Tennis, J. T. (2016). “Methodological Challenges in Scheme Versioning and Subject Ontogeny Research.” In Knowledge Organization 43(8): 573-580. Methodological Challenges in Scheme Versioning and Subject Ontogeny Research Joseph T. Tennis University of Washington Abstract: This paper describes the philosophical and operational challenges to methodology in subject ontogeny research. The observation that indexing languages, comprising thesauri, classification schemes, taxonomies, and ontologies change over time has been a concern of the first order. The need to question the methods and aims of subject ontogeny and scheme change are second order work requiring ontological and epistemic assumptions. Operational concerns for the study of scheme change and subject ontogeny are species of one two-sided issue: isolation quantification. While some foundational issues make the study of the topic difficult, there may be some helpful techniques including phenomenology. 1.0 Introduction The observation that indexing languages, comprising thesauri, classification schemes, taxonomies, and ontologies change over time has, in the past, been a concern of the first order (Tennis 2015b). That is, the methods brought to bear on studying indexing language change have grown out of attempts at building faithful representations of a universe of knowledge for a particular user group. Those methods include literary warrant comparisons (Hulme 1911-12), facet analysis (Ranganathan, 1967; Gnoli ed., 2008), user studies (Fidel 1994), domain analysis (Hjørland 2002; Tennis 2003), cultural and ethical warrant (Beghtol 2002; Olson, 2001; Smiraglia 2014), and discussions on the structure and nature of classificatory structures (Broadfield 1946; Olson and Schlegel 2001; Tennis 2016; Parrochia and Neuville 2013; Lee 2011; Dawkins 1976; Frické 2012, passim). In all of these cases, the methodological concerns focus on a faithful representation of the literature so that users can find items in the collection or make sense of the scope and range of the collection. This is coupled with the concern with creating a parsimonious divide among literatures – that is the differences that make a difference. Effective and efficient access is the desideratum. Even with the ethical and cultural warrant methodologies, there is a pragmatic concern directed toward the goal of the indexing language working effectively and efficiently in order to place appropriate and relevant literature in front of a user based on their queries. What is lacking in this approach is a reflection on the complexity of the work of maintaining indexing languages once they are built. Further, it is left to other methods, second order methods and foundational methods, to focus on the internal consistency of indexing languages through change, as well as the constant investigation into the foundations on which these indexing languages are built (Tennis 2016). First order literature looks at the practice of designing and implementing classification schemes. This literature is core to our understanding of classification and shapes the topics of our canonical texts (e.g., Berwick Sayers 1955; Ranganathan 1967; Vickery 1960; Hunter 2002). The twentieth century saw innovations and advancement in the form of faceted classification and analytico-synthetic classification design methods devised by S. R. Ranganathan and the Classification Research Group (CRG). This is first order because we need to have classificatiom schemes in order to study how best to improve them.