Some notes on definitions in empirical science