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.

Notes

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.

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).

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.

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.

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).

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).

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.

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.

¹⁸ One anonymous reviewer raised the issue that the finding (ii) could be an artifact of the restriction of our experiment 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.

Additional information

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 supported 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 Miękisz) and from the National Science Centre/Ministry of Science and Higher Education (grants no N N106 223538, awarded to Magdalena Smoczyńska 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).