A Standard Pashto

D. N. MacKenzie

Bulletin of the School of Oriental and African Studies, University of London, Vol. 22, No. 1/3 (1959), 231-235

1. Among modern Iranian languages, other than Persian, Pashto shares pride of place with Kurdish as regards both area of territory and number of speakers. Both languages, moreover, are prolific in dialects, but there any similarity ends.

The features differentiating one Kurdish dialect from the next are mainly morphological. The differences are also progressive, in the sense that when dialect II differs from dialect I in only one feature, the next further dialect III will differ from I in both this and some other feature, or features. It can be said, in other words, that the extent of the morphological differences between any number of Kurdish dialects is roughly proportional to the distances between them. One obvious effect of this phenomenon is that to this day no standard Kurdish has emerged as a literary vehicle with any wide scope or vogue.

The case with Pashto is quite the reverse. The morphological differences between the most extreme north-eastern and south-western dialects are comparatively few and unimportant. The criteria of dialect differentiation in Pashto are primarily phonological. With the use of an alphabet which disguises these phonological differences the language has, therefore, been a literary vehicle, widely understood, for at least four centuries. This literary language has long been referred to in the West as 'common' or 'standard' Pashto without, seemingly, any real attempt to define it.

The increase in literacy among Pashto speakers has given rein to a natural tendency to use phonetic rather than standard spellings. It is perhaps surprising to find support for this somewhat parochial behaviour among Pashtuns at a time when literate Kurds everywhere are thinking of a utopian 'unity' of their language. Yet this has been the effect of almost every innovation of recent years [1] and may be expected to continue if, and when, the Afghan Academy () proceeds to introduce a Latin alphabet. [2] If only on this account it seems opportune to attempt to define standard Pashto in more concrete phonemic terms than any adaptation of the Arabo-Persian script permits.

Standard Pashto Alphabets & Transliteration:

2. Of the 36 consonant signs of the standard alphabet seven, , appear almost exclusively in loan-words of Arabic origin and represent no additional phonemes of Pashto. They are mere 'allographs', marked in the transliteration by a subscript line. Three signs, , to some extent represent 'elegant phonemes'. [3] Of the remaining 26 only those numbered 1-5 require any further description as they alone are realized appreciably differently in different dialects. The details, already well known, may be summarized in the following table:

Dialects
SW (Kandahar)	SE (Quetta)	NW (Central Ghilzai)	NE (Yusufzai)
1. c [ts]	c	(s)	(s)
2. j [dz]	j	(z)	(z)
3. ž [3]	ž	ž	[4]
4.	[5]	[6] [j]	(g)
5.	[5]	[7]	(x)

It will be seen that, moving away from the south-western dialect, there is a steady depletion in the inventory of consonant phonemes, owing to coincidence with existing phonemes (in parentheses).

3. Penzl, [8] observing that 'the Kandahar dialect... is the only dialect which has a phonemic system corresponding to the prevailing orthography', makes the following inference. 'The correlation between the Kandahar phonemic pattern and the graphic pattern of the special Pashto symbols of the Arabic alphabet is so close that we must assume that these symbols were created in the area of the Kandahar dialect. Kandahar appears to be the cradle of the Pashto alphabet.' In short, the derivation of the signs 4, 5 from could only have occurred in the south-west, where they represent respectively.

This is in direct opposition to Morgenstierne's earlier hypothesis [9] that 'when the orthography of Pashto was fixed in the 16th century, the distinction between and x, g seems still to have been preserved even among the north-eastern tribes, who were probably the creators of Pashto literature'.

It would be rash to decide this question on orthographic evidence alone, but there is this to be said in support of the 'north-eastern' hypothesis. An earlier orthographic tradition than that now prevailing once existed. In the earliest known Pashto manuscript, written in A.H. 1061/A.D. (1651) [10] (with subscript dot) is written for j, (i.e. sign 4) for and (with central dot) for . These signs were still used in a MS Diwan of Mirza, [11] dated A.H. 1101/ A.D. (1690), but were abandoned shortly after.

In a MS Diwan of the (?) Yusufzay poet Najib, [12] written in A.H. 1108/ A.D. (1696-7), the copyist, Gul Muhammad Peshawari, was presumably responsible for the change from the old signs used in his model to the new. In one case only, when he noticed the difficulty too late to change the alphabetical order of the Diwan, was he obliged to preserve the sign at the end of the rhyme word. Elsewhere he changed this to , now used for both c and j. This suggests (a) that the older tradition was not exclusively Roshani, and (b) that the new signs still represented distinct phonemes in the north-east at this date. At the time of consciously disguising one distinction (viz. that between c and j, which still exists) the scribes would be unlikely to perpetuate other distinctions (between and other phonemes) if they were meaningless. Only later copyists are more prone to give up these distinctions.

4. While the north-eastern dialects show the most changes in the pattern of consonant phonemes they are more conservative with regard to vowels and semivowels. The distinction between all the vowel phonemes, either expressed or implicit in the standard orthography, is preserved, with the possible exception of i : ī, u : ū. Moreover they all appear, alone or in diphthongs, as morphological relevant final syllables [13]:

There are variations in the realization, e.g. -y > -e and -w > -o in final position following a long vowel, -ay > , but neither phonemic nor morphological system is thereby disturbed.

In the south-western dialects, on the other hand, at least one important change has taken place. While stressed e, o are preserved, in unstressed final position o commonly coincides with u and e with i, except when morphological confusion would result. For example, fem. sg. oblique becomes , but direct is preserved (fem. sg. dir. would coincide formally with indeclinable adjectives in -i); similarly fem. sg. obl. becomes , but dir. is preserved (a fem. sg. direct form in unstressed -i would create a new category of nouns). In the verbal system, however, 2nd sg. present is preserved, to avoid confusion with 3rd sg. . Thus there is modification in detail of certain nominal paradigms only.

5. One question of the phonemic structure of the western, if not of all, dialects remains vexed, namely the delimitation of the phonemes a and .

It seems certain that in unstressed, as opposed to medium- or loud-stressed, position there is no phonemic difference between a and . No two utterances are distinguished by these sounds alone. In the latest invaluable dictionary published by the Afghan Academy [14] an 'archiphoneme' unstressed A [a ~ ] is sometimes marked as having variously a or in different dialects. More often it is described phonetically with the equivalent of , Pashto . For example:

Inevitably both the unstressed endings fern. sg. dir. , masc. sg. dir. and the stressed endings masc. sg. dir. , fern. (old spelling ) are given as , respectively. No ambiguity arises in the dictionary, as the gender is there marked, but it has led to some confusion elsewhere.

For morphological reasons it is more convenient to consider unstressed A, whatever the realization, as an allophone of a. This is the analysis tacitly adopted, though not altogether consistently, in short grammatical sketch of western Pashto.[15] He is thereby enabled to define a number of morphological categories in simple phonemic terms. One small error in interpretation is the adoption of the spelling '(he) is', with the sign designed for the stressed fem. nominal ending , in place of , i.e. unstressed day.

Penzl follows the phonetic verdict implicit in the . Having established two phonemes e (i.e. ) and a, he limits their allophonic range severely. Thus, GPK, ch. ii, § 4.3, 'sometimes e appears in weak-stressed, and a in loud-stressed position, e.g. las "ten", "eleven"'; ibid., ch. ii, § 4.4, 'weak-stressed e varies among speakers of the Kandahar dialect in certain inflectional and derivational morphemes with a centralized allophone of a, e.g... or "tongue"'.

This hyper-distinction leads to others. In no other dialect recorded do more than two diphthongs of the type , occur. Final denotes both fern. sg., pl. etc., always stressed, and 2nd pl., stressed or not. It has various realizations, including the types and , but never both distinctively in one dialect. In view of the tendency in Kandahar to give up vocalic distinctions, except when they are morphologically relevant, Penzl's added contrast, GPK, ch. ii, § 10.3, of ei with ey is surely unreal. By introducing further the concept of A as unstressed a his range of diphthongs is reducible to two, as follows:

6. We have seen that the Kandahar dialect has preserved all the consonant phonemes expressed in the standard alphabet, but that while also preserving the full range of vowel phonemes it has put them to use in novel ways. The other dialects, particularly of the north-east, have abandoned a number of consonant phonemes but have generally confirmed the vowels in their morphological positions. It is an obvious inference that an older stage of Pashto, still current in the seventeenth century if the orthographic evidence is trustworthy, combined a 'south-western' consonant system with a 'north-eastern' vowel phoneme system.

It is this conceptual phonemic system, therefore, which is reflected in the verse of the classical period of Khushal Khan and Rahman Baba. Apart from the evident value of this 'Standard Pashto', in its discreet native dress, as a universal literary medium among Pashtuns, it appears to have another important application. It permits the description of Pashto morphology in more accurate and universal terms than does any single dialect. Moreover, once established, by a comparison of the main north-eastern and south-western dialects, it may well serve as the basis for a simple description of the regular phonetic divergences of other dialects.

A meeting of Pashtun scholars and writers from both Afghanistan and Pakistan, held in Kabul during August 1958, proposed a number of standardizations in the use of the present alphabet. These proposals, reported in full in the periodical Kabul, No. 465 of 23 September 1958, represent in the main a welcome return to the classical standard described above.

1. See G. Morgenstierne, Report on a linguistic mission to Afghanistan, Oslo, 1926, 10;
   W. Lentz, 'Die Paschto-Bewegung', ZDMG, xcv, 1, 1941, 118;
   Begam Jamal, 'Yaw xat', Abaseen (Karachi), June 1957.
2. See W. Lentz, Lateinalphabet für das Paschto, Berlin, 1937.
3. See H. Penzl, A grammar of Pashto: a descriptive study of the dialect of Kandahar [GPK], Washington, D.C., 1955, § 37.
4. See Morgenstierne, Etymological vocabulary of Pashto [EVP], Oslo, 1927, 105;
   Report... Afghanistan, 11.
5. See GPK, § 34.3; EVP, 106, s.v.
6. See GPK, § 35.2; EVP, 49, s.v.
7. See GPK, § 34.3; EVP, 77, s.v.
8. GPK, ch. i, § 4.4.
9. Report on a linguistic mission to north-western India, Oslo, 1932, 17.
10. See Morgenstierne, 'Notes on an old Pashto manuscript, containing the Khair-ul-Bayan of Bayazid Ansari', New Indian Antiquary, II, 1939-40, 567. 'Dorn's mser. of Akhund Darweza', there conceded to be older, is now B.M. Add. 27312 and is in fact an indifferent mid-eighteenth century MS.
11. B.M. Or. 4228.
12. B.M. Or. 4498.
13. See Morgenstierne, 'Archaisms and innovations in Pashto morphology', NTS, xii, 1942, 91.
14. , ed. Sadiqullah Rishtin et at., Kabul, 1330-3/1951-4.
15. D. A. Safeev, '', supplement to P. B. Zudin, Russko-Afganskij slovar, Moscow, 1955.

¯²{{{{²¯