Evans on Russell

Russellian referring expressions
=logical proper names
=demonstratives
= sentence containing such a term would be meaningless if the term has no referent.

Evans wants to say--
Referring expressions ~=Russellian referring expressions
Referring expressions have two sub-categories:
(1) Russellian singular terms &
(2) descriptive names

Q: how is a descriptive name different from a definite description?

Evans presents an argument that, by adopting a pecular definition of 'reference (as closely related to truth),' descriptive names(DN) and definite descriptions(DD) are referring expressions.
So this argument, if successful, shows that DN are referring expressions.
But it also shows that DD are referring expression, which is just not the case.
Hence Evans need to find a way to dissimilate DD and DN, and says that the DD is not referring expressions.

Evans considers three arugments from Russell to the conclusion that DDs are not referring expressions.
(A) one can fail to know that two DDs have the same referent. <--> assumes that REs are such as to render knowledge of coreference obligatory.
(B) the scope of negation <-->assumes that all REs are Russellian.
(C) if DDs are REs, they would be meaningless if empty.<--> question-begging

[2.4]

more on descriptive names!

rigid, but can be empty...