RRL, LXIII, 3, p. 297–313, Bucureşti, 2018
AND AGREEMENT IN ROMANIAN
1BLANCA CROITOR
2Abstract. Clefting and pseudoclefting are focalizing strategies, in which a constituent is focused by rephrasing a simpler sentence. In contemporary Romanian, only pseudocleft sentences are used. They are a subtype of specificational sentences. In this article we describe a few properties of pseudoclefts and specificational sentences in Romanian, from a contrastive perspective. We present the patterns of agreement in specificational, as well as in other types of copular sentences and present several previous analyses of agreement in specificational sentences. Then we propose our own hypothesis, which takes into account the richness of feature marking of the two DPs involved in a copular sentence.
Keywords: pseudoclefts, specificational sentences, agreement, focalizing strategies, copular sentences.
1. PRELIMINARY REMARKS
This article is based on the observation that in Romanian pseudocleft structures the verb agrees with the postcopular element, unlike in English:
(1) Ceea ce mă atrage la această facultate sunt / what CL.1SG.ACC attract.PRES.3SG to this faculty be.PRES.3PL *este profesorii.
be.PRES.3SG teacher.PL.DEF
‘What attracts me to this faculty is the teachers.’
Pseudocleft structures are a subtype of specificational sentences (see section 2.
below). The same type of agreement patterns are found in other types of specificational sentences in Romanian, unlike in English (see the translation):
1This research is supported by the project PN-II-RU-TE-2014-4-0372, Structura informaţională şi ordinea cuvintelor în propoziţie în limba română, financed by the Ministry of Education (UEFISCDI) from Romania. I would like to thank Ion Giurgea for his comments and suggestions on this article.
2 “Iorgu Iordan – Al. Rosetti” Institute of Linguistics, Bucharest, [email protected].
(2) Cel mai bun lucru la această facultate sunt /
DEF.M.SG more good thing about this faculty be.PRES.3PL
*este profesorii.
be.PRES.3SG teacher.PL.DEF
‘The best thing about this faculty is the teachers.’
Agreement with the postcopular element involves the person feature as well (again, unlike in English):
(3) a. Cel care a pierdut eşti tu.
DEF.M.SG which AUX.PERF.3SG lose.PPLE be.PRES.2SG you
‘The one who lost is you.’
b. Vinovatul eşti tu.
guilty.DEF be.PRES.2SG you
‘The culprit is you.’
We will provide an explanation for the agreement patterns above, taking into account several factors: semantic or structural factors, information structure, feature hierarchy. First, we will take into account the properties of pseudoclefts and specificational sentences in Romanian (sections 2 and 3), then we detail some of the constraints on the focalized element in Romanian pseudocleft structures, since this is a subject very poorly discussed in Romanian literature on pseudoclefts; in section 5 we present the patterns of agreement in pseudoclefts and in specificational sentences, in Romanian and in other languages; in section 6 we present some of the previous analyses of agreement and propose a new hypothesis.
2. ON PSEUDOCLEFT STRUCTURES IN ROMANIAN
Cleft sentences are complex sentences that put a constituent into focus and have the meaning of a simple sentence. In English, the main types of cleft structures are it clefts, wh- clefts and reversed wh- clefts, exemplified in (4)a–c, where the subject is focalized; in (5) the focalized element is the object, in the same three types of structures:
(4) a. It was a red car that was blocking her driveway.
b. What was blocking her driveway was a red car.
c. A red car is what was blocking her driveway.
(5) a. It is roses that she loves.
b. What she loves is roses.
c. Roses is what she loves.
Wh- clefts are also called pseudoclefts. Romanian does not employ it clefts3 (see also
3 As Romanian does not have an it pronoun (which could act as an expletive) or a presentational pronoun (such as there in English), the sentence corresponding to (4)a begins with copula followed by the DP, prosodically marked as a focalized element. The structure is not ungrammatical, but it does not sound natural and it is not used:
(1) A fost o maŞInă roşie ceea ce i-a
AUX.PERF.3SG be.PPLE a car red what CL.DAT.3SG=AUX.PERF.3SG blocat aleea.
block.PPLE driveway.DEF
‘It was a red car that was blocking her driveway.’
Şerbănescu 1996, Gheorghe 2013); it has wh- clefts (pseudoclefts) and reversed wh- clefts (also called reversed pseudoclefts)4. The precopular element can be of two types:
(i) formed with relative pronouns; (ii) formed with an overt head noun. In the first type, we find relative pronouns such as: ce, ceea ce (both are uninflected and have neuter, singular forms) ‘what’, cine (uninflected) ‘who’, or the definite determiner cel followed by relative clauses with ce and care – see below, (6)c, d. When cel is followed by ce, it has only masculine forms, singular and plural; when it is followed by care, it has forms for feminine as well, singular and plural. In Romanian, cel is used as a definite determiner in contexts in which N is not expressed: cel verde ‘the green one’, cel din stânga ‘the one from the left’5. The unexpressed N can be anaphorically linked, but not necessarily (see Giurgea 2010 for more details).
(6) a. Ce / ceea ce nu îmi place este what what not CL.DAT.1SG like.PRES.3SG be.PRES.3SG zgomotul de la restaurantul acesta.
noise.DEF from restaurant.DEF this
‘What I don’t like is the noise from this restaurant.’
b. Cine a venit ultimul a fost George.
who AUX.PERF come.PPLE last.SG.DEF AUX.PERF.SG be.PPLE George
‘The last one who came was George.’
c. Cei ce au venit ultimii au
DEF.M.PL what AUX.PERF.3PL come.PPLE last.PL.DEF AUX.PERF.PL fost profesorii.
be.PPLE teacher.PL.DEF
‘The last ones who arrived were the teachers.’
d. Cea care a plecat ultima
DEF.F.SG which AUX.PERF.3SG leave.PPLE last.F.SG.DEF
a fost Maria.
AUX.PERF.3SG be.PPLE Maria
‘The last one who left was Maria.’
The second type of pseudocleft, with an overt noun followed by a relative clause, is exemplified bellow (for English, see Akmajian 1970b, Mieszek 1974).
(7) a. Lucrul care nu îmi place este zgomotul thing.DEF that not CL.DAT.1SG like.PRES.3SG be.PRES.3SG noise.DEF
de la restaurant.
from restaurant
‘The thing I don’t like is the noise from the restaurant.’
4 According to Gheorghe 2017, in old Romanian the patterns were more varied, with respect to the structures involved and the focalizing strategies. The author concludes, based on the absence of basic pseudo-clefts in the corpus olf old Romanian texts, that the pattern from modern Romanian may be due to an external influence, from French or Italian and it is relatively recent. For old Romanian, see also Pană Dindelegan 2015.
5 The relative ceea ce is composed of the former distal demonstrative ceea (feminine, singular), literarily meaning ‘that’, and the relative ce ‘what’; ceea is related to the feminine form of cel (cea) and it could be a determiner itself. In GALR (2008, I: 282) it is considered a compound relative pronoun.
b. Femeia pe care a ales-o George woman.DEF DOM which AUX.PERF choose.PPLE=CL.ACC.F.3SG George este Maria.
be.PRES.3SG Maria
‘The woman George chose was Maria.’
c. Copiii care au venit ultimii sunt children.DEF which AUX.PERF come.PPLE last.PL be.PRES.3PL George şi Daniel.
George and Daniel
‘The children who arrived last were George and Daniel.’
It is generally accepted in the literature that pseudoclefts are a subtype of specificational sentences. Under older assumptions (see Akmajian 1970a, b), a structure like the one in (8) had two readings, specificational and predicational:
(8) What John is is silly. (Iatridou & Varlokosta 1998: 3)
A. specificational reading: ‘John is silly’ (a property is predicated of John)
B. predicational reading: ‘The fact that John is X (= a spokesman/POTUS/an ice cream truck driver) is silly’ (a property is predicated of a property of John)
The same two readings are available for similar Romanian structures, such as the one in (10):
(9) Ceea ce spune Ion e o prostie.
what say.PRES.3SG John be.PRES.3SG a stupidity
‘What John says is a stupidity.’
A. specificational reading: ‘[Of all the things that he could say,] John says the following:
something stupid.’
B. predicational reading: ‘John says something that, in my opinion, is stupid.’
More recently, structures like the one in (8) and (9) are considered pseudoclefts only if they have specificational reading, because only under this interpretation they are focalizing strategies.
3. ON SPECIFICATIONAL SENTENCES IN ROMANIAN
Specificational sentences are a subtype of copular sentences, with the structure NP1 + be + NP2, in which the prenominal noun (N1) is semantically a variable and the postcopular noun (N2) specifies the value of this variable. Therefore, N1 has a more general meaning, while N2 specifies what N1 refers to6.
6 In the unmarked word order, the noun with the more general meaning (the variable) is first. If the word order is reversed, the first DP is focalized:
(1) Lipsa BAnilor a fost problema.
lack.DEF money.GEN AUX.PERF be.PPLE problem.DEF
‘The lack of money was the problem.’
(10) a. Vinovatul eşti tu.
culprit.DEF be.PRES.2SG you.NOM
‘The culprit is you.’
b. Problema sunt copiii.
problem be.PRES.3PL child.PL.DEF
‘The problem is you.’
c. Candidatul republican este Donald Trump.
Candidate.DEF republican be.PRES.3SG Donald Trump.
‘The republican candidate is Donald Trump.’
According to the typologies that have been proposed, the other types of copular sentences with the structure NP1 + be + NP2 are: predicative, equative / equational, and identificational (Higgins 1973, Declerck 1988, den Dikken 2006, Heycock 2012, Roy 2013). In predicative sentences, the postcopular phrase contributes a predicate, and the subject (S) is its argument (the sentence denotes the inclusion of S in a class; a property of S): Mary is a lawyer., Daniel is curious. In equative or equational sentences, the prenominal and the postcopular phrases are of the same semantic type, namely referential (entity type): Mark Twain is Samuel Langhorne Clemens. In identificational sentences, the subject position is a deictic DP – a demonstrative, the pronoun it in English (a DP whose reference is obtained deictically, in the context of utterance): This is me., It is my new car.
The specificational sentence in which the first position (the precopular position) is occupied by a relative clause or cel + a relative clause (see above, (6)) is a pseudocleft structure.
4. CONSTRAINTS ON THE FOCALIZED ELEMENT IN ROMANIAN PSEUDOCLEFT STRUCTURES
As we have mentioned, clefting and pseudoclefting are focalizing strategies, in which a constituent is focused by rephrasing a simpler sentence. In Romanian, the target of the pseudocleft construction can be the subject (as in (11)), the direct object (as in (12)), and some prepositional objects (see (13, 14)):
(11) a. Dan a plecat la Cluj.
Dan AUX.PERF.3SG leave.PPLE to Cluj ‘Dan went to Cluj.’
b. Cel care a plecat la Cluj DEF.M.SG which AUX.PERF.3SG leave.PPLE to Cluj a fost Dan.
AUX.PERF.3SG be.PPLE Dan ‘The one that left to Cluj was Dan.’
(12) a. George a cumpărat un trandafir.
George AUX.PERF.3SG buy.PPLE a rose ‘George bought a rose.’
b. Ceea ce a cumpărat George a fost what AUX.PERF.3SG buy.PPLE George AUX.PERF.3SG be.PPLE
un trandafir.
a rose
‘What George bought was a rose.’
(13) a. Alex se bazează pe prietenii lui.
Alex CL.REF.3SG count.PRES.3SG on friend.PL.DEF his.GEN
‘Alex counts on his friends.’
b. Cei pe care se bazează Alex sunt DEF.M.PL on which CL.REF.3SG count.PRES.3SG Alex be.PRES.3PL prietenii lui.
friend.PL.DEF his.GEN
c. Pe cine se bazează Alex sunt prietenii lui.
on who CL.REF.3SG count.PRES.3SG Alexbe.PRES.3PL friend.PL.DEF his.GEN ‘The ones Alex counts on are his friends.’
(14) a. Maria are încredere în proiectul ei.
Maria have.PRES.3SG confidence in project.DEF her ‘Maria has confidence in her project.’
b. În ceea ce are încredere Maria este proiectul ei.
in what have.PRES.3SG confidence Maria be.PRES.3SG project.DEF her ‘What Mary has confidence in is her project.’
The pseudocleft structures in which the target is the indirect object (in the dative case) are restricted. The structures with the relative pronoun cine ‘who’, which would have the dative form cui, are ungrammatical. In addition, the structures with cel/cea care are restricted to animate nouns:
(15) a. Ion i-a telefonat profesorului.
John CL.DAT.3SG=AUX.PERF.3SG phone.PPLE teacher.DEF.DAT
‘John phoned the teacher’
b. Cel căruia i-a telefonat Ion DEF.M.SG which.DAT CL.DAT.3SG=AUX.PERF.3SG phone.PPLE John
a fost profesorul.
AUX.PERF.3SG be.PPLE teacher.M.SG.NOM.DEF ‘The one John phoned was the teacher.’
c. *Cui i-a telefonat Ion a
who.DAT CL.DAT.3SG=AUX.PERF.3SG phone.PPLE John AUX.PERF.3SG fost profesorul.
be.PPLE teacher.M.SG.NOM.DEF
‘The one John phoned was the teacher.’
(16) Ion a trimis guvernului o petiţie.
John AUX.PERF.3SG send.PPLE government.DEF.DAT a petition ‘John sent a petition to the government’
> a. *Cui i-a trimis Ion o petiţie who.DAT CL.DAT.3SG=AUX.PERF.3SG send.PPLE John a petition a fost guvernul.
AUX.PERF.3SG be.PPLE government.DEF
b. *Cel căruia i-a trimis Ion DEF.M.SG which.DATCL.DAT.3SG=AUX.PERF.3SG send.PPLE John o petiţie a fost guvernul.
a petition AUX.PERF.3SG be.PPLE government.DEF ‘The one John sent a petition to was the government.’
The pseudocleft structures in which the target is the adjective are restricted. Usually, the sentences are ungrammatical (see 17b)7, but if the adjective is accompanied by adjuncts that circumscribe the property it denotes (as in (18) and (19)), the pseudocleft structures become acceptable (more or less):
(17) a. Dan este inteligent.
Dan be.PRES.3SG intelligent.
‘Dan is intelligent.’
b. *Ceea ce este Dan este inteligent.
what be.PRES.3SG Dan be.PRES.3SG intelligent.
‘What Dan is is intelligent.’
(18) a. Dan poate fi naiv, dar este cu siguranţă inteligent.
Dan may be.INF naive but be.PRES.3SG with certainty intelligent ‘Dan may be naive, but he surely is intelligent.’
b. ?Ceea ce este Dan cu siguranţă este inteligent.
what be.PRES.3SG Dan with certainty be.PRES.3SG intelligent.
‘What Dan surely is is intelligent.’
(19) a. Dan este înainte de toate şiret.
Dab be.PRES.3SG before of all sly
‘Dan is first of all sly.’
b. %Ceea ce este Dan înainte de toate este şiret.
what be.PRES.3SG Dan before of all be.PRES.3SG sly
‘What Dan is first of all is sly.’
Spatial and temporal adjuncts can also be focalized by pseudoclefts, in some contexts; noun-headed relatives are preferred:
(20) a. Acum plouă la Constanţa.
now rain.PRES.3SG at Constanţa.
‘Now it is raining in Constanţa.’
b. Unde plouă acum este la Constanţa.
where rain.PRES.3SG now be.PRES.3SG at Constanţa.
‘It is in Constanţa that it’s raining now.’
(21) a. Mă întâlnesc cu Maria pe 10 ianuarie.
CL.ACC.1SG meet.PRES.1SG with Maria on 10 January ‘I will meet Maria on January 10.’
7 Compare with English:
(1) What Dan is is intelligent.
b. Ziua în care mă întâlnesc cu Maria este day.DEF in which CL.ACC.1SG meet.PRES.1SG with Maria is 10 January
10 ianuarie.
‘The day I meet Maria is January 10.’
c. *Când mă întâlnesc cu Maria este pe when CL.ACC.1SG meet.PRES.1SG with Maria be.PRES.3SG on 10 ianuarie.
10 January.
‘*When I meet Maria is January 10.’
In English, spatial and temporal adjuncts can be focalized in pseudoclefts with bound and free relative clauses:
(22) a. The place where I saw John was Boston. (Akmajian 1970b: (36)c) b. The time at which I met John was 4 o’clock. (Akmajian 1970b: (36)d) (23) a. Where I saw John was in Boston. (Akmajian 1970b: (37)c)
b. When I met John was at 4 o’clock. (Akmajian 1970b: (37)d)
In Romanian, modal adjuncts require an initial clause headed by a noun, but the acceptability of the structure depends on the type of modal adjunct:
(24) a. Maria citeşte tare poezia.
Maria read.PRES.3SG loudly poem.DEF ‘Maria reads the poem loudly.’
b. ?Modul cum / Felul în care citeşte Maria poezia way.DEF how / type.DEF in which read.PRES.3SG Mary poem.DEF
este tare.
be.PRES.3SG loud
‘The way Mary reads the poem is loud.’
c. *Cum citeşte Maria poezia este tare.
how read.PRES.3SG Mary poem.DEF be.PRES.3SG loud ‘How Mary reads the poem is loud.’
(25) a. Ion a plecat pe jos.
John AUX.PERF.3SG leave.PPLE on down ‘John left by walking’
b. *Modul / felul în care a plecat Ion manner.DEF way.DEF in which AUX.PERF.3SG leave.PPLE John
este pe jos.
be.PRES.3SG on down
‘*The way John left is by walking.’
c. *Cum a plecat Ion este pe jos.
how AUX.PERF.3SG leave.PPLE John be.PRES.3SG on down ‘*How John left is by walking.’
In some contexts, genitives can be the target of a pseudocleft structure, unlike in English (see Lees 1963: 378, via Mieszek 1974: 230):
(26) a. A lătrat câinele lui George.
AUX.PERF.3SG bark.PPLE dog.DEF LUIGEN George ‘George’s dog barked’
> b. *Al cui câine a lătrat este George.
AL who.GEN dog AUX.PERF.3SG bark.PPLE be.PRES.3SG George ‘*Whose dog barked is George’s.’
(27) I am afraid of Peter’s dog.
> a. *What dog I am afraid of is Peter’s.
b. *Whose dog I am afraid of is Peter’s.
c. *What I am afraid Peter’s of is dog. (Leese 1963: 378)
Some prenominal determiners prevent cleaving of the nominals they precede. The postcopular DP in pseudoclefts and specificational sentences must be referential, entity- denoting. Similar restrictions are found in English, as shown in the translations below (see also Lees 1963: 380, via Mieszek 232):
(28) a. Niciun student nu a venit. > *Cel care / *cine no student not AUX.PERF.3SG come.PPLE DEF.M.SG which who a venit (nu) a fost niciun student.
AUX.PERF.3SG come.PPLE not AUX.PERF.3SG be.PPLE no student ‘No student came.’, ‘*The one who came was no student.’
b. Fiecare vecin i-a ajutat. > *Cel each neighbour CL.3SG.ACC= AUX.PERF.3SG help.PPLE DEF.M.SG care / *cine i- a ajutat a which who CL.3SG.ACC=AUX.PERF.3SG help.PPLE AUX.PERF.3SG fost fiecare vecin.
be.PPLE each neighbour
‘Each neighbour helped them.’, ‘*The one who helped them was each neighbour.’
c. Ambii vecini i-au ajutat. > *Cei care both neighbour.PL CL.3PL.ACC=AUX.PERF.3PL help.PPLE DEF.M.SG which i-au ajutat au fost ambii vecini.
CL.3PL.ACC=AUX.PERF.3PL help.PPLE AUX.PERF.3PL be.PPLE both neighbour.PL ‘Both neighbours helped them.’, ‘The ones that helped them are both neighbours.’
d. Aceşti vecini i-au ajutat. > Cei care these neighbour.PL CL.3PL.ACC=AUX.PERF.3PL help.PPLE DEF.M.SG which i-au ajutat au fost aceşti
CL.3PL.ACC=AUX.PERF.3PL help.PPLE AUX.PERF.3PL be.PPLE these vecini.
neighbour.PL
‘These neighbours helped them.’ ‘The ones who helped them were these neighbours.’
In some sentences, the predicate nouns can be the target of focus in Romanian, unlike adjectives (see above). The acceptability of the sentences is higher if the two copulas are distant or if different copular verbs are used (compare a and b below):
(29) a. *Ce este Ion este arhitect.
what be.PRES.3SG John be.PRES.3SG architect ‘What John is is an architect.’
b. Ce a devenit el după câştigarea alegerilor
what AUX.PERF.3SG become.PPLE he after winning.DEF election.DEF.PL.GEN
e un monstru.
be.PRES.3SG a monster
‘What he became after winning the elections is a monster.’
In English, only predicate adjectives can be targeted by a pseudocleft sentence (see Iatridou and Varlokosta 1998), not predicate nouns (see Lees 1963: 380, via Mieszek 1974:
232):
(30) a. John is silly. > What John is is silly.
b. She is a teacher. > *What she is is a teacher.
When the focalized element is a 1st or 2nd person pronoun, the reflexive clitic in the relative sentence can be bound by it, unlike in English (see (32)). The sentence in (31)c is more colloquial (a lower registry), but still acceptable.
(31) a. Eu mă bărbieresc cu lama.
I CL.REFL.1SG shave.PRES.1SG with blade.DEF
‘I shave myself with a blade.’
b. Cel care se bărbiereşte cu lama
DEF.M.SG which CL.REFL.3SG shave.PRES.3SG with blade.DEF
sunt eu.
be.PRES.1SG I.NOM
‘The one who shaves himself with a blade is me.’
c. Cel care mă bărbieresc cu lama
DEF.M.SG who CL.ACC.1SG shave.PRES.1SG with blade.DEF sunt eu.
be.PRES.1SG I.NOM
‘*The one who shaves myself with a blade is me.’
(32) a. The one who shaves himself with a straight razor is not me. (Akmajian 1970b: (14)a) b. *The one that cut myself is me.
In reversed pseudocleft structures, the reflexive clitic can be bound by the 1st or 2nd person pronoun in English, as well as in Romanian:
(33) a. I am the one who cut myself. (Akmajian 1970b: (18)a)
b. Eu sunt cel care mă bărbieresc I be.PRES.1SG DEF.M.SG which CL.ACC.1SG shave.PRES.1SG
cu lama. – colloquial with blade.DEF
‘I am the one who shaves himself with a razor.’
b’. Eu sunt cel care se bărbiereşte I be.PRES.1SG DEF.M.SG which CL.ACC.3SG shave.PRES.3SG
cu lama. – literary language with blade.DEF
5. PATTERNS OF AGREEMENT
As we have mentioned, in Romanian specificational sentences the copular verb agrees with the postcopular nominal, in number and person. This type of agreement is found in pseudocleft structures as well:
(34) a. Ei se iubesc, problema sunt / they CL.REFL.3love.PRES.3PL problem.DEF be.PRES.3PL
*este părinţii ei.
be.PRES.3SG parents.DEF her
‘They love each other, the problem is her parents.’
b. Aeroportul din Arad în pericol: Motivul sunt taxele airport.DEF from Arad in danger reason.DEF be.PRES.3PL tax.PL.DEF suplimentare. (www.arq.ro)
supplementary.PL
‘The airport from Arad is in danger: The reason is the supplementary taxes.’
c. Principalul dezavantaj al ţărilor vest-europene main.DEF disadvantage GEN country.PL.GEN west=European.PL
rămân costurile ridicate ale forţei de muncă.
remain.PRES.3PL cost.PL.DEF high.PL GEN force.DEF.GEN of work
‘The main disadvantage of western European countries is the high costs of the workforce.’ (TV news, apud Croitor 2012)
(35) a. Ceea ce îmi place la Barcelona sunt / what CL.DAT.1SG like.PRES.3SG about Barcelona be.PRES.3PL
*este clădirile şi cafenelele.
be.PRES.3SG building.PL.DEF and coffee-shop.PL.DEF
‘What I like about Barcelona is the buildings and the coffee-shops.’
b. Ceea ce mi se pare a fi o problemă what CL.DAT.1SG CL.REFL seem.PRES.3SG to be a problem
sunt concedierile abuzive.
be.PRES.3SG firing.PL.DEF abusive.PL
‘What seems to me like a problem is abusive firings.’
In the examples above, singular agreement would be ungrammatical.
(36) a. Vinovatul sunt eu / *este eu.
culprit.DEF be.PRES.1SG I.NOM be.PRES.3SG I.NOM
‘The culprit is me.’
b. Problema eşti tu / *este tu.
problem.DEF be.PRES.2G you.SG.NOM be.PRES.3SG you.SG.NOM
‘The problem is you.’
c. Viaţa mea sunteţi / *este voi.
life.DEF my be.PRES.2PL be.PRES.3SG you.PL.DEF
‘My life is you.’
d. Sunt câteva lucruri pe care le ador în be.PRES.3PL some thing.PL DOM which CL.3PL adore.PRES.1SG in viaţa asta. Unul dintre ele eşti tu.
life.DEF this one of them be.PRES.2SG you.SG (http://circulmeu.ro)
‘There are a few things I adore in this life. One of them is you.’
(37) a. Vinovaţii suntem / *sunt noi. (3pl. & 1pl.) culprit.PL.DEF be.PRES.1PL be.PRES.3PL we
‘The culprits are us.’
b. Problemele lui Dan sunteţi / *sunt voi. (3pl. & 2pl.) problem.PL.DEF GEN Dan be.PRES.2PL be.PRES.3PL you.PL
‘Dan’s problems are you.’
(38) a. Cel pe care îl aşteptai sunt /
DEF.M.SG DOM which CL.ACC.M.SG wait.IMPERF.2SG be.PRES.1SG
*este eu.
be.PRES.3SG I.NOM
‘The one you were waiting for is me’
b. Cea care ştie totul eşti / *este that.F.SG which know.PRES.3SG everything be.PRES.2SG be.PRES.3SG tu.
you.NOM
‘The one who knows everything is you’
(39) a. Eu sunt / *este cel pe care îl I be.PRES.1SG be.PRES.3SG DEF.M.SG DOM which CL.ACC.M.SG
aşteptai.
wait.IMPERF.2SG
‘I am the one you were waiting for’
b. Tu eşti / *este cel care ştie totul.
you be.PRES.2SG be.PRES.3SG DEF.M.SG which know.PRES.3SG everything
‘You are the one who knows everything’
Italian, European Portuguese and to some extent Brazilian Portuguese, and Spanish behave like Romanian with respect to agreement (Moro 1991, 1997, 2000, 2006; Costa 2004; Heycock 2012):
(40) a. Il colpevolo sono io / *è io / *è me. (Italian)
‘The culprit is me.’
b. La causa della rivolta sono / *è alcune foto del muro.
‘The cause of the riot is some pictures of the wall.’ (Moro 2006)
(41) O assassino sou eu. (Portuguese)
DEF assassin be.PRES.1SG I
‘The assassin is me.’
In English, the agreement patterns are opposite, with singular agreement, 3rd person agreement:
(42) a. The real problem here is me / *am me / *am I / *is I.
b. The culprit is me / *am me / *am I / *is I. (cf. Heycock 2009)
(43) The cause of the riot is / *are some pictures of the wall. (cf. Heycock 2009) (44) a. What attracts people to this store is the roots. The CEO and the owners of the
store have very impressive backgrounds in the fashion world. (www.cervera.com) b. What could ensure a humiliating loss for Mr. Trump in November are his troubles with constituencies that have favored Republicans in recent elections.
(http://www.nytimes.com)8 There are some languages where both nominals can be sources of agreement, such as Faroese (analysed by Heycock 2009):
(45) Orsøkin til eldin var/ vóru tey brennandi kertiljósini í stovuni.
cause.DEF to fire.DEF was/ were the burning candles.DEF in room.DEF
‘The cause of the fire was the burning candles in the living room’ (Heycock 2009)
6. HYPOTHESES REGARDING AGREEMENT IN SPECIFICATIONAL (& PSEUDOCLEFT) SENTENCES
6.1. Previous analyses
In order to analyse which factors determine agreement in specificational sentences (including pseudoclefts), we need to understand the structure of the copular sentence.
Following Stowell’s 1978 proposal, it is accepted that the copular sentence has the underlying structure in (46), in which the two nouns form a small clause, with a subject and a predicate semantic role:
(46) be + SC [Subject Predicate] (see Stowell 1978) a. English: be [you problem]
b. Romanian: a fi [tu problema]
In a canonical predicative sentence, the subject of the small clause is raised in precopular position, as in (47) and (48):
(47) a. Tu eşti problema.
you be.PRES.2SG problem.DEF
‘You are the problem.’
8 Some native speakers consider the plural to be incorrect.
b. Eu sunt vinovatul.
I be.PRES.1SG culprit.DEF
‘I am the culprit.’
(48) a. You are the problem.
b. I am the culprit.
In a specificational sentence, also called inverse copular sentence, in Moro 1991, 1997, 2000), the predicate of the small clause is raised in precopular position, as in (49) and (50):
(49) a. Vinovatul sunt eu.
culprit.DEF be.PRES.1SG I.NOM
‘The culprit is me.’
b. Problema eşti tu.
problem.DEF be.PRES.2SG you
‘The problem is you.’
(50) a. The culprit is me.
b. The problem is you.
If we understand the underlying structure of the copular sentence, we can propose an explanation pf agreement in specificational sentences that takes into account the function of the two DPs from the underlying small clause. From a semantic point of view, the first DP has an attributive/predicative value, while the postcopular DP is the subject. Therefore, is it only ‘natural’ that the copula agrees with the second DP. Under this hypothesis, the agreement patterns we find in Romanian (and other languages) are typical subject – verb agreement instances. The fact that English does not have this type of agreement can be analysed in relation to the rigid word order from this language. In the DP be DP structure, the first DP is the grammatical subject, because it is in the canonical subject position, SpecTP (distinct from the subject of the underlying small clause).
A second hypothesis would take into account the information structure of the specificational sentence (see also GALR 2008, II: 210–211). From an informational point of view, the first noun is the theme (in GALR, the theme corresponds to the topic), while the second noun is the rheme, which means that the first noun is the subject, while the second noun is the predicate. Therefore, the agreement with the first noun would be a canonical instance of verb – subject agreement (as in English, for instance), while agreement with the postcopular noun is not (other factors have to be taken into account to explain the pattern of agreement found in Romanian).
A third hypothesis, the topicalisation hypothesis, was proposed by Heycock (2009).
The first noun is not raised (from the small clause) to subject position, but to a topic position, above InflP. The first noun is outside the domain of agreement and inaccessible for agreement, therefore the copular verb has to look for another noun, which is what happens in Romanian, Italian and other languages where we have agreement with the postcopular verb etc. In English, the first noun triggers agreement when it passes through the subject position (Spec,InflP). Then it raises further, to Spec,TopicP. The contrast between Italian, Romanian and English is due to the different moments when agreement is triggered (later in Romanian, Italian etc., after the first noun reached its topic position).
6.2. Our hypothesis
Our hypothesis takes into account the morphological marking of the features involved in agreement and is evidenced by equative and identificational sentences. In these two sentences, if one of the two DPs is plural, the verb is always plural. If one of the two DPs is 1st or 2nd person, it triggers agreement on the verb.
(51) Părinţii ei sunt / *este lumina . parent.PL.DEF her.GEN be.PRES.3PL be.PRES.3SG light.DEF sufletului său
soul.DEF.GEN his
‘Her parents are the light of her soul.’
(52) a. Asta sunt / *este eu.
this be.PRES.1SG be.PRES.3SG I.NOM
‘This is me.’
b. This is me. / *This am me. / *This am I. / *This is I.
c. Acea femeie eram / *era eu.
that woman be.IMPERF.1SG be.IMPERF.3SG I.NOM
‘That woman was me.’
d. That woman is me. / *That woman am me. / *That woman am I. / *That woman is I.
The patterns of agreement in these copular sentences points to a hierarchy of features, described in (53):
(53) Hierarchy of features:
Person: 1 > 2 > 3 Number: pl. > sg.
1st or 2nd person (singular or plural) > plural (3rd person) > singular (3rd person) / default Based of these pattern of agreement, we propose the more general hypothesis that in Romanian copular sentences, agreement is triggered by the nominal with most prominent or powerful number & person features (according to the hierarchy in (53)). Both DPs are equally accessible sources of agreement (given that word order is not rigid) and the copula agrees with the nominal whose features are more prominent or more marked (see also Croitor 2012). Thefore, the semantic or syntactic structure of the copular sentence does not necessarily play a role in Romanian. The hypothesis is stated in (54):
(54) In Romanian copular sentences, the verb agrees with the noun that is more marked for number and person.
7. FINAL REMARKS
In this article we presented the properties of Romanian pseudocleft structures, which are very little discussed in Romanian literature on copular sentences. We showed that they are a subtype of specificational sentences (displaying similar semantic structure). In pseudoclefts and in specificational sentences, verbal agreement is in the plural in
Romanian, if the postcopular nouns is plural, unlike in English, where the verb agrees with the first noun, in the singular. If the postcopular DP is a first or second peson pronoun, the verb agrees with it in Romanian, while in English the verb would be in the third person. In Romanian, similar agreement patterns are found in equative and identificational copular sentences as well. We examined previous analyses of agreement in specificational sentences, which take into account, for instance, semantic factors (the precopular noun is the predicate from the small clause underlying the copular sentence, while the postcopular noun is its subject) or rigid word order (which explains agreement with the first DP in English). We proposed our own hypothesis, according to which the agreement in copular sentences in Romanian (and other languages which display the same patterns) is with the most marked noun / DP. Most marked is to be understood as higher on the hierarchy of person and number: first person and second person are higher than third person, plural is more marked than singular. Under this hypothesis, semantic or structural factors are less important than feature hierarchy in copular sentences. This helps to explain not only agreement in specificational sentences (including pseudoclefts), but also in equative and identificational sentences.
REFERENCES
Akmajian, A. 1970a, Aspects of the Grammar of Focus in English, doctoral thesis, MIT.
Akmajian, A., 1970b, “On Deriving Cleft Sentences from Pseudo-Cleft Sentences”, Linguistic Inquiry, 1, 2, 149−168.
Akmajian. A., 1979, Aspects of the Grammar of Focus in English, New York: Garland.
Costa, J., 2004, Subject Positions and Interfaces: The Case of European Portuguese, Berlin, Mouton de Gruyter.
Croitor, B., 2012, Acordul în limba română, Bucureşti, Editura Universităţii din Bucureşti.
Declerck, R., 1988, Studies on Copular Sentences, Clefts, and Pseudo-clefts, Cornell University Press.
Den Dikken, M., 2006, “Specificational Copular Sentences and Pseudoclefts”, in: M. Everaert, H. van Riemsdijk (eds.), The Blackwell Companion to Syntax, Blackwell Publishing, 292–409.
GALR – V. Guţu Romalo (ed.), Gramatica limbii române, I–II, Bucureşti, Editura Academiei Române, 2005, 2008.
Gheorghe, M., 2013, “Pseudo-cleft constructions” in: G. Pană Dindelegan (ed.), The Grammar of Romanian, Oxford, Oxford University Press, 488–489.
Gheorghe, M., 2017, “(Pseudo)cleft constructions in old Romanian”, Diacronia, 6 (www.diacronia.ro).
Giurgea, I., 2010, Pronoms, déterminants et ellipse nominale, Bucureşti, Editura Universităţii din Bucureşti.
Heycock, C., 2009, “Agreement in specificational sentences in Faroese”, in: P. Svenonius, K. Bentzen, C. Heycock, J. í Lon Jacobsen, J. Bondi Johannessen, J. K. Parrott, T. Strahan, and Ø. A. Vangsnes (eds), Tromsø Working Papers on Language & Linguistics: Nordlyd 36, special NORMS issue on Faroese, CASTL, University of Tromsø, 1–22.
Heycock, C., 2012, “Specification, equation, and agreement in copular sentences”, Canadian Journal of Linguistics/Revue canadienne de linguistique, 57, 2, 209–240.
Higgins, F. R., 1973, The Pseudo-Cleft Construction in English, doctoral thesis, MIT.
Higgins, F. R., 1979, The Pseudocleft Construction in English, Volume 17 of Outstanding dissertations in linguistics, New York, Garland Publ.
Iatridou, S., S. Varlokosta, 1998, “Pseudoclefts crosslinguistically”, Natural Language Semantics, 6, p. 3−28.
Lees, R., 1963, “Analysis of the cleft sentences in English”, Zeitschrift zur Phonetik, Sprachwissenschaft und Kommunikationsforschung, Berlin, 16, 4.
Mieszek, A., 1974, “Some remarks on Pseudo-Cleft Sentences in English and Polish”, Poznań Studies in Contemporary Linguistics, 3, 225−242.
Moltmann, F., 2013, “Identificational Sentences”, Natural Language Semantics, 21, 43–77.
Moro, A., 1991, “The Anomaly of Copular Sentences”, Working Papers in Linguistics, 1, 1–29.
Moro, A., 1997, The Raising of Predicates: Predicative Noun Phrases and the Theory of Clause Structure, Cambridge, Cambridge University Press.
Moro, A., 2000, Dynamic Antisymmetry, Cambridge, MA, The MIT Press.
Moro, A., 2006, “Copular Sentences”, in: M. Everaert, H. van Riemsdijk (eds), The Blackwell Companion to Syntax, II, Blackwell Publishing, 1–23.
Pană Dindelegan, G., 2015, “Mecanisme discursive de emfază şi focalizare în limba veche. Cu referire specială la subiect”, in: M. Sala, M. Stanciu-Istrate, N. Petuhov (eds), Lucrările celui de-al cincilea Simpozion Internaţional de Lingvistică, Bucureşti, 27–28 septembrie 2013, Bucureşti, Univers Enciclopedic Gold, 310–322.
Roy, I., 2013, Nonverbal predication: Copular sentences and the syntax – semantics interface, Oxford University Press.
Stowell, T., 1978, “What was there before there was there”, in: D. Farkas et al., Papers from the Fourteenth Regional Meeting of the Chicago Linguistics Society, 457–471.
Şerbănescu, A., 1996, “Construcţii scindate”, Limba română, XXXV, 1, 3–10.
Blanca Croitor 18 314
