RE : [tied] Re: North of the Somme

> > > > Once we get into gadflies & such, Uralic also has *parma, in
> > > > Finnic, Mari and Khanty at least; and Samic *poaró. Mordvinic
> > > > puromo fits better together with the latter.

> > *parma is represented regularly by Finnic *parma, Mari *paarmo
> > and Khanty _puur@... Mordvinic *u and Samic *oa are both normally
> > from *o.
> >
> a/u 'ablaut' is a feature of the ar-/ur- substrate language (hence
> the name).

This isn't a/u however, it's a/o. Similar alternation BTW appears in loans from Indo-Iranian. It may be an adaptation issue, if the original phoneme was both open and rounded.

> > On the IE side it is German _Bremse_, Swedish _broms_ that seem
> > like possible connections (and Lithuanian _sparva_ ??)
>
> Good point. On the other hand, if you admit that, why would the semantic distance between "worm" and Phthirius pubis be too large?

No, I'm not saying it's "too" large. See next.

> > > > Also plain "fly":
> > > > F. *kärpä- (Livonian käärmi), Mo. karvo, Ma. karme
> > > > with an irregular (non-inherited?) cluster.
> > >
> > > Metathesis k - p?
> >
> > I would rather consider the possibility that the "(gad)fly" words
> > come from a root of the shape #kwarPa- (with #P some labial),
> > specifically "fly" from a de-labialized descendant #karPa, and
> > "gadfly" from a de-velarized #parPa. By the semantics we expect
> > these words to be closer related than the "worm" group.
> >
> > > It would seem we have two suffixes, -k and -m. -k is a NWB
> > > suffix too. -m is part of the Caland set.
> >
> > I'm not sure if plain -m works. Substrate loans in western Uralic
> > commonly include the correspondence of Mordvinic /v/ vs. /m/
> > elsewhere (for example "linden", "fog": F. lehmus, sumu ~ Mo.
> > levos, suv). However here we have /p/ in Central Finnic. Unless
> > the Livonian form with the expected /m/ means that *p is a later
> > (onomatopoetic) variant?
>
> How about my favorite phoneme: /n,W/, the nasal labio-velar?

Well if we want to derive them all from a single form. But that doesn't seem to be necessary. This case rather looks like related substrate languages having related, but distinct, invertebrate terminology.

> > > > Finnic has probably affectiv fronting.
> > >
> > > Affective... hm.
> > > Is that similar to the *vëlki/*volki alternation you mentioned
> > > to me?
>
> > No, this is a reasonably common Finnic-internal process, not
> > unlike the "expressiv palatalization" in Basque.
>
> Similar to the 'pejorative j" in the Scandinavian languages
> (and p-/pt- (< *p-/*pj-) alternations in Greek)?
> http://tech.groups.yahoo.com/group/cybalist/message/31015
>
> Torsten

That too. For example _otus_ "creature", _ötökkä_ "bug"; _tuhma_ "naughty", _tyhmä_ "stupid"; _tormata_ "to rush"_, _törmätä_ "to crash"; _harottaa_ "to be spred out", _häröttää_ "to be mess'd up"... and plenty of phonesthetic coinages too.

John Vertical