There appears an obvious fit between the application of ‘social media’ technologies to the making of scholarly editions in digital form and the markedly collaborative nature of the typical digital humanities project. Accordingly, it may be argued that the model of the collaborative project-based edition need only to be extended, to become ‘social’. This article questions that thesis, demonstrating the problems that can arise with collaborative projects applying digital methodologies to scholarly work through analysis of the Shakespeare Quartos and the European Virtual Museum Transnational Network projects and arguing that the term ‘collaboration’ needs critical examination. Indeed, to the extent that ‘collaboration’ may be closed, and may serve narrow scholarly purposes, it can be the antithesis of ‘social’. In place of project-based collaboration, this essay proposes that we see ‘social’ editions as grounded in communities, not in collaboration, and that the principle upon which they should be built is (following Shirky) ‘design for generosity’. This implies a different role for the editors and scholars from the academy: rather than the leaders of collaborations, we may become key participants in, and enablers of, communities. In turn, this mandates a loss of control: generosity means allowing others to use what is given freely, including in ways not foreseen, and even opposed, by those who created the data. For academics who are used to identifying control with assurance of quality, this is a difficult step. Wide adoption of the Creative Commons Attribution licence (without the ‘non-commercial’ restriction) would take us far towards these aims.