Abstract
This chapter investigates the acquisition of obstruent laryngeal contrasts in foreign language acquisition. The goal is to determine the alignment between the putative universal markedness relationships established through cross-linguistic investigation and the patterns found in second language phonology, particularly those patterns that appear to be independent of both the native and the foreign language systems. Cross-linguistic research has revealed that obstruent laryngeal contrasts are more common in nonfinal than in final positions; that when contrast is limited in final position, voiceless obstruents are the preferred segment type; and that the preferred repair for underlying voiced obstruents in final position is devoicing of the obstruent rather than any of the logically possible alternatives such as post-obstruent vowel insertion. A survey of studies on the acquisition of foreign language laryngeal contrasts supports a hierarchy of difficulty in the acquisition of foreign structures that is consistent with the principles established by typological investigation: learners from a wide range of native language backgrounds show earlier success in mastering final voiceless than final voiced obstruents, even when the native language has neither and the foreign language has both, while the opposite order of acquisition is not attested. Furthermore, devoicing is frequently found in second language phonology even in the absence of a phonological devoicing process in either the native or the foreign language. However, second language learners do exhibit some tendencies that are not predicted by established typological generalizations, such as effects of word size and of manner and/or place of articulation on the likelihood of obstruent devoicing. We consider possible explanations for these tendencies as well as the question of whether these tendencies ever become phonologized as categorical processes in established native language grammars.
