To lie to or to deceive? Possible evidence of (in)transitive alternation in Old Persian

In the entire Old Persian corpus, only in the Bīsotūn inscription do we find 25 occurrences of verb forms of the root drau̯g- (Indo-Iranian *d(h)rau̯gh- ‘(be)lügen’), usually translated as ‘to lie, to deceive’ (cf. Kent 1953: 191; Schmitt 2014: 170, ‘(be)lügen, trügen’). Despite the relatively small number of occurrences, three different constructions are documented for this verb: a) Nominative (18 occurrences; example 1); b) Nominative-Genitive (2 occurrences; example 2), where the genitive, which in Old Persian is a syncretic case, has a dative function; c) Nominative-Accusative (2 occurrences; examples 3a and 3b). (1) hau̯ adurujiya avaθā aθanha he.NOM lie.IMPF.3SG thus say.IMPF.3SG ‘he lied; thus he said’ (Schmitt 1991: DB 4.8); (2) hau̯ kārahyā avaθā adurujiya he.NOM people.GEN thus lie.IMPF.3SG ‘He lied to the people thus’ (Schmitt 1991: DB 3.80); (3a) kāram avaθā adurujiya people.ACC thus lie.IMPF.3SG ‘To the people thus he lied’ (Schmitt 1991: DB 1.78), cf. Kent (1953: 120), ‘thus he deceived the people’; (3b) taya imai̯ kāram adurujiyaša because this.NOM.PL people.ACC lie.IMPF.3PL ‘because these (men) lied to the people’ (Schmitt 1991: DB 4.34-35), cf. Kent (1953: 131), ‘so that these (men) deceive the people’. Both the genitive/dative and the accusative express the entity to whom one lies or who is deceived. The three remaining occurrences are forms of the participle in -ta- (example 4), which are not strictly relevant for the purposes of the present study. (4) nai̯šim ima vr̥navātai̯, not=he.ACC this.NOM convince.PRS.SBJV.3SG.MID duruxtam maniyātai̯ lie.PST.PTCP.ACC regard.PRS.SBJV.3SG.MID ‘(and) this should not convince him, (but) he regard it as false’ (Schmitt 1991: DB 4.49-50). Thus, from a semantic perspective, we can initially classify the verb drau̯g- as a verb of communication or, more generally, of ‘interaction’ that usually involves two animate entities and is characterized by a low degree of semantic transitivity. While the absolute use of the verb drau̯g- has not attracted the attention of scholars, the coexistence of the construction with the accusative and with the genitive/dative has been of greater interest, even though a convincing explanation for this data has not yet been provided. Indeed, the various scholars who have considered this issue have either limited themselves to labelling the different constructions as ‘transitive’ or ‘intransitive’ (see, among others, Schmitt 1991: 54, fn. 78), without giving a detailed account for the phenomenon, or interpreted this alternation as a case of stylistic variation, without considering either the morphosyntactic level proper, or the semantic one, whether in synchrony or diachrony (Schmitt 2016: 106). This latter reading has been mainly based on information provided by the Elamite and Babylonian versions. The aim of this paper is to demonstrate that the alternation between Nominative-Genitive(Dative) and Nominative-Accusative constructions is not a mere stylistic variation, but has a valid linguistic basis, and that the two expressions need to be considered as two different argument structure patterns. Various elements seem to lead to this interpretation, although the scarcity of Old Persian documentation in this case does not allow us to formulate a robust hypothesis – or even posit that there is just one plausible hypothesis. Nonetheless, it seems quite reasonable to start from the assumption that the verb semantics and the low degree of transitivity may have played a role in the production of the alternation between the genitive(dative) and the accusative as second argument. This has often been recognised, both at a synchronic and a diachronic level, by typological studies and research on transitivity (see, among others, Tsunoda 1981 and 1985; Næss 2007). Interestingly, on the genealogical side, comparable constructions in Young Avestan and Vedic evidence a second argument encoded by the accusative (example 5) and the dative (example 6) respectively, suggesting the possibility that the alternation of dative(genitive) and accusative for this verb developed in the Indo-Iranian branch of the Indo-European languages, but was preserved only in Old Persian. (5) yōi [pauruua] miθrəm [družiṇti] (Yt. 10.45) who.NOM.PL before Miθra.ACC betray.PRS.3PL ‘who [betrayed] Miθra [before]’ (cf. Skjærvø 2009: 127); (6) tā́bhyaḥ sá nír r̥cchād, yāḥ ||1|| this.ABL.F he.NOM away go.PRS.SBJV.3SG who.NOM naḥ prathamò ’nyò ’nyásmai drúhyāt (TS 6.2.2.1-2) =us.GEN first.NOM one another.DAT be deceitful.AOR.OPT.3SG ‘He who first among us will be deceitful to another will suffer loss of these [bodies]’ (cf. Kulikov 2012: 565). Finally, from an internal diachronic perspective, the various Middle Persian developments from the Old Persian drau̯g- are classified as transitive or intransitive.