Slow cortical potentials (SCP) are among the oldest candidates for an EEG-based language. Early experiments with biofeedback largely relied on these SCP and on the fact that people can, to a certain degree, learn to control them. Negative SCPs are typically associated with movement while positive ones are a sign for reduced cortical activity. SCP are among the lowest frequency signals that can be obtained via EEG. Their frequency ranges between no more than 0.1 and 0.5 Hz.
BCIs that use SCP are referred to as thought-translation-devices. They are successful in enabling people with severe ALS (more) to control the movement of a cursor on a screen, and do currently supply quite a few of these severely handicapped people with basic communication capabilities.
Users of SCP-based BCIs practice for weeks or months with a training-unit that offers only up-and-down choices combined with feedback (more). As soon as this works reasonably well, they switch to a language support programme (more). In practical terms, SCP-systems are clearly the most succesful ones so far, but they do have some problems (more).