Snap Judgments of Similarity between Drawings Intended to Portray Emotional Expression
The present chapter reports an experiment in methodology. In experimenting with methods of studying symbols, it seemed best to use some type of symbol which had no official warrant for meaning anything, and which could not be rated as a test. Pictorial material seemed to afford the desired freedom from established misconceptions.
There are certain peculiar advantages in studying line drawings of the human face intended to portray emotional expression just as there are other advantages in studying posed photographs or actual photographs and motion pictures of individuals in real situations. With line drawings one may distort the action of certain facial muscles, and even make combinations of antagonistic muscles which could not occur in real life. Furthermore, drawings or caricatures seem to be sufficiently removed from reality to enable even the most naïve individual to refrain from assuming that he is dealing with 'emotion.'
Drawings, photographs of assumed expressions, voluntary or habitual contractions of the facial muscles, and the like offer a rich field for investigation in social psychology as part and parcel of methods of communication between individuals. The problems met in this field are perhaps identical with those to be met in the study of verbal responses, but since the material is pictorial, one happily avoids many fallacies gratuitously introduced into the study of language by the philologists. But in obtaining this emancipation, one must also rid himself of the fallacies already introduced into the field by the experimental psychologists. To make this point clear, it is necessary briefly to review some of the previous experimentation in this field.
In dealing with pictures which observers have been asked to name, there is a subversive tendency to depart immediately from the reality in question -- the picture -- and to deal with the hypostatized reality of the thing named, to identify the verbal report of one observer with the verbal report of another observer, thus assuming the identity of the references in the verbal reports of the two observers. One finds experimenters grouping pictures according to verbal reports, as if the reports were the reality, when in fact the verbal reports, even if identical, may have had quite different references by two observers or by the same observer on different occasions.
Langfeld1 presented certain of the Rudolph pictures to the same observer on two different occasions and recorded the verbal interpretations. He reports that a picture intended to show "fear and horror" was interpreted by one observer as "anger" on one day and as "apprehension, terror" on another day; again, by another observer as "rage" and subsequently as