Originally Posted By: Steve
Originally Posted By: goofy
But this hypothesis is not falsifiable in your theory, because your theory doesn't account for regular sound change.


This is a logical fallacy called begging the question; that is, assuming that the question — “Are the sound changes that historical linguists consider inviolable laws regular?” — has already been answered affirmatively. And It most certainly hasn’t — at least in any way that a logician, empirical scientist or philosopher of science should consider logical or scientific. To the contrary, the answer to the question is evidently “No,” even though you’re having trouble accepting that answer.


I think we know that sound change is at least partly regular, because we have a mass of evidence showing regular sound correspondences. The reason we assume that sound change is invariably regular is because it lets us make predictions and falsifiable hypotheses. In other words, it's a useful assumption. If we assume that sound change isn't invariably regular, then we can't predict how the forms of words will differ across languages, and every hypothesis we make is as valid as any other. I have no interest in that sort of theory, it's not useful or interesting.

Quote:

On the other hand, I’ve just disproved the supposed laws specifying that Gc *p and *d never correspond to Gk and L *p and *d.


No, you've simply asserted that some words are related, without offering an explanation of the various historical processes that led to the different forms. Why do some of the Germanic words begin with p and some with f.