A series of multiple regression analyses was conducted on a corpus of eye movement data to examine whether the influence of properties of words n-1 and n+1 on the time spent fixating word n changes as a function of whether word n is associated with a punctuation mark (i.e., whether or not a punctuation mark separates word n from either word n-1 or word n+1). The results suggest that distributed processing is not significantly impaired. However, punctuation marks also carry word class information and word classes are not evenly distributed across positions relative to punctuation marks. Word class probability does modulate parafoveal-on-foveal effects.