Intermixture of Mongolian and Oirat in 17th Century Manuscripts

Abstract

The Oirat alphabet Clear Script was created in 1648, but few specimens of Oirat penmanship from the 17th c. have survived, and very little is known about the early history of its development and adaptation. Some information on the subject can be gathered from manuscripts that were discovered at the site of the ruined Dzungar monastery Ablaikit. Among these manuscripts are multiple fragments of the Buddhist canonical collection Kanjur and two folios from ritual texts composed by the Fourth Panchen Lama. These texts are written in Mongolian, but the scribes used graphemes from Clear Script, elements of the vocabulary and grammar of Written Oirat. Fragments of another manuscript found in Ablaikit, a small birch-bark copy of the Heart Sutra, contain a text written in Oirat with interpolations from Mongolian. Combined, the observations based on the study of these sources show that the transition from Mongolian to Clear Script was gradual, and for a period of time in the second half of the seventeenth century both writing systems were used by the Oirats.

Full Text

Clear Script (Oir. todo bičiq) was created in 1648 by the Oirat Buddhist teacher and scholar Zaya paṇḍita Nam mkha'i rGya mtsho (1599–1662) based on the Mongolian script (Mong. qudum bičig or mongγol bičig). In this new alphabet, the ambiguity of several Mongolian graphemes was eliminated, and new letters were introduced in order to clarify the pronunciation and bring the written text closer to the spoken language.¹ Clear Script was conceived as a writing system for all the Mongols, but eventually was adopted only by the Oirats (it is still in use today among the Oirat population of Xinjiang, China).²

Zaya paṇḍita and his disciples translated numerous Buddhist texts into Oirat,³ which means that todo bičiq was actively used from the mid-seventeenth century. However, no written sources have come down to us from the first decades of its existence, and most of the surviving manuscripts and xylographs date back to the eighteenth century or later (the earliest dated manuscript in Clear Script is the letter written by Galdan Bošoqtu Qān (1644–1697) to the Czar of Russia in 1691⁴). In the absence of early sources, it is difficult to estimate to which scale this ‘young’ system of writing was used in the seventeenth century, and how long it took for it to replace the Mongolian script in the Oirat cultural milieu.

Some data on the subject can be gathered from the manuscripts that were discovered at different times at the site of the ruined Dzungar monastery known as Ablaikit (Oir. abalaiyin keyid or abalayin süme ʻAblai’s monasteryʼ; its proper name remains unknown). This fortified monastery was built by the order of the Khoshut leader Ablai tayiǰi (fl. 1638–1671) in 1654–57, at the south-western foothills of the Altai mountains (modern East Kazakhstan Region). In 1657, its temple was consecrated by Zaya paṇḍita himself, who was invited by his brother Ablai tayiǰi. Ablaikit was last mentioned in a historical source in connection with the events of 1661 when the fortress suffered a siege.⁵ Its subsequent fate remains unclear. Presumably, the monastery fell into decay after the demise of its founder who died around 1672.⁶ Starting from the 1720s, travellers and explorers visited the deserted monastery while its walls were still intact and found the remnants of a large library. In 1734, around 1,500 fragments of manuscripts in the Tibetan and Mongolian languages were collected at the site by the Second Kamchatka (or the Great Northern) Expedition and delivered to St. Petersburg.⁷ The manuscripts were stored at the Academy of Sciences, but were not described or studied until 2014 when they were identified as fragments of three sets of the Buddhist canon Kanjur, one in Tibetan (204 folios) and two in the Mongolian language (1,245 folios). These materials (1,449 folios in total) are kept at the Institute of Oriental Manuscripts, Russian Academy of Sciences. Several dozens of fragments of the same manuscripts are preserved in a number of European collections.⁸

Fragments of the Mongolian Kanjurs found in Ablaikit come from two different manuscripts which will be referred to as Ms 1 and Ms 2.⁹ Ms 1 is of Mongolian (presumably, Southern Mongolian) origin, while Ms 2 was copied by Oirat scribes, most probably — in Ablaikit itself. The surviving folios of Ms 2 are of thick layered paper, 71×25 cm in size; the text (30–36 lines per page) is written with a calamus in black ink, on some folios red ink is used to draw the frame or highlight certain words. The Dzungar origin of the manuscript is obvious from several features: the prevailing style of frames (most folios of Ms. 2 have an Oirat-style frame, i.e. double lines that mark the right and left margins, with no framing on the upper and lower margins; see Pl. 2), the ductus (the graphic elements on the right and left side of the axis have an incline¹⁰), and the sporadic use of graphic elements of Clear Script, Oirat orthography and grammar (will be discussed below). Several details indicate that Ms 2 was copied from Ms 1. The two manuscripts have identical marginal markers (indicating the section of the Kanjur, and in some instances — the name of a particular text) even in those cases when the volumes are marked in an unusual or ‘random’ way.¹¹ The folios were delivered to St. Petersburg mixed together, which means that they had probably been mixed before they were found in Ablaikit where they had been stored within the same space at the temple library. The number of the surviving folios of Ms 1 (794 ff.) is almost twice as big as that of Ms 2 (479 ff.). All these observations suggest that Ms 2 was copied from Ms 1, and the copying process was never finished, which would mean that the work was carried out in the place where the folios were found, that is, in Ablaikit itself. Possibly, the copying process stopped when the monastery was abandoned. Based on this hypothesis, Ms 2 can be provisionally dated to the period from the late 1650s to the 1670s. It is remarkable that, although Clear Script was known to some of the scribes, the Kanjur was copied in Mongolian. Translating the Kanjur into one’s language is a potent move in terms of cultural empowerment, but, clearly, the goal of creating a copy of the Buddhist canon in todo bičiq was not pursued in this case. Possibly, there were no means to organize such a grand project, as creating an Oirat Kanjur would require a board of skilled editors to coordinate the work. Alternatively, it could indicate that the tradition of writing in Clear Script had not been developed enough by the time Ms 2 was being copied, or that the new alphabet was not viewed as a symbol of cultural identity at that stage.

In Ms 2, elements of Clear Script and the Oirat language, from single graphemes to several words in a sentence, occur irregularly, showing that some of the scribes were not simply acquainted with todo bičiq, but had developed a habit of writing in Oirat. These elements are found in those fragments of text that are written in the typical Oirat ductus characterized by oblique (rather than horizontal) transverse lines (this ductus prevails in Ms 2). The most often used grapheme is i: instead of the Mongolian ‘stick’ or ‘long tooth’ (an element used in Mongolian writing for both i and y in the middle position) the scribes prefer the Oirat i that has a ‘notch’ that ‘breaks’ the axis. Plate 1 shows a fragment of folio 126 recto (volume pa, Vinaya section) where the scribe uses the Oirat form of i exclusively (the Mongolian i is not used, the ‘long tooth’ designates the letter y only).¹² In Plate 2, in the margin of folio 195 recto (vol. kha of the Pañcaviṃśatisāhasrikā Prajñāpāramitā) the scribe used two Oirat graphemes — i (in qorin) and e (in yeren).

Pl. 1. Mongolian Kanjur. 17th c., paper, ink. Fragment of folio 126 recto, volume pa, Vinaya section, Shelfmark Mong. K 36; IOM, RAS.

Pl. 2. Mongolian Kanjur. 17th c., paper, ink. Fragment of folio 195 recto, vol. kha, Pañcaviṃśatisāhasrikā Prajñāpāramitā, Shelfmark Mong. K 29; IOM, RAS.

The text copied on folio 195 recto demonstrates a peculiar case of a change of hand which corresponds to the change of ductus. The first five lines are written in an ‘Oirat’ hand characterized by the incline of transverse lines and the use of the grapheme i from Clear Script. This hand continues up to the middle of line 6 (the words nogčigsen-e olan bolai:), when (starting from the words sayibar oduγsan-a olan bolai:) another hand takes on. The ductus changes considerably: the soft, sliding, rounded oblique lines give way to straight, thick, bold geometrical strokes, with stark contrast between thin and thick lines. The graphemes a and n in the final position have long vertical ‘hanging tails’ (as opposed to curved or horizontal ones), the ‘braids’ of l and m are thick vertical lines that go directly upwards and downwards (where space allows it). This kind of calligraphy was used in seventeenth-century Mongolian manuscripts (one famous example is the manuscript of the “Twelve Deeds of the Buddha” from the collection of St. Petersburg State University, Mong. E-13). This exemplary ‘uncial’ hand goes on in lines 7–13, until the ‘Oirat’ hand returns in line 14 — for just one line, only to be taken over again by the thick-bodied ‘uncial’ from line 15 onwards. A closer look at line 14 shows that the scribe was trying to mimic the ductus of lines 6–13 by introducing vertical ‘tails’ (in kemer-ün, iǰaγur-tan and köbegün) that he had not used in the initial lines of the folio. Not only did he copy the vertical shape of the ‘tails’, he tried to reproduce the decorative element (a thin crescent) on their ends. His first attempt resulted in a clumsy forked line (kemer-ün), while the following two (iǰaγur-tan, köbegün) closely resemble the original, although the delicate thinness of the rounded stroke is not achieved here. Apart from this experiment in imitating Mongolian calligraphy, the Oirat scribe holds on to his habitual ductus, and in the last word of line 14 uses two graphemes from Clear Script: i and q (inaqsi).

This little case study is a vivid instance of two writing traditions in interaction. While one can but fantasize about the circumstances under which the copying of this folio took place, a few details can be reconstructed with some certainty. The page demonstrates the work of two scribes copying a sutra written in Classic Mongolian. One of them is well familiar with Clear Script, his hand is used to writing in the Oirat style (hence the typical ductus), and he sporadically uses graphemes from todo bičiq. Whether these graphemes are reproduced mechanically, due to the formed habit of writing in Oirat, or introduced intentionally, is not clear. The other scribe is very well skilled in Mongolian calligraphy. It is possible that on this page he demonstrated the style of writing that he had mastered or considered to be most suitable for copying a canonical text. After writing a few lines to display his technique, the calligrapher allows the Oirat scribe to try and copy his style of penmanship, and line 14 shows the result of this experiment, demonstrating how hard it is to break the habit of moving one’s pen in a certain way and switching to a different ductus.

Among the Kanjur fragments found in Ablaikit and preserved at the Institute of Oriental Manuscripts there are two folios (shelfmark Mong. K 38) that belong to a collection of ritual texts composed by the Fourth Panchen Lama Blo bzang Chos kyi rGyal mtshan (1570–1662) and, possibly, other authors. These texts are not part of the Kanjur, and the size of the folios is smaller than that of Ms 2, however, the handwriting is very similar, and certain features of orthography and the use of grammar structures point to the Oirat origin of the manuscript: it is safe to say that it was copied by Oirat scribes, possibly, in Ablaikit or another site in Dzungaria. The folios (numbers 59 and 90) are of thick unpolished paper, 64.5×22 cm in size; the text (52 lines per page) is written with a calamus in black ink. The full facsimile, identification and transcription of these fragments have been published by Alla Sizova.¹³

The ductus of the handwriting on both folios is the same as the prevailing ductus of Ms 2 — the ‘slanting’ Oirat kind (see Pl. 3, in which a fragment of folio 59 recto is presented). Occasional use of graphemes from Clear Script (i, e, q) is present as well. The ritual texts on the folios of K 38 contain a number of dhāraṇī (formulas in Sanskrit) written with the help of Galik signs, that is, special letters used in Mongolian scripts to render foreign words with more precision. Here, apart from the Mongolian Galik, the scribes introduced two signs that are used exclusively in Oirat Galik — the graphemes that imitate the Tibetan subscript letters ya  (Tib. ya btags) and wa  (Tib. wa zur). These signs are used in such words as šāky^a (= Skt. śākya), bidy^ā (= Skt. vidyā; see Pl. 3, lines 17, 19), sw^ā hā (= Skt. svāhā; see Pl. 3, lines 16, 17, 18, 20 and 21), bôdhi sadw^ā (= Skt. bodhisattva; see Pl. 3, line 4), etc. The spelling of bôdhi sadw^ā is a good example of the fusion of two Galik traditions. The word bôdhi is spelled using the Mongolian Galik sign for ô  (this letter marks o in foreign words, as in Mongolian writing there is no distinction between o and u) — in Clear Script it would be spelled as bodhi using the letter o . The word sadw^ā is spelled as it would be in Clear Script — in Mongolian it would most probably be spelled as saduva.

Pl. 3. Collection of ritual texts composed by the Fourth Panchen Lama. 17th c., paper, ink. Fragment of folio 59 recto, Shelfmark Mong. K 38; IOM, RAS.

Apart from the use of individual graphemes from todo bičiq, the two folios of K 38 contain several examples of entire words written in Clear Script: ösȫ ‘malice’ (= Mong. ösiy-e), könȫkü ‘harmful’ (= Mong. könögekü), yamāru ‘which’ (= Mong. yambar), tegēd ‘after that’ (= Mong. teyin kiged). Several words are partly written in Clear Script, for example: duusuγsan ‘complete’ (cf. Mong. daγusuγsan, Oir. duusuqsan), duurisuγsan ‘resounding’ (cf. Mong. daγurisuγsan, Oir. dourisuqsan). There is an Oirat ‘trace’ in the lexicon of the text as well: the term ilaγun tegüsügsen (the Oirat equivalent of the term Bhagavan, a title used to address buddhas, corresponds to the Tib. bcom ldan 'das) is used here consistently (occurs eight times), while its Mongolian counterpart ilaǰu tegüs nögčigsen does not appear on the two surviving folios.

Finally, elements of Oirat grammar (case markers and verb suffixes) appear in the text alongside Mongolian ones. To mark the Genitive case, in several instances the suffix -yin is used instead of ‑un: busud-yin, takil-yin, nom-yin, γar-yin, iǰaγur-yin, tangγariγ-yin, yabudal-yin, simnus-yin. It is not written according to the rules of Clear Script (cf. Oir. simnusiyin, nomiyin, etc.), but the Oirat influence is apparent. In several cases the suffix -i is used to mark the Genitive instead of -u (boluγsan‑i, ǰirüken-i, amitan-i, burqan-i, čiγulγan-i, mön-i), but the use of -i for the Genitive is not uncommon in seventeenth-century Mongolian manuscripts, so in this case the connection with the Oirat tradition is arguable.

For Converbum Imperfecti, the suffix -ǰi/-či is used alongside -ǰu/-ču: üiledči, abči, γarči, tasulǰi, daruǰi. In Oirat, this is the only suffix for this type of converb, while in Mongolian manuscripts it is rather rare (Nicholas Poppe described this form as occurring under the influence of the colloquial language).¹⁴ A similar example is the suffix of the past tense -lai which appears in the text once (nomlalai). It is not known to have been used in Mongolian texts of the seventeenth century (Poppe describes it as a variation of -luγa/lüge that occurs in popular books of the modern period).¹⁵ In the Oirat language, -lai/-lei is one of the three main suffixes of the past tense.¹⁶

The fragments of manuscripts collected at the abandoned temple of Ablaikit in the eighteenth century are not the only sources that cast light on the book-related practices that took place at the monastery. An important contribution was made by the archeological expeditions that worked at the site in the years 2016–19.¹⁷ Over two hundred fragments of manuscripts on birch bark (the exact number has not been published) were discovered during the excavations at the ruined temple, most of them around the altar (closer to the northern wall of the temple), at the depth of 0.3–0.5 m from the daylight surface. Previously the site had been disturbed by treasure-hunters, and it is problematic to reconstruct the circumstances under which the manuscripts appeared in the occupation layer of the temple.¹⁸ It is possible that they were scattered inside the temple after Ablaikit was abandoned, but the fact that the fragments were found within the foundation of the altar suggest that they could have been ritually buried there during the construction of the monastery. Ceremonial burying of Buddhist texts inside sacral constructions (possibly, as part of consecration rituals) was practiced in Mongolia as a way of handling damaged or dispensable manuscripts that could not be simply discarded because they contained sacred scriptures.¹⁹

All the manuscripts discovered during the archeological excavations at the Ablaikit temple are written on birch-bark, and most of them are small, damaged pieces. The few fragments that still carry legible text contain dhāraṇī written in Tibetan script. There are three fragments with writing in Clear Script which have been identified as folios of the same manuscript — a small pothi format book (5×10 cm, black ink, calamus) that contained the Prajñāpāramitāhṛdaya sutra (widely known as the Heart Sutra) translated by Zaya paṇḍita.²⁰ If the manuscript was indeed ritually buried within the foundation of the altar, it could be dated rather precisely to the period from 1648 (the creation of Clear Script) to 1657 (the consecration of the temple), if not — to a longer period, possibly up to the 1710s. In either case, it is one of the earliest surviving specimens of Oirat penmanship.

The three folios of the Prajñāpāramitāhṛdaya are severely damaged, and very few complete words can be read with full certainty (Pl. 4 demonstrates folio 13 recto — the best preserved piece of text). It is beyond dispute that the text is written in Clear Script: graphemes exclusive to todo bičiq are clearly visible on all the three folios, the text contains Oirat vocabulary (e.g. ilaγun tögüsüqsen for Bhagavan) and elements of grammar — case markers (e.g. -ēce for the Ablative, -bēr for the Instrumental). For example, see the following words on folio 13 (Pl. 4): sw^a ha in (line 3), sadw^a (lines 4 and 5), cāna kürüqsen (lines 7–8), tegēd (line 10). However, the few undestroyed fragments of text contain elements of Classic Mongolian as well: in four instances Mongolian forms of case markers are used instead of Oirat ones. In two cases the suffix -un/-ün marks the Genitive instead of -iyin: b[o]disad[w^a-na]run (f. 2r, lines 5–6) and [sed]kilün (f. 2v, lines 3–4). The other two examples show the suffix ‑dur/‑dür used for the Dative-Locative instead of -du/-dü: yabud[al]-dur (f. 2r, line 8) and üzeqči-dür (f. 13v, lines 5–6).

Pl. 4. The Heart Sutra (Prajñāpāramitāhṛdaya) in Oirat. 17th c., birch bark, ink. Fragment of folio 13 recto. Regional Museum of History and Local Lore, Oskemen (Ust-Kamenogorsk), Kazakhstan.

The data gathered from these fragments of text is scarce, but it complements the observations based on the Ablaikit Kanjur (Ms 2). While in Ms 2 the scribes were writing in Mongolian and demonstrating their knowledge of Clear Script, in the birch-bark copy of the Prajñāpāramitāhṛdaya the situation seems to be reverse: the scribe was writing in todo bičiq, but was not very well skilled in writing in Oirat, hence the occasional interpolations from Mongolian. If the assumption concerning the early origin (1648–1657) of this manuscript is correct, it is possible that its text reflects the period when the practice of writing in the new script was still being adapted. That said, the irregularities in the use of case markers could simply come from the lack of experience and knowledge of this particular scribe: his handwriting is unsteady, the text contains mistakes and cases of irregular spelling. The fact that it is written on birch bark (an inexpensive material compared to paper) indicates that the manuscript was probably meant for private use or served as an amulet.

In the absence of other manuscripts that date back to the first decades after the introduction of Clear Script, the texts discovered in Ablaikit serve as a unique source of information on the early development of the Oirat writing tradition. Most notably, these texts show that the transition from Mongolian script to todo bičiq was not immediate for the Oirats: there was a period when the two literary traditions coexisted, and both scripts were used concurrently within the same environment. This material is especially significant as the manuscripts come from Ablaikit — a place so closely connected with the life and work of the creator of Clear Script, Zaya paṇḍita.

[1] For a thorough description of Clear Script and the classical Oirat language see Jamсa 1999 & Iakhontova 1996.

² Luvsanbaldan 2015: 295.

³ For the list of texts they translated into Mongolian and Oirat see Radnabhadra 1999: 62–67.

⁴ Krueger 1969.

⁵ Radnabhadra 1999: 75, 78.

⁶ For a detailed account of the life of Ablai tayiǰi see Tsyrempilov 2020.

⁷ The circumstances were described by Gerhard Friedrich Müller in Müller 1738.

 [8] There are fragments of other Tibetan and Mongolian manuscripts that were delivered from Siberia to St. Petersburg and Europe in the eighteenth century, but their origin is a matter of argument. A number of publications have come out on the topic in the recent years. A summary on the topic can be found in Baipakov et al. 2019: 181–281. The subject has been thoroughly studied by the curator of the Tibetan collection of the Institute of Oriental Manuscripts Alexander Zorin. See Zorin 2015. The most recent findings on the topic can be found in a special section of the volume “Tibetan Studies in St. Petersburg” (Tibetologiia 2021: 14–266).

 [9] The two manuscripts are described, and most of their fragments from different collections listed, in Iampol'skaia 2015.

¹⁰ This ductus was described as typically Oirat by Gyorgy Kara in Kara 2005: 151–152.

¹¹ For example, in both manuscripts the folios that belong to the nga volume of the Tantra section are marked as niγuča quriyanggui-yin dandir-a, part of the folios of the ca volume — včir erike-yin dandir-a, the pa volume — dandir-a yeke altan gerel. Within the Pañcaviṃśatisāhasrikā Prajñāpāramitā, the folios of the ka volume are marked as yum qorin tabun mingγatu, the kha volume — qorin tabutu nögüge, the ga volume — qorin tabun mingγatu, the nga volume — qorin tabutu. In the Vinaya section, the folios of the ja volume are marked as dulba (from the Tib. 'dul ba), all the other ones — vinay-a (in different spelling variations). In the Sutra section, the volumes ga, ja, da, na, dza and wa are marked as olan sudur, the volumes ma, ya, sha and i — eldeb, the volumes zha, ra and sa — eldeb sudur, while the folios from the volume ah have to varying margins — eldeb and ǰaγun üiletü. These volume markers match consistently in Ms 1 and Ms 2 and, as most of them have nothing to do with the content of the volumes, this cannot be a coincidence.

¹² In the same fragment, the grapheme t in the word metü is also written as it would be in todo bičiq (see lines 4, 10), but the phenomenon of using the initial form of t in the middle position occurs in Mongolian manuscripts of that period as well (in particular, in the Mongolian Kanjurs discovered in Dzungaria), so the use of this grapheme can be attributed to the archaic orthography of the manuscript.

¹³ Sizova 2022.

¹⁴ Poppe 1954: 96.

¹⁵ Poppe 1954: 92.

¹⁶ Iakhontova 1996: 86–87.

¹⁷ The work was initiated in 2016 by the Margulan Institute of Archeology and taken on “Archeological Expertise” LLC under the academic guidance of professor Karl Baipakov (1940–2018) in 2016–18 and Marc-Olivier Pérou in 2019. Archeologist Yelran Kazizov was in charge of the excavations throughout the whole period of 2016–2019.

¹⁸ Baipakov et al. 2019: 345–348.

¹⁹ Chiodo 2000: 2.

²⁰ For the full facsimile, transliteration and detailed commentary see Yampolskaya 2022.