SOAS Research Online

A Free Database of the Latest Research by SOAS Academics and PhD Students

[skip to content]

Chowdhury, Tarapada (1930) On the interpretation of some doubtful words in the "Atharvaveda". PhD thesis. SOAS University of London. DOI:

[img] PDF - Submitted Version
Download (50MB)


This is an attempt to find the value of a certain number of abscure words found in the Atharvaveda. The suggested readings and meanings are by no means absolutely certain; they are to be taken for what they are woth. The main criteria (besides context) in determining them have bean the following: (1) text-criticism, (2) linguistic principles holding in the Indo-European or Indro-Aryan languages or in Sanskrit itself, (3) uses, remarks, or paraller passages in later literature, and (4) facts otherwise known. Any or all of these may have been applied in dealing with a certain word, but their application is mainly illustrated as follows. (1) Test corruptions appear in two forms: (i) wrong accentuation end division of words, as in AKSYAM, TAYADARAM, MADHYAMSIR etc., or (ii) confusion or loss of letters and sounds, as in ALPASAYU-, ASVAKSABHA-, KAKUTSALA... GHRTASTAVAS-, DURADABHNA-, PRARTHA-, BHRMALA- etc. Both these combined in the case of ASUNGA- and UTTARADRAU-.(2) Help from other Indo-European languages is sought in finding the meanings of KAMALA- and ASIR- in MADHYAMASIR-, end in trying to get at the etymological connections of RD, KUMBA, and VRA-; that from the principles holding in the Indo-Aryan language in the case of ARATAKI-, KURUTINI-, KHADURA-, PARASVANT-, PAVASTA-, SAMUSPALA-, etc.; while the treatment of ADRUKSNA-, AVARJUSINAM KANAKNAKA-, GALANTA-, CITI-, TAULA-, NYA- and PESTRA- depends on the grammar of Sanskrit itself. The treatment of ALAJI-, OPASA-, KUMBA- etc., KHRGALA-, JABHYA etc. DURSA, and tintin depends on (3), while that of AHARJATA-, ANDIKA-, ayavana-, and muradeva on (4). The following linguistic phenomena have been discussed under the captions within brackets: (1) Alternance of -a and -u (AKSU-), (2) lengthening of the i- or u- vowel (ARDUKSNA-) (3) t d l (ARATAKI-, KURUTINI-, KHADURA). (4) alternance of m and v (TURMISAN (5) alternance of s and k (Kamala- (6) cerebralisation (ARATAKI-), (7) devoicing (adrogha- vita-), (8) deaspiration before ks (ADRUKSNA-) (9) loss of the third consonant in a group (KAKUTSALA-), (10) loss of unexploded t before a conjunct beginningwith s (SAMUSPALA-), and (ii) secondary suffix -ana (PARSANA-). Confusion of p ( ) and y ( ) is illustrated in ALPASAYU-; of v and r as second member of a conjunct in UTTARADRAN and UPAKVASA; of kt ( ) and tt ( ) in ASAMSUKTAGILA-; and of r and ra in BHRMALA-. In suggesting an emendation of the text there is always a tacit assumption that (1) the mss. Either originate from the same source or possibly be compared with an corrected by each other, and that (2) the tradition has been for the most part a written one. The results speak for themselves: but the following facts may be considered. (1) Most of the Mss. Come from Gujarat (12 out of the 20 authorities of S.P. Pandit, who, at p.16 of the introduction to his edition admits giving mostly the Gujarat tradition); (2) S.P. Pandit, Atharvaveda, vol.111 pp.216 and 315, remarks that the tradition of this Veda is derived from mss. and is not strictly oral. As to the time when these graphic confusions may have occurred, Buhler's Palaeographic Charts IV, col.XVI and VI, col.V show that about the sixth or seventh century A.C. the respective letters began to be sufficiently similar to be confused with each other.

Item Type: Theses (PhD)
SOAS Departments & Centres: SOAS Research Theses > Proquest
DOI (Digital Object Identifier):
Date Deposited: 16 Oct 2018 15:05

Altmetric Data


Download activity - last 12 monthsShow export options
Downloads since deposit
6 month trend
6 month trend
Accesses by country - last 12 monthsShow export options
Accesses by referrer - last 12 monthsShow export options

Repository staff only

Edit Item Edit Item