Much ado about nothing: inapplicable data as insertion-deletion events.

The treatment of inapplicable characters has proved especially vexing to systematists. Investigators have wrestled with alternative coding scenarios to capture both the presence and absence of a feature, and its variation when present, in a reasonable manner. Three basic issues have presented themselves: (i) impossible states at internal nodes; (ii) action at a distance among disparate parts of the tree; and (iii) "secondary" (i.e. aspect variation) characters overwhelming "primary" (i.e. character presence/absence) patterns in grouping taxa. Multiple methods have been proposed to deal with these issues in the context of standard character coding with varying levels of complexity. Here, I show that these issues can be dealt with in a direct fashion by treating presence/absence not as a character, but as insertion/deletion of a character with all its potential variation. This approach removes these three problems in simple, straightforward manner.