Rectifying Nuristani orthographic conventions

A persistent issue in Nuristani historical linguistics is the widely inconsistent orthographic variation among different linguists. Since the Nuristani languages are in general sparsely documented and ignored by linguists working within the broader Indo-Iranian (and Indo-European) linguistics, this post will attempt to rectify the different orthographical conventions, following the proposal laid out by Halfmann (2021, 2023, 2024, 2025).

Stress is marked with an acute accent, but is not indicated in monosyllabic words. Orthographic forms are enclosed in angle brackets.

(For abbreviations and sources, see this post.) 

Richard Strand

On his website, Richard Strand lays out a general transcription convention for all cited Nuristani word forms that is explained in detail here. He uses a phonemic transcription, which does not take into account allophonic variation and conditioned sound changes. Thus, he represents Kt. (SE) vuṣṭ "eight" as <ʹusṭ>, without representing the insertion of /v/ before initial u- and /s/ > /ṣ/ before /ṭ/.

Most significantly, he uses <a> to transcribe a central vowel written as ë (earlier <ə>), at least in Ash., Kt. (which becomes /i/ after š, č, and ǰ), and Pr., as in Kt. (SE) <sʹa> for së "year" and <ǰʹa> for ǰi "bowstring". However, he also uses the same sound to transcribe NKal. (N) /a/ [ɐ], hence NKal. (N) <sa> = sa "that". He uses the same symbol to indicate vowel length in Kt., so that Kt. (SE) <ćʹoa> = ċō "branch", but for other languages, vowel length is marked with <ː>, such as Ash. (S) <domʹoː> = domṓ "wind; front of neck".

The usage of <â> to transcribe /a/ in the same three languages is a notable departure from earlier authors, who indicate the same sound as <a>. Thus, Kt. (SE) <dʹâ> = da "take (imp.)", Ash. (S) <gʹâ> = ga "cow". In NKal., it generally transcribes /ā/, as in NKal. (N) <gâ> = gā "cow".

Macrons in earlier sources

Some authors, such as George Morgenstierne and Buddruss & Degener, often transcribe forms with macrons that are no longer seen as phonemic in later sources. Compare the forms for "six" in NKal. (N): <ṣū> (Morgenstierne 1954), <ṣu> (Degener 1998), <ṣu> (Strand 2011) = ṣu. In the Nuristani languages, the contrast between short and long vowels other than /a/ and /ā/ seems to be present only in Kt. (and possibly Ash.)

Transcriptions of dental affricates

The transcriptions of Nuristani dental affricates /ts/ and /dz/ also differ by author. The current transcriptions of these affricates are /ċ/ and /j̈/.

/ċ/ and /j̈/ are transcribed by Morgenstierne as <c> and <j>, by Degener as <ċ> and <j>, and by Strand as <ć> and <ź>.

v and w

There are competing transcriptions for the phoneme /v/ ~ /w/ (no Nuristani language phonemically contrasts /v/ and /w/), which is often realized as [w] ~ [β] ~ [v]. Morgenstierne and Buddruss & Degener prefer <w>, while Strand prefers <v>. For Kt., Halfmann (2024) prefers /v/, and for Ash., Pr., and NKal., he prefers /w/.

ň/ṇ and ř/ṛ

These symbols (ř is also transcribed <ŕ>) usually denote different phonemes in Katë, though ň/ř are written as n̆/r̆ in Halfmann (2024, 2025). In the other Nuristani languages except Prasun, the two symbols correspond to /ṇ/ and /ṛ/, but Strand (2024) indicates a distinction between ň and ṇ in <vʹoň> "irrigation channel". Note that the same distinction must be posited for NKal. (N) <u·eň> "id.", a cognate of the previous form. The sequence <vṛ> in NKal. is written as vř here.

Other symbols

For Ashkun, Morgenstierne (1929, 1934) makes use of <ä>, <ʌ> and <ọ̄>. Examples include Ash. (M) <'kʌsʌrʌ> ~ Ash. (S) <kasarʹa> "nose", Ash. (M/T) <sat'täis> ~ Ash. (S) <sâtʹâːs> "seventeen", and Ash. (T) <dọ̄s> ~ Ash. (S) <dʹost> "hand". It has been suggested that <ä> and <ʌ> reflect allophones of /a/, while <ọ̄> may reflect an allophone of /o/.

For Prasun, Buddruss and Degener (2017) makes use of <ọ̄> and <ö>, as in ǰọ̄ra "pair" and wyödi- "to laugh". Halfmann (2025) represents <ö> as /e/, as in wyödi- = w(y)ed-.

This blog's proposal

The following orthographic conventions will be normalized:

a = <a>, <ä>, <ʌ> (most linguists), <â> (Strand, except NKal. and Tr.)

ā = <ā> (most linguists), <âː> (Strand, Ash. (S)), <â> (Strand, NKal. and Tr. only) 

e = <e>, <ē> (most linguists), <ɛ> (Morgenstierne)

ë = <ə>, <e>, <ë> (most linguists), <a> (Strand, except NKal. and Tr.)

i = <i>, <ī> (most linguists), <a> (Strand, Kt. after š, č, and ǰ)

ī = <iː> (Strand)

o = <o>, <ō> (most linguists), <ọ̄> (Morgenstierne, Buddruss & Degener)

ō = <oː> (Strand)

ö = <ö> (Buddruss & Degener)

u = <u>, <ū>

ū = <uː> (Strand)

ü = <ü>, <ǖ>

ċ = <c>, <ċ>, <ć>

j̈ = <j>, <ź>

v = <v>, <w>

References

• Morgenstierne, Georg. (1929). The Language of the Ashkun Kafirs. Oslo, Norway: Norsk Tidsskrift for Sprogvidenskap 2.
• Morgenstierne, Georg. (1934). Additional Notes on Ashkun. Oslo, Norway: Norsk Tidsskrift for Sprogvidenskap 7.
• Morgenstierne, Georg. (1954). The Waigali Language. Oslo, Norway: Norsk Tidsskrift for Sprogvidenskap 17.
• Strand, Richard F. (1999-2024). Richard Strand's Nuristân Site. https://nuristan.info/
• Buddruss, G.; Degener, A. (2017). Materialien zur Prasun-Sprache des afghanischen Hindukusch. Harvard Oriental Series 80. Harvard University Press. Cambridge, MA. 
• Halfmann, Jakob (2021). Terminological Proposals for the Nuristani languages. Himalayan Linguistics.
• Halfmann, Jakob (2023). Lād "law": a Bactrian loanword in the Nuristani languages. London: Bulletin of the School of Oriental and African Studies.
• Halfmann, Jakob (2024). A Grammatical Description of the Katë Language (Nuristani) (PhD thesis). Cologne: University of Cologne.
• Halfmann, Jakob (2025). The Diversification of Indo-Iranian and the Position of the Nuristani Languages. Wiesbaden: Reichert.