I'm reminded of my freshman English class in college, when we were studying S.I. Hayakawa's Language in Thought and Action. Even when using a word to denote one specific entity, the dynamic process-entity referenced one time is not the same entity that is referenced another time. For example, the Kevin Randall writing this message today is not the same Kevin Randall that read Regina Reynolds' message earlier; innumerable changes have occurred in the state of my being in between those two events. But we accept certain levels of abstraction that will collapse various states of a dynamic entity into one or more specific identities for the purpose of communicating in any given context.
But of course it does become trickier when we get to changes in things that we generally think of as static, like statues. Or my favorite architectural work, Lyndhurst in Tarrytown, New York. Originally completed in 1838 and known as Knoll, it was doubled in size in 1865 and renamed Lyndenhurst, and renamed yet again (to its current name of Lyndhurst) in 1880. Knoll is no longer extant, having been replaced by Lyndhurst—although much of Knoll is contained within the elements that make up Lyndhurst.
I think we have to be flexible in the application of the WEMI model, and define the specific instances of each W, E, M, and I in the context for which the descriptions are being made, and whatever makes the descriptions best fit together and interact in the most friendly way with other descriptions.
Kevin M. Randall
Principal Serials Cataloger
Northwestern University Libraries
Proudly wearing the sensible shoes since 1978!