Abstract
In Japan, electoral campaigning in plurality districts often appears strangely ineffective to observers. In their explanations, political scientists have so far fallen back on the proximate cause, namely the strict Public Office Election Law (K), claiming that it is this law that prohibits different and more effective campaigning. This study, however, is based on the assumption that such an explanation could be insufficient since even those campaign tactics legally allowed are often executed and/or selected in an apparently ineffective way in terms of their appeal to voters. Especially when looked at from the perspective of structural learning theory, which assumes that actors adapt their behavior within stable structures by learning what works well and what does not, these observations invite further investigation