The purpose of this study was to investigate the temporal unfolding of loca
l acoustic information and sentence context using both cross-modal interfer
ence (CMI) and word-monitoring tasks. The timing of sentence context effect
s have important theoretical implications for models of language processing
(e.g., initial context independence vs. initial interaction). Yet, differe
nt tasks tend to yield different results. For both Experiments, stimuli fro
m an acoustically manipulated "goat-to-coat" continuum were embedded in sen
tences whose interpretation was biased toward either "goat" or "coat." In e
xperiment 1 (CMI), the primary task was listening to sentences for comprehe
nsion; the interference task was a word/nonword decision to an unrelated vi
sual probe that appeared at one of three positions within the sentence. Res
ults showed immediate effects of the acoustic manipulation, but only delaye
d effects of sentence context. These results were interpreted to indicate t
hat phonological processing is initially context-independent but is followe
d by rapid context integration. Experiment 2 used a word-monitoring task: R
esponse times were significantly longer when sentence context was incongrue
nt with the,monitoring target, showing an immediate effect of context. The
apparently contradictory results of the two experiments together support an
account of language processing in which phoneme categorization is initiall
y independent of sentence context unless an explicit judgment about the ide
ntity of the target is required.