Infinitesimal and intensional software

点击翻译按钮获取中文摘要

I will investigate theoretical and practical aspects of a new type of logic I call "infinitesimal logic".Like fuzzy logic, it uses a spectrum of truth values between completely true at one end and completely false at the other. But unlike fuzzy logic, the additional truth values are discrete and separated from each other, not unlike quantum mechanical energy levels. For example, the truth value just below standard truth can be considered as qualitatively less true - in fact, incomparably (even "infinitely") less true.One theoretical use is in the semantics (meaning) of rule based systems with negative conditions - such as Prolog with negation-as-failure. On the practical side, the intermediate truth values can help users express preferences and priorities.For example, suppose you want to fly to Ottawa and prefer an aisle seat. Infinitesimal logic allows you to formulate this so that there are three answers: an unqualified Yes, which means you can; a qualified Yes, which means you can get to Ottawa, but not in an aisle seat; and an unqualified No, which means you can't get to Ottawa.You can therefore make clear the fact that having  an aisle seat is incomparably less important than getting to Ottawa; for instance, that you are not interested in aisle seats to Calgary.