Who is speaking

A speaker is named only where the evidence supports it: the publisher's own metadata, the transcript's labels, or an introduction on air. A voice nobody can name stays unnamed.