Longitudinal study, corpus of biographical interviews. 100 hrs of recorded speech, more than 400.000 tokens transcribed, so far.
