Ni, Crain, and Shankweiler (1996) present evidence to suggest that the focu
s operator only, can guide how reduced relative clause sentences are initia
lly parsed. In this paper, we demonstrate that this does not hold for relat
ive clause sentences that start with a noun-phrase, verb, noun-phrase const
ruction. We report an eye movement study in which subjects read reduced and
unreduced sentences of this type with and without the focus operator only
There mere longer first-pass reading times in the critical region of reduce
d sentences than in the same region of unreduced sentences, regardless of t
he inclusion of only. Furthermore, readers spent less time re-inspecting po
rtions of text after being garden pathed when reading reduced relative clau
se sentences that contained the focus operator than when reading reduced re
lative clause sentences that did not. We conclude that subjects initially s
yntactically misanalysed reduced relative clause sentences with and without
only, and the inclusion of a focus operator facilitated recovery procedure
s rather than guiding initial parsing. These results are inconsistent with
the referential theory and undermine the conclusions of Ni et al. (1996).