The Limitations of Auditory-Phonetic Speaker Identification