Notes

n.1Skilling 2016, p. 36, n. 65, points out that verses similar to those found in this work are recited in contemporary Thailand during various types of merit-making activities.

n.2For a survey of references to the terms stūpa, mahāstūpa, caitya, and mahācaitya in early Buddhist literature and inscriptions, see Skilling 2016. For Tibetan perspectives on the culture of the stūpa, see Dorjee 1996. See the 84000 Knowledge Base article, “Stūpa.”

n.3There are many scholarly studies of these monuments. One relatively recent collection of essays on the Buddhist stūpas of South Asia is Hawkes and Shimada 2009. See Gómez and Woodward 1981 for a collection of essays focused on Borobudur.

n.4For instance, the practice of making such small clay stūpas or saccakas‍—a term that is etymologically related to the Tibetan word tsa tsa‍—is described in ādikarma literature such as the Ādikarma­pradīpa and Kudṛṣṭi­nirghātana. For a pioneering (abridged) translation of the relevant passage in the former, see Beyer 1974, p. 59.

n.5See Walters 1997 for a study that connects Buddhist narrative literature to the emergence of a culture of stūpa construction in early Buddhist India.

n.6Into this stūpa, one is instructed to insert the sacred stanza on dependent arising (pratītya­samutpāda) that begins with ye dharmāḥ…, “Those things…,” a practice that is well attested by archaeological and epigraphic record. For further information on the latter, see Boucher 1991.

n.7For more information on the parallels between The Avalokinī Sūtra, the Mahāvastu, and the Śikṣāsamuccaya, see sections i.­4–i.­6 of the introduction to the 84000 translation of The Avalokinī Sūtra.

n.8See Melzer 2010, pp. 55–57, for references to these Sanskrit manuscripts: four fragments in the Hoernle collection (likely from Kucha, Xinjiang) held at the British Library in London, three iterations (one complete) in the Buddhist manuscripts from Gilgit (Pakistan) held at the National Archives of India in New Delhi, and two fragments (likely from Bamiyan, Afghanistan) in the Martin Shøyen collection in Oslo.

n.9According to Melzer 2010, pp. 56–57, the complete version is found in Vira and Chandra, eds., Gilgit Buddhist Manuscripts, vol. 10, folio 1585 and continues on folios 1524–1528.3. Two fragmentary versions are also found in vol. 7, folios 3249.3–3250, comprising the ten opening stanzas, and vol. 10, folios 3356–3357, containing approximately seven stanzas.

n.10The preserved fragments of this Sanskrit manuscript (Cambridge MS Os.131) are held at the Cambridge University Library and are available online through the Cambridge Digital Library. This manuscript has largely escaped the notice of recent scholarship since it was first acquired by Cecil Bendall at the end of the nineteenth century, although Bendall’s original handwritten notes to the manuscript indicate that he had identified it as containing a version of The Verses on Circumambulating Shrines, among other works. We became aware of its existence thanks to a blog post by Dorji Wangchuk in which he cites Péter-Dániel Szántó. Szántó then gave us further information on the manuscript via personal communication, which helped us to make our own preliminary analysis of it.

n.11For references to these three verses as found in Sanskrit witnesses to the Śayanāsanavastu and Saṅghabhedavastu, see Melzer 2010, pp. 58–59. See also Matsumura 1985 and Belanger 2000.

n.12For more on these parallels, see Belanger 2000, pp. 12–14, and Melzer 2010, pp. 58–59.

n.13For English translations of these passages from the Divyāvadāna, see Rotman 2008, pp. 155–59, and Rotman 2017, pp. 176–80.

n.14In his brief discussion of this work, Skilling 2016, p. 33, also highlights these two claims as significant ones.

n.15Appendix A in Tropper 2010, p. 53, shows parallels between the Chinese of Taishō 700 and the partial Tibetan inscription of this text at Alchi in Ladakh.

n.16Taishō 699 was translated by Divākara during the Tang Dynasty (680 ᴄᴇ). An English translation has been published by Boucher 1991, pp. 8–10.

n.17Belanger 2000, p. 11, n. 48.

n.18For the Khotanese transliteration see Bailey 1951, pp. 72–74; and for a translation, see Bailey 1974.

n.19Denkarma, fol. 301.a; Herrmann-Pfandt, pp. 159–60; Phangthangma, p. 22.

n.20Following the Cambridge Sanskrit MS, which reads yo ’graḥ prajñāvatāṃ, and a handful of Kangyurs, such as London and Namgyal, which have shes rab ldan mchog. Most editions of the Kangyur, including Degé, Stok, and those collated in the Pedurma, read shes rab blo ldan, “possessing wisdom and intelligence.”

n.21Following Stok, which reads: lha klu de bzhin gnod sbyin dang// sha za de bzhin srin rnams kyis// thams cad du yang mchod par ’gyur. Stok is supported by the Cambridge MS, which reads devair nāgais tathā yakṣaiḥ | piśācair atha rākṣasaiḥ | sarvatra pūjāṃ prāpnoti…. The last three lines of the Degé read a bit differently here, lha klu gnod sbyin dri za dang // lha min mkha’ lding mi’am ci// lto ’phye che rnams mchod byed ’gyur, “…shall be honored by devas, nāgas, yakṣas, gandharvas, asuras, garuḍas, kinnaras, and mahoragas.”

n.22The Degé reads lha dang de bzhin mi rnams kyis, while the Cambridge MS has divyaṃ mānuṣyakaṃ tathā.

n.23The Tibetan reads blo ldan, “possessing intelligence,” while the Cambridge MS has vīro, “a hero.”

n.24Here and in verse 6, the Cambridge MS uses the same word, vicakṣaṇa (“clear sighted” or “bright”), whereas the Degé in verse 6 has mdzangs pa and here reads mkhas pa, “wise.” We have chosen “bright” here because it conveys both the physical and mental quality of the term.

n.25The Tibetan reads ’du byed stong par mthong na kyang, apparently reading dṛṣṭvā saṃskāraśunyatāṃ, whereas the Cambridge MS reads dṛṣṭvā saṃsāraśunyatāṃ, “seeing saṃsāra as empty.”

n.26The Tibetan has sang rgyas bstan la dga’ ba dang, while the Cambridge MS reads prasanno buddhaśāsane. This is the first of several instances in this Tibetan translation in which the Tibetan term dga’ ba, more commonly used to translate words meaning “joy,” appears to translate the Sanskrit term prasanna, which typically means “faithful,” “serenely confident,” or “trusting,” but which can also sometimes mean “pleased” or “glad.” Other occurrences of these terms are found in verses 49–55 and 57, while in the translation of verse 58 the Tibetan translates prasanna with dad, “faithful.”

n.27The Tibetan has de ni nor bu dag pa bzhin, while the Cambridge MS reads maṇivatpariśuddho sau, “…pure like a jewel.”

n.28Tib. ma ma and Skt. dhātrī may indicate female caregivers more broadly, such as wet nurses, nannies, and so on, and may even refer to mothers, though the former sense seems the more primary usage. The overall sense of the verse suggests something about the affluent circumstances into which one is born.

n.29For this line, S reads de la gnod ba mi nus te, while D has de la gnod par mi ’gyur te, but the Cambridge MS, which has nāsau laṃghayituṃśakyaḥ, would seem to support the former.

n.30S has lus brtan, while D reads lus bde, “comfortable body,” but the Cambridge MS has sthirakāya, which again supports the former.

n.31Tib. dpung mgo zlum zhing brtan par ’gyur. “Rounded shoulders” (Tib. dpung mgo zlum, Skt. saṃvṛttaskandha) are among the thirty-two auspicious marks of a great being. This line may indicate a variation, since the Cambridge MS reads sthirasthaḥ sarvadharmeṣu, “abides in steadiness in regard to all things.”

n.32The Degé has ’thab bral lha yi rgyal po, “king of the gods of the Heaven Without Strife,” while S has ’thab bral lha yi dbang po, “lords of the gods of the Heaven Without Strife,” and the Cambridge MS has suyāmo devaputraś, “the god Suyāma,” who is said to be the divine ruler of the Yāma Heaven or Heaven Without Strife.

n.33Tib. shas cher, Skt. samvegabahula. The concept of saṃvega (“agitation,” “intensity”) is one that is sometimes applied in Buddhist literature to persons who are practicing the Buddhist path with a persistent mindset.

n.34The Degé has’khor rnams tham cad, “all assemblies,” but the Stok has khon rnams tham cad…, which is indeed supported by the Cambridge MS, which reads sarvavaira.

n.35In Tibetan, this verse interrupts the preceding series of four-line verses with a verse of six lines. This six-line verse appears to be an elaboration of the extant Sanskrit versions from Gilgit and Central Asia, on which see Melzer 2010, p. 64. However, it is supported textually, at least in part, by the Cambridge MS, which includes the line praṇidhiśca pradakṣiṇaṃ (Tib. smon lam btab pa’ang bskor ba yin), “circumambulation is also the taking of a vow,” which none of the others have. Nonetheless, the Cambridge MS still does not seem to contain anything corresponding to the Tibetan bgrod dka’ bde dgur…, “easily” and “…that are difficult to reach.” Further work is required here to establish a textual basis for the underlying Sanskrit.

n.36In this and the following two verses, the Tibetan has bskor ba’i gom pa gcig, “a single step of a circumambulation,” but none of the extant Sanskrit versions, including the Cambridge MS, include a word corresponding to bskor ba, in keeping with what seems to be the original context of their utterance in the Vinaya. It is worth noting that the Tibetan translation of the Vinaya passage does not include the word bskor ba either. In the Tibetan, verses 46–48 are six-line verses with seven syllables per line.

n.37D kam bod za yi bu mo brgya; Skt śataṃ kāmbojikākānyāḥ; S tho gar yul gyi bu mo brgya (“one hundred Tocharian girls”).

n.38The Tibetan has gangs kyi glang chen, “elephants of the snows,” while the Sanskrit has haimavatā nāgāḥ, haimavat being another word for himālaya, which in Sanskrit means “abode of snow.”

n.39Here and in the next several verses, Tib. mkhas pa gang gis, “by a wise one,” and dga’ ba’i sems kyis, “with a thought of joy,” translate Skt. vidvān and prasannacittaḥ, respectively. For dga’ ba’i sems / prasannacitta, we have chosen to prioritize the Tibetan interpretation, although other renderings are possible. For more on these terms, see the introduction and notes n.­26 and n.­43 for further discussion.

n.40In the Tibetan, verses 49–55 are four-line verses with nine syllables per line, reflecting the fact that the meter of the Sanskrit verses also changes from eight syllables to eleven syllables per line.

n.41The Degé has pho drang, “palaces,” while the Stok has pha bong, “nuggets,” which works better with the comparison made in the verse and is closer to the Sanskrit piṇḍā.

n.42The final three verses in the Tibetan translation return to a pattern of four lines of seven syllables each, reflecting a return to anuṣṭubh meter in the Sanskrit.

n.43Here we have chosen to vary the translation of dga’ ba in the phrase sems dga’ ba ni mtshungs ’gyur na, which translates saṃaṃ cittaṃ prasādyeha, choosing “devout” as a translation for it here even though “joyful” would remain consistent with the Tibetan interpretation. The term again translates a form of the Sanskrit verb prasad, “to settle down,” “to become clear,” “to be pleased or glad,” “to have faith or trust,” of which prasanna is the past passive participle. For more on this verse and the general idea it expresses, see Schopen 1997, pp. 131–33 and note 75.

n.44Here the Tibetan translation of prasanna varies from dga’ ba to dad rnams, “those who have faith.”

n.45This verse has parallels or close variants in other canonical texts, including the Vinaya passages cited in the introduction, as well as Toh 189 (1.­25) and Toh 846a (1.­2).