Abstract. The notion of the logical form of an expression E is shown to be a semantic notion that can be derived from the notion of the structured meaning of E. This simple idea can be traced to categorial grammars, and it is implicitly used by Richard Montague. We argue that a most fine-grained tool for defining structured meaning can be found in Pavel Tichý’s Transparent Intensional Logic (TIL). Structured meanings of expressions are identified with abstract procedures, known as TIL constructions, expressed by the expressions. We construe concepts as closed constructions, and present a method of seeking the best semantic analysis (identical to the structured meaning) of a given expression. The method terminates in a complete lattice over the set of possible analyses; which analysis is the best one must, however, be relativized to a conceptual system. This relativization concerns primitive concepts of the conceptual system within which the semantic analysis is set. Our main thesis is: Every well-formed expression E of the language under analysis can be associated with a logical form that is unambiguously derived from the structured meaning of E. 1 The problem of adequate explication

It is far from clear what is meant by logic or what should be meant by it. It is nevertheless reasonable to identify logic as the study of inferences and inferential relations. The obvious practical use of logic is in any case to help us to reason well, to draw good inferences. And the typical form the theory of any part of logic seems to be a set of rules of inference. This answer already introduces some structure into a discussion of the nature of logic, for in an inference we can distinguish the input called a premise or premises from the output known as the conclusion. The transition from a premise or a number of premises to the conclusion is governed by a rule of inference. If the inference is in accordance with the appropriate rule, it is called valid. Rules of inference are often thought of as the alpha and omega of logic. Conceiving of logic as the study of inference is nevertheless only the first approximation to the title question, in that it prompts more questions than it answers. It is not clear what counts as an inference or what a theory of such inferences might look like. What are the rules of inference based on? Where do we find them? The ultimate end