1. The Lexicon (related to the Magas/Bhojakas)

The lexicon employed in the Purāṇic sections related to the Magas/Bhojakas is very peculiar. Particularly in relation to the ritual sphere, we find certain terms, that seem to be more closely connected to the Iranian environment than the Indian one. Even more interesting, this ‘foreign vocabulary’ appears to be drawn from different strata of Iranian languages, even though we cannot establish precisely which Iranian language each term belongs to. In fact, the terms have been ‘Sanskritized’ for their use in the Purāṇic texts; i.e., they have been rendered in a Sanskrit form.^[1] What is meaningful is that the Iranian loanwords were preferred to their Sanskrit equivalents, showing that the Iranian character of these Brāhmaṇas was very important to the author of the texts (and probably for the group of Magas/Bhojakas itself). Even if these loanwords were inserted in the Purāṇas in later times, the inclination towards Iranian culture (or Iranian roots) is clear. Moreover, it is not important that these words may have been later additions, possibly drawn from Pārsī teachings;^[2] the attempt to maintain a link with the Iranian and Mazdean milieu is evident here. Furthermore, the Iranian connection lies in other aspects of Śākdvīpīyas’ (or the Magas’/Bhojakas’) cult, especially in their calendar, which shows a clear resemblance to the Zoroastrian one, albeit with remarkable Śaiva influences.^[3]

Generally, when loanwords are accepted and integrated into another context, we must consider the fact that they could simply have crystallized ideas and expressions that were already known to the target environment. The geographically close Iranian religious tradition could have reached North India before the Magas’ migration. According to the chronology of the Purāṇic tradition, the most ancient sections of the Sāmba-purāṇa should have been composed before the first Sanskrit translations of the Avestan texts. On the contrary, it is very difficult to date the different parts of the Bhaviṣya-purāṇa, which are definitely more recent.

An important source for comparing the ‘Sanskritized’ Mazdean terminology with the Purāṇic one is the Sanskrit translation of the Avestan materials made by Nēryōsangh Dhaval, a Pārsī mōbēd, around 1200 AD. His translation of the Yasna has been widely studied, especially between the 19^th and 20^th centuries, and translated into other languages, either in whole or in part.^[4]

Nēryōsangh himself discloses the authorship of his work:

[…] idaṃ ijisnijaṃdapustakaṃ mayā nirīosaṃghena dhavalasutena pahalavījaṃdāt saṃskṛtabhāṣāyāmavatāritaṃ // [...]

“[…] Ce volume nommé le livre Idjisni (Izeschné)^[5] a été traduit par moi Niriosangha, fils de Dhaval, du livre pahalavi (Pehlvi) en langue Sanscrite […].ˮ (BURNOUF 1833: XV f.)^[6]

Different scholars did not all use the same manuscripts^[7] in preparing the translation of Nēryōsangh's text, and the methodologies applied to the analysis are those that were in use at the turn of the 20^th century; even the places in which the physical manuscripts were presumably held have changed names. SPIEGEL (1861: 14) points to four different manuscripts: one is preserved at the University Library of Copenhagen, number six in Westergaard's Catalogue (WESTERGAARD 1846); another of Nēryōsangh’s Sanskrit translations is part of the “II fonds d'Anquetil” series, at the Bibliothèque Royale in Paris, currently the Bibliothèque Nationale de France. The last two manuscripts used by Spiegel are kept at the Parisian library as well, one known also by Burnouf, but not preserved in good condition, and the other coming directly from Burnouf's private collection. MILLS (1910: XXVIII ff.) lists a different set of manuscripts: one, classified as S', has the name Meher Nawrozji Kutar on it. Another one is marked Oxford E¹ or J³, which Destoor Jamaspji Minpcherji Jamasp Asana, the High Priest of the Pārsīs in Mumbai at the time, gave to the library; it was probably the oldest manuscript accompanied by Nēryōsangh’s translation, but it was very damaged with age.^[8] The manuscript J*, with the original Yasna and a sentence-by-sentence Sanskrit translation, property of Kai Khosu Destoor Jamasp Asana, a successive High Priest of the Pārsīs in Mumbai, (and son of the donor of the previous manuscript), has been kept at the Bodleian Library since Mills consulted it. Finally, a copy of the manuscript J^[9], also donated by Destoor, is preserved at the same library. This is the Khordeh Avestā with a Sanskrit translation, dated by Destoor himself to around the 15^th century.

The model for Nēryōsangh’s translation was the Pahlavi version of the Yasna; the Pahlavi text he used was not identical to the one we have nowadays, especially not with regard to certain sections of the text and glosses. Sometimes Nēryōsangh decided to follow the Avestan version instead, but in some sections, the Sanskrit translation is completely different from the Avestan original, because the Gujarati author did not completely understand the Pahlavi version, or he rendered the text in a literal way, sacrificing the Sanskrit syntax and, in some cases, changing the meaning of some sentences.^[10]

Probably the most important aim, among the Pārsīs, was to preserve and transmit the tradition, and the accuracy of the translations and knowledge of the languages of transmission took second place. Nēryōsangh tries to adhere to his original text (i.e. the Pahlavi version), even from a syntactical point of view; the word order is as close as possible to the Pahlavi one, ignoring the fact that some of his choices are against the rules of Sanskrit syntax. He neglects the rules of saṃdhi, too. HARLEZ (1882: 468) states, ‘Ce qui prouve que Neriosengh a suivi la version pehlevie c’est qu’il en reproduit généralement les erreurs.’

Even his glosses do not correspond to those of the Pahlavi manuscripts; sometimes they are longer, shorter or different altogether. For example, in Yasna XXVIII, we find a long Sanskrit gloss that does not exist in the Pahlavi text.

The Sanskrit terminology is not always appropriate or precise. Nēryōsangh’s language is full of terms drawn from Gujarati, and in some cases Sanskrit terms are used in their Gujarati equivalents; in others, Gujarati words are ‘Sanskritised’. Sometimes the author had to adapt Iranian names to Sanskrit phonology; for example, instead of the voiced alveolar fricative z, absent in Sanskrit, he used the voiced palato-alveolar affricate j [dʒ]: cf. Horomijda, Ahuramajda.^[11] Nēryōsangh also appropriated some terms from the Sanskrit cultural milieu, to translate Iranian ones, and in this way, he actually added some new vocabulary to the original Yasna: for instance, he translates Avestan X^vaniraθa-^[12] as Jambudvīpa-, more fitting to the Indian tradition and more comprehensible to Indian readers.^[13]

Finally, SPIEGEL (1861: 12) postulates that different authors, at least two, carried out the translation of the Yasna into Sankrit. The closing of the text exhibits a different style from the rest of it (in particular chapter XLVII), and the Sanskrit restitution of some Iranian terms is different and, in some cases, is closer to the Persian (Middle, but also New Persian) word. Compared to Nēryōsangh, who–even if he did not have an excellent knowledge of the language–was probably able to read Sanskrit works without great difficulty, the second author did not have any familiarity with Sanskrit case system, verbal forms and constructions.

In any case, for the purpose of this research, the lexicon employed is particularly meaningful^[14] : the comparison of the names of the ritual objects and the divine names could highlight many elements of similarity or dissimilarity between the two traditions. For example, the same nomenclature could suggest that the time-space coordinates of the two works are quite similar. The Bhaviṣya-purāṇa mostly contains a ritual lexicon with Iranian elements and we know it is a late text that has been intermingled with many different materials; if the ritual terms had been the same in the Purāṇa and in Nēryōsangh’s translation, we could have postulated that this lexicon was actually produced and used in the same environment within a roughly contemporary period. On the contrary, the terms, especially those in the ritual vocabulary, are spelled differently.

Purāṇas	Nēryōsangh
avyaṅga- (abhyaṅga-)	aiviaṅghana-
varśman-	baresmana-
pūrṇaika-	–
patidāna-	–
homa-	hūma-
parahoma-	prāhūma-

The Avestan aiwyǡŋhana-, called kusti among the Pārsīs, is rendered as avyaṅga- (abhyaṅga-) in the Sāmba- and the Bhaviṣya-purāṇas, whereas Nēryōsangh’s aiviaṅghana- closely follows the Avestan spelling. Probably due to the low competence in Sanskrit language discussed above, Nēryōsangh decided to reshape the Avestan term barǝsman (Middle Persian barsom) as an -a stem (baresmana-), which was easier to manage than a consonant stem. In the Purāṇic text, the initial b- turns into a v-, a phenomenon typical of Middle Indic languages. Two ritual objects, the pūrṇaika- and the patidāṇa-, are not present in the translation of the Yasna; the first one was probably a sort of ladle for pouring water, while the latter clearly echoes the Avestan paiti.dāna-, the handkerchief used by priests for covering the mouth while officiating.^[15] The Avestan haoma-, Middle Persian hōm, Sanskrit sóma- is given as homa- in the Purāṇas and hūma- in Nēryōsangh’s text; paradoxically, the spelling of the Purāṇic term is closer to the Middle Persian hōm than Nēryōsangh’s one.

The same is valid for the parahoma-/prāhūma-, the restitution of Avestan parahaoma-, Middle Persian prāhōm.

Purāṇas	Nēryōsangh
Mihira	Mihira
Raśnu (Rājña, Rājā, Rājñā)	Rasna (Raśna)
Saośra (Srauṣa, Stoṣa, Toṣa, Strauṣa, Śroṣa)	Śrośa
Jaraśastra (Jataśastra, Jaraśabda)	Jarathustra
–	Hormijda–Majda

The names of the divine figures prompt some interesting considerations. The Avestan god Miθra-, in both cases, had already adopted the Middle Persian form Mihr. Mihira- could have been derived from the Bactrian version MIIRO (or MIORO), well known from the coins of the Kuṣāṇa king Kaniṣka.^[16] The names of the two attendants of Mithra on his morning crossing, Avestan Rašnu- and Sraoša-, Middle Persian Rašn and Srōš, are given in various Sanskritic forms in the Purāṇic version; in Nēryōsangh’s translation, the name Rasna- is rendered once again with an -a stem form, while Śrośa- follows the Middle Persian form. For the name Avestan Zaraθus̆ tra-, Middle Persian Z̆ artūs̆ t, Nēryōsangh opted for the form Jarathustra-, closer to the Iranian, maintaining the second part of the compound as -ustra-, which is generally spelled uṣṭra- (with retroflection) in Sanksrit. The absence of the fricative z in Sanskrit, which concerns both Jara[…]- and Hormijda-/Majda-, was discussed above.

Finally, Nēryōsangh’s translation contains some other interesting elements. The sun, hvar-, is translated as sūrya-, showing a perfect adaptation of the term to the Indian environment: for example, in VI, 36, we find Sanskrit […] sūryaṃca […] for Avestan […] hvareca […]. On the other hand, the word deva- is used for indicating demonic beings, the Iranian Daevas; in this case, the term is completely decontextualized, the Devas being the deities of the Indian pantheon (cf. SPIEGEL 1861: 9). Nēryōsangh seems not to follow any predetermined rule in the process of translation, and the content of his work can be confusing for non-Zoroastrian readers. For this reason, we can assume that this translation was composed for Pārsī Zoroastrians, who were familiar neither Avestan nor Pahlavi any more. The choice of Sankrit, the sacred language of the Hindus, was motivated by its perception as a more solemn and widespread language than the regional Gujarati. Moreover, this option allowed non-Zoroastrian Indians the opportunity to read the sacred texts of the Pārsīs and, in this way, provided them with access to the Avestan corpus.

Footnotes and references:

[back to top]

[1]:

I have discussed this topic extensively in PALLADINO 20??b.

[2]:

See HUMBACH 1978: 248.

[3]:

On this topic, see PANAINO 1996. Varāhamihira presents the Maga Brāhmaṇas’ calendar in Pañcasiddhāntikā 1, 23–25.

[4]:

In addition to SPIEGEL (1861) and MILLS (1910; 1912), we also have other editions of the Sanskrit translation of Nēryōsangh: see, for example, UNVALA 1918.

[5]:

Yasna (cf. BURNOUF 1833: 24).

[6]:

The text and translation are presented also in SPIEGEL 1861: 2; in BURNOUF’s edition, the Sanskrit text is in devanāgarī, whereas SPIEGEL gives his own transcription. For the Sanskrit text, I have opted for the IAST transcription from Burnouf’s edition.

[7]:

One of the digitalized manuscripts is also available on the Columbia University Libraries website at http://www.columbia.edu/cu/lweb/digital/collections/cul/texts/ldpd_8886507_003/.

[8]:

Actually, I do not know if all these manuscripts still exist, especially those that were in bad condition when SPIEGEL and MILLS published their editions. For the purpose of my research, I did not need to check the manuscripts of Nēryōsangh’s translation, but could rely on the edited texts. In the future, it could be a good topic of investigation, in order to update and refresh these commendable, but outdated studies.

[9]:

Bhaviṣya-purāna: mahātkuṣṭham

[10]:

He sometimes confuses the instrumental and the locative, or the third person singular, the second plural and the past participle. He sometimes takes the genitive for the nominative or the vocative; he does not understand the compounds and the inflected forms contain many mistakes. This is due to the fact that Nēryōsangh had a mediocre knowledge of Pahlavi, and confused the forms rendered with the same preposition or the same endings (cf. HARLEZ 1882: 467).

[11]:

“Fü r z, welchen Laut das Sanskrit nicht besitzt, wird j gesetzt, und so kehrt also auch hier die Schwankung zwischen j und z wieder, die wir schon im Pâ rsi gefunden haben.“(SPIEGEL 1861: 6).

[12]:

X^vaniraθa- is one of the seven climates (kars̆ var-; cf. BARTHOLOMAE 1961: 459) of the earth. It is described as a ‘splendid region’, and we find the expression […] imat̰ kars̆ varǝ yat̰ x^vaniraθǝm bāmīm […] in Mihr Yas̆ t vv. 15, 67 and 133 (cf. GERSCHEVITCH 1967 and HINZE 2014); Ras̆ n Yas̆ t 15 (cf. DARMESTETER 1883); Yasna 57, 31

(cf. MILLS 1887); Widēwdād 19, 39 (cf. CANNIZZARO 1990); and Hādōxt Nask I, 14.

[13]:

SPIEGEL 1861: 10.

[14]:

For the complete list of the ritual objects and divine names present in Nēryōsangh’s translation, see APPENDIX A.

[15]:

The purpose of this ritual object is to avoid contaminating the sacred fire with the priest’s involuntary spittings while reciting.

[16]:

About this very interesting topic, see WEBER 1857; THOMAS 1876; HUMBACH 1975; MACDOWALL 1975; idem 1978.