Indus signs discussion

George Thompson gthomgt at COMCAST.NET
Sat Feb 17 16:06:50 EST 2007

Dear Lars Martin,

I agree with your post up to a point.  To the extent that the issue 
rests on the basis of formal statistical criteria, yes, I agree that the 
matter isn't really Indological, or philological, or linguistic.  But 
the issue does not rest on that basis alone.  One issue that could 
legitimately be discussed on this list is the question of what language 
or languages may have been in use in IVC at the relevant period.  Asko, 
both in the recent paper that has been mentioned and in his 1994 book, 
makes the case for Dravidian. Many have  expressed their doubts about 
his use of the rebus principle, which is so loose and slippery as to 
make any interpretation of any sign seem possible.  But given our 
knowledge of South Asia as a linguistic area, we could discuss his claim 
that Dravidian is the best candidate.  We could discuss the relative 
viability of other candidates, such Sanskrit, or Munda languages, etc.

Being only a Vedicist, I myself would approach this issue by 
constructing a model of what a set of Vedic pictographs would look 
like.  In my opinion -- not controversial on this list, I would think -- 
a set of Vedic pictographs would not look very much like the set of 
pictographs that we seem to have in IVC. 

Well, we could ask other Dravidianists what they think of Asko's use of 
Dravidian in his interpretations of IVC pictographs.

I myself am not persuaded that the IVC signs reflect a logographic 
system,or a script, but I think that it is viable that Dravidian might 
have been a language of IVC.

I too would like to think that we could discuss such matters on this 
list without personalia.

Best wishes,

george Thompson

Lars Martin Fosse wrote:

>Like Dean Anderson, I am also sad that there is not going to be a more
>comprehensive discussion of the matter on either of the lists. But the
>discussion would under all circumstances be "asymmetric". Parpola works on
>the assumption that the Indus script is a script, whereas the
>Farmer-Witzel-Sproat theory (FWS) disputes that the Indus sign system is a
>script at all on the basis of formal statistical criteria, as well as other
>criteria. A discussion of the FWS would therefore primarily be a discussion
>of the formal character of script systems with special reference to the
>Indus sign system. 
>I occasionally receive "background noise" from people who claim that FWS has
>not received universal acceptance among people who hold competent opinions
>on script systems. Unfortunately, I have not been able to make such people
>offer a formal, statistically informed rebuttal of FWS. But clearly, that is
>where we have to start. Engaging Parpola before a more comprehensive
>discussion of script systems based on statistical theory has taken place,
>does not seem meaningful to me. The upshot of this is that the problem is
>not really Indological, or even philological or linguistic. As formulated by
>FWS, it mostly belongs to a different sphere.
>If the conclusion were that the Indus sign system really IS a script, then
>the obvious question would be: is there enough information in the material
>to allow us to make a meaningful interpretation? That again, is not
>necessarily a philological or linguistic question. The plethora of
>interpretations (the Indus script is "cracked" at least once a month,
>apparently), shows that the sign system can be construed as almost any
>number of languages ancient or modern, a phenomenon which is due to the
>statistical properties of the material. None of these solutions work a 100%,
>but some of them are good enough to convince at least the credelous. 
>May I politely suggest that the matter is referred to a group of people who
>have comprehensive and competent views on script systems? Such a group could
>have its own list dedicated to this particular problem. Candidates for the
>list - apart from Farmer, Witzel etc. - would be people working with such
>scripts as cuneiform, Mayan script etc., combined with statistically
>competent persons who could contribute to the study of the statistical
>properties of known script systems. I suggest that relevant persons are
>identified and invited to join such a discussion. 
>I personally feel that the "probability mass" for the moment favours FWS. It
>would be the job of a critic to reverse that situation by demolishing the
>FWS ideas on the statistical properties of script systems. Although other
>arguments in my opinion also would seem to favour FWS, the formal properties
>of the sign system are the "pièce de résistance". 
>Best regards,
>Lars Martin Fosse
> Lars Martin Fosse 
>Haugerudvn. 76, Leil. 114, 
>0674 Oslo - Norway 
>Phone: +47 22 32 12 19 Fax:  +47 850 21 250 
>Mobile phone: +47 90 91 91 45 
>E-mail: lmfosse at 

More information about the INDOLOGY mailing list