Agent-based modelling belongs to the wider category of computational modelling and is used to replicate real-world social or physical systems and simulate their behaviour in a computer environment, with a view to studying the rules that govern them. Although agent-based modelling has quickly gained recognition in the natural sciences, social scientists only started to explore its potential in the last few decades. This trend has gained momentum over the last ten years, and applications of agent-based modelling to the social sciences have been steadily increasing. Researchers dealing with complex phenomena (in both the social and natural sciences) are increasingly switching from purely analytical (i.e. mathematical) approaches (which often attempt to find a 'closed-form solution' to a problem) to computation-based approaches (which rely on algorithms and simulations). Whereas applications of analytical methods to study language-related issues have a long-standing tradition, systematic adoption of a complexity-based view of language matters and application of computational methods have yet to establish themselves. However, recent research in language studies shows attempts to address the complexity of language-related issues in a more systematic way. Particular attention is devoted to the fact that such issues involve numerous agents and variables, and causal links between these variables are often non-linear. Although it may be challenging to model such systems through equations, agent-based modelling offers a natural solution, in that it provides an easy way of replicating various degrees of diversity in a computer environment. These models can then be run many times with various settings to study the system's short-term and long-term behaviour and the way it responds to different initial conditions. Agent-based modelling lends itself very well both to studies that investigate the nature and functioning of languages and to sociolinguistic studies that focus on language-mediated interactions between individuals.