White 2014: The Modalized Complement Analysis of forget #
White 2014 maintains a uniformly factive denotation for forget by
positing a covert root modal in non-finite complements (the Modalized
Complement Analysis, MCA). This file verifies the MCA's coverage on its
own terms — the finite-CP / plain-infinitive asymmetry it was designed
for — and confirms Fragment-level consistency: the English Fragment's
split of forget into forget (negative implicative, infinitival) and
forget_rog (factive, finite) aligns with the frames the MCA
classifies correctly.
The MCA's structural prediction is formalized as
Williams2026.MCAPredictsModal (the reconstruction in the later
paper's §3.1.1, whose study file also prosecutes the gerund
refutation, Williams2026.mca_overpredicts_gerund — per the
chronological-dependency rule that comparison lives there, not here).
The pre-existence classifier used in the fragment-consistency section
is the post-2014 project-canonical apparatus
(Williams2026.SatisfiesPreExistence); its use here is the sanctioned
canonical-successor deviation.
MCA coverage of the canonical contrast #
Finite CP: the MCA correctly predicts no modal insertion.
Plain infinitive: the MCA correctly predicts modal insertion.