"Out of India"

[joe at sfbooks.com]

(A copy of this message has also been posted to the following newsgroups:
sci.archaeology, sci.lang)

In article <joe-2811960459210001 at josephb.tezcat.com>, I wrote:

>2.  Population changes in the Indus Valley
>--Vidyanath Rao has consulted the article I cited (which I noted at the
>time I had not read), and has described in a note to me a detailed
>analysis of its contents.  He finds that they do not support the
>conclusion I stated (which I understood to be the authors' conclusion),
>that there had been a significant population shift at Harappa between 800
>BC and Mature Harappan times.  I will post this note (with his permission)
>momentarily.

And here it follows.

Nothing between the dashed line and my .sig is me; it's all Vidyanath Rao
<vidynath at math.ohio-state.edu>, who apparently had his own reasons for not
wishing to post it but did not object to its *being* posted (if I
understand correctly).

Joe Bernstein

---------------------

    I finally got around to browsing the Hemphill, Lukacs and Kennedy
paper in Harappa Excavations 1986--1990. I will summarize the relevant
facts from there.

They give the results of cluster analysis and tree diagrams, as well
as principal component results for craniometric, dental non-metric and
cranial non-metric data. These three are not done on the same samples.

Cranometric:
Samples used are Harappa Cemetry 37A and 37B (Harappan phase),
Harappa Cemetry H open burials and pot/jar burials, Chatal Huyuk 
(Turkey, 5000-64 BCE), Tell al-Judiadah (Palestine?, 5000-64 BCE),
Kish (2900-2800 BCE), Tepe Hissar 2 and 3 (dates given as 3500-3000
and 3000-2000 BCE, but may have to be revised), Naqada (7000-5000 BCE),
Abydos (I dynasty), Badaria (Predynasitc) [last three Egypt?],
Napalese (Modern), Tibeans (Modern), Veddahs (modern), Sedement
(9th Dynasty), Mohenjodaro, and Timagarha (1400-800 BCE).

Cluster analysis of the Prehistoric polpulations from Indus valley
alone give Mohenjodaro as the outlier, with Timagarha and R37A
closely related to each other. The relationship of others differes a
bit when sex is taken into account. but in both cases, are more
closely related to Timagarha-R37A than to Mohenjodaro.

When modern populations are added, they group together, and more closely
related to Mohenjodaro than the rest. When all samples are used,
Cemetry H (Jar burials) are closest to Chatal Huyuk, Timagarha-R37A to
Tepe Hissar, Cemetry H (open burials) to Kish, and Mohenjodaro to
Badaria. The modern populations seperate out from all the prehistoric ones.

Dental Non-metric:
Samples used are Harappa, Chalcolithic Mehrgarh (4500 BCE),
Neolithic Mehrgarh (6000 BCE), Inamgaon (1600-700 BCE), Mahadaha (8000 BCE),
Timagarha and Sarai Khola (200-100 BCE).

Timagarha and Sarai Khola are closer to each other than others, 
Chalcolithic Mehrgarh and Harappa to each other, Neolithic Mehrgarh and
Inamgaon to each other and next to Ganga Valley.

Cranial Non-metric:
Samples used are Harappa, Egypt (4000 BCE), Ancient Palestine (700 BCE),
Modern Palestine, Modern Punjabi, Modern Burmese, Modern Bedouin,
Mahadaha, Lidar (2300-2000 BCE), Kamid el-Loz (500 BCE), and Sarai Khola.

Harappa is closest to Modern Palestine and then to Ancient Palestine;
this group is closest to the pair Kamid el-Loz and Lidar. The others
form the second major break-out, with further clustering as follows:
Mahadaha and Modern Punjabis pair and are closest to Modern Burmese;
this group's nearest neighbor is Sarai Khola; the next one out is
Egypt and all of the second group are closer to Modern Bedouins than
to those of the first group.

-----

I fail to see how anyone can conclude that there is a demagrphic break
in Indus Valley between 1700 BCE and 800 BCE. Timagarha is closer to
R37 than any of the three are to Mohenjodaro in craniometric analysis.
-- 
Joe Bernstein, writer, banker, bookseller joe at sfbooks.com
speaking for myself alone http://www.tezcat.com/~josephb/
But...co-proponent for soc.history.ancient, now back under
discussion in news.groups!