Partially Automated Techniques for Transcribing Naturally Occurring Continuous Speech