X immediately c-commands Y within tree root iff X c-commands Y
(in root) and there is no Z such that X c-commands Z and Z
c-commands Y (in root).
This is the "closest" c-command relation.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The Head Movement Constraint (HMC)
A head X can only move to the head Y that immediately c-commands X.
Formally: If X moves to Y, then Y immediately c-commands X's base position, with no intervening heads.
Equations
- One or more equations did not get rendered due to their size.
Instances For
A movement violates HMC iff it doesn't respect it
Equations
- Minimalism.violatesHMC m root = ¬Minimalism.respectsHMC m root
Instances For
Head-to-specifier movement always violates HMC.
From Harizanov (p.12, p.29): In head-to-specifier movement, the head X becomes a maximal projection in its derived position.
This violates HMC because:
- HMC requires
isHeadIn mover root isHeadInrequires¬isMaximalIn mover root- But head-to-spec movement is defined by
isMaximalIn mover result - Therefore the mover is not a head in the result
- Therefore HMC fails
Any head-to-spec movement violates HMC, by the very definition of what head-to-spec movement is.
Position-aware HMC: a head must be non-maximal AT ITS POSITION
Equations
- One or more equations did not get rendered due to their size.
Instances For
Position-aware HMC violation
Equations
- Minimalism.violatesHMC_positional m root pos = ¬Minimalism.respectsHMC_positional m root pos
Instances For
Head-to-specifier movement (positional) violates HMC
This version works correctly with multidominance:
- The mover is maximal AT ITS DERIVED POSITION
- This means it is not a head at that position
- Therefore HMC fails for the derived position
Unlike the global version, we don't need to claim global maximality.
Amalgamation: a post-syntactic (PF) operation
Unlike syntactic head movement, Amalgamation:
- Happens at PF, not in narrow syntax
- Respects the HMC (is strictly local)
- Results in phonological fusion without syntactic effects
From Harizanov (Section 3.3): Amalgamation "trades" the syntactic relation between a head X and the head Y of its complement for a PF relation of affixation. Since affixation requires adjacency, Amalgamation is strictly local and cannot skip intervening heads.
Example: French V-to-T "Jean ne parlait pas français" - V amalgamates with T at PF
- target : SyntacticObject
The element that amalgamates (the "target")
- host : SyntacticObject
The host (what it amalgamates to)
- is_local (root : SyntacticObject) : immediatelyCCommands self.host self.target root
Amalgamation is LOCAL: host immediately c-commands target. This is the defining property that distinguishes Amalgamation from syntactic head movement.
Instances For
Amalgamation cannot skip intervening elements.
This formalizes Harizanov's claim (Section 3.3, p.15): "Amalgamation-based displacement obeys the Head Movement Constraint"
The proof is immediate from the definition: Amalgamation requires
immediatelyCCommands host target root, which by definition means
there is NO z such that host c-commands z and z c-commands target.
This is what distinguishes Amalgamation from syntactic head movement, which can skip intervening heads (as shown by Bulgarian LHM and V2).
If there's an intervening element, the displacement cannot be Amalgamation
This provides a diagnostic: if we observe a head displacement that skips an intervening head, we know it must be syntactic movement, not Amalgamation.
From Harizanov (Section 3.3): The properties of Amalgamation-based displacement "differ substantially from those of Internal Merge."
Amalgamation respects locality (the host c-commands the target within root)
Diagnostic: does this movement involve syntactic head movement?
Equations
- Minimalism.isSyntacticHeadMovement m root = Minimalism.violatesHMC m root
Instances For
Diagnostic: is this compatible with Amalgamation?
Equations
- Minimalism.compatibleWithAmalgamation m root = Minimalism.respectsHMC m root