This study presents a computational approach to the problem of lexical access based on FrameNet and similar resources. FrameNet is compared to Lexical Conceptual Structures, WordNet, Levin's WordList, PropBank and VerbNet. Each of these publicly available lexica contains some features that make it comparable to FrameNet, even though they all lack in its fine-grained overall data description. In particular, the paper focuses on Dorr's Lexicon of Conceptual Structures. The underlying theoretical framework is introduced in some detail. One feature in particular is investigated: the notion of Causality and the way it is represented at the lexical level. The final part of the paper describes the effort of building one such resource for Italian in the late 1980's. That effort aimed at reconciling world knowledge and linguistic semantic and conceptual information needed to produce highly structured lexical entries as the ones provided by FrameNet and Lexical Conceptual Structures. Besides presenting the algorithm and its linguistic components, the author highlights those features that make it different and, in some cases, computationally more perspicuous than the others. (Cristiano Broccias)