The imperial kilns at Jingdezhen, Chenghua reign

Almost everything a buyer needs to know about a Chenghua porcelain is decided by six characters on the underside, and those six characters prove nothing. Da Ming Chenghua nian zhi, made in the Chenghua reign of the Great Ming, is one of the two most frequently copied inscriptions in the history of ceramics. Kangxi, Yongzheng and Qianlong potters wrote it on their finest work as homage. Nineteenth-century workshops wrote it out of habit. Twentieth and twenty-first century forgers write it because it is the cheapest feature of the object to imitate. No source consulted for this page will put a figure on how much marked Chinese porcelain carries a mark that does not match its date, and the compiler of the standard open reference on Chinese porcelain marks does not need one: his rule is that marks are helpful but, as a general rule, the last thing you should trust in dating a piece. So the first working rule of this field is the opposite of what a collector expects. The mark is what you look at last.

What replaced the mark as evidence came out of the ground. From the late 1970s the Jingdezhen Institute of Ceramic Archaeology, working the Ming imperial kiln site at Zhushan under Liu Xinyuan, recovered a quantity of broken porcelain that is now given at more than twenty million pieces. Those sherds are not debris. The imperial factory destroyed on site everything it would not send north, so what was buried at Zhushan is a stratified record of exactly what the kilns attempted in each reign, in the reign itself, with no possibility of a later hand having added anything. A Chenghua stratum with a half-finished doucai cup in it, outlined in cobalt and never enamelled, establishes the type in the reign more securely than any whole cup in any collection. The Palace Museum in Beijing built six exhibitions out of that principle between 2015 and 2019, setting excavated rejects beside imperial survivors reign by reign; the Chenghua edition opened on 25 October 2016 with more than 180 pieces.

The third thing to know is the scale, because it is the reason the prices are what they are. This is not a long production run of a famous ware. The palace bowls belong to roughly the decade before 1485, and Regina Krahl records that Chenghua fragments recovered at the kiln site amount to less than half the volume recovered from the much shorter Xuande reign. Around two dozen Chenghua pieces of mark and period, of any design, are recorded in private hands anywhere. The National Palace Museum in Taipei holds more than ninety per cent of the world's Chenghua doucai. Everything else is a museum object, a sherd, or a copy.

A factory, a decade, and six characters

There is no artist on this page and there was never meant to be one. The imperial kiln factory at Zhushan in Jingdezhen was established in 1369, at the start of the Ming, and was administered directly by the court rather than through the commercial kilns that surrounded it. Production was controlled by a ministry in the capital, by then Beijing, roughly seven hundred miles to the north. Court artists supplied drawn or woodblock printed designs. Work at the kilns was divided into specialities to increase efficiency and consistency, which is the exact opposite of the conditions under which a personal hand becomes visible. The man who threw a Chenghua wine cup did not paint it, and the man who drew the blue outlines did not lay in the enamels.

So an attribution here is not to a person. It is to a place, a factory and a decade, and it is made from the body, the glaze, the palette, the drawing, the foot and the mark taken together. Every one of those can be imitated and several of them have been imitated well. That is why a page about the Chenghua kilns has to be organised around evidence rather than around a life: the mark and what it does and does not prove, the excavated material and what it settles, the technique in enough detail to be diagnostic, the administrative system that produced and destroyed, the corpus and who counts it, and the money.

One consequence should be stated at the front because it governs everything that follows. Iconotheca's record of the Meiyintang chicken cup, sold at Sotheby's Hong Kong on 8 April 2014 for a hammer price of HKD 250 million and HKD 281.24 million with the buyer's premium, covers that single object: its chain of owners, its census, its sale and what its buyer did with it afterwards. This page does not repeat any of that. It describes the manufactory that made it, and the great majority of what follows applies to several hundred surviving objects rather than to one.

Da Ming Chenghua nian zhi, and how to read it

The standard imperial mark of the reign is six characters in underglaze cobalt blue, written on the base in regular script, arranged in two vertical columns of three and read from the top of the right-hand column downward and then the left. The first pair names the dynasty, Da Ming, Great Ming. The second pair names the reign era, Chenghua. The last pair, nian zhi, means made in the period. Vertical, right to left is the ordinary convention for Chinese text and derives from the bamboo strips on which it was originally written; where a mark runs in a single horizontal row on a rim or a foot it is still read right to left.

The characters sit inside a frame. Chenghua marks are found within a double circle and within a double square, and the framing device is itself a dating tool. The double square is an innovation of the late Chenghua phase and is characteristic of the doucai wares; the double circle is the form on the blue and white palace bowls, and a Chenghua palace bowl offered at Sotheby's in 2016 carried its six characters within a double circle. Neither frame is decoration. Each narrows a date within a reign that lasted twenty-three years and whose imperial production was concentrated in the last ten.

What the mark looks like is the part that matters and the part that is hardest to find written down. It is not elegant. The Chenghua reign mark is thick, often unbalanced, and immature, and that impression is consistent enough across genuine examples to be diagnostic. Jan-Erik Nilsson, who has published the standard open reference on Chinese porcelain marks for a quarter of a century, sets out seven characteristics of a mark true to the period. On the first character, meaning great, the beginning of the second stroke seldom extends much beyond the first stroke, so that it looks stubby, and where it does extend the beginning is fat, while the third and final stroke ends thickly. On the third character, cheng, the third descending stroke is not curved but straight and vertical. The fifth character, nian, is unusually squat and square. On the last character, zhi, the ninth stroke does not extend beyond the standing knife radical. The final tails on most characters are abrupt and sharp, like fish hooks. Where the surround is square it is thick with ink at each right angle. And the mark in general is faintly obscured, as if covered with a thin haze.

A researcher working from the National Palace Museum's own Chenghua holdings reaches the same conclusion by a different route, describing the brushwork rather than the shapes: the genuine marks lack the lifting-up and the wrapping of the brush tip that a trained calligrapher uses to close a stroke, and the four stages of a stroke, starting, connecting, turning and closing, are visibly incomplete. That writer draws the contrast that makes the whole thing memorable. The painting on a Chenghua doucai cup is among the most accomplished ever executed on porcelain. The six characters underneath it are not accomplished at all, and the gap between them is one of the marks of the reign.

Whose handwriting is it

Two explanations are on offer for why the mark looks the way it does, and neither has been proved.

The first is that a single calligrapher wrote all the marks on all the official porcelain of the reign. That is plausible on general grounds. The number of hands entrusted to write the imperial mark at the imperial kilns was small in any period, small enough that an individual handwriting can often be recognised, which is one of the few structural advantages the authenticator has over the forger. It is not, however, established for Chenghua by any document consulted here.

The second explanation belongs to Liu Xinyuan, the archaeologist who directed the excavations at the Jingdezhen imperial kiln site. His view, reported both in the Sotheby's catalogue note on the Meiyintang cup and, first hand, by Nilsson from a conversation at the excavations in the early 1990s, is that the distinctive and somewhat naive calligraphy of the Chenghua reign mark is the hand of the young emperor himself, written when his handwriting was not yet good and then copied thereafter. Zhu Jianshen came to the throne on 23 February 1464 at the age of sixteen. If Liu is right, the one trace of an individual person anywhere in the output of an anonymous state factory is six clumsy characters on the underside of a wine cup.

This record does not adopt that reading and does not dismiss it. It is a proposal by the man who dug the site, made without a document to support it, and it has never been tested against the marks themselves in any published study reachable here. Nilsson, who heard it directly, records it and immediately adds that he is not sure the single-calligrapher assumption can be made either. Both positions are set down because the honest state of the question is that nobody knows who wrote the Chenghua mark, and the entire market for the reign rests on being able to recognise it.

The apocryphal mark, and why it is usually not a forgery

The Chenghua mark is, with Xuande, the most copied reign mark in Chinese ceramics, and the reason is a compliment. Imperial porcelain of the Chenghua period was considered the highest quality ever produced, and later potters signed their best work with the earlier name as an act of reverence. A vast number of Kangxi pieces carry Chenghua marks for exactly that reason. The trade word for the practice is apocryphal, and it appears constantly in auction catalogue descriptions; what it means is that the mark records an intention and an homage rather than a date.

Two museum-catalogued cases show the practice being handled honestly. The Cleveland Museum of Art holds a pair of famille verte bowls with Xiwangmu and attendants, accession 1964.218, made in the Kangxi reign between 1662 and 1722 and carrying Chenghua marks from two centuries earlier; Cleveland catalogues them by what they are, giving the culture line as Qing dynasty, Kangxi reign, but Chenghua mark. That pair is discussed in the Iconotheca record on the Meiyintang cup and is the reference case for the whole class. The Victoria and Albert Museum in London holds a second, less well known example: a Kangxi period dish painted with a river scene, two figures under a rock and a pine, which carries a spurious Chenghua mark on its underside. The mark is not visible in a view of the interior, which is itself the point. A specialist looks at the drawing, the palette and the glaze first and turns the piece over last.

Fraud is a separate and continuing category and it converges on the first over time. An honest homage piece loses its context, acquires an owner who does not know what it is, then a dealer who suspects, and eventually a description that omits the word apocryphal. Meanwhile marks continue to be applied to new objects for the antique markets. A four-character version, Chenghua nian zhi, written in archaising seal script rather than the six-character regular-script form of the reign, circulates widely on much later porcelain and is one of the commoner tells; the reign-period wares do not use it.

The practical rule follows and it is unglamorous. A mark is a terminus for an intention, not for an object. The date is established from the porcelain body and its translucency, the character and surface of the glaze, the range and quality of the enamels, the tone of the underglaze blue, the drawing, and the finish of the foot. Kate Hunt of Christie's Chinese works of art department puts the same test in one sentence a collector can apply without a laboratory: read the mark in conjunction with the quality of the work of art, because an imperial mark on an object that is not of imperial quality is a problem, while a crude mark on a piece of ordinary minyao or people's ware is perfectly consistent with itself.

The waster heaps at Zhushan

The single most important body of evidence about this workshop is not in any collection. It is the broken porcelain the factory itself buried.

Imperial quality control at Jingdezhen was absolute and the disposal of failures was part of it. Pieces with the smallest flaw, and pieces surplus to an imperial order, were destroyed on site so that they could not reach the market. Sotheby's own history of Chinese imperial porcelain states it flatly: quality control by the court was so strict that any vessels with even the most minor flaws, and any surplus to imperial orders, were discarded and destroyed. The consequence for the archaeology is that the ground at Zhushan holds a dated, stratified, uncontaminated sample of what the kilns made in each reign, including everything they made and rejected.

Systematic recovery began in the 1970s under Liu Xinyuan and the institute that is now the Jingdezhen Imperial Kiln Institute, with extensive excavation from 1982. The Institute now gives the total recovered since the late 1970s as more than twenty million kiln relics, running from the late Tang to the Republic. A further joint season with the School of Archaeology and Museology at Peking University and other bodies ran from 2002 to 2004 and produced the site's architecture as well as its contents: the enclosure and courtyard wall on the north of the site, groups of gourd-shaped kiln furnaces facing steamed-bread-shaped stoves at the southern foot of Zhushan hill, and, crucially, small pits and piles of sherds and imperfect wares that had failed selection for the court and been dealt with centrally, with others scattered anywhere nearby. Quan Kuishan's report notes that most of the excavated wares are imperfect in colour and glaze and that several types were discovered there for the first time.

What this settles that a whole vessel cannot is a matter of logic rather than of connoisseurship. A complete cup in a collection has a history; a sherd in a sealed layer at the factory has a manufacture. Sherds of an unfinished doucai chicken cup, painted in underglaze blue only and still lacking its enamels, were recovered from the third and last Chenghua stratum, datable to the final years of the reign. That fragment does three separate jobs at once. It fixes the type to the reign independently of every surviving cup. It proves the sequence of the doucai process directly, because the piece was caught in the middle of it. And it shows that the complete design was drawn onto the body in pale underglaze blue before firing, even where the outlines would later be hidden under the enamels, which is the technical definition that separates doucai from wucai.

The material is not confined to Jingdezhen. Liu Xinyuan's survey of the Chenghua kiln site and its finds was published in A Legacy of Chenghua: Imperial Porcelain of the Chenghua Reign Excavated from Zhushan, Jingdezhen, the catalogue of a 1993 exhibition mounted by the Jingdezhen Institute of Ceramic Archaeology with the Tsui Museum of Art in Hong Kong; that catalogue is still the reference cited by auction houses when a piece is matched to an excavated fragment. The institute had already shown excavated material with the Fung Ping Shan Museum at the University of Hong Kong in 1992. In 2013 a selection of 118 pieces and sets of excavated and restored Chenghua imperial porcelain travelled to the Civic and Municipal Affairs Bureau gallery in Macau, arranged in four categories: doucai, blue and white, monochrome and coloured glazes, and imitations of Song dynasty wares.

Setting the rejects beside the survivors

Between 2015 and 2019 the Palace Museum in Beijing ran a series of six exhibitions built on a single curatorial idea: put the pieces the emperor received next to the pieces the same kilns made and threw away in the same year. The Chenghua edition, The Ming Imperial Porcelain: A Comparative Exhibition of Archaeological Findings at the Imperial Kiln Site in Jingdezhen and Chenghua Period Porcelain in the Collection of the Palace Museum, ran from 25 October 2016 to 26 February 2017 with more than 180 porcelains. The sixth and final edition, covering the Jiajing, Longqing and Wanli reigns, ran from 6 November 2018 to 22 February 2019 with 298 exhibits drawn from the palace collection and from the Jingdezhen excavations.

The exhibitions are worth citing not as events but as method. They put on public record why individual pieces failed, and the reasons are specific rather than aesthetic: deformity, incorrect inscription, wrong painting. One Chenghua blue and white dish with dragons and phoenixes was rejected because the painter had given one dragon six claws and another four. The imperial standard was five. Nothing was wrong with the potting, the glaze or the blue. A drawing error in a claw count was enough to have the object smashed and buried, and it stayed buried for roughly five hundred and thirty years until an archaeologist reassembled it.

Two conclusions follow that bear directly on any provenance. The first is that survival rate is not production rate. The number of Chenghua pieces in the world is a function of what the factory allowed out, not of what it made, and the factory was the first and harshest editor in the chain. The second is that the excavated corpus and the transmitted corpus overlap only partially, so a design known from sherds and unknown from whole vessels is not a fake type, and a design known from whole vessels and absent from the sherds is not thereby suspect either. Both statements sound obvious and both are routinely got wrong by people describing objects for sale.

Doucai: two firings and the arithmetic of loss

Doucai has been translated as contrasted colours, contending colours, fitted colours, colours which fit together and dovetailed colours. The National Palace Museum in Taipei, on the gallery label in its own pottery and porcelain rooms, gives it as contending colours and then describes the whole procedure in four clauses: outlining forms on porcelain in cobalt blue, covering the piece with transparent glaze and firing at high temperature, then filling the spaces with colours, then putting the piece into the kiln again to be fired at low temperature and completed.

The temperatures are what make it difficult. The first firing is a full porcelain firing. The second, which fixes the overglaze enamels, runs at roughly 850 to 900 degrees, and every colour that will not sit beside another in the same pass requires a further trip through the muffle kiln. On the chicken cups several enamels were superimposed to build the variegated plumage, and each superimposition meant another firing. The palette of the reign is iron red in more than one tone, greens ranging from a bright to a dark olive, yellow, and a transparent aubergine which can be laid over other colours and which therefore costs an extra firing wherever it appears, all of it sitting inside a shaded, pale underglaze blue.

The consequence is arithmetical rather than mystical, and Sotheby's states it plainly: the repeat firings necessary for the process would naturally have reduced the number of successfully completed examples. Scarcity in this ware begins inside the kiln, before any question of five centuries of breakage, and before the inspectors got anywhere near it. A Chenghua doucai cup that survives has passed the first firing, two or more enamel firings, the imperial inspection, and then five hundred years of handling. Each of those is a filter and only the last of them is chance.

The technique did not begin in the reign. A doucai dish with red and green enamels was excavated at the Jingdezhen kilns from the Xuande period, 1426 to 1435, and Xuande doucai was only recognised as a category at all once the Zhushan fragments made the comparison possible. What Chenghua did was perfect it. The Taipei label puts the sequence exactly: the technique began in the Xuande reign and was perfected by the time of the Chenghua Emperor, and the durability and pure translucence of the thin bodies, together with the contending colours, made these wares renowned at the time and the object of imitation in later generations. Cleveland's description of its own Chenghua doucai wine cup with children at play, accession 1957.61 and only four point eight centimetres high, is the same account written from the object: free, spontaneous drawing, perfect compositional balance over a very small surface, and a porcelain body that is thin, translucent and light in the hand.

The blue is domestic, and that is the point

The softness that specialists praise in Chenghua underglaze blue is a raw material fact before it is an aesthetic one, and it is one of the strongest dating arguments available.

Early Ming imperial blue and white was painted with imported cobalt, the ore the Chinese sources call sumaliqing, brought from the Middle East and used for the Yongle and Xuande wares. It is strong, saturated, and prone to run and to throw dark iron spots where it pools. From the middle and late Xuande period domestic ore came into use alongside it, and by the Chenghua reign the domestic material had taken the leading position. The ore is pingdengqing, also called pitangqing, mined at Leping in Jiangxi, close to Jingdezhen itself. It is higher in manganese and lower in iron than the imported ore, it diffuses much less in the glaze, and it fires to a quieter, greyer, more even tone. It is the standard cobalt of the Chenghua and Hongzhi reigns.

Nicolas Chow, describing a Chenghua palace bowl for Sotheby's, calls the result a very soft, watery cobalt blue and treats it as a deliberate break from the early fifteenth century, when the cobalt was imported and very strong and punchy. That is the right way round. The quiet blue was not a limitation that the Chenghua potters were unable to overcome; it was the effect they wanted, and it is inseparable from the other thing the reign did with its blue and white, which was to leave a great deal of the porcelain white. Where the Yongle and Xuande painters filled the surface, the Chenghua painters set a single scrolling stem against an open field and stopped.

The material argument has since been tested instrumentally. Collaborative research between the Jingdezhen Imperial Kiln Institute and Peking University, reported in 2025, confirmed the coexistence of imported and domestic raw materials from the Xuande period onward, which is exactly the transition the connoisseurial account describes. That is the useful shape of evidence in this field: a proposition arrived at by eye in the nineteenth and twentieth centuries and then confirmed by composition in the twenty-first, using sherds from the factory floor as the reference set.

How an order reached Jingdezhen

A Ming imperial porcelain begins as a document in Beijing and ends as a crate on a river. The intervening machinery is the reason this page describes a manufactory rather than a workshop.

The imperial kiln factory was established at Zhushan in 1369, in the second year of the dynasty, and by 1402 there were twelve imperial kilns at Jingdezhen. Production was controlled by a ministry in the capital, far to the north, and supervised locally by officials delegated to Jingdezhen from the imperial workshops at court. Court artists supplied the designs, drawn or printed from woodblocks, so that the pattern on a Chenghua dish was fixed in Beijing and executed in Jiangxi by men who had never seen the room the object was going to.

The quantities are the part that surprises people who know the reign only through eight-centimetre wine cups. A single palace order in 1433 called for 443,500 pieces, all with dragon and phoenix designs. In 1459 the court sent an order to the supervisors at Jingdezhen for 133,000 pieces, later reduced to 80,000 because the resources to make them were not there. Imperial wares were not only for the emperor's table: they went to the provincial courts of Ming princes and out of the empire as diplomatic gifts, which is one reason a fifteenth-century Jingdezhen stem cup can be found today in the ceramics gallery of the Tibet Museum in Lhasa.

The labour was industrial and specialised. Production employed hundreds and probably thousands of workers whose tasks were divided into separate trades to increase efficiency and consistency: chopping the firewood, purifying the clay, grinding the cobalt. By the sixteenth century there were hundreds of kilns at Jingdezhen with thousands of workers. The environmental cost is recorded too: mountains stripped bare for firewood and fern ash, hillsides quarried for clay, and the sky over the town dark with smoke. None of the people who did any of this is named. What we call a Chenghua chicken cup is the output of an administrative system operating to a specification, and the correct object of study is that system.

What the factory did to its failures, and why a provenance depends on it

The imperial factory operated regardless of cost and prioritised quality over commercial return, which is a sentence about accounting rather than about art. It could afford to destroy because it was not selling. Defective pieces, and pieces beyond the number ordered, were smashed on site and buried so that they could not enter the market. The practice ran until 1566, after which seconds were handled differently.

Three things follow for anyone assembling a chain of ownership on a Chenghua object, and they are not obvious.

First, there is no legitimate fifteenth-century commercial provenance. An imperial Chenghua piece did not leave Jingdezhen through a market; it left in an official consignment to the court or it did not leave at all. Any account of a Chenghua imperial porcelain that begins in a private Ming collection has an unexplained step in it, and the honest version of most Chenghua provenances is a blank of four centuries followed by a twentieth-century exhibition catalogue.

Second, the buried rejects are the reason a modern market in excavated Chenghua sherds exists at all, and the reason that market is legally and ethically fraught. Material smashed and buried by the state in 1485 and dug up in the 1980s belongs to the archaeological record, not to the trade. The published, curated route into that material is through the Jingdezhen Institute, the Palace Museum comparison exhibitions and the 1993 Tsui Museum catalogue, and a fragment offered with no excavation reference is a fragment with a problem.

Third, and most usefully for a buyer, the destruction is what makes the excavated corpus a control set. Because the factory buried what it rejected, and because it buried it where and when it made it, a questioned whole vessel can be compared against fragments of known reign, known kiln and known date. That is the closest thing this field has to an instrumental test, and it was created by an administrative decision to destroy value rather than release it.

The corpus, and who is entitled to count it

No complete catalogue of Chenghua imperial porcelain exists and none can, because the production records do not survive and because the census question and the authentication question are the same question. What exists instead are three partial counts that a reader should keep separate.

The first is the count of pieces in private hands. Regina Krahl, cataloguing for Sotheby's, records that only some two dozen porcelains of Chenghua mark and period, of any style or design, are recorded as remaining in private hands anywhere. That figure covers the whole reign and every shape, not one type, and it is the single most useful number on this page for anyone considering a purchase.

The second is the count within a design. The blue and white palace bowls, the classic type of the reign, were made over roughly a decade in the 1480s in about eleven or twelve distinct designs, ten of them floral and one showing fruit. Of the melon design, Krahl documents twelve complete bowls surviving worldwide, none of them apparently in the National Palace Museum in Taipei or the Palace Museum in Beijing. Of the daylily design, Sotheby's records only about three known, two of them in the British Museum. These are counts by connoisseurs of a type, not registers, and they move when a piece surfaces.

The third is the institutional concentration, which is the number that explains the market. The largest surviving body of Chenghua porcelain from the former imperial holdings is in Taipei, and a researcher working from that collection puts more than ninety per cent of the world's Chenghua doucai in the National Palace Museum. Taipei mounted 197 pieces in a Chenghua exhibition in 1977 and 207 in 2003, with a digitised catalogue for the later show. Against a private-hands figure of about two dozen for the entire reign, that is not a market with a thin float. It is a market in which the float is a rounding error.

The chicken cups are the best known sub-corpus and they are counted separately in the Iconotheca record of the Meiyintang cup, where the competing figures of sixteen, seventeen and nineteen and the disputed Beijing examples are set out. What belongs here is only the structural point: Taipei holds eight of them, more than any other institution, and the count of any Chenghua type is a specialist's opinion about authentication rather than an inventory.

The collections that define the field

Four collections decide what a Chenghua object is, and a fifth decides what it costs.

The Percival David collection is the one a European or American researcher will reach first. Sir Percival Victor David Ezekiel David, 1892 to 1964, a Bombay-born financier who read and spoke Chinese, began collecting around 1913 and first visited China in 1923. He acquired roughly forty pieces that had come out of the Forbidden City after 1927, when objects were disposed of by members of the Imperial Household Department, buying through dealers, at auction and from other collectors. He gave the collection to the University of London in 1950; it opened to the public at 53 Gordon Square in Bloomsbury on 10 June 1952 and served as the teaching collection for the study of Chinese art in Britain. Funding failed and Gordon Square closed at the end of 2007. The collection moved on long-term loan to the British Museum and opened in the purpose-built, bilingual Room 95 on 23 April 2009. On 14 November 2024 the trustees announced that they were giving the 1,700 pieces to the museum outright, a gift the museum valued at around one billion pounds and described as the highest value object donation in the history of British museums; the transfer completed on 2 April 2025. The terms require the collection to stay on public display together in perpetuity in a dedicated gallery, with loans to the Shanghai Museum and the Metropolitan Museum of Art contemplated, and the transfer required the consent of the Charity Commission. The doucai chicken cup of the Chenghua reign was one of two objects the museum named in announcing the gift.

The National Palace Museum in Taipei holds the Qing imperial collection, and with it the largest surviving body of Chenghua porcelain from the former palace holdings and, on one researcher's estimate, more than nine tenths of the world's Chenghua doucai. Its chicken cup, accession 005190 in the older numbering and K1B005192 in the current one, is four centimetres high and eight point three across, with two groups of a cock and hen and three chicks separated by China roses and orchids, three blue lines at the mouth and base, and a plain white interior. The museum's own record adds something no Western account carries: the cup can be traced through the Qing Imperial Workshop archives to the Yongzheng reign and comes with a Qianlong-period silk-embroidered box. That is a documented eighteenth-century provenance for an object of the fifteenth, which is more than almost any Chenghua piece in the West can show.

The Palace Museum in Beijing holds the other half of the imperial collection and is the institution that has done most to change the method. Its comparative exhibitions with the Jingdezhen material between 2015 and 2019 moved the centre of Chenghua scholarship from the vitrine to the excavation, and its own Chenghua examples remain contested at the margin, with two chicken cups excluded from the Sotheby's census by Western specialists.

The Meiyintang collection, built from the late 1950s by the brothers Stephen and Gilbert Zuellig and catalogued by Regina Krahl in four volumes between 1994 and 2010, is the private collection against which the modern market has been calibrated; its history and dispersal are set out in the Iconotheca record of its chicken cup. Its palace bowls came to Sotheby's Hong Kong in 2011 with estimates in the tens of millions of Hong Kong dollars, and the catalogue notes Krahl wrote for that sale and for the Pilkington collection sale of 2016 are, for practical purposes, the working literature on Chenghua blue and white in English.

The fifth collection is the ground itself. The Jingdezhen Imperial Kiln Institute, successor to the Jingdezhen Ceramic Archaeology Research Institute, holds more than fifty thousand exhibit-ready items, shows nearly seven hundred restored porcelains in its permanent display alongside thousands of sherd specimens, and since June 2022 has run an Ancient Ceramic Gene Bank with Tsinghua University, the Palace Museum and the Shanghai Institute of Ceramics. More than three thousand specimens have been digitised, including three-dimensional models with their archaeological records and analytical data, yielding close to 1.2 million individual data points. In one reported case seventeen comparative samples across the Ming and Qing were used to identify a dragon-patterned dish as late Qing folk-kiln work. The institute's own phrase for the procedure is a porcelain DNA test. That is where Chenghua authentication is heading, and it is a museum in a small Jiangxi city rather than an auction house that is building it.

The market, with the hammer separated

Chenghua is the most expensive reign in Chinese ceramics and it is traded on a float of roughly two dozen objects. Prices are therefore lumpy, estimates are unreliable, and the published figures need reading with care because almost nobody separates the hammer from the premium.

The clearest recent result is a blue and white palace bowl with a daylily design, 14.9 centimetres across, sold at Sotheby's Hong Kong on 8 October 2019 as lot 1002 in a two-lot sale, Two Ming Porcelain Masterpieces from an Important Collection. The estimate was published as in excess of HKD 50,000,000. The hammer fell at HKD 48,000,000. The price realised was HKD 56,738,000. Three things are worth noticing in that sequence. The bid was below the published estimate and only the buyer's premium carried the total above it, which is exactly the arithmetic that makes premium-inclusive reporting misleading. The winning bid was placed on the telephone by Peter Song of Sotheby's Beijing office for an unnamed client. And the reporting outlet's own headline rounded the result to HKD 56.3 million against a body figure of HKD 56,738,000, an internal inconsistency in a single article about a single lot; the body figure is the one used here.

Two Chenghua cups came up together at Sotheby's Hong Kong on 5 April 2017 in In His Majesty's Palm: Exquisite Porcelain Playthings. Lot 3102, a blue and white cup, was estimated at HKD 15,000,000 to 20,000,000 and sold for HKD 33,850,000. Lot 3101, a doucai stem cup, was estimated at HKD 8,000,000 to 12,000,000 and sold for HKD 13,300,000. Neither figure is broken out into hammer and premium by the source consulted and neither should be assumed to be a hammer price.

Estimates in this field are not a guide. The Meiyintang palace bowl offered at Sotheby's Hong Kong in 2011 as lot 56 carried an estimate of HKD 80,000,000 to 120,000,000; the melon palace bowl from the Pilkington collection offered in 2016 as lot 25 carried HKD 50,000,000 to 70,000,000. Neither lot page displays a result to a visitor who is not signed in, which is a recurring obstacle to research in this market and is recorded here as such rather than filled with a guess.

One recent result is reported in a way this record cannot reconcile. A Chenghua blue and white palace bowl of lily design, known as the Palmer Palace Bowl and offered at Sotheby's Hong Kong on 9 September 2025 from the Ise collection, is described in the single source consulted as carrying an estimate of HKD 40,000,000 to 80,000,000 and selling for HKD 35,145,000. If the selling figure includes the premium, the hammer would have been roughly HKD 28 to 29 million, well below the published low estimate. That is possible where a reserve sits under the estimate, but the source does not say so and no hammer is published. The figure is carried here with the discrepancy flagged rather than smoothed.

The reference point above all of these is the chicken cup sold on 8 April 2014, hammer HKD 250,000,000 and HKD 281,240,000 with the premium, then the highest price ever paid at auction for any Chinese porcelain. That sale has its own Iconotheca record and is not re-described here. What belongs on this page is the ratio it implies: a single doucai cup of the reign made more than four times the best palace bowl result of the following decade, which is a statement about the chicken cup's fame rather than about Chenghua porcelain, and confusing the two is the commonest error made about this market.

One older figure is worth setting beside them. Cleveland's Chenghua palace bowl with lily scrolls, accession 1967.64, passed through the sale of the Eumorfopoulos collections at Sotheby's London on 28 to 31 May 1940 as lot 238, illustrated in the catalogue, and reached Cleveland as a gift of John D. Rockefeller III combined with a purchase from the John L. Severance Fund. The 1940 price is not published in the sources consulted here. An object of that class crossing a London saleroom in the month of Dunkirk is a reminder that the modern price is a modern fact.

A short reign, a short window, and the court that ordered it

The wares this page describes were made for a particular household in a particular decade, and the court in question was not a calm one.

Zhu Jianshen took the throne on 23 February 1464 at sixteen and reigned until his death on 9 September 1487, aged thirty-nine. The dominant figure of the reign was Lady Wan Zhen'er, 1428 to 1487, who had been his nurse when he was a child, was thirty-four when he came to the throne, and became his consort. She bore him a son in 1466 who died within a year, and thereafter, on the standard account, forced other palace women to abort so that no rival heir could be born. Her agents ran illegal mining, foreign trade, pearl gathering and salt dealing, and the emperor rarely restrained her. She died in the same year he did. A secret police organisation, the Western Depot, was created in 1477 alongside the existing Eastern Depot and run by the eunuch Wang Zhi, who had served Lady Wan and who terrorised the Beijing elite until he was transferred in 1483. By the end of the century the palace employed over ten thousand eunuchs, close to the number of regular civil servants.

A persistent claim in the Chinese-language literature attributes the character of Chenghua porcelain to Lady Wan directly, holding that her taste and her demands drove the kilns to the standard they reached. That claim is repeated widely and this record has found no primary document behind it. It is recorded here as an argued position, not as a fact, and a reader should treat any account of Chenghua wares that leans on it as leaning on an attribution of motive rather than on evidence.

What is documented about the emperor's own commitments is religious. He supported Buddhism against the objections of Confucian officials, and in 1473 ordered the building of five pagodas in the Indian style west of Beijing. Ming imperial porcelain travelled as diplomatic gift as well as court supply, and blue and white of the period is held today in the ceramics gallery of the Tibet Museum in Lhasa, which is one visible end of that traffic. The connection between the emperor's Buddhist patronage and the design of particular vessels is not established by anything consulted here and is not asserted.

The reign's ceramic output is also, quietly, an antiquarian programme. Alongside the doucai and the blue and white, the Chenghua kilns produced imitations of Song dynasty wares, and a crackle-glazed conical bowl of Ge type in the Shanghai Museum, filed to the reign, is one of them. The Macau exhibition of excavated Chenghua material organised its 118 pieces into exactly four groups, and Song-style imitations were one of the four. A court factory that spends part of its capacity reproducing the ceramics of a dynasty five hundred years dead is making a claim about lineage, and the same claim is being made two centuries later, in reverse, every time a Kangxi potter writes Chenghua on the base of his best bowl.

The window closed early. Palace bowl production has been placed in a maturing phase from about 1481 and the finest wares confined to the decade before 1485, when the imperial kilns halted production; on one reading the whole classic phase runs to about seven years. The emperor outlived his kilns by two years. That is the fundamental reason Chenghua imperial porcelain is rare, and it has nothing to do with breakage.

Not only doucai, and not only cups

The reign is known in the market for eight-centimetre doucai wine cups, and the market's picture is too narrow.

The blue and white palace bowls are the type most scholars would defend as the reign's highest achievement. They are around fourteen to fifteen centimetres across, decorated on the exterior with a continuous scrolling design and usually left blank inside, and painted in the soft domestic cobalt with a great deal of white porcelain left showing. Nicolas Chow points to a compositional trick in them that he says was never repeated, even in the Qing: a motif is repeated three times and then the scroll simply breaks. Cleveland's example with lily scrolls, 14.5 centimetres across and 6.9 high, is a clean specimen of the type and is reproduced as plate 35a in Harry Garner's Oriental Blue and White of 1954.

The kilns also made monochromes at the top of their range. A Chenghua dish in Cleveland, accession 1989.288, is 21.5 centimetres across, glazed in a plain even yellow with no decoration whatever, and marked with the same six characters. Sherman Lee, Cleveland's director and the author of the standard survey, treated the refinement of monochrome porcelain as the major theme of the whole Chinese ceramic tradition and held that the Ming and Qing gave equal attention to monochrome reds and yellows as to painted wares. What the dish was for is not stated in the museum record and is not guessed at here.

Then there are the imperial forms that carry the state's own symbols. A Chenghua dish in the Shanghai Museum is painted with a five-clawed dragon among lotus scrolls with floral medallions around the cavetto; another Shanghai bowl carries a five-clawed dragon among clouds pursuing a flaming pearl; a third carries a winged mythical beast with a lotus spray. The claw count is not ornament. Five claws were the imperial standard and a miscount was grounds for destruction, as the rejected Chenghua dragon and phoenix dish from the Zhushan excavations demonstrates. When the 1433 order specified 443,500 pieces all with dragon and phoenix designs, it was specifying an emblem of legitimacy in an industrial quantity.

Set against that range, the doucai cups are a small, late, technically extreme corner of the output. They are also the corner the market has decided to pay for, which is why the reign's reputation now rests on the class of object most vulnerable to imitation and least likely to survive.

The reign that would not stay in its century

Chenghua wares were expensive within a hundred years of being made and they have never stopped being expensive, which is unusual and which has consequences for how the material reaches us.

The National Palace Museum's own record of its chicken cup cites the scholar Cheng Zhe for a valuation under the Wanli Emperor, reigning 1572 to 1620, of a hundred thousand cash for a pair. The Qianlong Emperor, reigning 1736 to 1795, wrote an ode in praise of chicken cups. A chicken cup turns up as a plot detail in Cao Xueqin's Hong lou meng, the great eighteenth-century novel, used by accident by a character who should not have had it. By the time the Qing potters were producing their homages, the object being honoured was already a literary reference.

The Qing revival is where the doucai technique reached its second peak and where the confusion begins. A pair of doucai cups in the Art Institute of Chicago, accession 1941.701a-b, made in the Yongzheng reign between 1723 and 1735, shows how far it went: floral medallions delicately outlined in pale cobalt, glazed, fired, then filled with soft translucent enamels that sit so exactly inside the outlines that the museum uses them to explain the term. The Yongzheng Emperor was himself an artist and connoisseur and took a personal role in overseeing the imperial kilns. And that pair carries a six-character mark in regular script inside a double square reading Da Qing Yongzheng nian zhi: its own reign, honestly. The Qing potter who signed Chenghua instead was not obliged to. He chose the earlier name.

Doucai continued at Jingdezhen through the Qing and is still being made there. The Jingdezhen China Ceramics Museum displays a Kangxi doucai dish with lotus under a bilingual label calling it a contrasting-colours plate. Reproductions of Chenghua cups are produced in quantity for the antique markets and sold both as declared replicas and as something else.

The result is a five-hundred-year continuum of objects that all say the same six characters and mean different things by them, and no instrument on the base can tell them apart. The reign mark on a Chenghua porcelain is the most valuable inscription in Chinese ceramics and the least reliable, and both halves of that sentence have to be held at once by anyone who intends to buy, catalogue or publish one.

Chenghua is the most expensive reign in Chinese ceramics and it trades on a float of about two dozen objects in private hands. Prices are lumpy, estimates are unreliable at the top, and almost no published figure separates the hammer from the buyer's premium. Read every number below with that in mind; where the hammer is unpublished this page says so.

The cleanest recent result is the daylily palace bowl of 8 October 2019, and it is instructive precisely because the two figures diverge: the estimate was published as in excess of HKD 50,000,000, the hammer fell at HKD 48,000,000, and the price realised was HKD 56,738,000. The bid was below the estimate; only the premium carried the total above it. Anyone comparing that HKD 56.7 million against a hammer figure from another market is comparing two different things.

The ceiling is set by a single object and should not be read as a level. The Meiyintang chicken cup made HKD 250 million at hammer and HKD 281.24 million with premium on 8 April 2014, more than four times the best palace bowl result of the following decade. That is a statement about the fame of one cup, not about the price of Chenghua porcelain, and Iconotheca's record of that object covers it in full.