Techniques for context-sensitive language learning are disclosed. For example, a language learning system may include an interface for communicating with at least one user, at least one sensor for collecting at least one form of data regarding the context in which the system is being used, and a processing device capable of making at least one adjustment to the communication with the user based on analysis of at least a portion of the data collected by the at least one sensor. The data may include audio data, visual information, biometric data, location, or velocity and the sensors may include a microphone, a camera, a biometric sensor, a global positioning system (GPS) device, or a velocimeter. The system may also use this data, alone or in combination with schedule data obtained from an external source, to determine the attention level of the user and to make corresponding adjustments to the communication. The system may further be capable of tracking changes to the data collected by the sensor and/or the number and/or type of errors made by the user and making corresponding adjustments to the communication.