Abstract
This experiment, following Rartlett's Picture-writing, was designed to show : how much agreement there would be in the choice of a sign as the most fitting in connection with a word; whether the reasons given for a choice of sign mould vary greatly, and would be rationalizations rather than a primary ground for so choosing; and whether the chosen signs were more easily recalled than the not-chosen.
Six cards were shown to the subject, 011 each of which was one word aiid two signs. The signs were not intended to represent any object, although none was arbitrary. The subject had to choose the more appropriate sign and give the reason for his choice. Afterwards he had to reproduce all the signs.
Agrcement in choosing one of the signs of each pair was surprisingly high, although the reasons varied considerably. The chosen signs were reproduced more accurately and frequently, the reproductions showing the conventionalizing effect of the reasons given.
Choice of a sign as most fitting seemed to depend upon certain formal structural qualities of a general character. The most fitting sign for a verbal setting seemed to combine a conventional structure with simplicity and regularity of design. Linking these results with those of an experiment of Harrower's, it is considered possible that a method may here be available for exploring forms of mental organization not so far tapped by standard intelligence tests.
