It is difficult to draw a clear distinction between language and cognition. As proficient language users and thinkers, adults are constantly using language as inner speech to carry out various thinking processes, including solving problems and making decisions. The language we know can shape our perception as well. For example, young Korean toddlers are more sensitive than their age-matched English counterparts to the distinction between loose- and tight-fitting containment, a concept marked in Korean but not in English (Choi, McDonough, Bowerman, & Mandler, 1999). On the contrary, there is evidence of a language-independent system for thinking. Using a preferential looking paradigm, Hespos and Spelke (2004) reported that infants responded to conceptual distinctions not marked in their language, a sensitivity lost as language develops.

This close interaction between language and cognitive thought processes makes it difficult to design tasks targeting a single construct. For example, many measures commonly included in standardized omnibus tests of language require short-term retention of verbal material in memory. Sentence recall is one such task. The memory demands of repeating sentences are reflected in findings that repetition accuracy is reduced for long sentences compared with short sentences (Willis & Gathercole, 2001). Such findings, however, do not negate the linguistic processing inherent in the repetition of sentences. Indeed, semantic processing of words and sentences proceeds relatively automatically in many cases. The linguistic processing of words and sentences in recall tasks is suggested by reports of better recall for nonwords with high than low wordlikeness (Gathercole, 1995) and for sentences than words (Baddeley, Hitch, & Allen, 2009). Despite recognition that multiple linguistic and cognitive processes may be tapped in many verbal tasks, there remains uncertainty regarding the specific relationships and the extent to which this knowledge can be applied to the assessment and intervention of children with language impairments.

In this issue, the cognitive processes supporting a variety of verbal tasks, including sentence comprehension and nonword repetition, are considered. One cognitive resource of particular interest is working memory, the ability to briefly hold and manipulate information in the current focus of attention. It has been suggested that working memory plays an important role in language learning (Gathercole & Baddeley, 1993) and may be impaired in children with specific language impairment (SLI; Montgomery, 2002). Performance on immediate memory tasks, however, has been found to be influenced by linguistic factors such as vocabulary size (Snowling, Chiat, & Hulme, 1991), phonotactic knowledge (Windsor, Kohnert, Lobitz, & Pham, 2010), and syntactic complexity (Marton, Schwartz, Farka, & Katsnelson, 2006). As well, theoretical debate persists regarding whether or not working memory is a domain-general cognitive process separable from other constructs such as language (MacDonald & Christiansen, 2002), executive functioning, or intelligence generally (Conway, Kane, & Engle, 2003). The articles in this issue address these questions by considering evidence related to the cognitive processes contributing to performance on a number of verbal tasks commonly used by speech-language pathologists in the assessment of children with language-related communication disorders.

In the first article, Archibald (2013) draws on a large database of performance results for school-aged children on standardized tests of language, short-term and working memory, and verbal and nonverbal intelligence to gain a better understanding of the underlying structure and relationships between these measures. One intriguing finding from the article is that measures commonly used in standard language batteries, such as following directions and repeating sentences, primarily loaded on a language processing factor, although following directions was also associated with working memory and repeating sentences with phonological short-term memory. As well, an interesting developmental pattern emerged, suggesting that older children draw on their established experience and knowledge and working memory when performing common verbal tasks whereas younger children must rely on their ability to analyze novel problems to complete these same tasks.

Next, Kidd (2013) provides a review of the literature pertaining to sentence comprehension. Kidd considered the suggestion that verbal working memory is not a conceptually distinct resource causally linked to language acquisition but a functional property of the language processing system. Although no firm conclusions could be drawn, it is widely acknowledged that long-term language knowledge must influence working memory functioning. Kidd argues that developmental and training studies are needed to better understand these resources and, in particular, how experience can influence functional capacity for sentence comprehension.

The remaining three articles address the issue theme through the lens of nonword repetition. Nonword repetition was first proposed as a relatively pure measure of short-term memory (Gathercole & Baddeley, 1993). Nevertheless, many researchers (see Gathercole, 2006) have since acknowledged the importance of additional factors that influence nonword repetition. Stokes, Moran, and George (2013) and Leclercq, Maillart, and Majerus (2013) take up the theme introduced by Kidd (2013) that language knowledge held in long-term memory accounts for differences in working memory tasks and apply this thinking to the case of nonword repetition. According to this view, poor nonword repetition reflects weakly established, or difficulty accessing, phonological representations from long-term memory. The findings from Stokes et al. (2013) suggested that the nonword repetition of late talkers was constrained by the ability to access phonological representations from long-term memory. Similarly, Leclercq et al. (2013) found that children with SLI showed a reduced benefit in repeating words than nonwords, especially under conditions of high syllabic complexity. These results might reflect more poorly developed lexical and sublexical knowledge or a reduced ability to access this knowledge. Both of these articles are important because they present methods for examining the nature of long-term representations in SLI, a resource that has been difficult to target empirically. Finally, Archibald, Joanisse, and Munson (2013) examined motor speech influences in nonword repetition. An intriguing finding from the article is that only language impairment (and not working memory impairment) was associated with reduced accuracy in the repetition of motorically complex words in school-aged children. The results provide further evidence of a motor component to SLI (see also Goffman, 2004).

This issue is aimed at providing further consideration of the cognitive processes that influence and interact with language abilities. In particular, the construct of working memory, its separability and its relationship with long-term linguistic representations, is examined. Across several of the articles in this issue, long-term, experience-based language knowledge emerges as a distinct resource influencing performance on tasks commonly used clinically to assess language and constraining performance in groups with SLI.

-Lisa M. D. Archibald

Issue Editor

REFERENCES