Elsevier

Cortex

Volume 116, July 2019, Pages 192-208
Cortex

Special issue: Research report
Regular and irregular inflection in down syndrome – New evidence from German

https://doi.org/10.1016/j.cortex.2018.08.010Get rights and content

Abstract

The study aims to explore whether regular inflectional morphology is affected in children/adolescents with Down syndrome (DS). German past participle forms were elicited for ten regular and ten irregular inflected verbs as well as for five novel verbs. Data were collected from a group of 21 monolingual German children and adolescents with DS (chronological age M = 11;03 years) and a group of 21 typically developing monolingual German children (chronological age M = 4;03 years) matched in chronological age to the nonverbal mental age of the DS group (mental age DS group M = 4;05 years). Data analysis indicated that eight children/adolescents with DS displayed a deficit in acquiring the regular participle marker -t and in applying it as default inflection in German participle formation. In contrast, a group of 13 individuals with DS performed similar to the typically developing control children. They had successfully acquired the regular participle affix -t and readily applied it as default inflection to produce participles for irregular verbs and novel verbs. The data indicate that the acquisition of regular inflectional morphology is not outside the scope of individuals with DS and succeeded in many affected individuals. However, a substantial number of individuals with DS displayed a selective deficit with regular default inflection. The occurrence of a selective deficit with regular default inflection in individuals with DS supports dualistic views to inflection, according to which two different cognitive components are involved in regular and irregular inflection that can be selectively affected in language disorders.

Introduction

Inflectional morphology is an area of language that is frequently affected in acquired or developmental language disorders (overview in Penke, 2008). Typically, such deficits do not affect all inflected forms in uniform but operate selectively, affecting for instance only one inflectional category (for example tense inflection or verbal inflection) while leaving other inflectional categories (such as agreement inflection or noun inflection) spared (Clahsen et al., 1997, Friedmann and Grodzinsky, 1997, Laiacona and Caramazza, 2004, Rice et al., 1995). A distinction that has proven to be of particular relevance in this respect is the distinction between regular and irregular inflectional morphology.

In many languages, inflectional systems display a distinction between regular and irregular inflected forms. Consider, for instance, English past tense formation where we find regular forms inflected with -ed (love–loved) and irregular ones like sing–sang that are idiosyncratic and largely unpredictable. According to an influential view in linguistics and psycholinguistics – most prominently defended by Pinker (1999) – the representations and mechanisms involved in the processing of regular and irregular inflectional forms are qualitatively distinct. So-called dualistic models of inflection (Clahsen, 1999, Penke, 2012a, Pinker and Prince, 1988, Pinker, 1999) assume that regular inflected forms are built by the application of a mental operation that combines a stem and an affix. Irregular forms, in contrast, are stored as fully inflected word forms in the mental lexicon. Under the dualistic view, regular and irregular inflection rely on two independent modules of the human language faculty: a computational component where regular affixation is carried out, i.e., grammar, and a storage component in which idiosyncratic learned information is stored and retrieved, i.e., the mental lexicon. If this dualistic view is correct, we should observe some language deficits that selectively affect only one of these two components while the other is spared. Thus, we should observe language disorders that either selectively affect the regular inflectional component with the irregular inflected forms spared or, vice versa, selectively affect stored irregular inflected forms leaving the regular submodule unimpaired. The search for such selective impairments of regular and irregular inflected forms in language impaired speakers has inspired much research in clinical linguistics over the last two decades and a number of different impairment syndromes have come into focus during this time (Penke, 2008 for overview). One of these syndromes is Down syndrome.

Down syndrome (DS) is a genetic disorder that is typically caused by a third exemplar of all or parts of chromosome 21 (trisomy 21). It is the most common neurodevelopmental disorder causing intellectual disability and occurs in approximately 1 in 900 live births in Europe (Loane et al., 2013). DS is associated with a number of physiological characteristics as well as intellectual disability that may range from mild to severe. Language is often reported to be a domain of particular difficulties in individuals with DS, considerably lagging behind expectations based on their nonverbal mental development (Abbeduto et al., 2007, Chapman et al., 1998, Fowler et al., 1994). Within language abilities, morphosyntactic abilities have been found to be particularly affected, leading to the assumption that language production in DS might be characterized by a generalized deficit in grammatical morphology (Chapman et al., 1998). Vocabulary skills, in contrast, seem to be less affected in individuals with DS and have often been found to conform to the nonverbal cognitive abilities reached (Næss et al., 2011, Roberts et al., 2008).

The reported dissociation between impaired grammatical skills and better retained lexical skills has furthered an interest on the issue how regular and irregular inflection are affected in individuals with DS. Under the dualistic view to inflection one might expect regular inflection to be affected in individuals with DS since it requires morphological processing that is assumed to be impaired in DS. Irregular inflected forms, in contrast, might be less affected in DS as they are supposed to be stored in the mental lexicon, a component often developed according to mental-age expectations in these individuals. Indeed, selective difficulties in regular verbal inflection with irregular inflected forms spared have been reported in two studies investigating English-speaking individuals with DS. Based on an analysis of spontaneous-speech recordings, Eadie, Fey, Douglas, and Parsons (2002) reported a significantly weaker performance of 10 children with DS (mean age 7 years) with the regular English past tense marker -ed in comparison to a group of 10 typically developing control children (aged 3 years) matched for mean length of utterance (MLU). For irregular past tense forms, in contrast, percentages of correctly produced forms did not differ between control children and children with DS. A deficit with the regular past tense marker -ed was also found by Laws and Bishop (2003) in an elicitation task performed with 14 individuals with DS (aged 10–19 years) and 11 typically developing children (age 4–7 years) matched for nonverbal mental age. Whereas the participants with DS achieved significantly lower mean percentages for correctly produced regular past tense forms compared to the control children, performance with respect to irregular past tense forms did not differ in the two participant groups.

While Eadie et al. (2002) and Laws and Bishop (2003) report selective difficulties in regular verbal inflection with irregulars spared, Ring and Clahsen (2005) did not find a selective deficit with regular past tense inflection in a group of eight English-speaking adolescents with DS (aged 12–14 years) who performed an elicitation task on regular and irregular past tense forms. Ring and Clahsen report that the participants with DS produced a significantly higher number of uninflected verb forms compared to a control group of 35 typically developing children (aged 4;11 to 6;11) matched for mental age. Once these uninflected forms were disregarded, there was no performance difference in the production of correctly inflected regular and irregular past tense forms between the two subject groups. Likewise, in a grammaticality judgement task performed with eight Greek-speaking adolescents (aged 12–18 years), Stathopoulou and Clahsen (2010) reported the regular, sigmatic and the irregular, non-sigmatic perfective past tense formation of existing verbs to be parallel to a group of 16 typically developing control children aged 5–7 years. Also, like the control children, participants with DS displayed a preference for regular, sigmatic past tense forms over irregular, non-sigmatic ones for rhyming and non-rhyming novel verbs, indicating that they were able to productively apply the rule for past tense formation.

Two factors might be responsible for the different outcomes regarding regular and irregular inflection in individuals with DS in these studies. One relates to the huge variability in language performance observed in individuals with DS (Fowler, 1995). In a longitudinal study (duration 2;06 years) on the acquisition of 14 grammatical morphemes in 12 English-speaking children with DS (aged 12–38 months) Rutter and Buckley (1994) found that only the progressive marker -ing was acquired by all children, whereas the regular past tense marking was only acquired by half of the children. Also, in the above-mentioned study on regular and irregular past tense inflection in English-speaking subjects with DS, Laws and Bishop (2003) reported that for half of their 14 participants with DS performance with respect to regular and irregular past tense forms did not differ significantly from the control group of typically developing children, whereas the whole group of participants with DS displayed a pronounced selective impairment with regular past tense forms compared to the control children. Both studies indicate that with respect to regular and irregular inflected forms individuals with DS do not perform as a homogeneous group but that some individuals display an impairment that might be missing in others. As the existing studies on regular and irregular inflection in DS have investigated relatively small numbers of participants (between 8 and 14), it might hence be that the divergent findings that have been reported in the literature are due to sampling biases, i.e., to studies accidentally including fewer or more participants in the DS group who have or have not acquired the specific inflectional phenomena under investigation. To address this issue, studies are needed that investigate larger subject groups while taking individual variation into account.

A second factor relates to the influence of language-specific factors on language impairments. Studies on the outcome of language disorders in different languages have cautioned us that differences regarding the typology or complexity of inflectional systems in different languages might lead to differences in the specific deficits observed in a particular language disorder (see overview in Penke, 2008). For instance, the selective deficit of regular inflection characteristic for English-speaking children with Specific Language Impairment (SLI) (van der Lely & Ullman, 2001) has not been evinced in German children with SLI (Clahsen, Rothweiler, Sterner, & Chilla, 2014). While the reason for this difference is still under discussion, it has been found that omission rates decrease with the number and the morphosyntactic relevance inflectional affixes have in a given language (Bates et al., 1987, Dromi et al., 1999). Thus, in English, an analytic language with a largely reduced inflectional component, inflectional markers tend to be omitted in language disorders, whereas in languages where inflectional systems are more elaborate and express more syntactic information, such as in German, omission rates are markedly lower compared to English (Penke & Westermann, 2006). To disentangle the influence of language-specific factors on the symptoms of language impairments cross-linguistic evidence is needed.

The investigation of German regular and irregular past participle formation that is undertaken in this study aims at addressing these issues by investigating a larger group of affected individuals and by providing data from a language and an inflectional system that exhibits different properties than the English and Greek systems that have been studied in previous research.

The system of past participle inflection (henceforth participle inflection) in German is similar to the English past tense and the perfective past tense in Greek in that it distinguishes between regular and irregular inflected forms. German participles are formed with a phonologically conditioned prefix ge-, a verb stem, and the endings -t or -n. In regular inflected verbs there is no change of the verb's base form in the participle and participles are suffixed with -t (see examples (1a) and (1c)). In contrast, irregular inflected verbs often show a modification of the stem vowel in the participle stem, the so-called ablaut, and take the participle ending -n (see examples (1b, d-f)). The system differs from English past tense formation in that both regular and irregular German participles have distinct endings that are separable from the verb stem. It also differs from the Greek regular, sigmatic perfective past tense in that the verb's base does not undergo further changes in deriving a verb stem to which regular inflection is affixed.

Neither the base vowel nor the phonological shape of the verb's base predict whether a German verb will be inflected regularly or irregularly. Thus, although the verbs blinken (1a) and sinken (1b) rhyme, one has a regular inflected participle form (geblinkt), the other one has an irregular participle form (gesunken). Also, as examples (1c-f) indicate, verbs with the base vowel /e/ can be regularly (1c) or irregularly inflected (1d-f). Moreover, in many cases the type of vowel change that occurs in irregular inflected participle forms is not predictable from the verb's base vowel. For example, irregular verbs with the base vowel /e/ display several different Ablaut patterns and further modifications of the base stem in the participle form (see (1d-f)). Due to this unpredictability of verb class (regular or irregular inflection) and the occurring vowel and stem changes in irregular verbs, dualistic approaches to inflection assume that irregular inflected participle forms are stored as fully-inflected forms in the mental lexicon. Regular participles such as geblinkt in (1a) are, in contrast, assumed to be built on the fly by combining the basic verb stem blink- with the participle affix -t via a productive affixation operation (Clahsen, 1999, Wunderlich, 1996). A central tenet of this dualistic view of German participle inflection is that participle formation with the regular participle marker -t serves as default operation to construct a participle. Whenever no stored participle form can be accessed in the mental lexicon, the default procedure applies and a participle is produced by suffixing the marker -t to the verb's base (Clahsen, 1999).

Regular and irregular German participles display a distinct developmental pattern in first language acquisition, resembling the acquisition of the English past tense (Cazden, 1968, Clahsen and Rothweiler, 1993, Marcus et al., 1992, Penke, 2006, Penke, 2012b, Weyerts and Clahsen, 1994). German participles will appear as early as the two-word stage, typically from about 18 months onwards (Clahsen and Rothweiler, 1993, Elsen, 1998, Lindner, 1998, Penke, 2006, Szagun, 2011). While regular inflected participle forms are produced correctly throughout development, the acquisition of irregular participle forms displays a very characteristic u-shaped developmental curve. The first irregular participles produced by children are typically correct. This stage is followed by a stage where children overapply the regular inflectional pattern to verbs that have irregular inflected participle forms in the target language, resulting in overregularization errors such as the participle form *gesinkt instead of the correct irregular form gesunken (see example (1b)). As a consequence of these overregularization errors, accuracy scores for irregular participles decrease. First overregularization errors in German child language typically appear by age 2;0 years (Clahsen and Rothweiler, 1993, Elsen, 1998, Lindner, 1998, Penke, 2006). The weeding out of these overregularization errors takes some time in language acquisition. As irregular participle forms are learned and stored in the mental lexicon during acquisition, correctness scores for irregular participles rise again and finally reach the adult correctness level (Clahsen and Rothweiler, 1993, Penke, 2012b, Weyerts and Clahsen, 1994).

While overregularization errors are a characteristic sign of child language, overapplications of irregular inflectional patterns are comparably very rare (usually an order of magnitude smaller) (Penke, 2006, Szagun, 2011, Weyerts, 1995, Xu and Pinker, 1995). Dualistic approaches to inflection regard the greater productivity of the regular inflectional pattern as evidence that the child acquires a productive mental operation that combines the verb's base and an affix (here the participle affix -t) to generate a participle (Clahsen, 1999, Penke, 2012a). Once the child has identified and acquired the regular participle affix -t, this affix can freely apply as default inflection to any instance of the category verb for which no irregular participle form has yet been learned and stored in the mental lexicon – hence overregularization errors result.

The aim of this study is to provide new evidence on the issue whether regular and irregular inflected forms are affected differently in children/adolescent with DS. The paper reports results from a study on regular and irregular German participle inflection conducted with 21 German-speaking children and adolescents with DS and a control group of 21 typically developing children.

Adopting the dualistic view, the regular participle ending -t is among the first affixes to be acquired by typically developing German children. It is assumed to be acquired when the first overregularization errors occur, usually at the beginning of age 2;0 years, indicating that this affix can now be used as default inflection in participle formation (Clahsen and Rothweiler, 1993, Weyerts, 1995, Wittek and Tomasello, 2002). The acquisition of regular participle inflection thus proceeds early enough to discover developmental delays or deficits in children/adolescents with DS since they will often achieve a nonverbal mental age that is above an age of three years. If individuals with DS with a mental age well above three years display problems with regular participle formation, these problems are indicative of a deficit in acquiring the regular default affix -t.

A deficit in acquiring the regular default affix -t should not only disturb the production of regular inflected participles but should also affect irregular inflected verbs as well as the inflection of novel verbs. Typically developing children use the regular participle marker -t as a default to produce participles when needed and no stored participle form can be accessed in the mental lexicon. As a consequence, they produce overregularized participles for irregular verbs for which the irregular participle form has not yet been acquired and stored in the mental lexicon. Likewise, they will produce regularized participles for novel verbs because a stored participle form is not available for verbs unknown to children (Weyerts, 1995, Wittek and Tomasello, 2002). Regularizations of novel verbs as well as overregularization errors with irregular verbs are, however, only possible if the regular participle affix -t has been acquired and can, hence, be used as default inflection to form a participle for a verb for which no stored participle form can be found in the mental lexicon. Consequently, a deficit with regular default inflection should effect the production of overregularization errors with irregular verbs and the production of participle forms for novel verbs.

To evaluate whether the production of participles in individuals with DS differs quantitatively and/or qualitatively from the performance of typically developing children, performance of individuals with DS is compared to participle production in a group of typically developing children matched in chronological age to the nonverbal mental age of the participants with DS. This comparison allows for differentiating a deficit in language acquisition in children/adolescents with DS from a language performance that is expected given the level of cognitive development achieved. If language performance in individuals with DS differs quantitatively and/or qualitatively from the performance of the mental-age-matched control children, this is indicative of a language impairment because, in this case, performance in individuals with DS cannot be accounted for by the level of cognitive development achieved by these individuals.

Section snippets

Participants

Twenty-one German-speaking children and adolescents with DS participated in this study. Their performance on participle production was compared to a group of 21 typically developing (TD) control children.

German-speaking children and adolescents with DS were recruited via self-help websites and publications for parents of children with DS. Participants were included in the study if they were monolingual German, if oral language was their primary means of communication, and if they produced at

Results

The results obtained by participants of the two subject groups are summarized in Table 3.

Discussion

Previous studies by Eadie et al. (2002) and Laws and Bishop (2003) on English-speaking individuals reported a selective deficit with regular past tense forms whereas performance with irregular past tense forms was comparable to the performance of typically developing children. In contrast to these two studies, the group of 21 German-speaking children and adolescents with DS investigated here attained an accuracy score for regular participles of over 90% that did not differ from the mean

Conclusion

Two different performance profiles were identified in an elicitation task on regular and irregular participles in German-speaking individuals with DS. A subgroup of 13 participants with DS achieved high correctness scores for regular inflected participles and productively used the regular participle marker -t as a default marker to generate participles when needed, similar to a control group of typically developing children. Performance of these individuals shows that the acquisition of regular

Acknowledgements

The work reported in this paper was supported by a German research foundation (DFG) grant [grant no. PE 683/3-1]. I am grateful to parents and children who undertook the effort in participating in this study. I would also like to thank the members of the research team, Alina Schmitz, Ruth Szustkowski, and Bernadette Witecy, for their assistance in data collection and analysis.

References (65)

  • H. Baayen et al.

    The CELEX lexical database (CD-ROM)

    (1993)
  • U. Bellugi et al.

    The neurocognitive profile of Williams syndrome: A complex pattern of strengths and weaknesses

    Journal of Cognitive Neuroscience

    (2000)
  • E. Berglund et al.

    Parental reports of spoken language skills in children with Down syndrome

    Journal of Speech, Language, and Hearing Research: JSLHR

    (2001)
  • D. Bittner et al.

    Introduction

  • G. Booij

    Morphology and phonology

  • J. Bybee

    Regular morphology and the lexicon

    Language and Cognitive Processes

    (1995)
  • J. Bybee

    Lexical, morphological and syntactic symbolization

  • C.B. Cazden

    The acquisition of noun and verb inflections

    Child Development

    (1968)
  • R. Chapman et al.

    Language skills of children and adolescents with Down syndrome: II. Production deficits

    Journal of Speech Language and Hearing Research

    (1998)
  • H. Clahsen

    Lexical entries and rules of language: A multidisciplinary study of German inflection

    Behavioral and Brain Sciences

    (1999)
  • H. Clahsen et al.

    Underspecification and lexical learning in early child grammars

  • H. Clahsen et al.

    Speeded production of inflected words in children and adults

    Journal of Child Language

    (2004)
  • H. Clahsen et al.

    The acquisition of agreement morphology and its syntactic consequences

  • H. Clahsen et al.

    Inflectional rules in children's grammars: Evidence from German participles

  • H. Clahsen et al.

    Linguistic markers of specific language impairment in bilingual children: The case of verb morphology

    Clinical Linguistics & Phonetics

    (2014)
  • E. Clark

    The principle of contrast: A constraint on language acquisition

  • J. Cleland et al.

    Relationship between speech, oromotor, language and cognitive abilities in children with Down's syndrome

    International Journal of Language & Communication Disorders

    (2010)
  • B. Davies

    Auditory disorders

  • E. Dromi et al.

    Verb agreement morphology in Hebrew speaking children with specific language impairment

    Journal of Speech and Hearing Research

    (1999)
  • M. Dryer

    Prefixing versus suffixing in inflectional morphology

  • P. Eadie et al.

    Profiles of grammatical morphology and sentence imitation in children with specific language impairment and Down syndrome

    Journal of Speech, Language and Hearing Research: JSLHR

    (2002)
  • H. Elsen

    The acquisition of past participles: One or two mechanisms

  • Cited by (2)

    View full text