This article focuses on categorization in translation as one of the major problems students of translation encounter when translating, a problem that plays its role on various levels of the translation process, notably that of the interpretation of the source language (SL)-decoding-and that of re-expressing the SL meaning in target-language (TL) forms-encoding. On these levels not only semantics and pragmatics in the strict sense play their parts, but also culture. First the author argues that in order to be able to translate properly and solve problems of translation adequately, the student needs to have access to a well-structered lexicon, and that such a lexicon should follow the principles of 'Natural Categorization'. The author provides an outline of the model of "Natural Categorizatoin" as opposed to that of "Classical Categorization" and discusses the consequences of both models for the structure of lexicon. Finally, the author presents a model of grammar, called TCM ('Two-Cycle Model of Grammar') and outlines the structure of its lexicon. The author argues that the lexicon in TCM meets the criterion of 'Natural Categorization' and, as such, can be of help to the student of translation in cases of translation problems.