Children with developmental language disorder (DLD) show significant difficulties mastering language yet exhibit normal-range nonverbal intelligence, normal hearing and speech, and no neurological impairment. Deficits in sentence comprehension represent a major feature of school-age children’s language profile. So do memory limitations, including deficits in verbal working memory, controlled attention, and long-term memory. Though there is general consensus that the memory and comprehension deficits of these children relate in some fashion, the relationship has historically been unclear. In this chapter, we present the first conceptually integrated and empirically validated model of the sentence comprehension abilities of school-age children with DLD that describes the structural relationship among all these abilities.

Literature has documented that working memory (WM) is one of the best predictors of literacy and word reading by children and adults. This chapter summarizes some of previous studies on word reading from the perspective of two distinctive modalities in the Baddeley’s multicomponent WM model (i.e., phonological and visuospatial aspects), reviews recent studies on WM in Chinese word reading, and suggests issues that can be further explored in WM and word reading

According to some researchers, different languages foster specific habits of processing information, which may be retained beyond the linguistic domain. In left-branching languages, for instance, the head is usually preceded by its dependents, and real-time sentence comprehension may require a different allocation of attention as compared to right-branching languages. Such sensitivity to the branching of languages may be so pervasive to also affect how humans process stimuli other than words in a sentence. In this chapter, we will review previous studies on the link between word order, statistical learning habits, and attention allocation, and specifically discuss the effects that branching habits may have on working memory processes, well beyond the linguistic domain. We will conclude by fostering a stronger cross-linguistic approach to the study of branching and working memory, and suggesting possible lines for future research.

Drawing on work from cognitive psychology, a vast body of research has examined the role of working memory (WM) in second-language (L2) development, processing, and use (e.g., Linck et al., 2014). Our ability to discern such relationships, however, may be obscured by the different measures of WM that are adopted and employed by L2 researchers. There exist a daunting number of WM tasks, and the variability in task design, implementation, and scoring adds to the challenge of accurately measuring WM capacity and interpreting findings. To this end, this chapter presents a methodological synthesis that surveyed the literature spanning 20 years (2001-2020) of WM-L2 research to describe the use of six common WM tasks. We coded a total of 329 unique samples on a range of features related to the WM tasks and reporting practices. Our findings suggest that, among the six most common WM tasks used in L2 research, task design and scoring procedures tend to diverge more than converge within and among the tasks. We also found neglected areas in reporting practices in the WM-L2 domain. Based on our findings, we provide insights into the use of the WM tasks and future directions to help researchers make informed decisions for measuring WM and interpreting findings critically.L38

The role of working memory in language learning has received considerable attention, but several pertinent issues remain. One of these concerns the directionality of the relationships between working memory and language learning. Another issue relates to different types of processing and working memory components involved in learning different aspects of a second language (vocabulary, grammatical sub-skills, e.g., subject-verb agreement, verb placement, word order, auxiliaries). In this chapter we review and integrate findings of previous studies, following the extraction and integration model (Thiessen et al., 2013), and apply these to second language learning. In so doing, we distinguish between statistical learning based on conditional relations of adjacencies (extraction) and statistical learning based on distributional patterns of non-adjacencies (integration). We propose how L2 children's gradual increase in knowledge of the second language increases the sensitivity of working memory to cues in ambient speech that, in turn, fosters further second language learning.

This chapter explores the dynamic relationship between working memory (WM) and grammar development across adult L2 learning. For over twenty years, WM has received considerable attention in research on adult second language (L2) development. One reason for this is that L2 learning requires both processing and storage to comprehend input and to extract intake for acquisition, so differences in WM capacity may explain differences in developmental rates. Most studies on WM and morphosyntactic development in adults support the “more is better” hypothesis (Miyake & Friedman, 1998); yet others did not yield evidence in its support (e.g., Foote, 2011; Grey, Cox et al., 2015). While linguistic targets and methods may explain many discrepancies, recent research (e.g., Serafini & Sanz, 2016) may also help us understand these differences as a reflection of changes in what constitutes a cognitively demanding task (i.e., what tasks recruit WM resources) across L2 learning

A distinguishing feature of the cognitive process of speech planning is its flexible balancing of speed, quality, and effort. Utterance planning strategies can vary adaptively depending on speaker goals and circumstances. For example, when speed is a priority, the planning process might sacrifice the quality of an utterance by engaging in more incremental, on-the-fly planning. A focus on utterance quality may require more time. But sometimes, speakers seem to plan utterances well in advance without sacrificing quality or speed. In this chapter, we focus on recent research that explores how working memory can foster the flexibility of speech planning strategies. We review the role that WM might play in individual levels of planning, including message planning, grammatical encoding (including lemma selection and structure building), and phonological encoding, and the extent to which the scope and quality of planning at these different levels could be subject to WM constraints or predicted by WM capacity. We conclude that WM is a (sometimes optionally invoked) part of a complex system of compensatory factors that can determine how speech planning unfolds.

We view working memory as a general resource in which attention can be allocated to any type of information and stimulus input. One vital skill that requires the use of working memory is the comprehension and production of language. In this chapter, we outline the basis of the embedded-processes model of working memory. We then discuss how the different parts of the model might be relevant to language use. Finally, we discuss how working memory is used in the acquisition of language as children develop or as individuals take on a second language. We wish to provide a basis for understanding the importance of attention and its interaction with long-term memory for the successful use of language, from understanding a simple sentence to producing a well-formed set of sentences for communication.

This chapter starts by providing brief accounts of both first and second language speaking, and then surveys empirical work, measurement issues, and theory on the use of second language speaking tasks – the sort of tasks, often with real-world connections, used in communicative language classrooms to nurture second language development and performance. The main section of the chapter is concerned with the relationships between working memory and performance on such tasks. Broadly it is argued that, as yet, there are not many systematic findings relating task characteristics, working memory, and actual performance. In contrast, the conditions under which second language speaking tasks are done (such as planning opportunities, repeated tasks) do show some interesting results. Based on such research, it is argued that working memory plays more of a role in the Formulation stage of speech production. Proposals are made regarding the areas where it would be most helpful to research working memory connections with second language speaking tasks.

To conceptualize the communicative role of working memory (WM), the Ease-of-Language Understanding (ELU) model was proposed (e.g., Rönnberg, 2003; Rönnberg et al., 2008, 2013, 2019, 2020). The model states that ease of language understanding is determined by the speed and accuracy with which the signal is matched to existing multimodal language representations. When matching is fast and complete, language understanding is effortless; this process may be facilitated by predictions based on the contents of WM. However, when the contents of the language signal mismatches with existing representations, WM is triggered to access knowledge in semantic long-term memory (SLTM) and personal experience from episodic long-term memory (ELTM) – promoting inference-making and postdictions in WM. The interplay between WM and LTM is fundamental to language understanding; its efficiency becomes apparent in adverse conditions and its breakdown may explain cognitive decline and dementia. Empirical support, limitations, and future studies will be discussed.

Many general linguistic theories and language processing frameworks have assumed that language processing is largely a chunking procedure and that it is underpinned and constrained by our memory limitations. Despite this general consensus, the distinction between short-term memory (STM) and working memory (WM) limitations as they relate to language processing has remained elusive. To resolve this issue, we propose an integrated memory- and chunking-based metric of parsing complexity, in which STM limitations of 7 ± 2 (Miller, 1956a) are relevant to the Momentary Chunk Number (MCN), while WM limitations of 4 ± 1 (Cowan, 2001) are relevant to the Mean Momentary Chunk Number (MMCN). Examples of concrete calculations of our new metric are presented vis-à-vis Liu’s MDD metric and Hawkins’ IC-to-word Ratio metric. Related methodology issues are also discussed. We conclude the paper by echoing some recently repeated calls -(O'Grady, 2012 & 2017; Gómez-Rodríguez et al., 2019; Wen, 2019) to include STM and WM limitations as part and parcel of the language device (LD; cf. Chomsky, 1957) in that their impacts are ubiquitous and permeating in all essential linguistic domains ranging from phonology to grammar, discourse comprehension and production.

The high working memory demands of writing are now well documented across the development of writing as well as in expert writing. The limited capacity of working memory of beginning writers and their effortful untrained writing processes indeed constrain learning to write. In experienced writers, operations of the writing processes are more complex and therefore continue to heavily engage working memory. In this frame, this chapter describes the theoretical models of the writing processes that describe the role of working memory. It then reviews research that examined how writing and the writing processes engage working memory. It is shown that managing the planning, translating, and revising processes involve the executive and nonexecutive components of working memory for storing verbal and visual spatial information that is processed during writing.