--- In firstname.lastname@example.org
, Piotr Gasiorowski <gpiotr@...> wrote:
> On 2007-10-01 11:01, tgpedersen wrote:
> > We have
> > ON em, ert (est), er, erum, eruþ, eru,
> > OE am, art, is, aru, aru, aru
> This is almost OK for Northumbrian (except that the 2sg. is arþ and
> the pl. is <arun> -- sorry, I didn't spot the typo in my own
> posting) and Mercian, where the forms are <eam, earþ, is, earon>.
> West Saxon, however, has
> eom, eart, is, sind(on)
> > What part of that can't be explained by analogy? Why do we need a
> > root *er-?
> How, then, do you explain WS eart (the diphthong <ea> represents the
> breaking of pre-OE *æ < *a)? It's a lonely a-form in WS: where does
> it come from? Finally, the Anglian final fricative in <(e)arþ> makes
> no sense in Anglian unless it's inherited.
First a digression: Some Germanic languages/dialects (German, Dutch,
Kent, Somerset) have/had initial z- and v- (voiced f) for s- and f-
(but not G- for h-). As I mentioned before, these could be integrated
in a specialized version of Verner for those dialects in which the
exception to the voicing for word-initial stops was left out, and the
rhotacization z -> r was factored out to take effect after the
modified Verner rule(s). Inítial rhotacization z- -> r- would be
blocked by what Glen used to call Paradigmatic Resistance.
Imagine further that PIE had alternative forms of 3pl., one the
well-known one with zero-grade root (for *es- that would be PIE
*senti, PGerm. sind), and one with the reduced-grade a that we find in
some Hittite verbs, eg 3sg. e:pzi, 3pl. appanzi, (for *es- that would
be PIE *asénti -> *azénti -> *ázn.ti -> *árunþ etc.); in the latter
case Paradigmatic Resistance, which worked only initially, wouldn't
For some (Northern?) PGerm. (or Hittite-primitive-like substrate in
Eastern England and Scandinavia?) the present of *es- would then be:
*em *es-tu/*es þu *is *arunþ *arunþ *arunþ
(the 1,2pl of ON look reconstructed from 3pl to me; already Gothic
passive had 3pl-based 'Einheitsplural')
from which I think it's possible to reach by analogy-derivation the
paradigms mentioned. WS would, given the political situation, have a
motivation to shibboleth-purge the offending *ar- stem from the plural
even after it had replaced its 2sg with an *ar-form.
What do you think?