The function of the south-Levantine Late Chalcolithic and Early Bronze Age basalt vessels bearing circumferential depressions: Insights from use-wear analyses

One of the most characteristic aspects of the Late Chalcolithic and Early Bronze Age periods in the southern Levant is the appearance of large assemblages of basalt vessels. These vessels, frequently meticulously made, appear sometimes a considerable distance from the raw material sources and are found mainly at habitation sites. While these and their prestigious value have been widely discussed in the past, their function is still obscure. In the current paper, we address their functionality through microscopic use-wear analysis. Emphasis was placed on basalt vessels with a distinct wear pattern–circumferential depressions, which appear along the perimeter of their interior bases. The documented traces were compared to results of an experimental study we conducted to characterize the effects of abrasion, grinding, and lubrication on basalt surfaces. The results of the comparative experimental study suggest that the circumferential depression was formed from a repetitive rotational activity using a narrow-ended tool. Further, it seems that two material types acted in combination as the circling device and processed material. One was hard and abrasive, such as stone, and the other was semi-resilient, such as wood or mineral powder. Water was likely used as a lubricant in the rotational process. While the actual function of the bowls bearing the circumferential depressions is not entirely clear, the use-wear analyses suggest that they may have been devices involved in craft industries, used for processing materials unrelated to food (minerals in particular). Whatever the exact function was, it clear that this use continued from the Chalcolithic through the Early Bronze Age, providing evidence for functional continuity between these two periods.


Introduction
The protohistoric southern Levant is characterized by notable economic and technological developments and important social changes [e.g., [1][2][3]  with either a flat base or a fenestrated stand were prevalent [e.g., 23, 70,71], occasionally exhibiting incised triangular decoration on the interior rim or extensive external geometric decoration [18,69]. They usually constitute the largest group of ground stone artefacts (between 48.7-62.5% for the published Late Chalcolithic assemblages included in the analysis). During the Early Bronze Age, the number of basalt vessels declined (down to between 8.9-22.7% based on the published Early Bronze Age assemblages included in the analysis), and more ordinary coarse vessels, mainly produced from limestone, became more common [e.g., [72][73][74][75][76][77][78]. This trend hints toward a recession in the basalt vessel industry. Accompanying this trend were changes in form and style. Some of the typical Late Chalcolithic forms, like the full or fenestrated pedestals, disappeared [however see Fig 2 in 79], and flat-based vessels with flaring walls and thick bases became dominant, alongside other distinct types such as the four-handled vessels [65 and references therein]. The vessels are also sporadically distinguished by their rare decoration formed in relief (also called a 'necklace' pattern, characterized by a 'crested ridge band' [70] or chain of knobs encircling the rim [e.g., 61,62,70]. Despite these changes, the specialized basalt vessel industry continued to form an important part of the Early Bronze Age I craft [e.g., 61,62,70]. All vessel types were found dispersed throughout modern day Israel and Jordan, although the four-handled vessels were by and large limited to the northern sites [65]. This wide distribution occurred despite the restriction of suitable basalt sources in the southern Levant to the Jezreel Valley, Galilee, Golan Heights, the Jordan Valley, and parts of northern Jordan and southern Syria [80][81][82][83][84][85].
The first stages of the vessel production likely used a hammerstone and chisel to roughly shape the external form of the vessel and hollow its interior, while the later stages used meticulous pecking and smoothing, which in many cases camouflaged the rougher production marks [23]. Pecking and stone-to-stone abrasion has been proven the most efficient and feasible techniques employed in a final stage of basalt tool manufacture [e.g., [86][87][88][89][90]. The production technology, especially when applied to produce vessels this large and symmetrical, with proportionally thin walls, great level of surface refinement, and sophisticated decoration, was incredibly time consuming and required a great deal of skills and experience. Therefore, it seems that these vessels were likely manufactured by specialists in workshops, located at or near basalt sources, and circulated through complex distribution networks [e.g., 3, 20, 23, 61-63, 70, 91-93].
Notably, a small fraction of the Late Chalcolithic and Early Bronze Age basalt vessel assemblages bear a specific feature: a circumferential channel-like depression (Fig 1, marked by arrows) that is located inside the bowls at the joint of the wall and base. This phenomenon is observed on up to only 14.3% of basalt vessel bases in a basalt vessel assemblage according to the selection of fully published sites (with the maximum number of four vessels with circumferential depressions identified in the assemblage of Tel Bet Yerah).
The current paper explores this unique feature through the application of use-wear analysis and experimental procedures. We present our observations and the results of the experiments we conducted and discuss several interpretive lanes regarding the mechanisms that may have formed this unique pattern. Finally, we address how this contributes to our understanding of the Late Chalcolithic and Early Bronze Age basalt vessels' function.
The depression is a shallow channel-like surface (Fig 2A), appearing along the perimeter of the vessel interior base where it adjoins the interior wall. In the most pronounced variation, the worn surface appears as a relatively deep circumferential depression with clear abrasion marks, parallel longitudinal striations, and highly reflective polish observable to the naked eye, especially on the bottom of the depression and its exterior margin ( Fig 2B). In the least pronounced variation, the worn surface is particularly shallow. The depression is clearly distinguished from the central area of the interior base (Fig 2C), which is usually elevated in comparison to the perimeter.
For the current study, we selected 15 vessels bearing a distinct circumferential depression, which originated from eight Late Chalcolithic and seven Early Bronze Age sites in Israel ( Fig  3). These vessels originated from the recent excavations of the Israel Antiquities Authority or its storage facilities located in Beth Shemesh. No permits were required for the study, which complied with all relevant regulations.
The vessels were photographed, and their metric attributes were documented. The morphometric attributes of the circumferential depressions were measured using a digital caliper with a precision of 0.01 cm to calculate the average (using the minimum and maximum measures) depth and width. Further, five base fragments exhibiting pronounced circumferential depressions (three dated Late Chalcolithic and two to the Early Bronze Age I) were selected for comprehensive use-wear analysis. Of these, three specimens represent flat-based V-shaped bowls from the Late Chalcolithic sites of Yehud (YH9 [119], YH42 [120]) and Giv'at HaOranim (GHO1 [121]), and two fragments represent an upright/flaring bowl with a narrow base from Early Bronze Age I contexts of Tel Bet Yerah (TBY37 [77]) and a four-handled bowl from Kabri (KB; Israel Antiquities Authority collection) dated to the Early Bronze Age I. All the sampled vessels were produced of fine-grained compact basalt with olivine inclusions tightly embedded in the groundmass and minor-to-no natural porosity.
We applied the methodological framework of use-wear analyses, in which the characteristics of wear were defined using a list of attributes and then replicated through experimentation to infer on the mechanisms of their formation. We followed the protocols and attributes outlined by Adams [103,104] for the low-power analysis and Dubreuil and Savage [88,108] for the high-power analysis, with minor adjustments in terminology ( Table 1).
The analyses were conducted at the Laboratory for Ground Stone Tools Research and at the Use-Wear Analysis Laboratory, both at the Zinman Institute of Archaeology, University of Haifa. An integrated observational approach, combining low-and high-power observations, was employed. For the low-power observations (at magnifications of 10-50x), a Zeiss DIS-COVERY V8 stereomicroscope was used with the aid of a fiber optic light to provide a side-light source when presenting topography. High-power observations (at magnifications of 50-500x) were conducted using a Zeiss Scope.81 metallographic microscope. Photographs were taken using Z-stacking with a range of 10 to 30 shots that were merged into a full depth images, depending on the topography of the surface.

The analysis procedure
The analysis began with a thorough microscopic examination of the vessel surfaces to detect possible residue. As no residues were detected, the selected archaeological artifacts were

PLOS ONE
The function of the south-Levantine Late Chalcolithic and Early Bronze Age basalt vessels washed in an ultrasonic tank filled with room-temperature water for 15 minutes to remove soil and crust (natural calcite deposits on the basalt surface). When dried, the microscopic observations used first low and then high-power observations. The unmodified breaks of the sampled archaeological pieces were observed first to assess post-depositional alternations. Following this, observations were applied to the circumferential depressions, the elevated central

Sinuous
Topography profile with rounded and gradual transitions between low and high areas

Uneven
Topography profile with abrupt, rapid, and inconsistent transitions between low and high areas

Regular
Topography profile with minimal variation in the elevation of adjacent grains LEVEL 2

Irregular
Topography profile with pronounced variation in the elevation of adjacent grains

Comet-shaped pit
A pit with a single worn side resulting from grains dragged in a single direction, indicating unidirectional abrasion

Polish
Visible alteration to the natural surface that increases its reflectiveness through leveling the microtopography and tribochemical reactions between working materials

Pecking
Conglomerates of steep shallow pits; the wear increases the dimension of low topography

Striation
An isolated, intermittent, or continuous track on the basalt surface created by grain dragged along the surface. May occur (regarding density) separated, close, or connected and (regarding spatial distribution) loose, covering, or concentrated area of the bases, the exterior surface of the bases, and the preserved exterior wall of each of the five vessels analyzed. The interior walls of the sampled vessels are not preserved so were not analyzed. Various spots exhibiting distinct features were photographed for each vessel to show the variations in traces, and these features were defined using the list of attributes presented in Table 1.

The experimental procedure
The experimental program designed for the current study was aimed particularly at replicating the mechanism for the formation of the traces observed on the archaeological specimens, associated with the formation of the circumferential depressions (surface abrasion, parallel striations, and polish). The preliminary observations conducted on the archaeological samples showed that the wear within the circumferential depressions is indicative of a friction between two parallel surfaces; therefore, all the experiments were conducted using grinding. We decided to test basalt against (1) high-asperity rock types to study the abrasion of basalt groundmass, (2) various types of plants to check how and to what extent grinding organic products of various hardness alters the compact basalt surface, and (3) lubricants to understand their influence on the abrasion mechanism and the development of micropolish. The basalt used for the experiment was collected from the Golan Heights. The region constitutes the westernmost part of the Transjordanian Harrat Ash Shaam volcanic province, which is rich in high-quality basalt. The area likely served as the raw material source for production of many of the Late Chalcolithic and Early Bronze Age basalt vessels [64,81,93,122].
Since the basic assumption on which the first stage of the experiment was designed is that use-wear on stone vessels is formed on top of production traces, we overlapped the various wear procedures through sequential operations on the same stones. The experiments therefore included two stages. The first was an abrasion experiment, which was aimed to produce a smooth surface to simulate the vessels' production wear on which the use-wear is assumed to develop. The second was a grinding experiment where the abraded stones with the smoothed surfaces were used as the lower grinding stones, and various materials were grinded on top of them. This stage simulated the formation of use-wear on top of production wear. In this experiment, a subset of experiments was included to test the effects of lubrication on the interaction between the upper and lower stone and the use-wear. Lubrication is hypothesized to enhance the rotational action, which was assumed to be involved in the formation of gloss and the depression on the archaeological tools (Table 2).
In the first stage of the experiments, two highly abrasive and locally accessible stones were selected to create the smooth surface: well-cemented beachrock and aeolianite sandstone (also called kurkar). Both are composed of gravel, shells, and quartz sand, which make them particularly efficient in smoothing hard rocks. Quartz sand was also used as an additive, and altogether these materials were tested in various combinations. The selection of contact materials was driven by the microscopic observations of the archaeological vessels and their surface characteristics, cross-referenced with ethnographic and experimental studies on basalt tools manufacture and use. The experiments were carried out until a smooth surface was formed on the basalt pieces. The smoothed surfaces were documented and then used as the lower stones in the second stage of the experiments where three plant materials were ground (Fig 4). The plants selected represent different levels of pliability and moisture; fresh rosemary stems represent a relatively oily plant and a fresh wood, dry oat flakes represent a soft and dry flour-generating plant, and dry flax seeds represent lipid-rich seeds with a relatively hard shell ( Table 2). In this stage, for the grinding of the plants, raw pieces of compact basalt slabs served as upper grinding stones. These experiments were carried out until a handful of the processed end-product was acquired.
Finally, the lubricant experiments involved abrading basalt against sand and water, sand and oil, and wet limestone. Sand was used since it provides the abrasiveness and surface fatigue factors, which according to our preceding abrasion experiment ( Table 2) was successful in enhancing smoothing, with minor influence on the reflectiveness. The experiment with limestone and water aimed to create a shiny surface using both a lubricant and a soft but compact stone.
The duration of the experiments was decided based on our observations, up to a point when wear was observed with the naked eye. Before conducting the microscopic analysis, all experimental stones were soaked in soap water for at least five minutes (up to two hours in the case of stones involved in the experiment with oily materials) and cleaned for 15 minutes in an ultrasonic tank filled with room-temperature water. The traces were documented and analyzed following the same protocol established for the archaeological vessels. The final stage of analysis compared the wear patterns on the experimental pieces with the archaeological vessels. The results of microscopic observations were also crossreferenced with past experiments targeting use-wear on basalt tools [e.g., 88, 101-104, 108-110, 115].

The experimental pieces
The experiments produced wear linked to the hardness and asperity of the contact materials and lubrication (Fig 5). The use of abrasive rocks, with or without the addition of sand, for the preparation of the experimental lower stones produced diagnostic wear. In each case we obtained working surfaces that appeared well-flattened with a relatively dull reflection visible to the naked eye. Depending on both the applied method (beachrock vs aeolianite vs aeolianite or beachrock with the addition of sand) and the original topography of basalt rock, this task took 3-120 minutes. The specific characteristics of the wear for each stone are listed in Table 3.
Abrasion against aeolianite generally produced traces indicative for contact with a hard material (Table 3), including flat plateaus, pits with sharp edges, and leveled surfaces. The abrasion of aeolianite with the addition of sand caused a greater reduction of the surface matrix but a more rounded topography as sand enabled a more fluid motion against the aeolianite and caused the collapse of sharp edges and grain removal. Beachrock caused a more significant reduction of the matrix, creating a rough and irregular topography.
All the abraded pieces were used in the subsequent experiments involving grinding plants and lubrication, and the details of the wear are shown in Table 4. Processing plant materials removed traces produced during the abrasion experiments. Grinding oat flakes generated traces typical for highly pliable products; the surface that was originally flat became sinuous, with pits with rounded edges. In addition, patches of a flat, bright, and striated polish formed on protruding surfaces and phenocrysts (Fig 5D, marked by arrows), resulting from occasional direct contact with the upper stone. Grinding rosemary changed the sinuous profile of the surface into a flat one, with pits showing relatively abrupt edges and polish typical of woody plants ( Fig 5H). Grinding flax rounded and smoothed the previously rough surface, and elevation variation caused by the beachrock was reduced. A polish with the characteristics typical for beachrock abrasion covered the surface, with isolated patches of duller and sinuous polish from the flax grinding ( Fig 5L).
Grinding with sand mixed with a lubricant (Fig 6A and 6B) allowed a more fluid motion and generated particularly well-smoothed surfaces (to an even higher level than with dry sand), clearly enhancing the abrasion. Traces on the macro and micro-scale are more or less the same for the experiments with water or oil, with minor differences visible only under a very high magnification (500x). The polish produced by sand and oil (Fig 6B) is less reflective than polish produced by sand and water (Fig 6A), suggesting that water is the component enhancing reflectivity. In the case of the limestone and water (Fig 6C), the polish differs and is slightly less reflective than the polish formed by water and sand, exhibiting a combination of flat and striated patches (Fig 6C, shown by the red arrow) and smooth and reflective sections (Fig 6C, shown by the white arrow).
To summarize, the experiments provided the wear reference collection that allows to proceed with the analysis of the archaeological artifacts. The most relevant conclusions include: 1. The use of sand generates a leveled topography with rounded transitions between plateaus and pits and less pronounced linear features (striations), while abrading compact abrasive rocks without the intermediate sand generates large plateaus with sharp edges and (in particular beachrock) is more likely to cause grain fracturing and irregular topography;  6. Grinding rosemary resulted, among others, in wear features indicative for basalt-to-wood interaction-in particular fine reticulation of opaque and bright polish on the most exposed protruding surfaces; 7. Grinding relatively resilient oil-rich organic products (rosemary and flax seeds) enhances the development of reticular polish on protruding areas of the basalt topography; 8. The presence of lubricant-both water and oil-enhances reticulation of the micropolish. Water additionally enhances its reflectivity; 9. The combination of abrasive mineral (sand or limestone powder) with lubricant both reinforces the abrasion of basalt surface and facilitates the development of reticular polish.

Archaeological samples
3.2.1 General observations. Based on the 15 examined vessels (including the five vessel fragments selected for use-wear analysis), the average depth of the circumferential depressions is 0.37 cm (standard deviation 0.33 cm), and the average width is 2.42 cm (standard deviation 0.59 cm). The average dimensions and the range of variation for each artifact are presented in Fig 7. The majority of the depressions oscillate between 1.40 cm and 3.50 cm in width and are up to 0.80 cm deep, with two particularly large exceptions, which are on average 3.75 cm wide ( Fig 7N) and 1.25 cm deep (Fig 7O).
The use-wear analysis, applied to the five selected vessel fragments (YH9, YH42, GHO1, TBY37, and KB), showed that the break areas of each of the vessels present little to no alteration, retaining the original structure of the raw material. It was therefore established that post-depositional processes did not significantly affect the appearance of production-or usewear.
The comparison of the surfaces of the exterior wall and base and the interior central elevated base to the circumferential depression revealed that the exterior walls and bases exhibit similar wear: abraded surfaces and pecking marks (Fig 8) that are less pronounced on the exterior base (Fig 8C) than the wall or the exterior wall-base joint (Fig 8A and 8B). Such wear is sometimes evident on the interior base as well, within the elevated center. The basalt topography in this area exhibits a more sinuous profile than within the circumferential depression, with pronounced and irregular pits. The pecking marks were largely abraded and leveled, likely during the final stages of vessel manufacture, aimed at forming clear smoothed surfaces [23,88]. They could be further worn down during the use-life of the vessel or through holding  Table 4).
https://doi.org/10.1371/journal.pone.0252535.g006 the vessel exterior or resting it against a surface. The use of the vessel interior for processing or storing would contribute to abrasion of the pecking marks visible on the central elevated base; therefore, they are evident to various levels in different vessels and entirely missing in the circumferential depressions.

The circumferential depression.
The circumferential depression differs remarkably from the other surfaces of the vessels. It exhibits no evidence of pecking, and it is more heavily abraded, with a flatter topography, smoothed and heavily striated surfaces, developed polish, and occasionally surface darkening. The traces tend to intensify towards its bottom and the exterior margin (Fig 2B) that adjoins to the vessel wall.
The five vessels sampled were described and discussed separately, focusing on characterization of the traces within the circumferential depression (Table 5). YH9 (Fig 9): the Late Chalcolithic base fragment shows a relatively superficial circumferential depression compared to the other vessels (Fig 7). The shiny appearance of the depression is notable and evident to the naked eye. On the macro-scale, traces in the depression are similar to those observed on the central elevated base (Fig 9A); however, in the depression they are

PLOS ONE
more pronounced (Fig 9B; Table 5). The polish is spreading on wide surfaces (Fig 9C), developed to a higher degree on elevated areas and associated with fine striations. The striations appear throughout the surface of the depression, parallel to its main axis. Our experiments suggest that such a moderate change in surface topography combined with the spreading of the polish may result from contact with a medium-hard material. YH42 (Fig 10): the Late Chalcolithic base fragment is nearly complete with partially preserved walls. The circumferential depression is the deepest among the analyzed vessels examined for use-wear (Fig 7). Compared to the central elevated base (Fig 10A), the surface within the depression exhibits a marked change, with wide striated flat surfaces and reduced pits ( Fig  10B; Table 5). The micropolish is generally developed to a relatively low degree, observed in patches on protruding topography (Fig 10C). Striations appear all over the circumferential depression, parallel to its axis. Based on our experiments, this combination implies that a relatively hard material was rotated along the depression, causing a massive reduction of the basalt groundmass. The prominent reduction compromised the formation of a well-developed polish, allowing its development only in isolated areas. GHO1 (Fig 11): the Late Chalcolithic complete vessel base preserved with minor parts of the vessel's wall. The width and depth of the circumferential depression are typical when compared to the other examined vessels (Fig 7). Contrary to the central elevated base (Fig 11A), the circumferential depression exhibits massively worn flat surfaces with macro-and microscale parallel striations oriented along the depression (Table 5). Significantly, comet-shaped pits (Fig 11B) indicate a unidirectional motion of abrasion (Fig 11B). The direction of the comets corresponds with the direction of the striations (Fig 11C), and none of these appear on the central elevated base. The polish is developed along the protruding surfaces between the macro-scale striations, indicating contact with a hard material. TBY37 (Fig 12): the Early Bronze Age vessel is relatively small, with about one third of the base and a minor part of the wall preserved. The circumferential depression is rather wide and

PLOS ONE
The function of the south-Levantine Late Chalcolithic and Early Bronze Age basalt vessels deep (Fig 7). In comparison to the central elevated base where clear pecking marks are observed (Fig 12A), the depression exhibits massive abrasion characterized by a flat surface with darkening and striations (Fig 12B; Table 5). The polish is remarkably bright with a reticular distribution pattern (Fig 12C). This configuration implies contact with a hard material. Further, the brightness of the polish and its distribution has parallels to the lubricant experiment using sand and water (Fig 5A), suggesting that water was used as a lubricant here. KB (Fig 13): the Early Bronze Age base belongs to a four-handled bowl (Fig 13). The walls are preserved to the same height (ca. 0.5 cm) along the perimeter, and the break appears relatively even, which may suggest deliberate removal [65]. The vessel is exceptionally large compared to the others, and the circumferential depression (Table 5), although relatively shallow (Fig 7), is distinct and regular. The analysis showed a remarkable difference between the wear on the central elevated base and the depression. The transition between these two areas is abrupt and very clear. The center of the base is elevated with a flat cross-section (unlike most of other pieces that exhibit a convex central base with gently sloping sides), and it is covered with pecking marks (Fig 13A shown by the arrows). Within the depression highly reflective

PLOS ONE
The function of the south-Levantine Late Chalcolithic and Early Bronze Age basalt vessels polish and darkening are visible to the naked eye ( Fig 13B). The surface inside the depression is extremely flat and striated; the parallel striations reflect a continuous rotary motion along the base's perimeter (Fig 13B). The polish is highly developed, extremely bright, and rough, spreading in a reticular pattern (Fig 13C). It is associated with parallel micro-scale striations (Fig 13C). Based on our experiments, these traces suggest contact with a hard material, possibly involving a lubricant that would enhance the reflectivity and form the reticular polish (as in the case of the lubricant experiments with water-see Fig 6A and 6C). The wear is similar to those observed for TBY37 (described above, Fig 12).
The observations conducted on the archaeological samples in the context of the experimental results lead us to formulate several preliminary conclusions regarding production and usewear development on the basalt vessels: 1. The appearance of the exterior and interior surfaces of the vessels is the result of their production and use-life, and unrelated to the post-deposition of the artifacts;

PLOS ONE
The function of the south-Levantine Late Chalcolithic and Early Bronze Age basalt vessels 2. Pecking marks are the most prominent wear feature associated with the vessel manufacture, more evident on the vessel exterior wall and exterior wall-base joint, relatively less obvious on the exterior base and interior central elevated base, and virtually absent within the circumferential depression (Figs 8, 12A and 13A). It is therefore suggested that both high finishing (smoothing) and the continuous handling and use of the vessel caused the reduction of the pecking marks; 3. The experiments involving sand resulted in leveled basalt topography with rounded transitions between the plateaus and the pits and mainly superficial intermittent striations. The traces have parallels on the exterior walls, exterior bases, and central elevated bases of the archaeological vessels. Therefore, we suggest that sand was used for surface abrasion in the final stage of the vessel production; 4. The wear observed within all the circumferential depressions is analogous: characterized by substantial reduction of the basalt matrix (surface darkening), leveled topography with pronounced plateaus, continuous, parallel striations (Figs 9B, 9C, 10B, 11B, 12B and 13B)

PLOS ONE
The function of the south-Levantine Late Chalcolithic and Early Bronze Age basalt vessels covering the high topography (plateaus) and distributed along the depression, and the presence of thick, opaque, bright micropolish, in particularly pronounced cases developing into a reticular pattern (Figs 9C, 10C, 11C, 12C and 13C). The transitions between high and low topography are abrupt or slightly rounded, and the bottoms of the pits are largely unmodified.
5. The surface leveling, the presence of continuous striations, and grain edge leveling/rounding are features indicative for abrasive wear formed by friction between two surfaces (grinding). Based on the experiments (Figs 5A, 5B, 5E, 5F, 5I, 5J and 6C), the flat topography of the circumferential depressions, associated with the wear development mainly on the protruding surfaces and prominent striations are a result of contact with a hard compact material of highly abrasive properties-likely a high-asperity stone; 6. The thick, opaque, and bright micropolish within the circumferential depressions exhibits irregular-to-sinuous morphology and occasionally develops reticulation. It spreads mainly on the elevated areas in between macro-scale striations, either in isolated patches on topographic peaks or in a covering manner (Figs 9C, 10C, 11C, 12C and 13C). It suggests that while the striations are the result of a contact with an abrasive stone, the micropolish resulted from an interaction of basalt with a different type of material. Based on our experiments (Figs 5G, 5H, 5K, 5L and 6C), this type of polish results from the contact with a semi-resilient material. While the experimental results are inconclusive in this matter, we are inclined to propose the involvement of a fine mineral powder as our main suggestion. Contact with wood may also be considered as it shares some of the polish characteristics;

PLOS ONE
The function of the south-Levantine Late Chalcolithic and Early Bronze Age basalt vessels 7. Based on our experiments (Fig 6), during the formation of circumferential depressions water was involved as a lubricant, enhancing both reticulation and reflectivity of the micropolish; 8. The wear intensity within the depressions (depth of the depression, the degree of surface leveling, the degree of micropolish development and the presence of surface darkening) differs between the vessels. It is suggested these differences are the results of varying intensity and duration of the activity that formed them;
The results of the current study indicate that these were not just storage or serving vessels; rather, at least some of these were intensively used as a platform for a specific activity. The results clearly show that all the analyzed vessels reveal an analogous pattern of wear distribution, where wear within the circumferential depressions is significantly different from the wear on other tested areas of the vessels. These other areas-exterior base, walls, and interior central elevated base-exhibit wear traces indicative primarily of the vessel manufacture. In contrast, the wear within the circumferential depressions exhibits a distinct orientation, following the axis of the depression, and the surface is flat and highly reflective. This implies that the activity that formed these traces was limited to the base perimeter, with slight or no effect on adjacent surfaces of the interior of the vessel. Considering the dimensions of the depression and the orientation of the traces within, a narrow-ended tool (ca. 1.5-3.5 cm in width, see Fig 7) was likely used and consistently performed a rotational movement. This activity used the vessel wall as a retaining barrier for the force of a rotating device (Fig 14A), leaving a distinct channel-like depression. The experimental results indicate that the traces in the depression were formed through friction between the basalt surface and two other types of material in contact: a hard and highly abrasive stone and a semi-resilient substance like a mineral powder (e.g., stones and minerals that were processed for a paste) or perhaps wood. Furthermore, our results indicate that water was added either as a component of a worked material or an additive, which was essential to enhance the formation of the depression.
The combination of these two wear types reflects an activity where a hard rotating tool was used to process a semi-resilient material (or vice-versa, a wooden tool was used to process a highly abrasive mineral) with water. This is particularly true of four of the analyzed vessels (YH42, GHO1, TBY37, and KB), which exhibit massive matrix reduction, continuous parallel striations, and poorly altered low topography. Moreover, last three attest for abrupt pit edges (Figs 11B, 12B and 13B), grain leveling (see Table 5), and surface darkening (Figs 11-13, shown by white arrows on the macro-photographs), which are typical of intensively abraded surfaces.
One vessel shows evidence for unidirectional rotation (GHO1, Fig 11B), exhibiting cometshaped pits, while others may have been worked bidirectionally, suggesting alternating motion of the rotating device. Notably, one archaeological sample (YH9, Fig 9) displays a prominent surface rounding with minor reduction of the basalt matrix, finer striations evident only under high magnification, and additional type of thin, translucent polish spreading within lower topographic areas (Fig 9C). The indications of rotational abrasive activity, however less pronounced, are still present in this sample; it is suggested that the rotating tool was likely made of more pliable material than in the other examples, exhibiting lower asperity. This would also explain the superficial nature of the depression in comparison to the other tested vessels (Fig 7A).
Looking at the characteristics of the wear, the results allow us to address the general functionality of the basalt vessels. In terms of context, the circumferential depression is not a widespread feature, appearing inconsistently on a small fraction of the Late Chalcolithic and Early Bronze Age vessels, with no association to a particular vessel type and size or geographic terrain. We acknowledge that context alone is a poor indicator of vessel function, especially when discussing vessel fragments. However, because basalt bowls are often assigned a prestigious value and ritualistic function based on their presence in mortuary assemblages, we consider it important to mention that the majority of basalt vessels and fragments were discarded within settlements-associated with domestic activity layers or deposited in subterranean complexes of unknown function [e.g., 97], and some show signs of intensive utilization. Many were uncovered inside shafts and waste pits located within or nearby the habitation areas of settlements [e.g., 23, 124,125], which are often filled with debris characteristic of domestic activities-pottery fragments, animal bones etc. The vessels are mostly found broken, and there is no clear link between these fragments and food processing or storage.
This contextual variation and the variation of the traces in the basalt vessels suggest that the vessels were used for an array of functions. For example, vessels are traditionally viewed as part of the food processing industry, and although the current study cannot definitively link basalt vessels to the processing of staple crops characteristic of the Late Chalcolithic and Early Bronze Age-cereals, legumes, and olives [e.g., 11,12,15,46], it is possible that the analysis of other vessels may reveal connections to food processing. Indeed, there are few examples where the basalt vessels occur alongside food processing and storage apparatuses, like grinding stones, pestles, cooking installations, pots, and storage jars [e.g., 95,126,127]. This includes one example of a four-handled basalt vessel fragment with a circumferential depression that was found in situ in association with several grinding slabs, a pestle, and multiple fragments of large storage and cooking vessels and small serving bowls [see Fig 39:1 in 128].
The dimensions of the depression reflect, in part, the dimensions of a pestle or any other tool that was operated in these vessels. However, the relatively rare Late Chalcolithic and Early Bronze Age pestles [e.g., 61,129,130] are all much wider than the identified circumferential depressions, so they could not act as the tools that created them. The lack of stone pestles in the assemblages and their incompatibility combined with evidence for activities related to abrasion, crushing, and rotary movements in many limestone and basalt vessels drive scholars to suggest that wooden pestles were used [e.g., 77]. Based on the use-wear analysis (Fig 5G and  5H), wood might potentially be one of the materials involved in forming the polish inside the circumferential depressions, but evidence for this is so far insufficient.
Alternatively, the analysis suggests that at least some of the basalt vessels were used to process materials unrelated to food in various craft industries. This is partially supported by contextual data. Basalt vessels in situ were occasionally found in a direct association with copper smelting installations and waste (e.g., Ashqelon Afridar, Ashqelon Barnea [40,61,96,131]) and it has been suggested they were used for the copper ore processing [61,131]. The example from Ashqelon Barnea includes one complete flaring bowl likely bearing a circumferential depression [see Fig 5:8 in 61]. In addition, some vessels may reveal evidence for pigment processing (with one published example bearing residue of a red pigment [77]). Pigments-especially red-brownish ones-were used for decorative purposes during the Late Chalcolithic and Early Bronze Age periods, in particular for painting pottery vessels, figurines, and ossuaries, and occasionally-during the Chalcolithic period-wall paintings [e.g., 16,[132][133][134].
The consistent character of the rotary phenomenon suggests that the vessels were used in an activity involving repetitive turning or stirring where wear features accumulated along the course of the depression. A related possibility views the circumferential depression as a type of intentional vessel design. It could be roughly formed in the production stage to facilitate a secure rotational movement, which would be especially helpful when working with a lubricant or a powder. The pre-formed depression would help restrict the tool's movement to the circumference and would explain the confined distribution of the wear (Fig 14B). It would also allow the processed material to accumulate with ease in the base perimeter. The majority of the circumferential depressions are merely a few millimeters deep (see Fig 7), so their formation could also result from use alone; however, pre-shaping would explain the peculiar regularity of the wear. It is important to emphasize though that even if the channel-like features were present before vessel use, the wear within them was formed entirely during further extensive and repetitive use that also certainly impacted shape and depth of the depressions.
The wear characteristics may relate to different crafts that required processing hard materials with the addition of water, such as minerals and clay. The closest example of similar micropolish comes from the stone abrader found in the Natufian mud-plastered grave at the site of Hilazon Tachtit [Fig 10 in 113, 135]. According to the authors, this type of wear was created from abrasion against a flexible yet rough surface like leather-hard clay (in this case mud-plastering), and it was compared to traces identified on pebbles used for pottery burnishing [136]. Considering this, the use-wear identified in the circumferential depression may be in some way related to clay/pottery manufacture.
The pottery industry during the Late Chalcolithic and particularly during the Early Bronze Age advanced with typo-morphological standardization, specialized production in workshops, and interregional distribution [e.g., 3,16,19,20,137,138]. Partially tournette-formed ceramic vessels [139][140][141] were attested already in the Late Chalcolithic period, and their number increased significantly during the Early Bronze Age [61, 139, 141 and references therein]. Still, it was not until the Early Bronze Age II when potter's wheels appeared in significant frequencies [e.g., 137,141,142].
An interesting related, yet still unsupported suggestion, is that the basalt vessels with circumferential depressions were involved in the pottery industry, used for rotating a vessel while coiling/shaping. The circumferential depression would be then formed by using the vessel base as a turntable on top of a smaller vessel or a narrow socket made of a hard abrasive material (i.e., the rim creating the circumferential depression while rotating; Fig 14C), respectively in place of upper and lower potter's wheels, which are conspicuously missing from the Chalcolithic material culture repertoire. Similar suggestion was made by Rosenberg [143] in relation to a perforated basalt vessel base from Ashqelon Barnea. This would explain the uncommon nature of the circumferential depressions within the vessels versus their broad geographic distribution across the Late Chalcolithic and Early Bronze Age southern Levant.
To conclude, it should be stressed that all of the tasks mentioned above relate to a processing activity that required rotary movement and perhaps benefited from the interaction between the 'active' tool and the vessel. The rotation was conducted with a high level of consistency and regularity, and it was probably enhanced by water and/or an intermediate material.
Based on the characteristics of the depressions and wear, this action was likely achieved using a compact and narrow abrasive object. While minor differences in the tools used or the materials processed may exist, the same phenomenon was applied to all vessels. This activity was performed in typologically varied basalt vessels, starting during the Late Chalcolithic and continuing through the Early Bronze Age. Considering this, these specific basalt vessels provide evidence for a direct techno-functional link between these two periods in the southern Levant, showing that concepts or traditions of vessel use were shared for a long period of time.