This epilogue reflects on the notions of ‘irregularity’ and ‘mismatch’ which form the leitmotiv for this issue on perspectivization in language use, and more specifically on perspective shift and perspective persistence. While endorsing the approach, it is suggested that, given the complexities involved (who orients to what aspects of context and how?) any stable form-function coding would probably fail to perform the communicative task at hand, so that adaptable structures and processes are predictable. It is also pointed out that the observed irregularities seem to show a high degree of persistence across time and space, and it is suggested that this ‘regularity’ may be the most useful focus for this type of research.