A conversation is a clear example of joint activity. However, little is known how the more basic sensori-motor processes underlying joint action apply to the complex interwoven between language production and comprehension during communication. The present project aims to contribute to our understanding of language as a joint action by exploring how the two fundamental processes that contribute to the success of any joint action, namely prediction and monitoring, are engaged during verbal communication. To do so, the temporal dynamics of the different levels of representation underlying speech production (conceptual, lexical and phonological), will be tested in the listener during prediction and monitoring of other’s speech. By means of five experiments, we aim at providing response to three specific research questions regarding prediction and monitoring of other’s generated speech. Firstly, we will explore how the language production, as a form of action, is involved in predicting other’s to be generated words. Second, we will explore the nature of those predicted representations. And finally, we will focus on exploring the monitoring processes engaged in detecting errors in other’s word production. Behavioral and EEG/ERP measures will help us to answer these questions. Considering the relation between language and action, the present work will help to determine how the basic sensory-motor skills apply to the complexities of language. Finally, answering these questions will contribute to enlarge our understanding of how language works during verbal communication.