Docs / Studies / Studies.Harley2017

The ‘Bundling’ Hypothesis and the Disparate Functions of Little v

Harley, H. (2017). In The Verbal Domain (pp. 3–28). Oxford University Press

Source Bibliography

Harley adopts Pylkkänen's view that the functions of little v may be bundled on one head or split across several, and proposes the hierarchy vP > ApplP > CausP1 > VoiceP > CausP2 (49). A pattern is a partition of the projections a structure realizes, whose parts are its heads, and the patterns of (49) bundle only neighbouring projections. Key's analysis of Turkish causatives follows by Vocabulary Insertion over these heads.

Main definitions

Feature: the projections of (49).
turkish, hiaki, korean, italian: the patterns of (49a–d).
vocabulary: Key's Turkish causative vocabulary.
ofCauseHead: Pylkkänen's root- and verb-selecting causative heads as patterns.

Implementation notes

A projection absent from a pattern lies outside the set it partitions; the projections (49) calls distinct are singleton parts.
Harley raises spanning only as an alternative to bundling, so being an interval of the realized projections is a theorem about the patterns of (49) rather than part of their type.
ofCauseHead is undefined on phase-selecting Cause, which would need a second Voice below CausP2.

TODO

Harley's diagnostics for Chol, Persian, Hiaki and Chemehuevi; Italian passives and light verb constructions; the footnote setting English apart from Pylkkänen's analysis.

References

Harley 2017, The ‘Bundling’ Hypothesis and the Disparate Functions of Little v
Pylkkänen 2008, Introducing Arguments formalized
Key 2013, The Morphosyntax of the Turkish Causative Construction

The hierarchy (49) and its patterns

Among (49a–d), only Italian puts Voice on the head of v.

theorem voice_mem_part_v L86–90

:

Feature.voice ∉ turkish.part Feature.v ∧ Feature.voice ∉ hiaki.part Feature.v ∧ Feature.voice ∉ korean.part Feature.v ∧ Feature.voice ∈ italian.part Feature.v

Built from , , ,

Key's Turkish causatives

A lexical causative allomorph applies to the head bearing Caus only if v is on that head, in every context.

theorem v_mem_of_applies L152–157

{s : Finset Feature} {P : Finpartition s} {ρ : DistributedMorphology.Root} {l r : List Terminal} (h : Morphology.Exponence.Applies (lexicalItem ρ) { focus := causTerminal P, leftCtx := l, rightCtx := r }) :

Feature.v ∈ P.part Feature.caus₁

Built from , , ,

In the lexical causative next to its root, the root's allomorph is inserted.

theorem subsetPrinciple_lexical L159–170

(ρ : DistributedMorphology.Root) :

DistributedMorphology.subsetPrinciple (vocabulary ρ) { focus := causTerminal lexical, leftCtx := [{Sum.inr ρ}] } = some (Exponent.lexical ρ)

Built from , ,

In the productive causative (49a), -DIr is inserted in every context.

theorem subsetPrinciple_productive L172–182

(ρ : DistributedMorphology.Root) (l r : List Terminal) :

DistributedMorphology.subsetPrinciple (vocabulary ρ) { focus := causTerminal turkish, leftCtx := l, rightCtx := r } = some Exponent.dir

Built from , , ,

Pylkkänen's causative heads

A causative head has a causee position iff its pattern puts v and Voice on different heads.

theorem hasCauseePosition_iff L197–201

{c : Pylkkanen2008.CauseHead} {P : Finpartition vCausVoice} (h : ofCauseHead c = some P) :

c.HasCauseePosition ↔ Feature.voice ∉ P.part Feature.v

Built from ,