The direction of Harappan writing

IN their attempts to "force fit" Harappan script into Sanskrit moulds, Rajaram and his collaborator ignore many known facts about Harappan inscriptions. One of the most glaring conflicts with the evidence comes in their claim that in most cases the scrip t is to be read from left to right, like Sanskrit.

Much evidence has accumulated over seven decades that this is the reverse of the case. Indeed, one of the few things that all Harappan researchers agree on concerns the usual right-left direction of the script. Writing direction in ancient scripts often varied in different contexts, but evidence of many sorts suggests that Harappan deviated from right-left patterns in less than seven per cent of inscriptions.

Some of this evidence arises from studies of inscriptions on pot sherds. As B.B. Lal showed in the 1960s, examination of overlapping lines on those inscriptions shows that the script was normally inscribed from right to left. Other evidence is apparent t o the untrained eye. Below, we give two examples from images in the Corpus of Indus Seals and Inscriptions compiled by Asko Parpola and his collaborators. The evidence in both cases has been known since the early 1930s.

One kind of evidence involves the spacing of characters. In seal impression M-66a (using Parpola's numbers), shown below, we see one of many cases where an engraver ran out of room when engraving the seal, causing a bunching of letters on the left. In th e seal, no room at all was left for the "jar sign" often found at the end of inscriptions. This forced the engraver to place it below the rest of the inscription, on the far left. Its placement would be inconceivable if the "jar sign" were a wildcard vow el beginning inscriptions, as Rajaram and Jha claim.

Other evidence shows up in Parpola's seal H-103a, shown below. The unusually long inscription in this case runs around three sides of the seal, with the top of the symbols pointing towards the nearest edge. This suggests that the inscription was to be re ad by turning it around in the hand to read its three parts. Only the top side of the inscription is filled with symbols, indicating that this is the first line. The inscription was hence to be read right to left, turning it clockwise to see the rest.

Further evidence comes from studies of initial and final sign sequences, from studies of repeating sign combinations, and other data. All this evidence has been discussed by a long line of researchers stretching from G.C. Gadd in 1931 to Gregory Possehl in 1996. None of this evidence is mentioned in Jha and Rajaram's book.