Introduction
The Museum of Far Eastern Antiquities (MFEA) in Stockholm holds a unique collection of archaeological finds from Neolithic and Bronze Age (c. 3000–600 BCE) sites in north/north-west China assembled by the Swedish geologist Johan Gunnar Andersson and his Chinese colleagues in the early twentieth century (Figure 1). Andersson worked in China helping establish the Chinese Geological Survey but soon became fascinated with archaeology. This led to the first collaborative archaeological project in China, resulting in the discovery of more than 50 prehistoric sites. These excavations are celebrated as the beginning of Chinese archaeology, and some of the finds were exported to Sweden with the agreement of the Chinese government to build the founding stock of the MFEA. Nevertheless, little is known about the full range of sites discovered during this fieldwork, and the Chinese members of Andersson’s team are rarely mentioned.
Map of sites discovered by Andersson’s team (Sommarström Reference Sommarström1956).
Figure 1 Long description
The map displays prehistoric sites in Northern China, highlighting regions such as Shensi, Hunan, and Shantung. Key locations include the Yellow River, Wei Ho, and various provinces. The inset map focuses on the Tao Ho Valley, detailing specific sites like Yuan Tsui, Pan Shan area, and Pai Tzu Ping. The map uses labels and markers to indicate significant archaeological locations.
In recent decades, the prehistoric groups of north-west China and their role mediating the transfer of goods and technologies between Central Asia and China have received increasing attention. Yet, none of this research has made use of the extensive MFEA collections, some of which are from sites that are no longer accessible. The present project is ‘re-excavating’ these materials from storage to make them accessible and answer questions about life along the proto-Silk Road, while also retracing the history of the 1920s fieldwork. The project aims to: 1) establish an inventory of the Andersson collections in a digital database; 2) reconnect artefacts with archival materials to establish original find locations and retrieval methods; 3) identify Chinese team members and their contributions; 4) map proto-Silk Road interactions and identities using ceramic petrography, use-alteration and residue analysis; and 5) investigate ancient foodways, changes in animal use and human impacts on local fauna using zooarchaeology.
Database and cataloguing
Beginning in 2017, the project team first resolved labelling and storage issues, moving collections from wooden containers to acid-free boxes. We then developed a database that allows for recording and linking: object finds, ceramic data, zooarchaeological data, archival materials and publications. The current interface is in English with entries including terms in Chinese and Swedish; the next step is to make it fully trilingual. All content is shared with the museum team who transfer non-sensitive data to their publicly accessible bilingual online catalogue.
By May 2025, we completed a general inventory of 680 boxes of material (ceramic sherds, stone tools, animal bones, small finds) from 67 find locations, 123 complete ceramic vessels and 87 boxes of archival records with thousands of pages. We have fully reconstructed excavation information for three sites, allowing us to pinpoint artefact find locations and archaeological contexts, for example at Machangyan (Figure 2). We are expanding this approach to other sites to recontextualise the entire collection.
Sherds, bones and soil samples reassociated with archival documentation from the site of Machangyan (figure by authors).
Figure 2 Long description
The illustration features a collection of archaeological artifacts and records from prehistoric sites of Northern China. It includes boxes of ceramic sherds, stone tools, animal bones, and small finds from 67 find locations. Additionally, there are 123 complete ceramic vessels and 87 boxes of archival records with thousands of pages. The map in the center shows the approximate positions of finding places at Machangyan, with specific locations marked. The image also includes detailed views of the artifacts and records, highlighting the reconstruction of excavation information for three sites.
Ceramics research
Ceramic technology
We have conducted macroscopic analysis on 651 ceramic sherds from 15 sites, dating them based on style and assessing production technique and raw material (see online supplementary material (OSM)).
A pilot study combining portable x-ray fluorescence and thin-section petrography revealed considerable diversity in ceramic production (Hein & Stilborg Reference Hein and Stilborg2019). Further analysis of 84 ceramic sherds from 10 sites showed inter-site variation in technological traditions despite outwardly similar wares, and diversification in raw-material choice and processing over time (Hein & Stilborg Reference Hein2024). We also identified ‘Double Wares’, ceramics consisting of a fine-ware and a coarse-ware part (Stilborg & Hein Reference Stilborg and Hein.2021) and examination of 188 ceramic thin sections from four Neolithic sites showed how potting communities were connected across large distances (Dammer Reference Dammer2023). Further analyses are under way and results will be added to the open-access Ceramic Petrographic Database (https://doi.org/10.6078/M79Z930G).
Vessel use
Analyses of residues (Figure 3) and food crusts were largely inconclusive (Keute et al. Reference Keute, Isaksson, Devièse and Hein2021). We abandoned residue analysis until more effective methods are available. Instead, we conducted production-trace and use-alteration analysis on 123 Middle Bronze Age (c. 1500–800 BCE) vessels noting variation in the use of similar vessel types between sites.
Authors Hein and Womack conducting use-alteration analysis (photograph by authors).
Figure 3 Long description
Two individuals, one wearing glasses and a black top, and the other with a beard and a red plaid shirt, are in a laboratory setting. The person in the black top is holding a ceramic artifact while wearing blue gloves. The other person is assisting or observing. They are standing next to a table with various tools and equipment, including a computer keyboard and a red chair. The room has white walls and overhead lighting.
Zooarchaeology
To date, faunal remains from three or four sites from each period (Neolithic and Early and Middle Bronze Age) have been analysed, identifying taxa, skeletal elements and portions of bones present, age data and taphonomic markings (Figure 4, see also OSM). Domesticates are most common (cattle, pig, sheep, goat, dog); wild taxa are primarily deer, with some small carnivores, birds, rhinoceros and bovids. Non-diagnostic samples are currently undergoing ZooMS analysis (zooarchaeology by mass spectrometry). To investigate the genetic history of aurochs and cattle, samples have been selected for ancient DNA analysis and radiocarbon dating, which will further help contextualise the collections of worked bone artefacts, deepening our understandings of the non-subsistence use of animals, craftsmanship and local economy.
Authors Brunson and Ko studying animal bones at Stockholm University (photograph by authors).
Figure 4 Long description
Two individuals are seated at a table, wearing gloves and examining animal bones. The table is cluttered with various tools, notebooks, and boxes containing more bones. The individuals appear to be studying the bones closely, possibly identifying taxa, skeletal elements, and taphonomic markings. The background includes wooden drawers and a door with a posted document. The setting suggests a laboratory or research environment.
Archives
The MFEA holds extensive image and document-based archives related to Andersson’s work. We identified documents containing information on Andersson and his team’s fieldwork, the people involved and negotiations regarding splitting the collections between the MFEA and Chinese partners. These include field and personal diaries, letters, object and shipping lists, site maps, landscape sketches, object drawings, personnel lists, business cards and book-keeping documents. We are currently reconstructing the network of people and institutions involved in the fieldwork (Hein Reference Hein and Stilborg2024) and have been able to establish that many sites were discovered by Bai Wanyu. Originally employed as cook and groom, he also contributed directly to the archaeological work, as did many locals (Figure 5). We are currently investigating these contributions and those by other individuals.
Andersson, Mr Doodha and two assistants packing jars, Lanzhou, 1924 (MFEA image archives #280).
Figure 5 Long description
Four men are engaged in packing large jars and sacks outside a building with traditional Chinese lattice windows. Two men are seated on the ground, handling the jars and sacks, while the other two men are seated on chairs, overseeing the process. The building features intricate lattice designs on the windows and doors, indicative of traditional Chinese architecture. The scene appears to be from an earlier time period, likely the 1920s, based on the clothing and setting.
Future research
With the database built and a first inventory complete, we are turning to systematic recording, analysis and publication of materials from sites where find location and retrieval conditions can be reconstructed. We are combining this information with research on ceramics and animal bones, providing insights into ceramic production, subsistence patterns and food preparation and consumption. In the coming years, the database will be made trilingual and shared with scholars and the interested public worldwide. Our research will enable the MFEA to update the narrative around the history of the collection, highlighting the international nature of the 1920s excavations and giving full credit to the contributions of Andersson’s Chinese colleagues.
Acknowledgements
We thank the MFEA team and particularly Anna Schottländer, Adam Norberg, Mia Broné, Monica Sagren and Magnus Johansson. Special thanks to Ole Stilborg, a founding member of the project, and colleagues at Stockholm University, especially Jan Storå, Anders Götherström, Sven Isaksson and Jan Romgaard. We thank our students Zhuang Yu, Dayna Thomas, Avnish Kulkarni, Pan Dixuan, Clarice McKee and Caden Vandygrift. We are indebted to our colleagues in Gansu and the people living around the sites for welcoming us. We thank Alan Farahani for creating the complex project database.
Funding statement
This research was supported by the John Fell Research Fund (EBD00020Ö and EBD00460), US National Science Foundation (231673), Jackie Lambert Research Fund, Meyerstein Foundation, American Philosophical Society, Wesleyan University and Furman University, National Social Science Fund of China (23AH018), Henan Province Research Support Program (2025-CXTD-03).
Online supplementary material (OSM)
To view supplementary material for this article, please visit https://doi.org/10.15184/aqy.2026.10295 and select the supplementary materials tab.
Author Contributions: using CRediT categories
Anke Hein: Conceptualization-Lead, Data curation-Equal, Formal analysis-Equal, Funding acquisition-Equal, Investigation-Equal, Methodology-Equal, Project administration-Lead, Writing - original draft-Lead, Writing - review & editing-Lead. Andrew Womack: Conceptualization-Supporting, Data curation-Equal, Formal analysis-Equal, Funding acquisition-Equal, Investigation-Equal, Methodology-Equal, Project administration-Supporting, Writing - original draft-Supporting, Writing - review & editing-Supporting. Katherine Brunson: Conceptualization-Supporting, Data curation-Supporting, Formal analysis-Equal, Funding acquisition-Equal, Investigation-Equal, Methodology-Equal, Project administration-Supporting, Writing - original draft-Supporting, Writing - review & editing-Supporting. Jada Ko: Conceptualization-Supporting, Data curation-Supporting, Formal analysis-Equal, Funding acquisition-Supporting, Investigation-Equal, Writing - original draft-Supporting, Writing - review & editing-Supporting. Michel Lee: Conceptualization-Supporting, Project administration-Supporting, Resources-Equal, Writing - review & editing-Supporting.
Open access
