binding (adj./n.)

A term used in logic, and frequently encountered in GRAMMATICAL and SEMANTIC THEORY, for the relation between a VARIABLE and a QUANTIFIER or other OPERATOR on which it is semantically dependent – normally the lowest operator CO-INDEXED with the variable and taking it in its SCOPE. In GOVERNMENT-BINDING THEORY it refers to a series of CONDITIONS which formally relate, or bind, certain elements of a sentence. Two kinds of binding are distinguished: A-binding and A-bar-binding (u-binding). The former obtains if a category (an ANAPHOR) is CO-INDEXED with a C-COMMANDING NOUN PHRASE in an A-position (= ARGUMENT-position). The latter obtains if a category (e.g. a variable such as a WH-MOVEMENT TRACE) is co-indexed with a c-commanding category which is in an A-bar position (a position other than subject, object, and object of a preposition), e.g. the clause-initial position occupied by a wh-phrase. The extension (or generalization) of the approach from the former to the latter is known as generalized binding. Elements which are not bound are FREE. Binding theory is one of the (sub-)theories of government-binding theory. It is primarily concerned with the distribution of NPs in a sentence, determining the situations in which they can or must be co-indexed with other NPs. The NPs are classified into ANAPHORS, PRONOMINALS and R-EXPRESSIONS (‘referring expressions’). The three principles of binding theory – binding conditions A, B and C – are: (a) an anaphor is A-bound in its governing category; (b) a pronominal is A-free in its governing category; (c) an R-expression is A-free (everywhere). The binding inheritance principle is a reinterpretation of the FOOT-feature principle of GENERALIZED PHRASE-STRUCTURE GRAMMAR within HEAD-DRIVEN PHRASE-STRUCTURE GRAMMAR.