I was considering the following two natural questions about the relationship between unambiguity and complementation for the class of context-free languages:
- Is the complement of an unambiguous context-free language also a context-free language?
- If a language is context-free, and its complement is context-free as well (i.e., a so-called strongly context-free language), is it the case that the language is unambiguous context-free?
The first question above is motivated by the fact that it holds for deterministic context-free languages, which are a strict subclass of the unambiguous context-free languages.