Twin Buddhist Hymns, Shared Title: The Structural Intertextuality and Tradition of the Composition of the Sanskrit–Tibetan Bilingual Jātakastava and the Khotanese Version

Abstract

In the history of Buddhist literature, there have been two texts sharing the title Jātakastava (Birth Stories Hymn), one in Khotanese and the other in Sanskrit and Tibetan. These two texts exhibit similarities in genre, framework structure, thematic selection, and narrative logic, while diverging in the arrangement of specific materials, textual length, authorship, dissemination regions, and translation approaches. This study examines the Sanskrit–Tibetan bilingual Jātakastava—a text that has received limited scholarly attention—and undertakes a comparative analysis with the Khotanese Jātakastava to reassess the relationship between these two texts bearing the same title, as well as their associated authors and translators. Through the inclusion of stories such as the Elephant King Leaping off a Cliff and King Maitrībala sacrificing his flesh to feed the rākṣasas, the Sanskrit–Tibetan bilingual Jātakastava reinforces the sequence of the six perfections (pāramitās) and exhibits a more pronounced thematic progression in its arrangement, thus representing a concise model of classical Sanskrit hymns. The Sanskrit–Tibetan bilingual text is attributed to Jñānayaśas, who may be associated with the Northern Zhou translator Shénà Yēshě 闍那耶舍. The text was translated into Tibetan by the scholar Chos skyong bzang po of the Zhwa lu school, whose word-for-word rendering style possibly reflects a grammatical and commentarial approach characteristic of the Zhwa lu school’s linguistic studies. The Khotanese Jātakastava may be seen as an expanded recension of this Buddhist hymn tradition as it existed in Dunhuang around the tenth century, while the Sanskrit–Tibetan bilingual version represents an abridged form that reached Tibet in the fifteenth century. The similarities and differences between the two texts reveal the inherent tension between localized dissemination and standardized compilation in the transmission of Jātaka narratives.

1. Introduction

Among the extensive corpus of Buddhist eulogistic literature, there exist two works sharing the title Jātakastava, corresponding to the Sanskrit–Tibetan bilingual version and the Khotanese version; while bearing the same appellation, they diverge in textual length and linguistic medium, thereby constituting two distinct textual entities.

Of the two works, the Khotanese Jātakastava is more well-known. As a representative example of Buddhist eulogistic literature, this manuscript was unearthed from the Dunhuang Library Cave and is currently housed in the British Library under Ch. 00274. In 1938, photographs of this manuscript and a transcribed synopsis of its content by H. W. Bailey were published in the Codices Khotaneses (H. W. Bailey 1938). In 1944, H. W. Bailey published a study concerning the colophon of this manuscript, followed by a full Latin transliteration of the text in 1945 (H. W. Bailey 1944, 1945). While there is a substantial body of scholarship on this work, the most comprehensive study to date remains the monograph published by Mark J. Dresden in 1955 (Dresden 1955). This monumental work includes a Latin transliteration, a complete English translation, grammatical analysis, and a general introduction, alongside an extensive effort to identify parallel texts for the Jātaka stories mentioned in the Jātakastava; in terms of scale, this study is substantial enough to stand as an independent volume. Regarding the latest scholarship in China, the late Professor Duan Qing completed a full Chinese translation of the Khotanese Jātakastava before her passing; this work is forthcoming from Zhongxi Book Company (Fan 2024, p. 12).

The significance of the Khotanese Jātakastava has drawn modern scholarly attention to its namesake, the Sanskrit–Tibetan bilingual Jātakastava (Tib. Skyes pa rabs kyi bstod pa). This text, attributed to Jñānayaśas, is preserved in the Tibetan Tengyur as a Sanskrit poem transliterated into Tibetan script, accompanied by a Tibetan translation. H. W. Bailey first brought this work to light in 1939 during his search for the Sanskrit original of the Khotanese version (H. W. Bailey 1939); he conducted a bilingual collation and preliminary philological study, outlining the text’s structural framework and identifying the sources of the Jātaka allusions contained therein. Building on this foundation, D. R. Shackleton Bailey published a second critical edition with an English translation and notes in 1954, reconstructing Sanskrit terms that had not been entirely restored in the earlier edition in 1939 (D. R. S. Bailey 1954). In 2019, D. A. Komissarov compared the two Jātakastava texts, arguing that this form of hymn is not derived from Buddhist literary tradition but is rather a poetic form known since at least the time of the Ṛgveda, employed here as a literary device to praise the Buddha’s former lives (Komissarov 2019). Chinese scholars have also paid close attention to the Sanskrit–Tibetan version while studying the Khotanese text. For instance, Duan Qing and Fan Jingjing examined the bilingual version alongside their research on the Khotanese Jātakstava. Fan Jingjing also systematized the Jātaka allusions, demonstrating the principle of selecting stories based on the six pāramitās, and presented Chinese translations of selected verses (Fan 2024).

A number of points here call for closer examination. Although the two Jātakastava texts exhibit differences in content, they share a striking similarity in stylistic structure. This prompts the following questions: Is it possible that there exists a common literary source or an adaptive relationship between the two? What are the similarities and differences in their translation or compilation strategies? Who exactly was Jñānayaśas, the author of the Sanskrit–Tibetan Jātakastava? Is he the same person as the scripture translator Shénà Yēshě 闍那耶舍 (also named Jñānayaśas) recorded in Chinese historical sources? And what were the specific temporal and spatial contexts of his activities? The answers to these questions await further investigation.

Focusing on the relatively understudied Sanskrit–Tibetan Jātakastava and comparing it with the Khotanese version, this paper attempts to reveal a unique aspect of hymn composition and Tibetan Buddhist translation—specifically regarding terminological standardization and poetic conversion. Ultimately, this study aims to provide a representative case for research into Buddhist literature and the history of sutra translation.

Methodologically, this study covers three dimensions—textual relationships, translation strategies, and authorship authentication—which differ from each other in the types of evidence applied and the reliability of such evidence, and attempts to clarify their boundaries. For the identification of textual relationships, the exact correspondence of particular stanzas or narrative structures is regarded as strong evidence, while stylistic similarity may only indicate a common paradigm. The analysis of translation strategies is based on observable linguistic phenomena and is limited to objective description as far as possible. The authentication of authorship and the dating and geographic contextualization, which rely on cross-verification of colophons and name-based inference, is of an exploratory nature, and should by no means be taken as a definitive conclusion.

2. A Comparative Analysis of Textual Structure and Content

2.1. An Overview of the Sanskrit–Tibetan Bilingual Jātakastava

The Sanskrit–Tibetan bilingual Jātakastava is preserved in the Tengyur of the Tibetan Buddhist Canon, specifically within the section of Hymns (bStod tshogs), and is cataloged as number 1178 in the Derge edition, spanning merely five folios (Ka. 255v4-257v7). Notably, this text is absent from the Peking and Narthang editions of Tengyur.

The images and e-text of this text can be found on the website Resources for Kanjur & Tanjur Studies (http://www.rkts.org). As illustrated in Figure 1, the text on the opening page begins from the fourth line, adopting an alternating layout: a line of Sanskrit transliterated into Tibetan script is followed by its corresponding Tibetan translation, and so on, thus forming a bilingual text. The Sanskrit portion consists of 14 lines, comprising 20 verses (śloka). The meter employed is Śārdūlavikrīḍita (meaning “Tiger’s Play”), which features 19 syllables per line (structured as 12 + 7), with four lines per verse. Figure 1, Figure 2, Figure 3, Figure 4 and Figure 5 contain the entire content of this text.

Figure 1. The Derge edition of the Sanskrit–Tibetan bilingual Jātakastava, fol. 255v.

Figure 2. The Derge edition of the Sanskrit–Tibetan bilingual Jātakastava, fol. 256r.

Figure 3. The Derge edition of the Sanskrit–Tibetan bilingual Jātakastava, fol. 256v.

Figure 4. The Derge edition of the Sanskrit–Tibetan bilingual Jātakastava, fol. 257r.

Figure 5. The Derge edition of the Sanskrit–Tibetan bilingual Jātakastava, fol. 257v.

2.2. Similarities and Differences in Structure and Material Selection

2.2.1. Structure

The overall frameworks of these two homonymous works exhibit a striking consistency, both adhering to the standard tripartite structure of the Indian Stotra (hymn) genre: an introduction, a main body, and a conclusion. This model is not an innovation of Buddhist literature but originates from earlier Indian literary traditions, with prototypes already visible in the Vedic hymns of the Ṛgveda (Komissarov 2019, p. 165).

The Sanskrit–Tibetan bilingual Jātakastava encompasses 15 Jātaka stories. Specifically, verses 1–3 serve as an introductory eulogy to the Buddha; verses 4–18 constitute the main body, comprising the 15 Jātaka narratives; and verses 19–20 function as a concluding dedication of merit. In contrast, the Khotanese Jātakastava consists of 169 verses in total: verses 1–10 form the introduction; verses 11–158 make up the main body, containing 51 Jātaka stories; and verses 159–169 serve as the conclusion.

The function of the introduction is to declare the author’s intent, express reverence for the Buddha, and include conventional self-deprecating remarks. In both works, the authors acknowledge their limited capabilities to adequately articulate the Buddha’s greatness through language, yet this does not deter their creative impulse. The introduction to the Khotanese Jātakastava is notably longer; it incorporates a special note from the translator that describes the specific historical and political context of the text’s production and dedicates the merit to the King of Khotan, Viśa’ Śūra.

The main body serves as a eulogy to the Buddha’s virtuous deeds and merits from his past lives, and exhibits a prominent formal feature: a concatenated narrative structure, akin to a string of pearls. Specifically, this is a poetic structure where each verse (or chapter) narrates the core plot of a single Jātaka story. There is no single plotline running through the entire text; rather, it is composed of a series of juxtaposed, relatively independent short stories.

In his analysis of the two extant Jātakastavas, Dmitrii A. Komissarov identifies this arrangement as a recognizable “iterative structure”: in the Sanskrit–Tibetan bilingual Jātakastava, each stanza employs a dependent clause, introduced by the subordinating conjunction yat (“since”), to distill the core events and characters of a Jātaka, followed by a main clause, introduced by tena (“for that reason”), to explicate the consequences of that virtuous deed (Komissarov 2019, pp. 157–59). This recurring yat…tena… syntactic pattern, recurring throughout the poem, enables numerous Jātaka stories to be clearly cataloged within a single poem, forming a kind of “list of Jātakas” (Komissarov 2019, p. 166). Although no such strict syntactic correspondence is found in the Khotanese Jātakastava, each stanza block is concluded by a fixed homage formula (e.g., “Therefore, to you homage”), similarly serving the function of demarcating different Jātaka stories (Komissarov 2019, p. 158). Komissarov further traces this form—characterized by its “frame + iteration + enumeration of virtuous deeds”—back to the Ṛgvedic hymns to Indra, arguing that Buddhist authors appropriated a pre-existing Indian literary tool rather than inventing a new genre ex nihilo within the Buddhist tradition (Komissarov 2019, pp. 163–66).

In the context of Buddhist poetry, this narrative approach manifests in two aspects: independence, reflected in how each verse or stanza condenses the essence of a Jātaka story (e.g., “Offering the Body to the Tigress” or “Cutting Flesh to Ransom the Dove”), and accumulation, which refers to the way this narrative method aims to generate a sense of cumulative merit and awe through constant repetition and enumeration. It demonstrates to the audience that the Buddha’s enlightenment was not accidental, but the result of countless sacrifices and virtuous deeds accumulated over innumerable kalpas.

The Sanskrit–Tibetan Jātakastava follows this pattern strictly, with each verse corresponding to a single story in a highly regular manner; in contrast, the structure of the Khotanese Jātakastava is more flexible, where a single Jātaka story may span from one to five verses. Whether one verse or five, the narration is not exhaustive but highly condensed, mentioning only the iconic characters and core plot points.

Building upon Komissarov’s formal analysis, the present study further conceptualizes the structure of this Jātaka poetry as a form of “thematic iteration.” Although the plots differ (some involve giving eyes, others giving flesh), the core behavioral pattern is iterative: the Buddha engaging in extreme generosity or moral discipline in every life. Narratively, this form reinforces core Buddhist doctrines by constantly repeating the pattern of “sacrifice and perfection.”

The primary function of the poetic conclusion includes the phalaśruti (lit. “hearing the fruits”, benefit of the fruits), which outlines the meritorious rewards to be gained from reciting or writing the hymn and dedicates this merit to specific recipients, such as the king or all sentient beings. Komissarov notes that this merit transfer is not a Buddhist invention, but a standard framing feature of Indian hymnography (stotra), whose function is to enumerate the merits attainable through composing, memorizing, or reciting the poem and to dedicate them to specific recipients (Komissarov 2019, p. 155). The concluding sections of both Jātakastavas embody this feature. The Sanskrit–Tibetan Jātakastava dedicates its merit universally to all sentient beings. In contrast, the conclusion of the Khotanese Jātakastava can be characterized as composite: it incorporates an interlude of direct eulogy to the Buddha by the author; a continuing Jātaka narrative; and a dedication of merit to the king, Zhang Jinshan, all family members, and sentient beings, making its content significantly more enriched.

2.2.2. Material Selection

To examine the similarities and differences in the selection and arrangement of materials between the two homonymous works, the Table 1 lists the themes of the 15 Jātaka stories contained in the Sanskrit–Tibetan Jātakastava, alongside the corresponding locations of these stories in the Khotanese Jātakastava.

Table 1. Correspondence of Jātaka stories in the two Jātakastava texts.

Sanskrit–Tibetan Jātakastava 1			Khotanese Jātakastava 2
Story No.	Verse Count	Jātaka Story	Story No.	Verse Count	Jātaka Story
1	4	Initial Aspiration of the Mind (King Mahāprabhāsa)	1	11–15	Aspiration of King Mahāprabhāsa
2	5	Offering the Body to the Tigress	20	74–75	Offering the Body to the Tigress
3	6	Elephant King Leaping off the Cliff (Hastin Jātaka)	—	—	—
4	7	King Candraprabha Offering His Head	15	57–59	King Candraprabha Offering His Head
5	8	Giving Away Wife and Children (Viśvantara Jātaka)	47	147–148	Giving Away the Queen
6	9	Parrot Extinguishing the Forest Fire	40	127–130	Partridge Extinguishing the Forest Fire
7	10	Rabbit Burning Its Body as Offering	3	20–23	The Rabbit
8	11	The Patient Ascetic (Kṣāntivādi-jātaka)	9	38–41	The Ascetic of Patience (Kṣāntivāda)
9	12	Six-Tusked Elephant Offering Tusks	2	16–19	The Six-Tusked White Elephant
10	13	Suicide Save Merchants	25	91–92	Merchants in a Shipwreck
11	14	Entering the Sea to Seek Treasures (Badaradvīpa Jātaka)	18	68–70	Prince Pāpaṃkāra
12	15	(King Maitrībala) Cutting Flesh to Feed the Rākṣasas	—	—	—
13	16	King Śibi Cutting Flesh to Feed the Hawk	51	158	Cutting Flesh to Ransom the Dove
14	17	Fish Jātaka	39	124–126	King Padmaka Transforming into a Red Fish
15	18	King Sunetra Offering His Eyes	34	112–113	King Sunetra Offering His Eyes

¹ (H. W. Bailey 1939, p. 852). ² (Dresden 1955, pp. 420–46).

A comparison of the two homonymous works reveals that all 15 stories found in the Sanskrit–Tibetan Jātakastava have corresponding parallels in the Khotanese version. Previously, Fan Jingjing suggested that two of the 15 stories in the Sanskrit–Tibetan text—specifically Story No. 3 (Elephant King Leaping off the Cliff) and Story No. 12 (King Maitrī-bala Cutting Flesh to Feed the Rākṣasas)—were absent from the Khotanese version (Fan 2024, p. 14). This difference may be considered from at least three perspectives.

Firstly, from the perspective of textual transmission, the Sanskrit–Tibetan bilingual version and the Khotanese version each underwent different paths of copying and translation. The latter is a local manuscript belonging to the Buddhist circle of the Western Regions, Dunhuang, and its selection of materials was likely influenced by local devotional traditions and the visual artistic environment. As for the former, its Sanskrit source text may have originated from a lineage in the Indian subcontinent or the Nepal region; during its northward dissemination, the compilers would add, delete, or adjust the stories according to the religious needs of different regions. The two stories of “the Elephant King Leaping off a Cliff” and “King Maitrībala” had long been recorded in the Jātaka tradition of India proper, so it is natural that the Sanskrit source text preserved them; however, these stories were not popular enough in the Khotan region of the Western Regions and thus did not find their way into the local manuscript.

Secondly, from the perspective of editorial intent, the compilers of the Sanskrit–Tibetan bilingual version may have deliberately sought to heighten the extreme manifestations of the “perfection of generosity” (dāna-pāramitā). The story of King Maitrībala cutting his own flesh to feed a rākṣasa is a quintessential “self-sacrificial giving” narrative, forming a thematic resonance with other tales such as the prince sacrificing himself to feed a starving tigress, King Candraprabha offering his head, and King Śibi cutting his flesh to feed a hawk. The inclusion of this story would make the narrative of generosity more intense and systematic. The story of the Elephant King Leaping off a Cliff, on the other hand, is a “life-forfeiting rescue” tale featuring an animal protagonist, forming a sequence with other animal Jātakas such as the Six-tusked Elephant Offering His Tusks and the Rabbit Burning Its Body as an Offering, thereby reinforcing the doctrine that “all beings are equal, and even animals can practice the bodhisattva path.” The compilers of the Khotanese version may have been more inclined to select stories that complemented the local cave murals and sculptural art, whereas the compilers of the Sanskrit–Tibetan version were more concerned with the doctrinal integrity of the overall structure.

Moreover, from the perspective of the Tibetan translation process, the primary audience of the Tibetan version was monks, who required systematic meditative texts. Therefore, supplementing the two stories to complete the six perfections (pāramitās) in their proper sequence was in line with practical needs. For example, the Sanskrit–Tibetan bilingual version contains two elephant-related stories, “Elephant King Leaping off the Cliff” and “Six-Tusked Elephant Giving Tusks,” appearing in stanzas 6 and 12, respectively. In contrast, the Khotanese version includes only the story of the “Six-Tusked White Elephant” (its second story). In the Hastin Jātaka, the core theme of the Elephant King who leaps off the cliff is “compassionate salvation of the world,” emphasizing unwavering compassion in the face of individual suffering.

Below is the content of the sixth verse of the Sanskrit–Tibetan Jātakastava:

Skt. yat pralayasitopalādribahalaṃ mātaṅgatāṃ bibhratā

kāntārāntaravartināṃ tanubhṛtām ucchittaye vyāpadām ।

śailendraṃ pṛthulopalaṃ calayatā muktas tvayātmā taṭān

manye matsariṇāṃ nitāntamalinās tena kṣatā vṛttayaḥ

∥6∥ (D. R. S. Bailey 1954, p. 23)

Tib. dkar rdo ba ri glang po nyid bzung

dgon par mi ’khyam pa lus bzung chad¹ ’phongs pa rnams la

ri dbang yangs pa rdo ba g.yo ba bor khyod kyis bdag ’gram du

sems ser sna dri ma can des nyams spyod pa rnams

∥6∥ (D1178, 256a4–6)

¹ chad] chang D

Because you, bearing the elephant-form [as heavy as] the white-stone mountain at the cosmic dissolution,

To utterly destroy the calamities of embodied beings dwelling deep in the wilderness,

Shook that lord of mountains with its broad boulders and thereby freed yourself from the cliff.

I think that the extremely impure behaviors of the envious ones have thereby been shattered.

(New translation from Sanskrit)

As above, the sixth verse of the Sanskrit–Tibetan bilingual Jātakastava is about the “Elephant King Leaping off the Cliff.” The Sanskrit verse is supposedly not directly quoted from the standard Jātaka story as found in Buddhist scriptures, but rather from an independent Buddhist hymn of praise, and is based on the narrative prototype of that Jātaka but undergoes artistic refinement and philosophical sublimation. The core actions in the verse, such as “moving the huge stone” and “self-released,” have no direct counterpart in the standard Jātaka story. Therefore, this verse can also be regarded as an independent hymnic text based on the Jātaka motif illustrating the evolution of Buddhist literature from narrative to hymnic praise.

It is worth noting that the Tibetan translation of this verse cannot be aligned with the Sanskrit original, and it shows discrepancies at some key points. For example, line a of the Tibetan lacks the core modifier “crushing” (pralaya) found in the Sanskrit; line b mistranslates “wild-dwelling” (kāntārāntaravartin) as the opposite—“not wandering in the wilderness”; line c completely omits the crucial clause “your self was released” (muktas tvayātmā) and instead erroneously adds “I am on the riverbank” (bdag ’gram du), clearly misinterpreting the Sanskrit tataḥ (“therefore”) as “riverbank/hillside”; and line d of the Tibetan lacks the two main constituents “I think” (manye) and “destroyed” (kṣatā), reducing the entire sentence to an incomplete noun phrase. The Tibetan translation of this text exhibits such translation problems at multiple points.

Another example of different material selection is Story No. 12, i.e., verse 15, in the Sanskrit–Tibetan Jātakastava, which corresponds to the story of King Maitrībala cutting his flesh to feed rākṣasas; the act of feeding the rākṣasas in this account is analogous to the 8th story (verses 35–37) in the Khotanese Jātakastava. The core theme of King Maitrībala is “making great vows” (praṇidhāna), emphasizing the offering of one’s own flesh and blood for the purpose of liberating sentient beings. In the Maitrībala Jātaka, five yakṣas have been reborn as emaciated and tormented hungry ghosts due to their past evil karma, and approach King Maitrībala and beg for his flesh and blood. The king nourishes them with his own body, restoring their strength (Vaidya 1959, pp. 43–54; Khoroche 1989, pp. 47–57; Āryaśūra 2020, pp. 86–107). It is precisely this episode that is described in the 15th verse of the Sanskrit–Tibetan bilingual version.

By contrast, in the story of the Great Ascetic of the Snow Mountains sacrificing his body to hear a śloka, although there are elements of “rākṣasas” and “flesh and blood,” the central theme is renouncing the body to hear the dharma, rather than compassionately appeasing hunger. Thus, this is another similar Jātaka story.

Khotanese Jātakastava (D. R. S. Bailey 1954, p. 426; newly translated):

na’tcapha-kala tta khu huṣka samudrra.

pāttāla gaṃtsä bihī pve’styu tva.re

gara-ke’ca mäñaṃdä vistārī mistä

bīysānai stīrū rakṣaysa miysirkä

∥35∥

At the time when the eon came to its end and the ocean lay dry,

In the depths of Pātāla there dwelt a rākṣasa.

Vast in extent, like a mountain chasm,

Immense, dreadful, and terrifying to behold.

śau śīlo dā udisāyä yakṣä ’eha

uysānā dīṣṭai gara-trraikhāña ṣṭāṇa.

ni himye pvaṇä khu dyai ba’tai ttūrrä

āci-ve’sāra bīśa’ rrijsā haskye

∥36∥

For the sake of a single śloka,

You cast yourself from a mountain peak into the mouth of a yakṣa.

Seeing its gaping jaws, its tongue flickering like flame, its sharp fangs,

You knew no fear like a great sage, you remained steadfast.

hatsạ-āspe hamye gyasto hvaṃdä.

śakrrä āstaṃna. phare pajsama yuḍäṃda

rāja-riṣa’yä ṣṭāna yuḍai duṣka kīra

ttīnā namasumä ba’ysa ṣadi jsa brrīya

∥37∥

Gods and men found refuge in you,

Śakra and the others as well. Multitudes paid homage to you.

As a royal sage you accomplished wondrous deeds.

Therefore, I bow before you, O Buddha, with faith and loving devotion.

The content of the 8th story in the Khotanese Jātakastava is the sacrifice of one’s body to feed a yakṣa to obtain a śloka, which should be the Jātaka of the Great Ascetic of the Snow Mountains. This is the difference between the two texts.

Furthermore, Fan Jingjing argues that the Sanskrit–Tibetan Jātakastava contains the story of Prince Kalyānakārin (Good Deeds)’s voyage across the sea in search of treasure (Fan 2024, p. 14). However, this observation can be further refined; more precisely, verse 14 in the Sanskrit–Tibetan Jātakastava depicts a figure who journeys into the depths of Badaradvīpa (the Jujube Island) and obtains a cintāmaṇi (wish-fulfilling jewel). Yet Badaradvīpa is not associated with the story of Prince Kalyānakārin; rather, it belongs to the narrative of another protagonist, Supriya. Supriya’s story is known as the Badaradvīpa-jātaka (“The Jātaka of Jujube Island”), and may be consulted in the second story of the Jātakamālā composed by Haribhaṭṭa, another work in the genre of devotional praise (Hahn 1992, pp. 31–37; Khoroche 2017, pp. 19–25; Straube 2019, pp. 57–66). In fact, this verse can be understood as a highly condensed summary of the latter half of the Badaradvīpa-jātaka, covering the episode from the expedition to Badaradvīpa in search of the jewel to Supriya’s return home and beneficent activities. Most significantly, the simultaneous appearance of the keywords Badaradvīpa and cintāmaṇi makes it highly likely that the verse alludes precisely to this Jātaka.

The Jātaka of Prince Kalyānakārin and the Badaradvīpa-jātaka may be regarded as two variants of the broader narrative theme of a treasure-seeking maritime expedition. The story of Prince Kalyānakārin is widely transmitted in Chinese Buddhist literature. For instance, the chapter “Prince Kalyānakārin’s Voyage Across the Sea” 善事太子入海品 (Shanshi taizi ruhai pin) in the Xianyu jing 賢愚經 (Sūtra of the Wise and the Foolish) recounts how Prince Kalyānakārin sailed across the ocean to obtain a maṇi jewel for the benefit of the world and its people, suffered persecution at the hands of his younger brother during the journey, but ultimately recovered his sight and returned safely (He 2020, pp. 76–90). Similarly centered on the theme of voyaging across the sea in search of treasure, the Khotanese Jātakastava contains three such stories, namely Nos. 18, 41, and 46, which recount the blinding of Prince Kalyāṇakāri by Prince Pāpaṃkāra (vv. 68–70), Sarvārthasiddha’s journey to the nāga palace beneath the sea to obtain the wish-fulfilling jewel (vv. 131–133), and Vīryabala’s seafaring quest for treasure, during which he courageously battles rākṣasas (vv. 144–146), respectively. Among these, Story No. 18 belongs to the Kalyāṇakāri-jātaka cycle.

In terms of the sequencing of the Jātaka stories, although the two works differ in the total number of stories included, they both commence with the narrative of “Mahāprabhāsa,” specifically the story of the taming of the elephant. This work opens with the episode of “Mahāprabhāsa’s Initial Aspiration of the Mind,” a point first noted by Fan Jingjing (Fan 2024, p. 14).

Beyond this, the ordering of the other stories appears to show little direct correspondence, yet their arrangement is not arbitrary but follows a thematic progression centered on the core ideas of Mahāyāna Bodhisattva thought. The stories in the Khotanese Jātakastava are more diverse; by contrast, those in the Sanskrit–Tibetan bilingual Jātakastava collectively highlight the most difficult Bodhisattva practice, i.e., the perfection of giving (dāna-pāramitā), and especially the spirit of selflessness that pushes giving to its limit. The opening story of King Mahāprabhāsa, who first arouses the resolve for awakening, provides the fundamental motivation for all subsequent deeds, i.e., every great feat originates from the initial Bodhicitta. The middle stories—giving the body to feed a tigress, cutting off flesh to feed a hawk, offering the head, offering eyes, and so on—are concrete practices driven by this great vow, with the level of difficulty escalating step by step. The story of the Kṣāntivādin (Patient Ascetic), placed toward the end, instead emphasizes the utter selflessness and freedom from anger on the mental level. Meanwhile, stories such as the parrot extinguishing a forest fire, the merchant who sacrifices himself to save others, and the elephant story manifest another facet of compassion, namely, active deliverance and skillful means in self-sacrifice. In other words, the ordering in the Sanskrit–Tibetan bilingual Jātakastava likely takes into account its function in religious ritual: first, establishing the resolve for awakening, then progressively extolling the various difficult practices of the Bodhisattva, and finally, through the state of selfless forbearance, inspiring reverence and emulation in the faithful.

3. Examination of the Identities of the Authors and the Translators

The identities of the authors and translators of these two works are a subject worthy of investigation, as they are crucial for determining the dates of the works’ formation and transmission. The Khotanese Jātakastava does not provide explicit information regarding its author, offering only details about the translator, named in verse 4 as Vedyaśīlä (Dresden 1955, p. 402). To date, no scholarly research has been conducted on this figure. According to the colophon, the Khotanese Jātakastava was commissioned for copying by Zhang Jinshan (H. W. Bailey 1944, p. 11), who was a historical figure of the 10th century; thus, it can be inferred that the work also dates to the same century. Fan Jingjing further speculates that the work was compiled and translated around 970 CE (Fan 2024, p. 12).

3.1. Authors

The name of the author of the Sanskrit–Tibetan Jātakastava appears as Jñānayaśas/Ye shes grags pa in the colophon. As shown in lines 5 and 6 of Figure 5, the original colophon reads as follows:

Skt. jātakastaba ācāryā jñānayaśasaḥ ∥

Tib. skyes rabs kyi bstod pa slob dpon ye shes grags pas mdzad pa’o ∥

The Jātakastava is composed by Acarya Jñānayaśas/Ye shes grags pa.

In fact, this name has been a subject of controversy in modern scholarship, and the Tibetan translation is one of the contributing factors to this confusion. Fan Jingjing translates this name as “Zhì Míng 智名” (Wisdom Fame) (Fan 2024, p. 14). In the Tibetan Tengyur, there are three works attributed to the Tibetan name Ye shes grags pa; however, only the Jātakstava records the original Sanskrit name in transcription. Some scholars posit that there were two distinct individuals named Ye shes grags pa. They identify the author of the Jātakastava as Jñānayaśas, while attributing the other two works—which are related to the Mahāmudrā tradition and differ significantly in ideology—to a different author. These scholars reconstruct the Sanskrit name of the latter as Jñānakīrti (Mathes 2006, p. 206; Szántó 2015, p. 759), meaning “Zhì Chēng 智稱” (Wisdom Renown). The correspondence between Sanskrit kīrti, Tibetan grags pa, and Chinese “稱” (chēng) is more common in Buddhist history. A prominent example is the scholar Dharmakīrti, whose Tibetan name is Chos kyi grags pa. Consequently, a divergence has emerged in academic circles supporting the theory of two distinct Ye shes grags pa figures. The present study concurs with this position.

In the Tibetan Tengyur, there is another monk named Ye shes grags pa, the author of the Tattvāvatāra (Realizing Reality) and the Pāramitāyāna-bhāvanā-krama-upadeśa (Instructions on the Stages of Practice in the Paramitayana), who was active during the 10th to 11th centuries. This monk is not the same person as the author of the Jātakastava. In the Tibetan translation of the Tattvāvatāra, there are subtle variations in how the author is addressed: the summaries and colophons of the first three chapters mention the author, and the name given is “slob dpon ye shes grags pa” (Ācārya Ye shes grags pa), whereas the colophon uses the name “slob dpon dpal ye shes grags pa” (Ācārya Glorious Ye shes grags pa). The colophon of the Pāramitāyāna-bhāvanā-krama-upadeśa refers to him as “slob dpon chen po dpal ye shes grags pa” (Great Ācārya Glorious Ye shes grags pa) (Derge No. 3922, Ki. 77b2–3). Both texts include the additional element “dPal” (Glorious), which has led to the reconstruction of Ye shes grags pa’s Sanskrit name as Śrī Jñānakīrti. The element “Śrī/dPal” may simply be an honorific title attached to Jñānakīrti. In studies concerning the Tattvāvatāra, scholars more commonly render the Sanskrit name of Ye shes grags pa as Jñānakīrti (Lu 2018, p. 150–51).

So, what is the specific identity of the author of the Jātakastava? Currently, very little is known. Judging from the Chinese phonetic transcription, is the author Jñānayaśas related to the Northern Zhou translator Jinayaśas 闍那耶舍 (Shénà Yēshě, active 564–572), who translated the Mahāmeghasūtra (Great Cloud Sutra)? Typically, the Sanskrit name of Jinayaśas can be reconstructed in two forms: Jinayaśas or Jñānayaśas. Biographical data on Jinayaśas is scarce; he is generally considered to be a monk from Magadha, Central India, during the Northern Zhou dynasty. His disciples, Yaśogupta (active 561–578) and Jñānagupta (523–600), who accompanied him to China, are perhaps better known. The Xu Gaoseng Zhuan (Continued Biographies of Eminent Monks, 續高僧傳) states the following: “The meditator Jinayaśas from Magadha; in Zhou language, ‘Zàng Chēng’ 藏稱” (CBETA 2026, R1, T50, No. 2060, p. 429c1). The Kaiyuan Shijiao Lu (Kaiyuan Era Catalogue of Buddhist Teachings, 開元釋教錄) notes the following: “The śramaṇa Jinayaśas; in Zhou language, ‘Zàng Chēng’, also known as ‘Shèng Míng’ 勝名” (CBETA 2026, R1, T55, No. 2154, p. 545a12). The name “Shèng Míng” corresponds to Jinayaśas, which is easy to understand: jina means “victory” 勝 and yaśas means “fame” 名 or “renown” 稱. However, how the name “Zàng Chēng” corresponds to “Jinayaśas” is unclear. Yaśas is transliterated as “Yēshě” 耶舍 and can be translated as “Chēng” (fame/renown). However, whether “Shénà” 闍那 is transliterated from jina (victory) or jñāna (wisdom), translating it as “Zàng” (store/treasury, 藏), is not straightforward. Therefore, we can only say that it is difficult to completely rule out a connection between the Northern Zhou translator Jinayaśas/Jñānayaśas and the author of the Sanskrit–Tibetan Jātakastava. Based on the onomastic evidence, there is a possibility that they are the same person.

Based on the aforementioned literature, it can be determined that Jñānayaśas and Jñānakīrti are two distinct Sanskrit names with similar meanings, referring to two different Indian monks. In other words, in Tibetan literature, the dharma names of two monks were translated as the Tibetan Ye shes grags pa; they were active in different eras, and their extant works belong to different genres. If the Northern Zhou translator Jinayaśas/Jñānayaśas and the author of the Sanskrit–Tibetan Jātakastava are indeed the same person, then it can be inferred that the composition of the Sanskrit Jātakastava likely dates to the 6th century. For the time being, we cannot rule out this possibility.

3.2. The Translator of the Sanskrit–Tibetan Jātakastava

The colophon of the Sanskrit–Tibetan Jātakastava records the translator’s name as Dharmapālabhadra, which serves as an alternative translation for the great Zhwa lu translator Chos skyong bzang po (1441–1527/28). This identification suggests that the translation of this Jātakastava dates to approximately the 15th century. Chos skyong bzang po translated numerous works, including several related to Vyākaraṇa (Sanskrit grammar), three of which are preserved in the Tengyur. However, the Tibetan translation of the Jātakastava does not adopt a polished poetic form; instead, it translates Sanskrit words on a one-to-one basis. This characteristic has led to doubts regarding the translator’s proficiency. H. W. Bailey pointed out errors and limitations in the Tibetan glosses of the Jñānayaśas version, and D. R. S. Bailey chose not to collate the Tibetan text, deeming it of little use (H. W. Bailey 1939, p. 852; D. R. S. Bailey 1954, p. 22). However, it is unreasonable to cast such doubts on Chos skyong bzang po, a master of Vyākaraṇa.

A more plausible explanation is that Chos skyong bzang po treated this translation as an exercise in vyākaraṇa (grammar, grammatical analysis). This Tibetan version functions essentially as a lexical concordance, focusing on the mastery of Sanskrit vocabulary. This method resembles the annotation format of modern descriptive grammar, aiming to analyze the grammatical structure of the source language (Sanskrit) through the corresponding target language (Tibetan).

This annotation style adheres to the Sanskrit word order, which deviates from Tibetan idiomatic expressions and results in the loss of information that should have been conveyed in Tibetan. However, this provides a valuable case study for examining information loss in cross-linguistic text transmission. Further investigation into similar cases is required, such as the work Daṇḍakavṛttastotra attributed to Nāgārjuna in the Tengyur, which is also a Sanskrit–Tibetan bilingual text (Hahn 2010, p. 462).

Michael Hahn (Hahn 1987, pp. 11–23) conducted a detailed Sanskrit–Tibetan collation of the Daṇḍakavṛttastotra (attributed to Nāgārjuna and preserved only in the Narthang and Peking editions of the Tengyur), demonstrating that its Tibetan translation likewise adopts a strategy of word-for-word correspondence, subordination of Tibetan word order to Sanskrit, and the disassembly and recombination of highly condensed compounds. Hahn’s findings indicate that this type of grammatical gloss-style translation is not an isolated case in the Tengyur but rather may represent a recognizable paradigm. The following combines Hahn’s research with an analysis of the sixth verse of the Sanskrit–Tibetan bilingual Jātakastava (see Section 2.2.2 above) to demonstrate the operational mechanism of such glosses.

The Sanskrit original of this verse is a typical vyājastuti compound long sentence with an extremely high degree of poetic condensation. The Tibetan translation by Chos skyong bzang po displays a processing strategy highly isomorphic with that analyzed by Hahn (1987, pp. 20–22) for the Tibetan translation of the Daṇḍakavṛttastotra and can be examined from three aspects: word order, grammatical annotation, and compound reconstruction.

First, the Tibetan word order accommodates Sanskrit, thereby disrupting natural Tibetan grammar. The Tibetan poetic lines strictly follow the Sanskrit four-line structure, and the internal word order within each line also maintains as much consistency with Sanskrit as possible. Taking the pāda c as an example, the Sanskrit word order śailendraṃ pṛthulopalaṃ calayatā muktas tvayātmā is rendered word-for-word as ri dbang yangs pa rdo ba g.yo ba bor khyod kyis bdag, with the verbs g.yo ba (corresponding to calayatā) and bor (corresponding to muktas) moved forward into the middle of the clause, leaving only bdag ’gram du (“oneself on the bank”) at the end. This completely violates the Tibetan SOV (Subject–Object–Verb) expression pattern. This phenomenon of “grammatically transparent yet pragmatically obscure” is precisely the same pattern described by Hahn in the Daṇḍakavṛttastotra: the Tibetan word order is forced to conform to Sanskrit, rendering readers unable to rely on Tibetan’s inherent grammatical rules to understand the syntax naturally (Hahn 1987, p. 20).

Second, grammatical morphological annotation takes precedence, while contextual meaning and grammatical relations are submerged. In the pāda b of the Sanskrit, kāntārāntaravartināṃ tanubhṛtām (“those embodied beings who move in the interior of the wilderness”) is a genitive plural modifying vyāpadām (“calamities, disasters”); the entire sentence means “to annihilate the calamities of wild-dwelling creatures in their plight.” Although Tibetan renders this word-for-word as dgon par mi ’khyam pa lus bzung (“those who do not roam in the wilderness, holders of bodies”), the genitive plural marker is completely lost, rendering the modifying relation ambiguous; moreover, vartin (“moving”) is translated as mi ’khyam (“not roaming”), a semantically reversed rendering. This does not stem from the translator’s unfamiliarity with Buddhist doctrine but rather results from an excessive focus on Sanskrit morphological form (ākāra) at the expense of contextual meaning (artha). This is entirely parallel to the mechanism of graphic misreading pointed out by Hahn in the Daṇḍakavṛttastotra, where vraṭa (“crowd, multitude”) was mistaken for vrata (“vow, religious observance”) and translated as brtul zhugs (“religious discipline, yogic practice”): training in vyākaraṇa emphasizes root-and-sound correspondence, easily leading to confusion or counterinterpretation of graphically similar words (Hahn 1987, p. 21).

Third, when confronted with poetically condensed expressions, the translator performs a “grammatical reconstruction.” The Sanskrit matsariṇām (“the greedy”) is a single noun, yet Tibetan disassembles it into sems ser sna (“mind + greed”), reducing the noun to its conceptual components for morphological annotation rather than employing a ready-made Tibetan equivalent. This is the same mechanism identified by Hahn, where pātitā-jñāna-sārā-draye (“the one who has caused the mountain [of ignorance/of the essence of knowledge] to collapse”) was reconstructed as mnga’ bdag (“the lord/master…”) and snying mchog gang gi (“the heart-jewel of…”): the translator tends to “break down [Sanskrit] highly condensed compounds into smaller, corresponding units” (auseinandergenommen und neu zusammengesetzt), in order to serve grammatical analysis rather than literary reading (Hahn 1987, p. 22).

More crucially, the complete logical structure of the fourth Sanskrit line (pāda d) disintegrates in the Tibetan translation: manye (“I think”) is absent, and kṣatā (“damaged, destroyed”) is omitted, making it impossible to identify the subject–object relation of the clause and the passive voice. This characteristic—morphological annotation taking precedence while textual coherence is compromised—is precisely the internal evidence that such translations function as vyākaraṇa exercises rather than literary works. The underlying purpose of the Tibetan translation is likely to analyze the grammatical structure of the source Sanskrit through the Tibetan in corresponding positions, akin to the morphological glossing conventions of modern linguistics, but the cost is the loss of information that Tibetan expressive conventions would normally convey, such as case marking and voice, as well as the broader textual logic.

In sum, the Sanskrit–Tibetan bilingual Jātakastava and the Daṇḍakavṛttastotra share the following three characteristics among Sanskrit–Tibetan bilingual texts in the Tengyur: (1) word-by-word disassembly of Sanskrit compounds, with word order accommodating Sanskrit and disrupting natural Tibetan grammar; (2) morphological correspondence taking precedence over contextual precision, easily producing graphic misreadings or counterinterpretations; and (3) grammatical reconstruction when facing poetically condensed expressions, transforming literary rhetoric into analyzable lexical units.

These parallel phenomena suggest that a certain paradigm of grammatical gloss-style translation had already formed within the fifteenth-century Zhwa lu pa tradition of vyākaraṇa. Therefore, Dharmapālabhadra’s translation of the Jātakastava should not be regarded as an isolated case of the kind criticized by Bailey as “not always to be trusted to give a correct explanation” (H. W. Bailey 1939, p. 852); on the contrary, we have good reason to believe that it is precisely a paradigmatic practice of using translation itself as a tool for vyākaraṇa.

4. Conclusions

The foregoing analysis has established that two texts bearing the title Jātakastava are attested in two distinct historical and geographical contexts: 10th-century Khotan (or Dunhuang) and 15th-century Zhwa lu Monastery (TAR). Despite this chronological and spatial separation, the two texts exhibit notable structural and thematic correspondences, suggesting that they may draw upon a shared hymnographic repertoire rather than having emerged independently. This finding points to the persistence and mobility of hymn literature across the Buddhist world, even in periods and regions where its South Asian origins were no longer directly traceable.

The two Jātakastava texts both adhere to the traditional Indian tripartite rhetorical structure, i.e., general–particular–general. The main body of the hymns employs a condensed, bead-string-like sequential narrative, reinforcing the cumulative nature of Bodhisattva practice through thematic iteration. In terms of subject selection, the Khotanese version with 51 stories essentially covers the Sanskrit–Tibetan bilingual version of 15 stories. However, the latter includes two additional stories—the Elephant King Leaping off a Cliff and King Maitrībala sacrificing his flesh to feed the rākṣasas—which reflect an editorial intention to emphasize the six pāramitās in their sequential order and to serve monastic contemplative practice. The former, by contrast, appears to have been more influenced by the indigenous beliefs and artistic environment of the Western Regions and Dunhuang area. Both versions open with the story of “Mahāprabhāsa’s Initial Aspiration of the Mind,” yet the arrangement in the Sanskrit–Tibetan bilingual version implies a thematic progression from the initial generation of bodhicitta, through extreme generosity, to selfless forbearance, thereby profoundly demonstrating the gradations of Mahāyāna Buddhist cultivation. As such, it more closely approximates a standard and refined exemplar of the classical Sanskrit hymn genre.

Regarding the authorship and translator identity of the Jātakastava texts, the current textual research still leaves room for further inquiry. Based on the colophon of the Khotanese Jātakastava, the conclusion that it was composed by the author around the 10th century may be tentatively retained for now. The author of the Sanskrit–Tibetan bilingual Jātakastava is attributed to one Jñānayaśas (Ye shes grags pa), and scholars tend to distinguish this figure from another Indian Ācārya bearing the same Tibetan translation of his name. This author may have been connected with the Northern Zhou translation monk Shénà Yēshě 闍那耶舍. If this hypothesis is held, the date of composition could potentially be traced back to the 6th century, though this awaits further confirmation.

Concerning the Tibetan translator, the Sanskrit–Tibetan bilingual Jātakastava was completed by Chos skyong bzang po, a great translator of the Zhwa lu school active in the 15th century. With respect to the stiff word order and literal word-for-word correspondence characteristic of this translation, some scholars have previously questioned the translator’s competence. However, comparative textual studies suggest that this may not be due to any deficiency on the translator’s part, but rather due to a distinctive paradigm of grammatical commentary translation that belonged to the Śabdavidyā tradition of the Zhwa lu school during that period. It is possible that the translator conceived the work as a pedagogical tool, employing Tibetan equivalents aligned with the Sanskrit to parse the latter’s grammatical structure. Nevertheless, this remains no more than a plausible conjecture, and further evidence from additional textual sources is still needed.