Wiktionary talk:About Irish

More on mutations

@Angr, Catsidhe, Embryomystic 1. The best place for saying which types of words mutate and which don't would probably be here, but before I write an entire section on it, I just want to make sure I have it right: nouns, verbs, some pronouns, and (most) adjectives (e.g. not mo, do, etc.) mutate, and everything else (e.g. adverbs (such as anois), prepositions, mé, tú, etc.) doesn't. Right?

2. Do adjectives used substantively with the definite article mutate as if they were nouns? Thus, is "the black one" (nom. sg.) an dubh/an dhubh, (gen, sg.) an dhuibh/na duibhe, (nom. pl.) na dubha, etc? And on a semi-side note, would the genitive plural be na (n)dubha or na (n)dubh? Esszet (talk) 22:02, 3 November 2017 (UTC)

@Esszet: Well, mo and do are determiners, not adjectives, so the generalization holds. And I can't think of a pronoun that mutates other than tú/thú, but that's synchronically not really lenition since it doesn't happen in the usual lenition environments. It's probably also important to say which parts of speech can undergo which mutation; for example, adjectives never undergo eclipsis in the modern language (though they did in Old Irish and so there may be exceptions in frozen expressions like place names), and the only forms that undergo t-prothesis are masculine singular nouns. To question 2, AFAIK Irish doesn't nominalize adjectives like that. "The black one" is an ceann dubh. —Aɴɢʀ (talk) 07:36, 4 November 2017 (UTC)

Also, at least the colours are both nouns and adjectives, often with extra senses as nouns: an dubh is "the black, the ink", an buí is "the yellow", an rua is "the red, the red-headed person", etc. --Catsidhe ^{(verba, facta)} 10:15, 4 November 2017 (UTC)

That's true, but those are real nouns that take nominal inflection. For example, the plural of rua (“red-headed person”, noun) is ruanna with a plural ending that adjectives never take. —Aɴɢʀ (talk) 10:20, 4 November 2017 (UTC)

While we're waiting on Embryomystic, I did get several hits for na hÉireannacha, which looks like a substantive adjective to me. Am I right? Esszet (talk) 23:28, 6 November 2017 (UTC)

I'd just say Éireannach is a noun. There are several Irish nouns suffixed with -ach, and not all of them have corresponding adjectives. —Aɴɢʀ (talk) 10:00, 7 November 2017 (UTC)

But the plural for the noun is Éireannaigh, not Éireannacha. Esszet (talk) 20:14, 7 November 2017 (UTC)

Oh, I see what you're getting at now. I get only 9 Google results for "na hÉireannacha" vs. 41,600 for "na hÉireannaigh". I'd say "na hÉireannacha" is an error, probably committed mostly if not only by learners. (Irish is unusual among the world's languages in having more nonnative learners than native speakers.) —Aɴɢʀ (talk) 23:19, 7 November 2017 (UTC)

Alright, one final qustion: can mutation ever be blocked by intervening words? I see from here that it's idir fhir agus mhná; is it thus do bhuachaillí agus chailíní, go anois ndeir agus gcreideann sé…, d'eile theach, etc.? Esszet (talk) 01:31, 10 November 2017 (UTC)

I think idir X agus Y is a bit of a special case. I'm not a native or even a fluent speaker, but I don't think your other examples are grammatical. I think you have to repeat the trigger word in those cases and say do bhuachaillí agus do chailíní and anois go ndeir agus go gcreideann sé, and eile comes after the noun so it's do theach eile. However I did read an article about English/Irish code-switching once that showed that when you insert fuckin’ into a noun phrase in Irish, you insert it before the noun, and lenition skips over it as if it weren't there. The example sentence recorded from a native speaker was Cá bhfuil mo fuckin’ sheaicéad? (“Where’s my fuckin’ jacket?”) with lenition of seaicéad to sheaicéad after mo. (The lack of lenition on fuckin’ itself is unremarkable since unassimilated foreign words beginning with d, t, f, s – basically those where the place of articulation gets lost under lenition – are immune to lenition anyway. I once watched an episode of Ros na Rún and noticed that the vocative of Tadhg was a Thadhg with lenition, while the vocative of Tom was a Tom without it.) —Aɴɢʀ (talk) 10:52, 10 November 2017 (UTC)

Actually I just noticed something in my own sandbox, from a text written by a native speaker in the 1920s: i n-áiteachaibh mar Chloich Cheannfhaolaidh, Ghaoth Dobhair agus pháirt de na Rosaibh (“in places like Cloughaneely, Gweedore and parts of the Rosses”), where Gaoth Dobhair and páirt are both lenited after mar even though they're separated from it. —Aɴɢʀ (talk) 13:05, 10 November 2017 (UTC)

Alright, then off I go to an Irish forum, I guess. I’ll let you know what they say. Esszet (talk) 20:00, 11 November 2017 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ You were right, triggers generally can't be left out, but where they can (and it appears that mar is one such case), mutation isn't blocked by intervening words. I'll incorporate all of this into the article sometime in the next day or two, hopefully. Esszet (talk) 21:43, 12 November 2017 (UTC)

Adjective templates

@Catsidhe, Embryomystic Mahagaja and I would like to migrate {{ga-decl-adj-1}}, {{ga-decl-adj-1a}}, {{ga-decl-adj-2}}, and {{ga-decl-adj-3}} over to {{ga-decl-adj}} because the latter a) shows lenition where it always occurs (e.g. the nominative feminine singular) b) is totally unified (in my opinion, it also looks better). Do you guys agree? Esszet (talk) 13:57, 24 February 2018 (UTC)

I have no objection to this. I particularly like the bit where it's unified. I'll have to study up on the unified template, though, so I can actually create new entries that use it. embryomystic (talk) 20:54, 24 February 2018 (UTC)

@Catsidhe? Esszet (talk) 22:21, 28 February 2018 (UTC)

Sorry. Yes, unification is good. I wanted to do a comparison between the old and new template results, but life got in the way. If there are infelicities, they can be cleaned up later. --Catsidhe ^{(verba, facta)} 00:52, 1 March 2018 (UTC)

@Catsidhe, Embryomystic, Esszet: {{ga-decl-adj-3}} has been completely migrated and deleted. —Mahāgaja (formerly Angr) · talk 22:41, 20 May 2018 (UTC)

I see you did it by hand, I’ll give that request a shot, but don’t be surprised if it isn’t taken up. Esszet (talk) 02:31, 22 May 2018 (UTC)

That isn't necessarily a bad thing. We're in no hurry to make the migration, and doing it by hand allows us to do other cleanup to the entries as well. —Mahāgaja (formerly Angr) · talk 14:09, 22 May 2018 (UTC)

Early Modern Irish / Classical Gaelic

(…) "Irish" is considered to include Early Modern Irish, i.e. all stages of the Irish language from the 13th century to the present. Wiktionary does not use the code ghc for what Ethnologue calls "Hiberno-Scottish Gaelic"; words attested in Irish authors from the early modern era are considered Irish, and those attested in Scottish authors are considered Scottish Gaelic (code: gd).

What’s the reason behind this? As stated, this makes this single language code and header in entries cover over 800 years of language development. And the so called Classical Gaelic or Classical Irish, ie. the highly standardized language used in bardic poetry represents very conservative norm, based on (late?) 12th century language (see eg. David Greene’s The é-future in Modern Irish, Ériu vol. 29, 1978: “it is generally accepted that the linguistic norm described in these tracts [ie. Irish Grammatical Tracts] is based on the language of the twelfth century.”) – and bardic poetry generally adheres to this norm. Early modern prose texts, even as late as 17th century, sometimes use archaic forms from this norm too.

Classical Gaelic allows many forms that are not possible in Modern Irish and is sometimes much closer to Middle Irish: it keeps use of accusative after ní fhuil and for direct object of a verb, it keeps the distinction between the static dative and accusative of motion after prepositions, absolute vs conjunct endings, it has remnants of infixed pronouns, etc. And it has phonology distinct from Modern Irish and Sc. Gaelic (different syllable count of many words, keeping distinctions between dh ⁊ gh and th ⁊ sh – at least for delenition). And it’s not a scarcely attested language either.

So the decision to group all Early Modern Irish terms and phrases together with Modern Irish makes Wiktionary mostly ignore older attested forms no longer used in any dialect of Goidelic (like classical é-future forms, eg. imeórad, imérad, alongside more conservative imairfead, imirfead, for modern imreoidh mé). Wiktionary thus doesn’t contain entries for a well-attested language stage contemporary with Classical Old Norse, Middle English, and earlier than Old Polish (which has its own entries and language code zlw-opl, but no About page).

This makes Wiktionary hard to use, if not just useless, to anybody trying to read classical texts.

There are obvious problems with classical Gaelic lexicography: since this language isn’t a common ancestor of modern Goidelic language, should lemmata attested in Scottish texts be included? Should we have separate Early Modern Scottish Gaelic (and E.M. Manx?) as well? What about 17th and 18th century prose which often has much more in common with modern languages than with the conservative classical standard? What orthography should be used? But most of those problems are relevant to Old and Middle Irish too.

But since, as far as I’m aware, we don’t have any Manx texts from that period, and AFAIK (though I’m no expert, so I’m happy to stand corrected!) not many texts from the earlier part of early modern era (13th–15th c.) that would show Scottish Gaelic features, I’d propose to add Classical Gaelic/Irish as a separate language to Wiktionary – focusing specifically on the bardic standard from both Ireland and Scotland, with forms attested in Dán Díreach and commented upon in grammatical tracts. This language wouldn’t be considered an ancestor of the modern languages (even if some Irish dialects may continue it directly) and modern Irish/Scottish/Manx forms shouldn’t be derived directly from it – but it could be helpful to list classical cognates for comparisons in etymologies.

The spelling used could either be based on the one used in tracts (eg. Gáoidhealg, imérad, slégur, a ttigh, céd, sgél, sgían) or normalized to pre-reform modern Irish spelling of Dinneen’s dictionary (Gaoidhealg, iméarad, sléagar, i dtigh, céad, scéal, scian).

Early modern prose from 16th c. forth would keep being considered modern (Irish/Sc. Gaelic/Manx) rather than classical.

What do you all think? // Silmeth ^@talk 21:29, 1 December 2021 (UTC)

Also not sure whom to ping, so picking @Mahagaja (who added the paragraph cited above) and other people active in this page and talk: @Embryomystic, @Marcas.oduinn, @Catsidhe // Silmeth ^@talk 22:21, 1 December 2021 (UTC)

All attested Early Modern Irish spellings are allowed to have their own entries; they should just be marked as obsolete forms or spellings and use the code ga and the header ==Irish== (unless they are found in texts from Scotland, in which case use gd and the header ==Scottish Gaelic==). We're a dictionary, not a reference grammar or textbook, so phonological and syntactic issues are outside our jurisdiction anyway (though they can be mentioned in Usage notes if necessary). —Mahāgaja · talk 07:40, 2 December 2021 (UTC)

Aside from the consequences for Classical Irish/Gaelic, Modern Irish (and perhaps Scottish Gaelic though I can't say) suffers from the presence of obsolete forms which readers can easily assume are current usage (anecdotally, I have heard this complaint from readers, especially from Irish learners using wiktionary as a supplememtal reference). While the issue for modern Irish can be solved with better marking of usage, I think seperating the two languages would be a lot less messy. AVIASENIAS (talk) 17:39, 2 December 2021 (UTC)

Can you provide examples of obsolete Irish forms listed at Wiktionary that aren't marked as obsolete? I think separating the languages would be hugely messy, as the vast majority of words are spelled the same today as they were in EMI, at least in the normalized editions that most people would be reading EMI texts from. And I think the fact that we list so many archaic and dialectal forms that readers of Irish literature are likely to encounter gives Wiktionary a big advantage over the standard dictionaries, which tend not to list them. —Mahāgaja · talk 18:20, 2 December 2021 (UTC)

@Mahagaja: I have changed alternative to obsolete in adhaigh (“night”) just today (which also listed nonsense *adhaighe and *adhaigheanta forms which I believe have never existed, certainly not in Classical Gaelic – and the most recent example of the word in Corpas RIA I could find was in a much older poem cited in an 18th c. text… all newer occurrences look to be misspellings of aghaidh – most of earlier 17th c. occurrences too). This is a word from 12–13th century that modern Irish speaker will not recognize (even though it is listed in FGB).

I found the merging ghc ⁊ ga discussion in which you wrote:

There's no reason 17th-century Irish can't be simply ga, with words no longer in use labeled "archaic" or "obsolete" (…). Speakers of Modern Irish have no more difficulty reading Geoffrey Keating than speakers of Modern English have reading Shakespeare.

but:

1. you picked 17th century Irish for the sake of your argument – the very end of over 500-years long “early modern” era, while the standard early modern language is based on the 12th century norm – contemporary with Early Middle English, at least 150 years before Chaucer!,
2. even when reading 17th century Keating, I believe speakers of Modern Irish (even stronger native ones) have more problems than English speakers reading Shakespeare, because 1. the familiarity with older forms of the language (and general language competence) among Irish speakers is weaker (English is just a stronger language of the two today), and 2. Keating tends to sometimes use archaic bardic forms (even though his texts are more “modern” than “classical”); reading Theobald Stapleton (contemporary with Keating) might be easier (because he tried to write in more vernacular style) – but don’t quote me on that ;-),
3. Early Modern English (15th–17th c., just ~200 years-long stage!) did not have this kind of highly standardized norm kept for half a millennium in literature,

so this argument doesn’t really work for the entirety of Early Modern Irish.

And as for “the normalized editions that most people would be reading EMI texts from” – that depends. If you read the Irish Gramamatical Tracts edited by Bergin, or the “first” tract edited by Eoin Mac Cárthaigh (The Art of Bardic Poetry: A New Edition of IGT I, 2014) you’ll get the text normalized to spelling closer to the manuscript than to modern pre-reform orthography (ie. é instead of éa in the likes of sgél, imérad; eclipsis of voiceless stops by doubling: a ccruth (but a bhfuil), length mark on vowels in long diphthongs: do-chúaidh, do Gháoidheilg, etc.). On the other hand the texts on léamh.org are normalized to more modern spelling (sgéal, taobh, modern eclipsis).

If we really wanted to limit the amount of “Irishes” on Wiktionary, it’d make more sense to me to merge Old and Middle Irish than Classical Gaelic and Modern Irish… (but I’m not sure why would we do that, as Wiktionary’s fundamental policy is to include “all words in all languages”) // Silmeth ^@talk 20:25, 2 December 2021 (UTC)

You're welcome to start a discussion at WT:RFM about splitting ghc off from ga and gd; the last discussion (the one you linked to) was 8 years ago. I agree my life would be easier if Old Irish and Middle Irish were merged, but since two separate codes exist it would be difficult to persuade people to expand the definition of one into the territory of the other. —Mahāgaja · talk 20:42, 2 December 2021 (UTC)

About the verb riastraim

This word is related to the noun riastradh and the adjective riastartha. I found the entry "riastradh 2" on Foclóir Gaeilge-Béarla (1977) by Niall Ó Dónaill and the entry "riastraim, -adh" on the 2nd edition of Foclóir Gaeḋilge agus Béarla (1927) by Patrick S. Dinneen. Both cited the past automonous form of the verb in the example riastradh uime and eDIL listed the Old Irish verb "ríastraid" but I couldn't find the 2nd person singular imperative form of the verb. What should I do? YukaSylvie (talk) 08:14, 9 October 2023 (UTC)

Dinneen says it's only used impersonally, and Ó Dónaill's entry implies that the only form in existence is the autonomous past indicative, and even then only in the phrase riastradh uime. I'd make an Etymology 2 at riastradh for the autonomous verb form and list it as {{only used in}} the phrase. —Mahāgaja · talk 08:24, 9 October 2023 (UTC)