Let us focus on the following argument: Is this inference sound? A theory on the nature of fictional characters laid by Amie Thomasson has the consequence that the truth of (1) and (2) does not guarantee the truth of (3). The reason would be that what looks one proper name, ‘Harry Lime’, is either two different names or one name that is used in one way in (1) and in another way in (2) and (3). No matter what route is chosen, (1) and (2) do not provide the conditions which made it possible for the sustitutivy of identity principle to deliver (3). This diagnosis is based on the fact that the occurrence of ‘Harry Lime’ in (2) and (3) is used to refer to an abstract entity while the occurrence of ‘Harry Lime’ in (1) would only refer to a character in the context of The Third Man’s plot. Whoever thinks it necessary to distinguish between what is narrated in a plot and what critics and sophisticated commentators remark on it seems to be forced to reject the inference from {(1), (2)} to (3). The words ‘Harry Lime’ do not make the same semantic contribution in each premise of the inference. I will to resist this line of criticism by giving a chance to an alternative approach that has that there is only one proper name in (1)–(3) and it keeps its meaning constant through its constituents. The main idea is that The Third Man’s script authors—Graham Greene, Alex Korda and Orson Welles—simply imagined a character and had to introduce the name to put up a plot.