A Cross-Linguistic Study of the Acquisition of Clitic and Pronoun Production

Abstract

This study develops a single elicitation method to test the acquisition of third-person pronominal objects in 5-year-olds for 16 languages. This methodology allows us to compare the acquisition of pronominals in languages that lack object clitics (“pronoun languages”) with languages that employ clitics in the relevant context (“clitic languages”), thus establishing a robust cross-linguistic baseline in the domain of clitic and pronoun production for 5-year-olds. High rates of pronominal production are found in our results, indicating that children have the relevant pragmatic knowledge required to select a pronominal in the discourse setting involved in the experiment as well as the relevant morphosyntactic knowledge involved in the production of pronominals. It is legitimate to conclude from our data that a child who at age 5 is not able to produce any or few pronominals is a child at risk for language impairment. In this way, pronominal production can be taken as a developmental marker, provided that one takes into account certain cross-linguistic differences discussed in the article.

Full text

Language Acquisition, 00: 1–35, 2015
Copyright © Taylor & Francis Group, LLC
ISSN: 1048-9223 print / 1532-7817 online
DOI: 10.1080/10489223.2015.1028628
A Cross-Linguistic Study of the Acquisition of Clitic
and Pronoun Production
Spyridoula Varlokosta
University of Athens
Adriana Belletti
University of Siena
João Costa
Universidade Nova de Lisboa
Naama Friedmann
Tel Aviv University
Anna Gavarró
Universitat Autònoma de Barcelona
Kleanthes K. Grohmann
University of Cyprus
Maria Teresa Guasti
University of Milano-Bicocca
The research reported in this paper was undertaken within COST Action A33 (PI and chair: Uli Sauerland; vice chair:
Heather van der Lely).
Spyridoula Varlokosta was the coordinator of the working group that undertook the research reported in this article
within COST Action A33 (WG1 – Binding) and contributed to test design, data collection and analysis, and writing
up. Adriana Belletti, João Costa, Naama Friedmann, Anna Gavarró, Kleanthes K. Grohmann, Maria Teresa Guasti, and
Laurice Tuller contributed to test design, data collection and analysis, and writing up; they are listed alphabetically. Maria
Lobo contributed to test design as well as data collection and analysis. The rest of the authors provided descriptions of the
pronominal systems in the languages tested, translated the experimental materials into their own language, or contributed
to collection and analysis of their own language data; they are listed alphabetically.
Correspondence should be sent to Spyridoula Varlokosta, University of Athens, Philology/Linguistics,
Panepistimioupoli Zografou, Athens, 15784 Greece. E-mail: svarlokosta@phil.uoa.gr
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/hlac.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

2 VARLOKOSTA ET AL.
Laurice Tuller
Université François-Rabelais, Tours
Maria Lobo
Universidade Nova de Lisboa
Darinka And
--
elkovi
´
c
University of Belgrade
Núria Argemí
Universitat Autònoma de Barcelona
Larisa Avram
University of Bucharest
Sanne Berends
University of Groningen
Valentina Brunetto
University of Siena
Hélène Delage
University of Geneva
María-José Ezeizabarrena
University of the Basque Country
Iris Fattal
Tel Aviv University
Ewa Haman
University of Warsaw
Angeliek van Hout
University of Groningen
Kristine Jensen de López
Aalborg University
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 3
Napoleon Katsos
University of Cambridge
Lana Kologranic
University of Zagreb
Nadezda Krsti
´
c
University of Belgrade
Jelena Kuvac Kraljevic
University of Zagreb
Aneta Mi˛ekisz
University of Warsaw
Michaela Nerantzini
University of Athens
Clara Queraltó
Universitat Autònoma de Barcelona
Zeljana Radic
University of Zagreb
Sílvia Ruiz
Universitat Autònoma de Barcelona
Uli Sauerland
Zentrum für Allgemeine Sprachwissenschaft [Center for General Linguistics]
Anca Sevcenco
University of Bucharest
Magdalena Smoczy
´
nska
Educational Research Institute, Warsaw
Eleni Theodorou
University of Cyprus
Heather van der Lely
Harvard University and Université Catholique de Louvain
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

4 VARLOKOSTA ET AL.
Alma Veenstra
University of Groningen and Max Planck Institute for Psycholinguistics
John Weston
Queen Mary, University of London
Maya Yachini
Tel Aviv University
Kazuko Yatsushiro
Zentrum für Allgemeine Sprachwissenschaft [Center for General Linguistics]
This study develops a single elicitation method to test the acquisition of third-person pronominal
objects in 5-year-olds for 16 languages. This methodology allows us to compare the acquisition of
pronominals in languages that lack object clitics (“pronoun languages”) with languages that employ
clitics in the relevant context (“clitic languages”), thus establishing a robust cross-linguistic baseline
in the domain of clitic and pronoun production for 5-year-olds. High rates of pronominal production
are found in our results, indicating that children have the relevant pragmatic knowledge required
to select a pronominal in the discourse setting involved in the experiment as well as the relevant
morphosyntactic knowledge involved in the production of pronominals. It is legitimate to conclude
from our data that a child who at age 5 is not able to produce any or few pronominals is a child at risk
for language impairment. In this way, pronominal production can be taken as a developmental marker,
provided that one takes into account certain cross-linguistic differences discussed in the article.
1. INTRODUCTION
In this study, we explore the acquisition of pronominal objects in 16 languages. The comprehen-
sion of pronominal elements has been intensively studied in a variety of typologically diverse
languages. These studies indicate that, in some languages, young children interpret pronominals
differently from adults. A clear asymmetry has been found between the interpretation of pronouns
and clitics, however. Children allow a local interpretation of the personal pronoun, whereas they
interpret clitics in a targetlike manner, i.e., not allowing a local antecedent for the (nonreflexive)
clitic, already at the age of 3 (Chien & Wexler 1990; McKee 1992; Baauw, Coopmans & Philip
1999;Varlokosta2000, among others). The comparison between pronouns and clitics in produc-
tion has not received as much attention. Specifically, third-person object pronominal elements
have been studied mainly in clitic languages (that is, languages that have clitics in addition to
nonclitic pronominal elements), without comparison to pronoun languages (languages lacking
clitic pronouns) (but s ee Bloom 1990;Grüter2006, 2007; Pérez-Leroux, Pirvulescu & Roberge
2008b, for a comparison between clitic and pronoun languages). These studies have not always
used the same methods; for example, some of the results for some of the languages stem from
spontaneous speech, which is not directly comparable to data from elicitation tasks.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 5
Here, we developed a single elicitation method that was used to test the acquisition of third-
person pronominal objects in 5-year-olds for 16 languages.
1
This allowed for a direct comparison
of languages that differ in how they express pronominal arguments. This methodology allowed
us to compare the acquisition of pronominals in languages that lack object clitics (pronoun
languages) with languages that employ clitics in the relevant context (clitic languages), thus
establishing a robust cross-linguistic baseline in the domain of clitic and pronoun production
at the same stage of language development.
2
This is particularly important given the relevance
of clitic production to detecting specific language impairments (SLI). Omission of third-person
clitics or lack of clitic production in obligatory contexts has been claimed to be a clinical marker
of SLI in French (e.g., Paradis, Crago & Genesee 2003), Greek (Stavrakaki & van der Lely 2010;
Tsimpli & Stavrakaki 1999; but, for different claims, see Manika, Varlokosta & Wexler 2011;
Varlokosta, Konstantzou & Nerantzini 2014), and Italian (Bortolini et al. 2006). We do not know
whether such omissions are limited to clitics or extend to all types of pronominals. In fact, differ-
ences in the expression of pronominal direct objects across several linguistic dimensions might
lead to different performance in their production in typical and atypical language-developing
children. These differences between clitics and pronouns can be found at the phonological, mor-
phological, syntactic, and semantic levels (Kayne 1975;Zwicky1977). On the phonological side,
clitics are unstressed monosyllabic elements, in contrast to pronouns, which may receive stress
and can include more than a single syllable. As for their acquisition, according to some theories,
less salient phonological elements may be harder to acquire (Gerken 1996; Vanderweide 2005).
This approach would predict better production of (stressed) pronouns compared to clitics. At the
morphological level, object clitics are typically homophonous with definite articles or are other-
wise related to a full pronominal form. Additionally, at the syntactic level, clitics are heads that
cannot stand on their own and need to attach to an appropriate host. A further syntactic difference
among pronominal elements relates to their distribution: Pronouns have the same distribution as
full nominals (DPs), whereas clitics need a “special position,” depending on the aforementioned
“appropriate host”—this can be the inflected verb (leading to preverbal proclisis or postverbal
enclisis in the languages included in our study) or another element (such as the second-position
clitics found in many Slavic languages). Object clitics can thus occupy a position different from
that of canonical object DPs. On the semantic side, clitics may refer to human and nonhuman or
inanimate objects, while pronouns typically refer to humans/animate.
These differences have lead Cardinaletti & Starke (1999) to propose a tripartition into clitic,
weak, and strong pronouns across languages. An essential feature of this tripartition is that weak
pronouns have an intermediate status between strong pronouns and clitics. Like strong pronouns,
they fill syntactic positions reserved for maximal projections in the clause structure (DP positions
for direct object pronouns), whereas clitics ultimately fill a head position (they incorporate within
the inflected verbal head in Romance languages, for example). Like clitics, phonologically weak
1
Our choice for testing only 5-year-olds was determined by the fact that our objective was to discover suitable test
methods that can be used for diagnosing language problems and impairment in children of school-entry (i.e., age 5 to
6) in these 16 languages, since at the age of school-entry, an assessment of language problems and impairment has great
practical importance.
2
Clitics and pronouns have the same pragmatic function because they have the same information structure status:
They are used as anaphoric forms when the antecedent is clear from the context. In Gundel, Hedberg & Zacharski’s
(1993) Givenness Hierarchy, elements that are in focus (in a given discourse) are highest on the hierarchy. These are
expressed as pronouns in pronoun languages and clitics in clitic languages.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

6 VARLOKOSTA ET AL.
pronouns do not carry independent stress (as mentioned before). Hence, weak pronouns may
be seen as phonological clitics but not as syntactic clitics, which have a head status. As indi-
cated by Germanic languages (such as German), strong and weak pronouns may be largely
homophonous. Hence, their distinction is only detectable through their different syntactic and
phonological behavior. Without committing ourselves to any specific analysis of cliticization (see
Belletti 1999; Sportiche 1996, among others), we can safely assume that pronominal clitics may
be considered the weakest form of pronouns in that at the end of the cliticization process the
clitic ultimately is a head and incorporates/forms a word with the verb that hosts it. In contrast,
both weak and strong pronouns maintain the distribution of a maximal projection and thus can be
considered less weak in the relevant sense. In this article we will consider clitics and pronouns,
without capitalizing on the further weak/strong partition, as the relevant distinction in the test uti-
lized ultimately hinges on the head versus maximal projection status of the pronominal element.
The tripartition proposed by Cardinaletti & Starke (1999) has been challenged theoretically by
Gabriel & Müller (2005) and empirically by Schmitz & Müller (2008), who show that within the
group of clitic pronouns a distinction must be made with respect to their complexity, which allows
to explain acquisition asymmetries between subject and object clitics in French. However, given
that the research reported in this article is about direct object clitics, we do not address this issue
further, since this would take the discussion too far afield, as the issues concerning subject clitics
in French are rather complicated (see Hamann & Belletti 2006 for further relevant discussion).
Previous studies on clitic production lead to the following generalizations:
i. Children omit clitics in some languages.
Broadly speaking, clitic omission up to at least 4 or 5 years is found to different extents
for Catalan (Wexler, Gavarró & Torrens 2004; Gavarró, Torrens & Wexler 2010),
(European) Portuguese (Costa & Lobo 2006), French (Jakubowicz et al. 1996;
Hamann, Rizzi & Frauenfelder 1996; Jakubowicz & Rigaut 2000), Italian (Schaeffer
1997), and Spanish (Fujino & Sano 2002), as well as bilingual Spanish in contact with
Basque (Ezeizabarrena 1996; Larrañaga 2000; Larrañaga & Guijarro-Fuentes 2011).
In other languages, such as (Standard Modern) Greek (Tsakali & Wexler 2003),
Romanian (Babyonyshev & Marin 2006), Serbo-Croatian (Ilic & Deen 2004), and
possibly Spanish (Wexler, Gavarró & Torrens 2004; Gavarró, Torrens & Wexler 2010),
children were found not to omit clitics from age 2, but see the following discussion.
ii. Children tend to place their clitics in the correct position from the onset of clitic
production.
For instance, Guasti (1993/1994) shows that Italian children place clitics preverbally in
declarative sentences, but postverbally in imperative and nonfinite contexts in a
targetlike way. Similar findings have been reported for other languages in which
proclisis is the dominant pattern for clitic placement (e.g., Marinis 2000, for Greek;
Ezeizabarrena 1996, 1997, for Spanish; Wexler, Gavarró & Torrens 2004, for Spanish
and Catalan). As for predominantly enclitic languages, such as European Portuguese
and Cypriot Greek, children are known to make placement errors, generalizing the
postverbal position for clitics arguably beyond age 3 and a half (see Duarte & Matos
2000, for European Portuguese; Petinou & Terzi 2002, for Cypriot Greek).
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 7
The data collected so far for all of these languages allow for some conclusions to be drawn
regarding the nature of the problems in the acquisition of clitics. First, when children omit clitics,
they do not do so because of the phonological deficiency of these forms. As shown in Jakubowicz
et al. (1998), children omit accusative clitics in French but do not omit determiners with the exact
same phonological form. Clitics are phonologically similar across languages, yet they are omitted
only in some of the clitic languages.
Second, the rate and nature of clitic omission may be language-specific. For this reason, it is
important to determine in which languages clitics are actually omitted, and at what rate. Take
Spanish, for instance: while Wexler, Gavarró & Torrens (2004) claim that there is very little or no
omission of clitics in the language, Castilla, Pérez-Leroux & Eriks-Brophy (2008) contend that
there is some, although the variety of Spanish examined in the two studies is not the same; the
former is a study of continental Spanish and the latter of Columbian Spanish.
3
Similar lack of
consensus on the status of omission within a language exists for Romanian (Avram 1999, 2001;
Babyonyshev & Marin 2006) and French (Jakubowicz et al. 1998; Pérez-Leroux, Pirvulescu &
Roberge 2008a). These different results—even for the same language—may stem from the fact
that the data have been collected with different methodologies or come from different sources
(spontaneous speech vs. experiments). One has to be cautious about a direct comparison between
languages when the methods used were not exactly the same; in other words, it is not entirely
clear if one can say that languages differ with respect to the acquisition of clitics on the basis of
data gathered with different methodologies.
As for object pronouns, good production has been reported in elicitation studies for English
(De Villiers, Cahillane & Altreuter 2006, but see Hyams & Wexler 1993 and Valian 1991,for
evidence from spontaneous data that there is little object omission in English before the age
4;06), Dutch (Spenader, Smits & Hendriks 2009), and Hebrew-speaking children (Ruigendijk
et al. 2010; Novogrodsky, Balaban & Friedmann 2010), but relatively high levels of omission
have been observed for German-speaking children (Jakubowicz et al. 1996, 1997; Ruigendijk
et al. 2010).
For the purposes of this study, we asked the following research questions:
What is the pattern of third-person object pronominal production at the age of 5 years, when
children who acquire different languages are tested with the same methodology?
In particular:
(RQ1) Are pronouns produced better than clitics or vice versa?
(RQ2) Is there a difference in clitic placement errors for languages with different clitic
positions?
3
Research by Fujino & Sano (2002), Ezeizabarrena (1996, 1997), Larrañaga (2000), and Larrañaga & Guijarro-
Fuentes (2011) suggests that there is more omission, based on naturalistic longitudinal corpora. However, the obligatory
contexts for clitic production in naturalistic data are often few or include different kinds of clitics, such as first-, second-,
and third-person pronominal clitics or reflexive clitics; thus, the conclusions based on naturalistic data should be read
with caution. Moreover, the (Basque Region in the North of Spain) Spanish variety of some of the children studied in
Ezeizabarrena (1996, 1997), Larrañaga (2000), and Larrañaga & Guijarro-Fuentes (2011) permits null objects in some
environments (see Franco & Landa 2003).
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

8 VARLOKOSTA ET AL.
We will use the term pronoun to cover both weak and strong pronouns.
4
This article is structured as follows. Section 2 provides the relevant background concerning
existing research on clitic and pronoun production tasks with young children, followed up in
section 3 by cross-linguistic issues of clitic placement, and core properties of the languages
included in the study, divided into clitic languages and pronoun languages. In section 4,the
methodology of the experimental task is described in detail and results are presented in section 5.
Section 6 contains the discussion of the findings.
2. BACKGROUND: CLITIC AND PRONOUN ELICITED PRODUCTION STUDIES
Experiments eliciting production of accusative clitics have been carried out in several languages.
As mentioned earlier, the results reported in the literature show a great deal of heterogeneity,
across languages, but also for the same language, and even in the same age range for a given
language. Most of these studies have been undertaken on 2-, 3-, 4-, and 5-year-old children.
Table 1 lists elicited production experiment results by age of the children studied, by particular
study and by language, including Catalan, French, Greek, Italian, Portuguese, Romanian, and
Spanish. The results are summarized in three rates: rate of clitic production (including clitics
produced with agreement errors); rate of object omission; and rate of production of full, lexical
DPs.
5
From Table 1, we see that in some languages clitic production (the first rate provided in the
grey cells) is quite high from very early on (for example, in Greek); in other languages this
rate is initially quite low (Catalan, French, Italian, and Romanian). Clitic production appears
to remain low for some time in French and in Portuguese, though different studies show quite
different rates for clitic production in French, even at age 3 and 4. This wide variation in results
for French continues in the studies of older children, ranging from just above 50% to well over
90% in 5- and 6-year-olds. Extremely varying rates of clitic production (and DP production and
omission) are reported for Spanish elicited production experiments throughout the ages studied,
with rates ranging from 45% to 100% in children aged 3 and 4. While, at first glance, differences
in rates might be taken to be related to differences in the dialect of the particular language under
investigation (Colombian Spanish versus Iberian Spanish or Canadian French versus continental
French), closer examination shows that varying rates are found for the same variety of a language,
particularly for French (e.g., Jakubowicz & Rigaut [2000] for 2-year-olds; Chillier-Zesiger et al.
[2006] for 4-year olds; Chillier-Zesiger et al. [2006] vs. Tuller et al. [2011] for 6-year-olds).
Notice also that the rates of clitic omission and production of a full DP also vary considerably
from study to study, including studies of the same age range and the same language (Italian:
4
Most studies for different languages have focused on third-person accusative clitics only. Little is known concerning
the acquisition of other clitics and pronouns in most languages. Some exceptions are Babyonyshev & Marin (2006)for
dative clitics in Romanian, Jakubowicz & Rigaut (2000) for reflexive clitics in French, Costa et al. (2008) for dative
and reflexive clitics in all persons in European Portuguese, Ezeizabarrena (1996) for the production of first- and second-
person Spanish clitics and pronouns from age 1;06 to 4, Gavarró & Mosella (2009) for indirect object clitics in Catalan,
Novogrodsky, Balaban & Friedmann (2010) for reflexive pronouns in Hebrew, and Ruigendijk et al. (2010) for reflexive
pronouns in Hebrew and German.
5
Note that these three rates often do not add up to 100%, as children also produced other answers irrelevant to clitic
production.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

TABLE 1
Elicited Production of Accusative Clitics in Children Aged 2, 3, 4, 5, and 6: Age x Language x Study (Mean Age, Age Range, and N of Children; % Clitics,
% Omission, and % Full DPs)
Catalan French Greek Italian Portuguese Romanian Spanish
2-year-olds W, G & T 2004: J & R 2000: T & W 2004: S 2000: B & M 2006: B & L 2001:
2;03 (1;10–2;11)
N = 8
2;04 (2;00–2;05)
N = 5MLU< 3
2;04–3;00
N = 15
2;05 (2;01–2;06)
N = 5
2;05 (2;00–2;11)
N = 12
(2;04–3;10)
N = 15
18% 79% 3% 0% 62% 20% 99% 1% 0% 22% 64% 14% 38% 60% 2% NA 13% NA
J & R 2000: T 2009: W, G & T 2004:
2;05 (2;04–2;07)
N = 7MLU>3
2;07 (2;04–2;10)
N = 5
2;08 (2;06–2;11)
N = 8
21% 9% 47% 14% 59% 27% 90% 8% 2%
3-year-olds W, G & T 2004: J & N, Forthcoming: S, E, S & V 2008: S 2000: C & L 2006: A 2001: W, G & T 2004:
3;06 (3;0–3;11)
N = 11
3;02
N = 12
3;06 (2;10–4;03)
N = 9
3;05 (3;01–3;11)
N = 11
3;06 (2;07–3;11)
N = 14
NA (2;00–4;00)
N = 11
3;07 (3;05–3;11)
N = 10
70% 22% 8% 39% 33% 20% 94% 4% <2% 62% 15% 23% 13% 72% 15% NA 28% NA 98% 0% 1%
P-L, P & R 2008a: M, V & W 2011: T 2009: C, L, C & S 2008: B & M 2006: C 2009:
3;05 (3;00–3;10)
N = 10
3;10 (3;01–6;00)
N = 27
3;07 (3;02–3;11)
N = 14
3;07
N = 24
3;06 (3;04–3;10)
N = 13
3;00 (2;09–3;03)
N = 39
13% 35% 51% 96% NA NA 86% 14% 0% 13% 67% 0% 93% 7% 0% 45% 22% 4%
A 2001: C & P 2010:
NA (2;01–5;00)
N = 14
3.0 (NA)
N = 34
NA 31% NA 33% 25% 1%
4-year-olds W, G & T 2004: C et al. 2006: S & L 2010: S 2000: C & L 2006: W, G & T 2004:
4;07 (4;03–5;01)
N = 12
4;00 (3;05–4;05)
N = 18
4;04 (3;11–5;03)
N = 17
4;06 (4;01–4;10)
N = 10
4;04 (4;00–4;09)
N = 7
4;07 (4;03–5;01)
N = 12
96% 4% 0% 69% 21% 9% 97% NA NA 89% 0% 11% 23% 68% 9% 100% 0% 0%
J & N, Forthcoming: T 2009: C & P 2010:
4;02
N = 12
4;04 (4;02–4;10)
N = 10
4;00 (NA)
N = 39
78% 9% 10% 86% 12% 2% 62% 15% 1%
(Continued)
9
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

TABLE 1
(Continued)
Catalan French Greek Italian Portuguese Romanian Spanish
P-L, P & R 2008a:
4;03 (4;01–4;08)
N = 10
42% 26% 31%
C et al. 2006:
4;09 (4;06–4;11)
N = 20
88%9%3%
5-year-olds P-L, P & R 2008a: S, E, S & V 2008: S 2000: Si 2008: C & P 2010:
5;02 (5;00–5;05)
N = 9
5;10 (4;11–5;11)
N = 9
5;06 (5;00–5;11)
N = 9
5;06 (5;00–5;11)
N = 21
5;00 (NA)
N = 30
53% 12% 30% 96% 4% 0% 91% 0% 9% 63% 20% 13% 73% 13% 0%
C et al. 2006: S & L 2010:
5;03 (5;00–5;06)
N = 19
5;01 (4;02–6;02)
N = 18
89% 6% 5% 97% NA NA
J, N, R & G 1998:
5;07 (5;06–5;11)
N = 20
79% 4% 17%
C et al. 2006:
5;09 (5;07–5;11)
N = 22
94% 4% 2%
10
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

6-year-olds C et al. 2006:
6;03 (6;00–6;05)
N =20
S & L 2010:
6;02 (4;07–8;03)
N = 12
Si 2008:
6;03 (6;00–6;05)
N = 10
90% 3% 7% 98% NA NA 69% 20% 6%
T et al. 2011:
6;07 (6;01–7;01)
N = 24
70% 10% 12%
G 2005:
6;07 (6;02–7;01)
N = 12
73% 7% 7%
J & N, Forthcoming:
6;08
N = 12
93% 0% 7%
Note.A(Avram2001); B & L (Bedore & Leonard 2001); B & M (Babyonyshev & Marin 2006); C (Castilla 2009
6
); C & L (Costa & Lobo 2006); C, L, C & S (Costa,
Lobo, Carmona & Silva 2008); C et al. (Chillier-Zesiger, Arabatzi, Baranzini, Cronel-Ohayon & Thierry 2006); G (Grüter 2005
7
); J & R (Jakubowicz & Rigaut 2000);
J & N (Jakubowicz & Nash, Forthcoming); J, N, R & G (Jakubowicz, Nash, Rigaut & Gerard 1998); M, V & W 2011 (Manika, Varlokosta & Wexler 2011); P-L, P &
R (Pérez-Leroux, Pirvulescu & Roberge 2008a
8
); S (Schaeffer 2000); Si (Silva 2008); S, E, S & V (Smith, Edwards, Stojanovik & Varlokosta 2008); S & L (Stavrakaki
& van der Lely 2010); T (Tedeschi 2009); T & W (Tsakali & Wexler 2004); T et al. (Tuller, Delage, Monjauze, Piller & Barthez 2011);W,G&T(Wexler,Gavarró&
Torrens 2004). NA = data not available.
6
Colombian Spanish.
7
Canadian French.
8
Canadian French.
11
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

12 VARLOKOSTA ET AL.
Schaeffer 2000 vs. Tedeschi 2009, for 3-year-olds; Romanian: Babyonyshev & Marin 2006 vs.
Avram 2001, for 3-year-olds). Sorting out which of all these differences are genuine is extremely
difficult. Not only are the SDs that accompany these means often very high (see references cited),
and the number of participants in some of them very small, but, more importantly, different tasks
were used.
Concerning experimental methodology, in the first studies conducted, the child was shown
some pictures, and prompts used to elicited clitic production were like that exemplified in (1):
(1) What is X doing to/with Y?
This was the method used by Jakubowicz and collaborators in various studies, by Avram
(1999), and, subsequently, by Smith et al. (2008), Stavrakaki & van der Lely (2010), and Tuller
et al. (2011). In Chillier et al. (2006) and Castilla (2009) the same kind of prompt was preceded
by a short story, building up to the outcome, as in (2):
(2) It’s late, it’s time to go to bed, there’s a blanket. What is papa doing to Pierre?
Schaeffer (1997, 2000) modified this method and introduced a second experimenter:
(3) Experimenter 1: What’s X doing?
Experimenter 2: I know what he is doing: He is verbing Y.
Experimenter 1: No, he’s not. You tell us: What is X doing to Y?
Schaeffer’s method was adopted by Babyonyshev & Marin (2004), Tsakali & Wexler (2004),
Wexler, Gavarró & Torrens (2004), Costa & Lobo (2006, 2007), Pérez-Leroux, Pirvulescu &
Roberge (2008a), and Manika, Varlokosta & Wexler (2011).
Bedore & Leonard (2001) resorted to a completion task, as exemplified in (4):
(4) Experimenter: The children wash the car and then ...
Child: [they] it-push.
All these elicitation methods share the property of focusing on the verb and rendering the
object referent as part of the background knowledge. However, they differ in the degree to
which the direct object referent is a well-established topic. The more salient the direct object
is in the discourse, the more likely it will be pronominalized, as opposed to being produced
as a full DP. On the other hand, in languages with a null object option, null objects are also
more likely to occur in this kind of context. Schaeffer’s method appears to have been partic-
ularly successful in eliciting pronominal answers instead of full DPs. Another feature all of
these experiments have in common is the fact that they not only elicit a direct object pronoun,
but they also elicit a subject pronoun. For Catalan, Greek, Italian, Portuguese, Romanian, and
Spanish, which are all (subject) pro-drop languages, this simply means that the elicited answer
will consist of the object clitic and the verb. In French, however, the elicited answer requires pro-
duction of a nominative clitic followed by an accusative clitic (a clitic cluster) followed by the
verb.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 13
Further variations in the elicitation method occur as a result of other language-specific prop-
erties. Thus, Costa & Lobo (2006) and Costa et al. (2008) added a strong island condition, for
reasons that are reviewed in Section 4. These experiments differ also as to the other elements
that were elicited besides definite third-person accusative clitics: Some experiments included
items eliciting reflexive clitics, or first- or second-person accusative clitics, and Pérez-Leroux,
Pirvulescu & Roberge (2008a) and Tedeschi (2009) included a “non-individuated” condition
(with a prompt, after correction as in the Schaeffer method, of “What did X do?” eliciting an
answer containing a new event “Bertrand is eating [the bone]”). Experiments with multiple con-
ditions could be argued to increase task difficulty, as, in any given condition, the child is in
the position of having to inhibit responses that are readily present due to another experimental
condition being present in the task.
9
Elicited production experiments specifically focused on how well children produce object
pronouns are rather sparse. Jakubowicz et al. (1996, 1997) found that object omission is very
frequent in the elicited production of German 2- and 3-year-olds with mean length of utter-
ance (MLU) below 3 (50% in three children with MLU < 3 in Jakubowicz et al. 1996; 55%
in seven children with MLU < 3 in Jakubowicz et al. 1997), and these omissions often result in
target-deviant constructions.
10
Ruigendijk et al. (2010) found also relatively high levels of object
omission in an elicitation experiment using pictures (“Here the woman wets the girl down, and
what does the woman do to the girl here? [she] dries/towels her”) with German-speaking 3-,
4-, 5-, and 6-year-olds (omission rates of 23%, 9%, 7%, and 10% respectively, and pronouns
produced at 40%, 55%, 76%, and 58%, and DPs at 3%, 13%, 9%, and 24%). These results for
German pronouns are in contrast with studies on pronoun production of English-, Hebrew-, and
Dutch-speaking children. De Villiers, Cahillane & Altreuter (2006) observed nearly flawless pro-
duction of pronouns, in their elicited production study, with English-speaking children between
4;06 and 7;02 years old (mean age 6;03).
11
Similarly, in their elicited production experiment,
Ruigendijk et al. (2010) and Novogrodsky, Balaban & Friedmann (2010) found that Hebrew-
speaking children produced very few omissions (4%, 13%, 2%, 4%, and 4% in 2-, 3-, 4-, 5-, and
6-year-olds respectively) and pronouns were produced at rates of 70% for the 2- and 3-year-olds,
and 87%–92% for the 4- to 6-year-olds (while DPs were generally marginal—3%–7%). Finally,
Spenader, Smits & Hendriks (2009) report that Dutch-speaking children aged 4;05 to 6;06 (mean
age 5;05) exhibited flawless pronoun production, as they rarely produced reflexives instead of
pronouns in the conditions that targeted a pronoun response (the f ocus of this study was on the
difference between pronouns and reflexives). However, it is worth noting that in some of these
9
Further discrepancies in the results are found in the studies using repetition (Eisenchlas 2003) or based in spontaneous
production, in which the criteria for determining when a clitic should be expected vary from study to study (see the results
in Guasti 1993/94; Lyczkowski 1999; Fujino & Sano 2002; Gavarró, Mata & Ribera 2006; Pirvulescu 2006, etc).
10
Similar conclusions are reached by Jakubowicz et al. (1996, 1997) on the basis of spontaneous language samples,
as well.
11
Although Hyams & Wexler (1993) and Valian (1991) report object omission in English before the age 4;06 (between
1;01 to 2;08 years [Valian 1991]; between 1;06 to 3;00 years [Hyams & Wexler 1993]) on the basis of spontaneous data,
both studies find that the level of object omission is not high compared to the level of subject omission observed in the
same period. For example, in Hyams & Wexler (1993), the proportion of missing subjects in Adam’s (between 2;05 to
3;00 years) and Eve’s (between 1;06 to 2;01 years) speech is 48%, compared to 8% of missing objects in the early period,
and 22% compared to 8% respectively in the late period.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

14 VARLOKOSTA ET AL.
TABLE 2
Elicited Production of Object Pronouns in Children Aged 2, 3, 4, 5, and 6: Age x Language x Study
(Mean Age, Age Range, and N of Children; % Pronouns, % Omission, and % Full DPs)
Dutch German Hebrew
2-year-olds J et al. 1996: R et al. 2010:
2;09 (2;05–3;01)
N = 3, MLU < 3
2;08 (2;04–2;11)
N = 7
5% 50% 45% 70% 4% 7%
2;06 (2;06–2;09)
N = 3, MLU > 3
55% 15% 30%
3-year-olds R et al. 2010:
3;06 (3;01–3;09)
N = 12
R et al. 2010:
3;05 (3;01–3;09)
N = 14
40% 23% 3% 70% 13% 4%
4-year-olds R et al. 2010: R et al. 2010:
4;06 (4;01–4;11)
N = 11
4;05 (4;00–4;10)
N = 15
55% 9% 13% 87% 2% 4%
5-year-olds S et al. 2009 R et al. 2010: R et al. 2010:
5;05 (4;05–6;06)
N = 83
∗
5;06 (5;01–5;11)
N = 11
5;06 (5;03–5;10)
N = 14
34% NA% 38% 76% 7% 9% 89% 4% 5%
6-year-olds R et al. 2010: R et al. 2010:
6;04 (6;01–6;08)
N = 10
6;04 (6;02–6;07)
N = 10
58% 10% 24% 91% 3% 4%
Note. R et al. (Ruigendijk, Friedmann, Novogrodsky & Balaban 2010); S et al. (Spenader, Smits & Hendriks
2009); J et al. (Jakubowicz, Müller, Kang, Riemer & Rigaut 1996); NA = data not available.
∗
Data for different age groups are not available.
conditions children produced a high proportion of full DPs rather than pronouns. This choice was
natural because their contexts presented two equally focused antecedents, and so using a pro-
noun would in fact be confusing. Table 2 illustrates the elicited production data from studies that
provided details (Jakubowicz et al. 1996; Ruigendijk et al. 2010; Spenader, Smits & Hendriks
2009).
To sum up, studies of elicited production of accusative clitics with 2- to 6-year-old children
show a great deal of heterogeneity, in methodology and in results, across languages and within
languages. Overall, it has been reported that accusative clitic production is quite high in some
languages from very early on (for example, in Greek), while in other languages this rate is at
least initially quite low (for example, in Catalan, French, and Italian). Studies of elicited pro-
noun production are quite few in number, and their results are often inconsistent: High rates
of object omission are observed in some of them, while very few omissions are reported in
others.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 15
3. SUMMARY OF THE SALIENT DISTRIBUTIONAL PROPERTIES
OF PRONOMINALS IN CLITIC AND PRONOUN LANGUAGES
In this section, we list some of the most salient distributional properties of the pronominal systems
of the languages tested in our cross-linguistic experiment.
Clitic languages are those languages that have clitic pronouns in addition to (strong and/or
weak) pronouns. The 11 clitic languages in our sample are Cypriot and (Standard Modern)
Greek; six Romance languages: Catalan, (European) Portuguese, French, Italian, Romanian,
and Spanish; and three Slavic languages: Croatian, Polish, and Serbian. We will only discuss
third-person accusative clitics/pronouns, as these are relevant for the elicitation task we have
created.
In all Romance languages and in (Cypriot) Greek, third-person clitics are usually
homophonous to definite determiners. Third-person singular clitics inflect for gender, number,
and case. Regarding placement, proclisis is generally observed with finite verbs and enclisis with
nonfinite ones and with affirmative imperative. Exceptions are Portuguese and Cypriot Greek,
where proclisis is syntactically determined. In French, proclisis is also observed with nonfinite
verbs. With negative imperative, proclisis or enclisis is observed, depending on the verbal form
used to express the imperative; in Italian, infinitives are used and both enclisis and proclisis are
possible; in Catalan and Spanish, the subjunctive is used and proclisis is evident; in French,
the clitic is proclitic in negative imperative (featuring the same order as with infinitives); in
Romanian, clitics are preverbal in infinitival structures introduced by a, and the clitic o is pre- or
postverbal depending on the verb form. In Italian, French, and Catalan, a clitic preceding the past
participle triggers participle agreement (agreement is not always audible in French or used when
audible). In addition to clitics, null objects are allowed in Portuguese, but these are not possible in
island contexts, like adverbial clauses. Some instances of null objects are observed in French, but
no island restriction applies in this case, and these instances of null objects appear to be lexically
determined. Null objects are also allowed in some environments (i.e., for inanimate objects) in
Basque Spanish (see Franco & Landa 2003). In Catalan, Portuguese, Italian, and Spanish, clitic
climbing is (optionally) observed with auxiliaries and restructuring verbs. Romanian (and some
varieties of Spanish) have clitic doubling, and finally, in Portuguese, mesoclisis (with future and
conditional) is found (for clitics in [Cypriot] Greek, see Agouraki 1997; Anagnostopoulou 1994;
Terzi 1999; for clitics in the Romance languages, see Belletti 1999; Kayne 1991; Sportiche 1996;
Uriagereka 1995). In Slavic languages, like Croatian and Serbian, clitics occur in second position,
and enclitic placement is typical. In Polish, clitics tend to occur in second position, but they can be
placed at different positions as well, due to the relatively free word order of the language; clitics
are canonically enclitics, but there is a tendency to find them in different positions (e.g., as procli-
tics) (for clitics in the Slavic languages, see Boškovi
´
c 2001;Browne1974; Dimitrova-Vulchanova
1999, Franks 2010; Franks & King 2000).
The five pronoun languages in our sample include four Germanic languages: German, Dutch,
Danish, and English; and the Semitic language Hebrew (see for Germanic languages, Cardinaletti
1999, Cardinaletti & Starke 1999; Laenzlinger & Shlonsky 1997; Schaeffer 2000;Zwart1991;
for Hebrew, Friedmann 2007). All of these languages have strong and weak pronouns, where the
difference is mainly in terms of presence versus absence of stress. Pronouns generally inflect for
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

16 VARLOKOSTA ET AL.
gender, number, and case. They are typically positioned where full DPs appear.
12
In Dutch and
German, pronouns scramble, while this process is more restricted for full DPs. Thus, in Dutch
main clauses, pronouns appear right-adjacent to the finite verb in verb-second position, and in
embedded clauses, they appear right-adjacent to the subject. In German, pronouns in embedded
clauses appear after the complementizer (second position or Wackernagel position),
13
while in
main clauses they are not necessarily adjacent to the verb in the second position.
4. METHODOLOGY
4.1. Participants
The participants were children with no diagnosed language, hearing, or speech pathologies. All
participants were monolingual native speakers of the language they were tested in. Generally,
20 children were tested for each language. The number of children, their ages, and their genders
are presented for each language in Table 3. We tested kindergarten and preschool children aged
5;00–5;11.
In order to establish what the target situation was, we also tested adult native controls in each
language: Catalan (27 adults), Croatian (10 adults), Cypriot Greek (8 adults), Danish (8 adults),
TABLE 3
Participants’ Information per Language
Language N
Age range
(months)
Mean age
(months)
SD age
(months) Girls Boys
Catalan 20 61–70 65 3 10 10
Croatian 25 60–70 63 3 12 13
Cypriot Greek 24 60–72 67 4 13 11
Danish 23 59–62 64 5 13 10
Dutch 20 61–72 67 4 12 8
English 19 63–71 67 3 11 8
French 25 62–70 66 3 13 12
German 22 59–69 64 3 11 11
Greek 20 60–71 65 4 11 9
Hebrew 21 59–71 64 4 8 13
Italian 20 61–71 67 4 6 14
Polish 31 61–71 68 3 15 16
Portuguese 20 60–71 66 3 13 7
Romanian 20 61–72 65 3 11 9
Serbian 25 60–71 67 3 10 15
Spanish 23 60–68 64 2 12 11
12
In Hebrew there are certain distributional differences between pronouns and full DPs. Specifically, unlike full DPs,
pronouns cannot appear postverbally after an unaccusative verb (Friedmann 2007).
13
It is actually weak/unstressed pronouns that appear after the complementizer in embedded clauses (Müller 2001;
Grohmann 1997; Laenzlinger & Shlonsky 1997).
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 17
Dutch (15 adults), English (5 adults), French (10 adults), German (8 adults), Greek (10 adults),
Hebrew (10 adults), Italian (13 adults), Polish (12 adults), Portuguese (10 adults), Romanian
(12 adults), Serbian (10 adults), and Spanish (20 adults). This was particularly necessary for the
languages in which the use of pronominals may be complicated by the presence of some other
option (e.g., null objects are an option in European Portuguese or in French, although under
different conditions).
4.2. Task and Procedure
To elicit object pronominals, we used a picture and a question that triggered a reply with a
because-clause containing a pronominal. The context strongly favors the use of a pronominal
by making the antecedent very salient in the discourse. First, a picture was presented in which
one animate character was performing an action on another character or object. The experimenter
described the picture in a sentence and then asked the child a why-question. For example, for
Figure 1, the experimenter’s introduction and the target sentence are given in (5). The corre-
sponding introductions and target clauses for this test item are provided for all languages in
Appendix A.
(5) Experimenter: “The boy is spraying the cat and the cat is wet. Why is the cat so wet?
The cat is wet ...”
Target response:“...because the boy is spraying him.”
The use of the because-clause was motivated by one of the tested languages, Portuguese, which
allows for null objects but not inside syntactic islands. To create a context in which a pronominal
is required (and a null object is ungrammatical)—even in Portuguese—we employed a strong
island, introduced by because. In such a context, according to Raposo (1986) and Costa & Lobo
(2006), null objects are ruled out; see example (6) from Costa & Lobo (2006).
FIGURE 1 Sample test picture.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

18 VARLOKOSTA ET AL.
(6) A: E a Maria?
what about Maria
B: O Pedro está triste porque o Zé
∗
(a) beijou.
the Pedro is sad because the Zé her
CL
kissed
The test started with two training items, which targeted object clitics or pronouns. Clitics
or pronouns were provided by the experimenter if the child did not produce them. No further
feedback was provided once the experimental session started.
Each participant was tested individually in a quiet room in the kindergarten/school or at home.
No time limit was imposed during testing, and no response-contingent feedback was given by the
experimenter, only general encouragement. The sessions were tape-recorded and transcribed in
full during the testing as well as after the test, by native speakers.
4.3. Materials
One of the most challenging tasks was to find a list of strongly transitive verbs that suited all the
languages studied and that could be used in sentences with direct objects. We specifically chose
verbs for which transitivity was not lost in the translation/adaptation across languages. The list
of the verbs used, short descriptions of the pictures, and the lead-in sentences for the English
experiment are presented in Appendix B. The list of the verbs used in the 16 languages are listed
in Appendix C.
The target sentences included singular pronouns of various genders (feminine, masculine, and
neuter, where applicable). Our pilot studies revealed that the gender of the target pronominal
element did not have any effect on clitic/pronoun production, and thus we decided not to bal-
ance for gender in the final testing, as this would have been very difficult, since gender is very
idiosyncratic across languages.
The pictures were controlled for cultural appropriateness. Before final testing, several pilot
studies were run to ensure that the pictures were clear and the verbs were appropriate. After
piloting, we changed some pictures and verbs that had elicited nontarget responses in some of the
languages, typically indirect objects and descriptions that did not use the target verb.
The test included 12 test sentences and five fillers. The fillers targeted sentences without object
pronouns.
4.4. Scoring
In clitic languages, we expected children to produce clitics, and in pronoun languages, we antic-
ipated use of pronouns (strong or weak). We did not distinguish between weak and strong
pronouns, as it was rather difficult to distinguish them consistently across different languages;
thus, weak and strong pronouns are collapsed together in the scoring. Alternative responses found
in the data were categorized in the following categories: use of the corresponding full noun phrase
(DP), failure to produce any direct object (i.e., omission), use of pronouns in clitic languages, and
production of other responses.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 19
We also scored errors of gender, number, and case. Sometimes participants changed the target
verb but employed a verb that still required a direct object. These were not counted as errors as
long as a clitic was produced in a clitic language or a pronoun in a pronoun language.
5. RESULTS
Responses were analyzed both in a quantitative and a qualitative way. Tables 4 and 5 provide the
mean percentages of scores in the clitic and pronoun languages respectively for the native adult
control data. The data in Tables 4 and 5 establish what the target situation was in each of the
languages tested.
Tables 6 and 7 provide the mean percentages of scores in the clitic and pronoun languages
respectively for the child data. Our test prompted production of pronominals, as the relevant ref-
erent was already mentioned in the previous discourse. As discussed earlier, pronominals take
different forms in various languages: clitics, weak or strong pronouns. In general, the most defi-
cient pronominal form of a language is used, unless the context requires otherwise. This means
that, in a language with clitics, these must be employed in situations like the one created by our
TABLE 4
Mean Percentages of Use of Clitics, DPs, Omission, Pronouns, and Other Structures in Clitic Languages
(Adult Data)
Language Clitics DPs Omission Pronouns Other
Catalan 99.7 0 0 0 0.3
Croatian 90.2 2.3 6.0 0 1.5
Cypriot Greek 100.0 0 0 0 0
French 93.3 6.7 0 0 0
Greek 98.4 0 0 0 1.6
Italian 91.0 0 0 0 9.0
Polish 84.0 14.6 0 0 1.4
Portuguese 88.3 1.7 3.3 0 6.7
Romanian 96.5 0 0 0 3.5
Serbian 97.5 2.5 0 0 0
Spanish 93.3 2.1 0 0 4.6
TABLE 5
Mean Percentages of Use of Pronouns, DPs, Omission, and Other Structures in Pronoun
Languages (Adult Data)
Language Pronouns DPs Omission Other
Danish 88.5 9.5 0 2.0
Dutch 89.0 10.0 0 1.0
English 95.0 5.0 0 0
German 92.7 2.1 1.0 4.2
Hebrew 98.4 0.8 0 0.8
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

20 VARLOKOSTA ET AL.
TABLE 6
Mean Percentages of Use of Clitics, DPs, Omission, Pronouns, and Other Structures in Clitic Languages
(Child Data)
Language Clitics (range; SD) DPs Omission Pronouns Other
Catalan 98.8 (92.0 − 100; 3) 0.4 0.0 0.0 0.8
Croatian 79.6 (33.0 − 100; 18) 5.0 6.4 1.0 7.7
Cypriot Greek 96.0 (83.0 − 100; 5) 0.7 0.6 0.0 2.7
French 90.0 (33.0 − 100; 19) 4.7 2.3 0.0 3.0
Greek 92.1 (67.0 − 100; 10) 2.5 0.4 4.2 0.8
Italian 92.9 (67.0 − 100; 10) 2.5 2.5 0.8 1.3
Polish 71.0 (0.0 − 100; 33) 3.0 18.3 3.0 4.8
Portuguese 19.3 (0.0 − 75; 25) 11.9 56.1 8.2 4.5
Romanian 90.8 (67.0 − 100; 11) 5.4 1.3 0.0 2.5
Serbian 77.7 (42.0 − 100; 14) 6.3 6.7 2.7 6.6
Spanish 94.2 (75.0 − 100; 7) 1.8 2.2 1.1 0.7
TABLE 7
Mean Percentages of Use of Pronouns, DPs, Omission, and Other Structures in Pronoun
Languages (Child Data)
Language Pronouns (range; SD) DPs Omission Other
Danish 70.0 (0.0 − 100; 30) 12.6 2.5 14.9
Dutch 66.2 (8.0 − 83.0; 25) 8.8 20.0 5.0
English 81.6 (25.0 − 100; 23) 18.0 0.0 0.4
German 89.0 (67.0 − 100; 11) 4.6 3.4 3.0
Hebrew 98.9 (92.0 − 100; 2) 0.4 0.0 0.7
task, while the use of pronouns would not be appropriate. In German, English, or Dutch, how-
ever, pronouns are expected to be employed, as these languages do not have clitics. Therefore,
our first analysis aimed at establishing whether children used the target pronominal form in their
language and whether they did so to the same extent in all languages investigated. This means
that, in clitic languages, clitics were the target form; in languages without clitics, weak or strong
pronouns were the target form, as indicated by the adult native control data in Tables 4 and 5
respectively.
To investigate the potentially different percentages of use for clitics versus pronouns, a t-test
was conducted comparing all clitic versus all pronoun languages. No significant difference was
observed, t(357) = 0.85, p = .30. The same percentage of pronominals was used in pronoun
languages (M = 81, SD = 24) and in clitic languages (M = 84, SD = 25). Thus, it is evident that
children at this age use the appropriate pronominal form at a high level.
We next compared the use of clitics among clitic languages with a one-way ANOVA with
languages as the independent variable and percentage of clitics as the dependent variable. This
revealed a significant difference, F(10, 242) = 33.68, p < .001, with further analysis, using a
post hoc Scheffé test revealed that the use of clitics is significantly lower in Portuguese than in all
the other languages, and it is lower in Polish than in Catalan, Cypriot Greek, Italian, and Spanish
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 21
(p < .001). Table 6 displays the mean percentages as well as the ranges and SDs of clitic use in
the clitic languages. It also reports the mean percentages of use of DPs, omission, pronouns, and
other structures.
In Portuguese, children mostly used null elements, some DPs, and pronouns. An ANOVA with
percentage of null elements as the dependent variable displays a significant difference among
clitic languages, F(10, 242) = 29.76, p < .001; a post hoc Scheffé test reveals that Portuguese
is different from all the other languages (p < .001), and Polish is also different from Catalan,
Cypriot Greek, French, and Greek (p < .05). A second ANOVA with percentage of DPs as the
dependent variable reveals a significant effect of language, F(10, 242) = 2.94, p < .01, due to
Portuguese differing from Catalan, Cypriot Greek, and Spanish in the use of DPs (p < .05). Then,
we compared the use of clitics in adult and child Portuguese and found that adults use more clitics
than children (M = 88, SD = 7 for adults and M = 19.3, SD = 25 for children), as confirmed
by an ANOVA with percentage of use of clitics as the dependent variable, F(1, 28) = 69.45, p <
.001. No difference was found in the use of clitics between children and adults in French (M =
90, SD = 16 for children and M = 93, SD = 10 for adults) and in Polish (M = 84, SD = 29 for
adults and M = 71, SD = 33 for children).
All children produced at least one clitic, except 9/20 children in Portuguese and 2/31 children
in Polish, who did not produce clitics at all. In most of the clitic languages, at least half of the
children produced 100% clitics: 11/20 children in Catalan, 14/24 in Cypriot Greek, 9/25 in
French, 10/20
in Greek, 11/20 in Italian, 9/20 in Romanian, and 15/23 in Spanish. In Polish,
Croatian, and Serbian, only 3/31, 5/25, and 3/25 children respectively produced 100% clitics.
Errors of gender, number, and case are very rare but are found in all languages except Cypriot
Greek. Their percentages range from 0% to 12% errors at most, calculated on the number of clitics
used. An ANOVA with percentages of errors in clitic languages as the dependent variable yields a
significant effect of language, F(10, 242) = 8.7, p < .001, and a post hoc Scheffé test revealed that
Greek has a significantly higher occurrence of errors than all other languages (p < .05), except
Serbian (for Greek see Varlokosta, Konstantzou & Nerantzini 2014). Errors in the placement of
clitics were found in Croatian (1.2%), Cypriot Greek (49%),
14
and Portuguese (72%).
Table 7 displays the mean percentages of use of pronouns, DPs, omission, and other struc-
tures in pronoun languages. The use of pronouns in pronoun languages was also not uniform,
with Danish and Dutch displaying the lowest use. A one-way ANOVA on the percentage of
pronoun use confirmed this observation, F(4, 101) = 8.48, p = .001, with Dutch differing sig-
nificantly from German (p = .02) and Hebrew (p < .001), and Danish differing from Hebrew
(p < .001). When children did not use pronouns, they used DPs, especially in English, as shown
by the ANOVA of the percentage of use of DPs, F(4, 101) = 3.50, p = .01. This variation
among languages in the use of DPs is essentially due to English being significantly different
from Hebrew (p = .02). Occasionally, children omitted pronouns altogether, especially in Dutch.
14
For further discussion of enclitic Cypriot Greek, in particular the diglossic situation in Cyprus and the differences
in placement to proclitic Standard Modern Greek, see Grohmann (2011). As more comprehensive testing with younger
and older children showed (Grohmann et al. 2012), the nontarget proclitic placement is not misplacement as such, but
rather a reflex of the complex sociolinguistic situation. This is approached under the umbrella term of “socio-syntax of
development” through competing motivations between Cypriot and Standard Modern Greek with the onset of schooling
by Grohmann & Leivada (2012).
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

22 VARLOKOSTA ET AL.
In this case, the ANOVA with pronoun omission as the dependent variable yielded a signif-
icant difference, F(4, 101) = 12.61, p < .001, as Dutch differs from all the other languages
(p < .001). In Dutch, some instances of omission of pronouns were grammatical in the adult lan-
guage, as some verbs are optionally transitive (kammen ‘comb,’ verven ‘paint,’ and likken ‘lick,’
as reported in van Hout, Veenstra & Berends 2011). However, the removal of these grammatical
omissions, with the percentage of ungrammatical omissions dropping from 20% to 12.5%, does
not affect the results, F(4, 101) = 7.5, p < .001, and Dutch is still significantly different from the
other languages (p < .01).
15
The number of children who produced 100% pronouns was high in Hebrew (19/21 children)
but lower in German (9/22), English (6/19), and Danish (5/23). All children produced pronouns,
except one child in Danish, who did not produce any, and two children in Dutch, who had a very
low pronoun production rate (8%).
Errors of gender, number, and case were also found, with differences among languages, mostly
in German. Their percentage ranges from 0% to 19% errors at most, calculated on the number
of pronoun uses. The ANOVA with percentage of errors as the dependent variable yielded a
significant difference, F(4, 101) = 23.81, p < .001, due to German, which differed from all the
other languages (post hoc Scheffé test p < .001).
16
Errors in the placement of pronouns were rare
and were observed only in Dutch (2.5%).
Given that there was some variation in the use of clitics among some clitic languages and in
the use of pronouns among some pronoun languages, we conducted an ANOVA with verb type as
independent variable and clitic/pronoun (production) as dependent variable to see whether there
are differences in the production of clitics/pronouns depending on the verb used, and if there
are, whether they are the same cross-linguistically. In this analysis we included only languages in
which the production of clitics/pronouns was less than 90% (Croatian, Danish, Dutch, English,
German, Polish, Portuguese, and Serbian). The analysis indicated no effect of verb type, F(11,
72) = .236, p = .994. By looking at the data, one may notice that the verb paint elicited less
clitic and pronoun production in Polish and Danish respectively, the verb eat elicited less clitic
production in Serbian, and the verb comb elicited less pronoun production in Dutch. Crucially,
though, none of these verbs elicited less clitic/pronoun production in the other languages, as
indicated by the statistical analysis.
In summary, the target pronominal form is used frequently by children. In Portuguese
clitics are used only half of the time; Portuguese-speaking children mostly use null elements.
Differences are observed also among pronoun languages, with Danish- and Dutch-speaking chil-
dren using fewer pronouns than children speaking other languages. In Dutch, a relatively high
use of null elements was observed, while in Danish a range of structures were used, mostly DPs,
but their use was not significantly different. Errors of gender, number, and case were rare, and
were mostly found in Polish and German. Errors in the placement of clitics were frequently
15
van Hout, Veenstra & Berends (2011) analyzed the Dutch omission data in more detail. A subject analysis shows
that there was quite some individual variation. About half of the participants did not omit objects or omitted them only
once, whereas four participants omitted objects at high rates. The authors do not have any further linguistic information
about these four participants.
16
Some of the gender errors in German were also observed in the adult data and were due to the mismatch between
grammatical and semantic gender in the noun girl (grammatical: neuter; semantic: feminine) in the test sentences with
the verbs comb and draw.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 23
observed in Cypriot Greek (but see fn. 8) and Portuguese, while the placement of pronouns was
less problematic, with some rare errors found only in Dutch.
6. DISCUSSION
The results of the production experiment across 16 languages reveal some general common ten-
dencies and also a few cross-linguistic differences. First, it is clear that at the age tested, children
are able to produce pronominal elements. This is evident by the high rates of clitics and pronouns
produced according to the target in all languages. This massive production of pronominals is
revealing in three ways:
i. It shows that there is a general cross-linguistic consistency in the results, in the sense that
it is possible to say that, at age 5, children are able to produce pronominals in all the
languages tested.
ii. It shows that children have the relevant morphosyntactic knowledge involved in the pro-
duction of pronominals; children know the placement and distribution of pronominals, and
most of the time opt for the target case form.
iii. It shows that children have the relevant pragmatic knowledge required to select a pronom-
inal in the discourse setting involved in the experiment; this is shown by the fact that
children do not freely vary between pronominals and DPs or forms that would be
pragmatically infelicitous.
In general then, it is legitimate to say that, at age 5, pronominal elements are acquired across
languages.
However, some cross-linguistic differences were found, as well. Pronominals were produced in
all languages but at different rates. The different rates make it possible to establish three language
groups: languages with clitics, languages with strong/weak pronouns, and languages with null
objects. Clitic and pronoun languages are those in which pronominals are produced at high and
similar rates. In Portuguese, which has null objects, the rate of clitic production is quite low,
and null complements are the preferred option. Two main observations can be drawn from these
cross-linguistic differences:
i. No difference was found between clitics and pronouns in production. This contrasts
with the literature on pronominal comprehension in which a difference has been estab-
lished between clitics and pronouns (see the Delay of Principle B Effect or the Pronoun
Interpretation Problem; e.g., McKee 1992; Baauw, Coopmans & Philip 1999; Baauw
et al. 2011;Varlokosta2000, among others) (nevertheless, see the discussions in Baauw
et al. 2011 and Conroy et al. 2009 for the possibility that experimental design and
choice of experimental materials and tasks may have affected the results observed in the
comprehension studies of pronominals).
ii. The fact that null objects are preferred in Portuguese confirms that, in this language, chil-
dren overuse the weakest form of the nominal hierarchy (i.e., null objects), producing
them even in a context in which the target grammar rules them out (Costa, Lobo & Silva
2009). However, it is worth emphasizing that there does not seem to be a problem with
pronominals in this language—they are produced correctly at a moderate rate. The only
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

24 VARLOKOSTA ET AL.
problem is the preference for an alternative strategy that is available in the language—the
null object option.
17
The comparison between the three language groups, including the observation that, in
Portuguese, children prefer the null-complement option, reveals a striking cross-linguistic pat-
tern in spite of the differences: Across languages, children consistently opt for the weakest form
available to them in the language they are acquiring. This is particularly clear in a language like
Italian, which has strong, weak, and clitic pronouns, and yet children go for the weakest clitic
form, or in a language like Portuguese, which has strong and clitic pronouns and null objects,
and children opt for the latter. This cross-linguistic preference for the weakest form is interesting,
since it again reveals that children know the relevant pragmatics associated with the discourse
context and that behavior is very consistent across languages. The option for the weakest form
is a sign of good pragmatic knowledge, because it indicates that children know that given ref-
erents are best mapped onto nonsalient expressions (Ariel 1990; Gundel, Hedberg & Zacharski
1993). The cross-linguistically consistent behavior reveals not only that the differences found are
superficial but also that children exhibit good knowledge of the pronominal paradigms in their
languages and of the set of alternatives that could be used in the context of a pronominal. Note
that, in order to select the weakest form (including the null object in Portuguese), one must accept
that this form competes with the stronger alternatives. Moreover, note that our claim that children
opt for the weakest form available in the language they are acquiring entails that we expect cross-
linguistic variation in the forms used by children across languages, as indicated by the results of
the experiment we report here.
Related to the point just made that children know the paradigms in their languages, it is inter-
esting to comment on their knowledge about null alternatives. Tuller (2000) and Cummins &
Roberge (2005) show that French has null complements. Null objects in French, however, appear
to be different from those available in Portuguese. In French, null objects are lexically constrained
and not licensed in the pragmatic context of the present test. Interestingly, French children did
not opt for a null complement in the context of this experiment. This indicates that children know
the pragmatics associated with the empty categories available to them, as well as the lexically
determined limitations of their distribution, whenever this is relevant, as in French. They may
fail to master the syntactic distribution of the null complement, as evidenced by the overuse of
null objects in strong islands by Portuguese-speaking children, but they do not randomly replace
obligatory pronominals by any available null expression in the language. Again, this leads to the
conclusion that their pragmatic knowledge is highly accurate at age 5.
Topic-drop is another construction that may license a null complement. In topic-drop, the
empty complement must be in sentence-initial position, which does not happen in the context
of elicitation of the present experiment, since we used embedded adverbial subordinate clauses.
The fact that, in Dutch, there were some instances of null complements may signal that Dutch
children occasionally overuse the topic-drop construction, in the same way Portuguese-speaking
children overuse the null object construction.
17
Children acquiring Portuguese are not adultlike. According to Costa & Lobo (2011), this may be due to the fact that
a late acquisition of the properties distinguishing pro and variables delays a steady knowledge of the context in which null
objects are ruled out. If the null object is not interpreted like a variable, it can be used in contexts in which pronominals
are accepted, such as islands.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 25
In short, the comparison between the three types of null-complement possibilities (lexically
restricted, as in French; null object, as in Portuguese; topic-drop, as in Dutch) suggests that chil-
dren know the specific properties of the null complements available in their languages but may
still overuse them in inadequate syntactic contexts, although not in inadequate pragmatic contexts,
which explains the good performance of the French children.
Summarizing, in this discussion we provided arguments that: (i) 5-year-old children’s knowl-
edge of pronominals is target-consistent across languages; (ii) children consistently opt for the
weakest alternative, in accordance to the scale pronoun > clitic > null, depending on what is
available in their languages;
18
(iii) children reveal a very accurate knowledge of the alternative
options for the discourse setting of the experiment; (iv) the preference for the weakest form
in languages with a null complement may disguise the fact that children are quite competent
in producing pronominals; (v) children’s knowledge of null-complement typologies is good,
although they sometimes misuse them extending null complements to inappropriate syntactic
environments.
We have been able to reach these conclusions through use of a common methodology in 16 lan-
guages. Our method involved use of a single task that did not produce any noise. The task was
short, eliciting a specific construction and not a variety of structures, thus freeing the child from
possible task-related processing difficulties, and at the same time allowing us to tap into their
knowledge more directly.
The overall results and the clarifications made regarding the nature of null complements—in
particular the clarification that the production of null complements does not mean that children do
not know pronominals—allow us to draw a firm conclusion: Pronominal production can be taken
as a cross-linguistic robust marker of language development. We see that 5-year-old children
produce third-person pronominal objects and that the shape and placement of pronominal forms
is targetlike. The ability to produce pronominals may be hidden by massive production of null
complements in languages allowing for it. However, it is legitimate to conclude from our data
that a child who at age 5 is not able to produce any or few pronominal forms is a child at risk
for language impairment. In this way, pronominal production can be taken as a developmental
marker, provided that one takes into account whether the language allows for null objects.
ACKNOWLEDGMENTS
We would like to thank two anonymous reviewers and the Guest Editor of Language Acquisition
for useful comments that helped us improve our paper.
18
One anonymous reviewer raised the issue that the finding (ii) could be an artifact of the restriction of our exper-
iment to third-person pronominals, and another result may arise if first- and second-person pronouns/clitics, including
reflexives, were tested with a different design. An indication that this finding is not an artifact of the restriction to third-
person pronominals is that Costa et al. (2008) elicited dative and reflexive clitics in first- and second-person in European
Portuguese, where the null option is not allowed, with a similar design to the design used in this study. It is hard to say
whether our findings for third-person pronominals and the findings in Costa et al. (2008) for first- and second-person
dative and reflexive clitics are the automatic outcome of the design used, namely elicitation tasks with pictures, as we are
not aware of other designs used to elicit pronouns/clitics.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

26 VARLOKOSTA ET AL.
FUNDING
Kleanthes K. Grohmann acknowledges financial support from the University of Cyprus for
the Gen-CHILD Project (8037-61017). The Danish study was supported by a grant from The
Danish Agency for Science and Technology and Innovation to Kristine Jensen de López in
support of the NASUD project Normal og Atypisk Sproglig Udvikling (Typical and Atypical
Language Development) grant # 09-063957. The Serbian study was supported by the Ministry of
Education and Science, Republic of Serbia, Project Fundamentalni kognitivni procesi i funkcije
(Fundamental Cognitive Processes and Functions), Grant # 179033. The Hebrew study was sup-
ported by the Lieselotte Adler Laboratory for Research on Child Development and by GIF grant
1113-97.4/2010. The Polish study was partially supported by public funding from the Faculty
of Psychology of Warsaw University (grant no BST 184724/09, awarded to Aneta Mie¸kisz)
and from the National Science Centre/Ministry of Science and Higher Education (grants no N
N106 223538, awarded to Magdalena Smoczy
´
nska and no 809/N-COST/2010/0, awarded to Ewa
Haman). Last, María-José Ezeizabarrena acknowledges financial support from the Ministerio de
Ciencia e Innovación for the Spanish data (Grant FFI 2012-37884-C03-02) and from The Basque
Government (IT-676-13).
REFERENCES
Agouraki, Yiorgia. 1997. On the enclisis/proclisis alternation. In Gaberell Drachman, Angeliki Malikouti-Drachman,
John Fykias & Ceila Klidi (eds.), Proceedings of the 2nd International Conference on Greek Linguistics [ICGL 2],
393–404. Salzburg: University of Salzburg.
Anagnostopoulou, Elena. 1994. Clitic dependencies in Modern Greek. Salzburg: University of Salzburg dissertation.
Ariel, Mira. 1990. Accessing noun-phrase antecedents. London: Routledge.
Avram, Larisa. 1999. Clitic omission in child language and multiple spell-out. Poster presented at the Generative
Approaches to Language Acquisition [GALA 1999], September 10–12. University of Potsdam.
Avram, Larisa. 2001. Remarks on the optional clitic stage in child Romanian. Bucharest Working Papers in Linguistics
3(1). 16–28.
Baauw, Sergio, Peter Coopmans & William Philip. 1999. The acquisition of pronominal coreference in Spanish and the
clitic-pronoun distinction. In Jan Don & Ted Sanders (eds.), UiL-OTS Yearbook 1998-1999, 1–19. Utrecht University.
Baauw, Sergio, Shalom Zuckerman, Esther Ruigendijk & Sergey Avrutin. 2011. Principle B delays as a processing prob-
lem: Evidence from task effects. In Angela Grimm, Anja Müller, Cornelia Hamann & Esther Ruigendijk (eds.),
Production-comprehension asymmetries in child language, 247–272. Berlin: Mouton De Gruyter.
Babyonyshev, Maria & Stefania Marin. 2004. Object clitics in child Romanian. In Alejna Brugos, Manuella R.
Clark-Cotton & Seungwan Ha (eds.), Proceedings of the 29th annual Boston University Conference on Language
Development [BUCLD 29], 49–60. Somerville, MA: Cascadilla Press.
Babyonyshev, Maria & Stefania Marin. 2006. Acquisition of Romanian pronominal clitics. In Anna Gavarró & Conxita
Lleó (eds.), Catalan Journal of Linguistics 5. 17–44.
Bedore, Lisa M. & Leonard B. Laurence. 2001. Grammatical morphology deficits in Spanish-speaking children with
specific language impairment. Journal of Speech, Language, and Hearing Research 44. 905–924.
Belletti, Adriana. 1999. Italian/Romance clitics: Structure and derivation. In Henk van Riemsdijk (ed.), Clitics in the
languages of Europe, 543–579. Berlin: Mouton De Gruyter.
Bloom, Paul. 1990. Subjectless sentences in child language. Linguistic Inquiry 21(4). 491–504.
Bortolini, Umberta, Barbara Arfé, Maria Cristina Caselli, Luisa Degasperi, Patricia Deevy & Laurence B. Leonard. 2006.
Clinical markers for specific language impairment in Italian: The contribution of clitics and non-word repetition.
International Journal of Language & Communication Disorders 41(6). 695–712.
Boškovi
´
c, Željko. 2001. On the nature of the syntax-phonology interface: Cliticization and related phenomena.North-
Holland Linguistic Series: Linguistic Variations 60. Amsterdam: Elsevier.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 27
Browne, Wayles. 1974. On the problem of enclitic placement in Serbo-Croatian. In Richard D. Brecht & Catherine
Chvany (eds.), Slavic transformational syntax, 36–52. Ann Arbor MI: Michigan Slavic Materials.
Cardinaletti, Anna 1999. Pronouns in Germanic and Romance languages: An overview. In Henk van Riemsdijk (ed.),
Clitics in the languages of Europe, 33–82. Berlin: Mouton de Gruyter.
Cardinaletti, Anna & Michal Starke. 1999. The typology of structural deficiency: A case study of the three classes of
pronouns. In Henk van Riemsdijk (ed.), Clitics in the languages of Europe, 145–233. Berlin: Mouton de Gruyter.
Castilla, Anny P. 2009. Morphosyntactic acquisition in monolingual 3-, 4-, and 5-year-old Spanish-speaking children. In
Victoria Marrero & Idaira Pinera (eds.), Linguistics: The challenge of clinical application. Proceedings of the 2nd
International Conference on Clinical Linguistics [ICCL 2], 527–533. Madrid, Spain: Euphonia Ediciones.
Castilla, Anny P., Ana Teresa Pérez-Leroux & Alice Eriks-Brophy. 2008. Syntax and the lexicon in early omission of
Spanish clitics. In Harvey Chan, Heather Jacob & Enkeleida Kapi (eds.), Proceedings of the 32nd annual Boston
University Conference on Language Development. [BUCLD 32], 72–83. Somerville, MA: Cascadilla Press.
Castilla, Anny P. & Ana Teresa Pérez-Leroux. 2010. Omissions and substitutions in Spanish object clitics: Developmental
optionality as a property of the representational system. Language Acquisition 17. 2–25.
Chien, Yu-Chin & Kenneth Wexler. 1990. Children’s knowledge of locality conditions in binding as evidence for the
modularity of syntax and pragmatics. Language Acquisition 1. 225–295.
Chillier-Zesiger, Laurence, Marina Arabatzi, Lara Baranzini, Stephany Cronel-Ohayon & Deonna Thierry. 2006. The
acquisition of French pronouns in normal children and in children with specific languageimpairment (SLI) .Ms.
University of Geneva.
Conroy, Anastasia, Eri Takahashi, Jeffrey Lidz & Colin Phillips. 2009. Equal treatment for all antecedents: How children
succeed with principle B. Linguistic Inquiry 40. 446–486.
Costa, João & Maria Lobo. 2006. A aquisição de clíticos em PE: omissão de clíticos ou objecto nulo? [The acquisition
of clitics in EP: Omission of clitics or null objects?]. In Fatima Oliveira & Joaquim Barbosa (eds.), XXI Encontro
Nacional da Associação Portuguesa de Linguística. Textos selecionados, 285–293. Lisboa: Associação Portuguesa de
Linguística.
Costa, João & Maria Lobo. 2007. Clitic omission, null objects or both in the acquisition of European Portuguese. In
Sergio Baauw, Frank Drijkonongen & Manuela Pinto (eds.), Romance Languages and Linguistic Theory 2005, 59–72.
Amsterdam: John Benjamins.
Costa, João & Maria Lobo. 2011. Assessing children’s knowledge of null objects in European Portuguese. Paper presented
at the Linguistic Symposium on Romance Languages [LSRL 41], May 5–7, University of Ottawa, Canada.
Costa, João, Maria Lobo, Jaqueline Carmona & Carolina Silva. 2008. Clitic omission in European Portuguese: Correlation
with null objects? In Anna Gavarrò & M. João Freitas (eds.), Language acquisition and development. Proceedings
of Generative Approaches of Language Acquisition 2007 [GALA 2007], 133–143. Newcastle-upon-Tyne: Cambridge
Scholars Publishing.
Costa, João, Maria Lobo & Carolina Silva. 2009. Null objects and early pragmatics in the acquisition of European
Portuguese. Probus 21. 143–162.
Cummins, Sarah &Yves Roberge. 2005. A modular account of null objects in French. Syntax 8. 44–64.
De Villiers, Jill, Jaqueline Cahillane & Emily Altreuter. 2006. What can production reveal about principle B? In Kamil
Ud Deen, Jun Nomura, Barbara Schulz & Bonnie D. Schwartz (eds.), Proceedings of the Inaugural Conference on
Generative Approaches to Language Acquisition – North America, 89–100. Honolulu, HI: University of Connecticut
Occasional Papers in Linguistics 4.
Dimitrova-Vulchanova, Mila 1999. Clitics in the Slavic languages. In Henk van Riemsdijk (ed.), Clitics in the languages
of Europe, 83–122. Berlin: Mouton de Gruyter.
Duarte, Inês & Gabriela Matos. 2000. Romance clitics and the minimalist program. In João Costa (ed.), Portuguese
syntax: New comparative studies, 116–142. New York: Oxford University Press.
Eisenchlas, Susana. 2003. Clitics in child Spanish. First Language 23(2). 193–211.
Ezeizabarrena, María-José. 1996. Adquisición de la morfología verbal en euskera y castellano por niños bilingües
[Bilingual children’s acquisition of Basque and Spanish verb morphology]. Bilbao: University of the Basque Country
dissertation.
Ezeizabarrena, María-José. 1997. Morfemas de concordancia con el sujeto y los objetos en el castellano infantil
[Subject and verb agreement morphemes in early Spanish]. In Ana Teresa Pérez-Leroux & William R. Glass (eds.),
Contemporary perspectives on the acquisition of Spanish, 21–36. Somerville, MA: Cascadilla Press.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

28 VARLOKOSTA ET AL.
Franco, Jon & Alazne Landa. 2003. Null objects revisited. In Rafael Nuñez-Cedeño, Luis López & Richard Cameron
(eds.), A Romance perspective on language knowledge and use: Selected papers for the 31st Linguistic Symposium
on Romance Languages (LSRL), Chicago, 19-22 April 2001, 311–26. Amsterdam: John Benjamins.
Franks, Steven. 2010. Clitics in Slavic. Glossos. The Slavic and East European Language Resource Center 10. 1–157.
Franks, Steven & Tracy Holloway King. 2000. A handbook of Slavic clitics. New York: Oxford University Press.
Friedmann, Naama. 2007. Young children and A-chains: The acquisition of Hebrew unaccusatives. Language Acquisition
14(4). 377–422.
Fujino, Hanako & Tetsuya Sano. 2002. Aspects of the null object phenomenon in child Spanish. In Ana Teresa Pérez-
Leroux & Juana Muñoz Liceras (eds.), The acquisition of Spanish morphosyntax, 67–88. Dordrecht: Kluwer.
Gabriel, Christoph & Natascha Müller. 2005. Zu den romanischen Pronominalklitika: Kategorialer Status und syntaktis-
che Derivation [On Romance pronominal clitics: Categorial status and syntactic derivation]. In Georg Kaiser (ed.),
Deutsche Romanistik – generativ, 161–180. Tübingen: Narr.
Gavarró, Anna, Meritxell Mata & Eulàlia Ribera. 2006. L’omissió dels clítics d’objecte i partitius en el català infantil:
dades espontànies [Omission of object and partitive clitcs in child Catalan: spontaneous data]. In Claus D. Pusch
(ed.), Zeitschrift für katalanistik, special volume La gramàtica pronominal del català – Variació, evolució, funció
[The grammar of Catalan pronouns – Variation, evolution, function], 27–46. Aachen: Shaker Verlag.
Gavarró, Anna & Marta Mosella. 2009. Testing syntactic and pragmatic accounts of clitic omission. In Jean Crawford,
Koichi Otaki & Masahiko Takahashi (eds.), Proceedings of the 3rd Conference on Generative Approaches to
Language Acquisition North America [GALANA 3], 25–35. Somerville, MA: Cascadilla Proceedings Project.
Gavarró, Anna, Vicent Torrens & Kenneth Wexler. 2010. Object clitic omission: Two language types. Language
Acquisition 17(4). 192–219.
Gerken, Lou Ann. 1996. Phonological and distributional information in syntax acquisition. In James L. Morgan &
Katherine Demuth (eds.), Signal to syntax: Bootstrapping from speech to grammar in early acquisition , 411–425.
Mahwah, NJ: Lawrence Erlbaum Associates.
Grohmann, Kleanthes K. 1997. Pronouns and the left periphery of West Germanic embedded clauses. In Abraham Werner
& Elly van Gelderen (eds.), German: Syntactic problems — Problematic syntax, 163–189. Tübingen: Niemeyer.
Grohmann, Kleanthes K. 2011. Some directions for the systematic investigation of the acquisition of Cypriot Greek: A
new perspective on production abilities from object clitic placement. In Esther Rinke and Tanja Kupisch (eds.), The
development of grammar: Language acquisition and diachronic change, 179–203. Amsterdam: John Benjamins.
Grohmann, Kleanthes K. & Evelina Leivada. 2012. Interface ingredients of dialect design: Bi-x, socio-syntax of develop-
ment, and the grammar of Cypriot Greek. In Anna Maria Di Sciullo (ed.), Towards a biolinguistic understanding of
grammar: Essays on interfaces, 239–262. Amsterdam: John Benjamins.
Grohmann, Kleanthes K., Eleni Theodorou, Natalia Pavlou, Evelina Leivada, Elena Papadopoulou & Silvia Martínez-
Ferreiro. 2012. The development of object clitic placement in Cypriot Greek and the Romance connection. In Sadrine
Ferré, Philippe Prévost, Laurice Tuller & Rasha Zebib (eds.), Selected Proceedings of the Romance Turn IV, 128–152.
Newcastle-upon-Tyne: Cambridge Scholars Publishing.
Grüter, Theres. 2005. Comprehension and production of French object clitics by child second language learners and
children with specific language impairment. Applied Psycholinguistics 26. 363–391.
Grüter, Theres. 2006. Object clitics and null objects in the acquisition of French. Montréal: McGill University dissertation.
Grüter, Theres. 2007. Investigating object drop in child French and English: A truth value judgment task. In Belikova
Alyona, Luisa Meroni & Mari Umeda (eds.), Proceedings of the 2nd Conference on Generative Approaches to
Language Acquisition North America [GALANA 2], 102–113. Somerville, MA: Cascadilla Proceedings Project.
Guasti, Maria Teresa. 1993-1994. Verb syntax in Italian child grammar: Finite and nonfinite verbs. Language Acquisition
3. 1–40.
Gundel, Jeanette K., Nancy Hedberg & Ron Zacharski. 1993. Cognitive status and the form of refeering expressions in
discourse. Language 69. 274–307.
Hamann, Cornelia & Adriana Belletti. 2006. Developmental patterns in the acquisition of complement clitic pronouns.
Comparing different acquisition modes with an emphasis on French. Rivista di Grammatica Generativa 31. 39–78.
Hamann, Cornelia, Luigi Rizzi & Uli H. Frauenfelder. 1996. On the acquisition of the pronominal system in French. In
Harald Clahsen (ed.), Generative perspectives on language acquisition, 309–334. Amsterdam: John Benjamins.
van Hout, Angeliek, Alma Veenstra & Sanne Berends. 2011. All pronouns are not acquired equally in Dutch: Elicitation
of object and quantitative pronouns. In Mihaela Pirvulescu, María Cristina Cuervo, Ana Teresa Pérez-Leroux, Jeffrey
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 29
Steele & Nelleke Strik (eds.), Selected Proceedings of the 4th Conference on Generative Approaches to Language
Acquisition North America [GALANA 4], 106–121. Somerville MA: Cascadilla Proceedings Project.
Hyams, Nina & Kenneth Wexler. 1993. The grammatical basis of null subjects in child language. Linguistic Inquiry 24.
421–459.
Ilic, Tatjana & Kamil Ud Deen. 2004. Object raising and cliticization in Serbo-Croatian child language. In Jacqueline
van Kampen & Sergio Baauw (eds.), Proceedings of 5th Generative Approaches to Language Acquisition [GALA 5],
235–243. LOT.
Jakubowicz, Celia, Natascha Müller, Ok-Kyung Kang, Beate Riemer & Catherine Rigaut. 1996. On the acquisition of the
pronominal system in French and German. In Andy Stringfellow, Dalia Cahana-Amitay, Elizabeth Hughes &Andrea
Zukowski (eds.), Proceedings of the 20th annual Boston University Conference on Language Development [BUCLD
20], 374–385. Somerville MA: Cascadilla Press.
Jakubowicz, Celia & Lea Nash. Forthcoming. Why accusative clitics are avoided in normal and impaired language devel-
opment. In Celia Jakubowicz, Lea Nash & Kenneth Wexler (eds.), Essays in syntax, morphology and phonology in
SLI. Cambridge, MA: MIT Press.
Jakubowicz, Celia, Natascha Müller, Beate Riemer & Catherine Rigaut. 1997. The case of subject and object omission in
French and German. In Elizabeth Hughes, Mary Hughes & Annabel Greenhill (eds.), Proceedings of the 21st annual
Boston University Conference on Language Development [BUCLD 21], 331–342. Somerville MA: Cascadilla Press.
Jakubowicz, Celia, Lea Nash, Catherine Rigaut & Christophe-Loic Gerard. 1998. Determiners and clitic pronouns in
French-speaking children with SLI. Language Acquisition 7. 113–160.
Jakubowicz, Celia & Catherine Rigaut. 2000. L’acquisition des clitiques nominatifs et des clitiques objets en Français
[The acquisition of nominative and object clitics in French]. The Canadian Journal of Linguistics 45. 119–158.
Kayne, Richard. 1975. French syntax. Cambridge, MA: MIT Press.
Kayne, Richard.1991. Romance clitics, verb movement, and PRO. Linguistic Inquiry 22. 647–686 (reprinted in: Richard
Kayne, 2000. Parameters and universals, 60-97. Oxford: Oxford University Press).
Laenzlinger, Christopher & Ur Shlonsky. 1997. Weak pronouns as LF clitics: Clustering and adjacency effects in the
pronominal systems of German and Hebrew. Studia Linguistica 51(2). 154–185.
Larrañaga, María Pilar. 2000. Ergative Sprachen, akkusative Sprachen. Der Erwerb des Kasus bei bilingualen Kindern
[Ergative languages, accusative languages: The acquisition of case in bilingual children]. Frankfurt am Main:
Vervuert.
Larrañaga, María Pilar & Pedro Guijarro-Fuentes. 2011. Clitics in L1 bilingual acquisition. First Language 31(4). 1–25.
Lyczkowski, David. 1999. Adquiéretelo: On the acquisition of pronominal object clitics in Spanish. Cambridge,
Massachusetts: Harvard University MA thesis.
Marinis, Theodoros. 2000. The acquisition of clitic objects in Modern Greek: Single clitics, clitic doubling, clitic left
dislocation. ZAS Working Papers 15. 259–281.
Manika, Sophia, Spyridoula Varlokosta & Kenneth Wexler. 2011. The lack of omission of clitics in Greek children with
SLI: An experimental study. In Nick Danis, Kate Mesh & Hyunsuk Sung (eds.), Proceedings of the 35th annual
Boston University Conference on Language Development [BUCLD35], 427–439. Somerville, MA: C ascadilla Press.
McKee, Cecile. 1992. A comparison of pronouns and anaphors in Italian and English acquisition. Language Acquisition
2. 21–54.
Müller, Gereon. 2001. Harmonic alignment and the hierarchy of pronouns in German. In Horst J. Simon & Heike Wiese
(eds.), Pronouns – grammar and representation, 205–232. Amsterdam: Benjamins.
Novogrodsky, Rama, Noga Balaban & Naama Friedmann. 2010. When the child says that daddy hugs him, whom does
he hug? On the development of the comprehension and production of pronouns in Hebrew. Language and Brain 10.
1–19. (in Hebrew)
Paradis, Johanne, Martha Crago & Fred Genesee. 2003. Object clitics as a clinical marker of SLI in French: Evidence
from French-English bilingual children. In Barbara Beachley, Amanda Brown & Frances Conlin (eds.), Proceedings
of the 27th annual Boston University Conference on Language Development [BUCLD 27], 638–649. Somerville,
MA: Cascadilla Press.
Pérez-Leroux, Ana Teresa, Mihaela Pirvulescu & Yves Roberge. 2008(a). Children’s interpretation of null objects under
the scope of negation. In Susie Jones (ed.), Proceedings of the 2008 Annual Conference of the Canadian Linguistic
Association. Vancouver: University of British Columbia, 1–10.
Pérez-Leroux, Ana Teresa, Mihaela Pirvulescu & Yves Roberge. 2008(b). Null objects in child language: Syntax and the
lexicon. Lingua 118(3). 370–398.
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

30 VARLOKOSTA ET AL.
Petinou, Kakia & Arhonto Terzi. 2002. Clitic misplacement in normally developing and language impaired Cypriot-Greek
speaking children. Language Acquisition 10. 1–28.
Pirvulescu, Mihaela. 2006. Theoretical implications of object clitic omission in early French: spontaneous vs. elicited
production. Catalan Journal of Linguistics 5. 221–236.
Raposo, Eduardo.1986. On the null object construction in European Portuguese. In Osvaldo Jaeggli & Carmen Silva-
Corvalán (eds.), Studies in romance linguistics, 373–390. Dordrecht: Foris.
Ruigendijk, Esther, Naama Friedmann, Rama Novogrodsky & Noga Balaban. 2010. Symmetry in comprehension and
production of pronouns: A comparison of German and Hebrew. Lingua 120(8). 1991–2005.
Schaeffer, Jeannette. 1997. Direct object scrambling in Dutch and Italian child language. Los Angeles: University of
California dissertation.
Schaeffer, Jeannette. 2000. The acquisition of direct object scrambling and clitic placement. Amsterdam/Philadelphia:
John Benjamins.
Schmitz, Katrin & Natascha, Müller. 2008. Strong and clitic pronouns in monolingual and bilingual acquisition of French
and Italian. Bilingualism: Language and Cognition 11(1). 19–41.
Silva, Carolina. 2008. Assimetrias na aquisição de clíticos diferenciados em português europeu [Asymmetries in the
acquisition of different types of clitics in European Portuguese]. Lisboa: Universidade Nova de Lisboa MA thesis.
Smith, Nafsika, Susan Edwards, Vesna Stojanovik & Spyridoula Varlokosta. 2008. Object clitics, definite articles and
genitive possessive clitics in Greek specific language impairment (SLI): Deficits and explanations. In Theodoros
Marinis, Angeliki Papangeli & Vesna Stojanovik (eds.), Proceedings of the 2007 Child Language Seminar, 30th
Anniversary, 146–156. University of Reading.
Spenader, Jennifer, Erik-Jan Smits & Petra Hendriks. 2009. Coherent discourse solves the pronoun interpretation problem.
Journal of Child Language 36(1). 23–52.
Sportiche, Dominique.1996. Clitic constructions. In Johan Rooryck & Laurie Zaring (eds.), Phrase structure and the
lexicon, 213–276. Dordrecht: Kluwer Academic.
Stavrakaki, Stavroula & Heather van der Lely. 2010. Production and comprehension of pronouns by Greek children with
specific language impairment. British Journal of Developmental Psychology 28. 189–216.
Tedeschi, Roberta. 2009. Acquisition at the interfaces. A case study on object clitics in early Italian. Utrecht: University
of Utrecht dissertation.
Terzi, Arhonto 1999. Cypriot Greek clitics and their positioning restrictions. In Artemis Alexiadou, Geoffrey Horrocks &
Melita Stavrou (eds.), Studies in Greek syntax, 227–240. Dordrecht: Kluwer.
Tsakali, Vina & Wexler Kenneth. 2003. Why children omit clitics in some languages but not in others: New evidence
from Greek. Paper presented at the Generative Approaches to Language Acquisition [GALA 2003], September 4-6.
University of Utrecht, Utrecht, The Netherlands.
Tsakali, Vina & Wexler Kenneth. 2004. Why children omit clitics in some languages, but not in others: New evidence
from Greek. In Kampen Jacqueline & Sergio Baauw (eds.), Proceedings of Generative Approaches to Language
Acquisition [GALA 2003], 493–504. LOT.
Tsimpli, Ianthi Maria & Stavroula Stavrakaki. 1999. The effects of a morphosyntactic deficit in the determiner system:
The case of a Greek SLI child. Lingua 108. 31–86.
Tuller, Laurice. 2000. Aspects de la morphosyntaxe du français des sourds [Aspects of morphosyntax in deaf French].
Recherches Linguistiques de Vincennes 29. 143–156.
Tuller, Laurice, Hélène Delage, Cécile Monjauze, Anna-Gaëlle Piller & Marie-Anne Barthez. (2011). Clitic pronoun
production as a measure of atypical language development in French. Lingua 121(3). 423–441.
Uriagereka, Juan. 1995. Aspects of the syntax of clitic placement in Western Romance. Linguistic Inquiry 26. 79–123.
Valian, Virginia. 1991. Syntactic subjects in the early speech of American and Italian children. Cognition 40. 21–81.
Vanderweide, Teresa. 2005. The acquisition of manner in pre-vocalic sequences: A cue is a cue. In Marina Tzakosta,
Claartje Levelt & Jeroen van de Weijer (eds.), Developmental paths in phonological acquisition. Special Issue of
Leiden Papers in Linguistics 2(1).137–161.
Varlokosta, Spyridoula. 2000. Lack of clitic pronoun distinctions in the acquisition of principle B in child Greek. In
S. Catherine Howell, Sarah A. Fish & Thea Keith-Lucas (eds.), Proceedings of the 24th annual Boston University
Conference on Language Development [BUCLD 24], 738–748. Somerville, MA: Cascadilla Press.
Varlokosta, Spyridoula, Katerina Konstantzou & Michaela Nerantzini. 2014. On the production of direct object clitics in
Greek typical development and specific language impairment: The effect of task selection. In Kleanthes K. Grohmann
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 31
& Theoni Neokleous (eds.), Developments in the acquisition of clitics, 188–211. Newcastle Upon Tyne: Cambridge
Scholars Publishing.
Wexler, Kenneth, Anna Gavarró & Vicent Torrens. 2004. Feature checking and object clitic omission in child Catalan and
Spanish. In Reineke Bok-Bennema, Bart Hollebrandse, Brigitte Kampers-Manhe & Petra Sleeman (eds.), Romance
languages and linguistic theory 2002, 253–70. Amsterdam: John Benjamins.
Zwart, C. Jan-Wouter. 1991. Clitics in Dutch: Evidence for the position of INFL. Groninger Arbeitenzur Germanistischen
Linguistik 33. 71–92.
Zwicky, Arnold M. 1977. On clitics. Bloomington: Indiana University Linguistics Club.
Submitted 19 July 2012
Final version accepted 20 February 2015
APPENDIX A
Examples of sentences used to accompany the picture in Figure 1, for all the languages tested.
(1) a. preverbal clitic
Catalan
El nen ruixa el gat i el gat està moll.
the boy sprays the cat and the cat is wet
Com és que el gat està moll?
how is that the cat is wet
El gat està moll perquè el nen el ruixa.
the cat is wet because the boy
CL.ACC .3SG wet.PRES.3SG
French
Le garçon arrose le chat et le chat est mouillé.
the boy sprays the cat and the cat is wet
Pourquoi est-ce-que le chat est si mouillé?
why is-it-that the cat is so wet
Le chat est mouillé parce que le garçon l’arrose.
the cat is wet because the boy
CL.ACC.3SG spray.PRES.3SG
Italian
Il bambino sta bagnando il gatto ed il gatto è bagnato.
the boy is wetting the cat and now the cat is wet
Perché il gatto è così bagnato?
why the cat is so wet
Il gatto è bagnato perche il bambino lo bagna/ sta bagnando.
the cat is wet because the boy
CL.ACC.3SG.MASC wet.PRES.3SG/ is wetting
Romanian
B
˘
aiatul strope¸ste/ud
˘
a pisica ¸si pisica e ud
˘
a.
boy.the sprays cat.the and cat.the is wet
De ce e pisica a¸sa ud
˘
a?
why is cat.the that wet
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

32 VARLOKOSTA ET AL.
Pisica e ud
˘
a pentru c
˘
a o strope¸ste/ud
˘
ab
˘
aiatul.
cat.the is wet because
CL.ACC .3SG.FEM spray.PRES.3SG boy.the
Spanish
El niño está mojando a-l gato y ahora el gato está mojado.
the boy is weting to-the cat and now the cat is wet.
MASC
¿Por qué está mojado el gato?
why is wet.
MASC the cat
El gato está mojado porque el niño lo ha mojado.
The cat is wet.
MASC because the child CL.AC C .3SG.MASC has wetted
Standard Modern Greek
To aγori vrexi ti γata ke i γ ata ine vreγmeni.
the boy wets the cat and the cat is wet
Jati ine vreγmeni i γata?
why is wet the cat
I γata ine vreγmeni jati to aγori tin vrexi.
the cat is wet because the boy
CL.ACC .3SG.FEM wet.PRES.3SG
b. postverbal clitic
Cypriot Greek
To aγori vreSitiγata tSeiγata e vremeni.
the boy wets the cat and the cat is wet
Jati i γata e vremeni?
why the cat is wet
I γata e vremeni jati to aγori vreSi tin.
the cat is wet because the boy wet.
PRES.3SG CL.AC C .3SG.FEM
European Portuguese
O menino molhou o gato e o gato está molhado.
the boy wet.
PAST. the cat and the cat is wet
Porque é que o gato está molhado?
why is that the cat is wet
O gato está molhado porque o menino o molhou.
the cat is wet because the boy
CL.ACC .3SG.MASC wet.PAST.3SG
c. second-position clitic
Croatian
Dje
ˇ
cak prska ma
ˇ
cku i ma
ˇ
cka je mokra.
boy sprays cat and cat is wet
Zaštojema
ˇ
cka mokra?
why is cat wet
Ma
ˇ
cka je mokra zato što ju je dje
ˇ
cak poprskao.
cat is wet because
CL.ACC .3SG.FEM be.AUX boy spray.PERF.3SG
Polish
Chłopiec ochlapuje kota i kot jest mokry.
boy sprays cat and cat is wet
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

ACQUISITION OF CLITIC AND PRONOUN PRODUCTION 33
Dlaczego kot jest mokry?
why cat is wet
Kot jest mokry bo chłopiec ochlapuje go.
cat is wet because boy spray.
PRES.3SG CL.AC C .3SG.MASC
Serbian
De
ˇ
cak poliva ma
ˇ
cku i ma
ˇ
cka je mokra.
boy wets cat and cat is wet
Zašto je ma
ˇ
cka mokra?
why is cat wet
Ma
ˇ
cka je mokra zato što je de
ˇ
cak poliva.
cat is wet because
CL.ACC .3SG.FEM boy wet. PRES.3SG
(2) pronoun
Danish
Dreng-en sprøjter på katt-en og katt-en er våd.
boy-the spray on cat-the and cat-the is wet
Hvorfor er katt-en våd?
why is cat-the wet
Katt-en er våd fordi dreng-en sprøjter på den.
cat-the is wet because boy-the spray.
PRES on PRON.COMMON.3SG
Dutch
De jongen spuit de poes nat en nu is de poes nat.
the boy sprays the cat wet and now is the cat wet
Hoe komt het dat de poes nat is?
how comes it that the cat wet is
De poes is nat omdat de jongen ‘m nat heeft gespoten.
the cat is wet because the boy
PRON.ACC .3SG.MASC wet has sprayed
German
Der Junge spritzt die Katze nass und die Katze ist sehr nass.
the boy sprays the cat wet and the cat is very wet
Wieso ist die Katze so nass?
why is the cat so wet
Die Katze ist so nass, weil der Junge die/sie nass spritzt.
the cat is so wet because the boy
PRON.ACC .3SG.FEM wet sprays
Hebrew
Ha-yeled hirtiv et ha-xatul ve-axshav ha-xatul ratuv.
the-boy wet
AC C the-cat and-now the-cat wet
Lama ha-xatul ratuv?
why the-cat wet
Ha-xatul ratuv ki ha-yeled hirtiv oto.
the-cat wet because the-boy wet.
PAST.3SG.MASC PRON. ACC.3SG.MASC
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

34 VARLOKOSTA ET AL.
APPENDIX B
Verbs, Pictures, and Lead-In Sentences for the English Experiment
Ver b
Picture
description Experimenter
cover princess covering
soldier
The princess covered the soldier. Now the soldier is warm. Why is the soldier
warm? The soldier is warm because the princess ...
comb mother combing
girl’s hair
Mommy is combing the girl, and the girl is beautiful. Why is the girl
beautiful? The girl is beautiful because mommy ...
dry boy drying hippo The boy is drying the hippo. The hippo will be dry soon. Why is the hippo
getting dry? The hippo is dry because the boy ...
wake up girl waking up boy The girl woke up the sleeping boy. Now the boy is crying. Why is the boy
crying? The boy is crying because the girl ...
draw drawer/man
drawing girl
The drawer/man is drawing the girl and the girl is happy. Why is the girl
happy? The girl is happy because the drawer/man ...
wet boy spraying cat
with water
The boy is spraying the cat, and the cat is wet. Why is the cat so wet? The cat
is wet because the boy ...
paint painter/man
painting house
The painter/man painted the house, and it became blue. Why is the house
blue? The house is blue because the painter/man ...
lick dog licking cat The dog is licking the cat, and now the cat is happy. Why is the cat so happy?
The cat is happy because the dog ...
tie bee tying up
grasshopper
The bee tied up the grasshopper, and the grasshopper cannot jump. Why can’t
the grasshopper jump? The grasshopper can’t jump because the bee ...
wash girl washing
giraffe
The girl is washing the giraffe, and the giraffe is clean. Why is the giraffe
clean? The giraffe is clean because the girl ...
eat boy eating a piece
of the cake
The boy ate a piece of the cake, and that piece of the cake disappeared! Why
did the piece of the cake disappear? The piece of cake disappeared because
the boy ...
catch girl catching
butterfly
The girl caught the butterfly, and the butterfly cannot fly. Why can’t the
butterfly fly? The butterfly can’t fly because the girl ...
Downloaded by [Tel Aviv University] at 00:32 05 September 2015

APPENDIX C
The Verbs that Were Used in the 16 Languages (in their Infinitival Form; in Cypriot Greek and Standard Modern Greek, that Do Not Have Infinitives, the Verb Is
Provided in the Third-Person Singular Present Form)
English Catalan Croatian Cypriot Greek Danish Dutch French German Greek Hebrew Italian Polish Portuguese Romanian Serbian Spanish
cover tapar pokriti Sepazi putte toedekken couvrir zudecken skepazi lexasot coprire przykry
´
c tapar acoperi pokriti tapar
comb pentinar
ˇ
cešljati xtenizi frisere kammen coiffer kämmen xtenizi lesarek pettinare czesa
´
c pentear piept
˘
ana
ˇ
cešljati peinar
dry eixugar sušiti skupizi ‘wipe’ tørre afdrogen sécher abtrocknen stegnoni lenagev asciugare wyciera
´
c secar ¸sterge
‘wipe’
brisati secar
wake up despertar buditi ksipna vække wakkermaken réveiller aufwecken ksipna leha’ir svegliare obudzi
´
c acordar trezi probuditi despertar
draw dibuixar slikati zoγrafizi tegne schilderen dessiner malen zoγrafizi lecayer dipingere malowa
´
c desenhar picta crtati dibujar
wet mullar mo
ˇ
citi vre
Si våd natspuiten arroser
‘water’
nassspritzen vrexi lehartiv bagnare moczy
´
c molhar uda polivati mojar
paint pintar li
ˇ
citi po
Jatizi male verven peindre anstreichen vafi licbo’a dipingere pomalowa
´
c pintar vopsi farbati pintar
lick llepar lizati γlifi slikke likken lécher ablecken γlifi lelakek leccare liza
´
c lamber linge lizati chupar
tie lligar vezati
ðini binde vastbinden attacher fesseln δeni likshor legare zwi ˛aza
´
c atar lega vezati atar
wash rentar prati luni vaske wassen laver waschen pleni lirxoc lavare my
´
clavarsp
˘
ala prati lavar
eat menjar jesti troi spise eten manger essen troi leexol mangiare zje
´
s
´
c comer mînca pojesti comer
catch caçar uhvatati pcani fange vangen attraper fangen piani litpos prendere złapa
´
c apanhar prinde uhvatiti atrapar
35
Downloaded by [Tel Aviv University] at 00:32 05 September 2015