We tackle methodological challenges in experimental research on
Condition C reconstruction in A′-movement. We point out shortcomings of
previous experimental work in that it (i) relies on arbitrary thresholds
for the presence/absence of a Condition C effect and (ii) fails to take
relevant non-syntactic factors into account. We present two experiments
on Condition C reconstruction in German that address these challenges.
Experiment 1 involves a design that allows us to test Condition C
effects without reference to
thresholds by comparing subjects and objects, which is particularly easy
in SOV-languages. Experiment 2 investigates to what extent the
subject-object contrast can be related to non-syntactic factors. Our
results show that there is a significant subject-object contrast,
suggesting that the base-position of A′-movement matters. At the same
time, once the referent of the R-expression within the wh-phrase is
previously established in the discourse, the difference between subjects
and objects becomes very small (confirming a suggestion in Bruening & Al
Khalaf 2019). Our results thus suggest that Condition C should be
treated as a soft constraint (cf., e.g., Gor & Syrett 2019) that
interacts with/can be overriden by plausibility and salience.
