Amongst other things, an individual’s linguistic repertoire is an asset whose value depends on the surrounding linguistic environment, and a changing linguistic environment may improve or worsen a person’s situation. To normatively evaluate such changes will require a ‘metric’ of linguistic (dis)advantage. Unfortunately, existing approaches to measuring the communicative value of particular languages - such as de Swaan’s (2001) Q-value - cannot be directly appropriated for this purpose, since rival theories of justice attach different weights to different kinds of communicative opportunities. This paper develops, describes and compares three alternative metrics, drawn from the ‘equality of what’ debate, and addresses two questions. First, can metrics derived from the resourcist, welfarist and capability views be applied to evaluate language policies? Second, which view, if any, best captures the phenomenon of linguistic disadvantage?