How to make manga more interesting to readers: Using lattices derived from rough sets

Abstract

In manga (comics) and Japanese animation, “symbol-base theory,” which has its roots in Azuma’s data base theory, is widely accepted. This theory suggests that the elements of a character can be arbitrarily combined, and the elements of appearance and inner personality correspond one-to-one with each other. Mastui and Gunji showed that a method for deriving lattices from rough sets can be used to disprove symbol-base theory in terms of the impressions readers have of characters. In this paper, I applied this method to manga. I hypothesized that manga that created a non-linear lattice is perceived to be more interesting than manga that created using a linear lattice and that manga created a non-linear lattice is perceived to have characters that are more easily distinguishable than manga created a linear lattice. To verify these hypotheses, I constructed three manga using the software “Koikatsu” and conducted an experiment on the web. All manga have the same story, only the character’s hair and tone are different. Character hairstyle and tone of voice create a linear or nonlinear lattice. Manga 1 and Manga 2 created linear lattices and Manga 3 created non-linear lattice. In the experiment, participants read the manga and were asked questions about their impression of it. After that, I conducted a one-way ANOVA on the results. The manga that did not have a one-to-one correspondence between hairstyle and tone and that created a non-linear lattice was perceived to be funnier than the other manga. Also, this manga had no disadvantages in terms of the perceived readability and the ease of distinguishing characters. These results question symbol-base theory.