FRAHANG Ī PAHLAWĪG, “a Pahlavi dictionary,” is rather a description than the title of an anonymous glossary of some five hundred mostly Aramaic heterograms (ideograms), in the form used by Zoroastrians in writing Middle Persian (Book Pahlavi), each explained by a “phonetic” writing of the corresponding Persian word. The work was previously known to Parsis, by the traditional pronunciation of the two words of the first lemma, as “Mōnā (or Mōnō) Xudā” (see below).

A one-page fragment of a similar work containing part of a list of verbs, and probably dating from the 9th or 10th century C.E., was discovered at Turfan (see Barr). It differs from the later Frahang in listing several different forms of the ideographically written verbs, followed only by the infinitive in Persian spelling, e.g., ʿZLWN, ʿZLWNd, ʿZLWNyy, ʿZLWNyt, ʿZLWNm, etc., and finally šwtn, i.e., šudan “to go.” The Frahang, on the other hand, in the majority of manuscripts known, lists no more than three forms of verbs, each transcribed separately, e.g. ʿZLWNtnˈ - šwtnˈ, ʿZLWNytˈ -šwytˈ, ʿZLWNym - šwym (i.e., Aram. ʿzl “he went,” + generalized Aram. -wn-, for šudan, šawēd, šawēm), and a single form of other parts of speech, e.g., L - mn (i.e., a shortened Aram. ly “to me” for man “I, me”); MRʿḤ - ḥwtʾy (i.e., a corruption of Aram. mrʾh “his lord” for xwadā “lord”). These primary elements of the glossary are then transcribed interlinearly, and more or less corruptly, into Avestan letters, i.e., into Pāzand, whereby the ideograms appear in their traditional mnemonic pronunciation. Because of the ambiguity of the Pahlavi script this is often far removed from the original Aramaic spellings; e.g., the above words appear as wazrōntan - šuδan, wazrōnīṯ - šuuə̄ṯ, wazrōnīm - šuūə̄m, ra - man, mōnā - xʷudā).

Originally the Frahang was divided into chapters, some without headings, each containing words grouped according to subject. In some mss. the order of these chapters has been slightly disturbed, but it was basically as follows: (1) supernatural beings, etc., (2) gētīg “worldly things,” (3) ābhā “waters,” (4) dānag mēwagīhā “grains, fruits,” (5) xwārišn “drinking,” (6) tarragīhā “vegetables,” (7) čahārpāy “quadrupeds,” (8) murwān “birds,” (9) animals, (10) handāmīhā “parts of the body,” (11) *gōkān “details (of the family?), (12) abarīgān mardōmān “superiors,” (13) ērīgān mardōmān “inferiors,” (14) aswārīh “riding,” (15) dibīrīh “writing,” (16) ayōšust “metals,” (17) pāymār “assignments,” (18-21) verbs, (22) šnāyišn frazām “the end of praise” (? verbs of being and dying), (23) čē nibištārīh rāy “what is for written correspondence,” (24) pronouns, (25) adverbs, etc., (26) adjectives, (27) hangām ud brīn ī sālān “the division of the year,” (28) names of the days and months, (29) numerals. The 30th and last chapter is a collection mostly of variant spellings of Persian words, the more modern explaining the archaic (e.g., kwp - kwh, i.e., kōf, later kōh “mountain”), and of synonyms (e.g., ʾpdwm - pswm, i.e., abdom = pasom “last”). Some later mss. include additional lists of verbal ideograms, none of them in general use. Another recension of the Frahang came into being when the lemmata were rearranged serially according to the shape of the letters. It was in this recension that the work first became known in Europe, when in 1711 Anquetil du Perron published (Zend-Avesta II, 476-526) the version prepared for him by his Parsi teacher. His manuscript is now kept in the Bibliothèque nationale, Paris. A century later Hoshangji Jamaspji Asa and Martin Haug published an edition of the text arranged in chapters. Then Carl Salemann published two mss. (S and S2) from St. Petersburg, one of each recension. A first text-critical edition of the Frahang was made by Heinrich F. J. Junker in 1912, who used several other manuscripts from Bombay (J, M, U, etc.), the Bodleian (O, for Ouseley), and at a late stage from Copenhagen (K25, later published in facsimile by Arthur Christensen).

The system of partial heterographic writing of Pahlavi actually arose as a result of the use of Aramaic, in a script without heterograms, in the Achaemenid chanceries. As similar glossaries explaining Sumerian heterograms are known already from Akkadian times, however, an Assyriologist, Erich Ebeling, set out to prove that many of those in the Frahang are derived from Sumerian or Akkadian. This led to a number of extremely far-fetched interpretations; e.g., the explanation of the ideogram ḤLLN - myš “sheep, ewe” as gd*lln*, from Akkadian gadû lalû “young he-goat (and) female kid”; in fact it is a corruption of Aramaic rḥl “ewe.” Unfortunately, most of these impossible interpretations were incorporated into Junker’s second edition of the Frahang, as well as those by Moḥammad-Jawād Maškūr (in modern Persian) and Mario Nordio, all in orthographic order. The most recent edition of the Frahang is that published from the Nachlass of his teacher, Henrik S. Nyberg, by Bo Utas. It is useful for the identification of the Aramaic words behind the ideograms for the more obvious Persian words, though even here there are aberrations (ḤLLN “sheep,” for example, by willfully ignoring the spelling with Ḥ-, is interpreted as ʾLLN(N), explained by Syriac ʾarnā “mountain goat”). But in the elucidation of the more obscure and often corrupted entries, through a strange misunderstanding of the nature of the Frahang as containing definitions of the Aramaic words, invented proverbs, etc., it descends further and sometimes ridiculously into untrustworthiness. There are pitfalls for both user and editor of the Frahang in its present form. One is that, besides ideograms of Aramaic origin, it contains scattered examples of historical spellings of Iranian words, no longer recognized as such; e.g. lʾs (for early Mid. Pers. *rāθ) - lʾh, i.e., rāh “way”; ctl, for *čiθr, - cyhl, i.e., čihr “seed, essence.” It also, however, contains (i) a high proportion of ideograms which are not otherwise attested in Book Pahlavi texts and whose reality can therefore not be tested (e.g., ŠWMʾLH - gwlbk, i.e., gurbag “cat,” cf. Aram. šwnʾrʾ) and (ii) several pseudo-ideograms, coined by later scribes from Arabic words, e.g. ʾSDl - šyl, i.e., Ar. asad + “phonetic complement"-r = šēr, properly šagr, “lion.” Ebeling expressly rejected this interpretation, preferring the unlikely “Akk. ugudilû, from Sumerian, ‘speckled lion’ - šēr.” Nyberg resorted to fantasy: “askar (uskur) - *šikkar “porcupine"*ʾRYH, i.e., Aram. ʾaryā - šagr (šēr),” with a series of completely invented forms.


A. C. Anquetil du Perron, Le Zend-Avesta, 3 vols., Paris, 1771.

K. Barr, “Remarks on the Pahlavi Ligatures and ,” in Indian and Iranian Studies Presented to George Abraham Grierson on His Eighty-Fifth Birthday, BSO(A)S 8, 1936, pp. 391-403.

A. Christensen, ed., Cod. Avest. Pahl. XII, Copenhagen, 1944, fols. 49-74.

E. Eberling, Das aramäisch-mittelpersische Glossar Frahang-i-Pahlavik im Lichte der assyriologischen Forschung, Leipzig, 1941.

H. Jamaspji Asa and M. Haug, eds., An Old Pahlavi-Pazand Glossary, Bombay, 1870.

H. F. J. Junker, ed., The Frahang i Pahlavīk, Heidelberg, 1912; 2nd ed. as Das Frahang i Pahlavīk in zeichengemäβer Anordnung, Leipzig, 1955.

M.-J. Maškūr, Farhang-e hozvārešhā-ye pahlavī, Tehran, 1346 Š./1967.

M. Nordio, Lessico dei logogrammi Aramaici in Medio-Persiano, Venice, 1980.

H. S. Nyberg, ed., Frahang i Pahlavīk, ed. from the posthumous papers … by B. Utas, Wiesbaden, 1988 (rev. D. N. MacKenzie, OLZ 86, 1991, cols. 68-71).

C. Salemann, “Ueber eine Parsenhandschrift der Kaiserlichen Oeffentlichen Bibliothek zu St. Petersburg,” Travaux de la 3ème Session du Congrès International des Orientalistes II, Leiden, 1878, pp. 493-592.

(D. N. MacKenzie)

Cite this article: