The reason this interests me is based on the requirements of dependency parsing. According to Universal Dependencies guidelines, verbs can act like this, but a copula is an auxiliary and cannot. This is because a copula cannot be the head of any dependent words. It is perceived as a supplementary word itself, governed by the predicate which is the head of the clause. By contrast, where a verb is used, it is the head of the clause and all other words in the clause are its dependents (direct or not).

The result is that dependencies in phrases like amal nondafrecṅdirccsa "for that I am present" are unintuitive to say the least. The construction is amal (SCONJ) no (PART) nda (AUX) frecṅdircc (ADJ) sa (PRON), where the emphatic pronoun sa emphasises the subject, and the empty particle no is used only to create the nasalising relative clause which affects the copula. According to UD, the emphasising pronoun and empty particle are dependents of the predicate, which is an adjective, even though the emphatic pronoun actually refers to the subject which is only expressed by the copula. If this were a verb it wouldn't be a problem, because we could say sa is dependent on the verb.

One solution would be to say that, actually, in constructions without an overt subject, the "copula" is in fact a verb, operating in a manner comparable to the substantive verb. But this seems very out of line with both the traditional grammar of the language and with the diachronic development of the Irish copula.

Another potential solution, which I prefer, is to treat the copula in cases like this, as if it were the subject rather than the copula. If its relation to the predicate was as a subject primarily, rather than primarily as a copula, then it could take dependents. However, others suggest that to do this it would have to be POS tagged as a pronoun rather than as an auxiliary.

I don’t know what Universal Dependencies require, never heard of them before, and I’m no structuralist and never did any dependency grammar (and in general, I don’t like strict linguistic frameworks; from stuff I’ve read about, Construction Grammar makes most sense to me so far…). So what follows might not make any sense… but I’d approach it differently, I think.

I’d try to treat copula + the predicative as a single verb-like unit bearing stress. In is fer-som the whole is fer is the “stressed verb” on which -som depends. Similarly in is fer in rí the subject in rí follows the “verb” is fer.

The copula being basically a clitic turning nominals into “verbs”. And those whole copula + predicative stress units then behave like intransitive verbs.

I don't think theories of universal grammar are of as much relevance to language learning as knowledge of the actual language concerned. I would say the copula is a verb. Claims that it is not a verb or that it might just be a particle may be of interest to academics, but have little value for learners of a language. Much of what passes for academic research revolves around semantic categories (shall we define X as this? or shall we define Y as that? and then once we have asserted those definitions, shall we then reanalyse the language on the basis of our semantic definitions?) in a way that is fundamentally not interesting and ultimately circular and self-referential.

No such oddness in Ulster, where it's 'Cad é an t-am atá sé?' Neither FGB nor de Bhaldraithe gives this version, but An Foclóir Nua does.

It's hard to trust that dictionary as it indexes from anything published in Irish whether the speaker was native, or competent. It's full of mistakes and direct translations from English without any regard to Irish syntax or semantics (it very often uses the present tense for future situations, for instance).

Okay, can you give us some examples of those then. Anyway, ‘Cad é an/cén t-am atá sé?’ is what they say in Donegal.

From a list a friend of mine has compiled (and keeps adding to every time he finds them):

, where 'gach duine' just doesn't make sense. You're addressing them, so you'd need the vocative: 'a dhaoine'.





Should be Dé Satharn, as it's a single Saturday (see CO 20017 1.1.5)

There's also lack of consistency in how it uses certain things, often in contrast to the standard form. For instance, it often mixes the direct/indirect relatives when used after 'uair' (which, granted, can be done in dialects but it should be direct in the CO). It's very inconsistent and doesn't even follow the caighdeán. Then there's lots of times where it directly contradicts de Bhaldraithte and Ó Dónall (looking at you, 'uiscebhealach'), though that's more the Coiste Téarmaíochta not giving a rat's ass about traditional Irish semantics/syntax and just calquing than anything.

But, the fact that it pulls from non-native sources for a dictionary is a cardinal sin in any language. That'd be like using my French to help compile a French dictionary, despite my semantics and understanding being based on English.

Wow! I'd never heard of this dictionary, but that "gach duine" thing is dreadful. Some good points there, Galaxyrocker. The NuaChorpas na hÉireann violates all basic academic principles by including works by non-natives. Corpora of English or any other language don't do this. Of course, they do this because they're promoting a standard largely used by L2 speakers, and if you remove works by L2 speakers then you're deleting from the corpas the vast majority of the texts. If you create a corpas based on L1 speakers only (I mean Gaeltacht L1 speakers, not neo-natives of the "Standard"), then a large proportion of that will not be in the "Standard"... This problem all stems from the decision to create a made-up standard. The obvious thing - the stark raving obvious thing - would be to make Conamara Irish the Standard, but the Gaeltacht minister can't think that far, it seems.

Strange, I don't know if the vocative a dhaoine really rings of an imperative to me, much less one directed specifically at "everybody". Maybe a dhaoine uile? But something about the vocative seems too polite to me to constitute an order.





Not for love of the caighdeán, but this has to be a hard pass from me. As it would be for any one dialect over the others.

Then you're openly supporting a fake standard made up by a committee.