Comment by tokenadult | Hacker News Reader

Comment by tokenadult | Hacker News Reader

Parent Full thread

Is there a real controversy over whether the Harappan script is writing

Yes. Again, this is a familiar issue to people who are familiar with linguistics, and one of the criticisms of the Rao et al. publication is that none of the authors are familiar with the methodology used in historical linguistics to solve comparable problems in other parts of the world. (I am able to read ancient Chinese and am intimately familiar with the history of the Chinese script, which is an illuminating example for comparing periods of early ornamentation, proto-writing, and full writing as that script developed with attested archeological finds.) I am also very familiar with the extent of the Indo-European languages and with the fact that many languages spoken in India (e.g., Tamil, a language spoken by many friends of mine) are not Indo-European. The links I already provided give much of the needed background information.

But here are a few more links, for onlookers who are curious about this issue and may be interested in responses to the Rao et al. publication.

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

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

http://horadecubitus.blogspot.com/2009/04/indus-what-did-rao...

http://earningmyturns.blogspot.com/2009/04/falling-for-magic...

(The link immediately above is by a former president of the Society for Computational Linguistics.)

http://www.mitpressjournals.org/doi/pdf/10.1162/coli_a_00011

(The review article linked above from the journal Computational Linguistics discussed the methodological issues involved.)

Why did you leave out Rao's rebuttal to Farmer and friends' unabashed ad-hominmen laced pieces that you have given?

Rao's reply in Computational Linguistics: http://www.cs.washington.edu/homes/rao/IndusCompLing.pdf

Also other articles at: http://www.cs.washington.edu/homes/rao/IndusResponse.html

Edit 1.

An aside:

Being a logic student who uses probabilistic machine learning, Farmer's use of the words refutation, proof etc. make me despair. This web page is an example:

http://www.safarmer.com/indus/simpleproof.html

Rao is much more careful and nuanced in his methods. Using time-tested logic and math. http://www.cs.washington.edu/homes/rao/ScienceIndus.pdf

" As clearly stated in the last sentence of the paper, our results provide evidence which, given the rich syntactic structure in the script (and other evidence as listed below), increases the probability that the script represents language."

It seems to me that at least 50% of this controversy is a culture clash between computational linguists and historical linguists.

It's a little odd to describe the Rao paper as using time-tested logic and math when the idea of conditional entropy as an indicator of whether a given symbol system is linguistic is in no way time-tested. I'm not a professional linguist, and it may well be a valid and interesting technique, but it is not time-tested.

How do you time-test models with a strictly limited and static set of test cases? Any scheme would suffer from over-training. Maybe we have to evaluate approaches using logic and math e.g. conditional entropy to attack problems spaces like this.

The Indus Valley script doesn't look like a mere form of ornamentation to me (knowing something about archaeology and historical linguistics). The problem though is that when we say "writing" we usually mean "encoding otherwise spoken language in durable form" and there is no real evidence that this is what the Indus Valley script does.

The problem is that there are plenty of kinds of markings that don't fall under either of these. Take for example Roman Numerals. You aren't encoding IX or VIIII as the word "nine" or equivalent in any language. Instead You are illustrating the number on a tally stick and the tally stick, rather than the spoken language, acts as the shared point of reference.

So I think one has to be careful about saying that these are the only two options. The script could have a communicative function and even a formally defined one (much like Roman Numerals do) without being writing in the sense of encoding spoken language in durable form.

> when we say "writing" we usually mean "encoding otherwise spoken language in durable form"

Isn't Chinese an obvious counterexample? The Bible say in the beginning was the verb but for Chinese characters the main view is that the speech encoding part has been added later.

But people writing Chinese are still encoding spoken Chinese in an ideographic system. It's not a phonetic system or even close. But it's still a written representation of the spoken language.

Reply on news.ycombinator.com