[ NEANDERBONK VOICE ]

> Who said that? Two speaker-attribution approaches, one microphone