The article is intended as a voice in the discussion on language aptitude, with particular regard to its two components: memory and language analytic ability. It will be argued here that – unlike memory, favoured by Skehan (2003) – it may be the language analytic ability manifested, among others, as considerable dexterity in re-trieving constructional schematizations to decode language innovation, which grows in importance with learner proficiency. It will also be stated that both capacities, the said ability as well as memory, should be considered in relation to working memory, which should be understood in terms of storage and processing considered separately and ascribed to individual differences, and not as a homogenuous storage-and-processing space.
To verify the above claim, a study was carried out in the years 2007-2008 in three groups of advanced EFL learners (N=60) at three different levels of language proficiency (B1/B2, N=20; C1, N=20; and C2, N=20). All testees were asked to solve two tests which required interpreting 32 (16/test) skeletal sentences containing schematic representations of events such as X verbed Y. The only given in each sentence was the verb, a product of noun-to-verb conversion1 like to bottle or to buoy. In Test 1 the constructions chosen for interpretation were highly schematic (in-, mono- and ditransitive); as a result, the testees had to deal with sentential constructions such as X bottled Y or X buoyed. Test 2, on the other hand, included examples of complex substantive constructions2 such as X buttered home (where the verb slot is reserved for verbs of motion) or X kept Y bungeed (where only the fi nal slot is open to inter-pretation). The present article presents a comparative analysis of the results of both tests on the three different levels mentioned above. Their interpretation and following conclusions are based on VanPatten’s Input Processing (VanPatten 1990, 2004), Cowan et al.’s model of working memory (Cowan et al. 2005), and Truscott and Sharwood-Smi-th’s Acquisition by Processing Theory (APT; Truscott and Sharwood-Smith 2004). Towards the end of the article, all this is related to the discussion of language aptitude: its components, with special regard to the afore mentioned two: memory and analytic language ability; the importance of the two components as regards different perspec-tives on language aptitude (CALP vs. BICS; Cummins 1983).