Are there Domain Specific Languages?

Turing complete languages can express unbounded computations over unbounded structures, either directly or by a suitable encoding. In contrast, Domain Specific Languages (DSLs) are intended to simplify the expression of computations over structures in restricted contexts. However, such simplification often proves irksome, especially for constructing more elaborate programs where the domain, though central, is one of many considerations. Thus, it is often tempting to extend a DSL with more general abstractions, typically to encompass common programming tropes, typically from favourite languages. The question then arises: once a DSL becomes Turing complete, then in what sense is it still domain specific?