"Out of India"

[Lars Martin Fosse]

>
>Lars Martin Fosse wrote:
>>It should be added that there are several algorithms for cluster analysis
>>that yield somewhat different results. It would be foolhardy simply to
>>produce a cluster analysis and accept the result as authoritative, several
>>attempts with different sets of criteria and different algorithms would have
>>to be done. 
>
>This adds another item to the list of 'I don't understand'; which algorithms 
>are more 'objective'? Has non-parametric statistics (eg maximum likelihood)
>been used in this area?

Generally speaking, most of the Indological studies that have used
statistics are of very poor quality. Most of the scholars who did this did
not bother to take a statistics course, the notable exception being
Trautmann, who based his study upon the chi-square test. Yardi acquired some
knowledge in analysis of variance, but his statistical modelling is very
bad, and his conclusions are in no way acceptable. I have given a detailed
critique of Yardi in my thesis. 

Apart from that, I definitely think that philologists should talk to
mathematicians (or statisticians) if they want to use numerical methods. But
they should also learn to use some of these techniques themselves. In modern
languages studies, a large number of very good statistical works have been
produced. There is no reason why we shouldn't be able to do this in our own
field.

Best regards,

Lars Martin Fosse