In both realist and non-realist conception of (scientific) theories, the standard view is that theories are identified by the objects they are about; it is just that the non-realist withholds reality for those objects. It is argued in this chapter that the notion of basic objects that are sustained through theory-change is suspect. This is because, during theory-change, the domains of explanation also change. Since a theoretical domain is identified by the structure of its basic objects, it follows that during theory change those objects also change. If they change, why should they be viewed as basic? The argument is illustrated with a detailed study of theoretical linguistics. It is shown that, as the study of language shifted from ancient Paninian concerns to contemporary generative grammar, the basic vocabulary and focus of theory changed so much that it is difficult to see that Panini and Chomsky are studying the same domain in their theories. More significantly, such radical shifts can be detected within the different phases of generative grammar itself. Finally, the phenomenon of lost domain is not restricted to linguistics. Current attempt in physics to somehow unify the quantum and relativity theories seems to result in a similar form of domain-shift. So, the notion of what theorists are talking about is not perspicuous.