John Mock & Kimberley O'Neil

Dards, Dardistan, and Dardic:
an Ethnographic, Geographic, and Linguistic Conundrum

The Karakoram presents exceptional ethnic, geographic, and linguistic diversity, where high mountains and intervening deep river gorges often mark the boundary lines of ethnic and linguistic features. This topography, however, has channeled rather than blocked human movement, so that there has been persistent contact between the various people speaking different languages. The present social and linguistic fabric of the Karakoram is a result of continual processes of convergence and divergence.

A lineal taxonomy of the languages of the Karakoram presents us with a snapshot of genetically related languages at one moment, but does little to illuminate the ongoing change due to contact, and gives no indication of the interrelationship between neighboring but genetically unrelated (i.e., not sharing a common ancestor) languages.

Early British efforts placed almost all the peoples and languages of the upper Indus River between Kashmir and Kabul into one group, distinguishing only the ethnically and linguistically Tibetan Balti people as a separate group. This unitary classification, however, obscured the distinct identities of all other peoples and their interrelationships. Not only was there a conflation of ethnic identity, but the resulting conflated ethnicity was then employed as a linguistic classification, which provided the basis for a geographic classification. Each term carried with it social, political, and ideological baggage, resulting in the conundrum of Dard, Dardistan, and Dardic. Looking at how and why this conundrum came about will help us understand the implications of choices we make in categorizing the diversity in the Karakoram.

No people in the Karakoram and vicinity today refer to themselves as Dards, their country as Dardistan, or their language as Dardic. The word dard itself is unknown in any languages of the area, except as a loan word from Persian via Urdu, in which it means "pain". Why should a term with no self-referentiality be so widely used by scholars today? The work of Fussman (1972), Tucci (1977), Müeller-Stellrecht (1979), Jettmar (1980), and Vohra (1982; 1983; 1988; 1989) offers representative examples of the varying usage of the term "dard" by scholars. Graham Clark has questioned the broad application of this term, pointing out that Dardistan is a "non-existent county" and the term Dard "a casual reification … lacking a firm basis either in the ideas of the people themselves, or in the classical sources, … resulting from misconceptions that have arisen from theoretical biases in the colonial literature" (Clark 1977: 325-45). In Clark's view, the ethnographic construction of the country of Dardistan, populated by Dards, is "the genesis of a pedigree" which operates by positing "a continuity between a past literature and a present people that is assumed rather than demonstrated". Schmidt, approaching the question from a linguistic perspective, notes "the term Dardistan suggests a political unity which has never existed" (Schmidt 1984:678). Allan, from a geographical perspective, notes that the promotion of a homogeneous Dardistan from Kashmir to Kabul in the late 19th century "reflected an attempt by supporters of imperial India's 'forward policy' to link the Indian northwestern frontier tracts to Kashmir" (Allan 1993:26). In these discussions of the term Dard, however, important source material from the Karakoram region itself has not been considered, namely the Rajatarangini and recently translated epigraphic material. Hence, it is appropriate to review this question including all available sources: Greek and Roman references; Karosthi and Brahmi epigraphic references; Sanskrit epic and Puranic references; Kashmiri references; British colonial administrative references; the emergence of Dardic as a linguistic classification; and in conclusion, to discuss the referential field for the three terms and their relative validity.

Greek and Roman References

In a well-known and much repeated story, Herodotus (4th century B.C.) mentions a war-like people on the frontier of India, near to whom are found gold-digging ants. Herodotus provides the name Dadikai for one of the groups living on India's frontier, which was then the seventh satrapy of the Achaemenian empire. Writing much later, Strabo (64 B.C. to A.D. 23) and Pliny (A.D. 23 to A.D. 79) repeat Herodotus' story and name the war-like people Dardae. Alexander, whose travels provide much of the data for classical geography of India, apparently did not meet any Dard people, but he did go to a place called Daedala. Curtius reports Alexander fought against people called Assakenoi in Daedala. Tucci assumes the Assakenoi were a Scythican tribe whose name derives from the word for horse (Tucci 1977:29). Herodotus' Dadikai may be the Persian name for the darada given in the Puranic lists, which Strabo and Pliny applied to the war-like people whom they equated with Curtius' Assakenoi. Hence, Herodutus' original citation appears to have been derived from Puranic sources. Finally, Ptolemy gives us a map that shows the Indus River arising in the country of the daradrai (map in McCrindle 1885), a term that appears to be received from Sanskrit epic and Puranic sources.

Sanskrit Epic and Puranic References

These Sanskrit references to Daradas, although they cannot be assigned any historicity, indicate that the Darada were known to those familiar with such texts. Singh cites references in the Vayu, Brahmanda, Markandeya, Vamana, and Padma Puranas (Singh 1972). Daradas are also mentioned in the Brhatsamhita, and in Manu, where they are classified pejoratively as Mlecchas. Mahabharata refers to them as degraded Kshatriyas (XII 35, 17-8 in Singh 1972). Rather than a specific people, the term Dard may have been used to characterize a fierce people, residing in the northwest, outside the boundaries of civilization. Their land is near to the "Strirajaya", the Country of Women. These fantastic and vaguely defined regions and the people who lived in them belong as much to the mythic landscape of ancient India as to the historiographic. David White, in discussing the European, Chinese, and Indian traditions regarding these people, points out that "they are a negativity, a blank space on the fringes of the conceptual map of these traditions' self-centered universes" (White 1991:117).

Epigraphic References

Three inscriptions on rocks along the Indus and Gilgit Rivers in the southern reaches of the Karakoram provide the earliest epigraphic references to Dard kings. One is found on rocks where the present-day road between Gilgit and Skardu crosses the Gilgit River, over a bridge known as the Alam bridge, now called the Farhad bridge. The inscription is in poor Kharosthi, and Fussman has read "daradaraya", meaning "King of the Dards" (Fussman 1978:1-6). The second inscription is found at Chilas Terrace, near to Chilas village along the Indus River, south of the junction of the Gilgit River and the Indus River. It has been discussed by Dani (1983) and more recently by Hinuber (1989). It is in Brahmi script. Hinuber publishes a transliteration srir daranmaharajavaisrava, which he interprets as daran-maharaja "great king of the Dards" (1989:57-8). A third inscription is immediately below the Thalpan bridge over the Indus River on the Thalpan side of the bridge. It is also in Brahmi script. Hinuber publishes a transliteration of daratsu maharaja sri vaisravanasena ssatrudamanah, which he translates as "The glorious Vaisravanasena, the subduer of enemies, great King in the land of the Dards" (1989:59). Hinuber interprets these Brahmi inscriptions as referring to the same king Vaiaravanasena, and dates them to the 4th or 5th centuries A.D. He remarks that this king "is the second oldest king of the Dards known by name, preceded only by the daradaraya mentioned at Alam bridge in a Kharosthi inscription" (1989:59). These inscriptions appear to be the only known self-reference to a Dard people.

Kashmiri References

Kalhana, in Rajatarangini (Chronicle of the Kings of Kashmir), first mentions "the Darada country" as the location of a Vihara built by king Surendra (Stein 1979:I, 93, 17). Surendra is the fifth king after Ashoka, whom Kalhana identifies as the familiar Ashoka of the 3rd century B.C. Daradas are next mentioned, along with Bhauttas and Mlecchas, as impure people during the reign of Mihirakula, calculated as from Laukika dates 2372-2442 (Stein 1979:I, 312-6, 46). Sircar gives dates of A.D. 515to A.D. 545 for Mihirakula, based on an inscription at Gwalior (Sircar 1965:424-5). Stein offers a footnote to the reference to Mihirakula's reign, which is worth quoting in full:

The Daradas are the modern Dards regarding whose territory and ethnography Drew, Jummoo, pp. 393 sqq., may be consulted. Their seats, which do not seem to have changed since the time of Herodotus, extend from Citral and Yasin across the Indus regions of Gilgit, Cilas, and Bunji to the Kishanganga Valley in the immediate north of Kasmir. The tribes inhabiting the latter valley are meant in most passages in which the Chronicle mentions the Daradas or Darads.

In Rajatarangini, these people are mentioned as residing to the north of Kashmir, and as frequently attempting to invade Kashmir and/or intrigue with factions in Kashmir. They are next mentioned during the latter part of the reign of Lalitaditya-Mukt pida, around A.D. 750, who "did not tolerate the continual wine-drinking of the Darads" (Stein 1979:IV, 169). Dards are next mentioned during the reign of Shankaravarman, A.D. 883 to A.D. 902 (Stein 1979:V, 152-155, 206). Kalhana goes on to name several Dard rulers: Acalamangala, during the reign of Ananta of Kashmir, A.D. 1028 to A.D. 1063 (VII, 167), Vidhyadhara Shahi during the reign of Harsa, 1089-1101 A.D. (VII, 913), Jagaddala during the reign of Uccala, A.D. 1101 to A.D. 1111 (VIII, 209), Manidhara during the reign of Sussala, A.D. 1112 to A.D. 1120 (VIII, 614), and Yasodhara during the reign of Jayasimha, A.D. 1128 to A.D. 1149 (VIII, 2454). During Kalhana's own time, the Darad ruler joined forces with an opposition faction and fought against Jayasimha, only to be defeated (VIII, 2764-2873). Clearly, Kalhana identified people living in the Karakoram region northwest of Kashmir as Darads. Whether they were one distinct ethnic group, or whether the term broadly signified the unruly wine drinkers living in the mountains cannot be determined.

Colonial Usages of the Term

Stein's note (Stein 1979:I, 312-316, 46ft) exemplifies the assumption of a link between the people referred to by classical Greek authors as "daradrai" and people in the Karakoram regions northwest of Kashmir. Stein refers to Frederic Drew, a colonial officer, as the ethnographic authority on the Dards as a people. Prior to Drew, however, the Dards were first mentioned by Mir Izzet Ullah, who was the assistant to the chief veterinarian of the East India Company, William Moorcroft. Izzet Ullah was sent by Moorcroft from Kashmir via Leh to Yarkand, in search of horses. His notes were translated from Persian into English and published in the Quarterly Oriental Magazine of Calcutta in 1825. In them he mentions "Dardi, an independent mountain tribe, three or four marches north from Dras, who speak the Pashtu as well as the Daradi language" (Izzet Ullah 1843:286). This would most likely refer to the Astor Valley, across the Deosai Plains from Dras, where today, as then, Shina is spoken and Pushtu would be most improbable. H.H. Wilson, sometime Boden professor of Sanskrit at Oxford University, prepared Moorcroft's own notes for posthumous publication. He added a footnote in reference to Izzet Ullah's report, linking the Dards to classical Greek and Sanskrit accounts, "Few people can be traced through so long a history in the same place as these, as they are evidently the Dáradas of Sanscrit geography, and the Daradae, or Daradrae of Strabo" (Wilson in Moorcroft 1841:II, 266). It is as if Wilson has discovered those very people whom classical geographers led him to expect to find inhabiting the Karakoram.

The connection between past and present thus established, the term became accepted through repeated usage. G.W. Leitner emerged as an unabashed advocate of the ethnographic and political reality of the Dards and Dardistan, writing that "the country is known, since my visit in 1866, as 'Dardistan', [even though] the name 'Dard' was not claimed by any of the race that I met" (Leitner 1983:59). Although Robert B. Shaw mentions in a footnote that "I have heard the Drás people of that [Dard] tribe apply [the name Dard] to their parent stock in Astor under the form Dardé" (Shaw 1878:27ft), no other writer on the area confirmed his assertion. On the contrary, John Biddulph, who spent many years in Gilgit as Political Officer, wrote "the name Dard is not acknowledged by any section of the tribes to whom it is so sweepingly applied" (Biddulph 1880/1977:156). Biddulph recognized Leitner's term Dardistan as "a name founded on a misconception" (Biddulph 1880/1977:8), but accepted the term as a convenient way of designating the difficult, diverse, and largely unknown Karakoram between Kashmir and the Hindukush Range. This usage of the term is curiously parallel to the Sanskrit usage, where it connoted nonspecific ferocious outsiders living in the mountains beyond the borders of civilization. So it is not surprising to find an appropriation of the more fantastic elements of those earlier sources. The 1919 Glossary of the Tribes and Castes of Punjab and North-West Frontier Province informed its readers that "the tribes which occupied the modern Kafiristan, Gilgit, and Chitral were called Pisacha or 'eaters of raw flesh', and traditions of ritual cannibalism still survive among the Shins of Gilgit, the Wai and Bashgal Kafirs and in Dardistan. Indeed the Dards of Gilgit had a reputation among the Kashmiris for cannibalism as late as 1866" (Ibbetson, Maclagan, and Rose 1919:25). This date of 1866 is the date of Leitner's visit to Gilgit, when he heard the well-known legend of the cannibal king of Gilgit, Shri Badat (Mock 1997). Leitner had succeeded in establishing a country, even though its boundaries were most elastic. "Dardistan, or the country of the Daradas of Hindu mythology, embraces, in the narrowest sense of the term, the Shiná-speaking countries (Gilgit, etc.); in a wider sense, Hunza, Nagyr, Yasin, and Chitral; and in the widest, also parts of Kafiristan" (Leitner 1893:1). He apparently felt no need to specify by what criteria his sense of the term was to be expanded.

Linguistic Usage

Leitner's Dardistan, in its broadest sense, became the basis for the classification of the languages in the north-west of the Indo-Aryan (IA) linguistic area (which includes present-day Afghanistan, northern Pakistan, and Kashmir). G.A. Grierson proposed a Dardic sub-family of the Indo-Aryan language family on terms seemingly identical to Leitner's neologism. "Following the example of the extended connotation of the word "Dardistan", it is convenient to give the general name "Dardic" to all Aryan languages spoken in this [whole mountainous] tract [between the Hindukush and the frontiers of India proper]" (Grierson 1919:1). Grierson initially termed these languages Pisacha, but dropped that term in favor of Dardic in the face of objections that Pisacha denoted a cannibal demon in Sanskrit. Grierson cited McCrindle's 1885 account of Greek and Roman references to Daradrai, Derdai, Dardae and Dardanoi, the Rajatarangini references to Darada, the Sanskrit epic references to Darad, and Leitner's coinage of Dardistan as sources for his use of the term Dardic. Grierson classified Dardic as a separate, third branch of Indo-Iranian, and identified three major groups within Dardic: a Kafir group, which included Bashgali, Wai-ala, Wasi-veri or Veron, Ashkund, Kalasha, Gawar-Bati, Pashai, Diri, and Tirahi; a Kho-war or Chitrali group; and a Dard group, which included Shina, Kashmiri, Kashtawari, Poguli, Siraji, Rambani, Kohistani, Garwi, Torwali, and Maiya. A Stammbaum "tree" form of Grierson's classification is reproduced in Masica (Masica 1991:461).

Georg Morgenstierne, who conducted extensive fieldwork in Afghanistan and what is now Pakistan during the early 20th century, revised Grierson's classification by distinguishing between Grierson's Kafir group and the rest of the languages, for which he reserved the term Dardic. Morgenstierne retained Grierson's designation of a third branch of Indo-Iranian for the Kafir languages alone, which "must have separated from the others at a very early date" (Morgenstierne 1961/1983:139), but moved Kalasha, Pashai, and Gawar-Bati to the Dardic group⁽¹⁾. Dardic he regarded as "simply a convenient term to denote a bundle of aberrant Indo-Aryan hill-languages" which "contain absolutely no features which cannot be derived from Old IA" (Morgenstierne 1961/1983:139). Morgenstierne doubted the coherence of Dardic as a linguistic group, remarking that "there is not a single common feature distinguishing Dardic, as a whole, from the rest of the IA languages" (Morgenstierne 1961/1983:139).

This classification is accepted by scholars today, whose further fieldwork has supported and refined Morgenstierne's view (Fussman 1972, Strand 1973, Edelman 1983, Schmidt and Koul 1983, Koul and Schmidt 1984). Strand has proposed the term Nuristani as a courteous replacement for the derogatory term Kafir (i.e., infidel). The following table is based upon Strand's classification (Strand 1973:302), which follows that of Morgenstierne ⁽²⁾. The classifications of language groups and languages marked by an asterisk (*) have been revised by subsequent fieldworkers. These revisions are discussed in the paragraphs following the table.

Dardic Languages (Northwest Indo-Aryan)

Morgenstierne's Chitral Group remains stable, and the close relationship between Khowar and Kalasha has been confirmed by recent work (Bashir 1988). The Kunar and Pashai groups have been largely inaccessible due to political instability in Afghanistan. The Kohistani, Shina, and Kashmiri groups, however, have received considerable attention, and Morgenstierne's classifications have been largely revised.

Buddruss has removed Dumaki from the Dardic languages entirely, and identified it as belonging to the Central Group of Indo-Aryan (Buddruss 1985:29).

Rensch (Rensch 1992) and Hallberg (Hallberg 1992) present data which divides Morgenstierne's Central (Kohistani) Group into a Swat Kohistani Group, consisting of Torwali and Kalami Kohistani (called Bashkarik by Morgenstierne), and an Indus Kohistani Group, consisting of Indus Kohistani (often termed Maiya, an apparent misappelation unknown to speakers of the language), Chilisso, Gowro (not to be confused with Gawri, a pejorative term for Kalami Kohistani), and Bateri. The relationship of Wotapuri and Tirahi to these two sub-groups within Kohistani has not been determined.

Koul and Schmidt note that Morgenstierne's classification "is relatively sketchy so far as Shina and Kashmiri are concerned" (Koul and Schmidt 1984:6). They note that "probably only Poguli (a miscellaneous transitional dialect) and Kashtawari can be considered true dialects of Kashmiri" (1984:19). Siraji and Rambani ⁽³⁾ (also miscellaneous transitional dialects) they do not consider to be dialects of Kashmiri, but to be creoles or possible dialects of Western Pahari (1984:20).

For Shina proper, Koul and Schmidt list seven dialects. Radloff, however, on the basis of phonological and lexical studies, identifies four geographical clusters of Shina: Northern; Eastern; Diamir; and Kohistan, which group around two distinct centers, Gilgit and Chilas (Radloff 1992:130-131). Sawi, spoken along the Kunar River in Afghanistan, belongs to the Shina Group and apparently diverged from Phalura (K.Decker 1992:78). Ushojo, a previously unknown language spoken in an eastern side-valley of Swat, has been identified as belonging to the Shina Group as a separate language from Shina proper (S. Decker 1992:72).

Clearly, an understanding of the languages of the region stretching from Kabul to Kashmir, and their relationship to the rest of the Indic languages, is a formidable task, the outline of which is still emerging. As the British Empire explored its forward frontier, the remarkable diversity in the region became apparent. Leitner's "Dardu" language was soon recognized as a completely separate language, Burushaski. This linguistic isolate, whose relationship to any other language has never been shown, is spoken in the Hunza and Yasin valleys⁽⁴⁾. Other non-Dardic languges of the Karakoram include Balti and Purik, both dialects of Western Tibetan spoken in Baltistan, along the upper Indus River. Wakhi is a Pamiri Iranian language spoken in the upper Hunza Valley and the upper Ishkoman and Yarkhun valleys, adjacent to Afghanistan's Wakhan, from where the Wakhi people migrated.

Koul and Schmidt have remarked on the difficulties presented even by attempting to marshal the bibliographic resources for the Dardic languages, and of the very real difficulties faced by anyone attempting to study them in a field which "stretches over 500 miles east to west and includes four major mountain ranges…and is difficult of access even today" (Koul and Schmidt 1984:5). They have, however, identified phonological, morphological, syntactic and lexical features "peculiar to the Dardic languages…[which]…may be considered to define them, not merely in terms of their geographical distribution, but as a linguistic grouping" (Koul and Schmidt 1984:7-9, 20). They propose these features in order to encourage debate over Morgenstierne's statement that "there is not a single common feature distinguishing Dardic, as a whole, from the rest of the Indo-Aryan languages".

Fussman points out that the term "Dardic and Nuristani languages is a geographic, not a linguistic expression. Taken literally, it allows one to believe that all the languages spoken in Dardistan are Dardic" (Fussman 1972:11). It also allows one to believe that all the people speaking Dardic languages are Dards. The circularity of the logic employed in these designations is evident: the people are Dards, hence the area they live in is Dardistan, and the languages they speak are Dardic, hence the people are Dards. A term used by classical geographers to identify the area inhabited by an indefinite people, and used in Rajatarangini in reference to people outside Kashmir, who, under various leaders, intrigued in Kashmiri politics, has come to have ethnographic, geographic, and even political significance today. Residents of the Karakoram region of Pakistan have currently appropriated the term Dardistan to serve as a rallying point for unity against domination by down-country Punjabi officials.

The indiscriminate and varied usage of these terms is further confused by the addition of a linguistic classificatory referent. Schmidt and Koul note what they term the fallacy of the label "Dardic". "In the absence of reliable comparative data about Dardic languages, a geographic or ethnographic label is frequently applied to a group of dialects. The dialects designated by this label are subsequently treated as a genetically homogenous group" (Schmidt and Koul 1984:3). That the ethnographic and geographic labels themselves are not determinative of any homogeneity is evident. Moreover, the confusion around these labels makes them poor choices for describing specific people or administrative regions today, a tendency more frequently found in Indian scholarship (Vohra 1983, 1989; Chohan 1983). One may state that such usage "is not an attempt to establish the continuity of the Daradas as an ethnic stock over this large span of time" (Vohra 1988:529), but such usage does continue the understandable confusion of earlier British writers, who identified the Darada of the epic and puranic literature and the Darada of the Rajatrangini with the various people living along the upper Indus River.

The term Dardic has gained linguistic currency, and will no doubt be with us for the foreseeable future. We should be aware, however, of the problems involved with it as a term. Schmidt and Koul mention that "tree" classifications, (which are largely phonological), may not be the best way to present the relationships of these languages among themselves and with other IA languages. They suggest analyzing the linguistic situation by eschewing neat groupings and distinctions, and instead seeking an underlying linguistic substratum (1984:7). Masica points out that "tree" diagrams represent a linear development model, in which languages, once they branch apart, do not come into contact with each other (1991:446). This is obviously not the case of the so-called Dardic languages. Wave theory models, which represent the diffusion of features through the drawing of isoglosses, allow us to more clearly see the patterns of linguistic influence among languages. This approach enables the inclusion of genetically different languages of the Karakoram in long contact with each other, such as Balti, Burushaski, and Wakhi. Masica's suggestion of employing "overlapping genetic zones" (1991:460) parallels Koul and Schmidt's call for a different approach to the analysis of these languages, which, hopefully, will be heeded by scholars. By mapping overlapping zones of similarity, cumbersome and misleading unitary group terms such as Dardic can be jettisoned. Recent work along these lines points to an interesting convergence of phonological and syntactic features among the genetically unrelated languages of the Karakoram (Tikkanen 1995). Mapping these similarities and divergences with speakers' own perceptions of their languages offers a way of depicting the languages that corresponds to both linguistic and cultural reality, avoids the use of the problematic terms "Dard" and "Dardic", and more clearly represents the relationships between these languages and cultures of the Karakoram and others of South Asia and Central Asia.

End Notes

1. The Nuristani languages (Morgenstierne's " Kafir" languages) include Kati, Wasi-weri, Ashkun, Kalasha-ala, and Tregami. These languages retain phonological features which cannot be derived from Indo-Aryan, but rather, only from Indo- European. See Masica 1991: 461-3 for a succinct discussion of the basis of this distinction.

2. Strand's classification that appeared in print (Strand 1973) should be regarded as an incomplete version. The table presented here is an updated version of that classification (Strand 1998, personal communication). Strand's most recent publication on the topic, "The Tongues of Peristân", is to appear in the The Gates of Peristan by Alberto and Augusto Cacopardo.

3. Peter Hook, who conducted field work in Ramban village, reports, he " could not discover anything called Rambani. The language of the place is called Zundhari…and the differences between Poguli and Zundhari did not strike me as decisive" (Hook 1997, personal communication).

4. Burushaski's grouping, toegther with Caucasion and Basque, withing a "Macro-Caucasian" phylum, part of a larger "Dene-Caucasin" family, has been proposed (Bengston 1991a, 1991b). Hermann Berger, however, remarks that "the structural similarity (of Burushaski) with Basque and Caucasian…is obvious, but what is still missing are really convincing etymologies…" (Berger 1985:37), a point of view he has recently confirmed (Berger 1995, personal communication).