Abstract

This thesis consists of three papers on the semantics of natural language and formal languages. Chapter one discusses how the possible interpretation of the noun phrases in a sentence is affected by the syntactic structure of the sentence. In particular, we focus on the phenomena related to principles B and C of the Binding Theory. We can explain all these phenomena, including the counterexamples that have been offered against standard Binding Theory, by viewing the binding principles as semantic principles that constrain the relation of presupposed co-reference, correctly identifying the context that is relevant for the application of the binding principles, and viewing binding theory as just one among several sources of semantic information that are used when interpreting the noun phrases in a sentence. Chapter two attempts to provide support for George Boolos' proposal of interpreting second-order quantification by means of plural qualification by arguing that several of the criticisms that have been offered against it are not correct. We consider the criticisms put forward by Charles Parsons, Patrick Grim and Michael Resnik. Chapter three distinguishes two main senses in which a formal language can be said to have been provided with an interpretation: [i] by having provided a model and a definition of satisfaction and truth in the standard way [ii] by having provided a translation into a natural language. We argue that the sentences of a formal language interpreted as in [i] do not mean anything. A formal language interpreted as in [i] models the way the truth of a sentence would be affected by two factors: the interpretation as in [ii] of the language, and a way the world might be. Viewing in this way the relation between interpreting a formal language as in [i] and as in [ii] allows us to justify the conceptual adequacy of the standard model-theoretic definitions of the properties of logical truth and logical consequence.