Exploring innovation studies through a blurb lenses: A textmetric approach to journal positioning

In this contribution, we articulate an argument to expand the realm of scientometrics and bring in journals more explicitly to the fore. Our analysis of the journals’ blurb evidence has generated novel insights about the outlets more focused on innovation studies and pointers that suggest this textmetric approach is of potential use for future research and in other areas. In particular, practitioners, including information scientists, administrators that monitor productivity and progress, and policy-makers who are concerned with the rhythm and direction of the science system, may benefit from efforts to open the academic journal black-box. Taking advantage of short self-promotional contents, referred here as blurbs, and by applying cosine similarity we are able to distinguish unacknowledged journal profiles and positionings: four clusters of journals emerged that have not been recognised before: “Innovation and technology”, “Sectoral economics”, “Organisational theory”, and “Strategic management”. Along the way, our research was able to validate the (heterogeneous) innovation-friendly journal sample we departed from and identify seven journals that belong to the core group of innovation studies (the “Innovation and technology” cluster, which is where Research Policy and Industrial and Corporate Change are located). By establishing and taking a journal as the benchmark in innovation-related research (Research Policy, is the first of a set of innovation-focused journals to enter the “academic marketplace”), we can also propose a semantic alternative to conventional metrics, i.e., a novel “soft” ordering of journals which can be used to guide journal-paper matching. This discourse-driven or substance-sensitive mapping of journals is quite different from the usual, citation-dependent rankings methods. We find that the “Innovation and technology” journals, where Industrial and Corporate Change is included, are the closest ones to Research Policy, followed by a hierarchy of other journals not connected to a particular journal cluster or correlated with impact factor. The evidence we obtained suggests it is possible to adopt a “journalytics” approach to scientometrics; this can be done by considering easy-to-read narrative meta-information (the small promotional materials we dubbed blurbs) to enhance our understanding of journals. The results found for the innovation studies field suggest that the methodology can also be useful to comprehend, compare and classify serial literature channels, i.e., the journals, in other academic fields. More should, and can, be done in the future about journals as objects of analysis and modern platforms of research distribution