Optional pragmatic processes or optional covert linguistic structure?*

The starting point for this paper is an asymmetry between overt linguistic indexicals and (alleged) covert indexicals. It has been proposed that the problem can be solved by making covert indexicals optional (Martí 2006) and that this provides a much better (linguistically respectable) account of the key phenomena than one that advocates an optional process of free (not linguistically mandated) pragmatic enrichment. I outline the debate and try to make the case that, at least for a theory of utterance comprehension, the pragmatic account is far preferable.