ABSTRACT
Many people think that Frege allowed that expressions could have sense yet lack reference. The question I wish to raise is how one could justify the claim that a systematic description of natural language will make essential use of such a view.
