Next: Conclusion Up: Discussion Previous: Situoid theory and modality Contents
The main difference between Barwisean situation theory and what we called situoid theory is our distinction of situations and situoids as entities existing at a point in time and entities having a temporal extension. Especially in the last section one may have noticed that this distinction may not have been such a good idea after all. Situations and situoids appear to behave similar in many ways. And we have to admit, that the distinction of Barwisean situations into situoids and situations did not enrich the theory at all. If we would take Barwisean situation theory and write ``situation or situoid'' every time he writes ``situation'', we would pretty much obtain what we have developed, from a functional point of view. Also our distinction of Barwisean states of affairs into states of affairs, pictures of states of affairs and infons gives not rise to a richer theory of situations.
But our goal was not to enrich situation theory, especially since most issues in situation theory are already answered. We integrated situations and situoids in a top level ontology, GFO, and to do this, several philosophical questions about the nature of situoids, situations and states of affairs had to be answered that were of no concern to Barwise and his followers.
Contrary to Barwise, we distinguished states of affairs, pictures of states of affairs and infons. Most of this distinction is based on tractatus. For this purpose, a new ontological category had to be found, the category of infons, information-carrying entities. This has an ontological impact, because infons and states of affairs behave differently. The same is true for states of affairs and their pictures. The pictures are special states of affairs, and they behave differently than states of affairs. While this is an ontological difference, our infons are as rich as Barwisean's, so we kept their expressive power needed for the various applications of situation theory, while adding ontological richness to the GOL-ontology. The decisions made by us are not the only possible, but we believe we tried to defend and support our views sufficiently to justify this change from situation theory.
The distinction of situations and situoids we made arose from some very basic assumptions in the GOL-ontology about time, a concept which situation theory is lacking, and the distinction between endurants and occurrents. This adds expressive power to the top level ontology we used as a background and provides us with the means of drawing fine-grained characterizations on the model of the world. In situation theory, a theory of time would have to be added to achieve the same level of modelling-opportunities.
While in situation theory, the phrase that situations ``are parts of the world that can be comprehended as a whole'' is taken as given and not explained further, we explained this phrase in great detail. Also, the implications of these explanations are only ontological in nature and motivation. They describe how situoids and situations behave in the world we want to describe with the means of the GOL-ontology. The axioms we developed reflect some of this knowledge about the behavior of situoids.
We showed how situoid theory in GOL can be applied to problems situation theory has been developed for, the most important being the meaning of natural language terms and phrases. And we have shown how the concept of modality can be introduced in the GOL-ontology based on situoids, situations and infons.
We believe that the expressive power of situation theory has not been lost, and that for every application of situation theory, situoid theory can be used as well, sometimes more easy because the background ontology of situoid theory provides additional knowledge, additional categories for characterizing problems.
leechuck 2005-04-19