ArticlePDF Available

Nuristani Theonyms in Light of Historical Phonology

January 2023

January 2023

Authors:

University of Wuerzburg

A contribution to the Proceedings of the conference "Roots of Peristan: The Pre-Islamic Cultures of the Hindukush/Karakorum" (Rome, ISMEO, 2022). This PDF-version contains comments giving updates and corrections to the published version.

Content uploaded by Jakob Halfmann

Content may be subject to copyright.

S E R I E O R I E N T A L E R O M A

n.s. 37

ROOTS OF PERISTAN

THE PRE-ISLAMIC CULTURES

OF THE HINDUKUSH/KARAKORUM

Proceedings of the International

Interdisciplinary Conference

ISMEO, Rome, Palazzo Baleani, 5-7 October, 2022

Part I

edited by Alberto M. Cacopardo & Augusto S. Cacopardo

ROMA

ISMEO

2023

ISMEO

ASSOCIAZIONE INTERNAZIONALE

DI STUDI SUL MEDITERRANEO E L’ORIENTE

SERIE ORIENTALE ROMA

FONDATA NEL 1950 DA GIUSEPPE TUCCI

DIRETTA DAL 1979 DA GHERARDO GNOLI

Scientific Board:

Timothy H. Barrett, East Asian History, School of Or. and African Studies, London

Alessandro Bausi, Äthiopistik, Asien-Afrika-Institut, Universität Hamburg

Peter Kornicki, East Asian Studies, Cambridge University

Daniel Potts, Ancient Near Eastern Archaeology and History, Inst. for the Study

of the Ancient World, New York University

Editor: Adriano V. Rossi

NUOVA SERIE

Vol. 37

R O M A

ISMEO

2023

S E R I E O R I E N T A L E R O M A

n.s. 37

ROOTS OF PERISTAN

THE PRE-ISLAMIC CULTURES

OF THE HINDUKUSH/KARAKORUM

Proceedings of the International

Interdisciplinary Conference

ISMEO, Rome, Palazzo Baleani, 5-7 October, 2022

Part I

edited by Alberto M. Cacopardo & Augusto S. Cacopardo

ROMA

ISMEO

2023

This volume is published with a grant from the MUR Project “Storia, lingue

e culture dei paesi asiatici e africani: ricerca scientifica, promozione e

divulgazione” CUP B85F21002660001.

ISSN 0582-7906

ISBN 978-88-66872-65-8

via Malladra 33 – 00157 Roma

e-mail: info@scienzeelettere.com

www.scienzeelettere.com

www.ismeo.eu

CONTENTS

Foreword by Adriano V. Rossi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Preface by Alberto M. Cacopardo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Scientific Committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

List of Speakers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviii

List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx

PART I

ETHNOGRAPHY

Alberto M. Cacopardo, Who Claimed the Golden Throne. The Political

History of the Eastern Kati and Their Settlement in Gobor . . . . . . . 3

Thomas Crowley, The Politics of Peristan . . . . . . . . . . . . . . . . . . . . . . . 53

Inayat Ullah Faizi, Fairies Commanding Resource Management: Myths

in Pre-Islamic Peristan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Ulrik Høj Johnsen, Captivating Roots—Connectivity across Time and

Place . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Tatch Sharakat Kalasha (Taj Khan Kalash), O may Jahil Baba, ‘Oh My

Ignorant Pagan Sister:’ A Traditional Kalasha Funeral and Feast in

the Twenty-First Century . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Max Klimburg, High Status and Sexuality Symbolism in Kafiristan . . . 131

Homayun Sidky, Differentiating Shamans from Other Ritual Intercessors

in Hunza: A Theoretical Perspective . . . . . . . . . . . . . . . . . . . . . . . . 151

Birgitte Glavind Sperber, The Pure Goats of the Kalasha . . . . . . . . . . . 165

Irmtraud Stellrecht, Gaining Renown in Hunza (Northern Pakistan):

Remembering Bóot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

Karl Wutt, Symbols in Pashai Culture . . . . . . . . . . . . . . . . . . . . . . . . . . 223

LINGUISTICS

Elena Bashir, Copular Use in Khowar, Kalasha, and Neighbouring

Languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

Almuth Degener & Irén Hegedűs, Spatial Morphemes in Nuristani . . . 277

Pierpaolo Di Carlo, Lineages, Rituals, and Gods among the Kalasha

of Birir. Textual Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295

Jakob Halfmann, Nuristani Theonyms in Light of Historical Phonology 317

Jan Heegård Petersen, Boundaries in Grammar, Landscape, and Cult-

ure: Kalasha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359

Anton I. Kogan, The Linguistic Evidence on the Ethnic Composition of

pre-Tibetan Ladakh and Baltistan: A.H. Francke’s ‘Dardic’ Hypo-

thesis Revisited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381

Henrik Liljegren, The Languages of Peristan through the Lens of Areal

Typology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391

Robert C. Tegethoff & Sviatoslav I. Kaverin, Documenting Grangali . . 433

PART II

HISTORY

Michele Bernardini, Timur and the Siyāhpūšān (1398-1399) . . . . . . . . 455

Paul Bucherer-Dietschi, A Recently Discovered Manuscript of Munshi

Syed Shah’s Second Visit to Kafiristan in 1888-1889 . . . . . . . . . . . . 463

Jürgen Wasim Frembgen & Ejazullah Baig, Animal Motifs on a Wooden

Box from Hunza (Karakoram) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479

Nile Green, Christian Missionaries, Pashtun Middlemen, and the

Attempt to Convert Kafiristan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491

Wolfgang Holzwarth, Eastern Hindukush and Karakorum, 1000 to

1880. With Special Regard to the Orbit of Kashmir . . . . . . . . . . . . . 517

Hermann Kreutzmann, Approaching Kafiristan. Colonial Exploration

and Cartographic Representation in the Pamirian Crossroads . . . . 559

John Mock, On the Title Tham of Hunza Rulers . . . . . . . . . . . . . . . . . . 591

Stefano Pellò, Bābā Siyar and the Soul of the Unbeliever: Mapping the

Persian Poetic Territories of Kāfiristān from Constantinople to Mada-

klasht . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613

Hidayat ur Rahman, Shah Khairullah: Relations and Policy towards the

Kafirs of the Hindu Kush . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631

Ilaria E. Scerrato, Heritage from Peristan: Observations on Zoomorphic

Decorations in the Graveyards of Swat Kohistan and Indus Kohistan 645

ANTIQUITY

John Mock, Tibetans in Gilgit and Wakhan. New Data, New Implications 669

John Mock, On the Onomastics of Shri Badat, the Cannibal King of Gilgit 687

Cristiano Moscatelli & Anna Filigenzi, Non-Buddhist Customs of Bud-

dhist People II: A Shifting Perspective on Wine, Goat Deities and

Connected “Dardic” Themes in Gandharan Art . . . . . . . . . . . . . . . . 703

Jason Neelis, From Khāśarājya to Daraddeśa—Steps towards a Re-

gional Macrohistory of Peristan in the First Millennium CE . . . . . . 727

Luca M. Olivieri & Matteo Sesana, A Tale of Two Cities. Notes on the

Two Barikot in Dir and Swat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743

Abdul Samad & Ruth Young, Pre-Islamic Archaeology and Heritage in

Chitral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759

Richard F. Strand, Kâmboǰâs and Sakas in the Holly-Oak Mountains:

On the Origins of the Nûristânîs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781

COMPARISONS

Augusto S. Cacopardo, The Kalasha of Chitral and the Brokpa of

Dah/Hanu. A Cultural Comparison . . . . . . . . . . . . . . . . . . . . . . . . . 811

Jadwiga Pstrusińska, Multidisciplinary Remarks on Some

Hindukush-Karakorum Ethnonyms in Eurasiatic Perspective and a

Comparative Analysis of One Case Study . . . . . . . . . . . . . . . . . . . . 837

Bernard Sergent, Les dragons du solstice d’hiver . . . . . . . . . . . . . . . . . 847

Michael Witzel, Indra among the Kalasha . . . . . . . . . . . . . . . . . . . . . . 859

Sergey A. Yatsenko & Sviatoslav I. Kaverin, Traditional Costume of the

Eastern Hindukush Area: Some Problems of Its Origin and Modification 875

Claus Peter Zoller, In Search for the Roots of Peristan . . . . . . . . . . . . . 903

Nuristani Theonyms in Light of Historical Phonology1

JAKOB HALFMANN

SUMMARY. This paper re-examines established ideas about the etymologies of the re-

ligious terminology of pre-Islamic Nuristan, in particular the names of gods. After a

detailed discussion of a number of selected terms, the paper concludes that the general-

izations made by Fussman (1977; 2012) about the pre-Islamic religion of Nuristan rep-

resenting an independently inherited survival of Proto-Indo-Iranian religion cannot be

upheld, since most of the relevant terms are in fact post-Vedic borrowings from Indo-

Aryan languages, which implies a closer connection with classical Hinduism than was

previously assumed.

The Nuristani languages have attracted attention ever since Georg Mor-

genstierne established their unique position as a likely independent third branch

of Indo-Iranian. In a frequently cited article, Fussman (1977) has tried to extend

this unique position of the Nuristani language family to the religious sphere,

claiming that the pre-Islamic religion of Nuristan, too, had preserved features

of ancient Indo-Iranian religion lost everywhere else and that these are assured

to be ancient by being associated with words that were independently inherited

from Proto-Indo-Iranian in the Nuristani languages. Others, e.g. Allen (1991)

and Tremblay (2015: 38-40), have continued research in this direction and Fuss-

man has reiterated his conclusions a few decades later, modifying only details

(Fussman 2012).

This article is intended as a re-examination of that view based on more

ample linguistic data and on the better understanding of Nuristani historical

1 This article is a significantly expanded version of my presentation at the “Roots of Peristan”

conference, which was based on a talk I had originally prepared for the Indo-European Research

Colloquium. I would like to thank Alberto Cacopardo for inviting me to repeat this talk in Rome

and for discussing with me the implications of its content, which has caused me to refine my ar-

gumentation in several points. I had already discussed the etymologies of three theonyms (Imró,

Mon, Bagíṣṭ) in my talk at the 2019 Cologne Nuristani Workshop, which finally appeared as a

paper in the workshop proceedings 4 years later (Halfmann 2023). There is consequently some

degree of overlap between the contents of that contribution and the present article.

phonology that we have today. Since Fussman’s theses are based on linguistic

arguments and rest on the separate linguistic status of the Nuristani language

family, this article will treat only the pre-Islamic religious terminology of the

languages in this family. This means that I will only marginally discuss terms

from the religion of the Chitral Kalasha, the only remaining non-Muslims in

the Hindu Kush, who speak an Indo-Aryan language. This limitation is some-

what artificial, but it is imposed by Fussman’s argument. It also requires an

anachronistic terminology, since the most precise available terms, Nuristan and

Nuristani, have only come into existence after the conquest and conversion to

Islam of Kafiristan, ‘the land of infidels.’ I must therefore speak of “pre-Islamic

Nuristan” and “the pre-Islamic religion of the speakers of Nuristani languages.”

These are oxymorons in themselves, but preferable to the term Kāfir ‘infidel,’

which would apply to a wider region not limited by linguistic affiliation.

BASIC ASSUMPTIONS

A basic tenet of historical linguistics is the idea that sound change is regular,

or, more precisely, that “[c]hange in pronunciation that is not conditioned by

non-phonetic factors is regular and operates without exceptions at a particular

time and in a particular speech community, with possible environmental re-

strictions”2 (Hock 2021: 46). This insight was the basis of the Neogrammarian

revolution, which established historical linguistics and etymology as a rigorous

field of study, in which hypotheses can be falsified.

The regularity of sound change is also what allows us to distinguish be-

tween words that have been regularly transmitted in a given language since the

time of a common ancestor language shared with other related languages and

words that entered into it from outside at a later time via borrowing. The latter

category of words will not have gone through the sound changes that had been

completed within the borrowing language before the time of borrowing. Instead

it will attest to the sound developments of the donor language and only from

the time of borrowing onwards it takes part in the sound changes of the bor-

rowing language.

With this basic principle in mind, there are a number of statements in Fuss-

man’s articles (1977; 2012) that cast some doubt on the soundness of his lin-

guistic argumentation. One example is the claim that “difficulty to explain the

vocalism” (la difficulté d’expliquer le vocalisme) or the lack of an accepted

etymology can be taken as arguments for the direct inheritance of a Nuristani

word from Proto-Indo-Iranian (Fussman 2012: 75). We cannot assume that a

word is necessarily inherited, just because the etymology of a word has not

been established and if the quality of a vowel is difficult to explain from its

2 I.e. restrictions to specific phonological environments.

318 Jakob Halfmann

etymological origin then this only means that the vowel changes in question

have not been worked out in detail. There is certainly no reason to believe that

vowel changes proceed in a more irregular fashion in some language families

than in others. Since sound change is generally an unconscious process, which

follows the same principles in any spoken language, it is entirely irrelevant

“that the old Kafirs did not have any notion of historical phonology and that

these historical-linguistic considerations were completely foreign to them”3

(Fussman 2012: 73). The same would apply to the average Frenchman or

-woman, and yet their language changes over time.

The fact that the sound developments of the Nuristani languages have not

yet been worked out in detail (despite the admirable foundations set down by

Georg Morgenstierne) makes it somewhat difficult to discuss the issues treated

in this paper. I will try whenever possible to provide parallel examples for a sound

change claimed to be regular, but in most cases I cannot give the full amount of

supporting data. This issue will hopefully be remedied in the near future, when

my notes on Nuristani historical phonology have been published, which will in-

clude a more complete set of supporting lexical data for each conclusion.

With regard to the theoretical foundations for the borrowing of theonyms,

the field of onomastics has apparently not advanced very far: Greule (2013:

17), while trying to establish “theolinguistics” as new field of study, claims

that “nothing is known so far about the borrowing of theonyms.”4

INITIAL OBSERVATIONS

A central linguistic fact that is often cited as evidence of the polar opposi-

tion between the religions of the speakers of Iranian and Indo-Aryan languages

is their divergent interpretation of the Proto-Indo-Iranian term *daiua-. In Indo-

Aryan languages, descendants of *daiua- generally mean ‘god,’ referring to a

positive concept of a god worthy of worship. In Iranian languages, on the other

hand, descendants of the same term *daiua- generally mean ‘demon,’ referring

to evil supernatural beings that must be warded off. In the light of this opposi-

tion, it is first of all interesting to note that the descendants of *daiua- in the

Nuristani languages—Katë de, Nuristani Kalasha de, Ashkun de, Prasun lu—

generally had a positive meaning in pre-Islamic times, akin to that found in the

Indo-Aryan languages. Though the Nuristanis are today all Muslims and no

longer worship the beings once referred to with these terms, the words survive

in some forms with positive associations. Thus, e.g., the Katë word dé-vani (lit.

‘god-associate,’ though no longer understood as such) can still be heard as an

honorific term of address.

3 “[…] que les anciens Kafirs n’avaient aucune notion de phonétique historique et que ces

considérations historico-linguistiques leur étaient tout à fait étrangères.”

4 “Über die Entlehnung von Theonymen ist bislang nichts bekannt.”

Nuristani Theonyms in Light of Historical Phonology 319

The Prasun form lu exhibits the likely very early sound change *d > l,

which makes it probable that the term is a genuine cognate of Old Indo-Aryan

deva- and Avestan daēuua- inherited from Proto-Indo-Iranian, rather than a

later borrowing from Indo-Aryan. The positive meaning of the Nuristani de-

scendants of *daiua- does not, however, necessarily point to a closer connec-

tion, either linguistic or religious, of Nuristani with Indo-Aryan, since the

Iranian meaning is an innovation probably brought about by Zoroastrianism

that diverges from the earlier meaning ‘god’ reflected in the wider Indo-Euro-

pean cognates (Latin deus, Lithuanian dievas etc.). The Nuristani languages,

as in many cases, simply do not participate in the innovation, which at most

means that they were not spoken within the sphere of influence of Zoroastrian-

ism at the relevant time.

This agrees with the fact that no traces of Zoroastrian teminology have so

far been detected in the general lexicon of the Nuristani languages. Such traces

can be found, e.g., in Khotanese, despite the Buddhist faith of the speakers at

the time of attestation, or in modern Sanglechi, despite the present-day Islamic

(Ismaili) faith of the speakers (Khotanese urmasyde, Sanglechi ormozd ‘sun’

< *ahura-mazdāh). The speakers of the Nuristani languages were clearly not

Zoroastrians at the time of European contact in the 19th century, but the lack of

any such Zoroastrian terminology probably even allows us to conclude that

there never was a significant number of Zoroastrians among them.

Another fundamental statement that can be made about the pre-Islamic ter-

minology of the Nuristani languages is that it does not appear to be particularly

influenced by Buddhism, which is perhaps somewhat surprising, since Bud-

dhism flourished in neighboring regions in pre-Islamic times.

What remains therefore, is the question of how much the pre-Islamic ter-

minology of Nuristan has in common with either reconstructed Indo-Iranian

religious terminology or with later forms of more specifically Indo-Aryan or

Hindu terminology. In the following sections, I will try to contribute to the ef-

fort to answer this question by treating first the etymologies of the names of

the most important deities and then some central terms for religious functions

and concepts. For reasons of time and space, I have not been able to discuss all

of the documented deities or religious terms in this paper, but I hope that the

included examples provide a clear enough picture of the overall situation.

NAMES OF DEITIES5

The available sources on the pre-Islamic pantheon of Nuristan are of vary-

ing reliability when it comes to the representation of the Nuristani sound sys-

tems. The earliest Western sources, written by British colonial officers with no

5 Abbreviations in this section: Kt.—Katë, NE—Northeastern dialect, SE—Southeastern dia-

lect, W—Western dialect; NKal.—Nuristani Kalasha, Z—dialect of Zhönchigal, N—dialect of Nis-

heygram; A.—Ashkun, W—dialect of Wama; Pr.—Prasun; IAKal.—Indo-Aryan Kalasha;

320 Jakob Halfmann

linguistic training, are especially unhelpful, but even more recent sources are

often inconsistent in their transcription of unfamiliar sounds. To establish the

correct phonological form of the different names I have relied, wherever poss-

ible, on sources produced by linguists and native speakers, but in some cases

we still only have approximations of the real phonological forms.

The deities treated in this paper are numbered from 1.) to 7.). In each case,

I first note the forms of the name attested in the different Nuristani languages,

then I move on to the characterstics of the deity and finally to the etymology

of the name.

1.)

Kt. NE Imró [<Iмрō> (Grjunberg 1994: 85); < > (Cacopardo, Schmidt

2006: 56 [69])];6 W/SE Imró (Strand 1999a; 1999b); NE Yum (in hymns) (Mor-

genstierne 1951: 179).

NKal. Z Yamrái (Tāza 2017); N Yamrä (Degener 1998: 571), Yamarä

(Strand 1999c).

A. Imrā (Morgenstierne 1951: 163).

Pr. Māra (Buddruss, Degener 2016: 717); Mənǰəm Māra, Mənǰəm Malik

(Buddruss, Degener 2016: 716).

According to Robertson’s (1896: 381) testimony from Kamdesh, “Imrá is

the creator of all things in heaven and earth. By the breath of his mouth he en-

dowed with life his ‘prophets’ Moní, Gish, Satarám, and the rest.” Andreev

summarizes the statements of his Eastern Katë informants with the statement

that “Imro is considered the highest deity, the chief of the gods, creator and

master of the universe” (Grjunberg 1994: 87). Giṣ is named as his brother, Vutr

or Arti-Vutr as their mother, Artyo as their grandfather (Grjunberg 1994: 86-

87). In hymns he is addressed as Kt. délū móči mror ‘lord of gods and men’

(Morgenstierne 1951: 179; Grjunberg 1980: 89) and described as dressed in

items made of gold (Grjunberg 1995a: 147-159). Surviving wooden statues

show him riding a horse and carrying a whip (Klimburg 2002: 54, fig. 3). He

was worshipped in temples “in every village” (Robertson 1896: 389), but es-

pecially in the great temple at Ušüt (Kštöki) in Parun (cf. Castenfeldt 1996 for

a reconstruction of the temple). Next to the temple there was a “famous hole”

leading to the underworld, where horses were sacrificed (Robertson 1896: 393;

Edelberg 1972: 37). The temple was called Pr. Mareš iré ‘Māra’s hall’ or

OIA—Old Indo-Aryan; MIA—Middle Indo-Aryan; Gandh.—Gandhari; Prs.—Persian; Av.—

Avestan; PIIr.—Proto-Indo-Iranian; PIE—Proto-Indo-European. Language names follow the con-

ventions of Halfmann (2021). Asterisks mark reconstructed or unattested forms, < > marks

transcriptions that are not necessarily phonemic.

6 Morgenstierne (1951: 163) noted the name as “Imˈrō (-řō?),” but it is never pronounced or

written with /r/ by native speakers. The two phonemes /r/ and /r/ are frequently confused in Mor-

genstierne’s transcriptions.

Nuristani Theonyms in Light of Historical Phonology 321

Mənǰəm iré (Zamān Khān, Strand 1998; Buddruss, Degener 2016: 564),7 but

also “great Mecca”—Pr. Öl Maká (Zamān Khān, Strand 1998), Kt. NE Al Mokó

(Grjunberg 1995a: 147-159),8 being equated to Mecca as the most sacred pil-

grimage site. In the same way, Imró himself had also become equated with the

monotheistic God of Islam (cf. Cacopardo, Schmidt 2006: 56-57 [69]). By the

time of conversion, the two were considered synonymous to such a degree that

Imró (beside Allāh and Ḫudā(y)) is still an acceptable word for ‘God’ among

the Muslim Katë today.

Among the Nuristani Kalasha Yamrái / Yam(a)rä was not the highest god

or the creator, but rather the god of death and the underworld (Klimburg 1999:

142-143).9 Curses such as Yamrä grā-dío tu ‘may Yamrä take you away’ have

been recorded in Nisheygram (Degener 1998: 435). He was even said to eat

corpses (Klimburg 1999: 142). In Ashkun-speaking areas he was also connected

with the passing of the seasons (Klimburg 1999: 143). Both Nuristani Kalasha

and Ashkun speakers associated worship of Yamrái/Imrā especially with the

Prasun and Katë speaking areas (ibid.: 141-142).

Morgenstierne (1951: 163) connected the names Kt. Imró, NKal. Yamrái,

Yamrä and A. Imrā with OIA Yama-rājan- ‘King Yama.’ This etymology is

convincing and impressively supported by the simplex Yum ~ OIA Yama- pre-

served in Katë religious hymns.10 Fussman (1977) drew his own conclusions

from this etymological connection and argued that Kt. Imró, NKal. Yamrái,

Yamrä and A. Imrā must be independently inherited Nuristani reflexes of the

mythological character called Yamá- in Old Indo-Aryan and Yima- in Avestan,

claiming more recently that “the only explanation that one could give today of

the name of Imrā and of the character is to assume a development from com-

mon Indo-Iranian.”11 From this he drew the implication that the role of Imró in

7 Buddruss, Degener (2016: 564) seem to imply that the temple of Ušüt (Kštöki) was distinct

from the iré ‘assembly hall’/‘dancing house’ (Tanzhaus) of the village and Klimburg (2002) also

treats the temple and the village assembly halls in separate sections, but at the same time the

temple is identified in its name as an iré and Klimburg (2002: 61) notes that there was only one

iré in each village, each of which “served also as a temple for the leading village deity.”

8 Grjunberg (1995a: 150) interprets Andreev’s transcription <Ал Мōкōсар> as a single word

Almukósar ‘paradise’ with variants Almoko and Moko, but the phrase transcribed by Andreev as

<Ал Мōкōсар бэ´нцjош`> is more likely to be read as/Al Mokō=stan bénċyoš/ ‘you appeared out

of the great Mecca.’

9 Klimburg’s (1999: 142-143) statements on “mara” are somewhat misleading, since NKal.

Z marə˜

´, N mará is simply a common noun meaning ‘death’ (cognate with Kt. NE mëre˜, SE marë

‘death’ and corresponding to OIA maraṇa- ‘death’) with no direct relation to either the Buddhist

Māra or the Paruni deity Māra.

10 The development of ya > i in Kt. Imró can be explained as a result of the syncope of pre-

tonic vowels followed by *yC > iC, cf. Kt. itë ‘pair, suit’ < *iukta-ka- ‘something joined together.’

The form Yum shows the regular development of *a in stressed monosyllables > Kt. u, cf. dut

‘tooth’ < *dant-a-.

11 “La seule explication que l’on puisse aujourd’hui donner du nom d’Imrā et du personnage

est de postuler une évolution à partir de l’indo-iranien commun” (Fussman 2012: 81).

322 Jakob Halfmann

pre-Islamic Nuristan could be used as crucial evidence for the reconstruction

of the original function of this mythological figure. Many researchers have

since followed this assessment and it is today generally accepted (cf. Söhnen-

Thieme 2009: 814; Degener 2002: 114; Cacopardo 2016a: 253, fn. 18).

In Halfmann (2023: 118), I argued by contrast that inheritance of these

names from Indo-Iranian times is very unlikely, based on their forms alone.

The simplex form *iamá can safely be reconstructed for Proto-Indo-Iranian,12

both in the meaning ‘twin’ and as the name of a particular mythological figure

named ‘twin’ (Mayrhofer 1996: 400-401). Notably however, the OIA word

rājan- has no direct cognate in Iranian and though Yamá-/Yima- was imagined

as a king in both the Vedas and the Avesta, the particular compound Yama-

rājan- ‘King Yama’ is a post-Vedic innovation only found on the Indo-Aryan

side.13 It is therefore essentially impossible that it should have been indepen-

dently inherited in this form in another branch of Indo-Iranian. This is also ex-

cluded by the development of the palatal affricate in rājan-: One of the basic

isoglosses separating Nuristani from Indo-Aryan is the preservation of the

contrast between Indo-European palatovelars (reflected as dental affricates in

Proto-Nuristani) and secondarily palatalized velars (reflected as palatal affri-

cates in Proto-Nuristani), as against the merger of the two sound classes in

Indo-Aryan. Since OIA rājan- derives from the Indo-European root *h3reǵ- ‘to

straighten, rule’ (cf. e.g. Latin rex ‘king’) with a palatovelar *ǵ, it should have

a Nuristani correspondence with reflexes of a dental affricate *ȷ , if it were an

inherited word. The names Imró, Yamrái etc. do not show such a development

and they even point to borrowing from Middle Indo-Aryan Gandhari rather

than Old Indo-Aryan, because rājan- is reflected not with a palatal affricate but

with a y (dropped from word-final -āy in Kt. and A.), the regular development

of intervocalic j in Gandhari. The Gandhari word rāya ‘king’ has also been bor-

rowed on its own, leading to Kt. aró, NKal erā ‘rich.’14 The prothetic vowels

a-, e- are loan adaptations, since all instances of original word-initial *r- in

these languages had developed into approximants or fricatives, as in Kt. ru;

NKal.(z.) ẓū, (n.) wru; A. ẓo ‘intestines’ < PII *rauta- (cf. Morgenstierne 1933).

12 The form Yima- is owed to particular sound developments of Avestan, which must be dis-

tinguished from the ancestor of all Iranian languages. Persian J

am-šēd and Bactrian Iam-šo show

that the Proto-Iranian form was still *iama-. Fussman’s (2012: 79) consideration of whether or

not Kt. Imró or Yum could be derived from Yima- is therefore misplaced.

13 The compound Yamá-rājan- is only used in the meaning ‘having Yama as king, Yama’s

subject’ in Vedic, whereas the compound Yama-rājan- with the meaning ‘King Yama’ is first at-

tested in the Bhāgavata Purāṇa (Böhtlingk, Roth 1855).

14 The connection of Kt. aró with Gandh. rāya is supported by the feminine form of the

Southeastern dialect arëni ~ OIA rājñī ‘queen.’ Synchronically, the connection between Imró and

aró is not apparent and Imró is no longer analyzable as a compound. Fussman’s (2012: 79-80)

attempts to prove that -ró in Imró does not mean ‘rich’ are therefore unnecessary. His observation

that “a form *Rōim” would not be unpronounceable (Fussman 2012: 80) is beside the point. The

phrase ‘rich Yama,’ which he intended to construct, would have the form aró Yum, not *Rōim.

Nuristani Theonyms in Light of Historical Phonology 323

The god referred to as Yama-rājan- in post-Vedic Indo-Aryan, is no longer

the Yamá- of the Rgveda, the first mortal who rules the idyllic realm of the an-

cestors (Söhnen-Thieme 2009: 807-808), but more and more exclusively the lord

of death and the (often dreadful) underworld (Söhnen-Thieme 2009: 810-811).

This image agrees very well with the reported pre-Islamic beliefs of the Nuristani

Kalasha about Yamrái/Yam(a)rä. An incongruence between the form of the name

and the concept of the deity appears only in the beliefs of Katë speakers about

Imró. From the linguistic perspective it seems therefore much more likely that

the divergent conception of Kt. Imró is an innovation that has taken shape after

the name was borrowed from Middle Indo-Aryan, rather than the continuation

of an independently inherited concept of Proto-Indo-Iranian *Iamá-.

Prasun Mā ra15 is conceptually identified with Kt. Imró, but the two names

are probably not etymologically cognate. Morgenstierne’s (1951: 163) connec-

tion of Pr. Mā ra with Kt. mror has to be rejected in light of the recognition that

mror is not a proper name but only a regular noun meaning ‘lord, master’ (bor-

rowed into Prasun as muar). A connection with the Buddhist mythical figure

Māra, an opponent of the Buddha, which had been entertained by Morgen-

stierne (1951: 163), has led Jettmar (1975: 72-73) to ideas of an anti-Buddhist

religious reaction in Kafiristan. Augusto Cacopardo (2016a: 252) calls this the-

ory “quite conjectural, and even doubtful.” Phonologically the connection is

equally doubtful: The preservation of an unstressed final a in Prasun seems out

of the question.

It is possible that the name Māra is a borrowing from Gandhari mahārāya

< OIA mahā-rājan- ‘great king’ (cf. Parkes 1991: 89). As the simplex form

rāya has certainly been borrowed into the Nuristani languages (see above) it is

not unexpected that we should also find the higher rank. The same etymology

could then explain the personal name Kt. Maró, NKal. Marā, which would join

the group of royal titles used as elements of personal names in the pre-Islamic

onomastic tradition,16 in spite of the absence of actual kings in Nuristan. The

stress on the first syllable could be due to a re-analysis in analogy to newer

compounds with -ra (~ Gandh. rāya) as the second element, such as the per-

sonal name Külǘ-ra. The field consecrated to Māra—šē ütnyog—is explained

by Zamān Khān as “the king’s field.”17 Seen from this perspective, šē ütnyog

might be a possessive construction parallel to Marē munȷ¨ ‘Māra’s meadow,’

but formed with the Persian word šāh or the Bactrian <ϸαο> /šā/ ‘king.’ If Māra

is originally not a name, but a title of honor, this could explain the existence of

15 In this article I follow the basic form identified by Buddruss, Degener (2016: 717), but

the placement of stress and the indication of vowel length on both vowels is shown in various

ways in the available transcriptions.

16 E.g. Kt. Mer (< Prs. mīr), Malik (< Arab. malik), Kon (< Turk. xagan ~ xān) and Aró ‘rich’

(etymologically < Gandh. rāya, but not synchronically understood as such).

17 “Šé ütnyok kay monōzë? Mer ütnyok. Mérë=stë tul.”—“What does še ütnyok mean? King

ütnyok. The king’s field.” (Zamān Khān, Strand 1998).

324 Jakob Halfmann

further Mā ras, such as Paškire Māra.18 It does not help, however, in establish-

ing whether Māra and Imró originally bore the same name or whether they

were originally separate deities—an epithet ‘great king’ could easily have re-

placed the original name.

Complicated mythological entanglements connect Pr. Mā ra with another

name—Mənǰəm Malik in Prasun, and Mir Mara or Minǰ Mara in the Indo-

Aryan Kalasha language (Parkes 1991: 87). The Indo-Aryan speaking Kalasha

of Chitral remember the temple of Ušüt (Kštöki) as a temple to their god Ma-

handéo (Parkes 1991: 75), but they also tell stories of a giant lying in the under-

world, on whose navel rests a pillar that supports the middle world and the

heavens (ibid.: 86-87). The pillar reportedly runs through the site of the great

temple, causing earthquakes whenever the giant gets hungry, either because he

shakes it (Morgenstierne 1973: 157) or because his stomach rumbles (Parkes

1991: 87). This story has a striking parallel in a Prasun myth recorded by Bud-

druss, in which the son of Mənǰəm Malik is abandoned together with his twin

sister, is raised by Paškire Māra and finally returns to fight his father, whom

he buries alive with his head at the top of the Parun valley and his feet at the

bottom, spreading the valley over him on top and placing a house (warəg) in

Ušüt (Kštöki) on his navel (Buddruss, Degener 2016: 124-139). Near the end

of the story the narrator begins to refer also to the son as Mənǰəm Malik. The

belief that movements of Mənǰəm Malik cause earthquakes is likewise reported

by the narrator (ibid.: 139, fn. 500), though he could not say whether or not

the house was to be identified with the temple of Māra, as in the Chitral Ka-

lasha narratives (ibid.: 139, fn. 499). Klimburg (2002: 58) was told a version

of the story, in which it is identical with the clan house (amə´l) of the Pəgailyé

clan, which would fit with Robertson (1896: 393-394) encountering the “mir-

aculous iron bar” of Imró “in a house in the village” nearby the temple, but

not inside it. It is unclear whether the name Mənǰəm Malik belongs originally

to the buried father or to the son, who rather appears to be the hero of the story.

A different version of the myth has been recorded by Andreev in his glossary

(interpretation of Kt. words in square brackets): “ныцычы пэ´р or ныцōчы пэ´р

[first element = nëċó ‘area in permanent shade’]—Former abode of the Jуш

[yuṣ ‘demon’], the evil spirit, who would shake the earth. Now this spirit lies

under the earth and an enormous iron stake (чиму"ри штȳjа) [čiméni štyu˜—

‘iron pillar’ ?] is driven into his body (into the heart), the end of which comes

out of the earth in Прäjсjунгыл [Prasyṹgël ‘Parun’]. When the jуш [yuṣ] be-

gins to stir and earthquakes happen, the seven cats located at his head begin to

drink his brain and the seven dogs at his feet begin to gnaw on his feet and he

calms down” (Grjunberg 1995b: 610).

18 Buddruss, Degener (2016: 750) record Paškire Māra and Piškire Māra who are probably

the same deity. Zamān Khān told Strand about “Pˈaǰǧir Mˈârâ”—presented as “the lower Imró”

in accordance with the equation Pr. Mā ra = Kt. Imró—who may or may not be the same (Zamān

Khān, Strand 1998).

Nuristani Theonyms in Light of Historical Phonology 325

Buddruss, Degener (2016: 139, fn. 499) suspect that Mənǰəm Malik was

worshipped in the temple of Ušüt (Kštöki) before Māra, but another possibility

is that the two characters are identical. This would explain why Māra is also

referred to as Mənǰəm Māra (ibid.: 414-417), why the temple is known both as

Mareš irē and Mənǰəm irē, why Mənǰəm Malik gives orders to the other gods

and sits on a golden throne among them in a different myth (ibid.: 33-37), and

why Katë hymns to Imró contain the terms Malik, “Mīžˈom” (translated as “Pra-

sun”) and “ṣṓro”19 (Morgenstierne 1951: 179-180). Artə, the name of a meadow

at the confluence of the two headstreams of the Pech river identified as the lo-

cation of Mənǰəm Malik’s head (Buddruss, Degener 2016: 139, fn. 497), may

be related to Arti-(Vutr) and Artyo, recorded respectively as the names of Imró’s

mother and grandfather by Andreev (Grjunberg 1994: 86-87).

Mīr and Malik are Persian/Arabic royal titles and their association or in-

terchangeability with Māra adds plausibility to the derivation from Gandhari

mahārāya. Mənǰəm Malik is translated by Buddruss (in Jettmar 1975: 79) as

“lord of the middle” (Herr der Mitte), implying a connection to Pr. munǰ, Kt.

mič ‘middle’ (~ OIA madhya-, Av. maidiia- ‘middle’), and perhaps more di-

rectly to OIA madhyama- ‘middle (adj.)’ (cf. Kt. maǰím- ‘middle (of siblings)’

~ OIA mādhyama-). If this is correct, it might be connected to the concept of

Kt. mičdéš, the middle world of humans, as opposed to the overworld (Kt. SE

úrdeš) and the underworld (Kt. SE ǘrdeš), known from Robertson’s (1896: 380)

testimony, but it could also refer to Parun (and Ušüt/Kštöki within it) as the

geographical and spiritual center of Kafiristan. Morgenstierne’s (1951: 180)

form “Mīžˈom” is a little more difficult to explain. It cannot be equivalent to

OIA madhyama- (Kt. o implies long *ā), but its context indicates that it was

probably the direct equivalent of Pr. Mənǰəm (Pr. ə can also result from *ā, cf.

19 Cf. Pr. ṣaro “an epithet of Māra” (Buddruss, Degener 2016: 257). Regarding Kt. “ṣō ro”

Morgenstierne (1951: 179, fn. 10) notes: “Here transl[ated] ‘pome-granate (flower).’ But the usual

meaning is ‘cattle.’” The original notebook has the Norwegian translation “o du med granat-

blomst-krans” (“O you with a pomegranate-flower wreath/crown”) for “ō ṣōro mīmol” and an

added pencil note “(cattle?)” above “ṣōro” (Norwegian National Library, Ms.fol. 4201 Kati VI).

The translation ‘cattle’ results from a confusion with ṣor ‘livestock’ which has a retroflex ap-

proximant r and is therefore unrelated. The translation ‘wreath/crown’ is likely that of mí-mol

(‘feather crest of a mí-moč feast-giver?;’ mol ~ OIA mālā- ‘garland, wreath, crown’), leaving the

unlikely ‘pomegranate flower’ for “ṣōro.” Perhaps this translation was a misunderstanding of

Pashto nangúrai ‘small pomegranate’ for nang-wə´ray ‘having honor?’ “Ṣō ro” could then be in-

terpreted as a transcription of ṣór-vo ‘having ṣor.’ Ṣor has the usual meaning ‘winter quarters (for

livestock),’ but also appears in personal names (Ṣór-Malik, Ṣór-Kon). It is derived by Strand

(1999b) from Iranian, with reference to Prs. šahr ‘town’ which has the older meaning of ‘realm,

rule’ (PIr. xšaθra-). Perhaps more likely is the interpretation of “ṣōro” as ṣór-o with a vocative

-o. In that case ṣor (in meanings other than ‘winter quarters’) might have been borrowed from an

Iranian word related to the titles šār and šēr, attested in Islamic sources on Afghanistan, and to

Bactrian <χαρο> xaro (all < PIr. *xšāθriia-) (cf. Marquart 1901: 79; Sims-Williams 2007: 277).

The translations ‘having realm/rule’ or ‘ruler!’ would both fit well with the other epithets of

Imró/Māra.

326 Jakob Halfmann

Pr. nəm ‘name’ ~ OIA nāman-). Since ž is not phonemic in Katë, Miǰóm, noted

as an alternative in Morgenstierne (1951: 184, fn. 8) and originally recorded in

Morgenstierne’s notebooks20 together with the Norwegian translation “Verden

i midten = Prasun” (“world in the middle = Prasun”), is most likely the correct

form. The first element is therefore probably equivalent to mič ‘middle.’ The

second element is less clear, but given the translation it might stand for ‘world,’

‘land’ or ‘earth.’ The Indo-Iranian term for ‘earth,’ OIA kṣam-, Av. zam- <

Proto-Indo-European *dég´ hom-, *dg´ hém has no known cognate in Nuristani

and the regular Nuristani development of the PIE consonant cluster *dg´ h is not

known. A somewhat shaky etymology might connect -ǰom, as the second el-

ement of *Mič-ǰóm > Miǰóm to ǰam- in NKal. Z ǰam-ḍuŋə´ ‘wild tuber, potato’

(ḍuŋə´ = ‘bulb, round object [classifier for fruits and nuts]’) (Tāza 2017) as de-

scendants of a Proto-Nuristani *ǰām(a) ‘earth’ and perhaps even to -ǰəmí in Pr.

wərǰəmí ‘person, human being’ (< *ǰām-ya-ka- ‘earthly’? cf. Proto-Celtic *gdo-

nyos ‘human’). This would result in the interpretation of Pr. Mənǰəm and Kt.

Miǰóm as ‘middle earth.’

It is not excluded that Mənǰəm Māra, possibly the *‘middle-earth great

king,’ and Kt. Imró were originally separate deities, who were identified with

each other for some reason—perhaps due to a superficial similarity of the

names or due to a shared association with the underworld—and that the name

Kt. Imró then became associated with features originally belonging only to Pr.

Mənǰəm Māra. But it is also possible that Mənǰəm Malik, Mənǰəm Mā ra, Mā ra

and Imró have always been the same deity (originally a borrowed Yama-rājan-).

The story about Mənǰəm Malik burying the giant (or being the giant) who holds

up the “middle earth” of men could simply be a new spin on the underworld

associations of Yama-rājan- that became popular at some point.21 In either case,

this myth could have made Mənǰəm Māra/Imró the closest equivalent to a cre-

ator and therefore the best candidate for identification with the God of Islam,

whence he might have acquired features of a ruling High God.

An unknown factor that may also play a role in the elucidation of the full

story of Māra and Imró is the Bactrian deity Iamšo. *Iamá in the Iranian tradi-

tion is not usually imagined as a god, but as a mythical hero king. The name of

the Bactrian Iamšo, on the other hand, is attested as part of theophoric

personal names such as <ιαμϸολαδο> Iamšolado ‘given by Iamšo,’

<ιαμϸοβανδαγο> Iamšobandago ‘slave of Iamšo’ (~ 4th-5th c. CE), which

clearly show that he was treated as a powerful deity (cf. Grenet 2012: 85-86).

He is identified by name on a Kushan coin, where he is depicted accompanied

by a bird sitting on his arm and with similar attributes as the Kushan rulers,

20 Norwegian National Library, Ms.fol. 4201 Kati VI.

21 Mənǰəm Malik (the son) also shares some mythical motifs with the Indian Yama, e.g. a

twin sister (~ OIA Yamī) or an abandoning mother and an evil stepmother. The story of Yama’s

mother and stepmother, which is only alluded to in the Rgveda, is further elaborated in the much

later Purāṇas (Söhnen-Thieme 2009: 812).

Nuristani Theonyms in Light of Historical Phonology 327

carrying arms (a sword and a spear) and wearing an elongated headdress (Gre-

net 2012: 86-87). The name Iamšo itself probably does not derive from the

usual Iranian name-epithet combination *Iama Xšaita (whence Persian

amšēd), since *Xšaita should have developed into Bactrian *šēd. Sims-Wil-

liams (1998: 196-197) instead assumes that it derives from *Iama Xšāu ā ‘King

Yama’ (cf. Bactrian ϸαο /šā/ ‘king’ < xšāu ā, which may be present in Pr. šē üt-

nyog ‘the king’s field’). Might ideas about a divine ‘King Yama’ have come to

Nuristan from both sides of the Hindu Kush?

At least we can conclude that a preserved concept of *Iamá independently

inherited from Proto-Indo-Iranian is not “the only explanation one could give”

(Fussman 2012: 81) of the relationship between name and function of Kt. Imró.

Given the late attestation of the compound Yama-rājan- ‘King Yama,’ the pres-

ence of Middle Indo-Aryan sound changes in the name and the presence of the

expected association between name and function of the deity in Nuristani Ka-

lasha, it is not even a likely explanation.

2.)

Kt. NE Mon [<мōн> < > (Cacopardo, Schmidt 2006: 58 [69])], SE/W

Móne (Strand 1999b; Morgenstierne 1951: 164), W-KT Múne (Strand 1999a).

NKal. Mādé (Degener 1998: 476; Tāza 2017).

A. Mādē, Mādē (Morgenstierne 1934: 98).

Pr. Mā ndi (Buddruss, Degener 2016: 713).

About this deity Robertson (1896: 399) reports “It would seem that Moní,

called emphatically ‘the’ prophet, ought to be ranked next to Imrá. He is wor-

shipped with more respect than enthusiasm, especially at Kámdesh and Braga-

matál. In Presungul he retains his rightful position in the Káfir Pantheon.

Traditionally, he is the god always selected by Imrá to carry out his orders to

exterminate demons, and so forth, and there are few stories related of him in

any other connection.”

The identification as the “prophet” appears already in the Court/Allāhdād

questionnaire, where the name Móne together with the title “prophet” was pro-

bably misread by the interviewer as a reference to Mānī-yi Naqqāš ‘Mani the

painter,’ the name by which the Manichaean prophet Mani was remembered in

the Persian tradition (cf. Holzwarth 1994: 192-193). The reasoning behind the

identification as a prophet was probably the idea that the second most important

“character” in Islam after God—who was equated with Imró—is the prophet

Muḥammad and that therefore Móne, the second most important god, must be

the prophet of Imró.

His strength and his role in the extermination of demons is mentioned also

by Azar (Cacopardo, Schmidt 2006: 58 [69-70]) who writes: “They believe that

he is the most powerful of all and that he has left no stone unturned to destroy

evil spirits and monsters.” A myth describing his destruction of the demons’

328 Jakob Halfmann

castle in the sky has been recorded in three different versions, both in Katë and

in Prasun (cf. Cacopardo, Schmidt 2006: 58, fn. 187).

According to Andreev’s notes, Mon is the son of Imró (Grjunberg 1994:

85). In Morgenstierne’s (1951: 164) account he is Imró’s “first creation and his

vizier.”

Andreev noted the following list of Mon’s forms, probably extracted from

the hymn published by Grjunberg (1995b) [interpetation of Kt. words added

in square brackets]: “Mon appeared on earth in seven forms: 1) шĭру (type of

mountain goat) [šurú ‘(markhor) buck’]; 2) bull; 3) ушэ´лык (bird…) [uṣéluk

‘vulture’]; 4) гаш billy-goat [gaṣ]; 5) in the form of a man from the Прäсjýн

tribe [Prasyṹ ‘Prasun/Paruni’]; 6) iц—bear [iċ]; 7) сун-марáнæ—name of a

bird [sun ‘gold’ + mare˜ ‘bird of prey, hawk, eagle’]” (Grjunberg 1994: 85). The

hymn also describes Mon as dressed in juniper (sëréċ), leaves of the vënzyúk

tree and gold (sun) and carrying an iron bow (Grjunberg 1995b: 607-608). The

form of a hump-backed bull and his golden quiver are also referenced in hymns

recorded by Morgenstierne (1951: 180, 187).

Klimburg (1999: 144-145) describes Mādé as the main deity for agricul-

ture, invoked in particular by women among the Nuristani Kalasha. He was

worshipped “all over Waigal as well as in Wama,” but was not important in the

other Ashkun-speaking areas.

A cognate to the Nuristani theonyms exists in the Indo-Aryan Kalasha term

Mahandéo (form from Trail, Cooper 1999) and both the Nuristani terms and

the Indo-Aryan Kalasha term were/are used with an epithet related to Old Indo-

Aryan kuśala- ‘adept, clever’ (Kt. NE kšúlë, SE kšǘlë; Pr. kšul; IAKal. kúšaɫa22).

Buddruss (1977: 37) already identified the Nuristani names as loanwords

from OIA Mahādeva- ‘the great god,’ but in the same year Fussman (1977: 30-

31), though acknowledging that the epithets related to OIA kuśala- must be

Indo-Aryan borrowings (due to PIIr. *ć > š), maintained that it is probable that

the names themselves, which he compared to the full Sanskrit noun phrase

mahān devaḥ, are “authentically Kafir” (authentiquement kafires), i.e. Nuristani

lexemes inherited from Proto-Indo-Iranian. In Fussman (2012: 72) he retracted

this assessment. I gave a more precise explanation as to why this is impossible

in Halfmann (2023: 119): “The Old Indo-Aryan adjective mahat- ‘great’ derives

from an extension of a well-known Indo-European root *meǵ-oh2- ~ *m ǵ-h2-

(cf. Mayrhofer 1996: 337-338), which should again lead to a Nuristani form

containing a *ȷ < *ǵ. Instead we find no reflex at all of the *-ǵ-, which is easily

explainable as a result of a dropping of an Indo-Aryan -h- < *-ǵ-h2- in the pro-

cess of borrowing into Nuristani languages, which do not have a phoneme /h/.

Through coalescence of the two adjacent a-vowels, we then get a long vowel

ā with its regular correspondence o in Katë.”

22 Translated by Augusto Cacopardo (2016b: 59) as ‘ingenious, clever,’ but by Trail, Cooper

(1999) strangely as ‘disagreeable.’

Nuristani Theonyms in Light of Historical Phonology 329

In a sound change that has affected all of the Nuristani forms as well as the

Indo-Aryan Kalasha form, nasal consonants cause nasalization on a following

vowel, which then solidifies into a segmental nasal consonant between the

vowel and a following plosive (NVC > NVC > NVNC). This “nasal conden-

sation” must have happened over a wide area in the Indo-Iranian frontier zone:

It is attested in the Indo-Aryan languages of Swat, but also in the toponym

Nangarhār < OIA Nagara-hāra- in the Kabul valley and in inherited and bor-

rowed Nuristani vocabulary. It must have taken place before the general loss

of single intervocalic plosives, since the resulting nasal + plosive clusters sur-

vive this loss. Later, the resulting -nd- clusters were reduced to -d- in Nuristani

Kalasha (nadí ‘river’ < *nandika ~ OIA nadikā-) and to -n- in Katë (SE naní

‘river’). This sound change explains the nasal reflexes before the -d- of

Mahādeva- and makes the derivation from a full noun phrase mahān devaḥ un-

neccessary.

The Northeastern Katë form Mon, which had previously confused re-

searchers,23 can be explained by a re-analysis in the way already suggested for

Pr. Māra above: Since the word de ‘god’ was independently present in Katë,

an older form *Māndé could have been re-analyzed in analogy to more recent

compounds as *Mān-de, which would also explain the retraction of the stress

(typically on the first element in recent compounds). From *Mān-de, *Mān

(> Mon) could be extracted, before the sound change nd > n again obscured

the morpheme boundary. The two variants Mon and Móne, were then varyingly

preserved in the different dialects. The sound change o > u in the environment

of nasals, peculiar to the dialect of Ktivi (Kantiwā), explains the form Múne of

that dialect (cf. W-KT muč ‘husband,’ mus ‘moon’ vs. W, NE, SE moč, mos).

The Prasun form is likely borrowed from earlier Katë *Mānde. It shows

the development e > i also seen e.g. in Pr. kiċ, borrowed from Kt. keċ ‘long

animal hair (esp. neck hair of male goats and markhors).’

Interestingly, Mahandéo does not agree with the regular developments of

Indo-Aryan Kalasha, which otherwise shows deletion or fronting of intervocalic

h (ga-méṣ/meṣ gak ‘buffalo’ < OIA mahiṣa-; ɫíik ‘to lick’ ~ OIA lih-; dhú-ik ‘to

give milk’ ~ OIA duh-; báuɫ(a) ‘dense, thick; the Pleiades’ ~ OIA bahula-

‘dense, thick; the Pleiades’), indicating that the name probably came into Nur-

istani and Indo-Aryan Kalasha from a third Indo-Aryan source (cf. Halfmann

2023: 120). In Halfmann (ibid.) I concluded that “[s]ince Mahandeo thus has

23 Palwal (1968: 71), who recorded the form Mon in the Northeastern dialect area, tried to

explain the final -e recorded in other sources as an instance of the oblique case ending. I previously

attempted to do the opposite, deriving Mon from a mistaken analysis of Móne as an oblique form.

However, all sources representing the Northeastern dialect agree in noting Mon and all sources

from the Southeastern dialect area agree on the form Móne, suggesting a genuine dialectal split.

The oblique case ending -e is attested only for the Western dialect and cannot necessarily be as-

sumed for the pre-history of the Eastern dialects (which have an oblique ending -ë), meaning that

it cannot be used to explain the divergence of the Eastern forms.

330 Jakob Halfmann

a more archaic form than one would regularly expect, it does not appear un-

likely that these divine names were originally learned borrowings from the lit-

erary language, which diffused from the centers of religious influence and

literary culture in the plains of Gandhara […].”

Generally speaking, such a solution is problematic, especially for words

with an uncertain etymology, because it could theoretically be invoked to ex-

plain anything and everything. As the article will show, however, there are se-

veral pre-Islamic religious terms with clear etymologies, but a phonological

form that is more archaic than would be expected according to the regular sound

developments of the languages in question, or even those of Middle Indo-

Aryan. Sanskritic borrowings in later Indo-Aryan are generally very common,

in the same way that learned borrowings from Latin are abundant in the Ro-

mance languages, and they are especially likely to be encountered in religious

vocabulary. Considering the status of Peristan as a “counter-civilization” with-

out money, states or writing (Cacopardo 2023), the assumption of literary bor-

rowings in the languages of this region may seem surprising, but we have ample

evidence that an Indo-Aryan literary culture existed in the Gandharan plains,

right next to Peristan. Here, first Middle Indo-Aryan Gandhari was written,

later this was increasingly influenced and finally supplanted by Sanskrit, the

literary language emerging from the older Vedic oral tradition.24 If religious

ideas diffused from these centers into Peristan—as linguistic influences cer-

tainly did—the associated terms could have been borrowed in sanskritic or

sanskritized forms.

The OIA epithet Mahādeva- ‘great god’ is first attested in the Atharvaveda,

where it is associated with Rudrá- (Böhtlingk, Roth 1855), a fearsome marginal

deity in the Vedas who later developed into Śiva, the supreme god of Śaivism. In

the Śatapathabrāhmaṇa Mahādeva- is mentioned as one of the “eight names” of

Rudra, which are later all popular epithets of Śiva (Bisschop 2009: 742). The name

Mahādeva- thus historically precedes Śiva-, which comes into more frequent use

as a name of Rudra starting with the Śvetāśvaropaniṣad (Bisschop 2009: 742).

The earliest depictions of Mahādeva are found on Kushan coins (early cen-

turies CE) which show some iconographical motifs later associated with Śiva

(e.g. a bull, a trident, multiple heads) together with the name of the Bactrian

deity Ooēšo (ibid.: 752). That these represent Mahādeva, can only be estab-

lished indirectly: The same iconographic features are used to depict Wyšprkr

(~ Av. Vaiiuš uparō.kairiiō), the Sogdian equivalent of Bactrian Ooēšo (~ Av.

Vaiiuš), who is explicitly identified with Mahādeva <Mɣʾtyβ> in a Sogdian ver-

sion of the “108 names of Avalokiteśvara” (Avalo-kiteśvaranāmāṣṭaśatakasūtra)

(Humbach 1975: 402-404; Grenet 2006: 92), dated to around the 8th century

24 Salomon (2001: 248) observes “a gradual movement towards sanskritization of Gāndhārī

whose roots go back to the first century [CE], but which seems to have intensified in the second

century, apparently during the reign of Kaniṣka and his Kuṣāṇa successors.”

Nuristani Theonyms in Light of Historical Phonology 331

CE (Henning 1946: 736). This also indicates that Mahādeva was a more usual

name than Śiva at this time. The name may have been borrowed before Śiva

became the more common term.

Though the myths from Nuristan do not mention the most iconic attribute

of Mahādeva (and Śiva), the trident (triśūla-), other attributes are clearly

shared: The Kushan Mahādeva, classical Śiva and Nuristani Móne etc. are all

associated with the hump-backed bull. Śiva and Nuristani Móne etc. share the

role as a destroyer of demons’ castles (Tripurāntaka) and as a bowman. The

story of Śiva destroying Tripura first appears in the Mahābhārata (Bisschop

2009: 745-746). While Śiva is a supreme deity, Móne takes second place as the

“prophet,” perhaps as a result of a deplacement by Imró.

3.)

Kt. NE Bagíṣṭ [<Bag(y)'iṣṭ> (Morgenstierne 1951: 165); <Багíшт> (Grjun-

berg 1994: 85); < > (Cacopardo, Schmidt 2006: 58 [69])], W Bagíṣṭ

(Strand 1999a), SE Bağíṣṭ (Strand 1999b).

P. Bagíṣṭe (Buddruss, Degener 2016: 623); Opkulú (?) (Morgenstierne

1951: 165).

A.w Bagiṣṭók (Buddruss, Degener 2016: 623).

Robertson (1896: 406) describes this god as “a popular deity” in Kamdesh,

noting that “he presides over rivers, lakes and fountains, and helps good men

in various ways in their struggle for wealth and power. It is more particularly

because the Káfirs believe that by sacrificing to Bagisht they will become rich

that they are assiduous in his worship.” Bağíṣṭ was not worshipped in temples,

but sacrifices to him were made on particular stones, often near the mouths of

rivers25 (Robertson 1896: 406). Andreev’s Eastern Katë informants characterize

him as ‘the god of dairy products’ (“božestvo moložn. produktov”) (Grjunberg

1994: 85), whereas Azar (Cacopardo, Schmidt 2006: 58 [69]) gives a more de-

tailed description: “About him they believe that he is always moving around

in rivers and ponds. His movements are associated with the element water. He

is a helper of the afflicted and a guide to those who have lost their way. He is

the one who blesses every undertaking.” Both the Kom respondents to the

Court/Allāhdād questionnaire and Azar (in a marginal note) identify him with

the Islamic prophet Ḫiżr, who has comparable functions in folk Islam (Caco-

pardo, Schmidt 2006: 58 [69]; Holzwarth 1994: 193-194).

Morgenstierne (1951: 167-168) recorded the myth of his birth in Bumboret.

Dísëni (see below) is said to be his mother (Cacopardo, Schmidt 2006: 59 [70];

Morgenstierne 1951: 165).

25 Robertson (1896: 406) mentions the mouth of the Ṣkurigël, a place in Ürmür (mouth of

the Ničingël), and a place in Mumōgrom (“Bagalgrom;” also at the mouth of a tributary).

332 Jakob Halfmann

Din Mohammad of Üċǘ (Dewa) in Parun connects Bagíṣṭe to the move-

ment of cattle and lists the physical characteristics of golden eyes, a black robe,

and a large battle axe (Buddruss, Degener 2016: 238-239). Bagíṣṭe himself

moves in the opposite direction from the herds—upvalley in autumn and down-

valley in spring (ibid.: 238-239)—probably due to his connection to the river

water which freezes upvalley in winter and melts and flows downvalley in

spring (cf. Jettmar 1975: 107-108). Din Mohammad gives Bagíṣṭe the epithet

pšlü, which is translated by Buddruss, Degener (2016: 239) as ‘hairy’ (“be-

haart”). The same epithet and its cognates Kt. W pṭílë, NE pṭélë, SE paƛ ë

´ are

also used of other deities, but they do not necessarily indicate that these were

imagined as physically hairy. In origin the word is probably an Indo-Aryan

loanword akin to OIA pattra-la-ka-, originally meaning ‘rich in leaves, leafy’

(cf. Strand 1999b). From this meaning arose the wider connotations of ‘sprout-

ing,’ ‘hairy’ and particularly ‘fertile,’ so that the epithet can be understood as a

reference to Bagíṣṭe’s connection with fertility and wealth.

Not much is known about the Ashkun cognate deity Bagiṣṭók. Klimburg

(1975; 1999: 141-155) did not record such a name in Wama and the only avail-

able information in Buddruss’s unpublished vocabulary is the gloss ‘name of

a god’ (“Name eines Gottes”). The -ók added at the end of the name is a dim-

inutive suffix. A cognate of Bagíṣṭ has not been recorded among the Nuristani

Kalasha (Klimburg 1999: 141-155).

In Halfmann (2023: 120), I claimed that the Katë forms of the name “can

easily be connected with Old Indo-Aryan bhagiṣṭha-, the superlative of bhaga-

‘dispenser (said of gods)’”, following a proposal originally made by Morgen-

stierne (1951: 165). A form *bhagiṣṭha- is unattested, but similar superlatives

of nouns are common in Old Indo-Aryan divine epithets. If the sequence -iṣṭ is

a reflex of the Indo-Iranian superlative suffix OIA -iṣṭha-, Av. -išta-, then its

retroflex form points to an Indo-Aryan origin, contrasting with the non-retroflex

inherited Nuristani reflexes in words such as Kt. NE kanštë

´ SE kanë

´štë ‘younger’

< PIIr. *kaništa-ka- ‘youngest,’ but falling in line with Indo-Aryan loanwords

such as Kt. NE ǰiṣṭ, SE ǰeṣṭ ‘leader,’ borrowed from OIA jyeṣṭha- ‘most powerful,

greatest, chief.’26 Since the preservation of a single intervocalic -g- (with only

the late dialectal lenition > ğ in Kt. SE) is incompatible with the regular sound

developments of the modern Indo-Aryan and Nuristani languages of the area,

I concluded that “a literary borrowing diffused from the [Indo-Aryan speaking]

plains is likely” (Halfmann 2023: 120), which would also sidestep the issue

that the pre-tonic correspondence Kt. NE/W a ~ SE a otherwise implies an ety-

mological long *ā.27

Buddruss (in Fussman 1977: 32) suggests an alternative derivation from

OIA bhāgya- ‘to be divided; luck, fate, fortune’ which could account for the

26 From the Indo-Iranian root *ȷ´yā- ‘to harm, to deprive’ (OIA jyā- Av. ziiā-) (Mayrhofer

1992: 602-603), which would regularly produce a dental affricate *ȷ¨ in Proto-Nuristani.

27 Pretonic short *a developed into Kt. NE/W ë / ∅ ~ SE a.

Nuristani Theonyms in Light of Historical Phonology 333

preserved -g- as the result of an original cluster -gy-, but he leaves the latter

half of the word unexplained. Tremblay (2015: 38) extends this proposal to a

reconstructed compound *bhāgya-rātar- ‘dealer of fortunes,’ but this could not

have led to the sequence -iṣṭ.

Since the Indo-Iranian root *bhaga- is generally more common in Iranian

languages, where it serves as the generic term for ‘god,’ one might consider

the possibility of an Iranian borrowing, as this would also make the preser-

vation of -g- easier to explain. Sims-Williams (2010: 39), while discussing

the Bactrian family name Bagatano, argues that this must be a derivation

from a personal name *Bagato with the common family name suffix -ano.

*Bagato, in turn, likely contains the word bago ‘god’ and an element -ato,

which could be connected with Avestan ašta- ‘messenger’ or, in light of the

Bactrian word kisato < *kas-išta-, perhaps with Avestan išta, a participle from

the root aēš- ‘to seek, to desire’ (Sims-Williams 2010: 39). A homonymous

participle is also formed from the root yaz- ‘to worship, to sacrifice.’ As a re-

sult išta can mean both ‘desired’ and ‘worshipped.’ If *Bagato derives from

*Baga-išta-, this would come quite close in form to the name of the deity,

but semantically it is not a match: *Baga-išta- would mean ‘desired by the

god,’ rather than ‘desired/worshipped god’ (cf. the name *Baga-dāta- ‘given

by the god’)—an appropriate meaning for a personal name but not for the

name of a deity. A theonym could only result from the ordering **Išta-baga-

(cf. OIA iṣṭa-deva-(tā-) ‘desired/worshipped god’ = ‘god that is the object of

personal devotion’). This together with the retroflex -ṣṭ- means that Morgen-

stierne’s etymology with the explanation as an Indo-Aryan loanword is still

the most likely.

Morgenstierne (1951: 165) identifies Kt. Bagíṣṭ with the deity <Opkul'ū>

of Parun without giving any further explanation. The names do not appear to

be cognate but there are apparent functional similarities between the two

gods—according to Klimburg (2002: 55) “Opkulū represented the notion of

wealth.” What does not fit this picture is the fact that Buddruss recorded a Pra-

sun text about a deity named Bagíṣṭe (Buddruss, Degener 2016: 238). The name

Opkuló is recorded separately in the glossary, but it does not appear in this or

any of the other texts, it is recorded only in the phrase Pr. Z Opkuló yasə˜ ‘belt

of Opkulo’ = ‘rainbow’ (ibid.: 557). Opkuló yasə˜ appears to be a possessive

construction, so that Opkuló, extracted from the phrase as the basic form by

Buddruss, Degener (ibid.), is more likely a possessive case form of the name.

It may be comparable to the special genitive encountered in Mārē munȷ¨

‘meadow of Mara’ (vs. more regular Māréš munȷ¨) (Buddruss, Degener 2017:

77). A direct parallel where a noun ending in -u has a possessive form in -o

may be found in Pr. I lō kodyum ‘god’s work, religious duty’ (from lu ‘god’).

The nominative could therefore be a form Opkulú, corresponding to Morgen-

stierne’s <Opkul'ū>, with the element lu ‘god’ at the end of the word. Since

opku- is not an otherwise identifiable word, the etymology of the name remains

obscure, but it seems more likely to be a Prasun word than a borrowing. By

334 Jakob Halfmann

contrast, it is not unlikely that Pr. Bagíṣṭe is a form borrowed from Kt. *Bagíṣṭ-

de. Such a compound is not attested in the available sources, but could easily

have existed judging from other Katë theonyms compounded with the noun -

de ‘god.’ In any case, Bagíṣṭe must have been borrowed into Prasun after the

sound change *b > w (cf. Pr. w- ‘to become’ vs. Kt. bu-). Consequently, it is

also possible that *Opkulú or Opkuló was generally considered to be the Prasun

name of the deity referred to by Din Mohammad as Bagíṣṭe.

Tremblay (2015: 38), building on the identification of Kt. Bagíṣṭ and

Pr. Opkulú as well as Fussman’s (1977) theory of pre-Vedic origins of Nur-

istani mythology, sees both as reflections of the Indo-Iranian *Apām Napāt ‘de-

scendant of the waters.’ He also offers an etymology for Opkulú based on this

connection, deriving it from a rather imaginative reconstruction “*Āp- (a)-

°kulā

˘-

u ānts ‘de la tribu des eaux.’” The Old Indo-Aryan word kúla- ‘household,

family, herd, tribe’ is most likely related to Sogdian wkwry ‘relatives’ (Mayr-

hofer 1992: 372-373). An l in this word would be expected only in the case of

borrowing from Indo-Aryan. Prasun was never in direct contact with Indo-

Aryan languages and whatever Indo-Aryan loanwords there are, likely came

in through the other Nuristani languages. An uncompounded reflex of borrowed

kúla- is found in Nuristani Kalasha kül ‘ancestral home’ (Tāza 2017), but absent

in Prasun. It seems hard to believe that this word was borrowed into Prasun at

a time when word-internal *p was still preserved (cf. Pr. āā(w) ‘water’), then

used to form a compound where it replaces one part of an ancient religious

phrase inherited from Proto-Indo-Iranian times, and finally lost entirely outside

of this compound.

In the Rgveda the word bhága- is used both as an epithet of various gods,

particularly of Savitŕ and Agní and as the proper name of a particular deity

Bhága, who is “closely associated with Varuṇa” (Boyce 1986) and particularly

connected with the granting of wealth (Oberlies 1998: 185-186). With its super-

lative ending, *bhagiṣṭha- looks more like an intensified epithet (‘the one is

most a dispenser’) than something that could have developed out of Bhaga

used as a proper name. Among the four gods Savitŕ , Agní, Váruṇa and Bhága,

Bagíṣṭ most clearly resembles the latter two: Bhága in his association with

wealth, and Váruṇa in his association with the water. The most attributes are

shared with Varuṇa as he appears in the post-Vedic Sanskrit epics. The Vedic

Váruṇa is “the highest lord among the gods in the Veda, who has all-knowingly

supported the sky, measured the earth and taken possession of all worlds, who

preordains the right way for the waters and the sun and to whose prescribed

order (vrata) even the gods are subject” (Moeller 1984: 181). In the epics, on

the other hand, “[h]e is no longer a heavenly god, no longer a god rivalling

Indra […]. He is lord of water” (Hopkins 1915: 117). The epic Varuṇa is also

gopati ‘lord of cows’ (Hopkins 1915: 117, 120-121), which parallels Bagíṣṭ’s

association with cattle. The mouths of rivers were considered tīrthas (sacred

“fords”) of Varuṇa (Moeller 1984: 183) and Bagíṣṭ’s sacrificial stones were

equally placed at river mouths.

Nuristani Theonyms in Light of Historical Phonology 335

Since the phonological form of the name points to an Indo-Aryan borrow-

ing, these connections with the post-Vedic Varuṇa, perhaps already identified

with the wealth-giver Bhaga, seem more directly relevant than any similarities

with the reconstructed Proto-Indo-Iranian *Apām Napāt.

4.)

Kt. NE Indër28 <Indr, ˈIdr> (Morgenstierne 1951: 163, 177), SE Indrë

(Strand 1999b).

NKal. N/Z Indr (Degener 1998: 438; Strand 1999c; Tāza 2017); N <Endr>

(Melabar 1978a; 1978b).

A. W Indrə (Buddruss n.d.; Strand 2008).

Pr. <endrə, indr, əndr> (Buddruss, Degener 2016: 548).

Indr was primarily a god of the Nuristani Kalasha and Ashkun-speaking

areas. He was “renowned as a protector of agriculture and in particular of vine-

yards, being a great lover of wine.” (Klimburg 1999: 146). His cult was cen-

tered in the Ashkun-speaking village of Wama in the Pech valley, where he had

a sacred garden and vineyard named NKal. Z Indr-küṇ, N Indrakún, A. W Indrə-

kūn29 (Tāza 2017; Degener 1998: 438; Buddruss n.d.). In Wama he was con-

sidered the main deity, founder of the village and “a great benefactor”

(Klimburg 1999: 146). He was also seen as a god that hated to be outsmarted,

killing his own son after he had tricked him (Klimburg 1999: 147). One physi-

cal attribute of his were entangled golden horns (Melabar 1978a: 55).

He was not worshipped in the Katë and Prasun-speaking areas, where only

a myth was remembered in which he competes with Giṣ/Imró, but loses (Mor-

genstierne 1951: 177-178; Robertson 1896: 387-388, 410). In Robertson’s

(1896) version he must retreat to “Tsarogul” (= Ċanúgël ‘Wama’). This myth

mirrors the one told on the other side, e.g. in the Nuristani Kalasha village

Nisheygram, in which Indr defeats Geṣ in a struggle over Indrakun, making

him retreat to Ktivi (Kantiwā), the homeland of the Katë people (Melabar

1978a: 53-57; Klimburg 1999: 146). Both probably served to explain the sep-

arate realms of Giṣ/Geṣ and Indr, with each side presenting their own deity as

28 Since all tr and dr clusters are broken up by an epenthetic vowel in the Northeastern dialect,

the same probably applied to the name Indër. Though this is not evident in Morgenstierne’s tran-

scriptions, it can be seen in the word Indërík ‘earthquake,’ which is still used today (see below).

29 The name is transcribed with a dental n by Buddruss (n.d. and in Degener 1998), but the

native speaker Tāza (2017) records a retroflex ṇ, which is probably more correct. The word NKal.

küṇ by itself means ‘garden; flat ground at a high elevation’ (Tāza 2017: s.v. kü

-ṇ) and probably

corresponds to Kt. kyuṇ “playing-ground” (Morgenstierne 1951: 167) (“playing-ground” of the

lake = ‘basin’ of the lake?). It may be connected (as a borrowing) to OIA kuṇda- ‘basin, pit.’ Kt.

NE kyuṇ, SE küṇ ‘projectile, stone for playing’ (ibid.: 167, fn. 8; Strand 1999b) is unrelated (~ OIA

khaṇdya- ‘to be broken off, fragment?’).

336 Jakob Halfmann

victorious. The Nuristani Kalasha believed that Indr and Geṣ were brothers

who came together from Indrstān, sometimes equated with India (Prs.

Hindustān), other times with Wama (Melabar 1978a: 49-51; Tāza 2017: s.v.

Indrstān).

The name of Indr is quite obviously related to that of Indra-, the Vedic “king

of the gods,” who appears in the Avesta as demon. In the form of the name itself

there is nothing that would clearly identify it as either a loanword or an inherited

word. Pace Fussman (2012: 75), the preservation of the consonant cluster -ndr-

in the Nuristani languages is not unexpected. The seeming preservation of a

final vowel in Kt. SE Indrë and A. W Indrə could be interpreted as evidence for

Sanskritic borrowing, but it is perhaps more likely an epenthetic vowel that de-

veloped in word-final position after the rare consonant cluster -ndr-.

Related names appear in the pantheon of the Chitral Kalasha, who speak

an Indo-Aryan language: Baɫimạín (Baɫimạíndr-) ‘the god to whom the winter

festival (čawmós) is dedicated’ (~ OIA *bali-tama- Indra- ‘strongest Indra’)

and Warín (Waríndr-) ‘the chief god of Birir valley’ (~ OIA apara- Indra- ‘un-

rivalled Indra’) (cf. Trail, Cooper 1999). Varín <Варī, Варīн> (but not Indër)

is listed among the gods of Kafiristan by Andreev’s Eastern Katë informants

(with gender ambiguity as “m. f.,” “f.-m.”) (Grjunberg 1994: 86). The form of

the name (with -ndr- > n#) shows that this deity was adopted from the Chitral

Kalasha.

Etymological connections with OIA Indra- can also be detected in Kt. NE

indërõ, SE indrõ, W-KT idrú ‘rainbow’ (Strand 1999a; 1999b) (~ OIA indra-

dhanuṣ- ‘rainbow,’ lit. ‘Indra-bow’) and Kt. NE indërík, SE indrë

´ṣ, W-KT idríċ

‘earthquake’ (Strand 1999b; Strand 1999a) (possibly ~ OIA *indreṣṭi- ‘Indra-

impulse’) (Turner 1962). Neither of the two phenomena were associated with

Indër in the beliefs of Katë speakers (Morgenstierne 1951: 163), the connection

is therefore purely etymological.

Another connection is found in the word indr-oċalóg, recorded as the NKal.

N word for ‘otter’ (Degener 1998: 438). Degener (1998: 438) assumes that the

word has been deformed due to a folk etymological association with Indr. The

word oċalóg by itself means ‘little calf’ and other languages in the region gen-

erally form their words for ‘otter’ by the semantic model ‘water-calf.’ Still, the

name of Indr in the compound is probably not a folk etymological deformation

of NKal. āw ‘water,’ with which it has no resemblance. Andreev recorded the

information that the Kafirs “avoid killing beavers ‘ōу уцéрук,’ since

they are considered heavenly dogs.”30 Persian saglāb means ‘beaver’ as indi-

cated in the translation, but Kt. NE ṓ-uċeruk actually means ‘otter’ (lit. ‘little

water-calf’). Whether or not the interpretation as “heavenly dogs” is original

or a translation mistake (an association with calves seems more likely), otters

Nuristani Theonyms in Light of Historical Phonology 337

30 “Избегают убивать бобров ‘ōу уцéрук,’ т.к. они считаются небесными

собаками.” (Grjunberg 1994: 85).

were clearly accorded some kind of sacred status in pre-Islamic Nuristan. Seen

from this perspective, it is not strange that they were interpreted in Nisheygram

as Indr-calves. The same sacred status of otters is probably also the background

for the stories about magical “calves” living in water springs (actually otters?),

discussed e.g. by Klimburg (1999: 159).

The attributes of the Indr of Wama are not quite the same as those of the

all-powerful Vedic war god Indra. They are rather more compatible with the

post-Vedic image of Indra as an increasingly weakened figure, a rain and fer-

tility god who can be outsmarted and has a weakness for intoxicants and pleas-

ure (cf. Moeller 1984: 113). The association with intoxicating substances is

already present in the Vedas where Indra is presented as a lover of the sacred

drink Soma, but it becomes stronger in later times, e.g. when Indra is overcome

by the demon Mada ‘drunkenness’ (Brockington 2014: 73). The Indr of Wama

is still a powerful ruling deity, but this, too, is not necessarily indicative of Vedic

(or even pre-Vedic) affinities, as Indra “maintained a degree of supremacy in

more popular belief longer than most Vedic deities” (ibid.: 67).

5.)

Kt. NE/SE Giṣ [<Гīш> (Grjunberg 1994: 85), < > (Cacopardo, Schmidt

2006: 57 [69])], (Morgenstierne 1951: 163; Strand 1999b); W Gëvíṣ (Strand 1999a).

NKal. Z/N Geṣ (Degener 1998: 433; Tāza 2017).

A. Gewíṣ (Buddruss n.d.).

Pr. Giwiṣ (Buddruss, Degener 2016: 663).

Robertson (1896: 400) describes Giṣ as “the most popular god of the Bash-

gul Káfirs.” He was the god of war, “the Káfir type of a true man,” “fierce and

sudden in his terrible onslaughts” (Robertson 1896: 401). He was worshipped

especially by warriors who went on killing raids.

In hymns he is called upon to subdue the enemies or to drink their blood

(Morgenstierne 1951: 188, 185). Andreev’s eastern Katë informants similarly

asserted that “after killing a Muslim they slightly lick the blood from the knife:

a ritual that takes its origin from Giṣ.”31 In myths, Giṣ could also appear as a

somewhat dull-witted trickster (Morgenstierne 1968: 530-532; Buddruss, De-

gener 2016: 232-233). An epithet of his was Kt. NE bëlím, SE balím (Morgen-

stierne 1951: 180, 188) (~ OIA *bali-tama-, a superlative formation of balin-

‘powerful, strong’).

In pre-Islamic times he was identified with the caliph Yazīd, the killer of

Husayn at Karbalāʾ (Hughes 1883: 424; Robertson 1896: 401; Herrlich et al.

31 “После убивания мусульманина слегка лижут кровь с ножа: обряд, ведущий

происхождение от Гīша.” (Grjunberg 1994: 86). Kt. lúy-aši-vo ‘having a blood mouth’ and NKal.

lai-āšə ‘blood-mouth’ are attested as praise epithets for warriors (Morgenstierne 1951: 189; Tāza

2017: s.v. lai-āšə).

338 Jakob Halfmann

1937: 335).32 He was said to have beheaded ʿAlī and played pace˜ with his head

(Robertson 1896: 401).33 In another version it is the head of Sanú, a Muslim

who wanted to convert the Kafirs (see below) (Morgenstierne 1951: 187). A

more conciliatory interpretation, found in the Court-Allāhdād questionnaire, in-

stead identifies Giṣ with ʿAlī himself, perhaps comparing their roles as great

champions (Holzwarth 1994: 193). After conversion, the value system surround-

ing the figure of Giṣ had to be inverted completely. The converted Azar (Caco-

pardo, Schmidt 2006: 57 [69]) writes: “they describe [Giṣ] as completely

opposed to Muslims […]. In reality it is Satan they call by this name.” The same

interpretation of Gëvíṣ as Satan (Šayṭān) was also given to the German Hindu

Kush expedition in western Nuristan (Herrlich et al. 1937: 311). In Parun, Bud-

druss was told that all the gods had converted to Islam, except Giwiṣ, who re-

fused (Buddruss 1983: 83).

While Giṣ/Gëvíṣ was the most popular god of the Katë speakers, he was

“much less admired” in Parun (Robertson 1896: 405). He was worshipped in

Wama, but not as a main deity (Klimburg 1999: 148, 154), and among the

Nuristani Kalasha Geṣ had the inverse role to that found among the Katë

speakers, being primarily remembered for his competition with and loss to

Indr (ibid.: 146).

For the etymology of the name, Morgenstierne (1951: 163) suggested a re-

constructed form *gav-eṣa-, pointing also to OIA gaveṣaṇa- ‘desirous of com-

bat/cows’ and gaviṣṭi- ‘cow-striving, battle.’ Fussman (1977: 31) accepted this

proposal, extending it in Fussman (2012: 75) to “*gaveṣa- + x” without ex-

plaining why and adding the puzzling remark quoted in section 2, which takes

the “difficulty of explaining the vocalism” as an argument against borrowing.

A better match, and an actually attested form, is suggested by Turner (1962:

T.4102) with OIA gaviṣá- ‘desiring cows/battle,’ which is composed of the

same roots go ‘cow’ and eṣ- ‘to seek, to desire’ and appears in the Rgveda as

an epithet of Indra. The epithet gaviṣá- seems to be textually unattested after

the Rgveda, though the abstract noun gaviṣṭi- and other combinations of go

32 It is interesting that the implications made with this identification only make sense within

the belief system of Shia Islam. Perhaps this shows that the Kafirs learned what they knew about

Islam from their Ismaili neighbors to the North and Northeast, with whom they had more friendly

relations, rather than from their arch-enemies the Sunni Pashtuns. The claim that Giṣ is the caliph

Yazīd may have been intended to horrify all Muslims and the Pashtuns in particular, but it could

hardly have shown this effect on Sunni Muslims, who have no particularly negative associations

with Yazīd.

33 This scene has entered the popular imagination together with Robertson’s identification

of the game in question as polo, appearing in this form e.g. in the film version of Rudyard Ki-

pling’s “The Man Who Would Be King.” Morgenstierne (1951: 187), probably influenced by Ro-

bertson, also gave the translation ‘polo’ for Kt. SE pac e˜ in the version of the story he recorded.

But polo (unlike Buzkaši) is not actually played in Nuristan. The pac e˜ game (NE pac én, W pac ev)

does not involve horses and has more in common with golf or hockey. An episode of Yazīd poking

the severed head of Husayn with his staff also appears in Shia narratives of Husayn’s martyrdom

and may have been the inspiration for the Kafir reinterpretation.

Nuristani Theonyms in Light of Historical Phonology 339

and eṣ- remain in use. The Nuristani forms are identified as Indo-Aryan loan-

words by the retroflex ṣ: The OIA root eṣ- derives from Proto-Indo-European

*h2eis- ‘to seek’ (Rix et al. 2001: 260) and PIE *s after *i should produce a

palatal š in inherited Nuristani vocabulary (cf. Kt. viš ‘poison’ < PIIr. *vĭ

-š-a-

> OIA viṣa- ‘poison’). Pace Fussman (2012: 75), there is in fact nothing unex-

pected in the vowel developments.

In light of the connection of the word gaviṣá- with Indra, the relationship

of Geṣ and Indr as described in Nuristani Kalasha mythology (brothers who

came together from Indrstān) may be interpreted as containing a shred of re-

membrance of them being originally identical. They also share the characteristic

of being an (unsuccessful) trickster. The lack of later attestation of gaviṣá- does

not necessarily mean that the term was borrowed in Vedic times, since the Vedic

oral tradition remained alive and could have reintroduced it into the religious

terminology. Still, if Giṣ is a development of Indra, his image as a powerful

war god represents an aspect that was more prominent in the earlier Indra. Per-

haps later developments in the character of Indra did not reach the more north-

ern parts of Kafiristan, but it is also possible that the martial aspect of Indra,

represented by the epithet gaviṣá-, may have come to the forefront under the

influence of the warlike territorial expansion of the Katë speakers in the cen-

turies before conversion and the increasing establishment of killing raids on

Muslim lands as a part of their culture.

The relationship with Dísëni (and her connection to OIA Devasenā-), as

well as the association of “peacock” feathers34 with killing feats, are aspects that

point also to possible influences from the war god Skanda-, a non-Vedic deity

whose worship became prominent during the Kushan empire, but faded out in

northern India after the end of the Gupta period (Clothey 2009: 638-639). His

name may be preserved in that of the Nuristani Kalasha war god and leading

deity of Waigal and Zhönchigal Traskə´n (Klimburg 1999: 150-151; Tāza 2017:

s.v. Traskən).

34 The peacock is the sacred animal of Skanda. Regarding the feathers used as killing rank

symbols in Kafiristan Alberto Cacopardo writes: “These were imported items, since peacocks are

not found in Nuristan” (Cacopardo, Schmidt 2006: 53-54, fn. 171). It seems more likely, however,

that the feathers used were actually those of the Himalayan Monal, a similarly colorful type of

pheasant that is native to Nuristan and often translated as ‘peacock.’ Andreev records: “вацы´—

name of some bird, highly respected by the Siahposh, who do not kill it. In the rainy season they

sometimes catch them when they run into the villages, but they immediately release them into

the wild, having tied a short thread around the neck. They acquire their feathers, which go into

headdresses during the worship of gods, from Kamdesh, where people kill these birds.” (Grjunberg

1995b: 612-613). It is also not quite correct that Kt. mol, the term used for the peacock feather

crest, “simply means feather” (Cacopardo, Schmidt 2006: 54, fn. 171). Andreev translates it as

“1) crest (on a bird) 2) plume of feathers, made from the beautiful feathers of some bird called

вацы´, and worn on the head at the time of a religious dance by men who have killed enemies.”

Originally it is a loanword corresponding to OIA mālā- ‘wreath, garland, crown.’

340 Jakob Halfmann

6.)

Kt. NE Dísëni [<дíсы нi> (Grjunberg 1994: 86), < > (Cacopardo,

Schmidt 2006: 59 [70])], w Dísëi (Strand 1999a), SE Dízëni (Strand 1999b).

NKal. Z Dísəni <Dīsřī> (Tāza 2017).

A. <däsäˈṇī> ‘ogress’ (Morgenstierne 1929: 255); <Dasäˈṇī-āma> ‘temple

of the goddess D.’ (Morgenstierne 1934: 90).

Pr. Dísni (Buddruss, Degener 2016: 652).

Robertson (1896: 410) states that “Dizane is a popular goddess and is wor-

shipped wherever I have been in Káfiristán. The Giché, or new year festival is

entirely in her honour, and she also has special observances during the Di-

zanedu holidays. […] When the men of a tribe are away raiding, and the women

collect in the villages to dance day and night to propitiate the gods and sing

their praises, Dizane is one of the chief deities they supplicate for help.” The

missionary Sayyid Šāh reports that “the Kafirs call upon her to destroy the

Muslims” (Hughes 1883: 422).

Robertson (1896: 396) describes a “very pretty little temple” to Dízëni built

by artisans from Parun at Kamdesh: “It is covered in carving, and has the

wedge-shaped roof so common in Presungul […]. Along both sides of the base

of the sloping roof poles are fixed, and support wooden images of birds, said

to be pigeons.”

She is said to have come forth from the Sudrém lake and consequently her

father is Sudrém (Morgenstierne 1951: 182; Grjunberg 1994: 86). Her mother

is said to be Naŋívutr and her son is Bagíṣṭ (Morgenstierne 1951: 164; Caco-

pardo, Schmidt 2006: 59 [70]). An epithet of hers is Kt. utáy ‘great (f.)’ (Mor-

genstierne 1951: 164)35 and she is called “strong and respected among the gods”

(Buddruss, Degener 2016: 32-33). She is described in hymns as carrying arms

(bow & quiver) (ibid.: 248-249) and accompanied by ringing bells (Morgen-

stierne 1951: 182-183).

Klimburg (2002: 55) reports that in Parun Dísni “dominated among the

goddesses, representing the image of an erotic fertility goddess.” In Prasun

hymns she is also addressed as Ṣu(w)ē Dísni (Buddruss, Degener 2016: 288-

289), using an epithet that is also connected with the goddess of childbirth Pr.

Pəṣaṣí, Kt. Nirméli (see below), who seems to be closely associated with Pr.

Dísni, Kt. Dísëni.

Her association with war is probably what led to her identification with

ʿĀʾišah, the third wife of the prophet Muḥammad (cf. Holzwarth 1994: 194;

Hughes 1883: 424): ʿĀʾišah was the only woman in the early history of Islam

who became actively involved in warfare and power struggles over the ca-

liphate.

35 Probably not ‘priestess’ as claimed by Morgenstierne (1951: 164), cf. the discussion of

utó below.

Nuristani Theonyms in Light of Historical Phonology 341

Dísëni was primarily worshipped in Katë and Prasun-speaking areas. Klim-

burg (1999: 149) records reports of her worship in Wama but writes that she

had “no or little importance in Waigal.”

Morgenstierne (1951: 164) phrased his proposal for the etymological origin

of her name as a question: “Has her name any connection with Skt. Dhiṣaṇā?”

Dhiṣáṇā- and dhiṣána- are two problematic Vedic terms that appear in widely

divergent meanings. Vedic philologists have variously interpreted dhiṣáṇa- as

‘potency of setting,’ ‘work of a priest’ or ‘padding,’ but it is also attested as the

name of an evil being and as an epithet of the male god Brhaspáti- (Mayrhofer

1992: 791-792). Dhiṣáṇā- refers to an implement used for preparing the sacred

drink Soma, but also (metaphorically?) to heaven and earth, and a number of

times in the Rgveda to a goddess associated with wealth (Böhtlingk, Roth

1855). The name of the goddess is usually explained as a personification of the

unclear noun dhiṣáṇa- (Mayrhofer 1992: 792). There are no cognates in Ira-

nian36 and no attestations of popular worship of a goddess Dhiṣáṇā-.

The identification of the Nuristani theonyms with Dhiṣáṇā- has been gen-

erally accepted,37 even though the meaning of the Vedic word is hardly clear

and even though it is impossible to reconcile with the attested forms, whether

via inheritance from Proto-Indo-Iranian or via borrowing from Indo-Aryan. In

order to explain the final -i one would first of all have to assume an original

form *Dhiṣaṇikā-. In the case of borrowing, we would expect Kt. NE ṣ, SE ẓ as

reflexes of the retroflex ṣ. In case of inheritance from Proto-Indo-Iranian

(< *dhišan-ī-kā- ?), we would expect a palatal sibilant NE š, SE ž. The initial

stress is unexpected in either case—stress has otherwise been generalized to

the penultimate syllable (which is today the ultimate after the loss of final syl-

lables) and is only retracted in cases of vowel contraction or epenthesis.

On the other hand, the presence of s after i as well as the stress pattern are

easily compatible with the assumption of a relatively recent compound. In fact,

the name Dísëni is completely transparent for modern Katë speakers as being

composed of di ‘sky’ (~ OIA div-a-) and sëní ‘soldier’ (pl.: ‘army’) (~ OIA

senya-, Av. haēniia-).38 Strand notes this segmentation, but identifies it as a

36 Alberto Cacopardo notes that the name Dísëni may have a relation to that of “an Avestic

demon” (Cacopardo, Schmidt 2006: 59, fn. 190), but this is based on a misinterpretation of Parkes’

(1986: 151) abbreviation “AV.” (= Atharvaveda, the source of dhiṣáṇa- in the meaning ‘name of

an evil spirit’) as a reference to Avestan (Av.).

37 Fussman (1977: 32-33) considered Morgenstierne’s doubt to be unjustified and declared

that the etymology could be taken as “certain.” The only argument he cites is that the function of

Kt. Dísëni as a fertility goddess together with the translation ‘ogress’ for A. <däsäˈṇī> supposedly

reflects the same “ambiguity” also seen in the Vedic attestations of dhiṣáṇa- and dhiṣáṇā-. As

Klimburg (1999: 149) shows, however, <däsäˈṇī> was no less a goddess in the Ashkun-speaking

region than elsewhere. The meaning ‘ogress,’ recorded once by Morgenstierne, therefore does

not make her more comparable to a demonic being.

38 This transparent meaning has even led to a proposal (seen on Facebook) to introduce it as

the word for ‘angel’ in the Islamic context, though it seems that this was not generally adopted.

342 Jakob Halfmann

“folk etymology.” If it is a folk etymology, then the word has been reshaped

by the folk etymological re-analysis to such a degree that today it is not distin-

guishable from a regular compound. The easier assumption is that the supposed

folk etymology is in fact the correct etymology.

The Katë dialectal forms are easily explained: intervocalic lenition of s > z

is a late innovation of the Southeastern dialect, found also in early Persian loan-

words (e.g. muzë

´rmon ‘Muslim’ < Prs. musulmān). The loss of n in the Western

dialect both in sëí ‘soldier’ and in Dísëi reflect the general Western sound

change n > r with subsequent loss or metathesis of r before i, which leaves be-

hind vowel hiatus. The forms of the other languages are quite possibly bor-

rowed from Katë.

If it is true that the name is a regular Katë nominal compound, this does

not tell us much about the origin of the deity. Still, the name ‘sky-soldier’ used

for a female goddess connected with war and sexuality has some resonance in

the religious world of Hinduism. It may be comparable to the Sanskrit khecarī-

‘female sky goer.’ This term is attested as an epithet of the yoginīs, “a ravening

horde of medieval tantric goddesses,” whose esoteric worship rituals contained

sexual elements (White 2009: 823). According to Hatley (2013: 26) “the arche-

typal yoginī is the autonomous Sky-traveller (khecarī).” The khecarī yoginīs

also had a military aspect—“Yoginīs were the semi-divine war goddesses of

many a medieval South Asian kingdom […],” divided by tantric sources into

“airborne (khecarī) and land-based (bhūcarī) divisions” (White 2003: 132).

The same term khecarī- has been calqued in Tibetan as maam <mkaḥ ḥgro

ma> /khandroma/ ‘female sky goer,’ spreading to Tibet together with tantric Bud-

dhism. There, it later became established as the Tibetan translation of Sanskrit

ḍākinī-. As Hatley (2007: 47) explains, “[w]hile tantric Śaiva sources generally

speak of the ḍākinī as a pernicious being, the term is often perfectly synonymous

with yoginī, especially in the yoginītantras of later Tantric Buddhism.” That

the source of the Tibetan calque was khecarī- and not yoginī- or ḍākinī- perhaps

indicates that this was previously the most common term. The Nuristani terminol-

ogy at any rate seems to be closer to the Śaiva conception, since reflexes of OIA

ḍākinī- generally refer to evil “witches” in the Nuristani languages.

A more obvious connection could be established between Dísëni and

Devasenā- ‘god-army,’ the wife of the war god Skanda-, who is generally ident-

ified with the birth goddess Ṣaṣṭhī- (cf. Kapp 1983: 317-318). It is relevant to

note in this regard that Dísëni was also declared to be the wife of the war god

Giṣ and that Dísëni is in several sources closely associated with the birth god-

dess Nirméli (see below). The name Devasenā- cannot be the direct source for

the compound Dísëni, and in the case of a re-analysis one would rather expect

the result *Dé-sëni, but it seems probable that it influenced the formation of

the compound in one way or another.

The birth goddess Ṣaṣṭhī- is noted by White (2003: 40-42) as one of the earlier

influences on the later yoginī cult. It is therefore possible that both the khecarīs

and Devasenā- played a role in the formation of Dísëni’s name. This could ex-

Nuristani Theonyms in Light of Historical Phonology 343

plain, why the first element of the compound is di ‘sky’ (representing OIA khe

‘in the sky’) and not de ‘god,’ but on the other hand also why Dísëni is an indi-

vidual goddess and not part of a horde. We could then understand the wooden

images of birds (pigeons) attached to the Dízëni shrine in Kamdesh and the clamor

of bells that accompanies her as influences from the developing khecarī/yoginī

cult—yoginīs were often imagined in the form of birds, including pigeons (White

2003: 189), and as “violently ringing bells and beating drums” (ibid.: 207).

Though it has roots in earlier beliefs, worship of yoginīs is a relatively late

phenomenon—according to White (2009: 823) “references to yoginīs begin to

appear frequently from the 7th century onward.” Interestingly, Uḍḍiyāna (the

Swat valley), which is not too far from Nuristan, is mentioned early on in the

yoginī-related tantric literature and is later identified as “the favored abode of

ḍākinīs and yoginīs” (White 2009: 823). Perhaps early forms of yoginī worship

came together with the cult of the goddess Devasenā- in forming both the name

and the character of Dísëni.

7.)

Kt. NE Nirméli [<Нīрмäлi> (Grjunberg 1994: 86), < >], SE Nirmë

´li.

Pr. Pəṣaṣí (Buddruss, Degener 2016: 753).

This goddess has only been recorded in the Katë and Prasun-speaking

areas. Robertson (1896: 411) describes Nirmë

´li as ‘the Káfir Lucina,’ equating

her with the Roman goddess of childbirth. She was the goddess responsible for

the “impure” processes of childbirth and menstruation and therefore she was

also the goddess of the seclusion house, into which women would withdraw

for childbirth and during menstruation (Morgenstierne 1951: 165, 175-176).

Here statues of her were also found (Cacopardo, Schmidt 2006: 24 [31]). She

has her seat next to Imró (Morgenstierne 1951: 176). A hymn recorded by Edel-

berg (1972: 73-74) in Parun identifies Pəṣaṣí, the equivalent of Kt. Nirméli, as

“the best of creatures.” A myth connects Nirméli with Dísëni, respectively as

the roots and trunk of a single tree (Robertson 1896: 386). Edelberg (1972: 37,

fn. 14) and, following him, Jettmar (1975: 105) have also seen an identification

with Dísni in Pəṣaṣí being called “dezeile,” but this transcription most likely

does not represent the name of Dísni, but the Katë word dëzéli ‘creatress’ (cf.

also the discussion of dëzélë ‘creator’ below). A goddess functionally analogous

to Nirméli/Pəṣaṣí is worshipped by the Indo-Aryan speaking Chitral Kalasha

under the name Dizálik (Cacopardo 2016b: 58). Her name is identified by the

palatal development as a Nuristani word,39 almost certainly a borrowing from

Katë, where dëzélik (NE) with a feminine diminutive suffix -ik means ‘little cre-

344 Jakob Halfmann

39 Proto-Indo-Iranian *dai ȷ´h- ‘to form, to shape’ > OIA deh-, which cannot produce a sibilant

z in this word. The retroflex ḍ in IA Kalasha is strange, but appears in the whole borrowed word

family, cf. ḍízik ‘to create,’ Díziɫa Dizáw ‘creator god.’

atress.’ It can therefore be assumed that Kt. dëzéli and dëzélik were common

epithets of Nirméli.

The word Nirméli itself has been connected by Morgenstierne (1951: 165)

with OIA nirmalikā- ‘pure (f.)’ (lit. “unsoiled”—nir- ‘privative prefix,’ mala-

‘dirt, filth’). This would of course be a very apt term for a goddess that remains

“unsoiled” despite being responsible for the “impure” female reproductive

functions, which makes the derivation quite convincing. However, in the light

of dëzéli ‘creatress,’ the name is perhaps also better considered a feminine im-

perfective participle (constructed in the same way with -li) formed with an

equivalent of OIA nir-mā- ‘to mete out, to create.’ This could explain why Imró

is given the corresponding masculine form nirmélë as an epithet (Grjunberg

1995a: 150, 154; Davidson 1902: 193)40—which is in fact translated by Da-

vidson (1902: 193) as ‘creator’—and why Nirméli is seated next to him. A pro-

ductive verb stem nirm- ‘to build, to create’ exists in Prasun (Buddruss,

Degener 2016: 730) and the same root is certainly also behind Kt. nirmó ‘fate’

(Strand 1999b) (~ OIA nir-mā-ta- ‘meted out’), which appears in the name of

the “nirma-gu” (= *nirmó-gu ‘fate bug’), a marriage oracle for unmarried girls

involving a beetle recorded by Palwal (1968: 79).

In all cases, the sound developments are surprising: The Old Indo-Aryan

consonant cluster -rm- was reduced to -mm- already in Middle Indo-Aryan

Gandhari (cf. Gandh. <ṇimaṇaradi-> /nimmanaradi/ ‘pleased by creating’

< OIA nir-māṇa-rati- (Baums, Glass 2002)), and is also simplified in the Nur-

istani languages. If the morpheme boundary after nir- is not in some way re-

sponsible, the easiest way to explain this fact would be to assume a Sankritic

borrowing transmitted from the lowlands. Since Nirméli, with both possible

etymologies, is clearly an epithet in origin, the original “proper name” of the

deity, if there was one, remains unaccessible to us.

The Prasun name of the goddess, Pəṣaṣí, must have a different origin. Mor-

genstierne (1949: 283), based on a wrongly noted phonological form “pəšāši”

and the incorrect meaning ‘a female demon,’ had suggested a connection with

OIA piśācī- ‘female flesh-eating demon,’ which is still maintained by Buddruss,

Degener (2016: 753). Parkes (1986: 154) was, however, right to point out that

“this etymology cannot be substantiated in view of Buddruss’ -ṣ-,” even if one

ignores the semantic distance.

Pəṣaṣí has the epithet Pr. Ṣu(w)ē, which is also associated with Dísni. An

explanation of this epithet and the name Pəṣaṣí itself may perhaps be found in

a connection with Ṣaṣṭhī-, the Hindu goddess of childbirth.

The name Ṣaṣṭhī- means literally ‘the sixth (f.),’ referencing her worship

on the sixth day after giving birth, when most of the danger for newborn and

mother has passed. Her worship is attested since Kushan times and Indian birth

40 Cf. also Palwal (1968: 76): “Nirmali, rightousness [sic], is the epithet of Imra.” The term

meant here is probably nirmélë.

Nuristani Theonyms in Light of Historical Phonology 345

goddesses resembling her have spread far outside of India (White 2003: 40-41).

White (ibid.: 41) stresses her special association with the number ‘six’ and there

is at least a superficial similarity between the word for ‘six’ in the Nuristani

languages (Kt. ṣu, NKal. ṣu, A., Pr. wuṣ(ū)) and the epithet Ṣu(w)ē. One might

think of a borrowing from Kt. ṣú-vay,41 which would mean ‘having news,

knowing (f.)’ but also ‘having six (f.),’ or ‘associated with six (f.)’ if it is the

feminine form of -vani rather than of -vo ‘having.’42

The Indian Ṣaṣṭhī- is closely associated with cats and has sometimes been

depicted as cat-faced (White 2003: 43). Though such an association is not ver-

bally attested in the few reports about the birth goddess of Nuristan, there is

in fact a good chance that Pəṣaṣí was generally depicted with cat whiskers.

Edelberg (1972: 37, fn. 14) relates a local report that the great temple of Māra

in Ušüt (Kštöki) contained “a big god sitting on a horse,” who was Māra

and “a small goddess (Paruni: banigya diz) sitting on a taboret,” who was

Pəṣaṣí/Dëzéli, placed to the left of Mā ra. The placement of Pəṣaṣí next to Māra

agrees with the belief that Nirméli had her seat next to Imró. Edelberg (1972:

37, fn. 14) suspects that the image of Māra referred to here is the surviving

statue KK1, now kept at the Musée Guimet (Edelberg 1960: figs. 2-6; Klimburg

2002: fig. 3). This statue and another one of a goddess sitting on a “taboret”

(Edelberg 1960: KK5, figs. 10-14; Klimburg 2002: fig. 18b, also now at the

Musée Guimet) are described by Edelberg (1960: 248-252) as not showing any

traces of smoke or weather influences, meaning that they were most likely not

originally placed near a hearth in a clan house or outside, but in the great

temple. Based on this information, it would seem the most reasonable to con-

sider the sitting goddess to be a depiction of Pəṣaṣí. This seated statue has in-

cised lines extending downward from her nose on both sides of the mouth that

look very much like whiskers, as well as a lozenge between her eyes. The same

features (downward whiskers, lozenge) are also found on at least one other

goddess statue documented in Kabul, also seated on a stool (Edelberg 1960:

KK6, figs. 12 & 14). Robertson (1896: 396) described a statue with very similar

features that he saw in Üċǘ (Dewa): “He is furnished with large circular eyes

with a dot in the middle; he has cat-like moustaches, and appears to be holding

his head in his hands.” While Robertson believed that this was a depiction of

Māndi, Klimburg (2002: 67-68, en. 29) considers that an “obvious mistake,”

since there was no temple to Māndi in Üċǘ (Dewa). He instead prefers to ident-

41 Regarding the substitution of Pr. ē for Kt. ay in loanwords, cf. Pr. Kimē ~ Kt. NE Kṣumáy,

SE Krumáy; Pr. Ṣiŋrē ~ Kt. Ṣiŋráy.

42 It seems that no traditions have been recorded that would connect the birth goddess of

Nuristan with the sixth day after giving birth in particular, but a period of twenty or twenty-one

days after giving birth is well attested in the sources as the period the new mother has to stay in

the seclusion house (Cacopardo, Schmidt 2006: 10 [10]). Worship of Janma-Ṣaṣṭhī (‘birth Ṣaṣṭhī’)

on the 21st day after giving birth, and the belief that the time of impurity ends on this day, allowing

the mother to leave the birth room, are equally attested in India (Kapp 1983: 310).

346 Jakob Halfmann

ify the described statue as Wuṣúm, the tutelary deity of the village. Still, since

the other statues with whiskers and a lozenge between the eyes are female,

what Robertson saw may also have been a depiction of Pəṣaṣí.

If it is true that Pəṣaṣí was depicted with whiskers, then it is not com-

pletely irrelevant that her name has some superficial resemblance to the word

for ‘cat,’ though this cannot be its source. In the Nuristani languages the word

for cat is Kt. pšaš, NKal. n pišā

´, A. pisãs Pr. pšig. The same root, though not

traceable to older Indo-Iranian languages, is also present in neighboring Indo-

Aryan languages, such as pisāso and pisōsō in Pashai dialects and pšǟ

˜šī in

Gawar-Bati.

The origin of the name Pəṣaṣí could be a borrowing from an Indo-Aryan

name corresponding to OIA *pra-ṣaṣṭhikā-. Ṣaṣṭhikā- with a -kā- extension is

attested as a form of the name Ṣaṣṭhī- and the prefix pra- can be used as an in-

tensifier, cf. Caṇḍā and Pra-caṇḍā- (names of the goddess Durgā) and sem-

antically bhaga- and *bhagiṣṭha-. An issue with this explanation is the

simplification -ṣṭ- > -ṣ-, which would have to be explained as a result of as-

similation to the preceding ṣ. It is a remote possibility, though difficult to prove,

that this assimilation was aided by interference from the similar and semanti-

cally related word for ‘cat.’

TERMS FOR RELIGIOUS IDEAS AND FUNCTIONS

1.)

Kt. NE/SE utó (m.), utáy (f.), W vëtó.

NKal. Z utā (Tāza 2017).

Pr. wutā (Buddruss, Degener 2016: 871).

A title of priests, but also of more wordly leaders, such as the Katë

´-uto

(Klimburg 2008: 400). Morgenstierne (1949: 277) doubtingly suggested a re-

lation to OIA hótrka ‘assistant of the hótr priest.’ Fussman (1977: 28-29) has

given clear arguments why this derivation is very unlikely. Strand (1999b)

has established that the word does not mean ‘priest’ as such, but simply

‘great.’

Considering the presence of Sanskritic borrowings with preserved single

intervocalic plosives, one may be tempted to connect this word with OIA

avatāra- ‘incarnation,’ which would otherwise be impossible since inter-

vocalic -t- should have been dropped. However, when considered together

with the common adverb Kt. NE/SE utō

´, W vëtó ‘(any) more; spare, extra’ a

derivation from an equivalent of OIA ut-tāra- ‘surpassing’ seems much more

likely. With this etymology, it may easily be an inherited word, though it

may have been borrowed from Katë into one or more of the other Nuristani

languages.

Nuristani Theonyms in Light of Historical Phonology 347

2.)

Kt. NE pṣë [<Pshe>, <Psse> “with retroflexed ‘sh’” (Palwal 1970: 25);

<> (Cacopardo, Schmidt 2006: 61 [71])]; SE/W pṣë; SE pṣë maṇlë

´ ‘seer’

(maṇlë

´ = ‘speaker’) (Strand 1999b; 1999a).

Pr. pəṣkí (Buddruss, Degener 2016: 752).

Title of the shaman or medium possessed by the gods in certain religious

contexts. A connection with OIA pas´-ya- ‘to see’ (only possible as a loanword)

is rightly rejected by Parkes (1986: 157) in view of the retroflex ṣ. This root

for ‘to see’ is in any case absent from the Nuristani languages. I cannot think

of a convincing etymology that would explain both the Kt. and the Pr. term,

but it seems possible that there is a relation with the verb root Kt. pṣu- ‘to

sleep’ < *pra-sup-, in which case it would not necessarily be an Indo-Aryan

loanword.

3.)

Kt. NE/W dé-bëlōlë; SE dé-balōlë.

Title of a priest responsible for the recitation of hymns. Literally ‘god

caller,’ a transparent formation of the word de ‘god’ and a masculine imperfec-

tive participle (ending -lë) of the verb SE balṓ-, NE/W bëlṓ- ‘to call [a spirit or

dog]’ (cf. Strand 1999b). This verb is related to Pashto baləl ‘to call,’ probably

a loanword from Indo-Aryan connected with the verbs collected by Turner

(1962) under T.9321 (Morgenstierne 2003: s.v. bal-/bol-).

The phrase has been wrongly segmented as “deb-lol” vel. sim. and sub-

sequently wrongly etymologized by previous authors (cf. Turner 1962: 6574).

Fussman (2012: 74) corrected his earlier fanciful etymology from *devya-lalla-

‘divine inarticulate noise’ (Fussman 1977: 28) to an analysis as SE dé-vallalë

‘god speaker’ at the advice of Pierre Reichert, but this is still clearly not the

correct verb root. He suggests that, since the sound change *v > b is otherwise

unattested, the title must be an especially ancient compound. This conclusion

(and even the assumption of such a sound change) does not follow, and in fact

the opposite is the case.

4.)

Kt. NE mí-moč [<mī moc> (Klimburg 2008: 399)]; W mí-moč, SE mǘ-moč

(Strand 1999b; 1999a).

NKal. N mu (Strand 1999c; Tāza 2017: s.v. ūrtā = mū); <mõu> (Klimburg

1999: 391).

A. W <gul-mõu> (Klimburg 1999: 391).

348 Jakob Halfmann

Title of a prestigious feast-giving rank. The Nuristani Kalasha title mu was

in use in Nisheygram, but in Waigal the corresponding title was the unrelated

urtā. What seems clear is that the Kt. and A. terms are compounds, the former

with the second element moč ‘husband; man (in compounds)’ (< PIIr. *martia-

‘mortal, man’) and the latter with the first element gol ‘country, valley.’ The

remaining element can be compared between all three languages. Strand

(1999b) connects these terms to OIA mukuṭa- ‘diadem,’ but this is semantically

doubtful and leaves the lack of a reflex of -ṭ- unexplained.

In monosyllables the correspondence Kt. NE/W i ~ Kt. SE ü can regularly

reflect older *ŭ

- after labials, e.g. Kt. NE/W min, SE mün ‘forehead’ ~ MIA

*munddha- < muddha- (with “nasal condensation” as described in section 3)

< OIA mūrdhan- ‘head, peak.’ In NKal. N we would normally also expect

ü as a reflex of older *ŭ

- (cf. mül ‘price’ ~ OIA mūlya- ‘price’), but unlike in

Kt. the sound change *u > ü has also affected unstressed syllables in NKal.

N (cf. müsülmān, borrowed from Prs. musulmān). In this context, NKal.

N mudá ‘neck’ ~ MIA *munddha-ka- (with “nasal condensation”) suggests

that the change *u > ü may not have taken place between two nasals. A. has

not undergone a change *u > ü at all. The root vowel of the feast-giver title

may therefore have been *u, if we can assume that a nasal originally fol-

lowed. Proto-Indo-Iranian single intervocalic *-n- becomes vowel nasaliza-

tion in Katë, Nuristani Kalasha and Ashkun when it ends up in word-final

position. This nasalization becomes non-contrastive in cases where a nasal

consonant precedes the vowel and is therefore usually not transcribed in these

cases.

It is therefore possible that the title derives from an equivalent of OIA muni-

‘sage.’ This word probably originates from an Indo-European root connected

with silence/muteness (Armenian mownǰ, Greek myndós ‘mute’) and there are

no Iranian cognates (Mayrhofer 1996: 362). In the meaning of a prestigious

title it is most likely to be an Indo-Aryan loanword.

5.)

Kt. SE drum, translated as Hind. dastūr ‘order, guideline, rule’ (Morgen-

stierne 1951: 187); SE ċár ō ba-drúm ye- ‘to become fatherless,’ anë

´n ō ba-drúm

ye- ‘to become motherless;’ nadrë

´m ye- ‘to menstruate’ (Strand 1999b).

Pr. naḍə´m, naḍúm ‘improper, indecent, wrong, unlawful.’

Of the above terms and expressions, Kt. drum had already been related to

OIA dharma- by Morgenstierne (1951: 187). The others almost certainly con-

tain the same element. Kt. se SE ċár ō ba-drúm ye- and anë

´n ō ba-drúm ye- ‘go

into the drum of the fatherless/motherless.’ Kt. SE nadrë

´m and Pr. naḍə´m con-

tain the negation na- followed by the same element. The development of the

vowel *a to ë in the stressed second syllable of disyllables is expected in Kt.

SE, as is the development dr > ḍ in Pr.

Nuristani Theonyms in Light of Historical Phonology 349

The term dharma- is a central one in Indian religions and it has received

different interpretations throughout the ages in Hinduism, Buddhism and Jain-

ism. It originates in the Vedic word dhárman- ‘support, foothold, order,’ which

took the form dhárma- from the Atharvaveda onwards. Indirect cognates

(~ Vedic dharmán-) exist in Iranian, but have very different meanings (e.g.

Middle Persian darman ‘medicine’). The meaning of the presumable Proto-

Indo-Iranian word *dhárman- would therefore have been very different and al-

most certainly free of the religious connotations of later Sanskrit dharma-. An

inherited reflex of this Proto-Indo-Iranian word in the Nuristani languages

could consequently not have led to the attested meanings. The sense that

emerges from the various phrases and expressions listed above is one of a gen-

eral order of the world—Hindi-Urdu/Persian dastūr—, which is interestingly

the exact same word the non-Muslim Chitral Kalasha use to describe their ‘tra-

ditional order’ (Cacopardo 2016b: 41-42),43 but also that of an individual fate

(e.g. that of an orphan), and that of a law that should not be violated. Violations

of this law seem to include menstruation (considered impure and polluting in

pre-Islamic Nuristan), but also acts like eating a ritual dish without praying be-

forehand—Pr. lust lēnjǝwógnī oyinīg naḍǝm rē-s ‘eating it without having

raised the hands (in pre-Islamic prayer) was unlawful/improper’ (Buddruss,

Degener 2016: 437-438).

The same breadth of meanings is included within the dharma-concept of

classical Hinduism: “In its basic meaning, the Hindu concept of dharma des-

ignates a comprehensive complex of rules for individual and social behavior.

Its observance not only guarantees the ritual purity of the person or group

[…], but also is significant for the maintenance of the social and cosmic order.”

(Strauch 2009: 736). This concept of dharma- and also the idea of dharma- as

individual fate and the obligations that go along with it (svadharma-), which

is apparent in the expressions referring to the drum of an orphan, are distinctly

post-Vedic conceptions (cf. Strauch 2009: 737).

CONCLUSIONS

The above survey has shown that the majority of the religious vocabulary

of pre-Islamic Nuristan, with the notable exception of some Prasun theonyms

with mostly unclear etymologies and some possibly inherited terms connected

with religious functions, has its origin in Indo-Aryan languages and in the re-

ligious terminology of post-Vedic Hinduism.

43 The man who recited the hymn to Móne which includes the word drum for Morgenstierne

and gave the translation dastūr was Čanlü, who was in fact a bilingual Chitral Kalasha man from

Urtsun (see Cacopardo A.M., Cacopardo A.S. 2001: 266, 269, figs. 111, 273).

350 Jakob Halfmann

Based on the explanation of the Nuristani theonyms as independent and

direct linguistic inheritances from Proto-Indo-Iranian, previous studies (esp.

Fussman 1977; 2012) had concluded that the pre-Islamic religion of Nuristan

must have been independently transmitted since ancient times and that it could

be used as evidence for the reconstruction of Proto-Indo-Iranian religion. Fol-

lowing on from this tradition of research, one may now be tempted to conclude

the opposite—that there is nothing pre-Vedic in the pre-Islamic religion of Nur-

istan, since the names are predominantly post-Vedic borrowings. But what does

the linguistic origin of the name of a certain god or religious function really

mean for religious history? What is the “essence” of a deity aside from its name

and epithets? Is it the rituals associated with its worship, the mythological

stories told about it, or its iconographic attributes? As a linguist, I cannot hope

to make a meaningful contribution to the anthropological or religious side of

these questions, but it seems likely to me that all of these elements are subject

to constant re-interpretation. The presence of borrowed names at least means

that at some point there was cultural contact and that influences came from the

side of the speakers of the donor language.

Considering the large amount of shared mythological motifs and attributes

between the Indian gods and their namesakes in Nuristan, there seems to be no

good reason to assume that these names are merely new labels slapped onto

pre-existing deities. From the perspective of general pricinciples of lexical bor-

rowing, the use of a borrowed name for a borrowed concept is the most usual

case. It is also common that terms for borrowed concepts are built up from pre-

existing lexical elements. What would be far more unusual, though perhaps not

impossible, is a borrowed name suddenly being used for a pre-existing concept

that has not changed at all. It would be even less expected that this should

happen on a large scale and in each case with complete replacement of the pre-

existing terms for the concept.

This is not to say that older beliefs and religious rituals could not have re-

mained in place in Nuristan next to the influences from the Indic sphere. After

all, there is reason to suspect that some of the socio-religious practices of Peri-

stan are both distinctive and of considerable age (cf. Cacopardo 2023: 66-67).

In the pre-Islamic religion of Nuristan it seems to be especially the rituals and

forms of worship (as opposed to the mythology) that are derived from older

local models: Already the Young Avestan Yašts (c. 1st millenium BCE) reference

the blood sacrifices of the daēuua-worshipping Vii āmburas, who burn juniper

(hapərəsī-) as part of their rituals, a practice that is widely attested in the pre-

Islamic religions of Peristan (extending much further east than Nuristan), and

seemingly only there (cf. Morgenstierne 1931: 40-41; Schwartz 1990). Inter-

estingly, we find inherited vocabulary precisely in this area—Kt. NE sëréċ, SE

saréċ, W sëvréċ, Pr. ṣoẓ ‘juniper’ is an exact cognate of the Avestan hapərəsī

(< PIIr. *sapr ćī-) used by the Viiāmburas. Another place to look for ancient in-

heritances is in the etymologically unclear Prasun theonyms, some of which

may not be calques. Still, their identification with the Indo-Aryan names used

Nuristani Theonyms in Light of Historical Phonology 351

in the other Nuristani languages probably means that any older features that

these deities may have had, will have been overlaid with interpretations fol-

lowing the Indo-Aryan models. The example of Māra also gives us cause to

consider influences from Bactria.

I would conclude that the study of the pre-Islamic mythology of Nuristan

more likely gives insight into popular expressions of Hinduism as practiced on

the frontier of Gandhara, than it provides input for the reconstruction of Vedic

and pre-Vedic religion. Though I would assume that it developed out of forms

of post-Vedic Hinduism, it still seems possible to uphold one part of Fussman’s

(1977) analysis, also endorsed by Cacopardo, Cacopardo (2001: 28), namely

that these were forms of Hinduism that had not undergone a “brahmanical theo-

logical elaboration” (“élaboration théologique brahmanique”).

Based on the likely timeframe for the involved deities and concepts, I

would hypothesize that the first few centuries CE, possibly up until the estab-

lishment of Islam, were probably a time of relative cultural openness to and

interaction with the wider world in Peristan (or at least in its Nuristani-speaking

parts). In this time, the predecessors of the Nuristani languages must also have

borrowed terms for ‘law’ and ‘judge’ from Bactrian (Halfmann 2023), which

suggests a certain degree of contact with Bactrian-speaking state institutions,

perhaps those of the Kushan empire.

The mythologies of the mountain-dwellers may have been marginal forms

of Gandharan Hindu mythology from the start, but their isolation from new de-

velopments in India due to the expansion of Islam means that they may have

in some form preserved ideas that were obscured by later re-interpretation in

the heartland of Hinduism, which could be helpful in studying the historical

trajectory of the religion. Their increasing isolation within the local context

and the interaction with the increasingly dominant Islam can probably explain

some of the more peculiar features that were documented at the time of Euro-

pean contact in the 19th century.

352 Jakob Halfmann

BIBLIOGRAPHY

Allen, N.J. (1991) Some Gods of Pre-Islamic Nuristan. Revue de l’histoire des religions,

208, 2, pp. 141-168. https://doi.org/10.3406/rhr.1991.1679

Cacopardo, A.M., R.L. Schmidt, eds. (2006) My Heartrendingly Tragic Story: Shaikh Mu-

hammad Abdullah Khan ‘Azar.’ Instituttet for sammenlignende kulturforskning.

Novus forlag. Oslo.

Azarnouche, S., C. Redard, eds. (2012) Yama/Yima : variations indo-iraniennes sur la

geste mythique (Variations on the Indo-Iranian Myth of Yama/Yima). Collège de

France – Publications de l’Institut de Civilisation Indienne 81. Paris.

Baums, S., A. Glass (2002) A Dictionary of Gāndhārī. <https://gandhari.org/n_diction-

ary.php/> accessed 20 March 2023.

Bisschop, P. (2009) Śiva. In Jacobsen 2009: Vol. 1, 741-754.

Böhtlingk, O., R. Roth (1855) Sanskrit-Wörterbuch. St. Petersburg.

Boyce, M. (1986) Apąm Napāt. In Encyclopædia Iranica, online version. <https:

//www.iranicaonline.org/articles/apam-napat/> accessed 20 March 2023.

Brockington, J. (2014) Indra in the Epics. Studia Orientalia Electronica, 94, pp. 67-82.

Buddruss, G. (1977) Nochmals zur Stellung der Nūristān-Sprachen des afghanischen

Hindukusch. Münchener Studien zur Sprachwissenschaft, 36, pp. 19-38.

Buddruss, G. (1983) Spiegelungen der Islamisierung Kafiristans in der mündlichen

Überlieferung. In P. Snoy, ed., Ethnologie und Geschichte: Festschrift für Karl

Jettmar, pp. 73-88. Beiträge zur Südasienforschung 86. Steiner. Wiesbaden.

Buddruss, G. (n.d.) Wamai-Vokabular. Unpublished Manuscript from the A.L. Grjun-

berg archives. St. Petersburg.

Buddruss, G., A. Degener (2015) Materialien zur Prasun-Sprache des afghanischen

Hindukusch. Vol. 1: Texte und Glossar. Harvard Oriental Series 80. Harvard Uni-

versity Press. Cambridge, MA.

Buddruss, G., A. Degener (2017) Materialien zur Prasun-Sprache des afghanischen

Hindukusch. Vol. 2: Grammatik. Harvard Oriental Series 84. Harvard University

Press. Cambridge, MA.

Cacopardo, A.M (2023) Peristan: The Counter-Civilization of the Hindukush/Karako-

rum. The State of Ethnohistorical Research. In A. Degener, E. Hill, eds., Proceed-

ings of the International Workshop “The Nuristani Languages in Synchrony and

Diachrony,” 1-2 November 2019, University of Cologne. International Journal of

Diachronic Linguistics and Linguistic Reconstruction, 19, 2022, pp. 61-78.

Cacopardo, A.M., A.S. Cacopardo (2001) Gates of Peristan: History, Religion and So-

ciety in the Hindu Kush. IsIAO. Rome.

Cacopardo, A.S. (2016a) A World in-between. The Pre-Islamic Cultures of the Hindu

Kush. In S. Pellò, ed., Borders: Itineraries on the Edges of Iran. Eurasiatica 5,

pp. 243-270. Edizioni Ca’ Foscari. Venice. https://doi.org/10.14277/6969-100-

3/EUR.

Cacopardo, A.S. (2016b) Pagan Christmas: Winter Feasts of the Kalasha of the Hindu

Kush. London.

Castenfeldt, S. (1996) A pre-Muslim Temple in “Kafiristan:” A Reconstruction Project.

In E. Bashir, Israr-ud-Din, eds., Proceedings of the Second International Hindukush

Cultural Conference, pp. 109-116. Oxford University Press. Karachi.

Clothey, F. (2009) Murukaṉ. In Jacobsen 2009: Vol. 1, 637-646.

Davidson, J. (1902) Notes on the Bas

-ẖgalī (Kāfir) Language. Calcutta.

Nuristani Theonyms in Light of Historical Phonology 353

Degener, A. (1998) Die Sprache von Nisheygram im afghanischen Hindukusch. Neuin-

dische Studien 14. Harrassowitz. Wiesbaden.

Degener, A. (2002) The Nuristani Languages. In N. Sims-Williams, ed., Indo-Iranian

Languages and Peoples, pp. 103-117. Proceedings of the British Academy 116. Ox-

ford University Press. Oxford.

Edelberg, L. (1960) Statues de bois rapportées du Kafiristan a Kābul après la conquête

de cette province par l’émir Abdul Rahman en 1895-1896, transl. A.D. Schlumber-

ger. Arts Asiatiques, 7, 4, pp. 243-286.

Edelberg, L. (1972) Some Paruni Myths and Hymns. Acta Orientalia, 34, pp. 31-94.

Fussman, G. (1977) Pour une problématique nouvelle des religions indiennes anciennes.

Journal Asiatique, 265, pp. 21-70.

Fussman, G. (2012) Qu’y a-t-il dans un nom : Imrā au Nouristan. In Azarnouche, Re-

dard 2012: 67-82.

Grenet, F. (2006) Iranian Gods in Hindu Garb: The Zoroastrian Pantheon of the Bactrians

and Sogdians, Second-eighth Centuries. Bulletin of the Asia Institute, 20, pp. 87-99.

Grenet, F. (2012) Yima en Bactriane et en Sogdiane : nouveaux documents. In Azar-

nouche, Redard 2012: 83-94.

Greule, A. (2013) Theonyme. Namenkundliche Informationen, 101/102, 2012/2013,

pp. 11-21.

Grjunberg, A.L. (1980) Jazyk Kati. Teksty, Grammatičeskij Očerk. Moscow.

Grjunberg, A.L. (1994) Materialy po ėtnografii i mifologii Nuristana iz arxiva

M.S. Andreeva. In Festschrift. Georg Buddruss zur Vollendung des 65. Lebensjahres und

zu seiner Emeritierung, pp. 83-107. Studien zur Indologie und Iranistik 19. Reinbek.

Grjunberg, A.L. (1995a) Nuristanskie teksty iz arxiva M.S. Andreeva. In J.V. Vasil’kov,

N.V. Gurov, eds., Sthāpakaśrāddham. Sbornik statej pamjati G.A. Zografa, pp. 144-

161. St. Petersburg.

Grjunberg, A.L. (1995b) Kafirskij gimn bogu Monu (iz arxiva M.S. Andreeva). Peter-

burgskoe Vostokovedenie, 7, pp. 605-620.

Halfmann, J. (2021) Terminological Proposals for the Nuristani Languages. Himalayan

Linguistics, 20, 1, pp. 28-64. https://doi.org/10.5070/H920150079

Halfmann, J. (2023) Advances in the Historical Phonology of the Nuristani Languages.

In A. Degener, E. Hill, eds., Proceedings of the International Workshop “The Nur-

istani Languages in Synchrony and Diachrony,” 1-2 November 2019, University of

Cologne. International Journal of Diachronic Linguistics and Linguistic Recon-

struction, 19. International Journal of Diachronic Linguistics and Linguistic Re-

construction, 19, 2022, pp. 113-133.

Halfmann, J. (2023) Lād “Law:” A Bactrian Loanword in the Nuristani Languages.

Bulletin of the School of Oriental and African Studies, 86, 3, pp. 505-510.

https://doi.org/10.1017/S0041977X23000836

Hatley, S. (2007) The Brahmayāmalatantra and Early Śaiva Cult of Yoginīs. Disserta-

tion. University of Pennsylvania.

Hatley, S. (2013) What Is a yoginī? Towards a Polythetic Definition. In I. Keul, ed.,

“Yoginī” in South Asia: Interdisciplinary Approaches. Routledge Studies in Asian

Religion and Philosophy 9. London.

Henning, W.B. (1946) The Sogdian Texts of Paris. Bulletin of the School of Oriental and

African Studies, 11, 4, pp. 713-740. https://doi.org/10.1017/S0041977X00089783

Herrlich, A., W. Lentz, W. Roemer, K. von Rosenstiel (1937) Geographische Bemer-

kungen. In A. Scheibe, ed., Deutsche im Hindukusch. Bericht der Deutschen Hin-

354 Jakob Halfmann

dukusch-Expedition 1935 der Deutschen Forschungsgemeinschaft, pp. 295-351.

Deutsche Forschung: Schriften der Deutschen Forschungsgemeinschaft (Neue

Folge) 1. Berlin.

Hock, H.H. (2021) Principles of Historical Linguistics. 3rd ed. Berlin. https://doi.org/

10.1515/9783110746440.

Holzwarth, W. (1994) Sich verständlich machen. Tak und Shamlar aus Kamdesh be-

antworten einen Fragebogen des Generals Auguste Court zum „kafirischen Lebens-

stil“. In C. Elsas, ed., Tradition und Translation. Zum Problem der interkulturellen

Übersetzbarkeit religiöser Phänomene. Festschrift für Carsten Colpe zum 65. Ge-

burtstag, pp. 179-199. Berlin.

Hopkins, E.W. (1915) Epic Mythology. Strassburg.

Hughes, T.P. (1883) Visit of an Afghan Christian Evangelist to Kafiristan. Church

Missionary Intelligencer, 8, pp. 418-425.

Humbach, H. (1975) Vayu, Śiva und der Spiritus Vivens im ostiranischen Synkretismus.

In Monumentum H.S. Nyberg, Vol. I, pp. 397-408. Teheran-Liège-Leiden.

Jacobsen, K., ed. (2009) Brill’s Encyclopedia of Hinduism. 6 vols. Leiden.

Jettmar, K. (1975) Die Religionen des Hindukusch. Die Religionen der Menschheit 4,

1. Stuttgart.

Kapp, D. (1983) Ṣaṣṭhī – Kult und Legende einer indischen Volksgöttin. Zeitschrift der

Deutschen Morgenländischen Gesellschaft, 133, 2, pp. 300-320.

Klimburg, M. (1975) Zur Religion des Ashkungebietes. In Jettmar 1975: 162-172.

Klimburg, M. (1999) The Kafirs of the Hindu Kush. Art and Society of the Waigal and

Ashkun Kafirs, Vol. 1. Stuttgart.

Klimburg, M. (2002) The Arts and Culture of Parun, Kafiristan’s “Sacred Valley.” Arts

Asiatiques, 57, pp. 51-68.

Klimburg, M. (2008) A Former Kafir Tells his “Tragic Story.” Notes on the Kati Kafirs

of Northern Bashgal (Afghanistan). East and West, 58, 1-4, pp. 391-402.

Marquart, J. (1901) Ērānšahr nach der Geographie des Ps. Moses Xorenacʿi. Mit histo-

risch-kritischem Kommentar und historischen und topographischen Excursen. Berlin.

Mayrhofer, M. (1992) Etymologisches Wörterbuch des Altindoarischen, Vol. 1. Indo-

germanische Bibliothek, 2. Reihe: Wörterbücher. Universitätsverlag Winter. Hei-

delberg.

Mayrhofer, M. (1996) Etymologisches Wörterbuch des Altindoarischen, Vol. 2. Indoger-

manische Bibliothek, 2. Reihe: Wörterbücher. Universitätsverlag Winter. Heidelberg.

Melabar, M.A. (1978a) The Role of Endr in the Mythology of Ancient Nuristan (part 1).

Afghanistan. Historical and Cultural Quarterly, 31, 1, pp. 49-58.

Melabar, M.A. (1978b) The Role of Endr in the Mythology of Ancient Nuristan (part 2).

Afghanistan. Historical and Cultural Quarterly, 31, 2, pp. 33-49.

Moeller, V. (1984) Die Mythologie der vedischen Religion und des Hinduismus. In

H.W. Haussig, ed., Götter und Mythen des Indischen Subkontinents, 1-204. Wör-

terbuch der Mythologie V. Stuttgart.

Morgenstierne, G. (1929) The Language of the Ashkun Kafirs. Norsk Tidsskrift for

Sprogvidenskap, 2, pp. 192-289.

Morgenstierne, G. (1931) Notes on Balochi Etymology. Norsk Tidsskrift for Sprogvi-

denskap, 5, pp. 37-53.

Morgenstierne, G. (1933) Neupersisch rūda und Verwandtes. Zeitschrift für die verglei-

chende Sprachforschung auf dem Gebiete der Indogermanischen Sprachen, 61, 1/2,

pp. 29-36.

Nuristani Theonyms in Light of Historical Phonology 355

Morgenstierne, G. (1934) Additional Notes on Ashkun. Norsk Tidsskrift for Sprogvi-

denskap, 7, pp. 56-115.

Morgenstierne, G. (1949) The Language of the Prasun Kafirs. Norsk Tidsskrift for

Sprogvidenskap, 15, pp. 187-334.

Morgenstierne, G. (1951) Some Kati Myths and Hymns. Acta Orientalia, 21, 3,

pp. 161-189.

Morgenstierne, G. (1968) Mythological Texts from the Kates of Nuristan. In Mélanges

d’Indianisme à la mémoire de Louis Renou. Publications de l’Institut de Civilisation

Indienne, Série in-8° 28. Paris.

Morgenstierne, G. (1973) The Kalasha Language: Texts and Translations, Vocabulary and

Grammar, Appendix. Indo-Iranian Frontier Languages 4. Universitetsforlaget. Oslo.

Morgenstierne, G. (2003) A New Etymological Vocabulary of Pashto, Compiled and

Edited by J. Elfenbein, D. MacKenzie, N. Sims-Williams. = Beiträge zur Iranistik,

23. Reichert. Wiesbaden.

Oberlies, T. (1998) Die Religion des Rgveda. Erster Teil: Das religiöse System des

Rgveda. Vienna.

Palwal, A.R. (1968) History of Former Kafiristan, Part Two: Polytheism of the Kafirs.

Afghanistan. Historical and Cultural Quarterly, 4, pp. 61-88.

Palwal, A.R. (1970) History of Former Kafiristan. The Kafir Religious Practitioners

and their Rituals. Afghanistan. Historical and Cultural Quarterly, 23, 4, pp. 24-36.

Parkes, P. (1986) Etymological Glossary of Kafiri Deities, Ritual Terms and Priests.

In K. Jettmar, ed., The Religions of the Hindukush, Vol. I: The Religion of the Ka-

firs, the Pre-Islamic Heritage of Afghan Nuristan, transl. A. Nayyar, pp. 151-158.

Warminster.

Parkes, P. (1991) Temple of Imra, Temple of Mahandeu: A Kafir Sanctuary in Kalasha

Cosmology. Bulletin of the School of Oriental and African Studies, 54, 1, pp. 75-103.

Rix, H., M. Kümmel, T. Zehnder, R. Lipp, B. Schirmer (2001) Lexikon der indogerma-

nischen Verben. Die Wurzeln und ihre Primärstammbildungen. 2nd ed. Wiesbaden.

Robertson, G.S. (1896) The Káfirs of the Hindu-Kush. London.

Salomon, R. (2001) “Gāndhārī Hybrid Sanskrit:” New Sources for the Study of the

Sanskritization of Buddhist Literature. Indo-Iranian Journal, 44, 3, pp. 241-252.

Schwartz, M. (1990) Viiāmburas and Kafirs. Bulletin of the Asia Institute, 4, pp. 251-255.

Sims-Williams, N. (1998) A Bactrian Deed of Manumission. Silk Road Art and Archae-

ology: Journal of the Institute of Silk Road Studies, Kamakura, 5, 1997/1998,

pp. 191-211.

Sims-Williams, N. (2007) Bactrian Documents from Northern Afghanistan, Vol. 2:

Letters and Buddhist Texts. London.

Sims-Williams, N. (2010) Iranisches Personennamenbuch, Vol. 2: Mitteliranische Per-

sonennamen, Fasc. 7: Bactrian Personal Names. Österreichische Akademie Der

Wissenschaften, Philosophisch-Historische Klasse, Sitzungsberichte 806. Verlag

der Österreichischen Akademie der Wissenschaften. Vienna.

Söhnen-Thieme, R. (2009) Yama. In Jacobsen 2009: Vol. 1, 807-815.

Strand, R.F. (1999a) kâtʹa vari (ktʹi vřâ·i vari) Lexicon (Last edited 2011). <http://nur-

istan.info/lngFrameL.html/> (no direct link) accessed 20 March 2023.

Strand, R.F. (1999b) kâmv’iri Lexicon (Last edited 2019). <http://nuristan.info/lngFra-

meL.html/> (no direct link) accessed 20 March 2023.

Strand, R.F. (1999c) kalaṣa alâ (nišei alâ) Lexicon (Last edited 2011). <http://nuristan.

info/lngFrameL.html/> (no direct link) accessed 20 March 2023.

356 Jakob Halfmann

Strand, R.F. (2008) saňu vi:ri Lexicon (Last edited 2011). <http://nuristan.info/lngFra-

meL.html/> (no direct link) accessed 20 March 2023.

Strauch, I. (2009) Dharma. In Jacobsen 2009: Vol. 2, 736-743.

Tāza, S. (2017) Farhang-i Zabān-i Nuristānī (Kalaṣə-alā): Maʿnā wa Tašrīḥ-i Luġāt,

bā Zabānhā-yi Pax´tō wa Darī. Kabul.

Trail, R., G. Cooper (1999) Kalasha Dictionary—with English and Urdu. Islamabad.

Tremblay, X. (2015) Les daēuuas vus du Pamir et d’Israël : naïves réflexions sur des

faits bien connus. Notes de lecture avestiques VI. In P. Swennen, ed., Démons ira-

niens. Actes du colloque international organisé à l’Université de Liège les 5 et 6

février 2009 à l’occasion des 65 ans de Jean Kellens, pp. 17-50. Liège.

Turner, R.L. (1962) A Comparative Dictionary of the Indo-Aryan Languages. 4 vols.,

3 Suppl. London.

White, D.G. (2003) Kiss of the Yoginī: “Tantric Sex” in Its South Asian Contexts. Chi-

cago.

White, D.G. (2009) Yoginīs. In Jacobsen 2009: Vol. 1, 823-827.

Zamān Khān, R.F. Strand (1998) Topics in Vâsi Ethnography. <http://nuristan.info/Nur-

istani/Vasi/VasiCulture/ZamanXan.html/> accessed 20 March 2023.

Nuristani Theonyms in Light of Historical Phonology 357

ResearchGate has not been able to resolve any citations for this publication.

Advances in the historical phonology of the Nuristani languages

Article

Full-text available

Jan 2022

Jakob Halfmann

This article illustrates the potential of historical phonology to generate insights for various other fields of research concerned with Nuristani languages. Three examples are discussed: The names of pre-Islamic deities, the areal spread of retroflex affricates and the unclear classification of Dameli. (Published as part of the proceedings of the 2019 Cologne Nuristani Workshop)

Terminological Proposals for the Nuristani languages

Article

Full-text available

Nov 2021

Jakob Halfmann

Recent years have seen an increase in the variety of language names used for the Nuristani languages (Indo-Iranian), which are spoken in Eastern Afghanistan and to some extent across the border in Northern Pakistan. This increase is driven by efforts to recognize local practices, but it has also created confusion and inconsistencies in many areas, making terminological justifications a necessary part of every publication regarding these languages. A unified terminology would not only be convenient for linguists and other researchers - in so far as scientific usage can influence colloquial practices, it would also increase the visibility and recognizability of the languages of this little-known language family on the national and international level, thus also potentially benefiting the native speaking communities. The proposals given in this paper aim to create an internally consistent terminology that is scientifically precise, accurate, and recognizable for speakers, while also being congruent with the naming principles put forward for the Glottolog database.

PAGAN CHRISTMAS. Winter Feasts of the Kalasha of the Hindu Kush. Gingko Library. London 2016

Cover Page

Full-text available

Oct 2016

Augusto Cacopardo

Vayu, Śiva und der Spiritus Vivens im Ostiranischen Synkretismus

Chapter

Jan 1975

Helmut Humbach

Principles of Historical Linguistics

Book

Oct 2021

Hans Henrich Hock

Les daēuuas vus du Pamir et d’Israël Naïves réflexions sur des faits bien connus

Chapter

Jan 2015

Xavier Tremblay

Some Paruni myths and hymns

Article

Lennart Edelberg

Yima en Bactriane et en Sogdiane: nouveaux documents

Book

Jan 2012

Frantz Grenet

Yama/Yima : Variations indo-iraniennes sur la geste mythique

Book

Jan 2012

The Brahmayamalatantra and early Saiva cult of yoginis

Article

Jan 2007

Shaman Hatley

The present thesis comprises a study of the Brahmay&amacr;malatantra, a scripture of early medieval tantric or "esoteric" &Sacute;aivism, and its cult of yogin&imacr;s: flying, shapeshifting female deities whose occult powers practitioners sought in visionary, transactional encounters. Composed prior to the ninth century, and perhaps considerably earlier, this unpublished work of one-hundred and one chapters and more than 12,500 verses constitutes one of the most significant sources for the study of early &Sacute;aiva ritual and goddess cults. After introducing the text, the tradition, and the figure of the yogin&imacr;, chapters 2 and 3 review the extant literary, art-historical, and epigraphic sources concerned with yogin&imacr;s, with a focus on the background and early development of their &Sacute;aiva cult. It is within this context that the Brahmay&amacr;mala is situated. Particular problems addressed include the relationship between the yogin&imacr; cult of the Brahmay&amacr;mala and the Brahmanical Mother goddesses (m&amacr;tr&dotbelow;), the post ninth-century temples of yogin&imacr;s, early tantric &Sacute;aiva literature, and the Buddhist yogin&imacr;tantras. Chapter 4 investigates the Brahmay&amacr;mala's form, textual strata, provenance, and social and geographic horizons, while chapter 5 examines the position the text articulates for itself within the &Sacute;aiva tradition. Part II of the dissertation consists of critical editions and translations of several chapters of the BraY&amacr;, which appear in print for the first time.

Nuristani Theonyms in Light of Historical Phonology

Abstract

Recommended publications

Nuristani Theonyms in Light of Historical Phonology

Lād “law”: a Bactrian loanword in the Nuristani languages

Yoginīs in the Flesh: Power, Praxis, and the Embodied Feminine Divine

Advances in the historical phonology of the Nuristani languages