Carnap’s reinvention of the Ramsey-sentence approach to scientific theories has been at the center of a new debate in recent years. Following Grover Maxwell, Psillos (2000a) argued that Carnap’s re-invention of the Ramsey-sentence had failed to lead to the desired neutral stance in the realism-instrumentalism debate, and ended, instead, to a form of structural realism which happened to be liable to Newman’s objection to Russell’s version of structural realism. The objection held that without putting suitable restrictions on the range of the variables of the Ramsey-sentence, a Ramsey-sentence approach to theories renders trivial and a priori true all ontological commitments to unobservable entities issued by scientific theories. By arguing that Carnap achieved the neutral stance, Friedman (2011) counter Psillos claim. He denied that any form of realism could be attributed to Carnap. In this paper, I provide a middle ground, where an unorthodox form of structural realism could be attributed to Carnap. I highlight parts of Carnap’s work which deal with the problem of designation of abstract terms and the relation of the language to the facts of the matter (in Carnap 1934; 1950; 1966), to argue that it was Carnap’s view about the practical methodological considerations, being at work in the construction (or choice) of the linguistic systems, which led him to the unorthodox form of structural realism. I also claim that the same practical considerations constitute the nub of a viable Carnapian answer to Newman’s objection.