ArticlePDF Available

On the syntactic specialization of Romanian demonstratives and the grammaticalization of the article CEL

Authors:

Abstract

The paper discusses the diachronic syntactic specialization of Romanian demonstratives according to the morphophonological weak/strong distinction, and the grammaticalization of the article CEL, specific to Romanian from a Romance perspective. It is shown that the reanalysis of the aphaeretic form of the distal demonstrative as the article CEL, through a grammaticalization process that regularly took place in the emergence of Romance determiners, strongly correlates with the diachronic specialization of Romanian demonstratives and with other syntactic changes taking place across-the-board in the Romanian DP.
ON THE SYNTACTIC SPECIALIZATION OF ROMANIAN
DEMONSTRATIVES AND THE GRAMMATICALIZATION
OF THE ARTICLE CEL
ALEXANDRU NICOLAE1
Abstract. The paper discusses the diachronic syntactic specialization of
Romanian demonstratives according to the morphophonological weak/strong
distinction, and the grammaticalization of the article CEL, specific to Romanian from a
Romance perspective. It is shown that the reanalysis of the aphaeretic form of the distal
demonstrative as the article CEL, through a grammaticalization process that regularly
took place in the emergence of Romance determiners, strongly correlates with the
diachronic specialization of Romanian demonstratives and with other syntactic changes
taking place across-the-board in the Romanian DP.
Keywords: diachronic specialization, demonstrative, determiner, reanalysis,
grammaticalization, Romanian
1. PRELIMINARY REMARKS
1.1. Aim and outline of the paper2
The syntax of modern Romanian demonstratives is driven by the weak/strong
distinction which associates each type of demonstrative with specific selection features and
1 Romanian Academy, “Iorgu Iordan – Al. Rosetti” Institute of Linguistics, Department of
Linguistics, Faculty of Letters, University of Bucharest, nicolae_bibi@yahoo.com.
This paper is supported by the Sectorial Operational Programme Human Resources
Development (SOP HRD), financed from the European Social Fund and by the Romanian
Government under the contract number SOP HRD/159/1.5/S/136077.
I am grateful to the Gabriela Pană Dindelegan, Alexandra Cornilescu, Camelia Stan, Adina
Dragomirescu, Adam Ledgeway, and Ian Roberts for all their comments on my work on the syntax of
the Romanian DP, including parts from the present paper, to Emanuela Timotin for setting up the
corpus for The Syntax of Old Romanian (in preparation), partly used here, and to Irina Nicula for
reading the final version of this text. The remaining errors are all mine.
2 For the glossing of the examples, we have used the symbols and conventions generally
accepted in the field, recently used in the 2013 OUP Grammar of Romanian (pp. xxviii–xxxi); of the
abreviations used more rarely or generally employed in the description of Romanian, we draw the
reader’s attention to the following ones: AL – freestanding genitival/possessive marker specific to
Romanian, AUX auxiliary, CL – clitic, DEF – definite, DOM – differential object marker, PERF
perfective, PS simple past, S strong, W – weak; the symbol = indicates cliticization.
RRL, LX, 1, p. 47–70, Bucureşti, 2015
48 Alexandru Nicolae 2
particular syntactic derivations (Cornilescu 2005). By contrast, the syntax of old Romanian
demonstratives is not driven by the weak/strong division – in other words, old Romanian
demonstratives are not syntactically specialized. The first goal of this paper is to investigate
the mechanisms by which Romanian demonstratives become specialized, and the factors
involved in this diachronic process.
Another quirk of modern Romanian (at least from a Romance perspective) is the
existence of the so-called “adjectival / demonstrative article” CEL, a definite determiner
with particular distributional and interpretative features. Old Romanian presents a
somewhat different picture: the forms out of which CEL grammaticalized have a dual
grammar, to some degree common to that of demonstratives, but nonetheless
idiosyncratic. The second goal of this paper is to account for the grammaticalization of
CEL, and to show that this process is associated with the syntactic specialization of
demonstratives and with other diachronic changes in the overall syntax of the
Romanian DP.
The paper is structured as follows: on the basis of previous literature, we present
a synchronic account of the syntax of demonstratives and of the article CEL in modern
Romanian (section 2), and then we turn to the diachronic specialization of
demonstratives and the grammaticalization of CEL (section 3); in section 4 we draw the
conclusions.
1.2. Period investigated
We follow the generally accepted periodization of the Romanian language (Gheţie
1997: 52–53), and distinguish between old Romanian (1500–15103/15214 to 17805) and
modern Romanian (1780 to the present-day). The old Romanian data are extracted from
original texts and translations, mostly focusing on the earliest writings.
2. THE SYNTAX OF DEMONSTRATIVES AND OF THE ARTICLE CEL
IN MODERN ROMANIAN
In contrast to old Romanian, in modern Romanian there is a strict morphological and
distributional specialization of demonstratives, and the determiner CEL has a robust
morphology and a constrained and limited distribution. The goal of this section is to present
the morphosyntactic features of Romanian demonstratives and of the determiner CEL,
against which we set the diachronic analysis that follows (section 3).
3 The earliest attested Romanian text, The Hurmuzaki Psalter (PH.1500-10), a religious translation.
4 The earliest attested Romanian original text, a letter sent by Neacşu Lupu from Câmpulung
to Johannes Benker of Braşov (DÎ.1521: I).
5 S. Micu and G. Şincai’s grammar Elementa linguæ daco-romanæ sive valachicæ (1780).
3 The Syntactic Specialization of Romanian Demonstratives 49
2.1. Demonstratives6
From a functional point of view, the demonstrative system of modern Romanian is
organized along the bipartite proximity distinction: acest(a) (‘this’) vs. acel(a) (‘that’),
behaving similarly to modern French and standard Italian, but contrasting with other
Romance languages (e.g. Portuguese, Spanish, Valencian, and Occitan) (Salvi 2011: 325).
From a morphological point of view, Romanian distinguishes weak and strong
demonstratives7; strength is achieved (i) by means of the final vocalic augment -a8, either
simply added to the weak form ((1a) vs. (1b)) or replacing the final segment -ă of the weak
form ((1c) vs. (1d)), or (ii) by word internal processes ((2a) vs. (2b)).
WEAK DEMONSTRATIVE STRONG DEMONSTRATIVE
(1) a. acest b. acesta
this.W.M this.S.M
c. această d. aceasta
this.W.F this.S.F
(2) a. acea [aʧa] b. aceea [aʧeja]
that.
W.F that.S.F
2.1.1. Distribution
From a distributional point of view, there are stark contrasts between the weak and
the strong forms9. The weak form is a prenominal determiner which occupies the DP-initial
position, selects a non-definite noun / nominal phrase, and precedes all other DP-internal
constituents (numerals, modifiers, the head noun, etc.); it may only be preceded by light
adverbials or by the universal quantifier tot (‘all’):
(3) a. doar aceşti trei copii
only these.W three children
‘only these three children’
6 Since we are interested mostly in the syntactic behaviour of demonstratives, our presentation
in this section makes use of the standard Romanian etymologically complex demonstratives acest(a)
(‘this’) and acel(a) (‘that’), which exhibit both strong and weak forms. In standard Romanian, the
etymologically simple demonstratives ăsta (‘this’) (ia) and ăla (‘that’) (ib) have only strong forms,
whose distributional behaviour is similar to that of their strong counterparts acesta and acela (see
Nicolae 2013a); they are mostly employed in spoken Romanian (Nicula 2008, 2009). See Niculescu
(1968) on the inventory of non-standard demonstratives.
(i) a. M.SG: ăsta F.SG: asta M.PL: ăştia F.PL: astea ‘this(S)’
b. M.SG: ăla F.SG: aia M.PL: ăia F.PL: alea ‘this(S)’
7 In glosses, the weak and the strong form will be distinguished by the symbols W and S,
respectively.
8 Agreement has not yet been reached with respect to the origin of the vocalic augment -a; see
Dimitrescu (1978 and references therein) for discussion.
9 The weak and strong forms of the demonstrative have been also associated with different
DP-internal information-structure functions, which lay beyond the interest of this paper (see, for
details, Tasmowski 1990, Manoliu 2000, Cornilescu 2005, Vasilescu 2009a, i.a.).
50 Alexandru Nicolae 4
b. toţi aceşti copii frumoşi
all these.W children beautiful
‘all these beautiful children’
By contrast, the strong form is a postnominal determiner which combines with a
definite noun. The noun is suffixed by the definite article ((4a) vs. (4b)) and obligatorily
precedes the demonstrative (cf. (4c)). Optional multiple definiteness is available in modern
Romanian, especially in the spoken language (4d) (Iordan 1956).
(4) a. fratele acesta
brother.DEF this.S
‘this brother’
b. *frate acesta
brother this.S
c. *acesta frate(le)
this.S brother(.DEF)
d. muncitorul ăla vrednic / vrednicul
worker.DEF that.S diligent diligent.DEF
‘that diligent worker’
Furthermore, the demonstrative is strictly adjacent to the definite noun, the insertion
of other constituents in between the noun and the demonstrative being completely banned
(5); even nominal arguments (5a) and relational (classifying and thematic) adjectives (5c)
become separated from the selecting head. Another particularity of the postnominal
demonstrative construction is that it freely allows the postnominal distribution of cardinal
numerals (5e).
(5) a. fratele acesta al meu
brother.DEF this.S AL my
‘this brother of mine’
b. *fratele meu acesta
brother.DEF my this.S
c. maşina aceasta nemţească
car.DEF this.S German
‘this German car’
d. *maşina nemţească aceasta
car.DEF German this.S
e. copiii aceştia doi
children.DEF these.S two
‘these two children’
f. *copiii doi aceştia
children.DEF two these.S
Another significant difference between weak and strong demonstratives concerns
nominal ellipsis and occurrence as a predicative: only the strong form may head DPs with
elided heads (6a) (vs. (6b)) and appear as postcopular predicatives (7a) (vs. (7b)).
5 The Syntactic Specialization of Romanian Demonstratives 51
(6) a. omul acesta / acesta
man.DEF this.S this.S
‘this man’ ‘this (one)’
b. acest om / *acest
this.W man this.W
‘this man’
(7) a. El este acesta. b. *El este acest.
he is this.S he is this.W
‘He is this one’
Let us now focus on the syntactic characterization of the weak vs. strong
demonstratives as results from the distributional characteristics reviewed above.
2.1.2. Syntactic analysis
The high domain of the Romanian DP consists of (at least) the following functional
projections (Cornilescu 2005, 2007, Tănase-Dogaru 2009, Cornilescu and Nicolae 2011a,
i.a.)10:
(8) DP > DemP > QP >
The D0-head of the DP projection accommodates interpretable definiteness, being
thus responsible for definiteness valuation/checking; demonstratives merge in Dem0 or in
Spec,DemP (cf. Giusti 1993, Brugè 2002)11, depending on their phrasal status
(head/phrase), and potentially undergo movement to D0/DP; the specifier of QP
accommodates cardinal numerals and other quantifiers.
Turning to the particular situation of Romanian demonstratives, the following results
can be drawn from the facts reviewed in the previous section:
(i) behaviour under nominal ellipsis (6) and in predicative position (7) shows that the
strong demonstrative is phrasal (XP), while the weak demonstrative is a head (X0)
(Cornilescu 2005);
(ii) different DP-internal operations are associated with each type of demonstrative:
while the distribution of the weak demonstrative is similar to its prenominal counterpart
from Romance or English and poses no special problems (9), the postnominal distribution
of strong demonstrative points to the fact that the definite noun undergoes movement to D;
the strict adjacency constraint of the postnominal demonstrative to the nominal head
illustrated in (5) further indicates that the type of movement involved is head-movement
10 Intensive research on the left periphery of the Romanian DP (Cornilescu 2007, Cornilescu
and Nicolae 2011a) in the split-D framework (Giusti 2005, Laenzlinger 2005 i.a.) has shown that the
functional domain of the Romanian DP is more complex: the D-area consists of at least a
higher/external deixis D and a lower/internal agreement D. The adoption of a simple, non-split D
projection suffices for the purposes of the present analysis.
11 UG actually provides two merger positions for demonstratives (Guardino 2012), a high
position like the one in (8) and a lower functional projection in the extended nominal domain. Of
these two positions, (old and modern) Romanian makes use only of the high position (see Cornilescu
and Nicolae 2015).
52 Alexandru Nicolae 6
across the phrasal demonstrative (Cornilescu 2005); head-movement is chosen over phrasal
movement (the general option in the Romanian DP, see Cinque 2004) as a last resort
strategy to bypass a locality problem (Cornilescu and Nicolae 2015), namely the
impossibility of one specifier to crossover another specifier (a constraint not encountered in
old Romanian, see Cornilescu and Nicolae 2011b).
(9) acest frate
this.W brother
a. [DemP acest [NP frate …]]
b. [DP acest [DemP acest [NP frate]]]
(10) fratele acesta
brother.DEF this.S
a. [DemP acesta [NP fratele]]
b. [DP [N+D fratele [DemP acesta [NP fratele]]]]
Retaining this analysis for the diachronic account that follows, in the next section we
briefly turn to the distribution and analysis of the article CEL.
2.2. The article CEL
The article CEL, characterized by Romanian traditional scholarship as a
“demonstrative / adjectival article” (GLR 1966, I: 107–108), is an innovation of Romanian
among the Romance languages (Niculescu 1965: 19–20, Iliescu 2006, 2009), which does
not have Romance counterparts (Reinheimer Rîpeanu 2001: 198, Vasilescu 2009b: 273), at
least as far as its distribution with postnominal modifiers is concerned. From an
interpretative point of view, in modern Romanian, CEL is devoid of demonstrative meaning,
i.e. it no longer encodes proximity distinctions.
2.2.1. Distribution12
CEL functions as a (last resort) freestanding definite article and values definiteness
when the DP-initial position is occupied by numerals which cannot bear the suffixal
definite article (Cornilescu 2004) (11); “quantifying adjectives”, which display a mixed
adjectival and quantificational behaviour (Pană Dindelegan 2003, Cornilescu 2009), have
the option of valuing definiteness either by CEL-insertion (12a) or by definite article
suffixation (12b).
(11) a. cei doi fraţi (vs. a'. *doii fraţi)
CEL two brothers two.DEF brothers
‘the two brothers’
b. cel de-al doilea frate
CEL second brother
‘the second brother’
12 For the full distribution of CEL, see Nicolae (2013b); in this paper, we limit ourselves to the
contexts which present a direct interest for the diachronic analysis that follows.
7 The Syntactic Specialization of Romanian Demonstratives 53
(12) a. cei (foarte) mulţi fraţi
CEL very many brothers
‘the very many brothers’
b. mulţii fraţi ai Mariei
many.DEF brothers AL Mary.GEN
‘Mary’s many brothers’
While the freestanding definite article usage of CEL is not peculiar from a cross-
linguistic perspective, its other context of occurrence (adjectival article) is highly specific
to Romanian (at least from a Romance perspective, see Ledgeway 2012: 113–115). In this
second context13, CEL is postnominal and precedes14 APs headed by qualifying adjectives
(13a), PPs (13b), agreeing past participles (13c), agreeing gerund (obsolete) (13d)
(Cornilescu 2004).
(13) a. casa cea nouă
house(F).DEF CEL.F new.F
‘the new house’
b. casa cea din deal
house(F).DEF CEL.F from hill
‘the house on the hill’
c. copiii cei pierduţi
child.PL.DEF CEL.PL lost.PL
‘the lost children’
d. lebăda cea murindă
swan(
F).DEF CEL.F dying.F
‘the dying swan’
Although CEL is preceded by a definite constituent, its distribution is not as similar to
that of strong demonstratives as may seem at first sight. First of all, CEL may be preceded
by phrasal constituents (compare to the ungrammatical counterparts with postnominal
demonstratives in (5b), (5d)):
13 The interpretation of the CEL-construction has been subject to much controversy; we refer
the reader to the discussion in Cornilescu and Nicolae (2011a; 2012: 1087–1093) for a review of the
relevant literature and a unifying proposal of interpretation.
14 The full distribution of CEL also includes its occurrence as a formative of the superlative (i)
and in the structure of PNs (ii). The grammaticalization of CEL as a superlative morpheme (i) took
place after its grammaticalization as an adjectival article (Brăescu 2015), and represents a Romance-
specific type of grammaticalization (Iordan and Manoliu 1965: 153), widely attested cross-
linguistically, namely DEFINITE > SUPERLATIVE (Heine and Kuteva 2002: 106) (see also Ledgeway
2012: 114, fnt. 44 for discussion). The presence of definite determiners in the internal structure of
proper names (ii) is also widely attested in Romance and cross-linguistically.
(i) a. cea mai tânără fată (ii) Ivan cel groaznic
CEL more new girl Ivan CEL terrible
b. fata cea mai tânără ‘Ivan the Terrible’
girl.DEF CEL more young
‘the youngest girl’
54 Alexandru Nicolae 8
(14) a. fratele meu cel mic
brother.DEF my CEL little
‘my little brother’
b. maşina nemţească cea roşie
car.DEF German CEL red
‘the red German car’
Secondly, while postnominal demonstratives directly precede relational adjectives, in
the CEL-construction relational adjectives are pied-piped along by the head noun (14b), and
cannot be directly preceded by CEL (15a), a distribution permitted with strong demonstratives
(see (5c) above). Another difference between postnominal demonstratives and CEL concerns
the availability of multiple definiteness, totally excluded in the CEL- construction (15b), but
possible in the postnominal demonstrative construction (4d). Also, in contrast to prenominal
demonstratives, CEL cannot (directly or indirectly) precede nouns (15c).
(15) a. *maşina cea nemţească
car.DEF CEL German
b. muncitorul CEL vrednic / *vrednicul
worker.DEF that.S diligent diligent.DEF
‘the diligent worker’
b. *cea maşină / *cea roşie maşină
CEL car CEL red car
CEL is also the licenser of definite nominal ellipsis in Romanian. Romanian keeps
distinct two very similar processes, nominal ellipsis and substantivization, by means of
different licensers (Cornilescu and Nicolae 2012, Nicolae 2013c, Dragomirescu and
Nicolae 2015), contrasting to other Romance languages which employ one and the same
element, the definite article (see Sleeman 1996). Nominal ellipsis, the discourse-
conditioned omission of the nominal head, is headed by the article CEL (16), while
substantivization, a lexically-conditioned process which involves the incorporation of a
silent but contentful noun, e.g. HUMAN, COLOUR, CATEGORY (Kayne 2005), is licensed by
the suffixal definite article (17). Traditional scholarship has also characterized CEL as a
“semiindependent pronoun” (Manoliu-Manea 1968) due to the fact that CEL is unable to
stand alone under nominal ellipsis, being obligatorily accompanied by at least a remnant
(owing to its clitic nature, see Nicolae 2013b: 311); once more, this sets it in contrast to
demonstratives (18).
(16) a. mărul roşu şi cel verde / *şi verdele
apple.DEF red and CEL green and green.DEF
‘the red apple and the green one’
b. câinele sănătos şi cel bolnav / *şi bolnavul
dog.DEF healthy and CEL sick and sick.DEF
‘the healthy dog and the sick one’
(17) a. verdele [
COLOUR]
green.DEF
‘the colour green’ (but not ‘the green apple’)
9 The Syntactic Specialization of Romanian Demonstratives 55
b. bolnavul [HUMAN]
sick.DEF
‘the sick man’ (but not ‘the sick dog’)
(18) Acela / *Cel a venit.
that.S CEL AUX.PERF.3SG come.PPLE
‘That one came’
2.2.2. Syntactic analysis
Taking stock of the properties reviewed in previous subsection, we can draw the
conclusion that CEL directly merges in the D-position of the higher functional nominal
domain, merger in Dem(P) being excluded as CEL does not have demonstrative value. The
fact that CEL may be preceded by phrasal constituents (14) and the inability of CEL to stand
alone under nominal ellipsis (18) further indicate that CEL is a head, not a phrase. If CEL
were phrasal, it would induce the same last resort type of head-movement and block phrasal
movement, just as in the case of the as the strong demonstrative, contrary to fact.
Hence, as a freestanding determiner preceding quantifiers, CEL merges in D0 and
types the phrase as definite; in this structure, movement of the noun across CEL is possible
(Cornilescu 2004), but rare (Nicolae 2013b: 315):
(19) cei doi oameni
CEL two people
[
DP [D0 cei [QP [CardP doi] Q [NumP/NP oameni]]]]
In the adjectival article construction15, CEL also merges in the same position, and
there is obligatory phrasal movement across CEL to the specifier of the DP:
(20) fratele (meu) cel mic
brother.DEF my CEL little
[DP [NP fratele (meu)] [D cel [FP [AP mic] … tNP
2.3. Summary
In this section, we have reviewed the distributional and syntactic properties of
demonstratives and of the article CEL in modern Romanian. The following facts have
emerged from the discussion:
(i) the weak proximal and distal demonstratives and the article CEL are heads, while
the strong demonstratives are phrasal;
(ii) in the extended nominal projection which assumes at least the projections DP >
DemP > QP,
(a) weak demonstratives merge in Dem0 and undergo head movement to D0;
15 Under the split-D hypothesis (see fnt. 10), the analysis of the adjectival article construction
is slightly different (see Cornilescu and Nicolae 2011a); this difference is immaterial for the purposes
of the diachronic analysis that follows.
56 Alexandru Nicolae 10
(b) strong demonstratives merge in Spec,DemP; N0-movement across the
demonstrative derives the postnominal demonstrative construction;
(c) CEL directly merges in D0; the adjectival article construction is derived
by NP-movement to Spec,DP across CEL.
3. THE VIEW FROM DIACHRONY
In the previous section, we have highlighted the fact that the syntax of
demonstratives in modern Romanian aligns along the weak/strong distinction and there is
virtually no optionality in this respect; in other words, modern Romanian demonstratives
exhibit a robust form – syntax correlation. In this section, we show that this correlation was
not active in older stages of Romanian: the distribution of demonstratives in old Romanian
indicates that the weak/strong distinction does not correlate with the head/phrase
distinction. The effect of this lack of specialization has repercussions on the DP-internal
movement operations: the adjacency constraint of postnominal demonstratives to a definite
noun is not well-established in old Romanian (see, for the initial observation, Cornilescu
and Nicolae 2011b: 214).
Similar considerations hold for the article CEL, which also displays a uniform syntax
in modern Romanian. The extensive examination of old Romanian texts actually shows that
in the earliest stages of old Romanian CEL stood a dual grammar, displaying both
demonstrative and article (distributional and interpretative) properties, and hence
illustrating Kroch’s (1989) notion of “grammars in competition” (see Roberts 2007: 319–
331 for an up-to-date discussion).
After a brief presentation of the origin of the Romanian demonstratives and of the
article CEL, we turn to their distributional and interpretative behaviour in old Romanian,
and then propose a diachronic scenario that accounts for the changes encountered in the
transition to modern Romanian.
3.1. The origin of Romanian demonstratives and of the article CEL
Romanian possesses both etymologically simple demonstratives and etymologically
complex demonstratives. The demonstratives ăsta (‘this’) and ăla (‘that’), used only as
strong forms mainly in the spoken language, originate from the vulgar Latin demonstratives
ĭstus (classical Latin ISTE) and, respectively, illum (classical Latin ILLE) in stressed position.
The demonstratives acest(a) (‘this’) and acel(a) (‘that’) are etymologically complex forms,
resulting from the combination between Latin eccum (variant of ECCE ‘behold’) and istum
and, respectively, illum (see the discussion in Dimitrescu 1978: 275).
Turning to CEL, its origin is the aphaeretic distal demonstrative acel(a) (‘that’)
(Dimitrescu 1975: 169). Iordan and Manoliu (1965: 145) correctly remarked that the rise of
CEL as a different form of the distal demonstrative took place after the 16th c., a hypothesis
verified by the present research.
11 The Syntactic Specialization of Romanian Demonstratives 57
3.2. Non-specialized demonstratives
In this section, we focus on non-aphaeretic demonstratives (acest(a) ‘this’ and
acel(a) ‘that’); in the next section, which specifically deals with CEL, we show that the
aphaeretic forms display an ambiguous demonstrative/definite article grammar.
3.2.1. Distribution
The weak forms of demonstratives may appear both prenominally (21) and
postnominally (22); the postnominal usage of weak demonstratives has been eliminated.
The internal make-up of old Romanian DPs with postnominal weak demonstratives is
sometimes very similar to that of modern Romanian postnominal demonstrative DPs with
strong demonstratives (22b).
(21) a. pentru acel bir (DÎ.1593: X)
for that.W tax
‘for that tax’
b. această sămânţă (PH.1500–10: 19r)
this.W seed
‘this seed’
(22) a. neamul acel (CP1.1577: 185r)
nation.DEF that.W
‘that nation’
b. mâncăriei acea porcească (CC2.1581: 18)
food.DEF.GEN that.w porcine
‘of that porcine food’
By the same token, the strong form may appear both prenominally (23) and
postnominally (24). The prenominal usage of the strong form has been also eliminated.
(23) a. aceasta a mea scrisoare (DÎ.1594: X)
this.S AL my letter
‘this letter of mine’
b. acestea cărţi creştineşti (CCat.1560: 2r)
these.S books Christian
‘these Christian books’
(24) a. după feciorulǔ acela (CC2.1581: 24)
after boy.DEF that.S
b. în iezerul acela (A.1620: 19v)
in mountain.lake.DEF that.S
‘in that mountain lake’
Authors like Dimitrescu (1978: 278) and Stan (2013: 35–36) have also remarked the
lack of distributional constraints on the selection of the weak vs strong form.
Furthermore, the strong demonstrative may also precede nouns suffixed by the
definite article, an option no longer available:
58 Alexandru Nicolae 12
(25) a. însuşǔ acela judecătoriulǔ dereptǔ (CC2.1581: 33)
himself that.S judge.DEF honest
‘that honest judge himself’
b. aceasta moşia vândut-am (DÎ.1595–96: XIII)
this.S property.DEF sell.PPLE=AUX.PERF.1SG
‘I sold this property’
Another feature of postnominal strong demonstratives is that they may (26) or may
not (27) be adjacent to the definite article. The latter distribution indicates phrasal
movement across the strong demonstrative, an option no longer available.
(26) in anulu acesta 1593 (DÎ.1593: CXIII)
in year.DEF this.S 1593
‘in that year 1593’
(27) fiiulǔ meu acesta mortǔ era (CC2.1581: 12)
son.DEF my this.S dead was
‘this sone of mine was dead’
3.2.2. Nominal ellipsis
Another disparity between old and modern Romanian demonstratives concerns
nominal ellipsis; both strong (28) and weak (29) forms may occur in DPs with elided heads.
In the transition to modern Romanian, the selection of the weak form in elliptical DPs has
been eliminated.
(28) a. audu şi înţelegu lucru ca acesta (DÎ.1599: XVIII)
hear.PRES.3PL and understand.PRES.3PL thing like this.S
‘they hear and understand a thing like this one’
b. Acela era răul sterpiciunei (SVI.~1670: 7v)
that.S was wickedness.DEF sterility.DEF.GEN
‘That was the wickedness of sterility’
(29) a. acel e frate mie (CT.1560–1: 74r)
that.W is brother me.DAT
‘That is my brother’
b. Aceste zise marele împărat Alexandru (A.1620: 74v)
these.w say.PS.3SG great.DEF emperor Alexander
‘The great emperor Alexander said these (words)’
3.2.3. Evidence for the commencement of specialization
However, in contrast to the data reviewed above, there is evidence for the fact that
the strong/weak distinction tends to become syntactically specialized. To begin with,
postnominal weak demonstratives have a low frequency, and they are mostly attested in
translations; in general, the distribution of weak vs strong forms in original documents is
more stable than in translations (Stan 2013: 35–36).
13 The Syntactic Specialization of Romanian Demonstratives 59
More importantly, the postnominal weak demonstrative construction (30) differs
from the postnominal strong demonstrative construction (31) in that only the latter allows
(and actually favours in old Romanian) the multiple expression of definiteness.
(30) a. mâncăriei acea porcească (CC2.1581: 18)
food.DEF.GEN that.W porcine
‘to that porcine food’
b. bucinulǔ acelǔ îngerescǔ (CazV.1643: 26v)
alphorn.DEF that.W angelical
‘that angelical alphorn’
(31) a. lumiei aceştiia înşălătoarea (CC2.1581: 18)
world.DEF.GEN this.S.GEN deceiving.DEF.GEN
‘of this deceiving world’
b. locul acesta sfântul (CV.1563–83: 17r)
place.DEF this.S sacred.DEF
‘this sacred place’
A similar behaviour has been observed in the situation the aphaeretic demonstratives
(see Vasiliu 2007 for the initial observation): apocopate (i.e. weak) forms of CEL disallow
multiple definiteness, while their strong counterparts actually favour it (see the next section
for details).
3.2.4. Summary
The old Romanian distribution of demonstratives reviewed above indicates that the
weak/strong distinction does not correlate with a clear phrasal status (head/phrase), nor with
uniform DP-internal movement options: (i) both weak and strong forms may head elliptical
DPs, a fact which shows that weak forms are not uniformly classified as heads;
(ii) movement to D is not obligatory for either type of demonstrative, as indicated by the
fact that both weak and strong forms may be preceded by other material, but is possible for
both weak and strong demonstratives, as shown by the fact that both types of form may
occupy the DP-initial position preceding the head-noun; (iii) strong demonstratives may be
preceded either by definite nouns or by complex phrasal constituents, this again testifying
to the fact that the head/phrase categorization is unclear. Despite this vacillating behaviour,
there are certain clear signs of specialization from the earliest (16th c.) texts, namely the
availability of multiple definiteness only with strong demonstratives.
3.2.5. Diachronic development
The passage to modern Romanian consolidated the weak/strong distinction by
specializing the weak form as a head which obligatorily undergoes movement to D0, and
the strong form as a phrase which occupies the Spec,DemP and no longer moves to the
D-position16. In other words, a grammar with more options has been reset to a grammar
16 The loss of the strong form’s ability to move to D does not imply that strong forms no
longer entertain any relation with the D-position; the relation between D and strong demonstratives is
mediated by Agree (coindexation in older terminology, see Roberts and Roussou 2003: 133).
60 Alexandru Nicolae 14
with fewer options in which each type of demonstrative is associated with an unambiguous
phrasal status and with unambiguous movement options.
3.3. The ambiguous grammar of CEL
3.3.1. Acel > CEL or acela > CEL?
The grammaticalization of CEL is supported by phonological processes of reduction
of the distal demonstrative acel(a) (‘that’); as known, morphophonological reduction is a
frequent process that takes place in the grammaticalization of determiners (Roberts and
Roussou 2003: 132).
The origin of CEL is actually the aphaeretic weak form of the distal demonstrative,
(a)cel, not the strong form acela simultaneously affected by the aphaeresis of the initial
vowel a- and the apocope of the final vowel -a (cf. also Giurgea 2012: 41–42). The
aphaeretic strong form (i.e. cela) as the etymological basis of CEL is ruled out by
phonological17 (see Dimitrescu 1978: 276) as well as distributional reasons; in particular,
notice that aphaeretic strong forms do not exhibit the so-called “semiindependent”
behaviour characteristic to CEL (the presence of a DP-internal constituent to their right)
(32a) and may precede definite or non-definite constituents (32b, c), just like the
non-aphaeretic demonstratives:
(32) a. ceia-u dzis că nu e acolo (DÎ.1593: CXII)
those.S=AUX.PERF.3PL say.PPLE that not is there
‘those (men) said that it is not there’
b. asculta (...) de [cealea de Pavelu grăitele] (CV.1563–83: 43v)
listen. IMPERF.3SG of those.S by Pavel spoken.PL.DEF
‘he listened to the those (things) spoken by Pavel’
c. pintru celea săbii (DÎ.1599–600: XXV)
for those.S swords
‘for those swords’
Further evidence for the fact that the grammaticalization of CEL proceeds from
simultaneously aphaeretic and apocopate forms (henceforth glossed as CEL18) is given by
multiple definite constructions. Vasiliu (2007: 75) correctly remarks that in postnominal
position the apocopate forms of old Romanian CEL disallow multiple definiteness (33),
while non-apocopate forms behave like the old and modern Romanian postnominal strong
demonstratives, i.e. multiple definiteness is allowed (34) – actually favoured in this stage of
Romanian –, but not always obligatory (35). Vasiliu’s observation is strongly supported by
17 This is obvious when we compare the singular feminine strong form (a)ceea whose
augmentation is achieved through a word-internal change with the singular feminine form (a)cea.
18 For convenience, we will gloss the aphaeretic and apocopate forms with CEL, bearing in
mind that these forms stand a dual demonstrative/definite article analysis, as will be seen in the next
subsection.
15 The Syntactic Specialization of Romanian Demonstratives 61
the fact this behaviour is systematic in the collection of the earliest attested original
Romanian writings (DÎ).
(33) a. înţelepciunea cea veacinicî (CazV.1643: IIv)
wisdom.DEF CEL eternal
‘the eternal wisdom’
b. fraţii cei mici (CC2.1581: 34)
brothers.DEF CEL little
‘the little brothers’
c. datoria cea veache (DÎ.1595–96: XII)
debt.DEF CEL old
‘the old debt’
d. cugetul lui cel rău (DÎ.1600: XLIV)
thought.DEF his CEL mean
‘his mean thought’
(34) a. la locul cela strimtul (DÎ.1521: I)
at place.DEF CEL narrow.M.DEF
‘at that narrow place’
b. iară popa cela greşitul (să
and priest.DEF CEL.M.SG trespassed.DEF.M.SG SUBJ
se/să facă călugăr) (CPrav.1560–62: 9r; Prav.1581: 206v)
SE become.SUBJ monk
‘and the/that priest who has trespassed should become a monk’
c. fiulǔ lui cela mai marele (CC2.1581: 12)
son.DEF his that.S more old.DEF
‘his older son’
(35) feciorulǔ lui cela mai mare (CC2.1581: 22)
son.DEF his that.S more old
‘his older son’
Hence, the aphaeretic non-apocopate forms are variants of the full demonstrative, and
an accurate diachronic analysis of the grammaticalization of CEL should mostly focus on
the forms simultaneously affected by aphaeresis and apocope.
3.3.2. Demonstrative distribution and interpretation
Despite the clearly different distribution of the aphaeretic and apocopate forms in
contrast to their non-apocopate counterparts with respect to multiple definiteness, there is
distributional and interpretative evidence that the aphaeretic weak forms also exhibit
demonstrative behaviour in old Romanian.
To begin with, these forms may directly precede non-definite nouns (36) / adjectives
plus nouns (37) / nouns plus adjectives (38) (Dimitrescu 1978, Giurgea 2013, Stan 2013), a
distribution no longer available for modern Romanian CEL, but available for old and
modern Romanian weak demonstratives:
(36) a. cei oameni ai săi (CC1.1567: 97v)
CEL men AL his
‘his people / those people of his’]
62 Alexandru Nicolae 16
b. au luat cel grâu (DÎ.1593: IX)
AUX.PERF.3PL take.PPLE CEL wheat
‘they have taken the / that wheat
c. pre cel sol (A.1620: 57r)
DOM CEL messenger
‘the/that messenger’
(37) a. cea bună nădejde (FT.1571–75: 2v)
CEL good hope
‘the/that good hope’
b. cea puţină credinţă (CC2.1581: 297)
CEL little faith
‘the/that little faith’
(38) a. toate cele lucrure bure (MI.~1630: 191r)
all CEL things good
‘all the/those good things’
b. cel fecior curvariu (Ev.1642 : 179)
CEL son fornicating
‘the/that fornicating son’
c. den celǔ lucru rău (CC2.1581: 17)
from CEL thing bad
‘from the/that bad thing’
Another demonstrative feature of these forms is represented by the direct
combination with relational adjectives (Brăescu and Dragomirescu 2014), a feature no
longer available in the modern Romanian CEL-construction:
(39) a. cea dumnezeiască viaţă (CC2.1581: 147)
CEL godly life
‘that divine life’ (dumnezeiesc < Dumnezeu ‘God’)
b. birăul cel rumânescu şi cu cel armenescu (DÎ.1593–97: XCVII)
mayor CEL Romanian and with CEL Armenian
‘the Romanian mayor with the Armenian one’
From an interpretative point of view, the aphaeretic and apocopate distal forms still
seem to preserve a spatial meaning (i.e. the meaning of a full distal demonstrative); in the
oldest Romanian writings, in the same text, we also find aphaeretic and apocopate proximal
forms; hence, a proximity opposition of aphaeretic and apocopate forms is still at play to a
certain degree:
(40) a. cele grele pedepse a tale (FT.1571–75: 3r)
CEL hard penalties AL your
‘your hard penalities / those hard penalties of yours’
b. ceastă lume (FT.1571–75: 2v)
this.W world
‘this world’
17 The Syntactic Specialization of Romanian Demonstratives 63
(41) a. celǔ feciorǔ micǔ (CC2.1581: 11)
CEL son little
‘the/that little son’
b. avuţiia cestui pămînt (CC2.1581: 49)
wealth.DEF this.W.GEN land
‘the wealth of this land’
To sum up, in contrast to the evidence for a distinction between the aphaeretic forms
with apocope and those without apocope presented in the previous subsection, in the
present subsection we have discussed certain characteristics that indicate that the aphaeretic
apocopate forms also present demonstrative features. We may safely conclude that the
forms out of which CEL eventually grammaticalized displayed dual demonstrative/definite
article behaviour; in other words, these forms had a double categorization, hence
illustrating the phenomenon of grammars in competition.
3.3.3. Nominal ellipsis
A final piece of evidence in favour of the idea that the CEL-forms of old Romanian
did not display the same behaviour of the modern Romanian CEL-construction is given by
nominal ellipsis. Recall that modern Romanian, in contrast to other Romance languages,
distinguishes definite nominal ellipsis from substantivization by using the article CEL for
the former process and the suffixal definite article for the latter (see examples (16)–(17)
above). By contrast, in the oldest Romanian writings, elliptical DPs are headed by the
suffixal definite article (42a); crucially, in the later editions of the same passage, the
definite article strategy is replaced by the CEL-strategy (42b) (Stan 2015: 62).
(42) a. nooa (CT.1560–1: 121v)
new.DEF
b. cel nou (BB.1688: 793)
CEL new
‘the new one’
This change took place across-the-board in the passage from the earliest texts of old
Romanian to later texts (Dragomirescu and Nicolae 2015; Nicolae 2015).
3.3.4. The grammaticalization of CEL
The grammaticalization of the aphaeretic and apocopate form of the distal
demonstrative as definite determiner is intimately related to the diachronic specialization of
the Romanian demonstratives.
Recall that the principal phenomenon that took place in the syntax of demonstratives
is the specialization of the weak forms as heads which select a non-definite nominal
complement, and of the strong forms as phrases which take a definite noun as their
complement, which subsequently raise to D via head-movement.
In what follows, we sketch the steps which have led to the reanalysis of CEL as a
definite article:
64 Alexandru Nicolae 18
(i) The aphaeretic and apocopate form of the distal demonstrative generally patterns
with the weak demonstrative forms. Hence, along with the weak demonstrative, it gradually
becomes specialized as a head.
(ii) Since it displays dual demonstrative/article features, two merger positions (i.e.
two structures) are available for CEL-forms: Dem0 (with potential movement to D0, just like
modern Romanian weak demonstratives) (43a) and D0 (43b); these two structures co-exist
for a period. The same scenario has been also advocated by Giusti (1998) for the emergence
of the Romance definite article.
(43) a. DP b. DP
ei ei
D
0 DemP D0
ei cel
Dem
0
cel
(iii) The semantic bleaching of the [+demonstrative] feature – also supported by the
almost complete disappearance by the end of old Romanian of aphaeretic and apocopate
proximal demonstratives (cest < acest(a)), with which the CEL-forms enter the proximity
opposition – triggers the complete reanalysis of CEL as D0-head; structure (43a) is no longer
possible, the only option for the merger of CEL being (43b).
In sum, the grammaticalization of article CEL is a familiar Move > Merge type of
reanalysis (Roberts and Roussou 2003: 136), often encountered in the emergence of
Romance determiners. Similarly to the transformation of the Latin distal demonstrative ille
into the Romance definite article (including the Romanian one, see Nicolae 2012), the
transformation of the Romanian distal demonstrative into a definite article involved
morphophonological reduction (acel > cel), semantic bleaching (loss of the demonstrative
property), and categorial change (demonstrative > article). This has prompted researchers
like Iliescu (2006, 2009) to qualify the development of CEL as an example within Romanian
of a recurrent Romance typological change: the production of a parallel new form from
almost identical material.
3.3.5. CEL in quantificational phrases
One final problem concerns the usage of CEL as a freestanding definite article in
phrases that contain morphologically defective quantifiers (see example (11) above). This
usage of CEL is related to an old Romanian structure which disappeared in the passage from
old to modern Romanian, namely the “low definite article”, first discussed in Cornilescu
and Nicolae (2011b). In this structure, the definite article may be suffixed on a noun
preceded by a non-definite adjective; in other words, definiteness valuation/checking
proceeds across a prenominal intervener:
(44) a. tinde (...) cătră noi [svântă mana ta] (FT.1571–75: 3v)
extend.PRES.2SG towards us holy hand.DEF your
‘extend your holy hand towards us with mercy’
19 The Syntactic Specialization of Romanian Demonstratives 65
b. cu [cinstită cartea mării tale] (DÎ.1596: CVI)
with honoured letter.DEF highness.DEF.GEN your
‘with your highness’ honoured letter’
This type of definiteness valuation/checking was reset by the end of old Romanian in
favour of a grammar which favours a more local type of valuation: the bearer of
definiteness must be hosted by the first noun/adjective in the DP.
The low definite article construction is also available with quantificational phrases in
old Romanian; the suffixal definite article checks definiteness across the intervening
quantifier:
(45) a. deade Dumnezeu [zeace cuvintele sale] (CCat.1560: 4r)
give.PS.3SG God ten words.DEF his
‘God gave his ten commandments’
b. arătarea [a dooa venireei lui] (CC2.1581: 600)
showing.
DEF second coming.DEF.GEN his
‘the showing of his second coming’
However, in contrast to prenominal adjectives, which are φ-complete and have the
option to be suffixed by the definite article and hence value definiteness in a local manner,
quantifiers (with few exceptions, see (12) above) are φ-incomplete (i.e. morphologically
defective) and cannot be suffixed by the definite article. Since non-local definiteness
valuation in no longer available, the presence of the suffixal definite article on a post-
quantifier constituent is excluded. A novel freestanding exponent of the D0 projection is
already available in the old language, namely CEL, and hence it is selected as a last resort
option to check/value definiteness in quantificational DPs. Hence, definiteness valuation in
DPs of the type in (45) is satisfied by the insertion of CEL (46):
(46) a. cele zece cuvinte ale sale
CEL ten worlds AL his
‘his ten words’
b. cea de-a doua venire a lui
CEL second coming AL his
‘his second coming’
The low definite article in quantificational DPs was already an archaism in old
Romanian (in contrast to the low definite article with prenominal adjectives, which was
relatively frequent, see Cornilescu and Nicolae 2011b). This construction is rare, and the
usage of CEL as a freestanding definite article devoid of (distal) demonstrative meaning is
quite frequent; for example, in the following fragment from the popular novel Alexandria
(A.1620), it is obvious that the quantificational DP “cei 6 filosofi” (‘the six philosophers’)
anaphorically refers back to the indefinite DP “6 filosofi” (‘six philosophers’), and CEL is a
definite article not a distal demonstrative:
(47) Şi luo Alexandru de la ei 6 filosofi şi ieşi
and take.PS.3SG Alexander from them six philosophers and leave.PS.3SG
66 Alexandru Nicolae 20
din ostrov [...] . Şi-i îmbrăcă
from island and=CL.ACC.3PL dress.PS.3SG
pre cei 6 filosofi […] (A.1620: 7v)
DOM CEL six philosopers
‘And Alexander took six philosophers from them and left the island. […] And he
dressed the six philosophers and …’
The non-ambiguous usage of the aphaeretic and apocopate form as freestanding
definite article in very early old Romanian documents further validates the idea that CEL-
forms have dual categorization in old Romanian, defended in section 3.3.1 above.
4. CONCLUSIONS
4.1. The reconfiguration of the syntax of demonstratives and the emergence of CEL
The main phenomenon that took place in the syntax of Romanian demonstratives is
the diachronic specialization of the weak and strong forms as heads and phrases,
respectively. This conferred non-ambiguous syntactic derivations of each type of form:
(i) the weak form is a head merging in Dem0 that selects a non-definite complement;
the short demonstrative values definiteness and moves to D0;
(ii) the strong form is a phrase which merges in Spec,DemP and selects a definite
complement; the D-position is targeted by the definite noun; the definite noun may only
undergo head-movement to D0, phrasal movement being blocked by locality constraints.
The fact that each type of demonstrative has particular selectional features (weak
forms select a non-definite complement / strong forms take a definite complement) is
further verified by the fact that in both old and modern Romanian multiple definiteness is
available only with strong forms.
The aphaeretic and apocopate distal demonstratives were included in the weak
demonstrative paradigm and unambiguously categorized as heads; the subsequent loss of
the [+demonstrative] feature triggered their reanalysis as articles. The result of this process
is a second definite article of Romanian, CEL, which has particular distributional features,
and a particular DP-internal information structure function.
4.2. Factors favouring the emergence of CEL
Due to space limitations, little has been said about the interpretation of the
non-quantificational CEL-construction (see footnote (12) above). Except for DPs containing
quantifiers and for elliptical DPs, CEL-insertion is mostly optional in modern Romanian
(see Nicolae 2013b); however, when CEL insertion takes place, the postnominal modifier
acquires a particular pragmatic interpretation, which is generally associated with discourse
prominence / saliency, and with the signalling of an identifying property of referent of the
CEL-containing DP. The question arises as to what are the factors that have led to the
emergence of this construction.
We believe that this construction emerged from the interplay of two other changes in
the internal syntax of Romanian DPs. The first phenomen is the specialization of the
21 The Syntactic Specialization of Romanian Demonstratives 67
prenominal adjectival position: in old Romanian, AP-movement to the left edge of the DP
was virtually unbounded, and any type of adjective could occupy the DP-initial position, as
richly documented by Brăescu and Dragomirescu (2014). Very often, AP-displacement to
the DP-left periphery is information-structurally driven. This change is part of a larger shift
witnessed in the change from Latin to late Latin and Romance, documented by Ledgeway
(2012: 210): “Though as a typological diagnostic the position of the adjective with respect
to the noun has been claimed on crosslinguistic grounds, as well as on the evidence of Latin
and Romance, not to correlate robustly with other word order patterns, there are nonetheless
some consistent Latin-internal patterns which incontrovertibly point towards an early shift
from an original head-final AN order to the head-initial NA order (…) that continues into
Romance”.
The other change has already been briefly discussed in section 3.3.5 above: the
strengthening of locality conditions on definiteness valuation in the passage from old to
modern Romanian, further complicated by the fact that the Romanian definite article is a
suffix with particular conditions of encliticization.
The combined effect of these two changes is the emergence of a novel construction
in which CEL functions as an escape hatch allowing a nominal phrase to target the leftmost
position of the DP, bypassing thus the intervention effects induced by information-
structurally marked modifiers, DP-periphery constituents themselves.
The emergence of a novel construction with a particular DP-internal information
structuring function is strategy that compensates for the gradual reduction of
“pragmatically-driven word order (…) resulting form the greater accessibility of topic- and
focus-fronting positions situated in the left edge of individual phasal projections”
(Ledgeway 2012: 281–282) – in other words, discourse configurationality – which has
taken places across-the-board in the passage from Latin to Romance, as extensively
documented by Ledgeway (2012).
CORPUS
A.1620 Alexandria. Ed. F. Zgraon, Bucureşti, Fundaţia Naţională pentru Ştiinţă şi Artă,
2005 (Cele mai vechi cărţi populare în literatura română, 11).
BB.1688 Biblia. Ed.: Biblia adecă Dumnezeiasca Scriptură a Vechiului şi Noului Testament,
tipărită întâia oară la 1688 în timpul lui Şerban Vodă Cantacuzino, Domnul Ţării
Româneşti, Bucureşti, Editura Institutului Biblic, 1977.
CazV.1643 Varlaam, Cazania, ed. J. Byck, Bucureşti, Editura Academiei RSR, [s.a.], 1–506.
CC1.1567 Coresi, Tâlcul Evangheliilor. Ed.: Coresi, Tâlcul evangheliilor şi molitvenic
românesc, ed. V. Drimba, Bucureşti, Editura Academiei Române, 1998, 31–187.
CC2.1581 Coresi, Evanghelie cu învăţătură. Ed. S. Puşcariu, Al. Procopovici: Diaconul Coresi,
Carte cu învăţătură (1581), vol. I, Textul, Bucureşti, Socec, 1914.
CCat.1560 Coresi, Catehism. Ed. Al. Roman-Moraru, in I. Gheţie (coord.), Texte româneşti din
secolul al XVI-lea. I. Catehismul lui Coresi; II. Pravila lui Coresi; III. Fragmentul
Todorescu; IV. Glosele Bogdan; V. Prefeţe şi Epiloguri, Bucureşti, Editura
Acadmiei Române, 1982, 101–5.
CP1.1577 Coresi, Psaltire slavo-română. Ed.: Coresi, Psaltirea slavo-română (1577) în
comparaţie cu psaltirile coresiene din 1570 şi din 1589, ed. S Toma, Bucureşti,
Editura Academiei RSR, 1976, 35–662.
68 Alexandru Nicolae 22
CPrav.1560–2 Coresi, Pravila. Ed. Gh. Chivu, in I. Gheţie (coord.), Texte româneşti din secolul al
XVI-lea, 218–31.
CT.1560–1 Coresi, Tetraevanghel. Ed.: Tetraevanghelul tipărit de Coresi. Braşov 1560 – 1561,
comparat cu Evangheliarul lui Radu de la Măniceşti. 1574, ed. F. Dimitrescu,
Bucureşti, Editura Academiei RPR, 1963.
CV.1563–83 Codicele Voroneţean. Ed. M. Costinescu, Bucureşti, Editura Academiei Române,
1981, 229–400.
Documente şi însemnări româneşti din secolul al XVI-lea, text stabilit şi indice de
Gh. Chivu, M. Georgescu, M. Ioniţă, Al. Mareş, Al. Roman-Moraru, Bucureşti,
Editura Academiei Române, 1979.
Ev.1642 Evanghelie învăţătoare. Ed. A.-M. Gherman, Bucureşti, Editura Academiei
Române, 2011, 153–480.
FT.1571–5 Fragmentul Todorescu (Carte de cântece). Ed. I. Gheţie, in I. Gheţie (coord.), Texte
româneşti din secolul al XVI-lea, 336–43.
MI.~1630 Manuscrisul de la Ieud. Ed. M. Teodorescu, I. Gheţie, Bucureşti, Editura Academiei
RSR, 1977, 153–170.
PH.1500–10 Psaltirea Hurmuzaki, ed. I. Gheţie şi M. Teodorescu, Bucureşti, Editura Academiei
Române, 2005.
Prav.1581 Pravila ritorului Lucaci. Ed. I. Rizescu, Bucureşti, Editura Academiei RSR, 1971,
161–83.
SVI.~1670 Varlaam şi Ioasaf. Ed.: M. Stanciu Istrate, Reflexe ale medievalităţii europene în
cultura română veche: Varlaam şi Ioasaf în cea mai veche versiune a traducerii lui
Udrişte Năsturel, Bucureşti, Editura Muzeului Național al Literaturii Române,
2013, 82–325.
REFERENCES
Brăescu, R., 2015, Gradarea în limba română. Perspectivă istorică şi tipologică, in preparation.
Brăescu, R., A. Dragomirescu, 2014, “Sintaxa adjectivelor relaţionale în limba română veche”,
Limba română, LXIII, 1, 27–47.
Brugè, L., 2002, “The position of demonstratives in the extended nominal projection”, in: G. Cinque
(ed.). Functional structure in DP and IP. The cartography of syntactic structures, volume I,
Oxford, Oxford University Press, 15–53.
Cinque, G., 2004, “A phrasal movement analysis of the Romanian DP”, in: A. Minut, E.Munteanu
(eds), Studia linguistica et philologica in honorem D. Irimia, Iaşi, Editura Universităţii
“A. I. Cuza”, 129–142.
Cornilescu, A., 1992, “Remarks on the determiner system of Rumanian: the demonstratives AL and
CEL”, Probus, 4, 3, 189–260.
Cornilescu, A., 2004, “În legătură cu conceptul de pronume semiindependent. Observaţii asupra
articolului adjectival cel”, in: G. Pană Dindelegan (ed.), Tradiţie şi inovaţie în studiul limbii
române, Bucureşti: Editura Universităţii din Bucureşti, 5162.
Cornilescu, A., 2005, “Demonstratives and minimality”, Bucharest Working Papers in Linguistics, 7,
1, 102–117.
Cornilescu, A., 2007, “Despre trăsăturile periferice şi cum le-am putea folosi”, in: R. Zafiu, C. Stan,
A. Nicolae (eds), Studii lingvistice. Omagiu profesoarei Gabriela Pană Dindelegan, la
aniversare, Bucureşti, Editura Universităţii din Bucureşti, 4356.
Cornilescu, A., 2009, “Measure phrases and the syntax of Romanian nouns and adjectives”,
Bucharest Working Papers in Linguistics, XI, 1, 35–66.
23 The Syntactic Specialization of Romanian Demonstratives 69
Cornilescu, A., A. Nicolae, 2011a., “Nominal peripheries and phase structure in the Romanian DP”,
Revue roumaine de linguistique, LVI, 1, 3568.
Cornilescu, A., A. Nicolae, 2011b, “On the syntax of Romanian definite phrases: Changes in the
patterns of definiteness checking”, in: P. Sleeman, H. Perridon (eds), The noun phrase in
Romance and Germanic. Structure, variation, and change, Amsterdam/Philadelphia, John
Benjamins, 193221.
Cornilescu, A., A. Nicolae, 2012, “Nominal ellipsis as definiteness and anaphoricity: the case of
Romanian”, Lingua, 122, 10, 1070–1111.
Cornilescu, A., A. Nicolae, 2015, “Diachronic variation in the syntax of Romanian demonstratives”,
in preparation.
Dimitrescu, F., 1975, Introducere în morfosintaxa istorică a limbii române, Bucureşti, Universitatea
din Bucureşti.
Dimitrescu, F., 1978, “Pronumele demonstrativ”, in: F. Dimitrescu (ed.), Istoria limbii române:
fonetică, morfosintaxă, lexic, Bucureşti, Editura Didactică și Pedagogică, 273–278.
Dragomirescu, A., A. Nicolae, 2015, “L’ellipse nominale avec article défini de l’ancien roumain au
roumain moderne: Le cas du participe passé”, in: É. Buchi, J.-P. Chauveau, J-M. Pierrel (eds),
Actes du XXVIIe Congrès international de linguistique et de philologie romanes, Strasbourg,
Société de linguistique romane/ÉliPhi, in press.
Gheţie, I., 1997, “Probleme teoretice și metodologice ale cercetării limbii române literare vechi”,
in: I. Gheție (ed.), Istoria limbii române literare. Epoca veche (1532–1780), Bucureşti, Editura
Academiei Române, 23–53.
Giurgea, I., 2012, “The origin of the Romanian «possessive-genitival» article al and the development
of the demonstrative system”, Revue roumaine de linguistique, LVII, 1, 35–65.
Giurgea, I., 2013, Originea articolului posesiv-genitival al şi evoluţia demonstrativelor în română,
Bucureşti, Editura Muzeului Naţional al Literaturii Române.
Giusti, G., 1993, La sintassi dei determinanti, Padua, Unipress.
Giusti, G., 1998, “The rise of a functional category. From Latin ILLE to the Romance article and personal
pronoun”, University of Venice working papers in linguistics, 8, 2, 53–71.
Giusti, G., 2005, “At the left periphery of the Romanian noun phrase”, in: M. Coene, L. Tasmowski
(eds), On space and time in language, Cluj-Napoca, Clusium, 23–49.
GLR 1966 – A. Graur, M. Avram, L. Vasiliu (eds), 1966, Gramatica limbii române, 2 volumes,
Bucureşti, Editura Academiei.
Guardiano, C., 2012, “Demonstratives, word order and the DP between syntax and semantics:
crosslinguistics remarks”, Studies in Greek linguistics, 32, 100–115.
Heine, H., T. Kuteva, 2002, World lexicon of grammaticalization, Cambridge, Cambridge University
Press.
Iliescu, M., 2006, “L’article adjectival roumain: un exemple de recurrence typologique ciclique”,
Revue roumaine de linguistique, LI, 1, 159163.
Iliescu, M., 2009, “Aspects de l’évolution de l’article défini en français et en roumain”, Travaux de
linguistique, 59, 2, 1323.
Iordan, I., 1956, Limba română contemporană, Bucureşti, Editura Ministerului Învăţământului.
Iordan, I., M. Manoliu, 1965, Introducere în lingvistica romanică, Bucureşti, Editura Didactică şi
Pedagogică.
Kayne, R. S., 2005, “Silent years, silent hours”, in: Movement and silence, Oxford, Oxford University
Press, 241–260.
Kroch, A., 1989, “Reflexes of grammar in patterns of language change”, Language variation and
change, 1, 199–244.
Laenzlinger, C., 2005, “Some notes on DP-internal movement”, GG@G [Generative grammar in
Geneva], 4, 227–260.
Ledgeway, A., 2012. From Latin to Romance. Morphosyntactic typology and change, Oxford, Oxford
University Press.
Manoliu, M., 2000, “Demonstratives, story-world and talk-interaction”, in: M. Coene, W. De Mulder,
P. Dendale, Y. D’Hulst (eds), Traiani Augusti vestigia pressa sequamur. Studia lingvistica
in honorem Lilianae Tasmowski, Padua, Unipress, 583–600.
70 Alexandru Nicolae 24
Manoliu-Manea, M., 1968, Sistematica substitutelor în româna standard, Bucureşti, Editura
Academiei Române.
Nicolae, A., 2012, “Gramaticalizarea articolului hotărât românesc. Noi rezultate”, Limbă şi literatură,
LIV, 1–2, 7–19.
Nicolae, A., 2013a, “Demonstratives”, in: G. Pană Dindelegan (ed.), The grammar of Romanian,
Oxford, Oxford University Press, 294–300.
Nicolae, A., 2013b, “The determiner CEL”, in: G. Pană Dindelegan (ed.), The grammar of Romanian,
Oxford, Oxford University Press, 309–318.
Nicolae, A., 2013c, Types of ellipsis in Romanian, PhD dissertation, University of Bucharest &
University of Cambridge (cotutelle).
Nicolae, A., 2015, “Nominal ellipsis”, in: G. Pană Dindelegan (ed.), The syntax of old Romanian,
Oxford, Oxford University Press, in preparation.
Nicula, I., 2008, “Utilizări pragmatice ale demonstrativului în limba vorbită actuală: asta vs aceasta”,
in: G. Pană Dindelegan (ed.), Limba română. Dinamica limbii, dinamica interpretării,
București, Editura Universităţii din Bucureşti, 127–132.
Nicula, I., 2009, “Dinamica pronumelor şi a adjectivelor demonstrative în limba română actuală.
Observaţii pe corpsurile de română vorbită”, in: G. Pană Dindelegan (ed.), Dinamica limbii
române actuale. Aspecte gramaticale şi discursive, Bucureşti, Editura Academiei Române,
181194.
Niculescu, A., 1965, Individualitatea limbii române între limbile romanice, I, Bucureşti: Editura
Ştiinţifică.
Niculescu, A., 1968, Observations sur les démonstratifs daco-roumains provenant du lat. ILLE, Revue
roumaine de linguistique, XIII, 5, 471–480
Pană Dindelegan, G., 2003, “Dificultăți de încadrare morfologică. La interferența mai multor părți de
vorbire: «mulţi», «puţini»”, in: Elemente de gramatică. Dificultăţi, controverse, noi
interpretări, Bucureşti, Humanitas, 65–74.
Reinheimer Rîpeanu, S., 2001, Lingvistica romanică. Lexic – morfologie – sintaxă, Bucureşti, All
Universitar.
Roberts, I., 2007, Diachronic syntax, Oxford, Oxford University Press.
Roberts, I., A. Roussou, 2003, Syntactic change. A minimalist approach to grammaticalization.
Cambridge, Cambridge University Press.
Salvi, G., 2011, “Morphosyntactic persistence’, in: M. Maiden, J. C. Smith, A. Ledgeway (eds), The
Cambridge history of Romance languages. I. Structures, Cambridge/New York, Cambridge
University Press, 318–381.
Sleeman, P., 1996, Licensing empty nouns in French, The Hague, Holland Academic.
Stan, C., 2013, O sintaxă diacronică a limbii române vechi, Bucureşti, Editura Universităţii din
Bucureşti.
Stan, C., 2015, “Some functions of the definite article in old Romanian”, in: G. Pană Dindelegan,
R. Zafiu, A. Dragomirescu, I. Nicula, A. Nicolae, L. Esher (eds), Diachronic variation in
Romanian, Newcastle upon Tyne, Cambridge Publishing Scholars, 55–65.
Tănase-Dogaru, M., 2009, The category of Number. Its relevance for the syntax and semantic
typology of the nominal phrase, Bucureşti, Editura Universităţii din Bucureşti.
Tasmowski, L., 1990, “Les démonstratifs français et roumains dans la phrase et dans le texte”, in:
P. Cadiot, A. Zribi-Hertz (eds), Aux confines de la grammaire: l’anaphore. Langages, 96, 82–99.
Vasilescu, A., 2009a, “Strategii pragmatice de reluare, gramaticalizate ca relaţii apozitive de tip
GN1GN2”, Limba română, LVIII, 2, 275284.
Vasilescu, A., 2009b, “Cel: categorie semilexicală”, in: R. Zafiu, B. Croitor, A.-M. Mihail (eds),
Studii de gramatică. Omagiu doamnei profesoare Valeria Guţu Romalo, Bucureşti, Editura
Universităţii din Bucureşti, 265287.
Vasiliu, L., 2007, “Atributul”, in: M. Avram (ed.), Sintaxa limbii române în secolele al XVI-lea –
al XVIII-lea, Bucureşti, Editura Academiei Române, 73–109.
... Another element that occurs in definite DPs in Romanian is 'Cel'. Cel has a controversial status in the literature between being considered a definite determiner (Nicolae 2015), a complex phrasal determiner (Marchis & Alexiadou 2009), or a free-standing article (Giurgea 2013). Its distribution, however, is non-controversial. ...
Article
This dissertation focuses on the syntactic aspects of Romanian-Serbian code-switching (CS). It explores a number of issues concerning several domains and theoretical mechanisms, especially the structure of the nominal domain, the structure and derivation of coordinated structures, cliticization (both second-position and verbal clitics), the nature of affixal articles, phases, and the mechanisms of Agree and case-licensing. In addressing these questions, a fundamental assumption is Bošković’s (2008, 2012) dichotomy which divides languages into NP (languages without articles) and DP (languages with articles). This distinction is especially relevant here, as the languages involved differ in this respect – Romanian having, and Serbian lacking articles. Chapter 2 focuses on the TNP-internal CS, examining the interaction between Romanian definite articles, Serbian nouns, and Serbian adjectives. By examining the requirements of these elements, I propose a new mechanism for article cliticization involving Agree and Affix Hopping that can account for both Romanian and CS constructions. Chapter 3 tackles the interaction between the nominal and the verbal domain through Left-Branch Extraction (LBE). Since the same nominal allows or disallows LBE in CS depending on its position, LBE is used to determine the points of CS, where CS within a phasal domain only affects that particular phasal domain, and not the entire structure. Chapter 4 investigates coordinated TNPs in CS, further examining the behavior of NP vs. DP elements and showing that NP elements are more flexible than DP elements in terms of the switches they allow. Chapter 5 focuses on clitics in CS, Romanian having verbal and Serbian second-position clitics. I show that word-internal CS is allowed as long as the elements involved do not form a morphosyntactic head (X0). Chapter 6 looks at case assignment in CS. Romanian Case-assigners are shown to behave differently than Serbian Case-assigners in CS, with Serbian verbs behaving differently in CS than they do in Serbian. Overall, while the findings illustrate relevant CS aspects, they highlight the functionality of analyzing elements outside of their input grammar, creating derivations that can exceptionally be found only in CS.
... Under a phase-based theory of movement (Chomsky 2000;2008), in order to extract out of a phase, constituents must move to the edge of that phase. Adopting the proposal that DPs are phases (Chomsky 2008), a 20 See, e.g., Nicolae (2015) for an analysis of weak demonstratives in Romanian as heads. 21 The determiner jiñ(i) may appear discontinuous to the noun (Vázquez Álvarez 2011: 337), but as discussed by Vázquez Álvarez (2011) and Little & Wiegand (2018), it also acts as a focus marker. ...
Article
Full-text available
Ch’ol (Mayan) exhibits asymmetries in what is available for left branch extraction. While both numerals and interrogative possessors are able to extract from absolutive subject position, only numerals may extract from absolutive object position. To capture this asymmetry, I provide evidence that objects with overt possessors always undergo object shift, blocking left branch extraction. This follows from the Freezing Principle (Ross 1974; Wexler & Culicover 1977), or a ban on extraction from a moved constituent. Objects with numeral modifiers do not obligatorily undergo object shift and therefore may extract from the object. In addition to numerals and interrogative possessors, I present and analyze possibilities for other elements to extract out of various positions. I situate this work within Agree-based theories of extraction (e.g. Rackowski & Richards 2005; van Urk & Richards 2015; Branan 2018) and discuss this proposal’s theoretical implications. Unless otherwise noted, all data comes from the author’s fieldwork.
... Finally, some mention of Romanian must be made which, in contrast to other Romance varieties, also presents two other analytic markers of definiteness in the DP, the so-called demonstrative and possessive articles cel (< ACCU ACCU--ILLE ILLE 'behold-that') and al (< AD AD 'to' + ILLE ILLE 'that'), respectively. Essentially, the former functions as a last resort strategy pressed into service whenever the canonical enclitic article is not available (Cornilescu 1992;Nicolae 2015b). This arises whenever the head noun is null, hence unable to host the enclitic article, e.g. ...
Chapter
Full-text available
The present chapter critically reconsiders the traditional typological distinction between Latin and Romance in terms of a syntheticity-analyticity opposition, according to which core grammatical categories and distinctions are marked morphologically in Latin but syntactically in Romance. After considering a wide selection of the Romance evidence for innovative analytic structures manifested in the emergence of a series of functional categories lexicalizing various functional heads within the nominal, verbal and clausal domains, a number of empirical and theoretical problems and limitations with this superficial dichotomy are reviewed. These highlight how the observed differences between Latin and Romance cannot be simplistically reduced to a synthetic-analytic opposition. Rather, it is argued that the observed rise of Romance analyticity should be considered an epiphenomenal development, ultimately the manifestation of a deeper change, but not, significantly, its cause, related to a change in the head directionality parameter from head-finality to head-initiality.
... Finally, some mention of Romanian must be made which, in contrast to other Romance varieties, also presents two other analytic markers of definiteness in the DP, the so-called demonstrative and possessive articles cel (< ACCU ACCU--ILLE ILLE 'behold-that') and al (< AD AD 'to' + ILLE ILLE 'that'), respectively. Essentially, the former functions as a last resort strategy pressed into service whenever the canonical enclitic article is not available (Cornilescu 1992;Nicolae 2015b). This arises whenever the head noun is null, hence unable to host the enclitic article, e.g. ...
Article
Full-text available
This article explores parallels in the dimensions of microvariation characterizing the functional structure and organization of the Romance nominal and clausal groups. Within a parameter hierarchy approach it is argued that observed synchronic and diachronic variation across both domains can be readily captured in terms of a single set of higher- and above all lower-level parametric options. This parallelism constitutes a welcome finding in that it points to how the available parametric space can be constrained and defined in terms of a set of common transcategorial principles and options.
Article
Full-text available
In this paper, we argue for the existence of two nominal peripheries (the n*-periphery and the d*-periphery), corresponding to the two phases (cf. Chomsky 2001, 2009, Legate 2003) inside the determiner phrase: the n*-phase, parallel to the vP (as in Svenonius 2004), and the d*-phase, parallel to the CP. The existence of the two peripheries is evidenced, in the first part of the paper, by the non-homogeneous behavior of Romanian pre-nominal adjectives, in what concerns properties like genericity, specificity, position with respect to cardinals a.o. The description of periphery adjectives is based on an integrated classification of adjectives from a threefold perspective: syntactic, ontological and combinatorial. In the second part, a typical Romanian structure is investigated in detail: the adjectival article construction (băiatul cel cuminte, boy.the that good 'the good boy'). It is shown that despite its post-nominal position, the adjective is in fact a pre-nominal periphery adjective, which merges above the lower D position in a split DP framework, and is given a focus interpretation. The definite nominal is subsequently attracted to the specifier of the higher D (cel). Movement of the noun is possible only if, after the linearization of the n*-phase, the definite noun is the edge constituent (cf. Chomsky's PIC).
Article
Full-text available
We discuss a type of variation in the pattern of definiteness valuation in Old Romanian (XVI th to XVIII th century), which has never been noticed before, and examine its significance for the evolution of the DP. In Old Romanian, the suffixed definite article variably occurs either on the first or on a lower [+N] constituent (noun or adjective), so that an indefinite adjective may precede the definite noun. In contrast, in Modern Romanian, it is always the first noun which bears the definite article, while, in case an adjective precedes the noun, the definite article occurs on the adjective. The existence of this lower (definite) article raises several questions (the contexts where it occurs, its significance for the emergence of the enclitic definite article, etc.), to which this paper provides tentative answers. We propose that the existence of a lower definite article combined with a tendency for economy made possible the extension of the use of the article to (pre-nominal) adjectives, gradually leading to stricter conditions in the valuation of definiteness in Modern Romanian (Local Agree). At the same time, the lower article is evidence that the Romanian enclitic definite article originates in a post-posed demonstrative.
Article
Cambridge Core - European Language and Linguistics - The Cambridge History of the Romance Languages - edited by Martin Maiden
Chapter
This chapter investigates the syntax of expressions of age and time. English sentences like At the age of seven, John... and it's already past six are argued to contain unpronounced YEAR(S) and unpronounced HOUR(S), respectively. Differences in this area of syntax between English and French or Italian may, in the spirit of comparative syntax work, be relatable to independent differences concerning number agreement. The proposal that few always modifies NUMBER is supported by the fact that in some cases (characterized by few being separated from number) few can to some degree of acceptability modify overt number. The study of expressions of age and time provides evidence in favor of postulating silent counterparts of the nouns year and hour (and age and time).
Article
I argue that the genitival agreeing marker al, used to introduce oblique-marked DPs and agreeing possessors, originates in a strong (i.e. non-suffixal) form of the definite article, Lat. illu(m)>*elu>alu, which in an unattested stage of Romanian behaved like present-day cel. I show that this form underwent four different reanalyses, yielding genitival al, ordinal al, alalt 'the other' and alde. I argue that the invariable genitival a of southern Balkan dialects and northern Romanian varieties comes from al by loss of inflection. I argue that the present-day distal demonstratives ăl(a)/al(a), aia, (a)hăl(a) etc. do not continue Latin ille (which prenominally is only continued by the article al), but represent an innovation (as first proposed by Iliescu 1967) due to the replacement of acest/cest by the forms aiest/aest/a(h)ăst/ăst, which triggered a similar replacement of acel by aiel/a(h)ăl/ăl. I then try to reconstruct the mechanism by which the reanalysis of al took place.
Article
The paper examines the syntax and interpretation of measure phrases (=MPs) inside Romanian DPs and APs. The MP construction is trans-categorial, so that an understanding of its semantic properties is welcome. Following Schwarzschild (2006), we suggest that MPs are means of measuring out monotonic dimensions in the lexical structure of adjectives, nouns, PPs, etc. Monotonic constructions constitute a family, prototypically represented by the cardinal numerals, but also by partitive and pseudo-partitive constructions. The existence of shared formal elements in all monotonic constructions suggests the presence of a shared functional category, called the Mon(otonicity) P(hrase). Its head, Mon 0 relates a lexical category in its complement with a MP in its specifier. Not all MPs receive a partitive monotonic interpretation. The interpretation of MPs inside DPs shows a difference between Partitive MPs, which track monotonic dimensions of objects (doi centimetri de sfoară 'two centimeters of rope'), and Attributive MPs, which describe dimensional non-monotonic properties of objects (găleată de zece litri 'ten liters bucket'). In the second part of the paper, a detailed description of the internal structure of DPs and APs which contain MPs is given, starting from the premise that the two interpretations of the MP, partitive and attributive, respectively, might correspond to two different configurations.
Book
The phenomenon of grammaticalization - the historical process whereby new grammatical material is created - has attracted a great deal of attention within linguistics. This is an attempt to provide a general account of this phenomenon in terms of a formal theory of syntax. Using Chomsky's Minimalist Program for linguistic theory, Roberts and Roussou show how this approach gives rise to a number of important conceptual and theoretical issues concerning the nature of functional categories and the form of parameters, as well as the relation of both of these to language change. Drawing on examples from a wide range of languages, they construct a general account of grammaticalization with implications for linguistic theory and language acquisition.