The Date of Śaṅkara: Between the Cāḷukyas and the Rāṣṭrakūṭas

The Date of Śaṅkara: Between the Cāl.ukyas and the Rās.t.rakūt.as∗ Kengo Harimoto 1 Introduction Śaṅkara is undoubtedly one of the most important figures in the philosophical, spiritual, or religious history of South Asia. When we think of his activity as a historical event, it is natural that we wish to define him in time as accurately as possible. This is why the subject of this article is one of the oldest in the history of indological studies. As shown below (p. 98), the earliest interest in the date of Śaṅkara I could trace was expressed in 1801. The fact that someone can still write on this subject matter after more than two hundred years is a definite example of indological experience: persistent difficulties in arriving at reliable dates for authors or texts. Much has already been said with regard to the date of Śaṅkara, the author of the Brahmasūtraśāṅkarabhās.ya (BSBh). There now is a general agreement that he lived sometime between the late 7th and the early 9th century.1 It is only with great hesitation that I present here a contribution to an already complex debate where almost every argument appears to have been presented. Yet I think it is possible to retry a once proposed but soon discarded approach (see pp. 99 ff.) to narrow down the possible date of composition of the BSBh. The approach is to identify personal names that appear in the BSBh with historic persons whose existence has been recorded in inscriptions. Since the first attempt, more inscriptions have become available, and the history of the relevant area, south India, has become clearer. With the knowledge we have today, I propose that the composition of the BSBh took place sometime in the period 756–772 ce. In the following, I will first examine major theories about the date of Śaṅkara. Then I will discuss three names that appear to be identifiable with those of some political/military figures contemporary with Śaṅkara. ∗ I thank Arlo Griffiths and Harunaga Isaacson for comments and suggestions in writing this article. My special thanks go to Annette Schmiedchen for valuable comments on the Rās.t.rakūt.as, as well as Ludo Rocher and Peter Bisschop for tracing writings of H. T. Colebrooke. 1 Cf. Vetter (1979, 11). 86 Journal of Indological Studies, No. 18 (2006) 2 Current theories Attempts to date Śaṅkara can be classified into four groups: (1) the view that places Śaṅkara in the late seventh century; (2) the one that places him in the first half of the eighth century; (3) the one that places him in the latter half of the eighth century; and (4) the traditional date, 788–820 ce.2 Among those, two views are most often cited: Nakamura’s 700–750 date; and the so-called ‘traditional’3 date, 788–820.4 The first three views have been propounded from various angles. The arguments based on relative chronology tend to find scholarly acceptance more readily, and may be followed up since evidence is verifiable. Here is a brief list of those who have set forth the three views and have found some support: • The first view that places Śaṅkara in the 7th century has been proposed by such scholars as Chintamani (1929), who assigns the date 655–6875 to Śaṅkara, and Kuppuswami Sastri (1937, lviii) who proposes 632– 664.6 Hacker (1964, col. 235) also favors the latter date. • Nakamura (1950, 63–121) has proposed the date 700–750. He states (1950, 86, n. 12) that he follows primarily the arguments of Chintamani (1929) yet reaches a different conclusion. His exposition has probably been the most thorough to date. It has been influential and was adopted by Mayeda (1980, 69–70) and Ingalls (1954, 292, n. 2). With a slightly different approach Thrasher (1979) reaches a similar date, and his argument is accepted by Potter (1981). • Morichini (1960) reviews Nakamura (1950) and suggests some corrections to facts used in the arguments. Kunjunni Raja (1960) takes these corrections into account and places Śaṅkara about 50 years later. The traditional date, 788/9–820/1, has been the most widely accepted. Most publications, especially by non-specialists, cite this date. The dating, however, has a complicated history of itself. It is difficult to trace where it came from. The first notice of this date might go back almost to the beginning of 2 More precisely, it should be 788/9–820/21 because a year in Indian calendars does not start on the same day as the Gregorian calendar. In the following, all the dates are in the common era unless otherwise noted. 3 Despite its being called ‘traditional’, as seen below, the dating is not as traditional as one would expect. (See pp. 93 ff.) Nor is it the traditional date since there are other traditions, such as the annals of Jagadgurus at the Śr.ṅgeri Mat.ha. (See n. 44 on p. 105) 4 Other dates outside the late 7th to the early 9th century have been proposed. Venkateswara (1916) proposes 805–897. I consider this dating has been sufficiently refuted by later scholars for being based on unreliable evidence. Absurd dates that push Śaṅkara far back to the 5th century bce have also been proposed. I do not suppose there is any merit in further mentioning such views. 5 Note that many scholars assign 32 years of lifetime to Śaṅkara. This comes from hagiographies, most notably the Śaṅkaradigvijaya. 6 It is not clear why Kuppuswami Sastri (1937) deviates from the view of Chintamani (1929) while claiming to agree with him. The Date of Śaṅkara (K. Harimoto) 87 indological studies. As far as I am aware, after Thrasher (1979), no serious attempt has been made to reconcile the diverse views. What follows is a brief history of the two widely accepted approaches—from relative chronology and from tradition. My analyses of their methodologies and conclusions will also be presented. 2.1 Approaches from relative chronology Chintamani (1929), Kuppuswami Sastri (1937), Nakamura (1950), Morichini (1960), and Kunjunni Raja (1960) all agree on basic chronology. Also, they disregard the traditional date 788–820 and all of them place Śaṅkara earlier than that period. Yet their conclusions differ significantly. Between the earliest year given for the birth year of Śaṅkara and the latest, there are 120 years of difference. Some of the facts agreed upon include that: (1) both Dharmakı̄rti and Kumārila precede Śaṅkara; (2) Śaṅkara precedes Vācaspati Miśra by several generations. The differences arise from various factors: differences in fixed dates from which the rest of the dates are calculated, how much weight is put on traditional accounts, and so on. Their discussions are the most easily followed if represented graphically. I reproduce the arguments of Chintamani (1929), Nakamura (1950), Morichini (1960) and Kunjunni Raja (1960) in figures 1 (p. 88), 2 (p. 89), 3 (p. 90) and 4 (p. 91) respectively. In the figures the following conventions are used: i) The distance between two thin horizontal lines is ten years, while time runs from top to bottom. ii) When the scholar whose view is represented considers that the floruit of a person is established, the person is inside a box with rectangular corners.7 iii) When the scholar considers that at least one year is established to fall within a person’s floruit, he is inside a box whose top and bottom are open. In most cases, the scholar estimates his first and last year. The year accepted as fixed is inside parentheses alongside the name. iv) The floruit of a person inside a box with rounded corners must be determined in relation to others. v) The height of the boxes represents the length of life/activity of the person. vi) Arrows signify the relationship. The person where the arrow originates knows the person the arrow points to, evidence being quotations, commentarial works, or sometimes teacherdisciple relationship. The last relationship generally determines the distance between the two authors, viz., the two authors should not be very far apart. However, such a relationship usually does not become manifest in the disciple’s writing. Some scholars nevertheless accept traditional accounts of such a relationship. vii) All such arrows must naturally go upward. viii) If 7 All the scholars, especially Nakamura, are inconsistent and ambiguous in distinguishing an author’s year of birth and the beginning year of his (writing) activities. In some cases, the date represents the first possible year when the author was alive, e.g., the date of Bhartr.hari in Nakamura (1950). 88 Journal of Indological Studies, No. 18 (2006) 600 Govinda Gauḍapāda Padmapāda Sureśvara Bhāskara Vidyānanda Jinasena (783) 800 Śaṅkara Maṇḍana Kumārila Prabhākara Dharmakīrti (639) Bhartṛhari (d. 650) Bhāvabhūti* (720) 700 Vācaspati (841) Figure 1: Chintamani’s dating of Śaṅkara (1929) an arrow has two arrowheads, then it means contemporaneity. Again, in most cases the contemporaneity is based on traditional accounts. This type of arrow should be horizontal in principle. ix) Morichini (1960) does not propose a date of Śaṅkara himself, but offers corrections to the scheme of Nakamura. The position of Śaṅkara in figure 3 represents the possible maximum deviation from Nakamura. Some observations are possible from the charts. First of all, it is immediately noticeable that expositions became more and more detailed in the course of time. More evidence was adopted by each subsequent scholar who took a similar approach. Also interesting to note is that the date of Śaṅkara shifts from early to later in each successive study. The most important point to note, however, is that all the scholars here have some freedom in determining where to place Śaṅkara, while still fulfilling all the conditions, viz., the fixed dates and chronological relationship. In the case of Chintamani (1929), from Kumārila on, everybody may in fact be placed much later without compromising the relative chronology. Even Bhavabhūti, to whom Chintamani gives the year 720, can be placed later since he did not have to die in the year his name is associated with. The reason why Chintamani 89 Sarvajñātman The Date of Śaṅkara (K. Harimoto) Maṇḍana Bhāskara Vācaspati (841) Śaṅkara Prabhācandra Kumārila Padmapāda Bhavabhūti * (740) Dharmakīrti (634–673) Kamalaśīla (700–750) Jinasena (783, 838) 800 Śāntarakṣita (680–740) 700 600 Bhartṛhari (530–630) Vidyānanda Figure 2: Nakamura’s dating of Śaṅkara (1950) Sureśvara Journal of Indological Studies, No. 18 (2006) Sureśvara Bhāskara Maṇḍana Prabhācandra (8 c.) Kumārila Padmapāda Jinasena* (838) Uṃveka Vidyānanda Dharmakīrti (634–673) Kamalaśīla (745–795) 700 600 800 Śāntarakṣita (725–785) Bhartṛhari (530–630) 90 Vācaspati (841) Śaṅkara Figure 3: Morichini’s correction to Nakamura’s dating (1960) Sarvajñātman 91 Uddyotana (779) Harisvāmin (638) Maheśvara Bhāvaviveka (500–550) Bhāskara Sureśvara Prabhākara Gauḍapāda Maṇḍana Kumārila Śaṅkara Bhartṛhari Vācaspati (841) Padmapāda Vasurāta Dharmakīrti Vasubandhu (400–450) Uṃveka Kamalaśīla (745–795) Diṅnāga 800 700 600 Śāntarakṣita (725–785) 500 The Date of Śaṅkara (K. Harimoto) Sarvajñātman Figure 4: Kunjunni Raja’s dating of Śaṅkara (1960) Haribhadra 92 Journal of Indological Studies, No. 18 (2006) (1929) places Kumārila et al. so early is because he accepts legendary accounts of Dharmakı̄rti, Kumārila, Śaṅkara and Man.d.ana meeting each other. Also, he maintains that Prabhākara, Man.d.ana, and Um . veka (whom he considers identical to Bhavabhūti), were all disciples of Kumārila. Under these conditions, after determining the dates of Bhartr.hari and Dharmakı̄rti based on Chinese sources, he does not have much choice. As for Nakamura (1950), he could have done the same as Chintamani, but it appears that the divergence from Chintamani’s dates stems from the fact that he wished to preserve the at the time generally accepted dates as much as possible, especially the date of Kumārila. He settles for the latest dates that the facts he accepts allow. By doing so, he disregards all the legendary accounts that Chintamani took into account. Yet he does not disregard the teacher-disciple relationship between Śaṅkara and Sureśvara or Kumārila and Um . veka and Man.d.ana. Also, Nakamura assumes a few more things: the life span of a person is invariably fifty years (except for those whose first and last years can be determined from other sources); if two authors are contemporaries the younger contemporary was born thirty years later (and died thirty years later, following the previous premise); but if one person is a teacher, then the disciple is twenty years younger. Morichini (1960) points out that the dates of Śāntaraks.ita and Kamalaśı̄la which Nakamura accepts are too early. He notes that their dates should be 45 years later than those Nakamura accepts. Also, Morichini points out that Bhavabhūti, who was alive around 720, may not be the same person as Um . veka. These two factors allow even more freedom in attributing a date to Śaṅkara. In principle, the date could be set 45 years later at the latest, even following all the other premises of Nakamura. This is because of the effect Kumārila’s date has on Śaṅkara’s date. Nakamura calculates 8 the date of Śaṅkara as follows: Um . veka/Bhavabhūti (690–740) is 20 years younger than Man.d.ana because he is a disciple of Kumārila and writes commentaries on works of both Kumārila and Man.d.ana; Man.d.ana is 20 years younger than Kumārila because he is Kumārila’s disciple; Sureśvara is one generation younger than Man.d.ana because he criticizes Man.d.ana;9 Śaṅkara is 20 years older than Sureśvara because he is Sureśvara’s teacher; and everybody has a 50 years of lifetime. Placing Śaṅkara at the latest possible date is exactly what Kunjunni Raja (1960) does. He chooses this date because of the silence on the part of Śāntaraks.ita et al. However, as Vetter (1979, 12) points out, this is not 8 Nakamura reaches this date based on the following: Yaśovarman of Kanauj was the patron of Bhavabhūti; around 740, he was defeated by Lalitāditya; Kamalaśı̄la who died in 750 knew Um . veka/Bhavabhūti. Nakamura accepts the identity of Bhavabhūti and Um . veka, the commentator on Kumārila’s Ślokavārttika, based on a colophon in a manuscript of Bhavabhūti’s Mālatı̄mādhava (Nakamura, 1950, 114). 9 I fail to understand this logic. Based on this fact, Sureśvara could be much later than Man.d.ana. The Date of Śaṅkara (K. Harimoto) 93 very convincing. The material Kunjunni Raja (1960) provides us indicates that Śaṅkara’s date must be somewhere in the period of from 650 to 800. Thus even using verifiable facts as evidence, datings based on relative chronology are considerably affected by many other variables.10 The 120 years divergence between the extreme views based on a similar methodology is a result of accumulation of differences in the base dates and assumptions. In the end, we indeed cannot go much beyond what Vetter (1979, 11) asserts: “Ś[aṅkara] muß zwischen 650 und 800 n. Chr. gelebt haben.” An earlier date in this period might seem preferable, following Thrasher (1979). His argument that appears to carry particular force is the fact that Um . veka mentions many variant readings of the Bhāvanāviveka of Man.d.ana Miśra. Thrasher therefore feels it necessary to assign much time between Man.d.ana and Um . veka. Since he considers Man.d.ana and Śaṅkara to be contemporaries (with which I agree) he assigns Śaṅkara’s floruit to “700 or slightly before (Thrasher, 1979, 139).” 2.2 The traditional date: 788–820 The so-called ‘traditional’ date is an entirely different story. Surprisingly, how this theory started to be accepted as traditional is hard to trace. This is because later scholars misrepresented predecessors’ contributions, sometimes mixing different contributions into one. This may be because they did not or could not check their sources. For example, Nakamura (1950) cites Pathak (1882) as the first promulgator of this view. But Pathak was certainly not the first to assume the year 788 as the birth year of Śaṅkara. Already in the same volume of the same journal, an editor (p. 263) notes that there had been a previous publication (Tiele, 1877) that had referred to the 788 date, but the editor misspelled the author’s name (‘Teile’). This misspelling was inherited by Chintamani (1929), and he went as far to say that Tiele (1877) referred to the year 820 as the year of Śaṅkara’s death as well. However, Tiele (1877) only mentions the year 788. The confusion reaches its peak with Kunjunni Raja (1960, 129) who states: 10 Another thing to be noted is the year 841 of Vācaspati Miśra which all four authors cite as the latest limit of Śaṅkara. The year 841 is based on the year ‘898’ in which Vācaspati says to have completed the Nyāyasūcı̄nibandha. Considering this year to refer to the Vikrama era, we arrive at 841 ce. However, this date has been highly contested, and the majority of scholars are now inclined to consider the year ‘898’ as Śaka year rather than Vikrama. This places the completion of the Nibandha in the year 976–77 ce. For a summary of the debate on the date of Vācaspati Miśra, see Acharya (2004, xii–xxii). The date of Vācaspati Miśra after all does not affect the dating of Śaṅkara much because his upper limit is determined in relation to other authors. Śaṅkara should not be much later than Um . veka who commented on Man.d.ana’s work. It is generally agreed that Man.d.ana is a contemporary of Śaṅkara. Um . veka’s commentary on the Ślokavārttika of Kumārila is quoted by Kamalaśı̄la who died in 794. Still, giving enough time between Śaṅkara and Vācaspati was a major concern, especially for Nakamura, and hence his dating of Śaṅkara was indirectly affected by the date of Vācaspati. 94 Journal of Indological Studies, No. 18 (2006) In 1877 Prof. Tiele in his Outlines of the History of Ancient Religions placed Śaṅkara between a.d. 788 and 820 on the basis of Yajnesvara Sastri’s Āryavidyāsudhākara which refers to Bhat.t.a Nı̄lakan.t.ha’s Śam . karamandārasaurabha. Even though the name Tiele was spelled correctly, the title of the book was wrong. It should be Outlines of the History of Religion (to the Spread of the Universal Religions). This mistake was another legacy of the remark by the editor of the Indian Antiquary, vol. XI. Moreover, Tiele (1877) did not mention any basis for ascribing the date 788 to Śaṅkara. The basis Kunjunni Raja (1960) cites comes from discussions in Telang (1884). Telang only speculated how Tiele might have come up with that dating. In the following paragraphs I will trace the process that lead to these confusions. This same process also led the period 788–820 to be established as the ‘traditional’ one. In 1882 Pathak publishes a short article (Pathak, 1882) in the June issue of the Indian Antiquary, vol. XI. He mentions an unidentified manuscript consisting of three folios in which the following is found: nidhināgebhavahnyabde vibhave śam . karodayah. | as..tāvars.e caturvedān dvādaśe sarvaśāstrakr.t | .sod.aśe kr.tavān bhās.yam . dvātrim . śe munir abhyagāt | kalyabde cam dranetrām kavahnyabde guhāpraveśah. | . . vaiśākhe pūrn.imāyām tu śam karah śivatām agāt | . . . As they stand, these stanzas are badly composed. One can assume that the fourth pāda means that Śaṅkara learned all the śāstras by the age of twelve, but we are missing a finite verb for the third pāda that should mean “Śaṅkara learned” (the four Vedas by the age of eight). Also, the nominative sarvaśāstrakr.t (the author of all the śāstras) does not seem suitble for describing Śaṅkara. In that place, too, we would like to have something that means “he mastered all the śāstras (sarvaśāstravid ?).” The sixth to the tenth pāda may be about the death (or disappearance) of Śaṅkara. However, there is something awkward about the seventh to eigth pāda: the subject is guhāpraveśah., which is unmetrical, and the sentence is surrounded by sentences where the subjects are in essense Śaṅkara. Despite all these clumsiness, Pathak’s understanding that the years mentioned, viz., Kali 3889 and 3921, refer to the years of the birth and death of Śaṅkara in Kali years probably seemed relatively reasonable and became widely accepted. And since Kali 3889 and 3921 roughly correspond to 788 and 820, later scholars cite those two dates. Surprisingly, nobody seems to have paid attention to the word vibhave in the second pāda. The word does not make sense when one tries to understand it either as vibhava or vibhu in their usual sense. But in the Hindu calendar the word vibhava is the name of the second year of the sixty years The Date of Śaṅkara (K. Harimoto) 95 of the Jovian cycle.11 Modifying the word year (◦ abde), the natural interpretation is to read the word as the name of a year. Kali 3889, however, is not the year vibhava, but is prabhava, the first year, in the cycle. There is one year discrepancy between the year expressed by numbers and the name of the year. The year vibhava comes next to prabhava. Following the publication of Pathak (1882), in the September issue of the same volume (p. 263) of the same journal an anonymous editor points out that the date 788 had been given by Tiele in 1877: . . . it is to be noted that Prof. Teile [sic.] (Outlines of the Hist. of Anc. Religions [sic.], p. 149) had, in 1877, given A.D. 788 as the date of the birth of Śam . kara. The two mistakes here becomes a major source of later confusions mentioned above (p. 93). In the following year Deussen quotes a portion of the Āryavidyāsudhākara12 in a footnote (1883, 37): tathā ca sampradāyavida āhur: nidhināgebhavahnyabde vibhave māsi mādhave | śukle tithau daśamyām . tu śaṅkarāryodayah. smr.tah. k iti | nidhināgebhavahnyabde (3889) navāśı̄tyuttarās..taśatyadhikatrisahasrı̄mite vars.a ity arthah. | kaliyugasyeti śes.ah. | tathā śaṅkaramandārasaurabhe nı̄lakan..thabhat..tā apy evam evāhuh.: prasūtah.13 tis.yaśaradām atiyātavatyām ekādaśādhikaśatonacatuh.sahasryām 14 ityādi | tis.yaśaradām . kaliyugavars.ān.ām ity arthah. | Note that the stanzas ascribed to “those who know the tradition (sampradāyavidah.)” here contain much of the first two pādas of what Pathak (1882) 11 The year names of the sixty years of the the Jovian cycle are still in wide use. Modern calendars published in India or Nepal most of the time include the year name. The year names are: prabhava (1st), vibhava (2nd), sukha (3rd), pramoda (4th), prajāpati (5th), and so on. For example, the year 2006 ce starts in vilambin (32nd year) in the northern system and pārthiva (19th year) in the southern system. There is no difference between the sourthern and the northern system till Kali 4009 according to the Pancanga program by Michio Yano. 12 Only the 1940 edition of the text has been available. The introduction to the edition states that the first edition was published in 1868. This is confirmed by the Library of Congress record. 13 The 1940 edition reads prasūta (Yajñeśvara Ciman.a Bhat.t.a, 1940, 192) without the visarga. It could just be a typographical error. At any rate, the word is not metrical. It is possible originally there was another word in that position. 14 Romanization and sandhis are adjusted to the modern conventions from the orginal in Deussen (1883). This passagte corresponds to Yajñeśvara Ciman.a Bhat.t.a (1940, 191–2). 96 Journal of Indological Studies, No. 18 (2006) cites, with the addition of the exact date. Certainly, similar stanzas that mention Śaṅkara’s ‘rise’ were circulating. They at least agree on the year when “the rise of Śaṅkara (śaṅkarodaya/śaṅkarāryodaya)” occurred.15 A comparison of the two versions of stanzas about the “rise of Śaṅkara” gives me the impression that the version Pathak quotes maintains an older form. The form śaṅkarāryodayah. in the Āryavidyāsudhākara is less straightforward than śaṅkarodayah. and is awkward (one would expect ācārya rather than ārya). It seems that ārya was added to fulfill the metrical requirement. Also, the word smr.tah. in the Āryavidyāsudhākara does not settle very well for a verse that itself is to be memorized and should be authoritative. Again, it appears to be added as a filler. Now the first two pādas of the stanzas Pathak quotes may preserve a somewhat older form, but the stanzas as a whole are still not well composed. This leads one to notice that the seventh and the eighth pādas from the same version are symmetrical to the first two pādas, while looking awkward surrounded by pādas where the subject is Śaṅkara. Taking into account that the eighth pāda is unmetrical, and that the compound candranetrāṅkavahnyabde, though not necessarily defective, spans beyond pāda boundary but is metrical by itself, I speculate that the original stanzas were something like: nidhināgebhavahnyabde vibhave śaṅkarodayah. | candranetrāṅkavahnyabde śaṅkarah. śivatām agāt k. This reconstruction gives explanations for some of the difficulties in Pathak’s stanzas mentioned above: the rest of the extant stanzas quoted by Pathak may be later additions. Still, the one year discrepancy between Kali 3889 and the year vibhava is persistent. The reading might be a simple mistake for prabhave, but obviously for quite some time the defective reading was transmitted uncritically. In the meantime, in the same year that Pathak (1882) was published, and again in 1884, Telang16 speculated about the source where Tiele could have gotten the idea. He proposes that Tiele got the information from the Āryavidyāsudhākara. The passing away of its author, Yajñeśvara Śāstrı̄ was a recent event in 1884. The portion of the Āryavidyāsudhākara he refers to is the same as Deussen does. Neither Deussen nor Telang mentions the other; but given the simultaneity of their publications, it is plausible that they communicated on the issue or that one took information from the other without acknowledgement. In the following years the described process of identifying the definite source of the 788–820 date became confused inextricably. Thus it was indeed Pathak (1882) who proposed the years 788 and 820 as a set. But the year 788 had been mentioned by Tiele (1877) five years earlier. Telang’s speculation notwithstanding, what was the basis for the dating? Tiele (1877, 139–40) mentions the date matter-of-factly: 15 16 Note that the Āryavidyāsudhākara does not mention the disappearance of Śaṅkara. See Telang (1882, 27, n. 3) and Telang (1884, 95). The Date of Śaṅkara (K. Harimoto) 97 Not long afterwards it [i.e., Buddhism] encountered a violent opponent in the celebrated teacher of the Mı̂mânsâ school, Kumârila-Bhat.t.a, and later still in the great enemy of all heresies, the orthodox S.ankarâchârya, who was born in 788 a.d. Even more interestingly, in the original Dutch version of the book published the year before, Tiele (1876, 150) does not mention Śaṅkara at all.17 Since Tiele was not an indologist, it is hard to imagine that he reached the date by himself. He must have had help from an indologist. In the preface he acknowledges his Leiden colleague H. Kern’s help in making corrections for the English edition (Tiele, 1877, xi). It is probable that the bit about Śaṅkara was introduced as per suggestion of Kern. Kern in a review (1877, 63) of Tiele (1876), the Dutch original, recommends to consult Colebrooke (1858) for religions of India. Other than that, I have not found anything that suggests his involvment in adding reference to Śaṅkara in Tiele’s book. Nor have I found Kern’s own exposition about the date of Śaṅkara. Could Colebrooke (1858) be the source where Tiele got the 788 dating for Śaṅkara? Now when we turn our attention to Colebrooke (1858), we only find the following on pages 212–3: I have had a former occasion of discussing the antiquity of this eminent person; and the subject has been since examined by ráma móhen ráya and by Mr. Wilson.18 I continue of opinion, that the period when he flourished may be taken to have been the close of the eighth or beginning of the ninth century of the Christian era; and I am confirmed in it by the concurring opinions of those very learned persons. This in fact was originally published as Colebrooke (1830) as part 4 of a series titled “On the Philosophy of the Hindus.”19 Now this “close of the eighth or beginning of the ninth century of the Christian era” appears suspiciously similar to the ‘traditional’ date (788–820). What was the “former occasion” and did he refer to the exact year 788? Now since Colebrooke cited above does not give a reference to the former occasion, we have to check Wilson (1819) to which Colebrooke refers. Wilson (1819, 11/xvi?)20 does refer to Colebrooke’s view: “Mr. Colebrooke is in17 The relevant portion reads: “Maar nog voor het einde der eeuw gelukte het den beroemden leeraar der Mı̂mânsâschool Kumârila, in een bloedige vervolging aan het Buddhisme in Indië den doodsteek te geven.” 18 Here is a footnote that simply says “Sanscrit Dict., first edit., pref. p. xvi.” 19 The article is reproduced three more times as part of Colebrooke (1837), Colebrooke (1871) and Colebrooke (1873). Deussen (1883, 37–38) refers to the above quoted passage in the 1871 edition (Colebrooke, 1871, 332). 20 I only had access to an Indian reprint of the dictionary. The publisher obviously decided to renumber the pages and to give the footnotes continuous numbers in the preface. As a result, I do not know the original page, but there is only one relevant page in the preface. 98 Journal of Indological Studies, No. 18 (2006) clined to give him an antiquity of about 1000 years” (Wilson, 1819, 11/xvi?). The statement has a footnote, but since the reprint of Wilson (1819) I can consult garbled the numbering of footnotes, it is not clear which of the footnotes on that page we should consult. The most likely footnote21 refers to the “Preface to the translation of the Dáya Bhága, [p. XI],” and it yields Colebrooke (1810, xi). The latter simply states “. . . Sancara-a’cha’rya, whose age cannot be carried further back at the utmost than a thousand years.” In 1810, when the book was published, this meant that Śaṅkara’s date cannot be pushed back earlier than 810. Again, this is close to the ‘traditional’ date, but Colebrooke does not mention the exact year. There are some more references to Śaṅkara’s date by Colebrooke.22 Among them is a brief mention in a note in Colebrooke (1805):23 “According to the traditions generally received in the peninsula of India, Śancara lived little more than eight hundred years ago.” This places Śaṅkara [a?] little earlier than 1005. Note that there are two hundred years of difference between the statements made in 1805 and 1810. From the year 1810 on, the opinion of Colebrooke is consistent, and his dating has become similar to the ‘traditional’ one. The “former occasion” may lie in sometime between 1805 and 1810. Colebrooke was certainly interested in the date of Śaṅkara even before 1805. He expresses his interest as early as 1801 in a note published in the Asiatick Researches, vol. 7, which is reproduced as Colebrooke (1872, 16). Given his interest and contacts between early indologists and pandits in India in his days (evidneced by his mention of traditions and Ram Mohan Roy), it is conceivable that he was exposed to the Kali 3889 tradition, and it is possible that he reported it somewhere, even though he did not base his opinion on it. However, I am unable to find evidence. After all, Tiele (1877) remains the first to have mentioned the year 788 as Śaṅkara’s birth year in the Western publication, and we do not know where it came from. There have been several versions of the stanzas about the rise of Śaṅkara that cannot be ascribed to a particular author or a work. With these gratifications, the birth date 788/789 may be called ‘traditional’. However, keeping the one year discrepancy between the Kali and Jovian year 21 The editor of the 1873 edition of Colebrooke’s Miscellaneous Essays confirms that this footnote is the correct one. He added a note to Colebrooke’s footnote cited above (n. 18). The additional note by the editor, Cowell, reads: “Cf. Preface to the two treatises on the Hindu law of inheritance.” 22 Other than the reference I mention below, references to Colebrooke’s dating of Śaṅkara include Windischmann (1833, 40–41), referred to by Deussen (1883, 37), who refers to “Transact. of the Royal As. Soc. I, p. 441,” “Col. Transact. I, p. 443,” and “Transact. II, p. 6” as Colebrooke’s. The first one (Colebrooke, 1827) (see the bibliography for its reproductions) simply states that Kumārila predates Śaṅkara. The second one should refer to the same publication but I do not find anything relevant in the pages of reproductions. The last one is the one reproduced in Colebrooke (1858, 212) above. 23 This is reproduced as Colebrooke (1858, 63), Colebrooke (1871, 104) and Colebrooke (1873, vol. 1, 94) The Date of Śaṅkara (K. Harimoto) 99 name does not seem to be a good tradition. The year recorded there itself is in jeopardy.24 3 Balavarman, Jayasim . ha and Kr.s.n.agupta As seen above, Śaṅkara’s floruit may be placed anywhere in the period of a century and a half, 650 to 800. This is a long time. This period is, as the figures on pages 88–91 show, also the period when many influential thinkers flourished. We would wish to know a more precise date of Śaṅkara. That would enable us to have a clearer picture of the important period in the history of Indian philosophy. There have been some attempts to determine the date of Śaṅkara positively based on place or proper names in the BSBh.25 They have been deemed unreliable, but there is one approach that may be worthy of reconsideration. That is the one D. R. Bhandarkar (1912) took. He paid attention to a passage in BSBh 4.3.5 where the names Balavarman, Jayasim . ha and Kr.s.n.agupta occur. He identified Balavarman in that passage with the Balavarman that appears in a copper plate grant from Kad.aba, Karnataka, reported by Rice (1883), and concluded that Śaṅkara should be assigned to the latter half of the eighth century.26 Already in the next year Narasimhachar (1913) criticized this approach, pointing out that D. R. Bhandarkar (1912) singled out Balavarman, while ignoring the remaining two names. He concluded, “When epigraphical or other evidence becomes available to prove the contemporaneity of the kings referred to with Śankarâcârya, the argument from the synchronism of Balavarmâ will be perfectly legitimate. Till then the names have perhaps to be looked upon as connoting imaginary persons like the words Devadatta and Yajñadatta or the letters A, B and C.” Chintamani (1929) and Nakamura (1950) were likewise critical of the approach. Chintamani (1929, 49) went so far as to say “There have been ever so many Balavarmans and Prof. 24 For other criticisms on the dating, see, for example, Nakamura (1950, 66). Apart from D. R. Bhandarkar (1912) mentioned below, Telang (1884) discusses place names Srughna and Pāt.aliputra and a personal name Pūrn.avarman in BSBh 2.1.18 (Telang considers the sūtra to be 2.1.17) to arrive at the conclusion that Śaṅkara flourished in the latter half of the sixth century. The problem with trying to identify Srughna and Pāt.aliputra with real places is that Śaṅkara took those place names from the Mahābhās.ya. In the Mahābhās.ya those two places are used as equivalents of Devadatta, etc., for places. For criticism of Telang (1884) and D. R. Bhandarkar (1912), see Venkateswara (1916, 153–55) and Nakamura (1950, 79-82) 26 In fact his goal was to support the view of Pathak that would later become the traditional date (788–820). However, D. R. Bhandarkar (1912) inadvertently contradicted Pathak’s view. Even if we grant that by late 8th century D. R. Bhandarkar (1912) meant the period close to the end of the century, if we follow Pathak’s view, Śaṅkara was still 12 or 13 years old at the end of the eighth century (800 ce). Writing the BSBh at the age of 13 seems very unrealistic. 25 100 Journal of Indological Studies, No. 18 (2006) Bhandarkar’s arguments cannot stand.” Nakamura (1950, 83–84) followed Narasimhachar’s (1913) observation. 3.1 Śaṅkara and names Here is the passage D. R. Bhandarkar (1912) called attention to: tathā ca loke prasiddhes.v apy ātivāhikes.u evam . jātı̄yaka upadeśo dr.śyate: “gaccha tvam ito Balavarmān.am tato Jayasim . . ham . tatah. Kr..sn.aguptam” iti | (Śāstri and Panśı̄kar, 1917, 992) Furthermore, in daily life, we encounter such an instruction with regard also to well-known leaders as the following: “You go from here to Balavarman, from him to Jayasim . ha, [and] from him to Kr.s.n.agupta.” This example appears in the following context: starting at BSBh 4.3.1, the topic is the devāyana through which one reaches Brahman after death. While reconciling various Upanis.adic accounts on the devāyana into one (BSBh 4.3.1–3), a question arises. What exactly are fire (arcis), day (ahar ), etc., in one sequence, and the world of Agni (agniloka), etc., in another, if those sequences in different accounts refer to one and the same sequence (BSBh 4.3.4)? The answer given by Brahmasūtra 4.3.4 is that they are leaders/guides (ātivāhikas). Śaṅkara presents this answer in contrast to two alternatives, viz., that arcis, etc., are signposts (mārgacihnas/mārgalaks.anas) to pass by, and that agniloka, etc., are the levels where the departed has experiences (bhogabhūmi s). The passage quoted above is part of the answer. It is in contrast to another example27 that appeared earlier in order to illustrate that arcis, etc., are signposts. This example, from the viewpoint of the proponent, is to illustrate that sentient beings (leaders), too, can be referred to in an itinerary. This passage makes most sense if Śaṅkara was actually using names of people who were alive in his time. The purpose of using this example is that readers are convinced that such an instruction is common, and therefore arcis, etc., too, are similar (in being leaders) to those people mentioned in the example. Illustrating the other possibility that arcis, etc., are signposts by an example is easier. Everybody can imagine a situation like the one Śaṅkara cites (“Go from here to that mountain, . . .”) even today. Note that in this example, there is an assumption that both the participants of 27 Śāstri and Panśı̄kar (1917, 991): yathā hi loke kaścid grāmam . nagaram . vā prati.s.thāsamāno ’nuśis.yate: “gaccha itas tvam amum . girim . tato nyagrodham . tato nadı̄m . tato grāmam . tato nagaram . vā prāpsyasi” iti “For example, in daily life someone who wishes to set out for a village or a town is instructed [thus]: ‘You go from here to that mountain, from there to a banyan tree, from there to a river. From there, you will arrive in the village or the town.’ ” The Date of Śaṅkara (K. Harimoto) 101 the conversation recognize the starting point (here) and the signpost (that mountain). However, finding a convincing example in which sentient leaders are referred to in an itinerary is harder. In order for the example to work, it is helpful that readers know that those in the example are alive and leaders of some sort. Otherwise, they would have to assume that these three persons are leaders somehow, and that they are known by the participants of the conversation.28 He does say that those people are well-known (prasiddha) as leaders (ātivāhikas). What happens if readers, like we, do not know those three persons as leaders? We have to take Śaṅkara’s words and believe that people do say such things. This weakens his argument to some degree. The three names in the example are unlikely to be imaginary for another reason. They are not like Devadatta, Yajñadatta, or Vis.n.umitra that everybody knows as hypothetical names for anybody, Sanskrit equivalents of ‘John Doe’. I am aware of no text where any of the names Balavarman, Jayasim . ha or Kr.s.n.agupta was used as an arbitrary name. In addition to Devadatta, etc., in similes and examples Śaṅkara mentions mythical figures, such as Manu, Vyāsa, Bhadrasena, Śvetaketu, etc. In these cases, too, we have an idea of who they are from other literary sources. Not only in similes and examples, when a proper name appears in the BSBh, we know who they are in most cases. The most likely explanation for the use of those three names in BSBh 4.3.5 is that he considered those persons, too, to be well-known. It has been more than ninety years since D. R. Bhandarkar (1912) tried to identify Balavarman in the BSBh with the one mentioned in the Kad.aba copper plates. Since then many more epigraphic records have become available, and the political history of India has become much clearer. Has the time “when epigraphical or other evidence becomes available to prove the contemporaneity of the kings referred to with Śankarâcârya (Narasimhachar, 1913, 54)” finally come?29 3.2 The Cāl.ukya connection If we can identify a period when three well-known (prasiddha) persons with the names Balavarman, Jayasim . ha, and Kr.s.n.agupta, lived at the same time, we may confirm that Śaṅkara was indeed referring to real people. In addition, the composition of the BSBh should fall in that period. When one looks 28 I venture to challenge readers of this article to come up with a similar example that fulfills Śaṅkara’s intention by replacing the three names. I for one find it quite difficult without using names of real people. 29 Another name Śaṅkara refers to, Pūrn.avarman (BSBh 2.1.18; ChUBh 3.19.2) might have been a historic figure. Śaṅkara refers to Pūrn.avarman in discussions on the nonexistence. In the two examples where the name Pūrn.avarman appears, the issue is the absurdity of talking about the king before him, who is non-existent. There is one famous Pūrn.avarman, but he is from Java. He is the oldest epigraphically attested ruler of Java. 102 Journal of Indological Studies, No. 18 (2006) for records of the first two names, Balavarman and Jayasim . ha, a noticeable correlation emerges. Those names both have connections to the Cāl.ukyas. The Cāl.ukyas first appeared in history in Karnataka as the Cāl.ukyas of Badami, and then while the main branch (that from Badami) was overthrown by the Rās.t.rakūt.as in the late 8th century, the Eastern branch (the Cāl.ukyas of Vengi) continued to thrive in present Andhra Pradesh. And then, in the 10th century, the Western Cāl.ukyas (the Cāl.ukyas of Kalyana), who may or may not be the decendants of the Cāl.ukyas of Badami, finally overthrew the Rās.t.rakūt.as and ruled a large empire covering most of South India until the 12th century. Of the two names, the name Jayasim . ha has more easily discernible connections with the Cāl.ukyas. The majority of inscriptions that record the name Jayasim . ha refer to persons of that name from various Cāl.ukyas dynasties.30 Four Cāl.ukya kings (one from Badami, two from Vengi, and another from Kalyana), as well as at least one prince (among the Cāl.ukyas of Badami) of that name from the sixth to the eleventh century are known.31 The date of Śaṅkara postulated from relative chronology falls within that period. Also, extending over modern Karnataka and Andhra Pradesh and sometimes into Kerala, the area where the Cāl.ukyas were influential covers much of the area where Śaṅkara was most likely from, and probably active. Unlike his date, there is no dispute with regard to where he came from, viz., south India.32 Traditionally, he is reported to have hailed from Kaladi, Kerala, and the Śr.ṅgeri mat.ha, considered to be the first one established by him, is in the middle of Karnataka. Being active in the area where the Cāl.ukyas were influential, it is well conceivable that Śaṅkara was aware of (a) political figure(s) called Jayasim . ha of the clan. Then when he uses that name in the BSBh, it would hardly be accidental, but probably be intentional. He probably meant to refer to a real person. Still, it is impossible to determine which of many Jayasim . has he was referring to. It should also be noted that there perhaps were more Jayasim . has among the Cāl.ukyas, given the common custom of sharing a small variety of names in a family.33 The 30 Note that this name later became popular among the kings in Rajasthan. Quite a few Maharaja Jai Singhs are known. 31 Among the Cāl.ukyas of Badami, Jayasim . ha (r. 500–520?) and a son of Pulakeśin II, called Dharāśraya Jayasim . ha, are known. See Dikshit (1980, 26–30, 123, 133, 312–313). Among the Cāl.ukyas of Vengi, two kings of the name Jayasim . ha are known: Jayasim . ha I (r. 641–673), II (r. 706–718), and another (Jayasim . ha III) in the 11th century. See Rao (1935) and Rao (1973, 318, 348, 408, 423). Among the Cāl.ukyas of Kalyana, a king of this name (r. 1015–43) is known. 32 The earliest mention of the locality of Śaṅkara in Western publications I have been able to consult was Duncan (1798, 5). For the acceptance of Śaṅkara’s southern origin, cf. Mayeda (1979, 4), Potter (1981, 117), etc. 33 The popular names among the Cāl.ukyas include: Jayasim . ha, Pulakeśin, Maṅgaleśa, Vis.n.uvardhana, Kı̄rtivarman, Vikramāditya. They might have employed the custom of naming brothers after the father’s brothers, the custom reportedly common in Kerala. Cf. Rāmanātha Śāstri (1931, xiv–xvi). Cf. also Pisharoti (1929, 678). To see how small variety The Date of Śaṅkara (K. Harimoto) 103 Jayasim . ha Śaṅkara referred to might not have left any record, especially during the period when the Cāl.ukyas conceded power to the Rās.t.rakūt.a rule. We may narrow down the period if we add the condition that this Jayasim . ha must have been contemporaneous with a Balavarman. As for the latter name, we learn that there were some Balavarmans whose records are available. From the south (Andhra Pradesh and Karnataka) six records that mention Balavarman are known (Rice, 1883; Ramesh and Murthy, 1977/78).34 None of them knows Balavarman clearly as a king, but one knows him rather as a vassal chief. Also, two of the inscriptions know him as a Cāl.ukya. The name Balavarman in those inscriptions might indeed refer to the same person (Ramesh and Murthy, 1977/78, 181). We can at least be certain that there was a certain Balavarman who was a vassal chief of the Cāl.ukya lineage after the Rās.t.rakūt.as overthrew the Cāl.ukyas of Badami from power. Since these two names both relate to the Cāl.ukyas, that they appear in the same place in BSBh 4.3.5 could be more than a mere coincidence. Did one of the Jayasim . has we know live at the same time as Balavarman, the Cāl.ukya? There are some problems. There no doubt was a Balavarman of Cāl.ukya heritage, but the only dated inscription mentioning Balavarman that Rice (1883) reports is suspected of forgery.35 Even if we adopt that dated (813)36 record as evidence, still the margin when Balavarman mentioned in the copper plate lived is very large.37 On paleographical grounds of names are recurring throughout different Cāl.ukya dynasties, Burnell (1878, 18–21) is convenient even though most details, especially the dates, have been superseded. 34 Outside south India, three kings among the Kāmarūpas of Assam are known: Balavarman I (early fifth century), Balavarman II (late eighth century) and Balavarman III (late ninth to early tenth century). 35 See Ramesh and Murthy (1977/78, 181). The origin of this suspicion is often ascribed to Fleet, The Dynasties of the Kanarese Districts of the Bombay Presidency from the earliest historical times to the Muhammadan Conquest of A.D. 1318, Bombay, Printed at the Govt. Central Press, 1882, p. 399, n. 7. However, the book has only 106 pages, and I do not find any relevant note in the book. Also, the book was published before Rice (1883). At least in Fleet’s later publication about spurious inscriptions (Fleet, 1901), the copper plate is not mentioned as spurious. One conceivable problem with the copper plate is the date recorded in it. It records Śaka 735 (expired), śuklapaks.a, 10th tithi of the month Jyes.t.ha, naks.atra Pus.ya, Monday. The 10th tithi of the śuklapaks.a of the month Jyes.t.ha in Śaka 735 (expired) is not Monday and its naks.atra is not Pus.ya. Also, the names in the genealogy in the grant are spelled differently from other inscriptions: Kakka for Karka; Inda for Indra; Kan.va/Kanna(Kan.n.a?) for Kr.s.n.a. (For the last one see note 41.) These differences are typical copying mistakes. It is thus conceivable that one suspects that the copper plate is not original but a replica. Note that this does not mean that the whole thing in the copper plate is fabricated. 36 See the previous note on the date of the copper plate. 37 The grant (Rice, 1883) on copper plates records that Rās.t.rakūt.a Govinda III granted a village on the application of Cākirāja to a Jaina temple, a monk of which successfully treated Cākirāja’s nephew-in-law (a son of his sister). This nephew-in-law of Cākirāja is called Vimalāditya. He is a son of Balavarman. Depending on how old the father was 104 Journal of Indological Studies, No. 18 (2006) Balavarman may be dated in the late 8th century (Ramesh and Murthy, 1977/78, 181), and this is the most we can say about the date of Balavarman. One might think that a rather well-known Jayasim . ha was contemporary with Balavarman. He is Dharāśraya Jayasim . ha, a son of Pulakeśin II and a younger brother of Vikramāditya I. His name appears in inscriptions as late as 739 (Mirashi, 1955, 127–145). However, in those inscriptions he is mentioned as the father of the sons who were responsible for the grants. The father of Dharāśraya Jayasim . ha, Pulakeśin II, had died in 642 or 643. It is highly unlikely that a son was living almost one hundred years after his father’s death. Accordingly, I discard the possibility that this Dharāśraya Jayasim . ha was the Jayasim . ha in BSBh 4.3.5. But there is another record 38 of a gift by a certain Dharāśraya Jayasim . ha in 727 or 728 (Śaka 649). We do not know much about this Jayasim . ha, but if the report is correct, he was probably not the same Dharāśraya Jayasim . ha, the youngest son of Pulakeśin II. And because of the popularity of the name Jayasim . ha among the Cāl.ukyas, it is conceivable that there were even more Jayasim . has. Even though we can be relatively certain that Śaṅkara had a real Balavarman and a real Jayasim . ha in mind, the best we can say is that Śaṅkara’s date falls in the 8th century. This does not help in narrowing down the date of the composition of the BSBh. Let us now take a look at the political situation of the area at the possible time. I have mentioned that we know of a Balavarman who was probably a vassal chief under the Rās.t.rakūt.a rule. The former feudatory Rās.t.rakūt.as overthrew the Cāl.ukyas from power slightly before 756.39 We naturally do not know much about the Cāl.ukyas in the west after the coup until they again overthrew the Rās.t.rakūt.as. Our Balavarman belongs to the period during which we know little about the Cāl.ukyas of the west. This may be why we have trouble identifying Jayasim . ha even though there were many Jayasim . has among the Cāl.ukyas. But now the name Kr.s.nagupta yet to be discussed demands attention. The first king of the Rās.t.rakūt.as after they completed the power transition from the Cāl.ukyas of Badami was Kr.s.n.a I40 (Kr.s.n.arāja in most inscriptions).41 There is a high likelihood that this Kr.s.n.a and Balavarman were contemporaries. Kr.s.n.a I is dated r. 756–772 (plusminus a year or two on each end),42 while Balavarman cannot be earlier when the son was born and how long they were active, the period that this Balavarman could be in charge can be between 30 and 80 years before the grant was issued. 38 See Indian Archaeology 1962–63, Review, p. 52 and Dikshit (1980, 310) 39 See Altekar (1967, 44–45), Schmiedchen (2001, 242). 40 There were two more kings with the name Kr.s.n.a among the Rās.t.rakūt.as: Kr.s.n.a II in the late ninth to the early tenth century; Kr.s.n.a III in the tenth century. 41 In the Kad.aba plate where Balavarman is mentioned, the name corresponding to Kr.s.n.a I seems to read as Kanveśvara or Kan.veśvara. Rice (1883, 14) reads it as Kanneśvara. 42 See Altekar (1967, 44–45). The Date of Śaṅkara (K. Harimoto) 105 than the demise of the Cāl.ukyas (756) and is highly unlikely to be later than the eighth century. Despite the addition of an appellation gupta in the name,43 I do not think there is much chance that Śaṅkara picked up the names Balavarman and Kr.s.n.agupta just by chance, given that there was a time when Balavarman and Kr.s.n.a were contemporaries. The time when they were active is inside Śaṅkara’s probable lifetime deduced from relative chronology. Also, they were active in the same area where Śaṅkara was active. It is most natural to think that Śaṅkara was at least inspired by the names of ‘leaders’ whom he was aware of. Based on the date of Kr.s.n.a I, I propose that the BSBh was composed sometime between 756 and 772. Identifying the three names, Balavarman, Jayasim . ha and Kr.s.n.agupta in the example in BSBh 4.3.5 with a Cāl.ukya vassal chief Balavarman, a Cāl.ukya descendant Jayasim . ha (we do not know who) and a Rās.t.rakūt.a king Kr.s.n.a I makes the example intelligible. The three names exemplify fire (arcis), day (ahar ), half month when the moon is waxing, etc., ending with Brahman. The order is from smaller to greater. Starting from a vassal chief and ending with the king of a large empire, which the Rās.t.rakūt.as had become, corresponds to that order. Nakamura (1951) observed from the BSBh that Śaṅkara was living in a time when various powers were in persistent conflict. Nakamura also noted Śaṅkara’s familiarity with the life of those who were in power and with the workings of kingdoms. He considered that Śaṅkara’s writing reflected the state of India after the demise of the Guptas. However, it seems more appropriate to ascribe what Nakamura observed specifically to the situation in Karnataka in the late 8th century when a former feudatory took over a kingdom.44 43 According to Apte’s dictionary, gupta is “an appellation usually (though not necessarily) added to the name of a Vaiśya.” The presence of gupta in ‘Kr.s.n.agupta’ in the BSBh clarifies that the personage referred is a human and not other famous mythical Kr.s.n.as. There certainly are persons whose names ending with -gupta but we regularly use without the appellation, such as Abhinavagupta or Prajñākaragupta. There can be many conceivable reasons why gupta is added even if the king is never given that name in inscriptions, including alteration in transmission of the BSBh. 44 Another thing to note is that the name Balavarman appears twice in the BSBh. In BSBh 2.4.1 (Śāstri and Panśı̄kar, 1917, 630) Śaṅkara says the following: sādr.śye saty upamānam . syāt—yathā sim . has tathā balavarmeti (When similarity exists, one may use a simile: “Balavarman is like a lion.”). This would be very flattering for Balavarman. In this regard, an oft-mentioned concluding verse—first referred to by R. G. Bhandarkar in R. G. Bhandarkar (1884, 80) and again in R. G. Bhandarkar (1928, 14–15)—of the Sam . ks.epaśārı̄raka of Sarvajñātman (an alleged disciple of Sureśvara, who himself was a disciple of Śaṅkara) requires our attention. It mentions that it was written when a certain Āditya of Manu’s lineage (manukulāditya) was ruling the earth. The word manukula appears often to refer to the Cāl.ukyas in inscriptions. Thus it is reasonable to consider this Āditya to be one of the Cāl.ukyas as the two Bhandarkars did (R. G. Bhandarkar 1884, D. R. Bhandarkar 1912). From what I have so far discussed, I agree with D. R. Bhandarkar in identifying the Āditya with Vimalāditya who is recorded as a grandson of Balavarman in the Kad.aba copper plate. Whoever the Āditya that Sarvajñātman referred to may have been, I get the impression that the sect of Śaṅkara may have been under protection of the 106 Journal of Indological Studies, No. 18 (2006) 4 Conclusion The narrowed down date I propose for the composition of the BSBh is between 756 and 772. This satisfies most of the relative chronology. The only difficulty is the one that prompted Thrasher (1979) to place Śaṅkara and Man.d.ana Miśra at an early date: the fact that Um . veka mentions variant readings of Man.d.ana’s Bhāvanāviveka. It would appear that there has to be some distance between Man.d.ana and Um . veka, whose lower limit is known in relation to Kamalaśı̄la who arrived in Tibet in 792 and who quotes Um . veka. However, many variants can creep in in a very short time. What is required is one copying by an incompetent copyist.45 The fact that Um . veka mentions variant readings of the Bhāvanāviveka should be considered proof only that he had access to more than one manuscript. References Diwakar Acharya. Vācaspati Miśra’s Tattvasamı̄ks.ā, The Earliest Commentary on Man.d.ana Miśra’s Brahmasiddhi, A Critical Edition with an Introduction and Critical Notes. Ph.D. thesis, University of Hamburg, Hamburg, 2004. Anant Sadashiv Altekar. Rāsht.rakūt.as and Their Times. Oriental Book Agency, Poona, second revised edition, 1967. D. R. Bhandarkar. “Can We Fix the Date of Śaṁkarâchârya More Accurately?” The Indian Antiquary, 41:200, 1912. R. G. Bhandarkar. Early history of the Dekkan: Down to the Mahomedan Conquest. Government Central Press, Bombay, 1884. Cāl.ukyas. There is another Cāl.ukya connection related to the date of Śaṅkara. Several authors refer to the record at the Śr.ṅgeri mat.ha, saying that it maintains that the birth or the year he founded the mat.ha was the 14th year of the reign of Vikramāditya. I have yet to verify how this tradition is kept, orally or inscribed somewhere. Nor am I aware if this is recorded in Sanskrit or in some other language. Cf. Rice (1897, 473) and Vidyasankar Sundaresan http://www.advaita-vedanta.org/avhp/dating-Sankara.html. The question has been who this Vikramāditya may have been. With the date I have arrived at, this Vikramāditya is probably Vikramāditya II of the Cāl.ukyas of Badami. His date of accession is 733 or 734, and the accession of his successor is disputed, but it should fall some time in the period 744–747 (Dikshit, 1980, 163–164, 182). The 14th year of his reign should be somewhere in 746 or 747. This is very close to the year 745, the year of Śaṅkara’s consecration that Rice (1897, 473) calculates from the annals at the Śr.ṅgeri mat.ha. (Rice himself thought that the 14th year of Vikramāditya referred to the Vikrama Sam . vat and therefore did not take this information into account when calculating the dates of Jagadgurus.) 45 One example I am aware of as a manuscript which is close to the composition but in a bad state is that of the Śrāddhakān.d.a of the Kr.tyakalpataru of Laks.mı̄dhara. The composition of the Kr.tyakalpataru is proposed to be 1125–40 and the manuscript is dated 1176. The manuscript contains many mistakes. See Harimoto (forthcoming, n. 22). The Date of Śaṅkara (K. Harimoto) 107 R. G. Bhandarkar. Collected Works of Sir R. G. Bhandarkar, Vol. II, Comprising Reports on Search for Sanskrit Mss. during 1882–91, Religious and Social Writings Etc., Edited by Narayan Bapuji Utgikar . Government Oriental Series. Bhandarkar Oriental Research Institute, Poona, 1928. A. C. Burnell. Elements of South-Indian Palæography, from the Fourth to the Seventeenth Century A.D., being an Introduction to the Study of SouthIndian Inscriptions and MSS . Trübner & Co., London, second enlarged and improved edition, 1878. T. R. Chintamani. “The Date of Śrı̄ Śaṅkarācārya.” Journal of Oriental Research, Madras, 4:39–56, 1929. H. T. Colebrooke. “On the Sanskrı̆t and Prákrı̆t Languages.” Asiatick Researches, or, Transactions of the Society Instituted in Bengal, for Inquiring into the History and Antiquities, the Arts, Sciences, and Literature of Asia, 7:199–231, 1801. Reproduced as Colebrooke (1837, v.2,1–34), Colebrooke (1872, v.1,1–34) and Colebrooke (1873, v.2,1–32). H. T. Colebrooke. “On the Védas or Sacred Writings of the Hindus.” Asiatick Researches, or, Transactions of the Society Instituted in Bengal, for Inquiring into the History and Antiquities, the Arts, Sciences, and Literature of Asia, 8:369–476, 1805. Reproduced as Colebrooke (1837, v.1,9– 113), Colebrooke (1858, 1–69), Colebrooke (1871, 9–122) and Colebrooke (1873, v. 1,8–132). H. T. Colebrooke. Two Treatises on the Hindu Law of Inheritance. A. H. Hubbard, Hindoostanee Press, Calcutta, 1810. H. T. Colebrooke. “On the Philosophy of the Hindus, part III.” Transactions of the Royal Asiatic Society of Great Britain and Ireland, 1:439–461, 1827. Reproduced as Colebrooke (1837, v.1,295–324), Colebrooke (1858, 188– 207), Colebrooke (1871, 295–324), Colebrooke (1873, v.1,319–349). H. T. Colebrooke. “On the Philosophy of the Hindus, part IV.” Transactions of the Royal Asiatic Society of Great Britain and Ireland, 2:1–39, 1830. Reproduced as Colebrooke (1837, v.1,325–377), Colebrooke (1858, 208– 242), Colebrooke (1871, 325–377) and Colebrooke (1873, v.1,350–401). H. T. Colebrooke. Miscellaneous Essays by H. T. Colebrooke, in two Volumes. Wm. H. Allen and Co., London, 1837. H. T. Colebrooke. Essays on the Religion and Philosophy of the Hindus. Williams and Norgate, London, 1858. This is a collection of articles excerpted from the 1937 edition of the Miscellaneous Essays. Only the reprint by Indological Book House, 1972 was accessible. 108 Journal of Indological Studies, No. 18 (2006) H. T. Colebrooke. Miscellaneous Essays by H. T. Colebrooke, in two Volumes, volume 1. Higginbotham and Co., Madras, second edition, 1871. H. T. Colebrooke. Miscellaneous Essays by H. T. Colebrooke, in two Volumes, volume 2. Higginbotham and Co., Madras, second edition, 1872. H. T. Colebrooke. Miscellaneous Essays by H. T. Colebrooke, a New Edition, With Notes by E. B. Cowell, in Two Volumes. Trübner & Co., London, 1873. Paul Deussen. Das System des Vedânta, nach den Brahma-sûtra’s des Bâdarâyan.a und dem Commentare des Çan̄kara über dieselben, als ein Compendium der Dogmatik des Brahmanismus. F. A. Brockhaus, Leipzig, 1883. Durga Prasad Dikshit. Political History of the Chālukyas of Badami. Shakti Malik, New Delhi, 1980. Jonathan Duncan. “Historical Remarks on the Coast of Malabar with Some Descriptions of the Manners of its Inhabitants.” Asiatick Researches, or, Transactions of the Society Instituted in Bengal, for Inquiring into the History and Antiquities, the Arts, Sciences, and Literature of Asia, 5, 1798. J. F. Fleet. “Spurious Indian Records.” The Indian Antiquary, 30:201–223, 1901. Paul Hacker. “Zur Geschichte und Beurteilung des Hinduismus, Kritik einiger verbreiteter Ansichten.” Orientalistische Literaturzeitung, 59:231– 45, 1964. Kleine Schriften, pp. 476–483. Kengo Harimoto. “The Original Skandapurān.a, Laks.mı̄dhara and Can.d.eśvara, Studies in the Skandapurān.a VIII.” Indo-Iranian Journal , forthcoming. Daniel H. H. Ingalls. “Śaṁkara’s Arguments against the Buddhists.” Philosophy East and West, 3(4), 1954. H. Kern. “Geschiedenis van den Godsdienst, tot aan de Heerschappij der Wereldgodsdiensten, geschetst door Dr. C. P. Tiele, Hoogleeraar bij de Remonstraanten, Amsterdam, P. N. van Kampen & Zoon, 1876.” In Verspreide Geschriften, Diversen en Nalezing (Eerste Gedelte), volume 13, pages 57–63. ’s-Gravenhage, Nijhoff, 1877. The collection was published in 1927. K. Kunjunni Raja. “On the Date of Śam . karācārya and Allied Problems.” The Adyar Library Bulletin, 24:125–148, 1960. The Date of Śaṅkara (K. Harimoto) 109 S. Kuppuswami Sastri, editor. Brahmasiddhi by Acharya Man.d.ana Miśra, with Commentary by Śaṅkhapān.i . Number 4 in Madras Government Oriental Manuscripts Series. Madras Government, Madras, 1937. Sengaku Mayeda. A Thousand Teachings, The Upadeśasāhasrı̄ of Śaṅkara, Translated with Introduction and Notes. University of Tokyo Press, Tokyo, 1979. Sengaku Mayeda. Vedānta no tetsugaku [Philosophy of Vedānta] . Heirakuji Syoten, 1980. Vasudev Vishnu Mirashi, editor. Inscriptions of the Kalachuri-chedi Era, Part 1 , volume 4 of Corpus Inscriptionum Indicarum. Government Epigraphist for India, Ootacamund, 1955. Giuseppe Morichini. “Early Vedānta Philosophy.” East and West, pages 33–39, 1960. Hajime Nakamura. Syoki no Vedānta Tetsugaku [Early Vedānta Philosophy], Syoki Vedānta Tetsugakushi [History of the Early Vedānta Philosophy] 1 . Iwanami Syoten, Tokyo, 1950. Hajime Nakamura. “Shankara Tetsugaku no Rekishi-teki Shakai-teki Tachiba [Śaṅkara’s Philosophy in the Historical and Social Contexts].” In Ui Hakuju Hakushi Kinen Ronbunshu, pages 331–372. Iwanami Shoten, Tokyo, 1951. R. Narasimhachar. “Śankarâchârya and Balavarmâ.” tuquary, 42:53–54, 1913. The Indian An- K. B. Pathak. “The Date of Śaṁkarâchârya.” The Indian Antiquary, 11:174– 5, 1882. K. R. Pisharoti. “The Three Great Philosophers of Kerala.” Indian Historical Quarterly, 3:676–92, 1929. Karl H. Potter, editor. Encyclopedia of Indian Philosophies, Advaita Vedānta up to Śam . kara and His Pupils. Princeton University Press, Princeton, New Jersey, 1981. Edeltraud Harzer, S. Subrahmanya Sastri, Allen W. Thrasher, contributors. S. K. Rāmanātha Śāstri, editor. Sphot.asiddhi of Ācārya Man.d.anamiśra with the Gopālikā of R . .siputra Parameśvara. Madras government, Madras, 1931. K. V. Ramesh and S. S. Ramachandra Murthy. “Three Inscriptions of Balavarmma.” Epigraphia Indica, 42:181–186, 1977/78. 110 Journal of Indological Studies, No. 18 (2006) B. V. Krishna Rao. “Revised Chronology of the Eastern Chalukya Kings.” Journal of the Andhra Historical Research Society, 9(4):1–32, 1935. Bhavaraju Venkatakrishna Rao. History of the Eastern Chalukyas of Vengi (610–1210 A.D.). Andhra Pradesh Sahitya Akademi, Hyderabad, 1973. Lewis Rice. “A Rásht.rakût.a Grant from Mysore.” The Indian Antiquary, 12:11–18, 1883. Lewis Rice. Mysore, a Gazetteer Compiled for Government, Volume 1, Mysore in General. Archibald Constable and Company, Westminster, 1897. N. S. Anantakr.s.n.a Śāstri and Vasudev Laxman Sas̄trı̄ Panśı̄kar, editors. Brahmasūtraśāṅkarabhās.yam, Śrı̄mad-Appayyadı̄ks.ita-viracita-Parimalopabr.m . hita-Śrı̄mad-Amalānandasarasvatı̄-pran.ı̄ta-KalpataruvyākhyāyutaŚrı̄mad-Vācaspatimiśrakr.ta-Bhāmatı̄vilasitam. Nirnaya Sagar Press, Bombay, 1917. Annette Schmiedchen. “Die Genealogiepassagen in den Herrscherkunden der Rās.t.rakūt.a-Dynastie, Norm und Abweichung in der Frühmittelalterlichen Sanskrit-Epigraphik.” In Akten des 27. Deutschen Orientalistentages (Bonn, 28. September bis 2. Oktober 1998): Norm und Abweichung, edited by Stefan Wild and Hartmut Schild, pages 241–253. Ergon Verlag, Würzburg, 2001. K. T. Telang. The Bhagavadgı̂tâ with the Sanatsugâtı̂ya and the Anugı̂tâ, volume 8 of The Sacred Books of the East. Clarendon press, Oxford, 1882. K. T. Telang. “The Date of Śaṅkarâchârya.” The Indian Antiquary, 13:95– 103, 1884. Allen Wright Thrasher. “The Dates of Man.d.ana Miśra and Śaṅkara.” Wiener Zeitschrift für die Kunde Südasiens, XXII:116–139, 1979. C. P. Tiele. Geschiedenis van den Godsdienst, tot aan de Heerschappij der Wereldgodsdiensten. P. N. van Kampen & Zoon, Amsterdam, 1876. C. P. Tiele. Outlines of the History of Religion to the Spread of the Universal Religions. Trübner and co., Ludgate Hill, London, 1877. S. V. Venkateswara. “The Date of Sankaracharya.” Journal of the Royal Asiatic Society, pages 151–161, 1916. Tilmann Vetter. Studien zur Lehre und Entwicklung Śaṅkaras, volume VI of Publications of the de Nobili Research Library. Gerold, Wien, 1979. The Date of Śaṅkara (K. Harimoto) 111 H. H. Wilson. Dictionary, Sanscrit and English. P. Pereira, Calcutta, first edition, 1819. Only the 1979 Indian reprint by Nag Publishers was accessible. F. H. H. Windischmann. Sancara, sive de Theologumenis Vendanticorum. Impensis T. Habichtl, Bonn, 1833. Yajñeśvara Ciman.a Bhat.t.a. Āryavidyā Sudhākara, or a Compendium of Hindu culture and Philosophy. Moti Lal Banarsi Dass, Lahore, new edition, 1940. (Keywords: Śaṅkara, Brahmasūtrabhās.ya, Śaṅkara’s date)