This article employs game theory to contribute to sociolinguistics (or the economics of language). From both the synchronic and the diachronic perspective, we are interested in the conditions (of language learning and literary production) that make some languages dominate others. Two results are particularly noteworthy: (i) Translations have an ambiguous effect on domination. (ii) We offer three different explanations of how a past language like Latin or Sanskrit can develop into a standard for literary production.