demonstrate that the language-based rebus principle was utilized. Dem- onstrating this successfully will actually amount to a partial decipher- ment. The material presently available will not, in my opinion, allow a full decipherment, or one covering most texts.
Screening and developing ideas rashly published in the first flush of enthusiasm in 1969, I have in 1994 presented coherent interpretations of more than twenty Indus signs. These interpretations based on the hypothesis that the underlying language is Proto-Dravidian are in accordance with the generally accepted theories of script and decipher- ment and make sense within the framework of the Indus Civilization and Indian cultural history. My main concern has been to find different ways to check the interpretations. One basic goal has been to achieve internal control comparable to that applied in solving crossword puzzles. I have targeted especially signs that come together in ligatures (complex signs formed by combining two or more simple signs or sign elements) or signs that together constitute compound words. If both signs of a potential compound can be interpreted, the result is controlled exter- nally by checking whether such a compound is actually attested in the known vocabulary of Dravidian languages. Semantically the results should make sense in their historical context, and at best they might even solve old problems. Personally I am convinced that this approach is correct, because it has been possible to go on expanding these interpretations systematically. I trust that the end of the road has not yet been reached, although the available material sets severe restrictions.
Without caring to demonstrate in detail what is wrong with these specific interpretations,21) Farmer and his colleagues dismiss them off-
21) I expect detailed criticism which points out specific faults in theory or in factual data. In the present case the rules are very few indeed, in accordance with the generally accepted theory, and do not change from case to case but are the same throughout, so I refuse to accept the implication that the general criticism leveled against all attempts (including that of Hrozn∞) applies here too: “by exploiting the many degrees of freedom in the ways that speech maps to scripts, it is possible by inventing enough rules as you go to generate half-convincing pseudo-decipherments of any set of ancient signs into any lan- guage — even when those signs did not encode language in the first place. The absurdity of this
〔 43 〕