Transitive causatives — den Dikken's affixal-particle analysis #

@cite{dendikken-1995} @cite{baker-1988}

@cite{dendikken-1995} chapter 5 extends the SC-in-SC template from triadic constructions (book chs. 3-4) to transitive causative constructions. Central thesis: so-called "applicative" and "causative" morphemes (Bantu -il-, Dutch ver-, Sanuma -ma, Indonesian -kan, French en-, a-, dé-) are affixal particles — the same syntactic SC heads as English particles, just morphologically bound to the verb.

The affixal-particle thesis (book p. 235 ex. 25) #

ver- = -ma = -kan = PARTICLE

The cross-linguistic homophony of "applicative" and "causative" affixes is dissolved: they are all affixal particles occupying the X position in the SC-in-SC template. The verb supplies the lexical content; the particle supplies the structural slot for SC3.

Causative D-structure template (book p. 239 ex. 32; p. 246 ex. 48) #

[VP V_caus [SC1 Spec_θ' [VP V_emb [SC2 Spec_θ' [XP X [SC3 NP_obj [PP P NP_causee]]]]]]]

V_caus = matrix causative verb (French faire; often empty for affixal causatives like Dutch ver- where the matrix CAUSE is non-overt — book p. 238).
V_emb = the embedded causativised verb (e.g. porter, manger, dormir) — supplies the lexical content of the causation target.
X = affixal particle slot (∅ in French faire; ver- in Dutch; -ma in Sanuma; -kan in Indonesian).
NP_obj = embedded direct object (the patient of V_emb).
PP = dative phrase whose object is the embedded subject (the causee).

Triadic ↔ causative parallel (book p. 246 ex. 47/48) #

Triadic:   [VP V_triadic [SC1 Spec_θ' [VP "BE"  [SC2 Spec_θ' [XP X [SC3 Theme  [PP P Goal]]]]]]]
Causative: [VP V_caus    [SC1 Spec_θ' [VP V_emb [SC2 Spec_θ' [XP X [SC3 EmbObj [PP P Causee]]]]]]]

Identical template. The only difference: the V slot in the lower VP holds the empty copula BE in triadics but the lexical embedded verb in causatives.

Two cross-linguistic correlations (book p. 241 ex. 35; p. 243 ex. 41) #

Both fall out of the structural assimilation:

I.  Case of *causee* in transitive causatives = Case of *Goal* in PD.
II. Case of *embedded direct object* in transitive causatives =
    Case of *Theme* in PD.

The causee occupies the PP-position SC3 has for the Goal in triadics; the embedded direct object occupies the SC3-subject position SC3 has for the Theme.

Cross-linguistic data anchored in this file #

French: Marie a fait manger des bonbons à ses enfants — matrix V is faire, X = ∅, V_emb = manger, dative à-PP holds the causee ses enfants. (Book p. 248 ex. 49.)
Dutch: Jan verschafte de kinderen eten — matrix V is empty, X = ver- (the affixal particle), V_emb = schaffen (provide). (Book p. 234 ex. 19b.)
Sanuma (Yanomami, Borgman 1989): kamisa-nö setenapi te niha manasi sa ta-ma-na-ni ke 'I made the non-Indian see the guan bird' — X = -ma, embedded V = ta (see), causee = setenapi. (Book p. 240 ex. 34b.)
Indonesian (Voskuil 1990): Parto menidurkan Ratna 'Parto made Ratna sleep' — X = -kan, embedded V = tidur (sleep), causee = Ratna. (Book p. 232 ex. 13b.)

Cross-references #

Phenomena/ArgumentStructure/Studies/Dendikken1995 — the triadic formalisation (book ch. 3). Same SC-in-SC template, V-slot filled with abstract BE rather than V_emb.
Phenomena/Constructions/ParticleVerbs/Studies/Dendikken1995 — the simplex PVC formalisation (book ch. 2.4). Particles in PVCs ≡ affixal particles in causatives at the syntactic level (book p. 235 ex. 25).
Theories/Syntax/Minimalism/Applicative.lean — Pylkkänen's high/low applicative typology. Per chronological dependency, this 1995 file does not import Phenomena/ArgumentStructure/Studies/Pylkkanen2008; the reverse direction is appropriate.

Scope (and what's deferred) #

Formalises the structural template (32)/(48) and verifies it satisfies IsSmallClause at every nested SC layer for both the French and the Dutch instantiations. The cross-linguistic affixal-particle thesis (eq. 25) is recorded as a structural identity claim about the X position's category.

NOT formalised (each pending substrate work):

Predicate Inversion + LF P-incorporation for causative DOC analogues (parallel to triadic; needs LF-reanalysis substrate, see SC/particles architectural target memo).
"Ergativisation" of V_emb (book §5.3.6) — den Dikken's claim that the embedded verb loses its external θ-role when embedded under V_caus.
Case-licensing chains for causee + embedded direct object across languages (would need a Case-feature substrate).
Clitic-climbing data (book §5.3.7) — requires a clitic-movement primitive in addition to Predicate Inversion.

Transitive causatives — den Dikken's affixal-particle analysis #

The affixal-particle thesis (book p. 235 ex. 25) #

Causative D-structure template (book p. 239 ex. 32; p. 246 ex. 48) #

Triadic ↔ causative parallel (book p. 246 ex. 47/48) #

Two cross-linguistic correlations (book p. 241 ex. 35; p. 243 ex. 41) #

Cross-linguistic data anchored in this file #

Cross-references #

Scope (and what's deferred) #

§1. Lexical items for the causative template #

§2. The X position: affixal particles across languages #

§3. PP and Spec primitives (parallel to the triadic file) #

§4. French causative example: faire manger des bonbons à ses enfants #

§5. Dutch ver--causative example: Jan verschafte de kinderen eten #

§6. Structural assimilation: French and Dutch share one shape #

§7. IsSmallClause witnesses (Phase 1.0 sorries) #

§8. The affixal-particle thesis as a categorial identity (book ex. 25) #

§9. Structural assimilation theorem (book ex. 47/48) #

§10. Bridge to Pylkkänen's applicative typology #