Talk:Old Irish

From Wikipedia, the free encyclopedia
Jump to: navigation, search

2003 edit[edit]

The page as it exists now is inaccurate; Old Irish is no longer spoken - Old Irish is the common ancestor of Irish and Scots gaelic, but the language in use in Ireland today is most definitly "New Irish". -- Jim Regan 01:05, 8 Aug 2003 (UTC)

Plus: the article is superfluous, as it has the same subject as the Irish language page, which is more complete.Erwin 10:08, 8 Aug 2003 (UTC)

Drastically different[edit]

"Old Irish possesses much more inflection than its descendants and also employs drastically different phonetic and grammatical structures"

Yes, it is more inflectional; it depends on what you mean by 'drastically different' -both languages might still recognise each other if they met on the street, even if the modern one is 'simpler'.

However, I cannot agree with the idea of "drastically different phonetic structures". I bet natives speakers and those who can pronounce Irish, would be within an hour able to speak with the same blas as OI. In that area, it has (for older speakers, anyway) seen the least change. Apart from the dental fricatives (there were 4), one can find all other phones (/R'/ included, but that is not to say it is the same /R'/ as 1200 years ago). Given that the phonetic table only has meaning if one places phones in their appropriate boxes, and the modern langauge uses the same sounds, (apart from dental fricatives, and some alveolar r's and gemminated consonants, all of which survived into the 20th century in Donegal irish, altho the dental fricatives were in sandhi only), how can one make this spurious claim? —The preceding unsigned comment was added by (talkcontribs) 16:30, 19 January 2006 (UTC)

Sorry I didn't notice this comment earlier! I agree that "drastically different" is an overstatement and have rephrased it. I do think the modern languages are phonologically much simpler than Old Irish, more in terms of overall phonotactics and syllable structure than in terms of individual phonemes. User:Angr 13:53, 6 July 2006 (UTC)


that's fair enough -I am not knowledgeable yet to comment on phonotactics


The wording: "/N/, /Nʲ/, /L/, /Lʲ/, /R/, /Rʲ/ represent fortis sonorants whose precise articulation is unknown, but which were probably longer, tenser, and generally more strongly articulated" is not very clear. "Longer", yes, that is a clear, universally understood parameter --- and about the only one clearly stated, for example, in Stifter's new book. But "tenser", "stronger" --- these are at best ad hoc terms used with different meanings by different authors for different languages. If you want to go by modern Irish as a model, then you should use precise articulatory phonetic descriptions thereof (from the literature), otherwise, it might be better not to even include such terms unless you want to say: "by stronger, I actually mean that the unlenited forms were unmarked, and that the lenited ones were approximants", or, "by tenser, I mean that laryngeal constriction accompanied the articulation", or whatever.Jakob37 13:40, 14 April 2007 (UTC)

Well, if we knew what the phonetic difference between the fortis and lenis consonants really was in Old Irish, I would have written that instead. Since we don't, I stuck to vague terms. Would it be better to write simply "...represent fortis sonorants whose precise articulation is unknown, but which were probably longer than their lenis counterparts, and probably differed from them somehow in articulation as well"? —Angr 13:54, 14 April 2007 (UTC)
One method used in historical linguistics is extrapolation: in this case, if we can fairly accurately define the modern Irish reflexes of these sounds, and if we know something about where they came from (Common Celtic and/or Indo-European), then we should be able to make a reasonable estimate of their nature in Old Irish (I haven't checked my references in the office, so this is just a general statement).Jakob37 04:48, 15 April 2007 (UTC)
Yes, but doing so here would be original research. We have to go by what's already been published, and AFAIK what's already been published is pretty vague on the specifics. —Angr 06:34, 15 April 2007 (UTC)

How sure are we that nasalized V actually existed?[edit]

I know this page isn't edited much but I wanted to ask. There are some phonemes that are described as being uncertain, but ṽ isn't one of them. How do we know that the supposed phonemic /ṽ/ wasn't just /v/, spelled differently simply because of its etymology? Irish has a long tradition of doing things like that, and in fact they still spell mh separately today even though the pronunciation has been merged with bh for 1000 years. Is there an Old Irish pronunciation guide somewhere, written when the language was still alive, that says that it's /ṽ/? Are there misspellings? Soap 13:04, 8 January 2011 (UTC)

There's good evidence it was something nasalized at any rate, because many dialects of Irish and Scottish Gaelic have nasalized vowels in position next to mh but not next to bh. It may have been a nasalized approximant like [w̃] or [ʋ̃], which is phonetically much easier to produce than a nasalized fricative, but the comparative evidence that it was somehow nasalized, and distinct from bh, is quite robust. —Angr (talk) 15:31, 8 January 2011 (UTC)
Addendum: while it's true Modern Irish uses relatively etymological spellings, in Old Irish, the spelling hadn't been around long enough that the pronunciation could have deviated much from it. Old Irish speakers would have had no reason to write dubae 'gloom' with b but dumae 'mound' with m if they hadn't pronounced them differently. —Angr (talk) 15:42, 8 January 2011 (UTC)
Thank you, that's a very helpful answer. Soap 12:00, 9 January 2011 (UTC)
Isn't [m̞] a simpler, more iconic and (handily) more vague method to represent relaxed [m]? I've also seen it represented as bilabial, by the way, instead of labiodental, in transcription: [β̃]. But a non-committal notation like [m̞] allows us to avoid going into the phonetic details: labial, labiodental or even labiovelar, fricative or approximant – we can't really determine the historical pronunciation with precision anyway. Just an idea.
(Of course, this trick could be used to mark lenited sounds in general, i. e., fricatives or approximants as far as they have a historical relationship with less relaxed consonants. It would eliminate the need for the "Greek character solution".) --Florian Blaschke (talk) 19:05, 11 April 2012 (UTC)

Posttonic internal short vowels[edit]

According to the analysis of Old Irish phonology generally taught nowadays (which is, for example, mentioned here – in German, page 77 – and, for those who can't read German, in Stifter's textbook which the Kommentar refers to, Sengoidelc), internal short vowels following the place of the stress (the interpretation of pretonic vowels is less clear) only distinguished two values, a rounded value, usually interpreted as /u/, and an unrounded value, usually represented as schwa. Only in absolute final position, all five vowel qualities were distinguished in short vowels. Internally, the written short vowels i, e and a should absolutely not be taken as full vowels, but only as ways to represent the neutral vowel orthographically in the absence of more appropriate or specific vowel symbols. Note that from Middle Irish on, unstressed short vowels had collapsed (been neutralised) into a single vowel quality – schwa – in all positions, even in absolute final position, and the full short vowels seen in the written representation are all merely graphic. The practice to write i between slender consonants, a between broad consonants, ai between a broad and a slender consonant and e (in Modern Irish, ea) between a slender and a broad consonant, and in Modern Irish, a at the end of a word after a broad and e after a slender consonant should absolutely not be taken on face value, namely as representing true full vowels /i/, /e/ or /a/ or as diphthongs! Students of Old Irish are discouraged from this interpretation from the beginning on; it is considered a fundamental error. The practice of vowel writing in Irish is likely derived from the phonetic phenomenon that makes a central vowel between slender consonants sound /i/-coloured, between broad consonants /a/-coloured, etc., and that has phonemically irrelevant, automatic glides appear whenever a front vowel and a broad consonant or a back vowel and a slender consonant occur in sequence. In fact, the orthography (the Modern Irish spelling even slightly more so than the Old Irish system) approximates the phonetic vocalic and diphthongal qualities quite closely and nicely and is therefore easy to understand (and helpful for one's reading and pronunciation once this fact is understood). The misconception that these phonetic vowels have phonological relevance, too, however, should absolutely not be propagated in Wikipedia. Currently, the article is not at all clear on this point. The transcriptions also need adaptation. While it is OK to write [ɪ] and [ɐ] in phonetic notation, correct phonological notation may only recognise these as automatic allophones of /ə/. --Florian Blaschke (talk) 19:50, 11 April 2012 (UTC)

aspirated unvoiced stops phonetic value[edit]

Presumably /p/ /t/ /k/ in the table are phonetic [ph] etc

I think it might be worth mentioning that very briefly, for several reasons. What do you think?

CecilWard (talk) 08:50, 11 June 2014 (UTC)

I don't feel strongly about it one way or the other. I can't think of a reason not to mention it, but neither does it seem so important that it requires mentioning. On a different topic, do you have a source for your assertion that "In some predominant Scottish dialects these two changes [i.e. eclipsis of /p t k b d g/ to /b d g m n ŋ/] did not take place across word boundaries"? It was my understanding that morphologically conditioned eclipsis was lost in Scottish Gaelic (except for a few remnants like bheil < bhfuil), not that it never occurred there. Angr (talk) 18:42, 11 June 2014 (UTC)

Tony, Ill wirte something substantial on htese points and oost it shirtly when imget my ducks inna row, have been realky ill in recent years somthings are a bit skowCecilWard (talk) 23:46, 13 June 2014 (UTC)

Lenition and orthography[edit]

The superdot is NOT used on just f and s - it's used in manuscripts for any lenition. Ch/ph/th are modern developments. It is also used frequently on b, d, and g, depending on the manuscript. It's often left off as well. As to the nasals - it is also used irregularly, but when it is used, it indicates lenition. I have no sources at the moment (other than the standard thurneysen's grammar) , but this is what I was taught as a graduate student in Old Irish. I will try to look it up this week.

"Notable Characteristics" Is possibly problematic.[edit]

This section here strikes me as wrong. I know a tiny bit of Hebrew, Ugaritic, Phoenician, etc.

A system of conjugated prepositions that is unusual in Indo-European languages (although found in many Semitic languages, e.g. Arabic), e.g. dím "from me", dít "from you", de "from him", di "from her", diib "from them" (basic preposition di "from"). There is a great deal of allomorphy here as well.

This is wrong I think. West Semitic, (Hebrew, Aramaic, Arabic, etc) uses the word Min to mean 'from.' In Hebrew, m'efo means 'from where.'

[1] [2]

HeirOfSumer (talk) 17:24, 19 June 2015 (UTC)

  1. ^
  2. ^

Initial Consonant Mutation[edit]

The article claims: "Initial consonant mutation must have been present in at least late Common Celtic (Proto-Celtic) because this distinguishing feature has survived with grammatical significance in both modern Welsh and Breton, and the extinct Cornish language also featured. Because the languages belong to the Brittonic branch of the Celtic language group (so-called "P-Celtic"), initial mutation must predate the split in the development paths of the Brittonic and Goidelic languages. No mutations are, however, attested in Gaulish material so a parallel evolution of the phenomenon in the neo-Celtic languages is also possible. Much of the complex allomorphy has been lost, but the rich sound system has been maintained, with little change, in the modern languages."

Considering the different sound changes that brought about mutations in the two branches, isn't this a bit misleading? I find it dubious that t>th tt>t could be in parallel with t>d tt>th, as in Old Irish vs Welsh.

Source on Welsh:

On the other hand, the lenition of /m/ to /ṽ/ is so unusual that it would be an astonishing coincidence for it to have happened separately in Goidelic and Brythonic. More likely, the lenition of /b d ɡ m/ probably happened in Proto-Insular Celtic, while the lenition of /p t k/ didn't happen until Goidelic and Brythonic had separated. —Aɴɢʀ (talk) 17:23, 12 April 2017 (UTC)