1 Introduction
When archaeologists began excavating large city sites of the Bronze Age (
c. 2000-220 BC) in China since the 1930s, bone-working workshop sites have been found in many of the major cities of this period (
Zhao 2023). From these ancient cities, various types of bone materials used in the bone-working production have been unearthed, sometimes in quantities of tons. However, for quite some time, these broken bones have not attracted sufficiently academic attention or been systematically analyzed. Bone-working in the early cities of this period is often regarded as an economic activity of limited technical and economic significance. The main reasons for this misconception come from two aspects. First, the primary products of the bone-working industry during this period were considered very cheap, especially compared to the precious bronze and jade artifacts. Thus, bone tools from Bronze Age site are no longer cherished as their counterparts from Paleolithic or Neolithic context. Secondly, many researchers seriously underestimate the actual demand of bone tools by the social members (especially commoners) in the Bronze Age society. Although bone tools and pottery were both low-value utilitarian daily goods for the common people, bone tools are generally perceived to be produced and consumed on a much smaller scale.
Since the 2010s, as the complex exploitation of animal resources in Chinese Bronze Age gradually has attracted more academic efforts, a growing number of zooarchaeologists also call for attention on the unique properties of large-scale bone industries in early urban context. Researchers notice that their production purposes, technological features as well as broader social interactions significantly differ from those of the preceding Paleolithic and Neolithic period. These new characteristics of bone-working activities not only can contribute to the understanding of how bone tools were produced, but also provide new perspectives for understanding the operation of urban handicraft economy in early China.
This study here presents the archaeological discovery and analysis of such a mass bone-working workshop from a major city site of the Late Bronze Age in central China. The archaeological context of this case is the Yuntang bone-working workshop, located within the site of Zhouyuan city in Shaanxi province, China (Fig. 1). Archaeological remains along with historical records have proved that Zhouyuan used to be one of three capitals of the Western Zhou dynasty (
Cao 2023;
Song 2017), which ruled the mid and lower Yellow River region during 1046-771 BC. It is one of the largest capital cities throughout the Chinese Bronze Age, covering about twenty-eight km
2 at its peak. A spectrum of craft activities has been confirmed within Zhouyuan (
Guo et al. 2021;
Lei 2014;
Ma 2009;
Xu 2002). The substantial archaeological remains demonstrate handicraft workshops for both elites and commoners at different loci of this city site. Their products include luxury goods such as ritual bronzes, bronze weapons, chariot accessories, jade, lacquerware and decorative roof tiles, as well as utilitarian goods like daily pottery, stone tools, shell tools, antler tools and bone tools.
Yuntang bone-working workshop was situated at the north-eastern corner of inner city of Zhouyuan in the early Western Zhou period. By the late Western Zhou period, with the expansion of the outer city, this location had become part of the central area of the city. This site has been subjected to two rounds of excavation. In 1976, the Zhouyuan Archaeological Team excavated an area of 350 square meters at the central section of this site (
Zhouyuan Archaeological Team 1980). They recovered more than ten tons of bone materials. They also conducted surface survey for this site, estimating the total area of the bone-working site is about 60,000 square meters (about 300×200 meters). The animal species of about 4000 kilograms bone materials from the largest pit 76H21 was preliminarily identified. Nevertheless, due to the political chaos at that time, no further in-depth research was carried out to explore the manufacture process and production organization. Much of the bone materials from this excavation was later backfilled.
In 2014, in order to promote the understanding of the city layout and handicraft economy of Zhouyuan city, the Zhouyuan Archaeological Team excavated a trench of twenty square meters (2×10 meters) about thirty meters north to the excavation location of 1976, recovering nearly two tons of well-preserved bone materials (Fig. 2). Meanwhile, to evaluate the preservation status of this site, the entire Yuntang locus was re-surveyed and it is confirmed that the area still with dense bone deposit at present is about 40,000 square meters (about 200×200 m). This study is mainly based on the materials from the excavation in 2014.
2 Archaeological context of Yuntang workshop
Dump deposit units are the primary archaeological features at Yuntang. Totally forty-five Western Zhou dump deposit units were identified within the 2014 excavation trench. According to the analysis of the stratigraphy and the chronology of pottery, the production at Yuntang spanned almost throughout the entire Western Zhou period, from about 1020 BC until the fall of Zhouyuan to Rong invaders in 771 BC. The artifacts recovered from these deposit units could be categorized into four types, according to their relationship with the bone-working industry, including bone materials, production tools, non-osseous raw materials for bone-working (such as turquoise for inlay), and domestic pottery shards.
Bone materials constitute the largest portion of all artifacts, either in weight or in number. The bones often densely pile up (Fig. 3). Within the excavation area of only twenty square meters (3-3.5 meters in depth), totally 1914.098 kilograms of bones was recovered. Most of the bone materials bear clear evidence of artificial modifications, especially saw marks (Fig. 4). Generally speaking, these saw marks are still sharp and smooth, and have not been significantly blunted by abrading or trampling.
According to their respective status in the bone-working production sequence, 1529.664 kilograms out of the 1914.098 kilograms of bone materials can be categorized into six major types, including complete raw material, semi-used raw material, blank, preform, finished product, offcut (Fig. 5). The production status of the other 402.134 kilograms of bones cannot be determined (but not affect the biological identification).
Specifically speaking, complete raw material weighs 28.290 kilograms. They are bones which generally retain their intact biological form without any significant artificial modifications.
Semi-used raw material weighs 269.331 kilograms, and they are the bones which are partly subjected to certain manufacturing process, especially the reduction treatment. Nevertheless, according to the production sequence, these materials have not been fully exploited. There are both considerable volumes of usable parts and unwanted parts in the bones.
Blank weighs 57.165 kilograms. They are defined as the beginning products from the reduction process. Usually, blanks are of sharp and flat surface produced by bronze saws, but have not been subjected to filing or grinding procedures. The majority of the identifiable blanks are fragile bone strips in different width.
Preform weighs 0.33 kilograms. Compared with blanks, preforms have been further processed towards the appearance of the final product, mainly through filing and grinding. They commonly lack the full final refinement such as polishing, engraving, inlay or other further decorative procedures.
Finished product weighs 1.487 kilograms. Finished products are the bone artifacts that have gone through the entire manufacturing process and are ready to use. Once a given bone product is highly polished, it usually means no more treatment would be conducted on the products.
Offcut weighs 1172.761 kilograms, contributing the largest portion (76 per cent) in all the bones available for determining their processing status in terms of weight. Offcuts are all the bone portions which were removed from the production system due to their lack of potential for providing usable blanks anymore. In the Yuntang assemblage, the offcuts mainly comprise the saw-off joint end, tuberosity, and other irregular-shaped parts of bones. In comparison, complete raw material, semi-used raw material, blanks, preforms and finished products all together only account for about 24 per cent in total weight. The percentage of finished products is even less than 0.1 per cent. Such composition pattern of bone materials complies with the nature of Yuntang locus as a workshop for manufacturing bone tools.
Through analyzing the characteristics of raw material, product as well as the production process, we can see that the bone-working industry at Yuntang exhibits the basic traits of a typical large-scale standardized production. Its technical and organizational features are reflected in three aspects. Firstly, the raw materials were simplified by concentrating on the use of animal bones from limited types of species and age. Secondly, within this bone-working workshop, the goal of production was simplified by focusing on manufacturing only two types of daily bone tools with massive and stable social demand. Thirdly, almost all the craftsmen strictly adhere to a set of uniformly regularized production templates (for each type of bone element) throughout the existence of Yuntang workshop for more than 200 years.
3 Raw material Pattern
3.1 Taxonomic composition
In terms of species, bone-working production at Yuntang almost exclusively depended on cattle bones. Bones of other species were only processed on occasion and in very small quantity. A total of 73,081 bone fragments (1914.098 kilograms) from the 2014 excavation were subjected to taxonomic identification. Of this assemblage 38,645 bones (1640.969 kilograms) are identified to genus. 34436 bones (273.129 kilograms) are either too highly processed or too fragmentary for identification. In total, thirteen taxa have been identified from the Yuntang assemblage (Table 1). Meanwhile, the NISP pattern are generally positively proportional to the pattern of MNI and weight.
Specifically, cattle bones dominate the entire faunal assemblage, reaching up to about 96 per cent in total NISP. The weight of identifiable cattle bones amounts to almost 1,600 kilograms. The MNI of cattle is also very high, numbering 826 individuals. The NISP of pig and sheep/goat remains at relatively the same order of magnitude, both around 1 per cent in total. Their MNI follows a similar pattern, staying around 1 per cent. The number for horse comes in the third level among the domestic animals, accounting for roughly 0.4 per cent in total NISP. Dog is the domestic animal with the smallest representation. Only twenty-one specimens were identified as dog, and the corresponding MNI is just four.
Compared with domestic animals, wild animals have an even smaller part in the assemblage. The wild species present in the Yuntang assemblage include deer, rabbit, bear, rhino, fish, bird, turtle and rodent, together accounting for about 1 per cent of total NISP. Meanwhile, from the perspective of diachronic comparison, the overall compositional pattern did not change significantly during operation of Yuntang bone-working production. The actual presence of large amounts of cattle bones implies that the bone-working industry might have a long-term and stable channel for raw material procurement, because the taxonomic pattern did not change significantly through a timespan of about 250 years.
Additionally, of the 19,417 species-identifiable bones with clearly artificial modifications caused by bone-working activities (marks of sawing, filling and grinding), cattle bones are 19,276, reaching up to nearly 99.3 per cent. Other animals like bear, rhino, horse and deer were only processed in very limited scale, and probably only occasionally. Although pig is in the second place in NISP and MNI, no pig bones carry clear manufacturing traces. They probably mainly came from the pork consumption by craftsmen at Yuntang workshop.
3.2 Body part representation
Regarding all cattle bones, craftsmen at Yuntang showed clear preference for the limb bones of cattle. The body part representation pattern of cattle in terms of %MNE is visualized in Fig. 6. The %MNE pattern for cattle bones reflects a highly unbalanced representation of bone element at Yuntang.
Specifically, calcaneus has the highest representation (left calcaneus NISP=826, right calcaneus NISP=763), thus the number of the left calcaneus is used as the denominator for calculating the relative body part representation of cattle. However, it should be clarified that although calcaneus, astragalus, carpal and tarsal also have high representation, they are rarely processed and should be classified as "offcut" in bone-working. Few manufacturing treatments have been observed on these elements. They were unsuitable for the bone-working production at Yuntang, either too small or too rugged. In fact, these bones were the "rider" of transportation. During the excavation, we found that many joints still apparently retained the status of articulation, like the ankle and wrist joints, though the ligament which used to connect them together have vanished (Fig. 7). Moreover, in most cases, the shafts of the limb bones have already been sawed off, leaving the joints only with the proximal or distal ends of the limbs, as well as the smaller carpals or tarsals in between. This phenomenon reflects Yuntang bone-working industry mainly used fresh bones as raw materials. In this case, the riders were conveyed together with the useful parts to the workshop. In addition, the time span, from the point when the fresh raw materials were dismembered from the animal carcasses, to the point when the bones were processed and corresponding offcuts were discarded into the deposit units, was relatively short. When the articulated joints entered into the dump deposit, the bone elements were still bundled together by the ligament tissues.
Aside from these riders, the %MNE of cattle limb bones, including humerus, radius, femur, tibia, metacarpal and metatarsal, is disproportionately higher than that of the axial bones. Their %MNE are all above 60 per cent. The %MNE of other elements are all lower than 5 per cent. Especially, the %MNE of axial bones is less than 1 per cent. In addition, over 90 per cent of these limb bones bear clear bone-working traces. Therefore, it can be seen that limb bones of cattle were the most important raw materials in bone-working production.
The polarized element pattern suggests that the procurement of cattle bones for bone-working production was a result of intentional selection. The living cattle were butchered at some other places within Zhouyuan city. Then fresh limb bones, along with riders, were immediately picked out and transported to Yuntang workshop, while butchery garbage like crania, mandibles and other axial elements were left at the killing location.
3.3 Survivorship of cattle
The aging of cattle from Yuntang is mainly based on metahphysis fusion, since the number of teeth and horn cores are too few to support the analysis. Fig. 8 shows the survivorship of cattle in the Yuntang assemblage. The ages of cattle are divided into five different stages, according to the fusion sequence of bone element. 100 per cent of cattle survived stage I (0-1 years old). For stage II (1-2 years old), only one unfused humerus distal and two phalanges I proximal were observed. Thus, 99 per cent of cattle survived stage II. Sizable slaughter seems to have started in stage III (2-3 years old). About 31 per cent of cattle were killed in stage III, while 68 per cent survived into stage IV. During stage IV (3-4 years old), 39 per cent of cattle were slaughtered. By this point, only 29 per cent of cattle were able to survive beyond four years old.
The survivorship of cattle at Yuntang largely coincides with a killing strategy mainly focusing on meat provision. Researchers have suggested that there was differentiation between the domestic animals consumed by nobles and commoners in the Western Zhou period (
Okamura 2004;
Lu 2015). Beef was mainly consumed by the noble class as a kind of luxury food, while pig and sheep were the primary meat resources for commoners. As the capital city, Zhouyuan was one of the most important habitations for the high-ranking nobles of the Western Zhou dynasty. In this case, nobles’ needs of beef for daily consumption, feasting, religious and diplomatic events may all have contributed to the demand for beef provision. Consequently, this also brought the bone-working industry a stable and adequate raw material supply at the same time. Probably one of the reasons why most of the large bone-working workshops in the late Bronze Age were exclusively found in capital city sites, was because only a noble population of considerable size could leave enough "trash" as raw bones for a massive bone tool production.
In general, for craft production, the simplification of raw material types is a necessary premise for practicing standardization at the starting point of the manufacturing sequence. Regarding the raw materials in bone-working, to a large extent, species equals shape, while age equals size of the raw material. The simplification of processed species types means that the possible variation of the shape of raw materials can be significantly restricted. The species of raw materials at Yuntang seem highly simplified, focusing on virtually only six types of cattle limb bones. Even the manufacture of the useful horns had been segregated to another locus in Zhouyuan.
Meanwhile, a concentrated age range of 2-4 years means that most of the bones have already reached their adult size, no matter whether their metaphyses were completely fused. Although individual differences still existed, the size of the same element in general would be more homogenized. The simplification of age range might not have been sought by the bone-working craftsmen on their own initiative, but probably was the result of butchering strategies for meat provision at Zhouyuan. However, a concomitant result is that the age range of the raw materials conveyed to the bone-working workshop was concentrated, thus de facto leading to the simplification of the size range of the raw bones.
The relative stability of the raw material types, in terms of species and age, provided the preliminary condition for the craftsmen to adopt a unified production procedure. Otherwise, they would have to consider and adjust the production sequence for each batch of raw materials with different shapes and sizes.
4 Product Pattern
Besides the simplification of raw material, the simplification of product types provided another prerequisite for regularizing production procedures at the final point of the production sequence. According to our observation on both finished products and blanks, it seems that bone-hairpins and bone needles were the dominant target products of the production at Yuntang.
Table 2 shows the statistics about the finished products at Yuntang. Among all the finished products, hairpin and needle constitute the top two categories in quantity. In total, 132 finished hairpins were found, accounting for 60.3 per cent (132 out of 219). The number of bone needles is forty-one, making up 18.7 per cent (41 out of 219) in total finished product. Other types of products account no more than 5 per cent. It should be noted that the discoveries of the finished bone products are fairly limited, and incomparable to the overall assemblage of the worked bones. This could be largely because of the nature of Yuntang as a workshop for production. Nevertheless, the relatively high frequencies of hairpins and needles still imply their high proportions in the actual product pattern.
Compared with the limited quantity of the finished products, the abundant bone blanks could be more informative for illustrating the product pattern. During the 2014 excavation, 57.165 kilograms of bone blanks (6152 pieces) were collected, 51.233 kilograms of them are strip blanks (5647 pieces), holding an overwhelming majority. This result largely coincides with our reconstruction of the production sequence through refitting analysis, in which we found producing strip blanks constitute the mainstream of most production templates.
The strip blanks can be further categorized into two groups, according to their smallest width at the midpoint (Sw-mid). The first type of strip blank has a Sw-mid value mainly around 7-11 millimeters. Their length of complete blanks varies from 110 millimeters to 210 millimeters (Fig. 9). Here we refer them to as thick strips. Totally 1369 pieces of thick strip were found. The second type of strip blank is much slimmer and extremely fragile. Their Sw-mid varies around 2-4 millimeters, with a length of 45-90 millimeters, so they are referred to as thin strips. Due to their tiny weight per piece, the total number of thin strip blanks are huge, counting for 4278 pieces. Fig. 10 shows part of thin bone strips from the deposit unit H4. The Sw-mid could be technically important for the bone-working craftsmen, because the manufacturing potential of a given strip blank was largely constrained by the Sw-mid. The thick strips should be used for manufacturing bone hairpins, while the thin strips could only be used for bone needles.
The reason why hairpin and needle dominated the production capacity at Yuntang and other bone-working sites in the Bronze Age China, could be closely related to the roles of these two bone tools in the particular social and cultural background of that time.
The majority of the bone hairpins found from Yuntang and other places in Zhouyuan city (
Luo 1998) are simple and plain in style and decoration, indicating them as products with low-value (Fig. 11). This kind of simple hairpin looks like long nails, having a pointed tip on one end, and a spreading cap on the other end, in order to prevent the pin slipping off the hair bun. It either has a cross-section of round or diamond shape, without any further surface decoration except polishing. In the collection of intact hairpins stored in the Zhouyuan Museum, 94 per cent (305 out of 325) of them are such simple hairpins, whereas hairpins with more complicated decoration only account for 6 per cent (20 out of 325).
According to archaeological discoveries and written texts, both females and males wore hairpins during the late Bronze Age and later historical period in China (
Sun 2014;
Wu 2015). Male usually wore only one or two hairpins at most, mainly to hold the hat on his hair bun. Female often used multiple hairpins at the same time, probably due to the sophistication of their hairstyles. Taking the tomb M81 of Western Zhou period in the Shaolingyuan cemetery for instance, this is a tomb for a woman of around twenty-five years old (Fig. 12). Twelve bone simple hairpins were found around her head, and it seems that they were dressed for certain hairstyle (
Shaanxi Archaeological Institute 2009). This is a typical commoner tomb of that time, with only narrow chamber and even no daily pottery offered. However, this occupant was still offered with twelve hairpins like, clearly showing the pervasiveness of using bone simple hairpin in commoner class. Meanwhile, simple bone hairpins are also widely distributed in the elite zones at Zhouyuan city. The simple hairpins were probably mainly used by commoners or low-ranking nobles, but this does not preclude their use by high-ranking class.
Besides the straightforward function of holding and decorating the hairstyle, there are also underlying social meanings about wearing hairpins in the Western Zhou society. Wearing hairpins was an important symbol for coming of age. The ceremony of coming of age for young men was called as
Guanli (hat ritual), and the ceremony for young women was called
Jili (hairpin ritual). By the age of fifteen, girls would change their hairstyles and could wear hairpins to signify that they have reached the marriageable age. Similarly, boys would start to wear hats (with the help of hairpins) at the age between twelve and twenty, also to indicate the coming of age for marriage. (
Dai 2006;
Peng 2004;
Sun 2014). In this case, though hairpin can hardly be considered as a kind of daily necessity for biological subsistence, it was one of the essential cultural necessities for an adult to carry out his or her normal social role in the Western Zhou society.
Bone needle is another important type of product at Yuntang (Fig. 13). Despite its trivial size and simple morphology, bone needle probably was also a type of tool with massive social demand in the Western Zhou society. The earliest bone needle currently known in China was discovered from the Late Paleolithic Upper Cave site in Zhoukoudian (stored in the National Museum of China), dated to around 34,000 BP (
Li et al. 2018). During the long period thereafter, bone needles were the most common stitching tools in different cultural and geographic backgrounds (
Huang 1993;
Shen 1994;
Xu 1994). This situation continued during the Bronze Age. Bone needles have been widely discovered in many Bronze Age sites, including the residential areas at Zhouyuan. In contrast to the frequent presence of bone needles, the discovery of bronze needles is very rare. Only one bronze needle has been discovered at Zhouyuan. From a broader perspective, the discovery of bronze needle is also limited throughout the Bronze Age in China (
Xu 1994). Scholars suggest that bone needles were still used as the primary stitching tool during the Bronze Age (
Shen 1994;
Xu 1994). As clothing is one of the basic subsistent goods, the social demand of bone needles, such a consumable item, could also be stable and massive.
In general, it is their nature as the inexpensive but indispensable daily necessities that provide the great and constant social demands on bone hairpin and needle. Additionally, the physical fragility of bone hairpin and needle also led to the situation that the demand perhaps could not be met by intermittent productions. The specific social customs of wearing hairpins in the Western Zhou society inevitably need to be fulfilled with substantial supply of product, and consequently would be transformed into the massive demand of the related craft products—bone hairpins. Furthermore, the bone-working production at Yuntang also demonstrates a simplified product pattern, providing another important technical prerequisite for designing a standardized production system. The production of other osseous products (e.g. horn) might have been spun off to other independent craft production locus, implying that the bone-working groups were highly specialized in certain products at Zhouyuan.
5 Characterizing the technology of Yuntang bone-working
The production procedures of bone-working at Yuntang can be grouped into five stages: 1) preparation: fleshing and degreasing; 2) reduction: extracting blanks through sawing; 3) refining: filing and grinding; 4) embellishing: polishing; 5) optional decoration: drilling, inlay, gilding, engraving and assembling. Here we will not present all the details of the operations in each stage, but will emphasize the organizational characteristics of the technical system in the Yuntang bone-working industry.
One of the distinctive features in the technical system at Yuntang was the introduction of massive standardized production. Through reconstructing the manufacturing sequence, two remarkable technical features can be generalized. One is the uniformity in reduction template designs, and another is the regularity in operational minutiae. Both of them reflected that the Yuntang bone-working industry was intentionally designed and controlled by particular specialists, and the production activities of the craftsmen were strictly regularized. Moreover, the long-term continuity of fixed technological traditions also implies that the particular knowledge about bone-working was inherited and persisted by generations of craftsmen at Yuntang.
5.1 Uniformed general template design
Among all these procedures above, reduction stage can be a representative step of standardized production. Reduction is the process to extract qualified blanks of different shape and size. It is the core stage in bone-working production at Yuntang, through which the originally biological, morphological features of bones are largely removed and transformed into artificially manufactured articles.
The suitable tools provide an important material condition for the craftsmen to be able to carry out the design of these templates. The bronze saw seems to be the primary tool for reduction. All the manufacturing traces left on the bones by reduction treatment are saw marks. No visible evidence of cutting, scraping or chopping has been observed in the Yuntang assemblage. Meanwhile, four pieces of bronze saws were found in 1976, while one was found in 2014. Such limited quantity may be related to the recircle of worn saws. These bronze saws are thin and narrow, with average width only around 7-12 millimeters and the thickness around 1-1.1 millimeters (Fig. 14). The bronze saws were probably installed on some supportive frame (wood or bamboo) to strengthen the straightness of the saw blade, rather than simply being held with a handle directly. Compared with stone knives or stone saws, bronze saws allow craftsmen to conduct more precise manufacturing actions, and the result of any movements also could be better predicted. The use of an axe or knife to process bones largely depends on instantaneous force. In comparison, sawing is a relatively low-speed and progressive process, thus the controllability of tools and actions was increased. This technical advantage allowed craftsmen to conduct more bold operations, rather than having to gradually grind a bone needle from a thicker bone piece, a process that may take longer.
Refitting analysis is employed as the primary method to reconstruct the reduction sequence (Fig. 15). The result shows that for each type of bone element, there were always one or two mainstream reduction templates. The majority of the bones were processed according to the mainstream scheme. Although some bone elements were also processed according to other template designs, these cases were secondary and only happened occasionally.
The high uniformity in reduction template design at Yuntang finally resulted in two phenomena in the worked bone assemblage. First, the uniformed reduction templates led to the high morphological repetitiveness of the bone debitage. As long as the bone debitage was from the same manufacturing step, it usually shared a similar form. Second, another direct and marked piece of evidence for the uniformity in template design is the fact that bone debitage from different units and periods can be put back into the same reduction sequence. Thus the progressive morphological change of the bone element through reduction can often be explicitly illustrated by using bone debitage from different units. For instance, a large number of candidates can often easily substitute for a given piece of bone debitage shown in the photo regarding to the reduction sequence, which will not affect the validity of the reconstructed reduction sequence.
Such high repetitiveness of bone debitage suggests that mainstream manufacturing schemes were used by the majority of craftsmen at Yuntang. It is precisely because of this high degree of repetitiveness that we are able to restore the reduction scheme of each bone element as comprehensively as possible through the refitting analysis, and finally reconstruct a relatively complete production flow chart at Yuntang workshop.
Moreover, from a diachronic perspective, the established mainstream templates persisted throughout the early to late Western Zhou period, a time span of more than 250 years. There was little significant modification in the reduction sequence of each bone element. This means multiple generations of craftsmen largely stuck to a fixed technical system. Meanwhile, the long-term continuity in reduction templates also implies that the knowledge about these template designs must have been intentionally passed down through different generations of craftsmen, and the later craftsmen always were able to learn and follow the technical tradition of their predecessors.
5.2 Regularized operational minutiae
Another important characteristic in Yuntang bone-working is the uniformity of operational minutiae by craftsmen. Operational minutiae refer to the details of behavior that do not affect the overall execution outcome of the production template, but do reflect the craftsman's action and gestures to achieve the requirements of the template. If craftsmen were only provided with a uniform general template, there may have been more than one way for them to accomplish the production task, as long as they could make the outcome similar to what the general template required.
Trivial manufacturing traces may better reveal to what degree the craftsmen's actions and gestures had been regularized, reflecting the uniformity of operational minutiae among the craftsmen at Yuntang. One typical example is the procedure that happened after the filing of strip blanks, which concerns how craftsmen removed the ‘hairpin butt’. For the strip blanks extracted from the shaft of limb bones, there were usually some spongy bones at the both ends of the blanks (Fig. 16). Spongy bones should not be present on the final products. However, the craftsmen did not saw off the spongy parts immediately after the extraction of blanks. Instead, the spongy parts were carried on to the filing stage. The craftsmen kept the spongy parts with the blanks until the cross-section of the blank was completely filed into the desired shape. During the filing stage, one end of the blank was processed into a pointed tip, while another end was filed into a spreading hairpin cap, like the situation shown in Fig. 16. Afterwards, the craftsmen sawed off the end with some spongy bones. The filed preform of the hairpin moved on to the grinding treatment, while the sawed-off spongy ends were discarded. These sawed-off spongy parts can be referred to as hairpin butts.
During the 2014 excavation, 573 hairpin butts were recovered through the sieving work. The cross-sections of these butts are either in rounded or diamond shapes, clearly suggesting that the products that were detached from them were cylindrical and diamond hairpins. One interesting phenomenon about these hairpin butts is that there was always a seam on each butt. The seam was created because the craftsmen without exception adopted a two-way sawing operation when sawing off the butts. Moreover, the craftsmen always sawed off the butts along the short axis of the diamond, and thus the seams were always parallel to the long axis of the diamond (Fig. 17). The width of these sawing seams ranges from 0.4-1 millimeters. Although the entire thickness along the short axis of the diamond is only 5-7 millimeters, the craftsmen never did this job using a one-way sawing. In fact, if the craftsmen wished, they could saw through the hairpin preforms with only about ten strokes. Instead, it seems that the craftsmen had great patience when conducting this operation. They only sawed into the hairpin preform for 2-3 millimeters from one side, and then switched the sawing direction. The presence of a central seam was observed on all hairpin butts, regardless of rounded or diamond-shape.
Such method was probably intended to guarantee the success of the sawing action. This is because the butt, if only sawed from one side, may break before the kerf passes completely through the hairpin preform. This may cause an irregular fracture on one side of the hairpin preform, leading to damage to the surface flatness and appearance of the finished products. On the contrary, when sawing from both sides, the final breaking point could be restrained within the kerf surface. In this way, the risk of affecting the quality of the hairpin could be reduced accordingly. The persistent presence of central seams on hairpin butts indicates that the craftsmen still need to follow certain uniform regulations even only for conducting such trivial action. To some extent, the production regulation not only defined what kind of result the craftsmen should achieve in particular steps, but also specified how they should perform in that step.
6 Scale of Yuntang bone-working
Due to the adoption of standardized production methods, the production efficiency of Yuntang workshop can be significantly improved. The promotion of production efficiency also would lead to the significantly larger product output. In order to further estimate the overall output at Yuntang, firstly it is necessary to extrapolate the possible overall number of cattle processed at Yuntang. The excavation area of 2014 is twenty square meters, and the MNI of cattle is 826. For the excavation in 1976, only one deposit unit pit 76H21 was subject to quantitative statistics. 76H21 is about thirteen square meters and its MNI of cattle from is 1306. Thus, the MNI of 2132 cows is confirmed within an area of about thirty-five square meters, and the density of cows per square meter would be sixty-one. Based on the systematic drilling in 2013-2014, the area with dense bone deposits at Yuntang is about 40,000 square meters. Even if we presume that the actual density of bone debitage beyond the excavated area is only one out of ten of that within the excavated area, which would be six cows per square meter, then the total potential MNI of cattle within the 40,000 square meters would be 240,000.
Based on the reconstruction of the production sequence, the number of possible products from each bone element can be estimated. The craftsmen could extract 6-8 hairpin blanks from each limb bone (including humerus, radius, femur, tibia, metatarsal and metacarpal). In this case, for one cow, the potential output of hairpins would be 72-96. In this case, the total possible output of hairpins would be 240,000 × 72-96 = 17,280,000-23,040,000. Divided by the time span of the Western Zhou period (c. 250 years), the annual output would be 69,120-92,160 hairpins (MNI = 960 cows).
The calculation here certainly does not intend to provide a series of precise figures for the output of bone-working production, and all these estimations beyond the excavated areas are only a very preliminary conjecture about the magnitude of annual output. Even considering the technical failures during the manufacturing operation, nevertheless, the bone-working production at Yuntang can be regarded as operating on a massive workshop level. If the consumption of hairpins by each commoner per year was about ten, the output of the Yuntang bone-working industry might be able to fulfill the demands of about 7000-10,000 people. The annual output seems far beyond the needs of craftsmen themselves. the scale of bone-working production was large enough to yield sufficient surplus, and probably the production was a type of commodity-orientated economy.
7 Conlusion
Although the bone-working industry has long been considered as a trivial economic phenomenon, this study shows that the bone-working industry can be considered as a kind of craft production running on the massive workshop level in early cities. Its production scale suggests that the primary economic role of this workshop was to produce large quantity of basic utilitarian goods for both commoners and elites, rather than only for few aristocratic groups. Reversely, the consistent demands of particular bone products may stimulate the application of a set of intentionally standardized production procedures. This is probably because standardization could lower the skill threshold in recruiting more craftsmen into the production system which were designed by few technical authorities. Along with the increase in the craftsmen size, the production efficiency and yield of products would also be increased. Previous research on the craft production in Zhouyuan city mainly emphasize the scale and importance of workshops for ceramic, bronze, jade, stone tools and etc. However, through examining the characteristics of bone tool production at Yuntang, it can be concluded that the massive bone-working industry was probably also occupied an important position in the economic fabrics of the Zhouyuan city. Meanwhile, the result of this study also clearly demonstrates the potential of bone-working remains for getting deep into many rarely touched but fundamental mechanisms underlying the operation of economy in the historical period.
The Author(s) 2026. This article is available under open access at journal.hep.com.cn.
PDF
(20119KB)
