There is no universal agreement regarding the meaning of the term “social software.” Clay Shirky, in his classic speech “A Group is its Own Worst Enemy,” defined social software as “software that supports group interaction” (Shirky, 2003). In this speech, this scholar of digital culture also observed that this was a “fundamentally unsatisfying definition in many ways, because it doesn’t point to a specific class of technology.” The example offered by Shirky, illustrating the difficulties of this definition, was electronic mail, an instrument that could be used in order to build social groups on the Net, but also to implement traditional forms of communication such as broadcasting, or noncommunicative acts such as spamming. In his effort to underline the social dimension of this phenomenon, rather than its purely technological aspects, Shirky decided to maintain his original proposal, and this enables scholars engaged in the analysis of virtual communities to maintain a broad definition of social software. Heterogeneous technologies, such as instant messaging, peer-to-peer, and even online multigaming have been brought under the same conceptual umbrella of social software, exposing this to a real risk of inflation. In a debate mainly based on the Web, journalists and experts of the new media have come to define social software as software that enables group interaction, without specifying user behaviour in detail. This approach has achieved popularity at the same pace as the broader epistemological interest in so-called emergent systems, those that, from basic rules develop complex behaviours not foreseen by the source code (Johnson, 2002). This definition may be more useful in preserving the specific character of social software, on the condition that we specify this carefully. If we include emergent behaviour, regardless of which Web technologies enter into our definition of social software, we will once again arrive at a definition that includes both everything and nothing. Emergence is not to be sought in the completed product, that may be unanticipated but is at least well-defined at the end of the productive cycle, but rather resides in the relationship between the product, understood as a contingent event, and the whole process of its production and reproduction. A peculiar characteristic of social software is that, while allowing a high level of social interaction on the basis of few rules, it enables the immediate re-elaboration of products in further collective cycles of production. In other words, social software is a means of production whose product is intrinsically a factor of production. Combining hardware structures and algorithmic routines with the labour of its users, a social software platform operates as a means of production of knowledge goods, and cognitive capital constitutes the input as well as the output of the process. If a hardware-software system is a means of production of digital goods, social software represents the means by which those products are automatically reintroduced into indefinitely-reiterated productive cycles. This specification allows us to narrow down the area of social software to particular kinds of programmes (excluding, by definition, instant messaging, peer-topeer, e-mail, multiplayer video games, etc.) and to focus the analysis on generative interaction processes that distinguish social software from general network software. Moreover, following this definition, it is possible to operate a deeper analysis of this phenomenon, introducing topics such as the property of hosting servers, the elaboration of rules and routines that consent reiterated cycle of production, and the relationships between actors within productive processes.