Abstract
The study demonstrated the effect of countability (countable vs. uncountable)as well as the effect of noun type (concrete vs. abstract). The result showed that article accuracy for concrete nouns was higher than that for abstract nouns in both countable and uncountable environments. The study also found that indefinite articles were more difficult than definite ones when both were modified by phrases or clauses. Furthermore, it was found that the proficient group was more accurate than the less proficient group in memorized-chunk structures, indicating a general tendency of the subjects to become dependent on them as their acquisition progressed. Finally, both groups were equally accurate in the selection of the definite article in subject position NPs introducing old information.
