Wiktionary talk:About Classical Gaelic

Classical Gaelic miscellanea
(moved from Beer Parlour and extended)

I’m preparing to start a stub for WT:About Classical Gaelic now that we’ve the  language code split.

Before I do that I’d like to ask ye about some preliminary stuff and see what you find about my ideas. There’ll be a few different issues here. I wasn’t sure if it fits any general discussion spaces, so I’ve chosen Beer Parlour as the most general one – hope that’s OK! I’ve been told to move it here. ;-)

This post is a general request for comments.

Would love your input (but no pressure! I know it’s a lot and ot all of you might want to chime in o those),, of course others are welcome to comment to.

What falls under Classical Gaelic and what doesn’t
So far we’ve had Middle Irish for anything up to ~1200 and then Irish or Scottish Gaelic for anything later. The  language code is not treated as the ancestor of Irish and Modern Gaelic (as it was a fairly standardized literary language that at the end of the period was quite different from the vernaculars). So the question is which texts should fall under the Classical Gaelic heading and which should not.

My proposal is:
 * 1) treat any text composed between the 13th and 15th centuries (inclusive) in Ireland and Scotland as Classical Gaelic,
 * 2) treat all poetry fulfilling dán díreach requirements from Ireland and Scotland, up to the 18th century, as Classical Gaelic,
 * 3) and for 16th century and later prose, decide depending on diagnostic features of the text in question.

The last will be fairly easy for Scottish texts, a bit more difficult for Irish ones. If a Scottish texts has a consistent use of: it’s (prose) Classical Gaelic and not Scottish Gaelic.
 * 1) plural verbs endings (cuirid for ‘they put’),
 * 2) full preverbs do-, a-, ad- (do-bheir, a-tá, ad-chím, etc.),
 * 3) no reduction of do in past tense and relative clauses (do ghrádhuigh agas do ghlac ‘which have loved and accepted’),
 * 4) no reduction of do before verbal nouns (tareis an fhuar chreidimh do chur ar gcul ‘after putting away the vain faith’),
 * 5) use of eclipsis (a gcriochaibh for ‘in bounds / lands / countries’),
 * 6) use of future tense separate from present,

So Carswell’s Foirm na n-Uirrnuidheadh would be Classical Gaelic, but 1767 Gaelic Bible would be Scottish Gaelic. As Donald Meek put it in Language and Style in the Scottish Gaelic Bible (1767–1807):

(…) Thus, a bardic poem was governed by regularions that defined its language and style to a very minute degree. Prose was less strictly controlled, but it was carefully regulated, with the verb-endings correctly used according to classical norms, and other spects of morphology (…) closely observed. This type of prose was written in Scotland well before Keating’s time, and occurs in the first Gaelic book ever printed, namely John Carswell’s translation of the Book of Common Order, published in 1567. (…)

Carswell’s version [of a Biblical passage] is noticeably different from the form of the verse in the 1767 translation (…).

Although stylistically variable, from the highly ornate prose of John Carswell to the leaner but consciously dignified prose of Geoffrey Keating, Classical Gaelic of Type A made few concessions to the sort of language actually spoken by the ordinary people, certainly in Scotland. Nevertheless, one senses in Carswell a transparency and an occasional lightness of touch (…) which seem almost to anticipate the need for a level of language capable of connecting with the non-classical vernacular language. (…)

This balance, between ‘differentiated register’ and the vernacular, was in due time fully achieved by employing another of Classical Gaelic, which we can call ‘Type B’ for convenience. [we just call this (Early) Modern Scottish Gaelic on Wiktionary] (…)

(…) Kirk’s Bible is in the ‘Type A’ style; the later Old Testament is in ‘Type B’. (…) The most salient differences are:  The pre-verb do, used to mark past tenses in the classical language, is not used in independent position in the Scottish Gaelic Old Testament, whereas it is fully preserved in Kird; thus sheas rather than do sheas (v. 3). The Scottish Gaelic text generally employs analytic forms of the verb, and does not use verbal inflections to indicate the person of the verb; thus chruinnich na Philistich rather than do chruinnigheadar na Philistinigh (v.1). The Scottish Gaelic text also marks nasalisation (eclipsis) in the Scottish manner, rather than in the classical manner (…); thus nam Philisteach rather than na Bhphilistineach (v. 2). There are noticeable differences in vocabulary, and the Scottish translation generally reflects a register more in keeping with Scottish use; thus ghlaodh e re slòigh Israeil rather than dfúagair ar sluaghaibh Israel (v. 8), (…) 

He calls the Gaelic of the later Bible translation “Classical Gaelic Type B” but honestly I don’t see why as it’s pretty much just a high register of modern Scottish Gaelic. Fairly close to the modern vernacular and quite distant from Carswell.

It’s a bit more difficult to make a list of such diagnostic features for Irish, as many of them are kept in one dialect or the other and show up in prose fairly late (even if the language is generally recognizably a modern dialect). I’d suggest a few, though: and perhaps more.
 * 1) points 2–3 from the Scottish list,
 * 2) occasional use of s-preterite forms in the past meaning,
 * 3) consistent use of the -idh ending in 3rd person present absolute verbs (and only restricting -ann to dependent/conjunct forms),
 * 4) other distinctions between synthetic absolute / conjunct endings (cuirmíd vs ní chuiream; cuirid vs ní chuiread),
 * 5) subject/object distinction in 1st sg., 1st pl. and 2nd pl. pronouns (mé vs mhé; sinn vs inn; sibh vs ibh),
 * 6) keeping the preposition re beside le (even if not distinguished in meaning),
 * 7) use of eclipsing ar to express perfect,
 * 8) use of eclipsing go in the meaning ‘with’,

I’ll need to collect more examples of Irish texts from 17th and 18th centuries and compare them to see how easy it is to classify them as classical or not.

Lemmatizing verbs
We need to choose the lemma form for the verb. We have a few options, each with its adventages, disadventages, and some precedence.

Personally I lean towards the 3rd person present indicative, but here’s the full list:


 * 1) 3rd sg. pres. indicative – the same form as we use for Old and Middle Irish,
 * 2) 1st sg. pres. indicative – as in Dinneen,
 * 3) imperative – like for new dictionaries of Modern Irish,
 * 4) verbal noun – like in medieval tracts.

1. is used in DIL (where most of citations actually are from classical texts, it’s actually more of a Early Modern Irish dictionary than an Old Irish one, despite mostly using the Old Irish-like spelling), and in vocab lists to some editions of classical texts (see glossary on the Léamh.org website, it’s also the form in Eoin Mac Cárthaigh’s The Art of Bardic Poetry. So I’d say it’s the standard modern practice for Classical Gaelic, and well established since the beginning of the 20th century.

2. is used by Dinneen – his dictionary deals mostly with Modern Irish, but it uses the pre-reform spelling which is often followed by the editions of classical texts (see next section), and contains a lot of historical usage (making it perhaps the most useful dictionary for reading classical texts). It’s also used in many editions of classical texts (again, see Léamh glossary, eg. the entry for cuirim).

Generally, editions of classical texts seem to choose either 1st sg. or 3rd sg. form.

I haven’t seen 3. used for classical vocab lists (and it isn’t used by DIL – the sole dictionary actually encompassing the language) – but it’s the form that’s generally used for modern languages (both Irish and Scottish Gaelic), except for Dinneen’s dictionary. Choosing it would be useful for people wanting to find classical forms of a verb when they know the modern Irish imperative – as the lemma would often be the same.

4. is what’s found in the actual inflection lists in grammatical tracts from 14th–16th centuries – the bardic schools treated verbal noun as the name of a verb. I wouldn’t choose it though as the verbal noun has its own separate morphology and syntax, so I’d rather keep it listed among the verb’s forms and then have it as a separate lemma with its own inflection table.

Spelling normalization of lemmata
The spelling in actual early modern manuscripts and printed books vary a lot and can be anything between early Old Irish practices to pretty much post-caighdeán modern Irish sometimes.

Still, modern editions generally use late (post-15th century) style spelling (using ao for the /əː/ vowel instead of áe or the like, marking lenition of b, d, g consistently, etc.) so I think we can stick to later practices too – only listing actual orthographic variants appearing in texts in the entry.

The question is which exact spelling we should choose. We have the Irish Grammatical Tracts i: Introductory likely from early 1500s which gives a spelling guideline and it’s followed by Mac Cárthaigh in his The Art of Bardic Poetry, it has features such as:
 * 1) no a between é and a broad consonant (eg. bél rather than béal for ‘mouth’),
 * 2) doubling of eclipsed voiceless stops (a ccríochaibh instead of i gcríochaibh, a ttír and not i dtír, etc.),
 * 3) use of sg, sb, sd over sc, sp, st,
 * 4) marking diphthongs as long (slúagh for sluagh, clíar for cliar) and long di-, tri-graphs also with length-mark (Gáoidhealg for Gaoidhealg).

On the other hand, many modern editions stick to forms that got popular later and are closer to (or identical with) Dinneen’s spelling (béal, i gcríochaibh, scéal, sluagh, Gaoidhealg).

Distinguishing prose/colloquial usage vs dán díreach standard
We’ll probably want a standardized way of representing forms like the preposition do-chum which were popular (and present throughout the whole period) but were explicitly considered faulty in dán díreach (and eg. avoided by Keating in his later texts, because “incorrect”). Similar thing with a-táim for a-tú (though later poems do use the former too).

I think we’ll want to label those as (non-standard) and note in usage notes (or in footnotes in case of inflected forms) that they were proscribed by the tracts and avoided in poetry. What do you think?

Reconsider whether we want Classical Gaelic to be a sister of Irish and Sc. Gaelic or their ancestor
It’s generally agreed upon that Irish and Scottish Gaelic have split during the Middle Irish period and started their own grammatical innovations during that time. But they both kept the same literary tradition wherever Gaelic political order with poets educated in bardic schools working as diplomats and public commentators was present. Gaelic was still popularly considered a single language at the time, up to at least the 18th century.

If we adopt the policy I suggested above, anything from Scotland and Ireland from the period 13th–15th century would classify as Classical Gaelic. And so it basically is the ancestor form of both – a language with some dialectal differences but generally a single literary standard allowing the use of many dialectal forms.

And most Scottish, as well as Irish, forms can be regularly derived directly from classical ones (even though some cannot, where the classical standard follows some Irish innovations not present in Scotland). So it can be useful in Etymology sections to derive some words from Classical Gaelic (and perhaps interesting to users of Wiktionary to see a “descendants” list under classical entries).

But since it was a classical standard not really representing any particular spoken dialect and during later times fairly removed from them, I’m not convinced that’s what we want to do.

So I’m mentioning this also as something to discuss further in the next weeks or months. // Silmeth @talk 20:00, 7 February 2024 (UTC)


 * Regarding lemmata, I'm in favour of using 3rd person present indicative forms of the verb as is typical for modern glossaries of historical varieties of Gaelic. I think it would be best to keep with that practice.
 * I am also in favour of either using Dinneen-like spelling of a normalisation of the spelling conventions used in the tracts for lemmata. Both would be relatively suitable at conveying the appropriate etymological information (and pronunciation of Classical forms) while not being overly archaic as with the spelling conventions of Old Irish. Either way, it would be good to include multiple spelling conventions within each entry to show the variation used in both historical and modern conventions to represent Classical forms.
 * Regarding making Classical Gaelic a etymological language, this could be effective for Modern Irish and Scottish Gaelic as it generally does contain forms which are etymological to the modern varieties and would thus be useful as an intermediate stage between Middle Irish and Modern Gaelic to represent such Early Modern etymological developments. But I also understand the reasons why it should not be marked as an ancestral variety of the modern varieties when in actuality it was a literary variety shared between Ireland and Scotland. Erisceres (talk) 12:04, 8 February 2024 (UTC)
 * Great opening remarks. Unfortunately my lack of knowledge of Classical Gaelic prevents me from having too many opinions, so just a few comments. Please feel free to correct any embarrassing mistakes.
 * Re lemmatisation, I'd say 3rd sg. present indicative for the sake of consistency. Taking the imperative as in the modern languages would also be justifiable, but there's no reason to, as the -idh ending as I understand it is basically the same in form and function as in the older language.
 * Re family tree, I'd say let it be the ancestor of the modern Goidelic languages. I understand the point that Classical Gaelic was quite far removed from the spoken language, but on the other hand that is the case with most literary standards which have evolved from an older literary standard. Written and spoken German differ a great deal in grammar and lexicon but I don't think you would call them sister languages. I doubt Classical Gaelic was transmitted from parents to children as living languages normally are. It was something you learned later in life. —Caoimhin ceallach (talk) 13:08, 9 February 2024 (UTC)
 * thanks for the input! I guess having it be the ancestor of modern Gaelic langs would also be consistent (as Thadh noted on Discord) with how we treat Old Ruthenian (ancestor of Belarusian and Ukrainian even though it’s a fairly standardized common chancellery language somewhat removed from the vernaculars which split fairly early). // Silmeth @talk 14:32, 9 February 2024 (UTC)


 * I’ve published a first draft of the About page. // Silmeth @talk 22:54, 12 February 2024 (UTC)