Harappan Civilization
Total Page:16
File Type:pdf, Size:1020Kb
~{ ,"r. ..... ~_Lu 5 ~ . Reprinted from Harappan Civilization Edited by GREGORY L. POSSEHL OXFORD & lBH PUBLISHING CO. New Delhi Bombay Calcutta _ \ Ii' L IRA V ATHAM MAHADEVAN 29. Terminal Ideograms in the Indus Script THE study of the Indus script has cOme of age with the direct borrowing or common descent. Pictograms and publication of two comprehensive, computerized ideograms, hy their very nature, are bound to have concordances which cover between them the entire resemblances even i.f tbey belong to different and known inscriptional material of the Harappan Civili independent writing systems. Sign sequences in the zation (Parpola et al., 1973, 1979; Ma.hadevan 1977). Indus script are unique, bearing no relation to any of These concordances present the texts and the rele the West Asian scripts. vant background data in a systematic manner, enabl 3) The lndus script is /l0/ related to the later ingscholars without direct acceSS to the original mate Indian scripts, namely, the Brahm. and the Khora rial to undertake analytical studies of the insc riptions sh\hi. The attempts to link fe atures like conjunct and to formulate or verify hypotheses regarding the consonants or medial vowel signs of the later Indian nature of the script and the typology of the underlying scripts with the supposedly simil"r [catllres o[ the language. Indus script have not been successful. Some positive results have already emerged aT 4) The most common supposition that the fre been confirmed by analytical studies based on the quent terminal signs of the Indus script represent concordances. The determination of the direction of grammatical suffixes, especially case endings, has nor writing (from right to left) and the segmentation of been confirmed by the concordances. A careful study the texts into probable " words" and " phrases" of the concordances shows that the most frequent through simple word-division techniques are among terminal signs are too closely related to their antece the mOre secure results obtained so far (Mahadevan dent signs and sign groups with which they occur in 1977, in press a). terminal positions in all contexts, to be variable case It is. h wever, significant that most of the initial endings. The relationship appears to be semantic resu lts fl owing from a careful study of the con cor- rather than grammatical. d(lfl ccs are negative in cha racter: ", 5) The Harappan language is not related to the I) The Indus script is not alphabetic or quasi Indo-European family of languages, as there is no alp habetic, judging from the number of individual evidence for prefixing or inflectional endings in the signs and their functional and distributional charac Indus script. Icrisr-ics . 6) The Harappan language is not related to 2) The Indus script is not closely related to any of Sumerian or other West Asian languages which place the contemporary pictographic scripts of the third the attribute after the substantive. The reverse word and the second millennia II.C., even though the Harap order of the Harappan language is proved by the pans were in contact with the West Asian cultures and occurrence of the numeraJs before the enumerated there could have beeD diffusion of ideas regarding objects in the Indus script. wri ting systems. The presence of a few common 7) A major negative conclusion emerging from pictograms or ideograms (e.g., ma.n, fish, mountain , an analytical study of tbe concordances is that none of river, rain, city, crossroads, house, plough, etc.) may the published c.Iaims of decipherment of the Indus be traced to such diffusion of ideas rather than to script is valid. Most of the attempts (especially those 3 12 1m va/ham Mahadevall which assign alphabetic or quasi-alphabetic values to most productive (Knorozov e/ oJ. , 1965, 1968; the signs) can be easily disproved by a simple Parpola el ai., 1969, 1973, 1979; Mahadevan 1977, in comparison of the frequency and distribution charac press a; Mahadevan and Visvanathan 1973). It is also teristics of the signs with those of the corresponding at this level that some measurable progress has been values in th e as.o;;um cd models. The morc sophisfi achieved in matters li ke determination of the direc cated attempts remain , at bcst, nnt proven. tion of writing, word division and delineation of the Shall one conclude therefore that the Indus script broad syntactical features of the texts. These studies cannot be deciphered at aU and that all further seem to indicate that the typology of the Harappan attempts are bound to be futile and a waste of time? language is non-Indo-European and resembles the There are two good reasons why such a wholly pes Dravidian languages closely. One has however to simistic attitude should not be adopted. Firstly, it is leave the computer behind at this stage when One 3n axiom of cryptology that, given adequate material, proceeds further to look fo r cl ues to find the meaning no code or ci pher can successfully resist decipher of the texts or phonetic values of the signs. ment fo r aU time. This is all the more true of ancient Emil Forrer (1932) pointed out that it was possible undeciphered scripts whose unintelligibility is a to acquire an objective comprehension of the con matter of accident rather than design. The corpus of tents of an inscription io an undeciphered script by Harappao inscriptions is growing steadily as new sites the observance of paraHel phenomena. Parallels can are being discovered and the known sites are taken up occur between a symbolic representation and a text again for more intensive excavation. It is, therefore, associated with it, between the written object and its reasonable to hope that in the near future the number designation, or between the written symbol itself and of inscriptions in the Indus script will be large enough its meaning. ParaUe ls can also be set up by observing to leod itself tn normal cryptanalytical procedures. the similarities in the standard form ul ae employed in The possibility that the Harappan language is totally ancient inscriptions. Forrer was able to show that lost without any surviving descendants in the Sub such comparisons revealed the basic grammatical fea continent is also too remote, considering the vast tures of the writing system even before its linguistic extent and the long duration of the Harappan Civili decipherment. zation. Secondly, it is quite likely that the Indian As 1 mentioned earlier, the method of parallel' is historical tradition, wit h its astonishing continuity particularly apt for a study of the Indus script on and vitality, has managed to preserve at least some account of the continuity of the Indian historical tra facets of the Lndus Civilization, thus providing valu dition. It is probable that even when the Indus script able clues for an understanding of the contents of tbe ceased to be a writing system, some of the more inscriptions in the Indus script. Since the pictorial important ideograms survived and evolved into traili motifs associated with these inscriptions, like the tional symbols of various kinds. Such survivals may depiction of Pasupali, phallic symbolism, veneration consist of iconographic elements and other religious of the pipal tree and the serpent are clearly seen to be symbols, royal insignia, emblems on coins and seals, connected with later .Indian tradition, there is no heraldic signs of the nobility, corporate symbols, good reason to deny the possibility of such intercon totem signs of clans and tribes and the like. L nection between the contents of the Harappan Pictograms and ideograms of contemporary picto inscriptions and the later tradition. graphic scripts may also fum ish valuable clues to the Any serious study of th.e Indus script must begin recognition of the probable objects or meanin gs (but with a formal or structural analysis oUhe texts. Such a not of course the sounds) depicted by similar signs in study will include compilation of a sign li st and a the [Odus script. The comparisons should not be in concordance, tabulation of sign frequencies and consistent with the results obtained from the fonnal statistical-positional analyses to determine the nature textual analysis of the inscriptions. of the script and the language. It is also necessary to In one of my earlier papers (Mahadevan 1972) I carry out a context-analysis of the inscriptions with suggested the possibility that parallels drawn from the reference to their background viz.., sites of occur Harappan substratum might occur in both the Indo rence, stratigraphy, types of inscribed objecL~ and the Aryan and the Dravidian languages. To recapitulate pictorial motifs associated with the inscriptions. It is brieRy, the method of bilingual parallels is based on at this level that the use of the computer has been the fo llowing assumptions: Terminal Ideogram s in Ihe Indus Scr;pl 313 I) T he Har" ppan seals, in accordance with uni t hey are found suffi xed. versal usage, gi ve the names and titles of the owners. Jt is also necessary to emphasize here the limita It is li kely that due to prolonged bilingual ism and tio ns of the me lhod I p ropose . T he tenlative linguistic racial fusion in the Subcontinent, the more important parallels suggested in this paper arc not to be re Hamppan names and ti tl es passed into the later lndo ga rd ed as a decipherment of the I ndus script. The Aryan languages as loan words o r loan tran lations. very d ive rsity o f the ta ter Indian parallels wo uld pre 2) It is possible that the later symbols, il crived cl ude us fro m assigni ng any specific phonetic values to from the Indus ideograms we re continued to be t he ideograms o f the I ndus script.