In this paper I bring together linguistic aspects of the motivations for language change that are generally kept apart. To this aim I will argue that the synchronic variation of want (to) reflects its historical developments and that it is possible to bring together the binary understanding of language variation (synchronic and diachronic) within a unified framework that reconciles form and meaning. I will show that the combination of such theoretical frameworks as mental space blending and Construction Grammar allows us to unify a broad range of complex cognitive mechanisms at work in language change and the resulting present-day polysemy.