Warning: Creating default object from empty value in /nfs/c03/h07/mnt/56080/domains/varnam.nationalinterest.in/html/wp-content/themes/canvas/functions/admin-hooks.php on line 160

The Indus Script – Decipherments

(Asokan inscription in Brahmi)

Read Part 1

When say “deciphering the Indus script” there are two aspects to it. The first is the structural analysis which looks at the signs which are used the most, the relationship between the signs etc and the second is assigning various sounds to the symbols to attempt a reading.


In 1968, the Russian linguist Yuri  Knorozov who assisted in the Battle of Berlin and later decoded the Mayan script found internal structures in the Indus seals using software analysis. Based on that he read the text as proto-Dravidian. One of the signs in the Indus script is that of a man carrying a stick. This for Knorozov  represented the posture of Yama or Bhairava hence he thought it was one of the predecessors of one of such gods. He also read the script and one such reading is ‘[day of the [god] -guardian honored leader, lightning of the cloud worthy hero’. The criticism of Knorozov is that while his analysis was useful, the reading was pure guess work[1].

But how did Dravidians, who currently live in the four Southern states, end up in the Indus Valley? According to one version, Proto-Dravidian speakers moved into the Indus Valley from Iran some time between 6500 and 3000 B.C.E. These people, who derived Proto-Dravidian from Proto-Elamite-Dravidian, developed the Indus culture over a period of 2000 – 4000 years[1]. When the Indo-Aryans arrived, sometime after the collapse of the Indus Valley, Dravidian was the dominant language.

When Dravidian was replaced by the Indo-Aryan language in Harappa, the Dravidian language did not just disappear. Instead, it left some words — a substratum — in the Indo-Aryan language, which a linguist can identify. Thus it is obvious that the language spoken by the Harappans was indeed Dravidian.

The work of Yuri  Knorozov, and by the Finnish group led by Asko Parpola  was continued by Bryan Wells who used structural analysis to identify syntactic units. This identification helped in finding out words at the subject-object-verb level. This structure was similar to Dravidian. Once that was done certain features of the language —- old Tamil names end with “an” and female names with “al” —-was used to map the symbols[1].

Now comes the tricky part. There is a symbol which looks like a fish but there are many words to describe a fish: one to describe fish generically and some to describe various species. Which one do you pick? Even among the various Dravidian deciphers there is no agreement on the reading. For example, what looks like a crown to one researcher looks like a mountain to another and a tent to the third.

Indo-Aryan version

When Indo-Aryan is mentioned as a possible language, it is shot down based on ‘chronological incongruity’. Remember how the horse evidence before 1500 B.C.E is suspect because horses were bought by the Indo-Aryans when they arrived in the subcontinent after 1500 B.C.E. According to similar logic, Sanskrit cannot be the language of Harappa since Indo-Aryans arrived only after the decline of Harappa. Due to this most books don’t even mention this possibility[4].

Such circular reasoning  to exclude Indo-European as a possibility has not prevented researchers from pursuing this angle. In 1934, G. R Hunter concluded that Brahmi was derived from Indus script. According to Hunter even scripts like Sabaean and Phonecian were derived from the Indus. John E Mitchiner looked at the one particular feature of the Indus script — the case endings — and concluded that it could not be Elamite or Dravidian, but only Indo-European[6].

Taking this further, Subhash Kak did a mathematical analysis of the Indus script and the oldest Indian script – Brahmi. When a table containing the ten most commonly occurring Sanskrit phonemes (from ten thousand words), was compared to the ten most commonly occurring Indus symbols and there was a convincing similarity, even though Brahmi was a millennium after the Indus script.  Surprisingly some of the characters, like the fish, looked similar too[6].

There are three possibilities here: (a) the similarity is random (b) scribes who used Brahmi used Indus signs without knowing how they read and (c) Brahmi was derived consciously from the Indus script. But when the probability of this happening by chance was computed, it was found to be quite low. Also among the ten most common signs of Indus and Brahmi there is striking similarity between four of five signs[6].

Kak also sees a particular Prakrit feature in the Indus script which is not found in Elamite and Dravidian. This particular feature — the gentive case marker — is used to specify ownership which could mean that the seals were used for trading purposes[2][6]. Frequency analysis of the Indus script found that one of the signs is a representation of the numeral 5. The Nagari script, used since 8 CE, also uses the same sign; in Brahmi, this sign means ‘pa’ – the first letter of ‘pancha’. Brahmi inscriptions found in Sohagaura on copper plates and caskets in Batthiprolu shows various compound signs, like in the Indus[7].

(To be continued)


  1. Bryan. Wells, “An introduction to Indus writing /–by Bryan Wells.” (Ann Arbor, Mich. :UMI,, 2001), ScientificCommons.
  2. Edwin Bryant, The Quest for the Origins of Vedic Culture: The Indo-Aryan Migration Debate (Oxford University Press, USA, 2004).
  3. Kamil V. Zvelebil, “Decipherments of the Indus Script,” in The Aryan Debate edited by Thomas R. Trautmann (Oxford University Press, USA), 254 – 271.
  4. Jane Mcintosh, A Peaceful Realm : The Rise And Fall of the Indus Civilization (Basic Books, 2001).
  5. Michel Danino, “A Dravido-Harappan Connection? The issue of Methodology.,” Indus Civilization and Tamil Language (2009): 70 – 81.
  6. Subhash C. Kak, “A FREQUENCY – ANALYSIS – OF – THE – INDUS – SCRIPT,” Cryptologia 12, no. 3 (1988): 129.
  7. Subhash C. Kak, “INDUS – AND – BRAHMI – FURTHER – CONNECTIONS” Cryptologia 14, no. 2 (1990): 169.
  8. Subhash C. Kak, “AN – INDUS-SARASVATI SIGNBOARD,” Cryptologia 20, no. 3 (1996): 275.

(Images via Wikipedia)

, , , ,

5 Responses to The Indus Script – Decipherments

  1. kalyanaraman November 11, 2009 at 2:47 am #

    Corpus of Indus script epigraphs is at http://sites.google.com/site/kalyan97/epigraphica-sarasvati

    A mleccha decoding of smith guild tokens is at


    namaskaram. kalyan

  2. froginthewell November 12, 2009 at 1:20 pm #

    Will you be addressing Michael Witzel’s claim that it was not a script at all? Thanks.

    • jk November 12, 2009 at 10:41 pm #

      froginthewell, that is an even fundamental question – the edifice on which these decipherments stand. Indus people had extensive trade relations with literate Mesopotamia. So why would someone bake a steatite seal – a process which takes few days — and send it abroad if no one can read it?

      But the real question now is: can a statistical analysis prove that a sequence of signs encode a linguistic system. I have added a link in the last post on Dr. Sproat’s lecture which argues against it. Dr. Rao et. al. argue that Indus does encode a linguistic system. This issue has not been resolved yet. There are new papers coming from both camps which should be interesting.

  3. S. M. Sullivan July 25, 2010 at 10:04 pm #

    Here is a link to an article about a new (June 2010) Indus script decipherment entirely unlike previous ones:


  4. Sujay Rao Mandavilli September 30, 2010 at 4:49 am #

    A very detailed structural analysis on the Indus script has been done by Michael Korvink (more than 200 pages)

Leave a Reply

%d bloggers like this: