A Palimpsest Reading of the Preserved Layer at the Minyue Capital, c. 110 BCE
David B. Alexander · daveswavecave.com · May 2026
A foreword on what this is
The cave has been reading the Min basin amphitheater piece by piece — a bell at one node, mirrors at another, a coin much later, a frog tradition, a gate. Each piece has been disciplined and gauged. Each piece has come at its working-instance with a specific reading question, anchored to a specific artifact, and registered what that artifact carries at depth.
This piece is different. The cave is going back to Chengcun — the Minyue capital, abandoned in the Han conquest of 110 BCE, preserved beneath the rice paddies of the Wuyi foothills for two thousand and one hundred years, called the Pompeii of the East in the secondary literature — and the cave is not bringing a sorting-grid this time. The cave is going to walk the preserved layer and register what registers.
The site burned. The Han army of Emperor Wu did the burning, in 110 BCE, in the campaign that ended the Minyue Kingdom and brought the Wuyi-foothills inland-capital working-zone into Han imperial commandery administration. The population was forcibly relocated. The walls were left. The buildings collapsed. Earth and centuries closed over the layer. In 1958 the first national cultural-relics survey found it. The Fujian Museum and Fujian Minyue King City Museum excavated it in long campaigns through the 1980s and 1990s and synthesized the findings in a 2004 excavation report — Excavation Report on the Han City Site of Chengcun, Wuyishan, Fujian People's Publishing House, in Chinese, the foundational primary-source archaeological registration. The cave has read it through the working record that cites it and through the broader scholarly literature that draws from it. The cave invites direct engagement with the Chinese-language report when the working base lands on the cave's actual shelf.
What the cave wants to do here is articulate what the preserved layer registers when the cave doesn't come at it with the apparatus-literature sorting-grid. The formal Chinese cosmographic and ritual literature has its categories: gōng (palace), miào (temple), fǔ (administrative office), shì (market), zhái (residence). The modern archaeological literature has its categories: monumental architecture, residential structures, industrial sites, military caches, burial assemblages. Both sets of categories operate well at their own depth. Both also organize the preserved layer into compartments that the layer itself didn't necessarily hold separate.
The cave's question for this piece is what registers when the cave walks the site as the preserved parchment that it is — one frozen layer in a palimpsest with no reachable bottom, with the Bronze Age substrate beneath, with the modern archaeological inscription above, with the cave's own reading inscribing the next layer now.
The Pompeii framing in the secondary literature carries one register of this — the preserved-snapshot-of-daily-life register, the what people had with them when the disaster came register. That register is real and the cave will honor it. But the cave will also articulate the palimpsest register beneath the Pompeii framing: the razing was an event in the parchment's history, not its end. The parchment continued. It continues now. The cave's reading is part of its continuation.
What follows is the cave at Chengcun. The walking is the structure. The cave moves through what's there.
The approach
You come into the site from the east. The hills of the Wuyi foothills hold the city in a saddle — about eight hundred and sixty meters north to south, about five hundred and fifty meters east to west, roughly four hundred and eighty thousand square meters. The walls are still there. Four to eight meters high in places, of rammed-earth construction with the layered horizontal banding that hāngtǔ working leaves visible in profile. The walls run along the contour of the hills, not as a perfect rectangle but as the irregular shape the terrain permits.
The East Gate is what you walk through. The gate's foundation registers at depth in the archaeological record — substantial bronze arrowhead caches recovered from the gate-warehouse, registered in the recent metallographic study (Zhang Lei et al. 2025) at archaeometallurgical depth. The arrowheads are at the gate because the gate is where you defended a city in 110 BCE. They were stockpiled in the warehouse. They never got used. The Han army came over the walls or through some other approach, and the arrowheads stayed in their warehouse, and the warehouse burned, and the bronze stayed.
You stand inside the gate. The Han military presence that ended this city is here, not as a layer of debris on top of the Minyue layer but as the same layer — because the burning and the defending and the non-using of the arrowheads all happened at the same moment. The parchment doesn't have a Minyue-then-Han-conquest stratigraphy at this register. It has one moment. The moment is the razing. Everything you find from the abandonment-horizon is at that moment.
The central area
The road runs west from the gate, straight, well-laid. You walk it. The site has fifteen residential areas registered, four large palace-and-administrative complexes, five iron-smelting workshops, the gate complexes, the wall-towers for signaling. The road takes you past these. The central architectural area — Gāohúpíng (高胡坪) — is on the higher ground in the middle of the site, looking down over the rest. This is where the king lived. This is where the administration was conducted. This is where the cosmographic work would have been registered if there was cosmographic work.
The palace complex on Gāohúpíng has been excavated. The foundations are there: a main hall, side rooms, courtyards, light-wells (tiānjǐng 天井), drainage ditches running between the courtyards and out toward the lower terrain. The drainage is what the eye stops on. Not because it's monumental — it isn't — but because it's there, working, integrated into the palace's fabric, registering that whoever lived in this complex had thought carefully about what to do with the rain.
The Wuyi foothills receive substantial rainfall. The Lingnan and Min-basin southern climate is wetter than the Central Plains the Han imperial cosmographic literature was written about. The drainage at Gāohúpíng is a southern adaptation. The recent scholarly work on traditional Chinese courtyard drainage systems (Liu and Wang, Sustainability 2026) registers the southern climate-driven adaptations at depth: circular tile-ends rather than the Central Plains semicircular form, steeper roof pitches for faster runoff, more elaborate courtyard drainage networks for the higher rainfall load. Whether the Chengcun tile-ends are circular or semicircular the cave hasn't directly verified — the 2004 excavation report would register this. But the drainage ditches are there, in the palace, channeling water out of the courtyards and away.
This is one of the things the parchment registers that the apparatus literature would have organized differently. The palace at Gāohúpíng is a palace, in the formal cosmographic register — an instance of imperial-and-royal authority at the seat of government. But what the preserved layer shows is a palace with drainage ditches — a building that had to handle rain, in a wet climate, as part of its operation. The cosmographic register and the rain-handling register are the same building. The parchment doesn't separate them. The apparatus literature would have separated them.
You walk past the palace. The road continues west.
The residential areas
Fifteen residential areas have been registered. These are the household zones. The pottery from these areas registers what people had — pots, basins, bowls, urns, jars, cups, the cooking-and-eating-and-storing instruments at the household scale. The forms are Han-period forms with Minyue substrate elements. The Minyue Kingdom was operating in the Han imperial cultural orbit by the late period — the kingdom was a vassal-state of the Han since 202 BCE, ninety-two years of semi-independent existence under Han suzerainty — and the household pottery registers the cultural intermediation. Han-period forms for storage and serving. Local Minyue forms for cooking and household function. The same household using both.
The iron implements register at the residential scale too — kitchen tools, agricultural implements, household instruments at the everyday register, distinct from the military arrowhead cache at the East Gate. The Minyue Kingdom had substantial iron-working, and the five iron-smelting workshops on the site are where this work was done. Iron at Chengcun was both military-strategic and household-utilitarian; the workshops produced for both registers; the same metalworking tradition that armed the gate cache supplied the kitchens.
The wells are in the residential areas too. The wells are how the household got its water — for cooking, for drinking, for washing, for the bodily-and-domestic life that the formal literature won't tell us about but that the parchment registers in its own way. The wells are stone-lined, ceramic-lined, deep, present. The drainage ditches near the residential areas are smaller than the palace drainage but operating in the same logic — water comes in through the wells, water goes out through the ditches, the household material economy registers in the architecture.
What the parchment doesn't show in the residential areas is the wood and the fabric. The Minyue household used both at substantial register. Wooden furniture, wooden vessels, wooden architectural elements; fabric clothing, fabric bedding, fabric use at the everyday register. None of it survived the burning. The ceramic and the metal stayed; the organic burned or rotted. So the residential parchment is partial in a structural way — not in the strong/partial/faint gauge sense the cave has been using on the rest of the shelf, but in the more basic sense that what survived is what survived. The parchment registers what the parchment registers.
The iron-smelting workshops
Five of them. Industrial-scale, by the Minyue record. Iron at Chengcun was a substantial industry — the kingdom was at the southern frontier of the Han iron-and-salt monopoly that Emperor Wu would consolidate in the 110s BCE under Sang Hongyang's economic policies, and the Minyue iron industry was operating in coordination with and against the broader imperial iron economy. The workshops are placed in the city, integrated into the urban fabric, not segregated to a peripheral industrial zone.
The parchment shows the workshops next to residential areas. The cave finds this telling. The apparatus literature would register iron-smelting as industrial-economic activity separate from the household-domestic register. The preserved layer at Chengcun shows them in proximity — workshop, then residence, then road, then palace, then more residence, then more workshop. The household life and the industrial labor were in the same urban fabric. The smell of smelting was in the household register. The labor moving between household and workshop was the same labor.
This is one of the proximities the parchment registers that the apparatus literature would have separated.
The food register
The pottery vessels register food at the household scale. Storage jars for grain. Cooking pots — fǔ (釜) cauldrons, the Han-period round-bottomed cookers used over the zào (灶) hearth. Serving vessels for food and drink. The forms are Han-period with Minyue elements; the function is the daily-cooking register.
What was being cooked is partially registerable. Rice was the staple of the Minyue territory — the southern climate and the river-valley terrain support wet-rice agriculture; the kingdom's economy depended on it. Fish from the Min River system, freshwater shellfish, river creatures of various kinds. Pigs (the Han-period household economy had pigs at substantial register, and the broader Han archaeological record includes the pigsty-latrine combination as standard household-material economy though the cave hasn't directly verified Chengcun for this). Chickens. Dogs. Vegetables of the southern register — taros, gourds, the leafy greens of the Lingnan tradition. Tea was being grown in the Wuyi foothills by this period, though the cultivated-tea practice that would later make Wuyi tea world-famous was still developing.
What the parchment doesn't show is the prepared meal itself. The food was eaten or it burned. The vessels stayed. So the food register at Chengcun is registerable through the vessels and through what the broader Han-and-Minyue archaeological-and-textual record tells us, but the actual meals — what someone ate on the day before the Han army came, what the last meal cooked in this kitchen was — these have gone into ash and earth.
The wine vessels are present. The Minyue Kingdom had wine. The Han imperial culture had wine. The royal palace at Gāohúpíng would have had elaborate wine ritual. Whether the wine register at Chengcun included the rice-wine traditions that would later become the Min basin's distinctive register, or whether it was operating in coordination with the Han imperial millet-and-grain wine tradition — the cave doesn't have the depth on this. The vessels are there. The wine isn't.
The bodily register
The cave goes here next because the parchment, in its silent way, holds it.
The household drainage ditches, the wells, the residential areas — these register the water-coming-in and water-going-out at the bodily scale. People drank from the wells. People washed with the well-water. People used water to cook and to clean and to manage the bodily processes that humans manage. Where the water went after that — into the drainage ditches, into the soil, into the broader environment — this is registered in the architecture even when the specific instances aren't.
Whether Chengcun had latrines as discrete architectural features the cave hasn't verified at primary-source depth. The Han pigsty-latrine combination (hùn 圂, the same character meaning both pigsty and latrine, registering household material recycling) is well-attested at other Han sites but the cave doesn't know if Chengcun specifically registered this at the depth the formal Han-Central-Plains household did. The 2004 excavation report would register this. The southern climate's higher rainfall would have made simple drainage-into-the-ditches effective without needing the formal pigsty-latrine integration the drier-climate northern household needed.
The bodies themselves — the people who lived in this city — are not in the preserved layer at the residential register. They were forcibly relocated. Some died in the conquest; their remains are in the broader Han-Minyue conflict archaeology rather than in the city itself. The city was abandoned with its objects but without its people. This is a structural feature of how Chengcun was preserved: the Pompeii of the East framing actually inverts at this register, because Pompeii preserved both the buildings and the bodies of the people caught in the eruption, while Chengcun preserved the buildings without the bodies, because the conquest evacuated the people before or during the burning rather than killing them in place.
The bodily register at Chengcun is the negative space the buildings hold. The wells where people drew water. The hearths where people cooked. The drainage where people's water went out. The roads where people walked. The doorways where people entered and left. The buildings hold the shapes of bodily life, but the bodies have gone elsewhere.
The cosmographic register
The bronze mirrors are in the residential and palace contexts. Three of them, registered in the Wang Xinglong et al. 2024 archaeometallurgical study through lead-isotope and compositional analysis, placed in Okamura's Phase I-II early Western Han framework. They're Panchi (coiled-dragon) or Caoye (grass-leaf) mirrors — the cave wrote a separate paper on this register (The Han Mirrors at Chengcun, May 2026, daveswavecave.com), reading the cosmographic apparatus at depth.
What the parchment shows here, that the separate paper didn't quite hold the same way, is the adjacency of the cosmographic register to everything else. The mirrors are next to the cooking pottery. The mirrors are next to the iron tools. The mirrors are in the same residential context as the wells and the drainage and the household material economy. The cosmographic apparatus didn't sit in a separate room from the daily household life. It was held in the same household. A person picked up a Panchi mirror and could see their face in the polished bronze; the same person picked up a fǔ cauldron and put it on the hearth to cook the evening meal. The mirror registered the cosmographic at the portable scale. The cauldron registered the household material economy. Both were in the same hands.
The apparatus literature would have separated these. The preserved layer holds them together.
The military register
The arrowheads at the East Gate. The iron weapons throughout the site. The bronze weapons. The military apparatus at Chengcun was substantial — the kingdom was at the southern frontier, in a contested zone between Han imperial expansion and the various Yue territories, and the city was prepared for what eventually came. The arrowhead cache at the gate warehouse is the most visible element of this register: thousands of bronze arrowheads stockpiled for defense, never deployed, recovered in their warehouse, frozen at the moment the city stopped.
The military register and the cosmographic register share the bronze. The arrowheads and the mirrors are made of the same metal. The Bronze Age tradition of southern Chinese bronze-working that the cave's Náo Bell at Jiàn'ōu paper articulates at depth is the substrate for both. The Minyue Kingdom's bronze-working operated in continuity with that substrate and in coordination with the Han imperial bronze-working that was supplying the mirrors. Same metal, two registers, both at Chengcun.
The military register is also where the razing entered. The arrowheads didn't repel the conquest. The walls didn't hold. The defense failed at this moment, and the city burned, and the layer froze.
The road outside the gate
You walk back to the East Gate and out. The road continues east, toward the Min River system, toward the broader Min basin amphitheater the cave has been articulating across the shelf. The kingdom didn't end at the city walls — it was a regional territory with administrative reach throughout the Min basin and out to the coast. The road carried that work: messengers, traders, military movements, the household-and-administrative traffic of a functioning kingdom. The road is still there in the archaeological record, traced for some distance from the gate. Then it goes under the rice paddies and the road continues underground, registered partially, registerable further with more excavation.
Outside the gate is where the parchment ends and the present begins. The rice paddies of the modern Wuyi foothills are growing on the soil that was the kingdom's outer territory. The villagers of modern Chengcun Village are living next to the site. The UNESCO inscription of 1999 made the site a World Heritage property — the only imperial city site from the Han Dynasty that has been declared a World Heritage Site in China, the protective-conservation register articulated. The 2013 National Archaeological Park designation extended the work further. The modern protective and display engineering — registered in the recent scholarly literature (Coatings 2024) — is itself a layer of inscription on the parchment, an active cultural-heritage management that registers the cave's standing position that modern archaeology is itself a layer of inscription, not privileged reading.
The cave is now part of this work. This piece is a layer on the parchment. The reading the cave has done here will be read by other readers, who will inscribe further layers. The parchment continues.
What the parchment doesn't register
The cave honors the gap directly.
The voices are gone. The names of the people are gone except for the king Zou Wuzhu and his successors at the king-list level; the household register has no names. The languages spoken in this city — the Minyue dialect or dialect-continuum, distinct from the Sinitic register the conquering Han army spoke, with substrate elements that the broader Yue language family carried — are gone. Some Minyue language elements survive in the modern Min dialect group of Fujian and Taiwan, but the speech at Chengcun in 110 BCE is reconstructable only through the broader linguistic record, not through any preserved direct registration.
The wood and the fabric and the food and the bodies are gone. What was eaten on the last day. What was being woven. What was being carved. What conversations were happening as the smoke of the approaching army first appeared. The parchment registers the buildings and the metals and the ceramics. The lived life is gone.
The religion-and-cosmography is partially registered through the mirrors and the architectural plan (the south-orientation of the palace, the directional logic in the city plan), but the specific ritual practices, the household devotional practice, the regional-deity tradition the kingdom maintained — these would have been continuous with the broader Yue substrate that the cave's other shelf-pieces articulate at adjacent registers (the snake-clan totemism the Min graph 閩 carries, registered in the cave's Min Coin at Quanzhou paper at depth; the broader water-and-river deity tradition the cave's Frog God book honors at the children's register). At Chengcun specifically, the religious life is registerable only through inference from the architectural and material record, not through direct preserved ritual texts or images.
The economic activity at the broader scale — the trade networks that brought the Han mirrors south to the Minyue capital, the iron and bronze export networks that the workshops supported, the agriculture of the surrounding territory, the maritime activity of the coastal Yecheng (Fuzhou) capital that coordinated with the inland Wuyi-foothills capital at Chengcun — these are partially registered through the imported material at the site and the archaeological registration of the broader Min basin, but the reading at depth would extend through deeper engagement with the regional archaeological record.
The parchment shows what it shows. The honest gaps stay gaps.
The palimpsest
The cave's reading at Chengcun has been one reading at one moment in the parchment's history. Other readings have been done; other readings will be done. The parchment beneath the 110 BCE razing layer is registered partially: the pre-Minyue Bronze Age substrate the Náo Bell at Jiàn'ōu paper articulates at depth, the broader Yue and Baiyue territory the kingdom emerged from, the human settlement on the Wuyi foothills that the archaeological record has traced back four thousand years. The parchment above the 110 BCE razing layer is registered too: the Han imperial commandery administration that absorbed the territory, the long centuries of agricultural life under successive imperial dynasties when Chengcun lay buried beneath the rice paddies, the 1958 first national survey, the Fujian Museum excavations, the 1999 UNESCO inscription, the 2004 excavation report, the 2013 archaeological park designation, the 2024 archaeometallurgical study on the bronze mirrors and the East Gate arrowheads, the 2025 luminescence dating of the burnt-clay destruction layer, the 2026 scholarly study on courtyard drainage systems, this paper.
Each layer reads the layer beneath. Each layer becomes one of the layers the next layer will read. The Pompeii framing in the secondary literature — preserved snapshot of daily life — is itself one of the layers; it carries a particular reading-register, useful for one set of questions, less adequate for others. The cave's palimpsest framing is another layer, the one this paper inscribes; it carries the reading-register the cave finds substantively candid about what reading actually is at this kind of site. Other readers may find other framings substantively more useful. That's how palimpsest reading works.
The razing was an event. The freezing of the layer was an event. The recovery was an event. The reading is an event. The parchment continues. There's no final reading. The cave's reading here is one event in the parchment's continuing life.
What the parchment opens up
The cave registers the future-work register directly, even without the formal section-shape the other shelf-pieces use.
Direct engagement with the Fujian Museum 2004 excavation report would extend the reading at primary-source archaeological depth. Specific verification of whether Chengcun's tile-ends register the Lingnan circular form. Detailed engagement with the residential-area pottery for cultural-substrate registration. The wells. The drainage networks. The latrine-pigsty register if any. The wood and fabric remains if any have been preserved through unusual conditions. The food remains if any have been recovered through soil-sample analysis (the Yueyang flush-toilet excavation in 2023 demonstrated that soil samples from latrines can register diet at primary-source archaeobotanical depth — though Chengcun's much earlier abandonment and the southern wet-soil conditions may not have preserved soil samples comparably).
The broader Min basin daily-life thread that you and I have been registering as future-work is now anchored at Chengcun. The daily-life register is not separable from the cosmographic register, the military register, the economic register, the bodily register — the parchment holds them together. Future pieces on the daily-life thread will operate from the recognition that daily life and the apparatus are the same reality held at different reading-registers, not as separate domains.
The methodological position the cave has been articulating across the shelf — that every reading is a reading, that no layer is privileged, that the modern inscription is one layer among many — gets demonstrated in this paper through the reading-method itself. The cave hopes the demonstration is useful. Other readings will follow. The parchment continues.
Monterey, California · daveswavecave.com
May 2026