Response to question on the so-called Indus script

[Guest guest]

(If anyone responds to this post please copy as well to my email

address, saf , since due to time constraints I normally

only post on Indus issues in the Indo-Eurasian Research List, which I

run with the Indologists Michael Witzel, Lars Martin Fosse, and

Benjamin Fleming: <Indo-

Eurasian_research/>)

 

Dear List,

 

After a long time (several months?) without receiving any emails from

the Indology List I just got a bunch in the Digest all dated 10

July -- including one written by my friend and collaborator Michael

Witzel on the Rao et al. paper (now thoroughly trashed by the

computational linguist community).

 

Michael actually wrote that post back on April 24th!

 

Be that as it may, let me respond quickly to one email again

officially dated to July 10th (but actually written on April 28th)

from Philipp Ernest:

 

> Posted by: " Phillip Ernest " yadbhavisyati yadbhavisyati

> Fri Jul 10, 2009 9:15 pm ((PDT))

>

> I'l be interested to know if the following article has been

> responded to anywhere yet:

>

> The Collapse Melts Down: A Reply to Farmer, Sproat, and Witzel; by

> Massimo Vidale; East and West, v. 57 nos. 1-4, December 2007, pages

> 333-366.

 

We never took the time to try to answer it, Phillip. Vidale so badly

misrepresented the arguments in our 2004 paper that it would take

forever to do so -- a wasted effort since there is a lot of major new

evidence coming out on things Indus that continues to make our case

even stronger.

 

Vidale in any event has never personally studied the Indus symbols --

which is quite obvious if you read his invective-filled paper (where

he doesn't significantly discuss even one Indus symbol, as I

recall). Most of his paper consists of masses of archaeological

data, none involving the Indus Valley, that are totally irrelevant to

the issue.

 

One of our best-known arguments (about the fact that if this *were* a

" script " the Indus would be the *only* literate civilization in the

world that never left behind even a *single* artifact behind on

durable materials that anyone would be tempted to call a " text " )

Vidale disposes of in one sentence, simply mentioning a bit obscurely

(not noting that we develop this argument at great length vs. the old

script model) that this problem doesn't bother him.

 

Well, it certainly bothers a lot of other people, and we talked at

some length about this problem at the key Indus conference held in

Kyoto, Japan, 6 weeks ago.

 

At that conference, I presented a general update on the weird

defenses of the old " Indus script " thesis that have been shown up in

the Indian press and occasionally (as in the Rao) in scholarly

journals since we published our paper in 2004. I spent most of my

time talking about Rao et al., since that paper has gotten so much

publicity, while no one has paid any attention at all to Vidale's

polemical piece.

 

Parpola by the way was there in Kyoto, and in discussion of his own

paper, which only peripherally pertained to Indus symbols,

acknowledged that due to our criticisms he no longer thinks that

Indus symbols were part of anything more than a " proto-script. " Well,

that's a little progress, although we already presented strong

evidence, which he ignores, even against the " proto-script " view in

our well-known 2004 paper; see e.g. page 33 in that paper)

 

Here again is a link to that paper, which has been downloaded over

100,000 from my server alone (and this only tells part of the story,

since it is posted on multiple servers) since it was first published

in December 2004:

 

" The Collapse of the Indus-Script Thesis "

http://www.safarmer.com/fsw2.pdf

 

Here is the most recent version of our refutation of Rao et al.,

mentioned in Michael's note written on April 24th. Note also the link

at the top of this version to the new plot that Richard Sproat and I

created a month ago that debunks the claim made by Rao et al. that

" conditional entropy " can distinguish writing from nonlinguistic

symbols in *any* situation -- a fact that is actually pretty obvious

to any computational linguist, none of whom reviewed Rao et al.

before it was published. (Note also that there are no computational

linguists in Rao's group: he is a specialist in robotics, not

linguistics.)

 

I presented that plot in Kyoto -- where Rao et al. didn't have a

single defender (including by the way Parpola and Kenoyer, who were

earlier reported in the press to support the paper, as Michael

pointed out; Kenoyer in Kyoto, however, told us that he tried in fact

to *discourage* publication of the Rao paper):

 

http://www.safarmer.com/Refutation3.pdf

 

Here's a direct link to the plot referred to in the note at the top

of that paper:

http://www.safarmer.com/more.on.Rao.pdf

 

Here is an abstract of the talk I gave in Kyoto (presented on behalf

of me, Michael Witzel, and Sproat) 6 weeks ago:

 

" The Collapse of the Indus-Script Thesis Five Years later: Massive

Nonliterate Urban Civilizations of Ancient Eurasia "

 

http://www.safarmer.com/Kyoto2009.pdf

 

The title rightly suggests that what was discussed in the talk goes

way beyond just the Indus Valley; we also discussed other non-

literate civilizations *surrounding* the Indus -- a topic also

touched on by Daniel Potts in Kyoto. Daniel, from the University of

Sydney, is of course one of the leading specialists on Gulf and

Iranian archaeology, and is a full supporter of the non-script model,

which he thinks is " obvious " . The idea of a large " Script Free Zone "

reaching far East of Elam to India and Central Asia from the early

3rd millennium on until well into the first millennium is an idea

that Michael and I first proposed at another conference in Japan in

2005.

 

Our full paper from the most recent conference will be published

early in 2010 (by the Research institute on Humanity and Nature

[RIHN], in Kyoto -- sponsor of the most important current Indus

archaeological excavations) and will be made available online at that

time via the Indo-Eurasian Research List.

 

I'd like to add that much new evidence was presented in Kyoto of

striking regional variations in the so-called script which confirms

predictions about the symbols that we made as early as 2003, over six

years ago. The fact is that the more things that are dug up in new

sites (like the exciting new finds made at Farmana by V. Shinde et

al., who in Kyoto gave me new photos of new seal finds with

previously unknown regional symbols), the more the traditional script

thesis fades as a serious model. (The new data from Farmana and other

smaller new regional sites are really exciting! I discussed some of

these in our presentation in Kyoto and plan myself to visit Farmana

in December, when the excavation season begins again.)

 

NB that Michael Witzel, Richard Sproat, and I also have a

comprehensive FAQ in the works on the so-called Indus script story,

summarizing all the arguments and counter-arguments on it, a bit

about how we got stuck with the script model, etc. It is meant as a

" one-stop-shop " to kill the thesis off once and for all (it is long

overdue).

 

Finally, on August 6th, I'd like to note that Richard Sproat will be

giving the invited keynote address at a major computational

linguistic conference in Singapore, where he will summarize the wide

range of interesting statistical issues raised by our work in

distinguishing speech-encoding (i.e., " writing " ) from non-speech-

encoding ( " nonlinguistic " ) sign systems, in elucidating why it is so

easy (trivial even) to produce " pseudo-decipherments " of symbols of

any type, including those that aren't linguistic, etc.

 

For other materials on the so-called Indus script issue, which arose

in odd ways from unrelated work of mine on brain-cultural studies,

see the many links at my website:

 

http://www.safarmer.com

 

Regards,

Steve Farmer