Even if speech perception and production have been revealed to share a common recruitment of both discrete auditory and motor areas, this overlap being also common to reading and listening, no study has investigated the involvement of larger networks in the three tasks yet. So, we first identified the multimodal bilateral brain areas conjointly activated and asymmetrical during listening, production and reading of word-list using fMRI in 144 healthy right-handers (27 years ± 6 years). Such a selection made it possible to unravel 14 regions of the left hemisphere including motor, premotor and inferior parietal cortical areas. On the right, 7 regions were selected, including the posterior Human Voice Area (pHVA). To characterize the network organization within these 21 regions, we then analysed resting-state functional connectivity in 138 of the same participants. It allowed us to segregate a network of executive areas in relation with task completion from a bilateral WORD_CORE network composed of (1) all left areas supporting the action-perception cycle, in which articulatory gestures are the central motor units on which word perception, production, but also reading, would develop and act together according to the motor theory of speech; (2) the right pHVA, acting as a prosodic integrative area, underpinning the intertwining across hemispheres between prosodic (pHVA) and phonemic (left SMG) processing. The present results show that word processing, whatever the language modality involved, is based on a network of plurimodal areas hosting processes specific to each hemisphere and on their cooperation built upon synchronisation at rest.