Admonitions of the Instructress to the Court Ladies

A painted silk handscroll illustrating a 3rd-century Confucian text of moral instruction for palace women, traditionally ascribed to the 4th-century master Gu Kaizhi but almost certainly a Tang-dynasty copy of a lost original. It left the Qing imperial collection in 1900, during the international occupation of Beijing that followed the Boxer Rebellion, in the possession of a British Indian Army officer who sold it to the British Museum in 1903 for 25 pounds.

Conservators then cut the silk into flat panels and remounted it in a Japanese album format, a decision now treated by British Museum conservators themselves, working with Chinese specialists in a five-year study completed in 2014, as a documented conservation error. The object is displayed six weeks a year and kept in total darkness the rest of the time. It sits at the centre of China's ongoing case for the return of art removed from the country during the century of foreign military intervention, a case the British Museum's director has publicly declined to act on.

What survives, physically

The object the British Museum holds under number 1903,0408,0.1 is a painted silk handscroll, now cut down and mounted flat on panels rather than kept in roll form. Nine painted scenes survive of an original twelve; the missing first three are presumed lost before the scroll left China, since a 12th-century paper copy made from it already appears to reconstruct rather than faithfully transmit them. Each surviving scene pairs a painted illustration with a block of calligraphy quoting the didactic text it illustrates.

Attached to the painting, and inseparable from any honest account of the object, is nearly a millennium of accretion: seals, colophons, a title inscription and an added painting, contributed by a sequence of documented owners from the Song dynasty to the Qing. The physical object today is therefore not simply a 4th-to-8th-century painting; it is that painting plus the record, in ink and seal-paste, of everyone who is known to have possessed it since at least the 11th century.

The text it illustrates, and what it was for

The paintings illustrate the Nushi zhen (Admonitions of the Instructress to the Court Ladies), an 80-line didactic poem composed around AD 292 by the Jin-dynasty official and poet Zhang Hua (232-300). Zhang wrote it, by the traditional account, to rebuke the empress consort Jia Nanfeng (257-300), whose conduct scandalised the Western Jin court, and more broadly to set out the virtues expected of women attached to the imperial household: humility, self-restraint, discretion, and the subordination of personal desire to duty.

The painted scenes function as a visual conduct manual rather than as narrative for its own sake. One shows Lady Feng stepping between a charging bear and the Emperor Yuan of Han, an exemplary act of self-sacrificing courage; another shows a court lady seated at her toilette, a scene traditionally read as a warning that outward beauty requires inward self-examination to match it; a third shows the Emperor waving away a court lady approaching his sedan chair, illustrating Lady Ban Jieyu's historical refusal to ride beside Emperor Cheng of Han in order to avoid impropriety and the appearance of undue influence. Each scene is a specific, textually anchored moral example aimed at an audience of imperial consorts and attendants, not a generalised depiction of court life.

This is an ordinary, well-documented didactic function, not a hidden one: the object exists because a Confucian moral text needed a visual teaching aid for palace women, and every scene traces to a passage of Zhang Hua's poem. No esoteric or concealed programme is argued for it in the literature consulted for this record; the moral content is the entire content, stated plainly by the text panels themselves.

Attribution: an argument, not a fact

No painting securely documented as Gu Kaizhi's own hand survives anywhere. The corpus attributed to him, including this scroll, the Nymph of the Luo River (surviving only in five later Song-Ming copies, a separate composition and a separate physical object in every surviving instance) and Wise and Benevolent Women (a 13th-century Song copy), is a reconstruction built by connoisseurship, catalogue tradition and, in the modern period, technical analysis. The British Museum's own online catalogue dates this object to the 5th-7th century, declining to commit to a Tang date; this record follows that caution rather than repeating the popular shorthand of 'the Tang copy' as settled fact.

Among named specialists, Shane McCausland, who organised the British Museum's 2001 international colloquium on the scroll and edited the resulting volume, argues it as a Tang-dynasty tracing copy on the basis of integrated textual and pictorial features that diverge from what is independently known of Jin-period painting. Kohara Hironobu has argued that the calligraphy accompanying the scenes was not added until after 1075, in a hand deliberately archaised to look older than it is, which would separate the date of the text from the date of the painting and push at least part of the object later than the Tang. Wen Fong has read the same calligraphy differently, finding it consistent with the style of the monk-calligrapher Zhiyong, active in the late 6th to early 7th century, which would argue for an earlier date. Chen Pao-chen, of National Taiwan University, has made the case in print for treating the painting as carrying genuine early, pre-Tang pictorial content rather than being a straightforward later tracing. Yang Xin, of Peking University, has proposed a different origin altogether: an original work commissioned at the Northern Wei court of Emperor Xiaowen (r. 471-499), rather than a copy of a Gu Kaizhi original at all.

None of these positions is settled, and this record does not arbitrate between them. What is documented, not argued, is that the calligraphic and pictorial evidence is read differently by named scholars working from the same object, and that the attribution to Gu Kaizhi himself is, at best, an attribution to a lost original he may have composed, transmitted through an unknown number of intervening copies before this one was made.

Through the Song, Ming and Qing collections

The earliest secure sighting is in the painter-connoisseur Mi Fu's Hua shi (History of Painting), written around 1103, which places the scroll with the court eunuch Liu Youfang. By 1120 it is catalogued as a Gu Kaizhi in the Xuanhe huapu, the painting catalogue of Emperor Huizong's imperial collection, meaning it had entered Song imperial hands by then.

When Kaifeng fell to the Jurchen Jin in 1127 and Huizong's collection scattered, the scroll's custody becomes uncertain again; a seal associated with the Jin Emperor Zhangzong appears on one section, but scholars are not confident it was applied at the time rather than added from elsewhere later. The chain resumes with named Ming-dynasty owners: the official Wang Ao, the powerful and later disgraced minister Yan Song (whose confiscated property reverted to the Ming court after his 1562 fall), the official Gu Congyi, and finally the collector Xiang Yuanbian, who left roughly 50 of his own seals on the object, among the heaviest documented interventions by any single owner before the Qianlong Emperor.

After Xiang Yuanbian's death in 1590, the trail goes dark again until the scroll surfaces in the 18th century in the collection of the Tianjin merchant An Qi, from whom Qianlong acquired it.

Qianlong's remaking of the object

The Qianlong Emperor (r. 1735-1796) did not simply own the scroll; he physically remade it. In and after 1746 he had it remounted, added a title inscription in his own hand, painted an orchid onto the mount, and over the course of his sixty-year reign impressed 37 personal seals onto it, an act of ownership so dense it amounts to a form of authorship in its own right. He commissioned the court painter Zou Yigui (1686-1772) to add a closing composition, Pine, Bamboo, Rock and Spring, as the final section of the mounted scroll. He kept it, alongside three other favoured works he grouped as the 'Four Beauties,' in his private Pavilion of Tranquil Ease.

This is worth stating plainly because Western commentary on the object sometimes treats the Qianlong material as clutter obscuring an earlier, purer painting. Qianlong's seals and colophon are themselves primary evidence: of imperial taste, of the object's status at the summit of the Qing collection, and of exactly the kind of intensive, physically transformative engagement with a scroll that a Chinese collector considered a mark of honour rather than damage.

1900: removal during the occupation of Beijing

By the 1890s the scroll was kept, under Empress Dowager Cixi's de facto rule, at the Summer Palace complex west of Beijing. When the Boxer Rebellion broke out and the international legations in Beijing came under siege, an Eight-Nation Alliance force, comprising troops from Britain, France, Japan, the United States, Russia, Germany, Austria-Hungary and Italy, occupied the city from mid-August 1900. Between mid-August and October 1900 the Summer Palace was burned and systematically looted by soldiers of the occupying powers, an action distinct from, though following on from, the earlier 1860 burning of the same palace complex during the Second Opium War.

Captain Clarence Arnold Keatinge Johnson (1870-1937), an officer with British Indian cavalry (sources describe him variously as of the 1st Bengal Lancers or 1st Bengal Cavalry) stationed in the Beijing area during the occupation, acquired the scroll in this period. Johnson's descendants later told the British Museum that he received it as a personal reward from a Chinese noblewoman whose family he had escorted to safety. No document corroborates this account, and it should be read as an undocumented family tradition rather than an established fact. What is documented is the setting: an occupied capital, a burning and looted palace, and a foreign officer leaving China in possession of one of the most treasured objects in the recently displaced Qing imperial collection. Whatever the specific transaction, Johnson's acquisition took place inside, and cannot be meaningfully separated from, the wider looting of the Summer Palace by the forces of the occupying alliance.

1903: the sale to the British Museum

Johnson returned to Britain and in 1902 brought the scroll to the British Museum, seeking only to have the jade toggle fitted to its wrapper appraised; by the museum's own account he had no idea what the painting itself was. Sidney Colvin, Keeper of the Department of Prints and Drawings, and his assistant Laurence Binyon recognised the object's significance. The Trustees purchased it from Johnson in 1903 for 25 pounds sterling.

This figure is widely repeated and this record could not find it corrected or superseded in any museum or scholarly source consulted; it is treated here as confirmed. Twenty-five pounds in 1903 is approximately 3,900 to 4,000 pounds in 2026 terms on the UK composite consumer price index, a domestic-servant's or junior clerk's annual wage of the period, paid for what is now one of the most closely studied paintings in the British Museum's entire collection. Colvin's own contemporary condition note is blunt: the silk was, he wrote, 'much worn and rotted, and bears many traces of ancient repair.' Later British Museum conservation research attributes further damage to rough handling during the object's transit from China to London, on top of the wear Colvin had already recorded on arrival.

The 1910s remounting, and why it is now regarded as an error

Around 1912-1915, the British Museum had the scroll dismantled from its Chinese roller-scroll form and remounted flat, using a Japanese byobu (folding-screen) technique, carried out with the involvement of Japanese specialists including the block-cutter Sugizaki Hideaki and the printer Urushibara Mokuchu (who worked concurrently on a woodblock facsimile of the scroll, discussed below). The painting sections were separated onto one mount and the accompanying colophons onto another; Zou Yigui's closing painting was detached and mounted as an independent album leaf.

The intention, by the standards of the time, was conservation: to stabilise a fragile, damaged silk object using techniques the museum's own staff and outside advisers judged sound. The consequence, judged by the standard a Chinese scroll-mounter would have applied, was close to the opposite. Traditional Chinese scroll-mounting is itself an intensive, retouching craft, expected to be periodically renewed by a trained specialist working with the object's original structure and materials in mind; flattening a silk scroll onto rigid panels in the Japanese album manner is a different physical treatment built for a different kind of object, and it exposed the painting to a flat, evenly distributed light and handling regime the rolled Chinese format was never meant to receive. The 2015 British Museum Technical Research Bulletin study of the scroll's condition and treatment history from 1903 to 2014 documents this sequence of interventions in detail as part of tracing how the object arrived at its 2013-2014 condition.

By the time the British Museum began the joint study that led to the object's 2014 redisplay, in-house conservation staff, including Qiu Jinxian, the museum's first Chinese-trained senior conservator of East Asian painting (in post since 1988), were explicit that the Japanese remounting technique used in the 1910s was not the appropriate treatment for a Chinese painting on silk of this kind and date, not least because Japanese scroll-mounting practice, though itself derived from Tang-dynasty Chinese methods, had diverged from the methods Chinese conservators had continued to develop through the intervening dynasties. When Qiu and her colleagues undertook fresh conservation work on the scroll in 2013, they judged it too fragile to remount as a rolled scroll at all, and instead refined the flat, panel-mounted presentation the 1910s remounting had already imposed, rather than attempting to reverse it.

The 2013-2014 joint study and redisplay

Following roughly a decade of preliminary research, the British Museum convened an expert meeting in July 2013 bringing together East Asian painting conservators, Chinese scroll-mounting specialists and art historians to assess the object's condition and agree a conservation and display strategy. The project, completed in June 2014 with the scroll's redisplay in a new purpose-built case, is documented in the museum's 2015 Technical Research Bulletin article on the object's condition, treatment and housing from 1903 to 2014.

The resulting case is a purpose-built, double-sided steel structure roughly 4 metres long and weighing about 900 kilograms, with a hydraulic lifting mechanism and drawer access for handling the mounted panels. Rather than mechanical air conditioning, the team chose passive humidity buffering using ProSorb conditioning cassettes, judged less prone to the fluctuations and maintenance intrusions of an active system, monitored with Hanwell radiotelemetry sensors. Display materials were vetted with Oddy testing to screen for pollutants that could react with the silk or pigments over time.

The most consequential decision was on light. The scroll is now shown for a strict, fixed six-week window each year against an annual exposure budget of 30 kilolux-hours, lit in the gallery by LED units built into the case's own pedestals, with manually operated rather than motorised blinds chosen deliberately for their mechanical reliability. Outside its display weeks, the case is kept in total blackout. This is a conservation regime built around the acknowledged, and by 2013 well understood, vulnerability of light-degraded silk and pigment, not a display schedule chosen for dramatic effect.

Reproduction: the 1912-1913 woodblock facsimile

Concurrent with, and connected to, the 1910s remounting, the British Museum Trustees in 1911 commissioned a hand-coloured woodblock facsimile of the entire scroll, cut by Sugizaki Hideaki and printed by Urushibara Mokuchu, with accompanying text by Laurence Binyon. Produced in London in 1912-1913 from 34 separate blocks, the edition ran to 100 copies, sold at 7 guineas (7 pounds 7 shillings) each; the British Museum holds a copy under number 1913,0708,0.1. Copies were presented to leading sinologists of the day, including Edouard Chavannes and Raphael Petrucci, functioning as an early, deliberate act of scholarly circulation: the facsimile put a high-fidelity surrogate of the scroll into other hands and other libraries well before photographic reproduction or digitisation could do the same job.

The restitution question, and where it stands

The Admonitions Scroll sits inside a live, unresolved argument over Chinese cultural property removed from the country during roughly a century of foreign military intervention, from the Opium Wars through the 1900 occupation of Beijing to the Second World War. In August 2023, the Chinese state-run Global Times published a formal editorial call for the British Museum to return, without payment, Chinese relics it holds that were acquired through what it characterised as improper channels, part of a wider wave of restitution demands following revelations that a British Museum curator had stolen and sold items from the museum's own storerooms. UNESCO figures cited in that coverage put the number of Chinese objects in the British Museum's collection at roughly 23,000, with about 2,000 on long-term display; the Admonitions Scroll is regularly named, alongside Chinese ceramics and bronzes, as one of the collection's most prominent Chinese holdings with a documented colonial-era removal.

The British Museum has not issued a response naming the Admonitions Scroll specifically among the sources checked for this record. Institutionally, its position has been set out in general terms: in May 2025 the museum's director, Nicholas Cullinan, publicly ruled out any change to the museum's policy against permanently deaccessioning objects, which in the case of the British Museum would in any case require an Act of Parliament to alter the governing statute. Speaking about restitution demands broadly (in remarks focused on the Parthenon Marbles and the Benin Bronzes, not on this object specifically), Cullinan said he was 'more comfortable with loaning items' and proposed partnership and loan arrangements, of the kind negotiated with Greece, as an alternative to permanent return, describing the museum's aim as 'connecting countries rather than putting up barriers.' No export licence, restitution claim, or legal proceeding specific to the Admonitions Scroll has been identified in the sources consulted for this record.

What is happening to it now

The scroll remains on the British Museum's register as object 1903,0408,0.1, held in Room 91a. Under the conservation regime adopted in 2014, it is shown for one fixed six-week window a year and kept in total darkness for the rest; the museum's published schedule for 2026 sets its display from 13 July to 24 August. Outside that window, study access is by appointment through the museum's Asia department rather than public display.