In this article, which examines the cyclic evolution of future markers in Western Romance (mainly French and Spanish), I make use of the "satellite model" in the version proposed by Koch and Oesterreicher (1996) to capture the complex interplay between functional change, synchronic variation and sociolinguistic evolution. This model conceives of linguistic cycles as push-chains. Thus, I will argue, young future markers originally arise from argumentative patterns that are aimed at validating announcements concerning the speaker's projected actions. The rationale behind these mechanisms is pragmatic efficiency rather than the functioning of the language system itself. Thus, linguistic systems usually contain more items than are technically needed to keep the language system operative. In the categories of the satellite model, a certain number of younger constructions ("satellites") exist side-by-side with a canonical construction (i.e., a functionally and sociolinguistically unmarked item), which one of the former may eventually oust from its privileged position. As I will show by sketching the evolution of the numerous future markers of peninsular Spanish from the seventeenth to the twentieth century, this perspective allows a fresh look at the evolution of cycles. More often than not, the competition between the canonical form and its satellites does not cause the former's replacement by the latter; rather, it normally ends with the obsolescence of one the satellites involved. The eventual replacement of the canonical form by a satellite that is, the completion of a full cycle - represents a very special (and relatively rare) case.
