Re: [tied] Re: Latin barba in disaccord with Grimm's Law?

>Germanic ia also instrumental in ruining Warnov-Ringe's (CPHL project)
>perfect tree models.

But we already know that tree models fail to capture the reality of what
happened in IE - continued influence from neighbouring families on another,
wave efffects, and convergence as well as divergence.

So tree models may be neat, but they are not "perfect" for our purposes.

Peter