Hearing an utterance of a word or construction that is interesting, and then writing it down or asking to record it is what was and still is done very often in linguistic research. This can be combined with participant observation, fieldworker distributed questionnaires or rapid and anonymous surveys. Transcribing or recording on the spot is the best way to gain data for words and phrases in natural conditions and getting around the observer's paradox. The biggest problem is that usually an informed consent must be acquired if the noted or recorded utterance is to be used in research (cf. Milroy and Gordon 2001: 14).
Created with the Personal Edition of HelpNDoc: Produce electronic books easily