Our detailed analysis provides new information on ornaments and the ways in which they were likely used for prehistoric infants. This multi-disciplinary research combined a wide array of methods (GIS, micro-CT, photogrammetry, use-wear and residue analysis, experiments) to provide as much information as possible about an extremely rare discovery. To ensure the robustness of our analyses, we compared the results obtained from different methods, such as our visual analyses of the GIS data and the photogrammetry model. This allowed us to ascertain the accuracy of our interpretations of the relative position of ornaments and human remains. We also calculated the angles and the roughness around the perforations of C. rustica to complement our use-wear analyses. Interestingly, the results of the roughness and summed angle analyses aligned well with the results of our qualitative microscopic analyses, which demonstrate the reliability of the optical assessment for measuring use-wear intensity. In addition, the computation of average roughness allowed us to explore potential correlations between it and other values such as the size of the perforation area, as others have hypothesized that bigger perforations might indicate longer use (Taborin, 1993). Here, the statistically significant negative correlation between roughness and perforation size supports this hypothesis along with previous research that noticed a similar pattern (Perlès, 2018). This suggests that wider perforation may be a sign of longer wear. Interestingly, average roughness correlated also significantly with thickness, which may be a sign that repeated usage and friction tend to round thicker shells’ perforation edges, whereas it will instead break thinner shells’ edges, which will then remain relatively fresher. This interpretation needs to be explored further using experimentation.
In the discussion below, we first place the ornaments within their social context. Then, we summarize the characteristics of the burial, and finally refer to ethnographic research to support our interpretation of how the beads were used by the infant and its community, in life and in death.
The Use of Columbella rustica
Columbella rustica lives on submerged rocky shores in warm waters and is commonly found along modern Mediterranean shores (Bertolini et al., 2016; Cartonnet, 1991; Cristiani, 2012; Taborin, 1993). As Arma Veirana is located < 20 km from the present-day Mediterranean coast, it is possible that the shells were collected locally from the nearest source (i.e., the beach in Albenga); however, further studies are needed to confirm or refute this possibility.
Ornaments made from this species have been found in multiple European sites, spanning the Aurignacian to the Neolithic (e.g., Álvarez-Fernández, 2010; Biagi et al., 1987; Taborin, 1993) and it was the most popular taxon for shell ornaments in the Mediterranean region during the Mesolithic (e.g., Cristiani et al., 2014; Cvitkušić, 2017; Newell, 1990; Perlès & Vanhaeren, 2010; Stiner, 2014). Columbellae have been found in several Mesolithic sites from the Ebro Valley of the Iberian peninsula (e.g., Álvarez-Fernández, 2006; Martínez-Moreno et al., 2010) to the Balkans (e.g., Benghiat et al., 2009; Cristiani et al., 2014; Cristiani & Borić, 2017), suggesting the presence of a wide-ranging and intricate network of social relationships connecting those southern European regions. In fact, here we argue that the groups who buried the infant at Arma Veirana were likely part of this social network, as the columbellae found in the burial conform to Early Mesolithic ornamental norms through their taxonomy, chaîne opératoire, ochre coverage, and use-wear patterns, which are detailed below.
While their spatial distribution is extensive, columbellae are often found in relatively small numbers in Early Mesolithic sites across the Italian peninsula and around the Adriatic Sea (e.g., Álvarez-Fernández, 2006; Mussi, 2001; Taborin, 1993). Only a few sites have yielded more than 40, including Vlakno Cave (n = 352), Vela Spila (n = 338), and Pupićina Cave (n = 94) in Croatia (Cristiani et al., 2014; Cvitkušić, 2017), Franchthi cave in Greece (n = 141 in the Lower Mesolithic, Perlès, 2018), and Grotta di Pozzo (n = 45, Brunelli et al., 2016), Grotta Continenza (n = 73, Colombo & Serradimigni, 2015), Romagnano Loc III (n = 76, Borrello & Dalmeri, 2004), and Grotta della Serratura (n ~ 500, Mussi, 2001). The Croatian sites of Vlakno and Vela Spila also have yielded many unperforated columbellae and about half of the shells found at Serratura were unperforated, suggesting that the beach near those sites may have been an important source for this raw material for ornament making. Therefore, while the people of Arma Veirana may have collected their shells within Liguria, it is also possible that they did so through some connections with people living in the Adriatic or the Tyrrhenian region. Future research will help us answer this question.
The average size of modern columbellae ranges between 11 and 14 mm in length, depending on the source. The size of the specimen found at Veirana falls on the higher part of that range, a pattern that has been observed at other sites (e.g. Perlès, 2018). The bigger size of the shells used for the burial can be explained by the fact that bigger shells are easier to perforate (Perlès, 2018). However, it should be noted that the fact that those shells, buried with an infant, are not smaller than specimens found in other archaeological assemblages, which suggests that they were not a “child” version of adult ornaments, as hypothesized at other children burials (e.g., Vanhaeren & d’Errico, 2001; White, 1999).
When used as ornaments, columbellae are usually perforated on the side opposite their natural opening. Perforations are often large—although size varies per site—and quadrangular or subcircular in shape with straight edges or inwardly expanded cone fractures (e.g., Benghiat et al., 2009; Dalmeri & Fiocchi, 1998). The upper areas of the perforations are sometimes chipped, which has been identified as a likely result of using direct percussion to create the perforation (Bertolini et al., 2016; Cristiani et al., 2014, 2020). The Arma Veirana assemblage fits this description perfectly.
Direct percussion is a perforation method commonly associated with C. rustica (e.g., Álvarez-Fernández, 2006; Benghiat et al., 2009; Micheli, 2004; Stiner et al., 2013; Perlès, 2018), although some have hypothesized that indirect percussion could have been an alternative technique used to perforate such shells (Cristiani et al., 2014, 2020; Perlès, 2018). Because most shells from Arma Veirana are worn on the top part of the perforation, this prevented us from identifying the required diagnostic crushing on a large portion of the ornaments. To remedy this, we are currently performing more systematic experiments to confirm our interpretation of the analyzed assemblage.
The shells found in the Arma Veirana burial exhibit only minimal ochre staining and do not show black staining, which is common in Early Mesolithic burials (Grünberg, 2016). At Arma Veirana, those low levels suggest that the shells themselves were likely not colored. Instead, we suggest that one of the following alternative explanations best explain the origin of the ochre: 1) the ochre specks transferred onto the shells from the decorated wrap or an ochred string; 2) the ochre traces are remnants of the arrangements they were used in prior to being sewn on the wrap; or 3) only a select few shells were colored to create a striking visual contrast between beads of different colors. Modern C. rustica found on the Mediterranean shore are naturally brown to red or gray (Pauc & Pauc, 2006), but shoreline activity bleaches the shells over time, leading to white specimens (e.g., Benghiat et al., 2009; Perlès, 2018). Most of the Arma Veirana columbellae show signs that they were collected from thanatocenosis on the beach (sponge and bioerosion marks); therefore, some of them may have been white at the time of collection. While a few archaeological specimens still show a faint outline of their original colors, it is possible that some specimens were selected for their color or that bleached shells were subsequently recolored. However, it does not appear that all the burial shells were intentionally colored red, given that coloring the shells would have left more residues within the spires than what has been identified in the assemblage.
The type and location of the use-wear documented on Arma Veirana columbellae also fit the pattern found in other archaeological assemblages. At most Mesolithic sites, perforated C. rustica are well worn, with rounding found especially at the top of their perforation and on the inside part of their outer lip (e.g., Bertolini et al., 2016; Cristiani & Borić, 2017; Perlès, 2018), which is concordant with what is seen in the Veirana assemblage. The long use life of C. rustica beads seen at most sites was only possible because of the shells’ durability, which may have been one of the reasons why they were selected in the first place (see Stiner, 2014). Some have hypothesized that the shells were attached through the lips so that the ventral part would show (Bertolini et al., 2016; Cristiani, 2012), but our 3D reconstruction of the burial shows that the visible part of the shells might instead have been the side between the natural and anthropogenic openings, a pattern also identified in the Neolithic burial of Avignon (Zemour et al., 2017).
The Glycymeris Pendants
While the columbellae found in the burial fit well within the ornamental norms of the Early Mesolithic documented above, the fit of the four Glycymeris perforated pendants found in this burial remains mysterious, as similar ornaments are extremely rare in the European prehistoric record. Glycymeris shells have been very popular throughout prehistory; they were found in Middle Paleolithic sites (Zilhão et al., 2010) as well as Middle-Eastern and African Middle Stone Age sites more than 90,000 years old (Bar-Yosef Mayer et al., 2009; Vanhaeren et al., 2013). However, in almost all archaeological instances, Glycymeris shells were used as complete valves, sometimes strung through their perforated umbo or the distal part of their valves were used as half-moon pendants, likely attached using a string around their middle (e.g., Álvarez-Fernández, 2010; Anfossi, 1972; Bar-Yosef Mayer et al., 2009; Borić & Cristiani, 2019; Borrello, 2004; Micheli, 2004; Peresani et al., 2019; Rivière, 1887; Tripković et al., 2016; Vanhaeren & d’Errico, 2011). In European sites, up to recently, single Glycymeris perforations were always found on the umbo (Taborin, 1993, p. 284).
An extensive literature review shows that perforated pendants with a shape similar to the ones found at Arma Veirana have been found further West in Liguria at the Balzi Rossi (Mussi, 2001), in the Late Epigravettian in northeastern Italy at Riparo Tagliente (Accorsi Benini, 1972; Cilli et al., 2006) and Riparo Dalmeri (Borrello & Dalmeri, 2004), and in the Epipaleolithic of l’Aven des Iboussière in France (d’Errico & Vanhaeren, 2000); however, in those cases, the pendants were made on bone or stone rather than on shell. Perforated Glycymeris fragments with a similar shape have only been found in a handful of contexts: the Late Epigravettian of Grotta del Romito—here, as a pendant in making as the perforation was never finished (Martini et al., 2007)—the Capsian of the Grotte du Polygone, in Algeria (Cornaggia & Girod, 1965), and in the Bronze Age Sardinian occupations at Padru Jossu (Borrello, 2004) and Anghelu Ruju (Puddu, 2014). However, all those specimens are either older or much younger than those recovered at Arma Veirana, thus preventing us from linking them directly to the group who buried the infant.
While the shape of the Arma Veirana pendants is unique for their date, ornaments made on half-moon-shaped Glycymeris fragments have been found in association with Late Epigravettian burials in the nearby site of Arene Candide (~ 35 km from Arma Veirana). However, these pendants differed in their method of suspension; they were hung with a string wrapped around their middle (Cardini, 1980), whereas the Arma Veirana pieces were all hung through their perforation. Moreover, the pieces found at Arene Candide were thinner than the ones found in direct association with the burial, which suggests that the Veirana pieces came from bigger shells, despite being buried with a younger individual, an interesting pattern that is discussed below. The pieces found at Arene Candide—including the > 200 still unpublished specimens mentioned by Cardini (1946)—are characterized as semi-lunate fragments. Similar semi-lunates have been documented in multiple Late Epigravettian sites throughout southern Europe and Northern Africa (Cornaggia & Girod, 1965), which suggests a certain level of connectivity between all those locations. However, without other contemporaneous similar specimens, it is difficult to understand how the Arma Veirana pendants fit within the Early Mesolithic cultural context. To this date, they are certainly one of a kind.
As mentioned in the introduction, personal ornaments are thought to communicate identity, gender, or status (e.g., Cristiani et al., 2014; Cvitkušić, 2017; Kuhn & Stiner, 2007), as well as to help maintain social relationships (Wiessner, 1982), and protect from evil (Miller, 2009). In the perspective of beads as a communication device, the taxa combination seen in the assemblage studied here suggests that the group who buried the infant maintained strong ties with the highly connected population of Early Mesolithic southern Europe (C. rustica), while keeping a certain individuality (Glycymeris sp.).
Reconstruction of the Burial Process
The analysis of the perforated shells combined with the taphonomy of the burial suggests that the infant was wrapped in a piece of textile, fur, or hide that was decorated with several arrangements of perforated shells and pendants. The wrapped infant was deposited in a small pit with her legs folded over her abdomen and was covered with sediment.
The 3D position of the beads found near the abdomen suggests that they were placed between the abdomen and folded legs. Combined with our interpretation that the infant was wrapped, it is possible that those shells were sown on the wrap itself or on a piece of clothing such as undergarments, which was then covered by the wrap. Unfortunately, we cannot identify which of those possibilities is the correct one. However, we can assume that, by folding the legs of the infant over the abdomen, the group who buried the infant likely hid some of those ornaments, thus removing the beads’ aesthetics from the burial. This suggests that the beads were not buried with the infant to serve as funerary decoration, but were rather part of a decorated garment or baby sling that was likely used during the infant’s life (see Miller, 2009), similarly to other burials of children that have yielded abundant ornaments (Boric et al., 2014; Cardini, 1980; Cristiani & Borić, 2017; Henry-Gambier, 2001; Vanhaeren & d’Errico, 2001). This is different from the burial of the infants at Krems-Wachtberg, who were interred with beads that did not bear any use-wear and were thus likely made specifically to serve as grave goods for the burial (Teschler-Nicola et al., 2020).
The 3D position of the beads curved around the humerus and next to the cranium suggests strongly that the shells were sewn on the wrap used to bury the infant. Most of those beads were aligned tightly close together, which suggest that the shells were likely strung together and the string was then sewn on the wrap. Furthermore, the widespread location of ochre residues on several shells suggests that the strings used or the wrap itself were likely treated with some ochre, which transferred to both the internal and external parts of the shells, as inferred for other archaeological assemblages (Rigaud, 2011; Velliky et al., 2018). Ochre was found in higher quantities on shells located near the head and the abdomen, which suggests that ochre may have been preferentially applied near the head and the abdomen as shown in the artistic reconstruction of this burial (Fig. 24). However, as the human remains were not tinted red nor the sediment around them, it is unlikely that big quantities of ochre were applied to those regions during the burial. Alternatively, the higher level of ochre found on shells located near the head could result from a different type of stringing and ochre arrangement or from their longer usage as they were significantly more worn than the ones found near other body parts. It is possible that those beads accumulated ochre throughout their lengthy use in different arrangements when worn by other people.
Fig. 24 Artistic reconstruction of the Arma Veirana burial (drawing by Mauro Cutrona) Full size image
This brings us to our interpretation that most ornaments found in the burial had been used for an extended period of time by the group who buried the infant before the infant was even born. This interpretation is supported by the extensive use-wear recorded on most shells, the numerous rounded fractures found on the siphon canal of columbellae, and the location of facets on a few of them.
Vanhaeren and d’Errico (2011, p. 69) have argued that the intensity of use-wear traces on ornaments can be an indicator of how long they were used. In the Arma Veirana burial, the extensive degree of wear and low surface roughness measured on the columbellas could not have been produced during the short life of the buried infant (40–50 days). Therefore, the level of wear probably developed while the beads were used over time by other members of the community before being inherited by the infant during her short life.
This interpretation is supported by the location of worn facets found on a few columbellae, which were likely created through friction against another hard surface (Fig. 16). While friction between shells can produce such facets in other taxa (Vanhaeren et al., 2013), the position in which we found the facetted shells suggests that the facets were created before the beads were “given” to the infant, as they are located on surfaces that would not have been in contact with other shells. While we cannot reject the possibility that some hard objects made on organic matter were part of the arrangement when the infant was alive and did not preserve in the burial, the most parsimonious interpretation is that the facets were created while the shells were worn in different arrangements by other members of the group/family prior to being sewn on the wrap in which the infant was buried.
In addition to the facets and use-wear, the presence of numerous rounded fractures in the assemblage of columbellae suggests that the shells were used for an extended period of time after the fracture occurred. The sources of the fractures are unknown; they could be the result of the previous use of the ornaments in specific arrangements involving compression between shells or they could simply be collateral damage incurred during perforation. However, such fractures are relatively rare in contemporaneous assemblages of columbellae, which suggests that they were likely not produced by natural processes prior to being collected.
The pendants as well were likely worn by the community for a considerable amount of time before they were given to the infant. The location of use-wear on those pendants suggests that they were worn with their ventral side against the body, which is consistent with how they were found in the burial. This suggests a certain continuity in how those pendants were worn by the group and by the infant. Therefore, it is possible that the shells transferred to the baby in life or death may have acted as a link between her and her close relatives, a possibility we discuss further below.
Microscopic analyses show that the Arma Veirana pendants were collected as broken fragments on the beach. We cannot prove that the foragers who collected those fragments were looking for a specific taxon; however, it is possible that they were selecting for a specific color, shape, finish, smoothness, and thickness, as all pendants are roughly similar in size and finish. This in itself points to the symbolic aspect of these pendants, which means that people might have overlooked utilitarian concerns—thinner fragments would have been easier to perforate—in search for a certain subjective beauty or meaning. While most Arma Veirana pendants have natural red coloring on their ventral side, it is unlikely that they were selected for that characteristic because the use-wear patterns show that the ventral side was hidden from view. Instead, the pendants were possibly selected for their white dorsal color, especially as the low amount of ochre residues found on those fragments suggests that they were not intentionally colored red. In fact, the ochre traces found on the pendants are so sparse that it suggests that they could have come from an ochred string used to bind a previous arrangement or simply from contact with a few ochre specks in the sediment.
Excavation around the burial pit uncovered an eagle owl talon that had briefly been used as an ornament and left in a small pit adjacent to the burial, which was interpreted as an offering (Hodgkins et al., 2021). Pendant #3484 was found at the same elevation as the eagle owl talon (~ 6 cm directly above the burial) and may have been another offering placed there intentionally (as per Arias’s (2012) definition). The reasoning behind this interpretation is the following: (1) Identical pendants have not been found at any contemporaneous sites in the region and beyond, as we discussed above. Therefore, the coincidence of having one pendant directly above a buried individual interred with three identical pendants makes it very unlikely that the 4th pendant was unrelated to the three buried ones or that it was lost there by mistake. (2) This pendant was found immediately adjacent to the only large rock (> 10 cm in length) documented above the remains. Rocks of similar size or bigger were found at all elevations around the burial pit, suggesting that they are natural occurrences in the cave. However, they are completely absent from the 6 cm of sediment found immediately above the human remains (see S1 video). The fact that the pendant was found adjacent to a big rock and at the same elevation as the eagle-owl talon, therefore, suggests that the pendant was deposited on top of the burial fill after the funeral, and that it was then covered by natural accumulation of small roofspall and sediment over time. (3) This pendant was more altered than the three recovered in the pit, which could have been due to a longer exposure to atmospheric conditions that produce a characteristic white patina on certain bivalves (Manca, 2016). (4) We did not find signs of soil disturbance around the 4th pendant nor the land snails that were found in high quantities in the sections of the burial that had been disturbed by taphonomical processes (e.g., near and inside the cranium and around the ribs). All these combined suggest that the pendant was deposited intentionally on top of the covered burial. Interestingly, if our interpretation is correct, this would suggest that the burial remained undisturbed for a relatively long time, which may even suggest that the cave remained unoccupied during that time. More research is needed to confirm this interpretation, however.
Finally, the placement of the ornaments on the Arma Veirana infant body is interesting as it is focused on the region around the head and the abdomen, where shells also have more ochre traces than near other body parts. The head and abdomen are two body parts that are often decorated with ornaments in prehistoric burials. This is especially true for the abdomen/pelvis region in children, as seen in the children burials of Krems-Wachtberg, Grotta dei Fanciulli, Sungir, Bogebakken, and Vlasac (e.g., Cristiani et al., 2014; Cristiani & Borić, 2017; Grünberg, 2016; Henry-Gambier, 1995; Riel-Salvatore & Gravel-Miguel, 2013; Vang Petersen, 2016). As different body parts have different symbolic meanings in modern forager societies (Vanhaeren & d’Errico, 2011), this may suggest that these two lines were purposely sewn above the abdomen for a symbolic purpose such as to protect and strengthen their vital body parts (as seen in Miller, 2009). On the other hand, decorated headwear are common in Upper Paleolithic burials of individuals of all ages. Those have been interpreted as implements to broadcast social information, due to the high visibility of the head when viewed from a distance (Riel-Salvatore & Gravel-Miguel, 2013). Combined with the intense use-wear of the shells found near AVH-1’s head, this suggests that the shells used to convey social belonging may have been intentionally reused for as long as possible to retain and strengthen their symbolized social identity.
Therefore, AVH-1 was likely interred with ornaments composed of beads that had already been used by her community. Combined with our analysis, which showed that shells with very different levels of wear were found within single arrangements, this suggests that perforated shells were recycled on a regular basis and remixed into new arrangements. In addition, the fact that shells placed near the head are significantly more worn than others suggests that certain shells were specially selected to adorn that body part and thus may have had a special significance. In the next section, we refer to ethnographic research to explain why we believe that the wrap on which the beads were sewn was likely used as a baby sling while the infant was alive.
Exploring Community Intentions Surrounding the Burial
The high number of beads found in the AVH-1 burial is impressive. Up until recently, this would have been seen as a marker of high status. However, one ethno-anthropological research into the role of material culture in Indigenous societies reveals that in some modern Amazonian societies, forager groups perceive body decorations and ornaments as materializations of the parental care toward a child (Miller, 2009). During the first years of life, such care are crucial for the health of the child and in this context, ornaments represent a reflection and an extension of this care, and protection from evil. Not surprisingly, in those societies, infants and children are always well adorned. Among the beads that are used to decorate and protect their bodies, the majority are “second-hand” items, i.e., beads that have been donated by the parents, grandparents, and relatives as an act of care toward the child. Accordingly, ornaments play a key role in “building” children’s bodies through social relationships and protection from diseases. They hence become material evidence of the network of relationships linking the child to the members of the community, which are necessary for the child to become “human.”
In certain modern forager populations, such decorations are placed on baby carriers and slings (Vang Petersen, 2016). As archaeologists are increasingly discussing the possibility that baby carriers and slings were widely used in prehistory (Langley & Suddendorf, 2020; Nowell & Kurki, 2020; Suddendorf et al., 2020; Taylor, 2010), re-analyses of infant burials have hypothesized that some infants were, in fact, buried in such carriers. For example, grave 7 of La Vergne contains the remains of two adults next to the remains of a child that were placed in a box decorated with multiple perforated marine shells and animal teeth (Laporte & Dupont, 2019; Laporte et al., 2021) that could be interpreted as a carrier. Recent re-analysis of the infant of Abri Labattut provides another example, as this research shows that the infant was buried with multiple cowrie shell ornaments that were too big (~ 2.5 cm in length) to have been worn as jewelry or on the clothes of such a small individual, and which use-wear suggests that they had been attached to a fixed object, possibly a blanket, baby carrier, or other (Henry-Gambier et al., 2019, p. 199). Coincidentally, this is reminiscent of the Arma Veirana infant, who was buried with three Glycymeris pendants that were as big as the cowries of Labattut (mean ~ 3.3 cm in length vs. ~ 3.2 cm for Labattut) and were associated with an even younger individual (45–50 days old vs. < 10 months old for Labattut). As we tried to understand how the pendants would have fit on a piece of clothing used for a ~ 2-month-old infant, the possibility that those would have adorned a carrier seemed more logical and practical.
Moreover, as discussed in the introduction, recent research has shown that ornaments were likely used to produce sensorial experiences for the infants (Rainio & Mannermaa, 2014). Artifacts from Mesolithic and Neolithic burials have been interpreted as rattling ornaments (Larsson, 1984, p. 20; Rainio et al., 2021; Rainio & Mannermaa, 2014; Rainio & Tamboer, 2018). Based on these, we should not rule out the possibility that the pendants from AVH-1, standing out for their way of suspension, their dimension, and developed traces, might have produced a rhythmic sound when hitting each other.
Going even further, Vang Petersen hypothesizes that the ornaments used on carriers could have symbolized “amulets” used to protect the child. Moreover, he highlights that among the Pueblo Indians of the American Southwest, the carrier of a child who dies is burned with them to avoid passing over the “defective” amulets to someone else (Vang Petersen, 2016, p. 119).
While it is impossible to prove, we hypothesize that a similar scenario could explain the presence and 3D location of the ornaments found in association with the Arma Veirana infant. As mentioned above—and similarly to the documented use-wear on some of the ornaments associated with the child of La Vergne (Laporte et al., 2021)—the use-wear documented on the ornaments shows that they had been worn for long periods by members of the community prior to the burial (and indeed the birth) of the infant. Some of those beads could have previously been used for other baby carriers for infants who survived to adulthood and thus, whose ornaments could be passed on to the new baby girl. However, as she did not survive, it may have been deemed safer to bury her with her adorned carrier than to reuse the ornaments that failed to protect her in life. We believe that, given the ethnographic and archaeological examples detailed above, this hypothesis is possible.
Alternately —although, not mutually exclusive—the presence of the pendants in the burial could have also embodied the long-lasting connection to and care from parents, relatives and community members to the child. This could explain why the group decided to part with beads that had been curated for a considerable amount of time rather than use new shells they could have obtained from the nearby coast. In fact, this strongly suggests that the beads found here were more than just decorations, as shells used only as burial embellishment would not have needed to be durable (Benghiat et al., 2009). Finally, another possibility is that the adults intentionally parted with their own ornaments to adorn a ceremonial shroud, providing a link between the dead and the living, as was probably done in the Late Epigravettian of the Arene Candide with intentionally broken pebbles used as spatulas to decorate the deceased with ochre as part of larger funerary rituals (Gravel-Miguel et al., 2017). Distinguishing between all these possible interpretations will require more research.