Why researchers publish in journals not indexed in mainstream databases: training, bridging and gap-filling

Although journals indexed in mainstream Journal Indexing Systems (JIS), i.e. Web of Science (WoS) and Scopus, have more visibility, reputation and are more valued in evaluation, in developing countries researchers keep publishing in journals not indexed in mainstream databases, but indexed instead in alternative JIS such as Scielo or RedALyC. The conventional explanation to this behaviour is that developing countries' research often does not have sufficient quality so as to be published in journals indexed by WoS or Scopus. We conducted 30 interviews to researchers in Colombia working in chemistry, agricultural sciences, and business and management asking the reasons for publishing in different types of journals, in particular those indexed by mainstream and alternative JIS. The answers provided by the researchers were classified into three reasons. The first reason is that journals in alternative JIS offer a space for training in publishing, both as an introduction to academic publishing and as a step-stone towards publishing in WoS. The second reason is that journals indexed by Scielo and RedALyC have a knowledge-bridging function, providing a link between articles covered by mainstream JIS and articles of regional communities with limited access to WoS or Scopus journals. The third reason is that alternative JIS journals have a knowledge-gap filling function, allowing the publication of topics that are not well covered in WoS-indexed journals, such as locally relevant agricultural products or regional history. We conclude that scientometric indicators based on mainstream JIS underrepresent the contribution of research from developing countries - as they do not value these training, knowledge-bridging and gap-filling functions. We discuss these findings in the light on universalistic versus particularistic conceptualisations of science.