Jikji

Jikji is the short name of a Korean Buddhist book whose full title is Baegun hwasang chorok buljo jikji simche yojeol. Put simply, it is an anthology: a collection of short passages, sayings and verses by Buddhas and famous Chan (in Korean, Seon; in Japanese, Zen) masters, chosen to help students understand how to awaken to their own true nature. The title echoes the Chan teaching “直指人心，見性成佛”, “point directly at the human mind; see its nature and become a Buddha”, a formula recorded, for example, in the Song-dynasty Chan dictionary Zuting shiyuan.

Jikji is famous not mainly for what it says but for how it was made. According to the colophon printed at the end of the book, it was printed in the seventh lunar month of 1377 at Heungdeoksa, a temple outside the town of Cheongju, using cast metal type. No older book printed with movable metal type is currently known to survive. UNESCO recognised this when it added the book to the Memory of the World Register in 2001.

The monks printed the work in two volumes. The first volume is lost. The second survives, missing its first leaf, and belongs to the Bibliothèque nationale de France (BnF) in Paris, which has put a digital copy online. A year after the metal-type printing, in 1378, the text was carved onto wooden blocks and printed again at Chwiamsa temple in Yeoju; copies of that woodblock edition survive in Korea.

Written Chinese characters, which educated Koreans used for books at the time, are not an alphabet of a few dozen letters. Each character stands for a word or part of a word, so a printer needs thousands of different ones. To print with movable type, workers had to make a separate little piece of metal for every character that appeared on a page, and often several copies of common characters.

How the type for Jikji was cast is still debated. The explanation most often given is wax casting: a character is carved on a small piece of beeswax, the wax model is enclosed in a clay mould and melted out by heating, and molten metal is poured into the hollow. In this proposed method, individually carved wax models are consumed during casting, and differences between the models can produce variations in the resulting letterforms. Some bibliographers have argued instead for sand-mould casting. Modern reconstructions show how the work could have been done, but they do not settle which method the printers actually used.

The printer then arranged the pieces of type in a frame, line by line, to match the text. The pieces were held steady, inked with a brush, and a sheet of thin paper was laid on top and rubbed gently from behind. Korean printers did not use a screw press like the one Gutenberg later used in Europe. After enough copies were made, the frame was taken apart and the same type could be reused for the next page.

Jikji shows that this was hard work for a temple workshop. Its lines are sometimes crooked, the ink is darker in some places than others, a few characters are printed upside down, and at least one character that failed to print was later written in by hand. Where the printers ran short of type, they filled gaps with smaller type and a few wooden pieces. These little mistakes are actually useful: they help scholars show that the book was printed from movable pieces and not carved on a single woodblock.

Jikji is a textbook for Buddhist practice. Its compiler, the monk Baegun, selected passages from the lives and sayings of earlier teachers, drawing on Chan histories such as the Jingde chuandeng lu (Record of the Transmission of the Lamp). The first volume began with the seven Buddhas of the past and the Indian and Chinese patriarchs; the surviving second volume continues with later Chinese masters. Many passages are short stories of a student and a teacher meeting, in which the teacher’s surprising answer helps the student suddenly understand.

The central idea is that the Buddha is not somewhere far away. Seon teaching holds that every person’s mind is originally pure, and that by seeing one’s own nature clearly a person discovers that his or her mind is the mind of the Buddha. The book’s title, “pointing directly to the essence of mind”, sums up this message.

People sometimes call the book Jikji simgyeong, “the Jikji sutra”, but strictly speaking the name is inaccurate. A sutra (gyeong) is a scripture believed to record the Buddha’s own teaching, whereas Jikji is a yojeol, a set of selected essential passages compiled by a later monk.

Baegun Gyeonghan (born 1298 or 1299, died 1374) was a Seon monk of the late Goryeo dynasty (918–1392). In 1351 he went to China, then ruled by the Mongol Yuan dynasty, where he met the Linji (Korean: Imje) Chan master Shiwu Qinggong (石屋淸珙) and the Indian monk Zhikong; he returned to Goryeo in 1352. Shiwu gave him a one-fascicle collection of Chan sayings, and Baegun later expanded it into the two-volume anthology we call Jikji, drawing on Chan histories such as the Jingde chuandeng lu. He completed the compilation in 1372; his postscript gives his age then as seventy-five. He died at Chwiamsa temple in Yeoju in 1374, three years before Jikji was printed.

Buddhism was the leading religion of Goryeo and enjoyed royal patronage. Temples were major centres of learning, and they had long experience with printing. Goryeo had carved the entire Buddhist canon on woodblocks twice, most famously in the thirteenth century, when the Tripitaka Koreana was produced during the Mongol invasions. Printing Buddhist texts was also considered an act of merit, a good deed that could benefit the sponsor and others.

The colophon at the end of Jikji records that the printing at Heungdeoksa was carried out with the help of Baegun’s disciples, the monks Seokchan and Daljam, and paid for by a Buddhist nun named Myodeok. The involvement of a nun as patron is a reminder that women took an active part in Goryeo religious life and in sponsoring books; Myodeok also supported the woodblock edition of the following year.

Jikji was printed late in the Goryeo dynasty, under King U. Metal-type printing had already been known in Korea for more than a century, so the Heungdeoksa colophon does not record a new invention; it documents the use of this technology in a local Buddhist publishing project, away from the capital.

Printing began in East Asia with woodblocks, where a whole page is carved in reverse on a single board. Korea preserves one of the oldest known woodblock prints, a Buddhist dharani sutra found inside the Seokgatap pagoda at Bulguksa and dated to before 751. Movable type came later. The Chinese scholar Shen Kuo described how a craftsman named Bi Sheng made movable type from baked clay in the 1040s, and wooden movable type was later used as well.

Goryeo appears to have taken the next step by casting type in metal. The scholar Yi Gyu-bo recorded that a 50-volume state ritual code, the Sangjeong yemun (also called Sangjeong gogeum yemun), was printed in 28 copies with cast type after the court had fled to Ganghwa Island in 1232; no copy survives. Another text, a Chan song of enlightenment with verse commentary (Nammyeong cheonhwasang songjeungdoga), survives in a woodblock edition of 1239 whose afterword says it was recut from a cast-type edition. These records show that metal type existed in Korea more than a century before Jikji, even though Jikji is the oldest surviving book actually printed with it.

After the founding of Joseon, the state took up metal type on a large scale. The royal foundry cast a font in 1403 (the Gyemija type) and a much improved one in 1434 (the Gabinja type) under King Sejong. These official fonts were more uniform and better finished than the temple type used for Jikji, and they were used to print government, Confucian, historical and other books.

In Europe, Johannes Gutenberg developed printing with cast metal type in Mainz in the mid-fifteenth century; his 42-line Bible was completed around 1454–1455, roughly 78 years after Jikji. Gutenberg’s system combined a type mould for rapid casting, a metal alloy, oil-based ink and a screw press, and it suited an alphabet of a few dozen letters. It spread rapidly in Europe. Korean publishers, for their part, continued to use both metal type and woodblocks.

Almost nothing is known about the metal-printed Jikji for the five centuries after 1377. It reappears at the end of the nineteenth century in the collection of Victor Collin de Plancy, the first French diplomatic representative in Korea, who served in Seoul from 1888 to 1891 and again from 1896 to 1906. Collin de Plancy collected Korean books and ceramics; he is generally thought to have bought Jikji during his second posting, but how and where he acquired it is not documented. By the time he owned it, the first volume was already gone and the second had lost its first leaf. A French note on the cover identifies it as the oldest Korean book printed with cast type and gives the date 1377. According to the BnF, the book was shown to the public in the Korean pavilion of the 1900 Universal Exhibition in Paris.

Maurice Courant (1865–1935), who worked under Collin de Plancy at the French legation, catalogued Korean books in his Bibliographie coréenne, and in a supplement published in 1901 he included Jikji and recorded its colophon: printed with cast type at Heungdeoksa in 1377.

In 1911 Collin de Plancy sold much of his collection at an auction at the Hôtel Drouot in Paris. Jikji was bought for 180 francs by Henri Vever, a jeweller and collector. Vever died in 1942, and under his will the book was given to the Bibliothèque nationale in 1950; it physically entered the collections in 1952 and is kept as Coréen 109 in the Manuscripts Department.

For decades the book sat quietly in the library. Its importance became widely known through Park Byeong-seon (1923–2011), a Korean historian who worked at the BnF from 1967. She studied the volume and its colophon, and in 1972 it was displayed at the library’s exhibition for UNESCO’s International Book Year, where it was recognised as the oldest surviving book printed with movable metal type.

One question remained: where exactly was Heungdeoksa? The colophon said only that the temple lay outside Cheongju, and the temple had long disappeared. In 1984 land development in Uncheon-dong, Cheongju, uncovered an unnamed temple site, and in 1985 archaeologists from Cheongju University excavated it. A bronze gong fragment inscribed with the words “Heungdeoksa of Seowon-bu” and a bronze bowl lid inscribed “Heungdeoksa” confirmed the location.

The site was designated a national Historic Site in 1986, and the Cheongju Early Printing Museum opened next to it in March 1992. Since 2000 the museum has made Jikji the central theme of its permanent exhibition. Cheongju has built its identity as a “city of Jikji”, hosting festivals and supporting projects to reconstruct the lost type.

In 1975 Park Byeong-seon also traced, in a BnF storeroom at Versailles, the royal protocol books (uigwe) that French troops had taken from the Oegyujanggak archive on Ganghwa Island in 1866. Her work helped start the long campaign that brought 297 of those volumes back to Korea in 2011.

On 4 September 2001 the second volume of Jikji was inscribed on UNESCO’s Memory of the World Register, which lists documentary heritage of world significance; the recommendation came from the programme’s International Advisory Committee, which met in Cheongju in June that year. In 2004 UNESCO approved the UNESCO/Jikji Memory of the World Prize, funded by the Republic of Korea through the city of Cheongju. First awarded in 2005, it is intended as a biennial award for individuals or institutions that help preserve documentary heritage and make it accessible.

In Korea, Jikji is presented as a symbol of the country’s early achievements in printing, alongside the Tripitaka Koreana woodblocks and the fifteenth-century Korean alphabet. Facsimiles have been published in Korea since 1973, and the Cheongju Early Printing Museum and its multilingual Jikji Global website present the book with images based on the BnF’s digital copy.

From 2011 to 2016 the Cheongju Early Printing Museum ran a project to restore the type and text of both volumes, using the 1378 woodblock edition as a guide for the lost first volume, and a master craftsman has recast the whole two-volume text by wax casting. Such reconstructions are experiments: they test how the monks might have worked, but they cannot tell us for certain what the lost first volume of the metal-type edition looked like.

The book also plays a part in public debates about global history. It is often used to challenge the idea that printing with movable type was a purely European invention, and to encourage a more connected, many-centred picture of how technologies developed.

Jikji is at the centre of a long-running debate. Many Koreans argue that it is an essential part of Korean heritage and should return to the country where it was made. The BnF, which holds it, treats it as one of the most precious items in its collections, conserves it with minimal handling, presents it as a treasure of the world of printing and makes it available online to everyone. Its passage through a Paris auction in 1911 and a bequest to the library is documented, but those later transactions do not show how Collin de Plancy first obtained the book in Korea.

The case differs from that of the Oegyujanggak uigwe. Those books were seized by French forces during a military expedition in 1866, and after negotiations that began in the 1990s they were returned in 2011 under a renewable loan arrangement. Jikji was not taken in war, and according to the Korean studies scholar Jamie Jungmin Yoo, Korea’s heritage authorities have acknowledged that its return cannot be legally compelled; Yoo describes the library’s title as defensible in law but morally contested. Korean policy has therefore tended to focus on research cooperation, exhibition and digital access; in 2023 the BnF and Korea’s Cultural Heritage Administration signed a three-year cooperation agreement, and the BnF displayed the original in its exhibition on Gutenberg’s Europe.

The debate raises broader questions that museums worldwide now face: how to weigh legal title against the cultural ties of source communities, what it means for an object to be “world heritage” while it is held in one country, and whether digital access can reduce the need for physical return. There is no consensus answer, and readers will find passionate and reasonable arguments on both sides.

Specialists are careful to call Jikji the oldest extant book printed with movable metal type, not the first. Written evidence shows that Goryeo printers used cast type by the 1230s, during the Mongol wars, and the Encyclopedia of Korean Culture argues that type was already in use at the capital Gaegyeong before the court moved to Ganghwa in 1232. The Sangjeong yemun printing is known only through Yi Gyu-bo’s record, and the 1239 songjeungdoga woodblock edition only tells us that a metal-type edition once existed. Jikji’s value lies in being physical proof, with a precise date and place in its own colophon.

Claims of earlier surviving metal type have been controversial. In 2010 a set of metal characters called the jeungdoga-ja was presented as Goryeo type that might predate Jikji. In April 2017 the heritage committee of the Cultural Heritage Administration voted not to designate 101 of them as a Treasure: comparisons of script and typesetting did not show that they had printed the songjeungdoga, and their provenance was unclear, although no proof of forgery was found and carbon dates from ink traces fell in the Goryeo period. In October 2025 the head of the Korea Heritage Service told the National Assembly that the case would be examined again, so the question is not closed. Note that this is a question about pieces of type: Jikji remains the oldest surviving book generally recognised as printed with movable metal type, and the type used to print it does not survive.

Another question is whether knowledge of East Asian metal-type printing reached Gutenberg. Some writers have suggested that it travelled across Eurasia during the Mongol era, but no documentary evidence of such a transfer has been found, and the technical differences between the Korean and European methods (casting, ink, press, script) are substantial. The question remains unresolved. The international research project From Jikji to Gutenberg, based at the University of Utah, is studying the technical evidence while describing the two developments as apparently independent.

Finally, metal type and woodblocks coexisted in Korean publishing, and their uses differed: type could be taken apart and reused for other texts, while stored woodblocks allowed later impressions of the same text. Explaining why Korean and European print cultures developed differently requires evidence about particular institutions, readers and book markets, and comparisons built on a single model of a “printing revolution” can mislead.

The key primary evidence is the colophon (gangi) on the last leaf, which states that the book was printed with cast type (juja) at Heungdeoksa outside Cheongju in the seventh month of the seventh year of Xuanguang (宣光七年), that is, the third year of King U’s reign (1377), and names the monks and the patron involved. Because the colophon is part of the same printing as the rest of the surviving volume, its date is generally accepted as reliable.

Physical features support the claim of movable-type printing. The same character never appears in an identical form twice on one page, yet the same pieces of type reappear on other pages, as one would expect when a limited stock of individually cast type is reused. Lines are slightly askew, ink density varies, some characters are inverted or only partly printed, and there are many small specks around the characters. A woodblock edition followed in 1378. One explanation offered is that stored woodblocks made later impressions easy, whereas a page of movable type had to be set again once its type had been taken apart; this is an interpretation of the circumstances, not a claim that the metal type could not be reused. The 1378 edition includes a preface by the scholar Yi Saek, and its colophon says that the disciple Beomnin and others oversaw it at Chwiamsa.

The surviving metal-type volume has 39 leaves, of which the first is missing; it measures 24.6 by 17.0 cm, has eleven lines to a page with 18 to 20 characters a line, and is sewn with five stitching holes in the Korean manner. The original first leaf is still missing; the present cover and a blank leaf between the cover and the text were added in a later rebinding, probably centuries after printing. Conservation, imaging and, since 2021, physico-chemical analysis by the BnF with Korean and French partners allow scholars to study the shapes of individual characters, the paper and the ink. The casting method is still debated: the UNESCO nomination file and most standard accounts attribute the type to wax casting, while the bibliographer Kim Sung-soo has argued from the uneven heights of the type that sand-mould casting is more likely. Reconstructions such as Cheongju’s should therefore be seen as informed hypotheses.