Firstly, the study was mostly qualitative, and still needs good quantitative research to substantiate many suppositions made. Also, the speaking tasks were done for homework, making it difficult to monitor for planned versus unplanned speech. Further, the transition from declarative to procedural knowledge requires considerable practice and time (Anderson 1983, 1985). It is virtually impossible to give students enough practice over five weeks, and the duration is much too short to expect any major changes. Lastly, and perhaps most importantly, the study did not have a control and noncontrol group to measure differences