Technology that analyzes written human language displays compelling possibilities for computer assisted language learning (CALL). Applications may be designed to examine second language students’ free text production in order to suggest improvements, draw attention to selected linguistic elements, provide examples from native language use, etc. However, language technology is not free from issues. Output from the tools is occasionally inaccurate, and the tools’ emphasis on language structure stands in possible contrast with pedagogies that foreground the social nature of language.

These difficulties cannot be resolved by improving technology alone. It is equally important to find out in empirical detail how students handle possibly problematic directions from particular tools, and how tasks and teacher guidance may affect students’ tool usage. This thesis provides detailed analyses of applied CALL activities with a language tool, in order to discern how usage of that particular tool occurs, and how complex interrelationships of tool and context of use direct students.

The thesis makes six interrelated contributions, ranging from particular empirical results to implications for general theory and methodology in CALL. On the level of theory, the thesis 1) provides an argument to reintroduce language technology in CALL, and 2) suggests grounding in sociocultural theory for investigating second language classroom CALL as it unfolds. From these standpoints, it 3) develops methodology and empirical studies on the use of a particular tool in the hands of students in class. The studies result in 4) illuminations of problems in use and means to avoid these problems by attending to classroom setting. Returning to general theory, the results provide 5) evidence that the introduction of novel tools demands more than merely providing technology for students and teachers, suggesting 6) a need for more detailed considerations of how CALL tools are introduced to students.