In many languages, antipassive morphology is comprised of aspectual morphology (Polinsky 2008). This thesis presents an analysis of the syntax and semantics of antipassives in Inuktitut by exploring the link between aspectual morphology and antipassive morphology. It resolves the longstanding question as to the factors governing the distribution of the antipassive morpheme, showing that the presence of the antipassive morpheme is determined by the meaning of the construction, i.e. it does not merely change the grammatical function. It is proposed that the antipassive construction has imperfective viewpoint in contrast to the ergative construction. Antipassive morphology is obligatory with punctual telic verbs, i.e. achievements, which are verbs that have perfective viewpoint by default. Antipassive morphology is thus necessary to convey imperfective viewpoint for verbs that are by default perfective. Using a modified Reichenbachian (Reichenbach 1947) framework, it is shown that imperfective viewpoint does not allow for telic interpretations. Instead, punctuality determines the types of viewpoint, which coincide with the aspectual meaning of the antipassive marker. Viewpoint contrasts in Inuktitut are encoded not only in morphology but in changes of case and agreement configurations. They are derived using a version of the Minimalist Program (Chomsky 2008). Imperfective viewpoint can either be default, in which case there is inherent case on the internal argument, or derived through the addition of antipassive morphology, in which case the construction closely parallels a nominative-accusative structure. Perfective viewpoint is encoded through absolutive case on the internal argument, either in an ergative construction or in a canonical intransitive construction with unaccusative verbs. The thesis provides insight into the relation between case-agreement configurations and aspectual contrasts in language and the nature of those aspectual contrasts. It also provides a new approach to the relation between lexical aspect and viewpoint by considering the role of punctuality.

In many languages, antipassive morphology is comprised of aspectual morphology (Polinsky 2008). This thesis presents an analysis of the syntax and semantics of antipassives in Inuktitut by exploring the link between aspectual morphology and antipassive morphology. It resolves the longstanding question as to the factors governing the distribution of the antipassive morpheme, showing that the presence of the antipassive morpheme is determined by the meaning of the construction, i.e. it does not merely change the grammatical function. It is proposed that the antipassive construction has imperfective viewpoint in contrast to the ergative construction. Antipassive morphology is obligatory with punctual telic verbs, i.e. achievements, which are verbs that have perfective viewpoint by default. Antipassive morphology is thus necessary to convey imperfective viewpoint for verbs that are by default perfective. Using a modified Reichenbachian (Reichenbach 1947) framework, it is shown that imperfective viewpoint does not allow for telic interpretations. Instead, punctuality determines the types of viewpoint, which coincide with the aspectual meaning of the antipassive marker. Viewpoint contrasts in Inuktitut are encoded not only in morphology but in changes of case and agreement configurations. They are derived using a version of the Minimalist Program (Chomsky 2008). Imperfective viewpoint can either be default, in which case there is inherent case on the internal argument, or derived through the addition of antipassive morphology, in which case the construction closely parallels a nominative-accusative structure. Perfective viewpoint is encoded through absolutive case on the internal argument, either in an ergative construction or in a canonical intransitive construction with unaccusative verbs. The thesis provides insight into the relation between case-agreement configurations and aspectual contrasts in language and the nature of those aspectual contrasts. It also provides a new approach to the relation between lexical aspect and viewpoint by considering the role of punctuality.

La présente thèse analyse la décomposition sous-événementielle et l'interprétation aspectuelle des événements, en se focalisant sur les prédicats dérivés du basque, tels que dantzatu “danser” du nom dantza 'danse', amets egin 'rêver' du nom amets “rêve”, etxeratu 'revenir à la maiso' de etxera “à la maison” ou amatu “devenir mère” de ama “mère”. La thèse est centrée sur la configuration syntaxique représentant la structure événementielle des prédicats et la lexicalisation de ces configurations par des mots réels. J'admets que les prédicats se décomposent en composants plus basiques (par exemple, Levin and Rappaport Hovav 1995), et que cette décomposition est syntaxiquement représentée (Hale and Keyser 1993). Dans cette vue, les événements exprimés par les prédicats sont formés de sous-événements, qui peuvent être aussi bien des processus que des états (Harley 1995 2005, Cuervo 2003, Folli & Harley 2005, Ramchand 2008) et de compléments rhématiques (Ramchand 2008). Les rhèmes sont des compléments qui mesurent et décrivent les sous-événements dont ils sont les compléments. Ils peuvent être de différents types: des PPs, des DPs quantifiés, et même des radicaux non catégorisés. Partant de Harley (2005), je propose une ontologie des radicaux qui les classifie selon qu'ils nomment un Evénement, une Chose ou une Propriété. Les radicaux qui dénomment des Evénements et des Choses sont les compléments de sous-événements de type processus; tandis que les radicaux qui dénomment des Propriétés sont les compléments des états. A l'intérieur de chaque classe, les propriétés des mesures associées aux radicaux comme [+/-croissant], [+/- limite minimale] et [+/- limite maximale] déterminent les propriétés aspectuelles de l'événement tout entier, à savoir, s'il est ponctuel ou duratif, s'il est télique ou atélique. La thèse examine également les interactions entre l'interprétation aspectuelle interne des prédicats (aktionsart), et l'aspect externe (Smith 1997 [1991]). Cette relation est analysée en tenant compte de la catégorie lexicale à la base du prédicat et des différentes configurations de réalisation des prédicats en basque. / This dissertation analyses the subeventive decomposition and aspectual interpretation of events, paying special attention to derived predicates in Basque, like dantzatu ‘to dance’ from dantza ‘dance’, amets egin ‘to dream’ from amets ‘dream’, etxeratu ‘to go home’ from etxe-ra ‘to home’ and amatu ‘to become a mother’ from ama ‘mother’. The discussion is mainly concerned with the syntactic configuration that represents the predicates’ event structure and the lexicalization of these configurations by means of actual words. In this dissertation, it is assumed that predicates can be decomposed into more basic components (e.g. Levin & Rappaport Hovav 1995) and that this decomposition is syntactically represented (Hale & Keyser 1993). To be more precise, in the analysis made in this dissertation, it is assumed that the events conveyed by predicates consist of smaller subevents (Harley 1995 2005, Cuervo 2003, Folli & Harley 2005, Ramchand 2008) (which can be either processes or states) and Rheme objects (Ramchand 2008a). Rhemes are complements which describe and measure the particular subevent they are complementing. They can be of different types: PPs, quanticized DPs and even a-categorial Roots. Building on Harley (2005), I propose an ontology of Roots which classifies different types of Roots depending on whether they name an Event, a Thing or a Property. Event and Thing naming Roots are the complements of the process subevent, whereas Property naming Roots are the complements of states. Within each class, properties of the measure associated to each Root like [±incremental], [±lower bound] and [±upper bound] are going to determine the aspectual properties of the entire event, i.e. whether it is durative or punctual, and telic or atelic. Additionally, this dissertation considers the interaction of the aspectual interpretation of the predicates (commented in the previous paragraph) with viewpoint aspect (Smith 1997 [1991]). This relation is going to be analyzed in combination with the lexical category and the different configurations in which a predicate can surface in Basque.