This work investigates the differences between two widespread types of causative constructions: the so called lexical causative as in Gianna opened the door or the corresponding Italian sentence Gianna ha aperto la porta, and the periphrastic causative, as in the Italian Gianna ha fatto aprire la porta (Gianna had the door opened/made the door open), assuming as theoretical framework a neo-Davidsonian Events Semantics. Within this framework both these constructions are best analysed as denoting two events in a causal relation, therefore the question of pinpointing the sources of their differences is essential to answer to those critiques that on the their basis deny the lexical causative denotes more than one event. I consider these differences in a systematic way in their morpho-syntactic, semantic, pragmatic and conceptual aspects. The main evidence for decomposing the lexical causative semantically in Cause (e, e') is adverbial modification, since many adverbs can be interpreted as modifying only the event e or the event e' in the causal chain. The be-eventive analysis of the lexical causative solves these ambiguity issues but also creates a puzzle to the extent that not all adverbs can modify only the effect. The solution based on the morphological distribution of adverbs into those that can and those that cannot modify below VPs, is that the lexical causative is a silent morpheme embedding roots or some other lexemes, but not VPs. Other possible approaches to this puzzle prove empirically incorrect. Adverbial modification data have also been the source of the debate whether the lexical causative should denote a super-event conceptually analysed in its sub-events, rather than two events in a causal chain. The bi-eventive analysis proves to be the best solution. As for of the periphrastic causative, I follow Folli and Harley (2007) in "updating" its account with the most recent decompositional finding in syntax. I depart from them, however, in the specific syntactic analysis and in the semantic treatment that in their analysis is not developed, accounting thus for old and new data. Fare is a light verbs embedding only verbalized material independently possible for a verb (not roots). This embedding pattern accounts for all the different fare constructions. In particular, my account of fare par is a rehabilitation of those theories that in a more lexicalist approach claim that in this construction fare embeds a passive, and it is possible within a purely syntactic approach only because I assume disjoint projections for the causative morpheme and the Agent theta role. The assumed syntax has semantic repercussions to the extent that the faire infinitif construction is analysed as [Cause (e, e') & Agent (e, x) & Cause (e', e"), while the faire par construction as [Cause (e, e') & Cause (e', e"]. The semantic difference between the two constructions thus boils down to the Agent of fare causing the Agent of the embedded verb to do something in faire infinitif versus the Agent of fare simply causing some event in faire par. The fact that fare can embed a root as verbalized by the silent causative morpheme or as VP means that when a verb undergoing the causative alternation is embedded under fare, e.g. Gianna ha fatto aprire la porta (Gianna had the door opened/Gianna made the door open), the resulting sentence is ambiguous between a faire par construction without da phrase expressed or fare+VP, as the English translation shows transparently. Both these constructions express some kind of indirectness in the way Gianna gets the door open. In the former the indirectness is grammatically conveyed by the presence of the embedded lexical causative silent morpheme, and the sentence is typically interpreted as Gianna having somebody opening the door. In the latter the indirectness is derived a la Grice from the meaning of the lexical causative. The meaning of the silent lexical causative morpheme is causation in presence of spatiotemporal proximity between the causing and the caused events, assuming a definition of spatial proximity between events as spatial proximity between the entities participating as theta roles in these events, typically Agent, Theme and Instrument. Fare, on the other hand, simply means causation, but triggers the conversational implicature of non-spatiotemporal proximity between causing and caused events. The challenge is how to characterize spatiotemporal proximity considering that in many of the given situations the causal chain does not seem intuitively proximate and still is perfectly expressed by the lexical causative. Another problem is that most of these situations can be described by both the lexical and the periphrastic causatives, although one or the other might be preferable. (Abstract shortened by UMI.) [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]