Spoken word access is typically investigated by means of on-line experimentation, yet the lexicon itself contains a wealth of structure from which we can make inferences about processing. I review existing studies of the statistical structure of the English lexicon and conclude the following: (a) the lexicon is shot through with partial but statistically significant correspondences between form and meaning/function, reflecting the brain's predisposition for creating topographic mappings; (b) constraints within the phonology of English concerning the ordering of segments reflect the demands of incremental processing on the part of the listener; (c) conversational ­ "fast" ­ speech involves phonological changes that accentuate important aspects of the structure of the lexicon and protect the intelligibility of speech. Regarding the vexed issue of whether or not top-down processing exists in speech perception, I will argue that statistical studies of the lexicon can complement on-line experimentation and that they are at least consistent with a genuine flow of information from "higher" to "lower" levels of representation and processing.