The Indus Script — A real language or not?

An article in NewScientist Thursday alerted me to a recent controversy over the Indus script, a set of symbols associated with the Indus Valley civilization of eastern Pakistan and western India. The Indus valley civilization is dated in the timeframe of 2500 to 1900 BCE, according to writer Ewen Callaway (see “Scholars at odds over mysterious Indus script.”)

The basic controversy is over whether the Indus script really represents a language or is merely a set of religious or political symbols. Advocates on both sides have used computational analysis to support their conclusions.

The viewpoint that Indus script is not a language in part rests on the observation that “most of the inscriptions contain fewer than five characters, few of the characters repeat, and many of the symbols occur very infrequently,” writes Callaway.

This viewpoint is put forward in a 2004 article for the Electronic Journal of Vedic Studies by Steve Farmer and colleagues — see “The Collapse of the Indus-Script Thesis: The Myth of a Literate Harappan Civilization.”

Farmer notes that there were other “nonlinguistic symbol systems in the Near East that served key religious, political, and social functions without encoding speech or serving as formal memory aids.” His paper points to evidence “that the Harappans’ lack of a true script may have been tied to the role played by their symbols in controlling large multilinguistic populations.” (The Indus civilization is sometimes called Harappan, referring to one of the key archaeological sites associated with the culture.)

However, the cocksureness of Farmer’s tone (just consider the title of his article, referring to the “collapse” of the opposite viewpoint and his calling it a “myth”) sets off my “expertitis” meter. The readings are high in this case, so I am interested to find recent research that argues in favor of the Indus-Script thesis that in fact “Indus inscriptions were tightly bound to language,” in Farmer’s words.

Finnish professor of Indology Asko Parpola of the University of Helsinki analyzes and critiques Farmer’s work in “Study of the Indus Script.” It’s worth noting what he has to say, as his work is a key target of Farmer.

Rajesh Rao, a University of Washington computer scientist, has more recently published research based on use of artificial-intelligence pattern-analysis software to study the Indus script.

A writeup in Wired describes Rao’s findings (see “Artificial Intelligence Cracks 4,000-Year-Old Mystery“):

[Rao’s team] fed the program sequences of four spoken languages: ancient Sumerian, Sanskrit and Old Tamil, as well as modern English. Then they gave it samples of four non-spoken communication systems: human DNA, Fortran, bacterial protein sequences and an artificial language.

The program calculated the level of order present in each language. Non-spoken languages were either highly ordered, with symbols and structures following each other in unvarying ways, or utterly chaotic. Spoken languages fell in the middle.

When they seeded the program with fragments of Indus script, it returned with grammatical rules based on patterns of symbol arrangement. These proved to be moderately ordered, just like spoken languages.

Wikipedia offers some useful articles related to this question of the Indus script — see “Indus script,” “Undeciphered writing systems,” and “Decipherment.”

A refreshingly readable discussion of why decipherment is so hard is available from Cecil Adams at The Straight Dope — see his article “How come we can’t decipher the Indus script?” This is the first time I have encountered Cecil Adams, and I am pleased to be able to start following the work of a fellow know-it-all.

AB — 25 April 2009