**2. Date and Place of Compilation of the** *D. ak¯ arn ¯ . ava*

S´astr ¯ ¯ı and Chaudhuri presented pioneering studies of the date and place of compilation of the *D. ak¯ ar¯ n. ava*. <sup>14</sup> The *D. ak¯ ar¯ n. ava* includes a number of verses in Apabhram. sa (or a form of Eastern Prakrit rather than Apabhra ´ m. sa according to ´ Tagare15), which had many Bengali words and expressions such as *tumi* (for *tvam*), as found by S´astr ¯ ¯ı and Chaudhuri. Chaudhuri also said that they were based on the East Bengal dialect,16 and he concluded that the Apabhram. sa text in the ´ *D. ak¯ ar¯ n. ava* was composed in Bengal. Chaudhuri also said that the Apabhram. sa language in the ´ *D. ak¯ ar¯ n. ava* was in a "moribund form", and thus was never earlier than the *Dohako´ ¯ sa*s (by which he means no earlier than the 12th century). The Nepalese-German Manuscript Preservation Project (NGMPP) A138/9 can be considered as the oldest among the extant Sanskrit manuscripts of the *D. ak¯ ar¯ n. ava*. Although its date of production is not reported in the colophon, from the letters used in it, S´astr ¯ ¯ı speculated its date of production to be the 12th century, and Chaudhuri hypothesized it to be in the 13th century. As S´astr ¯ ¯ı pointed out, some pages contain marginal notes (mostly transcriptions of chapter titles and mantras in the text) in the Dbu med script or running Tibetan hand. Chaudhuri speculated that this manuscript might have been a Sanskrit document that the translators had used while translating the *D. ak¯ ar¯ n. ava* into Tibetan. It is difficult to decide only with the short marginal notes whether the translators (officially Jaysena and Dharma yon tan) really used this manuscript. However, it is almost certain that this manuscript was not the principal Sanskrit document used for the translation because, as shown in the critical apparatuses in Chapter 5, some differences can be found between this manuscript and the Tibetan translation (although most of them are minor ones).

<sup>14</sup> (S´astr ¯ ¯ı 1915, pp. 165–66; Chaudhuri 1935, pp. 16–20).

<sup>15</sup> (Tagare 1948, p. 20).

<sup>16</sup> For the use of Bengali words and expressions, " ... it contains a number of songs in vernacular. What that vernacular is, it is difficult to say, but I venture to throw out a suggestion that many of them are in Bengali." (S´astr ¯ ¯ı 1915, p. 166) and " ... there we have specially many Bengali words and expressions, e.g., tumi, la¯ı, chay, yemanta, kaj, p ¯ ai, p ¯ uv, ke, juvanisayala, mantasayala, etc., ¯ ... The language of the present work seems to be based on East Bengali dialect. The pronunciation of some of the letters belongs rather to East Bengal than to West Bengal, e.g., the pronunciation of d. a as ra, and consequently, the pronunciation of ra as d. a, are found particularly in East Bengal dialect ... The deaspiration of some aspirated words and consequently the aspiration of deaspirated words show East Bengal tendency . . . " (Chaudhuri 1935, p. 19).

As Maeda pointed out, the names of some Buddhist scriptures are mentioned in Section 4 of Chapter 50 in the *D. ak¯ ar¯ n. ava*. They are the *Samaja ¯* (= *Gudyasamajatantra ¯* ), *Sam. varottara* (= *Sarvakalpasamuccaya*),<sup>17</sup> *Sam. vara* (= *Sarvabuddhasamayoga ¯* ), *Vajrad. aka ¯* , *Abhyudaya* (= *Herukabhyudaya ¯* ), *Tattvasam. graha*, *Vajrabhairava*, *Sam. put.a* (= *Samput.odbhava*), and *Laghusam. vara* (= *Cakrasam. vara*).<sup>18</sup> Among these, the *Vajrad. aka ¯* , *Herukabhyudaya ¯* , and *Samput.odbhava* are the latest tantras. (Among the three, the *Herukabhyudaya ¯* is probably the earliest.) They were composed from perhaps the late 9th to the 10th century CE, sometime between when the *Cakrasam. vara* (around the 9th century CE) was composed and when Bhavabhadra (or Bhavabhat.t.a), Advayavajra, Gayadhara, and Ratn ¯ akara´ ¯ santi flourished (around ¯ the end of the 10th century to the 11th century CE).<sup>19</sup> In fact, the *D. ak¯ ar¯ n. ava* includes various teachings that were originally derived from the scriptures mentioned above and that were more developed than the original teaching in them.20

The passage in Section 4 of Chapter 50 of the *D. ak¯ ar¯ n. ava* examined in the previous paragraph does not mention the *Kalacakratantra ¯* (hereafter *Kalacakra ¯* ). The period of the compilation of the *Kalacakra ¯* can be dated to the early 11th century

<sup>17</sup> Szánt<sup>ó</sup> points out that the term *Sam. varottara* or *Sa´ m. varottara* is a designation for the *Sarvakalpasamuccaya* (Szántó and Griffiths 2015, p. 368).

<sup>18</sup> (Maeda 1988, pp. 70–71). I present the unpublished text of my edition (*D. ak¯ ar¯ n. ava*, 50.4.51c–53): *karma tes.u ca vijñeyam. samaje sa ¯ m. varottare // sam. vare vajrad. ake c ¯ abhyudaye tattvasa ¯ m. grahe / vajrabhairave sam. put.e anyes.a¯m. ca yathavidhi // eva ¯ m. trilaks.abhidh ¯ an¯ ad¯ ak¯ r.s.t.am. laghusam. varam / akar¯ adik ¯ s.akar¯ anta ¯ m. pat.ales.u yathakramam // ¯* ("The ritual regarding those [fifty letters] is to be recognized in the *Samaja ¯* , *Sam. varottara*, *Sam. vara*, *Vajrad. aka ¯* , *Abhyudaya*, *Tattvasam. graha*, *Vajrabhairava*, and, among others, *Sam. put.a* in accordance with the prescription. In the same way, the *Laghusam. vara*, [which was] extracted from the *Abhidhana ¯* of three hundred thousand stanzas, is [connected with the fifty letters] from the letter *a* to the letter *ks. a* in [its] chapters in order"). The *D. ak¯ ar¯ n. ava* also mentions the *Khasamatantra* (50.4.54d).

<sup>19</sup> The *Herukabhyudaya ¯* was translated into Tibetan by Advayavajra. The *Herukabhyudaya ¯* contains a passage that is more archaic than the *Cakrasam. vara* (Sanderson 2009, pp. 213–14; Szántó 2012a, p. 37). However, it is very likely that the compilation of the whole text of the *Herukabhyudaya ¯* postdates the *Cakrasam. vara*: The *Herukabhyudaya ¯* teaches the fundamental mantra of Vajravar¯ ah¯ ¯ı, which cannot be found in the earlier texts of the Sam. vara tradition. The *Vajrad. aka ¯* and the *Sam. put.a* were translated by Gayadhara. Ratn ¯ akara´ ¯ santi mentions the ¯ *Vajrad. aka ¯* and the *Sam. put.a* in his *Muktaval ¯ ¯ı* (Skt ed. (Tripathi and Negi 2001, p. 18, l.8–p. 19, l.9)), a commentary on the *Hevajratantra*. For a recent analysis of the date of Advayavajra, see (Isaacson and Sferra 2014, p. 71). Bhavabhadra (*bha ba bha dra*), who has been said to be identical to Bhavabhat.t.a, the author of the commentaries of the *Catus.p¯ıt.ha* and *Cakrasam. vara*, wrote an extensive *Vivr.ti* commentary on the *Vajrad. aka ¯* (D 1415). Whether they might be the same person or not, it is certain that Bhavabhadra is not after Gayadhara (the first half of the 11th century) ¯ because Bhavabhadra's *Vivr.ti* commentary on the *Vajrad. aka ¯* was translated into Tibetan by Gayadhara. ¯ Isaacson and Sferra analyzed Bhavabhat.t.a (identical to Bhavabhadra) to have flourished "around 1000 CE" (Isaacson and Sferra 2015b, p. 477), and Szántó, in the "first half of the 10th cent" (Szántó 2015, p. 320). According to Szántó, the *Samput.odbhava* is quoted by Durjayacandra, who might have been active around 1000 CE (Szántó, pp. 402–3). See also Sugiki (2007, pp. 14–19) for the chronology of the scriptures belonging to the Sam. vara tradition.

<sup>20</sup> I have clarified this in several publications. See especially (Sugiki 2007, p. 17, chaps 2, 4, 5, and 7; Sugiki 2018a, 2018b). This monograph also deals with some examples.

according to Newman, Wallace, and Sferra.21 However, there are also passages that suggest a possibility that the extant version of the *D. ak¯ ar¯ n. ava* was completed after the appearance of the *Kalacakra ¯* .

As discussed in detail in Chapters 4.1.4, 4.2.2, 4.2.3, 4.3.1, and 4.3.2 in this monograph, the *D. ak¯ ar¯ n. ava* 15, Vajrapa¯n. i's *Laghutantrat.¯ıka¯*, and the *Kalacakra ¯* , as well as Pun. d. ar¯ıka's *Vimalaprabha¯*, include similar lists of females born of various castes (or d. akin ¯ ¯ıs of lineages), animals living on the ground and in water, sky-going creatures such as birds, and kinswomen. The *Laghutantrat.¯ıka¯* is a commentary on the *Cakrasam. vara*, composed partially based on the Kalacakra's system. The compilation ¯ of the *Laghutantrat.¯ıka¯* is datable to around the end of the 10th or the beginning of the 11th century, when the *Kalacakra ¯* assumed its earliest form or was still a work in progress. The *Vimalaprabha¯*, a commentary on the *Kalacakra ¯* , was composed in the first half of the 11th century: it may be contemporaneous with or is slightly later than the *Kalacakra ¯* . 22

Comparison of these four texts shows the following: (1) The lists in the *D. ak¯ ar¯ n. ava* are more extensive than the lists in the *Laghutantrat.¯ıka¯*, the *Kalacakra ¯* , and the *Vimalaprabha¯*. For example, the list of thirty-six ground-going animals in the *Laghutantrat.¯ıka¯* and the *Kalacakra ¯* with the *Vimalaprabha¯* commentary are divided and expanded into the two separate lists of thirty-six animals living on the ground and thirty-six water creatures in the *D. ak¯ ar¯ n. ava*; <sup>23</sup> (2) A confusion or peculiar logic can be found in some of the lists of creatures in the *D. ak¯ ar¯ n. ava*. For example, *bila¯d.¯ı*, which seems a deliberate recasting of *bid. al¯ ¯ı* ("female cat"), appears in the list of sky-going creatures such as birds in the *D. ak¯ ar¯ n. ava*, whereas a "(female) cat" (spelled as *vid. ala ¯* or *bid. al¯ ¯ı*) is included in the list of ground-going animals in the *Laghutantrat.¯ıka¯*, the *Kalacakra ¯* , and the *Vimalaprabha¯*; <sup>24</sup> (3) Some deities' names in the lists in the *D. ak¯ ar¯ n. ava* are more similar to those in the lists in the *Laghutantrat.¯ıka¯* and Pun. d. ar¯ıka's interpretation of those deities' names in his *Vimalaprabha¯* than to those in the lists in the *Kalacakra ¯* ; <sup>25</sup> (4) Some of the names of the human castes and animals assume their Middle-Indic forms in the *D. ak¯ ar¯ n. ava*. For example, Sau ´ n. d. in¯ı in the *Kalacakra ¯* and the *Vimalaprabha¯* is Kalyapalin ¯ ¯ı in the *Laghutantrat.¯ıka¯* and is Kallaval¯ ¯ı (a Prakrit form of Kalyapal¯ ¯ı) in the *D. ak¯ arn ¯ . ava* for the "wine-selling woman".26

<sup>21</sup> (Newman 1998, p. 343; Wallace 2001, pp. 3–4; Sferra 2015, p. 341). However, they investigated the date from slightly different viewpoints.

<sup>22</sup> For the chronology of the *Laghutantrat.¯ıka¯* and the *Vimalaprabha¯*, see particularly (Cicuzza 2001, p. 13; Isaacson and Sferra 2014, p. 97, footnote 18; Sferra 2015, p. 343; Isaacson and Sferra 2015b, p. 477).

<sup>23</sup> For details, see Chapters 4.2.3 and 4.3.2 in this monograph.

<sup>24</sup> For details, see Chapter 4.2.2 in this monograph. See also the *D. ak¯ arn ¯ . ava* 15.128a and its footnote.

<sup>25</sup> For details, see particularly Chapters 4.2.2 and 4.3.2 in this monograph.

<sup>26</sup> For Saun ´ . d. in¯ı–Kalyapalin ¯ ¯ı–Kallaval¯ ¯ı, see the *D. ak¯ arn ¯ . ava* 15.89b and its footnote.

The four results of the comparison mentioned above show the possibility that the lists in the *D. ak¯ ar¯ n. ava* were developed after the lists found in the *Laghutantrat.¯ıka¯* and the *Kalacakra ¯* . It is also worth noting that in Section 7 of Chapter 50 in the *D. ak¯ ar¯ n. ava*, the Lord, who provides the discourse of an aspect of the passage of time, is named "Kalacakra" ("Wheel of Time"). ¯ <sup>27</sup> However, the third result of comparison mentioned above also shows a possibility that the lists in the *D. ak¯ ar¯ n. ava* and the *Kalacakra ¯* were separately and independently evolved from the list in the *Laghutantrat.¯ıka¯* (or the list in some lost text that is similar to the list in the *Laghutantrat.¯ıka¯*.) If this is the case, it is obscure which of the lists in the *D. ak¯ ar¯ n. ava* and the *Kalacakra ¯* were developed earlier.

Isaacson and Sferra state that the *Sam. varodayatantra* (hereafter *Sam. varodaya*) was probably composed in Nepal after the compilation of the *Kalacakra ¯* in India.28 The *Sam. varodaya*, as well as the *Kalacakra ¯* , is not mentioned in the passage in Section 4 of Chapter 50 in the *D. ak¯ ar¯ n. ava* examined earlier. However, this does not mean that the *Sam. varodaya* was not in existence at the time when the *D. ak¯ ar¯ n. ava* was compiled. As I argued in my earlier publication, the discourses of the death signs introduced in the *D. ak¯ ar¯ n. ava* are very likely to be extended or developed versions of the discourses of the same topic found in the *Sam. varodaya*. The teachings of the chakras (*cakra*) and inner channels (*na¯d.¯ı*) in the *D. ak¯ ar¯ n. ava* are more detailed and appear to be more developed than the teaching of the same topic in the *Sam. varodaya*. <sup>29</sup> It seems that the *Sam. varodaya* also antedates the *D. ak¯ ar¯ n. ava*, or at least, it can be stated that the *Sam. varodaya* includes discourses that were composed prior to the parallel or similar discourses found in the *D. ak¯ arn ¯ . ava*. 30

<sup>27</sup> *D. ak¯ ar¯ n. ava* 50.7.61ab [my unpublished edition]): *ity aha bhagav ¯ an k ¯ alacakr ¯ akhya ¯ m.* (for *-khyas*) *tu tathagata ¯ h.* / ("Thus taught the Blessed One, a Tathagata called K ¯ alacakra."). The system of the wheel of time ¯ in the *D. ak¯ ar¯ n. ava* is different from that of the *Kalacakra ¯* . The former system is more in line with the systems traditional in the Sam. vara literature. For details, see (Sugiki 2007, chaps 4, 5 and 7).

<sup>28</sup> (Isaacson and Sferra 2015a, p. 315).

<sup>29</sup> For details of the comparison between these discourses found in the *Sam. varodaya*, the *D. ak¯ ar¯ n. ava*, and other tantras belonging to the Buddhist Yogin¯ıtantra traditions, see (Sugiki 2007, pp. 17–18, 110–13, 257–82, 316–59). Among them, see particularly (Sugiki 2007, pp. 317–21) for the textual relationship between the *Sam. varodaya*, the *D. ak¯ ar¯ n. ava*, and the *Var¯ ah¯ ¯ıkalpa*, where I showed that the text of one of the teachings of the death signs was very likely transmitted directly from the *Sam. varodaya* to the

*D. ak¯ arn ¯ . ava* and from both of them to the *Var¯ ah¯ ¯ıkalpa*. <sup>30</sup> Although it is not strong evidence, the following may support the possibility that the *Sam. varodaya* postdates the *D. ak¯ ar¯ n. ava*: The *Sam. varodaya* was translated into Tibetan (D 373/P 20) by Gzhan la phan pa mtha' yas (Indian) and Rgyus ban de smon lam grags (Tibetan). Their translation was finalized by Gzhon nu dpal (late 14th–15th century), who used Vanaratna's (mid-14th to mid-15th century) Sanskrit manuscript of this scripture. Ratnaraks.ita, who composed the *Padmin¯ı* commentary on the *Sam. varodaya*, was active around 1200 CE. After the downfall of Vikrama´s¯ıla in the beginning of the 13th century, Ratnaraks.ita came to Nepal (Lewis 1996, p. 156). It is not certain when Gzhan la phan pa mtha' yas and Rgyus ban de smon lam grags, who were the two translators of the *Sam. varodaya*, were active. However, it is very likely that Ratnaraks.ita flourished after Padmavajra and Jayasena, who were a commentator and a translator of the *D. ak¯ ar¯ n. ava*, respectively. As discussed below in this monograph, Padmavajra and Jayasena most likely flourished in the 12th century.

Anupamaraks.ita (late 10th to early 11th century according to Sferra)'s commentary on the *Namasa ¯ m. g¯ıti*, entitled the *Amr.tabindupratyaloka ¯* , <sup>31</sup> A´svaghos.a's commentary on the *Mahak¯ alatantra ¯* , entitled the *Rudrakalpamaha´¯sma´sana ¯* (the date of compilation of which is relatively late),32 and Indrabhuti's ¯ *Vr.tti* commentary on the *Cakrasam. vara* (which is one of the latest commentaries on the *Cakrasam. vara*) <sup>33</sup> mention the *Mkha' 'gro rgya mtsho*, *D. a ki rgya mtsho'i rgyud ¯* , *Rdo rje mkha' 'gro rgya mtsho*, and *Badzra da ki rgya mtsho ¯* , the Sanskrit of which may be *D. ak¯ ar¯ n. ava*, *D. ak¯ ar¯ n. avatantra*, and *Vajrad. ak¯ ar¯ n. ava* (the last two), respectively, and quote passages from them. However, in the extant version of the *D. ak¯ ar¯ n. ava*, I cannot find passages that are evidently comparable to those that the three authors quoted under the name of the *Mkha' 'gro*

<sup>31</sup> \**Aryamañju´ ¯ sr¯ınamasa ¯ m. g¯ıti-amr.tabindupratyalokav ¯ r.tti* (*'Phags pa 'jam dpal gyi mtshan yang dag par brjod pa'i 'grel pa bdud rtsi'i thigs pa sgron ma gsal ba zhes bya ba*), D 1396. The translators are *bod kyi lo tsa ba ¯* ("Tibetan translator") Dharmak¯ırti (who is different from the famous logician Dharmak¯ırti around the 7th century), Sugata´sr¯ı, and Rdo rje rgyal mtshan, who was active in the 13th century CE. For the date of Anupamaraks.ita, see (Sferra 2000, p. 48, footnote 101; Sferra 2015, p. 344). In his *Amr.tabindupratyaloka ¯* , Anupamaraks.ita mentions the *Kalacakra ¯* and quotes passages from it several times. He is also the author of the *S. ad. angayoga ˙* of the *Kalacakra ¯* system, whose text and translation were published by Sferra (Sferra 2000).

<sup>32</sup> \**Sr´ ¯ımahak¯ alatantrarudrakalpamah ¯ a´¯sma´sanan ¯ ama ¯ t.¯ıka¯* (*Dpal nag po chen po'i rgyud drag po'i brtag pa dur khrod chen po zhes bya ba'i 'grel pa*), D 1753. Translators' names are not informed in its colophon. The author is Tantric A´svaghos.a, a person different from the famous Buddhist poet A´svaghos.a in the ancient India.

<sup>33</sup> The *Dpal 'khor lo sdom pa'i rgyud kyi rgyal po bde mchog bsdus pa zhes bya ba'i rnam par bshad pa*, whose Sanskrit is transcribed (or insufficiently back-translated) as *shri tsa kra sa mba ra ta ntra ra dza su ka ra ha ¯ sa mu tstsha* [?] *na ma b ¯ r. tti*, D 1413. The Sanskrit title is perhaps *Sam. varasamuccaya*, *Sukhavarasamuccaya*, or equivalent. The names of the translators are not informed in its colophon. Gray (Gray 2007, pp. 23–24) speculated that the author of this commentary, Indrabhuti, had been active around the 10th ¯ century, using Taran ¯ atha's history of Buddhism. However, there are multiple authors who are named ¯ Indrabhuti. I consider the date of this "Indrabh ¯ uti" to not be as old as Gray speculated. I have a ¯ plan to discuss the contents of Indrabhuti's ¯ *Vr.tti* in detail in the future. Here, in this monograph, I note some remarks for considering its date: (1) Indrabhuti's ¯ *Vr.tti* mentions the names of several scriptures of the Sam. vara tradition composed around the 10th century, such as the *Abhidhanottaratantra ¯* , *Yogin¯ısam. caratantra ¯* , and *Vajrad. aka ¯* , and quotes passages from them, and most of those passages can certainly be found in the extant versions of those Tantras; (2) Some "Indrabhuti" of the Sa ¯ m. vara tradition is mentioned in Kr.s.n. a's *Sam. varavyakhy ¯ a¯* (D 1460, 6r3: i ndra bhu ti'i gzhung nyid dang/ ¯ ... ). However, it is not certain whether this "Indrabhuti," whom K ¯ r.s.n. a mentioned, is identical to the author of the *Vivr.ti*; (3) In his *S´uramanojñ ¯ a¯* commentary on the *Cakrasam. vara* (D 1405), Bhavyak¯ırti mentions Indrabhuti and quotes his teachings five times (D 1405, 5v7–6r2, 8v1–v3, 10r4, 14r7, and ¯ 14v1–v2). His *S´uramanojñ ¯ a¯* was translated into Tibetan by Dharma´sr¯ıbhadra and Rin chen bzang po (active between the mid-10th and the mid-11th centuries). However, Indrabhuti's ¯ *Vr.tti* does not contain passages that are evidently equivalent to the passages Bhavyak¯ırti quoted as Indrabhuti's ¯ teachings. (Contradictory ideas can also be found between them.) Some "Indrabhuti" of the Sa ¯ m. vara tradition might have been in existence at the time when Bhavyak¯ırti flourished. However, this does not necessarily mean that the extant version of Indrabhuti's ¯ *Vr.tti* was also in existence in that age; (4) Indrabhuti's ¯ *Vr.tti* has the idea of the four classes of tantras, namely, *bya ba* (\**kriya¯*), *spyod pa* (\**carya¯*), *rnal 'byor* (\**yoga*), and *rnal 'byor bla na med pa* (whose Sanskrit is generally \**yoganiruttara*) (D 1413, 6r and 38r). This is identical to the fourfold classification of tantras that gradually became general in Tibet in and after the 12th century (Dalton 2005, pp. 158–59).
