1. The phonology of Indus is briefly and with some modification presented in the Mahesvarasutras at the beginning of the Panini's grammar Astadhyayi : 1. a  i u, 2. R L,  3. e o,  4. ai au, 5. ha ya va ra, 6. la, 7. 5na ma 3na  Na na, 8.  jha bha, 9. gha Dha dha, 10. ja ba ga Da da, 11. kha pha cha Tha tha ca Ta ta, 12. ka pa, 13.s'a Sa sa, 14. Ha

a. The brevity is confined to noting only the short and pure vowels. It is later elaborated with the remark that the vowels may be distinguished as short, long and protracted as well as pure and nasalised. The Indus vowels were at least distinguished as short and long. 

b. The modification was carried out in three respects:

i. The enumeration of the vowels was begun with the most usual vowel a. In the original numeral list, however, the vowels began with u, which was indicated by Panini himself by highlighting u in the indication of their length and by naming each series of the stops with the vowel u (ku cu Tu tu pu for k-series, and so on). 

ii. The whole list of the vowels was broken into 4 groups and, each time, beginning from the right, they were read from left to right. 

iii. By the time of Panini, the original Indus vowels o and e had become obsolete. They had continued only upto the RV. On the other hand, ra had been reduced to R due to accent on the neighbouring syllable, and its dental form L too had come out. However, as the Sanskrit records show, it remained restricted to a single syllable kL in kLptį.

2. Thus the original Indus order of vowels, read from right to left, was this: Ai  Au   ai  au  e  o  i  u. From here we can proceed towards the Mahesvarasutras.

a. It was broken into 4 groups, beginning from the right: 4. Ai Au  3. ai au  2. e o  1. i u. Then these groups were read, each time from left to right, putting a at the head, and reading R L  for e o. The original diphthongs too were reduced as e o and ai au. The new order was this: 1. a i u 2. R L  3. e o 4. ai au.

i. Because the Indus script ran from right to left, the number of strokes in the numeral signs increased towards the left: 

ii. The diacritical marks too followed the same pattern: 

b. A few instances of the correspondence of the Vedic U with the OP au and Av. ao in OP gaud-,Av. gaoz- and Vedic gUh- shows that there were long diphthongs aU and aI besides Au and Ai in Indus. Thus the graphic 29  of the Indus texts may be transcribed as ra-U and as ga-I rather than as rAu and gAi respectively. 

3. The consonants had been grouped as semivowels, nasals, voiced stops, voiceless stops and sibilants, with  voiced stops, voiceless stops and sibilants, with different shapes and sizes of the numerals increasing towards the left.

4. there were at least two ha-phonemes in Indus, one was voiced , the other voiceless .Thus the voiced ha was put at the end of the voiced consonants, and the voiceless ha was put at the beginning of the voiceless consonants in the right to left order of the Indus syllabary. The voiced ha was as much of the vowels as much different from the voiceless ha as they were put at the two ends of the consonants : the voiced ha was put contiguous with the vowels as if it were also the un- represented a of Indus. Its h element was often less audible than the following a. It might be then equal to the aleph of the semitic languages or the vocalic h in the Persian bachche'. It was a consonant only to the extent that it could not combine with the following vowel. This ha is naturally the last of the line of the semi-vowels :

a. As the Mahesvarasutras run , the next in order were the nasals put in the following ascending order :

b. This group was followed by the voiced stops, first the unaspirates , then the aspirates in the ascending order of the numerals.

c. They were followed by the voiceless stops in the same order of the unaspirates and aspirates, but with the special marks for p and k.

(I). The group originally began from the right with pa and ka, in some way or other the very remarkable of the group, marked with smaller strokes, followed by ta Ta ca tha Tha kha pha cha, running from right to left in the ascending order of the numerals marked with a little bigger strokes:

(II). It was redrafted by putting kha in place of cha to keep kha pha and ka pa at the two ends of the remaining voiceless stops. Because kh ph and k p do not have the corresponding homorganic sibilants, they were removed from the group which has the corresponding sibilants.

d. Even the order of the sibilants was changed from Sa  sa  s'a to s'a  Sa sa.

i. The morphological analysis of some Vedic forms suggest that these sibilants had their corresponding voiced forms in their early history. 

In A-dhvam from Az-dhvam for As+dhvam and in s'as'A-dhi from s'as'Az-dhi for sas'As+dhi, z has been lost without a trace. 

In e-dhi (Av. z-di) from az-dhi for as-dhi, az has been reduced to e. 

In a-sto-Dhvam from a-sto-z-dhvam for a-sto-s-dhvam, -z- has disappeared after cerebralising the following dh. 

When S was preceded by a k in Indus, it was palatalised and, when followed by a voiced dh, it was also voiced to z. 

In tA-Dhi from taz-dhi for takS-dhi, the secondary -z- has disappeared after cerebralising dh and lengthening the preceding vowel. 

Though these phonetic changes relate to the Vedic language, these voiced sibilants (Z'a, za, Za) may be supposed to have existed   in Indus.

From the examples like az-dhi = e-dhi and Saz for (SaS) + dha = So-DhA, Saz + das'a = So-Das'a , as well as taz' (for takS) + dhi = tA-Dhi, we see that az', aZ and az behaved differently.

 

e. The last ha of the Mahesvarasutras is truly a consonant, a fricative sound , also voiced. The Sanskrit phoneticians do not know any voiceless h. On the other hand, the phoneticians of the RV-pratis'akhya (14, 28), say that the pronunciation of h as voiceless is faulty. It is then clear that they knew a voiceless h. In the Indus language, that is to say, there was certainly a voiceless ha, dialectically pronounced s (as in Sanskrit sapta) and h (as in Iranian hapta).

(I) If Hermann Beckh (Neue Wege Zur Ursprache, Page 24) is right, Indus may be supposed to have third ha, voiced and fricative, corresponding to the voiceless fricative ha. This may explain the varying correspondences of the Vedic sapta : Avestan hapta, Vedic hima : Avestan zima.

Copyright: INDUS SCRIPT 2001-02
Site Concept By Sumit Mishra