Written by Afsheen Sharifzadeh, a graduate of Tufts University focusing on Iran and the Caucasus. This article aims to illuminate an ancestral relationship between Pashto and Old Avestan, and explores areal Eastern Iranian features in Pashto as well as the prolonged and intimate contacts with Indic.

Background
The ancestor of Pashto in the Old Iranian stage appears to have been akin to the language of the Zoroastrian Gathas (​​𐬔𐬁𐬚𐬁 gāθā “hymn”). The problem of Pashto’s proximity to Old Avestan is complicated by uncertainties regarding the (1) historic distribution of Avestan and thus the location of the cradle of Zoroastrianism as mentioned in the Gathas (𐬀𐬌𐬭𐬌𐬌𐬀𐬥𐬀 𐬬𐬀𐬉𐬘𐬀𐬵 Airyanəm Vaēǰah ‘Abode of the Aryans’; in concordance with the early proposition of Wilhelm Geiger and Josef Markwart, in the opinion of this author the most likely locale is Khwārazm) (2) issues related to Eastern-Western Iranian classification, since Avestan appears to have been more ancient than this split but Pashto is classified as Eastern (3) ethnic and linguistic continuity between the speakers of Old Avestan or related dialects and the the people who starting in the 3rd century onwards are referred to by the exonym Afghān (Bactrian: αβγανανο abganano), and (4) questions related to when these people left Khwārazm(?) and whence they arrived and expanded into their present distribution in the Hindu Kush mountains and the northwestern peripheries of the Indus river valley. Moreover, with the extant corpus of Old Avestan supposedly dating to 1500-1000 BC in an unknown location, and the first attestation of Pashto appearing only in 1651 AD, the answers to these questions are obscured by the sands of voiceless millennia. In this humble setting, as in many others like it–reconstructive linguistics and foreign loanwords may illuminate avenues into the past, along the same paths traversed by the ancestral Pashtuns, when we would otherwise be forced to admit outright defeat. 

(1) Map showing the approximate locations of the “sixteen perfect lands” mentioned in the Avesta. Note the proposed location of Airyanəm Vaēǰah in Chorasmia (Khwārazm) (2) Current distribution of Pashto with superimposed political borders

Darmesteter’s monumental work (1890) proved that Pashto is, indeed, an Iranian language, descended from Avestan or another Old Iranian dialect, having separated from Persian before the Pahlavi or Middle Persian period. Grierson cites Darmesteter and states (1921:5) that Pashto is a Medic, or Non-Persic, or eastern Iranian language. The Indic-look features are accounted for by the fact that it has borrowed largely and freely from North Western India but, in its essence, it is an Erānian tongue (p. 9; see also Meillet 1922:44-6). The connections with Saka and other Eastern Iranian languages appear to be areal features rather than ancestral; Khotanese Saka is sufficiently distant from Pashto and closer to modern Wakhi language in the Wakhān Corridor. 

Linguists have more recently placed Avestan in an intermediate group somewhere between Western and Eastern Iranian where it was previously classified as Eastern; it appears to have been recorded prior to the split which gave rise to these branches. It was long thought that Avestan represented “Old Bactrian”, but this notion had “rightly fallen into discredit by the end of the 19th century”. It is the opinion of this author, in agreement with Elton L. Daniel, Bahram Farahvashi and Naser Takmil Homayoun, that the homeland of Avestan (Airyanəm Vaēǰah) was most likely in Khwārazm. Indeed, Ali-Akbar Dehkhoda once called Khwārazm “the cradle of the Aryan tribe.” 

Contrarily, Pashto is manifestly an Eastern Iranian language, joined in that clade by historically prominent literary languages such as Khwarezmian, Sogdian, Bactrian and Khotanese Saka, with which it once must have formed an areal group. The modern Eastern Iranian languages that join Pashto such as Pamiri languages, Munji-Yidgha, Ossetian and Ormuri-Parachi by contrast appear to be descendants of historically unattested Eastern Iranian dialects whose remoteness allowed them to resist the domination of Persian. This would suggest that Pashto, like the aforementioned modern Eastern Iranian tongues, was relatively isolated. If we are to accept Mallory’s proposal regarding the Yaz culture in Margiana (Marw) as the archaeological remnants of the early “Eastern Iranian” culture as described in the Avesta, the ancestor of Pashto once traveled by way of Margiana into Bactria and thence perhaps the Badakhshan valley and Hindu Kush before arriving in the Indus valley. From the 3rd century CE onward, they are mostly referred to by the name Afghān (Abgān). 

Interestingly, it appears the name Pashto is derived from the same Old Iranian stem that gave Pārsi “Persian” and perhaps also Parthian, namely *Parsuwā ‘frontier’, “borderlands’ (cf. 𐎶𐎠𐎭 *Māda ‘Media”, ultimately from Proto-Indo-European *médʰyos (“middle”)  meaning ‘middle country’, ‘middle kingdom’) and Pashtun from *Parswāna-, with the basic stem *Parsū (Untersuchungen zur Geschichte von Eran, Göttingen and Leipzig, 1896-1905, II, p. 177; cf. Morg[enstierne], par. 40b). Perhaps this demonym was once used by a conglomeration of Iranian tribes inhabiting a geographic and/or ethno-linguistic frontier in Transoxiana, wherefrom the name spread widely over different regions following demic diffusions. It is more tenable however that the term has been applied broadly by different groups in different periods; for example, when Western Iranian speakers had already established themselves in the southern Zagros mountains at the gates of Elam and the Persian gulf; a true ‘frontier’ in more than one sense. This hypothesis is buttressed by the parallel usage of Māda ‘Media’ (lit. “middle lands”) by Iranic speakers further north of *Parsuwā.

Although there are numerous different dialects, Pashto is essentially one language (with one possible exception, Waṇecī). Moreover, modern broadcast channels have closed any gap of unintelligibility among dialects. Due to overlaps among various isoglosses, it is difficult to establish a satisfactory classification of Pashto dialects, thus the dialects presenting a further development of common Pashto vocalism may belong either to the “soft” or the “hard” group. Moreover the modern “hard” pronunciation of ҳt as xt is restricted to the northeastern dialects and evidently of recent origin, as evidenced inter alia by the orthography. By contrast the western pronunciation centered in Kandahar, used as standard in this article, retains the archaic phonology ʂt.

Stage	Notes	Contact Languages	Localization	Era
Old Iranian	Akin to Avestan	The unknown language of BMAC	Khwarezm ?Bactria-Margiana?	Mid 2nd millennium BC to early 1st millennium BC
Middle Iranian	Development of  “Eastern Iranian” and later contacts with East Iranian areal group	Sogdian, Khwarezmian, Bactrian, (Khotanese?)	Yaz culture of Bactria -Margiana? ThenBadakhshān valley?	Mid to late 1st millennium BC to 1st millennium AD
Middle Iranian	Contacts with Dardic	Dardic languages	Kashmir, Khyber Valley, Kohistān	1st millennium AD
Middle Iranian and Neo-Iranian	Contacts with Indo-Aryan; prescriptive influences of Sassanian Pahlavi and later classical Persian; contacts with Tajiks Fārsiwān	Lahnda dialects (Western Punjabi); literary Persian, Afghan Persian (Tajik)	Khyber Valley, Punjab and Sindh, later Peshāwar and Kandahār	1st millennium BC – present

Phonotactic and lexical influences from Indic (Indo-Aryan) appear to be both ancient and profound, and have in turn given Pashto the impression of an Indic language. Lexical items are diverse and include flora, fauna, quotidian objects and concepts as well as abstractions (rōǧ “health”, max ‘face’ from Skr.  mukha, byal from Skr. bheeda “away, separated”). Words with retroflex dentals ʈ, ɖ must be regarded as borrowings from Indo-Aryan; conversely retroflex r and n were originally internal innovations but were likely reinforced by Indo-Aryan and included among later borrowings. This is not to say that either (1) all instances of these phonemes in Pashto must occur in borrowed Indic words, or (2) all examples of ɳ / ṛ / etc. came from Avestan rather than Indic. Moreover, Pashto /ʂ/, /ʐ/ ɳ / and / ṛ / are all developed from Avestan. Pashto ʂ is a reflex of Avestan sr, rs, r, š corresponding to PIE *kś. The Pashto liquid / ṛ / developed from Av. rt, rd, and the nasal from rn. ʈ and ɖ, and kh, however, developed from Indic languages. There must have been prolonged and intimate contacts (perhaps spanning centuries) between ancestral Pashto speakers and speakers of so-called Lahndi or western Punjabi dialects in the “North Indian cultural realm”, as well as with other unknown prakrits close to Sanskrit. Indeed, some Indo-Aryan lexica in Pashto are closer to Sanskrit than to northern Prakrits. This suggests a relatively early date of arrival into the Hindu Kush region. The influence of Dardic has been comparatively meager; given Pashtun dominance in the region the trajectory has primarily been from Pashto to Dardic. 

“It is impossible…to suggest in any but the most general way how the retroflexes have developed in the Iranian words in which they are found in such languages as Pashto and Yidgha. Bilingualism, involving Indo-Aryan languages, must be the answer, but no historical or social details can be given to elucidate the process.” -Emenau 1965

Distinctive cultural materials among Pashtuns, including bangṛi ‘bangles’, the partug qamis, and the tradition of facial jewelry must have been adopted within this milieu of exchange with the peoples of Sindh and Punjab. Pashtun women use a variety of jewelry such as paizwan (suspended below the nasal ala), natkai (large nose rings), chargul, pita and maikhakay (small nose ornaments) as well as sangley (pazaib) worn as ankle bracelets. Despite these influences, Pashto remains unmistakably Eastern Iranian character, and like Ossetian has retained a remarkable number of characteristics from Old Iranian–ergativity, use of prepositions with postpositions, gender distinctions, initial consonant clusters– in comparison to Persian. Moreover it seems clear that despite its many superficial resemblances to Indic languages, Pashto is an Iranian language (one of the most conservative); and that the only features in it which cannot be explained by direct reference to Avestan are the stops /ʈ/ and /ɖ/. 

Paṣ̌tūn haγa na daγ či Paṣ̌tō wāyi, lekin haγa či Paṣ̌tō ləri 

“A Pashtun is not he who speaks Pashto, but he who has Pashtunwali”

List of Pashto words and their Avestan Equivalents

A selection of Pashto lexical items with self-apparent Avestan correspondences is listed below: 

anā – “grandmother’ G. < Av. 𐬵𐬀𐬥𐬁 hanā ‘old woman’

āra ‘need, request’; āriya ‘at variance with, disputing’ G. < Av. *arǝθa ‘process, dispute’

bar, ‘victorious’ < Av. upara- ‘higher, superior’

bǝl “other, second’ < Av. bitya–

ǧel ‘thief’ < Av.  gadha– 

ǧuna – hair, color [of the skin]  < Av. gaona– hair, color of the hair

ǧarai– windpipe, throat < Av. garah-

ǧaw ‘noise, brawl’ < Av. gav- ‘shouting’

jinai ‘young girl’ < Av. janai– ‘woman’

kašr ‘younger, junior’ <Av. kasu–

lar ‘lower, below’ <Av. adhara

las ‘ten’ < Av. dasa–

mīna ‘love, affection’, mayən ‘in love, a lover’ < Av.. mayā– ‘pleasure, bliss’; mayah ‘coition” 

maira ‘stepmother’ <Av. māthryā

marǧe ‘bird’ < Av. merega–

ōṣ̌a ‘tear’ < Av. arsu-

ōspīna, ōspana ‘iron’ G. <Av. *ayo spaēnǝm *haosafna-

ōmǝ ‘name of a plant’ G. < Av. haoma–

ōr ‘flame’ G. < Av. aθr 

ōrǝ “cloud” G. < Av. awra

paī ‘milk’ < Av. payah–

pēẓandel ‘to recognize’ <Av. paiti-zan “to recognize”

plān ‘wide’ < Av. pathana–

pēǧla ‘maid, virgin’ < Av. *payō–gatā 

pāṇa ‘leaf’ < Av. parena–

pōr ‘debt’ G. < Av. pāra “guilt”

pse ‘general name for goats and sheep’ < Av. pasu

spai ‘dog ‘ < Av. span–

sara ‘together with’ < Av. sar– ‘union’

saṛai ‘man’ < Av. sarəidya ‘fellow’ 

star ‘big’ < Av. *stura ‘large, big’

stōrai ‘star’ < Av. stār– 

starga ‘eye, planet’ < Av. stere

seẓai ‘lung’ < Av. sushi 

ṣ̌e “ good” < Av. sraya- “more beautiful” , srao– ‘good’ 

trə “paternal or maternal uncle’ < Av. turiya–

tōr ‘black’ < Av. taθra

tar ‘until, to, from “ablative” < Av. tarō

taš ‘empty’ < Av. tusa “to be empty’

taṣ̌tēdal ‘to flee’ < Av. tarshti– flight

uṣ̌ ‘camel’ < Av. ustra– 

wana ‘tree’ < Av. vanā

wrōst ‘rotten, decayed’ < Av. *frista– ‘to decay, putrify’ 

waṛai ‘wool’ < Av. varenā–

wəṛ ‘successful, winning’ Cf. Av. wərəd ‘to augment’

wāwra ‘snow’ < Av. vafra–

wāyəl ‘to speak’ probably < *wafya cf. Av. vaf– ‘to sing’

xpal ‘own, self’ < Av. xwaēpaiθya– ‘own’

yašna ‘boiling’ < Av. yaēšyat– ‘boiling’ 

zmaka ‘ground, earth’ < Av. zam–

zimai ‘winter’ < Av. zim–

zṛə ‘heart’ < Av. zərəd  < *zrdaya