The only complete syntax of Avestan which is still usable today is H. Reichelt’s Awestisches Elementarbuch (Heidelberg 1909, pp. 218-387, indicated by “R” in the references). Each chapter features a list of older manuals and other related studies. This outdated and insufficient work has been only partially replaced. J. Kellens offers a systematic study of the verbal syntax in Young Avestan (hereafter Y. Av.) in Le Verbe avestique (Wiesbaden; 1984 “VA” in references, condensed in Compendium linguarum iranicarum, pp. 50-53). J. Kellens and E. Pirart have analyzed some aspects of Old Avestan (hereafter O. Av.) syntax: the use of cases (not including nominative and accusative), construction of the relative clause, use of modes and tenses, and particles (Textes vieil-avestiques II, Wiesbaden, 1990, pp. 1-189; “TVA II” in the references). The questions of syntax which were revisited after Reichelt was published are described in three Avestan “Forschungsberichte” in Kratylos, the first written by J. Duchesne-Guillemin (7, 1962, pp. 3-36), the following two by Kellens (16, 1971 [1973], pp. 16-17 and 36, 1991, pp. 22-25). Many works on Indian syntax (referenced below), starting with B. Delbrück’s classic Altindische Syntax (Halle, 1888), can be transposed without much difficulty to Avestan. From a comparatist perspective, J. Haudry’s work (Emploi des cas en védique, Lyon, 1977) contains some very interesting remarks on Avestan.

Number (R, pp. 220-22). Avestan has three numbers: singular, dual, and plural. The dual becomes progressively less and less used and loses ground to the plural. Elliptical uses of number are quite rare and often uncertain. There is one confirmed example of the elliptical dual in O. Av.: Y. 33.8 təuuīšī, which presupposes the *utaiiūitī member of the traditional dvandva. The plural of š´iiaoθana-applies to the entire manah-/vacah-/ š´iiaoθana-triad (Kellens in Corolla Iranica, pp. 101-108). On the other hand, there is positively no ellipsis of other subjects in Y. 47.1 ahmāi dąn... mazd å,because dąn and mazd ådo not necessarily belong to the same syntagm. There is a definite tendency toward number concordance (Y. 44.9, etc. mōi daēnąmdaēnā” vs. Y. 34.13 daēn “my åsaošiiaṇtąm “the daēnās of the saošiiaṇts,” Kellens, loc. cit.).

Use of Cases. The nominative (R, pp. 224-26) is the case of the subject.The accusative (R, pp. 227-32) is the case of the object and the case of spatial or temporal limit. Although in Y. Av., it can be governed by an agent noun and even an action noun, in O. Av. it is used adnominally only after a superlative, which is a substitute for a verbal form (Y. 51.1 bāgəm aibī.bairištəm “very much furnishing a share”) or after the agent adjective in -ra- (Y. 46.5 yəˊ aṧauuā drəguuaṇtəm vīcirō “knowing [how to] tell the difference between the drəguuaṇt and the aṧauuan”).The instrumental case (R, pp. 232-39) has been the object of broad discussion, in which theological concerns also are present. The fact (identified as early as 1892 by W. Caland and K. F. Geldner in KZ 31, pp. 256 and 319 respectively) that the instrumental of the name of the Gathic entities is abnormally frequent appeared to be significant and deserved a comparatist explanation. First, H. Pedersen (KZ 40, 1907, pp. 134 ff.), and then E. Schwyzer (IF 47, 1929, pp. 214 ff.) considered the instrumental as a possible substitute for the nominative, accusative, and vocative in the rection of transitive verbs. The skepticism that this hypothesis provoked (already in H. Humbach, Gathas I, p. 68), was justified in Kellens-Pirart (TVA II, pp. 3-20). They rejected the substitutive functions as well as the adnominal use and closely held to the uses listed by Delbrück for Indian: the instrumental of the personal noun is sociative, indicating accompaniment with the subject. The instrumental of the concrete or abstract noun expresses means (including the internal means: Y. 31.8 məˊŋˊhī manaŋhā “I think with the thought”), reason for the action, place of passage, and duration (Y. 51.16 vīštāspō... nąsaṱ vanhəˊuš padəˊbīš manaŋhō yąm cistīm “the idea to which Vīštāspa came by way of the good Thought”), expected gain (Y. 50.7 yaojā... auruuatō jaiiāiš “I want to harness my steeds out of self-interest in victories”).

The dative (R, pp. 239-45; TVA II, pp. 20-30). The most represented categories are, for personal nouns, the commodi et incommodi dative (Y. 33-12 us mōi uzārəšuuā “stand up for me”); for concrete and abstract nouns, the dative of purpose (Y. 34.11 aṱ tōi ubē hauruu åscāxᵛarəθāi.ā amərətatåscā “both health and immortality are to feed you”), which is particularly frequent in nominal sentences; and the expression of time to come (yauuōivīspāi “for eternity”). There is no certain example of the ethical dative in O. Av., and there is no evidence for the double dative (H. Hettrich, MSS 43, 1984, pp. 80 ff.). Sometimes the logical subject of a primary transitive verb appears in the dative in the causal construction: Y. 41.1 stūtō... mazdāi.. ācā vaēdaiiamahī, “we make the praises known to Mazdā” (E. Tichy, Sprache 26, 1980, pp. 14 ff). The dative is frequent in adjectival rection, e.g., the degrees of vaŋhu- + dative which produce “good for.”

The genitive. According to Reichelt (pp. 251-61), the genitive’s main adnominal uses are the partitive genitive (sometimes substituted for the subject or the object of the verb), subjective genitive (Y. 50.8 arədraxˊiiācā nəmaŋhā “by the homage of the competent-sacrificer”), objective genitive (Y. 57.3 yasna yazatanąm “by the sacrifice to the gods-worthy-of-the-sacrifice”), possessive genitive, genitive of material (Haδōxt Nask [H.] 1.5 gəˊuš xᵛarəitinąm “meals of beef (meat)”), genitive of quality (Yt. 13.2 aiiaŋhō kəhrpa xᵛaēnahe “in the form of a metal beautiful to behold”), and genitive of time (θβarštahe zrū; see P. Tedesco, Language 36, 1960, p. 134). Kellens-Pirart (TVA II, pp. 30-36) are mainly concerned with creating a list of the possible positions in relation to the determinant.

The ablative (R, pp. 245-51; TVA II, pp. 36-39). This is mostly found in nominal or adjectival rection (e.g., as complement of the comparative: Vispered [Vr.] 7.3 vātō... hubaoδitarō aniiaēibiiō vātaēibiiō “a wind more fragrant than the other winds”) and is quite frequent as the complement of the verb. It expresses substituted object (Y. 32.12 aṧāṱ varatā... drujəm “he preferred the Druj over Aṧa”), cause (Y. 51.12 aodərəšcā zōišənū vāzā “two draft animals shivering because of the cold”) and, with the ā postposition, place of direction with the nuance of “(all the way) to” (Y. 30.10 aṱ asištā yaojaṇtē ā hušitōiš vaŋhəˊuš manaŋhō “two steeds will be harnessed to ride to the good residence of the good Thought”). This use is the origin of the -δa forms in Y. Av. and of turns of phrase such as Yt. 14.31 varəsəm zamāδa saiianəm “a hair fallen to the ground” (Finck, KZ 40, 1907, p. 123). The adverbial phrase of place of origin requires in Y. Av. the preposition haca, but the ablative governed by it in O. Av. has an unclear meaning, including the frequent aṧāṱhacā. A double ablative occurs in O. Av. with the preposition hanarə “without” (Y. 31.15 hanarə.. aēnaŋhō pasəˊuš vīrāaṱcā “without hurting the cattle and the people”).

The locative (R, pp. 261-66; TVA II, pp. 39-44). The most frequent functions are the expression of location (aiβi gāme “in winter,” Y. 32.3 būmiiå haptaiθē “on the seventh part of the earth,” or Y. 43.2 xᵛāθrōiiā “in wellbeing” with the ā postposition, which in Y. Av. tends to merge with the ending--hence Vd. 3.7 plur. dātāhuua “between the teeth” or sing. zaṇtauua “in the territory”), point of view (Y. 46.13 yəˊ zaraθuštrəm marətaēšū xšnāuš “the one who pleased Zarathushtra in the presence of men”), emotion (Y. 33.2 tōi vārāi rādəntī ahurahiiā zaošē mazdå “they will meet the demands of Ahura Mazdā, with his great approval”), exclamation (uštā “According to wish!” “Happiness!”).

The vocative (R, 226-27; TVA II, pp. 44-52). How can one associate another person with the one who is being addressed in the vocative? Neither dvandva nor asyndeton nor coordination of vocatives is allowed. E. Risch (MSS 17, 1964, pp. 51 ff.) and R. Zwolanek (Vāˊyav índraś ca, München, 1970) confirm conclusively Caland’s thesis (KZ 30, 1890, pp. 542 ff.), which states that the noun coordinated with the vocative is in the case which is required by the syntactical logic: e.g., Y. 29.8 nəˊ mazdā... aṧāicā “for us, O Mazdā and Aṧa.” This type of invocation has been re-examined recently: J. S. Klein (MSS 40, 1981, pp. 13 ff.) interprets it as a way of invoking complementary divinities; and, according to S. W. Jamison (MSS 49, 1989, pp. 13 ff.), these divinities’ name normally forms a dvandva. It is quite obvious now that this type is represented, in one way or another, by Y. 30.9 mazdåscā ahuråŋhō (Humbach, MSS 41, 1982, pp. 95 ff.; J. Narten, Aməṧa Spəṇtas, pp. 55 ff.; and Kellens-Pirart, TVA III, pp. 51 ff., despite Gershevitch, in Studia Grammatica Iranica, pp. 83 ff.). Risch (op. cit., pp. 57 ff.) and Zwolanek (op. cit., pp. 30-34), who hold that the vocative may be extended by one or several nouns in the instrumental, are criticized by Kellens-Pirart (TVA II, pp. 50 ff.).

The agent function. According to Jamison (Sprache 25, 1979, pp. 129 ff.), the case for this function used to be exclusively the instrumental, but she provides few examples in Avestan. On the other hand, the materials compiled by Kellens-Pirart (TVA II, pp. 30 ff.) show that the agent use of the verbal in -ta- is exclusively in the genitive in O. Av. (Y. 48.2 aŋhəˊuš... vistā ākərəitiš, “the norm discovered by the state [of thought]”). On the possibility of a dative as agent of the negative verbal in -ta- (Y. 31.1 aguštā... aēibiiō “not destined to be heard by them”), see Jamison (op. cit., p. 139) and Kellens-Pirart (TVA II, p. 59).

Syncretism of cases. Reichelt indicates, in each section, the situations (fairly frequent in Y. Av.) where one case is substituted for another. It would be risky to come forward with any definite statements on this matter before an exhaustive study is available; the late aberrations and errors in transmission need to be distinguished first from facts caused by the natural evolution of the language. The list of the prepositions and the rules governing them are described by Reichelt (pp. 266-79). Regarding pronouns (R, pp. 279-96), see K. Hoffmann (EIr. III, pp. 58-59).

Syntax of the verb (R, pp. 296-349, Kellens, VA, passim, and TVA II, pp. 65-98). Avestan has two voices (active and middle; see Hoffmann, EIr. III, p. 59), three tenses (present, aorist, perfect), and five personal modes (indicative, injunctive, subjunctive, optative, imperative). The opposition between the three tenses is present only in O. Av., since in Y. Av. all the verbal functions were organized within the system of the present tense.

The present indicative. This is the mode of the actual present, regardless of the clause it appears in. While the duration of the process in question does not matter in Y. Av., it seems that in O. Av. only verbs with a stative meaning produce a present indicative. There is no evidence of a historical present function.

The present injunctive. In Y. Av., this expresses the actual past, regardless of the duration of the process in question and the clause it appears in. Thus Y. Av. is notable, in comparison with other ancient Indo-Iranian dialects, in that it expresses the past using a verbal form without augment. The use of the present injunctive with the particle to express the inhibitive prohibition takes place in O. Av. as well as in Y. Av. (H. 2.17 mā dim pərəsō yim pərəsahi “stop interrogating the one whom you are interrogating”). In O. Av., the present injunctive is totally uninvolved in the expression of the past: it seems that it is exclusively produced by momentative verbs, and hence it expresses an instantaneous action which recurs a certain number of times in an indeterminate period, which necessarily includes the present. The only imperfect found in O. Av. is “to be” (Y. 39.9 as = ās < (a@ ā+) ās), and the few forms in Y. Av. seem to be used in different ways. If we disregard the ah “to be” and i “to go” forms, which do not produce an injunctive, some forms are a dialectal peculiarity of the passages in which they appear (Y. 19-21, Vd. 18-19), while others seem to have a real syntactical function. They are in opposition to the present injunctive, in relation to which they express anteriority: Yt. 5.68 tąm yazata jāmāspō yaṱ spāδəmpairi.auuaēnaṱ “Jāmāspa made a sacrifice to her when he had seen the army.”

The present subjunctive. When this appears in the independent clause or in the main clause, its syntactical function varies according to the person: the first person form has a volitive value, expressing the intention of the one who is speaking about performing the action (Y. 14.1 vīsāi vəˊ... staotā “I want to serve you as the one who praises you”); the second person form has, in practice, the value of an imperative (Y. 71.16 frapāraiiåŋˊhēuruuānəm tarō cinuuatō pərətūm “make your soul go over the bridge of the Cinvaṇt”); the third person form has a consecutive value: it indicates that the completion of the action depends on the previous fulfillment of another action (Yt. 12.3-4 imaṱ uxδəm vacō framruiiå ... daēuuaiiāzō zafarə auua.gəuruuaiiąn “recite this word, (so that) the worshippers of demons shut their mouths”). The present subjunctive is used in interrogatives, when interrogation applies to a circumstance which has not started to happen yet (Vd. 5.10 kuθa tē vərəziiąn aēte yōi mazdaiiasna “how do you think the Mazdeans should act?”). In a relative clause, the present subjunctive expresses either a process which will be carried out only if the action of the main clause is completed (Y. 65.2 dāiiå mē... frazaiṇtīm yā mē frāδaiiāṱ nmānəmca ... “may you grant me a progeny that will bring greatness to my house”), or it expresses the hypothetical present. In the latter case, the principal action is either prescriptive or interrogative (Y. 65.7 māδa nō apō... yō nō... iririxšāite gaēθanąm “waters, do not be to the one among us who would seek to submit our belongings to abandonment”); or, being itself in the present subjunctive, it is conditioned by the action of the relative clause (Y. 11.5-6 yō mąm taṱ draonō zināṱ... nōiṱ ahmi nmāne zānāite āθrauua “the one who would steal my ritual portion, a priest shall not be born in his house”). The same structure of hypothetical present is found in sentences with a conjunctive clause introduced by yezi “if” and yaṱ “in case, when.” The present subjunctive is required after certain conjunctions, but only if the process is not considered as accomplished: yezi nōiṱ “otherwise, if not,” yaṱ or yaδa “so that, in order that,” pasca yaṱ “after,” āaṱ yaṱ or yauuaṱ “as soon as,” para ahmāṱ yaṱ “before,” vīspəm ā ahmāṱ yaṱ or aṱciṱ ahmāṱ yaθa “until,” yauuata or vispəm ā ahmāṱ yaṱ “as long as,” yauuaṱ “of the same quantity as.” Most of these functions are valid in O. Av., except that the aorist subjunctive is used in the conjunctive clause of final purpose introduced by yaθā, when there is a conditional relationship with the main clause, which expresses the hypothetical present.

The present optative. In the great majority of its occurrences, the present optative is used in the main clause or in an independent clause with a prescriptive value which ranges from wish to most imperative order (Y. 8.6 vasō.xšaθrō hiiāṱ aṧauua “may the righteous be free”; Yt. 10.2 miθrəm mā janiiå “do not break the contract”). The present optative is not only frequent in the comparative conjunctive clauses introduced by yaθa, yaθa yaṱ, or mąnaiiən ahe yaθa(na) “as” (Yt. 8.55 tištriiō... pairikąm ādarəzaiieiti... mąnaiiən ahe yaθa hazaŋrəm narąm ōim narəm ādarəzaiiōiṱ "Tištriia immobilizes the witch as a thousand men would immobilize a single man”), it even has an intrinsic comparative value in the independent clause, main clause, and relative clause (Yt. 13.107 yeŋˊhe nmāne aṧiš fracaraēta “the one in whose house, one would say, Aṧi circulates”). The present optative can also express the iterative aspect of the past (decisive study by Hoffmann, Aufsätze, pp. 605 ff.): the action started, finished, started again, finished again, and so on for an indeterminate number of times in the past (Yt. 10.73 miθrəm... yō... auuarōiṱ vācim “Miθra who, incessantly, raised his voice”). The forms which perform this function frequently are augmented, like the one in the example above. The negation is with forms having a prescriptive value and nōiṱ with the others. In O. Av., the present optative appears in the main/independent clause (where it cannot receive a negation) with a prescriptive value and in the relative clause to express the imaginary present. The use in a conjunctive clause is suppletive to the aorist optative.

The present imperative appears only in the independent clause and has an exclusively prescriptive value. In contrast with the present optative, which expresses generic prescriptions, the present imperative is used to issue orders which have to be carried out hic et nunc. In Y. Av., the negation is . In O. Av., the imperative is incompatible with negation; see the present injunctive and the aorist injunctive.

The aorist indicative is only found in O. Av., where it nearly always expresses the recent past. The aorist injunctive in O. Av. expresses in an independent/main clause a punctual action, accomplished once and for all hic et nunc. In a dependent clause (whatever its type), it contrasts with the main verb and conveys a relation of anteriority to it, expressing the past in relation to the present, and the anterior past in relation to the past. With the particle, the aorist injunctive expresses the preventive prohibition (Y. 48.5 mā nəˊ dušə.xšaθrā xšəˊṇtā “may the beings of evil power have no power over us”). The aorist injunctive disappeared in Young Avestan, although it should be noted that the great Yašts sporadically use some forms which are derived from it, in order to express the past (Kellens, VA, pp. 397 ff.).The aorist subjunctive also belongs exclusively to O. Av. (except for bū: buua- “to become” in periphrastic phrases which add a strong future meaning to the present subjunctive in the independent/main clause). In the independent/main clause, the first person (just as in the present subjunctive) has a volitive function (Y. 50.7 aṱ vəˊ yaojā zəuuīštiiəˊṇg auruuatō “I want to harness for you the fastest steeds”). The third person, marked with the particle, expresses the future of necessity with an exhortative nuance (Y. 47.6 hā zī pourūš išəṇtō vāurāitē “she must push away the many who seek to come”). The aorist subjunctive in the main clause and the aorist subjunctive in the relative clause are in the conditional relationship of hypothetical present: Y. 33.2 aṱ yəˊ akəm drəguuāitē... varəšaitī... tōi vārāi rādəṇtī ahurahiiā... mazdå “the one who hurts the deceiver (and the one who...), they will meet the demands of Ahura Mazdā” (in Y. Av., this function is governed by the present subjunctive). The same applies in sentences made up of a main clause and a conjunctive clause, structured by the hiiaṱ... aṱ “if... then” correlation. Sometimes, the aorist subjunctive in a dependent clause connotes a virtual action, which exists only in the speaker’s mind and therefore is extraneous to tenses, modes, or categories (actual, hypothetical, unreal) which set the degree of reality of an action (Y. 51.6 aṱ ahmāi akāṱ aš´iiō yəˊ hōi nōiṱ vīdāitī “(Mazdā gives) worse than bad to the one who does not serve him”). The aorist subjunctive is required after parā hiiaṱ “before,” yezi “since (causal),” when the process in question belongs to the future, and yaθā “in order to.” The only certain example of a final clause in O. Av. Is Y. 44.1 yaθā nəˊ ā vohū jimaṱ manaŋhā “in order for him to come to us with the good Thought.” The aorist optative. In the independent/main clause, its function is the same as the function of the present optative; and in the relative clause it seems to be interchangeable with the aorist subjunctive to express a virtual action. It is very rare in a conjunctive; but, in two instances, we can hypothesize that the aorist optative of the main clause and the aorist optative of the conjunctive clause are in a conditional relation of unreal present. In Y. Av., the aorist optative survives with a precatory function (Y. 60.5 vainīṱ…... sraošō asruštīm “may obedience win over disobedience”). The aorist imperative. This appears only in O. Av., and its function is identical to that of the present imperative. However, the present imperative frequently follows closely the aorist imperative, and it seems that the fulfillment of the command expressed by the latter is a prerequisite for the fulfillment of the command expressed by the former (Y. 33.11 sraotā mōi mərəždātā mōi “hear me, (then) pity me”).

The perfect. As in Vedic or Greek, the indicative expresses the acquired state and the resultative past. It is frequent in enumerations of temporal opposition, where it represents the past in relation to the present (present indicative) and to the future (future indicative present in Y. Av. and aorist subjunctive in O. Av.): Y. 19.10 fraca vaoce fraca mruiiē fraca vaxšiiete “which was proclaimed, which is proclaimed, and which will be proclaimed.” The same contrast can be expressed with participial forms. The perfect subjunctive and the perfect optative are only found in Y. Av. The present subjunctive appears in structures expressing the hypothetical present, and it indicates anteriority in relation to the present subjunctive and is in contrast with the latter. The perfect optative expresses the unreal past, as is usual in ancient Indo-Iranian dialects (Yt. 13.12-13 yeiδi zī mē nōiṱ daiδīn upastąm... frauuaṧaiiō... drūjō astuuå aŋhuš åŋˊhāṱ “if the Fravaši had not given me their help, the material life would have belonged to the Deception”).

The impersonal modes. The present participle can be used in absolute constructions, such as the ablative absolute introduced by paiti, and in verbal phrases, where it or the personal verb play the role of the aspectual or modal auxiliary. The most frequent of these phrases consists of a combination of the middle present participle of xšā (xšaiiamna-) with a personal verb to express the idea of “to be capable.” Already in O. Av., the aorist participle either expresses anteriority in relation to the personal verb or is used in order to express the idea of final purpose.

The infinitive. Its functions and formations are in general similar to the ones in Vedic. It is actually used only in O. Av.

The sentence. (For Vedic, see H. Hettrich, Hypotaxe im Vedischen, Berlin, 1988). The simple sentence (R, pp. 349-63) does not create any specific problems, and several issues regarding the conjunctive clauses (R, pp. 372-87) are presented above in the use of modes. The relative clause (R, pp. 363-71) has been extensively studied in a large number of recent works. Kellens-Pirart (TVA II, pp. 53-64), while studying the construction of the relative clause in O. Av., remark the frequency of the sa-/ta- ... ya- correlation and emphasize the fact that the simple deictic is necessarily correlative when there is a relative clause. In the absence of a correlation, the antecedent is either in direct contact with the relative pronoun or is introduced in the relative clause and attracted to the case of the relative pronoun. B. Forssman (MSS 45, 1985, pp. 55 ff.) studies the different modes of the “prolongation” (Fortführung) of a relative clause by another relative clause: if the antecedent changes, by a secondary relative clause or coordinated relative clause; if the antecedent does not change, by an interpolated relative clause, relative clause in asyndeton, coordination of a verbal syntagm, or addition of a new sentence beginning with a personal pronoun, which is a substitute for the relative pronoun. N. Oettinger (MSS 42, 1983, pp. 177 ff.) reconstructs the history of the yá-... cí- °ca/yá ... ká- °cid correlatives and concludes that the first variant is the original one. His observations are extended, from a comparatist viewpoint, by Klein’s (MSS 44, 1985, pp. 105 ff.). According to Oettinger (IIJ 29, 1986, pp. 45 ff.) the relative pronoun of a nominal relative clause in O. Av. can be attracted to the case of its antecedent, if the latter is in the accusative or in the instrumental; this would be the origin of the use of the instrumental functioning as a nominative or a vocative in Y. Av. However, Kellens-Pirart (TVA II, pp. 62 ff.) considers that this attraction is so exceptional that the excerpts illustrating it become suspect and, therefore, the use of the eżāfa is not proven in O. Av. The use of the eżāfa, i.e., an unmotivated relative pronoun between the noun and its attribute or its determinative genitive, is, on the other hand, frequent in Y. Av. H. Seiler’s hypothesis (Relativsatz. Attribut und Apposition, Wiesbaden, 1960), according to which the eżāfa has a specifying value, does not account for all of its occurrences, as Duchesne-Guillemin showed in decisive manner in his “Forschungsbericht” (pp. 24-25).

Particles. These always have been neglected, and Reichelt does not mention them at all. There now exists J. S. Klein’s study (Toward a Discourse Grammar of the Rigveda, Heidelberg, 1985) for Vedic and Pirart’s (in Kellens-Pirart, TVA II, pp. 99-193) for O. Av. The latter’s conclusions are conveniently summarized in the TVA II lexicon under each specific lemma. This allows us to mention below only the most important usage. The aṱ particle. G. E. Dunkel (KZ 101, 1988, pp. 53 ff.) sees in it a mark of antithesis, apodosis, and emotional exclamation. Pirart emphasizes more strongly the correlative function. The usual order is hiiaṱ ... aṱ, but some particular conditions may produce the reversal into aṱ... hiiaṱ.The ā particle. Even if generally the Avestan ā or the Sanskrit (Sk.) āˊ, is not coordinating (Dunkel, IIJ 24, 1982, pp. 98 ff.), the nōiṱ ... naēdā “neither... nor” form has nōit + ā as the second member (F. B. J. Kuiper, IIJ 30, 1987, pp. 209 ff.). Hence, Pirart (pp. 131 ff.), who identifies the functional equivalency with Sk. ná... ná u, believes that ā is equivalent to Sk. u.The °cā particle: Pirart’s study confirms the validity in O. Av. of several rules laid down by Klein for Sanskrit. When a coordinated member is made up of several words, the particle °cā is always attached to the first of these words (p. 142). °cā carries a differential value, so that the two adjectives or the two relative pronouns which it brings together cannot represent the same thing or person (p. 160). When there are more than three coordinated members, a subdivision has to be carried out; in the case of exactly three members, the only permissible schemes are Acā Bcā Ccā, A Bcā Ccā, A B Ccā, Acā B C.

Miscellaneous. The particle, when it is not prohibitive, is equivalent to Sk. sma and gives the value of a perfect to the indicative of a secondary present (Pirart, pp. 178 ff.). There is no other particle apart from the disjunctive one “either... or “ (ibid., pp. 180 ff.).


Bibliography (in addition to references given in the text).

Compendium linguarum iranicarum, ed. by R. Schmitt, Wiesbaden, 1989.

Corolla Iranica. Papers in honour of David Neil MacKenzie, ed. by Ronald E. Emmerick and Dieter Weber, Frankfurt, 1991.

Studia Grammatica Iranica, Festschrift für Helmut Humbach, (Münchener Studien zur Sprachwissenschaft, Beinheft N.F. 13), ed. by R. Schmitt und P. O. Skjærvø, München, 1986.

K. Hoffmann, Aufsätze zur Indoiranistik, Wiesbaden, 1975.

H. Humbach, Die Gathas des Zarathustra, 2 vols., Heidelberg, 1959.

J. Narten, Aməṧa Spəṇtas im Avesta, Wiesbaden, 1982.

Search terms:

 زبان اوستایی avestan language  zaban e avestae


(Jean Kellens)

Originally Published: July 20, 2002

Last Updated: August 17, 2011