In line with current assumptions about how the grammar of a language is 'created' through history, the modal auxiliaries in English and Dutch ('can'/'kunnen', 'may'/'mogen', etc.) are known to have emerged from full lexical verbs, many centuries ago, and to have gradually acquired auxiliary (i.e. grammatical) status through a process of 'grammaticalization'. According to current views, this process is normally unidirectional (i.e., linguistic forms develop from lexical to grammatical, and from less to more grammatical, but not vice versa). Yet, in the history of both Dutch and English, there is a phase in which the evolution of the modal auxiliaries is reversed (a return to a more autonomous, lexical status, at least partially). In English this phase is situated mainly in the Middle English period, ending around 1500 (after 1500 the modals – at least the central ones – grammaticalized again further). In Dutch, however, this phase only started after 1650, in the New Dutch period, and it seems to be continuing until today. This project involves a systematic diachronic corpus investigation of these phases of re-autonomization of the modals in the histories of Dutch and English. The aim is to achieve a better view of the factors and forces which have effected and affected these processes in both languages, and to compare them, and, thus, to contribute to a better understanding of the processes of and degrammaticalization in general.