This study uncovers Japanese middle constructions based on the approach of “distributed morphology”. The findings reveal that adjunct is obligatory in Japanese middles. Two types of grammatical elements contribute to the adjunct: suffix and adverbs. The suffix yasui corresponds to English “able”. The case of the subject must be nominative, i.e., が. Once verbs are attached by the suffix yasui, their part of speech transits from verb into adjective. The new lexicon predicates an inherent property of the subject. Regarding middles with adjuncts rendered by adverbs, two subtypes are confirmed: the na-adjective formed adverb 簡単に kantan ni, and the i-adjective formed adverb よく yoku. The former is produced by the na-adjective 簡単 with the copular に. The latter is formed by the i-adjective よい with the predicate く ku. The mechanisms of the constructions rendered by the two are similar. Furthermore, unlike English middles, where non change-of-state verbs are ruled out, there is no distinct lexical category of middle verb Japanese. Rather, six groups of verbs are compatible: (a) motion verbs; (b) change-of-state verbs; (c) action verbs; (d) perception verbs; (e) stative verbs; and (f) accomplishment verbs. Crucially, such generosity does not result from the adjuncts. It is the “potential form” of verbs that enables psychological and perception verbs to be licensed in Japanese middles.