Character Coding and Inapplicable Data

Abstract

Inapplicable character states occur when character complexes are absent or reduced in some of the taxa. Several approaches have been proposed for representing such states in a character matrix so that the inapplicable condition has no effect on the placement of taxa and/or the applicable states are independent and not redundant. Here we examine each of these approaches and demonstrate that all have shortcomings. Coding inapplicables as “?” (reductive coding), although flawed, is currently the best way to analyze data sets that contain inapplicable character states.