At present, there are few attempts to provide external explanations for the patterns subsumed under the heading of “translation universal”. This paper discusses the possible cognitive basis for the patterns/processes that have been variously referred to as simplification/generalization, normalization, standardization, sanitization, and exaggeration of target language features. The framework that is adopted is that of cognitive grammar, and the claim is that all of the above arise from the existence of asymmetries in the cognitive organization of semantic information. The paper also proposes that the converse case is true: cases involving a lack of conspicuous cognitive asymmetries will demonstrate the opposite effect in translated text. In closing, the argument is placed in a larger perspective by adopting Croft’s (1990) scalar notion of generalization in a discussion of explanation in Translation Studies.