Tan and Ward (2000) kept track of outloud rehearsals during item list learning to distinguish between the nominal and functional serial position of items in a list. What did they mean by this, how did they demonstrate the difference between the two, and what were the implications of their findings for understanding models of short-term memory?