The aim of this paper is to offer an inductive account of the emergence of auxiliary do throughout the Early Middle English period. The previous accounts mainly attributed the rise of auxiliary do to various grammatical changes occurring in Middle English, such as the loss of the V2 constraints, re-categorization of pre-modal verbs, the loss of morphological cases, and the loss of V-to-I movement. In this paper, it is argued that these grammatical changes in Middle English as well as the rise of auxiliary do can be understood as a consequence of the change of phasal status in the native speaker's internalized grammars from monodic phase (CP) to dyadic phases (CP, vP) with the introduction of TP and vP in the Early Middle English. Moreover, it is argued that this phasal change also led to the loss of the V2 constraints, the loss of V-to-I raising, the general decline in verbal inflection, and re-categorization of pre-modal verbs. Finally I propose to demonstrate that auxiliary do is base-generated in the light v of vP in the framework of the minimalist program and also try to show that auxiliary do arose with the introduction of vP in the Middle English clause structure, which explains how the different sentences with the same meaning such as I did not speak, I not spoke and I spoke not appeared simultaneously at that time.