The present chapter bridges two lines of neurocognitive research, which are, despite being related, usually discussed separately from each other. The two fields, second language (L2) sentence comprehension and artificial grammar processing, both depend on the successful learning of complex sequential structures. The comparison of the two research directions will be taken as the starting point for the attempt to study L2 sentence comprehension using a miniature language model. The experiments that will be presented made use of the event-related potential (ERP) method, which provides a sensitive tool for the fast and multidimensional processes that characterize online processing of sequential structures. The first part of the article provides an overview of ERP studies on first and second language sentence comprehension and compares these to findings from the field of artificial grammar processing. In the second part, two studies in which native and nonnative sentence processing in the model of a miniature version of Japanese was investigated is presented. The findings of these studies suggest that miniature languages can serve as useful tools for capturing the impressive degree of plasticity as well as certain limitations in language comprehension mechanisms applied by learners.