoniazzi et al. 1990; Giusberti et al. 1997).
This kind of funerary practices, which involve dislocation and commingling of human remains, have been interpreted as part of a cult of the ancestors, where the individual is depersonalised through fragmentation and confusion with other ones, aimed at reinforcing social cohesion and collective memory (for Central and Northern Italy cf. Conti et al. 2006; Cocchi Genick 2009, 2014; Miari 2013, 2018; Todaro and Girella 2013).
Besides any possible ritual interpretation, the available literature regarding these funerary contexts mostly describes practices of selection, displacement and commingling of human remains as post mortem interventions performed on skeletonised remains. On the contrary, as regards peri mortem treatment of the corpses, there is not much evidence from Neo/Eneolithic Italian funerary contexts. This can be due to the absence of available documentation and anthropological examination for the sites excavated during the last centuries. In fact, most recent analyses are radically changing this scenario. The taphonomic and anthropological analyses carried out on the isolated Early Eneolithic cranium recently retrieved in the Marcel Loubens Cave allowed stating that some peculiar funerary practices, such as cleaning from soft tissues, were performed on the individual immediately after death, i.e., peri mortem (Belcastro et al. 2021). Similar results come from the recent revision of the human skeletal remains from the ‘Tana della Mussina’, where it was possible to detect the presence of some peri mortem cut marks on a mandible (Cavazzuti et al. 2020). A recently revised and still under study cranium from ‘Fornace Cappuccini’ provided some evidence of peri mortem lesions, interpreted as scrape marks (Amitrano 2018; Belcastro et al. 2021).
In this framework, the results of the anthropological revision of the human remains from the Farneto rock shelter hereby presented can confirm the existence of intentional cleaning practices and dismembering of corpses soon after death, in a peri mortem period. While a particular predilection for crania has been often highlighted by literature so far (Miari 2013; Miari et al. 2017; Cavazzuti 2018; Cavazzuti et al. 2020; Belcastro et al. 2021), the careful study of the whole Farneto rock shelter skeletal remains assemblage also allowed to detect traces of disarticulation and scarnification not only on cranial districts, but also on postcranial bones.