different sequence of two signs is repeated in the ten-sign text K-10. One sign is repeated three times, but not in a row, in the ten-sign text M-634, and a different sign is similarly repeated three times but not in a row in the six-sign text 1093.13) Two further signs in addition to those already mentioned occur twice but not in a row in the eleven-sign text M-1169 and the eight-sign text M-357 respectively. The last mentioned text is a “bar-seal,” a category considered particularly crucial for the script thesis by Farmer et al. (p. 33).
Lost Texts When Farmer et al. wonder how a script with so many single-occur- rence signs could possibly have worked over a wide area, they speak as if our present corpus of texts would represent everything there ever was. But thousands of seals come from Mohenjo-daro alone and yet less than one tenth of that single city has been excavated. The number of single- occurrence signs would surely be reduced if the whole city was exca- vated.
Indeed, an integral part of the thesis of Farmer et al. is the claim that the types of inscriptions we know from the excavated material is every- thing there ever was. They categorically reject the much repeated early assumption that longer texts may have been written on “birch bark, palm leaves, parchment, wood, or cotton cloth, any of which would have perished in the course of ages” (Marshall 1931: I, 39).
Alexander’s historians mention cloth as writing material used in the Indus Valley. Cotton has been cultivated there since Chalcolithic times, and is supposed to have been one of the main export items of the Harappans. Yet all the millions of pieces of cotton cloth produced by the Harappans have disappeared, save just a few microscopic fibers preserved in association of scrap pieces of metal. Along with seed finds, however, those fibers do preserve the information that cotton was
actually cultivated and processed. In the same way, the thousands of ______________________________
13) I.e., Marshall 1931: III, pl. 106, no. 93. For the other texts quoted here see Joshi & Parpola 1987 and Shah & Parpola 1991.
〔 38 〕