VAGARIES IN THE DELIMITATION OF CHARACTER STATES IN QUANTITATIVE VARIATION - AN EXPERIMENTAL-STUDY

experimental study on the delimitation oi character states in continuo us variation indicates that (1) the way data are presented influences the assignment of character states and (2) states in the same data set are delimited in various ways by different individuals. Forty-nine in dividuals were given a set of graphs denoting variation of 10 characte rs in the genus Kalmia (Ericaceae) and outgroups, all identification h aving been removed from the graphs. The variation was represented in o ne of three ways: as 95% confidence intervals on a linear scale, as 95 % confidence intervals on a log(10) scale, or with bars showing SD x 2 on a linear scale. No two individuals scored a set of graphs in the s ame way, and only one character in one representation was scored ident ically by all individuals; the scoring ibr this character was complete ly different when the ordinate was changed from linear to logarithmic. Together; the 49 individuals delimited states within each character b etween 9 and 16 different ways. In general, variation represented by 2 x SD bars elicited the largest numbers of different scorings, yet wit h a relatively low number of states; the complexity of the patterns in the graphs in this representation was greatest, Expert knowledge appe ars to be of dubious value in delimiting states in such variation, and if such characters are to be used in phylogenetic analyses, states co uld be delimited by people who know nothing of the details of the stud y being scored; in any case, presentation of data and an explicit prot ocol to follow when delimiting stales are essential. In converting dat a of this type into character states, psychological factors are partic ularly likely to come into play. Other implications of our experiments include thr severe underdetermination of some phylogenetic hypotheses by observation and the heterogeneous nature of morphological data.