I once heard a conjecture that a primitive L-function does not have multiple zeros except the central point of the critical strip.

Question:Why it is reasonable to conjecture a primitive L-function does not have multiple zeros except the central point of the critical strip? Is it possible to explain why the central values behave so different?