A Diachronic Analysis of that-t Effect between Complement and Relative Clauses

This paper proposes to account for contrastive that-t effect between complement and relative clauses from diachronic perspective. The complementizer that carried a finiteness feature denoted by a mood feature including a force feature in OE and early ME, so it did not trigger that-t effect. But as a force feature was substituted for a demised mood feature from late ME, it came to trigger that-t effect. The spread of that from a complementizer to a relativizer was related to the loss of inflections and the change of gender system. If a subject from a relative clause undergoes wh-movement, the derived structure has exhibited counter that-t effect from OE through ME to ModE. Regarding wh-subject movement, the deletion of that in the complement clause or the retention of that in the relative clause reflects the computational condition to avoid syntactic-perceptual ambiguity. Since a wh-moved subject from a complement clause carries an interrogative force feature, the complementizer that with a declarative force feature must be deleted on C. In case of a wh-moved subject from a relative clause, since the relativizer that carries a [-wh]-declarative force feature, it is compatible with a declarative force feature on C. This means that that-t effect is a specific rule which is applied only to wh-interrogatives in ModE.