This thesis tells the tales of auxiliation in Swedish. It describes how the auxiliary features of the aspectual auxiliaries bruka ‘use to’, tendera ‘tend to’, börja ‘begin, start’, fortsätta ‘continue’, sluta ‘stop, cease’ och hota ‘threaten’ evolve, as seen in a text corpora covering the 11th to the 21st century. A grammaticalization perspective is employed, which not unexpectedly has implications for the way data is interpreted and presented, but also less expectedly for how the method is understood and applied. The method consists of Swedish auxiliary tests, designed to reveal auxiliary features. The discussion of their validity involves auxiliation theory as well as a more general theoretical discussion of how the category and the function of auxiliaries have been conceived. The main point made is that a valid test should have an explicit link to auxiliary function. Establishing the nature of such a link, however, is no trivial matter and so the plausibility of the link turns out to be the main factor in judging the validity of auxiliary tests. In the early stages of the auxiliation processes, action nominals were often used as complements of the verbs under study. Later on these verbs started taking infinitive complements, lost their semantic restrictions on their subjects and started occurring in a type of passive construction where the nonfinite verb carries the semantic relationship to the arguments of the clause. The occurrence in this type of passive construction is perceived as the most plausible indication of auxiliary function available for subject raising auxiliaries, since the auxiliary and the nonfinite verb in these clauses do function together as one single predicate. The evolution of these aspectual auxiliaries essentially follows the predictions made by auxiliation theory. That said, they all show individual deviations from those predictions, depending on such factors as their semantics, their source construction and during which period in the history of Swedish their auxiliation took place.

Fondée sur une approche contrastive, cette étude a quatre objectifs : (a) considérer l'inférence des langues en contact dans le développement diachronique de l'anglais et bâtir une hypothèse sur l'origine de l'emploi périphrastique de DO ; (b) s'appuyer sur les faits avérés en diachronie et en étymologie pour montrer que la grammaticalisation de DO n'est pas le résultat d'un processus de désémantisation ; (c) partir d'une démarche fondée sur l'invariance pour montrer que tout comme sa contrepartie lexicale, la forme auxiliée de DO a bien une signification en langue naturelle ; (d) formuler une analyse critique des acceptions courantes en terme de catégorisation binaire ou ternaire entre le lexical et l'auxiliaire ; ce qui permet de montrer d'une part que la proforme DO SO peut effectivement reprendre des prédicats statifs purs, et d'autre part qu'en tant qu'opérateur de ‘thèse', DO a une fonction énonciative qui rend compte non seulement de son occurrence, mais aussi de sa non-occurrence dans le linéaire.