Pahlavi Texts, Part I — SBE Vol. 5 (trans. E. W. West, 1880)

West opens by explaining why these volumes exist at all, and the reason is a gap rather than a discovery: "though we must look to the Avesta for information regarding the main outlines of the Parsi religion, it is to Pahlavi writings we must refer for most of the details relating to the traditions, ceremonies, and customs of this ancient faith." The Avesta gives the outline; the detail survives only in the commentary literature. A reader who wants to know what was actually done, rather than what was believed, has to come here.

He then divides his own material by how far it can be trusted, in the introduction, before a reader reaches a line of it. One class translates an Avesta text that still exists and can be checked against it. The other consists of "translations and Zands of Avesta texts no longer extant" or "the opinions and decisions of high-priests of later times, when the Pahlavi language was on the decline" — and of those he says flatly that their "authority is more open to dispute." An editor grading his own contents is rarer than it should be.

The sentence that follows cuts at the living community and at the scholar equally: such writings "would hardly be considered of indisputable authority by any Parsi of the present day, unless they coincided with his own preconceived opinions." He is describing confirmation bias in a tradition's selective use of its own literature — in 1880, in a volume that tradition's members were expected to read. Note that the observation is not offered as a reason to dismiss the texts. His very next move is to say what they are good for instead: "for outsiders they have the inestimable value" of preserving details "which would be vainly sought for elsewhere."

THE WRITTEN LANGUAGE WAS NEVER THE SPOKEN ONE. This is the fact a reader most needs and is least likely to guess. Pahlavi mixes Persian with Semitic (Huzvâris) words, "so that the language has the motley appearance of a compound language" — but West argues it "was never spoken as it was written." The Semitic words were logograms: "the Persians would write malkan malka, 'king of kings,' but they would read shahan shah." His own analogy is the one to keep: they were "merely a Pahlavi mode of writing their Persian equivalents (just as 'viz.' is a mode of writing 'namely' in English)." A Pahlavi manuscript is a notation, not a transcript.

His proof of that is an argument from disappearance, and it is elegant. When the Persians adopted the modern alphabet, the entire Semitic component of their written language vanished at once — which is only explicable if nobody had been saying it. "As the Semitic words were merely a Pahlavi mode of writing their Persian equivalents, they disappeared with the Pahlavi writing, and the Persians began at once to write all their words, with their new alphabet, just as they pronounced them." A change of script does not delete vocabulary people actually use.

THE ALPHABET HAS FOURTEEN LETTERS, and the count is the whole difficulty. Several letters carry many sounds, joined letters resemble other single letters, and West reports "some compounds of two letters which have from ten to fifteen sounds in common use." Three letters serve five long vowels and about five short ones. Reading Pahlavi is therefore not decoding but disambiguation, and West sets out the full table of ambiguities rather than asking to be taken on trust.

THE ADMISSION AT THE CENTRE OF THIS INTRODUCTION. Where the compilers of the traditional glossary could no longer recover a word's pronunciation, "they adopted the most obvious reading of the letters before them, which thenceforth became an artificial word to be handed down to posterity, by successive generations of writers, WITH ALL THE AUTHORITY OF OLD TRADITION." He names the resulting forms: anhûmâ for aûharmazd, yahân for yazdân, madônad for maînôk, shatan for shatrô. This is a claim that some of what a tradition receives as ancient is the residue of an earlier act of guesswork — stated in the front matter of a volume of the Sacred Books of the East, and offered with examples so a reader can check it.

Having said that, he does not simply substitute his own guesses. He goes outside the manuscripts for a control: the rock inscriptions and coin legends of the Sasanian kings give "some means of ascertaining the ancient pronunciation of a few Huzvâris words, independent of the opinions of comparative philologists." The earliest, for Artakhshatar son of Pâpak (A.D. 226–240), is cut in Greek and two Pahlavi scripts at once. A trilingual inscription on rock, datable to a reign, is evidence of a kind a manuscript copied for a millennium cannot be — and West says plainly that it can settle questions "left in doubt by the manuscript alphabet."