Perso Indica

The Tarjuma-yi kitāb-i Bārāhī is a Persian translation of Varāhamiria’s Bṛhatsaṁhitā, also known as Vārāhīsaṃhitā, done by ‘Abd al-‘Azīz Šams-i Tahānisarī who lived in the time of Fīrūz Šāh Tuġluq (r. 752-790/1351-1388). His name is alternatively given as ‘Abd al-‘Azīz Šams-i Bahā Nūrī, but since Tahānisarī is clearly spelled in the oldest preserved manuscript, this seems to be the correct version, associating his origins with the city of Thanesar, an important place of Hindu learning. According to the preface of the Tarjuma-yi kitāb-i Bārāhī, ‘Abd al-‘Azīz Šams-i Tahānisarī was also the author of a Ta’rīḫ-i Fīrūz-šāhī. The two historiographical works known by this title belong however to other authors, and Šams-i Tahānisarī should not be confused with Šams-i Sirāj ʿAfīf, the author of one of them (Jalali - Ansari 1985, pp. 163-164). We may yet speculate if Šams-i Tahānisarī’s Ta’rīḫ-i Fīrūz-šāhī is not in fact the same as the anonymous Sīrat-i Fīrūz-šāhī, whose author was certainly well versed in astronomy (for these astronomical parts, see Sīrat-i Fīrūz-šāhī 1999, pp. 301-320).

The translation was done by order and under the patronage of Fīrūz Šāh Tuġluq and is in line with other translations commissioned by that ruler, especially the Dalāʼil-i Fīrūz-šāhī (Badā’ūnī 1868, p. 249). Šams-i Tahānisarī therefore probably belonged to the entourage of the ruler and was somehow related to the court. He was well versed in astrology and astronomy. The assertion that he was a teacher of Sanskrit at one of the madrasas founded by Fīrūz Šāh (Ismāʽīlpūr 1380/2001, p. 1709) could not be confirmed.

The Bṛhatsaṁhitā was originally written in the 6th century by the famous Indian astronomer and astrologer Varāhamihira and commented upon by Utpala or Bhaṭṭotpala in the 10th century. Šams-i Tahānisarī used that commented version for his translation and refers in some places explicitly to the commentator (see, for example, Ms. London, British Library, India Office, 1262, f. 99b). While the Sanskrit text is composed in verses and contains only few parts in prose, the Persian text is in prose. The Sanskrit version of the text comprises one hundred and six chapters in Bhat’s edition and covers a vast range of prognostic sciences including astrology and all kinds of portents. The book also contains a lot of cultural and social information: chapters three to twenty and thirty-nine to forty-one are related to astrology and astronomy, chapters twenty-one to thirty-eight deal with different natural phenomena like earthquakes and rainbows. Other chapters are devoted to architecture (fifty-two, fifty-five), water springs (fifty-three), arbori-horticulture (fifty-four) and fauna (sixty to sixty-six), but also lamps (eighty-three), toothsticks (eighty-four) and precious stones (seventy-nine to eighty-two) are included (for a complete list, see Shastri 1996, 1, pp. 27-28).

The Persian translation closely follows the structure of the Sanskrit version. It does however omit eight chapters of the Bṛhatsaṁhitā because they were considered objectionable and related to unbelief (kufr). The chapters concerned are chapter forty-one (Indra’s banner), forty-two (lustration), forty-six (royal ablutions), forty-seven (crowns), fifty-four (temples), fifty-six (idols), fifty-seven (wood for making idols), and fifty-eight (veneration of idols). Moreover, the translator declares in the preface of the book that he left out all other passages dealing with kufr. His omissions and adaptations will be looked at in detail below. While kufr is the only issue Šams-i Tahānisarī makes mention of, modifications to the text have also been made with regard to women and sexuality. Other potentially delicate subjects like omens taken from the genitalia of men, the smell of semen or from urinating have been translated without restrictions. Although the numbers of the omitted chapters have been skipped in the translation, the chapter numbers in the India Office manuscript differ from Bhat's edition of the Sanskrit text by one, or rather two, since chapter twenty-seven has been annexed to chapter twenty-six and chapter thirty-eight is missing (in this article, the chapter numbers of Bhat's edition are used).

The Bṛhatsaṁhitā is very different from texts on astrological and prognostic methods from the Muslim world. This disparity results both from the dissimilar methods and concepts of prognostication used in India and from the fact that the entire range of prognostic methods is considered to be a part of jyotiṣa. Jyotiṣa, often translated as astrology, traditionally comprised mathematical astronomy (tantra), horoscopy (horā) and natural astrology (aṅgaviniścaya or śākha) (Gansten 2010a, p. 281). Varāhamihira, who also uses the term saṃhitā, applies it to the entire field of astrology as well as only to natural astrology (Shastri 1996, 2, pp. 430-431). This last part of the Bṛhatsaṁhitā consists of all sorts of omens taken from bodily marks, the movements of birds, signs of lamps, toothsticks and so on. While some of these resources, such as bodily marks, were used in the Islamic world for prognostication, there they do not belong to the field of astrology, but constitute separate branches of knowledge like physiognomy. This difference of classification is however only one problem the translator faces in his text. The other, more important challenge lies in the fact that many of the concepts and techniques described by Varāhamihira are completely unknown in the Muslim world or very different from practices there. Even in the field of prognostications based on the astral constellation, where we find a lot of shared principles, not all the concepts are commonly used by somebody trained in Islamic astronomy and astrology. The comparison between the Sanskrit and Persian version of some passages and chapters allows us to look more precisely at the translator’s methods of addressing foreign concepts and terms.

Chapter twenty-one deals with “the pregnancy” (garbha) of clouds, a method used to determine rainfall (Ms. London, British Library, India Office, 1262, ff. 81b-88b). As with most headings, the Sanskrit title of that chapter, megh garbha-lakṣaṇa, has first been transliterated and then translated into Persian as “abrī ki bārvar šawad kudām waqt ast,” which means “when is the time at which a cloud becomes fruitful.” In the text of the chapter, the Sanskrit expression garbha has always been kept transliterated in Persian script, as well as quite a number of other Sanskrit terms, which are in each case explained in Persian. The concept of megh garbha-lakṣaṇa meant that a cloud is supposed to give rain one hundred and ninety-five days after its formation and that this period is regarded as pregnancy (Shastri 1996, 2, pp. 496-499). This concept was certainly familiar to Indian readers, where the analogy between clouds and women was famously used by Kalidasa in his play Śakuntala and had become common in poetry (Kalidasa 2003, p. 73), but it was not known in Persian prognostication and meteorology and must have sounded rather strange to Persian readers. The key term garbha is translated in Persian not as “ābistan” (pregnant) but as “bārvar” (fructiferous), which was probably considered more appropriate for a cloud. Nevertheless, later in the same chapter, the destruction of the cloud fetus (garbha) is explicitly compared to the miscarriage of a woman (mānand-i ʽauratī ki pīš az āmadan-i mudda bačča-rā biāfkanad). While the translator does not comment upon this concept at all, he is very careful in providing the reader with additional information needed to understand the text. The Persian version therefore begins with a detailed description of the two types of Indian months (from full moon to full moon, pūrṇimānta, and from new moon to new moon, amāvasyā) and the names of each month. Since these names are later used to identify the moments of rain, they are essential to allow the reader to follow the text.

Chapter thirty-six on “aerial cities” (gandharva-nagara) offers another example of Shams-i Tahānisarī’s specific approach to foreign concepts (Ms. London, British Library, India Office, 1262, ff. 129a-130b). Again, the title is first given in Sanskrit, and then translated into Persian as “the rules of fog that give the air the appearance of forts, cities and houses.” The chapter deals with predictions deduced from such appearances, with both its direction and its color as main factors to be taken into consideration. The Persian version begins with a short explanation of that phenomenon and the Sanskrit term for it. It then continues with a translation of the Sanskrit text, with the Sanskrit term gandharva-nagara again kept throughout the translation. Since the different cardinal directions are of some importance for the prediction, the translator adds a paragraph on the directions used by Indian scholars at the end of the chapter.

The translation of chapter fifty-three on architecture poses more complex problems of adjustment and adaptation (Ms. London, British Library, India Office, 1262, ff. 158a-171b). It is a very long chapter dealing with the science of house building, which was known as vāstu-vidyā or vāstu-mandala and which was supposed to create a harmony between architecture and the universe (Narayanan 2010, p. 318). This science does not have any counterpart in the Muslim world. It describes, for example, the ideal measures for the houses of different social groups and office holders, different types of houses with and without verandas and halls, and the division of the ground plan of a house into eighty-one squares in which deities are to be placed (Narayanan 2010, p. 321). To render this chapter in Persian, Šams-i Tahānisarī translates some passages verbatim, especially those on measures, but skips the parts dealing with deities. The introduction in which the science of house building is related to Brahmā has thus been omitted, as well as the entire passage dealing with the placement of deities and short passages referring to evil spirits dwelling in the corners of a house. Rituals to be performed when starting to construct a house have also been omitted or transformed: the offering of worship has been Islamized into giving alms (ṣadaqa) to the needy. However, the mention of a temple (but-ḫāna) has been kept in the context of evil effects supposed to result from its neighborhood. Short explanations are interspersed with the text, usually introduced by “bidān ki” (know that). The Persian text adds, for example, that a house in which a specific relation between length, breadth and height is not observed is not good, but also information about the measuring units, their name in Sanskrit and their subdivision. The translator further explains that Vaiśyas and Śūdras are social groups (ṭā’ifa az hinduān) in India. At the end of this lengthy chapter, he adds some advice from his own expert knowledge: he suggests that the owner of the house should only move in at an auspicious time, giving details about fortunate and unfortunate constellations.

As these three examples show, the text often retains the Sanskrit vocabulary. This is especially true of all terms for which no Persian equivalent exists: while “rainbow” can easily be translated into Persian, “aerial city” cannot. The translator therefore usually first transliterates the Sanskrit term, explains it once and then uses it throughout the text. As a result, the Tarjuma-yi kitāb-i Bārāhī is a precious and extensive source for the reproduction of Sanskrit technical terms in Persian, with no attempt to find or create Persian equivalents for them.

The examples further demonstrate the translator’s awareness of the difficulties a reader unfamiliar with Indian concepts was facing. His explanations deal however mostly with technical aspects: he gives information on measures, directions, or social classifications, but does not refer to the background or contexts of the prognostication at all. This is especially evident in the chapter on architecture, in which he constantly writes about the different sizes of houses in relation to the number of verandas, the height of the building and the social affiliation of the owner. Nevertheless, Šams-i Tahānisarī does not comment on or explain these correlations, he just takes them as given. For him, it was clearly sufficient to be able to apply them.

The chapter on architecture is moreover indicative of the way the translator deals with religious issues and with what he calls kufr. As he announces at the beginning, Šams-i Tahānisarī has censored the text and omitted many passages dealing with temples, deities and their names, religious hymns, devotional practices and veneration. Although he does not define kufr as everything related to the hunūd or hinduān and does not call the Hindus infidels, it becomes clear from the use of these two terms that he applies them to designate a group religiously distinct from the Muslims. The translation of certain beliefs and practices is thus followed by expressions like “these are the words of the hunūd, but Muslims shall not do so” or “this is a custom of the hinduān which is not appropriate for the Muslims” (Ms. London, British Library, India Office, 1262, ff. 61a, 92b). The hinduān are also said to sing hymns for Indra. Expressions like “at a place where they burn the hinduān,” or “ascetics from a group of the hunūd and fire-worshippers” further point to a religious connotation of these two terms (Ms. London, British Library, India Office, 1262, ff. 2a, 150a, 52a). The Tarjuma-yi kitāb-i Bārāhī thus substantiates Carl Ernst’s observations on a shift in the use of the term Hindu from an ethnic designation to a religious connotation in the literature of the Delhi Sultanate. Accordingly, Šams-i Tahānisarī’s concept of kufr can likewise be related to the terminology of Islamic legal texts from that period in which Hindus are classified as ḏimmī or kāfir (Ernst 1992, pp. 23-26, 28).

When censoring the text, Šams-i Tahānisarī sometimes just leaves out single words, while in other places entire paragraphs or even chapters (see supra) have been omitted. Examples of his censorship are the passage dealing with the sun as the source of the universe and describing the creation of the universe out of a golden egg in the first chapter, the myths related to the sage Agastya at the beginning of chapter twelve, as well as hymns sung for the invocation of balances used to weigh seeds in chapter twenty-six. More interesting than those parts in which religious issues have just been left out are however those in which Šams-i Tahānisarī translates such issues, sometimes with modifications. We thus find in chapter eight the enumeration of the gas (eras) together with their lords Vishnu, Brahma, Indra, Agni and others. While the names of deities have been kept here, an interesting explanation is added for Agni: he is explained to be an angel (firišta) responsible for fire. Something similar happens with Durga, who is described as a devil (šaiṭān) in chapter forty-four (Ms. London, British Library, India Office, 1262, ff. 41b, 139b). In these instances, Indian deities have thus been “Islamized” by transforming them into beings well known in the Islamic world. A similar adaptation can be observed when worship and veneration practices are replaced by almsgiving (ṣadaqa), death at a holy place while doing yogic exercises by death on the Islamic ḥajj or at a place of pilgrimage (ziyāratgāh), and when trees standing in temples or on burial grounds are translated as trees populated by dīvs and parīs (evil and good demons) (Bhat 1981, pp. 491, 650, 496. Ms. London, British Library, India Office, 1262, ff. 170a, 212b-213a, 171a). But there is also another way of dealing with difference: in chapter sixty-eight, the Sanskrit text refers to metempsychosis. In this case, Šams-i Tahānisarī translates the text, but adds a comment saying that this is what people believing in metempsychosis think, and that such beliefs are futile and shall not be observed by Muslims. Some of the above-mentioned examples of the use of the terms hinduān and hunūd belong to the same range of argument.

The most sophisticated treatment of religious issues is to be found in chapter forty-six on the so-called utpāta (also known as adbhuta), that is anything contrary to nature (Ms. London, British Library, India Office, 1262, ff. 138b-141a). Utpātas are regarded as inauspicious portents created by the gods because of men’s misdeeds. To avert the negative effects of these portents, the gods have to be placated (Shastri 1996, 2, pp. 364-366, see also Gansten 2010b, pp. 744-745). The chapter describes altogether twelve sorts of utpāta, related to fire, crops, rainfall and so on. The Persian translation is much shortened and leaves out many portents. It renders in most detail explanations about the first sort of portents, dealing with idols of gods in temples. Šams-i Tahānisarī adapts the explanation of the utpāta to an Islamic audience by reducing the number of enraged gods (devatāḥ) to one: according to his version, utpātas are caused by the fact that men are not faithful in their religion (mustaqīm dar dīn-i ḫwīš nabāšand), tell many lies and therefore rouse the ire of the exalted god (ḫudā-yi taʿāla). Interestingly, the translation leaves open the option of this religion being something other than Islam. The same is true in the description of remedial measures to be taken: here again, the translator writes that the ruler and the population shall pray, do good deeds, give up their evildoing and shall practice the religious duties of their respective faith. This is somewhat surprising, since Hindu practices are regarded as kufr by Šams-i Tahānisarī. But if that is the case, why does his translation depict obedience to them as a means to placate the god? The translator’s reluctance to translate anything related to Indian deities and veneration practices becomes apparent again in the list of offerings—gifts of gold, food, cows and more in the Sanskrit version, and ṣadaqa in the Persian— that people should donate to avoid the effects of the utpāta. Šams-i Tahānisarī here omits to mention oblations into a sacred fire, and while he mentions that a cow shall be milked in such a way that her milk flows onto the ground, he leaves out that this should happen inside a Śiva temple. Yet, describing the portents related to idols of gods, he writes about temples, Durga who starts dancing, sheds tears, speaks and so on. Why does Šams-i Tahānisarī not omit this paragraph, as he does with so many others in which anything related to kufr is mentioned? To understand his motivation, one should take into consideration that Durga is classified here as a šaiṭān, and that all these utpāta take place, according to the translation, in the wilāyat-i kāfirān, the realm of unbelievers. I therefore assume that the fact that in the context of the utpāta, the idols are doing what they are not supposed to do and that their actions have negative effects induced Šams-i Tahānisarī to translate this paragraph, and that this is also the reason why he translated just this part of the chapter.

The only other topic which has deliberately been changed in Šams-i Tahānisarī’s translation are references to women and eroticism. Chapters seventy-four to seventy-eight are explicitly dedicated to this topic: starting with the praise of women (seventy-four), these chapters also deal with the winning of affection (seventy-five), erotic recipes (seventy-six), the preparation of perfumes (seventy-seven) and the union of man and woman (seventy-eight). Varāhamihira was the first to include chapters belonging to the field of kāmaśāstra (erotology) into jyotiṣa. Much of his information seems to have been taken from the kāmasūtra (Shastri 1996, 2, pp. 435, 456). In the Persian version of the text, chapters seventy-six to seventy-eight have hardly been changed at all. However, chapter seventy-four in particular has been greatly shortened. While in the Sanskrit version, women are described as jewels and gems, as clean and pure and as providing delight and worldly pleasures to men, all this has been omitted in the Persian version. Here, their greatest advantage is to produce offspring (Ms. London, British Library, India Office, 1262, ff. 225b-226a). While references to sexuality have not generally been left out, they are only described in the context of marital life. It was therefore not a problem to keep chapter seventy-eight. But while in chapter nineteen, one of the effects of a year presided over by the moon is, according to the Sanskrit version, that “charming ladies will always be delighting their handsome lovers with amorous sports,” the Persian version points here to love between spouses (Ms. London, British Library, India Office, 1262, f. 79a). Such alterations are probably due to changes in the attitude towards women and sexuality in the 14th century compared to the 6th century, but it is impossible to decide if such different views were typical of Muslims only, or applied to the entire society.

In conclusion, Šams-i Tahānisarī’s attitude towards the translation of source materials combines different approaches. He largely incorporates the technical lexicon of the original text in the Persian version. This method consists in leaving all technical terms untranslated and transliterated and explaining them in Persian. At the same time, the translation is selective; it removes parts of the text and adds glosses to it, and by doing this produces a composite effort to adapt the original source to the Muslim environment of the 14th century. Šams-i Tahānisarī provides additional information necessary for non-Hindu readers to understand the argument, especially in the field of measures, directions and so forth. He omits most references to Indian religions and even gives them a pejorative bias when he translates them. Moreover, he adjusts the image of women and sexuality to his own moral standards.