Abstract
Background: Children with language impairment often exhibit significant difficulty in the use of grammatical morphology. Although English-speaking children with language impairment have special difficulties with verb morphology, noun morphology can also be problematic in languages of a different typology.
Aims: Hungarian is an agglutinating language with multiple suffixation, in which both regular-class and irregular-class nouns contain the same recognizable grammatical markers, but the two classes differ in their morphophonology and productivity. Such typological characteristics provide a good basis for evaluating processing accounts of language impairment such as the morphological richness account.
Methods & Procedures: We examined the production of Hungarian irregular and regular noun morphology through elicited production of nouns with plural, accusative case and plural plus accusative case suffixes in an older (8–10 years) and a younger (4–7 years) group of children with language impairment and two verbal control groups matched on vocabulary size. The children's accuracy was scored both in terms of grammatical function (whether plural and/or accusative case was appropriately marked) and morphophonology (whether the production reflected the phonotactic form required for the stem plus suffix).
Outcomes & Results: The younger children with language impairment were less accurate than the younger verbal control children when two suffixes (marking plural and accusative case) were required, at least when noun stem classes were regular. All groups showed significant overgeneralization of stem forms with correct selection of suffixes. However, there were strong word frequency effects in the language impairment, but not in the verbal control groups.
Conclusions & Implications: Much of the data were consistent with predictions of the morphological richness account. However, there was also evidence suggestive of differences between the language impairment and verbal control groups in their representations. In particular, the children with language impairment seemed to rely more (though not exclusively) on memorized items in the lexicon.
Acknowledgements
This research was supported by Research Grant Number R01 DC00458 from the National Institute on Deafness and Other Communication Disorders, National Institutes of Health (USA) to Laurence B. Leonard and by Research Grant Number OTKA TS 049840 from the Hungarian National Science Foundation to Csaba Pléh. A´gnes Lukács was a grantee of the Bolyai János Research Scholarship of the Hungarian Academy of Science. The authors are grateful to the children in Simon Antal Primary School in Vac, in the Dr Nagy László Institute of Special Education in Kószeg, in the ELTE Special Preschool and Early Intervention Centre and in the Zőlderdö Preschool for speech therapy and nature preservation for their participation. They also thank the speech therapists in both institutions for their help with screening and organization. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
Notes
1. For ease of exposition, we use standard Hungarian orthography and do not give phonetic transcriptions. Hungarian orthography is fairly transparent, geminates are marked by double consonants (also by doubling the first letter in a consonant digraph), and accents above vowels mark length. However, not every accented vowel is phonetically equivalent to their short counterpart, so we present the phonetic symbols for Hungarian vowels and non-transparent consonantal letters here. Vowels: a [c], á [a:], o [o], ó [o:], u [u], ú [u:], e [3], é [e:], i [i], í [i:],ö [ø], ő [ø:], ü [y], ű [y:]; and consonants: c [ts], cs [tf], dzs [d3], g [g], gy [
], j [j], ly [j], ny [n], r [r], s [c], sz [s], ty [ç], zs [
]. Our description is based on Kiefer (Citation1998), Nadasdy and Siptar (Citation1994), and Törkenczy (1994). Linguistic accounts of the alternations are not presented; we provide only the information relevant to understanding the behaviour of the inflected forms examined in this study.
2. The 15 older children with LI and their VC pairs were also part of a larger group of children who participated in a separate study on verb agreement in Hungarian (Lukács et al. forthcoming).
3. The authors are grateful to Professor Dorothy Bishop for providing us with the TROG for this purpose. To date, 600 typically developing children have been tested as part of the standardization process. The scores of the children with LI in this study were compared against the values obtained for the typically developing children participating in the standardization.
4. The authors are grateful to Péter Halácsy for the frequency calculations for bound allomorphs.
5. Results for statistical tests in error analysis are only given where differences are significant.
6. This is due to the phonotactics of the accusative: overgeneralized accusative forms result in phonotactically more well-formed sequences than overgeneralized plural forms in most cases.