Abstract
The acquisition of the formal structure of an artificial language was examined using variants of the language that featured no cues, single extralinguistic or linguistic cues, and correlated pairs of these cues to word class. While performance on a grammatical judgment test was generally better for single cues than no cues and equal for all types of single cues, performance for correlated pairs of cues only exceeded that of corresponding single cues when both cues were fairly transparent as markers. In addition, distinctiveness of the markings played a role in ease of acquisition. These results clarify previous artificial language learning results, and are in accord with data from natural language learning.
Get full access to this article
View all access options for this article.
