In my Old Testament notes I have written this:

dá, dhá: dá and dhá are combinations of the preposition ag/do governing the verbal noun and a third-person pronoun object. PUL used dá in passive senses (dá dhéanamh, “being done”) and dhá (’ghá in the original) in a transitive context (dhá dhéanamh, “doing it”). Where the latter is given in the original as ’á, this has been edited here as á. Both would be likely to be written á and pronounced /ɑ:/ by later speakers of the WM dialect, and á is also the usage of the GCh in both meanings. PUL often uses am in the original text where the verbal noun takes a first-person singular pronoun object (corresponding to do mo in GCh). This is edited here as ’om. Dom is also given in this meaning in Genesis 49:29. ’Ot is found before vowels and often before mh (see ’ot mhaslú in Psalm 68). ’Ghat’ in Genesis 37:10 is edited here as dhot (corresponding to do do in GCh). Dúr in Genesis 50:23, Isaiah 46:3 and Ieremias 29:7 couples do with a second-person plural object of the verbal noun. Rarer examples exist, where the passive sense relates to a non-third-person object, such as dod bhrú in Deuteronomí 28:33 and dár dtarrac in Caoineadh 5:5.

ag: “at”. Ag is used with pronoun objects of the verbal noun in active meanings: compare beidh mé ag úr n-iompar, “I will be carrying you”, with sibhse atá dúr n-iompar, “you who are being carried”, in Isaiah 46:3-4. See also ag úr n-árdú féin in Uimhreacha 16:3, ag úr n-iomadú, ag úr ndísciú and ag úr gcur ar neamhní in Deuteronomí 28:63, ’gúr spáráil in Iob 16:6, ’gá ollmhú in Iob 27:17, ag úr gcrá in 1 Samuél 10:18 and ag úr mealladh in 4 Ríthe 18:32 and 2 Paralipomenón 32:11. PUL was insistent that there was a difference between ag and do in such constructions, but it seems the two are aligned in the modern dialect, with the forms being ’om, ’od, á, ár, úr and á. (The form ’ot in place of ’od is found not only before vowels, but also in other combinations such as ’ot lorg.) In this edition, active constructions with ’ghá in the manuscript are transcribed here as dhá (in line with the editorial policy of Coiste Litríochta Mhúscraí). But where an unlenited ’gá in such an active construction is found in the manuscript, it is retained. See under dá, dhá.

snoím, snoí: Dhom shnoí (Psalm 138), “wasting away; growing thin, emaciated”, (spelt gham’ shnoídhe in the manuscript, literally “wasting myself”). Dá shnoí amach in 2 Paralipomenon 21:19 indicates that the distinction beween ’ghá and dá in such constructions was artificial and inconsistent. Compare dúr gcloí under cloím, cloí.

cloím, cloí: Dúr gcloí (Esechiel 24), “pining away, wasting away”, literally “wearing yourself down”. Compare dhom shnoí under snoím, snoí.

They are just some notes on some phrases I found in Peadar Ua Laoghaire's Old Testament manuscripts. I haven't looked at those notes for several years, and I'm not sure what I would make of the issue today.

I’m fairly sure the distinction between active/passive meaning is a much much later unetymological orthographical convention. I’m pretty sure there was no such distinction in classical language, and nothing(?) like that in OIr. or MIr.

The question really is whether the later progressive forms /əm, əd/ (and also Scottish gam, gad, Ulster (ag) mo, (ag) do, (ag) ár /(ə) mə, (ə) də, (əɡ) æːr/, Connacht /ɡə mə, ɡə də/, Manx dy my, dy dty /də mə, də də/ – German GnaG gives a nice overview) come from ag or do. IMO it’s clear that Ulster and Scottish ones come from ag, Connacht could be either (since do and go merged, hence both ag mo and do mo would end up with /ɡ/) and Manx seems to show do here (in later Manx gy changed to dy – but apparently the dy my forms are attested in older texts that generally keep clear distinction between gy and dy).

Ó Nualláin’s Studies in Modern Irish vol. 4 on prose of Keating has:


I believe in Old Irish generally oc (later ag) was used for the progressive sense (though in OIr. times it was far from mandatory, as McCone puts it in The Early Irish Verb, p. 22: The optional status of the progressive in Old Irish may be compared with the ‘used to’ habitual of Modern English, which is likewise optional (e.g. ‘he used to go home every evening’ or ‘he went home every evening’). These rather marginal Old Irish periphrastic progressives will be ignored hereafter.) – do, I think, was restricted to the purpose ‘to do’ meaning.

But then I’ve read that those prepositions generally got mixed before verbal nouns (later constructions like an teach atáim do thógáil, a thógáil show this, though an taigh a tha mi a’ (=ag) togail in Sc. Gaelic – but then introduction to Stories from Keating’s History of Ireland claims it’s a “a solecism unknown to the speech” that in Sc. Gaelic is based on analogy with non-relative construction, see also this post – I’m somewhat suspsicious about this claim too).

In this response to my speculations Cionnfhaolach gave some examples of Old Irish using do in the purpose meaning too (‘to go to talk to them’, ‘has come to convert you, to be converting you’). An Lon Dubh wrote in that thread there is no agreement on the issue and generally it seems it came from confusion between ag and do in the classical times.

So my own opinion – but really just based on random discussions on the Internet, and a few examples of older usage – because nobody seems to have dealt with it in details – is that some dialectal forms continue ag and some possibly do (and I think that Manx gives strongest evidence for that), perhaps without a way to find a definite answer. But because of that I’d really like to know what evidence do people claiming it’s definitely do have.

Silmeth, I agree that ag seems more logical, as the meaning of these sentences is progressive, not purposeful (usually).

tá sé ag déanamh
tá sé ag a dhéanamh = á dhéanamh

Why wouldn't that just be the progressive with a possessive included?

I'm not sure tigh a thógáilt vs. tigh do thógháilt is relevant. A is just the worn down form of do there˙

Beidh mé ag úr n-iompar in Ua Laoghaire's Bible is quite explicitly not "do bhur". But whether that stemmed from his own ideas about Irish grammar is another thing entirely.

I find it relevant because it is the relativized version of the progressive (a)táim ag tóg(bh)áil an tighe ‘I am building the house’.

Munster uses straightforward passive in the relative: an tigh atá á thógáil agam, but other dialects seem to do something weird and make a direct relative clause with the antecedent being the object of a verbal noun: an teach atá mé a thógáil – but with do instead of ag, except in Scotland (and perhaps part of Ulster) where it’s ag.

This construction is a later one (acc. to Introduction in Stories from Keating’s… it appears there only twice), the earlier one being given as an teach agá dtaoi thógbháil though I don’t really understand it – it uses the eclipsing ‘indirect’ relative pronoun (so as if an teach ag a bhfuil tú thógáil in modern language), but I don’t understand where (or whether) the genitive/possessive a ‘its’ has hidden here nor why thógbháil is lenited (is it because of this invisible a?).

Anyway – it shows to me at least one context where ‘infitive-like’ construction an teach do thógbháil ‘to build a house’ gets seemingly confused with the progressive ag tógbháil an tighe and the earlier do thógbháil an tighe of purpose.