Sacred Texts  Zoroastrianism  Index  Previous  Next 
Buy this Book at

Pahlavi Texts, Part IV (SBE37), E.W. West, tr. [1892], at

p. xxviii p. xxix


ATTENTIVE readers of the Sacred Books of the East have had ample opportunities of becoming acquainted with the Zoroastrian scriptures, so far as these have been preserved by the Parsis. In vols. iv, xxiii, and xxxi they have translations of all the texts extant in the original language of the Avesta, excepting a few fragments which are not yet collected. And in vols. v, xviii, and xxiv they have translations of later Pahlavi texts, showing how faithfully the old doctrines and legends were handed down by the priests of Sasanian times to their immediate successors. But they will also have noticed that the translators of these texts are well aware of the fact that the texts themselves are mere fragments of the religious writings of the Zoroastrians, which owe their preservation to the circumstance that they were those portions most usually committed to memory by the priesthood, such as the liturgy, sacred myths, and ceremonial laws. The object of the present volume is to add to those fragments all the accessible information, that can be collected from Irânian sources, regarding the contents of the whole Zoroastrian literature in Sasanian times.

It has been long known that this literature was contained in twenty-one Nasks, or treatises, named either from the nature of their contents, or from their initial words, and each having one of the twenty-one words of the Ahunavair attached to it as a kind of artificial reminder of their proper order and number while enumerating them. Very brief statements of the contents of each Nask have also been accessible in manuscripts of the Persian Rivâyats, such as those translated in pp. 419-438 of this volume. And the existence of a much longer account of the Nasks in the Dinkard was ascertained by Haug, who published

p. xxx

some extracts from it in 1870, when describing several of the Nasks in the Index to the Pahlavi-Pâzand Glossary. He was unable to do more, on account of the defective state of all modern manuscripts of the Dinkard, in which a large portion of the text of the description of the Nasks, in the eighth and ninth books, is missing in various places without any hint of the omissions. These defects were owing to the abstraction of 52 folios of this part of the Irânian manuscript of the Dinkard, after it was brought to India and before any copy of it had been written; and, even now, two of these folios are still missing, as stated in pp. 262, 270. The importance of recovering these 52 missing folios was due to the fact that they contain the text of Dk. VIII, Chaps. VII, 5-XIX, 36, XXXI, 31-XXXVIII, 19, XLIV, 34-XLVI, 5, and Dk. IX, Chaps. I, 1-XI, II, XII, 15-XLVII, 17, or nearly half the text of the two books.

Regarding the early history of the Dinkard there exists a detailed statement in the last chapter of its third book, which can now be translated with greater precision than was possible in 1867, when Haug published its Pahlavi text, with an English translation, in his introduction to the Farhang-i Oîm-aêvak, or Zand-Pahlavi Glossary. In this historical statement it is evident that §§ 1-8 refer to the traditional history of the Zoroastrian scriptures generally, considered as the original source of the information contained in the Dinkard; but §§ 9-13 may be accepted as the actual history of the compilation of the work itself, the facts of which may, very possibly, have all been within the personal knowledge of the writer of the statement. The Pahlavi text of this statement, as preserved in the manuscripts B and K (see pp. xxxv-xxxviii and 2), may be translated as follows:—

‘1. About the Dinkard scripture (nipîk), from the Exposition of the Good Religion, there is this:—The Dinkard scripture is a work which is adorned with all wisdom, and a publication of the Mazda-worshipping religion. 2. And, first, the work—which was derived from the good religion of those of the primitive faith, and which was the knowledge revealing the good religion of the

p. xxxi

prophet (vakhshvar) Spîtâmân Zaratûst, whose guardian spirit is reverenced, and his first disciple through asking and hearing the sane reverenced guardian spirit—is information which is a similitude of enlightenment on every subject from the original light. 3. And those original questions and the decision of the exalted ruler Kaî-Vistâsp to have them written were its origin, and he ordered them to deliver the original to the treasury of Shapîgân 1, and to distribute copies provided. 4. And, after that, he sends a copy to the fortress of documents, to keep the information also there.

‘5. And during the ruin that happened to the country of Irân, and in the monarchy, owing to the evil-destined villain Alexander, that which was in the fortress of documents came to be burnt, and that in the treasury of Shapîgân 2 into the hands of the Arûmans, and was translated by him even into the Greek language, as information which was connected with the ancients (min pêsînîgân padvastakŏ).

‘6. And that Artakhshatar, king of kings, who was son of Pâpak, came for the restoration of the monarchy of Irân, and the same scripture was brought from a scattered state to one place. 7. The righteous Tôsar of the primitive faith, who was the priest of priests 3, appeared with an exposition recovered from the Avesta, and was ordered to complete the scripture from that exposition. 8. He did so accordingly (ham-gûnakŏ), to preserve a similitude of the splendour of the original enlightenment in the treasury of Shapîgân 4, and was ordered to distribute copies of the information provided.

‘9. And after the ruin and devastation that came from the Arabs, even to the archives (dîvân) and treasures of the realm, the saintly 5 Âtûr-farnbag, son of Farukhŏ-zâd, who became the leader of the orthodox, brought those copies, which were scattered on all sides, and new resources, back from dispersion into union with the archives of his residence; and, through observance and consideration for the Avesta and Zand of the good religion, he made the sayings of those of the primitive faith again a similitude of the illumination (fîrôkŏ) from that splendour.

p. xxxii

‘10. Through the awful displeasure (or defect) and ruin (or injury) that happened to Zaratûst, son of Âtûr-farnbag, who became the leader of the orthodox, even those archives came to devastation, that scripture to dilapidation and dispersion, and the statements (vâkîh) also to obsoleteness, perversion, and corruption.

‘11. And, after that, I, Âtûrd, son of Hêmêd and leader of the orthodox, have likewise written, from their fragments (sûbâragânŏ), a new means of giving assistance to the Mazda-worshipping religion, with much prayer, investigation, and trouble. 12. From whatever was recovered from those dilapidated (visândakŏ), decayed, worn out, and dust-mingled (khâk-âmêg) archives—and these, too, brought back by taking away, carrying off, and seizing—it is selected, owing to 1 the assistance of the counselling wisdom of the mighty spirit, for the rediffusion of the words and deeds of the ancients, and of the evidence of the Avesta, for those of the primitive faith. 13. And the increase of knowledge from the good religion, arranged and prescribed in its chapters, is a lustre from encountering that splendour from the enlightenment of the original light primarily composed for the exposition of the good religion, and this which is named is a resemblance by adoption of the thousand chapters of that great original Dinkard 2. 14. It is perfected by the sacred beings, and transmits the powerful effect which has come upon even that which is the perfect religion of the sufferers 3 in this age, and also the coming of the assistance of the soul to the knowledge 4 of the orthodox; and even reunion with 5 the rest of Irân is acquaintance with the exposition of the Mazda-worshipping religion, and the reproviding of more resources of a like origin, which will be also due to those whom the Supreme has provided, the disciples of Aûshêdar 6, son of Zaratûst, for asking again a declaration of 7 the good religion from Aûshêdar.’

From this statement it appears that the compilation of the Dinkard was commenced by Âtûr-farnbag, son of Farukhŏ-zâd, one of the leaders or supreme high-priests of the Mazda-worshippers, and was revised and completed

p. xxxiii

by Âtûrd, son of Hêmêd, one of his successors. From the Mâdîgân-î gugastak Abâlis we learn that Âtûr-farnbag had a religious disputation with Abâlis in the presence of the Khalîfah Al-Mâmûn, who reigned A.D. 813-833; he must therefore have been compiling the Dinkard during the first half of the ninth century. In the Sikand-gûmânîk Vigâr, IV, 107, IX, 3, X, 55, he is also mentioned as a compiler of the Dinkard., but the details there quoted must have been taken from its first two books which are still missing. It is likewise stated at the beginning of both its fourth and fifth books that their contents are derived from his statements, and a similar acknowledgement is made with regard to some of the contents of Chap. CXLII of the third book; so that the evidence of his authorship is very complete. With regard to Âtûrd, the completer of the Dinkard, we may safely identify him with the Âtûrd, son of Hama, mentioned in Bd. XXXIII, 11 as a contemporary of Zâd-sparam, who flourished at the latter end of the ninth century (see S.B.E., vol. xviii, p. xiv). We have, therefore, every reason to be satisfied that the whole of the Dinkard was compiled during the ninth century.

The history of the transmission of the text of its last seven books, through the last thousand years, down to the present manuscripts, is equally satisfactory, owing to the preservation of a series of colophons appended to the text, of which the first and most important may be translated as follows

‘Completed in great joy and full of gratification this last portion of the manuscript of the incomparable, priceless, and unequalled Dinkard, at the place where it was found and happily disinterred 1 by us in Asûristân, within the happily prosperous, odoriferous, precious, well-thriving, and glorious Bakdâd of Good Rectitude 2; from a copy which, as regards the religion, is just as the leaders of the saintly and orthodox, who were of the family of the saintly Âtûrd, son of Mâraspend, (who re-explained knowledge, by five or six well-destined ones, from the pure revelation

p. xxxiv

which is the all-embellished learning of learnings) and the successive leaders of the orthodox (who again provided at different times [ahamvâr] for its restoration, through manuscripts at various places, to maintain reading and investigation therein) had written.

‘I, Mâh-vindâd, son of Naremâhân, son of Vâhrâm, son of Mitrô-âpân, like an adopted son for his own possession, who wrote it, am letting it forth on the day Dên of the month Tîr, the victor, of the year 369 after the year 20 of that Yazdakard, king of kings, who was son of Shatrô-ayâr [2nd July, 1020] 1; in reliance on the pure good religion of the Mazda-worshippers, as regards remembrance of Zaratûst, the Spîtâmân with the righteous guardian spirit, and of the genuine achievement of Âtûrd, son of Mâraspend; and as regards remembrance of the righteous utterance of blessings for the whole embodied existence by the desirers of righteousness, who are thinkers of good thoughts, speakers of good words, and doers of good deeds; in the worldly existence, through completely-wishful kind regard of the practices of righteousness, they shall unite with the union of the renovation of the universe, and spiritually their pure souls and guardian spirits attain to the supremely great position and eminence, and complete acquirement of recompense, which are in the light that is endless, constantly beneficial, and full of glory, which they shall obtain. This is especially for those saintly and supremely learned men, Âtûr-farnbag, son of Farukhŏ-zâd, and Âtûrd, son of Hêmêd, by whom this priceless 2 Dinkard scripture was selected so learnedly and (with a pure perception of the spiritual lord, in seizing the cream of the fortunate commentary of 3 the good religion) so truly amicably, and fully affectionately for the good creatures and religion, with great advantage for us moderns, and concealed for me who, through eagerness for righteousness, like an adopted son, have happily disinterred this scripture; and even he who reads, and shall make use of it, is reliant and free from doubt about it; and him who shall take a copy from it, and preserves it with propriety, they shall appropriately connect with it.’

p. xxxv

(This is followed by a long succession of aphorisms, and the colophon winds up with some threats against those who shall misuse the manuscript.)

As this colophon mentions only the 'last portion' of the Dinkard, and is appended to the text of Dk. III-IX, it is probable that the first portion of the work, Dk. I, II, had already become separated from the rest within 150 years of its revision and completion. And if Mâh-vindâd did not copy from the original manuscript of Âtûrd, he must certainly have done so from a very early transcript.

The second colophon was written by Shatrô-ayâr, son of Êrdîshîr, son of Aîrîk, son of Rûstâm, son of Aîrîk, son of Kubâd, son of Aîrân-shah, who completed his copy on the day Aûharmazd of the month Spendarmad in the year 865 after the 10th year of Yazdakard [3rd October, 1516], having transcribed it from a copy written by Mar’zapân, son of Spend-dâd, son of Mar’zapân, son of Mitrô-âpân, son of Spend-dâd, son of Mitrô-âpân, son of Mar’zapân, son of Dahisn-aîyyâr, son of Rôg-vêh, son of Shâh-mard. The date of Mar’zapân's copy may be approximately fixed by observing that his father's first cousin wrote a copy of AV. and Gf., mentioned in K20, in the year 690 of Yazdakard, while his great great granduncle wrote a similar copy, mentioned in MH6, in the Pârsî year 618. If this Pârsî year be reckoned from the era of the 10th year of Yazdakard, as seems probable 1, these dates give 52 years for three generations; and Mar’zapân, living one generation later than the writer of A.Y. 690, may perhaps have written his copy of the Dinkard about A.Y. 707 [A.D. 1338]; so that there was probably another copyist, intermediate between him and Mâh-vindâd-î Naremâhân, of whom no record has been preserved. Shatrô-ayâr concludes his colophon by quoting a long passage from the first colophon, as already stated in p. xxxiv, n. 1, and by acknowledging his obligations to three other persons whom he names. This colophon is the last that now remains attached to the manuscript B, but it was formerly followed

p. xxxvi

by a third colophon, written by the actual writer of B, and preserved in copies transcribed from B since its arrival in India.

This third colophon was written by Mâh-vindâd, son of Vâhrâm, son of Êrdîshîr of Tûrkâbâd, who completed his copy, from that of Shatrô-ayâr, on the day Âvân of the month Khûrd in the Pârsi year 1009 after the 20th year of Yazdakard [21st December, 1659, N. S.]. This copy, which constitutes the manuscript B, was afterwards approved by Vâhrâm, son of Mâh-vindâd, son of Rûstâm, son of Anôshak-rûbân, son of Rûstâm of Tûrkâbâd, who blesses the writer of the second colophon, on the day Tîstar of the month Vohûman in the year 1038 of Yazdakard [18th August, 1669, N. S.]. It was also finally seen and approved by Rûstâm, son of Gûstâsp, son of Êrdîshîr, who likewise blesses the writer of the second colophon; and the approximate date of this approval may be guessed from the fact that Rûstâm Gûstâsp is known to have copied one manuscript in A.D. 1706, and another in 1741.

Regarding this manuscript B, written in 1659, it appears from Mullâ Fîrûz's Avîgêh Dîn (Bombay, 1830) that Mullâ Bahman, son of Mullâ Behrâm, a Parsi priest of Yazd, brought this manuscript of the Dinkard from Irân to Surat in 1783, and, having shown it to Aspandiârji Ratanji-shâh, he lent it to Kâusji Rustamji, then Dastûr of Surat, and allowed him to have it copied. Mullâ Bahman had great difficulty in obtaining the return of his manuscript, and when it was returned many folios were missing. It was after this loss of folios that Aspandiârji had several other copies transcribed from the defective manuscript, to be sent to various persons, and all these copies were therefore equally defective.

This manuscript B, thus defective, afterwards came into the possession of Mullâ Fîrûz, who was high-priest of the Kadmi Parsis in Bombay; and, after his death in 1830, it descended to his successor. In 1875 it belonged to Dastûr Sohrâbji Rustamji, high-priest of the Kadmis, through whose courtesy, and that of Dastûr Dr. Jâmâspji Minochiharji, it was then lent to me long enough to

p. xxxvii

enable me to copy and collate two-thirds of Dk. III and to collate Dk. IV-IX; and Dastûr Jâmâspji, afterwards, kindly supplied me with a copy of the remainder of Dk. III.

The manuscript has been bound in its defective state, and contains 322 folios, originally fourteen inches high and ten inches wide, written 20 to 22 lines to the page. When complete it appears to have consisted of 392 folios, all numbered in Persian words, but with several blunders, including one of fifty folios, so that the last folio was really numbered 442. Of the 70 folios not bound with the rest of the manuscript, fourteen were lying loose in the volume; forty-three belonged to Dastûr Rustamji Kaikobâdji of Nausâri, with a copy of which I was kindly supplied by Dastûr Dr. Peshotanji Behramji of Bombay, who also enabled me to collate it with the original folios; and seven folios were lent to me by Dastûr Dr. Hoshangji Jâmâspji of Poona, for the purpose of copying. The remaining six folios have not been discovered; they comprise the first folio of the manuscript, containing the commencement of Dk. III, which was probably lost before the manuscript arrived in India; also one folio in Dk. VII, two in Dk. IX (see pp. 262, 270 of this volume), and the last two folios of the manuscript, containing the third colophon and final approvals (see p. xxxvi).

I am likewise much indebted to the kindness of Professor Kielhorn, who gave me a modern copy of Dk. IV-IX (with the text in its defective state) which had been prepared at Poona, so that it was only necessary to collate this copy with the original text of the manuscript B. With the aid of all this liberal assistance I was enabled to obtain the whole text of the Dinkard, known to exist, in the course of a few months; that it has since taken as much as sixteen years to find opportunities for translating and publishing rather more than one-fourth of its contents, will not surprise any one who is acquainted with the nature of the work that had to be done.

The only known manuscript, independent of B, that contains any portion of the Dinkard, is the old codex K

p. xxxviii

brought from Persia by the late Professor Westergaard in 1843, and now No. 43 of the Irânian manuscripts in the University Library at Kopenhagen. This codex contains about one-fifth of the text of the Dinkard in two detached portions, together with other Pahlavi texts. The first portion occupies fols. 177-261, and comprises Dk. VI, of which one-eighth is missing, with Dk. III, Chaps. CLX and CCLXXXIII, and a colophon, all written in the district of Tûrkâbâd by Mitrô-âpân, son of Anôshak-rûbân, son of Rûstâm, son of Shatrô-ayâr, son of Mâh-vindâd, son of Vâhrâm, son of Gûshisn-ayâr, son of Mitrô-âpân, and completed on the day Gôs of the month Mitrô in the Pârsî year 943 after the 20th year of Yazdakard [10th May, 1,594, N. S.]. This copyist appears to have been a great-uncle of the writer who approved the manuscript B in 1669, ten years after it was written; and the original from which he copied was, no doubt, descended from Mâh-vindâd-î Naremâhân's manuscript of 1020, as he appends to his colophon all the latter part of Mâh-vindâd's colophon (see p. xxxiv, n. 3). The second portion of the text of the Dinkard, contained in the manuscript K, is written by another hand on 42 additional folios, and comprises the last two chapters of Dk. III, the whole of Dk. V, and the first three-tenths of Dk. IX (as mentioned in p. 172, n. 1, of this volume). This manuscript supplies several short passages in the Dinkard, which are omitted by B, especially in the first portion of the text described above. It has also afforded much assistance in the translation of Dk. IX, Chaps. I, 1-XXXI, 17.

Regarding the authorship of the summary account of the Nasks, contained in Dk. VIII, IX, it may be reasonably assumed, in default of any positive information, that the compiler was Âtûrd, son of Hêmêd, the last editor of the Dinkard. And, as nothing is said about any previous treatise being consulted, it may be safely supposed that he had access to the Avesta texts and Pahlavi versions of all the Nasks he describes, fully three centuries after the Muhammadan conquest of Persia. The only Nask he could not obtain was the Vastag, and the Pahlavi version

p. xxxix

of the Nâdar was also missing; under which circumstances, the fully detailed accounts of these two Nasks, given in the Persian Rivâyats, must be viewed with suspicion, until better evidence of their authenticity has been discovered than is at present available.

The survival of so much of the sacred Zoroastrian literature, during three centuries of Muhammadan rule, indicates that the final loss of nearly all this literature was not so directly attributable to the Arabs as the Parsis suppose. So long as a considerable number of the Persians adhered to their ancient religion, they were able to preserve its literature almost intact, even for centuries; but when, through conversion and extermination, the Mazda-worshippers had become a mere remnant, and then fell under the more barbarous rule of the Tartars, they rapidly lost all their old literature that was not in daily religious use. And the loss may have been as much due to their neglecting the necessary copying of manuscripts, as to any destructiveness on the part of their conquerors; because the durability of a manuscript written on paper seldom exceeds five or six centuries.

The statements of the Dinkard, about the classification and subdivisions of the Nasks, are corroborated and supplemented by those of Zâd-sparam (see pp. 401-405). The division of all literature into three classes of knowledge, religious, worldly, and intermediate, is one that would naturally suggest itself to any classifier 1, but the names employed (which are transcribed from the Avesta, and do not exactly correspond with these three meanings) must have originated at a period when the Avesta language was still spoken. That such a classification cannot be very strictly carried out in practice is already admitted in Dk. VIII, Chap. I, 13.

p. xl

The further division of the literature into twenty-one books, seven in each of the three classes, is a much more artificial arrangement, and can, perhaps, be best explained as an attempt to make the twenty-one words of the Ahunavair serve the purpose of a reminder for enumerating the Nasks in their proper order. This arrangement was probably made at some period when the scattered Avesta literature was being collected and re-arranged, the Pahlavi versions being then supplied, and the present Pahlavi names of the Nasks appointed. This may possibly have been the work of 'composition and preservation' attributed to Âtûrd, son of Mâraspend, in Dk. VIII, Chap. I, 22, when 'the Nasks were enumerated' (see Dk. IV, 27, in p. 415), which occurred in the fourth century.

Why the established sequence of the Nasks, detailed in Dk. VIII, Chap. I, 12, should differ from the successive sequences of their three classes, given in §§ 9-11, is very imperfectly explained; but some of the reasons for the difference may perhaps be guessed. If the notation proposed in p. 7, n. 3, be adopted, the established sequence is G2-4; H1-7; G5; L6; G7; L7, 1-5; G6, 1; in which the only Nasks that are out of their order in the classes are G1, 5-7 and L6, 7. The placing of G6, 1 next after L5 (that is, the Hâdôkht and Stôd-yast next after the Vendîdâd) may perhaps have been owing to the constant use of these three Nasks in the liturgy, in which either the Vendîdâd, or the Hâdôkht 1, was frequently interpolated in the recitation of the Stôd-yast which comprised by far the larger portion of the present Yasna and Vîspêrad. But this position of the Stôd-yast, at the end of the list of Nasks, was probably considered derogatory to its sacred character by most of the writers of the Persian Rivâyats, who have, therefore, restored it to its original place at the head of the Gâthic Nasks. Dk. VIII, Chap. I, i5, states that G5 was placed after 117 because the Vastag was connected with the Vistâsp-sâstô, probably by the nature of its contents. And, possibly, the sequence L6, G7, L7 of the Kitradâd, Spend, and Bakân-yast, between the Vastag

p. xli

and Nîkâdûm, may indicate some similar resemblance of contents; especially as the contents of the Kitradâd and Bakân-yast were so far from being strictly legal that these Nasks were placed in a sub-class by themselves, and the connection of the Spend with the Gâthas appears to have been merely historical. The Persian Rivâyats place the Spend next after the Vastag, thereby bringing the two imperfectly Gâthic Nasks together, as well as the two imperfectly legal ones; but then they also transpose the Ganabâ-sar-nigad and the Hûspâram, for which there seems to be no justification.

With regard to the names of the Nasks, it is evident that several of the Persian names, used in the Rivâyats, are more or less irreconcileable with the Pahlavi names in the Dinkard, and some others are improbable readings of the Pahlavi forms. In this translation the Pahlavi forms have been followed, as clearly more authentic than the Persian corruptions, and some few of the names have been read differently; while in other cases the most probable readings have been merely suggested in foot-notes, not on account of the Persian reading being justifiable, but because the evidence for the suggested reading is less complete than would be desirable.

In dealing with this account of the Nasks it is always necessary to remember that the compiler of the Dinkard relies entirely upon their Pahlavi versions, as he states distinctly in Dk. VIII, Chap. I, 3; he occasionally mentions the Avesta texts, as in Chaps. VI, 1, XII, 1, and it is abundantly evident, to the practised translator, that Avesta phrases often underlie the Pahlavi passages which seem to be quoted at length from the original Nasks, especially in Dk. IX; but, for some of the details mentioned, there may be no older authority than a Pahlavi commentary, and this should ever be borne in mind by the sceptical critic in search of anachronisms.

Owing to his complete reliance upon the Pahlavi versions, it is impossible to ascertain with certainty whether any particular statement, made by the compiler of the Dinkard, was contained in the Avesta text; his summary, therefore,

p. xlii

throws little or no satisfactory light upon the origin of that text. A few of the details he mentions (such as those contained in Dk. VIII, Chaps. XIII, 17-20, XLIII, 24 and Dk. IX, Chaps. XXXII, 17, XXXIX, 13-16, LIII, 3) evidently refer to Sasanian times, and may be reasonably supposed to have originated in the Pahlavi versions of those times. But vaguer prophecies of good or evil, such as are common in all religions at all times, may have often occurred in the Avesta texts themselves.

It is evident, however, that all the Nasks have accumulated around the Gâtha centre of the Stôd-yast, and that this Gâtha centre in the earliest Sasanian times was neither more nor less extensive than it is at present. The age of Gâthic composition had so long passed away in the time of the earliest Sasanian monarchs, that the sages whom they appointed to collect and re-arrange the sacred literature, were unable to fully understand many of the stanzas they had to translate into Pahlavi, much less could they have added to their number. How far they may have been able to write ordinary Avesta text is more uncertain, but any such writing was probably confined to a few phrases for uniting the fragments of old Avesta which they discovered, or for interpolating opinions of their own. All such compositions, however, would have been hazardous, as forming no part of their duties, which seem to have been confined to the arrangement of the fragmentary Avesta texts, and their translation into Pahlavi with explanatory comments in that language. It appears from the traditional statements, mentioned in p. 415, that this work was completed, and the Nasks were fully arranged, by Âtûrd son of Mâraspend, in the reign of Shahpûhar II (A.D. 309379); but the Pahlavi versions were certainly revised, and some further commentaries added, after the suppression of the heresy of Mazdak, as late as the reign of Khûsrôî I (A.D. 531-579).

That the Avesta texts themselves were not written, to any great extent, in Sasanian times, is shown by the quantity of Pahlavi commentary necessary to adapt them to the altered circumstances of those times. The Gâthic

p. xliii

[paragraph continues] Nasks, being strictly religious, required only some explanations, with little extended commentary; because the religion had to be maintained without sensible modification. Of the Hadha-mãthric Nasks we know but little. But the strictly Legal Nasks consisted chiefly of the commentary which is always necessary to adapt ancient laws to modern ideas.

With regard to the mode of describing the Nasks, adopted in the Dinkard, it is evident that the compiler intended, in the first place, to give merely a very short account of the general contents of each Nask, to be followed by a detailed statement of the particular contents of each chapter (see Dk. VIII, Chap. I, 23, 24). But, when he had fully carried out this intention with respect to the first three Nasks, his work came to a premature conclusion, which has deprived us of much valuable information regarding the rest of the Nasks. The descriptions of these other Nasks vary in extent, but may be roughly classified as follows:—Of the Nâdar and Vastag there is no description whatever. Of the Dâmdâd, Radŏ-dâd-aîtag, Kaskîsrôbô, Vistâsp-sâstô, Bakân-yast, and Stôd-yast the description is very short, averaging 80 Pahlavi words for each. Of the Pâgag, Baris, Kitradâd, Spend, and Hâdôkht the description is rather longer, averaging 358 Pahlavi words for each; but, as such a description is still far too brief to be satisfactory, the compiler must have intended to add a detailed account of each chapter of all these Nasks. On coming to the strictly Legal Nasks, however, he adopted a different plan, by giving a much more voluminous statement of the contents of certain selected chapters; thus the very long description of the Nîkâdûm, Ganabâ-sar-nigad, Hûspâram, and Sakâdûm averages 3670 Pahlavi words for each. This change of plan is somewhat modified in the case of the Vendîdâd, where the description of 1272 Pahlavi words is only moderately long. While the first three Nasks, the Sûdkar, Varstmânsar, and Bakŏ, after a very short description averaging 65 Pahlavi words for each, are again described in detail, as already mentioned, to the average extent of 8647 Pahlavi words for each.

p. xliv

From these descriptions, and their connection with certain Avesta texts and Pahlavi writings, it is now possible to form a more or less adequate conception of the contents of Nasks I-IV, X, XIII-XIX, XXI, and also some idea of those of Nasks VI, XII; but the accounts of the remaining six Nasks, most of which belonged to the Hadha-mãthric or scientific class, are very unsatisfactory.

With reference to the total extent of the Nasks, when they were all extant, it is obvious that the length of descriptions, drawn up on the same plan, ought to bear approximately some definite proportion to the lengths of text described; so that, if the extent of the text of one Nask be known, and the proportion it bears to the length of its description be ascertained, this proportion becomes a rough means of estimating the probable extent of other Nasks, from the length of their descriptions drawn up on the same plan. Three years ago an attempt was made 1 to estimate the total extent of the Nasks in this way, based upon the assumptions that the Nasks still extant were three in number, that the length of the description of the Vendîdâd was a fair average one for estimating the extent of Pahlavi version in all the lost Nasks, and that the proportion of Avesta text to Pahlavi version in the Nîrangistân was also a fair average for estimating the extent of their Avesta texts. These assumptions were carefully made, as the least liable to objection, and the total extent of the Nasks in Sasanian times, thus estimated, amounted to 133,000 words of Avesta text and 844,000 of Pahlavi version.

Since the completion of the translation of Dk. IX it has, however, become possible to estimate the probable extent of the first three Nasks from the proportion between the actual extent of the first three fargards of the Bakŏ (Yas. XIX-XXI) and the length of their description. It has also been thought no longer reasonable to neglect the actual length of the Nîrangistân as a basis for estimating the extent of the Pahlavi versions of the strictly Legal Nasks XV-XVIII;

p. xlv

and the Bakân-yast has been identified with the Yasts still extant 1. These additional considerations have led to a new estimate of the probable extent of each Nask separately, based upon the best data available in each case, as stated in detail in the foot-notes to the names of the Nasks in the Extant Fragments (pp. 451-488 of this volume). These estimates are here collected, for the sake of convenient reference, as follows:—















„ 3





„ 4











„ 2





„ 3





„ 4





„ 5





„ 6





„ 7

























„ 1





„ 2





„ 3





„ 4





„ 5










„ 1








Total extent of 21 Nasks




This total is about 2½ times as great as that of the former estimate, but, as nearly the whole of this increase is in the four strictly Legal Nasks, whose length is well attested by that of the extant Nîrangistân, there is little probability that further investigation will lead to any reduction

p. xlvi

of this estimate. No probable alteration of the estimate of the extent of the Hadha-mãthric Nasks, which is the most uncertain, would materially affect the total.

Another matter of interest to the readers of translations from the Pahlavi, especially to those who are aware of the ambiguities of the original text, is the degree of confidence they can place in the correctness of the translation. In the case of the Dinkard it is fortunately possible to consult manuscripts written in Persia, and descended through only four or five intermediate copies from the work of the original writer, so that the text is remarkably free from copyists’ errors. The eighth and ninth books also contain very few of those involved sentences, with long parenthetical clauses, which, owing to the habitual absence or misplacement of stops, are very perplexing to a translator. The chief difficulties of the text arise from its synoptical character, and the consequent want of connection between its sentences; there being often too little context to define the meaning of a doubtful word. The number of words of doubtful meaning in Pahlavi is, however, fast diminishing, in proportion to the advancing study of the texts; and the certainty of a translator, as to the correctness of his work, is increasing in a like proportion. At any rate, the reader may safely rely upon the general accuracy of these translations, even if a few errors should hereafter be discovered.

As an instance of such possible errors I will here correct one that exists in my translation of the Epistles of Mânûskîhar, which was pointed out to me by Môbad Tehmuras Dinshawji Ankalesaria, in a letter dated 28th October, 1.88. In Ep. II, ii, 9-11, there occurs an illustration of what should be done when commentators differ, derived from the use that can be made of different observations of the stars, and containing three names that were difficult to identify. These names were doubtfully read as corruptions of the names of three of the lunar mansions, but it now appears that they were the names of three sets of astronomical tables (zîk); so that Shatro-ayârân, Hindûk, and Ptolemêôs should be read, instead of Satvâharân, Avênak, and Padramgôs; both sets of readings expressing the same

p. xlvii

[paragraph continues] Pahlavi letters. With these alterations the passage may be translated as follows:—

Ep. II, ii, 9. ‘And there may be a position of the stars, settled even by computers of the stars, when they would take that of the sun and moon from the tables of Shatro-ayâr, that of Saturn from the Hindû tables, and that of Mars from the tables of Ptolemy, and the position comes out very good, and they are able to speak of the maturity of strength undoubtedly brought on. 10. That this is to be seen as an occurrence is a conjunction which is not possible; because, if the tables of Shatro-ayâr be exact, yet, since its Saturn and Mars are not from the tables, the effect is not a good configuration; if the Hindû tables be correct, yet, since its sun, moon, and Mars are not from those tables, the effect is not good; and if the tables of Ptolemy be correct, yet, since its sun, moon, and Saturn are not from those tables, the effect is not good; on account of which the conjunction is not correct in any way; they believe it possible, however, for a firm mind to accomplish this auspicious labour. 11. But they say the just and wise are making the decision that this would be a very good position, because that which is in the tables of Shatro-ayâr is truly issuing from him, the great Shatro-ayâr; and that of Shatro-ayâr, being better through the tables of Ptolemy, remains that employed.’

In conclusion, it is desirable to make some remarks upon the transliteration of Pahlavi, because it is necessary to express not only the various sounds of the letters of a very deficient alphabet, but also the mode of writing several abbreviated compounds which are quite as essential to the correct orthography of Pahlavi as the forms of the separate letters themselves. For this purpose italics are used to indicate not only a few differences of sound from the usual English pronunciation of consonants, but also different letters having the same sound, and letters abbreviated in the writing of compounds. When the abbreviated letter is already italicised, the preceding short vowel (which is not expressed in Pahlavi writing) is also italicised to indicate the abbreviation, or an apostrophe is introduced between the two consonants when no short vowel sound intervenes. Hyphens are used both to connect the components of compound words, which are often written

p. xlviii

separately, and also to separate words that are written together in Pahlavi. The application of these rules will be best understood by reference to the following list of transliterations which have been found necessary:—


a (initial), â, h, kh, zd.

2, a (privative), aê, âê, (final), h (final).


f, p, v

d, t.

g, k, z.

l, r.


or s, + .

s, sh, + , g-a.




l, n, ô, ŏ, r, û, v.

l, r.

d, ê, g, g, ga, î, y.



( + , or + ). aêk, af, ag, ap, av, az, âf, âg, âk, âp, âv, âz, haf, hak, hap, haz, khef, khek, khz.


3, aê, aê (final), êh (final).

adîn, dîdû (better bîdû).

lp, rag, raz, r’g, rp, r’z.

f, dîp, gîf, sag, sak, sap, sig, sip, sp.

sk, yâf, yâv.


dag, d-k, êg, êk, gag, gak, gug, îk, îv, îz, yaz, yez.

dên (better bên), yên.

p. xlix


A.D. for Anno Domini; âf. for Âfrîngân; Ar. for Arabic; AV. for Ardâ-Vîrâf nâmak, ed. Hoshangji and Haug, 1872; Av. for Avesta; A.Y. for Anno Yazdakardi; B for Bombay MS. of Dinkard, written in Irân, A.D. 1659, see pp. xxxv-xxxvii; B29 for MS. No. 29 in the University Library at Bombay; Bd. for Bundahis, as translated in vol. v of this series; Bk. for Book; B.P. for Bahman Pûngyah, see p. 418, n. 3; Byt. for Bahman Yast, as translated in vol. v of this series; Ch. or Chald. for Chaldee; Chap. for chapter; Dd. for Dâdistân-î Dînîk, as translated in vol. xviii of this series; Dk. for Dinkard; Dv. for Dîn-vigirgard; ed. for edited by or edition; Ep. for Epistles of Mânûskîhar, as translated in vol. xviii of this series; Farh. Oîm. for Farhang-i Oîm-aêvak, or Zand-Pahlavi Glossary, ed. Hoshangji and Haug, 1867; fol. for folio; G for gâthic; Gah. for Gahanbâr or Gâhanbâr; Gen. for Genesis; gen. for genitive; Gesch. der Sas. for Geschichte der Perser und Araber zur Zeit der Sasaniden, 1879; Gf. for tale of Gôst-î Fryânô, ed. West and Haug, 1872; Gld. for Geldner; H for hadha-mãthric; Haug's Essays for Essays on the Sacred Language, Writings, and Religion of the Parsis, by M. Haug, 2nd ed.; Hn. for Hâdôkht Nask, ed. Haug, 1872; Ibid. for ibidem; J2 for Jâmâspji's Yasna MS. with Pahlavi, now in the Bodleian Library, by the same copyist as K5; K for Kopenhagen MS. No. 43, written A.D. 1594, see pp. xxxvii-viii; K1, K5, K20, K35 for Iranian MSS. Nos. I, 5, 20, 35 in the University Library at Kopenhagen; L for legal; l. for line; ll. for lines; Mf4 for the Mullâ Fîrûz Library's Yasna MS. with Pahlavi, descended from an ancestor of K5; MH6, MH10 for MSS. Nos. 6, 10 of Haug's Collection in the State Library at Munich; Mkh. for Dînâ-î Maînôg-î Khirad, as translated in vol. xxiv of this series; MS. for manuscript; n. for foot-note; NS. for new style; Ny. for Nyâyis; O225 for MS. No. 225 of Ouseley's Collection in the Bodleian Library at Oxford; OM for Olshausen and Mohl's Fragmens relatifs à la religion de Zoroastre; p. for page; Pahl.

p. l

for Pahlavi; Pâz. for Pâzand; Pers. for Persian; pp. for pages; Pt4 for Peshotanji's Yasna MS. with Pahlavi, similar to Mf4; Riv. for Rivâyat; S.B.E. for Sacred Books of the East; Sd. and Sg. for Sad-dar and Sikand-gûmânîk Vigâr, as translated in vol. xxiv of this series; Sîr. for Sirôzah; Sls. for Shâyast-lâ-shâyast, as translated in vol. v of this series; Sp. for Spiegel; Vend. for Vendîdâd; Vig. for Vigirkard-i Dînîk, ed. Peshotan, Bombay, 1848; Vîsp. for Vîspêrad; vol. for volume; W. or Westerg. for Westergaard; Yas. for Yasna; Yt. for Yast; ZA. for Zend Avesta; Zs. for Selections of Zâd-sparam, as translated in vol. v of this series.


xxxi:1 Both MSS. have Shaspîgân here, but see p. 423, n. 4.

xxxi:2 So in K, or perhaps Shîzîgân; B has Shaspîgân.

xxxi:3 So in K.

xxxi:4 So in K; B has Shapân.

xxxi:5 B has Âtûrd inserted here by mistake.

xxxii:1 Assuming that m stands for min.

xxxii:2 Both MSS. have zak rabâ bûn Denô-kardŏ.

xxxii:3 K has dênô-î bûrdârân.

xxxii:4 B has 'arising.'

xxxii:5 B has 'and the reunited selection for.'

xxxii:6 See Dk. VIII, Chap. XIV, 12.

xxxii:7 B omits 'a declaration of.'

xxxiii:1 Khûs-kand might be the name of a place here, but cannot be so in the next paragraph.

xxxiii:2 The angel Aharîsvang (Av. ashis vanguhi).

xxxiv:1 The remainder of this colophon, so far as it is here translated, is also quoted in the second colophon.

xxxiv:2 Here written ar’gŏ, but it is an-ar’gŏ in the second colophon.

xxxiv:3 Reading dên farukhŏ zand shîr-hâkŏ-î, but this is doubtful. From this point the whole of the rest of this colophon, including the aphorisms, is also found in K.

xxxv:1 Observe the use of the phrase 'Pârsî year' in the third colophon and in the manuscript K (see p. xxxviii).

xxxix:1 Professor Darmesteter has suggested to me the very similar apportionment of the old Hebrew literature, mentioned in Jeremiah xviii. 18, thus:—'For the law shall not perish from the priest, nor counsel from the wise, nor the word from the prophet.' And in Ezekiel vii. 26, thus:—'Then shall they seek a vision of the prophet: but the law shall perish from the priest, and counsel from the ancients.'

xl:1 So long as it was preserved.

xliv:1 In the Sitzungsberichte der philosophisch-philologischen and historischen Classe der k. b. Akademie der Wissenschaften zu München, 1888, pp. 441, 442.

xlv:1 This had been done, long ago, in a Persian Rivâyat, quoted in B29, fol. 164, which states that the following sixteen Yasts were in the Bayân-yast Nask, namely, the Hôrmezd, Âbân, Mâh, Tîr, Gôs, Mihir, Srôsh, Rashn, Fravardîn, Bahirâm, Râm, Dîn, Âshasang, Âsd, Zamyâd, and Khurshêd Yasts.

Next: Observations