The think-aloud method is used in many empirical studies and is regarded as an important tool within the field of second language acquisition research. It is used to gain insight into cognitive processes. However, few studies include a critical reflection on the cognitive basis which this method is based upon. Furthermore, central terms such as ‘introspection’, ‘verbal protocol’ or ‘think aloud-protocol’ are used in different ways which makes it difficult to compare results among studies.
After clarifying central terms, this paper discusses the think-aloud method and its critics and compares it to other forms of verbalisation which, although similar on the surface level, show crucial differences with regard to what kind of data is elicited. Special emphasis is laid upon the question in which language think-aloud protocols should be carried out in second language research.