Nonword repetition in specific language impairment: More than a phonological short-term memory deficit

SUSAN E. GATHERCOLE 0 0 University of York , York, England The possible role of phonological short-term memory in the nonword repetition deficit of children with specific language impairment (SLI) was investigated in a study comparing serial recall and nonword repetition of sequences of auditorily presented CV syllables. The SLI group showed impairments in both serial recall and nonword repetition relative to typically developing children of the same age, however the SLI deficit in nonword repetition was greater and persisted even when differences on an independent measure of short-term memory were taken into account. These findings cannot be readily explained in terms of a sole deficit in short-term memory, and point instead to differences between the serial recall and nonword repetition paradigms as potential factors contributing to this disorder of learning. - The capacity to repeat a novel phonological form such as woogalamic is one of the most basic and important language abilities. Every word we now know was once unfamiliar to us, and was learned, in part, via such a repetition attempt. The evidence linking nonword repetition and language learning abilities is now extensive. In particular, individuals who perform poorly on nonword repetition typically struggle to learn the phonological form of language. The evidence for this is now extensive. Individual differences studies of typically developing samples of children have established highly specific links between nonword repetition and knowledge of vocabulary of both the native language (e.g., Gathercole & Baddeley, 1989; Gupta, 2003) and foreign languages (e.g., Cheung, 1996; Masoura & Gathercole, 1999). Childrens nonword repetition abilities are also highly associated with the speed of learning the phonological forms of new words under experimental conditions that control exposure to the novel tokens, although not to nonphonological aspects of learning such as the acquisition of semantic features (e.g., Gathercole, Service, Hitch, Adams, & Martin, 1999; Gupta, 2003). Finally, severe deficits of nonword repetition have been found to characterize several groups of children with particularly marked impairments of language learning, including individuals with specific reading disabilities (e.g., Snowling, 1983), Downs syndrome (e.g., Laws, 2004), and specific language impairment (SLI; e.g., Gathercole & Baddeley, 1990). The co-occurrence of deficits in nonword repetition and SLI in particular has now been extensively documented (see, e.g., Conti-Ramsden, 2003; Dollaghan & Campbell, 1998; Edwards & Lahey, 1998; Montgomery, 2004). SLI is a relatively common developmental condition in which a child fails to develop language at the typical rate despite normal general intellectual abilities, adequate exposure to language, and in the absence of hearing impairments. Affected children have the greatest problems in learning word forms and the grammatical structure of language, with acquisition of semantics and pragmatics relatively spared (Leonard, 1998). Current interest in nonword repetition and SLI was sparked principally by Gathercole and Baddeleys findings in 1990 that a group of children with SLI had impairments in repeating lengthy nonwords that were even greater in magnitude than the language deficits that formed the basis for their diagnosis. The nonword repetition impairment has subsequently been established in many independent studies to be a hallmark of SLI (see Roy & Chiat, 2004, for review), and has been hailed both as a clinical indicator of SLI (Conti-Ramsden, 2003; Dollaghan & Campbell, 1998) and as a phenotypic marker of the genetic basis for SLI (Bishop, North, & Donlan, 1996) associated with abnormalities of chromosome 16q (SLI Consortium, 2004). The impairment is particularly compelling as it captures the language learning difficulties of individuals with SLI in a simple paradigm that mimics word learning. The established links between nonword repetition and language learning has led to widespread interest in understanding the cognitive processes that underlie nonword repetition. Nonword repetition was first proposed as a relatively pure index of phonological short-term memory (STM) capacity (Gathercole & Baddeley, 1989, 1993). According to this view, repetition of nonwords requires more reliance on the temporary storage of phonological representations in STM than items such as words or digits because of the reduced availability of long-term lexical knowledge to support the unfamiliar phonological forms. It was further suggested that the nonword repetition deficit in SLI arises from an impairment of phonological STM (Gathercole & Baddeley, 1990) such that limitations in storage capacity result in reduced repetition accuracy. Reports that children with SLI do indeed perform poorly also on more conventional serial recall measures of STM, such as digit and word span, are consistent with this position (e.g., Archibald & Gathercole, 2006; Montgomery, 1995). Research with other populations also indicates that an impairment of STM does disrupt language learning. Experimental studies of normal adult participants have established that the learning of new words is impaired under experimental conditions known to disrupt phonological STM (e.g., Papagno & Vallar, 1992). Furthermore, neuropsychological patients with acquired deficits in STM appear to be unable either to repeat nonwords or to learn the phonological forms of new words, despite retaining normal abilities to learn semantic associations (see, e.g., Baddeley & Wilson, 1993). Together, this evidence converges on the view that nonword repetition taps phonological STM, that STM mediates the phonological long-term learning of new words, and that at least some of the language learning difficulties associated with SLI may arise from an impairment of STM. On this basis, Baddeley, Gathercole, and Papagno (1998) proposed that the primary function of phonological STM is to support word learning. One potential problem for the proposal that children with SLI have an impairment of STM and thus for the account of phonological STM as an important influence on language learning is that SLI deficits in nonword repetition tend to be more marked than those in standard serial recall measures of STM. For example, in a recent study of 20 children with SLI from 7 to 11 years in age, Archibald and Gathercole (2006) found that, whereas 14 of the children showed a deficit on measures of serial recall of digits and word lists, every child had a deficit in nonword repetition. Furthermore, the absolute magnitude of the deficits was greater for nonword repetition than serial recall, although the stimuli employed in these tasks differed substantially (i.e., in length, familiarity, and phonological properties). The discrepancy in the magnitude of the SLI deficits may reflect better memory for words than nonwords (see, e.g., Hulme, Maughan, & Brown, 1991), which would convey an adva (...truncated)