Has Economics Finally Become an Immature Science? Mapping Economics at an Epoch of Fragmentation, by Combining Historical Perspectives and New Quantitative Approaches
Partecipanti al progetto
Descrizione del progetto
- Università degli Studi di SIENA (Alberto Baccini; principal investigator)
- Università degli Studi di TORINO (Mario Cedrini; associated investigator)
Continuously growing in size and diversity, economics appears as a more and more heterogeneous discipline, tending towards a state of “immaturity”, to borrow from Kuhn. Due to specialization, scientific progress occurs at the frontier, not within one paradigm, but within many local paradigms. Mainstream economics is now populated by a plurality of research programmes that significantly deviate from the neoclassical core. The resulting pluralistic landscape is a theoretical challenge to the established view of the mainstream of the discipline as monolithic.
This research project aims to provide the set of “science maps” – representing the structure and interconnections between research programmes in historical perspective – required to navigate the discipline at a time of increased fragmentation. The originality of the analysis stems from explicitly integrating qualitative and quantitative techniques, in the awareness that quantitative analysis can be safely used to formulate and test theories only when supported by qualitative research. The four main tools here employed to draw such maps are selected based on their ability to mimic the possible inherent properties of the puzzle to explain, that is fragmentation. This latter can be “content-based” (c-b) or “content-neutral” (c-n), that is of mainly institutional nature; and it can result from scholars’ explicit quest for differentiation from the core (“explicit structure”, es), or be the latent result of specialization as research strategy (“latent structure”, ls). Upon completion, the research team will have produced:
1) "History of economic thought" (HET) maps (c-b, es), representing – in a (non-traditional) “intellectual history” perspective, focusing on communities of scholars as unit of analysis – the history of economic theory as sequence of research programmes having origins and developing in localized contexts;
2) "Interlocking editorships" (IE) maps (c-b, ls), focusing on the basic tools of scientific communication in economics (the dataset will include editorial boards – observed every ten years – of all economics journals from 1926 to 2016). IE maps depict relationships between different economics journals in different moments of time by scrutinizing through network analysis tools the crossed presence of scholars on their editorial boards;
3) "Gatekeepers of economics" (GoE) maps (c-n, es), which focus explicitly on editors. While IE maps associate overlapping journal boards with scientific proximity, GoE maps emphasize the academic-power dimension of ties between economics journals;
4) “Topic-modeling” (TM) maps (c-n, ls), which result from applying unsupervised text-mining techniques to the full content of the whole collection of articles published in the economics journals stored in the digital library JSTOR. Automatically detecting recurring themes (topics) in the form of co-occurring words, TM maps are able to decode the latent structure of economics in historical perspective.
The set of maps thereby obtained will show evidence of the cross-fertilization between the “distant reading” perspective typical of quantitative analyses of the history of economics (the originality the method stemming from considering the full text of economic articles, rather than metadata or specific features) and the “closer” reading of established qualitative approaches to the evolution of the discipline. Acting as virtual traits d’union between HET and TM maps, the “intermediate” reading perspective provided by maps of the dissemination of knowledge (IE and GoE maps) also remedy the lack of accurate data that impedes robust historical bibliometric analyses.
Once integrated into a coherent whole, the maps will contribute to the general purpose of identifying “fragmentation markers” for the discipline, to be employed to test alternative theories about the evolution of economics. A non-secondary, concrete aim of the project is to highlight possible major faults or biases involved in human-assigned classification systems like the Journal of Economic Literature codes. Aimed to both produce and facilitate further knowledge development, the proposed mapping exercise and its innovative methodology are expected to have significant scientific impact. The project will give life to an interactive open-science platform making all datasets and maps (the general “latent” – topic-guided – map of the economics discipline in historical perspective; “IE” and “GoE” maps of journal networks over time, and hybrid maps of specific communities of scholars with various “zoom” levels, combining HET and TM techniques) freely available to scholars under CC license for further explorations. The platform can represent a preliminary step towards developing open, innovative online teaching and learning tools and resources, characterized by interactivity, complexity, and implementable contents, stimulating critical reasoning in learners.