The present paper is focused on the study of those relations that auxiliary verbs can establish among themselves when chained in a sequence. Regarding those sequences, which in Spanish can be considerably long, the literature has displayed primarily interest in formulating a set of principles that can predict possible relative orderings among auxiliaries. On the contrary, our paper delves into a less walked path: the description of relations established within an auxiliary chain. We will start from the traditional definition of auxiliary verb as a unit that modifies the ‘main’ or ‘lexical’ verb, and proceed to show that such a conception makes the wrong predictions when it comes to explain those internal relations, for it only accounts for a subset of the cases. This explanatory problem is common to both traditional and more formal models. In our opinion, the distinction between between <em>lexical </em>and <em>functional</em> auxiliaries that we propose in this work, in the context of a dynamic computational model that includes and derives this distinction, allows us to overcome these shortcomings of traditional analyses.