What is a "Duplicate Key"?

The term "Duplicate Key" is used to describe an index in a table that allows multiple records to have the same key. The internal ordering of the identical keys in the index is not guaranteed.
A consequence of this is that it is not possible to do a SeekBy and be guaranteed you will end up with a specific record matching that key, this record may change over time as the table is being modified.

When a key is not a duplicate key, it is called a "Unique Key". They are sometimes also called "Identity keys" because these types of keys can be used to identify a specific record in that table.

Remark

If a unique key is also a "null key", the uniqueness ONLY holds for records that are in the index. Records that are not in the index are not required to have key segment values that would make them unique if they would be in the index.