Notes on Nolan (the Relative Clause)

THE RELATIVE CLAUSE

Below are listed the many instances where a direct relative is used where an indirect relative would be more logical:

A. Temporal Clauses. Direct relative particle a is used in temporal clauses where the indirect relative would be more logical, like ón lá úd a ghoibh sé orthu and an fhaid a bhíos ag cuardach. Nuair also takes the direct relative. However the indirect relative clause is also found with other temporal phrases, such as um an dtaca go raibh sé and le línn ha haimsire ‘na rabhadar.

This leads to the surprising discovery, that since nuair takes the direct relative, temporal clauses often contain both direct and indirect relatives in the same clause, such as nuair a tháinig an t-am ‘nar mhithid do é.

Note that in a double relative clause with is dó’ le we have both direct and indirect relatives side by side, but the logical order is inverted by putting the indirect relative with the is dó clause. For example in ag breithniú na haimsire ‘nar dhó’ leó a bheadh sé ag teacht we have a fusion of two clauses:

1) na haimsire ba dhó’ leó (direct relative), and

2) na haimsire ‘na mbeadh sé ag teacht (indirect relative)

but once fused together in a double relative clause, the indirect relative precedes the dó’ clause.

B. Modal Clauses. The direct relative is used where the indirect relative would be more logical in modal clauses of the type mar a bhí, with the doxology used as an example: mar a bhí ar dtúis, mar atá anois, is mar a bheidh go brách.

In the case of mar the use of the direct relative distinguishes meanings:

mar tá: because there is

mar atá: as there is

mar a bhfuil: where there is

Chómh haicillí agus d’fhéadfad é is an example of a direct relative do being used with indirect relative modal meaning.

Note that although conas takes the direct relative in Munster, the indirect relative is used with cad é an chuma…?

C. The direct relative is used where the indirect relative would be more logical following the proleptic a or (“however”): bhí iúnadh a gcroí orthu a fheabhas do dheineadar an gnó.

D. The direct relative is used where the indirect relative would be more logical following proleptic de and other prepositional pronouns used proleptically. The example given is b’fhéidir gur déinide a déanfar an guí an teachtaireacht do chur tímpall uaitse, where a déanfar is equivlaent to an chuma na ndéanfar.

E. The direct relative is used after proleptic amhlaidh and GÓN claims this is logically superfluous (e.g. in is amhlaidh a dhein sé…). It doesn’t appear superfluous to me.

[F. Sar, “before”, takes the indirect relative in Munster (where Keating had used the direct relative).]

G. Emphatic elliptical sentences also have a direct relative where there is no nominative or accusative relationship and therefore an indirect relative might have been expected: is liomsa a bhaineann an chainnt sin.

H. In relative clauses that follow a comparative or superlative, where the comparative/superlative is not relative itself, the meaning is modal and therefore is another instance of the direct relative being used for an indirect meaning. Is feárr is eól dómhsa é ná mar is eól duitse é.

Where the comparative or superlative is itself relative, the relative that follows it is indirect: níl éinne is feárr gurb eól do é ná mar is eól dómhsa é. Here is feárr is relative, and followed by the indirect relative gurb eól. Another example is tá in easnamh fós air an ní is mó na bhfuil gá aige leis.

Where the dative relationship is expressed before the comparative/superlative, however, the indirect relative does not follow the comparative/superlative: is dómhsa is feárr is eól cá luíonn an bhróg orm. An exception to this is clauses with and cad: an example from Niamh is bhíodh an formad ann, leis, féachaint cé aige dob fheárr na mbeadh an t-ollmhúchán déanta.

I. GÓN also says that sentences like cad é an rud a bhí Bruadar ar aigne a dhéanamh?, which he says are rarely found, are an instance of the direct for the indirect relative. It seems straightforward to me, but maybe this sentence can be essentially genitive, with a dhéanamh referring obliquely to an rud. Many of these various types seem straightforward and only become less so by overanalysis.

Other direct relative particles

Do as a relative particle: GÓN shows that the use of do as a relative particle has developed by confusion from the perfective particle (do chuaigh) and the verbal prefix (do-gheibhim). As these were often used without do, the particle came to take over relative functions. This is similar to way the relative atá and adeir aided the spread of a as a relative particle. Originally, lenition alone was enough to give relative force, later on do and a took over this function. It is also possible that in phrases like cad d’oireann dóibh, the relative particle is derived from the final consonant of cad, i.e. cad oireann became cad d’oireann.

A dh’ is also used as relative, either as a combination of a and do or maybe the dh’ is just phonetic padding. An example of an accusative relationship is an té a dh’éiríonn go moch bíonn an rath air. This usage bears a comparison to the use of do and a dh’ as the particule governing verbal nouns: scéal a dh’ínsint.

Finally, the direct relative can simply be understood, with lenition alone showing the relationship: sin é mhíll me.

Indirect relatives

A (with eclipsis) and ar (with lenition):

The relationship is not nominative or accusative. Genitive examples include sid é an fear ar chodlas na thigh aréir and sid í an bhean a bhfuair a fear bas inné.

A and ar as indirect relatives are more commonly used in Munster Irish with sar and mar (in the meaning of “where”). Sula raibh uain ag an marcach é thabhairt fé ndeara: sula here includes the indirect relative particle. Fan mar a bhfuil agat. Is mairg don fhear san trína ndéanfar Mac an duine do dhíol.

Go and gur

Dative examples include is mó duine go mbíonn an tseóid úd aige. A genitive example is ní dó’ liom go mbeidh an fear eile sin le fáil is mó go mbeidh a bhean na cabhair is na cúnamh aige ná mar a bheidh sí agatsa.

Some uses of go appear irregular, because a usage is found that is equivalent to a genitive, and therefore what GÓN calls a virtual genitive. Na hoibreacha gur thug m’athair dom iad le déanamh could be rephrased with a ndéanamh at the end.

In some senses, supplying an unexpressed ’na thaobh will bring out the virtual genitive relatinship: Íosa éigin a fuair bás & go raibh Pól dhá rá [’na thaobh] go bhfuil sé beó.

Na and ’nar

An example is an bhean na bhfuil an t-éadach corcra uirthi.

These can also be virtually genitive: an t-aimhleas na mbeadh duine lánceapaiathe ar é dhéanamh is virtually genitive because é dhéanamh can be replaced by a dhéanamh.

Na and go can be found in the same sentence: brostú chun na háite na mbeidh aoibhneas síoraí againn, & go bhfanfaidh sé againn.

The fact that go is a conjunction in subordinate clauses as well as an indirect relative particle means that in some sentences you could parse go in a number of ways. For example: is aoibhinn don té go dtabharfairse teagasc do, & go múinfir as do dhlí é. This could be the conjunction go with ellipsis of a rá after the ampersand, or you could parse it as a relative particle used with an accusative relationship (i.e., one that should be a direct relative) under influence of the earlier indirect relative in the first clause.

In sid é an té gur leis é and cad é an ceanntar ’narbh as é, the prepositional pronoun has to follow the relative (you can’t say le gur or anything like it). GÓN links this with his earlier discussion of the copula, pointing out that the prepositional phrase is the predicate that follows the copula.

Go as a realtive developed from ag+the indirect relative a, although the connection with ag is no longer felt, and so goaige is frequently found.

In Keating, gach mnaoi dhíobh ag ar mhair a fear ag teacht in Éirinn dóibh does not mean what it would appear to mean in modern Irish, where ag ar mhair a fear would mean “with whom her husband lived”. Keating was using the earlier grammar before ag+a had turned into go, and he means gur mhair a fear (“whose husband lived”).

The conjunction go has also influenced the development of go as an indirect relative, as it is often hard to tell which the writer is intending when he writes go. Take these examples:

Cad é mar dhuine é seo go smachtaíonn sé gaoth & farraige & go ndeinid siad rud air. The first go is a conjunction, and the second appears to be relative.

Cá bhfuil an seómra bídh go n-ithead an Cháisc i bhfochair mo dheisceabal? Here go appears to be a conjunction used with the subjunctive, but by supplying an additional unexpressed ann it would become relative.

Féachaint an bhfaigheadh sé aon rud go bhféadfadh sé greim a bhreith air. Here go could be a conjunction, with the clause meaning “that he might take hold of it”, or a relative particle, with the clause meaning “that he might take hold of”.

A final point on the development of go as a conjunction is that as go and are the affirmative and negative conjunctions leading into subordinate clauses it was natural that an affirmative relative go would develop in contrast to the negative relative .

On the development of the relative na and nar, GÓN points to the way that prepositions combined with the indirect relative to form ina, óna, go n-a (where go means “with”), trína, lena, etc. An example from Catilína is gach treabhchas daoine lenar mian iad féin a bheith os cionn na n-ainmhithe eile, where lenar is used for what would once have been lér’ in older Irish. This constant use of na(r) became relative in feel, producing things like, from Séadna, san áit nar bhain an órdóg leis an dtalamh.

That na(r) has nothing to do with the preposition i or at least not any more is shown by the use of ann after it, e.g. in the phrase saothar a dhéanamh na mbeidh tairbhe ann, where the preposition force of the n- is no longer felt, and na is just the indirect relative.

In his further remarks on the relative, GÓN notes that whereas in Ulster they say goidé tá tu a dhéanadh? this becomes in Munster cad tá agat á dhéanamh?

Double relatives

GÓN claims to be the first to notice the double relative construction in Irish. He gives as an example cé is dó’ leat do scríbh an leitir? This fuses together two separate relative sentences:

Cé (hé an té) is dó’ leat?

Cé (hé an té) do scríbh an leitir?

The two sentences are connected, in that the second of these relative sentences is dependent in thought on the first, and thus combined in the one overall double relative sentence given above.

In tógfar uaidh gach a measann sé atá aige, we have a similar fusion of two relative sentences:

Tógfar uaidh gach a measann sé.

Tógfar uaidh gach a bhfuil aige.

We saw earlier how in sentences like ag breithniú na haimsire ‘nar dhó’ leó a bheadh sé ag teacht the indirect relative is shifted to the dó’ clause, which is then followed by a direct relative. The same process is at work here. As gach a measann sé already has the indirect relative, it is followed by atá aige, even though the separate sentences nested in this double relative sentence show that gach a bhfuil aige with an indirect relative has been subsumed.

Treble, quadruple and quintuple relative clauses

There can be multiple relative clauses nested within the one sentence. Take this: is minic nuair is sia is dó’ leat a bhímse uait gurb ea is giorra bhím duit. This subsumes the following three clauses:

1. an uair is sia

2. an uair is dó’ leat

3. an uair a bhímse uait

The meaning of the first and third clauses is dependent in thought on the second clause, is dó’ leat.

A quadruple relative sentence can be created thus: cad é an uair aduart is sia is dó’ leat a bhímse uait? And you can even have a quintuple relative sentence in cad é an rud is dó’ leat aduairt sé do mheas sé ab fheárr a dh’oirfeadh don scoil? (“What do you think he said he thought would suit the school best?”) This includes the following five relative clauses:

1. Cad é an rud is dó; leat?

2. Cad é an rud aduairt sé?

3. Cad é an rud do mheas sé?

4. Cad é an rud ab fheárr?

5. Cad é an rud a dh’oirfeadh don scoil?

The dependence in thought in this relative sentence is such that the meaning of clause 5 is dependent on the other 4.

As mentioned previously, an indirect relative clause later in the sentence is often transferred to the is dó’ clause. Takes these 5 nested relative clauses (to produce a sentence meaning “in what way do you think he said they thought things would work out to their benefit?”):

1. Cad é an chuma is dó’ leat?

2. Cad é an chuma aduairt sé?

3. Cad é an chuma do mheasadar?

4. Cad é an chuma ab fheárr?

5. Cad é an chuma na raghadh an scéal i dtairbhe dhóibh?

By transferring the indirect relative clause to the first nested question, the final sentence then becomes cad é an chuma nar dhó’ leat aduairt sé do mheasadar ab fheárr a raghadh an scéal i dtairbhe dhóibh? [Compare ag breithniú na haimsire ‘nar dhó’ leó a bheadh sé ag teacht and tógfar uaidh gach a measann sé atá aige above.]

Another example given is ní baol ná go raghadh an tóir sa treó baíll ba lú nar dhó’ leó a tiocfí suas leó, where the indirect relative is also transferred from sa treó baíll na dtiocfí suas leó to the dó’ clause.

Exceptions and abnormalities in the double relative

A large number of cases of unexpected usages are listed.

1. In an té adeir sibhse gurb é úr nDia é (“he whom you say is your God”, from John 8:54), the gur clause (subordinating rather than relative) requires explanation. Without it, you would have had to say an té adeir sibhse is úr nDia (bearing mind that we saw in an earlier chapter that where the relative particle governs the copula as its subject, no additional pronoun or subpredicate is inserted). This construction, while correct, is also a little unusual. GÓN also argues that although adeir sibhse is accusative (“he whom you say”), it seems equivalent to a genitive (“he of whom you say”) and GÓN states that the double relative does not permit a genitival relationship in the first clause, but requires some kind of conjunctional clause to intervene, e.g., an té nar dhó’ libh na thaobh go… [Reading this passage as a whole, GÓN appears to be saying that if you said something like an té go ndeir sibhse na thaobh go..., it would be a viable sentence, but would not be double relative, because once you introduce the genitival relationship into the first clause the second go must be parsed as a conjunction and not a relative.]

2. Many apparent exceptions arise from confusion between go as a relative and go as a conjunction. In canad is toil leat go n

ollmhóimís é? the go appears to be a conjunction, but is actually relative, thus correctly forming a double relative sentence. The nested clauses are:

a) canad is toil leat?

b) canad go n-ollmhóimís é (which is equivalent to canad na n-ollmhóimís é.)

3. In many sentences go and could be parsed either way, as conjunctions or as relative particles, thus making it unclear if they are exceptions to the double relative construciton or not. For example, in buairt nár mheasas riamh gurbh fhéidir a leithéid do theacht ar mhnaoi, gur may be genitive relative (“a sorrow the like of which”), but nár could also be interpreted as the negative conjunction (“a sorrow so great that I did not…”). If nár is a conjunction, then gur must also be parsed as a conjunction, as there could be no double relative in that case. Finally, if nár is interpreted as a negative genitive relative (a truncation of nár mheasas riamh ’na thaobh), then gur must be a conjunction, as, as was stated above, the double relative cannot be used with a genitive relationship in the first clause.

The key issue in double relative sentences is therefore that the direct relative must be used in the first clause. In rud ba dhó’ le héinne nárbh fhéidir a dh’fháil, we see that ba is direct relative and nár is also a relative here (a genitival relative linking up with a dh’fháil), and so this is a double relative sentence. But once you say rud gur dhó’ le héinne nárbh fhéidir a dh’fháil, you no longer have a double relative sentence. This is because gur in the first clause is indirect relative (in a genitival relationship implying gur dhó’ le héinne ’na thaobh), and so there is no double relative here and the nár must be parsed as a conjunction. [The situation, as discussed several times above, where in a multiple relative sentence a later indirect relative is transposed to the ’ clause makes the first clause in such sentences appear dative and therefore indirect, but this does not violate the rule being presented here, as separation out of the nested relative clauses shows the first underlying clause to be direct.]

4. The example is given of cad é a mhinicí do fuaras é san áit nár mheasas a gheóbhainn é, there seems to be a problem that san áit a gheóbainn é has the direct relative, whereas as a separated clause this would be quite wrong, as san áit requires the indirect relative. If we said san áit do mheasas ná faighinn é, it would be quite normal, but when the negative relative is shifted by a process of inversion to the first relative clause, the direct relative of that clause is then shifted to the second clause.

Having noted that with a dó’ clause followed by a temporal or local clause with the indirect relative it is normal to transfer the indirect relative to the dó’ clause and follow with the direct relative, there are examples where the succeeding temporal or local clause retains the indirect relative too: chun gach tíre ’nar dhó’ leis go bhfaigheadh sí a bheag nú a mhór d’aon rud i bhfuirm nirt is an example of where chun gach tíre ba dhó’ leí becomes indirect relative, and yet we still have a go clause afterwards. Another way of parsing it is that the first clause is incomplete and the go is the conjunction to complete the sense.

In an rud gur mheasais gur rud fónta é, this cannot be a double relative, because either the first gur is a conjunction, in which case the second must be a conjunction too, or the first is genitive relative (short for gur mheasais ’na thaobh), which also precludes a double relative as the first clause in a double relative must be a direct relative (except for cases of inversion as mentioned above). To make this into double relative, you have to say an rud do mheasais ba rud fónta.

GÓN argues that use or non-use of the double relative makes for great subtley in PUL’s Irish. Fro example in bímíd go minic ag gáirí nuair ba cheart gur ag gol a bheimís, this is not double relative, and the meaning of nuair is rather “whereas”“we often laugh whereas in all reason we ought to weep”, creating a contrast betwen laughing and crying. If the stress were upon laughing at the very time we ought to be crying, then the double relative would be called for: bímíd go minic ag gáirí an uair ba cheart a bheimís ag gol.

To summarise, therefore, the double relative tends to occur after phrases such as is dó’ le and do mheasas, when they are themselves relative. There are 8 forms of the double relative sentence:

1. Where there is a direct relative in both clauses (i.e., the relatives are both nominative/accusative in nature). E.g. cé is dó’ leat do scríbh an leitir?

2. Where a comparative or superlative stands in a direct relative in the first clause, with an indirect relative in the second. E.g. níl éinne is feárr gurb eól do é ná mar is eól dómhsa é.

3. Where a direct relative in both clauses stands for the indirect. E.g. is dómhsa is feárr is eól cá luíonn an bhróg orm, where because the dative relationship has already been expressed before the comparative clause, both relative clauses are then direct.

4. Where there is a direct relative in the first clause and a direct relative standing for an indirect temporal in the second. Bímíd go minic ag gáirí an uair ba cheart a bheimís ag gol

5. Where there is a direct relative in the first clause and an indirect relative in the second in interrogative sentences. An example from the subsequent section on interrogatives in relative clauses is cad chuige aduairt sí ná raibh aon mhaith inti?

6. Where there is an indirect relative in the first clause and a direct relative in the second by a process of inversion. The underlying nested relative clauses have a direct relative in the first and a temporal indirect in the second. E.g. bhí gach éinne ag breithniú na haimsire ’nar dhó’ leó a bheadh an tÁrdrí ag teacht abhaile, where the indirect relative of the temporal clause na haimsire ’na mbeadh has been switched to the dó’ clause.

7. Where there is an indirect relative in the first clause and a direct relative in the second by a process of inversion. The underlying nested relative clauses have a direct relative in the first and a local indirect in the second. An example is cad é a mhinicí do fuaras é san áit nár mheasas a gheóbhainn é, where the underlying clause is san áit do mheas ná faighinn é.

8. Abnormal double relative clauses with indirect relatives in both clauses. E.g. chun gach tíre ’nar dhó’ leis go bhfaigheadh sí a bheag nú a mhór d’aon rud i bhfuirm nirt, where the indirect relative has been imposed on the dó’ clause via the process of inversion without replacing the subsequent indirect relative with a direct relative.

Negative relatives

, nách, and nár are negative relative pronouns as well as being conjunctions. GÓN arranges examples by case, showing these negative relatives are either direct or indirect:

a) Nominative (or genitive): ar ball do theastódh rud éigin uathu nárbh fhéidir a dh’fháil. Whether this is parsed as nominative or genitive depends on how a dh’ is parsed. If it is just the particle governing the verbal noun, once as a and repeated as do (and lenited), then the relationship is nominative. If a dh’ is understood as the possessive particle, padded out phonetically by dh’, then the relationship is genitive. [This ambiguity stems from the fact that the negative relative does not differentiate between the direct and indirect relatives. The affirmative form mian is ea é gur fuiriste a dh’fháil must be genitive relative following the indirect gur; the affirmative form mian is ea é dob fhuiriste a d’fháil must be nominative following the direct dob.] An example that is clearly nominative is sin rud ná raibh ann lem línnse.

b) Genitive. An unambiguous genitive is rud ab ea é nár ghá labhairt na thaobh.

c) Dative. Dheineadar roinnt cainnte, cainnt ná raibh puínn suime ag éinne acu inti.

d) Accusive (or genitive). Sarar deineadh an teitheadh sin bhí cruatán is brúth is feidhm ar Ultaibh ná féadfaidís a sheasamh puínn eile aimsire. In this sentence, the a could be interpreted as the particle governing the verbal noun (=do sheasamh). This would make the relative accusative. [GÓN states that in a sheasamh, seasamh would then stand in the dative. This seems a confusing point, but he seems to be referring to the fact that do is a dative preposition, and so when it is used as the particle governing the verbal noun, it puts the verbal noun “in the dative”. This may have etymological justification, but it seems easier for me to regard do or a governing the verbal noun as separate from the preposition do.] An alternative parsing would be to take cruatán is brúth is feidhm as a noun phrase that is considered as a whole and referred back to by a genitive a sheasamh. This would make the relative genitive and seasamh would be the accusative object of the verb féadfaidís.

An interesting example is this do measadh sinn a thabhairt ar an slógadh so le bréig, le geallúint nár measadh a chómhlíonadh. He argues that a could be the preposition do (i.e., the particle governing the verbal noun), and if measadh (the second measadh) in the sentence is autonomous n meaning, that would make the relative nár accusative in meaning. He adds that measadh could be seen as passive in meaning, which would make the relative nár nominative. By viewing measadh as either autonomous or passive, GÓN seems to be saying that the autonomous form can be either active or passive in meaning (“people thought” or “it was thought”). Yet PUL stated on a number of occasions that the autonomous form is always active, regardless of its translation into English. Another way of parsing the sentence is to have a as the possessive pronoun. Although bréig and geallúint are feminine, GÓN argues this need not prevent their being referred back to by a chómhlíonadh. This would make the relative nár genitive, and cómhlíonadh would be accusative (or nominative), rather than dative, governed by the verb measadh.

Some instances of the negative relative are virtually genitive. For example in níl bata sa tigh sin nárbh éigin seisreach do gabháil chun é thabhairt abhaile ón gcoíll, chun é thabhairt is equivalent to chun a thabhartha.

Comparative and Superlative adjectives.

Even when adverbial in meaning (GÓN calls them virtual adverbs), Irish comparatives and superlatives are adjectives in formal grammar and the double relative construction is frequently required to express what would in English be comparative and superlative adverbs.

No-one knows better than you how to do that” would become níl éinne is feárr gurb eól do conas é siúd a dhéanamh ná mar is eól duitse é, where feárr is to be parsed as an adjective, although equivalent in sense to an adverb or adverbial phrase. “The work he knows best” would therefore be an obair is feárr atá ar eólas aige.

Interrogatives and the relative

Interrogatives followed by a prepositional phrase, like cad na thaobh, cé dho and cé leis are elliptical. For example cad na thaobh ná rabhais anso inné? is equivalent to cad é an rud ná rabhais anso inné na thaobh, which makes the genitive relative. The subject of the unexpressed copula is an rud … na thaobh, the predicate is cad, and in the full form of the sentence é is the temporary subject.

Note how the prepositional phrase is normally placed next to the interrogative in the truncated sentence. An underlying cé hé an té go bhfuilir dhá thagairt sin do? therefore becomes cé dho go bhfuilir dhá thagairt sin? and cé hé an té gur leis an peann? becomes cé leis an peann?

These sentences require the indirect relative (cé dho go bhfuilir…), although GÓN adds that the Waterford form allows the direct relative, e.g. in sentences like cé leis atá sé ag cainnt?, but this is contrary to West Munster, Connaught and Ulster usage and, he argues, “a corruption of the true idiom”, which shoudl be cé leis go bhfuil sé ag cainnt?

A further reason to object to this “corruption” is that it destroys the distinction between single and double relative sentences, which depends on the distinction between the direct and indirect relative usages. Note for example the difference in nuance between cad chuige aduairt sí ná raibh aon mhaith inti? (a double relative sentence meaning “what did she say she was no good at?”) and cad chuige go nduairt sí ná raibh aon mhaith inti? (a single relative sentence meaning “why did she say she was no good?”) The latter sentence is an elliptical form of the full sentence, cad é an rud go nduairt sí ná raibh aon mhaith init chuige? As the double relative sentence requires the direct relative in the first clause (other than in cases of inversion), the difference in nuance is clearly seen.

Note the distinction between:

cathain aduairt sé a bheadh sé ann? (Double relative: “when did he say he would be there?”)

cathain aduairt sé go mbeadh sé ann? (Single relative: “when did he make the statement that he was going to be there?”)

and

cad na thaobh adeir sibh go mbíonn buaireamh oraibh? [“in connection with what do you say you are troubled?”]

cad na thaobh go ndeir sibh go mbíonn buaireamh oraibh? [“why do you say you are troubled”?]

Notes on Nolan 10

h) Proleptic de can be used to stand for whole clauses:

Níor mhiste deimhin a dhéanamh de go ndéanfaidís Gaeil Alban do dhísciú, where de tees up the subsequent go-clause.
Another example: ná dein iúnadh dhe go nduart leat: ní foláir sibh a bhreith an tarna huair.

Proleptic de is used with second comparatives. For example, is feárrde bia nú deoch é ‘ chaitheamh go réidh, where the -de anticipates é ‘ chaitheamh go réidh (food and drink are the better for consuming them slowly). Ní fheadar an feárrde iad ar thugas dóibh – note that the subject is iad, and the predicate is feárrde… ar thugas dóibh.

The -de may be retrospective where the explanatory clause precedes: dá mbeadh Méibh chómh honóireach leis dob fhusaide é.

These phrases are all of this type:
is feárrde me é – I am the better for it
is feárrde thu é – you are the better for it
The final é does not change – as it is anticipated by the -de.

Nolan adds that in the words móide and miste the -de appears to have become calcified and now lacks appreciable force. Sometimes you can expand the phrase to show the second comparative: ní miste dhuit dul abhaile láithreach is equivalent to ní measaide dhuit an scéal dul abhaile, where an scéal is the real subject and dul abhaile should be part of the predicate. However, Nolan concedes that in practical terms it is sufficient to take miste dhuit as the predicate and accept dul abhaile as the subject.

i) Acu
Acu is also proleptic in ceoca. Ceoca is feárr leat Gaelainn nú Béarla? Here the subject of the copula is (an ceann) is feárr leat, and ceoca is the predicate, and the acu included in that word stands proleptically for Gaelainn mú Béarla, which are also part of the predicate. The principal verb is not expressed, as the copula that appears is part of the subject phrase.

The English word “whether” is translated by ceoca when it introduces a noun clause (for example, where there is an alternative; where there is no alternative, the interrogative particle an is sufficient). But where “whether” introduces an adverbal clause, it is translated in Irish by peoca.

Ní fheadar ceoca ‘ thiocfaidh sé nú ná tiocfaidh.
Peoca ‘ thiocfaidh sé nú ná tiocfaidh fanfadsa.

j) Other prepositional pronouns
Air, leis, uime and others can all be proleptic.
Bhíos ag brath air go mbeithása anso rómham.
Ní raibh aon choine agam leis, & a fheabhas a thuig sé an obair, go dteipfeadh air mar do theip.
Is uime do thánag isteach san uair seo, chun t’onórasa.

k) Amhlaidh
Is amhlaidh a bhí náire air rómhamsa – we can see that the subject is not expressed, as the full sentence would be something like is amhlaidh mar a bhí an scéal aige bhí náire air rómhamsa.

Amhlaidh may also be retrospective where the explanatory phrase comes before: duart leis é ‘ dhéanamh go mear, & is amhlaidh do dhein.

Amhlaidh is followed by the direct relative, but the relative is logically superfluous, as shown by the full form of the sentence. Nolan points out that the direct relative particle has probably spread with proleptic amhlaidh from its use with retrospective amhlaidh, where it is needed.

Should Vocatives Decline?

Whether vocatives should decline or not is a constantly raised question. It seems most vocatives don’t decline, as “metaphorical” nouns do not take a morphological vocative. Let’s look at vocatives under a number of categories.

1. Collective nouns do not take the vocative.

This is clear from PUL’s letter to Gearóid Ó Nualláin about the occasion when he was confirmed at 13 and three congregations joined together for the ceremony, with the priest addressing them as a phobail!

PUL explained his confusion. Why was the priest saying, a phobail? You can only say a phobal! in Irish, as collective nouns do not take a vocative. This is logical as only a person or persons should ever take the vocative, and collective nouns do not have the same sense of personhood about the people being addressed. PUL assumed on that occasion that maybe because three congregations were together on one occasion, the priest was using the plural. However, Ó Nualláin pointed out this would still be wrong: pobail is nominative plural, not vocative plural, and the vocative corresponding to ‘ye congregations!’ is a phobla! /ə fobələ/. PUL also wrote in a letter to Pádraig Breathnach that the Irish speakers he knew would interpret a phobail! as some kind of name of a person or thing, as it couldn’t be a collective noun.

2. Inanimate objects do not take the vocative.

For example, in Sliabh na mBan bhFionn, we find a thromán! where the weight in a spindle is being called out to. The notes to this at the back of Ag Séideadh agus ag Ithe, which includes Sliabh na mBan bhFionn, indicate that inanimate nouns cannot be declined for the vocative case. However, note that where inanimate objects are addressed in terms that can apply to humans, you would decline the vocative. E.g. if you referred to the spindle weight as a mhic ó! Also names of inanimate things that are personified take the vocative, e.g. a bháis!

3. Usage in addressing animals.

T. F. O’Rahilly showed in an article in Ériu that usage was varied when addressing animals. E.g. PUL had a éan uasal! in his Aesop a Tháinig go hÉirinn, but also a mhada ruaidh!, with the adjective showing the latter was a declined vocative. Also, animals addressed in terms that could apply to humans, e.g. a mhic ó!, would receive declension in the vocative.

4. Metaphorical use of nouns referring to people.

The key problem is therefore metaphorical nouns referring to people. Standardized grammars state these cannot be declined, and it seems the earliest statement of this as a rule is T. F. O’Rahilly’s article, “The Vocative in Modern Irish”, in Ériu, volume 9 (1921-23), p85ff. He shows non-declension of 1st declension nouns in the vocative was common even in bardic verse, e.g. a bhéal cumhra, and so the failure to decline the metaphorical vocative of 1st declension nouns (the only declension that has the vocative singular) is of long standing. He quotes some examples of older poems where there was no decline vocative, but when Pádraig Ua Duinnín edited them, he put the vocative in (changing a stór to a stóir, for example). Clearly Ua Duinnín felt the non-use of the vocative was wrong, and yet by the same token, such usage goes back centuries.

So Ua Duinnín’s assertion in his dictionary that both a stór and a stóir were found in the vocative of that metaphorical noun needs to be seen in the light of this. Possibly he felt that was correct, but he had evidence of historical usage showing the contrary was normal. O’Rahilly also quotes PUL’s statement to Pádraig Breathnach that a stór! is the only correct vocative of that noun. PUL clearly did not believe that metaphorical vocatives declined.

O’Rahilly’s article therefore runs counter to the explanation given by Gearóid Ó Nualláin in his Studies in Modern Irish, Part 1. Ó Nualláin was a great scholar, and the first headmaster of the Munster training school in Ballingeary, but O’Rahilly was a great scholar too and so the issue needs to be examined on its merits.

In this regard, Gearóid Ó Nualláin’s comments in his Studies in Modern Irish Part 1, 219, are interesting. He puts forth the view here that the sex of the person affected the vocative declension. As it is first-declension nouns—masculine—that decline for the vocative, where these are used to refer to females, the tendency, or so he says, is for the vocative not to be declined. He gives an example of a shólás na ndobrónach!, referring to the Virgin Mary in the Litany in An Teagasg Críostaidhe. This is “sense construction”, he says, because she is feminine, and so a vocative a shóláis is not used. Another example given is a rún! referring to a woman, and not a rúin! Yet PUL told Pádraig Breathnach quite specifically that a stór! and a chumann! were the only correct forms, whether referring to a man or a woman. This seemed an ingenious argument by Ó Nualláin, but not ultimately the correct one.

Another argument put forward by Ó Nualláin is that where the vocative is part of a larger noun phrase, like a shólás na ndobrónach!, you could also parse this as the “Bracketed Construction”, where noun phrases lose their declension (just like hata fhear an tí, where fear an tí is taken as a phrase and the genitive drops out; GÓN shows that the Bracketed Construction is optional in the genitive, as is proved by PUL’s works, and hata fir an tí is also correct). However, PUL’s statement that a stór! is correct would mean it is not necessary to explain non-use of the vocative by the Bracketed Construction.

Looking at the Litany of Jesus in PUL’s An Teagasg Críostaidhe, we find these vocatives:

A Íosa, a Shaibhreas na bhfíoraon!

A Íosa, a Fhíorsholas!

A Íosa, a Sholas na gconfesóirí!

Correctly, therefore, PUL does not decline the metaphorical vocative. Interestingly, Shán Ó Cuív’s transcription in Leitiriú Símplí (LS) inserts the vocative in the first of these (as if from a shaibhris na bhfíoraon!), but he does so incorrectly.  Similarly, in the second example, a fhíorsholas! he transcribes as if from a fhíorsholais! This is also incorrect. Finally, the LS version correctly allowed the original to stand in its transcription of the third of these vocatives. It is possible he was influenced by Ó Nualláin’s views in his approach to these transcriptions.

In the Litany of the Blessed Virgin Mary, we find these vocatives:

A Scáthán an chirt!

A Shoitheach Sprideálta!

A Rós dhiamhair!

A Thúr Dáibhid!

A Árc na Connartha!

A Shólás na ndobrónach!

Some of these are phrase nouns, and all of them refer to a female, and this may be why in none of these cases did the LS edition try to reinsert the vocative, but it would have been incorrect to try to do so anyway.

While I am a fan of Ó Nualláin’s, I am forced to acccept that he came up with an ad hoc explanation to explain the lack of vocatives that he saw .

There are also some counter examples to deal with: PUL wrote a ghrá dhil in his Aithris ar Chríost. I am wondering if this  is equal to a ghrá ghil, but in any case the vocative of the adjective is declined.

5. Non-metaphorical vocatives of persons.

These should be decline in the 1st declension, but counterexamples can be found, e.g. a Árdaingeal in Soisgéal as Leabhar an Aifrinn, and a leanbh in Táin Bó Cuailnge. There is also a tendency for feminine nouns and adjectives to decline in the vocative under analogy with the 1st declension. Examples from PUL include a óinsigh!, a chailligh ruainnigh and a thoice bhig in Séadna; a chábóig gan chiall in An Bealach Buidhe; a chuil bhig in Bricriú; a spioraid shailigh, referring to the Devil in Soísgéal Naomhtha do réir Mharcuis (spioraid is only masculine in the phrase An Sprid Naomh); and a chroich shúigh “sooty old potrack” in An Craos-Deamhan.

Clearly, vocatives are hard to find in common use, as generally speaking there are only a few common ones, eg a fheara! (or a fhearaibh!). And where a vocative is called for, the nominative may be found therefore, as some researchers have claimed (see Gaeilge Chorca Dhuibhne) that few vocatives survive in Munster Irish.

Notes on Nolan 9

Chapter 2 – prolepsis

Prolepsis is the anticipation of what comes later in the sentence and is a significant feature of Irish grammar. Words than can be used proleptically include:

a) pronouns é, í, iad, ea. For example, the use of the subpredicate in copula sentences like is é ainm atá air ná Séadna is proleptic, as the é anticipates the name Séadna.

The pronoun can also function as a subsubject, anticipating the subject in sentences like cad é an rud é sin, where the subject is an rud (is) é sin.

Also note the function of é in sentences like: bhí áthas orthu é ‘ bheith le rá acu go bhfeacadar an rí.

b) sid é: sid é is mó a choisceann sólás ó Dhia ar theacht chút, a dheacracht leat iompáil chun úrnaithe.

c) : cuireann sé áthas orm tu ‘ bheith chómh maith is ‘taoi. Here is proleptic. This produces a by-form where there is no overt prolepsis: tá áthas orm tu ‘ bheith chómh maith is ‘taoi.

Note that ní raibh uain aige (rud a dhéanamh) is a by-form of the proleptic ní raibh sé d’uain aige (rud a dhéanamh).

Is air a bhí an iúnadh ná raibh Séadna a teacht would be a by-form of is air a bhí an iúnadh nuair a fuair sé ná raibh Séadna ag teacht.

d) so: cad a thug so dhómhsa máthair mo Thiarna do theacht ag triall orm?

e) san: so, sid é and é seo normally refers to what afterwards, with san, sin é and é sin referring to what went before. Usually therefore, san is not used proleptically, although it can be: ná cuirimís san de mhasla ar ár nglóíre go dtéifimís ón gcrois.

Similarly é sin can be used proleptically: cad é sin dúinne ceoca ‘táthar socair air nú ná fuiltear?

Note the use of proleptic é sin to anticipate a first or second-person pronoun: cad é sin dúinne thusa, a Íosa, a Mhic Dé?

GÓN argues that é sin has such a strong proleptic quality in such questions than even when it really refers to what has gone before an additional sin may be added at that end in proleptic reference: cad é sin d’éinne eile sin?

f) a: this is the key proleptic word. It is used before a verbal noun/noun to anticipate the object when that is a whole clause. For example: bhí sé tar éis a adhmáil do Shiobhán go raibh a croí dá shníomh le buairt. Also note how i dtaobh governs a rá in this sentenc: bhí buile ar Mhicil i dtaobh a rá go mbeadh sé de phláinéid ar Shéadna go dtug sé geallúint phósta do Shadhbh.

The proleptic a can be followed by a genitive in apposition, such as after a lán and a thuilleadh. Or followed by a partitive de, as in a mhalairt de, a bheag de, a mhór de.

Often the proleptic a is required where nothing of the kind appears in English: é dhá leogaint air gur theastaigh uaidh é féin do ghlanadh. However, bagairt, leogaint and aithint occasionally do not have the proleptic particle as expected: ní miste liom leogaint duit bheith ag imirt do chuid cleas ar dhuine éigin eile. After a vowel, it is frequently unclear if it is just a case of the proleptic a being swallowed up, as in dob fhuiriste aithint air go raibh súil aige.

A is also used proleptically with nouns of quality, quantity, time, etc. As in bhí iúnadh a gcroí orthu a fheabhas do dheineadar an gnó. This means “they were surprised at the excellency of it (the way in which) they accomplished the business”. GÓN notes in an aside that the direct relative is used after phrases like a fheabhas, although logically the indirect should have been expected (“the way in which”).

Note that the noun of quality, quantity, etc, would be followed by a noun in the nominative case: a luíghead airgead. The fact that the argument is in the nominative means that personal pronouns can be used as the arguments of the proleptic a.

Finally a is used proleptically before a noun of quality and condition to anticipate a clause with : bhí iúnadh orm, agus a fhuaire a bhí an aimsir, é ‘ bheith amu’ fén spéir in aon chor. Note here too that the a fhuaire is followed by the direct relative, where the indirect might have been expected.

The a can also take a partitive de, as in this example where the a anticipates a gur clause: a luíghead dá fhios a bhí ag an sagart bocht úd gurbh ar Bhríd naofa ba cheart do a bhaochas a bheith aige. The luíghead causes the use of the partitive de for the same reason that we say beagán aráin, but beagán den arán ab fheárr a bhí in Éirinn.

The abstract nouns governed by the proleptic a need to be studied: you can say a fheabhas, a fhuaire, a theó, whereas the nouns maithe, maitheas, fuacht and teas cannot be used in this way.

g) : meaning “however, in spite of”. GÓN states that this is an extension of the proleptic a, with the preposition de attached to it. He believes it originates from phrases like i gcath dá thruime (“in a battle of its seriousness”) and extended therefrom to dá thruime cath (“however serious a battle”).

Proleptic a always lenites, regardless of the gender and number of the noun it relates to.

has spread from strict to less strict relationships. In dá fhaid a ragham ar aghaidh is ea is giorra ‘ bheimíd don bhaile, the measurement is exact – the further we go, the nearer we will be.

In dá fhaid an lá is ea is giorra an oíche — this would be talking about the lengthening of daylight hours in the summer — the phrase appears to mean that the entire length of the daylight hours is exactly the same as length by which the night is shortened. However, it makes more sense to say that the increase in the length of the day is equivalent to the length by which the night is shortened. And in any case, no exactitude is really implied.

Finally, we come to usage that expresses not meant to be exactly equivalent or measurable. Dá fhaid a leogfar in aisce léi é is ea is dána ‘ leanfaidh sí dhe – here dá fhaid would be a time expression, measurable in days, whereas an increase in boldness cannot be measured in days. But it merely measn “the longer she is allowed to get away with it, the bolder she will remain”, with no real attempt at quantification.

Understanding the Copula

Understanding the Copula

The copula is one of the most awkward areas of Irish syntax to master. This is not helped by the fact that the promoters of an artificial ‘Standard Irish’ have muddied the waters with a confused explanation that in turn requires numerous exceptions to make it fit the language. Gearóid Ó Nualláin was a man with a forensic grammatical mind whose works on Irish grammar are the only detailed resource available, and his excellent presentation of the copula in his Studies in Modern Irish Part 1 repays careful reading, although it is undoubtedly not an easy read. Ó Nualláin’s understanding of the copula dovetailed with that expressed by Peadar Ua Laoghaire in Papers on Irish Idiom. I have also cast an eye over Ó Nualláin’s later New Era Grammar of Modern Irish, which gave more comments on the copula, and have updated this document where appropriate.

The most important point about the copula is that the function of the copula, as seen throughout the entire history of the Irish language, is to identify the predicate. It cannot stand next to the subject. This is not the explanation given in all textbooks, which often claim that copula constructions can be understood by a rule that claims the copula cannot stand next to definite nouns without the intervention of a pronoun, but Ó Nualláin shows this rule to be incorrect. It is true that in the copula of identification, a definite predicate does require a pronoun to be inserted between it and the copula, but as Ó Nualláin points out this results from the confusion of copula types and was not present in Old Irish, and there are many sentences where a definite noun does stand next to the copula. Another error in some textbooks is to claim there is a difference between the logical subject and predicate and the grammatical subject and predicate, but Ó Nualláin shows too that this is merely the importation of an anglicising influence on Irish, where the way things are said in English is held to exercise a dominant influence on the explanation of Irish grammar. These points will become clearer later in this essay.

The copula is always unstressed and the stress in copula sentences is always on the predicate, as it is the point of the copula to highlight the predicate. Consider this: if you ask an bhfuil sé anso? the answer can be . The substantive verb may answer the question unaided. But the copula cannot answer a question unaided, as it is always unstressed; it can only answer with the help of the (stressed) predicate. An leabhar é sin? The answer cannot be is; it can only be is ea, where the neuter pronoun is the predicate. A copula without a predicate is simply an impossibility in Irish.

The Copula of Classification

Let us start with the copula of classification. The basic form is Verb-Predicate-Subject (VPS), but Ó Nualláin is at pains to show that there is no “rule” that says this order is always followed, as there are examples of copula sentences where the subject comes first, for example, in rhetorical sentences (an example of which is given below). The important point is to understand that the copula points to the predicate, and so where the predicate does not directly follow the copula, a temporary predicate (subpredicate) is inserted to “hold” the place of the predicate. This then fulfils the rule that the copula points to the predicate.

Is ainmhí capall: “a horse is an animal”. Here ainmhí is the predicate and capall is the subject. “A horse” is the thing you are commenting on, and you are classifying it in terms of its being “an animal”.

You can make this emphatic this: ainmhí is ea capall. As the predicate ainmhí does not immediately follow the copula, a temporary predicate or subpredicate in the form of the pronoun ea is inserted to hold its place, and thus prevent the copula from standing next to the subject capall. This form then becomes PVpS, where the little p shows the subpredicate.

The forms that these sentences take in dependent clauses are worthy of notice. Is deimhin gur ainmhí capall, “it is certain that a horse is an animal”, is straightforward enough. But a double go is required for the emphatic forms in dependent sentences (the forms where ea holds the place of the predicate in classification sentences): is deimhin gur ainmhí gurb ea capall.

Further detailed points

The predicate is not always a noun. Is maith é sin has maith as an adjectival predicate. Is liomsa an leabhar san has the prepositional phrase liomsa as the predicate. Is maith an buachaill tu also has maith as the predicate, the subject being an buachaill tu—such subjects are elliptical, containing an implied relative clause (“the boy that you are”). Finally, the predicate in copula of classification sentences may be a name, as in Éamann a athair, an example chosen from Séadna. It seems to English speakers that Éamann should be an identification sentence, but it really is a name stated in a general sense, meaning “a person called Éamann”, and so it is a classificatory sentence in Irish. This shows the verb can be omitted; another example is Nasaret ainm an bhaile sin.

Most of these subforms can be emphasised with ea too (maith is ea é sin agus ní holc), but in some cases it would be odd to do so. Is breá an lá é is already emphatic enough, and does need further emphasis.

Irish grammarians are frequently concerned to impose English norms on Irish, and sometimes insist that in sentences like fear darbh ainm dò Seán Seán is the ‘logical predicate’, merely because in English “a man whose name was Sean” has Sean in the predicate. In the Irish sentence, Seán is the subject—the logical subject and the grammatical subject. Peadar Ua Laoghaire was concerned to point this out too, claiming that grammarians did not understand the grammar in the way that native speakers did: native speakers would always have felt that the copula pointed to the predicate, and the fact that the logic of English sentences was the other way round is really not relevant to a discussion of the Irish language. You can show the logic of the Irish sentence by translating in a more roundabout fashion: “a man for whom Sean was his name”. You can have Seán as the predicate by saying fear gur Seán ab ainm dò.

Often a long predicate has to be split up: fir ab ea iad ná leogfadh a gcroí ná a n-aigne dhóibh fanúint sa bhaile. Here fir is the fundamental noun of the predicate, ea is the subpredicate that holds the place after the copula, iad is the subject and ná leogfadh a gcroí etc is the rest of the predicate. A final form of the copula of classification is where the subject comes first and is recapped by a pronoun later on, often for rhetorical effect: an teagasc so a thugaimse ní liom é, where the subject is an teagasc so a thugaimse and é is the subsubject, or an additional recapping subject. In any case, the subject does not directly follow the copula.

The Copula of Identification

The general form of the copula of identification

The explanation of the copula, that in some sentences (particularly those with pronouns) the subject can go next to the copula, and that the main thing is that a definite noun cannot go next to the copula, has arisen because of the general form of the copula of identification: is é an saol so an t-earrach, “the spring is this world”, is an example from Jesus’ explanation of one of the parables. Here it is clear that in modern Irish the definite predicate in the copula of identification requires a pronoun to be inserted between it and the copula.

However, Ó Nualláin takes a historical view: in Old Irish the pronoun was not inserted in such circumstances, and the fact that a pronoun is now required in modern Irish in such sentences is the result of the confusion of copula types. For example, another type of copula of classification is sentences of the type is é ainm a bhí air ná Séadna: this form will be explained further below, but the proliferation of copula of classification forms where a pronoun is required has influenced the general form is é an saol so an t-earrach, such that the pronoun cannot be left out nowadays. In the interests of clarity, an t-earrach is the subject and an saol so is the predicate and é is the subpredicate. (This essay will end with a summary of the main reasons for stating that a definite noun can go next to the copula, with examples. I will first set out the various forms of the copula of identification.)

It is important to allow the logic of Irish grammar, which is different from English. Take the sentence sé an t-uabhar a thosnaigh an t-olc, “it was pride that started the evil”. “It” in English stands for the subject, but in Irish an t-uabharis the predicate and the é is the subpredicate, standing for the predicate. As mentioned above, the fact that the predicate is used with a subpredicate results from the confusion of copula forms centuries ago. The subject here is (an rud) a thosnaigh an t-olc. Consequently, in many ways Irish is more logical than English, and in translation the key thing is to identify what the predicate is. For example, in the sentence, “the man who stole the chalice was the one who found the key”, the information being imparted, in other words, the predicate identifying the subject, is “the one who found the key”, and so the sentence will be b’é an fear a fuair an eochair a ghuid an chailís. This more or less follows the English word order. But in “he thinks it is the men themselves who are responsible for this ugly custom”, the information being imparted, the predicate identifying the subject, is “the men themselves”, and so the translation is is dó’ leis gurb iad na fir féin fé ndeár an nós gránna san. The important thing is not the English word order, but the logic of the sentence. The subject is the noun that is being identified or defined; the predicate is the information that defines or identifies the subject, and it is this predicate that it is the function of the copula to highlight.

As the copula verb is often omitted, the general form of the copula of identification may be abbreviated as a PS sentence, where the predicate precedes the subject. As the subpredicate is only inserted to prevent the subject from standing next to the verb—and in the general form of the copula of identification this has been confused into a requirement for a subpredicate to prevent the predicate from standing next to the verb—once the verb is gone, the subpredicate is not needed either. An example of PS sentence is Gormfhlaith an chéad duine do bhuail uime, where Gormfhlaith is the predicate.

Finally, for rhetorical emphasis, it is possible for the subject to come first and be referred back to by a temporary subject pronoun or subsubject, otherwise keeping to the general form of the copula of identification. This would then be SVpPs: an t-arán a thabharfadsa uaim is é mo chuid feóla féin é chun beatha an domhain.

Copula forms where the real predicate comes later or earlier on

However, the basic VpPS form of the copula of identification is not possible in all circumstances, whether for rhetorical or other reasons, which is why the predicate often has to go later on or earlier on in the sentence. This is the reason why a subpredicate was inserted in the first place, to give VpSP (which subpredicate has then spread to VpPS sentences even where the predicate does come after the verb). Take a look at these examples where the predicate is thrown forward (I mean, to a later position) in the sentence:

Sé is mian leis an Eaglais fearg Dé do mhaolú: the (=is+é) contains a subpredicate so that the subject (an rud) is mian leis an Eaglais does not go next to the copula. The subpredicate anticipates the full predicate at the end, fearg Dé do mhaolú. The form is VpSP, which is why a subpredicate is now inserted in VpPS sentences such as is é an saol so an t-earrach.

B’é ab’fhada leis go raibh sé amu’: here also the é is a subpredicate holding the place next to the verb for the real predicate go raibh sé amu’. The subject is (an rud) ab fhada leis.

Sé mo thuairim ná tiocfaidh sé in aon chor: we can see clearly here that é does not refer to tuairim, which is a feminine noun, but is a subpredicate holding the place next to the verb for the real predicate ná tiocfaidh sé in aon chor The subject is mo thuairim.

Is é ainm a bhí air ná Séadna: some of these sentences where the predicate comes later have . Ó Nualláin comments in his New Era Grammar that this is more vivid and rhetorical than sentences with ná and should only be used where rhetoric is justified. É is the subpredicate and the real predicate is Séadna. Ó Nualláin argues that this has spread from the rhetoric use of ná go by analogy (ní deirim ná go bhfuil an ceart agat has helped to produce the affirmative form is é ’deirim ná go bhfuil an ceart agat); I am simplifying his argument, but the precise analogical derivation of is not too important to this presentation of the copula. Peadar Ua Laoghaire was emphatic that the definite article could not be used with the subject in sentences of the type, giving as an example in Papers on Irish Idiom, is é duine a bhí ann ná Tadhg. The reason why the definite article cannot be used with duine here is because “in these construction the word isé fully defines duine … hence the definite article cannot be used in the Irish”. (Ua Laoghaire also states that he prefers not to use the article with the definite predicate in the general form of the copula of identification in sentences such as dob é céad duine (or an chéad duine) a tháinig é, but he accepts that in such circumstances either way was commonly found.)

B’iad beirt iad san ná Maolmhórdha agus Sitric: this example is analogous to the one above, but whereas ainm a bhí air contained an explicit relative clause, many other such sentences only have an implied relative. Iad here is the subpredicate, beirt iad san is the subject with an implied relative beirt is iad san (“the pair that they are”) and Maolmhórdha agus Sitric is the real predicate. This shows that the sentences are generally affirmative sentences containing a relative clause in the subject.

Sé cuis ná héisteann sibhse le briathraibh Dé mar ní hó Dhia sibh: sentences with mar are similar to the sentences. The mar is logically superfluous. É once again does not refer to the feminine noun cúis as it is the subpredicate referring to the real predicate mar ní hó Dhia sibh, and preventing the subject (cúis ná héisteann sibhse le briathraibh Dé) from standing next to the verb.

Is é an namhaid an peaca: because the normal form of the copula of identification (is é an saol so an t-earrach) is of the form VpPS, and because the predicate is sometimes thrown forward in the sentence (producing VpSP), it sometimes requires contextual evidence to work out which order the predicate and subject are in. So in this sentence, it could either be a VpPS statement (“I am speaking of sin, and sin is the enemy”), or a VpSP statement, where the predicate is thrown forward, meaning “I am speaking of the enemy, and the enemy is sin”, which is what the context in the sermon of Peadar Ua Laoghaire’s that Ó Nualláin takes this phrase from shows it to mean. This is a more rhetorical form. In fact, while its existence is frequently found in Irish literature, Gearóid Ó Nualláin claims to be the first to have noticed it. The fact that is é an namhaid an peaca can mean two different things is shown by the two intonation patterns. In VpPS, an namhaid receives the stress, as the predicate is always stressed. In the rhetorical form VpSP there is pause after is é and another pause after an namhaid and then the emphasis is on an peaca, showing to any hearer that an peaca is the predicate.

Naturally, once the verb is omitted, VpSP sentences, where the predicate is thrown forward, simply become SP sentences, with no need for any subpredicate. An example is tír gan teanga tír gan anam: it is clear from the meaning of the sentence that tír gan anam is the predicate and not the subject, although contextual information is always required to determine whether such sentences are PS or SP. (Such sentences may seem to be copulas of classification, but tír gan teanga is not any individual land, but a class or kind of land—the identification of classes is an identification copula, as shown in sentences like is é rud é ná athrú ana-mhór, where the copula of identification is used to identify something here “as the kind of thing known as a great change”.)

Finally, in some copula of identification sentences, the predicate comes first for rhetorical effect or emphasis. This means that a subpredicate must be inserted after the verb. Tosach an uilc is é is usa do chosc: here the predicate is tosach an uilc and the subpredicate é prevents the copula from standing next to the subject (an chuid den olc) is usa do chosc. Another example would be clann na ríthe agus na n-uasal is iad a thagadh. These sentences normally require the fundamental noun of the subject to be understood: in this case the subject is (na daoine) a thagadh, and the predicate is clann na ríthe agus na n-uasal.

In all the sentence types in this section, the predicate is either positioned later in the sentence or earlier (or even in a preceding sentence), requiring a subpredicate to be inserted—and it is from these sentences and the habit of using the subpredicate with many types of copula of identification that the subpredicate spread even to the general form of the copula of identification where the predicate is in its normal place, which is why is é an saol so an t-earrach requires a subpredicate today (and not because the predicate is a definite noun).

Pronoun predicates and subjects

This brings us to one of the great misunderstandings of the copula: it is claimed in some accounts that first- and second-person pronouns even if they are the subjects must stand next to the copula. Examples include (is) mise Gearóid Ó Nualláin. Some grammarians have insisted that mise is the logical subject here purely because “I” stands in the subject in English. However, both Gearóid Ó Nualláin and Peadar Ua Laoghaire were emphatic that in Irish mise is the predicate here and that there should be no attempt to foist English grammar onto Irish. Peadar Ua Laoghaire wrote in The Leader, Febuary 7th 1903 (an article later printed in Papers on Irish Idiom):

In such Irish sentences as Is mise an rí, Is tusa fear an tí seo, Isé Tadhg an cléireach, etc., the words mise, tusa, the é of Isé are the words which express the information intended to be conveyed. To say that mise, tusa, é, or any word following is in such constructions, could be the subject of the sentence, is to state what is impossible from the very nature of is.

“That is all very well when Is mise an rí is an answer to the question, Cé hé an rí? but suppose it is used as an answer to the question, Cé hé thusa? how will the case stand? e.g., Q. Cé hé thusa? A. Mise an rí. Is not mise the subject here, and is not an rí the information given?”

In answering a question a person is at liberty to omit all except the information, i.e., the thing which is not in the question, but which the question asks for. Hence, in the example given, to the question Cé hé thusa? the answer can be simply An rí. Now, from the very essence of is in the sentence Is mise an rí, the information is in the word mise. But what sort of answer would the word Mise be to the question, Cé hé thusa?

The truth is this. The sentence Is mise an rí is not an answer to the question Cé hé thusa? It is an answer to the question Cé hé an rí? The true answer to the question Cé hé thusa? is An rí mise, or An rí. The only thing that can be said in order to justify the use of Is mise an rí as an answer to the question Cé hé thusa? is, that it conveys indirectly the information asked for, by answering, not the question Cé hé thusa? but the question Cé hé an rí.

“Oh, I have, over and over again, heard native Irish speakers say, Is mise fear an tí seo as Irish for ‘I am the man of this house,’ Is mise Tomás for ‘I am Thomas,’ etc., where you have the subject immediately after Is, just as in English.”

I can only say that, if you imagine that mise is the subject in the examples you have given, you are hopelessly wrong. In each of these examples mise contains the information. It is the English parallel that blinds you. In English the information may be put before or after the verb to be. In Irish it is essential that it should come immediately after is. The moment you appeal to English parallels I have no more to say.

This category of sentences comprises those with predicates including mise, tusa, sinne and sibhse; third-person pronouns fortified by a demonstrative (é seo, iad san, í siúd, etc); and pronouns accompanied by féin. Examples include an tu san?, “is that you?” Deir sé gurb é sin Pádraig Ó Cealla, and b’é féin árdollamh Uladh. In all these examples, the pronoun appears in the predicate.

Ó Nualláin points out that where it is definitely intended that the pronoun be the subject, it is possible for these pronouns to appear correctly as the subject of the copula (with VpPS being the syntax, in line with the general form of the copula of identification). Examples include Maois agus Elias iad san; mo Dhia thu, mo chuid ’en tsaol thu; sí cainnt an tSlánaitheóra féin í sin. The quotation from Peadar Ua Laoghaire shows that an rí mise is perfectly good Irish. Consequently, it is incorrect to state that in sentences like (is) mise Gearóid Ó Nualláin the subject is adjacent to the copula (whether expressed or omitted).

Standard accounts of the copula with the first- and second-person pronouns tend to make a distinction between sentences with unstressed pronouns (me, tu), which it is argued are therefore the subjects of the copula, and sentences with stressed pronouns (mise, tusa), which it is argued are therefore the predicates, as the predicate does, after all, take the stress. For example, in section 16.35 of the Irish edition of the Christian Brothers’ Grammar there is table showing that (is) mise an dochtúir and an sibhse na gardaí nua? are examples of sentences where the pronoun is the predicate. The very next table show that in an tu an dochtúir? the pronoun is the subject. Section 16.40 shows that an tusa Séamas? has the pronoun as the predicate, whereas an tu Séamas? has the pronoun as the subject. This therefore has the pronoun always next to the copula, with the emphasis or lack of it on the pronoun, as shown by a suffix, determining whether it is the subject or the predicate of the sentence. Yet it is then stated in smaller font at the end of section 16.40 that the pronoun does in fact come last in phrases like mo cheól thu and mo ghrá í sin.

Clearly, most presentations of the copula are beset with exceptions; it is unsurprising that Irish schoolchildren find the copula hard to learn when the grammar books find it difficult to explain the “rules” themselves. It seems likely that there is a genuine preference in Irish for the first- and second-person pronouns to come first in copula sentences, although counter-examples exist too. The grammatical explanations then rush to keep up with the linguistic reality, often only doing so with a morass of exceptions to the “rules”. One possibility is that sentences wheremise and tusa come first are actually, or can actually be, SP sentences, as it is generally the case with the copula of identification that the order of the subject and predicate in the sentence needs to be determined; this would only apply in cases where the verb were not given, to prevent the pronoun from standing directly next to the copula. Ó Nualláin accordingly has mise cailín an Tiarna as an SP sentence in his New Era Grammar, despite its being exactly the same sort of sentence as mise Gearóid Ó Nualláin parsed as a PS sentence in Studies in Modern Irish Part 1. In sentences where the verb is left out, it is therefore down to context to work out which is the subject and which the predicate, but in any case, a general tendency for the first- and second-person pronouns to go first is seen. As the verb is omitted, in this particular case, mise cailín an Tiarna could be parsed either way.

However, in Peadar Ua Laoghaire’s translation of the Gospels, we find is me Mac Dé: the verb is not omitted, and according to the theory that the copula cannot stand next to the subject, me must be the predicate, although it is not stressed (it does not appear as mise). Consequently, the consideration that some of these sentences might be SP sentences does not fully resolve the problem of copula use with personal pronouns, and the issue of the emphatic form of personal pronouns used as the predicate does not apply here either—in any case, the use of the emphatic pronouns to determined which is the subject and which the predicate seems a red herring, as shown in the sentence offered by Peadar Ua Laoghaire above, an rí mise, and as shown  by mise cailín an Tiarna if that is parsed as an SP sentence. In the final analysis, it is hard to deny that Irish has a preference for first-and second-person pronouns to be used in the predicate. Why should it generally be the case that Irish prefers to have the first- and second-person pronouns as the predicate and not the subject, contrary to usage in English? My conclusion is that in a dialogue between two people, one of the parties is “I” and the other is “you”. The two parties to the conversation have identified each other as real, flesh-and-blood human beings. Consequently, it is illogical of the English language to “identify” one of the interlocutors by a name or other definite noun. Such a name or definite noun is just a label and does not serve to “identify” the person being spoken to any more concretely than the fact that he already appears as a concrete, real human being does. It makes more sense to identify the name (essentially a label) as applying to a concrete person present in the conversation than it does to identify a concrete person present in the conversation by a label. Consequently, there is nothing illogical in the fact that Irish prefers first- and second-person pronouns to be predicates—the real human beings present in the conversation who provide the embodiment of the identification for the labels, but which cannot be identified as labels themselves, because they are already present and seen for what they are, as real human beings.

Ó Nualláin does not give any examples where a first-person pronoun is used as the subject of the copula in Irish in a PS sentence, however, and it is likely that such usage is rare; the only example I have is an rí mise, given by Peadar Ua Laoghaire above. It is less rare, but still rare, for second-person pronouns and pronouns with demonstratives to stand as the subject in PS sentences in Irish, but clearly they may do so. Mo ghrá thu would be one well-known example that appears to violate the “rules” imposed by the Standard grammars. Here thu is the subject and mo ghrá is the predicate. Another good example is Peadar Ua Laoghaire’s translation of Jesus’ words at the last supper: sé mo chorp é seo. Ó Nualláin states that all other translations of this phrase are wrong, include the is é seo mo chorp that is used in An Bíobla Naofa (a translation done after Ó Nualláin’s death, but in any case including the same incorrect translation that Ó Nualláin rails against as an error; it is curious that Pádraig Ó Fiannachta’s grasp of the copula was insufficient to prevent this anomaly from creeping into an important sentence in the Bible). The form used in An Bíobla Naofa is theologically untenable, as it has é seo as the predicate and mo chorp as the subject, and therefore implies “my body is here and nowhere else”, whereas Ó Nualláin points out that Christ’s Body is in heaven and also in every other consecrated particle in the world, and so cannot be in the one piece of bread alone. Is é mo chorp é seo merely states that “this is my body”, without insinuating anything about The truth is th/st/pemand rong/strongis. The sentence its not being anywhere else also.

The made-up rule about definite nouns not standing next to the copula

Having examined all the forms of the copula of identification we are now in a position to say that there is no such rule preventing definite nouns from standing next to the copula. The reason a pronoun is inserted in the copula is to prevent the subject from standing next to the copula. This is seem in the copula of classification (ainmhí is ea capall) and also in the copula of identification (sé mo thuairim ná tiocfaidh sé in aon chor); the reason why a pronoun is now required to prevent the predicate from standing next to the copula of identification is confusion between the forms centuries ago.

This is not just a semantic argument, or a different way of presenting the copula. There are many examples where a definite noun does stand next to the copula without a pronoun. These include:

a) where a definite noun is used as an adverbial phrase: is dócha gur an fhaid a bhí an dealús air a dhein sé é. Here an fhaid is definite, and stands directly adjacent to the copula without a pronoun between them. Another adverbs are also definite, e.g. is anois é, where anois is definite, and yet is not separated from the copula by a pronoun.

b) definite prepositional phrases: is i dTeamhair a bhíodar an uair sin. Here i dTeamhair is definite, but stands next to the copula without a pronoun.

c) relative clauses where the copula stands next to a definite noun: gurb é Íosa is Críost ann. Here Críost stand next to the copula without a pronoun. The reason is that the subject governing the copula here is the (unexpressed) relative particle: gurb é Íosa (a) is Críost ann. Because the relative particle is nominative to the verb here, the copula is not followed by the subject, but by the predicate, and so no subpredicate is required. Where the relative particle expresses any other case relationship, then a subpredicate is required. For example, in ar chuma nách é gach éinne a thuigeann, “in a way that not everyone understands”, the negative relative is in an accusative relationship, gach éinne is the subject and so cannot stand next to the copula without a subpredicate pronoun. In an té gurb é a dhúil bheith in’ aonar, the relative is in a genitive relationship (expressing “whose”) and a pronoun is required. In chómh fada agus is é do leas é the relative particle is in a dative relationship, as is required in temporal clauses (“the length of time in which”), and so a pronoun is required. In conas mar is é úr leas é the relative particle is also in a dative relationship, as required in modal clauses (“the way in which”), and so a pronoun is required.

The only time a pronoun is not required in such relative clauses is where the relative particle governs the following copula as its subject, in which case the subject does not succeed the copula and so no subpredicate is required to keep the subject separate from the copula. Ó Nualláin quotes one sentence from Peadar Ua Laoghaire’s Sgothbhualadh where a negative relative clause governs a succeeding copula as its subject and yet a pronoun is inserted—bhí a lán nithe nárbh é an lá ar áilleacht againn—but adds that this reflects the influence of the common phrases rud nách é and rud nárbh é.

Finally Ó Nualláin points out that in questions like cad is ainm duit? where it seems as if the subject ainm duit comes next to the verb, the sentence is to be understood as elliptical. The subject is really (an ainm a is) ainm duit and the predicate cad?, and consequently the subject of the is that is expressed in the sentence is the unexpressed relative particle in the elliptical phrase.

Where a pronoun intervenes in questions like cé hé an fear é sin?, is the predicate and the é is a temporary subject standing for the real subject, which follows, an féar é sin.

Notes on Nolan 8

Questions like an é Tomás Ó Ceallaigh do bhí ann? follow the ordinary syntax of the copula. However, what? and who? questions form a general exception to the syntax of the copula. This is because there is no real predication: the question asks for the predicate, but doesn’t really give one. The interrogative takes the place of the predicate, and is placed first in the sentence, and the copula is generally not expressed.

In some questions, like cad é an rud é sin? general syntactical rules relating to the copula are not violated. But where there is a separate copula that follows the question word, it is not the main verb at all. For example cad is ainm duit? This appears to have the subject ainm duit directly next to the copula, but it is an elliptical construction, containing an unexpressed relative.  Cad? is the predicate and (an ainm a is) ainm duit is the subject, but because there is really a relative clause in there, the subject of the expressed is is the implied relative particle and the predicate of the expressed is is ainm duit.

Notes on Nolan 7

p41

Nolan points out that copulas of identification an classification often come together: deir cuid acu gurb é Ieremias é nú duine de sna fáidhibh.

The reason why names are often indefinite in copula sentences is because Pádraigjust means a person bearing the name Pádraig. So deir sé gur Pádraig é siúd leis is telling you something general – it is a person called Pádraig. The same can be said about deir sé gur Pádraig is ainm do. Whereas ‘sé Pádraig a bhí ann means the man himself – not a person bearing the name – but the real man – and so is definite, a true proper name. In ‘sé ainm atá air ná Pádraig the term is more definite.

Other similar examples include is é rud é ná athrú ana-mhór and is athrú ana-mhór é. The second is indefinite – the matter is “a great change”. The first is identifying the class of thing something belongs to – and identifying it as the class of “great changes”, the kind of thing that is called a great change.

There is no rule in Irish requiring definite nouns to be separated from the copula by a personal pronoun. That Type I now requires a pronoun is purely in order to assimilate type I to types II, III and IV – not to separate the copula from a definite noun. The real rule is the subject cannot come immediately after is. Apart from the confusion of types of copula of identification, another proof that the pronoun is required to prevent the subject from coming after the copula is the fact that even in copulas of classification a pronoun is inserted to perform the same function, as in ainmhí is ea capall.

Other examples of how definite nouns can come after the copula include the use of an fhaid and similar adverbial phrases.  Is dócha gur an fhaid a bhí an dealús air a dhein sé é. And other examples with an iomad, anois, inniu can be found: is anois é. (The reason you have to say b’é faid an turais a chuir tuirse orm is precisely because Type 1 copulas of identification have been assimilated to the pattern of Types 2, 3 and 4.)

Also definite prepositional phrases do not require a pronoun and can stand next to the copula, adding fuel to GÓN’s point of view: is i dTeamhair a bhíodar an uair sin. Finally, in relative clauses such as gurb é Íosa is Críost ann, if it were the case that definite nouns could not stand next to the copula, then Críost would require a pronoun before it, but it doesn’t, as it is predicate here, the subject being the unexpressed relative particle that governs is.

Nolan adds that it is only where the relative is the subject of the copula in such sentences that no pronoun is required; where the relative is genitive, dative or accusative the pronoun is inserted, showing that the real point of the copula is not to prevent definite nouns from standing next to the verb, but to keep the subject from doing so. An example of a genitive relative is níl éinne ó bhaol ag teacht os cómhair daoine ach an té gurb é a dhúil bheith in’ aonar, where thre relative means “whose” here. An example of a dative relative is an fhaid is é grásta Dé atá dhá iompar, where the temporal clause requires a dative relationship (“the length of time in which…”). In conas mar is é úr leas é, the relative is also dative in a modal clause (“the way in which”). An example of an accusative relative is ar chuma nách é gach éinne a thuigeann, “in a way that not everyone understands”, where the relative is governed by thuigeann.

GÓN then points out that even when the relative is subject to the copula, the pronoun is occasionally expressed when the relative clause is negative, as in bhí a lán nithe nárbh é an lá ar áilleacht againn from PUL’s Sgothbhualadh. This can be understood as reflecting the influence of phrases such as rud nách é and rud nárbh é, the peculiarity here being that after the é the predicate is then is then repeated.

Notes on Nolan 6

p29

Identification copulas, Type IV, PVpS: these have the predicate brought forward for emphasis or rhetorical effect. In this case, the temporary predicate pronoun is retrospective rather than anticipatory. Tosach an uilc is é is usa do chosc. Another example from PUL’s Aesop is An bás a cheapas don éan is é is trúig bháis dom féin and from PUL’s sermons, an t-uabhar is é ‘ chuireann duine ag formad lena chómharsain. You often need to supply an implied an rud or na daoine before the relative clause of the subject phrase. The pronoun often assimilates to the subject, as in clann na ríthe agus na n-uasal is iad a thagadh. The pronoun may also refer back to a material predicate in a preceding sentence.

Type V, PS: the predicate and the subject are juxtaposed with no verb.  For example tosach an uilc is usa do chosc and Gormfhlaith an chéad duine do bhuail uime.

Type VI, SP: the subject and predicate are juxtaposed with no verb. This is an abbreviated form of Type II, easily confused with a copula of classification sentence. An example is Eagla Dé túis na heagna. This could be a Type V sentence meaning “to be truly wise we must fear God”, but in other contexts it would be an attempt to explain what the fear of God is — which makes it Type VI, equivalent to is é rud eagla Dé ná túis na heagna. Tír gan teanga tír gan anam – this sentence clearly defines tír gan teanga and so is an SP sentence, Type VI. It can’t be a copula of classification, because you are not talking about any individual country, but about a class of country that has no language, so this is not classification of an individual, but the identification of classes. It stands for ‘sé rud tír gan teanga ná tír gan anam.

Type VII, VPS: this is where the predicate is a 1st or 2nd person pronoun, a 3rd person pronoun with a demonstrative, or any pronoun and féinMise Gearóid Ó Nualláin. Deir sé gurb é sin Pádraig Ó Cealla. An tu san? B’é féin árdollamh Uladh. GÓN takes strong issue with the view that mise or tusa must be the logical subject in such sentences. Take for example, the question cé thusa? It is argued by some that in the reply mise an bás mise must be subject. However, GÓN cannot agree, unless these sentences be construed as examples of type VI, and in any case where the verb is expressed before mise, then it must be the logical predicate, because the function of the copula in Irish is to identify the predicate. The comparison with what would be the subject in an equivalent English sentence is not a valid argument for him, as it only serves to anglicise Irish. Also, has type VIII as a type of sentence where mise is the logical subject, and so is making an distinction here.

Type VIII: VpPS: this type differs from Type I only in that the subject is a pronoun of the 1st or 2nd person or a 3rd person pronoun strengthened by a demonstrative. This shows that the idea that these cannot be subjects is false, and also shows that misewas the predicate in Type VII sentences. These pronouns can be subjects if we definitely want them to be the subjects in our minds.

‘Sí cainnt an tSlánaitheóra féin í sin: the writer clearly wants cainnt an tSlánaitheóra féin to be the predicate in this sentence.

‘Sé mo chorp é seo: in the words of the consecration, which GÓN holds are usually incorrectly translated, it has to be phrased like this, with é seo as the subject and mo chorp as the predicate. To say is é seo mo chorpsa -leaving aside the fact that GÓN says the emphatic suffix is out of place – would mean “my body is here and nowhere else”, which would be theologically unsupportable.

Finally an example where thu is the subject is from PUL’s translation of Imitatio Christi: féach, mo Dhia thu, mo chuid ‘en tsaol thu. So the theory that first and second person pronouns must come after the copula (albeit omitted here) is false.

Type IX, VpSP: this type has a proleptic pronoun ea, unlike types II and II that had é, í and iad. In this type ea anticipates a definite predicate, but it is only found in poetry, and GÓN’s examples caome from old works, including Keating. One example is ‘sea ‘duairt sí – éist liom go fóill. It seems that this this is equivalent to is é rud aduairt sí – éist liom go fóill.

Type X, SVpPs: here the real subject comes first, and a pronoun refers back to it at the end. This is usually a case of a clunky subject clause. An t-arán a thabharfadsa uiam is é mo chuid feóla féin é chun beatha an domhain.

Notes on Nolan 5

p27

Identification Type III, VpSP: this differs from Type II (where the subject contains an expressed or implied relative clause or genitive that can be resolved into one) by having  a definite noun as the subject, possibly with a demonstrative particle or an adjective in tow.

An example is ‘sé an namhaid an peaca. In context, this may be a Type I identification copula (“sin is the enemy (of man)”), remembering that Type I is VpPS. But in the context that Gerald Nolan takes this example from (Sermon 238 of PUL’s Seanmóin is Trí Fichid), namhaid is the subject, making it Type III. That this is so is clear from the context, as an namhaid has been previously mentioned.

Nolan points out that Irish literature is full of Type III identification copulas, but these had not been noticed in Irish grammars before Nolan pointed them out. This usage is more rhetorical. Type III can be regarded as an abbreviated form of Type II, in other words is é an namhaid an peaca means the same as is é rud an namhaid ná an peaca. Nolan argues that Type I is distinguished from Type III by the intonation: in Type I is é an namhaid an peaca (“sin is the enemy”), the sentence is enunciated quickly with no pause, and the greatest stress is on the word namhaid. In Type III is é an namhaid an peaca (“the enemy is sin”) the enunciation is slower, there is a pause after é and another pause after namhaid and the greatest emphasis is on the word peaca. 

Nolan rejects the analysis of other grammarians distinguishing between logical and grammatical predicates, who would argue that the logical predicate in these sentences is peaca but the grammatical predicate is namhaid. He then quotes a sentence from Geoffrey Keating ‘sé an ceárd úd an nádúir dhaonna, which he says is a Type III sentence, and explains the fact that the predicate pronoun is é, despite referring to an nádúir by the fact that the proleptic pronoun is frequently assimilated in gender to the subject where its gender is different from that of the predicate.

Finally, he gives an example of Type III that mixes in features of Type II (ie, includes the particle ): is é an cosc go léir is an cosc is mó orainn ná deinimíd aon iarracht ar dhul ar bhóthar fíréantachta na naomh.