Page:Folk-lore - A Quarterly Review. Volume 4, 1893.djvu/91

This page has been validated.
Report on Folk-tale Research.
83

The list of incidents is valuable too. It is the first attempt to compile what has long been wanted. If it be imperfect, that is unavoidable; and the imperfection cannot balance our indebtedness to the author. The chief defects are, so far as I have tested it, of three kinds. First, incidents are defined too specifically. For example, the incident, found in drolls, of the fool who tried to get into some article of his clothing by jumping, should be indicated as Jumping into clothes, rather than by the mention of an article only found in some variants as the object of the hero's perspiring efforts. Second, the alphabetical order should be subordinated to some sort of logical order. Thus, I find Candle-lighting election under C, and Kingship test under K. These are both variant forms of one incident, which relates the supernatural designation of the hero to the office of king or pope; and the first may, in fact, be included in the second. What is wanted is a general heading, such as King, Designation [or Nomination—not Election] of, to be followed by sub-divisions into By animals, By bell-ringing. By candle-lighting, and so forth. The third kind of defect arises, I think, usually from too great a desire for compression. Compression is undoubtedly one of the chief matters to be aimed at, but not at the sacrifice of perspicuity. Who could tell that the Thrown in water incident was that of the hapless queen thrown into the water, or otherwise put away, during the king's absence, to make room for her uglier step-sister? Zigzag transformation hardly expresses the incident better yclept by Mr. Nutt Transformation fight. These blemishes, however, are all susceptible of amendment; and a committee of the Society, taking this list as a foundation, could easily compile a standard list adaptable to all our wants.

These practical efforts are so important that they will perhaps draw away the student's attention from the paper by which they are preceded. Such a result is much to be deprecated. Taken together with M. Cosquin's paper,