An Cionnfhaolach wrote:
Domhnaillín Breac wrote:
An Cionnfhaolach wrote:
Question, has the -t- in Oitir been lenited? It seems odd if it hasn't.
Does it? I thought significant borrowing from Old Norse only occurred after lenition ceased to be a productive process in Irish and was fully grammaticalised (ca. 7th cent. CE, IIRC).
Hmm, did you read that somewhere?
I've read it multiple places. My recollection of the 7th century as the endpoint of the original process came from having read Ranko Matasovic's essay "Insular Celtic as a language area" in the book
Celtic languages in contact a couple days ago, but I didn't have an opportunity to consult it before posting. I see that he doesn't give an exact date for the completion of the process but proposes that it took place "within the period of 350-550 CE". Kenneth Jackson, in
Language and history in Early Britain dates lenition in Irish to "the second half of the 5th century", placing it before other changes in that period (such as vowel affection). So it turns out that was being extremely conservative in naming such a late endpoint. (For reference, the beginning of the "Viking Age" is dated to the 790s. The first recorded Viking raids in Ireland occurred in 795 CE.)
An Cionnfhaolach wrote:
What do you mean by productive process and grammaticalised exactly in this context?
As a "productive process", I mean one that is applied across the board to all words in the language, native or borrowed. For instance, I would consider the change of [w] > [v] a productive process in contemporary German because there are no longer any native words with [w] in the language and foreign borrowings with [w] are pronounced with [v] by the majority of native speakers (though with the growth of bilingualism in English, an increasing number use [w], so this rule may cease to be productive in the near future).
From that point of view, lenition is clearly no longer productive in Irish. When Irish borrowed the English word
motor in the 19th century, it did so in the form
mótar. It did not become *
móthar. So clearly sometime between then and when Latin
cathedra was borrowed as
cathaír (mod.
cathaoir), the process stopped operating. (Jackson
et al. examined a great deal of evidence of this sort in order to narrow down that date to the before the end of the 5th century CE.)
However, as a
grammatical process, lenition is still active in contemporary Irish. The
t in
mótar may not ever be lenited, but the
m is in instances where the grammar of Irish requires this, e.g.
feithicil mhótair,
den mhótar,
mo mhótar. Originally, lenition was a
phonetic process which, as you say, applied universally to single consonants in intervocalic position, whether word-internally or otherwise (just as the change to [v] affects all instances of /w/ in contemporary German, regardless of position or source). But eventually this phonetic change became
grammaticalised into an initial mutation used to express one of a number of grammatical relationships between the word in question and the one(s) before it. As such, it applies
analogically to words which entred the language even as much as a millennium-and-a-half since the the original period of lenition ended.
An Cionnfhaolach wrote:
Lenition of internal consonants (phonetic development) being flanked by two vowels occurs right through the Early-Modern, Modern Irish period; e.g. Old Ir. saegal > saoghal
Nope; the orthography only makes it look that way. Old Irish
saegal was actually pronounced */seɣəl/. In Old Irish, the
ponc séimhithe was only used with the letters
f and
s (and even then not consistently). [Source: David Stifter,
Sengoídelc, p. 20.]
An Cionnfhaolach wrote:
dearmad > dearmad; Munster: dearmhad > dearúd (the fact that there is dialectal variability here, it would suggest that the Munster Irish form is a later development since dearmad is also the Old Irish form.
That's late Old Irish at best. The usual Old Irish form is
dermat. [See
the entry in eDIL.] Note the use of
t for a consonant already pronounced [d̪] in Old Irish! The spelling
m is ambiguous here, since after liquids Old Irish
m represented both /m/ (e.g.
format, mod.
formad) and /ṽ/ (e.g.
formór, mod.
formhór). So, in fact, the variation could be very old indeed.
BTW, I hope you enjoyed this digression for its own sake, since ultimately none of it matters in this case. The proximate source of the name is Old Norse
Óttar. So we're not dealing with a single consonant anyway. (Indeed, the conventional Old Irish spelling of the name is
Ottir.) So even
if the name had been borrowed early enough to be lenited (which it wasn't, Ottir Dub only having come to Ireland in 914 CE), it still would've ended up with
t and not
th.