University of Amsterdam
In 1771, a French scholarly adventurer by the name of Abraham-Hyacinthe Anquetil-Duperron shocked the Republic of Letters with his translation of Zend-Avesta, Ouvrage de Zoroastre. Published in three volumes with a long series of appendices and a book-length introduction about Anquetil’s travels in India, it offered the first known example of a monotheistic text with no direct relation to Judaism, Christianity, or Islam.
The shock was not that there was a Persian prophet called Zoroaster; his name had been known in Europe since antiquity. (Modern-day readers may know him as Nietzsche’s Zarathustra or Mozart’s Sarastro.) He had been portrayed as an ‘Eastern sage’ from Pico della Mirandola and Marsiliano Ficino to Jacob Brucker’s mid-18th century history of philosophy, and Hyde in Oxford already possessed an Avestan manuscript which had inspired Anquetil’s journey to India in the first place. The shock was that the Zend-Avesta was not a work of prisca theologia or philosophia perennis but a liturgy, at once strangely familiar and uncannily remote, written in a dead language that no one in Europe except Anquetil could read.
Anquetil was duly attacked as a fraud. There was reason for scepticism, for the learned world was grappling with other hoaxes. It had been a mere decade since the Scottish crook MacPherson had won literary fame with his volumes of the songs of Ossian, the ‘Nordic Homer’ – and though MacPherson had been discredited and never showed the ‘old manuscripts’ that he had purportedly found in Scotland, he could still count Goethe and Napoleon among his admirers even after his death. Anquetil himself had been led astray, together with Voltaire, by manuscripts of a supposedly ‘Vedic’ text with Christian undertones, fabricated by a Jesuit missionary, which was published in 1778 as L’Ezour-Vedam: ou, Ancien commentaire du Vedam, contenant l’exposition des opinions religieuses & philosophiques des Indiens. The spectacular story of Anquetil’s oriental manuscript hunt, which included a duel, a flight through the jungle, and wheeling and dealing with the English and French in the middle of the Seven Years’ War, also raised eyebrows among more sedentary scholars.
The most malicious attack came from William Jones, then a budding young orientalist who had just finished a Persian grammar. In an anonymous Lettre à M. A*** du P***, he argued a) that Anquetil hardly knew enough modern Persian to communicate with his informants in Surat, and these Parsi priest themselves hardly understood the dead languages of their holy books (this was true); b) that Anquetil was an impolite and unreliable rascal (this was half-true); and c) that what he presented as the Zend-Avesta was too ‘absurd’, ‘tedious’, and ‘repetitive’ to be a genuine holy text.
Johann Friedrich Kleuker, Anquetil’s German translator, made short ado of that. He retorted that presumed ‘absurdity’ and ‘tediousness’ did not prove that something was a modern forgery: “Muβ alles, was alt ist, weise seyn?” Quite the contrary, it showed that Jones’s ‘poor judgement’ was guided by anachronistic criteria, according to which even some laws of Moses would have been quite as absurd. (Besides, why would a forger write an absurd, tedious, and repetitive text?) Also, Jones was not in a position to judge, as a lawyer dallying in Persian love poems; and his criticism, marred by ‘chauvinism, bitterness, partiality, and passion’, contained “keine Data”.
What united Anquetil, Jones, and Kleuker is that they were all looking for confirmations of universal Christendom. This is why something as ‘absurd’ as the Zend-Avesta, for Jones, could not be genuinely old and venerable; and this is why, for Anquetil and Kleuker, every religion has its points cachés. Edward Gibbon, in The Decline and Fall of the Roman Empire, saw no objections to using Zend-Avesta as a historical source about the Persians, the more because “there are some remarkable instances in which Zoroaster lays aside the prophet, assumes the legislator, and discovers a liberal concern for private and public happiness”. True, the “motley composition” also contained “the most abject and dangerous superstition”; but then for Gibbon the Bible too was absurd, tedious, and repetitive.
All this did not solve the problem that no one could check Anquetil’s translation. Avestan, the original language of the Zoroastric corpus, was a dead language related to Old Persian and more remotely to Sanskrit (all authors mentioned in this essay refer to it as ‘Zend’). It is still unclear whether the corpus dates from the 6th century B.C.E. or even further back; at any rate, the alphabet in which it is written down dates from a much later period, and is derived from Pahlavi, a writing system for Middle Persian. The manuscripts that Anquetil gathered in Surat also contained commentaries and interpolations in Middle Persian and glosses and translations in Sanskrit, the latter attributed to the 15th/16th-century sage Nariosengh and intended to convert Indians to Zoroastrianism. Apart from Anquetil’s own notes, there was no grammar or vocabulary of Avestan; his manuscripts were deposited at the Bibliothèque du Roy, and for lack of Avestan printing type, they could not be reproduced and circulated. When Anquetil died in 1805, the case was still undecidable.
Between 1819 and 1823, Rasmus Rask travelled through Persia and India on a philological expedition funded by the Danish crown. He came back with a burning hatred of Asia and everything Asian, and devoted most of his efforts to Germanic languages afterwards, but he also wrote a tract that confirmed the Zend-Avesta’s ‘age and authenticity’ (Om Zendsprogets og Zendavestas Ælde og Ægthed, 1826). He also harshly criticized Anquetil’s translation, though, arguing that Anquetil rendered the letters of the Avestan alphabet sloppily and attributed the wrong (or not exactly accurate) sound to half of them. To prove his point, he put his own and Anquetil’s version side by side in a table, exaggerating the difference by putting Anquetil’s reproduction of the letters as they occur in the manuscript next to his own standardized, aligned version meant to serve as the basis for a typeface. Even after that scrutiny, though, at least one leading orientalist remained sceptical: A.W. Schlegel wrote in his Réflexions sur l’étude des langues asiatiques, as late as 1832, that “ceux qui affirmaient, Anquetil à leur tête, et ceux qui niaient, marchaient également dans les ténèbres”.
The issue was finally resolved by Eugène Burnouf in the 1830s. In his Commentaire sur le Yaçna (1833), he subjected the first 13 pages of Anquetil’s main manuscript to a word-by-word analysis, collating the Avestan text, the Sanskrit translation, Anquetil’s translation, and different versions, establishing the most likely meaning and grammatical role of each word and phrase. Each paragraph begins with a line of Avestan, then a line of Sanskrit (when available), then Anquetil’s version, and then pages of Burnouf’s own commentary. Eventually, these 13 pages of Avestan took him 592 pages of commentary + cxcii pages of notes. It is no surprise that a projected second or even third volume never saw press.
Burnouf also did something else. He had a lithographic reproduction made of the entire 562-page manuscript – a new means of reproduction that was faster and cheaper than engraving. Even so, it took 14 years to complete, partly because the lithographer died halfway. Like Burnouf’s Commentaire, it had only a limited circulation: the print run for the reproduction was 100. After all, this was still a text that almost nobody could read, if not with the help of Burnouf’s own Commentaire. But at least theoretically, it made Burnouf’s judgement apprehensible.
Along the paper trail from Anquetil to Burnouf, we can trace the development of linguistics as a discipline. In William Jones’s attack on Anquetil, language was not yet considered an object of study in its own right: “Ne savez-vous pas que les langues n’ont aucune valeur intrinséque?” Kleuker, Anquetil’s German translator, argued that Avestan had “alle Eigenschaften einer alten unausgebildeten Natursprache”, bearing resemblance to ‘Scythian, Iberian, Georgian, Celtic, and Latin’. What mattered for Rask, writing in the 1820s, was the structure of the Avestan language and its relation to other ancient oriental languages, and its place in a wider genealogy of races, families, stems, branches, languages, and dialects. Burnouf’s Commentaire is first of all an exercise in comparative grammar applied to text edition, filled with some 100 references to Franz Bopp. Bopp, in turn, included Avestan in his Vergleichende Grammatik, of which the first volume appeared in the same year as Burnouf’s Commentaire – at first drawing upon Anquetil, Rask, and earlier publications by Burnouf, but from the second delivery onward, there are regular footnotes to ‘Yaçna’.
Along this paper trail, the Zoroastric texts themselves underwent several transformations. From liturgical texts, they became the subject of philological-theological speculation, and then were turned into language material and linguistic data. For Anquetil, they were texts whose contents should be translated and elucidated; for Burnouf, they were samples that required faithful reproduction, technical commentary, and dissection. They even changed title: Anquetil had published them under the muddled collective title Zend-Avesta, while Burnouf used the more accurate designations Vendidad-Sadé and Yaçna for parts of the corpus.
The story of the Zend-Avesta controversy is thus also a story about the generation of expert judgement in the study of language. Passing an expert judgement means claiming a kind of professional authority that overrules ‘amateurs’ and ‘traditions’, on the basis of accumulated insight and a ‘methodical’ or ‘systematic’ approach. In a simple formula, expert judgement means being able to correct others, like Burnouf (and also Rask and Bopp) were doing. It also required a form of presentation – rhetorical, visual – by which something is presented as knowledge, as in Burnouf’s line-by-line dissection of the Yaçna, or his folio source publication of the Vendidad-Sadé. It required a certain narrowing down of the subject, excluding such theological agendas as espoused by Jones and Anquetil, but it was – and is – also frequently at odds with arbitrary disciplinary boundaries. Finally, expert judgement like Burnouf’s required significant maintenance and resources. It took institutions like the Collège de France and the École des Langues Orientales to produce the Sanskrit and Avestan printing types required for his editions – they had only been cut a few years earlier – and his facsimile could hardly have been produced anywhere else than in the metropolis.
The paper trail does not end there. Burnouf sent a copy of the Commentaire to Henry Rawlinson, who was then stationed as a British agent in Kandahar and Baghdad, working on his decipherment of the cuneiform inscriptions which he had transcribed at Behistun. Old Persian, one of the languages of the Behistun inscription, was after all a near cognate of Avestan. The influence of Burnouf’s Commentaire on Rawlinson’s eventual report, which filled two volumes of the Journal of the Royal Asiatic Society in 1846–49, is evident most of all in his comparisons of Old Persian to Avestan and Sanskrit, and in his sentence-by-sentence analysis of the Persian text in ch. IV.
In the seventy years between Anquetil and Rawlinson, not everything had changed. They were both adventurers and amateur scholars working on private initiative, though with official sanction, and reporting back to the Royal Academy in Paris and the Royal Asiatic Society in London. There is a touch of Indiana Jones to both of them. The major difference is that in Rawlinson’s case, the feedback loop functioned much faster: while he was busy in Kandahar and Baghdad, Burnouf and other scholars in Europe were also making progress; and within years of the publication of his report, he had a rival in deciphering cuneiform, Edward Hincks, who had never left the British Isles. But even in an era of state-funded expeditions, and in a slightly later era of linguistic fieldwork, the problem remained that expert judgement requires autopsy, and institutions could not initially provide it. To see things for yourself, you had to sit with pundits in Surat or scale rocks in Behistun.
 For a particularly acerbic account of “The Ossian Controversy”, see Hugh Trevor-Roper, The Invention of Scotland (New Haven: Yale University Press, 2008), ch. 6. It is ironically appropriate that Trevor-Roper himself had been victim of one of the biggest historical hoaxes of the 20th century, the ‘Hitler diaries’.
 The complicated story of the Ezour-Vedam, created in the 1730s by Jean Calmette S.J. and translated into broken Sanskrit and Telugu by his fellow missionaries, is in Urs App, The Birth of Orientalism (Philadelphia: University of Pennsylvania Press, 2010), ch. 7
 Johann Friedrich Kleuker, Anhang zum Zend-Avesta vol. II.2 (Leipzig/Riga: Hartknoch, 1783), p. 74
 Ibid., pp. 72, 111
 Edward Gibbon, The History of the Decline and Fall of the Roman Empire, vol. I (London: Strahan, 1776), ch. viii, p. 205
 Ibid., pp. 206-7
 Wikipedia, the Encyclopaedia Britannica, and several other sources refer to a “Dissertation on the Authenticity of the Zend Language” by Rask in vol. III of the Transactions of the Literary Society of Bombay (1821 [actually 1823]). There is none, at least not in that volume. The Transactions of the Royal Asiatic Society 3:3 (1834), pp. 524-540, posthumously published a letter by Rask to the L.S.B. President, Mountstuart Elphinstone, reacting to a paper in vol. II.
 August Wilhelm Schlegel, Réflexions sur l’étude des langues asiatiques (Bonn/Paris: Weber/Maze, 1832), p. 67
 William Jones, Lettre à M. A*** du P***, dans laquelle est compris l’examen de sa traduction des livres attribués à Zoroastre (London: Elmsly, 1771), p. 11
 Kleuker, Anhang zum Zend-Avesta, vol. II.2, pp. 9-10
How to cite this post
Solleveld, Floris. 2018. What Zarathustra said: The sixty-year controversy regarding Anquetil-Duperron’s Zend-Avesta. History and Philosophy of the Language Sciences. https://hiphilangsci.net/2018/03/27/zarathustra/