2 The Epistemological Dimension of Knowledge OrGANIZATION 1 Richard P. Smiraglia Ph.D. University of Chicago 1992. Visiting Professor August 2009 School of Information Studies, University of Wisconsin Milwaukee, USA ABSTRACT Epistemology is the science of knowledge. Knowledge organization is the science of the order of knowledge. The domain of knowledge organization is a discourse community in which inquiry takes place concerning knowledge, and its various orderings or sequences. The products of the domain are ordered segments of knowledge, and the rules either for discovering their natural orders, or the rules for imposing a useful sequence. This paper brings together some results from meta-analysis of the knowledge organization domain. Knowledge organization is a coherent domain with an ontological base, its extension, along a dimension stretching from concept theory (or semantics) to applied knowledge organization systems and an underlying teleology, which is the attempt to grasp the order of knowledge. The domain has common hypotheses first, that knowledge is made up of concepts; and second, that these concepts can be ordered in diverse and useful ways. There is epistemological consensus forming a dimension stretching from empiricism to rationalism. Finally, there are consistently marked dimensions within the domain, such that a) there is a trajectory from theoretical to applied on a conceptual continuum; and b) there is another trajectory from empirical to rational on an epistemological continuum. Keywords: Knowledge Organization. Discourse Community. Dimension of Knowledge. 1 What is epistemology in knowledge organization? Epistemology is the study of knowledge; i.e., it is the study of that which is known. Epistemology also is the division of philosophy that investigates the nature and origin of knowledge. Therefore, epistemology is the science of knowledge; the self-conscious inquiry of knowledge. Through rigorous inquiry epistemology tells us where knowledge comes from. The central problems for epistemology are the definition of knowledge, and the means of its acquisition. Historically there are just two main schools of epistemological thought: rationalism and empiricism, which arise from mathematics and logic, and the natural sciences, respectively. In this sense, knowledge can encompass that which we can logically discern, and knowledge can encompass that which we can observe. Knowledge organization is the science of the order of knowledge. The domain of knowledge organization is a discourse community in which rigorous, self-conscious inquiry takes place concerning that which is known, and its various orderings or sequences, both those that are natural or heuristic, and those that are imposed. The products of the domain, then, are ordered segments of that which is known, and the rules either for discovering their natural orders, or the rules for imposing a useful sequence. 1 This paper is based on a keynote address to the Second Brazilian Conference on Knowledge Organization and Representation, Rio de Janeiro, Getulio Vargas Foundation FGV, imay 27, 28 and 29, 2013 IRIS Recife v.2 n.1 p. 2-11 jan./jun. 2013
3 All applied knowledge organization is a form of discourse, in which the structures and rules are objects of communication, and which takes place in a cultural milieu or among actors in various cultural milieus. A very important component of the science of knowledge organization then must be epistemology, which is the science of knowledge itself. In the domain of knowledge organization we have come to rely on a framework set forth by Birger Hjørland, who lists four basic epistemological stances (or positions) (1998, 608): Empiricism: derived from observation, perception, and experience; Rationalism: derived from the employment of reason over sensory experience; Historicism: derived from cultural hermeneutics; and, Pragmatism: derived from the consideration of goals and their consequences. That which we know from our own experience of it, and in particular that which is known through the positivist sciences, is what we call empirical. We have solid evidence for the empirical, and we can point to the evidence as a means of prediction. That which we know from reasoning about it, and in particular that which is known through humanistic scholarship, is based in rationalism. There is no evidence, per se, for the rational; rather there are explanatory statements that seem to be logical when taken together. Historicist epistemology interprets evidence through a cultural lens, relying in particular on past experience. Pragmatism is knowledge derived from assumptions about the best means to an end. Pragmatic solutions work in the moment but do not necessarily rely on empirical evidence, and therefore do not necessarily pass the test of time. Rational solutions also often ignore empirical evidence and thus frequently yield unworkable schemas. Epistemology is an essential tool of knowledge organization and a dimension is an expression of the extent of a space. Therefore, in knowledge organization, epistemology represents one dimension, because it is how we can measure or express the space within our domain ranging from the empirical to the rational, the two primary epistemological stances. Each dimension both bisects the space and provides a trajectory to transit through or across it. In other words, the dimensions of our domain define what we call its intension (its depth) and its extension (its external parameters)(these axes first articulated by Tennis 2003). An hypothesis is that in knowledge organization there always is tension along the epistemological dimension stretching from the empirical to the rational. This is demonstrated in the modes of thought used by knowledge organization researchers. In thinking of knowledge organization as a scientific domain, there is a fundamental theoretical dimension represented by our chief research questions. Our main concern is with the identification of concepts that represent the totality of that which is known. A second critical concern is with the order of these concepts. Therefore our main research questions always must be some form of what is knowledge? and what is its order? This paper brings together some results from meta-analysis of the knowledge organization domain. These analyses demonstrate the evolution of the domain from its original goal of creating a universally-applicable knowledge organization system, to its current post-modern embrace of multiple goals. Among the multiplicity of goals we find an embrace of domain-specific ontologies that can be combined as necessary using faceted techniques, leading to the perception of multiple, but representable, universes of knowledge. In knowledge organization there always is work to solidify concept theory, which is at the core of the discipline. Knowledge organization as a domain demonstrates coherence across time and across geo-political boundaries, particularly as it concerns its theoretical foundations. There are consistently marked dimensions within the domain theoretical versus applied on one continuum, humanistic versus scientific on another, which is another way of IRIS Recife v.2 n.1 p. 3-11 jan./jun. 2013
4 stating the epistemological dimension. All of these dimensions serve to maintain constructive and dynamic tension within the domain, which in turn keeps the research front constantly in a state of renewal, constantly stretching the intension and the extension (the boundaries) of the domain. 2. Visualizations from meta-analysis of knowledge organization as a DOMAIN Bibliometric tools and meta-analysis can be used to visualize the extension and intension of KO as a domain. Analysis of the domain of knowledge organization is relatively easily generated from the formal publications in biennial ISKO conference proceedings, articles in our journal Knowledge Organization, and proceedings of biennial regional chapters, whose meeting years alternate with the ISKO international conferences. Knowledge organization is devoted to the conceptual order of knowledge. In the broadest sense, knowledge organization is the arena in which the heuristics of ordering knowledge are studied. Specifically, knowledge organization is the research community devoted to classification and ontology, thesauri and controlled vocabulary, epistemology and warrant, as well as the development of applied systems for all of the preceding (often, especially in North America, resource description is also considered to be a part of knowledge organization). There is a long tradition of the activities and tools of knowledge organization; classification, taxonomy, and typology, for example, have always been key to the development of scholarship. Specific applications for information storage and retrieval, such as indexes, bibliographic classifications, etc., have been part of the practice of librarianship and the research agenda of information since the late 19th century. The formal knowledge organization domain, represented by the International Society for Knowledge Organization (ISKO) and its chapters, councils and publications, dates from 1989 (Dahlberg, 2010). The classical geopolitics of KO as a domain traditionally has consisted of anchors in six countries. Ingetraut Dahlberg, founder of the modern form of the domain, is from Germany. Much of the scholarship in the first decades of the domain came from North America and the UK. Essential theoretical grounding found its base in the Royal School in Copenhagen, and the beginnings of epistemological and ontological thought in the domain came from Italy. New-ISKO chapters holding regional conferences or universities generating research are shifting the emphasis away from northern hemispheric dominance to a more global perspective. The countries of affiliation of authors at the 2012 Mysore ISKO International Conference are shown in Figure 1. The figure shows how the geopolitics of knowledge organization have been shifting. In 2012 Brazil accounted for almost a third of the research. It is not clear whether there are geopolitical influences in the domain, but it is logical to assume that even if authors are working on the same set of problems are they doing so from different cultural perspectives. IRIS Recife v.2 n.1 p. 4-11 jan./jun. 2013
5 Figure 1. Countries of affiliation ISKO Mysore 2012. A meta-analysis of knowledge organization based on 17 studies of knowledge organization literature incorporated 3494 source papers, of which 1100 appeared in journals such as Cataloging & Classification Quarterly, Library Resources & Technical Services, or Library Quarterly, 444 appeared in Knowledge Organization, but 600 appeared in ISKO conference proceedings, and 1350 were papers in ISKO regional conference proceedings (the results of the meta-analysis are gathered in Smiraglia 2012b.) The distribution of knowledge organization research is visualized in Figure 2. The figure makes it clear how important conference proceedings are as a venue for communication in the domain. Fifty-six percent of the research over time has been presented in conference papers. Only 13% of the total over time has appeared in the domain s formal journal. The extension of the domain is consistently represented as including theoretical foundations, such as classification and ontology, and epistemology, which lies at the heart of both. The intension is represented by development and testing of applications. The international nature of the domain means that there are occasional overlapping emphases on multicultural and multilingual issues. Geographical diversity does not preclude domain coherence. Thematic coherence is demonstrated, but granularity shifts over time. Figure 3 is a visualization of themes in the papers in the 2012 Mysore ISKO conference. Figure 4 shows those themes distributed by countries of affiliation. Figure 5 shows the distribution of methodologies in the 2012 Mysore ISKO conference. IRIS Recife v.2 n.1 p. 5-11 jan./jun. 2013
6 Figure 2. The distribution of knowledge organization research across venues. Figure 3. Distribution of themes in 2012 Mysore ISKO conference. IRIS Recife v.2 n.1 p. 6-11 jan./jun. 2013
7 Figure 4. Distribution of 2012 themes by country of affiliation. Figure 5. Distribution of methodolgies in 2012 Mysore ISKO conference. IRIS Recife v.2 n.1 p. 7-11 jan./jun. 2013
8 Figure 3 shows both the extension (classification, ontology, interoperability) and the intension of the domain in the diversity of themes brought forward to the international conference. Figure 4 shows that there is little geographical influence among the themes. Rather, in most places most of the themes are in play. Figure 5 shows the epistemological dimension of knowledge organization, with roughly half of the contributions (the upper half) representing empirical research and the other half representing rationalist perspectives. Another approach to visualizing the domain is citation analysis and its close relative author co-citation analysis. In Smiraglia 2012b, a series of author co-citation maps is presented to show how, although the intension keeps shifting, the extension of the domain remains concept theory (or semantics) and applied systems (the search for interoperability). Figure 6 is a visualization of interconference author co-citation from the 2012 ISKO conference in Mysore. Figure 6. Inter-conference author co-citation from ISKO 2012 Mysore. IRIS Recife v.2 n.1 p. 8-11 jan./jun. 2013
9 I use this method to visualize coherence within the domain; it shows us how much co-citation there among the authors who contributed to the conference. Notice that few authors are co-cited, that they are among the most cited, and that they represent, for the most part, the two traditional dimensions of knowledge organization semantics, or concept theory (represented in the upper cluster) and knowledge organization systems and interoperability (in the lower cluster). Notice the lone cluster around Ranganathan, representing the resurgence of his influence on faceted systems. His position near the dense cluster of currently cited and co-cited authors is a visualization of how his ideas are leading the research front. Figure 7 is a network map of author co-citation among the authors of the 2012 ISKO conference in Mysore created using Gephi 0.8.2. Figure 7. Network map of author co-citation from ISKO Mysore 2012. IRIS Recife v.2 n.1 p. 9-11 jan./jun. 2013
10 Again we see Ranganathan apart from but leading the pack, as it were, influencing the research front. In this visualization we see better the interconnections among the co-cited authors. The network of connections shows, as did the two-dimensional representation in figure 6, that the domain is coherent and well connected theoretically in its extension and in its intension. 3 Coherence along an epistemological trajectory My recent definition of a domain (Smiraglia 2012a, 114) is: A domain is best understood as a unit of analysis for the construction of a KOS. That is, a domain is a group with an ontological base that reveals an underlying teleology, a set of common hypotheses, epistemological consensus on methodological approaches, and social semantics. Knowledge organization clearly is a coherent domain with an ontological base, its extension, along a dimension stretching from concept theory (or semantics) to applied knowledge organization systems. There also is an underlying teleology, which is the attempt to grasp the order of knowledge. The domain has, therefore, a set of common hypotheses first, that knowledge is made up of concepts; and second, that these concepts can be ordered in diverse and useful ways. There is epistemological consensus forming a dimension stretching from empiricism to rationalism. Finally, as we have seen clearly here, knowledge organization has social semantics. That is, there are formal venues for research, and the knowledge published in them is clearly cumulative. There also is a geopolitical type of social semantics in which the domain functions scientifically globally, but at the same time engages cultural diversity. Finally we also can see that knowledge organization as a domain demonstrates coherence across time but there are consistently marked dimensions within the domain, such that a) there is a trajectory from theoretical to applied on a conceptual continuum; and b) there is another trajectory from empirical to rational on an epistemological continuum. These dimensions not only bisect the intellectual space, they also serve to maintain constructive and dynamic tension within the domain, which in turn keeps the research front constantly in a state of renewal. Faced with different universes of knowledge, KO as a domain approaches the problems of analysis and concept designation from within the dynamic tension demonstrated here. IRIS Recife v.2 n.1 p. 10-11 jan./jun. 2013
11 REFERENCES Hjørland, Birger. 1998. Theory and metatheory of information science: a new interpretation. Journal of documentation 54: 606-21. Smiraglia, Richard P. 2012a. Epistemology of domain analysis. In Smiraglia, Richard P. and Hur-Li Lee, eds. Cultual frames of knowledge. Würzburg: Ergon-Verlag, p. 114. Smiraglia, Richard P. 2012b. Universes, dimensions, domains, intensions and extensions: knowledge organization for the 21st Century. In A. Neelameghan and K.S. Raghavan eds. Categories, contexts, and relations in knowledge organization: Proceedings of the Twelfth International ISKO Conference, 6-9 August 2012, Mysore, India. Advances in knowledge organization 13. Würzburg: Ergon Verlag, pp. 1-7. Tennis, Joseph T. 2003. Two axes of domains for domain analysis. Knowledge organization 30: 191-95. IRIS Recife v.2 n.1 p. 11-11 jan./jun. 2013