Summary form only given. The increasing amount of hyperlinked human knowledge available on the Web has shifted the research focus to the search engines that will deliver the needed information quickly. This demand has triggered a revisitation of the long-standing issue of semantics, which was never addressed adequately by information theory or linguistic theory. Fuzzy sets and fuzzy logic, on the other hand, provide a powerful mathematical tool in modeling "semantics." The organization and the manipulation of information, knowledge base construction and discovery,. information retrieval and logical inferencing all deal with semantics in a fundamental way. Unless we experience a breakthrough in understanding semantics, especially related to machine automation, it will be difficult to make a quantum jump in accessing, creating, and modifying information. This paper intends to bring to center stage the complex problems of semantics confronted while dealing with natural language. The paper demonstrates the ability of fuzzy sets and fuzzy logic to serve as a mathematical vehicle to model semantics in a significantly improved way. This will include taking a closer look at the PNL, Precisiated Natural Language, proposed by Zadeh.