A hitherto unknown Jesuit confessionary in Japanese language and script (c. 1595)

It is always exciting to make new discoveries in libraries which hold old books. A printed book in Utrecht University Library, which had not been catalogued since its entry in a university library catalogue from 1754, was rediscovered and turns out to be a unique copy of a Japanese confessionary that sheds new light on the first decade of Jesuit printing in Japan.

The Jesuit Mission Press in Japan

The introduction of a press from Europe in 1590 finally enabled the Jesuits in Japan to print everything necessary for their missionary endeavors. Each of the about 30 different titles from the years to come known to be extant today qualifies as a bibliographical rarity. Few of these so-called kirishitan-ban are preserved in more than three copies, and every second one is known in only a single copy. Many of them were introduced in detail by Ernest Satow in his pioneering study on The Jesuit Mission Press in Japan (1888), but a number of further prints could be located in a series of discoveries extending into the mid-twentieth century. Hardly any new addition to the list of extant kirishitan-ban was possible afterwards, however, so that the only print not already noted in the third edition (1957) of Johannes Laures’ Kirishitan Bunko is the Compendium manualis Navarri (1597) found in 1985. All other findings during the last decades involved either further copies of already known titles or rarely the resurfacing of previously known prints whose exact whereabouts had however become unknown during the course of the last century. Fidesno quiǒ (1611), for instance, thus only became readily available to scholarship after its rediscovery in 2009. Against this backdrop it is unsurprising that the discovery – or rather: identification – of a hitherto unknown Jesuit print from Japan is an event of considerable interest in the global history of the book.

Part from the appendix

The newly identified Utrecht confessionary

The most recent case of such a new identification occurred in the beginning of November 2021, when the author of this article came across the following entry in the Auctarium catalogi bibliothecæ Trajectino-Batavæ (1754; p. 30):

“Compendium Christianæ Doctrinæ, lingua & charactere Japonico, ex Christian. Ravii donatione.”

[A compendium of Christian doctrine in Japanese language and script, from a donation by Christian Raue.]

The attempt to find out more about this most promising entry soon led to a recent publication by Bart Jaski, keeper of manuscripts and curator of printed books (rariora) at Utrecht University Library, stating in a footnote that “Ravius also donated […] V oct 852 rar, a compendium of Christian doctrine printed in Japan, which Reland studied, as he added its title in Latin” (Jaski 2021: 322, n. 5). A short email exchange later, several photos of the book made clear that V oct 852 rar constitutes a confessionary in Japanese language and script, printed by Jesuit missionaries in Japan towards the end of the sixteenth century.

The first printed page here corresponds to the first printed page in another such confessionary printed in 1598, which has been preserved in a single copy at the Biblioteca Casanatense in Rome (CCC M.VIII 41). The words “Salvator mundi” found on that page have commonly been treated as the title of the later confessionary. The next page, however, which is occupied by a title page in the later edition, containing in Latin the simple title of Confessionarium as well as an indication of its place and time of publication, is left empty here. Also, while the main text beginning on the following page reads more or less the same in both editions, it is clearly not the same in terms of, for instance, page layout, printing types and script choice. In fact, a preliminary comparison of the two versions of the confessionary has brought to light about one thousand textual differences – the majority of which is, however, merely orthographic in nature: For many words the 1598 edition introduces Chinese characters to replace kana-only spellings, while the opposite is much rarer. The addition of diacritics is commonly observed as well. Substantial differences between the texts on the other hand are only rarely seen. Leaving grammatical and stylistic revisions aside, one of the most notable amendments is certainly when the Buddhist term nyūmetsu ‘entering Nirvana; death (esp. of Buddha)’ (fol. 33r, 33v) to refer to the death of Jesus Christ is replaced with the neutral and in this context thus arguably more appropriate term shikyo ‘death’ in 1598 (fol. 21v).

Now, the output of the Jesuit Mission Press in Japan is typically divided into an earlier and a later stage, which in the case of prints in Japanese script corresponds to ca. 1591–1593 and 1598–1611 respectively. While several prints dating from the five years between 1593 and 1598 have been preserved, they are all printed in the Roman alphabet throughout. The Utrecht confessionary now may help to fill this gap of five years to some extent.

Printing types

In terms of printing, the Utrecht confessionary clearly still belongs to the earlier type. The printing types appear to be same as those seen in the only two works in Japanese script representing the earlier stage: the Doctrina of 1591 and the volume of 1593 commonly referred to under the title On baptism and preparation for death. Thus, the confessionary comprises besides hiragana only relative few Chinese characters (less than 200 in total), no ligatures apart from ones for taru and tamō, and no examples of kana with added handakuten either (for syllables starting with p-, as e.g. in pāteru nōsuteru ‘Lord’s Prayer’). These three early prints in Japanese script also share another characteristic, namely the lack of a title page and any explicit indication of the respectively year and place of publication. In all likeliness this is not due to text loss, but rather there were no title pages to begin with, as it seems now. The Japanese-language imprimatur following the end of the main text of the confessionary (40v, cf. below) is reminiscent of its counterpart in the 1593 volume On baptism as well.

The above-mentioned features of the earlier Utrecht confessionary are in stark contrast to the later one printed in 1598, which belongs to the later stage both in terms of the printing types used and in that it has a title page giving among others the year and place of publication. The confessionaries in Utrecht and Rome are textually very close to each other, but still they represent different editions of what might be considered to be the same work. In fact, from the annual letters of the Jesuits it had long been known that 1598 was not the first time for a confessionary to have been printed in Japan. As pointed out by Laures (1957: 57–58) in his treatment of the 1598 confessionary, an earlier print is already mentioned in a letter dated 1595. It is thus plausible and likely that the Utrecht confessionary represents such an earlier edition, dating either from 1595 or maybe slightly earlier – and whatever the exact date may be, this copy constitutes the only surviving one of that earlier edition.

Unnumbered page with the words 'Salvator Mundi'

Structure and content

The printed text spans 50 double pages of c. 12.8 × 19 cm (the first unnumbered, the rest counted as 1–49), preceded and followed by a single unprinted leaf. The overall structure of the confessionary is thus as follows:

  • 1 unprinted leaf, containing Christian Raue’s dedication to the library on the recto
  • 1 unnumbered leaf, containing the words “Salvator mundi” and the monogram “IHS” on the recto in print
  • 1–40: main text of the confessionary; fol. 40v contains an imprimatur stating in Japanese that: “This volume is put into print with the permission of the superiors following examination.”
  • 41–45: first appendix, a glossary of Chinese characters, arranged in the order of their appearance in the main text, indicating the page numbers throughout
  • 46–49: second appendix, glossary of Christian terminology as used in the main text (typically Latin or Portuguese loanwords, which are found on virtually every single page, but written in Japanese script throughout)
  • 1 unprinted leaf, containing the handwritten signature “Joaõ” (and again “J”) on the recto

The provenance of the confessionary

The dedication by Christian Raue (Ravius, also Ravis; 1613–1677) is of interest for the provenance of the confessionary and is thus worth quoting in full. It runs as follows:

“In nomine DEi. | Codicem hunc Chinensem | impressum | Florentissimæ Bibliothecæ Publicæ | VLTRAIECTINAE | ex suâ | æternæ sui affectus erga hanc | Academiam memo- | riæ | L. M. Q. D. D. D. [= libenter meritoque dat, dicat, dedicat] | Christianus Rauius Berli- | nas. | 15 Jan. 1644.”

"In the name of God. This printed Chinese book gives, devotes and dedicates with pleasure and as a favour to the most prosperous public library of Utrecht out of his own [library] to the eternal remembrance of his affection towards the university here – Christian Raue of Berlin. 15 January 1644."

Raue's dedication

At the time of this dedication, Raue was professor of Oriental languages in Utrecht. From the many other stations of his career two later ones are of special interest here: In 1649 he became fellow and librarian at Magdalen College, Oxford, whereafter Queen Christina of Sweden appointed him professor of Oriental languages at Uppsala University. Later he became royal librarian in Stockholm. Now, incidentally three of the most substantial Jesuit manuscripts from Japan came into the collections of Magdalen College (Ms. 228) and of Christina (now at Biblioteca Apostolica Vaticana, Reg. lat. 426 and 459) at some time, the latter until 1650 at the latest. Yet, their exact provenance is still uncertain. The fact that Raue was owner of a Jesuit print from Japan no later than 1644 now opens up the possibility that Raue was also in some way involved in the acquisition of these Jesuit manuscripts from Japan – even if this is little more than speculation at this point.

While it is uncertain how Raue got into the possession of the confessionary – and possibly further items from Japan – it seems most plausible to assume that he purchased them in the Netherlands. In fact, a large number of Jesuit prints and some manuscripts can be demonstrated to have been in Dutch collections in the seventeenth century already, many of which were auctioned at some point. Their owners included Joseph Scaliger (1540–1609), Reinier Pauw (1564–1636) and his son Adriaan Pauw (1585–1653), Ernst Brinck (1582–1649), and especially Leiden professor Jacobus Golius (1596–1667) – under whom Raue had studied Arabic in 1637–38. Note also that John Selden (1584–1654), one of the few known English owners of Jesuit prints from Japan at the time, also happens to have been a supporter of Raue. It was also Selden who redeemed the manuscripts and books which Raue had left with a London merchant when leaving for Sweden.

The description by Reland

Coming back to the text of the dedication we may note that the language of the confessionary was misidentified by Raue as being Chinese. Recall that unlike the confessionary of 1598 this edition does not have a title page in Latin indicating its country of origin. The situation is different in the brief explanation following the dedication and written in a different hand, which Jaski (2021: 322, n. 5) has already identified as being that of Adriaan Reland (1676–1718). It reads as follows:

“Compendium doctrinae | Christianae | lingua et charactere Japonico.”

[A compendium of Christian doctrine in Japanese language and script.]

Title desription by Reland

It goes without saying that this corresponds perfectly with the description of the confessionary in the 1754 catalogue. Interestingly, the language of the text is correctly identified as Japanese here for the first time. The Christian contents as such, regardless of the language of the text, may have been easy to tell owing to the first printed page, with the monogram “IHS” in the center being sufficient to clearly identify it as a Jesuit print. However, it is also not impossible, that Reland was able to decipher at least some of the Latin and Portuguese loanwords in the text or – more likely – in the second appendix at the end of the volume as they are written in hiragana throughout.

Specimens of writing from Japan had been known in Europe since the sixteenth century already. The first European print to feature hiragana dates to 1570, even if this was limited to a handful of words. Inventories of hiragana in the form of the so-called iroha poem appeared shortly thereafter, namely in the works of Blaise de Vigenère (1523–1596) and Claude Duret (died 1611), with Andreas Müller (1630–1694) in the late seventeenth century pointing out some of the errors in these earliest sources. Another, somewhat fuller overview of the Japanese syllabaries was provided by Engelbert Kaempfer (1651–1716) – but it was only in 1727 and thus a decade after the death of both its author and Reland himself that it finally appeared in print. Reland was certainly aware of at least some of these publications, but more importantly Reland was himself also in the possession of some prints and manuscripts in and on Japanese, in part likely acquired from employees of the Dutch East India Company. For one, we can tell from his 1718 auction catalogue that he had been in the possession of the later of the two grammars of Japanese written by João Rodriguez (ca. 1561–1633), printed in Macao in 1620: the Arte breve da lingoa Iapoa. On fol. 7r–v it provides an overview of hiragana together with romanizations. (Having been sold at least twice during the eighteenth century, Reland’s copy of the Arte breve eventually found its new home at the Biblioteca Nacional da Ajuda, call no. 50-XI-3.) A manuscript apparently deriving from Reland’s collection (Staatsbibliothek zu Berlin, Ms. or. fol. 429; Jaski 2021: 443, A fol 42) likewise contains an account of both hiragana and katakana. He could thus simply have compared the booklet with the specimens of Japanese (as well as of Chinese) he had at his disposal in order to correctly identify the script and language of the confessionary.

Fragment found in the bookbinding

The fragments found in the binding

At times fragments of the products of the Jesuit Mission Press in Japan also figure as waste paper in the bindings of other books. Incidentally the known cases include the copy of the 1598 confessionary kept at the Biblioteca Casanatense, which is especially notable as it features fragments preserving otherwise unattested prints.

The Utrecht confessionary is of interest in this respect as well. It features two related fragments, one each in the front and back cover. Of these, the fragment found in the back cover coincides entirely with a portion of text on fol. 41r (most of lines 1–8 and 11; lines 9–10 are barely visible due to folding) of the Doctrina in its 1591 edition, as a comparison with the single extant copy kept at the Biblioteca Apostolica Vaticana (Barb.or.153.pt.A) shows. It even contains exactly the same misprint in line 4 (giving adari for expected atari). The text is in dialogue form, with the pupil asking about the Ten Commandments and the teacher replying that there are ten in total, listing them afterwards. The question as well as the reply up to commandments 1, 2 and 4 are visible for the most part in the fragment.

The fragment found in the front cover is of even greater interest and significance. It is likewise closely related to the 1591 edition of the Doctrina, coming from its chapter on the sacraments, specifically about the Sacrament of Penance – a perfect fit thus for the binding of a confessionary. However, while the visible portion of text is very similar to fol. 68r (lines 1–8) of the Doctrina, it is not entirely identical with it either. While the text as such is the same, the first three line breaks occur in different places, and the exact choice of kana is different as well in several instances. This may either mean that there were several variant editions already in 1591, or that we are dealing here with a fragment from another (presumably slightly later) edition of the Doctrina, which however was still mostly identical to the 1591 edition. In either case the fragment gives us a glimpse at an otherwise unknown edition of this key text published by the Jesuit Mission Press in Japan.

Part of the appendix


The author would like to express his deepest gratitude to Dr. Bart Jaski and Drs. Frans Sellies at Utrecht University Library for their prompt and helpful replies to my inquiries in early November 2021 and overall for their most generous support.

Author: Sven Osterkamp

Professor of Japanese language and literature

Ruhr University Bochum