Oh Jens, it's so hard to get you to accept the facts.
It is painfully obvious (so painful that I could at any moment have
dangerous heart murmurs) that the rule that exists in IE, that "the accent
moves a syllable towards the end of the word if the flexive has an allomorph
that forms a syllable", is not intuitive. It's a complex and _learned_ rule.
If it is so complex, how did it come to be so? Again, more obviousness:
The "complex" rule must have been once _less_ complex. Why did it
become more complex? Certainly not because of any conscious processes
since accent has no inheirant semantic weight on its own.
So we conclude that the accent patterns were once less complex and
automatic unconscious processes worked to complicate the accentuation.
(Say yes if you agree at this point)
Now, you assume, and I stress ASSUME, that the acrostatic (initial accent)
must be the most ancient because it appears to represent this most
uncomplicated system that we are looking for. Strangely, we both agree
that some farout stage of IE had an initial accent. However, there is no
simple way in which to derive the hystero- or proterodynamic accent
patterns from an initial accent system without an intermediary system.
While we would both derive the IE accent system from a single initial accent
pattern of ancient, the acrostatic pattern seen in IE CANNOT be ancient
or representative of this earlier stage because it is TOO regular to have
completely resisted change for aeons! To believe this introduces unnecessary
hypothesis.
But that intermediary system that we need is in fact, a regular penultimate
accent pattern because this is the only accent pattern that can describe the
hystero- and proterodynamic types effectively.
This is a self-evident conclusion, I'm afraid, so, yes, we DO know that the
cause of the accent mobility is an earlier penultimate accent.
- gLeN
_________________________________________________________________
Add photos to your messages with MSN 8. Get 2 months FREE*.
http://join.msn.com/?page=features/featuredemail