? Why "Burden of proof" ?

[Larry Trask]

On Thu, 12 Aug 1999, Patrick C. Ryan wrote:

> But surely taking as a null hypothesis a formulation like: "No
> languages are related", which has been disconfirmed ---
> *repeatedly*, is useless?

No.  It's merely that, in seeking a relation between some *particular*
languages, the null hypothesis is "These *particular* languages are not
related."

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk