Chomsky(1998) proposed that there is no feature movement for feature checking. According to Chomsky(1998), Kim(1998b), and Yang (1998), there are two kinds of uninterpretable feature: uninterpretable formal features and uninterpretable selectional features. Uninterpretable formal features induce Agree whereas uninterpretable selectional feature induce Move. In this paper, I will focus on the uninterpretable formal features with respect to the Activation Hypothesis which has been recently proposed. In the main sections, I will analyze data in detail through the new concept of Agree and I will propose some new ideas in analyzing them. With this in mind, In section 2, I will deal with computational operation in the narrow syntax. Section 3 will focus on agreement phenomena with respect to feature activation. In section 4, I will try to identify the relation between Phi-features and Activation. Finally, section 5 will discuss the remaining issues about Agree and Activation. As a result, through this paper, I will try to derive the conclusion that the introduction of the new concepts "Agree and Activation" will help us to have an explanatory adequacy for driving computational operations.