controls, but not on regular forms (Clahsen & Almazan, 1998; Penke & Krause, 2004; Zukowski, 2004). However, no study to date has reported that individuals with WS outperform mental‐age controls, either on regulars or irregulars. Furthermore, the results of studies that have employed a larger number of participants (such as Thomas et al., 2001) show no interaction between group and regularity. As pointed out by Brock (2007), all studies that have investigated the performance of individuals with WS on regular versus irregular inflections suffer from “ceiling” effects, in that most of the participants in all the studies perform at ceiling on regulars, which makes it impossible for any group differences on irregulars to be found.
A large body of research has provided evidence for impaired morpho‐syntactic abilities in individuals with WS (Joffe & Varlokosta, 2007; Karmiloff‐Smith, Brown, Grice, & Peterson, 2003; Karmiloff‐Smith et al., 1998; Mervis & Klein‐Tasman, 2000; Stojanovik, Perkins, & Howard, 2001, 2004; Thomas et al., 2001; Volterra, Caselli, Capirci, Tonucci, & Vicari, 2003). Some of these studies show that not only do the individuals with WS not have superior language abilities, but that their language abilities may be on a par with those of children with diagnosed language impairments. For example, Stojanovik et al. (2004) compared the performance of a group of participants with WS aged between 6 and 12 and a group of participants with specific language impairment (SLI) or developmental language disorder (DLD) on a range of receptive and expressive standardized verbal measures, including measures of morpho‐syntax, and found similar performance across the two groups on all standardized language measures. The children with WS were completely indistinguishable on measures of morpho‐syntax from children with clinically diagnosed language impairment. Difficulties with the understanding of different morphosyntactic structures were also reported by Joffe and Varlokosta (2007). The participants with WS in their study, with a mean chronological age of 8 years 9 months, performed significantly less well on standardized and non‐standardized measures of grammatical ability than younger typically developing controls. The children with WS in their study had difficulties with the understanding and production of more complex syntactic structures like wh‐questions, and with the comprehension of passives, and performed lower than nonverbal mental age‐matched controls. Further evidence of specific difficulties with passive contractions relative to verbal and nonverbal mental age, and specific difficulties with passives when the verbs are psychological verbs, comes from a study by Perovic and Wexler (2010).
Some evidence suggests that the older the individuals with WS, the better their language abilities. The seminal studies by Bellugi and colleagues (Bellugi et al., 1988, 1994) had adolescents in their studies. Jarrold, Baddeley, and Hewes (1998) also showed that the verbal advantage in individuals with WS becomes more prominent as they get older. More recent studies, such as Musolino, Chunyo, and Landau (2010), which argue that individuals with WS have acquired the same core aspects of the computational component of language as neurotypical individuals, also report data from adolescents and adults with WS.
Evidence from older children and adolescents with WS also shows that they are able to extract patterns in language in an artificial grammar learning (AGL) paradigm, but possibly by relying on exemplar‐based rather than rule‐based processing strategies when processing language‐like stimuli. AGL performance is related to a person’s ability to predict language input based on the semantic and statistical context (Conway, Bauernschmidt, Huang, & Pisoni, 2010). A recent study by Stojanovik et al. (2018) reported that although the participants with WS (aged between 6 and 18) were outperformed by typically developing controls, the differences between the groups disappeared when nonverbal mental abilities were taken into account. Furthermore, the individuals with WS seemed to rely more on sequence familiarity rather than grammaticality when accepting or rejecting test sequences, and were able to extract patterns from the artificial grammar stimuli only when the training stimuli were presented with prosody. Such a finding fits well within the hierarchical framework of speech segmentation cues proposed by Mattys, White, and Melhorn (2005), in which it is suggested that prosodic cues are the lowest weighted cues in adult speech segmentation and the most critical ones at the onset of language development (lower than segmental and lexical cues). It appears that individuals with WS seem to be “stuck” in the lowest part of the hierarchy where they need prosodic cues in order to be able to make reliable acceptability judgments. And because of their strong preference for local processing, they need the prosodic cues to help them decide which sequences are acceptable. This suggests a possibly atypical processing strategy in WS.
There have been fewer studies of infants and toddlers with WS compared with ones on children, adolescents and adults, and although informative, the evidence is inconclusive. There is evidence for an asynchronous relationship between language and other cognitive abilities due to poorer language than cognitive abilities (Stojanovik & James, 2006), but also due to stronger language than cognitive abilities (Mervis & Bertrand, 1997). It is generally agreed, though, that the onset of language acquisition in WS in the early stages is delayed (Paterson, Girelli, Gsodl, Johnson, & Karmiloff‐Smith, 1999; Semel & Rosner, 2003; Stojanovik & James, 2006). The median age at which infants with WS have a vocabulary of 10 expressive words is 28 months, compared with 13.5 months in neurotypical infants (Mervis, Robinson, Rowe, Becerra, & Klein‐Tasman, 2003). Based on longitudinal data from Hebrew‐speaking children with WS, Levy and Eilam (2013) reported that infants with WS started to use combinatorial language on average 24 months later than their typically developing peers.
The reasons for the early language delay are still unknown. A recent study of statistical learning of young infants with WS aged between 8 and 20 months showed that they were able to discriminate between statistically defined “words” and “part words” in an AGL paradigm (Cashon, Ha, Estes, Saffran, & Mervis, 2016). This shows that infants with WS were able to detect statistical regularities in the speech stream, which suggests that the delayed early lexical acquisition of infants with WS may not be the result of their inability to segment words from the speech stream.
Early language acquisition in WS is generally an under‐researched area, and more longitudinal studies are needed in order to find out how language is acquired in WS and the relationship between language and other cognitive abilities.
Despite a large body of research into the morpho‐syntactic abilities of people with WS, there have been fewer studies investigating actual communication or pragmatic skills in this population. There have been reports that good social communication skills are a “hallmark” of the syndrome (Jones et al., 2000). Mervis, Klein‐Tasman, and Mastin (2001), using the Vineland Adaptive Behaviour Scales (VABS; Sparrow, Balla, & Cichetti, 1984), reported that communication in WS is a relative strength. In particular, Jones et al. (2000) argued that superior social‐communication skills distinguish this population from populations with other developmental disorders, such as DS and autism. In a series of tasks, Jones et al. (2000) reported that children with WS include a higher number of inferences about the affective state and motivation of story characters in comparison to typically developing children and children with DS. In the same study, individuals with WS provided a greater number of descriptions of affective states and evaluative comments during an interview task, and were more likely to ask questions of the interviewer. This was interpreted as showing that individuals with WS are “hypersocial.”
There have also been clinical and parental reports, as well as a number of research studies, that have shown that individuals with WS have problems in establishing friendships and have social difficulties, such as disinhibition and social isolation (Davies, Udwin, & Howlin, 1998).
Some studies have shown that individuals with WS are not sensitive to the needs of the conversational partner (Udwin & Yule, 1991). Further to this, Stojanovik et al. (2001) reported a high level of conversational inadequacy in a pilot study of a group of children with WS, in which participants with WS were found to have a tendency not to provide enough information for the conversational partner. Laws and Bishop (2004) reported pragmatic language impairment and social deficits in a group of older children and young adults with WS, using the Children’s Communication Checklist (Bishop, 1998). The checklist ratings showed pragmatic language deficits, evident from inappropriate initiations of
