jats:titleAbstract</jats:title> jats:pWhile in recent decades historians have, to some degree, engaged with anthropological debates over the nature of ‘religion’, they have tended still to take the key term ‘belief’ somewhat for granted. This article suggests that we have inherited an Enlightenment legacy of thought on ‘belief’ that tends to treat it as credulity when applied to the general masses (taking Edward Gibbon as an important exemplar of that legacy). More recently, the theorist Bruno Latour has written about belief in terms of J. L. Austin’s speech-act theory, a useful theoretical move, despite the fact that Latour himself in some ways mirrors Gibbon’s perspective, particularly as it applies to the pre-modern. Using a range of examples to discuss change over time, via a case study of southern France about 1000 to 1300, this article argues that social historians may find it useful to consider these issues in terms of performative ‘belief acts’, where the contextual setting — the ‘conditions of felicity’, to use Latour’s and Austin’s phrase — are amenable to historical analysis.</jats:p>