One of the most powerful ideas of the past quarter century is that our personal and social identities and their various components (gender, age, class etc.) are not natural properties, but are socially constructed through the combination of our actions and others' interpretations of them. Research in sociolinguistics has shown that language is critical to this process, and our interlocutors take into account not only what we say but also how we say it. For example, studies have shown that speakers using colloquial expressions (like pronouncing the word 'drinking' as drinkin') are perceived to be friendlier than speakers saying the same thing using more formal expressions (i.e. saying drinking). On the other hand, the latter are often perceived to be more competent than those using colloquial language. Although identity construction through language is a fundamental notion in the humanities and social sciences, we do not yet have a precise characterization of the cognitive processes involved. As a result, these influential ideas have remained isolated from work in cognitive science, computer science and artificial intelligence. The goal of SMIC is therefore to construct a mathematically explicit, computationally implemented theory of the identity construction process based on the hypothesis that identity construction is very similar to other kinds of linguistic communication, i.e. hearing drinkin' and thinking that the speaker is friendly is the same basic cognitive process as hearing drinkin' and thinking about imbibing liquid. Modeling linguistic communication is a central concern of formal pragmatics, and recent developments in this field have created exciting new experimental, mathematical and computational tools for studying linguistic meaning in context. SMIC aims to take advantage of these developments to build the model, and, in doing so, unite diverse lines of research across the social, cognitive and information sciences.