Introduction: The sources of c-command

Of the core syntactic relations in UG, none is more gregarious than c-command. It plays a key role in at least three different domains: binding, linearization and movement. Consider how.

All three principles of the binding theory exploit c-command in their definition of binding, binders being expressions that both c-command and are co-indexed with their dependents. More concretely, anaphors must be locally bound by their antecedents, pronouns cannot be locally bound by their antecedents, and R-expressions cannot be bound at all. In addition, pronouns interpreted as variables (“bound pronouns”) are (typically) c-commanded by their antecedents.

Similarly, most (if not all) versions of the Linear Correspondence Axiom (LCA) are defined in terms of asymmetric c-command: thus α precedes β just in case α asymmetrically c-commands β.

Lastly, movement also crucially invokes c-command. For example, ECP-based accounts define antecedent government in terms of binding and the latter, as noted, is defined in terms of c-command. In addition, chains are defined in terms of c-command (links in a chain c-command one another) as is a central well-formedness condition on movement and/or chains, the minimality condition. Consider the latter, as it will be a focus of what follows.

Minimality restricts operations in the configurations in (1).

(1) Minimality: A movement operation cannot involve X1 and X3 over an X2 which is identical to X3:

… X1 …. X2 …. X3 ….

A key feature of the above restriction is that it only applies when the relevant Xs are in a c-command configuration; in particular, X2 blocks X3 just in case it c-commands X3.