This finding is consistent with previous studies, which have often reported small and non-significant correlations between working memory and grammar measures in SLI (see, Introduction). The results throw further doubt on strong versions of claims that working memory deficits alone can fully account for normal language development (Baddeley et al., 1998) and for the language impairments Oligomycin A in SLI

(Gathercole and Baddeley, 1990). It might be argued that an absence of a correlation between working memory and grammar (or indeed the potential absence of clear and consistent working memory impairments, as discussed above), contradicts the PDH (Bishop et al., 2006). However, the PDH claims that PF-02341066 manufacturer the primary, core, deficit in SLI is of procedural memory, which is mainly responsible for the grammatical impairments in the disorder. Working memory and other non-procedural functions that depend in part on the affected brain structures underlying procedural memory are expected to co-occur probabilistically with these core deficits. The likelihood of such co-occurrence depends on factors

such as the anatomical proximity of those portions of the affected structures (e.g., frontal/basal-ganglia circuits) responsible for these functions to those portions that underlie procedural memory (and in particular, to those portions that underlie those aspects of procedural memory that subserve grammar) (Ullman and Pierpont, 2005). Indeed, as we have seen above (see, Introduction), procedural memory seems to depend more on BA 44 and premotor frontal regions, and working memory more on other prefrontal areas, including BA 46 and BA 45/47. Thus, although the PDH expects that the neural abnormalities underlying procedural memory may often extend to these frontal C-X-C chemokine receptor type 7 (CXCR-7) regions subserving working memory (and

the portions of the basal ganglia they are connected to), such abnormalities, and their accompanying functional deficits of working memory, are not expected to be a core feature of the disorder, and are unlikely to constitute the primary cause of the language problems in SLI (Ullman, 2004, Ullman, 2006a and Ullman and Pierpont, 2005). The findings reported here may also help inform other explanatory hypotheses of SLI. The observed memory deficits, in particular of visuo-spatial procedural memory, contradict strong versions of hypotheses that posit that only deficits of language, in particular of grammar, occur in SLI (Rice, 2000 and van der Lely, 2005). The correlation between declarative memory and grammatical abilities in SLI is also problematic for such hypotheses. Additionally, this correlation is not expected on the view that the language problems in SLI are explained by phonological deficits (Joanisse, 2004).