Clay sealings from Perlis, Malaysia, and the wider world of the Bodhigarbhālaṅkāralakṣa-Dhāraṇī

Abstract

Multiple copies of a particular clay sealing bearing the Buddhist Bodhigarbhālaṅkāralakṣa-dhāraṇī (mantra) inscription were discovered in Gua Berhala—a cave in Perlis, Malaysia. These sealings can be roughly assigned to the tenth century and they appear to have been stamped with an identical seal. However, critical reading of the textual rendition of the dhāraṇī had not yet been done despite several attempts to study it. Therefore, based on several fragments of these sealings, this article provides a detailed reading and translation of the dhāraṇī and considers the cultural significance of their production. The article also examines the textual structure of this Perlis dhāraṇī and compares it with similar dhāraṇīs preserved in a palm-leaf manuscript and other materials found across Asia. This includes a survey on the wider transmission of the dhāraṇī in the continent. In this comparative exercise, the physical characteristics of the Perlis sealing appear to be unique and express a distinct artistic style, while its textual tradition is slightly compressed compared with others, with no identical equivalent found elsewhere. This observation suggests that Perlis, with its proximity to the Bujang Valley, participated in the wider network of dhāraṇī culture rooted in Eastern India and was just not a passive recipient of this practice.

Introduction

Towards the end of the nineteenth century CE, certain types of Buddhist clay sealings¹ were discovered by Vaughan Stevens in a cave in Perlis—a small state located on the north-west coast of Peninsular Malaysia.² Later on, more of these sealings were picked up by Ivor H. N. Evans and H. D. Collings.³ In 1964, Alastair Lamb reported further discoveries in the same cave, called Gua Berhala, in addition to another cave nearby, known as Gua Kurong Batang.⁴ These sealings are usually circular or pear-shaped and come in several designs. Some are impressed with the figures of a 12-armed Avalokiteśvara and some show a seated Bodhisattva with his right foot on a lotus cushion or a Buddha seated in dharmacakra mudra (preaching the wheel of law gesture) surrounded by eight Bodhisattvas.⁵ Some of these sealings were also stamped with inscriptions on the reverse side. One particular type of sealing that is of interest to this article is the one stamped with an inscription containing a dhāraṇī found in Gua Berhala.⁶ These sealings were assigned as Type 1 tablets in Lamb's paper (Figure 1). He described them as:

Pear-shaped tablets, of sun-dried red clay, with an inscription of 16 lines in a deeply recessed circular cartouche. In some specimens of this type the inscribed area had been coloured with a red pigment (haematite?). The reverse side of these tablets, as also of many other red tablets from this site, showed impressions from some kind of bamboo matting. These tablets were by far the most common type in Gua Berhala, though only one complete specimen was discovered. A photograph of the inscription on this type of tablet was sent to Professor G. Cœdès in Paris; but he has not, at the moment of writing, been able to provide a translation of it, though he reports that it does not contain the ye dharmma formula.⁷

Figure 1. The shape of the sealing stamped with the Bodhigarbhālaṅkāralakṣa-dhāraṇī recovered in Perlis, Malaysia—a hand copy from Plate 9 in A. Lamb, ‘Mahayana Buddhist votive tablets in Perlis’, Journal of the Malaysian Branch of the Royal Asiatic Society 37.2 (1964), pp. 47–59.

Earlier, in around the mid-1930s, fragments of this particular sealing, discovered by Collings, were also sent to Cœdès for interpretation, who was the Director of the École française d'Extrême-Orient in Hanoi. Cœdès wrote:

A, E, F, K, L, N, Q and R are fragments of an imprint from one mould, A, E, K, N, Q correspondingly to the upper part of the text; F and R to the left and L to the right side. Unfortunately, the gathering of these fragments does not enable me to reconstruct the complete text. All that I can say is that it begins with the invocation: Namo Bhagavato, ‘Homage to the Blessed One!’ I can also read the word Tathāgata (an epithet of the Buddha) on two fragments. The text was evidently some sort of prayer or incantation in Sanskrit or more probably in Prākrit. As regards the probable age of these tablets [including other types of sealing], they can hardly be anterior to the 10th century A.D.⁸

Even earlier, in 1864, the British Museum acquired six similar sealings with the acquisition notes as follows:

Book of Presents [512]: Department of Oriental, British, Medieval Antiquities and Ethnography. British Museum 11 November 1864. Mr. Bliss has the honour of reporting to the Trustees that six terra cotta impressions of seals with inscriptions in an old Pali character supposed to be Buddhist prayers and two fragments of terra cotta impressions of seals representing Buddhist deities found in a cave in the district of Patania in the vicinity of Penang, have been received in the department for the Trustees, offered as presents by Mr. W. E. Jevons of 6 Rumford Place Liverpool. The subject of these inscriptions requires further investigations to elucidate their contents and copies have been made for that purpose.⁹

The sealings preserved in the British Museum are more intact compared with those documented in the Bujang Valley Archaeological Museum (Figures 2 to 9) and the Kota Kayang Museum (Figure 10), though not all of them are complete without certain parts at the side broken off. The characters in the sealings in most cases are clearly visible but, yet again, none of them is complete without certain parts of the inscription having been chipped off or effaced. At the moment, we do not think that the British Museum's sealings were found in a cave in or near Penang due to a lack of any reports implying their archaeological contexts. These sealings most probably originated from Perlis and were sold in Patania, which could refer to the present-day Sungai Petani—a town in southern Kedah, just north of the Province Wellesley in Penang.¹⁰

Figures 2 to 9. Fragments of clay sealings stamped with the Bodhigarbhālaṅkāralakṣa-dhāraṇī found in Gua Berhala in Perlis, Malaysia. Photos by the authors.

Figure 10. One of the three fragments of clay sealings stamped with the Bodhigarbhālaṅkāralakṣa-dhāraṇī preserved at the Kota Kayang Museum in Perlis. Photo by the authors.

The stamped inscriptions of all these sealings from the British Museum and those studied by Lamb and Collings are identical. They appear to have been impressed by the same stamp and most likely came from the same cave. The inscription does begin with namo bhagavate (not bhagavato). However, these earlier studies or records do not provide the complete reading of this inscription.

It was not until 2013 that Nasha Rodziadi Khaw and Mohd Mokhtar Saidin published the first reading of the inscription.¹¹ As no single complete sealing was available to them, the reading is based on a collation of eight fragments of the tablets (which are reproduced in Figures 2 to 9). They found that the reading of this inscription resembles the textual content of certain sealings found in Hund, Pakistan.¹² Unfortunately, the reading of the Hund sealings was incomplete due to the fragmental nature of the tablets and effaced characters. However, they managed to identify the reading as a certain dhāraṇī, similarly to other north Indian clay sealings preserved in the British Museum in London and in the Ashmolean Museum in Oxford through the work of Simon Lawson.¹³ Nevertheless, none of these papers was able to identify the dhāraṇī. But it has now become clear that the dhāraṇī that appears on the sealings of Perlis, Hund, and the British and Ashmolean Museums is the Bodhigarbhālaṅkāralakṣa-dhāraṇī or the dhāraṇī of the Hundred Thousand Ornaments of the Essence of Awakening.¹⁴

This dhāraṇī seems to have been circulated quite widely in Asia. Its adaptations are preserved in the Chinese Buddhist canon as well as in various Tibetan Kanjurs that include a version thought to have been translated from Chinese.¹⁵ Furthermore, its Sanskrit original was transmitted in different lengths and textual traditions.¹⁶ As such, this Perlis sealing would merit a more careful study in order to contribute to the repertoire of textual data of this dhāraṇī. In this research article, we will present a critical reading of the Perlis sealing. However, we will not conduct a systematic comparative study on the textual tradition of this dhāraṇī—the most recent of which was undertaken by Ingo Strauch in 2009—as our main aim is to present what was found in Perlis. Nevertheless, we will compare the structure of this Perlis dhāraṇī to a similar dhāraṇī named the Ārya-lakṣa-nāma-dhāraṇī found in a twelfth- to thirteenth-century Nepalese palm-leaf manuscript along with readings of other sealings found in India, as well as a gold foil that was recovered in Java, in present-day Indonesia. We will also conduct a general survey on the geographical distribution of this dhāraṇī, which will yield important insights into the extent to which this dhāraṇī has reached out to people in Asia and how Perlis came to be part of it.

The Perlis sealing

Perlis is the smallest and northernmost state of Malaysia, bordering the provinces of Satun and Songkhla of Thailand to the north and the state of Kedah to the south. Perlis was originally part of Kedah, which covered most of the north-western part of the Malay–Thai Peninsula before the nineteenth century CE. The state is situated approximately 110 kilometres from the Bujang Valley—a historical and archaeological complex in Kedah where numerous Hindu and Buddhist epigraphs, icons, and religious shrines have been unearthed.¹⁷ The Bujang Valley was the economic and political centre of Ancient Kedah—a maritime polity located at the entrance of the Strait of Melaka consisting of several riverine and coastal settlements on the west coast of the Malay–Thai Peninsula.¹⁸ Evidence of commercial activities and industry can be found in Bujang Valley that dates back to as early as the second or third century CE before it developed into an entrepôt from the seventh century onwards.¹⁹ Ancient Kedah was also situated at the entrance of several riverine trans-peninsular routes, such as one following the Muda River that crossed into Pattani on the east coast of the Peninsula in present-day Thailand to the Gulf of Siam²⁰ and another one from northern Kedah via Perlis to Songkhla, tracking between the Thammarat and Kedah Singgora ranges (Figure 11).²¹ Thus, the region was bustling with human activities, not only of commerce and trade, but also of culture as well as religious thought and practice. As such, Perlis, given its proximity to Bujang Valley and being one of the settlements of Ancient Kedah, might have actively participated in these exchanges in one way or another (more of this below).

Figure 11. Malay–Thai Peninsula with archaeological site of Bujang Valley, and Gua Berhala in Perlis.

As far as we know, this particular Bodhigarbhālaṅkāralakṣa-dhāraṇī sealing of Perlis (hereafter the Perlis Bodhi sealing) was only discovered in Gua Berhala, in multiple copies, despite the fact that, in the same cave, there were other types of sealings that were also found in other areas in the Peninsula. These Perlis Bodhi sealings are pear-shaped, possibly emulating the structure of a caitya or stūpa.²² Given the variations in the sizes of these pear-shaped sealings, the lumps of clay would be made by hand and the seal impressed upon them.²³ They were dried either under the Sun or through the heat of a small fire. Some specimens were well dried and some were not. In general, the length of a complete tablet is approximately 14.5 centimetres, the widest width is around 8.9 centimetres, and the diameter of the inscription area is around 5.1 centimetres (Figure 1).²⁴ The thickness is around 4.1 centimetres and the sealing weighs roughly 400 grams. The sealings presented in Figures 2 to 9 used to be preserved at the Bujang Valley Archaeological Museum in Kedah. Another three fragments are kept at the Kota Kayang Museum in Perlis (the largest fragment among them is presented in Figure 10). As mentioned earlier, an additional six of them are preserved at the British Museum, but similar specimens could be kept at other museums as well.²⁵

In terms of age, these pear-shaped Perlis Bodhi sealings exhibit certain characteristics of a group of sealings classified by Cœdès as the Type II Phra Bimb (or Phra Phim, พระพิมพ์, literally ‘sacred imprint’) among the sealings found in Siam.²⁶ According to Cœdès, this group of sealings is found mainly in the caves of the Malay–Thai Peninsula and most of them are made of clay, are pear-shaped, and feature the characteristics of Mahāyāna Buddhism. Inscriptions, usually stamped on the reverse sides of the tablets, consist of the ye dharmā formula in the Nāgarī script. Based on the stylistic ground of the art of Śrīvijaya and the script used, Cœdès argued that they could be dated to the tenth century CE.²⁷ Syed Ahmad Jamal and Othman Mohd. Yatim also indicate that the Perlis Bodhi sealing is an example of Śrīvijaya art in Peninsular Malaysia.²⁸ The art of Śrīvijaya might indicate an Indianised art style resembling the Pāla art of Bengal.²⁹ Even though the Perlis Bodhi sealings were not part of the Cœdès analysis above, based on artistic styles, the design of the Perlis Bodhi sealing points date to around the tenth century CE.

On palaeographic grounds, the script used in the Perlis Bodhi sealings shows closer affinity to the Siddhamātṛkā, which is, at times, referred to as ‘early Nāgarī’.³⁰ The script of the sealings appears to closely resemble the Siddhamātṛkā script of the Vimaloṣṇīṣa-dhāraṇī that was stamped on the clay sealings from Pejeng Village in Bali, Indonesia.³¹ On palaeographic grounds Arlo Griffiths estimates that the Bali sealings date to between 800 and 1000 CE.³² Another similar script sealing, also from Pejeng Village in Bali but with a ye dharmā inscription, now preserved in the National Museum of Thailand in Bangkok, is also estimated to date to between the ninth and eleventh centuries CE.³³ Siddhamātṛkā is a script that was developed in about the seventh to tenth centuries CE around northern India and was used as an epigraphic script not only in Northern and Eastern India, but also in the west and even the south, such as in the eighth-century Paṭṭadakal Pillar biscript inscription detailing the donation of a Śiva temple.³⁴ In Southeast Asia, this script is not widely seen in inscriptions found in the Malay–Thai Peninsula or Mainland Southeast Asia.³⁵ However, it seems to be more frequently used in (Java) Indonesia.³⁶ Therefore, based on palaeographic grounds, in the absence of carbon dating, the Perlis Bodhi sealings could be dated to the period around the tenth century CE or slightly earlier. It is also noteworthy that the sealing stamp could have been made earlier than the sealings themselves. As such, the sealings could be far younger than the palaeographic dating, as the stamp could have been used over a longer period of time.

Readings of the Perlis Bodhi sealing

The inscription of the Perlis Bodhi sealing is presented in 16 lines within a circular cartouche in the Sanskrit language. The readings collated from the nine sealings (Figures 2 to 10) were transcribed as follows.³⁷

Transcription:

The parentheses indicate that the reading is close but not entirely clear or is partial on extant fragments of the sealings.

Edition:

namo bhagavate vipula-vadana-kāṃcanotk[ṣ]ipta-prabhāsa-ketu-mūrdhe tathāgatāya | namaḥ śākyamunaye | oṃ bodhi bodhi bodhi bo[dhi] | sarve tathā[ga]tagocare | dhara dhara hara hara prahāra mahābodhicittadhare | culu 2 śataraśmisaṃcodite sarvatathā[gata]bhāṣite | guṇa guṇavate³⁸ | buddhaguṇāvabhāse mili gaganatala-³⁹ sarvatathā[gatā]dhiṣṭhite | nabhastale | śama praśama sarvapāpapraśamane sarvapāpaviśodhane hulu 2 bodhimārgasaṃprasthite sarvatathāgataprastiṣṭhite śuddhe svāhā | namaḥ sarvatathāgatavyavalokite.

According to Lawson, the syntax of this basic mantra (from oṃ bodhi to the end), which sings praises, seems to be composed in feminine vocative (although not entirely the case here).⁴⁰ However, for consistency, we read the dhāraṇī in this manner and offer the following translation:

Homage to the Blessed One, the Tathāgata, whose broad face⁴¹ glows with a golden radiance and is surmounted by radiant flames⁴²

Homage to the Sage of the Śākyans

Oṃ! Awaken awaken awaken awaken!

O you who are the domain of all Tathāgatas

Hold hold! Take take! Strike!

O holder of the thought of the great awakening

Culu culu!⁴³

You, who are impelled by a hundred rays

Who is spoken of by all Tathāgatas

O Virtue!

O you who possess virtues⁴⁴

One who manifests the virtue of a Buddha

Mili!

O you [who are established in] the vault of the sky

Sustained by all the Tathāgatas

You are the overarching canopy of the sky!

Calm, calm! Appease!

O you who appease all evil! O you who cleanse all evil!

Hulu hulu!

O you who set out on the path of awakening

The one who is established by all Tathāgatas

O pure one, Svāhā!

Homage to you

The one who is gazed upon by all Tathāgatas.

Lawson is of the opinion that the basic mantra is addressed to a female deity who is the personification of prajñāpāramitā.⁴⁵ If it is in this context, the praise could be directed at prajñāpāramitā, the ideal of the Perfection of Wisdom itself, which is also in feminine gender. Jacob Kinnard argued that this dhāraṇī described by Lawson in clay sealings is a ‘textual object’.⁴⁶ He added that ‘stamping such verses in clay, carrying them about, and ensconcing such dhāraṇī stones in stūpas constitute the habitus of the sculptural representation of texts’. He was alluding to the fact that this dhāraṇī, the Bodhigarbhālaṅkāralakṣa-dhāraṇī, is a kind of Prajñāpāramitā dhāraṇī.⁴⁷

Comparative structure of the dhāraṇī

In terms of the structure and content of this Perlis dhāraṇī, as we can see above, it begins by paying obeisance to the Buddha and is followed by the root mantra (mūlamantra) starting with oṃ bodhi until śuddhe svāhā. This dhāraṇī ends with a single line of the so-called ‘heart’ (hṛdaya) of the mantra but without the ‘lesser heart’ (upahṛdaya) section (see below).⁴⁸ This structure will become clearer if we compare the Perlis dhāraṇī to a slightly longer version preserved in a twelfth- to thirteenth-century Nepalese Sanskrit palm-leaf manuscript, MS Add. 1680.8.3, which is now kept at the Cambridge University Library.⁴⁹ In this manuscript, the section that contains this dhāraṇī is called the ārya-lakṣa-nāma-dhāraṇī (the Exalted Dhāraṇī Named A Hundred Thousand), which is embedded in the group of manuscripts called the Dhāraṇīsaṃgraha (Dhāraṇī Collection).⁵⁰ This ārya-lakṣa-nāma-dhāraṇī is presented as the Buddha speaking to his attendant monk, Ānanda, and it can be divided into three parts. First, is the mantra itself, which closely resembles the dhāraṇī on the Perlis Bodhi sealings. The second part is the Buddha describing the merit of writing this dhāraṇī and interring it into a caitya. The last part is the Buddha indicating for whose sake he declared this dhāraṇī. This three-part dhāraṇī is also found elsewhere but with varying degrees of dissimilarities, such as in various Tibetan Kanjurs,⁵¹ on a stone slab inscribed in Sanskrit that was previously housed in the Provisional Museum in Cuttack, but now preserved in the Odisha State Museum in Bhubaneswar, India,⁵² as well as in the Chinese adaptation of this dhāraṇī.⁵³ The Tibetan critical edition of this dhāraṇī edited by Gregory Schopen has additional Parts 4 and 5. Part 4 is about the consequences of not reading and worshiping this dhāraṇī and Part 5 is a conclusion translated by Schopen as ‘The Dhāraṇī of the Hundred Thousand taken from “The Hundred Thousand Ornaments of the Essence of Awakening”’. In fact, according to the Tibetan tradition, this ‘longer’ version is just an ‘extract’ of a much longer sūtra that was also taught by the Buddha himself.⁵⁴

The three-part dhāraṇī of the Nepalese manuscript can be presented and translated as follows. The Sanskrit text here is based on the version edited by Gergely Hidas except for a part that has the word hṛdaya where we think a sentence could have been omitted before the word and, because of that, we have selected the reading from the Cuttack stone lab.⁵⁵

The Sanskrit text in the manuscript:

[1]

[18]⁵⁶ namo bhagavate vipula-vadana-kāñcanotkṣipta-prabhāsaketumūrdha-tathāgatāya namaḥ śākyamunaye [tathāgatāyārhate samyaksambu]ddhāya | tadyathā bodhi 2 sarva-tathāgatagocare dhara 2 hara 2 mahābodhi-pratiṣṭhite | mahābodhi-cittadhare culu 2 sahasrara[śmi]saṃcodite sarvatathāgatābhiṣikte guṇi guṇavati | buddhaguṇāva-*bhāse | gaganatalapratiṣṭhite sarvatathāgatādhi[ṣṭhi]te svāhā | nabhastale śama 2 praśama 2 sarvapāpa-*⁵⁷praśamane | sarvapāpaviśodhane hulu 2 mahābodhimārgasaṃprasthite | sarvata[thāgatapratiṣṭhi]te śuddhe svāhā | [mūlamantraḥ | sarvatathāgatagocaravyalokite⁵⁸ jaya jaya svāhā |]⁵⁹ hṛdayam |

kuru 2 jayamūle svāhā | upahṛdayam |

[2]

yaḥ kaścid bhikṣur vā | bhikṣuṇī vā | upāsako vā upāsikā vā | anyo vā | yaḥ śrāddhaḥ kulaputro vā | kuladuhitā vā | imāṃ dhāraṇīṃ likhitvā ekam api caityaṃ kariṣyati tasya caityasyābhyantare | imāṃ dhāraṇīṃ sthāpayiṣyati | tenaikacai<14r>tyaṃ kṛtvā lakṣacaityāḥ kṛtābhavanti | tena sarvacaityāḥ pūjitā bhavanti | divyai gandhadhūpair mālyavilepana-cūrṇacīvaracchattra-dhvajapatākābhiś ca | na kevalaṃ caityāḥ pūjitā bhava++ buddharatnaṃ pūjitaṃ bhavati | evaṃ dharmaratnaṃ pūjitaṃ bhavati | evaṃ saṃgharatnaṃ pūjitaṃ bhavati | vividhair upakaraṇaiḥ pūjitā bhavanti |

[3]

evaṃ mayānanda saṃ[kṣi]ptena deśitaṃ mandapuṇyānām aśraddhānāṃ mithyādṛṣṭikānāṃ vaimatikānāṃ naimittikānām ānantaryakāriṇām alpāyuṣkāṇāṃ narakapreta-tiryagyama++[parāyaṇānāṃ] sattvānām arthāya | ≤ iyaṃ dhāraṇī bhāṣitā || ||

Ārya-lakṣa-nāma-dhāraṇī samāptā || ||

Translation of the Sanskrit text:

[1]

Homage to the Blessed One, the Tathāgata, whose broad face glows with a golden radiance and is surmounted by radiant flames. Homage to the Sage of the Śākyans, the Tathāgata, the Arhat, the Perfectly Awakened One.

[The dhāraṇī] is as follows:⁶⁰

Awaken awaken! O you who are the domain of all Tathāgatas.

Hold hold! Take take! O you who are established in the great awakening!

O holder of the thought of the great awakening!

Culu culu! O you who are impelled by a thousand rays!

O virtue! O you who possess virtue!

O one who manifests the virtue of a Buddha!

O you who are established in the vault of the sky

Sustained by all Tathāgatas, Svāha!

You are the overarching canopy of the sky!

Calm calm! Appease appease!

O you who appease all evil! O you who cleanse all evil!

Hulu hulu! O you who set out on the path of awakening

who are established by all Tathāgatas

O pure one, Svāha! [This is] the root mantra.

You are gazed upon in all the domains of Tathāgatas! Win win, Svāhā! [This is] the heart.⁶¹

Kuru kuru! O you are the root of victory,⁶² Svāhā! [This is] the lesser heart.

[2]

‘Whosoever monk or nun, or lay male or female disciple, or other devout son or daughter of a good family, after having written this dhāraṇī, would also make a single caitya [and] place this dhāraṇī inside that caitya—by having made that single caitya, a hundred thousand caityas [in effect] have been made. By that, all [those] caityas are venerated [as if together] with celestial fragrances, perfumes, flowers, ointments, aromatic powders, robes, umbrellas, flags, and banners.⁶³ Not only are the caityas venerated, [even] the Jewel of the Buddha is [in effect] venerated. In this way [also] the Jewel of the Dharma is [in effect] venerated, [and] the Jewel of the Community is [in effect] venerated by articles of these various kinds.’

[3]

‘Thus, O Ānanda, [this] was taught by me in brief, for the sake of living beings whose merit is weak, who are without faith, who hold wrong views, who are consumed by doubts, who interpret signs and omens, who have committed acts of immediate retribution, who have short lifespans, and who [are destined] to go to the hells, [realms of] hungry ghosts, animals [and the world of] Yama.’⁶⁴ This dhāraṇī has been spoken.

The Exalted Dhāraṇī Named A Hundred Thousand is completed.

As we can see from the reading of the manuscript above, the Perlis Bodhi sealing dhāraṇī contains the root mantra (mūlamantra), which is in close agreement with the manuscript and a portion of the ‘heart’ (hṛdaya) sentence, and it ends there without the ‘lesser heart’ (upahṛdaya). If the Perlis dhāraṇī were to be compared with other sealings carrying the same dhāraṇī, such as those found at Nālandā (Bihar) and Ratnagiri (Odisha) and those in the Eastern Gangetic Plains, then these Indian sealings contain the three sections of the mūlamantra, hṛdaya, and upahṛdaya but do not actually use these three words.⁶⁵ Furthermore, most of these other sealings end with the ye dharmā stanza instead of moving on to Parts 2 and 3 (see below). Therefore, the Perlis dhāraṇī is slightly shorter compared with the others. At the moment, we do not find readings of other sealings that are either identical to the textual tradition of this Perlis dhāraṇī or similar in shape and size to the stamped inscription. As such, these Perlis sealings were most likely produced locally. This is supported by Lamb's observation that the red clay of the sealings appeared to be the same as the material that made up much of the floor of the cave.⁶⁶ Nevertheless, while the sealing might have been a local production, the stamp used for the impression could have been made somewhere else earlier on, as stamps are a highly movable type of artefact.

The merits of stamping this dhāraṇī can be seen in Part 2 of the manuscript reading above. Venerating caityas or stūpas is considered a meritorious deed capable of yielding good karmic fruits in the future. The ritual of venerating caitya has been a focal point of many Buddhist traditions.⁶⁷ In the Tibetan tradition, dhāraṇī is seen as a relic of the Buddha's Dharmakāya (Dharma-body), which is valued as much as or even more than a corporeal relic.⁶⁸ Moreover, as mentioned in the longer dhāraṇī itself, writing this dhāraṇī on a sealing and interring it in a caitya are regarded as venerating the caitya and the three Jewels. The beauty is that, even if it is just a sealing, the merit generated would be multiplied 100,000-fold. Thus, it would be no surprise that those who seek merit, whether for future benefits or for protection, would participate in this ritual of stamping this dhāraṇī on clay tablets and then placing them into a caitya. Examples of this dhāraṇī found in a caitya or stūpa are those sealings found in Nālandā and Ratnagiri (see below).

This ritual of stamping multiple copies of sealings and making aspirations for future benefits can be seen in an archaeological record of a tablet from Pagan, Myanmar, dated circa 1113.⁶⁹ The tablet depicts the Buddha seated with legs suspended, while the verso has nine lines of Pāli incised in Mon script. Luce interprets the inscription as follows:

[This] thera, like Nāgasena, takes keen delight in learning (or who shines out in wisdom). By [this] work [in making] a thousand Buddhas, [may he become] a Buddha in [his] future existence. The work was done by the monk named Sumedha, with his own hand, for the sake of Deliverance.⁷⁰

Sumedha here refers to the monk who stamped those 1,000 tablets of ‘Chitsagôn’. In this case, the merit generated by Sumedha by stamping these multiple tablets was considered vast enough for him to make an aspiration to become a future Buddha.⁷¹ In the Tibetan tradition, the Buddha is recorded to have said to a brahmin (who wishes to obtain a child) that ‘by this Bodhimaṇḍālaṃkāra-nāma-dhāraṇī you will increase the roots of auspicious dharma and all your wishes will be fulfilled’. And, together with it, the Buddha expounded on the ritual and benefits of introducing it in a stūpa.⁷² Similarly, this could be the case for those who want to seek merit of their own with the Perlis Bodhi sealings. They either stamped those sealings themselves or caused it to be done through donations. Moreover, impelled by the notion ‘the more the merrier’, those who could afford it would pay for multiple copies of these sealings to be made. This would partly explain why numerous copies of these sealings were found in the cave.

The next question is: Why are the dhāraṇī sealings found in the cave in Perlis rather than in a caitya? In our opinion, there could be several possibilities. First, there could have been a caitya inside the cave that was made by heaping up soil or mud. The sealings were made on-site and placed inside the caitya or stacked up next to it, but the soil-made caitya did not survive the test of time.⁷³ A second possibility is that there could have been a caitya somewhere outside the cave, but too many sealings were made or sponsored for the caitya and the additions were either brought back to or remained in the cave for keeping, as one could not just throw such sealings away after a ritual. On top of that, an isolated cave that was far away from human activities was considered clean and pure for the keeping of such sealings.⁷⁴ Third, the pear-shaped sealing with a slightly elongated top that ends with a narrow tip might have been meant to emulate a caitya, even though it is in a two-dimensional miniature form. Therefore, stamping the dhāraṇī onto this caitya-like clay sealing would have been considered equivalent to installing this dhāraṇī into a proper caitya.⁷⁵

The title of the dhāraṇī

So far, we have discussed the practical aspects of the dhāraṇī but, before we move on to its transmission, we would like to re-examine its title, which may give us further perspective regarding its purpose. The Sanskrit title Bodhigarbhālaṅkāralakṣa-dhāraṇī seems to be recorded only in the Tibetan texts as the Indic title that is paired with the Tibetan title byang chub kyi snying-po'i rgyan 'bum gyi gzungs, according to a formulaic style that, at least ideally, places the paired Indic–Tibetan titles at the head of every translated text.⁷⁶ Other dhāraṇīs in the Tibetan tradition that carry this mantra (Part 1) but with slight variations are the Bodhimaṇḍalakṣa-dhāraṇī (byang chub snying po 'bum kyi gzugs or PT350) and the Bodhimaṇḍālaṅkāra-nāma-dhāraṇī-upacāra (byang chub kyi snying po'i gzungs-kyi cho-ga or PT555).⁷⁷ The Sanskrit versions of this dhāraṇī found in Indian scripts do not assign a name to this dhāraṇī, except for the Nepalese manuscript, which, as we have seen, calls it the Ārya-lakṣa-nāma-dhāraṇī (the Exalted Dhāraṇī Called a Hundred Thousand), and one of the Sanskrit texts in the Siddhamātṛkā script brought by Kūkai to Japan, which records the dhāraṇī under the title 菩提莊嚴陀羅尼 (Dhāraṇī Adorned with Awakening).⁷⁸ Its Chinese adaptation is 百千印陀羅尼經 (Sūtra on the Dhāraṇī of a Hundred Thousand Seals, T1369a and T1369b) (more about these texts below). As we can see, there is inconsistency in assigning (or not assigning) a name to the dhāraṇī. The title Bodhigarbhālaṅkāralakṣa-dhāraṇī is not verified by any Sanskrit evidence, so it is possibly a back-translation by the editors of the Kanjur.⁷⁹ Nor is the compound bodhigarbha independently verified in available lexicons.⁸⁰ The Sanskrit title feels awkward and unnatural, and this enforces the idea that it may be a back-translation.⁸¹

Bodhi—awakening or enlightenment—is a goal and inspiration that framed Buddhist practice over many centuries. It is the motivation that energises an individual's spiritual development through ritual and the contemplative practice of bodhicitta—the aspiration and determination to realise bodhi. If the dhāraṇī proper does not use the title internally or is used to explain the term bodhigarbha, it does employ the word bodhi several times in the mantra (Part 1), as well as the terms ‘great bodhicitta’ (mahā-bodhicitta) and ‘great path to bodhi’ (mahā-bodhimārga). It seems safe to say that the notion of bodhi is a driving force in the dhāraṇī.

What, then, is bodhigarbha, and how does it relate to the dhāraṇī? If an Indic form ever existed, it is rare or absent in extant documents. What sense can we make of the title, either in its Indic form or in Tibetan and its Chinese adaptation? Does the dhāraṇī itself offer any suggestions or solutions? As seen earlier, in Part 2, we read: imāṃ dhāraṇīṃ likhitvā ekam api caityaṃ kariṣyati tasya caityasya abhyantare imāṃ dhāraṇīṃ sthāpayiṣyati… as ‘after having written this dhāraṇī, one would also make a single caitya [and] place this dhāraṇī inside that caitya—by having made that single caitya, a hundred thousand caityas [in effect] have been made’. By that, it means that making a single caitya is equivalent to making an ornament of 100,000 caityas dedicated to awakening. The caitya is the womb or shrine (garbha) that holds the dhāraṇī dedicated to bodhi and hence is a bodhi-garbha in the same way as a caitya that holds or enshrines corporeal relics (saśārīra) is a dhātu-garbha.

The Tibetan title, byang chub kyi snying po'i rgyan ’bum gyi gzungs, consists of standard equivalents such as byang chub as bodhi, rgyan as alaṅkāra (ornament), and ‘bum as lakṣa (100,000).⁸² The difficulty lies in the polysemic term snying po, which has several possible equivalents.⁸³ If the Sanskrit bodhigarbha is not met in Buddhist or Indian literature, in contrast, Tibetan byang chub (kyi) snying po circulated in Tibetan writing, especially of the Nyingmapa school, and was often associated with the Tathāgatagarbha (Buddha nature) literature (in Tibetan, Tathāgatagarbha is de bzhin gshegs pa'i snying po).⁸⁴ But we do not think that the term bodhigarbha in this dhāraṇī is used in this context. As such, we might interpret the title Bodhigarbhālaṅkāralakṣa-dhāraṇī as ‘the dhāraṇī [which when recited, stamped, or copied creates] an array of one hundred thousand caityas dedicated to awakening’.⁸⁵

Transmission of the dhāraṇī

The presence of this Bodhigarbhālaṅkāralakṣa-dhāraṇī in Perlis indicates that the Malay–Thai Peninsula has taken part in the wider world of dhāraṇī transmission, in this case, since towards the end of the first millennium. The transmission of dhāraṇī of different genres is wide and complex, and it is beyond the scope of this article. However, we will conduct a review and general survey on the transmission of this particular dhāraṇī in order to give us an initial idea of its geographical distribution and how Perlis came to be part of it.

At the moment, we do not know when and where this dhāraṇī was composed or was first written down, either as an independent dhāraṇī later compiled to be part of a longer text or originating from a larger text. The basic assumption is that the original language is Sanskrit, which suggests a (greater) Indian provenance. So far, from the information that we have, this Bodhigarbhālaṅkāralakṣa-dhāraṇī, in various degrees of variations in its textual tradition, was widely transmitted, spanning several regions across Asia, in different languages, adaptations, scripts, materials, kingdoms, and cultures (Figure 12 and Table 1). This dhāraṇī is found from Eastern India⁸⁶ to the greater Gandhāra region of Pakistan and Afghanistan. It also moved eastward, perhaps via the Silkroad and Central Asia into the Tarim Basin, heading to central China, Inner Mongolia, and even into the Korean Peninsula and Japan. To the north of India, it was found in the Kathmandu Valley and the Tibet Plateau. To the south-east, the dhāraṇī has found its way to the Malay–Thai Peninsula and farther down to Java Island in Maritime Southeast Asia.

Figure 12. Map showing the provisional transmission of the Bodhigarbhālaṅkāralakṣa-dhāraṇī in different regions in Asia. The shaded areas indicate where the dhāraṇī is known to have circulated. Place names are given in modern ‘international’ forms.

Table 1. Provisional transmission of the Bodhigarbhālaṅkāralakṣa-dhāraṇī in different regions and its various presentations

This table is provisional because we have no doubt that further instances of the dhāraṇī will come to light over the course of time; M = mūlamantra, Hṛ = hṛdaya, U = upahṛdaya, Ye = ends with the ye dharmā stanza, H = height, W = width, T = thickness, D = diameter, NA = not available, p = partial, s = condensed version.

^a One-third is missing from the beginning.

^b Only a few words.

^c Widest width.

^d The inscription starts with the ye dharmā verse.

^e Schopen writes: ‘These tablets are written in ‘early medieval Nāgarī characters’ and probably date from the sixth to about the ninth centuries’—see G. Schopen, ‘The 'Bodhigarbhālaṅkāralakṣa and Vimaloṣṇīṣa Dhāraṇīs in Indian inscriptions: two sources for the practice of Buddhism in medieval India’, in Figments and Fragments of Mahāyāna Buddhism in India, More Collected Papers, (ed.) G. Schopen (Honolulu, 2005), p. 331.

^f There is no colophon to indicate when the text was translated from Sanskrit to Tibetan (Schopen, ‘The Bodhigarbhālaṅkāralakṣa and Vimaloṣṇīṣa Dhāraṇīs’, p. 318).

^g Text illegible.

^h Parts of the text were reconstructed by I. Strauch, ‘Seals, sealings and tokens from Gandhāra. By Aman Ur Rahman and Harry Falk. (Monographien zur indischen Archäologie, Kunst und Philologie, Band 21). Pp. 222. Weisbaden, Reichert, 2011’, Journal of the Royal Asiatic Society 22.3–4 (2012), pp. 605–606, based on the estimated size of the seal and the extant reading of the broken seal.

ⁱ S. van Schaik, ‘The Tibetan Dunhuang manuscripts in China’, Bulletin of the School of Oriental and African Studies 65.1 (2002), p. 129.

^j Between the dhāraṇī and the ye dharmā (Pratītyasamutpādahṛdaya) is the Uṣṇīṣavimala-dhāraṇī. The ye dharmā formula is the penultimate formula in PT350.

^k A longer Tibetan translation from Sanskrit. An edition of the Bodhimaṇḍālaṃkāra-nāma-dhāraṇī-upacāra (PT555) is being prepared (C. Scherrer-Schaub, ‘Some dhāraṇī written on paper functioning as dharmakāya relics: a tentative approach to PT 350’, in Tibetan Studies, Proceedings of the 6th Seminar of the International Association for Tibetan Studies, Fagernes 1992, (ed.) P. Kvaerne (Oslo, 1994), ii, p. 714).

^l In this Korean manuscript, the 百千印陀羅尼經 is longer compared to T1369a, and it could be in combination with other texts.

India

In India, this Bodhigarbhālaṅkāralakṣa-dhāraṇī was discovered mainly in Odisha and the Eastern Gangetic Plain. In Odisha, stone slabs and sealings carrying this dhāraṇī were found concentrated in an area now called the Buddhist Diamond Triangle of Ratnagiri-Udayagiri-Lalitagiri archaeological complex.⁸⁷ One stone slab with unknown provenance, previously preserved in the Cuttack Provisional Museum in Odisha (thus the name ‘Cuttack stone slab’), was dated approximately to the tenth century, with partial records of the mantra (Part 1) in Sanskrit including Parts 2 and 3, and a portion of Part 4.⁸⁸ Another stone slab inscription dated to around the ninth century, also in Sanskrit, was found in the Caityagṛha complex of Udayagiri. This stone slab bears Part 1, the mantra, and ends with the ye dharmā formula.⁸⁹ On the eastern side of this complex, a further stone slab bearing a 21-line inscription was also recovered. The inscription starts with the ye dharmā verse, followed by the Bodhigarbhālaṅkāralakṣa-dhāraṇī (Part 1) and two other dhāraṇīs.⁹⁰

On another site, just a few kilometres to the east of Udayagiri, four terracotta sealings stamped with this dhāraṇī dated to the ninth to tenth centuries were found in Stūpa 2 in Ratnagiri and another four oval-shaped terracotta sealings, also dated to around the ninth century, if not earlier, were recovered in Stūpa 253 nearby. The dhāraṇī of Stūpa 253 sealings also ends with the ye dharmā stanza.⁹¹

In the Eastern Gangetic Plain, in the modern states of Bihar and Uttar Pradesh, several terracotta sealings carrying this dhāraṇī were discovered in Nālandā, Śrāvastī, and other areas. In Nālandā, two terracotta sealings with legible inscriptions stamped with this dhāraṇī were recovered among 603 sealings in Site 2 of the southern monastery.⁹² According to Schopen, these two sealings, 2-42 and 2-193, were written in early medieval Nāgarī script and probably date from the sixth to the ninth centuries.⁹³ Ghosh provided the readings of these two sealings earlier, but the dhāraṇīs in them end differently; Sealing 2-193 ends with the ye dharmā formula, while 2-42, after the upahṛdaya section, has instead a condensed version of Part 2 with just one sentence, eka-caitya-kṛtena lakṣā(kṣaṃ) kṛtā bhava[n]ti (by making one caitya, 100,000 caityas are made), then followed by the word pratītya-samutpādatā with five illegible letters after that.⁹⁴ As the ye dharmā formula is considered to be the heart (hṛdaya) of pratītya-samutpāda or the law of Dependent Origination,⁹⁵ sealing 2-42 also ends with the essence of the ye dharmā but in a different way.

During his excavations at Śrāvastī in Uttar Pradesh, Alexander Cunningham found some burnt sealings in Stūpa No. 5, and one of them has a small stūpa image and 18 lines of writing on it. Originally, this sealing was interred inside a miniature stūpa but it was disinterred during the excavation process.⁹⁶ At that time, Cunningham managed to read only some parts of that sealing, which was stamped with the Bodhigarbhālaṅkāralakṣa-dhāraṇī. This sealing is now preserved in the British Museum. In the British Museum register, on the same page as the entry of the Śrāvastī sealing, there are also another four entries for sealings with this dhāraṇī but of unknown provenance. They are slightly different in design compared with the Śrāvastī one but they also have a small stūpa image in the middle of the tablets. These five sealings were recorded on the same page, coincidentally or not, and this has led Lawson to suggest that these four sealings might have come from the same area—and that is in Śrāvastī.⁹⁷ All of them end with the ye dharmā formula. The British Museum preserves another six sealings of unknown provenance that are stamped with this dhāraṇī.⁹⁸ Hidas indicates that these six sealings might have come from the Eastern Gangetic Plain.⁹⁹ Further to that, in the Ashmolean Museum in Oxford, there is also another group of six oval-shaped sealings stamped with this Bodhigarbhālaṅkāralakṣa-dhāraṇī but regrettably they are also of unknown provenance. One of them (Museum no. 2330) was read by Lawson and it also ends with the ye dharmā stanza.¹⁰⁰ Based on the paleographic grounds, the sealings kept at these two British museums were estimated to date from the eighth to eleventh centuries.¹⁰¹

From the accounts above, it is likely that, by the second half of the first millennium to the early centuries of the second millennium, perhaps peaking between the eighth and eleventh centuries, Buddhist communities in Odisha and the Eastern Gangetic Plain in India were quite familiar with this dhāraṇī and its ritual. However, it is noteworthy that at least three types of dhāraṇī sealings have been recovered from a stūpa site in Chotti Barauni, south of Gwalior in central India (Madhya Pradesh). Unfortunately, they are much abraded and have not been read so far. Although the dhāraṇī cannot be identified, these sealings show that the practice of stamping dhāraṇī was also followed in the Vindhyas.

Nepal

Farther to the north of the Gangetic Plain in Nepal, as mentioned earlier, the Bodhigarbhālaṅkāralakṣa-dhāraṇī is found in a bundle of 46 palm leaves called the Dhāraṇīsaṃgraha, in Ms. Add. 1680.8. There are around 59 different texts in this manuscript bundle. On a paleographical basis, this manuscript dates to the twelfth to thirteenth centuries.¹⁰² According to Hidas, manuscript Add. 1680.8 carries the features of both Eastern Indian and Nepalese manuscripts. While the hook-topped script is likely to indicate Nepalese origins, the two surviving illustrations including the string-hole decoration point towards Eastern India.¹⁰³ This suggests that there was mutual influence or exchange between the Kathmandu Valley and Eastern India with regard to this dhāraṇī collection. This is very likely, as we have seen earlier that Eastern India was familiar with dhāraṇī culture, too.

Tibetan plateau

In Tibetan traditions, the Bodhigarbhālaṅkāralakṣa-dhāraṇī, under the Tibetan title byang chub kyi snying po'i rgyan ’bum gyi gzungs, is preserved in various versions in different Tibetan Kanjur editions, such as in the Derge, Lhasa, and Peking editions. Schopen has edited this dhāraṇī based on these editions, including the Sanskrit readings from the Cuttack stone slab and the Nālandā sealings. He has discussed at length this dhāraṇī in his work that was first published in Wiener Zeitschrift für die Kunde Südasiens in 1985, vol. 29 (pp. 119–149) and was reproduced in 2005 (pp. 314–344); we refer to the latter version here. Moreover, Cristina Scherrer-Schaub also published further work on this dhāraṇī in the Tibetan tradition in 1994 (pp. 711–727). As such, we will not discuss further the dhāraṇī in the Tibetan tradition here except by mentioning briefly the possible date of its transmission.

Unfortunately, according to Schopen, the Tibetan Kanjur editions that he consulted do not have a colophon to indicate when this dhāraṇī was translated from Sanskrit into Tibetan.¹⁰⁴ Nevertheless, the earliest possible indication that this dhāraṇī was known to the Tibetan tradition is through its listing in the Lhan kar ma catalogue, which was compiled in around the first quarter of the ninth century, as 'Phags pa byaṅ chub sñiṅ po rgyan 'bum gyi gzuṅs in 250 Ślokas.¹⁰⁵ A further sign indicating this date is the discovery of the Tibetan Dūnhuáng manuscripts of the Pelliot collection that bear this dhāraṇī.¹⁰⁶ This group of Dūnhuáng manuscripts is estimated to date from the eighth to ninth centuries.¹⁰⁷ Therefore, if the date estimation is correct, then the Bodhigarbhālaṅkāralakṣa-dhāraṇī would have circulated in central Tibet, or at least have been attested there, in the eighth to ninth centuries, if not earlier.

Pakistan (Gandhāra)

In 1996, some 20 clay sealings and miniature stūpas were recovered from a site in Hund and among them are fragments of three sealings stamped with the Bodhigarbhālaṅkāralakṣa-dhāraṇī. Hund is a Gandhāran archaeological site, situated on the right bank of the river Indus in the Swabi district of Khyber Pakhtunkhwa in present-day Pakistan. On paleographical grounds, they are estimated to date from the seventh to eighth centuries.¹⁰⁸ As the sealings are broken and certain parts of their surfaces are erased, it is difficult to ascertain whether the stamped dhāraṇī ends with the ye dharmā formula. Besides these clay materials, a broken piece of metal seal measuring 3.9 by 3.1 by 2.0 centimetres that bears the Bodhigarbhālaṅkāralakṣa-dhāraṇī was also recovered in the Khyber Pakhtunkhwa region. The script is identified as the Śāradā and is estimated to date to around the sixth century.¹⁰⁹ However, only about a quarter of the seal is extant. Based on the readings available, and assuming the unbroken seal to have been slightly oval-shaped, Strauch has reconstructed the text of the seal. Part of the surviving portion and the reconstructed text indicates that the dhāraṇī ends with a ye dharmā stanza.¹¹⁰

Afghanistan (Bactria)

In the north of Afghanistan, perhaps in the region of Qunduz, two steatite rectangular stamps, measuring 5 by 6 centimetres, were recovered. One is now preserved in the British Museum (OA 1880.168)¹¹¹ and another is in a private collection. Strauch has reviewed these two stamps in detail and has provided the readings of the Bodhigarbhālaṅkāralakṣa-dhāraṇī on the stamps, which also end with the ye dharmā formula.¹¹² A suggestion has been made that the British Museum stamp is of Kashmiri origin but Strauch argues that, based on the characteristics of the stamp and also the lack of its acquisition information, the suggestion is difficult to sustain. In turn, he is of the opinion that both these stamps hailed from north Afghanistan, and perhaps even from the same workshop tradition.¹¹³ He further hypothesises that the script used in these stamps is of those in the transitional phase from late Gupta Brāhmī to Proto-Śāradā, thus dating the stamps to the time from the middle of the sixth to the beginning of the seventh centuries.¹¹⁴ If this is the case, the Afghanistan stamps, perhaps together with the broken metal steal mentioned above, are among the earliest surviving artefacts of this dhāraṇī.

China

Moving east from Central Asia, tracking by the edges of the sandy Taklamakan desert, the dhāraṇī found its way into the Tarim Basin. A small fragment bearing the Sanskrit version of the Bodhigarbhālaṅkāralakṣa-dhāraṇī in Uyghur letters was recovered from Turfan oasis. It is now preserved in the Turfan Collection of Berlin (Ch/U 6357b verso).¹¹⁵ Unfortunately, the other half-page of the fragment is missing. The surviving portion shows that the dhāraṇī ends in the mūlamantra section. The date of the fragment at the moment is unclear.

Farther down from Turfan to the south-east, in another oasis containing a crescent lake and singing sand dunes, are the Dūnhuáng (敦煌) caves that once preserved the Tibetan paper scroll manuscripts that carry this Bodhigarbhālaṅkāralakṣa-dhāraṇī. The texts that bear this dhāraṇī in these scrolls, as mentioned earlier, are entitled byang chub snying po 'bum kyi gzungs (Bodhimaṇḍalakṣa-dhāraṇī)¹¹⁶ and a longer byang chub kyi snying po'i gzungs-kyi cho-ga (Bodhimaṇdālaṅkāra-nāma-dhāraṇī-upacāra), or PT350 and PT555, respectively, in the Pelliot Tibetan collection now preserved in Paris.¹¹⁷ They also probably date to the eighth or ninth centuries. At the moment, we do not know whether these manuscripts were brought from central Tibet or translated from Sanskrit into Tibetan in Dūnhuáng.

Moving farther down the south-east direction via the Héxī Corridor (河西走廊), the dhāraṇī reached the ancient capital of China—Cháng'ān (長安), present-day Xī'ān (西安). The Chinese adaptation of the Bodhigarbhālaṅkāralakṣa-dhāraṇī appears as 百千印陀羅尼經 (Bǎi qiān yìn tuóluóní jīng or Sūtra on the Dhāraṇī of a Hundred Thousand Seals) in the Taishō Tripiṭaka (T), T1369a based on the Koryŏ or Gāolí (高麗) edition compiled in 1151 CE, and T1369b based the Míng (明) edition compiled in 1601 CE. In the record, 百千印陀羅尼經 was translated from Sanskrit by Śikṣānanda (實叉難陀) between the first year of Zhèng Shèng (證聖) and the fourth year of Cháng'ān during the Táng Dynasty (唐), and that was between 695 and 704 CE.¹¹⁸ According to the footnote in T1369b, because of the differences between these two versions (even though they both seem to have been translated by Śikṣānanda), they are reported separately in the Taishō Tripiṭaka. Both of these Chinese versions only have Parts 1 and 2 of the dhāraṇī text. Similarly to the Tibetan versions and the Turfan fragment, the mantra section of the dhāraṇī in 百千印陀羅尼經 was not translated into Chinese, but was transposed from the Indian writing system into the Chinese in an attempt to maintain the sounds of the original. The Chinese adaptation continues with the same message of writing this dhāraṇī and placing it inside a caitya. In this case, the Chinese text has the word tǎ (塔), which generally refers to a pagoda. However, stamping this dhāraṇī on clay sealings did not seem to have become common practice among Chinese Buddhists. Instead, they either hand-copied the dhāraṇī onto paper or block-printed multiple copies.

Far to the north-east of Cháng'ān, a Sanskrit paper fragment of a block print bearing this Bodhigarbhālaṅkāralakṣa-dhāraṇī was recovered in Khara-Khoto (黑水城) in Inner Mongolia.¹¹⁹ It was printed in an Indian script and Olga Lundysheva assigned it as a Pāla script.¹²⁰ However, only half of this white-paper block print survives. Nevertheless, it is enough to deduce that it contains Part 1 of the dhāraṇī, or up to the upahṛdaya section. Interestingly, the text was written on the upper portion of the block, occupying around a quarter to a third of the printed area. Below the printed text is decorated with multiple stūpa-shaped designs. This pattern could symbolise the power of this dhāraṇī—that is, a written dhāraṇī in a stūpa is equivalent to having built multiple stūpas and the merits that come along with it. Based on the type of paper, characters, and features of this block print, Lundysheva estimates that it dates to around the twelfth century.¹²¹

Korea

As we have seen above, one version of 百千印陀羅尼經, T1369a, listed in Taishō Tripiṭaka, is based on the Korean Koryŏ (高麗) edition. The Chinese started to catalogue and compile Buddhist texts and treatises as early as the fourth or fifth centuries, which led to the increasing formation of notions of canon and canonicity. The initial circulation of the Chinese Buddhist canon was rather limited, as it was copied by hand. However, towards the end of the first millennium, the first printed version of the Chinese canon called the Kāi Bǎo Zàng (開寶藏) by the Song Dynasty became available and was distributed widely in China as well as in Korea and Japan.¹²² The first Korean Koryŏ (高麗) edition, which was based on the Song's Kāi Bǎo Zàng, was then compiled and finished in around the eleventh century.¹²³ The 百千印陀羅尼經, which was part of the Kāi Bǎo Zàng, was then transmitted to Korea. Besides being preserved in the Koryŏ canon (Tripiṭaka Koreana),¹²⁴ the 百千印陀羅尼經 (Baeg-cheon-in-dalanigyeong) was also found in the golden manuscript commissioned by the Koryŏ royal court in the thirteenth century.¹²⁵

Japan

Farther to the east on the islands of Japan, at around the beginning of the ninth century, Kūkai (空海) brought 42 Sanskrit texts written in Siddhamātṛkā script (御請來目錄,Go-shōrai mokuroku) back from Táng China after two years of sojourn and, in 806 CE, he presented a catalogue of these texts to the imperial court.¹²⁶ He also brought back notebooks of copied texts—the Sanjūjō sasshi (三十帖冊子). One of the 42 Sanskrit texts, 菩提莊嚴陀羅尼 (Bodai shōgon darani), numbering 29 in his notebook, bears the mantra of the Bodhigarbhālaṅkāralakṣa-dhāraṇī in Siddhamātṛkā script.¹²⁷

Based on these short accounts, the adaptations of Bodhigarbhālaṅkāralakṣa-dhāraṇī were transmitted to China in around the seventh to eighth centuries and, at the turn of the first millennium, the dhāraṇī and its adaptations were widely distributed in China, including Korea and Japan.

Southeast Asia

On the other side of India, towards the south-east direction, perhaps sailing down along the coast facing the Bay of Bengal and through the Strait of Melaka, the Bodhigarbhālaṅkāralakṣa-dhāraṇī made its presence in the Island of Java, Indonesia. In 1993, a gold foil, measuring 5.2–5.8 by 25.5 centimetres and weighing 7.9 grams, with five lines of Siddhamātṛkā script, was discovered at the south-east corner of the central chamber of the northern Candi Plaosan Lor main temple in Central Java.¹²⁸ The gold foil, datable to the eighth and mid-ninth centuries, was inscribed with the Bodhigarbhālaṅkāralakṣa-dhāraṇī and Griffiths has provided its reading and translation.¹²⁹ The textual tradition on this gold foil is in agreement with the Perlis reading but with some differences; for example, the gold-foil version has a shorter text on paying obeisance to the Buddha and it finishes with the upahṛdaya section while the Perlis one ends earlier at the hṛdaya section. Therefore, even within Southeast Asia, the textual tradition of the dhāraṇīs varies. Indonesia is home to numerous archaeological sites where a variety of Buddhist dhāraṇī, mantra, and gāthā inscriptions were recovered.¹³⁰ However, at the moment, the Candi Plaosan Lor gold foil is the only epigraphic material known to bear the Bodhigarbhālaṅkāralakṣa-dhāraṇī.

Śrīvijaya was the leading power in the straits around the late seventh to eleventh centuries. It exerted its dominance over a string of coastal and riverine settlements in the Malay–Thai Peninsula, most of which were part of the Ancient Kedah polity. Its main port located in the Bujang Valley came under the political, economic, and cultural influence of Śrīvijaya, possibly serving as one of the main centres for exchange and production. Numerous premodern cultural remains related to Buddhism were discovered in Ancient Kedah, including images, stūpas, and inscriptions containing Buddhist texts. We have mentioned earlier that the Perlis sealings exhibit certain features of Śrīvijaya art that, in turn, resemble the Pāla Indian art style. Several scholarly works have been published on the historical artistic relationships between South and Southeast Asia, especially the link between the Pāla regions (750–1174 CE) and Śrīvijaya in their mutual exchanges as well as local adaptation in the development of art styles.¹³¹ There was also a strong religio-political connection between the Pāla Kingdom and Śrīvijaya, according to Hermann Kulke:

The P[ā]la [K]ingdom of Bihar and Bengal in the ninth and tenth centuries was the most powerful state of northeast India, and its Buddhist art strongly influenced Śailendra art and architecture in Central Java …. The donation of a vihāra (monastery) at Nalanda (in Bihar) was certainly the best choice for [Śailendra Dynasty's King] Bālaputra's (reigned from the mid-ninth century) ritual policy, enhancing not only his status as a newcomer in Śrīvijaya, but also Śrīvijaya's fame in Buddhist Southeast and East Asia.¹³²

In view of this, the Śrīvijaya features of the Bodhi Perlis sealing could safely indicate its root to be Eastern India.¹³³ The sealing also exhibits the characteristics of local adaptation; for example, the Bodhigarbhālaṅkāralakṣa-dhāraṇī was stamped on a pear-shaped sealing, and this design combination is not found in any other regions.

The influence of Northern and Eastern India on Southeast Asia goes beyond just artistic materials, extending to ritual practice, too. Skilling, through the example of the ye dharmā stanza, states that

[r]itual practices of enlisting the stanza developed across South and Southeast Asia, peaking in the late Pāla period, during which the stanza became integral to the installation of relics and the consecration of sites, painted scrolls, manuscripts, and images of Buddhas, bodhisattvas, and other figures. The practice seems to have been largely restricted to Northern India and to Southeast Asia.¹³⁴

Even though the Perlis Bodhi sealing does not end with the ye dharmā stanza similarly to sealings found in India, it is found in the vicinity of sealings with the ye dharmā stanza. For example, the sealing with an image of a Buddha seated in dharmacakra mudra surrounded by eight Bodhisattvas stamped with five ye dharmā impressions on its reverse side was also found in Perlis.¹³⁵ Near Perlis, the ye dharmā stanzas are found inscribed on other materials in Kedah, such as the Bukit Choras stone tablet and the Bukit Meriam tablet.¹³⁶

Another shared ritual practice between Eastern India, including Odisha and Southeast Asia, with reference to the Perlis Bodhi sealing was the stamping of dhāraṇīs on prepared clay surfaces. We have discussed earlier that the Bodhigarbhālaṅkāralakṣa-dhāraṇī sealings were found in Odisha and over the Eastern Gangetic Plain. Stamped sealings of other types of dhāraṇī, such as the Vimaloṣṇīṣa-dhāraṇī, were also widely found across northern India.¹³⁷ Generally, the letters on these sealings are small or even tiny, making the dhāraṇī difficult to read. More often than not, they were not meant to be read; rather, their production was driven by the quest to accrue merit. Skilling points out that ‘Merit is a much-desired personal and social commodity and a currency of multiple value systems. A donor gained physical and spiritual blessings in the present life and in future rebirths’.¹³⁸ Further to this, the merit made is believed to be transferrable to family members, including departed relatives and those beyond the family circles, too. Moreover, this merit-making ritual is not known to be restricted to only those who regarded themselves as Buddhists. Therefore, the ritual of stamping dhāraṇī could have been a means for travellers in Perlis, merchants, residents, and seafarers to seek merit at that time, either for blessing, safety at sea, protection, or the granting of boons for themselves or those they cared for.

Conclusion

The discovery of the Perlis Bodhi sealings demonstrates that the Sanskrit Bodhigarbhālaṅkāralakṣa-dhāraṇī circulated in Perlis in around the tenth century. Multiple copies of the dhāraṇī were stamped onto distinctive stūpa-like pear-shaped clay sealings and were stored in a cave. The textual tradition of the Perlis dhāraṇī is unique, too, and is generally slightly shorter compared with its counterparts found elsewhere. This indicates that the Perlis sealings could have been designed and produced locally, and are thus standing witnesses to the expressions of a mature and active Buddhist culture in the Malay–Thai Peninsula during that time in the context of Ancient Kedah's development as a port polity. It is not surprising that Perlis, as part of Ancient Kedah, which was strategically located along the ancient maritime trade routes of South and East Asia, participated in a particular dhāraṇī culture that once spanned today's South, East, and Southeast Asia. Exchanges between the regions of Eastern India including Odisha and the Malay–Thai Peninsula, especially with regard to this Bodhigarbhālaṅkāralakṣa-dhāraṇī, are particularly noteworthy. The presence of this type of dhāraṇī ritual demonstrates in part that, at around the turn of the first millennium, people in Perlis sought to accrue merit, too. The ideology of merit-making in Buddhism by way of specific ritual practices is shared across regions and cultures. Even though the rituals might involve different expressions in different locations, the quest for merit in desiring a better life, protection, cleansing of bad deeds, and gaining spiritual attainment has remained a common aspiration to many over the centuries and is still relevant today.