In psycholinguistics, written production has received only limited attention compared to spoken production, especially on the word level. The goal of the current project is to build a computational model of the processes underlying written picture naming based on errors in mono- and bilingual writing. For this, I will extend the interactive two-step model of spoken word production (Dell et al., 1997) to the written modality by adding a letter representation layer, based on the verbal theory of writing by Bonin and colleagues (2015). A first study will model the role of phonology and orthography in written picture naming. In a second study, the model will be extended with a motor layer to simulate typewriting, with the focus on explaining the motor error patterns. Finally, in a third study, the model will be extended to a bilingual context, studying the interactions between the native and second language's lexicons, phonology, and orthography during second language production. Each model will be built based on (existing) corpora of writing errors and will be tested with new experimental written picture naming data. The modelling of errors in written picture naming can be a first step toward a better understanding of the mechanisms underlying writing.