Please Wait...

Paippalada-Samhita of The Atharvaveda (Volume Three)

About the Book

The Paippalada-Samhita of the Atharvaveda was originally the most prominent branch of the Atharvaveda and was known as such to Yaska, Panini, and author of the Mahabhasya and even later. The Paippalada-Samhita often reveals its closeness to the common people. Its decline in the middle ages seems to have been related to that and the parallel rise of the Saunakiya-Samhita as the main text the Atharvaveda. A mutilated and hopelessly corrupt Sarada manuscript of the Paippalada-Samhita was known from the Paippalada-Samhita was known from the seventies of the nineteenth century through the efforts Rudolph von Roth. But with its lacunae and underminable readings (ed. L.C. Barret, 1905-1940) it hardly generated any decisive philological research into the Atharvaveda. The situation changed with late Durgamohan Bhattacharyya’s discovery, after years of search, of complete and much better manuscripts in the Oriya script which present edition is based The discovery was hailed ‘as one of the greatest events in Indology’ (Ludwig Alsdorf). Durgamohan Bhattacharyya (d. 1965) edited the first four Kandas (Sanskrit College, 1964, 1970) resulting in a ‘precious store of material... (being) spread out before our eyes’ (karl Hoffmann) but did not have the opportunity to complete the work. The book consists of twenty kandas. The first two volumes of the present Asiatic Society edition (Kandas 1-15, 3771 verses in 510 hymns; 16, 1963 verses in 155 hymns) were published in 1997 and 2008 (BI Series no. 319). This volume comprises the seventeenth and the eighteenth kandas in 55 hymns (496 mantras) and 82 hymns (663 mantras) respectively.

About the Author

Dipak Bhattcharya’s (b. 1940) work with the Paippalada-Samhita began in 1964 with the translation (Bhavan’s Journal XI.2) of a new hymn of the text (9.4) and has continued. He was awarded the Colette Caillat Prix for the year 2009 by the Institute de France for his previous works on the first and the second volume of the Paippalada-Samhita (Asiatic Society 199 and 2008) and to aid him in carrying out work on the remaining volumes. He was fortunate to have continuous advice from late Professor F.B.J. Kuiper while editing a part of the Paippalada-Samhita as a ZWO Fellow (1981-82) at Leiden. His publications on Indology are more than fifty and include, apart from the two volumes of the Paippalada Samhita, two books Paippalada-Samhita, two books on Vedic Philology and the edition of collections and papers on various branches of Indology. Bhattacharya retired as Professor of Sanskrit, Visva Bharati in 2005 after serving there for twenty-years.


The present publication is the Third Volume of The Paippalada-Samhita of the Atharvaveda, of a series of four-volume edited work on the Atharvaveda. This Samhita consists of the seventeenth and eighteenth Kandas with 55 and 82 hymns respectively. The 55 hymns are made of 496 mantras and the 82 hymns of 663 mantras making a total of 1159 mantras. The work is based on a critical edition of the manuscripts of Paippalada-Samhita originally found in Orissa by the Late Durgamohan Bhattacharyya, the celebrated father of the present Editor, who discovered these manuscripts from Orissa with great effort.

Apart from giving an account of the Atharvaveda hymns and meticulously analysing the rituals, charms and prayers and the concept of the universe of the Athrvavedins, the Introduction to this work gives us a critical idea of the whole literature on it and also analyses the various debates relating to the different versions of the Paippalada-Samhita and the implications of these different versions. The author has also gone into the question of the early association of the Atharvavesa hymns with some particular region. He disputes its association with either Kashmir region or the Kulu valley or the Ahmedabad region. As he says, “though a manuscript of the Paippalada-Samhita was first discovered in Kashmir, there are reasons to suppose that the Atharvaveda had not been introduced in that place before the 15th century.” On the other hand, the idea of its early association with Karnataka region, as advocated by Durgamohan Bhattacharyya and before him by V. Raghavan is more acceptable. From that region, the journey to the East started. As the author says, ‘There are enough evidences of medieval movements from Karnataka, northeast to it and towards eastern India in particular. Thus, the Calukyas of Pistapura, the Gangas of Orissa, many officers appointed by the Palas, the Senas of Bengal, Nanyadeva’s dynasty in Mithila and the eastern Kadambas were of Karnata origin...It makes one think of the possibility of the early Angirasi settlements having grown under the impact of the Karnata movement.” (p.cix). The scholarly method through which this conclusion is made draws our attention to the manifold implications of how to study manuscripts and reconstruct history through such studies.

This third volume of the Paippalada-Samhita is being published a little late but we feel confident that this work will be highly useful to the students of the Atharvaveda and of the AVP.


The presented third volume of the Asiatic Society edition of the Paippalada-Samhita consists of its seventeenth and eighteenth kandas with 55 and 82 hymns respectively. The 55 hymns are made of 496 mantras and the 82 hymns of 663 mantras making a total of 1159 [=496 + 663 (402+261:18.1-56+57-82:)] manritras. But, for reasons stated below, the above number does not reflect a correct estimate of the volume and the quantity of material in .the two Kandas.

In 1984 (BHAITACHARYA: 174) the number of mantras in the eighteenth kanda had been provisionally mentioned as 676 while in the Introduction to the first volume (1997: xxii) it was given as 734. I owe an explanation for this. The first one was caused by the fact that the Bengali transcript (Introduction, Vol. I: xvii) consisted of only the first 56 hymns of the eighteenth Kanda and lacked the 26 hymns to the Fathers (18.57-82).I noted their existence in the Orissa manuscripts as late as 1980. A rough estimate of the number of mantras with an average of eight per hymn led to the number 208 that appeared in the 1984 paper in the Indo- Iranian Journal. The actual number later turned out to be 261. Even with the wrong estimate the total number should have been, (402 + 208=) 610. But this was not to be so; for, in the Bengali. transcript the number of mantras in the first 56 hymns had been counted as 468. These 468 plus the estimated 208 made up the number 676. In the final count both the numbers changed. The numbers in the two parts of the kandas have been settled at 402 I and 261 making a total of 663.

The differences in the result of counting took place because there are different ways of counting the mantras without meaning any difference in content. The Bengali transcript followed the numbering in the V.V.R.I. edition of the AVS (Vishva BANDHU) while I tried to follow the indications in the manuscripts. They differ. The following is an example. In the V.V.R.I. edition the number of mantras in AVS 15.3 (AVP 18.29) is given as 11. Since the two versions are not identical the general norm of the AVS edition was followed in the Bengali transcript to arrive at the number 8. But none of the three AVP manuscripts mentions the total number of verses. In fact the 'hymn' being a prose passage there are no verses at all here. The number of sentences too is not given in the mss. In the absence of metres there was no compulsion of showing any verse count. So the passage has been shown as one in the present edition. In this way the total number of hymns came down without meaning any decrease in material content. If the way of the Vulgate had been followed the number of mantras would have been much higher.

In this connection I should mention here that I checked the Vulgate's actual numbering of the sentences in some such passages from a so far unknown padapatha manuscript of it and found, as far as examined, that the numbering in the existing editions was correct. Sometimes the new manuscript helped in determining which reading is the most common one in the Vulgate and in comparing the same with the AVP reading and in determining the common Atharvavedic convention about hymn and anuvaka units (below 1.2, 1.3). I indicated the new Vulgate ms with the sign Gi or A full description of the new ms may be published later.

1.2.Kanda 17 and its new material
1.2.1.AVP-AVS correspondences

Barring its second anuvaka i.e. 17.7-11 = AVS 10.7, the known mantras of the seventeenth kanda correspond to the twelfth kanda of the Vulgate. The kanda begins with the corresponding verses of AVS 12.1. The third anuvaka i. e, 17.12-15 is new, The fourth, i.e. 17.16-20, corresponds to AVS.12.4. The fifth anuvaka i.e. 17.21-26, again, is new. So is the sixth anuvaka i.e.. 17. 27-43 describing an unknown Anadudvrata 'The vow about the bull'. The long prose passage narrates how Indra had killed Vrtra through the said rite. The seventh anuvaka that is 17.44-49 corresponds to AVS 12.2 and the final one to AY5 12.3.

Table 1

AVP-AVS correspondence: Seventeenth Kanda

AVP anuvakaAVP hymnsNo. Of versesKAVS
Total 496

In its twelfth kanda the AVS has a 'decad’-division that had been noted by WHITNEY. In Gi too the verses in each decad' are independently numbered. The serial number of the 'decad' is not given but each one is separated from the next by the sign The same process is to be observed in kanda 18 too (below §1.3.3 and fn.13). The units at present' regarded in modern studies and editions as hymns and employed in indicating location are actually anuvakas in the AVS tradition. This is evident from Gi's statement at the end of the last verse of 12.1: bhaumas tryadhika sasthih// prathamonuvakah// 'The first anuvaka praising the earth has sixty-three (verses). ‘The authenticity of this convention of regarding the 'decads' as hymns and the bigger units as anuvakas is proved by the fact that in the AVP the 'decads' have been treated as kandikas and the bigger units as anuvakas. This is also proved by their numbering in K and by Sayana in the eighteenth kanda.

There is no 'decad' division in the prose passages of the fifth and the sixth anuvaka that is 17.21-43 found in the AVP only. K-deviates in the sixth anuvaka that is 17.27-43 in Or. The anuvaka forms a single story and is divided into seventeen prose passage understood as kandikas in Or. K does not show the kandika division in 29-43. But the original kandika number occasionally occurs. Obviously the story was given in its correct form in some mother K-manuscript. The story is in easy prose (see below) making its plot-unity very much apparent. That might have been the cause for missing the division into small passages by the copyist who had understood the meaning of the story.

The existing conventions of indicating the location of AVS mantras belonging to the twelfth and the eighteenth kanda with reference to the anuvaka and of the AVP parallels with reference to the kandika are unnecessarily different because their, unit division is the same in the existing traditions of both the schools. There should be a common way of indicating the location of verses. It should be shown either with reference to the 'decads' or to the anuvakas in both the AVP and the AVS: Apparently, the meaning unity in the anuvakas should favour the latter. But in kanda 18 there is some discrepancy in this regard (see below 1.3.3). I prefer the former.

Some of the new mantras belong to the Angirasa category. The third anuvaka (forty mantras) prays against hideous beings, mostly non-human but real, like insects, small vertebrates etc. Such hymns resemble RV. 191 in contents and employment and should be awarded an intermediate status between Angirasa and At harvana, otherwise called ghora and santa, because the prayers therein do not intend to harm human beings but are meant for their well-being. In the fifth anuvaka the first two hymns comprise twenty sentences partly in prose and of Angirasa character. Its remaining hymns (forty-six mantras) deal with dreams and nightmares.

1.2.2 The anadudvrata, Ahinas Asvatthi and other material The sixth anuvaka i.e. 17.27-43 describing an unknown anadudvrata 'vow about the bull' is of purdkalpa-arthavada character. The long prose passage recounts. the archetype how Indra had killed Vrtra by recovering his thunderbolt that had slipped down from his hand. It can lay better claim as the beginning of Vedic prose than the prose passages in the Black Yajurveda. The antiquity of the composition can be proved by the occurrence of the adverbial pair yat and tat ('that' and 'so') which went out of use after the Atharvaveda and will not be found in the Yajurveda. There are other indications of the archaic character of the sixth anuvakaboth ritually and stylistically.

The story has an anecdote where one Ahinas Asvatthi claimed the rite to be a krtya but refrained himself from censuring the knowledge of the rite lest he lost his accrued merits. The first four words of the sentence comprising Ahinas Asvatthi's observation na tad brahmanam nindani yad enam asrnon etc. (17.35) 1 should not censure the Brahmana that he listened to it' were purportedly cited by Vamana on P.7.1.39 as na tad brahmanad nindiimi. The use of ablative for accusative in brahmanad, supposedly after the relevant rule of Panini, is wrong and does not express the meaning of the sentence. The grammarian has obviously missed the sense of tad and the very reading brahmanam allowing us to draw the conclusion that he only knew the example handed down by a defective grammatical tradition but not the ritual context. WACKERNAGEL-DEBRUNNER (AiG III: 501) thought that the wrong citation in Kasika originated in some defective Vedic text tradition, but a flawless Vedic tradition still exists.

Sample Pages

Add a review

Your email address will not be published *

For privacy concerns, please view our Privacy Policy

Post a Query

For privacy concerns, please view our Privacy Policy


Related Items