Atkinson on phoneme inventories in Science
Michael Cysouw
cysouw at EVA.MPG.DE
Wed Apr 27 00:15:34 UTC 2011
Dear All,
given that we are now diving into a discussion of Atkinson's article, I would like to share with you a first very rough draft of a criticism of his approach.
I think his attempt is brave, and actually I have been looking for years for such a nice global effect in the typological distribution of the world's languages. However, I have never been able to find anything which I would find conclusive, so I was rather shocked to see that it might have been as simple as three WALS chapters on phoneme inventories. Of course, I don't believe it, but well, the proof is in the pudding...
Just to summarize my current thinking (please see the attachment for the full current argumentation): I think the global effect is an artifact of the way the WALS data is used by Atkinson.
I welcome any suggestions for further steps: please send them to me personally. Also let me reiterate: I really welcome the kind of approach as proposed by Atkinson! However, I think we should be very conscious about what the data really mean. And it seems to me that Atkinson overshot this one.
best
michael
[see attached file: cysouwATKINSON.pdf]
—————————————————————————
Michael Cysouw
Research Unit "Quantitative Language Comparison"
Faculty of Languages and Literatures
Ludwig Maximilians University Munich
http://www.lmu.de/qlc/
Postal Address:
Postbox 129
Geschwister Scholl Platz 1
D-80539 München
Visitors' Address:
Schellingstrasse 9, Room 301
D-80799 München
work: +49 89 2180 6123
fax: +49 89 2180 99 6123
home: +49 89 4805 8805
mobile: +49 160 94 777 076
email: cysouw at lmu.de
web: http://web.mac.com/cysouw/
AIM/MSN: cysouw at mac.com
skype: michaelcysouw
jabber: michaelcysouw at gmail.com
—————————————————————————
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cysouwATKINSON.pdf
Type: application/pdf
Size: 1762162 bytes
Desc: not available
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20110427/c37f7e0d/attachment.pdf>
More information about the Lingtyp
mailing list