Editing Indus Valley Civilisation (section)

=== Possible writing system ===
{{Main|Indus script}}
[[File:The 'Ten Indus Scripts' discovered near the northern gateway of the Dholavira citadel.jpg|thumb |upright=1.35|Ten [[Indus script|Indus characters]] from the northern gate of [[Dholavira]], dubbed the [[Dholavira#Sign board|Dholavira signboard]]]]

Between 400 and as many as 600&nbsp;distinct Indus symbols<ref>{{cite book |last=Wells |first=B. |title=An Introduction to Indus Writing |series=Early Sites Research Society (West) Monograph Series |volume=2 |location=Independence, MO |year=1999}}</ref> have been found on [[stamp seal]]s, small tablets, ceramic pots and more than a dozen other materials, including a "signboard" that apparently once hung over the gate of the inner citadel of the Indus city of Dholavira. Typical [[Indus inscriptions]] are around five characters in length,<ref>{{cite book |last1=Mahadevan |first1=Iravatham |author-link=Iravatham Mahadevan |title=The Indus Script: Text, Concordance And Tables |date=1977 |location=New Delhi |publisher=Archaeological Survey of India |url=https://archive.org/details/masi77indusscripttextsconcordancestablesiravathammahadevanalt_443_h |page=9}}</ref> most of which (aside from the Dholavira "signboard") are tiny; the longest on any single object (inscribed on a [[Indian copper plate inscriptions|copper plate]]<ref>{{cite journal |last1=Shinde |first1=Vasant |last2=Willis |first2=Rick J. |year=2014 |title=A New Type of Inscribed Copper Plate from Indus Valley (Harappan) Civilisation |url=https://ancient-asia-journal.com/upload/1/volume/Vol.%205%20(2014)/Paper/63-1-725-1-10-20141008.pdf |journal=Ancient Asia |volume=5 |doi=10.5334/aa.12317 |doi-access=free |access-date=27 January 2024 |archive-date=27 January 2024 |archive-url=https://web.archive.org/web/20240127191053/https://ancient-asia-journal.com/upload/1/volume/Vol.%205%20(2014)/Paper/63-1-725-1-10-20141008.pdf |url-status=live }}</ref>) has a length of 34&nbsp;symbols.

While the Indus Valley Civilisation is generally characterised as a literate society on the evidence of these inscriptions, this description has been challenged by Farmer, Sproat, and Witzel (2004)<ref>{{cite journal |author1=Farmer, Steve |author2=Sproat, Richard |author3=Witzel, Michael |url=http://www.safarmer.com/fsw2.pdf |archive-url=https://web.archive.org/web/20050207073634/http://www.safarmer.com/fsw2.pdf |archive-date=7 February 2005 |url-status=live|title=The Collapse of the Indus-Script Thesis: The Myth of a Literate Harappan Civilization |date=2004 |journal=Electronic Journal of Vedic Studies |pages=19–57 |issn=1084-7561}}</ref> who argue that the Indus system did not encode language, but was instead similar to a variety of non-linguistic sign systems used extensively in the Near East and other societies, to symbolise families, clans, gods, and religious concepts. Others have claimed on occasion that the symbols were exclusively used for economic transactions, but this claim leaves unexplained the appearance of Indus symbols on many ritual objects, many of which were mass-produced in [[Moulding (process)|moulds]]. No parallels to these mass-produced inscriptions are known in any other early ancient civilisations.<ref>These and other issues are addressed in {{harvp|Parpola|2005}}</ref>

In a 2009 study by P.N. Rao et al. published in [[Science (journal)|''Science'']], computer scientists, comparing the pattern of symbols to various linguistic scripts and non-linguistic systems, including DNA and a computer programming language, found that the Indus script's pattern is closer to that of spoken words, supporting the hypothesis that it codes for an as-yet-unknown language.<ref>{{cite journal |display-authors=4 |first1=Rajesh P.N. |last1=Rao |first2=Nisha |last2=Yadav |first3=Mayank N. |last3=Vahia |first4=Hrishikesh |last4=Joglekar |first5=R. |last5=Adhikari |first6=Iravatham |last6=Mahadevan|date=May 2009 |title=Entropic Evidence for Linguistic Structure in the Indus Script |journal=Science |volume=324 |issue=5931 |page=1165 |doi=10.1126/science.1170391 |pmid=19389998 |bibcode=2009Sci...324.1165R |s2cid=15565405|doi-access=free }}</ref><ref>{{cite news |title=Indus Script Encodes Language, Reveals New Study of Ancient Symbols |agency=Newswise |url=http://newswise.com/articles/view/551380/ |access-date=5 June 2009 |archive-date=11 August 2009 |archive-url=https://web.archive.org/web/20090811035516/http://newswise.com/articles/view/551380/ |url-status=live }}</ref>

Farmer, Sproat, and Witzel have disputed this finding, pointing out that Rao et al. did not actually compare the Indus signs with "real-world non-linguistic systems" but rather with "two wholly artificial systems invented by the authors, one consisting of 200,000&nbsp;randomly ordered signs and another of 200,000&nbsp;fully ordered signs, that they spuriously claim represent the structures of all real-world non-linguistic sign systems".<ref>[http://www.safarmer.com/Refutation3.pdf A Refutation of the Claimed Refutation of the Non-linguistic Nature of Indus Symbols: Invented Data Sets in the Statistical Paper of Rao et al. (Science, 2009)] {{Webarchive|url=https://web.archive.org/web/20190717170734/http://www.safarmer.com/Refutation3.pdf |date=17 July 2019 }} Retrieved on 19 September 2009.{{full citation needed|date=May 2019}}</ref> Farmer et al. have also demonstrated that a comparison of a non-linguistic system like [[medieval]] [[Heraldry|heraldic signs]] with [[natural language]]s yields results similar to those that Rao et al. obtained with Indus signs. They conclude that the method used by Rao et al. cannot distinguish linguistic systems from non-linguistic ones.<ref name="RAO">[http://www.safarmer.com/more.on.Rao.pdf 'Conditional Entropy' Cannot Distinguish Linguistic from Non-linguistic Systems] {{Webarchive|url=https://web.archive.org/web/20190717163301/http://www.safarmer.com/more.on.Rao.pdf |date=17 July 2019 }} Retrieved on 19 September 2009.{{full citation needed|date=May 2019}}</ref>

The messages on the seals have proved to be too short to be decoded by a computer. Each seal has a distinctive combination of symbols and there are too few examples of each sequence to provide a sufficient context. The symbols that accompany the images vary from seal to seal, making it impossible to derive a meaning for the symbols from the images. There have, nonetheless, been a number of interpretations offered for the meaning of the seals. These interpretations have been marked by ambiguity and subjectivity.<ref name="RAO" />{{rp|69}}

Photos of many of the thousands of extant inscriptions are published in the ''Corpus of Indus Seals and Inscriptions'' (1987, 1991, 2010), edited by [[Asko Parpola]] and his colleagues. The most recent volume republished photos taken in the 1920s and 1930s of hundreds of lost or stolen inscriptions, along with many discovered in the last few decades; formerly, researchers had to supplement the materials in the ''Corpus'' by study of the tiny photos in the excavation reports of Marshall (1931), MacKay (1938, 1943), Wheeler (1947), or reproductions in more recent scattered sources.