It is unclear whether there are differences in the neural mechanisms of word planning when it arises after processing of verbal versus nonverbal material. Twenty participants took part in an EEG experiment where they completed verbal- and nonverbal-context picture-naming tasks. The goal in both tasks was to name a target picture that was presented after constraining and nonconstraining contexts. In the verbal settings, the contexts were provided as sentences (e.g. constraining: “The farmer milked a…”; nonconstraining: “The child drew a…”; target picture: COW), while in the nonverbal settings, the contexts were provided as two priming pictures (e.g. constraining: nest, feather; nonconstraining: cradle, television; target picture: BIRD). The target pictures were named faster following constraining contexts in both tasks. This indicates that word planning starts before the target picture onset in the constraining condition. In the verbal-context task, we replicated alpha-beta desynchronization in the constraining relative to the nonconstraining condition before the target picture onset. In the nonverbal-context task, we did not find any alpha-beta desynchronization. Interestingly, we found three ERP components that showed significant differences between the conditions. Even though this suggests that word planning occurs in a different manner before the target picture onset following nonverbal constraining contexts, participants do engage in conceptual preparation, which speeds up word planning. These results were mirrored when we computed correlations with naming times in the pre-picture interval. In the verbal settings, we only found significant correlations between naming times and the desynchronization in the alpha- and beta-bands. In the nonverbal settings, we only found significant correlations between naming times and the ERP waveform. In conclusion, we show that the alpha-beta desynchronization is associated with the word planning process, but conceptual preparation alone is not enough to elicit it.