|Languages:||Avestan language, Middle Persian|
|Time:||400 - 1000 CE|
As a side effect of its development, the script was also used for Pazend, a method of writing Middle Persian that was used primarily for the Zend commentaries on the texts of the Avesta. In the texts of Zoroastrian tradition, the alphabet is referred to as din dabireh or din dabiri, Middle Persian for "the religion's script."
The development of the Avestan alphabet was provoked by the need to correctly represent recited Avestan language texts. The various text collections that today constitute the canon of Zoroastrian scripture are the result of a collation that occurred in the 4th century, probably during the reign of Shapur II (309 - 379). It is likely that the Avestan alphabet was an ad hoc innovation related to this - "Sassanid archetype" - collation.
The enterprise, "which is indicative of a Mazdean revival and of the establishment of a strict orthodoxy closely connected with the political power, was probably caused by the desire to compete more effectively with Buddhists, Christians, and Manicheans, whose faith was based on a revealed book." In contrast, the Zoroastrian priesthood had for centuries been accustomed to memorizing scripture - following by rote the words of a teacher-priest until they had memorized the words, cadence, inflection and intonation of the prayers. This they passed on to their pupils in turn, so preserving for many generations the "correct" way to recite scripture. This was necessary because the priesthood considered (and continue to consider) precise and correct enunciation and cadence a prerequisite of effective prayer. Further, the recitation of the liturgy was (and is) accompanied by ritual activity that leaves no room to attend to a written text.
The ability to correctly render Avestan did however have a direct benefit: By the common era the Avestan language words had almost ceased to be understood, which led to the preparation of the Zend texts (from Avestan zainti "understanding"), that is commentaries on and translations of the canon. The development of the Avestan alphabet allowed these commentaries to interleave quotation of scripture with explanation thereof. The direct effect of these texts was a "standardized" interpretation of scripture that survives to the present day. For scholarship these texts are enormously interesting since they occasionally preserve passages that have otherwise been lost.
The 9th - 12th century texts of Zoroastrian tradition suggest that there was once a much larger collection of written Zoroastrian literature, but these texts - if they ever existed - have since been lost, and it is hence not known what script was used to render them. The question of the existence of a pre-Sassanid "Arsacid archetype" occupied Avestan scholars for much of the 19th century, and "[w]hatever may be the truth about the Arsacid Avesta, the linguistic evidence shows that even if it did exist, it can not have had any practical influence, since no linguistic form in the Vulgate can be explained with certainty as resulting from wrong transcription and the number of doubtful cases is minimal; in fact it is being steadily reduced. Though the existence of an Arsacid archetype is not impossible, it has proved to contribute nothing to Avestan philology."
The Pahlavi script, upon which the Avestan alphabet is based, was in common use for representing various middle Iranian languages, but was not adequate for representing a religious language that demanded precision since Pahlavi was a simplified abjad syllabary which only contained a handful of consonant characters (most with multiple pronunciations), and left most vowels unexpressed. Pahlavi script had at most 22 characters (the number varied by region and epoch), and as "Book Pahlavi", the most common form of the script, had only twelve letters representing about 24 sounds.
In contrast, Avestan was a full alphabet, with explicit characters for vowels, and allowed for phonetic disambiguation of allophones. The alphabet included many characters from cursive Pahlavi, while some are characters that only exist in the Psalter Pahlavi variant (in cursive Pahlavi have the same symbol). Some of the vowels, such as appear to derive from Greek minuscules. Avestan o is a special form of Pahlavil that exists only in Aramaic ideograms. Some letters (e.g.), are free inventions.
Avestan script, like Pahlavi script and Aramaic script also, is written from right to left. In Avestan script, letters are not connected, and ligatures (the "standard" ones being) are "rare and clearly of secondary origin." Fossey lists altogether 16 ligatures, but most are formed by the interaction of swash tails.
Words and the end of the first part of a compound are separated by a dot (point). Beyond that, punctuation is weak or non-existent in the manuscripts, and in the 1880s Karl Friedrich Geldner had to devise one for standardized transcription. In his system, which he developed based on what he could find, a triangle of three dots serves as a colon, a semicolon, an end of sentence or end of section; which is which is determined by the size of the dots and whether there is one dot above and two below, or two above and one below. Two above and one below signify - in ascending order of 'dot' size - colon, semicolon, end of sentence or end of section. One above and two below signify 'turned end of sentence' and 'turned end of section'.
In total, the Avestan alphabet has 37 consonants and 16 vowels. There are two main transcription schemes for Avestan, the older style used by Christian Bartholomae, and the newer style used by Karl Hoffmann.
The following list shows the letters as ordered and transcribed by Hoffmann (1996), based on Bartholomae:
Not represented in the above table are the semi-vocalic glides ii and uu, which in the Bartholomae system are transcribed as y and w. Later, when writing Middle Persian in the script (i.e. Pazend), another consonant was added to it to represent the[l] phoneme that didn't exist in the Avestan languages.
The script has been proposed to be encoded in the Unicode Standard by Michael Everson and Roozbeh Pournader, and was accepted by the Unicode Technical Committee on 2 September 2007, to be included in Unicode 5.2.
The accepted Unicode range is U+10B00 through U+10B35 for letters (ii and uu are not represented as single characters) and U+10B38 through U+10B3F for punctuation. Two of the 64 spaces are not used.