i. TERMINOLOGY AND CONVENTIONS

Human language can be reduced to writing in two basic ways, by the use of symbols which express the sounds of speech or by the use of symbols which directly express the meanings conveyed by the spoken sounds. Many writing systems combine elements of both methods. For instance, the writing system used for English and other modern European languages is almost entirely based on the phonetic principle, but a few symbols such as numerals or the ampersand “&” represent complete words or concepts without reference to the sounds of which the words are composed. Such symbols may be referred to as “ideograms” or as “logograms” (from Greek logos “word”) and their use as “ideographic” or “logographic” writing.

The use of ideograms is well established in the cuneiform scripts of the ancient Near East and will be described in the next section of this article (see below, ii.). However, ideographic writing is only a marginal feature in the Old Persian cuneiform inscriptions, which make use of a small number of ideograms for key ideological terms such as “king,” “god,” and “earth.” These ideograms do not appear to be pictographic in origin, nor do they bear any discernible relationship either to the Old Persian phonetic writings of these words or to symbols used for words of similar meaning in other languages of the ancient Near East. Apparently they are freely invented, like the Old Persian script as a whole, which imitates the structure and appearance of Mesopotamian cuneiform whilst avoiding the borrowing of individual signs.

A particular type of ideogram or logogram commonly found in the cuneiform scripts of the ancient Near East is that which is sometimes referred to as a “heterogram” (from Greek heteros “other”). The term refers to a graph borrowed from another language (in which it may have been either ideographic or phonetic), e.g., Sumerian lugal “king” used as a way of writing Akkadian šarru “id.” or Akkadian ina “in” as a way of writing Hittite anda “id.” Heterograms are also referred to by terms which identify the source language: “Sumerograms,” “Akkadograms,” etc. The ideograms used in the Middle Iranian scripts derived from Aramaic (see HUZWĀREŠ) are all heterograms, or more precisely “Aramaeograms,” i.e., alphabetically written Aramaic words used as symbols for the correponding Iranian terms. A typical example is the Aramaic preposition mn “from,” which stands for az in Middle Persian, až in Parthian, ač or čan in Sogdian.

In transliterating Iranian texts it is conventional to distinguish the ideographic elements by the use of capital letters. Thus XŠ indicates the Old Persian ideogram for “king” (corresponding to the phonetic spelling x-š-a-y-θ-i-y, i.e., xšāyaθiya), and MN indicates the Aramaeogram which is used to represent the various Middle Iranian words for “from.” Where a “phonetic complement” is attached to an ideogram, usually in order to indicate an affix or grammatical inflection, this is transliterated in lower-case letters like other phonetic elements of the script, e.g., XŠ-m or XŠ-y-m (both representing xšāya-θiyam, accusative singular).

Since the Aramaic letters aleph (ʾ) and ayin (ʿ) have no obvious upper-case equivalents, scholars have generally been content to use the same symbols for these letters in transliterating ideographic and phonetic spellings: e.g., Middle Persian ʿL = ō “to,” MLKʾ = šāh “king.” A slight ambiguity arises when an aleph occurs at the boundary between an ideogram and a phonetic complement: is MLKʾn in Middle Persian MLKʾn MLKʾ = šāhān-šāh “king of kings” to be interpreted as MLKʾ + -n or as MLK- + -ʾn? An alternative system introduced by D. N. MacKenzie in his A Concise Pahlavi Dictionary (London, 1971, pp. xii-xiii) overcomes this problem by employing A and O as the upper-case equivalents of ʾ and ʿ, thus OL, MLKAn MLKA. MacKenzie also proposed that the letter derived from Aramaic he (which in Pahlavi is only used in ideograms) should be transliterated as E, rather than H or Ḫ, thus allowing the letter ḥeth to be represented by h/H, without diacritic; Aramaic ṣ is represented by c/C. This system has been widely adopted and extended (with some adaptations) to other ideographically written Middle Iranian languages. Nevertheless, many scholars still prefer to spell the ideograms in a manner which makes their Aramaic origins more transparent, and no single system has yet been accepted as a standard.

In Sogdian script the Aramaic letter l, which represents the fricatives /δ/ and /θ/ in Sogdian words, is conventionally transliterated as δ. In ideograms, however, the same letter is transliterated as L in accordance with its Aramaic value, e.g., Lʾ “not” (= Sogdian nē and nā) from Aram. lʾ. However, usage is not consistent in this respect: Aram. ṣ is usually represented by c/C (rather than *c/Ṣ), b by β/B, and ḥ by x/Ḥ or x/X (or in older works by γ/G). Thus ʾGRZY, ʾḤRZY, ʾXRZY and AXRZY all transcribe the same ideogram (Aram. ʾḥr “afterwards” + relative particle zy, representing Sogd. rty “then; and”).