Unity of science has been a very popular topic in history. The twentieth century saw the development of a radical version of this vision, the project of unified science advanced by logical positivists of the Vienna circle, and a series of attempts to build unity on a model of theory-reduction. Critics of this strategy, however, have opposed such efforts – and successfully illustrated the somehow mythical nature of the underlying logic – by emphasizing the irreducible plurality of sciences and the resulting, fundamental disunity of science. The significant difficulties associated with defending convincingly any “unity-of-science” project on traditional bases (ontological, theoretical, or methodological reduction unifications) directed the attention of both supporters and critics towards more flexible versions of unity (such as the proposal of unification as a “regulative ideal”) and to the theoretical and practical foundations of interdisciplinarity – unity being seen in terms of integration between different disciplinary approaches.

Some recent proposals for unifying the social sciences have originated within the economics profession. Our aim is to investigate the reasons why economists are increasingly debating this issue, and the (theoretical) feasibility of such projects, in the light of the current state of the relationships between economics and contiguous disciplines. The paper proposes a general theoretical framework in order to distinguish a set of possible options for integration between social sciences. As pointed out by various philosophers of science, interdisciplinarity (in its general meaning) is old news, and even the most integrative forms of disciplinary interaction seem to emerge from established disciplines. We therefore use disciplines as unit of analysis, and adopt the so-called “nation” metaphor in order to investigate, using the analogy with Dani Rodrik’s “world political trilemma” (whereby democracy, national sovereignty and global economic integration are mutually incompatible; it is possible to combine any two of the three, but never have all three simultaneously and in full), three ideal types of disciplinary integration, or three different roads to the realization of the unity of social science (which we here call, respectively, “reductionism”; “federalism”, and “complexity”).

The framework is then applied to three main “integration” projects, with a view to throwing light on their theoretical underpinnings and expected outcomes: 1. David Colander’s (and others’) idea of a ‘transdisciplinary’ social science, based on cross-fertilization of methods and approaches; 2. arguments advanced by influential philosophers of social sciences and economics (like Don Ross, Harold Kincaid) that regard disciplinary adaptation (of economics to sociology and psychology, and vice versa) as a possible solution to the unsolved problem economics has with the “social” dimension; 3. Gintis’ proposal for a new theoretical framework using the evolutionary perspective and game theory as bridges for unifying otherwise incompatible disciplinary approaches to human behavior.

While throwing light on the features of disciplinary relations as conceived in each of these projects, we speculate about their “implicit” origins, which we generally identify in the disunity characterizing the current fragmented state of economics. We then discuss the potential impact of these integration projects by focusing on the issue of pluralism in both the social sciences – which some proposals explicitly want to counteract – and in particular within economics, at a time of pervasive specialization, continuous creation of scientific research niches and declining cross-science research programs.