From: Arnaud Fournet
Message: 60391
Date: 2008-09-27
----- Original Message -----
From: "Andrew Jarrette" <anjarrette@...>
>
> >>> To start with, I think such a thing like -tl- is
> >>> impossible in PIE. This cluster becomes -tr-.
>
> >> Then what of Germanic words like OE (Anglian) <ne:þl>
> >> "needle", <boþl> "dwelling", and Lithuanian words like
> >> <arklas> "plough"? Don't these go back to PIE *-tl-?
>
> > I have needle from OE naedle
>
> The usual citation form is <næ:dl>; attested spellings
> include <nædl>, <naeðl>, <nedl>, <nethl>, and <netl>. In
> the Épinal-Erfurt Glossary (8th century) Latin <pictus acu>
> is glossed <mið naeðlae asiuuid>. <Ne:þl> (or <ne:ðl>)
> would be an Anglian form corresponding to <næðl>.
>
> > and booth seems to be a Danish loan-word.
>
> The word in question isn't <booth>; it's the OE word
> underlying the second element of such place-names as
> <Harbottle>, <Newbottle>, etc. The usual citation form is
> <botl> 'a dwelling, habitation, building'. I'm not aware of
> any <ð> or <þ> variants of this word, but given the
> cognates, including ON <ból>, it can only be from *boþla-.
>
> Brian
>
The citation <boðle> (dat. sg.) occurs in Bede's writings, and
according to A. Campbell, the place-name <Bothel> in Cumberland is
representative of this word. I used this example because of its
clear connection to a PIE root; unfortunately it is not common in
earlier Anglian texts. However there are many other earlier-Anglian
words, such as <a:ðle> d.sg. "disease", <we:ðl> "poor", <mi:ðlum> d.
pl. "horse's bit", also <seþel> "sedes" (beside <setel> "cathedra"),
which do not have a clear PIE source, which also demonstrate this
Anglian sound-sequence (A. Campbell says the change *-þl- > -tl- has
no parallel in Anglian, and the spirant *ð/þ is always retained
before *l in such Anglian words, in early Old English). The forms of
all these words with -dl- (usual after long vowels) or -tl- (usual
after short vowels) are found in West Saxon and Kentish; later
Anglian has -dl- in these words alongside the earlier forms
(according to A. Campbell).
=======
But do you (Arnaud) still say that PIE had no *-tl- sequences in any
words, that they always > *-tr-?
AJ
=======
I have not changed my mind.
All these examples have the peculiarity of having a laryngeal before the
suffix.
s-neH1-t-
buH2-t-
arH3-t
But this may not have any importance.
I also disagree that arklas is traceable to PIE.
this would mean that PIE was neolithic,
something I disagree with.
I suppose the alternation -tel-/-tr- which I consider was once allophonic,
became allomorphic in a second time.
So out of -tel- you can recreate -tel-/-tl-
and out of -ter- you can recreate -ter-/tr-
The only thing that really proves something is that
Semitic *tel is reflected by PIE treyes "three".
there is not a single word in Indo-European that is **tl- "three".
This is a cognate, not a word that can be created about any time in the
history of each Indo-European language.
Arnaud