The paper deals with the role played by morphology in core syntax within a generative minimalist framework: more specifically it deals with the theory of valuation of agreement or phi-features (that is, person and number features) and the valuation of tense or t-features (that is, features like [+/-present]), and there is a second part where I focus on the computation of both kinds of features according to the neuroimaging literature. The core idea that I aim to defend is that agreement features and tense features are each valued by a distinct head, namely 'v' and T. I argue that the number of steps in the 'real' processing or computation of agreement features and of tense feaures must be the same, irrespective of whether the morphological realization is 'zero' (as is frequently the case in a language like English) or not (as is typically the case in a Romance language like Spanish).
