April 30, 2009

Indus script encoded language (not)

A paper in Science aims to support the thesis that the Indus script represents a language, rather than a non-linguistic symbolic system as claimed by Steve Farmer.

Farmer, Sproat and Witzel completely reject this paper (pdf) in a quick response. I tend to agree with them. I didn't really see any strong evidence that conditional entropy can distinguish between linguistic and non-linguistic symbol systems in the Science paper, and the brevity of the Indus Script inscriptions argues strongly against the linguistic hypothesis.

(Another technical refutation at the Language Log)

Science doi:10.1126/science.1170391

Entropic Evidence for Linguistic Structure in the Indus Script

Rajesh P. N. Rao et al.

Abstract

The script of the ancient Indus civilization remains undeciphered. The hypothesis that the script encodes language has recently been questioned. Here, we present evidence for the linguistic hypothesis by showing that the script’s conditional entropy is closer to those of natural languages than various types of nonlinguistic systems.

Link

17 comments:

Maju said...

For me it is obviously an script. Why would Egyptians, Sumerians (and Elamites) as well as Chinese have their own scripts then and not Indus Civilization?

The entropy argument only adds to it but in fact, when reading this it was the first time I ever heard of anyone claiming that Indus script is not a script at all. I am flippant.

Thisis said...

I didn't really see any strong evidence that conditional entropy can distinguish between linguistic and non-linguistic symbol systems in the Science paper
Having conditional entropy similar to that of natural language is a good sign. It does show that Indus symbols have some order very similar to that of natural languages. The Rao et al paper also shows how Farmer et al were wrong in concluding the opposite, by just only calculating unigram frequencies! (see the supporting information of Rao et al)

Brevity does not mean anything! Even in the common era(CE), Indians mainly wrote on perishable materials like palm leaves (for example)

Read the discussion here between Richard Sproat and others and see Sproat bluffing!

leoeris said...

Choke.

Dienekes said...

Having conditional entropy similar to that of natural language is a good sign. It does show that Indus symbols have some order very similar to that of natural languages.I said that I did not see evidence that conditional entropy can distinguish between linguistic and non-linguistic systems, and that is true.

"Similar" doesn't mean anything; the point is to distinguish between linguistic vs non-linguistic systems, since this is what you are trying to test.

Discovering that DNA/protein sequences, FORTRAN, and two wholly artificial sets don't have the same conditional entropy as natural language tells us absolutely nothing.

Perhaps, if Rao et al. compared the Indus Script with a wide assortment of historical man-made symbolic systems, they could make a case for a systematic difference in conditional entropy. Until then, I see no real merit in the effort.

AP said...

I can't comment on linguistics, but Dr. Witzel says: "There are many problems here, including the fact that the first attestation
of Old Tamil came nearly two thousand years after the Indus civilization disappeared,
philological evidence that the Indus region was not Dravidian speaking in early historical times
[4], and evidence based on Indus sign orders discussed in our paper that cannot be reconciled
with purely suffixing languages like Old Tamil [2]."

In this he is referring to his own work which stands on very slender grounds. He is married to the concept that Aryans entered from the Northwest. He is an expert linguist, and his linguistic analysis shows that the oldest intrusion into Rg Vedic is Munda not Dravidian. Therefore, he is trying to prove that Indus language was not Dravidian but Munda. He has therefore gone to a ridiculous extent to change not only the language distribution of India but also its geography. I have never seen any text or material that shows Munda presence in the northwest. Its presence has always been in the East. Its related languages are all in the east. To make sense of his linguistic distribution he has rearranged country's geography too by putting Kikata south of the Punjab. Kikata was and always has been south of Magadha in eastern India in all texts that I have seen.

As to when the first Tamil is attested, he is referring to the Sangam period, I think. Now historically we have seen the Tamil Kingdoms referred to in Asokan inscription, therefore, 1900bc-250bc=1650 years.
My analysis of historical material shows Tamil in at least 500bc a little before the Nanda of Magadha. But if go by some references in middle-eastern texts, Tamil words are seen in the Solomon period. Solomon is traditionally dated by biblical text to the period of Sheshonq (Shishak) to 1000bc.

*[Ophir. The region whence Solomon obtained gold. (1 Kings ix, 26-28; x, 11 : 2 Chron. viii, 18), and " Almug trees," and gems, came also from Ophir : these trees have been thought to represent the sandal wood called Vulgu in Tamil (Pterocarpus santalinus), which would come from India. The ancient Sabean traders may have reached the Indus as early as 1000 B.c., and an overland trade with India in the 9th century B.c. appears to be shown by the representation (on the " black obelisk " of Shalmaneser) of apes, an elephant, and a rhinoceros. From Tarshish (Tarsus) also, Solomon may have obtained, by such overland trade, the "ivory, apes, and peacocks" (1 Kings x, 22 : 2 Chron. ix, 21) brought from India (see Peacock): for these in the Hebrew bear names also known in India, as well as in Egypt (Hebrew hab, Tamil and Sanskrit hab, Egyptian eb, " elephant" : Hebrew Koph, Greek Kepos, Latin Cepus, Tamil Kapi, Egyptian Kafi, " ape" : Hebrew Tukki, Tamil Tikai, " peacock ") : the name of the peacock especially is important, for elephants and apes would be known in Abyssinia and Nubia, whereas the bird (which is represented on the frescoes of Knossos in Krete) is of Indian origin"
http://books.google.com/books?id=OJsYAAAAIAAJ&dq

onur said...

Solomon is most probably a fictional character fabricated around the 7th-6th centuries BCE in Judah. In fact, composition of the earliest books of the Hebrew Bible probably doesn't go back beyond the 7th century BCE.

Maju said...

Maybe. I was actually thinking why are not the Sumerian exotic commercial relations, Meluhha specifically, mentioned, as it is normally assumed to be IVC.

Whatever the case I see absolutely no reason to dismiss such an elaborate set of signs as something else than writing. It must be mentioned that, before Champollion, many also had that despective attitude regarding the Egyptian hieroglyphs.

And IVC has little to evy to Egypt or the long ago eroded Sumer, except maybe the interest of the public.

Kaiser Tufail said...

Anyone ever wondered the medium on which the symbols appear? These are seals and, it stands to reason that they carry the names and titles of persons. It is a simplistic approach but is logical and is based on common sense. Indus 'script' is just names & titles. I agree with Sproat Witzel & Farmer.

sujayrao2009 said...

THIS IS A RELATED POST AND WILL THROW LIGHT ON THE SUBJECT

Here is the complete , comprehensive solution to the so-called Aryan problem
Part one is a high level overview. Part two is much more interesting
This is one of the longest research papers published in a peer-reviewed journal since independance.
> http://www.scribd.com/doc/27103044/Sujay-NPAP-Part-One
> http://www.scribd.com/doc/27105677/Sujay-Npap-Part-Two
> Mirror:
> http://www.docstoc.com/docs/25880426/Sujay-NPAP-Part-One
> http://www.docstoc.com/docs/25865304/SUJAY-NPAP-Part-Two
Links to the journal
Part one http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1324506
Part Two http://ssrn.com/abstract=1541822
SUJAY RAO MANDAVILLI

SUJAY RAO MANDAVILLI said...

Even those who beleive that the Indus script was not closely tied to speech agree that the Indus valley was, in many respects, the msot literate civilization in the ancient world with literacy levels far beyond what was acheived in Mesopotamia or Egypt. this can be inferred from the ubiquity of the script, the presence of signboards, large number of inscriptions etc. It was a common man’s civilization and there is no known evidence of Royal Authority. They had many acheivements in Engineering and Urban development which surpassed those of Mesopotamia or Egypt.

It was four times the size or Egypt/Mesopotamia, & had larger cities than either. About the question of whether there was an additional script or not which was used by a small number of people, for adminstrative usage, we know
(a) less than 10% of the IVC has been excavated
(b) There were contacts with West Asia through out the period of the IVC.
(c) It was four times the size or Egypt/Mesopotamia and was advanced in many ways. Even those who beleive that it was not closely tied to speech beleive that cunieform was intentionally not imported. By the same reckoning, it could have been.
(d) There was remarkable uniformity over an extremely large area and is it amazing how this uniformity was achieved in Eighty years. While pre-IVc cultures were highly localized a remarkable level of homogeneity was enforced with breathtaking rapidity in eighty years over an area that was nearly half the size of Modern India.. remember the primitive transportantion systems way back then…
(e) It could have been used by smaller groups of people on perishable or non=perishable materials or both

We would also hope, research on the Gangetic plains – between 1900 Bc and 600 BC becomes mainstream..
We have opinions of large number of mainstream and western scholars that there were a large number
of cities and towns there and there was literacy in the Gangetic plains too. There are many practical difficulties in researching the Gangetic plains and India specific research strategies must be the starting point… Unless there is a collaborative approach between western and Indian scientists, the entire field is doomed. if this is done, we will be greatly adding to human knowledge.

SUJAY RAO MANDAVILLI said...

(a) Logographic scripts which are not tightly bound to speech were suitable for ancient trade based civilizations
(b) Logographic scripts which are not tightly bound to speech were suitable for multi-ethnic civilizations like the Indus
(c) Logographic scripts can permit mass-literacy or mass quasi-literacy
(d) Since 5-10% of the IVC is excavated and there are 4000 examples of the indus script on seals, signboards, this gives an average of one example
per 125 people which is unheard of old world civilizations and far in excess of other civilizations
(e) Cuneiform would have been suitable where a script was imposed by royal authority
(f) Primarily non-linguistic systems can have a liguistic component
(g) The ‘lost manuscript hypothesis’ can be in another script , not necessarily in the same script. It could have been used by small groups of people
(h) An amazing fact is how a remarkable homogeneity spread over a 1.5 million square kilometres in less than 100 years. How did this happen?

Sujay Rao Mandavilli

SUJAY RAO MANDAVILLI said...

Even if a small linguistic component is added - rebus principle or punning (Witzel Kyoto, 2009 or Sproat in his presentations) or acriphony is added, it qualifies for full literacy. I assume some 'sound coding' would have been useful to them atleast on some occasions.. the longest seal is 17 characters non-analomous and 26 characters analomous. I have never said that what Farmer is saying is necessarily fully wrong, but even Parpola has been reading them mostly as logograms with a linguistic component. So how much of what Farmer is saying is new apart from the fact that he popularized the idea? These men have been saying almost the same thing and fighting with each other?Till 2900 BC Egypt and Mesopotamia were considered proto-literate even if their texts are shorter(not non-literate!!!!)- even if there is small difference between the 2 maybe the Indus system was more expressive than Egyptian proto-literate- because conditional entropy, order of signs, combinations probably did play a major role in meaning in the Indus script (Korvink). Terminologies pertaining to literacy cannot be changed unless all scholars agree - and any demands to change terminology must be met with suspicion, naturally. Only a very small portion of the IVC has been excavated, you know, 5% maybe! Even Farmer agrees “Judging from modern examples and research in the linguistic history of South Asia, the Indus Valley was probably intensely multi linguistic throughout its history. This may have provided the Indus emblem system with an advantage over ordinary writing as a means of providing the civilization with social cohesion. The fact that the majority of inscriptions rely on a surprisingly small core of symbols suggests that the meaning of Indus signs could have potentially been known by almost or all (ALL!!) of the population, resulting in a pervasive quasiliteracy far beyond that achieved in Mesopotamia or Egypt.” No other civlization mass produced writing or ("writing"!!).
I can instead cite Farmer and declare it the most literate civilization on erth. And he and I could be saying the same thing. I say such terms must be avoided.

Making fun of ancient people is absolutely disgraceful.
Sujay Rao Mandavilli

sujay rao said...

INDUS SCRIPT WAS TRUE WRITING

Please find my two papers below and circulate amongst the skeptics, particularly!

To state the obvious, the Indus script was a logo-syllabic script and a lost corpus did exist.

http://www.scribd.com/doc/46387240/Sujay-Indus-Script-Final-Version-Final-Final

Published in the ICFAI journal of history and culture, January 2011

http://www.scribd.com/doc/111707419/Sujay-Indus-Reintroducing-Lost-Manuscript-Hypothesis

Published in International journal of philosophy and journal sciences , November 2012

I am also introducing logo-syllabic thesis B in this paper

The paper is very self-explanatory! does anybody still beg to differ?

Sujay Rao Mandavilli

sujay rao said...

please find my fifth research paper to be published in a peer reviewed journal. this deals with literacy in post-harappan india. this is a logical and self-explanatory paper. to state the very obvious, literacy always existed in post-harappan india

http://www.scribd.com/doc/127306265/Sujay-Post-Harappan-Literacy-Final-Final-Final

sujay rao said...

Literacy in pre-Buddhist India (before 600 BC)

Please find my collection of papers on literacy in Pre-Buddhist India

Before mature phase of Indus valley civilization (before 2600 BC)

- There are some potters marks but none qualify as full writing

Indus valley civilization (2600 BC to 1900 BC)

1. The reconfirmation and reinforcement of the Indus script thesis (very logical and self explanatory paper)


http://www.scribd.com/doc/46387240/Sujay-Indus-Script-Final-Version-Final-Final

2. The reintroduction of the lost manuscript hypothesis (the case for this thesis has obviously become much stronger in the recent past)


http://www.scribd.com/doc/111707419/Sujay-Indus-Reintroducing-Lost-Manuscript-Hypothesis

Post-Harappan India (1600 BC to 600 BC)

1. Literacy in post-Harappan india (obviously literacy in post-Harappan India existed in certain pockets & were limited to very small sections of society- alphabetic scripts were brought from West Asia and the Indus script also continued – this a very logical and self-explanatory paper and anyone can cross-verify the conclusions)

http://www.scribd.com/doc/127306265/Sujay-Post-Harappan-Literacy-and-origin-of-Brahmi

Sujay Rao Mandavilli


sujay rao said...

I am publishing my sixth research paper directly online as it is an extension of my previous papers. Kindly read pages 4 to 18 as it contains a detailed discussion of the term ‘Aryan’. This paper explains why the Dravidian, Vedic and Paramunda Indus theories are not tenable.

http://www.scribd.com/doc/136268397/The-demise-of-the-Dravidian-Vedic-and-Paramunda-Indus-myths

Methods to reconstruct the languages of the Harappans were presented in the present and previous papers.

The older papers were written taking the 19th century school of Indology as a base and working backwards. These may appear to be outdated now (at the end of our very long journey). However, the fundamentals are still correct

Part one

http://www.scribd.com/doc/27103044/Sujay-NPAP-Part-One

Part Two very,very important!

http://www.scribd.com/doc/27105677/Sujay-Npap-Part-Two

the first 5 papers were published in peer-reviewed journals — Preceding unsigned comment added by 182.72.239.115 (talk) 17:51, 16 April 2013 (UTC)

Sujay Rao Mandavilli sujayrao2012@gmail.com

sujay rao said...

PLEASE FIND ALL MY SIX PAPERS - THIS IS THE COMPLETE SET

The Demise of the Dravidian, Vedic and Paramunda Indus myths

I am publishing my sixth research paper directly online as it is an extension of my previous papers. Kindly read pages 4 to 18 as it contains a detailed discussion of the term ‘Aryan’. This paper shows why the Dravidian, Vedic and Paramunda Indus theories are not tenable.

http://www.scribd.com/doc/136268397/The-demise-of-the-Dravidian-Vedic-and-Paramunda-Indus-myths

Methods to reconstruct the languages of the Harappans were presented in the present and previous papers. We hope other scholars take up the exercise of reconstructing the languages of the Indus Valley civilization!

The older papers were written taking the assumptions of the 19th century school of Indology as a base and working backwards. These may appear to be outdated now (at the end of our very long journey). However, the fundamentals are still correct.

Part one

http://www.scribd.com/doc/27103044/Sujay-NPAP-Part-One

Part Two very,very important!

http://www.scribd.com/doc/27105677/Sujay-Npap-Part-Two

(These comprise the complete and comprehensive solution to the Aryan problem)

for those who have trouble reading part two in the above link use the link below:

part two
http://www.docstoc.com/docs/25865304/SUJAY-NPAP-Part-Two




Literacy in pre-Buddhist India (before 600 BC)

Literacy in pre-Buddhist India (before 600 BC)

Please find my collection of papers on literacy in Pre-Buddhist India

Before mature phase of Indus valley civilization (before 2600 BC)

- There are some potters marks but none qualify as full writing

Indus valley civilization (2600 BC to 1900 BC)

1. The reconfirmation and reinforcement of the Indus script thesis (very logical and self explanatory paper)



http://www.scribd.com/doc/46387240/Sujay-Indus-Script-Final-Version-Final-Final

2. The reintroduction of the lost manuscript hypothesis (the case for this thesis has obviously become much stronger in the recent past)



http://www.scribd.com/doc/111707419/Sujay-Indus-Reintroducing-Lost-Manuscript-Hypothesis

Post-Harappan India (1600 BC to 600 BC)

1. Literacy in post-Harappan india (obviously literacy in post-Harappan India existed in certain pockets & were limited to very small sections of society- alphabetic scripts were brought from West Asia and the Indus script also continued – this a very logical and self-explanatory paper and anyone can cross-verify the conclusions)

http://www.scribd.com/doc/127306265/Sujay-Post-Harappan-Literacy-and-origin-of-Brahmi

Sujay Rao Mandavilli