Varieties of Middle Persian Part 2

Varieties of Middle Persian Part 2: Changing Letters in Changing Times?

By Tommy Benfey, 21 June 2022

As discussed in my previous blog on Middle Persian, most people who study Middle Persian tend to focus on the Book Pahlavi corpus—the group of texts largely or entirely composed and transmitted by Zoroastrian priests, which reach their final form in the 9th or 10th century CE. There are other varieties of Middle Persian, however, which differ substantially from Book Pahlavi. In this blog, I’ll be focusing on one of the key differences between Book Pahlavi and Documentary Middle Persian—the variety of Middle Persian in which the documents that survive from the sixth, seventh, and eighth centuries were written. These documents are incredibly interesting and important sources for late Sasanian and early Islamic history, but remain understudied and undercited. Aside from easing the way to studying these documents for those with a background in Book Pahlavi, the following discussion should also be interesting for anyone wanting to know the history of the languages and scripts of ancient and medieval Iran. In particular, I’ll be discussing an apparently novel feature of the documentary script—a new shape for the letter T. But scholars continue to debate whether this really constitutes a new development, or whether this T is actually something very old.

1. One of the starkest differences between the documentary and Book Pahlavi scripts, already noted by Olaf Hansen in the 1930s, lies in their respective renderings of the letter T. 

Book Pahlavi is notorious for its many ambiguous letters. Below you see a few lines from a well-known Book Pahlavi text, the Judgments of the Spirit of Wisdom (Dādestān ī Mēnōy ī Xrad), from the late 15th-century manuscript K 43. 

 
From P. O. Skjærvø’s Pahlavi Primer

Here I’ve highlighted several signs that are each essentially a line—sometimes vertical lines (as in the right- and leftmost examples) and sometimes a line at a slight tilt (as in the middle two). Despite their identical appearance, these signs can actually serve quite distinct purposes. While the two rightmost signs that I have highlighted are both the consonant W, the next one (proceeding from right to left) is an N, and in the final word in the line (again, reading from right to left), we have a W followed immediately by an R, and then, at the end of the word (the leftmost sign in the line), a sign with no phonetic value, the “word-final stroke,” which simply denotes the end of a word. So, to summarize, even in this brief snippet of text, we can see the same sign doing four rather different things: denoting the consonants W, N, and R, and marking the end of a word. 

2. Mercifully, however, the Book Pahlavi script does have a few unambiguous letters, and its T is among them. I have highlighted its occurrences in the text below. 

 
From P. O. Skjærvø’s Pahlavi Primer

In the documentary script, however, we encounter a different letter T to what we had seen in Book Pahlavi. All too often, this T looks very similar or even identical to the sign we looked at above: the sign that can already be W, N, R, or the marker of the end of a word. I have marked a few examples of T in the document reproduced below, Berlin 1, which comes from the vicinity of the city Qom, in the early 680s CE. 

 
Weber 2008, plate 1

The top T I have highlighted is indistinguishable from the W, N and R; the second looks a bit different, as it has a notable additional stroke; and the one on the bottom is the T familiar from Book Pahlavi. Most of the Ts we encounter in the documents resemble the first two I have highlighted: either they are identical to the W/N/R sign (like the top T), or they (like the T in the middle) resemble a W/N/R with an additional stroke at the beginning—and this additional stroke, as it happens, resembles yet another ambiguous letter, the G/D/Y. 

It is unclear where this other kind of T came from, which is distinct from the typical T we see in Book Pahlavi. It could simply be a simplified version of the latter, or it could be a survival from an earlier stage in the development of the Middle Persian script, as Dieter Weber has suggested. 

3. Both the Book Pahlavi and documentary scripts ultimately derive from Imperial Aramaic. Imperial Aramaic had, essentially, two T’s: one called ṭēt (and typically transliterated as ṭ), and the other called tāw (and typically transliterated simply as t). As shown in the table below, we see these two distinct T’s in the Middle Persian and Parthian inscriptions of the early Sasanian era—one T derives from the Aramaic ṭēt, and the other from the Aramaic tāw. However, the sign based on the Aramaic ṭēt is rare already in the Middle Persian inscriptions, and falls entirely out of use in Book Pahlavi (as well as the “Psalter script” which is often regarded as Book Pahlavi’s predecessor). 
 

Weber 1993, 191 (based on table in D. N. Mackenzie, A Concise Pahlavi Dictionary, London: Oxford Unversity Press, ix). 

But in the documents, we encounter two T’s—one resembling the Book Pahlavi T, and one that is not encountered in Book Pahlavi. Weber has suggested that the latter kind of T derives not from the Imperial Aramaic tāw, like the Book Pahlavi T, but rather from the other T that is attested in inscriptional Middle Persian—the one based on the Imperial Aramaic ṭēt. 

The key question, then, which remains open, is whether this second kind of T is new or old: whether it descends from the usual T of Book Pahlavi (shown in the bottom line in the table above), or whether it descends from an entirely different T (in the top line), which was otherwise thought to have gone out of use centuries before these documents were written.

Bibliography:

Dieter Weber. 2008. Berliner Pahlavi-Dokumente: Zeugnisse spätsassanidischer Brief und Rechtskultur aus frühislamischer Zeit. Wiesbaden: Harrassowitz.

———. “Kalligraphie und Kursive: Probleme der Pahlavi-Schrift in den Papyri und Ostraca.” In Medioiranica: Proceedings of the International Colloquium organized by the Katholieke Universiteit Leuven from the 21st to the 23rd of May 1990, edited by W. Skalmowski and A. van Tongerloo, 183-192. Leuven: Peeters.

P. O. Skjærvø. 1996. “Aramaic Scripts for Iranian Languages.” In The World’s Writing Systems, edited by Peter Daniels and William Bright, 515-35. New York; Oxford: Oxford University Press.