Introduction to the special issue on syntax and verbal short term/working memory in developmental disorders

Research on developmental disorders has provided a detailed description of the linguistic and cognitive characteristics of individuals with these disorders. To adequately describe and explain the language and its relation to cognitive abilities of individuals with developmental disorders, researchers have employed theoretical apparatus and tools developed within different disciplinary frameworks and approaches. In particular, they have mostly used theoretical frameworks provided by linguistics and cognitive psychology (especially the cognitive psychology of memory).

On the one hand, linguistics has been a fruitful framework within which to interpret the language difficulties of individuals with developmental disorders. Studies have targeted specific linguistic domains, i.e. phonology, morphology, syntax and/or pragmatics, and have examined performance in detail using linguistically sophisticated tasks. Linguistically-orientated research has revealed crucial clinical linguistic markers for developmental disorders. By so doing, this research has contributed to differential diagnosis of these developmental disorders or revealed shared clinical phenotypes among different disorders (Caselli et al., 2008; Rice, 2000; Sukenik & Friedmann, 2018, among others).

On the other hand, developments within the multicomponent model of memory (Baddeley & Hitch, 1974) have offered new insights in the interpretation of language difficulties in individuals with developmental disorders. Specifically, within this model, the phonological loop temporarily stores verbal information and thus it can be directly associated with linguistic performance. The ability to temporarily store verbal information has been referred to as Verbal Short Term Memory (VSTM). In addition, the ability to keep verbal information in mind while manipulating it has been referred to as Verbal Working Memory (VWM). Within this approach, language impairments have been understood as a secondary consequence of a primary impairment in the domain of VSTM or VWM or in another component of this multicomponent model. The majority of these studies indicate severe impairments in the domain of VSTM/VWM or in other memory domains, and their authors interpret language impairments in terms of impaired memory abilities (Archibald & Gathercole, 2006; Marton & Schwartz, 2003, among others).

Notably, the above lines of research have been developed to a great extent independently, since linguistic approaches were served mostly by linguists and memory approaches by psychologists. However, recently there have been studies which simultaneously investigated linguistic and verbal memory skills and highlighted associations or dissociations in cognitive domains of individuals with developmental disorders. For example, an early and influential study by Conti-Ramsden et al. (2001) examined a large number of English-speaking children with a history of Developmental Language Disorder (DLD) on linguistic and memory measures. Specifically, these researchers tested third person singular, past tense, nonword repetition and sentence repetition. While vulnerable performance was shown in all these domains, the most adequate marker for DLD identification was the sentence repetition task. In addition, a landmark study by Bishop et al. (2006) investigated phonological STM and tense marking in a large DLD sample and found minimal correlation between phonological STM impairments and tense marking. This study revealed that the grammatical and STM difficulties in individuals with DLD manifest independently. Therefore, a single clinical marker (STM limitations or a grammar deficit) may not adequately describe and account for the diversity of clinical profiles of individuals with DLD.

In addition, interpretations of impairments in verbal memory and/or language and their relationship in developmental disorders have been to a great extent ‘disorder specific’. Specifically, in some developmental disorders a closer relationship is presumed than in others. For example, with respect to Down syndrome, a number of studies show poor language and verbal memory abilities. Næss et al. (2011) performed a meta-analytic review of language and VSTM skills in children with Down syndrome. They found that children with Down syndrome have broad language impairments associated with verbal short term memory impairments. However, with respect to other disorders, DLD, for instance, controversies are more evident.

Taking into account the aforementioned research situation, the present special issue of First Language aims at contributing to the discussion of the relation between VSTM/VWM and language in developmental disorders in the following ways.

First, it focuses on the relation of expressive and/or receptive syntax to VSTM and VWM. Crucially, syntax requires demanding computational processes which are considered to be cognitively challenging. Impaired syntactic abilities have been extensively reported for many developmental disorders. Investigating whether and to what extent such impairments are associated with verbal memory provides a better understanding of how disorders manifest.

Second, empirical data are reported from different disorders. New insights are provided by raising methodological issues regarding the study of syntactic abilities in these disorders, by studying bilingual individuals alongside monolinguals, and by investigating syntax and verbal memory in understudied – from this particular perspective – disorders such as autism. Furthermore, bringing together data from different disorders allows inspection of the syntax–verbal memory relation beyond each particular disorder profile.

Thirdly, a cross-linguistic approach is adopted, as the empirical data come from a number of different languages (German, Greek, English, French, Hebrew, Italian, Russian). This provides a unique opportunity to see whether impairments extend beyond each language’s specific characteristics (in addition to each disorder’s characteristics).

Finally, investigating the language and VSTM/VWM abilities of individuals with developmental disorders promotes understanding of the relation between these cognitive systems and raises questions concerning the nature of these cognitive mechanisms per se. In this way current research provides crucial feedback for the reformulation of theories of language and memory as well as of human cognitive architecture.

More specifically, six empirical articles are included in this special issue. They investigate the relationship between syntax and VSTM/VWM in different disorders, namely high functioning autism (Meir & Novogrodsky), Down syndrome (Penke & Wimmer), deafness (Volpato), specific learning difficulties (Stanford & Delage) and developmental language disorder (Talli & Stavrakaki, Zebib et al.).

A couple of these studies report no interdependence between syntax and verbal memory in an absolute or partial way. Specifically, Meir and Novogrodsky reported no association between syntax and verbal memory in a sample of Hebrew-speaking monolingual and Russian–Hebrew speaking bilingual children. These researchers investigated whether variation in the language abilities of individuals with high functioning autism can be predicted by verbal memory. In their sample, some individuals with autism showed intact syntactic abilities, while others scored at-risk for language disorder. Notably, children with autism who were at-risk for language disorder and children with no risk showed the same level of performance on the VSTM/VWM tasks. This finding can be interpreted as indicating no impact of verbal memory on syntactic abilities. In addition, this study examined whether there was an effect of bilingualism on sentence repetition and verbal memory. It showed that, when vocabulary was controlled for, there was no effect of bilingualism on either sentence repetition or verbal memory.

A partial relation between VSTM and syntax is reported by Penke and Wimmer. These researchers explored whether there was a task effect on the (presumed) relation between VSTM and syntax in German-speaking individuals with Down syndrome. The results indicated that while VSTM predicted syntactic performance on two sentence–picture-matching tasks it did not predict performance on a picture-pointing task. These findings indicate that the (presumed) strong relation between VSTM and syntax in Down syndrome may not actually be that strong as intervening factors such as task demands have not been adequately considered up to now.

By contrast, other researchers showed a close interaction between syntax and verbal memory. In particular, Volpato showed a strong effect of verbal memory on syntax. She investigated whether relative clause comprehension in Italian-speaking children with cochlear implants is related to measures of VSTM and VWM. The results indicated a significant involvement of verbal memory in relative clause comprehension, as forward digit span was found to be a significant predictor of syntactic comprehension for some types of relative clauses. Volpato argues that the weak VSTM of children with cochlear implants affects phonological processing and representations and leads to impaired syntactic comprehension.

By exploring the relation of VWM to the comprehension of complex sentences containing intervention in a sample of English-speaking individuals with specific learning difficulties, Stanford and Delage argue – like Volpato – in favor of the VWM contribution to poor syntactic performance. More specifically, they suggest that structures with moved object and an intervening subject require VWM capacity. Individuals with specific learning difficulties who have VWM limitations show impaired syntactic performance on structures with intervention.

Two of the articles included in this special issue focus on language and verbal memory in the challenging context of bilingualism, and they provide evidence for a differentiated effect of VWM vs VSTM on syntax. Specifically, Talli and Stavrakaki investigated possible correlations between VSTM and VWM abilities and lexical and syntactic abilities in Greek-speaking monolingual and bilingual children (Albanian or Russian as an L1 and Greek as an L2) with DLD. These researchers found that VSTM was a significant predictor of syntactic performance for both monolinguals and bilinguals, while VWM was a significant predictor of syntactic performance only for bilinguals. They also found that both VSTM and VWM predicted performance of both DLD groups to a larger extent than the performance of typically developing groups on the syntactic measures, suggesting a closer developmental interaction of memory and syntactic abilities in individuals with DLD than in typically developing individuals. Individual analysis was also performed and poor syntactic performance co-occurred with low performance in VSTM for most but not all children with DLD.

Zebib et al. investigated whether sentence repetition is related to VSTM and VWM in bilingual individuals with DLD with Arabic, Portuguese or Turkish as their L1 and French as their L2. The results indicated that VWM functions as a predictor for sentence repetition in the Bi-DLD group, but not in the typically developing group. This finding is taken to show that individuals with DLD rely more on VWM than individuals with typical development. These authors performed individual analysis which showed that performance on sentence repetition tasks is independent of minimal VSTM or VWM spans.

In sum, the present issue brings together studies on the relation between syntax and VSTM/VWM in individuals with different developmental disorders, speaking different languages. In the light of new data, this issue raises the question of the relation between syntax and verbal memory and provides feedback for (re)considering this relation as well as each function per se (syntax, VSTM, VWM).

This special issue would not have been possible without the valuable encouragement and the endless support of Chloë Marshall, whom I would like to thank for bringing this issue to fruition. Furthermore, this issue would not have been possible without the authors’ contributions. I would also like to thank them. Finally, I would like to thank Maria Teresa Guasti, Chloë Marshall, and Nick Riches, who accepted my invitation to act as discussants and provided thought-provoking commentaries. Guasti argues for a weak relation of language and verbal memory (VSTM/VWM) in individuals with developmental disorders. In particular, she suggests that any involvement of memory has to do with the sequential/linear aspect of language and not with the predominant feature of language, which is its hierarchical organization. Marshall points out that VSTM/VWM cannot be measured independently of language and thus any claims concerning the language–memory correlation may not be correct. She doubts whether continuing to explore this relationship is profitable. Riches develops a skeptical account of STM and WM and claims that there is no need to posit such mechanisms. He proposes that future studies should address the causal relationship between STM/WM and language abilities by performing experimental rather than correlational research and by conducting longitudinal studies. To summarize, all discussants raise crucial theoretical and methodological issues. These issues can give impetus for performing further studies in the field of developmental disorders in the near future.