, 1998, Lee et al., 2002 and Ojemann and Dodrill, 1985). Studies of list-learning tasks have often found immediate recall to be impaired in SLI (Dewey and Wall, 1997, Lum et al., 2010 and Nichols et al., 2004), though in other studies normal performance has been reported (Riccio et al.,

2007 and Shear et al., 1992). Delayed recall more often seems to be spared in SLI find more (Baird et al., 2010, Riccio et al., 2007 and Shear et al., 1992), but sometimes shows impairments (Nichols et al., 2004). Studies have also been mixed with respect to delayed recognition for verbal information, alternatively reporting impaired (Nichols et al., 2004, Riccio et al., 2007 and Shear et al., 1992) or normal (Baird et al., 2010) performance in the disorder. Story recall seems to result in impaired immediate recall, but largely normal performance after a delay (Baird et al., 2010 and Merritt and Liles, 1987). Likewise, fast mapping tasks have yielded both deficits (Rice et al., 1990) and normal performance (Dollaghan, 1987). Importantly, most declarative memory task paradigms are subject to various confounds that may contribute to any observed deficits. In particular, at least the list and story learning paradigms depend heavily on working memory. Additionally, because verbal working memory tests involve language, the language deficits themselves in SLI could contribute to impaired performance

on these tasks. However, neither working memory nor language deficits have

been controlled for in any previous studies that we know of, and thus check details it Thiamet G remains unclear whether SLI is indeed associated with impairments of verbal declarative memory, once these factors are accounted for. Finally, a number of studies have examined associations between measures of memory and language. To date, most research has focused on associations between measures of phonological short-term memory or working memory with tasks probing grammatical processing. In this literature, it has generally been found that non-word repetition tasks only weakly correlate with elicitation tasks assessing past tense knowledge (Bishop et al., 2006, Botting and Conti-Ramsden, 2001 and Norbury et al., 2001). Correlations of a larger magnitude have been observed on tasks assessing phonological short-term memory or working memory with tasks of sentence comprehension (Montgomery, 1995, Montgomery, 1996, Montgomery, 2000, Montgomery, 2004 and Montgomery and Evans, 2009). We are aware of only one study examining associations between language and declarative or procedural memory in SLI. In Tomblin et al. (2007), initially separate groups of adolescents with and without SLI were then re-organised into other groupings. In one analysis, all the participants (SLI and TD) were organised into two groups comprising those who scored either high or low on vocabulary tests.