With this article the author wants to argue that it is possible and desirable to develop an extended register analysis. One characteristic of such analysis is the determination of register from features of the text, i.e. the analyst is interested in the situation in which the text is embedded to the extent that it makes a difference to the text. The aim of such analysis is to offer insights into a pre-translational text analysis, and link it up with a coherent and active theory of language, thus opening discussion between translation studies and linguistics. One of the central ideas in register analysis is that functions rather than structures are preserved in translation, with these functions taking the specific form of sub-variables of field, tenor and mode. These functions would be seen as related to lexico-grammatical realization in a relatively controlled way.