Written records describe the people who could write, which is a narrow group in most periods.
was developed to reach the rest.
An interviewer records a person describing what they did and saw.
Historians were slow to accept the method for an obvious reason.
Memory is , and an interview held fifty years later records the present as much as the past.
The reply that settled the argument was not a defence of accuracy.
It was that a document is not either, and its are harder to notice.
A tax record tells you what an official chose to count.
An interview tells you what a person chose to say, and you can hear them choosing.
Method has tightened considerably since then.
Interviews are recorded in full, questions are published, and are archived with the audio.
The audio matters because a pause carries information that a transcript deletes.
has become the difficult part rather than accuracy.
A person who agreed to speak in one decade may be identifiable in a searchable archive in the next.
Archives now offer layered , which is a modest technical answer to a genuine ethical problem.