Speaker: Ioannis Katochoritis (MIT)
Title: To Unlock is to (Re)Merge: Locality Domains, Intervention and Minimal Compliance
Time: Tuesday, April 28, 1:00pm – 2pm
Location: 32-D461
This is Part II of a distant mid-March syntax square. Some things in the world have changed in-between, but not the need to undo locality domains. Recall:
Rackowski and Richards (2005) propose a Phase Unlocking operation: phases can be made transparent for extraction if they first Agree with a higher probe, which may then attract a goal from within the phase’s domain. If phases are by default potential movable goals that intervene (Abels 2003), then prior Agree with the phase allows the probe to ignore that phase and Agree with an embedded goal, as per Richards’ 1998 Principle of Minimal Compliance (PMC) in (1).
(1) Once a probe P Agrees with a goal G, P can ignore G for the rest of the derivation.
However, the unlocking program raises some questions: 1. Are both unlocking and successive-cyclicity required to escape a phase? 2. How is unlocking compatible with the Phase Impenetrability Condition? 3. How does the PMC allow to ignore an Agreed-with locality domain? 4. Is unlocking an exceptional mechanism or a variant of some broader strategy to obviate intervention?
To address these questions, I suggest something like (2):
(2) Unlocking requires (re)merge of the containing phase XP to the specifier of the probing head H, before an XP-embedded goal YP can subextract to an outer SpecHP.
Hence, it is not mere Agree, but (re)merge of the phase with the probing head that makes it transparent for subextraction, in what yields a derived multiple-spec configuration. I will argue that (2), especially when extended to external merge to encompass c-selection via sisterhood, may reconcile unlocking with Phase Theory, structurally derive the PMC, the Weak PIC, certain A-movements and subextraction asymmetries, as well as unify syntactic strategies of obviating intervention.
The discussion will also be relevant to (and maybe arguing against) the recent view (see Halpert & Zeiljstra 2025) that there are no designated phase heads with special status, and all phase-like locality effects should be reduced to Relativized Minimality (Rizzi 1990).