[This is the original English version of a paper presented in Dari at the Symposium “Nûristân In the Course of History,” held at Afghânistân's Ministry of Information and Culture in Kâbul on 9 September 2013.]

Basic Processes in the Evolution of the Nûristânî Languages

Richard F. Strand

Cottonwood, Arizona, U.S.A.

1.    Introduction: The Five Nûristânî Languages

There are five languages belonging to the Nûristânî branch of the Indo-Irânian (Âryâî) linguistic family.  The one with the most speakers, spanning the north and east of Nûristân Province, is Kâmkʹata-vari, with dialects Western Kâtʹa-vari, Eastern Kâtʹa-vari, Kâmvʹiri, and Mumvʹiri.  A smaller language akin to Kâmkʹata-vari is Vâsi-vari, spoken in the Pârûn Valley.  The sound system of Vâsi-vari has changed much from its earlier form, when it was closer to that of Kâmkʹata-vari.  Together these two languages constitute the Northern Group of Nûristânî languages.

Three closely related languages comprise the Southern Group of Nûristânî languages: Âṣkuṇu-Saṇu-vîri, spoken in the Alingar and middle Pech Valleys, Kalaṣa-alâ, spoken in the Vâigal Valley and environs, and Tregâmî, spoken in the Tregâm (Kaṭâr) Valley off the lower Pech River.

2.    Background: The Comparative Study of the Indo-Irânian Languages

When we say that the Nûristânî languages belong to the Indo-Irânian linguistic family, which in turn is an eastern branch of the Indo-European family of languages, we are making an explicit statement about the descent and evolution of these languages over the course of some 6,000 years, from the time that the ancient Proto-Indo-European language was spoken, most likely in the Volga River Valley, through its breakup and the migration of its later Âryâî speakers into their present homelands.  By comparing the sounds used to produce related words in the languages of the region, we can be somewhat precise in establishing which sounds were used in the word that was the ancestor of the related words.  Such “reconstructed” words can then be analyzed to show the overall sound system of the reconstructed “proto” language and the changes that the sounds underwent to reach their pronunciations in the descendant languages.

For example, the Proto-Indo-European word for “ten” can be reconstructed as dek̂m̥ by comparing Latin dekem, Greek deka, Gothic taihun, Sanskrit daša, Kâmkata-vari duć, and Pashto las.  (See Table 1.)  It is clear that all of these have a similar structure, beginning with d, t or l followed by a vowel followed by a consonant, either k, h, (ه) š (ش), ć (څ), or s (س), followed by vowel and a nasal m or n.  By finding many more sets of related words, or “cognates,” we can show that in Gothic (an ancient Germanic language akin to English) a t appears and in Pashto an l appears wherever there is a d in the other languages.  Similarly, Gothic shows h, Sanskrit shows š, Kâmkata-vari shows ć, and Pashto shows s where we find k in Latin and Greek.  Finally, the final m or n of Latin and Gothic disappear after the final vowel of the Greek and the ancient Indo-Irânian languages.  The regularity of sound correspondences among these languages shows that the original sounds were d, k̂ (a “palatalized” k sound made with the tongue pushed forward), and a syllabic m̥.  d changed to t in the Germanic languages and to l in Pashto; k̂ changed to h in Germanic, š in the Indo-Âryan languages, ć in the Nûristânî languages, and s in the Irânian languages; and syllabic m̥ lost its nasality in Greek and the Indo-Irânian languages, leaving a.

2.1.    18^th Century "Discovery" of Sanskrit and Avestan Leads to Reconstruction of Proto-Indo-Irânian

This so-called “comparative method of linguistic reconstruction” was developed over two centuries of Western linguistic scholarship that began with the European “discovery” of the ancient languages of India and Iran in the latter 18^th Century A.D.  By comparing words in Sanskrit (and later Avestan) with words in the ancient European languages Latin, Greek, Gothic, Celtic, and Slavonic, linguists demonstrated that they had all “sprung from a common source,” as the English philologist William Jones stated as early as 1786 A.D.

After further decades of study, linguists showed that Sanskrit and Avestan formed a close-knit group.  This group was named the Indo-Irânian or Âryan branch of the Indo-European languages.   An early form of Sanskrit was the ancestor of all the modern languages of the Indo-Âryan branch of Indo-Irânian languages.  The Irânian languages constitute another branch of the Indo-Irânian languages.  Avestan, the language of Zarathushtra, is the oldest recorded Irânian language, slightly older than the Old Persian inscriptions of the 6^th Century B.C.

Sanskrit and Avestan showed a consistent development of Indo-European k̂ into a spirant sound (š or s) and the merger of the Indo-European vowels e and o with a.  On the basis of these developments 19^th Century linguists reconstructed a tongue-end spirant (fricative) sound of unspecified quality, ś, as the Indo-Irânian sound from which Sanskrit š and Avestan s developed.  Thus the ancient Indo-Irânian word for “ten” was at that time reconstructed as daśa.

2.2.    20^th Century Study of Nûristânî Leads to New Understanding of Proto-Indo-Irânian

The Nûristânî languages did not come to the attention of linguists until the end of the 19^th Century A.D., and it was not until the pioneering field research of the Norwegian linguist Georg Morgenstierne that the true relationship of the Nûristânî languages within the Indo-Irânian family was established (Morgenstierne 1945).

From the Nûristânî evidence, it is clear that the sound corresponding to the Sanskrit and Avestan spirants is a dental affricate sound in Nûristânî (ć, pronounced [ts] [څ]).  An affricate is made with the end of the tongue pressed against the upper teeth or gum ridge and then released into a spirant sound.  Among the Indo-Irânian languages there are numerous examples of affricates being weakened to spirants by losing the initial tight closure of the tongue (for example, Pashto ćalor ‘four’ becomes salor in the East), but examples of spirants changing to affricates are rare.  The Nûristânî evidence compels us to assume that the original Indo-Irânian sound was not a spirant, but an affricate, and that the Nûristânî languages preserve a sound closer to the ancient sound than either the š of Sanskrit or the s of Avestan.

What kind of affricate sound could develop into either a dental spirant s (made by forcing air past the tip of the tongue toward the upper teeth) or a lamino-alveolar spirant š (made by forcing air between the upper front of the tongue [the “blade” or “lamina” of the tongue] and the back of the gum ridge)?  We would expect a dental affricate (ć) to precede the Irânian dental spirant s and a lamino-alveolar affricate (č, pronounced [tš] [چ]) to precede the Indo-Âryan lamino-alveolar spirant š.  Is there any evidence for choosing between dental and lamino-alveolar?

Again, there is abundant evidence among the Indo-Irânian languages for choosing a lamino-alveolar affricate (č [چ]) over a dental one (ć [څ]) as the original sound.  Speakers of Pashto, the Pamir languages, and several Indo-Âryan languages pushed their tongues forward to change an original Indo-Irânian alveolar affricate into a dental one; for example the word for ‘four’ is čâr in Dari but ćalor in Pashto.  I know of no counter-examples of ć changing to č in the region, except after the front vowel i in the Irânian languages Munjî, Ishkâshmî, and Wâkhî.  We may infer that the original Indo-Irânian sound was č, which lost its affrication in Indo-Âryan to become š but was dentalized in Irânian and Nûristânî to become ć before losing its affrication in Irânian to become s.  In light of the Nûristânî evidence, our example of the ancient Indo-Irânian word for “ten” must be reconstructed as dača.

3.    Basic Phonetic Processes: Fronting and Backing

The fronting of the tongue, moving its end from contact with the alveolar ridge to the teeth, is characteristic of the evolution of the Irânian and Nûristânî languages.  Contrarily, the backing of the tongue is characteristic of the Indo-Âryan languages.  But underlying the front or back motion of the tongue is the front or back tensing of the vocal folds (the “vocal cords”).

3.1.    Front and Back Phonation

Muscular tensing of the front of the vocal folds produces “anterior phonation,” which is higher pitched, while muscular tensing of the back of the vocal folds produces “posterior phonation,” which is lower pitched (Catford 1977: 102-103).  Irânian and Nûristânî speakers tense only the front of their vocal folds during speech.  Contrarily, most Indo-Âryan speakers tense the back of their vocal folds to produce unaccented phonation and the front of their vocal folds to produce accented phonation.

3.2.    Lingual Fronting and Backing

The level of vocal-fold tension may be increased to encompass muscles of the larynx and those connecting the larynx to the base of the tongue.  If such tension is from anterior phonation, the tongue is pushed forward.  Contrarily, increased tension from posterior phonation pulls down on the base of the tongue, causing it to retract, and it may further cause the jaw to be pulled back and open.

3.3.    Prognathizing

An extreme type of lingual fronting occurs with “prognathizing,” which is a jutting out of the lower jaw while keeping the tongue's tip pressed firmly against the back of the lower teeth.  This posture places the tongue’s blade against the back of the upper teeth during the production of dental consonants.  Prognathizing occurs in the Southern Nûristânî languages and in other languages spoken near the Pech-Kunaṛ confluence, including Sâfî Pakhto and some Pashaî dialects, as well as in other languages across the Indo-Irânian-speaking region.

4.    Evolutionary Consequences of Fronting and Backing

All languages evolve, but the underlying processes that drive such evolution may change over long periods of time.  Today’s Nûristânî languages are no exception, but the processes that have driven recent change are primarily ones of speeding up speech through anticipation, whereby phonetic features of following sounds are anticipated to the preceding syllable.  The earliest changes – the ones that determined the ultimate relationships between the Nûristânî, Irânian, and Indo-Âryan languages – were primarily ones of lingual and glottal fronting and backing.

4.1.    Fronting

Progressive fronting of the tongue over generations of speakers has produced an evolutionary sequence of stop consonants that began in Proto-Indo-European with labialized dorso-velar kʷ (ﻛﻮ) delabialized to plain dorso-velar k (ک), which was fronted to dorso-palatal k̂ (ﻛﻲ), which was further fronted to lamino-alveolar č (چ), which was fronted by prognathizing to lamino-dental or apico-dental ć (څ).  Each one of these sounds may be found today in Kâmkata-vari: kʷ (ﻛﻮ) in Kti-vřâ·i-vari kuštiʹuv ‘hammer,’ k (ک) in Kâmviri kom ‘Kom,’ k̂ (ﻛﻲ) in Kâmviri vâik ‘lamb,’ č (چ) in puč ‘five,’ and ć (څ) in duć ‘ten.’

Such fronting happened twice in early Indo-Irânian.  (See Table 2.)  The first took early Proto-Indo-European k (ک) to late Proto-Indo-European k̂ (ﻛﻲ) to Indo-Irânian č (چ).  At that point the ancient Indo-Âryan speakers relaxed the č (چ) to š (ش), while the ancient Nûristânî and Irânian speakers continued fronting č (چ) to ć (څ).  Then the early Irânians relaxed the ć (څ) to s (س), while most Nûristânîs maintained the tense affrication of the ć (څ) until today.

The second Indo-Irânian fronting occurred later, after the split-up of Indo-Âryan and Nûristânî-Irânian speakers, but the fronting affected all the Indo-Irânian languages.  Proto-Indo-European kʷ (ﻛﻮ) became k (ک), which before the front vowels i (ﻱ) and e (ې) became č (چ) in Indo-Irânian and further became ć (څ) in Pashto before relaxing to s (س) in eastern Pakhto (early PIE kʷetuores ‘four’ > late PIE ketuores > I-I. k̂atvâras > early Irânian čatvâra > Pashto ćalor > Eastern Pakhto salor).

A third fronting occurred much later, after the Nûristânîs reached their current abodes.  This affected Âṣkuṇu-Saṇu-vîri and, to a lesser extent, Vâsi-vari, which are united via the Pech and Pârun valleys.  Details of these changes may be found on the author's website (http://nuristan.info/Nuristani/NuristaniEvolution.html).

The extreme lingual fronting that accompanies prognathizing is the reason that the lamino-alveolar affricates č (چ) and ǰ (ج) have changed to the dental affricates ć (څ) and ź (ځ) in the evolution of the Indo-Irânian languages.  But prognathizing may come and go.  It caused the aforementioned change in ancient Nûristânî and Irânian, and it appears today in the Southern Nûristânî languages; but in Kâmkata-vari the dental affricates are currently pronounced without prognathizing, with the tip of the tongue hitting the back of the upper teeth.

The shift to exclusively anterior phonation in Irânian and Nûristânî affected the Indo-European so-called “aspirated” consonants, which were produced with posterior phonation and partially constricted vocal folds.  The early Irânian and Nûristânî speakers increased front tension so much that it precluded back tension; and so the aspirated consonants lost their posterior phonation, becoming anteriorly voiced and merging with their counterpart unaspirated voiced consonants.  The strong lingual fronting of Nûristânî also affected the evolution of the vowel systems of those languages.

Successive waves of tongue fronting have apparently emanated out of the Nûristânî-Irânian-speaking region for over five thousand years.  The first two waves encompassed the Indo-Âryan languages during the Indo-Irânian stage of development, so that all the Indo-Irânian languages were affected by them.  But waves of tongue fronting emanated even farther through the ancient Indo-European-speaking communities, causing many or all of the same changes that occurred in the Irânian and Nûristâni languages to occur in the Slavic, Romance, and Germanic languages at later times.

4.2.    Backing

The backing of the tongue in Indo-Âryan probably arose out of the generalization of posterior phonation, which is used to some extent in all Indo-Âryan languages.  As stated before, posterior phonation tends to pull the tongue back.  Through certain sound changes a series of tongue-backed (“retroflex”) consonants, which contrasted with dental ones, arose in early Indo-Âryan.

The posterior voicing and lingual backing of the Indo-Âryan languages had a strong influence on the Irânian and Nûristânî languages that border them.  But because speakers of the latter two language groups use exclusively anterior phonation, they can only back their tongues in imitation of their Indo-Âryan-speaking neighbors.  All the Nûristânî languages have backed (“retroflex”) consonants, as does the Irânian language Pashto.

Very early in the development of Indo-Irânian there was a split between the Irânian and Indo-Âryan languages, as a single group, and the Nûristânî languages.  The former languages backed Indo-European s to š after u, but the Nûristânî languages did not.  Perhaps the Nûristânî pronunciation of u at that time was already somewhat fronted, as in modern Saṇu-vîri.  Thus the word for ‘mouse’ in Kâmkata-vari is musʹa, which contrasts with the backing of the s (س) to š (ش) in Fârsi muš and to ṣ (ښ) in Sanskrit mûṣaka-.  This split between Nûristânî and the other Indo-Irânian languages must have taken place before the split between the Irânian and the Indo-Âryan languages.  Later, the Nûristânî languages sided with Irânian in fronting Indo-Irânian č to ć.  Even later, the Nûristânî languages sided with Indo-Âryan in the further development of their consonants, in ways that are beyond our topic here.  Details may be found on the author’s website at http://nuristan.info/Nuristani/NuristaniEvolution.html.

4.3.    Possible Underlying Causes of Fronting versus Backing

Thus we see that the ancient split of the Indo-Irânian languages into its Nûristânî, Irânian, and Indo-Âryan branches was the result of successive waves of glottal and lingual fronting and backing.  What drove speakers to emphasize these postures of speech?

We may take a clue from folk opinions in Afghanistan about the sound of local languages.  Dari sounds “sweet,” while Pashto sounds “harsh.”  In Dari and even more so in Irânian Fârsi, the tongue is fronted along with strong anterior phonation, producing a raised pitch during speech that is characterized as “sweet.”  In Pashto, a backed tongue, particularly in the East, produces lower-pitched sounds, such as ṣ (ښ) and x (“kh”, [خ]), that sound “harsh” to non-accustomed hearers.  Likewise, the posterior phonation of Indo-Âryan often sounds “harsh” to the ears of outsiders.

It is widely recognized that the opposing modes of “sweetness” versus “harshness” are expressive of conciliation and belligerence, respectively.  At the root of these contrasts are the extremely ancient metaphors of feminine, associated with high pitch, versus masculine, associated with lower pitch.  Prognathizing, with its jutted-out jaw, is probably at its base also a signal of belligerence, employed by otherwise “sweet”-talking Irânians and Nûristânîs to express insolence.  This is not to suggest that today’s speakers are necessarily belligerent or nice, only that such attitudes prevailed at the time when such postures were adopted, and the postures have persisted.  Perhaps encoded in the ancient frontings and backings that define the linguistic branches of Indo-Irânian are waves of confrontation and conciliation that echo ancient ethnic encounters.

5.    Conclusion: the Unique Heritage of the Nûristânî Languages Should Be Nurtured

Thanks to the research of linguists led by Professor Georg Morgenstierne, the position of the Nûristânî languages as a kind of “eldest brother” to the Irânian and Indo-Âryan languages is now clear (see Figure 1.)  Our study of the Nûristânî languages has altered our understanding of the evolution of the Indo-Irânian languages and contributed to our understanding of the processes – such as the ones discussed here – that drive the evolution of the Indo-European languages in general.  It is beyond my topic here to discuss the grammatical, stylistic, and folkloric traits that give the Nûristânî languages an expressiveness rarely perceived by outsiders, but suffice it to say that such traits highlight the uniqueness of the Nûristânî languages within the Indo-Irânian linguistic family.

Figure 1.  Family-Tree Diagram of the Indo-Irânian Languages.

Over the years of my contact with Nûristânîs I have noticed a tendency among them to disparage their native languages in favor of the literary languages Dari and Pashto.  Rather than disparage their languages, Nûristânîs should proudly celebrate them as their unique, 5,500-year-old heritage.  The younger generation of Nûristânîs needs an invigorated interest in studying and sustaining their native languages by developing literacy in those languages.  Nûristânîs must keep in mind that language is the vehicle of culture, and that once a language disappears, so does the culture that it represents.

Bibliography

Catford, John C.   1977  Fundamental Problems in Phonetics. Bloomington, Indiana: Indiana University Press. 

Jones, William  1786  "The Third Anniversary Discourse, on the Hindus," delivered to the Asiatick Society, 2 February 1786.

Morgenstierne, Georg  1945  "Indo-European k' in Kafiri." Norsk Tidsskrift for Sprogvidenskap, 13: 225-238.

Strand, Richard F.  2008, 2010  “The Evolution of the Nuristâni Languages.”  http://nuristan.info/Nuristani/NuristaniEvolution.html.