HUZWĀREŠ

a term describing the use of Semitic word masks in Middle Persian texts, written in the official orthography of the Sasanian state and surviving in Zoroastrian texts, and a small number of inscriptions, and letters.

 

HUZWĀREŠ, a term describing the use of Semitic word masks in Middle Persian texts, written in the official orthography of the Sasanian state and surviving in Zoroastrian texts (in manuscripts) and a small number of inscriptions (on rocks) and letters (on papyrus). The only other surviving text written in this way is the Christian Pahlavi Psalter fragment from Turfan. Huzwāreš combines with the archaizing orthography and the insufficient differentiation of many letters particularly in (Book) Pahlavi script to make this script very difficult to decipher.

Origin of the term. Various fanciful etymologies of this term have been proposed, mainly out of misconceptions about the form and the age of the word. Huzwāreš is a Pāzand form corresponding to Middle Persian huzwā-rišn, or rather uzwārišn, (the “h” is not etymological), the verbal noun of the verb uzwār- “to disclose, to comprehend” (see MacKenzie, 1971, p. 85; in Manichean Mid. Pers. izwār-, the quality of the initial vowel being dependent on the consonants following). The term describes the process of reading as “comprehension” (see Tavadia, p. 18; Sundermann, 1984, p. 496, n. 36) and is modeled on the Aramaic verb prš (praš “to understand” and parreš “to explain, to interpret,” the latter used in Ezra 4:18 to describe the reading of a letter to the Persian king; see Schaeder and Polotsky). A more precise definition of the term, that is, whether it refers to the entire Pahlavi script (as seems to be the case in Ebn al-Nadim, see below) or just to the heterograms (or their reading), cannot be gained from the word itself and must remain unresolved.

The nature of huzwāreš. A normal Pahlavi text contains a large number of Aramaic heterograms (The terms “Aramaeogram,” “ideogram,” and “logogram” are also frequently used; see IDEOGRAPHIC WRITING i.) for all the word categories of the language side by side with “phonetically” written words, e.g., both DYNʾ and dʾtstʾnˈ (i.e., dādestān “judgement,” q.v.). Verbal and nominal inflectional endings can be attached to the heterograms (e.g., YNSBWNyt with 3rd sing. pres. -yt = /stānēd/ “he/she/it takes”; GBRʾn with plural ending -ʾn = /mar-dān/ “men”), as can letters giving an indication of the reading to be used (rˈ> = /tar/ “across, through,” though in this case the phonetic element has no apparent function, since * does not occur). These phonetic supplements can be called complements. In compound words a mixture of heterogram and “phonetically” written part can occur (e.g., mgwGBRʾ = /moγ-mard/ “magus”).

The present convention distinguishes the heterograms from the “phonetic” elements by transliterating the heterograms in capitals and the rest in lower case letters, e.g., MLKʾn MLKʾ (= /šāhān šāh/ “king of kings”). Some editors also separate the complement from the heterogram with a hyphen. In a transcription no indication is given as to the way a word is written, though it might be useful to continue the method used, for example, by Marijan Molé, which marked heterograms with a preceding ʾ, e.g., ʾu-š guft Purušāsp ʾkū for <÷Pæ gwptˈ pwrwšʾspˈ ʾYK>.

The use of heterograms is not confined to Sasanian Middle Persian. It is also found in Parthian texts (in the entire archive from Nisa, on coins, and in the Parthian versions of the inscriptions of the first Sasanians) and to a more limited extent in Chorasmian (Henning, 1965) and in Sogdian.

Controversy over the linguistic reality. The use of Semitic word masking the writing of Middle Persian is attested by Ebn al-Nadim; he quotes the 8th-century author Ebn al-Moqaffaʿ, who explains the term “zwārašn” (i.e., zwʾršn for uzwārišn; see below) with the following examples: for /gōšt/ “meat” one writes BSRʾ (actually BSLYʾ), for /nān/ “bread” (Ebn al-Nadim, ed. Tajaddod, p. 17, tr. Dodge, pp. 26; see Schaeder, pp. 206-8). This is not entirely dissimilar to the use of Sumerian heterograms in Akkadian texts, but there has been a long controversy on the linguistic reality behind the practice, that is, whether the words so written were just the graphic representation of Middle Persian words or were actually loanwords in Middle Persian.

Nearly all the editions of Middle Persian texts made in the 19th century followed the Parsi tradition that saw the heterograms as representing “real” words and transcribed them accordingly. This made Middle Persian a mixed language rather like New Persian with its large number of Arabic loanwords and resulted in, for example, text titles like šāyast lā šāyast (West, tr., p. 239) instead of the present reading šāyast nē šāyast. Though not the first one to recognize the heterograms as such, Theodore Nöldeke played an important part in (re-)establishing this fact. In 1879 he translated Kār-nāmagī Ardašir and gave non-heterographic readings in the annotations, but he did not present a complete reading of the text. In 1887 he described the heterographic system. The first non-heterographic reading of a Pahlavi text was done by Carl Salemann in 1887 (Salemann, in Grundriss I, p. 252, sec. 8, n. 1). Salemann (Grundiss I, p. 250) pointed to the lack of Aramaic loanwords in New Persian and in the Armenian, Syrian, and Greek sources dependent on Middle Persian. There is also internal evidence for the heterographic character of the Aramaic elements in parallel passages that vary between heterographic and “phonetic” writing (Humbach, 1973, p. 121).

At the start of the 20th century new discoveries gave clear support to the view that the heterograms were indeed part of the writing system rather than the language. The Middle Persian and Parthian texts in Manichean script found in Central Asia contain no heterograms whatsoever, but obviously represent the same languages as the ones used in the Sasanian bilingual inscriptions and Pahlavi texts. Similarly in the case of Sogdian, heterograms are restricted to texts written in the traditional Sogdian script (apart from the occasional use of ZY “and” in Sogdian texts written in Syriac script, where its shape was assimilated to Syriac ʿy or yz; see Schwartz, 1991, pp. 162-63). A particularly telling demonstration of the purely graphic character of the heterograms is provided by the Sogdian “Tale of the Pearl-borer” (Henning, 1945, pp. 465-69), of which two copies exist, one in Sogdian script, with the usual sprinkling of heterograms, and one in Manichean script without any. Nevertheless, as late as the 1970s, Wolfgang Lentz reverted to the idea that the heterograms were indeed loanwords, at least in the written language of the late and post-Sasanian periods (Lentz, 1975; idem, 1977). This is, however, hardly a tenable proposition.

Bo Utas (1984, p. 59) refers also to “the artificial shape of many of these Aramaic pseudo-words,” though he warns that a number of Aramaic words may indeed be loanwords. However, the fact that the heterograms were word masks does not preclude that at some stage they were read with the values of their individual letters. Shaul Shaked (p. 76) points out that the interchange of letters such as d/t or b/p in cases such as GLLTʾ for Aramaic gldʾ and KWKPʾ beside KWKBʾ for Aramaic kwkbʾ, which follow the orthographic patterns found in the case of non-heterographic spellings, indicates “a mode of transmissioŋthat must have been partly oral.” There was a reliable traditional reading of many heterograms, on the basis of which (for example) Martin Haug was able to read yehevûnd for the highly ambiguous YHWWNt. The New Persian term Gabr (Zoroastrian) may have arisen “as a contemptuous term for the people who wrote ‘GBRʾ’ instead of ‘mard’” (Sims-Williams, personal communication; see GABR for other views), in which case it demonstrates a correct reading of the heterogram involved. However, none of this indicates that the scribes knew Aramaic. The need to reproduce the text accurately via oral dictation often led to the use of an artificial pronunciation which was easy to convert into the correct written form. An example is the reading bwmn for BRḪ “son” taught to Anquetil-Duperron (q.v.; Skjærvø, 1998, p. 315). Similar artificial pronunciations were also applied in the case of difficult, non-heterographic spellings, e.g., ʾnhwmʾ for ʾwhrmzd (Lentz, 1977, pp. 1074-75).

Origin of the practice. The precondition for the origin of the practice is the use of Aramaic in the Achaemenid chanceries (discussed by, among others, Schaeder, Polotsky, and Henning; see also Sundermann, 1985, pp. 104-6). Bilingual scribes in the Achaemenid period simultaneously translated and wrote down in Aramaic a text spoken in a local Iranian language, restricting the Iranian words in the text to personal names, technical words for which they had no Aramaic equivalent, and Iranian loanwords common in Aramaic. Their intention was still to write Aramaic, since this allowed them to communicate with scribes bilingual in Aramaic and a second language different from their own. In the post-Achaemenid period, with the collapse of the central authority, this need faded; and, although Aramaic remained the language of writing, the scribes introduced more and more Iranian words into their Aramaic. At what stage the scribes became entirely ignorant of Aramaic is unclear, but certainly in the Sasanian period the Aramaic heterograms were in the main opaque, though a certain amount of systematizing made verbs (characterized by the ending -WN) and certain nouns (those ending in -Ḫ) recognizable as such. This becomes clear when the heterograms used in the Middle Persian inscriptions and in Pahlavi texts are compared with those in the Parthian inscriptions: Mid. Pers. YMLLWN (and YMRRWN) versus Parth. YMLL and YMLLW for the verb guftan “to speak”; Mid. Pers. ʿBYDWN versus Parth. ʿBD and ʿBDW for the verb kardan “to do” (see Skjærvø, 1986, pp. 431-38; idem, 1998, pp. 298 ff.); and Mid. Pers. BRḪ versus Parth. BRY “son,” Mid. Pers. YDḪ versus Parth. YDʾ “hand.” In fact the heterograms say nothing about the knowledge of Aramaic that a particular scribe may have had at any time. Once the transition from writing Aramaic to writing Iranian had been made, there was no need for an Aramaic-speaking scribe. He just had to be trained to use the complex system. There is therefore no reason to believe that Sasanian scribes had any knowledge of Aramaic whatever, since they were not employed to work bilingually. Still unresolved is the question as to how suddenly the change from Aramaic to Iranian was made. Walter Bruno Henning (1958, pp. 30-35) assumed it was gradual. Bo Utas (1984, p. 64) thinks it must have been the result of a deliberate decision, “an invention, made at one time and in one place.”

Date. The practice is generally seen as arising some time in the Arsacid period. Franz Altheim and Ruth Stiehl (pp. 64, 287-88, 304) contended that the use of heterograms did not arise until the Sasanian era or very shortly before, and thus they regarded all previous texts as Aramaic. This theory does not take account of the separate development of the Parthian and Middle Persian heterograms found in the bilingual Sasanian inscriptions, which must be the result of a certain period of development (generally from Parthian to Middle Persian and not the other way round, as Altheim and Stiehl, p. 289 would have it).

Henning (1958, p. 25) pointed to heterograms ending in -Ḫ, e.g., in BRḪ, which is used simply as a mask for the term “son” (Parth. puθr > puhr, Mid. Pers. pus), though the form writes Aram. breh “his son.” If the text is Aramaic, then a possessive is intended, which means the Aramaic genitive Particle zy/dy and a name should follow. Since this is often not the case in coin legends from the late 2nd century B.C.E., a date is thus established for heterographic use. Helmut Humbach, in his treatment of Aramaeo-Iranian and Aramaeo-Indian inscriptions dating from the reign of Aśoka (q.v.; in the first half of the 3rd century B.C.E.), wants to date the process even earlier, that is, to the middle of the 3rd century B.C.E. (Humbach, 1972, with reference to his work of 1969). Eduard Y. Kutscher (p. 398) agrees with an even earlier date, namely the 5th century B.C.E., proposed by Hans Heinrich Schaeder.

Many of the arguments about the date of the heterograms center in particular on the form of the verbal heterograms. Some are regarded as old in the sense of Imperial Aramaic, some as younger Eastern Aramaic forms either introduced into an already existing heterographic practice or indeed dating from the very start of the practice to the late period, when Eastern Aramaic is otherwise attested. Some forms are demonstrably fabrications of the scribes ( has its m- from a participle but -n from the 3rd pers. pl.), and Christopher Toll has suggested that all the inflected heterograms, even the realistic forms, are the result of systematization. A combination of both views is justified, which means that, after the scribes’ loss of an active knowledge of Aramaic, a process began which systematized the verbal heterograms in particular and made them easier to recognize. There are also indications that the scribes developed subsystems of verbal heterograms marked for tense structure and used these forms consistently at least in the inscriptions (Utas, 1984, p. 60; Skjærvø, 1986).

Lists of heterograms. Since Sasanian scribes had generally no knowledge of Aramaic, there was an obvious need for a teaching tradition in the heterograms used. The late Frahang ī Pahlawīg (q.v.; ca. 1600? latest edition by Nyberg, 1988) must go back in some form to this tradition, and the Frahang fragment from Turfan (see Geldner and Barr), containing only seven verbs in as many as fourteen different forms, proves the tradition’s existence for roughly the late Sasanian period. The Frahang ī Pahlawīg has been augmented and corrupted over the centuries. It contains about 1,000 entries, about 630 of which are heterograms. D. N. MacKenzie (1991, p. 69), notes that “only about half of the ideograms listed in the Frahang actually occur in surviving texts, be it inscriptions (ca. 170), Psalter (ca. 120) or Zoroastrian literature (ca. 200).” Some are “pseudoideograms, nowhere actually used, which are clearly based on Arabic words.” Shaked rejects the last claim and points out that many heterograms otherwise not attested may still be genuine, especially if found in Aramaic literature. “It does not seem plausible that recent scribes could or would invent Aramaic forms at the time in which the book was compiled. The store of ideograms listed here . . . must have largely come down in school traditions from an earlier date, when they were still in use” (Shaked, p. 75). Given the known loss of Pahlavi texts, the Frahang may be the only trace of a heterogram, but obviously only a list compiled from the extant Pahlavi texts can show the existence of a heterogram and give an example of its use. Moḥammad-Jawād Maškūr has made the first step in this direction. The works of Abramyan, Philippe Gignoux, D. N. MacKenzie, and Henrik Samuel Nyberg and the glossaries to text editions contain information on whether a word is written heterographically or “phonetically,” but a comprehensive dictionary of Pahlavi with an orthographic history of each lemma is still a long way off.

 

Bibliography:

R. Abramyan, Pehleviĭsko-persidsko-armyano-russko-angliĭskiĭ slovar’, Erevan, 1965.

Franz Altheim and Ruth Stiehl, Die aramäische Sprache unter den Achämeniden I: Geschichtliche Untersuchungen, Frankfurt am Main, 1963.

Kaj Barr, “Remarks on the Pahlavi Ligatures *** and ***,” in Indian and Iranian Studies Presented to George Abraham Grierson . . . , BSO(A)S 8/2-3, 1936, pp. 391-403.

James Darmesteter, Étude Iranienes, 2 vols., Paris, 1883, I, pp. 35-36.

Ebn al-Nadim, ed. Tajaddod, pp. 15-17; tr. Dodge, pp. 24-26.

Karl Friedrick Geldner, “Bruchstück eines Pehlevi-Glossars aus Turfan, Chinesisch-Turkestan,” SPAW 2, 1904, pp. 1136-37.

Philippe Gignoux, Glossaire des Inscriptions Pehlevies et Parthes, Corpus Inscr. Iran, Suppl. Ser. 1, London, 1972. Walter Bruno Henning, “Sogdian Tales,” BSO(A)S 11, 1945, pp. 465-87.

Idem, “Mitteliranisch,” in HO I, IV, Iranistik 1, London and Cologne, 1958, pp. 20-130.

Idem, “The Choresmian Documents,” Asia Major 11, 1965, 166-79.

Helmut Humbach, Die aramäische Inschrift von Taxila, Abhandlungen der Akademie der Wissenschaften und der Literatur, Geistes- und sozialwissenschaftliche Klasse 1, Wiesbaden, 1969.

Idem, “Additional Notes on the Aramaic Inscription of Taxila,” MSS 26, 1969, pp. 39-42.

Idem, “Late Imperial Aramaic and Pahlavi,” MSS 30, 1972, pp. 47-50.

Idem, “Beobachtungen zur öberlieferungsgeschichte des Awesta,” MSS 31, 1973, pp. 109-22.

Eduard Yechekzel Kutscher, “Aramaic,” in Thomas A. Sebeok, ed., Current Trends in Linguistics VI, 1970, pp. 347-412, esp. pp. 393-99 (“The A[ramaic] Logograms of Middle Persian”).

Wolfgang Lentz, “Note on Some Terms Used in Connection with the Aramaic Elements in Middle Persian,” in Acta Iranica 6, Liège, 1975, pp. 313-16.

Idem, “Mitteliranische ‘Ideographie’ im Lichte von Erfahrungen mit Sprachkontakten,” in Wolfgang Voigt, ed., 19 Deutscher Orientalistentag vom 28. September bis 4. Oktober 1975 in Freiburg im Breisgau, ZDMG Suppl. 3/2, Wiesbaden 1977, pp. 1061-83.

D. N. MacKenzie, “Notes on the Transcription of Pahlavi,” BSO(A)S 30, 1967, pp. 17-29.

Idem, A Concise Pahlavi Dictionary, London, 1971.

Moḥammad-Jawād Maškur, Farhang-e hozvārešhā-ye Pahlavi, Tehran, 1346 Š./1967.

Marijan Molé, La legende de Zoroastre selon les textes pehlevis, Paris, 1967.

Theodore Nöldeke, “Geschichte des Artaschir-i Pâpakân aus dem Pahlevi übersetzt, mit Erläuterungen und einer Einleitung versehen,” Bezzenbergers Beiträge zur Kunde der indogermanischen Sprachen 4, 1879 (Festschrift for Theodore Benfey), pp. 22-69.

Idem, Aufsätze zur persichen Geschichte, Leipzig, 1887, repr. Graz, 1974, pp. 150-58.

Henrik Samuel Nyberg, A Manual of Pahlavi II: Ideograms, Glossary . . . , Wiesbaden, 1974.

Idem, Frahang ī Pahlavīk, ed. with transliteration, transcription, and commentary from Nyberg’s posthumous papers by Bo Utas with the collaboration of Christopher Toll, Wiesbaden, 1988; reviewed by Prods Oktor Skjærvø in Kratylos 35, 1990, pp. 95-99, and by D. N. MacKenzie, Orientalistische Literaturzeitung 86, 1991, pp. 68-71.

Hans Jacob Polotsky, “Aramäisch prš und das ‘Huzvaresch’,” Le Muséon 45, 1932, pp. 273-83.

Carl Salemann, “Mittelpersische Studien,” Bulletin de l’Académie Impériale des Sciences de St. Pétersbourg (BASP) 31, 1887, pp. 417-50; also publ. in Melanges asiatiques tirés du BASP 9, 1887, pp. 222-42.

Idem, “Mittelpersisch,” in Grundriss I, pp. 249-332.

Hans Heinrich Schaeder, “Iranische Beiträge I,” Schriften derKönigsberger Gelehrten Gesellschaft, Geisteswissenschaftliche Klasse 6, Heft 5, Halle, 1930, pp. 199-296.

Martin Schwartz, “A Page of a Sogdian Liber Vitae,” in Ronald E. Emmerick and Dieter Weber, eds., Corolla Iranica, Papers in honour of Prof. Dr. David Neil MacKenzie, Frankfurt, 1991, pp. 157-66.

Shaul Shaked, “A Dictionary of Aramaic Ideograms in Pahlavi,” JAOS 113, 1993, pp. 75-81.

Prods Oktor Skjærvø, “Verbs in Parthian and Middle Persian Inscriptions,” in Rüdiger Schmitt and Prods Oktor Skœrvø, eds., Studia Grammatica Iranica: Festschrift für Helmut Humbach, MSS, Beiheft 13, Munich, 1986, pp. 425-39.

Idem, “Verbal Ideograms and the Imperfect in Middle Persian and Parthian,” in Charles Henri de Fouchécour and Philippe Gignoux, eds., Etudes irano-aryennes offertes à Gilbert Lazard, Studia Iranica, Cahier 7, Paris, 1989, pp. 333-54.

Idem, “Aramaic in Iran,” Aram 7, 1995 [1998], pp. 283-318.

Werner Sundermann, “Ein weiteres Fragment aus Manis Gigantenbuch,” in Orientalia Jacques Duchesne-Guillemin emerito oblata, Acta Iranica 23, Leiden, 1984, pp. 491-505.

Idem, “Schriftsysteme und Alphabete im alten Iran,” AoF 12, 1985, pp. 101-13.

Jahangir C. Tavadia, Die mittelpersische Sprache und Literature der Zarathustrier, Iranische Texte und Hilfsbücher 2, Leipzig, 1956.

Christopher Toll, “Die aramäischen Ideogramme im Mittelpersischen,” in Werner Diem and Abdoljavad Falaturi, eds., 24Deutscher Orientalistentag vom 26 bis 30 September 1988 in Köln, ZDMG Suppl. 8, Stuttgart, 1990, pp. 25-45.

Bo Utas, “Verbal Forms and Ideograms in the Middle Persian Inscriptions,” Acta Orientalia 36, 1974, pp. 83-112.

Idem, “Verbal Ideograms in the Frahang ī Pahlavīk,” in Wojcieh Skalmowski and Alois van Tongerloo, eds., Middle Iranian Studies: Proceedings of the International Symposium by the Katholieke Universiteit Leuven, Leuven, 1984, pp. 57-67.

Edward William West, tr., Pahlavi Texts I, SBE 5, Delhi, 1970, pp. xiv-xv.

(D. Durkin-Meisterernst)

Originally Published: December 15, 2004

Last Updated: March 23, 2012

This article is available in print.
Vol. XII, Fasc. 6, pp. 585-588