Documentation

Linglib.Phenomena.ArgumentStructure.Passive

The Passive Construction #

The Phenomenon #

English has an active-passive alternation:

The passive:

  1. Promotes the object to subject position
  2. Demotes the subject to an optional by-phrase
  3. Uses auxiliary "be" + past participle

The Data #

(1a) John kicked the ball. ✓ active transitive (1b) The ball was kicked. ✓ passive (short) (1c) The ball was kicked by John. ✓ passive with agent

(2a) Mary chased the cat. ✓ active transitive (2b) The cat was chased. ✓ passive (short) (2c) The cat was chased by Mary. ✓ passive with agent

(3a) *The ball was kicked the game. ✗ passive with object (impossible) (3b) *Was kicked John. ✗ passive missing subject

@cite{collins-2005} Data @cite{collins-2005} #

C-command asymmetries (§3) #

In passive, the derived subject (= original object) c-commands the by-phrase (= original subject). Evidence:

(10a) Every letter₁ was signed by its₁ author. ✓ bound variable (10b) *Its₁ author was arrested by every cop₁. ✗ no binding into subject (10c) No letter was signed by anyone. ✓ NPI licensed in by-phrase (10d) *Anyone was fired by no manager. ✗ no NPI licensing of subject

Particle placement (§3.1) #

(15a) The cat was let out. ✓ passive with particle (16a) *The cat was let out the dog. ✗ no post-particle object in passive (16b) *The cat was let the dog out. ✗ no intercalated object in passive

Auxiliary selection (§3.3) #

(23a) John has kicked the ball. ✓ have + past participle (23b) The ball was kicked. ✓ be + past participle (23c) *John has kicking the ball. ✗ have requires participle (23d) *The ball was kick. ✗ be-passive requires participle

source
def Phenomena.ArgumentStructure.Passive.data :
StringPhenomenonData

Passive construction data.

Pure empirical data with no theoretical commitments. Theories interpret this via their Bridge modules.

Equations
  • One or more equations did not get rendered due to their size.
Instances For
    source
    def Phenomena.ArgumentStructure.Passive.cCommandData :
    StringPhenomenonData

    C-command data establishing that the derived subject in passive c-commands the by-phrase. This is predicted by smuggling: the object raises to Spec-TP (c-commanding everything below), while the external argument remains in Spec-vP (inside the by-phrase).

    Equations
    • One or more equations did not get rendered due to their size.
    Instances For
      source
      def Phenomena.ArgumentStructure.Passive.particleData :
      StringPhenomenonData

      Particle data showing that in passive, the verb + particle form a constituent (PartP) that excludes the object. The object has been smuggled out of PartP to Spec-TP, stranding the particle with the verb.

      Equations
      • One or more equations did not get rendered due to their size.
      Instances For
        source
        def Phenomena.ArgumentStructure.Passive.auxiliaryData :
        StringPhenomenonData

        Auxiliary evidence for PartP as a syntactic constituent. Both have and passive be select a participial complement (PartP). Have c-selects PartP directly; passive be requires PartP to move to Spec-VoiceP for licensing (@cite{collins-2005}, conditions 24-25).

        Equations
        • One or more equations did not get rendered due to their size.
        Instances For
          source
          def Phenomena.ArgumentStructure.Passive.idiomData :
          StringPhenomenonData

          Idiom chunks preserved under passivization show that the object originates in the same VP-internal position in both active and passive. This supports UTAH: the external argument is generated in Spec-vP in both active and passive.

          Equations
          • One or more equations did not get rendered due to their size.
          Instances For