1 Indo-Iranian People in the Ancient Near East: An Overview of the Studies
The subject of Indo-Iranians in the cuneiform sources from the ancient Near East, specifically related to the Hurrian kingdom of Mittani, is striking in its bibliographical overabundance, given the limited amount of material attested. Since a detailed discussion of the bibliography is impossible here, in what follows we will summarize the milestones of the debate and the main solutions suggested without making any claims for completeness. For this purpose, we decided to conventionally divide the studies that have been conducted since the first identification of this linguistic material into two stages. During the first stage, five different solutions were suggested for how the material related to Indo-Iranians could be classified linguistically: as 1) Iranian, 2) Indo-Aryan, 3) Indo-Iranian before the split into the two branches, 4) both Iranian and Indo-Aryan, or 5) a third branch of the Indo-Iranian group. The beginning of the second stage can be dated roughly to the middle of the 20th century and is represented in the person of Manfred Mayrhofer. Through his many contributions to the field, it became clear that the Indo-Aryan hypothesis best accounted for the data.
1.1 From the Discovery to the Mid-20th Century
The acknowledgment of the presence of Indo-Iranian people in the ancient Near East in the second millennium BCE predates the discovery and interpretation of Hittite texts. The letters sent by Tušratta of Mittani to Amenhotep III and Amenhotep IV, found at Tell el Amarna beginning in 1887, contain names of the kings and nobles that had already been recognized as Indo-Iranian1 when Winckler (1907) pointed out the presence of Vedic god names in the cuneiform tablets found at Boğazköy that preserved the treaty between Šuppiluliuma I and Šattiwaza of Mittani (which led him to the misidentification of the ethnonym and toponym Hurri, read as Harri, with the Indic ārya-, the self-designation of the Indo-Iranians).2
Despite some sporadic criticism on whether the onomastic material was Indo-Iranian,3 Winckler’s discovery seemed to provide definitive confirmation of the presence of a group of Indo-Iranians in the kingdom of Mittani who belonged strictly to the ruling class, namely, an Indo-Iranian dynasty settled in an area in which a non-Aryan (Mittanian) community lived.4 Further evidence emerged from the Hittite hippological treatise by Kikkuli (CTH 284), in which some Indo-Aryan technical terms could be recognized,5 and a Hittite ritual text (CTH 395) in which the name of the Indic god Agni was found.6 These discoveries provided further support for the previous identification of the Kassite Sun god name Šuriya with the Vedic sū́riya-. Moreover, an Indo-Iranian etymology seemed to be available for relevant words and place-names in texts concerning Mittani from different areas (Alalah, Egypt, etc.)—for example, mariyanni7 and Waššukanni8—and some terms referring to horse colors with a likely Indo-Aryan etymology were later identified in texts from Nuzi.9
Thus, the collected evidence of personal names, divine names, place-names, and technical terms confirmed the presence of Indo-Iranian people throughout the ancient Near East. However, several difficulties emerge in trying to define the linguistic identity of such Indo-Iranians more thoroughly. Most of the personal names of the Mittanian rulers were initially regarded as specifically Iranian rather than Indo-Iranian or Indo-Aryan, although they showed some features seemingly pointing to a stage before the diversification of the Iranian dialects (in particular, the initial *s was not changed to h before a vowel).10 In contrast, the divine names found in the treaty between Šuppiluliuma I and Šattiwaza, who belonged to the Mittanian dynasty, seemed to be unequivocally Indo-Aryan,11 as did the glosses in Kikkuli’s hippological text.
Different solutions were suggested to account for such a puzzling situation. Meyer (1908b) regarded this Indo-Iranian stock as belonging to a stage predating the branching of the Indo-Iranian group into the two distinct Indo-Aryan and Iranian subgroups. This theory was quickly challenged by Jacobi (1909:726), who, while accepting the Iranian status of the personal names of the Mittanian rulers, remarked on the unequivocally Indic character of the deities in the Šattiwaza treaties and suggested that the Mittani Aryans were an Iranian tribe coming from the east of Iran and highly influenced by the Vedic culture. However, his scenario was disputed by several scholars, who accepted Meyer’s position.12 Konow (1921:60), much like Jacobi, explained the presence of Vedic gods and glosses in Hittite texts as the result of “the pre-historic expansion of Indian civilization” in the form of “a peaceful propaganda,” while Mironov (1933) opted for an unlikely scenario involving a mixed Indo-Aryan and Iranian community. A different position was staked out by Lesný (1932), who regarded the Mittani Aryan as neither an Indo-Aryan nor Iranian language but rather as the only known member of a third branch of the Indo-Iranian group.13 Feiler (1939) regarded all of the Mittani Aryan linguistic material, including the personal names of the Mittanian rulers, as closer to the Indo-Aryan group.14
We choose to regard Belardi’s comprehensive contribution, published in 1951, as the final chapter of this first stage of research. After carefully reevaluating all of the available data and the various hypotheses, he concluded that the linguistic identity of these Indo-Iranian relics could not be determined, although they could not be considered Iranian.
1.2 From the Mid-20th Century to the Present
From the middle of the 20th century until recently, the debate on the Indo-Iranians in the ancient Near East and their language was dominated by Manfred Mayrhofer, who published, beginning in 1959, thorough and fully documented studies of the linguistic evidence from linguistic, philological, geographical, and historical perspectives.15 Mayrhofer tackled the entire lexical evidence, including personal names, divine names, place-names, and technical terms occurring as loanwords from different Near Eastern areas. Moreover, he considered language contact between the Indo-Aryan superstrate and the Hurrian substrate and dealt with the Hurrian adaptation strategies used for Indo-Aryan words (see § 3 below).
Mayrhofer’s studies are crucial for the recognition of the essentially Indo-Aryan character of all of the lexical material attested in the cuneiform sources, although that solution was previously proposed by other scholars, as mentioned above, and by Mayrhofer’s contemporary Thieme (1960), who examined the material from a cultural perspective. Thieme revived Konow’s (1921) arguments on the specifically Indo-Aryan nature of the gods in the Šattiwaza treaties (although cautioning that Indo-Aryan and Proto-Aryan cannot be distinguished from a strictly linguistic point of view).
Mayrhofer’s Indo-Aryan solution was criticized by Kammenhuber (1968) and Diakonoff (1972), who challenged several Indo-Aryan etymologies suggested for personal names and loanwords and pointed out that no evidence for a living Indo-Iranian language can be found in Near Eastern documents of the 15th and 14th century BCE, claiming, in particular, that 1) the glosses in the Kikkuli text were just Hurrianized fossils mechanically reproduced by the scribes; 2) the Indo-Aryan deities in the Šattiwaza treaties, also in a Hurrianized form, do not occur elsewhere in Mittanian documents; 3) the Indo-Iranian personal names of the Mittanian kings were throne names (Šattiwaza originally bore the Hurrian name Kili-Teššub); and 4) no traces of Indo-Iranian influence can be found in Hittite. Despite Thieme’s (1960) analysis, even the gods in the Šattiwaza treaties offer no compelling evidence for whether the Near Eastern Aryans were still Indo-Iranians or already Indo-Aryans. Furthermore, they contended that any Indo-Iranian people who existed in the ancient Near East would not have been those who later reached India because no Hurrian elements can be found in Old Indic texts. They concluded that Indo-Iranians probably never penetrated the Near East in the second millennium BCE, instead only interacting with a group of Hurrians (the Mittanians sensu stricto) in a border area, which Diakonoff believed was Armenia (1972:116–120). These contacts, however, were highly significant as they gave rise to dynastic names, the adoption of some deities, and the introduction of technical terms related to horse training.
Mayrhofer defended his positions in his first monograph (Mayrhofer 1974a) and later publications. His results have been generally accepted by later scholars,16 who significantly added to them with fine-grained analyses of the linguistic material, on which the discussion in the following sections is based.17
2 Sources
As mentioned, there are no full texts in the Indo-Aryan from the ancient Near East but rather only a set of lexical items relating to texts with various findspots, associated with the kingdom of Mittani. The nature of the lexical material attested—personal names, divine names, place-names, glosses (technical terms), and sporadic loanwords—seems to suggest that they are relics rather than proofs of the existence of a living community of Indo-Aryan speakers in the kingdom of Mittani.
The Amarna letters from Mittani provide us with a number of personal names, mostly borne by the rulers of the kingdom and some officials. Other personal names and sporadic loanwords can be found in texts from Alalah (level IV) and Kassite Babylonia. The documentation from the archives of Boğazköy is the main source for divine names—particularly the (Hurrianized) names used to refer to the deities who witnessed the treaty between Šuppiluliuma I and Šattiwaza of Mittani, to which should be added Agni, who is mentioned in a Hittite ritual—and technical terms referring to the training of horses, which appear as glosses in the hippological treatise by Kikkuli. Texts from Nuzi also provide other Aryan personal names and technical terms belonging to the sphere of horse training, while other loanwords are found in texts from different areas.
3 Linguistic Analysis
In the following sections, we will provide an overview of the Indo-Aryan lexical material found in cuneiform sources. It is organized into two main groups: onomastics (with three subgroups) and loanwords. Since discussing each lexeme for which an Indo-Aryan etymology has been suggested would unnecessarily burden the text, we opted for a selection of the most relevant material, especially focusing on their Indo-Aryan etymology and the strategies of adaptation they show in the target languages in which they appear.
3.1 Onomastics
As mentioned, proper names represent the great majority of the Indo-Aryan lexical material from ancient Near Eastern sources. Most are personal names, although there are also some divine names (to which one may add those occurring in theophoric personal names) and perhaps a couple of place-names. Their etymology is often unproblematic and can be easily traced back to Indo-Aryan roots, although some unclear or ambiguous cases exist.18
Most of the personal names found in Indo-Iranian sources are built through nominal compounding.19 They are generally found uninflected, in their original stem form (Hittite case endings may be added sporadically in names occurring in Hittite texts). Conversely, most of the divine names and place-names feature more complex strategies of adaptation that involve Hurrian suffixes, although some elements still lack compelling explanations.
3.1.1 Personal Names
The etymology and structure of the Indo-Aryan personal names known from cuneiform sources have been more or less extensively discussed by almost all of the scholars who dealt with the issue of the Indo-Iranians in the ancient Near East. In the most recent comprehensive analysis of the material, Gentile (2019) has shown that these personal names are morphologically and semantically consistent with those attested in Vedic and Avestan. Most are built through composition, except for some single-stem names (e.g., Wāzzi, to be compared to the Vedic vājín- ‘racer’). The majority are possessive compounds (bahúvrīhi), for example, Tušratta ‘whose chariot is vehement’ (var. Tuišeratta, Tušeratta), < *tvaiśa-ratha-, or the names having -atti ‘guest’ (< *-atthi)20 as their second member, such as Intarratti ‘having Indra as his guest’, Mittaratti ‘having Mitra as his guest’, and Tewatti ‘having a god as his guest’ (< *daiwa-). The exocentric éntheos type is also found, although less frequently (e.g., Abiratta, ‘facing chariots’, < *abhi-ratHa-), and the endocentric determinative compounds (tatpuruṣa) are well attested, such as Indaruta ‘helped by Indra’ (with -ūtá as the second member; cf. the Vedic avi-/ū- ‘help’)21 and Yašdata ‘given by the sacrifice’ (< *yaj́a-dāta). Finally, several names are built as verbal governing (or synthetic) compounds, in which the ‘verbal’ element can be either the first or the second member (cf. especially Šattawaza ~ Šattiwaza < *sāti-vāj́a- ~*sāti-vāj́a- vs. Wašašatta < *vāj́a-sāta-, both meaning ‘having reached the prize’).
The meanings of Indo-Aryan names in ancient Near Eastern sources generally cover the most relevant semantic field of Indo-Iranian names. Some names that refer to war and military valor are compound names that include nouns like aš(šu)wa ‘horse’ (= the Vedic áśva- < *Haćwa-) and ratta ‘chariot’ (< *ratHa- lit. ‘the one who has wheels’), such as Aššuzzana ‘delighting in horses’ (< *Haćwa-cana), Biriyaššuwa ‘whose horse is dear’ (< *priHa-Haćwa-), Abiratta ‘facing chariots’, and Tušratta ‘whose chariot is vehement’.22 Other names, with -atti ‘guest’ (see above), relate to hospitality. Rightness, truth, and order are the basis of compound names with arta ‘truth’ (= the Vedic r̥tá-)—for example, Artamanya ‘thinking of R̥tá-’ (< *Hr̥ta-manya-), Artatama ‘whose abode is the R̥tá-’ (< *Hr̥ta-dhāman), and Artaya ‘acting according to the R̥tá-’ (< *Hr̥tayant-). These names are relevant to the dialectal position of the language among the Indo-Aryan branch (cf. names like Biryamašda, built with *mazdha- ‘wisdom’, or Zantarmiyašta and Zirdamiyašda, with *miyazdha- ‘sacrifice’, which preserve the original /azd(ʰ)/, regularly changed to /eːd(ʰ)/ in Vedic).23 From the perspective of language contact, however, their informative value is limited.
The fact that the Indo-Aryan names of the Mittanian kings were throne names24 is extremely relevant from a sociolinguistic point of view, but there is insufficient data to allow the reconstruction of a definitive scenario. From a strictly linguistic perspective, their spelling shows consistent strategies of graphic adaptations compared to their Indo-Aryan corresponding forms, with only sporadic examples of variation (e.g., Tušratta, Tuišeratta, Tušeratta), and no other peculiar features can be observed. In this respect, divine names, place-names, and loanwords—dealt with in the following sections—provide more interesting data, since their adaptation often involved Hurrian morphemes.
3.1.2 Divine Names
As mentioned, Indo-Aryan divine names were among the first Indo-Aryan materials identified with relative certainty. Most are found in the list of divine witnesses in the treaty between Šuppiluliuma I and Šattiwaza of Mittani. They are the most problematic divine names to analyze linguistically because their bases almost perfectly match the corresponding Vedic divine names, but the Hurrian suffixal elements they display are difficult to elucidate. The four names were written as follows:
KBo 1.1+ rev. 55′–56′ (with variants from KBo 1.3+ rev. 24′)d.MEŠMitraššil d.MEŠUruwanaššil (var. d.MEŠArunaššil) dIntar (var. dIndara) d.MEŠNašattiyanna
Their identification with Mitra, Varuṇa, Indra, and the Nāsatyā (the elliptic dual standing for the names of the twin gods Aśvin and Nāsatya)25 is straightforward, and their order perfectly matches the list found in RV 10.125.1cd.26 The first two names probably reflect the dvandva compound Mitrā-Varuṇā (both terms of which are in the dual form),27 which could explain the plural determinative applied to both names. However, the puzzling suffix -ššil remains obscure. A tentative solution, already suggested by Friedrich (1943), would see in the suffix the reflex of the Hurrian numeral šini ‘two’ + the Hurrian plural suffix -lla, which would be consistent with an original Indo-Aryan dual. Fournet (2010:7), while maintaining an origin from the Hurrian šini, explains the final -l slightly differently, identifying it as the Hurrian pronoun -lla ‘they, them’.28 In either case, such a suffixal chain is unique, and none of the analyses suggested can be confirmed. The name of Indra does not pose particular problems (for occurrences in personal names, see § 3.1.1 above). In d.MEŠNašattiyanna, reflecting the dual form Nāsatyā, the final suffix -nna is generally explained as the Hurrian plural marker.
Some other possible Indo-Aryan divine names have been identified in cuneiform sources: Agni, Sū́riya, and the Maruts. The former occurs uninflected (dĀgni—dupl. dAgni—functionally corresponding to a genitive) in the Hittite ritual of Hantitaššu (CTH 395), whose belonging to the Hurrian milieu is debated,29 and with Hittite endings in the omen text KUB 8.28 (nom. dĀgniš), in the fragmentary ritual KBo 13.147 (nom. dAgniš, dat. ANA dAgnī), in the tablet catalog KUB 30.51+, recording the existence of a mugawar of Agni (gen. dAgniyaš), and in the fragmentary historical text KBo 3.46+ (nom. Agniš), which deals with Muršili I’s campaign against Hurrians and contains some possible Indo-Aryan personal names.30 The god surely corresponded, both formally and functionally, to the Vedic fire-god Agni,31 but whether it should be regarded as a cognate or a true Indo-Aryan loanword is debated.32
As for Sū́riya, the name occurs as Šuriyaš in Kassite onomastics in cuneiform sources, in the personal name Šuriātti ‘having Sū́riya as his guest’ at Alalah (see § 3.1.1 above), and in a Kassite-Akkadian vocabulary (BM 93005), in which the Akkadian column equates it with Šamaš. Therefore, the identification of the Kassite divine name with the Vedic Sun deity seems unavoidable.33
Finally, the divine name Marattaš is found in the same vocabulary, matching the war god Ninurta. The same theonym occurs in Kassite onomastics as Maruttaš ~ Murutaš. The name has been compared to the Vedic Maruts, although this correspondence is not unanimously accepted.34
3.1.3 Place-Names
The only toponym that has been almost unanimously accepted as deriving from the Indo-Aryan dialect attested in the ancient Near East is Waššukanni, the capital city of the kingdom of Mittani. Friedrich (1925:121) explained this city name as the Old Indic *vasu-gaṇī- ‘containing the multitude of the Vasu’ (name of a class of deities), while Kretschmer (1927:93–94) recognized the Old Indic vasu- ‘good’ in the first part of the name and suggested the Old Indic jána- ‘humans, family, folk’ for the second part; Waššukanni, that is, *vasu-jani-, would thus mean something like ‘provided with noble population.’ Mayrhofer (1959b:2 fn. 4), however, rejects such etymology because the Hurrian reflex of the Old Indic j usually appears as z. Therefore, the preform should be reconstructed as *vasu-ka- (adapted with the Hurrian suffix -nni), with a ka- suffix that can be found in other Indo-Aryan words.35
Quite unconvincing is the Sanskrit etymology suggested by Fournet (2010:11; 2012:241–242) for the obscure toponym Mittani (also Maiteni at Nuzi), which is traced back to the Old Indic verb mith- ‘unite’: *m[a]ithāṃ (accusative) + Hurrian -nni would thus mean ‘union’ or ‘united kingdom,’ “en cohérence avec la présence simultanée de Hourrites et d’Indo-iraniens” (Fournet 2012:242).
3.2 Loanwords and Technical Terms
After scholars became aware of the existence of Indo-Aryan personal names and divine names in the ancient Near East, Indo-Aryan etymologies were suggested for several words of unclear origin attested in Akkadian, Hurrian, and Hittite texts. Some of these words are attested throughout the Near East; others are technical terms recorded in specific areas whose circulation outside the texts in which they occur cannot be evaluated. Most evince a clear Hurrian intermediation, signaling the major role of the kingdom of Mittani in the transmission of such material toward Anatolia and Syria.
The term mari(y)an(n)u, used to refer a high-ranking social class and typically translated as ‘charioteer, warrior, nobleman’ (vel sim.), is widely attested throughout the Near East, with occurrences at Amarna, Alalah, Boğazköy, Nuzi, and Ugarit (mryn), as well as in Egypt (mrjn, cf. Takács 2008:417–418). This noun is often traced back to the Old Indic márya- ‘young man, member of the Männerbund’,36 adapted into Hurrian through the suffix -nni,37 although some scholars defend a genuine Hurrian etymology based on the existence of Urartian mariahini ‘mare-men’.38 If it was an Indo-Aryan loanword in Hurrian, it was very productive (type IV, according to the classification employed in this volume: loanwords showing morphological integration and base productivity) because several derivatives are attested: the collective nouns mariyannardi (Mittani) and mariyanzari (Boğazköy) ‘group of mariannu’ and the morphologically unclear mariyannade (Alalah) and mariyannui (Boğazköy). A possible synonym of the far more common mari(y)an(n)u was martiyanni, only attested at Nuzi and also an Indo-Aryan loanword < mártiya- ‘man’ + the Hurrian suffix -nni (unless it is a scribal mistake for mari(y)an(n)u).39
Another widespread noun with a possible Indo-Aryan etymology is maninnu, mannin(n)i- ‘necklace’, attested in Akkadian texts from Amarna, Qatna, and Alalah, as well as in Hittite texts from Boğazköy. Its base may match the Vedic maṇí- ‘id.’, which was adapted into Hurrian as usual through the productive suffix -nni. Furthermore, the Hurrian noun wadurānni ‘bridewealth’, attested in the Mittani letter and at Alalah, is regarded by Mayrhofer (1996:161–162) as reflecting the Indo-Aryan *wadhū-rā- ‘bride-gift’ (cf. the Vedic vadhū́- ‘bride, young woman’. For the element °rā- ‘gift’, Mayrhofer compares the Vedic śatá-rā- ‘with a hundred gifts’), which was adapted into Hurrian through the suffix -nni. Mayrhofer (1965b) also proposed that the Akkadian mištannu ‘pay, reward’, attested in the treaty between Idrimi and Pillia from Alalah (AT 3), reflects a Hurrian word adapted with the suffix -nni from Indo-Aryan *miždhá- ‘pay, price’, with preservation of the original cluster -žd- (vs. the Vedic mīḍhá-) as in the personal names Biryamašda, Zantarmiyašta, and Zirdamiyašda (see above).40
According to Yakubovich (cited in Giorgieri 2010a:938 fn. 38), the Hurrian niġ(a)ri ‘dowry’ could match the Old Indic ni-har- ‘to gift’, although the absence of the suffix -nni, often employed in adapting foreign words into Hurrian, could argue against this possibility. Another Hurrian word with a possible Indo-Aryan etymology is maganni ‘gift’ (magannu in Akkadian texts), which can be compared to the Vedic maghá- ‘id.’.41 Less certain is the supposed Indo-Aryan origin of k/gat(t)inni, meaning unknown (attested at Alalah, Amarna, and Nuzi), which was perhaps related to the Vedic khādí- ‘bracelet, ring’.42
Some technical terms attested at Boğazköy, Nuzi, and Alalah concern hippology. Some Indo-Aryan glosses have been identified in the treatise on horse training by Kikkuli that was found in the Hittite capital city. All show the same structure—a numeral + wartanna- ‘lap, turn’ (= the Hittite wahnuwar)—and refer to the number of laps or turns made by the horses.43 They include aikawartanna ‘for one lap’ < *aika-wartana-; tierawartanna (also tierurtanna) ‘for three laps’ < *tri-wartana- (although the outcome tiera- remains largely unexplained);44 panzawartanna ‘for five laps’ < *panća-wartana-; šattawartanna ‘for seven laps’ < *sapta-wartana-; and nawartanna ‘for nine laps’ (haplologic from *nawawartanna < *nawa-wartana-).45 In the debate on the exact Indo-Iranian dialect attested in the kingdom of Mittani, these glosses provide crucial data by showing the Indic character of the language, although with some archaisms: see especially aika- ‘one’, which preserves the diphthong later monophthongized in the Old Indic éka-, where the suffix -ka points to Indo-Aryan (vs. the Iranian aiwa-), and šatta- ‘seven’, matching the Indo-Aryan saptá (with a trivial assimilation) rather than the Iranian hafta.
The etymology of LÚaššuššanni- ‘horse trainer’, the professional title held by Kikkuli in the incipit of his treatise, has been much debated.46 The aššu-element of this noun is generally regarded as a reflex of PIE *h1ek̂w-o- ‘horse’, but there is no consensus on whether it represents the Vedic áśva- (which, however, appears as ašwa or aššuwa in personal names, as well as in the noun ašuwaninni from Alalah) or an Anatolian outcome of this root47—compare the Luwian á-sù- or á-zú- (the latter being the most likely reading), Lycian esb(e)-, and perhaps the first element of the Pisidian toponym
Recently, Burgin (2017) suggested that the Hurrian adjective time/ari ‘dark’, attested in the bilingual ‘Song of Release’ from Boğazköy, in which it refers to the earth in the phrase timerre eženi = the Hittite tankuwai taknī, and which should correspond to the Kassite timiraš (a color of horses), is an Indo-Aryan loanword from *tam-r- (cf. the Vedic tamrá- ‘dark’, támas- ‘darkness’).
Some Akkadian adjectives referring to horse colors that are attested in tablets from Nuzi are terms deriving from Indo-Aryan words that were adapted into Hurrian with the suffix -nni: babrunnu (cf. the Old Indic babhru- ‘brown’), barittannu (cf. the Old Indic palitá- ‘gray’), and pinkarannu, pinkarami (cf. the Old Indic piṅgalá-, piñjara- ‘golden yellow’).53 Other Akkadian technical terms in the Nuzi texts that were borrowed from Hurrian and have a possible Indo-Aryan etymology are akkan(n)u ‘wild donkey’ (perhaps deadjectival with a semantic shift; cf. the Old Indic agha- ‘bad’ and the Vedic aghāśva- ‘with a bad horse’), amkamannu ‘?’ (whose base could reflect the Old Indic aṅka- ‘crook’ or aṅga- ‘member’), wirrarikkunni ‘?’ (perhaps related to the Old Indic vīra- ‘man’), zilukannu ‘?’ (with zilu- perhaps a variant of the Vedic jīra- ‘fast’), zirra(ma)nnu, perhaps meaning ‘(very) fast’ (cf. the Vedic jīrá- ‘fast’), etc.54
The noun ašuwaninni is attested at Alalah on a tablet dealing with the delivery of wood to build chariots (AT 422), so an etymology from Indo-Aryan *aśva-nī- ‘horse-leading’ (+ Hurrian -nni) has been suggested, possibly denoting a part of the chariot; compare the Vedic aśvanāya- ‘horse leader, horse shepherd’ and the Vedic phrase áśvam nī- ‘drive the horse (off the wagon)’.55 The Hurrian term aratiyanni, also from Alalah (AT 425, a list of chariots),56 may be the adaptation of an Indo-Aryan word matching the Vedic rathíyà- or ráthiya- ‘belonging to the chariot, part of the chariot, etc.’, with the suffix -nni and a prosthesis of a- because Hurrian does not allow initial r-.57 Similarly, the Akkadian word urukmannu (a decoration of a shield), attested at Nuzi and Amarna, may depend on a Hurrian word related to the Vedic rukmá- ‘gold ornament’, with a prothesis and the suffix -nni.58
4 Concluding Remarks
Despite some etymologies that remain uncertain, it is clear after more than a hundred years of study that the language of the Indo-Iranian onomastic material and loanwords found in second-millennium documents throughout the ancient Near East has Indo-Aryan characteristics, although it is not identical to the language of the Vedic texts because it preserves more archaic features (e.g., the retention of the diphthongs /ai/59 and /au/, the cluster /zd(ʰ)/), besides sporadic innovations such as šatta- ‘seven’ with assimilation vs. the Vedic saptá-. This is consistent with the references to Indo-Aryan gods that match Vedic gods formally and functionally in the Šattiwaza treaties.
The historical implications of the presence of Indo-Aryan people in the ancient Near East are not yet fully understood. The fact of their presence demands new scenarios for the migration of the ancient Indo-European populations. Scholars have suggested the following hypotheses, which are summarized by Mayrhofer (1966):
-
The Indo-Aryans of the Middle East, having left the territory of Mittani, moved east to colonize northwestern India. Mayrhofer rules out this scenario as unlikely.
-
The Indo-Aryans of the Middle East came from India. The solution is unconvincing but not entirely rejected.
-
Migrating Indo-Aryans, probably in Iran, separated into two groups. This is regarded as the most likely scenario by Mayrhofer.60
Regarding the role of Indo-Aryans in the kingdom of Mittani, it is difficult to say anything conclusive. Cultural and linguistic contacts with Hurrians surely existed, and probably the Indo-Aryan element was perceived as highly prestigious, given that Mittanian rulers adopted Indo-Aryan throne names. Nevertheless, there can be little doubt that the dynasty was not of Indo-Aryan blood: the onomastics point to Hurrian and, except for proper names, the lexical material is almost entirely restricted to the sphere of technical terms. As Kammenhuber concluded, there is no evidence for a living Indo-Aryan speaking community in the Near East in the 15th to 14th centuries BCE, and the Aryan endoethnonym is never attested in cuneiform sources.61 For the time being, little more than this can be stated with relative confidence. Only the discovery of new documentary sources could shed further light on the situation.
See Bezold and Wallis Budge 1892:144 (under W(?)idya) and 146 (under Wyašdata); Rost 1897:113, 216; and especially Hommel 1898, Hommel 1899:425, Scheftelowitz 1902:270–273, and Bloomfield 1904:8.
Winckler 1910:291, a correspondence definitively dismissed by Hrozný 1929b:91–92.
See Sayce 1909 and especially Clark 1917, who tried to show that any alleged Indo-Iranian element identified in such names could relate to linguistic elements found in Semitic, Hittite, and ‘Mittanian’ (i.e., Hurrian) names.
See Bloomfield 1904:8: “on the one hand an Aryan dynasty with Aryan names rules in Mitani; on the other hand there is no indication of Aryan nomenclature outside of this dynasty.” After Winckler’s discovery, see, for example, Meyer 1908a:17–24, Meyer 1908b, Jacobi 1909, Kennedy 1909, Konow 1921, Forrer 1922:247–249 (who argued that these Indo-Iranians corresponded to the Manda people [i.e., the Medes] mentioned in Hittite and Akkadian texts), Friedrich 1928:146, and Schmökel 1938.
See Hrozný 1919:xi–xii; Jensen 1919.
See Hrozný 1929a.
Winckler 1910:291.
Friedrich 1925:121, Kretschmer 1927:93–94, Sturtevant 1928:213–214.
First by von Soden 1957:336–337.
See Bloomfield 1904:10–11, who, as an alternative, also suggested the possibility that these names belonged either to a non-Iranian dialect close to Iranian or an Iranian dialect that preserved the initial *s (possibly Median).
See Jacobi 1909, Konow 1921, and Hrozný 1929b:104. A further problem—the fact that these deities did not occur in any of the letters sent to Egypt by the Mittanian kings, in which a different pantheon was found—is ably explained by Konow (1921), with further remarks and corrections by Thieme (1960), by invoking their specific role as the tutelary deities of contracts, peace, etc.
Cf., e.g., Oldenberg 1909, Keith 1909, Kennedy 1909, and, later, Porzig 1927.
This is in line with one of the suggestions by Bloomfield 1904:10–11.
Followed by Dumont 1947 and Hauschild 1962:34.
See Mayrhofer 1959a, 1959b, 1960, 1965a, 1965b, 1966, 1969, 1974a, 1974b, 1982, 1983, 1996, and 2007. On Mayrhofer’s contribution to the clarification of the issue of the Indo-Aryans in the ancient Near East, see especially Raulwing 2013.
But see also the reply by Kammenhuber (1977).
Among others, see Burrow 1973, Hodge 1981, Derakhshani 1998, Raulwing 2000, Witzel 2001, Sadovski 2009, Fournet 2010, Fournet 2012, García Ramón 2015, Sani 2017, Kroonen, Barjamovic, Peyrot 2018:2, and Gentile 2019.
See, e.g., the discussion on the name Biriyaššuwa ~ Biridašwa (probably meaning ‘to whom (his) horse is dear’) in Gentile 2019:142–143.
See especially Schmitt 2000 and Sadovski 2013.
See Mayrhofer 1966:22. Cf. also Pinault 1998.
On this name, see García Ramón 2015.
On these names, see also Sadovski 2009.
See also the loanword mištannu in § 3.2 below.
This is clear from the fact that the original Hurrian name of Šattiwaza, Kili-Teššub, appears in the colophon of the tablet containing his oath to Šuppiluliuma I (KBo 1.3+ rev. 45′).
On the elliptic dual in Indo-European languages, see Wackernagel 1924, I:82–83.
See Dumézil 1952:9. For the function of these deities in the Šattiwaza treaties, see especially Konow 1921 and Thieme 1960.
Various explanations have been suggested to reconcile the cuneiform spellings a-ru-na- and ú-ru-wa-na- with the Vedic Varuṇa, but no conclusive solution can be given (see Thieme 1960:303–304).
Goetze’s explanation (quoted by Thieme 1960:305 with fn. 13) as “some (indefinite) belonging to Mitra-gods,” from Mitra=š (plural indefinite) + -we (genitive) + -l(an) (accusative), appears even less formally and semantically convincing.
See the discussion in Miller 2004a:447–452.
See Kitazumi 2020.
See especially Álvarez-Pedrosa 2016.
Cf., e.g., Kammenhuber 1968:150–155 vs. Mayrhofer 1974:14. See also Carruba 2000.
See Ancillotti 1981:124–125 and Mayrhofer 1982:77, with references.
For a discussion on this divine name, see Ancillotti 1981:97–98. Cf. also EWAia, II:322.
See also Mayrhofer 1960:141 fn. 40 and Hauschild 1962:25 fn. 1.
On the márya- in Indo-European context, see Falk 2002.
Cf. Giorgieri 2000a:211.
For a comprehensive overview of the relevant bibliography, see Richter 2012:244–245.
For all of these forms, see Richter 2012:245.
See also Mayrhofer 1982:73–74.
See Mayrhofer 1960:143, with references.
See Mayrhofer 1960:145–146.
On the hippological interpretation, see Starke 1995 and Raulwing 2005.
Cf. EDHIL:878.
See Hrozný 1919:xi–xii and Jensen 1919. A full treatment can be found in Kammenhuber 1961:293–302.
For a compilation of the hypotheses, with bibliographical references, see HED A:222–223.
See especially Starke 1990:502 fn. 1852.
See Starke 1995:119.
See Wittmann 1964:147–148. For the meaning of this verb, see Sasseville 2020:226–227, with references.
See Starke 1995:117–118 and Carruba 2000:56–59.
See Sasseville 2020:85 for a thorough discussion.
See CAD Š/3:379–380.
Cf. EWAia, II:126–127.
For these terms and others, see Kronasser 1957.
See Mayrhofer 1960:140.
Perhaps the same as the obscure eratti(a)nni from Amarna (EA 22).
See Mayrhofer 1960:144–145.
See Mayrhofer 1960:145. For other possible Indo-Aryan loanwords in Hurrian attested in Akkadian texts, see Fournet 2012.
But cf. the personal name Tewatti < *daiwa-atthi-.
See also Diakonoff 1972 and Burrow 1973:125.
See also von Dassow 2014:12–13.