Sacred Texts  Hinduism  Index  Previous  Next 
Buy this Book at

The Grihya Sutras, Part 2 (SBE30), by Hermann Oldenberg, [1892], at

p. 1


p. 2 p. 3




The Grihya-sûtra of Gobhila differs from those of Sâṅkhâyana, Âsvalâyana, Pâraskara, Hiranyakesin in one essential point: while these texts presuppose only the same Vedic Samhitâs on which also the corresponding Srauta-sûtras are based, viz. the Rig-veda-Samhitâ, the Vâgasaneyi-Samhitâ, and the Taittirîya-Samhitâ; the Sûtra of Gobhila, on the other hand, presupposes, beside the Samhitâ of the Sâma-veda 1, another collection of Mantras which evidently was composed expressly with the purpose of being used at Grihya ceremonies: this collection is preserved to us under the title of the Mantra-Brâhmana, and it has been edited at Calcutta (1873), with a commentary and Bengali translation by Satyavrata Sâmasramin 2.

Prof. Knauer of Kiew, to whom all students of the Grihya literature are highly indebted for his very accurate edition and translation of Gobhila, has been the first to

p. 4

examine into the relation in which the two texts, the Mantra-Brâhmana and the Gobhilîya-sûtra, stand to each other. He has very kindly enabled me to make use, before they were published, of the results of his investigations, which he has laid down in the introduction to his translation of Gobhila. While I wish, therefore, to acknowledge the obligation under which Prof. Knauer has thus laid me, I must try, on the other side, to state my own opinion as to the problem in question, which in some points differs from, or is even opposed to, the theory by which Prof. Knauer has tried to solve it.

To begin with that side of the question regarding which there can scarcely be any doubt: it is certain, I believe, that Gobhila supposes the Mantra-Brâhmana to be known to the students of his Sûtra. The reasons which show this are obvious enough 1. By far the greater part of the Mantras of which Gobhila quotes the first words, are not found in the Sâma-veda nor, for the most part, in any other Vedic Samhitâ, except in the Mantra-Brâhmana, in which they stand in exactly the same order in which they are referred to by Gobhila. The descriptions of the Grihya sacrifices by Gobhila would have been meaningless and useless, and the sacrificer who had to perform his domestic ceremonies according to the ritual of Gobhila, would have been unable to do so, unless he had known those Mantras as contained in the Mantra-Brâhmana. And not only the Mantras, but also the order in which the Mantras stood, for Sûtras such as, for instance, Gobh. II, 1, 10 ('With the two following verses he should wash,' &c.), would have no meaning except for one who had studied the Mantra-Brâhmana which alone could show which 'the two following verses' were.

There are, consequently, two possibilities: either the Mantra-Brâhmana existed before the Gobhilîya-sûtra, or the two works have been composed together and on one common plan. It is the first of these alternatives which Prof. Knauer maintains; I wish, on the other hand, to call

p. 5

the attention of Vedic scholars to some facts which seem to me to render the second more probable.

A great part of the Mantras which have to be recited, according to Gobhila, at the performance of the Grihya ceremonies, are not given in the Mantra-Brâhmana, but they are either found in the Sâma-veda-Samhitâ and then their Pratîkas are quoted by Gobhila, or they are cited by Gobhila in extenso. Thus for the ceremonies described in the first Prapâthaka of Gobhila, such as the morning and evening offerings and the sacrifices of the full and new moon, the Mantra-Brâhmana gives, with one single exception, no Mantras at all 1; but those Mantras, most of which consist only of a few words, are given by Gobhila only. It is scarcely to be believed that in a Samhitâ which had to contain the Mantras required for the performance of the Grihya sacrifices, the Mantras belonging to the two daily and the two fortnightly sacrifices, which occupy one of the first places among all Grihya ceremonies and are treated of accordingly in all Grihya-sûtras, should have been omitted, unless that Samhitâ was intended to stand in relation to another text by which that deficiency was supplied: and the Gobhilîya-sûtra exactly supplies it. Prof. Knauer thinks that those Mantras were omitted because they had already found their place in the Srauta ritual; but we must not forget that in the Srauta ritual o f the Sâma-vedins neither the Agnihotra nor the Darsapûrnamâsa sacrifices, which are performed without the assistance of priests of the Udgâtri class, are treated of. Moreover the one Mantra to which we have already alluded 2, the single one which corresponds in the Mantra-Brâhmana to the first book of Gobhila, seems to me quite sufficient to show that it was not the intention of the compiler of that text to disregard that group of sacrifices; he gave that Mantra only, because the other Mantras, consisting of but a few words, were given in extenso in the Gobhila text. The Mantra of which we speak, belongs to the description

p. 6

of the paryukshana of the sacrificial fire. The sacrificer pours out water to the south, the west, and the north of the fire, with the Mantras, 'Aditenumanyasva,' 'Anumatenumanyasva,' 'Sarasvaty anumanyasva'; then he sprinkles water round the fire once or three times with a longer Mantra, 'Deva Savitah prasuva yagñam prasuva yagñapatim bhagâya. Divyo gandharvah ketapûh ketam nah punâtu. Vâkaspatir vâkam nah svadatu.' This last one is the Mantra given in its entirety in the Mantra-Brâhmana, while Gobhila 1 has only the first words of it. To assume here that the author of the Mantra-Brâhmana knew only of that one Mantra, and that at the time of Gobhila the custom of the Sâma-vedins had undergone a change, so that they used four Mantras instead of the one, would be, in my opinion, an artificial and not very probable way of explaining the facts; a much more natural supposition would be, I believe, that the Sûtra and the Mantra-Brâhmana describe one and the same form of the ceremony, so that the Brâhmana, by omitting the short Mantras, which were given in the Sûtra in their entirety, implicitly refers to the Sûtra, and the Sûtra, on the other hand, by quoting only the first words of the longer Mantra, refers to the Brâhmana in which the full text of that Mantra was given.

Among the numerous ceremonies described by Gobhila, which could furnish the occasion for similar remarks, we select only two: the rites performed in the evening of the wedding-day 2, and the sacrifice on the full-moon day of Âsvayug3. The bridegroom, having carried away his bride from her home, takes her to the house of a Brâhmana, and when the stars have appeared, he makes six oblations with the six verses lekhâsandhishu pakshmasu (Mantra-Br. I, 3, 1-6): these are given in the Mantra-Brâhmana, and Gobhila has only the Pratîka. Then follow two short Mantras: the bride, to whom the polar-star has been shown, addresses that star with the words: dhruvam asi dhruvâham patikule bhûyâsam amushyâsâv iti;

p. 7

and when she sees the star Arundhatî, she says, ruddhâham asmi. As the full wording of these Mantras is given by Gobhila, they are omitted in the Brâhmana. Finally the bridegroom recites over the bride the Rik dhruvâ dyaur dhruvâ prithivî, &c.; this we find in the M.-B. (I, 3, 7), the Pratîka only being quoted by Gobhila. If one were to suppose here that in the two texts two different stages in the development of this ceremony are represented, so that only the Mantras lekhâsandhishu and dhruvâ dyauh would belong to the more ancient form of it, while the Mantras dhruvam asi and ruddhâham asmi would have been introduced at a later time, it may perhaps not be possible to disprove, in the strictest sense of the word, such an opinion. But I think the data we have given point to another solution of the problem which, if not the only admissible, is yet the most probable and natural one. Gobhila gave the full wording of the shorter Mantras with which the description of the ceremony could be interwoven without becoming obscure or disproportionate; the longer Mantras would have interrupted, rather tediously and inconveniently, the coherency of his ritual statements; so he separated them from the rest of his work and made a separate Samhitâ of them. It is true that there are some exceptions to the rule that all long Mantras are given in the Mantra-Brâhmana and all short Mantras only in the Sara: on the one hand, there are some Mantras of considerable extent that are given by Gobhila and omitted in the Brâhmana, thus, for instance, the Mantra yady asi saumî used at a preparatory ceremony that belongs to the Pumsavana 1. On the other hand, a number of short Mantras which Gobhila gives in extenso, are found nevertheless also in the Mantra-Brâhmana: such is the case, for instance, with many of the Mantras belonging to the worship of the Fathers, Gobhila IV, 2. 3, Mantra-Br. II, 3.

p. 8

[paragraph continues] It appears then, that allowance must be made for a certain inconsistency or carelessness in the distribution of the material between the two texts: and such an assumption will easily be allowed by any one who does not entertain very exaggerated ideas as to the care and reflection which presided over the composition of the Sûtra texts.

I will add only a few words concerning a second Grihya ceremony, which calls for the same sort of comment as the rites which have just been discussed. For the offering on the day of the full moon, Gobhila prescribes (III, 8, 2) first the verse â no mitrâvarunâ, second the verse mâ nastoke. The Mantra-Brâhmana (II, 1, 8) has the second of these verses only, not the first: conversely, the first verse alone, and not the second, is to be found in the Samhitâ of the Sâma-veda (I, 220). We could hardly assume, as I think, that the Mantra-Brâhmana presupposed another form of the rite differing from Gobhila's; we should be much more inclined to consider the leaving out of that matter, which was contained in other texts of the Sâma-veda, as a proof that the compiler of the Mantra-Brâhmana assumed that those texts were known 1.

And this brings me to one of Prof. Knauer's conjectures concerning the Mantra-Brâhmana which I have not yet touched. According to tradition we consider the Mantra-Brâhmana as belonging to the Sâma-veda; in the Calcutta edition it is designated as the 'Sâma-vedasya Mantra-Brâhmanam.' Prof. Knauer thinks that it is doubtful whether the Mantra-Brâhmana belonged to the Sâma-veda originally. He conjectures 2 'that it existed already in the

p. 9

period during which the separate schools were as yet in the process of sifting, when there were as yet no Sâma-vedists in the later and stricter sense of the term 1.' For out of 249 Mantras of the Mantra-Brâhmana there are only four which are found in the Sâma-veda 2, as Prof. Knauer has shown, while a much greater number of these Mantras. occur in the other Vedic Samhitâs. I should be inclined to conclude the other way: just because the author of the Mantra-Brâhmana presupposed a knowledge of the Samhitâ of the Sâma-veda, but not of the other Vedas—or in other words because he destined his work for Sâma-vedins, he did not need to repeat what was in the Sâma-veda, but was compelled to incorporate in his compilation the Mantras out of the Rig-veda or of the Yagur-veda 3. Moreover, I would draw the same conclusions from the Mantras cited by Gobhila which are absent in the Mantra-Brâhmana, as I did from the Mantras which occur in the Mantra-Brâhmana, but are not to be found in the Sâma-veda. Those Mantras are all to be found in the Sâma-veda with the exception of those which Gobhila has in extenso, and which therefore could be omitted in the Mantra-Brâhmana. If we examine the thirteen Mantras collected by Prof. Knauer (p. 29), we find that in the case of nine of them the passage of the Sâma-veda (always of the first Ârkika of the Sâma-veda) where they are to be found is quoted by Prof. Knauer.

p. 10

The four other cases are:

rikam sâma yagâmahe, Gobh. III, 2, 48.

tak kakshur devahitam, III, 8, 5.

sam anyâ yanti, III, 9, 7.

pragâpataye, IV, 7, 36.

[paragraph continues] Of these Mantras the first is contained in the Sâma-veda (I, 369) just as the nine first-mentioned ones; the second is quoted by Gobhila in extenso; the third is to be found in the Âranyaka division of the Sâma-veda I (vol. ii, p. 292, ed. Bibl. Ind.); in the fourth finally the text is corrupt; it is intended for the verse out of the Mantra-Brâhmana Pragâpate na tvad etâny anyah. Thus the four apparent exceptions all vanish, and we have in the Mantras which are absent in the Mantra-Brâhmana a new proof that this text belongs to the literature of the Sâma-veda 1.

Thus, according to my view, we may describe the origin of the Mantra-Brâhmana as follows. The Sâma-veda contained in its Samhitâ a much smaller number of Mantras applicable to the Grihya rites than either the Rig-veda or the Yagur-veda; the peculiar character of the Sâman texts, intended for musical recitations at the most important sacrificial offerings, was quite remote from the character of formulas suitable for the celebration of a wedding, for the birth of a child, for the consecration of fields and flocks. Hence it is that, to a much greater extent than Âsvalâyana or Pâraskara, Gobhila mentions Mantras for which a reference to the Samhitâ was not sufficient; and this led to the compiling of a separate Samhitâ of such Grihya-mantras, which presupposes the Grihya-sûtra, just as the latter presupposes this Samhitâ. The almost perfect agreement of the Mantra-Brâhmana with Gobhila furnishes a valuable

p. 11

warrant for the good preservation of the two texts; of small discrepancies I have noted only two: Mantra-Brâhmana I, 6, 15, the formula âgantrâ samaganmahi is given for the ceremony of the Upanayana, while Gobhila does not prescribe this Mantra for this ceremony, although other Grihya texts do; and secondly, the Mantra-Brâhmana II, 5, 1-7 does not consist of six verses as Gobh. IV, 6, 5-6 allows us to assume, but of seven verses.

In concluding this introduction notice is to be drawn to the fact that the text of Gobhila has preserved for us the traces of a division differing from the one into four Prapâthakas which is handed down by tradition: in a number of places certain Sûtras or the last words of certain Sûtras are set down twice, a well-known way of indicating the close of a chapter. This repetition, besides occurring at the end of the first, third, and fourth Prapâthaka (not at the end of the second), is to be found in the following places which become more frequent towards the close of the work: I, 4, 31; III, 6, 15; IV, 1, 22; 4, 34; 5, 34; 6, 16.


3:1 The term 'Samhitâ of the Sâma-veda' ought to be understood here in its narrower sense as denoting the so-called first book of the Samhitâ, the Khanda-ârkika or collection of Yoni verses (see on the relation between this collection and the second book my remarks in the Zeitschrift der Deutschen Morgenländischen Gesellschaft, vol. xxxviii, pp. 464. seq.). Prof. Knauer in his list of the verses quoted by Gobhila (p. 29 of his translation of the Gobhilîya-Grihya) states that Sâma-veda II, 1138 (= I, 276) and 1139 is quoted in Gobhila III, 9, 6, but an accurate analysis of the words of Gobhila shows that the verse II, 1139 is not referred to, so that only the verse II, 1138 remains, which occurs also in the first book of the Samhitâ. The 'dvika' of which Gobhila speaks in that Sûtra is not a dvrika, but, as the commentators rightly understand it (see Knauer's edition of the text, p. xii), it is a dyad of Sâmans or melodies, the two Kâvasha Sâmans which are based on the text I, 276, and are given in the great Sâma-veda edition of Satyavrata Sâmasramin, vol. i, pp. 566, 567.

3:2 In the same way the Grihya-sûtra of Âpastamba stands in connection with a similar collection of Grihya verses and formulas, the Âpastambîya-Mantrapâtha.

4:1 Cf. Knauer's Introduction, pp. 24, 31 seq.

5:1 Cf. Knauer's translation, Introduction, p. 25.

5:2 Mantra-Brâhmana I, 1, 1.

6:1 Gobh. I, 3, 4.

6:2 Gobh. II, 3, 17 seq.

6:3 Gobh. III, 8.

7:1 Gobh. II, 6, 7. It is possible, though we have no positive evidence for this conjecture, that such statements regarding preparatory or auxiliary ceremonies may here and there have been added to the Sûtra collection in a later time. The Khâdira-Grihya (II, 2, 20) has instead of that long Mantra only a few words which in the Gobhilîya-sûtra stand at the end of it.

8:1 Any one who holds the view that the ritualistic formulas, which are not contained in the Mantra-Brâhmana, represent later extensions of the ceremonies in question, will do well to notice how in any one of the offerings of the Srauta ritual which we possess, both in the old description of the Samhitâ and Brâhmana texts, and in the more recent description of the Sûtra texts, Mantras have been added in more recent times to the former ones. I think that it would be difficult to draw from such observations any argument of analogy calculated to support Dr. Knauer's opinion as to the relation of the Mantras in Gobhila and in the Mantra-Brâhmana.

8:2 Introduction to his translation, p. 23.

9:1 Besides the reasons given below in opposition to this conjecture, I may be permitted to point out that this hypothesis is contrary to the whole chronology of the Grihya literature which we endeavoured to arrive at in the general introduction. It is a priori extremely improbable that there was a Grihya Samhitâ at a time when there was as yet no Sâma-veda.

9:2 Viz. (according to Prof. Knauer's alphabetical list of the Mantras of the Mantra-Brâhmana) imam stomam arhate, M.-B. II, 4, s = Sv. I, 66; II, 414; tat savitur varenyam, M.-B. I, 6, 29 = Sv. II, 812; bharâmedhmam, M.-B. II, 4, 3 = Sv. II, 415; sakema tvâ, M.-B. II, 4, 4 = Sv. II, 416.

9:3 Notice that of the four exceptional cases which we put together in the previous note, three cases are Mantras which are found only in the second Ârkika of the Sâma-veda, not in the first (cf. above, p. 3, note 1). The fourth verse (M.-B. II, 4, 2) is to be found in the first Ârkika, it is true, but it stands closely related to two verses which are not to be found in that Ârkika (M.-B. II, 4, 3. 4). This explains why it was put into the Mantra-Brâhmana, as well as those two verses.

10:1 One will not object that the Mantras in question which are absent in the Mantra-Brâhmana are all to be found in the Rig-veda as well as in the Sâma-veda. Since almost all the verses of the Sâma-veda are taken from the Rig-veda there is nothing astonishing about this. Before one could conclude from this that the Mantra-Brâhmana belongs to the Rig-veda he would have to answer the question, How is it that the verses in question are always verses of the Rim veda which are repeated in the Sâma-veda? Why are there not among them verses which are not to be found in the Sâma-veda?

Next: I, 1