Developmental Science 7:1 (2004), pp 88 –102 © Blackwell Publishing Ltd. 2004, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. Blackwell Publishing Ltd PAPER Syntactic complexity in maltreatment The impact of child maltreatment on expressive syntax at 60 months Inge-Marie Eigsti 1 and Dante Cicchetti 2 1. Department of Clinical and Social Sciences in Psychology, University of Rochester, USA 2. Mount Hope Family Center and Department of Clinical and Social Sciences in Psychology, University of Rochester, USA Abstract Although child maltreatment has often been described as leading to language deficits, the few well-controlled investigations of language acquisition in maltreated children have focused on language content rather than form, or have used qualitative rather than quantitative measures. This study examines syntactic complexity in 19 maltreated and 14 nonmaltreated preschool-aged children. Mother–child dyads participated in play sessions that were transcribed and scored for the presence of morphosyntactic forms in child speech and for specific sentence constructions in maternal speech. Findings indicated that child maltreatment was associated with language delay in both vocabulary and production of syntactic structures. There were also qualitative dif- ferences in characteristics of maternal utterances between maltreating and comparison groups. Because maltreatment initially occurred before age 2, this study highlights the long-lasting negative influence of maltreatment on language development and also provides the first demonstration of child language delays and differences in maternal speech within a single maltreatment sample. Introduction Although the process of language acquisition in norm- ally developing children is a highly robust phenomenon, clinicians and educators have often noted that children who have been maltreated exhibit language delays and deficits. However, because of the clear impact of mal- treatment on socio-emotional development (Cicchetti & Toth, 1995; Finkelhor & Kendall-Tackett, 1997), well- controlled studies of linguistic development in mal- treated children have focused on language content rather than form. Furthermore, those studies examining language structures have used qualitative rather than quantitative measures (McFadyen & Kitson, 1996; Beeghly & Cic- chetti, 1994; Lynch & Roberts, 1982; Allen & Wasserman, 1985; Elmer, 1981). To date, it has been difficult to un- cover the mechanisms that contribute to the range of devel- opmental outcomes observed in maltreating environments (National Research Council, 1993). A central principle of developmental psychopathology is that the study of perturbations in development can illuminate the con- straints on normal development (Cicchetti, 1984, 1993). Thus, research on atypical routes to language learning, such as may be found in maltreated populations, can lead to a better understanding of both the process of maltreat- ment and also of the robustness of the acquisition process. The current study investigates the language of a sam- ple of 19 maltreated and 14 nonmaltreated preschool- aged children from similar socioeconomic backgrounds, comparing the syntactic complexity of their spontan- eous utterances. Mother–child dyads participated in free play sessions that were subsequently transcribed and scored for specific morphosyntactic forms. The charac- teristics of maternal utterances addressed to the child were also examined. The study addresses the questions of whether maltreatment is correlated with language delays, particularly in the domain of syntactic develop- ment, and whether these delays are related to the quality of maternal utterances. Resilience of language acquisition and the role of child-directed language input Acquisition of a native language appears to involve a type of learning that is highly constrained. The learning Address for correspondence: Inge-Marie Eigsti, Department of Psychiatry, Columbia University, and the Sackler Institutes for Developmental Psychobiology at Columbia and Weill Medical College of Cornell University, 1300 York Avenue, Box 140, New York, NY 10021, USA; e-mail: ie2004@columbia.edu