For some time I’ve been writing notes on things I’ve found interesting in Goidelic languages (and especially Munster Irish) on a Polish linguistic forum, then I’ve decided I no longer want to participate in that forum because of some of its users, and removed my Goidelic notes (but most of them can still be found in their original Polish form using the Wayback Machine).

I’ve backed them up and wanted to continue them somewhere – I considered creating an Irish-related blog, or moving them to some other general language-related forums (like unilang for example). But I’ve decided it might be best if I translate them to English and continue them here, mostly for two reasons:

• it is a forum for learners of Irish – so hopefully some of the notes might be helpful to some people here, or at least interesting for someone,
• some people here are much better at Irish and Scottish than I am or ever will be (and there are native speakers here!), so I believe there will always be someone here to correct me, should I fail to write the truth.

For most users here those notes probably won’t be interesting, as they are just points that surprised or amused a beginner Polish learner of the Irish language, but then I tried to compare all the reviewed features among all the dialects I could find any information about (so perhaps some comparisons with Scottish or Manx will be interesting). Still, as I am trying mostly to learn to read Munster Irish, that was the main focus of those posts.

I hope those also might ignite some fresh language-related discussions among all the tattoo translation posts. ;-)

As as note of caution, I’ve been writing those on a linguistic forums, so I did not refrain from using linguistic terminology without explaining it much.

I’ll be using the colours to mark languages – Irish in green, Scottish Gaelic in sky blue and Manx in red, Old and Middle Irish and Gaedhealg Chlasaiceach examples will remain in the default font colour.

---

At the beginning a few resources. For learning Goidelic languages I’d recommend to start with the Standard Irish on Duolingo (but although it is an easy starting point, it has a few pretty serious errors, so do not use it as the only resource – you can read more about what’s wrong with Duolingo and what to use instead in a lengthy reddit.com/r/gaeilge post), and read the only available Internet (but very good and quite exhaustive) Irish grammar (GnaG – Gramadach na Gaeilge), as well as make friends with the teanglann.ie site which provides some of the best Irish-English and English-Irish dictionaries, and recordings of the pronunciation of most words made by native speakers, and some information about inflection.

One can learn Scottish from the lessons on taic.me.uk (also a quite good grammar description), additionally the articles written in a looser style on the Akerbeltz wiki are worth reading.

If you’re interested in Manx… the Internet seems to be lacking. Partial grammar description is to be found in the English Wikipedia article, there also exists an online dictionary (thankfully it does have usage examples). There is also learnmanx.com with some resources but it lacks any proper grammar or dictionary.

---

First a little note about the arís word. I once (on the original Polish forum) criticized the creators of the Irish Standard – a.k.a. Caighdeán Oifigiúil – because they chose the form arís as the standard one over common dialectal aríst, so in my first note I decided to rehabilitate them a bit. (But don’t worry, I will criticize the standard again because of the written form of a particular grammatical construct in the next one…)

It seems that although it really is common in some (and perhaps even the majority) of Irish regions to use aríst for ‘again’ in the speech, arís, without /t’/ at the end, is etymologically correct according to eDIL.



So, to summarize: in Old Irish there was no simple ‘again’, but they used this word (meaning ‘return track’) with a possessive pronoun, instead of I am here again they said something along the lines of I am here in my return with the form mo(ḟ)risi, you are here in your return with do(ḟ)risi, etc.

The possessive form of the 3rd person pronouns (of all genders and numbers) is a (it just triggers different mutations of the next word depending on the gender and number), so the most frequent form (‘his/her/its/their return’) was a(f)risi, which has later been extended to other persons and gained the meaning of ‘again’. The final /t’/ had to appear later in dialects and make its way to the majority of Ireland, but arís is still there, and it is more common in literary texts (Wikisource finds it 10 times more often than aríst). So there is some reason for this form, and it’s not just arbitrary as I thought before.

Anyway, the change of the ending must have happened quite early, since Manx also has reesht, and Scottish a-rithist. That is, equivalents of Ir. aríst.

Now one thing I find very confusing in the Caighdeán. A short introduction (which on this forum is probably unnecessary, but I’ll keep it) follows.

Goidelic languages have VSO word order. Additionally all of them have gerund (verbal noun) constructions similar to English continuous tenses.

It goes like this:
Táim ag ithe (Munster)
Tá mé ag ithe (the rest)
Tha mi ag ithe
Ta mee g’ee
be.pres pron.1.sg at eat.gerund
lit. ‘Am (I) at eating’

which means ‘I’m eating’.

Verbal nouns take their object in genitive, thus ‘he is eating the apple’ is (I’ll limit this to just Irish and Scottish):

Tá sé ag ithe an úill
Tha e ag ithe an ubhail
be.pres pron.3.sg.masc at eat.gerund art-def.masc.gen apple.sg.gen
lit. ‘Is he at eating of the apple’

Interestingly: if the apple is indefinite, Irish dialects keep the genitive and the sentence just loses the definite article, ‘he’s eating an apple’:
Tá sé ag ithe úill
Is he at eating of apple
But Scottish uses nominative-accusative in this case (and not a genitive, as before):
Tha e ag ithe ubhal
Is he at eating apple

Irish also would lose the genitive (and use nom-acc.) if the indefinite noun was further attributed (eg. ‘he’s eating a small apple’ tá sé ag ithe úll beag).

When we intend to use a personal pronoun as the object, a problem appears, because the pronouns do not decline in the Gaelics, they have only nominative-accusative forms (so one cannot say that the bear is at eating of me).

Because of that the Proto-Irish began using the possessive pronouns here. Possessive pronouns are put before the noun they govern, in contrast to the genitive (similar to English eating of me vs my eating) and the sentence ‘the bear is eating me’ sounded somewhat like:

*Tá an mathghamhain / béar ag m’ithe
be.pres art-def.masc.nom bear.nom.sg at pron.1.sg.pos eat.gerund
lit. ‘Is the bear at my eating’

EDIT: and, as explained later, also something like *Tá an mathghamhain / béar do m’ithe ‘is the bear to my eating’, as it seems ag and do were interchangeable when governing verbal nouns at some point in the older language.

The problem is, the Gaels seem to not like clusters of prepositions and short pronouns when they stand next to each other, especially when they fulfil a special grammatical role. And they contract them. By chance, each of the major dialects contracted them differently.

A very simplified picture follows.

In Scotland they were pretty consistent (so we’ll begin here):
gam (← ag mo + lenition, ‘at my’)
gad (← ag do + lenition, ‘at thy’)
ga (← ag a + lenition, ‘at his’; no lenition, ‘at her’)
gar (← ag ar + n- with vowel, ‘at our’)
gur (← ag (bh)ur + n- with vowel, ‘at y’all’s’)
gaN (← ag aN, where N is a nasal depending on the following word, ‘at their’)

And so in Scottish Gaelic:
‘the bear is eating me’, ‘the bear is eating thee’, ‘the bear is eating him’, ‘the bear is eating her’, ‘the bear is eating us’, ‘the bear is eating y’all’, ‘the bear is eating them’ is respectively:
Tha am mathan gam ithe, Tha am mathan gad ithe, Tha am mathan ga ithe, Tha am mathan ga h-ithe, Tha am mathan gar n-ithe, Tha am mathan gur n-ithe, Tha am mathan gan ithe

In Ulster – the part of Ireland closest to Scotland – it keeps being pretty simple. Ag typically disappears, the third person pronouns get lengthened and we get:

‘the bear is eating me’, ‘the bear is eating thee’, ‘the bear is eating him’, ‘the bear is eating her’, ‘the bear is eating us’, ‘the bear is eating y’all’, ‘the bear is eating them’:
Tá an béar (ag) m’ithe, Tá an béar (ag) d’ithe, Tá an béar á ithe, Tá an béar á hithe, Tá an béar (ag) ár n-ithe, Tá an béar (ag) bhur /mur/ n-ithe, Tá an béar á n-ithe

In Ulster bhur is pronounced /mur/ and sometimes also written as mur.

In Munster something similar happens – from ag a- remains, the pronoun typically keeps its consonant, third person pronouns are lengthened:
‘the bear is eating me’, ‘the bear is eating thee’, ‘the bear is eating him’, ‘the bear is eating her’, ‘the bear is eating us’, ‘the bear is eating y’all’, ‘the bear is eating them’:
Tá an béar am ithe, Tá an béar ad ithe, Tá an béar á ithe, Tá an béar á hithe, Tá an béar ár n-ithe, Tá an béar bhur n-ithe, Tá an béar á n-ithe

But Connacht is problematic. Here ag + a vowel is lenited, becomes agh + V. Then this /gh/ [ɣ] is reanalyzed as /dh/, because lenited /d/ is pronounced the same. Ag mo, ag do, etc. start to be pronounced similarly to Scottish as /gə mo/, /gə do/… Also do ‘to, for’ in Connacht becomes pronounced as /gə/ (as if spelt go). Connacht grammarians (or so I think I’ve read somewhere, but now I cannot find any actual source on it) of the 19th century decide that it’s not really ag + pronoun, but do (=‘to, for’) + pronoun. The confusion might have been reinforced by the fact, that do actually often governs the verbal noun (but not in this very construction).

On the other hand, as noted at the end, in the Middle Irish (and perhaps Classical Gaelic?) ag and do were commonly used interchangeably in all constructions involving verbal nouns (and ag finally won everywhere in continuous progressive sentences), so some of these Connacht contractions might actually come from do (if they contracted before ag won over do here), so there might have been no confusion here actually…

And so we have in writing:
‘the bear is eating me’, ‘the bear is eating thee’, ‘the bear is eating him’, ‘the bear is eating her’, ‘the bear is eating us’, ‘the bear is eating y’all’, ‘the bear is eating them’:
Tá an béar do m’ithe, Tá an béar do d’ithe, Tá an béar dhá ithe, Tá an béar dhá hithe, Tá an béar dhár n-ithe, Tá an béar dho’ur n-ithe, Tá an béar dhá n-ithe

But do mo (before a vowel do m’) is read with [g], like go mo, do do (do d’) is read like go do etc., and dhá is read just as if written, more etymologically, ghá… (EDIT: or not more etymologically, as it actually make sense that the lengthening comes from contraction of two vowels: do + a, and ag + a would just give short *gha?).

Also in Connacht all possessive pronouns in the plural are commonly pronounced the same, as /ə/, and all the plural contractions here (dhár, dho’ur, dhá) as /ɣaː/.

There is also a similar passive construction, ‘the bear is being eaten by me’ is literally ‘the bear is at its eating at (=by) me’ Tá an béar á ithe agam, with the same contractions. But some writers supposedly made a written distincion between /ɣaː/ in progressive written ghá (as if from ag) and in passive written dhá (as if from do).

As an effect of all this, the creators of the official standard believed the 19th century grammarians, creating this monstrosity (all the sentences in the same order as earlier):
Tá an béar do m’ithe, Tá an béar do d’ithe, Tá an béar á ithe, Tá an béar á hithe, Tá an béar dár n-ithe, Tá an béar do bhur n-ithe, Tá an béar á n-ithe

Which makes no sense at all, as it does not agree with any dialect, with any dialect’s pronunciation, nor with etymology (but it somehow keeps the richness of the language, as it revives the usage of do in the progressive sentences).

Interestingly, it seems in Manx this construction died out and one just uses the nominative-accusative pronoun as an object, ‘the bear is eating me’:

Ta’n maghouin g’ee mee
or something like that…

Nice comparison of different forms from Munster to Scotland and a general description of a progressive affirmative sentence with a verbal noun in Irish can be found in the GnaG.

On the Fòram na Gàidhlig there is a subject about pronoun object in Scottish Gaelic.

What I’m missing for a fuller picture is how similar passive sentences are constructed in Scotish and Manx or if they exist there at all – did there also the constructions of ag + pronoun and do + pronoun mix? But judging from the fact that on Mann ag+pronoun disappeared and Scottish seems to ignore the genitive more often, there is little chance it remained as a separate construction…

EDIT: but as it was also discussed on this forum, it seems that in Middle Irish ag and do were used quite interchangeably in progressive, passive, and infinitive-like construction (where do won and changed to a in modern language), so some of those contractions might actually come from do and not ag. It doesn’t change the fact that the Caighdeán do mo, do do, do bhur forms are not actually used anywhere.

EDIT2: It seems Scottish has two similar passive constructions. One with aig, equivalent to Irish progressive passive.

The other is perfective with a different preposition – air ’after’ ← OIr. iar, used in this form only with verbal nouns in the perfective expressions (but it has the same form as air ‘on’ coming from merging of OIr. for and ar, equivalent to Irish ar) – thanks Lughaidh for pointing this out.

So ‘the bear is (has been) eaten’ is tha am mathan air a ithe lit. ‘is the bear after its eating’ and ‘the bear is being eaten’ is tha am mathan ga ithe (which can also mean active ‘the bear is eating it/him’). I am not clear if one can provide the agent of such a passive action, as one can in Irish – if agam at the end for ‘by me’ makes any sense in Scottish.

Ulster doesn't use "bhur" but "mur"


afaik "air" before a verbal noun in ScG rather means "tar éis/i ndiaidh" : tha e air falbh = tá sé i ndiaidh imeacht = He has just left.

_________________
Is fearr Gaeilg na Gaeltaċta ná Gaeilg ar biṫ eile
Agus is í Gaeilg Ġaoṫ Doḃair is binne

Interesting, haven’t read too much about Ulster specifically and never come across it. Thanks.

Now I see Wiktionary mentions it.


Makes sense, then tha e air ithe ‘he is on eating’ = ‘he has eaten’ while with a possessive refering back to the subject you get the passive: tha mi air m’ ithe ‘I am on my eating’ = ‘I am / have been eaten’.

Historically that "air" comes from the preposition "iar" (+ eclipsis) which meant "after". If I remember well it's "er" in Manx, and it's still followed by the eclipsis.

_________________
Is fearr Gaeilg na Gaeltaċta ná Gaeilg ar biṫ eile
Agus is í Gaeilg Ġaoṫ Doḃair is binne

My only gripe with this is the recommendation for Duolingo. I definitely would not recommend anyone to start there. The quality is highly questionable, and even some of the commentators, while helpful at times, don't know as much as they think they do (i.e. one repeatedly gets in arguments with me saying cuid is never necessary, but only a matter of stress; needless to say, he's an American learner of the language who I also noticed used "d'iarr" in place of "d'fhiafraigh"...sadly he teaches it too) It can be alright once you get some grammar and vocab, as you can use it to practice what you've already learned, but otherwise I'd say stay away.



Personally, I lean towards the latter explanation. It just seems to fit Occam's Razor more to say that the do was already pronounced as go before it fell out of use, and thus there's no reason to propose a lenition of the ag + V structure. It'd simply follow the rules of the pronoun do completely perfectly. This could also make sense wit ag, as the a'[/] is often dropped before vowel sounds -- [i]tá mé g'ithe and, even when before a consonant, when it's the 'g that's dropped, you're left with just a schwa sound. So, regardless of which part gets dropped, it still seems like it could be analyzed as do fairly easily.




Which makes sense since I don't think many of them were/are native speakers?

That makes even more sense. After going through MacBain’s, faclair.com and Wiktionary I see it seems to exist in this form only in this construction, and besides as iar- in a few compounds like iar-aois ‘after-age’ (and similarly in Irish).

As for Manx, the online fockleyreen lists eg. oie er oie ‘night after night’, strangely, without any sign of eclipsis or n-prothesis, but there is also er n’aase (= iar n-fhás(adh/aigh)?) ‘grown’ with n-.

Anyway I’ll edit the post to correct this.


Yeah, we’ve talked about it briefly already on Reddit. I’ll edit to add a note of caution, and a link to explanation on r/gaeilge. I agree it has a few very annoying mistakes, but I still find Duolingo (and explanations in its lessons) a good first introduction to a Goidelic language (but absolutely not as an only resource to learn Irish).


Now I lean towards it more too, to be honest. I first wrote that note a few years back when I had no idea that do was actually in use there, so the whole situation was very confusing to me, and back then á seemed likely to come from *agha too.

I should probably just rewrite it, giving a bit more proper historical background. Maybe one day (when I finish reading Stifter’s Sengoídelc, and hopefully have a fuller picture then)…

The t-suffix is known as the intrusive t. Intrusive t becomes especially common after c.1750/ 1800, where it was often suffixed to nouns, verbal nouns and adverbs ending in -n, or -s. Hence, canmhain > canmhaint (Ir. canúint; cf. Sc. Gaelic Cànan), aríst, féachain > féachaint etc...

Intrusive t also occurs on verbal terminations ending in -aimís and -aidís in the Irish of east Cork, Waterford, Tipperary, Clare, and west and north Kerry: e.g. do théimíst, dá nglanaidíst etc...

The loss of f in frithissi is not a unique feature; the phenomenon is found throughout the development of the Irish language. The loss of f is attributed to what is known as hypercorrection. Since, lenited f has no sound, confusion arose as to whether f was in fact part of the original word, and thus the initial f often drops out. Hypercorrection also occurs in reverse, whereby f is often added to words which etymologically began with a vowel. In these instances, the f is known as an inorganic f: e.g. Old Ir. co n-accae > f(e)aca (nícon accae > ní fheaca); Old Ir. anaid > fan (nícon anaid > ní fhanann). The phenomenon continues to manifest itself in Modern Irish: iolar/ fiolar 'eagle' and aill 'cliff', Munster Irish: faill (cf. an fhaill, where an aill and an fhaill are indistinguishable).

The loss of personal endings in regards to frithissi/ aḟ- is not uncommon and also occurs with rarely used personal pronouns, e.g. féin. The use of the 3rd singular here as an adverb, meaning 'again', is part of the tendency for 3rd personal sing. of prepositional pronouns to gain adverbial force in general, even in Old Irish itself: íarum literally: 'after him/ it', adverb: 'afterwards'; ann: literally: 'in him/ it', adverb: 'there'. It occurs later in Middle Ir. with riam, literally: 'before him/ it', later used as an adverb for 'before'.

_________________

Please wait for corrections/ more input from other forum members before acting on advice

Yep, the CO often standardises ar as iar-, as it was in proper Old Irish, often used as a prefix meaning 'former'/ 'aforementioned' e.g. iar-scoláire 'former pupil', forainm iar-thagrach 'aforementioned pronoun'. However, iar becomes ar even in the Old Irish period, ar still exists in largely petrified phrases in Modern Irish, e.g. diaidh ar ndiaidh, ar ndóigh arna ...



á does indeed come from agha (ag + a). When a consonant is flanked on either side by vowels, the consonant becomes lenited and devoiced, leading to the vowels collapsing together causing diphthongisation, or compensatory lengthening. Hence, Old Ir. occa mmarbud 'killing them' > agha > aa = á marú. You see this phenomenon everywhere in Irish:

ag marbud > marbhaghadh (i.e. ag marú)
Dearmad, Munster Irish: dearmhad: dearúd.
Canmhaint > canúint
Suimeamhail > Suimiúil

It also occurs in Old Irish:

ad-cluinith- 'hear', is reduplicated in the preterite, whereby the first consonant is duplicated: cocluin-. The medial -c- becomes lenited and eventually devoiced: cochli-, resulting in diphthongisation: cúala.

There are exceptions to the rule: seachtain (<seachtmhain), is one that immediately comes to mind.

However, Scottish Gaelic tends not to follow this rule:

Old Irish: canmain > Cànan
Cuntabairt > Ir. cuntúirt, Sc. G cunnart
Feramail > Ir. Fearúil, Sc. G. fearail
Mathg[h]am[h]ain > Ir. mathúin, Sc. G mathan

See O'Rahilly, T. F., Irish dialects past and present, pg. 126-127, for further examples.

This of course follows Sc. Gaelic's propensity to shorten long vowels; the same phenomenon exists in Ulster Irish, but not to the same extent.

This may explain why Scottish Gaelic doesn't have the lengthened forms in the 3rd person sing. and plural:

'gam
'gad
'ga
'gar
'gur
'ga







The chapter on Connacht Irish and Classical Irish in Stair na Gaeilge doesn't seem, as far as I can see (I only took a quick look), to shed any light on this issue. However, I agree, I think it is probably do, realised as dho, as well.

Verb + subject + ag + adjectival pronoun (object) + verbal noun seems to be the form used in Old Irish: See Thurneysen, pg. 524. But there is a lack of any significant discussion.

See also Aislinge Óenguso :

Blíadain lán dó os sí occa aithigid fon séol sin (https://iso.ucc.ie/Aislinge-oenguso/Ais ... -text.html). First line of Section 2.

This is pure speculation on my part, but I reckon the use of do here is by analogy with other similar semantic instances where do is used with the verbal noun. A clue might be in Thurneysen:



See also: Aislinge Óenguso, section 1:

Luid Óengus do gabáil a llámae

Chuaigh Aonghas chun a láimh a ghabháil

Óengus went to take her hand,

dia tabairt cucci inna imdai.

chun í a thabhairt isteach sa leaba chuige

to take her to him into his compartment.

Section 3

co tuidich dot accaldaim 'tagadh sí chun cainte leat' = [literally: go dtagaidh dot'agallamh] 'who has come for your conversing'.

Section 11

“Timmarnad duit ó Ailill ocus Meidb dul dia n-accaldaim...”

“Tá ord agat ó Oilill agus Meadhbh dul chun cainte leo...”

“A request for you from Ailill and Medb to go to talk with them”.


...ro-fitir aní dia congarar.”

mar tá a fhios aige cad chuige a ghlaotar [ghartar] air.”

he knows why he was summoned”.

_________________

Please wait for corrections/ more input from other forum members before acting on advice

My first thought was to check those two chapters as well, actually. I found little about the syntax of the various verbal noun clauses in either of them; it mostly discussed the formation. In fact, it seems to be that way through the entire book, judging from the index. I also looked in Ó Curáin's The Irish of Iorras Aithneach, but wasn't able to find much there in the way of syntax of the verbal noun; the few examples listed under the verbal noun section of the book on verbs (2) detailed mostly uncommon, and in some cases obsolete, uses of the VN.