Introduction

One of the most important events in human prehistory was the transition from a hunter-gatherer economy to an agricultural lifestyle. The nature of this process is still the subject of intense debate in archaeology, population genetics, and palaeoanthropology (Cavalli-Sforza et al. 1994; Guilaine 2003; Broushaki et al. 2016). The Neolithic period marked the emergence of the first sedentary societies and a major economic, social, architectural, technological, symbolic, and biological change (Verhoeven 2002; Byrd 2005). At the end of the Epipalaeolithic (c. 12,000–10,000 BP), which coincided with an optimal global warming period associated with increased precipitation (Bar-Matthews et al 1997, 1999; Byrd 2005), hunter-gatherer (HG) populations from the ‘Fertile Crescent’ (Mesopotamia, Taurus and Zagros Mountains, and Levant) underwent significant economic changes (Bar-Yosef 2005; Goring-Morris and Belfer-Cohen 2011; Zeder 2011), adopting more sedentary lifestyles and domesticating plants and animals (Zohary et al. 2012; Fuller et al. 2014). This new economic and demographic lifestyle was introduced into Europe by agricultural populations over the next five millennia, displacing or mixing with local Mesolithic HGs (Price 2000) and creating a specific genetic scenario (Yang and Fu 2018).

During the Pre-Pottery Neolithic A –PPNA (12,000–10,500 years BP), the first small sedentary hamlets (approximately 1 ha and 100 inhabitants) appeared (Byrd 2005). During the subsequent Middle and Late Pre-Pottery Neolithic B (PPNB) (10,000–8000 years BP), there was a significant increase in the size of the villages: up to 30 ha and an oscillating population of around one thousand individuals. The socioeconomic swerve and increase in population size associated with the Neolithic led to an increase in social complexity (Watkins 2008; Belfer-Cohen and Bar-Yosef 2000). The first farming communities had to adopt new long-term co-residence patterns and develop ways of regulating access to the land and transmission of the property, while coping with the stress of having to live in highly populated areas. At the end of the PPNB, most archaeological sites were abandoned as part of a phenomenon generally known as the “PPNB collapse”, resulting from a climatic deterioration, overexploitation of local resources, or a reorganization of the settlement system and dispersal from aggregate villages to small and dispersed settlements across the south-central Levant (Rollefson and Kohler-Rollefson 1989; Kuijt 2000).

Precisely during this period, the initial manipulation of clay occurred (Schmandt-Besserat 1977). During the Pre-pottery Neolithic (PPN) period, there was a wide range of versatility in the use of objects made of clay for different purposes, such as architectural (Bar-Yosef and Gopher 1997; Kenyon 1981; Finlayson et al. 2011; Kuijt and Finlayson 2009), or producing relatively small portable objects such as clay human and animal figurines, containers, and geometric tokens, as well as the decoration of human skulls (Schmandt-Besserat 1977; Moor 1995; Wengrow 1998; Nakamura and Meskell 2004; Özdoğan 2009; Gibbs 2015). These first manifestations of clay processing represent “an unprecedented exploration of its tactile properties” (Verhoeven 2002). During PPNB (and, by extension, during PN), the presence of figurines is typical in funeral practices (Verhoeven 2002). These mud/clay plastic properties (as well as those of other materials such as wax and resin) allowed, as a result of their manipulation and processing, the imprint of the epidermal or papillary ridges and the interpapillary grooves of the fingers on the object, forming what is known as finger dermatoglyphs (Králík and Nejman 2007). Palaeodermatoglyphs are specifically those fingerprints recovered and identified in an archaeological context. There is palaeodermatoglyphic evidence before the Neolithic, such as fingerprints on adhesive used by Homo neanderthalensis (Koller et al. 2001); the Venus of Dolní Věstonice I, a fire clay representation of a robust woman dated 25,000 BP (Upper Paleolithic) in Moravia (Králík and Novotný 2003); and the fired clay objects (worked and unworked) from the site of Pavlov I (Upper Paleolithic, Pavlovian culture, circa 25–26,700 years BP) (Králík and Novotný 2005; Králík et al. 2008a, b).

Although most of the archaeological evidence of palaeodermatoglyphs comes precisely from the remains of pottery, bricks, and oil lamps; figurines; statues, cuneiform tablets; daub fragments; and even Chinese loan contracts with the fingerprints of witnesses to all party members in contracts—made of clay (Wilton 1938; Cummins 1941; Rees-Jones 1978; Kamp et al. 1999; Lloyd 1999; Berry and Stoney 2001; Branigan et al. 2002; Stinson 2004; Chapman and Gaydarska 2006; Hruby 2007; Králík and Einwögerer 2010; Mara et al. 2010; Floris 2012; Donida-Labati et al. 2012; Lapp and Nicoli 2014; Sanders 2015; Minguez et al. 2016; Bennison-Chapman and Hager 2018; Lichtenberger and Moran 2018; Fowler et al. 2019, 2020; Kantner et al. 2019; Papadopoulos et al. 2019; Strozyk et al. 2019; Coban et al., 2020; Sero et al. 2021), ancient fingerprints have also been identified on other materials such as organic materials used as an adhesive (Koller et al. 2001), wax of a fifteenth century seal (Strouhal 1999), and on a solidified cream in a Roman metal box discovered in London (Evershed et al. 2004; Králík and Nejman 2007), as well as ancient written texts (Bartsocas 1982) and corrosion products of metal artefacts (Peška et al. 2006) and painted motifs in the site of Los Machos rock shelter (Southern Iberia) (Martínez-Sevilla et al. 2020). Paleodermatoglyphics aids researchers in reconstructing the biological profile—sex and age estimation (David 1981; Acree 1999; Králík and Novotný 2003)—of past craft producers (Králík et al. 2008a, b; Sanders 2015; Minguez et al. 2016; Faux-Campbell 2018; Martínez-Sevilla et al. 2020), which allows for deepening knowledge of social organization and labour specialization of past societies (Králík et al. 2008a, b; Kamp et al. 1999; Sanders 2014, 2015; Bennison-Chapman and Hager 2018; Lichtenberger and Moran 2018; Papadopoulos et al. 2019; Fowler et al. 2019, 2020; Kantner et al. 2019; Strozyk et al. 2019). The latest biological and paleodemographic studies have stated that there is a distinction between ridge density and ridge breadth between male and female individuals and their ages (Králík and Novotný 2003; Králík and Nejman 2007; Nayak et al. 2010a,b; Krishan et al. 2013). Therefore, the finding of fingerprint traces on ceramics and/or utensils can shed light on the organisation of the labour.

In the specific case of the Neolithic, few studies have studied or described the existence of dermatoglyphs, all of which are imprinted on clay objects (mostly bricks and pottery), and most of which correspond to isolated partial fingerprints (Floris 2012; Bennison-Chapman 2014; Bennison-Chapman and Hager 2018). Bricks made of clay with involuntary fingerprints from the workers who produced them have been recovered from the Neolithic site of Jericho (6–7000 BCE) (Berry and Stoney 2001). Nevertheless, beyond the anecdotal description of the presence of palaeofingerprints on these Neolithic bricks, we do not have more information that allows us to ascertain or make inferences about the biological profiles of the workers who made these bricks. Fingerprints have also been documented in 6500–7000 BP pottery fragments of a Neolithic site (Cseplák, 1982; Gyenis 2000). Synchronically, in the Neolithic site of Těšetice-Kyjovice (approx. 6700 BP; Moravian Painted Ware culture) from the Chech Republic, 2 out of 7 pots had complete fingerprints, probably belonging to children (Králík and Hložek 2007). Finally, in the Neolithic site of Koutroulou Magoula, located in Fthiotida, Central Greece, dating to the first two centuries of the sixth millennium BC, almost 40 human fingerprints have been recovered, some of which have been erased on purpose (Papadopoulos et al. 2019).

This study aims to investigate the biological profile of the crafter of the animal-shaped figurine recovered at Tell Halula (a PPNB archaeological site from the Middle Euphrates, Syria) through the analysis of the breadth of the ridges of the fingerprints present on its surface. Tell Halula is located in the Euphrates Valley, specifically 20 km east of the city of Manbij, Syria (36.416667°N, 38.166667°E) and reaches 8 hectares at the time of greatest concentration. It is located at the confluence of three different ecological zones and offers a great variety of resources: the river, steppe, and low mountains. The site is a slightly circular Tell (360 × 300 m), with some 11 m of archaeological deposits (Molist 1998). One of the peculiarities of Tell Halula is its long period of occupation, from the middle PPNB to the recent PPNB, and which spatial organization would have been provided with ordered rectangular-shaped multicellular residential buildings (with 3–5 rooms with mud-brick walls), arranged along a north–south axis, with small spaces for circulation. In addition, the site has some of the earliest evidence of monumental and/or collective architecture (Molist 2001). The burials were found below the main floor at the house entrance, with an average of 10 burials for each house (Guerrero 2006).

Material

The assemblage of clay figurines from the Tell Halula site consists of 45 figurines, 38 of which were in the site’s facilities (Manbij, Syria), while the other 7 were exhibited in the Aleppo museum (Ayobi 2013, 2014). The figurines were classified according to their form into zoomorphic, anthropomorphic, and geometric categories by Rania Ayobi (2013, 2014). Anthropomorphic figurines are relatively scarce in Tell Halula since they only represent 11% of the total figurines among which a female human representation with a rather worn and irregular state of conservation can be highlighted (Ayobi 2013). The bibliography referring to the zoomorphic figurines from the site of Tell Halula is scarce (Molist et al. 1996; Ayobi 2013, 2014). In the initial publication of the excavation campaigns of 1991–1992, reference is made to the discovery of zoomorphic figurines (bovids and ovicaprids), in which it is mentioned that they would come mostly from the levels of sector II and cell IA (Molist et al. 1996). They were in a poor state of preservation and quite fragmented (Molist et al. 1996, p. 130), so a consolidation process with Paraloid B-72© was necessary (González and Pérez 2013). Finally, in addition, some cracks will be reinforced with prefabricated putty with clayey sediment and acrylic emulsion (González and Pérez 2013).

The figurine analysed in the present study, HL72 (Ayobi 2013, 2014), is a zoomorphic figure which is part of the second set of figurines exhibited in the Museum of Aleppo. Unfortunately, the exact context or burial of the piece HL72 and its exact dating is unknown, beyond the fact that it would be dated between the MPPNB and the beginning of the LPPNB (Ayobi 2013, p. 60). Anyway, in the site of Tell Halula, the figurines have been found in funerary contexts inside the burials, as well as other objects including beads, shells, flint and obsidian tools, macrolithic tools, bone tools, etc. During the early Neolithic, individuals in Tell Halula were systematically buried inside the household areas. These graves were frequently located under the floor of the front section of the house’s main room. Tools and burial goods, such as figurines, are frequently found around the lower limbs or feet, located at the bottom of the tomb, in unstable positions (Molist et al. 2009; Kuijt et al. 2011; Ortiz López 2016; Ortiz et al. 2013; Alarashi et al. 2018).

All zoomorphic figurines recovered from the Tell Halula site are medium-sized, representing quadruped animals. Most of them display cylindrical necks and, in some well-preserved cases, even horns, as well as cylindrical bodies formed by flanks. In most cases, despite poor preservation, the tail or remnants of the tail can be observed. Concerning the extremities, they are commonly short and cylindrical; in some cases, the front ones seem to be longer than the rear ones. Precisely in one of these animal figurines (HL72), the presence of several fragmented and/or partial fingerprints has been identified. This figurine represents a complete bovine (67 mm long by 35 mm high) with a broad, rounded conical snout (Figs. 1 and 2). In the neck region, it has a crest that continues to the lower part of its front legs (Ayobi 2013). The figure has sub-horizontal and broken rounded horns, below which small ears are situated. The neck is broad and horizontal, with a straight back and a small dorsal point (or pronunciation) near the neck. The flanks (i.e. the belly) are slightly swollen and concave. The animal’s legs are rounded and conical, presumably formed by the pressure of the toes, as are the hind legs. All four legs are bent and face backwards. Given the position and section of the horns, it has been proposed that it probably represented a bovine in movement (Ayobi 2013). The three fingerprints are located on the ventral part (Fig. 2). The first is a complete fingerprint located on the front legs, probably because of the manipulation and compression of the clay to the legs. The second complete fingerprint is located in the hind legs and, as with the former fingerprint, it is probably a printing product of the work of clay to shape the hind legs. Finally, a third partial fingerprint is located in a slightly more central position as a result of the manipulation of this object.

Fig. 1
figure 1

Detail of the Tell Halula figurine in both sides (from Ayobi 2013; Creative Commons CC-BY-NC-ND)

Full size image
Fig. 2
figure 2

Detail of the ventral region of the zoomorphic figurine with the fingerprints overprinted (image by the coauthor MM)

Full size image

Methods

The analysis of the fingerprints was done from the original digital photographs (Ayobi 2013). Although several studies have used 3D scanning technology to obtain fingerprint images (Dong and Chantler 2004; Counts et al. 2016; Delvaux et al. 2017; Papadopoulos et al. 2019), in the present study, this was not possible due to the development of the war in Syria, since the figurine is hosted at the Museum of Aleppo (Syria). The photography was originally made with a zenithal plane of the ventral/lower part of the figurine where the putative fingerprints are located. The calibration of the image was done digitally using Adobe Photoshop CC 2021 (22.3.1) software, using the known total length of the figurine initially measured manually by Ayobi (2013). Using the same software, the photographic parameters of light and contrast have been changed by applying different filters: contrast enhancement, colour inversion, and brightness reduction to obtain a better view of the figurine details. The two complete fingerprints were analysed.

Fingerprint determination

The first step was to determine whether the marks imprinted on the object were indeed fingerprints (Fig. 3) by identifying specific print minutiae or commonly called Galton’s details, which are variations, irregularities of direction, discontinuities, and branching (Králík et al. 2002). Although no systematic studies have been published on the variability of the minutiae (Gutiérrez et al. 2007), minutiae are nowadays used by law enforcement agencies for the identification of persons. In the present study, we have followed the typology of minutiae used by the Spanish Scientific Police (Antón-Barberá and De Luis 1993).

Fig. 3
figure 3

Different types of minutiae used by the Spanish Scientific Police (Antón-Barberá and De Luis 1993; Gutiérrez et al. 2007)

Full size image

Age estimation: mean epidermal ridge breadth (MRB)

The breath of the epidermal ridge (ERB) is of biological, medical, and genetic interest (David 1981). The metric study of epidermal ridge breadth enables the determination of the biological profile of the person who left the fingerprint (Králík and Nejman 2007; Gutiérrez-Redomero et al. 2011; Mundorff et al. 2014; Dhall and Kapoor 2016; Rivaldería et al. 2016). However, when measuring the ERB in archaeological contexts, some considerations must be taken into account. The first is precisely how ERB is defined and, thus, computed. In a fingerprint directly from the epidermis, the ERB is computed as the distance between the centre of one epidermal furrow and the centre of the next one along a line at right angles in the direction of the furrows (Penrose 1968; Králík and Novotný 2003). However, in the archaeological context, it is usually not possible to study the fingerprint directly—except in cases such as mummified individuals (Prominska et al. 1986; Spindler 1998; Králík and Nejman 2007), but to study its negative overprinted when the skin comes in contact with the object (Králík and Nejman 2007). Instead, however, there are different types of imprints depending on the surface/material in which they were formed. In plastic materials, such as ceramics (and in the present case of a figurine made of clay), negative 3D imprints are formed, and the fingerprint is observable due to the contrasts (under lateral light) between the lighter and darker sections (Králík and Nejman 2007). In this scenario, ERB is calculated as the distance from the edge of the shadow of the ridge to the edge of the shadow of the adjacent ridge (Králík et al. 2002; Králík and Novotný, 2003; Králík et al. 2002). Normally, what is calculated is the actual mean of several ERB, i.e. the mean ridge dreadth (MRB) (Králík and Novotný 2003; Fowler et al. 2019). MRB was obtained by calculating the distance between 10 consecutive ridges (as previously described) in a straight line perpendicular to the direction of the epidermal ridges following standard procedures (Kamp et al. 1999). The total distance was divided by 10 to obtain the mean ridge breadth and its standard deviation. The same fingerprint has been measured from two different locations to reduce the possibility of variance in the patterns of the same fingerprint since differences may exist depending on the region of the finger (Sánchez-Andrés et al. 2018).

The MRB is used to estimate the person’s age at the time they left their fingerprint by applying linear regression equations (Loesch and Czyzewska 1972; Kamp et al. 1999; Králík and Novotný, 2003; Fowler et al. 2019). In the archaeological context, however, it is essential to take into consideration which material the fingerprints were formed on since the taphonomy process and the inherent behaviour of each material may alter the appearance and size of the prints (Králík and Nejman 2007). Thus, the metric analysis of fingerprints on ceramics must take into consideration the shrinkage of the ceramics during the drying and burning process (Králík and Novotný, 2003; Králík and Nejman 2007; Fowler et al. 2019, 2020). This is a variable that should be taken into account when calculating the MRB and estimating sex and age as it could alter the conclusions drawn from it (Fowler et al. 2019). The degree of shrinkage is very variable (it can range from 0 to 30%; Králík 2000; Fowler et al. 2019). Thus, a shrinkage correction factor should be introduced when estimating the age from the MRB (Králík 2000). Králík (2000) suggested a shrinkage rate of 7.5%, whereas other authors have applied slightly lower shrinkage rates of 2% and 6% (Fowler et al. 2019). In the present study, we have applied the KAmod2 formula proposed by Fowler et al. (2019) that assumes the original KA formula (Kamp et al. 1999) by correcting it with 2% and 6% shrinkage corrections (Sanders 2015; Fowler et al. 2019) instead of the 7.5% shrinkage correction proposed by Králík and Novotný (2003).

$$\mathrm{KAmod}20.02:\mathrm{age}\;(\mathrm{months})=614\mathrm x\ast1.0204\ast\mathrm{MRB}\;(\mathrm{mm})-112\;(\mathrm{shrinkage}\;\mathrm{of}\;0.02)$$
$$\mathrm{KAmod}20.06:\mathrm{age}\;(\mathrm{months})=614\mathrm x\ast1.0638\ast\mathrm{MRB}\;(\mathrm{mm})-112\;(\mathrm{shrinkage}\;\mathrm{of}\;0.06)$$

Sex estimation: ridge density

Numerous studies agree that the density of ridges is causally related to sex (Ohler and Cummins 1942; Acree 1999; Gutiérrez-Redomero et al. 2008; Nayak et al. 2010a,b; Krishan et al. 2013; Sharma et al. 2021). The age estimate in a palaeodermatogliphic is based on the calculation of the number of epidermal ridges in a 25-mm2 square (5 × 5 mm) (Acree 1999; Fowler et al. 2019). In our case, the 25-mm2 square was drawn using Adobe Photoshop CC 2021 (22.3.1) and creating a new layer that was placed over the desired study area (i.e. that preserved a good pattern or epidermal ridges and was easy to count). The number of ridges was obtained by drawing a diagonal line from one of the corners to the opposite corner and counting all the ridges that were crossed (Gutiérrez-Redomero et al. 2008; Nayak et al. 2010a,b; Eshak et al. 2013).

Results

The analysis of the three possible fingerprints was performed to determine their human origins, following the previously described methodology (Králík et al. 2002). Therefore, the putative fingerprints were meticulously analysed to determine the possible presence of minutiae (Galton 1892; Králík et al. 2002). Despite the fragmentary nature of the remains of the fingerprints, a detailed analysis allowed us to identify the presence of 3 minutiae in fingerprint 1 (Fig. 4), which are separated from each other by 18 mm. The third one was discarded due to its fragmentation, which prevented testing for sex and age.

Fig. 4
figure 4

Detail of Fingerprint 2 with the grooves highlighted and the 25-mm2 square overlapped (left) and the identification of a bifurcation, termination, and an independent ridge minutiae

Full size image

MRB estimations for both analysed fingerprints are listed in Table 1. The MRB of the first fingerprint was 0.52 ± 0.076; the second fingerprint had very similar values, with an MRB of 0.50 ± 0.11. These MRB values were used to estimate the person’s age at the time of impressing the fingerprints, using the formula modified with 2% and 6% shrinkage published by Fowler et al. (2019) (Table 1). The estimation of the age from Fingerprint 1 is ~18.3 years (with a 2% shrinkage) and  ~19.5 years with a 6% shrinkage (Table 1). For Fingerprint 2, the estimated age is slightly lower, oscillating between  ~16.8 years (with a 2% shrinkage) and  ~17.9 years (with a 6% shrinkage) (Table 1). If both fingerprints were imprinted by the same individual, their age would range between 17 and 19.5 years.

Table 1 MRB and ridge density of both fingerprints
Full size table

Concerning the density of ridges, two measures were taken at slightly different locations for each fingerprint to minimize possible regional differences. The results obtained for the four different regions analysed were roughly similar (Fingerprint 1: 6 and 7; mean: 6.5 ± 0.5; Fingerprint 2: 7 and 7, respectively; mean: 7.0 ± 0.0; Table 1). A cross-plot of the mean ridge breadth and ridge density at 2% and 6% shrinkage was performed to better estimate both the sex and age of the individual (Fowler et al. 2019; Lass 2019). In this bivariate axis, as can be seen in Fig. 5, Tell Halula has relatively low values for ridge density (values more extreme than those reported in other studies), while the values of the MRB fall within the variability observed in other sites (Kamp et al. 1999; Králík and Novotný 2003; García Casanova 2018; Sánchez-Andrés et al. 2018; Fowler et al. 2019).

Fig. 5
figure 5

Age and sex correlation matrix combining mean ridge breadth and ridge density data. The value corresponding to Tell Halula is represented by the red rounded mark (Fowler et al. 2019)

Full size image

Discussion

The study of the imprints of the Tell Halula figurine fills a gap and contributes to a better understanding of the Neolithic society of the PPNB in the Middle Euphrates region. The final study sample is low; as of the 45 figurines initially recovered figurines from the Tell Halula site, only one had putative fingerprints, representing 2.2% of the initial sample. Clay figurines are burial goods interred with the dead, in addition to other objects including personal ornaments (butterfly beads composing diadems, necklaces and bracelets), sea shells, flint and obsidian tools, macrolithic tools, bone tools, ocher, limestone spheres, and pieces of galena (Guerrero et al. 2009; Molist et al. 2010; Kuijt et al. 2011; Ortiz López 2016; Ortiz et al. 2013; Alarashi 2016; Taha et al. 2017; Alarashi et al. 2018). The context and the potential link to any particular funeral (and hence to any particular biological profile of age and gender at death) are unknown. This is often a common situation (Coqueugniot 2003, p. 35): “in many cases, we have no information relating to the context in which the figurines were discovered and, at most, we have an inventory of them, published as an appendix to the archaeological study of the site”. Despite this, it has been referenced that the piece would be dated in the transition between the MPPNB and the LLPNB (Ayobi 2013). An increase in the number of clay figurines in the burials of Tell Halula of both adults and children has been documented in the transition from the MPPNB to the LPPNB (Kuijt et al. 2011). This situation fits with what is observed at a general level during the LPPNB in the Near East, in which animal figurines become more numerous representing indeterminate quadrupeds (Coqueugniot 2003). In any case, it is not possible to associate the HL72 figurine with any specific burial and we cannot discard that it might be “found in a context of abandonment, rejection, and not in their primary situation of use or voluntary deposit” (Coqueugniot 2003, p. 43).

In our study, two human fingerprints have been determined thanks to the presence of unequivocal minutiae such as terminations, bifurcations, or island points as determined by Galton (1892). Although low sample sizes are frequent in paleodermatoglyphic analysis (Králík and Novotný 2003), this kind of study can still yield satisfactory results (Králík and Novotný 2003; Minguez et al. 2016; Martínez-Sevilla et al. 2020). However, low sample sizes can obscure manufacturing process distortions due to the manipulation of the object, its manufacture, or unintentional contact with different pressure intensities (Králík and Novotný 2003), which can even affect the intensity ridge count (Minguez et al. 2016) and, thus, make it more complicated to determine the biological profile because of the huge overlapping in the intensity ridge values between males and females—males have average values ranging from 10.9 to 12.99, whereas females range from 12.61 to 15.61 (Sanders 2015). Thus, given the fragmented and poor state of preservation of ancient fingerprints (Králík and Novotný 2003), it is important to conduct detailed and thorough studies, considering all factors, to obtain an accurate estimate of the biological profile (Králík and Novotný 2003). For this reason, both the MRB and the epidermal ridge density were calculated to reconstruct the biological profile.

MRB may vary considerably among fingerprints from different fingers (Mundorff et al. 2014), regions from the same fingerprint (Krishan et al. 2013; Dhall and Kapoor 2016; Gutiérrez-Redomero et al. 2013), and even areas of the same region (Minguez et al. 2016). For example, several studies have highlighted that the MRB of a fingerprint from the thumb is the largest of all fingers (Mundorff et al. 2014). However, given the fragmentary nature of archaeological fingerprints, it is not possible to determine to which finger it corresponds. In the specific case of the fingerprints from Tell Halula, both MRB and ridge intensity values were quite homogeneous (Table 1) for the two analysed fingerprints, suggesting that both could have been imprinted by the same person. Moreover, we may assume it could correspond to a thumb, since the imprints have been located at the bottom of the figurine in a region where pressure is normally applied consciously with the thumb to create concavities and shape the ventral part after the front legs and before the hind legs (Fig. 2).

In this sense, the MRB of fingerprint1 is 0.53 ± 0.076, while the second fingerprint has an MRB value of 0.50 ± 0.11. The MRB estimate for the two fingerprints is quite similar, so they could belong to the same person. In this case, the average MRB would be 0.515, a relatively high value that may suggest a young adult individual: most published studies tend to agree that an MRB >   ~0.50 mm belongs with a high probability to adult individuals (Cummins et al., 1941; David 1981; Králík et al. 2002; Králík and Novotný 2003; Minguez et al. 2016; Fowler et al. 2019, 2020; Lass 2019). MRB increases with age (García Casanova 2018) following a logarithmic function (Minguez et al. 2016), in which the increases in MRB from adolescence are smaller, reaching a horizontal asymptote in adulthood (Minguez et al. 2016; García Casanova 2018; Fowler et al. 2019). In a recent study of 546 resentful individuals in Catalonia (Spain) aged from 6 to 89 years, the increase in MRB was observed to stop at the age of 35 (García Casanova 2018). Furthermore, the correlation between MRB and age was significantly stronger in males (García Casanova 2018). Based on the data published by Králík and Novotný (2003), Fowler et al. (2019) proposed the differentiation of three age groups based on the MRB: pre-pubescent children (<0.37 mm), adult males (>0.52) and a third ambiguous group including all adolescents and adult women and men. Therefore, the calculation of the MRB allows the sex of the fingerprint to be determined reliably only when it is  ~ ≥0.52. Values below this threshold are highly uncertain, as they may belong to both an adolescent—both male and female, an adult woman, or even in some cases to an adult male. However, while some studies indicate that these differences are statistically significant (García Casanova 2018), others affirm that the differences are not significant (David 1981). Therefore, according to Fowler et al. (2019), through the MRB, we could only reliably assess the age group of a fingerprint if it was  <0.37 (child) or  ≥0.52 (adult male). According to García Casanova (2018), an MRB  ≥0.47 indicates an adult older than 36 years (at 85% accuracy) (Martínez-Sevilla et al. 2020). Thus, given the MRB value for Tell Halula’s fingerprints (0.515), we can assume it would be an adult and, likeliest, a male individual (as will be discussed later). When applying the Kamod2 formula for age estimation published by Fowler et al. (2019), age ranges between 16.8 and 18.3 years with a 2% shrinkage and between 17.9 and 19.51 with a shrinkage of 6% are obtained. In any case, if both fingerprints belong to the same individual and if extreme values are selected, it can be assumed that the person would have an age ranging from 16.8 (lower value: Fingerprint 1 with a 2% shrinkage) to 19.51 (Fingerprint 2 with a 6% shrinkage), i.e. a young adult. This estimate of young adult age contrasts with that obtained using the current Mediterranean population as a comparative sample (over 36 years old) (García Casanova 2018). In any case, both methodologies tend to agree that it would be an adult individual.

Once discrimination by age was performed, sex estimation was performed. The MRB presents a marked sexual dimorphism (Mundorff et al. 2014; García Casanova 2018): MRB tends to be significantly higher in males than females (David 1981; Fowler et al. 2019; Mundorff et al. 2014: García Casanova 2018; Martínez-Sevilla et al. 2020). However, it should be noted that this sexual dimorphism is only observed in adult individuals and late adolescents, since no significant differences in the MRB are reported between boys and girls up to 12/15 years (Loesch and Czyzewska 1972; García Casanova 2018). As Fowler et al. (2019) state, MRB values of  ≥ ~0.52 correspond exclusively and without any doubt to an adult male individual. This threshold coincides with that of many studies conducted on current populations. In a study with a large sample with a population of California (n = 500), Mundorff et al. (2014) set several thresholds depending on the finger considered. For the thumb (as we have assumed our fingerprints may be from this finger), the threshold value of the MRB is precisely  ~0.52 (0.517 for the right thumb and 0.534 for the left one) (Mundorff et al. 2014). In the European Mediterranean population study, an MRB  ≥0.49 corresponds to a male adult (>35 years old, as previously discussed) with a probability of 72% (García Casanova 2018; Martínez-Sevilla et al. 2020). Therefore, we can assume that it would be a male individual.

Although the MRB values alone allow for a fairly accurate estimation of age and sex, given that they are only two fingerprints (presumably from the same individual), the ridge density was also calculated. Studies in current populations have highlighted that males tend to have a lower ridge density than females (on the other hand, an expected situation, as they tend to have a higher MRB) (Gutiérrez-Redomero et al. 2008; Dhall and Kapoor 2016). Acree (1999) stated that a given fingerprint possessing a ridge density  ≤11 ridges/25 mm2 was likeliest to be of male origin. On the other hand, in a study of a population of Spanish origin, the differences between the sexes are not so marked: in the radial area, a ridge count  ≤16/25 mm2 was more likely to correspond to a man’s fingerprint (Gutiérrez-Redomero et al. 2008). However, in an Indian population, Dhall and Kapoor (2016) found that an RD < 11/17 (depending on the region) had a probability of 1 to correspond to a male individual. Fowler et al. (2019) argued that values below 13 on a 25-mm2 square are 95% likely to be male. Therefore, although differences among populations exist, what is undeniable is that men tend to have a lower ridge count: an RD lower than 12/14—depending on the population and region—would correspond to males (Gutiérrez-Redomero et al. 2008; Nayak et al. 2010a,b; Krishan et al. 2013; Dhall and Kapoor 2016). Ridge density in the fingerprints from Tell Halula ranges from 6.5 to 7 (mean: 6.75) (Table 1). This is an extremely low value that corresponds to a male, as the MRB value had previously suggested.

There are scarce references to fingerprints in early Neolithic sites in the Near East (Bennison-Chapman and Hager 2018). In the Neolithic site of Boncuklu Höyük (Central Anatolia, dated from the mid-ninth to the mid-eighth millennium cal. BC), small geometric clay objects appeared with traces of fingerprints on most of them. These objects were manufactured mostly by adult women (Bennison-Chapman and Hager 2018). This contrasts with the results derived from the analysis of Tell Halula, according to which the figurine was probably done by an adult male. However, although the Aceramic site of Boncuklu may have been coterminous and contemporary with the Neolithic site of Tell Halula, they are quite distinct from each other (Baird 2012). It is therefore difficult to make any extrapolation or comparison between the two sites. Another study of Near East fingerprints was conducted by Sanders (2015), although with a much more recent chronology: Tell Leilan is an Early Bronze Age site (3400–1700 BCE) where the study of fingerprint impressions on pottery enabled the analysis of gender roles throughout more than one millennia. Despite being a more recent chronology, it should be noted that they concluded that the ceramics in that area of the Near East pottery production was not a gendered task before the rise of urbanism. This could make the seemingly opposite results of the present study in Tell Halula compatible with those reported for Boncuklu Höyük. However, several studies (O'Shea 1981; Testart 1982; Eshed et al. 2004; Hedges et al. 2013; Macintosh et al. 2014, 2017; Robb and Harris 2018) have shown that from Natufian to the beginning of the Neolithic, there was a division of labour. Musculoskeletal differences in the joints have been found (Eshed et al. 2004). According to Smith et al. (1996), men performed jobs that required more muscular strength, and Eshed et al. (2004) complemented this argument by saying that the increase in workload increased as the Neolithic period progressed, as muscle insertions became more pronounced. In addition, those joints necessary for grinding grain were found to be more degraded only in women from the Natufian period, as was the case at Abu Hureyra (Syria) (Molleson 1994). There was a significant change in the variety of duties performed by men and women with the adoption of agriculture and the development of complex societies. While agreeing that there have been changes in customs, questions remain about how to approach the division of labour. While experts contend that women and children were of lower rank than men based on the examination of the grave good assemblage variance by osteological sex and age (Jeunesse 1997; Hofmann 2009), others (Hofmann 2009; Masclans Latorre et al. 2021) have claimed that, rather than being created based on the data, these interpretations were based on contemporary ideas of binary gender hierarchy and values.

In conclusion, our results indicate that there is a high percentage of probability that the individual who handled the artefact was an adult male. However, given the low sample of fingerprints in the collection of figurines from Tell Halula, little more can be said, as not many inferences can be made in terms of social structure and division of labour, as other studies with larger samples have been able to address (Sanders 2014; Fowler et al. 2020). The problem of having such a small sample size and the fact that so limited studies have been carried out around the world on the fingerprints of Neolithic artefacts make it difficult to know exactly who manufactured these small objects and what they were used for. Therefore, more studies are needed to understand both the symbolism of the objects and by whom they were manufactured.

Geographical coordinates

The Tell Halula site (36°25′N, 38°10’E, 337 m.a.s.l.) is localized in the Mid-Euphrates region (Governorate of Rakka, Syria) about 105 km east of Aleppo and 25 km northwest of Membij.