Butchering technology and metal adoption in the Bronze Age southern Levant: SEM evidence from Tell Aphek, Israel

1 Introduction

Distinguishing natural (e.g., gnawing, trampling, weathering) from anthropogenic (e.g., butchering marks, polishing) bone surface modifications is a fundamental requirement in zooarchaeology and forensics (Egeland et al., 2024). This task is often challenging without the aid of advanced analytical techniques. During the 1980s, scanning electron microscopy (SEM) emerged as a transformative tool for addressing these challenges (Figure 1). Prior to its utilization, researchers relied on low-power optical microscopy, which limited the ability to distinguish among different types of surface alterations caused by stone and metal tools, carnivore teeth, or post-depositional processes. SEM, with its ability to provide high-resolution, three-dimensional images at magnifications of several thousand times, revolutionized the field by revealing microfeatures such as internal striations, fracture and impact morphologies, and tool-edge signatures undetectable by eye and under light microscopy.

Although Brothwell (1969) was among the first to advocate the integration of SEM in zooarchaeological research, the dedicated application of SEM to bone surface modification analysis was pioneered during the 1980s by Shipman (Potts and Shipman, 1981; Shipman and Rose, 1988), Rose (1983), and Olsen (1988). They provided clear criteria for distinguishing anthropogenic modifications from natural taphonomic processes (Behrensmeyer et al., 1986; Olsen and Shipman, 1994). This was achieved through the identification of microstriations and internal groove structures within butchering marks that were diagnostic of different tool technologies or materials. SEM was vital for reconciling debates over hominin foraging and hunting behavior (Andrews and Fernandez-Jalvo, 1997; Domínguez-Rodrigo, 1997). Advances in specimen preparation, such as sputter coating samples with gold or carbon for conductivity, and improvements in backscattered electron imaging further enhanced image clarity and interpretive precision. By the end of the 20^th century, SEM had become an indispensable component of microscopic bone analysis, particularly when coupled with groove morphology evaluation, offering a level of objectivity and reproducibility unmatched by earlier techniques. Subsequent advances in forensic science to study sharp and blunt force trauma utilized SEM element detection applications to identify the raw material of the weapon (Bai et al., 2007; Gibelli et al., 2012; Palazzo et al., 2018). In this manner, traces of metal residues (e.g., Fe, Cu, Zn) preserved in mark surfaces made it possible to determine the material and other attributes of the implement responsible for the damage. This methodological framework has recently been adopted into animal butchering research with a similar aim and promising results (Marginedas et al., 2024).

The SEM analysis of anthropogenic bone surface modifications, specifically butchering marks, has provided valuable insights into past subsistence, diet, butchering patterns and style, and livestock management. It allows researchers to identify the raw materials and technological sophistication of the implements used in these activities. The approach was expanded upon and became prolific in HJG's extensive investigations into the spread of metal technology in Eastern Europe and the Near East (Beller et al., 2021b; Greenfield, 1999, 2000, 2002, 2006, 2013, 2017; Greenfield et al., 2022a; Greenfield and Chaput, 2021; Okaluk and Greenfield, 2022; Sousa and Greenfield, 2021, 2023). In this paper, we revisit the application of SEM to bone surface modification, highlighting its methodological practicality and demonstrating its utility for reconstructing the diffusion of metal technology in the Bronze Age southern Levant through the examination of butchered assemblages from Tell Aphek, Israel.

2 Context of this study

The Bronze Ages (3500–1200 BCE) are transformative periods for the southern Levant. They are generally defined by the emergence and flourishing of early complex societies in the Early Bronze Age II-III (EB; 3000–2500 BCE). In turn, these societies are characterized by elite architecture, public buildings, fortified urban settlements, craft specialization, and new technologies (Milevski, 2011a; Prag, 2014; Richard, 2014), including bronze metallurgy (Genz, 2000; Levy et al., 2002; Philip, 1989). Long-distance exchange linked the region to Egypt and Mesopotamia, facilitating the movement of a variety of commodities, such as copper and tin. After a hiatus (EB IV, 2500–2000 BCE), complex societies flourished during the Middle Bronze Age (MB; 2000–1550 BCE). Despite Egyptian conquest and hegemony during Late Bronze Age (LB; 1550–1200 BCE), inhabitants of the southern Levant retained vibrant urban communities.

During the Bronze Age, soft alloys of low As/Sn–bronze tools are initially produced and used (Philip, 1988, 1991). By the later phases of the EB, harder alloys, such as 10% Sn-bronze, became widespread across the Near East, and their use increased during the MB (Avilova, 2009; Moorey, 1994; Philip, 1991; Rosenfeld et al., 1997; Yener, 2000). This stronger form of bronze was versatile, suitable for many more applications of everyday activities, such as butchering implements for processing food. Research on the origins of bronze metallurgy has focused mainly on preserved metal artifacts, primarily from recovered remains found in burial sites or hoards (Montanari, 2018; Nigro et al., 2018; Philip, 1988). For the most part, these were objects heavily biased toward prestige, warfare, or status items, thereby skewing our understanding of the broader usage of early metals (Batiuk and Harrison, 2017; Philip, 1988, 1989). Additionally, it is difficult to assess the significance of various technologies because metal objects could be and were often recycled, melted down, and recast when they lost utility (Genz, 2000). This practice is observed in other early metal-producing regions (Bray, 2022; Kiss and Romhányi, 2023; Radivojević et al., 2019).

Bronze daggers appear early in the EB across the Near East in general and southern Levant specifically (Genz, 2000), while bronze knives emerged later (Moorey, 1994; Philip, 1988, 1989). Similar to other regions where daggers served ceremonial and prestige roles (Nakou, 1995), daggers of EB southern Levant, although much rarer, also tend to occur in symbolic contexts (Montanari, 2014), and thus are not associated with butchering. The reason is relatively straightforward: knives are used for slicing and have one sharp edge, while daggers are used primarily for stabbing and have two sharp edges that converge to a point. Knives have more uniform thickness and sharpness along the length of the blade, while daggers tend to be increasingly wider, however slight, along the length (Greenfield, 2006). In the southern Levant, bronze knives begin to appear during the 500-year-long EB IV/MB I period, albeit it in small quantities, and become more common in the late MB (Moorey, 1994; Philip, 1989). By the LB, bronze technology had generally permeated most levels of society, becoming more accessible to people of lower socioeconomic status (e.g., peasants) (Greenfield, 2013).

The aforementioned patterns complement those conducted by lithic specialists who demonstrate that the replacement of stone with metal tools in general is a slow process that extends through the length of the Bronze Age. However, the rate of decline depends on the development of equivalent metal tools (Greenfield, 2013; Manclossi et al., 2019, 2018; Rosen, 1997). Hence, chipped stone tools persist through all periods. This trend is highlighted by sites where stone tools continue to be used through the Iron Age and Classical period for quotidian household activities, including butchering, since they are sharper, cheaper, and more readily accessible than metal tools—i.e., at the central Anatolian sites of Çatalhöyük and Göltepe where obsidian was locally available and plentiful (Greenfield and Chaput, 2021; Greenfield and Marciniak, 2019). As demonstrated elsewhere, quotidian metallurgy for butchering appears later and later the farther one travels from Anatolia (Eriksen, 2010; Greenfield, 2017; Jones, 2011; Nicholson and Shaw, 2000; Olsen, 1988; Stocks, 2003; Vandkilde, 1996). Even in the southern Levant, it appears that there is both geographic and temporal variability in the acceptance rate of the new technology. Another tentative pattern is that quotidian metallurgy appears later in smaller sites (periphery) than in the larger urban centers (Greenfield et al., 2022a,b). Therefore, as it is not entirely clear when metal technology, (i.e., bronze) becomes widespread and are utilized at every level of society, each region must be investigated separately. Such investigations have been conducted at a number of southern Levantine sites (Table 1). However, a site with a continuous Bronze Age sequence (i.e., EB-MB-LB) has yet to be studied in this manner. One potential site is Tell Aphek.

Table 1. Previous studies on the spread of metallurgy based on butchered assemblages from the southern Levant.

Tell Aphek is located on the southern Sharon Plain of central Israel, just on the northeast outskirts of the modern metropolis of Tel Aviv (Figure 2). The site has been inhabited since the Chalcolithic (4500–3500 BCE) with notable deposits from the Early, Middle, and Late Bronze Ages, as well as the Iron I and II (1200–586 BCE) Ages. Throughout these turbulent chrono-cultural periods, Aphek served as a prominent settlement and an important regional political and economic center (Beck and Kochavi, 1993). Occupation also occurred during the Roman (63 BCE-330 CE) and Ottoman (1516–1918 CE) periods. The collective faunal assemblage from Tell Aphek presents a unique opportunity to reconstruct the nature of raw material adoption and the spread of metallurgy during periods of dramatic societal development. In this study, we evaluate recently collected data on butchering marks from each Bronze Age occupation level to discern if and when metal tools replace stone tools in the butchering practices.

3 The archaeological site of Tell Aphek (Antipatris)

Tell Aphek first underwent salvage excavation in 1934–1935 by Ory (1936), followed by another short salvage excavation under Eitan (1969) in 1961. These efforts facilitated the Yarkon springs development projects. It was not until 1972 that the site underwent systematic excavation, this time directed by P. Beck and M. Kochavi of Tel Aviv University (Beck and Kochavi, 1993). Two large volumes have been produced which detail the project background, chronology, excavation methodology, and findings (Gadot and Yadin, 2009; Kochavi et al., 2000).

Archaeology and historical texts attest to the long-term occupation at Tell Aphek (Table 2) (Kochavi, 1981). It was a developing urban settlement during the EB I-II, with evidence of occupation uncovered in multiple areas of the site (Beck, 1985a), including streets, houses, and a city wall (Gal and Kochavi, 2000; Yadin and Kochavi, 2000). There is also well-preserved evidence of MB occupation, as a prominent urban center was established at the site (Gal and Kochavi, 2000; Yadin and Kochavi, 2000). Its strategic location allowed it to facilitate trade and exchange from Egypt to Mesopotamia, contributing to the settlement's prosperity. Among the public and residential buildings is a complex palatial structure that was probably the seat of an administrative or centralized authority (Yadin and Kochavi, 2000). Fortifications of mudbrick walls atop stone foundations with towers and gates have been discovered, indicating the need for increased defense (Beck, 1975, 1985b).

Table 2. Summarized chronostratigraphy at Tell Aphek.

In the Late Bronze Age, Egyptian occupation in Canaan established an outpost at Tell Aphek (Gadot, 2010). The site was later settled by the Philistines during the Iron Age I and II. By the Hellenistic period, it was known as Pegae. During the Roman occupation of Judea (and wider Levant), King Herod the Great constructed the city of Antipatris. The city was later destroyed by an earthquake in 363 CE. Settlement of Tell Aphek was reignited by the Crusaders, who built a castle on the hill known as “Migdal Aphek” or Tower of Aphek. Finally, the Ottoman ruler, Suleiman the Magnificent, is believed to have constructed the fortress of Binar Bashi in the 16^th century CE on the site.

The importance of Tell Aphek to the study of early metallurgy is its rich and abundant assemblage of butchered fauna from the Bronze Age levels. Further details of the butchering patterns and activities can be found in Beller et al. (in press). The same butchering marks can inform as to the nature of adoption of metal technologies in important urban centers during these crucial periods of the southern Levant.

4 Methods

4.1 Number of butchering incidences

The butchering data were recorded with several quantification measures. First, we used the traditional count of each bone as a separate specimen, commonly known as Number of Identified Specimens (NISP) (Lyman, 1994, 2018). Second, we counted each individual slice, regardless of how many occurred on an individual bone. This count reports all slices present and avoids issues of under-counting. Third, the Number of Butchering Incidences (NBI) was included (Greenfield and Brown, 2016; Greenfield et al., 2016a). This recently introduced measure reflects the number of times a butchering activity occurred on the same specimen. For instance, a bone with two separate slices, one on either end, would be an NISP of 1 but have an NBI of 2.

4.2 Butchering efficiency index

In addition, we calculated the butchering efficiency index (BEI) for each period. The BEI refers to the average frequency of slices with respect to slicing incidences (Greenfield and Brown, 2016; Greenfield et al., 2016b), and is calculated using the equation:

BEI for period=Total slices for periodNumber of slice incidences for period(1)

There is an inverse relationship between the BEI and butchering efficiency (Equation 1). The lower the BEI, the higher the butchering efficiency. Conversely, the higher the BEI, the lower the butchering efficiency. If most scholars are correct in the assumption that efficiency increases with the adoption of a new technology (e.g., from stone to metal tools), then there should be fewer slices/incidences (lower BEI) over time. Alternatively, but not mutually exclusive, is that BEI is related to butcher specialization, as greater skill in a craft will mean greater efficiency (Apel, 2008; Forte et al., 2023). Skilled butchers have better knowledge of where to strike and more experience with the amount of force needed (Potter, 2005; Willis and Boehm, 2015), thus impacting the bone less and leaving fewer marks. In the Bronze Age southern Levant, one may envision high skill butchers as specialized and low skill butchers as household members (Cheney, 2020).

4.3 Scanning electron microscopy

This study draws on the innovative analytical procedures and experimental work that established the relationship between implement material and groove morphology over the past 50 years (Figure 1). This type of evaluation has been previously applied to adoption of metal technology into the butchering process across the southern Levant (Table 2).

The two main raw materials from which butchering implements are made, stone and metal, leave behind grooves of detectable differences (Fisher, 1995; Greenfield, 1999, 2002, 2006, 2013; Potts and Shipman, 1981; Walker, 1978; Walker and Long, 1977). These are most apparent and best identified on isolated single slices, where the cross-section morphology of the groove is indicative of the raw material (Figure 3). Intersecting slice marks tend to obscure the necessary details. The criteria for discerning stone from metal slicing marks are well-established and presented in detail elsewhere (e.g., Greenfield, 1999, 2002, 2006) and supported by recent artificial intelligence and geometric morphometric analyses (Cifuentes-Alcobendas and Domínguez-Rodrigo, 2021; Domínguez-Rodrigo, 1997; Domínguez-Rodrigo et al., 2020; Maté-González et al., 2019). A brief summary follows.

Slice marks made by a chipped stone implement (e.g., flake, blade) appear uneven and asymmetrical in cross-section (Figures 3A–D). As the two sides meet at the apex, one rises relatively steeply and the other will ascend gradually or in a series of adjacent ridges. A unifacial (but not necessarily retouched) flake or blade will create grooves with parallel striations. Striations will be of uneven thickness and length. Bifacial stone implements will be more symmetrical than unifacial stone implements, but the groove created by the implement will still exhibit striations. As stone implements dull, their grooves will become more sinuous and variable in width.

Metal implements (e.g., dagger, knife) leave behind slice profiles with their own distinct attributes. These appear cleaner, even, and symmetrical in cross-section (Figure 3E), giving the apex a distinct point. The sides will be uniform, smooth, and devoid of striations. The force applied to the implement can create grooves of various sizes and depth, but with a characteristic V-shaped profile. As metal implements become dull, the point of the apex flattens (Figure 3F).

When analyzing the groove morphology, it is also necessary to rule out the possibility of their creation by natural origin (e.g., carnivore gnawing, trampling), especially in cases of deep prehistory. As mentioned earlier, these taphonomic activities display several distinctive microscopic features that are best considered as a suite to be effectively diagnostic. For instance, trampling marks are typically produced by sediment abrasion on the bone surface under the application of force (animal's weight) (Olsen and Shipman, 1988), leading groove trajectory and morphology to be among the most reliable variables. These grooves show irregular or sinuous trajectories, rounded outer edges, and broad, shallow morphologies (Domínguez-Rodrigo et al., 2009). Meanwhile, their cross sections are commonly wide and open, with rounded or flattened apices, but occasionally can be sharp (Domínguez-Rodrigo et al., 2009). Internal microstriations may be present but are often discontinuous, irregular, or even absent depending on post-depositional time (Behrensmeyer et al., 1986). Carnivore teeth, in contrast, frequently produce broader and more irregular grooves, punctures, and scores. These modifications generally exhibit U-shaped or flattened cross sections from a crushing and dragging movement rather than slicing (Greenfield, 2006; Shipman and Rose, 1983). The internal surfaces of such grooves tend to be rough and uneven, lacking the continuous parallel microstriations characteristic of butchering marks. This variability reflects differences in force, tooth morphology, and bone density during feeding behavior.

All faunal remains from Tell Aphek were examined for evidence of butchering (Table 3). Marks attributable to butchering were first distinguished from those that may derive from other ancient and modern taphonomic sources (e.g., scratches, weathering, preparation or cleaning damage). Specimens exhibiting butchering were isolated from the larger assemblages and the morphology of their butchering grooves was subjected to further evaluation. This was performed by a combination of eye, a hand-held illuminated magnifier (25x), and SEM (Figure 4).

Table 3. Characteristics of butchering activities, as updated from Beller et al. (2022, p. 197).

The application of SEM involves first creating clean molds of the grooves using a dental (e.g., Exaflex^®) hydrophilic vinyl polysiloxane impression compound (base + catalyst). Molds from such dental compounds create accurate copies of the original morphology of slices that can be obtained quickly, easily, at little cost, and with very little damage to the bone (Greenfield, 2006; Rose, 1983). It should be acknowledged that repeated molding can start to erode features of the mark and even modify the groove morphology (Valtierra et al., 2024). Where possible, molds are best applied to flat surfaces (e.g., flat bones) rather than curved (e.g., shafts). It is also best to avoid heavily weathered and degraded surfaces, as bone debris of flakes may be removed by the mold and impair detail. We targeted slice marks that occurred in isolation (no intersecting grooves) and on well preserved bones. Once dry, the molds are then independently evaluated under a light optical binocular microscope to ensure that the mold had been properly taken. A selection of silicone molds is trimmed to 1–3 cm to be examined by SEM as the chamber is relatively small (Figure 4). These are now examined at variable pressure while, in the past, they were typically sputter coated with gold to properly reflect the electrons at high vacuum pressure.

The molds were then observed under the SEM (Cambridge Instruments SEM Stereoscan 90 and Zeiss EVO Series 10 SEM) at the Near East and Biblical Archaeology Laboratory as part of the Earth Materials and Archaeometry Center at the University of Manitoba (Figure 4). The molds present as inverted reliefs or profiles of the original grooves (Frahm, 2014; Greenfield, 1999; Olsen, 1988; Rose, 1983), allowing the diagnostic features discussed above to be viewed in three dimensions within the SEM. If the mold is viewed from an overhead position, the shape of the apex and diagnostic criteria on both sides of the groove can be viewed, evaluated, and measured. This multi-level microscopic analysis ensures that the characteristics of stone vs. metal slices are visible, enabling greater confidence in or confirmation of the raw material identification. These were analyzed in a single-blind manner where we evaluated the molds in isolation of a priori knowledge of the naked eye and illuminated magnifier results. Afterward, we compared the results of the three techniques.

5 Results

5.1 Description of the butchered assemblage

The entire butchered assemblage from Bronze Age Tell Aphek contains 281 bones (Table 4). There is an uneven distribution by period, as some periods contain more faunal remains than others due to greater exposures of their levels. This is most apparent with the MB IIA (n = 114, 40.6%) and LB II (n = 104, 37.0%). These are distantly followed by the MB IIB (n = 27, 9.6%) and the EB I (n = 19, 6.8%). No butchered bones originate from Chalcolithic contexts. In addition, there are butchered specimens from the Iron Age and later periods at the site which are published elsewhere (Beller et al., accepted a; Horwitz, 2013).

Table 4. Distribution of butchered bones, butchering incidences, and butchering strokes by period.

Total NISP amounts (butchered + non butchered) are extracted from Hellwing (2000) and Horwitz (2013) faunal reports.

^aFrom Table 21.1 in Horwitz (2013).

The combined butchered assemblage is dominated by traditionally domestic taxa (Table 5). These primarily appear as Bos taurus/cattle (n = 106, 37.7%), followed by Sus scrofa dom./pig (n = 22, 7.8%), Ovis aries/sheep (n = 21, 7.5%), and Capra hircus/goat (n = 19, 6.8%). The indeterminate categories of Ovis/Capra (n = 52, 18.5%), and large (n = 28, 10.0%) and medium (n = 17, 6.0%) mammals, are commonly represented. Wild taxa are much fewer in number (n = 15, 5.3%), but have a wide diversity represented (seven taxa). These are mainly represented by Dama dama/fallow deer (n = 9, 3.2%), with specimens occurring throughout occupation. All other wild taxa have one representative specimen each.

Table 5. Identified taxa among the butchered assemblage.

5.2 Taphonomy

A full description of the taphonomy is found in Beller et al. (in press); Beller et al. (accepted a), but some observations are made here with specific reference to the Bronze Age specimens. For every period, light weathering is the dominant condition (Figure 5). The majority (72.2%) of bones are lightly weathered and only 1.1% are heavily weathered. A substantial number (26.7%) exhibit medium weathering. These patterns suggest that the bones were discarded and buried soon after consumption or use as tools, but not immediately.

The preservation quality of an assemblage can be evidenced by the sizes of element fragmentation (Table 6). At Tell Aphek, just over half (n = 141) of the butchered bones represent ≤ 25% of the whole element. Additionally, 31% (n = 87) represent ≤ 10%, and 79% (n = 223) represent ≤ 50% of the whole element. Only 13% (n = 37) bones are complete. These numbers can skew the presentations of elements and taxa. Complete specimens are typically small compact bones, such as tarsals, carpals, astragali, and phalanges. These tend to be more resistant to fragmentation and degradation.

Table 6. Fragmentation patterns of the butchered assemblage.

The butchered assemblage has not undergone significant heat treatment, as indicated by the thermal alteration (color) patterns (Figure 6). A total of 46 bones (16.3%) displays thermal alteration with colorations of black (n = 22, 7.8%) and brown (n = 21, 7.5%) being the most prevalent burning patterns. Of note are burning patterns during the MB IIA where 32 (28.1%) of the 114 bones were subjected to some level of heat treatment (burning).

The butchered assemblage contains a relatively high number (n = 90, 32.0%) of bones that underwent additional modification after carcass division or consumption (Table 7). Of these, it includes bones identified as tools (n = 83, 92.2%), game pieces (n = 4, 4.4%), ornaments (n = 2, 2.2%), and vessels (n = 1, 1.1%). The “tools” are mostly the fractured remains of bones that exhibit high polish and probably functioned as handles, awls, or spatulas. These are generally distributed across the periods proportionally to the number of butchered bones, as most occur in the MB IIA and LB II.

Table 7. Worked bone patterns of the butchered assemblage.

5.3 Chronological distribution of butchering marks

A total of 321 butchering incidences are distributed across the 281 bones (Table 4). Their distribution is skewed temporally, reflecting a corresponding pattern to the butchered bone counts. The most abundant are from the MB IIA (n = 129, 40.2%) and LB II (n = 118, 36.8%), followed by the EB I (n = 21, 6.5%). The total strokes, often several per incidence, generally adhere to the same distribution pattern as the NISP and NBI, although the amount in the EB I is significantly increased (15.5%).

There is little diversity among the butchering mark types over time (Figure 7 and Table 8). Slices are the most abundant (n = 293, 90.9%), while chops (n = 12, 3.7%) and scrapes (n = 5, 1.6%) are more infrequent. A lone saw mark, made through repeated slicing in a single groove, was also identified. The dominance of slices is typical of butchered assemblages of the southern Levant, but other contemporary sites tend to contain a greater diversity of types (e.g., bashes, nicks, shaves), albeit in relatively low frequencies (e.g., < 5%) (Beller and Greenfield, 2023, 2024; Beller et al., 2021a, 2022; Greenfield and Brown, 2016; Greenfield et al., 2016a; Okaluk and Greenfield, 2022).

Table 8. Distribution and types of butchering marks.

5.4 Butchering process

All five stages of the butchering process are represented among the butchered assemblages (Figure 8) (Rixson, 1989). Disarticulation is the most prevalent activity in each period, almost always exceeding 50% frequency. Most disarticulation occurs on the distal and proximal portions of bones or along the mandibles. Fileting typically appears in high frequencies (>20%) as well and is found predominantly on long bone and rib shafts. Skinning and dismemberment occur in much lower frequencies. The former being represented by perpendicular slices around lower limb bones and crania, while the latter by chops that divide natural portions indiscriminately. Toolmaking and headgear (antler/horn) removal only have a few occurrences across the entire butchered assemblage. Other activities present at contemporary sites, such as slaughter, and marrow and brain extraction, are not represented within the assemblage.

5.5 Raw materials of butchering implements

Identifying the adoption and utilization of implement raw materials in the butchering process is also achieved from an examination of the butchering marks (Figures 9, 10 and Tables 8, 9). While the Bronze Age sequence is not perfectly continuous, it is apparent that the introduction of metal for the butchering process occurs sometime between the EB II and MB IIA. Scrutinizing the butchering patterns from the EB and MB subperiods, as well as those of later periods, reveals the potential effects of raw material (stone vs. metal).

Table 9. Occurrences of raw materials for implements in the butchered assemblage.

Throughout the EB I-II, stone dominates (>90%) the raw material type for butchering implements. Only a single specimen was tentatively identified as a metal incidence, but it is possibly intrusive or misidentified. In general, it can be concluded that there is no evidence for the use of metal butchering implements at Tell Aphek during the EB.

There is a dramatic change at some point between the EB II and MB IIA. The absence of faunal material in the intervening periods as well the low sample size for the EB complicates drawing firmer conclusions. As the intervening periods are missing at Tell Aphek, we have to rely on the data from other sites to determine when metal first appears. Sites that include remains from the later EB III and EB IV/IB demonstrate that metal slices on bones made by metal implements are either very rare or non-existent, and limited to the latter half of the EB IV/IB (Table 1) (Greenfield et al., 2022a,b).

At Tell Aphek, by the mid-MB (IIA and IIB), metal implements are extremely common in the butchering process. In fact, metal butchering implements dominate the frequencies of raw material (MB IIA = 85.3% and MB IIB = 84.8%). This dominance continues in every period after the MB. For example, the LB II has the second highest frequency of butchering incidences, of which 89.8% were made by a metal implement.

The characteristics of butchering marks and associated activities suggest that there are, at minimum, three main implement types (Table 10). These are chipped stone flakes/blades, metal axes, and metal knives. The stone flakes/blades and metal knives created the slices and scrapes, used mainly for disarticulation, fileting, skinning, and tool making. Meanwhile, the metal axes caused the chops, all of which were performed for dismemberment.

Table 10. Raw material of implements used to make butchering marks.

5.6 Butchering efficiency

The long-term perspective allows us to monitor changes in butchering efficiency across the Bronze Ages. As discussed earlier, the calculation of the BEI in this study only involves slices, as these are present in all periods in abundance (Equation 1, Table 11).

Table 11. Butchering efficiency index for slices at Tell Aphek.

The higher the BEI value, the lower the butchering efficiency.

The results indicate a gentle increase in butchering efficiency from the EB to the MB. The NBI sample sizes of many periods are under 10 and difficult to compare. The EB I (15.9 BEI) is the largest sample at 21 NBI. Even when all the EB phases are combined (373 NBI/31 total slices = 12.0 BEI), they indicate lower efficiency than during the MB. This pattern is found in the study of contemporary assemblages as well, where an overall increase in efficiency is observed at Tall Zirā'a (Greenfield et al., 2022b) and Tell Halif (Greenfield and Brown, 2016).

The relatively higher butchering efficiency during the MB IIA (6.4 BEI) and MB IIB (6.3 BEI) remains consistent even into subsequent periods. For instance, in the LB II, the BEI is 6.3.

6 Discussion

An examination of the butchering marks, specifically the slices, allows us to trace the adoption of new technology in both time and space, and speculate on the expertise of the butchers. This study of the butchered assemblage from Tell Aphek provides insight into these aspects through an examination of groove morphology by scanning electron microscopy. Based on these results, we suggest five main takeaways.

• The Early Bronze Age as a continuation of the Stone Age: The butchered bones from the EB contexts retain butchering marks made almost entirely by stone implements as only one incidence may be categorized as metal, and any single outlier is highly suspect. The high frequency of chipped stone tool marks during the EB indicates the butchers' reliance on stone implements and is consistent with the analytical results from other contemporary industries (Beller et al., 2021b; Manclossi et al., 2019; Rosen, 1983). Consequently, the EB inhabitants remained in a Stone Age milieu. The general abundance of high-quality chert in the southern Levant provides people with raw material that can readily be reduced into formal slicing tools and/or expedient (ad hoc) flakes for various purposes (Figure 11) (Milevski, 2011b; Mozel, 2000; Rosen, 1983). To date, all studied chipped stone material from the Bronze Age levels is made from chert (Mozel, 2000). By contrast, obsidian does not occur within the southern Levant and must be imported, typically from Anatolia (e.g., Rosen et al., 2005; Shimelmitz and Adams, 2014). As a result, obsidian lithics were rare and probably costly items to use for such quotidian activities as animal butchering. The fact that stone continues to be used in the region, even until the Ottoman period (Greenfield and Brown, 2016), demonstrates its utility and convenience.
• Metal tools provide greater butchering efficiency. There is a noted increase in butchering efficiency that coincides with the introduction of metal technology. This pattern can be explained by two related factors: the introduction of metal technology and/or presence of specialized butchers. While correspondence between the increased butchering efficiency and the introduction of metal into the butchering process is apparent, it is difficult to demonstrate that specialized butchers are ubiquitous at these sites. Increases in butchering efficiency are evident at other Bronze Age sites within the southern Levant, although their values and timing vary. For example, there is only a minor increase in butchering efficiency across the EB-MB divide at Tall Zirā'a (in Jordan), as metal tools typically account for just over 50% of the implements (Greenfield et al., 2022a,b). Also differently, at Tell Halif, it is not until after the LB that metal tools become dominant and butchering efficiency increases (Greenfield and Brown, 2016).
• New (metal) tools affected the butchering activities. The introduction of metal into the butchering process has a profound effect on the butchering practices. The most notable example is that of the occurrence of dismemberment. This activity is evidenced by chops marks from metal axes (Greenfield, 2013; Greenfield et al., 2022a). These tools effectively sever carcasses, reducing the need for extensive disarticulation. For example, on Bos taurus, dismemberment occurs on humeri and innominates after the EB, whereas more extensive disarticulation occurred during the EB. Similarly, on Capra hircus innominates, Ovis/Capra and Sus scrofa dom. ribs and vertebrae, and Sus scrofa dom. crania, dismemberment only begins with the MB. The lack of chop marks prior to the MB is possibly a matter of indeterminate mark morphology left on bones by chipped or ground stone axes. A recent comparative study of chop marks by different raw materials demonstrated that stone chop marks are more irregular and often resemble a bash with conchoidal fractures rather than the typical metal V-shaped impression (Okaluk and Greenfield, 2022).
• Metal adoption was staggered throughout the Bronze Age. During the Bronze Age, the southern Levant experienced significant developments in metallurgy. The spread of bronze tools is not uniform across the southern Levant. It begins with ornaments, weapons, and jewelery among the elites during the EB and MB I (Genz, 2000; Philip, 1989). From using butchering data as a proxy for metal use, it is evident that peasants did not have access to even the simplest metal tools until quite late during the Bronze Age. It seems that by the MB IIA, metal tools begin to be available to the inhabitants of the largest most densely populated cities, as evidenced by this current study. By the MB IIB, inhabitants in smaller and more remote settlements began to acquire access to metal tools, but the adoption process was not fully completed until the LB (e.g., Tall Zirā'a) (Beller and Greenfield, 2023; Greenfield et al., 2022b).
• Large-scale metal production proceeded widespread adoption. Stone and metal tools differ significantly in availability, efficiency, and production requirements, influencing their use in butchering (Table 12) (Buck, 1982; Dewbury and Russell, 2007; Disa et al., 1993; Greenfield, 2002). Stone tools are generally locally sourced, easily produced, and require minimal energy or specialization, allowing household members to make sharp-edged flakes or blades as needed. However, these tools demand frequent sharpening through time-consuming retouching, making them less efficient for heavy-duty tasks. In contrast, metal tools cut through soft tissue more efficiently and require less sharpening (e.g., retouching on stone). Although recyclable, metal tools need specialized skills and facilities for production or repair, which are often unavailable locally. Additionally, the raw materials for metal tools may need to be imported, adding logistical challenges. While metal tools are more effective for butchering, their production is complex and resource-intensive compared to the simpler, more accessible stone tools. These factors were probably overcome by the mid-MB, as metal production had reached a scale where it could service the largest urban centers. Subsequently, it took longer for them to be adopted in the smaller centers, towns, and villages across the southern Levant.

Table 12. Comparative attributes of stone and metal tools for butchering purposes.

Table based on collective observations from Buck (1982); Dewbury and Russell (2007); Disa et al. (1993); Greenfield (2002).

7 Conclusion

The Bronze Age in the southern Levant experiences the rise of early complex societies, characterized by urban development, elite structures, craft specialization, and advancements in metallurgy. Initially, soft alloys of low As/Sn-bronze were used to make rare and prestige objects. This was later followed by stronger tin-bronze alloys (10% tin) throughout the region, enhancing their utility for quotidian tasks, such as butchering. Although metal tools gradually replaced traditional stone implements, the rate of adoption into wider society is not uniform and does not occur with the onset of the EB.

This study examined butchering marks on faunal remains from Tell Aphek, to trace the integration of metal technology into traditionally quotidian activities in the southern Levant. It was discovered that stone tools continue to be pervasive during the EB, suggesting limited or no access to metal and an enduring reliance on readily available chert (Tables 9, 10). Metal tools are only prolific in the butchering process beginning with the MB IIA.

When placed into the wider context of the southern Levant, the results from Tell Aphek contribute to a scenario wherein metal technology does not reach every settlement or every socioeconomic group simultaneously. Rather, metal tools are gradually integrated into society throughout the late MB. Initially limited probably to elites in larger cities, metal adoption in smaller settlements lagged, with this new technology only becoming more accessible to peasants and rural areas by the MB IIB and LB (Beller and Greenfield, 2023; Greenfield et al., 2022b; Greenfield and Brown, 2016). The staggered adoption of metal technology was probably related to the scale of metal production, as it required complex resource extraction and intensive labor compared to expedient or ad hoc stone tool manufacture (Avilova, 2009; Genz, 2000; Levy et al., 2002).

Metal technology had a profound effect on many industries and domains of Bronze Age society, including animal processing and butchering. The axes were effective at dismemberment, allowing faster carcass division and reducing extensive disarticulation. Tools lasted longer and were more durable, and butchering efficiency was greatly improved, meaning fewer strokes with knives to remove flesh. However, the correlation between metal tools and specialized butchers remains unclear and probably varied from site-to-site in the region.

Studying ancient faunal assemblages is crucial for understanding past diets, culling profiles, subsistence strategies, and technology. Butchering marks, in particular, are useful for reconstructing the butchering practices, such as the manner in which animals are slaughtered, skin removed, bodies dismembered and disarticulated, and meat fileted off the bone. These same butchering marks leave behind diagnostic features of the raw material that created them, allowing archaeologists to track the spread of metal technology. By examining the interplay between metal and stone use with zooarchaeological data, we gain insights into the complexities of cultural and technological evolution in the southern Levant. As similar studies are added to the growing regional dataset (i.e., Table 1), a more complete scenario of metal adoption will be attained.

SEM has become an indispensable tool for the study of bone surface modifications, particularly when the aim is to distinguish the raw material and technological attributes of butchering implements. Traditional low-power or even high-power optical microscopy is often insufficient for confidently differentiating between cutmarks produced by metal or chipped stone implements, carnivore damage, and post-depositional abrasion. SEM overcomes these limitations by providing high-resolution, three-dimensional imagery that captures striations and groove morphology with exceptional clarity. Furthermore, the use of silicone molds permits repeated, non-destructive examination of the same grooves, enhancing reproducibility and allowing multiple analysts to assess the same microfeatures.

The application of SEM to bone modification studies is particularly valuable in regions and periods where metal adoption was gradual and staggered, such as the Bronze Age southern Levant. Because metal objects were frequently recycled or preserved primarily in elite or ritual contexts, butchered faunal remains offer one of the few avenues for assessing the availability and use of metal tools. SEM-based evaluation of tool-induced grooves, therefore, serves as a powerful means to monitor evaluation for technological change within quotidian activities, precisely the domain often overlooked by traditional artifact-based approaches. At Tell Aphek, SEM analysis was critical for detecting the abrupt transition from stone to metal butchering implements between the EB and MB. Without SEM, many metal slices, especially those produced by dull blades, would have been indistinguishable from very sharp stone tools. Thus, SEM continues to be an integral technique for archaeologists.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Palmira Saladie, Institut Català de Paleoecologia Humana i Evolució Social (IPHES), Spain

Reviewed by: Francesc Marginedas, Institut Català de Paleoecologia Humana i Evolució Social (IPHES), Spain

Gabriel Cifuentes-Alcobendas, Institut Català de Paleoecologia Humana i Evolució Social (IPHES), Spain