Introduction

The Yangtze River Delta region in Southern China is one of the areas where rice cultivation originated1,2,3. In the early Holocene, rice cultivation began in this area (Shangshan culture, 11000–8600 BP), followed by the Hemudu culture (7000–5500 BP), the Majiabang culture (7000–5800 BP), and the Songze culture (6000–5300 BP). In the Late Neolithic period, the Liangzhu culture (5300–4500 BP), with rice cultivation as its economic base, emerged in the present-day Liangzhu area in Hangzhou city, Zhejiang province4,5,6,7,8,9,10,11. The Archaeological Ruins of Liangzhu City, formed during this period, had several features unlike the previous Yangtze River Delta cultures: urban sites surrounded by large enclosures and moats, water use systems with dams and canals, jade artifacts symbolizing religious beliefs, ritual facilities represented by altars, and cemeteries indicating social stratification. These indicate the establishment of a complicated social system; therefore, the Liangzhu culture is considered to have formed one of the earliest cities and states in East Asia4,7,10,11,12,13,14. In recognition of their historical significance, the Archaeological Ruins of Liangzhu City were added to the 2019 World Heritage List UNESCO15.

Within Liangzhu, the core area of the Liangzhu culture, several cemeteries were found at sites such as Fanshan, Yaoshan, Jiangjiashan, and Bianjiashan, and human skeletons were excavated from some of them4,7,10,16. Separate from these buried human remains, many human bones were unearthed in a scattered state from deposits in canal and moat features at the Zhongjiagang, Bianjiashan, Putaofan, Meirendi, and Huoxitang sites in Liangzhu17,18 (Fig. 1). These bones were found jumbled together with pottery fragments and animal bones, suggesting that they had not been formally buried, but rather discarded into the canals and moats. Interestingly, many of them show evidence of working, such as splitting, grinding, perforating, and polishing.

Fig. 1
Fig. 1
Full size image

The Archaeological Ruins of Liangzhu city and canal and moat sites where worked human bones were unearthed.

Worked prehistoric human bones have been found worldwide, dating back to the Pleistocene, such as the Mousterian skull fragments used as tools19 and the Magdalenian skull cups20. In the Holocene, in West Asia, there is much archaeological evidence of worked human bones, such as the modified skulls from the Pre-Pottery Neolithic period in Turkey (12000–8500 BP), which suggest skull cult21,22the Bronze Age human bone tools from Iran23and the Neolithic skull cups from Spain (6000–3000 BP)24. In North America, human bones were worked for trophies and ancestor worship in the Middle Woodland period (2200–1500 BP) in the Midwest25and Southern Texas and Southern Ontario archaeological sites yielded human bone artifacts26,27. In Teotihuacan in Mesoamerica (1900–600 BP), many human bones were found as pins, buttons, rings, and bracelets28. Numerous ethnographic evidence of working the bones of kin and friends into tools and ornaments are found throughout Oceania as well29,30. In the Japanese Archipelago in the Far East, human bone ornaments were excavated from several sites of the Early to Final Jomon period (7000–2500 BP)31,32.

However, in China, although worked skulls (ca. 3000 BP) were excavated from a Shang Dynasty site in the early historic age33no worked human bones have been found at Neolithic sites. In the Yangtze River Delta, where the Liangzhu archaeological sites are located, many human remains have been found at Neolithic sites, such as Hemudu (7000–5500 BP), Tianluoshan (7000–5500 BP), Majiabang (7000–6000 BP), and Guangfulin (5900–4400 BP). All consist of buried human remains without any work traces34,35,36,37,38,39,40. It is likely that worked human bones first appeared in Chinese prehistory during the Liangzhu culture. This raises the question, why were human bones worked on even though there was no cultural tradition of working with human bones? and why was Liangzhu the place for that?

Cases of human bone working worldwide often reveal social relationships, such as kinship or conflict between the worked and working parties, serving as motivations for the workings. However, our results indicate that human bone workings at Liangzhu were characterized by the use of human remains, detached from such social relationships. In this study, we conducted anthropological and osteoarchaeological research into worked human bones excavated from Liangzhu to illuminate the reasons and background for their appearance in the earliest state society of East Asia.

Results

Taphonomy and type classification of worked human bones

Of the 183 human bones surveyed, 52 (28.4%) were worked bones, and 131 (71.6%) were non-worked bones. Regarding the worked human bones, traces associated with manufacturing—such as splitting, grinding, perforating, and polishing—were identified (see Methods). There were no signs of violence, no cut marks on the bone surface that could be interpreted as disarticulation marks, and no evidence of burning. In addition, no bone healing was observed around the worked traces. There were also no gnawing or bite marks from animals such as rodents or carnivores41,42.

The worked human bones were classified into types A-F, based on shape and bone type – (A) skull cup, (B) mask-like facial skull, (C) small plate-shaped skull fragment, (D) skull with posterior perforations, (E) mandible with flattened mandibular base, and (F) long limb bone with worked traces at the end of the diaphysis (F1: narrowly flattened end, F2: transverse cut end, and F3: cracked end). The number of worked human bones by type and the number of non-worked human bones are summarized in Table 1.

Table 1 Number of human bones.
Full size table

  • Type A: Skull cup (n = 4).

These bowl-shaped calvariums, cut or split horizontally at the glabella level, referred to as “skull cups”21,22were recovered from Liangzhu. Based on size and cranial suture closure, all of them belonged to adults. One finished artifact had polished and flattened bowl rims43however, three were unfinished, with incomplete rim adjustments. Delamination marks caused by splitting and traces of grinding during shaping were observed on the edges of the unfinished products (Fig. 2). Moreover, skull cups, as grave goods, were excavated from the tombs of high-class burials of the Liangzhu culture at the Fuquanshan site in Shanghai44 and the Jiangzhuang site in Jiangsu Province11 in the Yangtze River Delta.

Fig. 2
Fig. 2
Full size image

Skull cup (Type A). Scale bar is 5 cm.

  • Type B: Mask-like facial skull (n = 4).

These were split along the coronal sutures. Three were adults and one was an adolescent. The edges were roughly split and there were no traces of elaborate shaping attempts, hence, they appeared unfinished (Fig. 3).

Fig. 3
Fig. 3
Full size image

Mask-like facial skull (Type B). Scale bar is 5 cm.

  • Type C: Small plate-shaped skull fragment (n = 21).

Small plate-shaped frontal, parietal, temporal, and occipital bones were excavated. Their edges were crudely shaped and unpolished, with only splits and cuts. They appeared to be unfinished or failed products (Fig. 4).

Fig. 4
Fig. 4
Full size image

Small plate-shaped skull fragments (Type C). Scale bar is 5 cm.

  • Type D: Skull with posterior perforations (n = 1).

Only one child’s skull was found with two posterior perforations of the parietal bones (Fig. 5). The skull was almost complete, except for the missing right temporal bone, however, a mandible did not accompany it. The child’s age was estimated to be 6–8 years as the roots of the upper first molars were at the just before completion stage and the upper lateral incisors were in the beginning stages of eruption. The margins of the two perforations were polished and smooth (Fig. 5: upper right). Several linear marks, possibly abrasion marks, were observed on the cranial crown surface (Fig. 5: lower right). It is unlikely that these were surgical perforations, as the margins of the perforations had been worked and there were no healing changes around the perforations. Since the perforation processing was finished, it was a completed product. However, no similar examples have been found, hence, the intent and purpose of use are unknown.

Fig. 5
Fig. 5
Full size image

Skull with posterior perforations (Type D). Scale bar is 5 cm.

  • Type E: Mandible with flattened mandibular base (n = 3).

Three mandibles were found, with the base of the left and right mandibular bodies worked flat. There are no similar cases, therefore, it is difficult to determine their purpose. A magnifying glass examination of the flattened surfaces revealed several parallel linear marks caused by polishing (Fig. 6). There were no work traces, except on the mandibular bases. Two were adults and one was an adolescent. One adult had large areas of the mandibular bases worked on both sides, and the shaping appeared to have been completed. The other two had only small areas worked on; likely were unfinished or failed products.

Fig. 6
Fig. 6
Full size image

Mandible with flattened mandibular base (Type E). Scale bar is 5 cm.

  • Type F: Long limb bone with worked traces at the end of the diaphysis (n = 12).

Three humeri, one radius, one ulna, six femurs, and one tibia were excavated, with worked diaphyseal ends. These were classified as follows: narrowly flattened end (type F1, n = 3), transected end (type F2, n = 4), fractured end (type F3, n = 4), and with features of both F2 and F3 (type F2/F3, n = 1). The morphological characteristics of F1 suggest a utilitarian function, such as scraping or engraving (Fig. 7). The presence of flaking and linear scars, observed under magnification on their flat ends, does not contradict the interpretation of these objects as utilitarian tools. F2 and F3 were considered unfinished or failed products, as the working stopped when the epiphysis was removed.

Fig. 7
Fig. 7
Full size image

Humerus with narrowly flattened end (Type F1). Scale bar is 5 cm.

  • Others: Worked human bones other than types A to F (n = 7).

Four skulls and three femur fragments with worked traces were found. One skull was the lower half without the cranial crown and mandible, and two were the back half of the skull without the facial area. These were the remains of types A and B, respectively. Another skull was the posterior of a child’s skull, removed around the greater occipital foramen, and the worked traces were partially found on the edges. Two femur fragments were proximal, and one was distal, all of which showed cut scars on the diaphyseal side, yet no other worked traces, suggesting that they are remnants from type F.

Sites

Most of the worked human bones were recovered from the Zhongjiagang site (n = 42, 80.8%), followed by the Bianjiashan site (n = 4, 7.7%), the Putaofan site (n = 3, 5.8%), and the Meirendi site (n = 3, 5.8%). The largest number of non-worked human bones was also excavated from the Zhongjiagang site (n = 89, 67.9%), followed by the Bianjiashan site (n = 26, 19.8%), the Putaofan site (n = 13, 9.9%), the Huoxitang site (n = 2, 1.5%), and the Meirendi site (n = 1, 0.8%).

The proportion of worked human bones to the total number of human bones recovered at each site was as follows: 32.1% (42/131) at the Zhongjiagang, 13.3% (4/30) at the Bianjiashan, 18.8% (3/16) at the Putaofan, 75% (3/4) at the Meirendi, and 0.0% (0/2) at the Huoxitang. Excluding Meirendi and Huoxitang, where the total number of human bones was fewer than 10, a comparison among Zhongjiagang, Bianjiashan, and Putaofan shows that the proportion of worked human bones was highest at Zhongjiagang. Worked and non-worked human bone composition ratios were compared across the three groups using Fisher’s exact test with Monte Carlo simulation (100,000 replicates), due to small expected frequencies in some cells. The test revealed a significant difference among the groups (p = 0.016).

Outside the core of Liangzhu area, the only other reported Liangzhu culture sites where worked human bones have been excavated are the Fuquanshan site in Shanghai and the Jiangzhuang site in Jiangsu province, where skull cups from tombs of the later Liangzhu culture have been founded.

Age, sex, and skeletal part compositions

The human bones included individuals of various ages – children, adolescents, young adults, middle adults, and old adults; however, no infants were present. Distinguishing between young, middle, and old adults was difficult for many adult bones, therefore, we grouped them all together as adults. The number of worked human bones, classified by age, was 46 adults (88.5%), four adolescents (7.7%), and two children (3.8%). The non-worked human bones consisted of 110 adults (84.0%), 16 adolescents (12.2%), and five children (3.8%). Fisher’s exact test revealed no significant difference at the 0.05 level between the age distribution of worked and non-worked human bones.

The number of worked human bones, by sex, was eight females (15.4%), four males (7.7%), and 40 undetermined sex (76.9%). The number of non-worked human bones consisted of six females (4.6%), 14 males (10.7%), and 111 undetermined sex (84.7%). Thus, although there was a higher proportion of female bones in the worked human bones, Fisher’s exact test did not detect a significant difference at the 0.05 level.

The number of worked human bones, by skeletal section, was 37 skull bones (71.2%), five upper limb bones (9.6%), and 10 lower limb bones (19.2%). The number of non-worked human bones consisted of 55 skull bones (42.0%), three vertebrae (2.3%), 29 upper limb bones (22.1%), and 44 lower limb bones (33.6%). Thus, the percentage of skull bones in the worked human bones was significantly higher than in the non-worked human bones (p = 0.0040).

Paleopathological findings

There were six worked skulls and one non-worked skull with orbital roofs, of which two worked skulls had cribra orbitalia (CO). Lower canines were present in two worked mandibles and two non-worked ones, and multiple linear enamel hypoplasias (LEH) appeared in all of them, suggesting that these individuals were repeatedly exposed to poor nutritional conditions during early life. Periosteal reaction (PR) was found in two of the 15 worked long limb bones and 14 of the 55 non-worked long limb bones. Fisher’s exact test revealed no significant difference at the 0.05 level in the frequency of CO, LEH, or PR occurrences between the worked and non-worked human bones.

Radiocarbon dates of the human bones

The calibrated ages of 14 C dating for the worked human bones ranged from 5040 to 4880 cal BP to 4810–4580 cal BP, with a concentration between ca. 4800–4600 cal BP (Fig. 8). Using OxCal’s interval command, the event duration was estimated to be between 200 and 391 years (1σ). The date ranges for the worked and non-worked human bones overlapped (Fig. 8).

Fig. 8
Fig. 8
Full size image

Calibrated radiocarbon ages of human bones.

Supplementary Information

The specimen ID, skeletal name and position, sex and age, worked/non-worked, type of worked bones, finished/unfinished, presence of CO, LEH, and PR, excavation site, dating data, and the number of each human bone are listed in Table S1. The number of human bones by site is compiled in Table S2. The number of human bones by age and sex is shown in Table S3. The occurrence of CO, LEH in the lower canines and PR in the worked and non-worked human bones is summarized in Table S4.

Discussion

Characteristics of worked human bones

The Zhongjiagang site, where particularly large numbers of worked human bones were unearthed, was Liangzhu’s workshop area18hence, it was inferred that workshops for working human bones existed at Zhongjiagang. Numerous animal bone products were unearthed from Liangzhu45; however, these were elaborately finished arrowheads and ornaments that did not resemble the worked human bones, suggesting that the intent for working the bones and their actual use were different from animal and human bones. While clarifying the similarities and differences between the two is crucial for understanding the nature of worked human bones, research on animal bone artifacts is still in progress and remains incomplete. Further investigation is therefore required.

Most worked human bones were produced between 4800 and 4600 cal BP (Fig8), corresponding to the middle to late phases of the Liangzhu cultural period8,46. This period also coincides with the decline of hydraulic infrastructure and a reduction in precipitation at Liangzhu47,48,49. These incidents may need to be taken into consideration when examining the possibility that the working of human bones had social or ritual aspects. However, given the considerable variation in the preservation of archaeological bones across different areas at Liangzhu14,18 the possibility that worked human bones from other periods once existed in regions unfavorable to bone preservation cannot be excluded.

There were no signs of violence on the worked bones, suggesting that the raw materials were not obtained through acts of killing. No cut marks were found at ligament or tendon attachment sites, which would indicate intentional dismemberment using sharp tools. This pattern suggests that the bones may have been collected after the natural decomposition of soft tissues. However, the specific means by which the bones were obtained remain undetermined.

Since there was no statistically significant difference in age and sex composition between the worked and non-worked human bones, it is unlikely that either child or adult bones or female or male bones were intentionally selected when obtaining material for working. On the one hand, there was a marked difference in the composition of human bone parts between the worked and non-worked human bones, suggesting that head bones were primarily selected. This is evident from the various types of worked skulls (types A–E), whereas the worked limb bones were less diverse (type F).

Examples of worked skulls exist worldwide; whereby, some are believed to have religious or ritualistic functions, such as the skull cult22. In China as well, sacrificial pits containing multiple human skulls have been discovered at the Shimao site—an urban settlement located in the Shaanxi region and dated to 4250–3750 cal BP50,51. Orschiedt stated that to be a skull cult, a religious context and the repeated performance of certain acts must be provable21. In Liangzhu, several worked skulls were found, hence, the act was considered repetitive. No worked skulls were unearthed from religious sites, such as altars and cemeteries; however, skull cups were excavated as grave goods from tombs at the Fuquanshan site in Shanghai and the Jiangzhuang site in Jiangsu from the Liangzhu culture period. These findings suggest that the skull cups from the Liangzhu culture period had religious, ritualistic, or prestigious purposes.

Nonetheless, these tombs did not contain any worked human bones other than skull cups. In addition, no features could be associated with religious goods, such as the jade objects unearthed from Liangzhu10. In short, there was no obvious evidence linking worked human bones to religious beliefs or rituals, except the skull cups such as type A. Types C (small plate-shaped skull fragment) and F1 (limb bone with narrowly flattened diaphyseal end) suggest that they may have been made for practical purposes. However, the shapes of types B (mask-like facial skull), D (skull with posterior perforations), and E (mandible with flattened mandibular base) did not allow us to envision specific use; there are no analogs made of other materials, such as animal bones or stones, hence, it was difficult to estimate the intent behind their production. Some of the worked bones, which were not practical or ritualistic objects, may have been made with demonstrative or artistic motives.

The presence of multiple worked human bones of each type indicated that the working of human bones followed standardized procedures. Nevertheless, many of them were not finished and appeared to be incomplete or failed products. The proportion of unfinished products (n = 42, Table 1) to all worked human bones (n = 52) excavated at Liangzhu was 80.8%. It is significantly more than the percentage of unfinished human bone products from Teotihuacan in Mesoamerica (15.6%)27. This high proportion prompts several questions regarding the reasons for abandonment. Were these items left unfinished due to technical failure? Did their incompleteness carry ritual or symbolic meaning? Or was production interrupted due to shifting social priorities? Although this study cannot provide definitive answers to these questions, they remain important avenues for future research to explore in order to deepen our understanding of the nature and significance of worked human bones at Liangzhu.

Hence, the characteristics of the worked human bones excavated from Liangzhu can be summarized as follows:

  1. 1.

    Many human bones were worked on at the Zhongjiagang site, where the workshop was located.

  2. 2.

    The production dates ranged from about 5000–4550 cal BP, with a concentration between 4800 − 4600 cal BP.

  3. 3.

    The age and sex of the bones did not impact material selection.

  4. 4.

    Skulls were selectively worked on.

  5. 5.

    The skull cups may have been used for religious or ritualistic purposes, however, there were many items whose functions were unclear.

  6. 6.

    There was a standardization of working.

  7. 7.

    There were many unfinished products.

Urbanized ancient Liangzhu society and human bone working

Given the number of worked human bones, the standardization in their manufacture, and the duration of production extending over two centuries, it is reasonable to interpret human bone working at Liangzhu as a sustained cultural or customary practice, rather than a temporal or incidental activity. Ethnographic and archaeological cases from other regions frequently associate the working of human bones with ancestor veneration, wherein the remains of ancestors or close kin are used as raw materials imbued with symbolic significance22,25,29,30. In contrast, the presence of numerous unfinished products at Liangzhu suggests the possibility that human bones were neither scarce nor particularly valued as materials. While it remains important to refrain from dismissing the possibility that specific emotional or symbolic associations—such as kinship ties—may have informed the selection or treatment of some remains, the evidence warrants serious consideration of alternative motivations distinct from ancestor worship in the practice of human bone working at this site. Since there was no evidence of premortal violence, the materials did not come from people killed in wars or rituals (this differs from the evidence of ritualistic violence found on human bones from the time of the emergence of social stratification in Andean civilizations52.

At the Fanshan and Yaoshan cemeteries in Liangzhu, the tombs of high-class elites had opulent grave goods10,14,16. Renfrew stated that the presence of high-ranking persons’ burial places and rich sumptuary goods was one of the features of a “city”53. The large amount of worked human bones and scattered human remains from the canal and moat deposits can be placed at the opposite end of the burial spectrum of high-ranking people. The presence of paleopathological bone changes, which indicated low health and poor nutritional status in the worked and non-worked human bones, suggested that their origins differed from those of high-ranking people.

Worked human bones first appeared in the Yangtze River basin during the Liangzhu culture period. As many of these products were discarded unfinished in the canals and moats, it suggested a transformation in the sociocultural perceptions about the dead during this period. Before the Liangzhu culture, human bones in the Yangtze River basin settlement sites were generally found only in burial contexts. This reflected the social structure of small-scale communities, where members were interconnected through kinship or other social relationships within the community, ensuring that all the deceased individuals were subjects of commemoration. However, Liangzhu society was complex, with a large population10,111214,. Individuals could not know all the members of society, which resulted in weaker social ties. Diluted human relationships in modern cities are a classic theme in sociology, represented by the “community lost perspective”54,55; such relationships between persons might have existed in the early urban society of East Asia, although the population and social system were different from those of modern cities.

The treatment of human remains reflects cultural perceptions of death and the deceased. The presence of many unfinished worked human bones recovered from the canal and moat deposits may symbolize a society that lived alongside socially distant “others.” In ancient Liangzhu, the coexistence of two distinct concepts of corpse treatment—namely, burial and use as raw material for production—was perhaps made possible by the rise of an urban society, where the dead increasingly included unrelated strangers who were not regarded as subjects of commemoration. This distinction may reflect new social realities of urban life in which the anonymous dead became more available and more utilitarian in character.

This study proposed a hypothesis explaining the emergence of worked human bones in ancient Liangzhu, focusing on urbanization, population growth, and the concept of the “other.” However, who, specifically, were these “others”? Were they “strangers” originating from outside the community or lower-status individuals who, due to social stratification, were not afforded burial? This study was unable to determine their identities. While isotopic or DNA analyses could elucidate geographical or kinship origins, these were beyond the scope of the present research. Future biomolecular studies are encouraged to verify the hypothesized ‘non-local’ or ‘low-status’ source of these worked human bones.

Methods

Materials

This study is based on all cranial and limb bone elements identified as human, which were recovered in a scattered state from Zhongjiagang (canal features), Bianjiashan (possibly moat-related features), Putaofan (moat features), Meirendi (canal features), and Huoxitang (moat features) at Liangzhu. A total of 183 specimens—including worked bones—were identified as human based on gross morphology, assigned individual specimen IDs, and are currently curated at the Zhejiang Provincial Institute of Cultural Relics and Archaeology.

Investigation of worked traces and classification of worked human bones

We investigated the human bones for the presence/absence of worked traces, such as split marks (including conchoidal fractures), grind marks, perforations, and polish marks (e.g., parallel striations from abrasion), referring to the descriptions of artificial damages to bones by Lyman56 LeMoine57 White et al.58 and Hargrave et al.25. The worked traces were examined by gross observation and microscopic observation with a 20x magnifying lens and a Dino-Lite digital microscope (Opto Science). In addition, human bones with worked traces were typified and classified based on shape and bone type. Furthermore, we looked for evidence of violence, such as stab and sword wounds59,60 cut marks at ligament and tendon attachments indicative of dismemberment41,56,61 traces such as micro-polish indicative of tool use42,56 evidence of burning62,63 and bone healing changes around the worked traces.

Age and sex determination

The age of the Individuals was estimated from the stage of tooth formation64 and eruption65closure of cranial suture53metamorphosis of the pubic symphyseal surface66and fusion of the epiphyses of limb bones67,68,69. They were classified into the following growth stages, based on Buikstra and Ubelaker70: infant (0–3 years), child (3–12 years), adolescent (12–20 years), young adult (20–35 years), middle adult (35–50 years), and old adult (50 + years). Cases that were difficult to classify were classified as adults.

The sex was determined from the shape of the adult coxae and skulls, using the sex determination method described in White et al.53. In cases where this was not possible, sex was estimated based on the size of the bones or the development of muscle attachments.

Paleopathological investigation

We investigated the occurrence of CO, LEH, and PR on the diaphyses of the long limb bones and other abnormal bone changes. Although we must be cautious in associating the paleopathological traces with health71 trends in their appearance frequency reflect some degree of living conditions72.

CO is a porotic bony change of the orbital roof caused by iron deficiency, anemia, or vitamin B12 or C12 deficiency during growth73,74,75. We examined the presence/absence of CO in the seven skulls with orbital roofs preserved, referring to Nathan and Haas76 and Hirata77.

LEH is an enamel defect caused by non-specific systemic stress due to malnutrition or disease during infancy78,79. In previous studies of archaeological remains, the upper and lower central incisors and canines were the subjects of LEH investigation (Goodman et al.80. Corruccini et al.81; Cucina and İşcan82; however, since incisors and upper canines were not present in the materials, we examined LEH only on the lower canines.

PR is an abnormal bony change on the surface of long limb bones caused by infection, poor health, or mechanical overload83,84. We examined the 70 long limb bones for PR, referring to Ortner83.

Radiocarbon dating

Radiocarbon dating was conducted on 9 worked and 43 non-worked human bones. Approximately 0.2–0.5 g of cortical bone was collected from each specimen using a diamond cutter, and collagen was extracted by through with alkali washing and gelatinization for purification85,86. Extracted collagen was introduced into Elementar’s vario ISOTOPE select elemental analyzer. Purified carbon dioxide, after combustion, was introduced into the vacuum glass line and heated along with hydrogen gas and iron catalyst to create graphite87. The 14 C ages of the graphite were measured using an accelerator mass spectrometer (AMS) owned by the University Museum of the University of Tokyo. Calibrated 14 C ages were calculated in OxCal 4.488 with IntCal2089. on the size of the bones or the development of muscle attachments.

Ethics statement

This study analyzed human skeletal remains recovered as archaeological heritage from the Liangzhu site in China, dating to approximately 4500 to 5000 years ago. As research involving ancient human remains does not include living individuals or any personally identifiable information, it is exempt from ethics review under the policies of the authors’ affiliated institutions.

We obtained permission to study these human skeletal materials, including the worked human bones, from the Zhejiang Provincial Institute of Cultural Relics and Archaeology (Zhejiang, China), which is responsible for the excavation and curation of the remains from the Liangzhu site.