Perso Indica

The Tarjuma-yi Sālōtar is the earliest Persian adaptation of Indic sources on hippology (aśvaśāstra) produced in South Asia that is known so far. Sālōtar (sometimes written as Sālhōtar) is the Persianized form of Śalihotra, which is the title of treatises dealing with aśvaśāstra. Śalihotra is the name of a person deemed to have been a Vedic sage. The actual Śalihotra, author of a Sanskrit text on hippology, certainly lived before the 10th century. Several treatises on aśvaśāstra are attributed to him. The Śalihotra was also translated into Tibetan and Hindi. Other Sanskrit texts by different authors were also known by this title, such as the Aśvaśastra by Nakula (ca 1000) and the Śalihotra by Bhoja (ca 1050) (on the Sanskrit texts, see Maurer 2006, pp. 108-109; Meulenbeld 2000, pp. 558-579; on the Persian translations, see Speziale 2018, pp. 207-215).

The Tarjuma-yi Sālōtar was prepared by ‘Abd Allāh ibn Ṣafī who writes in the preface that he did the translation (tarjuma) in the city of Gulbarga by order (farmān, in one manuscript ḥukm) of Sultan Aḥmad Walī (r. 1422-1435) of the Bahmanī dynasty, who was well known for his patronage of the arts. The year given for the translation in the manuscripts is however 810/1407-1408, that is, before the accession to the throne of Aḥmad Walī. Gulbarga was at that time the capital of the Bahmanī sultanate. In the preface, ‘Abd Allāh ibn Ṣafī writes that he has rendered into Persian (fārsī kard) the translation (īn tarjuma-yi Sālōtar) done by Durgādāsa, the son of Sargrāsī (Durgarāsī ibn Sargrāsī in the Persian text), who was a resident (sākin) of Almala. Almala is probably the village of the same name which is nowadays located in Maharashtra and at the time the translation was produced was part of the Bahmanī sultanate.

It is likely that Durgādāsa did an intermediate translation of the Sanskrit text into a vernacular language that was used by ‘Abd Allāh ibn Ṣafī to create the Persian version. Either Durgādāsa or ‘Abd Allāh ibn Ṣafī seem to have included other material in the text, or used and collated various works on aśvaśāstra, since the Persian text refers at one point explicitly to Sālōtar and some ḥukamā’ as its source, and shortly afterwards emphasizes that the next chapter has been taken from the books (kutub) of Sālōtar (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 6, p. 135). The way the source is referred to at the beginning of the chapter on horses proper for kings, jami‘-i nusḫahā-yi Sālōtar, meaning “all the copies of the Sālōtar”, also suggests that different Indic works were used to produce the Persian version (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 21, f. 70b). For concerning the part on prognostication, the contents of the Persian translation include material that is similar to Nakula’s and Bhoja’s treatises but do not match them exactly.

The Tarǧuma-yi Sālōtar can be divided into four parts: an introduction, consisting mainly of a story about primordial winged horses, a part on omens and prognostication, a part on veterinary medicine and a final section, containing once more prognostic information but also instructions about how to train a horse. The extant manuscripts of ‘Abd Allāh ibn Ṣafī’s translation show a high degree of variation. A number of textual variants are found in manuscript copies but also variants in the order and number of chapters. As regards the division and the order of chapters, the majority of manuscripts is organized in a rather similar way, with some slight changes, including the oldest dated copy (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 6). Two manuscripts (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 21 and Ms. Montreal, Library of McGill University, BWL W55/2) differ significantly from that order and include a lot of additional material, which is also partially found in another manuscript (Ms. Cambridge, Cambridge University Library, add. 307).

In the Sanskrit tradition, the wing story is contained in the Aśvaśastra by Nakula and some later Sanskrit treatises on horses, such as Godāvaramiśra (late 15th-16th c., see Meulenbeld 2000, p. 562). According to this tradition, horses were originally created with wings and were flying among the clouds. Indra was very impressed by them, but considered them to be even more useful without wings. For that reason, he asked Śalihotra, the master of horses, to chop off their wings. The latter did so, but in exchange for this loss, he taught men the secrets of horses to ensure that they treat them well (Nakula 1952, pp. 206-207; Maurer 2006, p. 111). The Tarǧuma-yi Sālōtar adopts this story and gives it a more Islamic frame. In this version, horses were also formed with wings, and were able to fly around by their own strength. Before the creation of Adam, Iblīs, i.e. Satan, warned the horses of men’s authority and their use of horses as mounts. For this reason, the horses tried to prevent the creation of Adam. Later, when Adam was created, God wanted his prophets, saints and believers to be able to ride on horses. Śalihotra, who had a deep knowledge of horses, therefore cut their wings and cauterized them, and God turned them into beings without wings. Similar to the Sanskrit tradition, Śalihotra wishes afterwards to guarantee the horses’ well-being and therefore spreads knowledge about them to mankind. The Persian version of the wing story shares elements with the Sanskrit tale, but also reinterprets it and adapts it for a Muslim readership. The most noticeable feature of the Persian version is that horses are considered primordial to men and that they try to take an active part in the creation process by opposing God’s plans. Instead of Indra, Iblīs and Jibrīl – the angel Gabriel – are the main agents of the story, Jibrīl on behalf of God, and Iblīs as his opponent. Interestingly, several manuscripts include a second version of this story, with the prophet Solomon being ordered to cut the horses’ wings because they had disobeyed him. In this version, the traces of the Indian background are barely distinguishable (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 21, ff. 37b-38a, Ms. Montreal, Library of McGill University, BWL W55/2, pp. 221-223, Ms. Cambridge, Cambridge University Library, add. 307, ff. 25a-26a).

Chapters on omens and prognostication taken from horses are contained in several Sanskrit treatises on hippology. They belong to the vast domain of natural astrology (aṅgaviniścaya or śākha) which covers a broad range of prognostic sciences, and is also dealt with in the Bṛhatsaṁhitā by Varāhamihira (6th century), a Sanskrit treatise that was translated in Persian in the second half of the 14th century (see Orthmann 2017). Neither the Nakula nor the Bhoja Śalihotra however comprise the same choice of omens in the same order as the Persian version by ‘Abd Allāh ibn Ṣafī. His version especially emphasizes bad omens and commercial aspects: it is important to know a horse’s character to decide whether to buy it or not. Many prognostications therefore contain an acquisition recommendation. Furthermore, a separate chapter on buying horses is included at the end of the book, with the recommendation based on omens taken from the movement of the horse’s tail or its legs. The insertion of Islamic elements is also evident in this chapter, since the horse to be observed has to be turned towards the qibla, the direction of prayer.

The largest part on omens is covered by the description of the āvartas, called by the Hindi term bhauṅrī in the Persian text, a kind of specific curls or hair formations in eight different shapes. Although the āvartas are of major importance in the Sanskrit versions, too, the commercial aspect of the prognostication is peculiar to the Persian version. In contrast, the part on deformities, describing e.g. horses with a kind of horn looking like a third ear, or horses with five legs, closely follows the Śalihotra by Bhoja. The internal organization of the chapters of this part as well as their length vary in the manuscript copies. Each prognostication usually includes the Indic term for the deformity or the bhauṅrī described, and then provides an explanation of its meaning in Persian. The consequences of buying such a horse are often exemplified by referring to some specific person, mostly a ruler, who possessed a similar horse and subsequently suffered losses or death. The names of these persons are standardized, with the same ruler associated with the same type of horse. Many manuscripts include illustrations of the deformities and the bhauṅrīs described in this part, or were supposed to include them, with spaces left for the depictions.

In the chapter on the character of horses, these are subdivided into Brahmin, Kśatriya, Vaiśya and Śudra horses, a classification contained in Nakula’s Aśvaśastra and in other texts such as Godāvaramiśra (Meulenbeld 2000, p. 562). However, in the Persian version, the assignment of a horse to one of the four categories is done by looking at the way it drinks water. In Nakula’s text, it is based rather on the horse’s character (timid, wild, coarse, etc.) and forms a part of the chapter on training horses. The Tarǧuma-yi Sālōtar also contains a chapter on training horses. This is located at the end of some manuscripts and certain copies are illustrated with drawings of the courses (kāwa) used to train horses. The chapter focusses on different courses the horse has to be trained on successively. It might have been influenced by Persian and Arabic traditions as well as by Indian ones. A close parallel to these courses is contained in Faḫr ad-Dīn Mudabbir’s Ādāb al-ḥarb wa al-šajā‘a, a book dedicated to Sultan Iltutmiš (r. 1210-1236), where we find a long list of equestrian training courses. These are explicitly associated with Muslim masters (Mudabbir 1346, pp. 206-216). There was also an Indian tradition of training horses, reflected in Nakula’s text in information on the condition of the training ground, the best time for training a horse and the qualities of people suited to train it. The translation uses some Indian terminology for the courses (kāwa) and for small bells (ghuṅgr) employed in the training, but does not give Indian names for the different courses.

Chapters on the gait of horses (raftār-i aspān) and horses proper (lāʼiq) for kings are included in two manuscripts (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 21, Ms. Montreal, Library of McGill University, BWL W55/2). The latter chapter is explicitly described as having been taken from the Sālōtar. In the version by Nakula, we find a parallel to this text, although it is much shorter, suggesting that the translator may have used another source or additional material. The comparison of a horse’s gait with that of specific animals (a peacock, a young elephant, etc.) is not included in Nakula’s or Bhoja’s version, but can be found in the Bṛhatsaṁhitā (chapter 66).

One last part of the section on prognostication deserves mention: the chapter on the provenance of horses. This chapter attributes good or bad qualities to horses depending on the region they come from. Places mentioned here are inter alia Khurasan, Bukhara, Hormuz and Aden, pointing to the key role of the trading networks with the Muslim world for the import of horses to India. This topic is not included in Nakula’s and Bhoja’s treatises, which were written before the arrival of the Muslims in India; a chapter on the regions from which horses come is however contained in the later treatises on horses by Godāvaramiśra and Basava (ca 1684-1745) (Meulenbeld 2000, pp. 562, 558). This could suggest that elements taken from the Muslim environment were integrated in the translation to update the contents of the text to the features of contemporary horse markets, linking the text again to commercial aspects, which dominate the entire section on prognostication of the Persian version.

The chapters on medical topics are grouped together and look at the causes, symptoms and treatment of horse diseases. Important textual variants and variants in the titles of chapters are found in the Persian manuscript copies. Topics of medical interest are also discussed in other parts of the text, such as in the chapters on the horse’s testicles (bayża) and the knowledge of a horse’s age (‘umr) by its teeth. A number of Indic terms are used in the medical chapters, especially concerning the terminology of diseases. A first group of chapters deals with diseases of the stomach, for which the Indic term sūl (colic, stomach-ache) is used and glossed by the Persian compound term dard-i šikam (stomach pain). The Indic names of each disease are provided at the beginning of paragraphs (faṣl), as in the thirteenth faṣl dealing with vāyu sūl, that is to say, the stomach troubles associated with the Ayurvedic humour vāyu, the wind. Arabic technical terms are sometimes used, such as the compound expression ḥabs al-bawl for strangury. A following group of chapters (eight or nine according to manuscript copies) is devoted to the types of fever. Different terms are used here to refer to fever, the Indic jar and the Persian tap and larza (tremor). Indic names of fevers are given at the beginning of each faṣl; the first deals with sanpāt jar – sanpāt is the association of the three humours of Āyurveda – that is characterized by a bad smell coming from the horse’s mouth.

Humoral pathology occupies an important place, and a group of chapters is devoted to the troubles associated with the wind. The term vāyu is largely used in these chapters as in other parts of the text, although the Persian equivalent bād is employed in the table of contents provided at the beginning of the translation. The translator’s choice is rather uncommon. South Asian writers of Persian medical texts seldom use vāyu, they commonly use bād to refer to Ayurvedic wind (see Speziale 2018b). The use of bād is well established in the Persian treatises dealing with Ayurvedic materials produced during the 14th century, such as the Šifāʾ al-maraż, written in 790/1388 by Šihāb al-Dīn Nāgawrī (see Speziale 2014). This could possibly suggest that ‘Abd Allāh ibn Ṣafī was not himself a physician. Weights of drugs used in compound formulas are always provided in dirham, a choice that shows the intention to adapt the translated materials for the Muslim readership.

The copies known so far of ‘Abd Allāh ibn Ṣafī’s translation were produced during the Mughal and the colonial periods. Several manuscript copies were illustrated with drawings of horses and some include drawings of horsemen, horse keepers and riding courses. The oldest dated copy, made in 1058/1648, further testifies to the process of Islamisation of the translated materials (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 6, pp. 154-169). After the colophon, this copy is illustrated with more than thirty coloured drawings of the horses of Muslim prophets. In each drawing, a colour or a type of horse is associated with a specific prophet. The prophets’ horses illustrated here are those of Ibrāhīm, Isma‘īl, Šu‘ayb, Dānyāl, Ilyās, Yūnus, Hūd, Yūsuf, Ayyūb, Ya‘qūb, Isḥāq, Nūh, Sulaymān, Mūsa, Zakariyā, Ṣāliḥ, ‘Isā and Muḥammad. One drawing depicts the Burāq, the winged creature ridden by Muḥammad during his heavenly journey, and eight drawings illustrate the horses of Muḥammad (d. 11/632). There are also coloured drawings of the horses of Ḥiżr, a figure evoked in the Quran, the imams Ḥasan (d. 50/670) and Ḥusayn (d. 61/680) as well as the “friend of God” (ḫudā kī yār), i.e. the Sufis. The drawings are explained by captions that look similar to the text of an Islamic prayer to the prophets – and including Ḥasan and Ḥusayn – that is found at the end of other copies of ‘Abd Allāh ibn Ṣafī’s translation (Ms. Hyderabad, Andhra Pradesh Oriental Manuscript Library and Research Institute, ṭibb-i ḥaywānāt 21, ff. 157b-158b; Ms. Montreal, Library of McGill University, BWL W55/2, p. 352, defective at the end). This prayer consists of a series of repetitive invocations to the prophets and their horses, and its structure resembles that of the masters’ silsila (genealogy) which is recited by the Sufis. It is likely that this invocation was associated with the guilds of Muslim horse keepers, as in the case of the genealogy (kursī-nāma) of the Muslim elephant keepers that is found in the Kursī-nāma-yi mahāwat-girī; this is an anonymous Persian treatise on the elephant, based on Indian materials, where it is stated that the first elephant keepers were the three grandsons of the prophet Noe (see Speziale 2018, pp. 108, 220-221).

Manuscript copies show several features of the circulation of the text in the colonial environment. The style of the coloured drawings of horse keepers in one copy kept in Montreal show Western influences (Ms. Montreal, Library of McGill University, BWL W55/1, pp. 105-108). Certain manuscripts were read and collected by members of the colonial elite. In the copy kept in Cambridge, the Latin names of some of the drugs mentioned in the text have been written in the margin of some folios, suggesting that the text was used and annotated by a Western reader. The annotation “C. Patrick, Surat, 1801” found at the top of two folios of the manuscript could be by the reader (Ms. Cambridge, Cambridge University Library, add. 307, ff. 60b, 95b). An English letter, dated August 20th 1840 and prefixed to the copy kept at the British Library, explains that this copy was sent to Reverend John Wilson D. D., president of the Bombay Branch of the Asiatic Society, by “Tirmal Ráo son of Ráo Bahadoor Venkut Ráo principal Sudr Ameen of Dharwar”, a city nowadays located in Karnataka (Ms. London, British Library, add. 14,057, f. 1a).