KO KNOWLEDGE ORGANIZATION Contents

Knowl. Org. 40(2013)No.2

KO KNOWLEDGE ORGANIZATION

Official Bi-Monthly Journal of the International Society for Knowledge Organization ISSN 0943 – 7444 International Journal devoted to Concept Theory, Classification, Indexing and Knowledge Representation

Contents

Articles ISKO News

Martin G. Channon. Conference ISKO-France, 10-1 October 2013, Paris. The Unification of Concept Representations: Contexts, languages and cultures in knowledge An Impetus for Scientific Epistemology...... 83 organization...... 147

Daniel Martínez-Ávila and Rosa San Segundo. Publication of ISO 25964-2, the standard for Reader-Interest Classification: Concept and interoperability of thesauri...... 147 Terminology Historical Overview...... 102 Book Review Carlos Henrique Marcondes. Knowledge Organization and Representation in The Disorder of Things: Metaphysical Foundations of the Digital Environments: Relations Between Ontology Disunity of Science by John Dupré. Massachusetts; and Knowledge Organization...... 115 London: Harvard University Press, 1993, 308p. ISBN0-674-21261-4 (Hb); and Human Nature and Teija Oikarinen and Terttu Kortelainen. the Limits of Science by John Dupré. Oxford; Challenges of Diversity, Consistency, and Globality New York: Oxford University Press, 2001, 201p. in Indexing of Local Archeological Artifacts ...... 123 ISBN 0-19-926550-X (Pb)...... 149

Urszula Sienkiewicz and Izabela Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs in the Area of Technical and Engineering Sciences...... 136

Knowl. Org. 40(2013)No.2

KNOWLEDGE ORGANIZATION KO

KNOWLEDGE ORGANIZATION Rebecca GREEN, Assistant Editor, Dewey Decimal Classification, Dewey Editorial Office, Library of Congress, Decimal Classification This journal is the organ of the INTERNATIONAL SOCIETY FOR KNOWLEDGE ORGANIZATION (General Secretariat: Vivien Division , 101 Independence Ave., S.E., Washington, DC 20540-4330, USA. E-mail: [email protected] PETRAS, Humboldt-Universität zu Berlin, Institut für Bibliotheks- und Informationswissenschaft, Unter den Linden 6, 10099 Berlin, Germany. José Augusto Chaves GUIMARÃES, Departamento de Ciência da In- E-mail: [email protected]. fromação, Universidade Estadual Paulista–UNESP, Av. Hygino Muzzi Filho 737, 17525-900 Marília SP Brazil. Editors E-mail: [email protected] Richard P. SMIRAGLIA (Editor-in-Chief), School of Information Stud- Birger HJØRLAND, Royal School of Library and Information Science, ies, University of Wisconsin, Milwaukee, Northwest Quad Building B, Copenhagen Denmark. E-mail: [email protected] 2025 E Newport St., Milwaukee, WI 53211 USA. Barbara H. KWASNIK, School of Information Studies, Syracuse Uni- E-mail: [email protected] versity, Syracuse, NY 13244 USA. E-mail: [email protected] Joseph T. TENNIS (Book Review Editor), The Information School of María J. LÓPEZ-HUERTAS. Universidad de Granada, Facultad de Bib- the University of Washington, Box 352840, Mary Gates Hall Ste 370, lioteconomía y Documentación, Campus Universitario de Cartuja, Bib- Seattle WA 98195-2840 USA. lioteca del Colegio Máximo de Cartuja, 18071 Granada, Spain. E-mail: E-mail: [email protected] [email protected] Nancy WILLIAMSON (Classification Research News Editor), Faculty Kathryn LA BARRE, The Graduate School of Library and Information of Information Studies, University of Toronto, 140 St. George Street, Science, University of Illinois at Urbana-Champaign, 501 E. Daniel Street, Toronto, Ontario M5S 3G6 Canada. MC-493, Champaign, IL 61820-6211 USA. E-mail: [email protected] E-mail: [email protected] David J. BLOOM (Editorial Assistant), School of Information Studies, Marianne LYKKE, e-Learning Lab, Center for User-driven Innovation, University of Wisconsin, Milwaukee, Northwest Quad Building B, 2025 Learning and Design, Department of Communication, Aalborg Univer- E Newport St., Milwaukee, WI 53211 USA. sity, Kroghstraede 1, room 2.023 Denmark 9220 Aalborg OE. E-mail: Melodie Joy FOX (Editorial Assistant), School of Information Studies, [email protected] University of Wisconsin, Milwaukee, Northwest Quad Building B, 2025 Ia MCILWAINE (Literature Editor), Research Fellow. School of Li- E Newport St., Milwaukee, WI 53211 USA. brary, Archive & Information Studies, University College London, Daniel Martínez Ávila (Editorial Assistant), Department of Library and Gower Street, London WC1E 6BT U.K. Information Science, University Carlos III of Madrid, C/Madrid 126 E-mail: [email protected] 28903 Getafe – Madrid, Spain. Jens-Erik MAI, Royal School of Library and Information Science, Co- E-mail: [email protected] penhagen Denmark. E-mail: [email protected] Editors Emerita Widad MUSTAFA el HADI, Université Charles de Gaulle Lille 3, URF IDIST, Domaine du Pont de Bois, Villeneuve d’Ascq 59653, France. Hope A. OLSON, School of Information Studies, University of Wis- E-mail: [email protected] consin-Milwaukee, Milwaukee, Northwest Quad Building B, 2025 E Newport St., Milwaukee, WI 53211 USA. H. Peter OHLY, Prinzenstr. 179, D-53175 Bonn, Germany. E-mail: [email protected] E-mail: [email protected] Clare BEGHTOL, Faculty of Information Studies, University of To- K. S. RAGHAVAN, DRTC, Indian Statistical Institute, Bangalore 560 ronto, 140 St. George Street, Toronto, Ontario M5S 3G6, Canada. 059, India. E-mail: [email protected] E-mail: [email protected] M. P. SATIJA, Guru Nanak Dev University, School of Library and In- Ingetraut DAHLBERG, Am Hirtenberg 13, 64732 Bad Konig,̈ Ger- formation Science, Amritsar-143 005, India. many. E-mail: [email protected] E-mail: [email protected] Aida SLAVIC, UDC Consortium, PO Box 90407, 2509 LK The Hague, Editorial Board The Netherlands. E-mail: [email protected] Jonathan FURNER, Graduate School of Education & Information Dagobert SOERGEL, Department of Library and Information Studies, Studies, University of California, Los Angeles, 300 Young Dr. N, Mail- Graduate School of Education, University at Buffalo, 534 Baldy Hall, box 951520, Los Angeles, CA 90095-1520, USA. Buffalo, NY 14260-1020. E-mail: [email protected] E-mail: [email protected] Jesús GASCÓN GARCÍA, Facultat de Biblioteconomia i Docu- Renato R. SOUZA, Applied Mathematics School, Getulio Vargas mentació, Universitat de Barcelona, C. Melcior de Palau, 140, 08014 Foundation, Praia de Botafogo, 190, 3o andar, Rio de Janeiro, RJ, Barcelona, Spain. E-mail: [email protected] 22250-900, Brazil. E-mail: [email protected] Claudio GNOLI, University of Pavia, Mathematics Department Li- Maja ŽUMER, Faculty of Arts, University of Ljubljana, Askerceva 2, brary, via Ferrata 1, I-27100 Pavia, Italy. E-mail: [email protected] Ljubljana 1000 Slovenia. E-mail: [email protected] Knowl. Org. 40(2013)No.2 83 M. G. Channon. The Unification of Concept Representations

The Unification of Concept Representations: An Impetus for Scientific Epistemology*†

Martin G. Channon

Project Cosmology, 2441 Guard Hill Rd., Middletown, VA 22645 USA,

The author is a real estate investor and small businessman in Maryland (USA). He has been developing an online effort to produce a (patent pending) unification of scientific schematics, Project Cosmology. He is largely self-taught in science and philosophy. Other publications include a recent paper in Physics Essays v. 24 (2011): 117-23, “The Falsification of an Accidental Universe” and “The Stowe Table as the Definitive Periodic System” in Knowledge Organization v. 38 (2011): 321-27.

Channon, Martin G. The Unification of Concept Representations: An Impetus for Scientific Episte- mology. Knowledge Organization. 40(2), 83-101. 80 references.

ABSTRACT: For virtually every major category of phenomena, science provides some standard schematic (e.g., the cross-section of the earth). The most notable exception concerns the cosmos as a whole. Project Cosmology (www.projectcosmology.net) is devoted to the presentation of such an holistic schematic. This is to be achieved by plotting the standard schematics for constituent phenomena within a three-dimensional coordinate system, time on the vertical axis and space on the other two. This produces a unification of schematics. As is discussed, this approach has the effect of allowing, more generally, an interactive unification of all graphical concept representations (schematics, graphs, formulae, tables, etc.). The result is a 3D, scientific, graphical user interface (GUI), one that is intended to map all knowledge. It can be characterized as a graphics approach to knowledge organization. It will be for scientific concepts what the Human Genome Project is to human DNA. The project is having the effect of revealing unnoticed gaps in knowledge, inconsistencies among the different sciences and apparent regularities throughout and across the various disciplines. Any such regularities would be laws relating to laws (i.e., laws relating to knowledge). The project, then, may facilitate the development of scientific epistemology (something already in process). This unification of concept representations is based on a cosmological perspective that provides a one-to-one correspondence between major entity and aspect classifications.

* Terms in brackets, e.g., [The Atom (LF)  H (RF)], indicate link paths at the website. “LF” refers to the left HTML frame, and “RF” refers to the larger of the right frames. (In this example, “H” refers to hydrogen in the Periodic Table.)

† I am indebted to Kevin Boyack (mapofscience.com) for having alerted me to the concept of “mapping knowledge.” He also provided a helpful review of an early draft of this paper. Sergey Bederov (ParallelGraphics, Moscow) is largely responsible for the graphics presented at the web site. I have also had assistance from various academic consultants, Wendell Hill and Alan Migdall (University of Maryland), John Harshman (Field Museum of Natural History) and Andrew Hopkins (University of Sydney). Robert Cochran provided a great deal of assistance with computer-related technical issues. I also thank the anonymous referees for their insightful and construc- tive comments; these have had the effect of improving the project and this discussion.

Received 29-11-2012; Revised 13-1-2013; Accepted 17-1-2013

1.0 Introduction as a whole. This discussion, then, is concerned with the description of a proposed schematic to fill this role. This In addition to the common graph, science makes much use is to be produced by plotting the standard schematics for of schematics, and for each major phenomenon, there is constituent phenomena along a timeline (part of a 3D typically some diagram (e.g., the cross-section of the coordinate system) in the order in which the phenomena earth, the phylogenetic tree, the embedding diagram for have typically developed. the gravitational field). There is, however, one critical ex- In this context, the term “cosmos” is being given a very ception to this rule; we have no schematic for the cosmos broad interpretation; as would be indicated by the range 84 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations

Figure 1. Simulated cross-section of the Hubble Sphere. of scientific disciplines, it includes time, space, particle Note also that something such as a concept, a lan- phenomena (e.g., atoms, planets), life, and civilization. As- guage, or a machine would be a feature of civilization, as trobiologists generally believe that life and civilizations are religion and art. Furthermore, mind in its most devel- are common in the universe; we have no evidence at this oped state is, as far as we know, exclusively a feature of time, but the relevant science supports this suspicion civilization. From this perspective, then, there are three (e.g., Gilmour and Sephton 2004). Life and civilization primary divisions to knowledge: the physical, life and civil are now widely viewed as critical considerations for cos- disciplines. Note that these correspond precisely to the mology (e.g., Harrison 2000; Davies 2004); this holistic primary categories commonly used in systems theory sense of the term is increasingly the accepted interpreta- (e.g., Bertalanffy 1968; Laszlo 1996). There are, of tion. This view has perhaps been best articulated by the course, alternate schemes for such metadisciplines, but editors of the online Journal of Cosmology, according to those used here appear to be the emerging de facto choices whom “Cosmology … is the study and understanding of of both cosmology and systems theory. Also, from this existence in its totality, encompassing the infinite and perspective, there are three primary phenomena or sys- eternal, and the origins and evolution of the cosmos, gal- tems: the Metacluster (expanding aggregate of galaxy axies, stars, planets, earth, life, woman and man” clusters), the biosphere, and civilization. Note the one-to- (http://journalofcosmology.com/About.html). Cosmol- one correspondence between disciplines and phenomena ogy, in this sense is the all-embracing science; all other (aspect and entity classifications). disciplines are subordinate. It is important to note that Notice also that something such as the cross-section this list (time, space…) is comprehensive; it includes all of the Hubble sphere, Figure 1, an alternative schematic known phenomena. (Our lack of evidence in support of for the cosmos, produces a relatively uninformative re- the apparent consensus among astrobiologists might well sult; it would only tell us about the distribution of galaxy be the simple result of our present inability to inspect clusters; it says virtually nothing about other phenomena. exoplanets, except with the crudest of techniques; cur- Each dot is a galaxy cluster (collection of galaxies). This rently, we detect them primarily by their gravitational ef- represents the Hubble Sphere for the typical observer fects on host stars.) over cosmological time. For the present epoch only, the Knowl. Org. 40(2013)No.2 85 M. G. Channon. The Unification of Concept Representations

Figure 2. A unified schematic. axis would extend to a lookback time of 1.37 1010 yr tational field. Above this, we plot the standard, disc-like, 10 (1.37 10 light years in distance). Notice that as a sche- cross-sectional schematics for particle phenomena (e.g., matic for the cosmos, this tells us only about the distribu- the cross-section of the Earth). In the upper section, we tion of galaxy clusters. It indicates nothing as to other plot a 3D version of the phylogenetic tree or cladogram, phenomena. Figure 1 would be a schematic primarily of and within that, a suggested schematic for civilization. relevance to physical cosmology, one aspect of the more These phenomena are accordingly labeled along the right, general subject area. providing a comprehensive list of all known phenomena Contrast this to Figure 2, the proposed schematic for (an all-inclusive entity classification). Along the left, we the cosmos, something that can apparently accommodate label the corresponding disciplines (providing a one-to- any level of detail relating to all phenomena. This Unified one aspect classification). At the web site, on-click events Schematic results from plotting the standard schematics for the list of phenomena load classification tables to a for natural phenomena in a 3D coordinate system, time second (right side) frame. Individual schematics, e.g., for on the vertical axis and space on the other two. Positions the hydrogen atom, are called out using links in the classi- along the vertical axis correspond to typical formation fication tables. On-click events for the list of correspond- times over the course of cosmological evolution. The ing disciplines will eventually load 3D concept maps to frame-like structure is the coordinate system, the stan- the right frame. These will lead to summations of con- dard schematic for space and time. (Dimensional units, cepts. Additionally, several standard tables and graphs are meters and years, are mouse-over effects at the web site; found to fit well as panels to the coordinate system, e.g., Figure 2 is a screen shot.) The lowest schematic, the em- the geologic time scale (left side panel) [Time log (yr since) bedding diagram, is the standard schematic for the gravi- (LF)]. The system can apparently accommodate any level 86 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations of detail, encompassing the entire body of science con- leading scale (e.g., the typical schematics for the solar sys- cepts. tem). Further, they eliminate the reliance on graphs There are, however, several alternative ways of charac- drawn only to a crude level of detail, neglecting critical terizing the project. One of these would be to say that it at- microscopic aspects (e.g., the fine structure for atomic tempts to present a unification of the various standard energy levels). They also allow us to get past outdated schematics used throughout the sciences. However, this graphics, e.g., the so-called modern Periodic Table. This is approach has proven amenable to the inclusion of all other little changed from the one developed by Mendeleev graphical concept representations. Thus the unification prior to the advent of modern atomic theory. It is greatly would relate to all of these. Notice that the unification of inferior to the Stowe Periodic Table, Figure 3, something concepts has been one of the broadest features of scientific progress; in this case, we are extending the effort to the manner in which concepts are organized and presented. Furthermore, this approach is apparently condu- cive to the systematic, graphical presentation of all concepts from the various special disciplines (even philosophy and the humanities). Thus another characterization would be that this effort concerns an attempt to map all knowledge. In any case, the result is a scientific, GUI, one devoted to the access of knowledge. It constitutes a graphics approach to knowledge organization. It is the author’s con- tention that this approach will be to the expository alternative (i.e., explanation in words) what the pervasive GUI has been to the command-line interface (CLI) for computer use. By way of analogy, the effort is the epistemic equivalent of the genome project. The justification for the project has to do, first, with the fact that schematics facilitate understanding. Likewise, they facilitate retention and review; the human mind works most effectively with images (hence the supremacy of the GUI in relation to the CLI). Furthermore, this project tends to highlight gaps in knowledge. It has revealed, for example, that there is no existing schematic for civilization, a clear indication that we do not have a rigorous conceptualization of this phenomenon. A comparable situation would exist if we had no schematic cross-section for the earth; this could only mean that we had a poor understanding of its geological structure. Since the effort to map any terrain invariably leads to its better understanding, the effort to map all knowledge would surely have similar results. Figure 3. The Stowe Periodic Table. As is further discussed below, the effort is, indeed, having the apparent effect of identifying unnoticed patterns that is based on the fundamental parameters of quantum throughout and across the various disciplines. This sug- mechanics (Channon 2011). If this project does nothing gests the possibility of metalaws for science, a scientific else, then, it dramatically illustrates that the future of sci- epistemology. (Over time, the program will be developed entific illustration is three-dimensional, interactive graph- to include analytic capacities, e.g., horizontal slices through ics; the 2D, printed schematic is to scientific visualization the cladogram for purposes of identifying biosphere struc- what the slide rule is to mathematics. The parameters are ture at a given historical time). the three quantum numbers, n (shell), s (spin), and m (ori- In this connection, we might briefly note the advan- entation). These are the fundamental determinants of tages of 3D, interactive graphics in science. First, they eli- atomic structure and properties. Notice the perfect, over- minate the reliance on 2D, static diagrams for phenom- all symmetry. All classes, groups, and “blocks” likewise fall ena that are, in fact, three-dimensional and in motion (2D into perfect rings, columns or levels, providing, for the representations of 4D phenomena). Likewise, they elimi- first time, precise, quantitative meanings for “period,” nate the need for schematics drawn to a false and mis- “group,” and “block.” The 3D, interactive version (www. Knowl. Org. 40(2013)No.2 87 M. G. Channon. The Unification of Concept Representations projectcosmology.net) provides simple, on-screen controls approach, then, would be one that somehow incorporates for manipulation and isolation of rings, columns and lev- a reference to the standard schematics for constituent phe- els. Each individual symbol is a link to extensive data. In nomena. This, of course, requires an identification of the time, each symbol will also provide access to energy level phenomena themselves, and that is somewhat contentious. diagrams, 3D, interactive orbital schematics, and lists of The categories used here should be viewed as a first ap- formulae for quantum mechanics. Due to budgetary con- proximation; they cannot be considered definitive. These straints, only hydrogen is fully developed at this time. phenomena are listed both in Figure 1 and Table 1. The However, this sufficiently illustrates that the system can derivations of this list are discussed below. There are sev- potentially accommodate all concepts related to a phe- eral apparent patterns. First, these phenomena all appear to nomenon and in a graphical format. be systems. Second, they tend to form hierarchies, as is of- This is a project that must cover every intellectual dis- ten noted in systems theory. Third, there are some ten cipline, every topic and every concept. Furthermore, these phenomena for each metacategory (physical, life, and civil), must be graphically rendered in 3D, interactive format, with an elemental unit in first position and an all inclusive something that often requires creativity and much trial phenomenon in final position. and error. Properly developed for existing knowledge, the Some definitions are required. In the Unified Schematic, project would require a very extensive collaborative effort galaxy groups and clusters are referred to as “Galaxy Sys- extended over at least several years. It would also require tems.” Likewise, “Ellipsoids” refers to stars, planets, and a significant budget, at least several million dollars. In- planetary satellites. These are grouped together, since they deed, this is a job that will never truly be completed (un- are essentially similar phenomena; a star seems so different less intellectual activity as whole is somehow finished). only because it is, so to speak, a planet so massive that However, any degree of success would be worthwhile. gravitational pressure ignites thermonuclear reactions. No- This project is, then, an extremely ambitious endeavor, tice also that other categories of phenomena have similarly far beyond the budget and abilities of the few individuals striking differences. Some atoms, for example, are highly currently working on it (primarily the author). Due to the stable, while others emit radiation. Also, the typical planet inherent enormity of the task, the most that can be pro- is, like a star, a radiative “gas giant.” “Ellipsoid systems” is vided at this time is the illustration of a potential. This a term introduced here as a reference to planetary systems, project, then, is currently very much a work in progress. stellar clusters, binaries, and globular clusters. The term Some of the schematics are cartoon-like. Others are grie- “institutions” is used in the strictest, sociological, interpre- vously superficial. A careful review will undoubtedly find tation; it is a reference to government, industry, the numerous mistakes. Likewise, a fully adequate discussion schools, the family, and the church. of the project should touch on literally every intellectual The next issue concerns the manner in which to ar- discipline and would be, therefore very lengthy. It inevi- range the individual schematics. The most sensible ap- tably has numerous implications throughout science and proach seems to be that of placing them along a timeline. philosophy, perhaps most notably, epistemology and There is another, critical recommendation for this choice. knowledge organization. Properly developed, this discus- A complete schematic for the cosmos must include not sion also would require an extensive collaborative and re- only the material phenomena discussed above, but also an view effort. Hopefully, this will be presented at some explicit reference to time and space. (For the benefit of point in the near future. The present treatment has been those who are unfamiliar with physics, space and time are prepared by the author working alone. It is, in effect, little physical phenomena with dynamic properties. The gravi- more than a lengthy summary; it is inevitably imperfect. tational force, for example, is a consequence of the de- These comments are not intended to discourage criti- formation to space induced by mass.) The standard cisms, quite the contrary. They are simply to emphasize schematic for time is a scaled axis, and, by including this, that we cannot reasonably expect at this time anything we incorporate the appropriate schematic for time. In other than a “first approximation.” physics, however, this schematic is typically in the form of a space-time frame (a 2D, Cartesian system), and, in 2.0 Itemizing phenomena; the fully-formed criterion using this, rather than a single axis, we include the necessary schematic for (one-dimensional) space. Notice, how- The first step in constructing a schematic for the cosmos ever, that in placing something such as the schematic would be a consideration of the principles involved. There cross section for a planet at a position on the timeline may be various approaches, but the one developed here is (i.e., such that the timeline is a “surface normal”), we im- based on the point of view that the cosmos is more than plicitly evoke two spatial axes, since a cross section is stars, planets, and galaxies. A proper schematic must take two-dimensional. In adding two spatial axes, we produce, all phenomena into consideration. The seemingly obvious overall, a three-dimensional space-time frame. 88 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations

Years since Event time Event N from Error Years big bang (log form) T Event number Eq. 1 for Eq. 1 ago (truncated) T ln Tys  N ys T  (truncated) (truncated) ln 10

Civilization 30 0 10000000000 10 10 30.701 0.701 Religion 29 0 to 100 9999999950 9.999999998 9(9)8 29.424 0.424 Technology 28 135 9999999865 9.999999994 9(9)4* 28.600 0.6 Science 27 322 9999999678 9.999999986 9(8)86* 27.596 0.596 Law 26 383 9999999617 9.999999983 9(8)83* 27.361 1.361 Communities 25 427 9999999573 9.999999981 9(8)81* 27.208 2.208 Art 24 708 9999999292 9.999999969 9(8)7* 26.439 2.439 Language 23 4600 9999995400 9.999999800 9(7)8* 23.159 0.159 Institutions 22 5000 9999995000 9.999999783 9(7)78* 23.021 1.021 Mind 21 10-12,000 9999989000 9.999999522 9(7)5* 21.884 0.884 Biosphere 20 ? ? ? ? ? ? Ecosystems 19 ? ? ? ? ? ? Multicellularity 18 ? ? ? ? ? ? Organ systems 17 5.4 108 9460000000 9.975891136 9.976 16.777 -0.223 Organs 16 5.5 to 5.9 108 9430000000 9.974511693 9.975 16.634 0.634 Tissues1 15 7.5 108 9250000000 9.966141733 9.966 15.865 0.865 Populations 14 1.1 109 8900000000 9.949390007 9.949 14.692 0.692 Cells 13 2.7 109 7300000000 9.863322860 9.863 11.658 -1.342 Organelles2 12 3.465 109 6535000000 9.815245592 9.815 10.771 -1.229 Polymers 11 3.5 to 3.9 109 6300000000 9.799340549 9.799 10.534 -0.466 Metacluster 10 ? ? ? ? ? ? Galaxy Systems 9 ? ? ? ? ? ? Galaxies 8 4.6 109 5390664637 9.731642314 9.732 9.725 1.725 Ellipsoid systems 7 ? ? ? ? ? ? Ellipsoids3 6 9.5 109 98300355.15 7.992555087 7.993 5.318 -0.682 Molecules 5 ? ? ? ? The Atom 4 9.9 109 240994.4191 5.382006985 5.382 4.061 0.061 Atomic nuclei 3 ~1010 106 -6 -6 2.682 -0.318 Hadrons 2 ~1010 1014 -14 -14 2.299 0.299 Elem.particles 1 ~1010 1040 -40 -40 1.724 0.724 Grav. field 0 ~1010 0   0 0

Table 1. Events, event times and derivation of the scaling equation (1010 yr idealized interval).

* Innovation in mathematical notation (e.g.): 9(2)5 9.95, 9(3)7 9.997 . Note that 9(9)6 and 9(7)566 have the same number of significant figures. 1. The event time commonly attributed to multicellularity has been assigned to tissues, since the first multicellular organisms would have had about this level of development. 2. The event time commonly attributed to the first cells has been assigned to organelles, since the earliest “cells” would actually have had about this level of development. 3. The event time commonly attributed to galaxies has been assigned to ellipsoids, since the event time given for galaxies in astrophysics texts actually pertains to proto-galaxies and this corresponds to the first stars.

Knowl. Org. 40(2013)No.2 89 M. G. Channon. The Unification of Concept Representations

Note that the concept of “space” in science (or at least formed condition for phenomena is not always obvious. mathematics) has now become quite general and even This is particularly true for civil phenomena such as art. sometimes metaphorical. In the most general usage, it is This fully-formed criterion, then, needs to be supple- so-called, “coordinate space,” and a corresponding coor- mented with additional principals related to identifying dinate system (e.g., the typical Cartesian system) can have the fully-formed condition. any set (or system) of quantifiable parameters. These might be vectors, as in “vector space,” or even functions, 3.0 Cosmological event times as in “function space.” The typical coordinate space concerns physical distance. In the Unified Schematic, the 3.1 Physical science events diagrams in the lower half are to be read as in the physical space of the lower panel. In the upper half, diagrams Let us now consider these systems, establishing events (e.g., the cladogram) must be read as in a yet-to-be speci- times where possible. For physical science phenomena, fied coordinate space of the overhead panel. This specifi- event times are as cited in textbooks; the physics com- cation remains one of the major outstanding deficiencies munity has a passable consensus as to these dates. Arbi- for the project. trarily, the primary references are Zeilik and Gregory The assignment of event times is often problematic. (1997) and Harrison (2000). Early in the history of the This is particularly true of civil phenomena. Consider cosmos, there was a build up of structure from the sim- technology, for example. The earliest stone tools date to plest to the more complex (elementary particles then about 2.3 to 2.6 Ma. (Kibunjia 1994; Kimbel et al. 1996; hadrons, then atomic nuclei ... ). Furthermore, according Semaw et al. 1997; Wood 1997). However, our ancestors to contemporary theory, smaller astronomical structures undoubtedly used more easily worked materials prior to developed first; there is a step-by-step process in which that date, and a stick or a bone used for any purpose is a the various phenomena separate from a general nebula, form of technology. But these would not be specifically first ellipsoids, then ellipsoid systems, galaxies, galaxy human tools; animals are also known to use sticks and groups, galaxy clusters ... (e.g., Silk 1999). At least in gen- stones e.g., Boesch and Boesch 1984; Shuster and Sher- eral, then, physical science phenomena develop in a hier- man 1998). An even more extreme case is that of the archical fashion. (A qualification is required; heavy atoms wasp, Ammophila urnaria, known to use a pebble as a form after the first generation of stars. However, light at- hammer. This was first noticed by S. W. Williston (1892) oms form earlier.) and George and Elizabeth Peckham (1898) and has been The event times for phenomena have been listed in widely discussed since (e.g., Frisch 1940; Brockmann Columns 3 and 4 of Table 1. Column 5 transposes these 1985). There does not seem to be any clear threshold be- numbers into purely logarithmic form. Given that the to- tween animal and human technology. Therefore, to trace tal time interval is measured in billions of years, the short the origin of technology, we need to look back so far that periods between civil science events necessitates very fine we are no longer discussing the features of civilization, discriminations. However, as we look downward in the nor even our distant ancestors (unless we are to look back list, large numbers of significant figures become progres- as far as insects). sively more meaningless. Hence, truncated values are In order to assign event times methodically, then, we shown in column 6. need to develop some alternative principle. Event times will be assigned, then, according to when the phenomena 3.2 Life science events first appear as fully formed (all components in place). This is recommended here as a way to normalize an event As to the life- and civil-science events, the policy followed time. (This procedure would be comparable to using the here is that of the so-called “principle of mediocrity,” the width at half maximum to characterize a Gauss curve.) belief that earth and human civilization are typical exam- We can combine the above considerations, then, into ples of such phenomena. This notion has been justifiably what can be termed the “fully-formed Criterion:” criticized (Deutsch 2011; Kukla 2010), but it does appear to at least articulate the apparent, intuition-based consen- The standard schematics of phenomena will be arranged sus among astrobiologists. The reasoning behind it is partly within a 3D coordinate system according to fully-formed motivated by the well-established Copernican principle, ac- event times. cording to which, we cannot attribute to ourselves any special position in the universe. Further still, various biological The application of this principle is not always straight- principles would probably apply to life regardless of where forward. It is, however, a good deal more workable than it develops; natural selection, for example, seems likely to the alternative. In particular, what constitutes the fully- be universal, as does the operation of biogeochemical cy- 90 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations cles. These considerations are admittedly speculative, but have been organisms such as the cnidaria (jellyfish, sea this is about the best we can do for now. In any case, it is anemones, corals, hydra, etc.). The corresponding event intuitively more sensible to assume that we are not special, time is 550 to 590 Ma. The most primitive organ systems rather than otherwise. Furthermore, while we have no ex- seem likely to correspond to the bilaterians. The event perience with life elsewhere, we have had plenty of experi- time is 540 Ma (e.g., Knoll and Carroll 1999). “Multicellu- ence in analyzing a wide diversity of phenomena (atoms, larity” is defined here as referring to life forms consisting stars, etc.); our intuition is well informed. The opposing of multiple organ systems. It is not clear as to when this principle, “the rare earth hypothesis,” has also been subject first appeared; hence no event time is provided. The ful- to extensive criticism, most notably an inconsistency with ly-formed event times for ecosystems and the biosphere the constantly increasing number of observed exoplanets. are also unclear. (See, for example, Jean Schneider’s Exoplanet.eu, CNRS/ As was the case for physical science phenomena, the LUTH, Paris Observatory.) Another principle commonly life science phenomena largely developed in a hierarchical cited in this connection is Nick Bostrom’s (2002, 57) “self- fashion. Ecosystems and the biosphere are possible ex- sampling assumption,” according to which we should think ceptions to this rule. However, if these phenomena are of ourselves as random observers from a “suitable refer- fully-formed only after “Multicellularity” appears, then ence class.” Revisiting the Principal of Mediocrity, it is there would be a complete hierarchical pattern. worth mentioning, as noted by Guillermo Lemarchand (2006, 458) that “From a Lakatosian epistemological point 3.3 Civil science events of view, this hypothesis is within the ‘hard core’ of the research programs which main purpose is the search for life Some previous consideration has been given to the iden- in the universe (e.g., exobiology, bioastronomy, astrobiol- tification of the major categories of civil phenomena, ogy, SETI).” Imre Lakatos (1970) has argued that the ‘hard most notably Durant (1939), Schrecker (1948), Childe core’ of a research program includes any hypothesis that is (1983) and Quigley (1961). However, while there is much widely viewed by the experts as valid, despite the possible overlap in the recommended categories, their writings lack of real supporting evidence. On the basis of these seem not to have inspired any consensus. Also, each of various considerations, we can cautiously assume that life these authors seems to have taken something less than a and civilization will develop on any Earth-like planet very methodical approach in determining the categories. In much as they have here. order to provide such a methodical determination, a sur- Unless otherwise indicated, the event times cited for life- vey article of some form would be very helpful. Fortu- science phenomena are taken from Schopf (1999). This nately, such a survey (of sorts) exists, the Propaedia for the book is a popularization, but it is effectively a compre- Encyclopedia Britannica (Goetz 1987). This is a 744-page, hensive survey of science related to the origin of life formal outline for the Britannica and, as such, it can be (with some consideration of evolution as a whole). used to quickly narrow in on portions relating to civil Schopf is a stellar figure in his field; his (popularized) phenomena. The Propaedia has been examined in detail, “survey” would be widely respected. resulting in the following list of civil phenomena: mind, Life seems to have first appeared 3,500 to 3,900 Ma institutions, language, art, communities, law, knowledge, (million years ago). This event time is attributed to poly- technology and religion. These phenomena are also in- mers, since these would have been the first distinctly bio- cluded in Table 1. logical phenomena. The first cell-like structures, “proto- As mentioned, establishing event times for civil sci- cells,” seem to have appeared at about 3,465 Ma. These ence systems is especially problematic. What follows, would have had about the developmental level of an or- then, is at best, a first approximation. In as far as possi- ganelle; hence the event time for organelles [Organelles ble, the discussion will attempt to establish event times (LF)]. The fully developed cell is the eukaryote. Brocks et based on the fully-formed criterion. Otherwise, a more al. (1999) have found the earliest evidence of eukaryotes intuitive approach is taken, one that focuses on the event thus far, 2700 Ma. Populations are associations of inter- times for paradigmatic examples. breeding organisms. These would have first developed Mind consists of two primary domains, the conscious when eukaryotes developed meiosis (sexual cell division). and unconscious. At a finer level of detail, it is commonly This gives an event time of 1,100 Ma. The earliest multi- said to involve concepts (thoughts), sensations, emotion, cellular organisms appeared at about 750 Ma. This event memory, personality, the will, and self-awareness. These time is attributed to tissues, since the earliest multicellular are all interdependent; it is apparently another system. organisms would have had about this level of develop- Most of these aspects would seem to have been well- ment. The event time for organs would correspond to the developed during prehistoric times. The development of first organisms with multiple tissue systems. These would concepts, however, would have been largely mythological Knowl. Org. 40(2013)No.2 91 M. G. Channon. The Unification of Concept Representations and unsystematic up to the point of the Neolithic Revo- when all constituent institutions were separate and present lution. It was at this time that we first developed settled in modern form. Completing this analysis for each institu- communities and farms. This would have required, at mi- tion would involve much research, and is beyond the scope nimum, systemized principles of engineering and law of the present effort. However, intuitively, one suspects (even if passed along via oral tradition). Arguably, this that industry, academia, the family and church have been marks the point in time at which systemized concepts in present in basically modern form for some time. This general began to be developed. A plausible event time for would not be the case for government. A plausible event mind, then, would correspond to the Neolithic revolu- time for fully-formed government would correspond to tion, 8,000 to 10,000 B.C. the earliest point in time at which there was a full separa- Language consists most obviously of vocabulary and tion of powers. Principles relating to the separation of grammar, both of which would have had prehistoric ori- powers were most notably articulated by Montesquieu gins (e.g., Corballis 2003). However, fully-formed, it also (1748). However, these principles were already in place in involves writing. The first true writing system (to be distin- the Roman Republic, circa 500 BC (e.g., Fairlie 1923). It is guished from proto-writing) was Sumerian cuneiform. A sometimes said that ancient Athens had a separation of plausible event time for language would then be 2600 B.C. power, but it was poorly developed (Fairlie 1923). A plau- (e.g., Kramer 1988). sible event time for institutions, then, is 500 BC. Technology is usually spoken of as various tools (and In Western culture, art came of age with the Renais- techniques). However, this is seemingly an inadequate ana- sance, circa 1300 (e.g., Jenkins 1973). This is when art was lysis. The various forms of technology complement one first pursued as “art for art’s sake.” Oil painting, a paradig- another, working together (with operators) to achieve vari- matic form of art, became common in the 15th century ous results. It is a monolithic phenomenon, the various (e.g., Owen 1987). A plausible event time for art would parts of which we identify as tools and techniques for spe- then be 1300 to 1500. The development of art for art’s sake cific purposes. Again, it is difficult to say when this system may have occurred earlier in Asia, but an effort to explore was fully formed. One option would be to use the event this was unsuccessful. (Jenkins also says that the assertion time for the full set of “simple machines” (lever, inclined “art for art’s sake” is a more recent event, ~1800.) plane, wheel and axel, screw, pulley and wedge). The most Of the various kinds of communities (e.g., clans, tribes, recent of these was the screw, developed circa 700 B.C. villages, nations, religious denominations, professional as- (e.g., Dalley and Oleson 2003). This is nicely consistent sociations, etc.) the concept seems to apply paradigmati- with another approach, one based on a paradigmatic ex- cally to political affiliations, in particular, the nation-state. ample, in this case the compound machine (two or more The first of these seems to have been the Dutch Republic independent but synchronized mechanism). The first of of 1581 (e.g., Geyl 2001). these appears to have been a device for working gems, cir- Knowledge today is largely and paradigmatically science. ca 550 BC (Lu 2004). Tentatively, then, we will use an event However, science is not simply a set of principles. There is time of 700 to 550 BC. also a method and an institutional component. The most Law is said to be most essentially a system of rules. His- recently developed component seems to have been the me- torians of law appear to agree that law in its modern form thod. The Scientific Revolution is usually said to have be- did not come into being until Hugo Grotius published his gun in the early 1500s. However, it was a bit later, 1686, De Jure Belli ac Pacis (On the Law of War and Peace) in 1625 that Robert Boyle published A Free Enquiry into the Vulgarly (e.g., Golding 1967). This separated jurisprudence from Received Notion of Nature. Here we find the first articulation Canon law. This roughly coincides with the beginnings of of the importance of repeatable experiments and the pub- modern constitutional law ~1600, the paradigmatic exam- lished presentation of results. The event time for fully ple of law. The event time for law, then, is taken to be 1625 formed science, then, is taken to be 1686 (even if the rec- to 1700. ommendation was not fully implemented until later). Institutions do not exist independently of one another. Religion is not simply a belief system. There is also the The family is said to provide the supply of labor used in ministry, a community and the sanctuary. Religion would industry and other institutions (a decidedly utilitarian inter- be fully formed when all of these are in place. Histori- pretation). The government regulates the various institu- cally, religious belief systems have been presented as tions (among other things). The schools provide the train- knowledge, although over the last few hundred years, it ing used by the work force. These entities, then, form what has become increasingly clear that they are in conflict might be referred to as an institutional complex (system). with science. The correction, however, is well underway. Initially, at least some of these institutions were mixed. In At least in the West, organized religion now tacitly, if not Sumer, for example, the king was also the high priest (e.g., explicitly, presupposes the validity of science. The Cate- Brisch 2008). This complex would have been fully formed chism of the Catholic Church (United States Catholic Church 92 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations

1997, 159), for example, now teaches that “methodical unreadable manner. We need, then, some variation of this research in all branches of knowledge, provided it is car- logarithmic scale. Ideally, the events would be spread out ried out in a truly scientific manner ... can never conflict in a uniform fashion. with the faith.” The argument remains simply over Notice that each event in Table 1 is given an event whether or not particular scientific theories are correct. number, N, and for most of these, we have known event Belief systems, then, are slowly adjusting to science. This times. Events with known event times are plotted in Fig- process can have only one outcome: theology and science ure 4, where placement of event numbers evenly along will eventually be consistent (even if this means a drastic the vertical axis reflects the effort to spread out events alteration of religion). This acceptance of science in gen- uniformly along the timeline. It is this graph that we need eral, despite problems with specific theories, seems to to model in order to get the proper scaling equation. have begun its development over the twentieth century. Graphs of this type are crudely of the form y  a x , Hence the event time for religion is 1900-2000. It is per- where a is some constant slightly greater than 1. How- haps worth mentioning that religion is undeniably a ma- ever, in the present case, no such simple formula will jor feature of civilization; it cannot be ignored. It is also work. The primary reason for this is that the graph is ex- worth noting that religion can be meaningfully defined, tremely sensitive at the high end (a minimum of ten sig- without making reference to the supernatural and the an- nificant figures to distinguish between the upper two thropomorphic (James 1902; Durkheim 1915). The points); any exponential function that gives good results common supposition that it will be sloughed off is a con- along the lower portion of the graph is way off at the jecture based typically on a naïve understanding of the top, and vice versa. Also, there is a severe change in cur- phenomenon. (The quotation above from the Catholic vature in the vicinity of event twenty-one; even more so Catechism—that science cannot conflict with faith—is, of than around events four to six. (The change in direction course, an article of faith itself, but it also expresses a of the curve is more noticeable in the vicinity of points commitment to science on the part of the Church.) four to six, i.e., the radius of curvature, R  1 K is smal- Again, it should be noted that the above list of phe- ler. However, the actual curvature, i.e., the rate of change nomena might be incomplete, or otherwise in error, and in direction, Kdds  , is greater near point twenty-one.) the event order and event times are almost certainly sub- A more complicated fit is required. Neglecting the deriva- ject to correction. Also, there is the strong possibility of a tion (basically just a fit), this is cultural bias in determining event times for civil phenomena; some of what is attributed to Europe may have oc- 340.2809 NT8.047 10 8.029 10  curred earlier in Asia or the Middle East. Indeed, virtually (1) + 10.1783exp 2554.47 10 T 0.484597 0.7505 every step in the development of the list and the assign-  ment of event times is debatable. However, if we are to produce a unification of scientific schematics, we must The output of this equation and the errors for each event start with some first approximation. appear in columns seven and eight of Table 1. To get a corresponding expression in terms of years ago, we can T 4.0 The scaling equations substitute T ln 1010 10ya  ln10 in equation (1),

0.2809 T The first part of the Unified Schematic to develop must ln 1010  10 ya 34  N 8.047 10 8.029 10 be the timeline itself, and this is not quite straightforward. YA ln10  The problem concerns an appropriate scale. As a first  0.484597 T (2) thought, we might use a strictly proportional scale. In this ln 1010  10 ya   + 10.1783exp 2554.47 10  0.7505 case, the total time period is spread out evenly along the ln10  time line (about ten inches of length for some ten billion  years). This has the effect of bunching up the first five physical science events in an unreadable manner at the In the development of these equations, we have used the bottom of the timeline. It also bunches up all of the civil typical time period for the development of a civilization, science events in an unreadable manner at the top of the 1010 yr, rather than the period as we know it, 1.37 1010 yr. line. The use of a timeline becomes pointless. Because of (Recent studies consistently show the peak in the star 9 this effect, as it relates to physical science events, astro- formation rate at a lookback time of 7.7 10 yr (Hop- physics typically uses a logarithmic scale ( 50 to 10 kins and Beacom 2006; Mannucci et al. 2006; Ota et al. 50 10 corresponding to 10 to 10 years). This works for 2008; Verma et al. 2007; Yüksel et al. 2008; Bouwens et these events, but it causes the life and civil science events al. 2008). This corresponds to a time of 610 9 yr after the to bunch up at the top of the line, many in a completely moment, t  0. To this we add on the commonly cited Knowl. Org. 40(2013)No.2 93 M. G. Channon. The Unification of Concept Representations

Figure 4. A unified schematic. time for the formation of the solar system and the evolu- 5.1 Frame and panels tion of life, 4.5 109 yr (e.g., Tilton 1988) to get the total typical time interval.) The use of these equations, in lieu The coordinate system is built in the standard fashion; of something such as a logarithmic scale, is not entirely the angles between the line of sight and the x, y and z satisfactory; they are too complicated. The author is con- axes of the cube are  /2, 0,  /2 respectively (i.e., the li- tinuing to explore alternatives. ne of sight is normal to the front face of the system). Three additional parameters, “field of view” (effectively a 5.0 Building the unified schematic measure of the lens used to view the scene) the size of the cube (the length of an edge) and distance of the cube We are now prepared to start putting things together. from the viewer are set at the least arbitrary values, re- This is done by first developing the coordinate system spectively, 0.1, 1 and 10 (in arbitrary programming (3D space-time frame), then adding to it the schematics “units”). Logarithmically, these correspond to the se- for physical-, life- and civil-science phenomena. quence, -1, 0, 1. 94 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations

The bottom panel of this coordinate system has spatial structure, wherein we find links that call out the schemat- dimensions in both directions, and this suggests that it be ics for atomic states from the unified schematic. used to present the standard schematic for two-dimensional, The primary schematic in biology is the phylogenetic Euclidean space, the usual x-y grid [Space |log(m)| (LF)]. At tree. The modern version of this is a cladogram, a circu- the web site, this schematic includes buttons that allow the lar schematic with branching, radial lines indicating di- schematic to morph into other forms of two-dimensional verging groups of organisms. Distance along the radial space (hyperbolic and spherical). component is used as a measure of rRNA base sequence The left side panel has time along one direction and changes. The angular aspect can be used as a measure of this suggests that we use it to incorporate the standard, diversity (the number of species or other taxonomic two-dimensional schematic for time, the geologic time- groupings). The typical cladogram does not ordinarily ha- scale, modified to include astrophysical and archeological ve a temporal axis (a limitation due to working in two di- increments [Time log (yr) (LF)]. This unified time scale is mensions), and is, therefore, technically imperfect. Since very elaborate. When rigorously developed, much of it is time is measured along a vertical axis in the Unified necessarily on a microscopic scale (if the overall size is a Schematic, a vertical component has been provided, giv- single page and everything is developed to scale). These ing the cladogram three dimensions. The cladogram can portions are accessed via buttons that cause a zoom ef- be brought out by clicking on the link for the biosphere fect. (Text shapes in the time scale will provide this effect [Biosphere (LF)]. Over time, more biology schematics if they turn blue upon mouse-over.) At some point in the will be included. At present, only the cross-sectional near future, the historical and degenerate (heat death) schematic for the protocell (an organelle-level phenome- time scales will be incorporated non) [Organelles (LF)] and ecosystem are included [Eco- The rear panel has a spatial dimension in one direction systems (LF)  Temperate deciduous/mixed (RF)]. and time in the other. This suggests that it be used to No standard schematic for civilization yet exists, so an present the standard, two-dimensional schematic for effort is made here to develop one [Civilization (LF)]. We space-time [Space-time (LF)]. The right side panel is arbi- can follow the same principle as is used for the Unified trarily used to present graphs for the primary cosmologi- Schematic as a whole: combine the existing schematics cal parameters, temperature, energy, density and radius for constituent phenomena. It appears that these are usu- [TE, , , r (LF)]. The overhead front panels have yet ally represented with branching diagrams reminiscent of to be developed. (There is, in fact, a way to use the front the phylogenetic tree (e.g., the branches of law, science, panel without interfering with readability of the graph as religion, language), hence the tree-like schematic for civi- a whole, but discussion of this would require a digression lization. This schematic is a “bare-bones” version. that is extraneous for present purposes.) 5.3 Lettering 5.2 Phenomena schematics The list of terms on the right in the Unified Schematic Within the coordinate system, the lowest diagram is the (e.g., “The Atom,” “Cells,” “Languages”) are primarily in- standard schematic for the gravitational field, the embed- tended to label the individual schematics. However, these ding diagram [Gravitational Field (LF)]. The gravitational labels also have secondary functions (mouse-over and on- field is a deformation of space. Therefore, strictly speak- click events). Most importantly, these terms will call clas- ing, this schematic should be understood as an extension sification tables to the right frame. The on-click event for of the schematic for space (the lower panel). This con- atoms has already been mentioned. Clicking on the link nection will be developed graphically at some point in the for atomic nuclei [Atomic Nuclei (LF)] calls the table of future. nuclides, and so forth for other classes of phenomena. The cross-sectional schematics for particle phenomena For some categories of phenomena, such classification are stacked, one on top of another, above the embedding tables already exist. The rest will be developed over time. diagram, in the order in which the corresponding phe- Note that these labels also have mouse-over events. For nomena have developed. Sizes are according to the abso- example, a mouse-over for [The Atom (LF)] will bring lute values of the logarithms of the radii (meters). These out spatial and temporal scales along the far left and lo- schematics for particle phenomena are called out to a wer edges of the Unified Schematic. These have blinking face-on position by first loading a classification table to indicators for values corresponding to the atom. the right (HTML) frame. Clicking on [The Atom (LF)], The lettering on the left side, the list of disciplines, is for example, loads the Stowe Periodic Table to the right intended to compliment the lettering on the right; as pre- frame. Clicking on symbols therein will load energy level sently conceived, each discipline corresponds to one of diagrams to the right frame. Links therein zoom to fine the primary classes of phenomena. These left-side terms Knowl. Org. 40(2013)No.2 95 M. G. Channon. The Unification of Concept Representations are intended to call lists of subdisciplines to the right side so be used for non-integer values of N. This would be frame. These subdisciplines, in turn, will eventually break especially helpful in evolutionary biology, where diver- down into topic areas, and then, finally, into summations gence times often have enormous margins of error. of concepts, e.g., lists of formulae. In some cases, formu- (These equations cannot be solved explicitly for T in lae are temporarily available via small, red buttons, [F terms of N, but numerical procedures can be used.) (LF)]. Over time, these simple table-of-contents-type lists As another example of such unnoticed regularities, the will be replaced with 3D concept maps; these will repre- effort to produce this Unified Schematic has revealed an sent relationships among disciplines, topics and concepts apparent relevance of 3D classification tables for the main a more realistic and useful fashion. The typical table of jor categories of particle phenomena, although many of contents is a purely one-dimensional concept map, and is these have been poorly developed (Channon 2011). This therefore quite limited. Thus even the lists of disciplines suggests that we should be looking for classification ta- and topics will be developed graphically. bles for other phenomena. The act of doing so would In addition to this outline of the Unified Schematic, it undoubtedly produce important insight, just as Men- would be helpful to at least illustrate its potential for han- deleev’s efforts did. dling detail. The hydrogen atom has been chosen for this It is particularly interesting, then, to imagine the pos- purpose. As developed, this includes all categories of sibility of a “periodic table” for the sciences (one of the concepts relating to quantum mechanics: the Periodic major classes of phenomena). This would suggest that Table, an energy level diagram, orbital schematics, graphs, disciplines occur in a patterned fashion, as do atoms in a data table, formulae and nomenclature. We have here the Periodic Table. Any such table for the sciences would enough detail to illustrate that the Unified Schematic can probably have significant implications for all disciplines, accommodate all details relating to the hydrogen atom. just as the Periodic Table had implications for atomic Hopefully, this is an indication as to how well it will han- theory. It would also constitute an important step toward dle the details of other phenomena. scientific epistemology. Furthermore, it appears that these 3D classification 6.0 Findings; unnoticed regularities, gaps, and in- tables are best derived using a certain principle, one that consistencies in science involves identifying the three fundamental parameters for a category of phenomena, assigning each to an axis and Consider again the previously mentioned, hypothetical then plotting symbols (e.g., “H” for Hydrogen). This was scenario in which we had no schematic cross-section for the approach used by Timmothy Stowe and is perhaps a the earth. In this case, geology would have to be grievously pattern or law-like statement relating to the classification deficient in certain respects. Given this scenario, a deliber- of phenomena in general (Channon 2011). If so, this ate effort to produce such a schematic would quickly lead would be a law relating to laws, another step toward sci- to important insights. The same seems to be happening in entific epistemology. connection with the effort to produce this schematic for This project is also having the effect of identifying gaps the cosmos. It is apparently having the effect of identifying of various kinds. Most importantly, it brings into stark re- various unnoticed regularities, gaps in knowledge and in- lief the fact that we have no holistic discipline for the civil consistencies among the different sciences. sciences, a parallel to physics and biology. Yet civilization is Consider, for example, the scaling equations developed so major a phenomenon, that approximately 60% of scien- above. The original motivation for these concerned the tific disciplines are concerned with studying its various as- need to distribute events along the timeline in a readable pects. (The physical sciences cover approximately 15% and manner; in this application, it was necessary to specify life sciences constitute about 24%.) If such an holistic dis- event number in terms of event time. However, these cipline were to be developed for the civil sciences, the set equations can also be used to specify event time if event of such disciplines would be complete. number is known. Notice, then, that if we had been igno- This assessment is based on the categories of disci- rant of one or more of the event times cited in Table 1, plines presented in the 2006 Survey of Earned Doctorates by these equations could have been used to successfully spe- the National Opinion Research Center (NORC) at the cify their values. This suggests that they can act as laws University of Chicago (Hoffer et al. 2007). Presumably, of nature. If so, they would provide a useful supplement these categories have been carefully prepared, but it is not to methods such as carbon dating, biostratigraphy, and likely that the procedure used was “scientific.” However, nucleocosmochronology (once we are clear on event the assessment of actual numbers of doctorates earned identification and event order). This would be a method would have been rigorous. This count shows that 68.5% relevant to all disciplines. (The other methods have more of doctorates were in the civil sciences. The physical sci- restricted applications.) Hopefully, these equations can al- ences include 10.3% and the life sciences, 21.2%. These 96 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations numbers are similar to those for the disciplines noted Another such gap in knowledge concerns the lack of a above. The computer and information sciences, and doc- consensus as to even the large-scale structure of civiliza- torates earned, are grouped with the civil sciences, as are tion. A similar situation would exist if physics had over- the disciplines and doctorates for the social sciences (in- looked the need to provide a description of the large- cluding psychology, corresponding to mind), engineering, scale structure of the universe (the expanding aggregate education, humanities, and “Other.” For present pur- of galaxy clusters). This would be why we have no stan- poses, any discipline concerned with some feature of dard schematic for civilization. All of this suggests that civilization is a “civil science.” Note that mathematics is the analysis of civilization is still at a preliminary stage of technically a kind of language, and this is a civil phenome- development. non. Engineering relates to technology, another civil phe- Finally in this connection, note that the unification of nomenon. theories and disciplines has been one of the defining fea- Even worse, some of the disciplines associated with tures of scientific progress. We so often read about the civil science phenomena are commonly thought of as be- unification of forces, but at an even grander level of ing largely unrelated to one another. Engineering, for ex- concern, biology, chemistry, and physics were once vie- ample, the discipline that corresponds most closely to the wed as radically distinct. However, biology has been rec- study of technology, is not typically thought of as being a onciled with chemistry and this with physics. What we are sister science to the humanities. Yet art and technology witnessing now, perhaps, is the unification of the civil are both primary classes of civil phenomena, with the disciplines with the rest of science (Whewell 1847; Has- humanities devoted to the former, and engineering to the kell 1972; Wilson 1998; Henriques 2008). “Unification” latter. Both study artifacts, possibly a super-class of civil and reconciliation as used herein refer to consistency phenomena. Engineering, while typically viewed as exclu- among the disciplines. The traditional understanding of sively an applied science, still presents an accumulation of reductionism, e.g., the explanation of biological phenom- knowledge and would be, therefore, at least arguably, the ena simply in terms of chemical processes and proper- discipline corresponding to the study of technology. En- ties, is indeed a defeated effort (Kitcher 1999). gineering is perhaps also thought of as being radically dif- This project has also revealed that there is no consen- ferent from the other civil sciences (sociology, psychol- sus as to a scientific schematic (i.e., map) for mind, pre- ogy, etc.) because it supposedly concerns things that are sumably a major phenomenon. The “mental iceberg” il- not “natural.” This would be a mistake; civilization is lustration for the conscious and unconscious is a meta- closely analogous to a beehive or an ant colony. It reflects phorical cartoon at best. The fact that psychology has yet the behavior of a natural species, and is, therefore, very to produce an agreed-upon schematic for mind is surely much a part of nature. In particular, technology is a part significant. Psychologists are not unaware of this deficit, of this natural phenomenon, and is, therefore, in the but they are also, perhaps, not fully conscious as to how strictest sense, a natural phenomenon itself. Human arti- much of a deficit it is. If a science is unable to provide a facts become something other than natural only if the realistic, structural description (map) of the phenomenon term “nature” is defined as relating to things nonhuman. it is studying (e.g., the interior of the earth), it cannot be Engineering is also thought to be unrelated to the other said to have reached a mature understanding of its sub- civil sciences because it makes liberal use of physics con- ject matter. cepts, and this is, admittedly, an important consideration. Another revealed gap in knowledge concerns the sys- Nevertheless, the apparent rule is that disciplines are cate- tematic identification of, and distinctions between, the ba- gorized (and associated) on the basis of the phenomena sic classes of phenomena. This is nowhere to be found in with which they are concerned. Thus cellular biology and science. Yet it is something so fundamental, that it should zoology fall within the purview of biology, as chromody- be common knowledge to grade-school graduates. The list namics and astronomy are classified as physics. For the provided here (Table 1) is surely not the last word on the same reason, the humanities should be associated with the topic, but it is at least a start. We need to get clear on this other civil sciences, since it is concerned primarily with art, as soon as possible. What could be more basic to science one of the primary components of civilization. (Defini- than identifying the major categories of phenomena? tions for the term vary.) Likewise, institutional studies (e.g., There is also considerable confusion in regard to the usage agriculture, business, medicine) would fall under this cate- of the term “institution.” It is often used in reference to gory, since institutions are a major component of civiliza- things that are only institution-like (e.g., money and mar- tion. Furthermore, physics concepts are also used in biol- riage). The absence of classification tables for most of ogy (“biophysics”) and even the social sciences (e.g., these categories of phenomena would be another of these econophysics). Nevertheless, this doesn’t make biology and gaps in knowledge. Finally, in physics, there is confusion economics subdisciplines to physics. concerning elementary particles and hadrons. (Hadrons, Knowl. Org. 40(2013)No.2 97 M. G. Channon. The Unification of Concept Representations such as protons and neutrons, are particle systems, consist- with a focus on how we acquire it. It was thus a reduction ing of either two or three quarks. Elementary particles― to psychology. This effort is generally viewed as a failure, bosons and fermions―are, as the term “elementary” sug- most notably because it eliminates the normative, a con- gests, truly basic, not made up of smaller particles, or so it cern with what knowledge ought to be (e.g., Almeder 1999; appears at this time. Yet mesons―a type of hadron―are Bonjour 1994; Foley 1994; Fumerton 1994; Putnam 1982); often spoken of as “bosonic” and categorized accordingly. this has long been the fundamental and distinctive concern There are similar sorts of confusion regarding other parti- of epistemology. Quine himself has also noted the circu- cles. The mistake is the result of focusing on “spin,” rather larity of attempting to use the empirical sciences to validate than structural form and hierarchical position. A compara- themselves, a variation on Hume’s problem (Quine 1990). ble error would be to view homonuclear diatomic mole- Another effort along these lines is cooperative natural- cules, e.g., dinitrogen, as types of atoms.) ism, the view that psychology is generally relevant to In a related manner, the project is also having the ef- epistemology (Goldman 1991; Haack 1995; Harman fect of revealing inconsistencies between the various sci- 1974; Kornblith 1994; Stich and Nisbett 1980). While this ences. For example, the astrophysics community has been is not terribly problematic, neither does it constitute a using a time scale that is expressed in terms that conflict transformation of epistemology; it is rather, an expres- with the usage relating to the geologic time scale. Also, sion of the sense that developments in other disciplines the increments of the archeological time scale are slightly simply have some relevance to epistemology. inconsistent with the geologic time scale (the Calabrian A third variation is substantive naturalism, the view and Stone Ages overlap). Indeed, until now, there has ap- that normative epistemic statements reflect or corre- parently been no attempt to unify these various time spond to natural facts, statements about the world (e.g., scales in detail. (Wikipedia collaborators are making a Kim 1988; Lycan 1988; Maffie 1990; Steup 1995). This is start at this.) not problem-free, but it is widely embraced by epistemologists. However, some argue that it is difficult to dis- 7.0 Conclusion; the advent of scientific epistemology tinguish between this and the positions of more traditional epistemologists (Chisholm 1982; Cleve 1985). Sub- The primary goal of the project, then, is to graphically stantive naturalism is apparently little more than classical and systematically map all knowledge, thereby identifying epistemology (e.g., Feldman 2012). It does not go very far patterns, gaps and inconsistencies amongst the laws of in developing epistemology as a scientific discipline. science. Most importantly, it is a search for regularities, However, while knowledge is only indirectly related to throughout and across the various disciplines. Any such psychology, it is of direct concern to the information sci- regularities would be laws relating to laws (i.e., laws relat- ences. Indeed, in a recent survey of some 50 leading theo- ing to knowledge), and therefore, they would be the sub- rists in the field, half provided definitions for “information ject matter of epistemology. If knowledge is anything like science” which used the term “knowledge,” often repeat- other phenomena, then there should be such regularities. edly (Zins 2007a). Zins, himself, in his contribution to the This suggests that epistemology might develop into a sci- list, surprisingly mentions “knowledge” more often than ence. Just as the other sciences have come out of phi- “information” (Zins 2007a, 339). Zins (2007a, 340-41) fur- losophy, so epistemology may now be separating. The ther cites “Six Conceptions of Information Science," all of goal of the project, then, might be alternatively character- which invoke a critical reference to knowledge. He also ized as an effort to facilitate the science of science itself, maintains that “Information science is one of six knowl- scientific epistemology. (Perhaps needless to say, episte- edge fields ... these are philosophy of knowledge (epistemology has long been concerned with the effort to ar- mology), philosophy of science, history of science, sociol- ticulate law-like statements. The deductive-nomological, ogy of knowledge, methodology of science, and informa- covering-law, model is certainly an attempt to characterize tion science” (Zins 2007a, 339). Quite plausibly, scientific scientific explanation in terms of a law-like statement. epistemology would also include information theory. In a The point of the present discussion, however, concerns related article, one that concerns the “data-information- the effort to develop laws of a more particular nature, knowledge-message phenomena,” all of the commenting e.g., the possibility that there are patterned (3D) classifi- theorists (forty-five in total) implicitly identify knowledge cation tables for all categories of phenomena.) as one of the fundamental concepts of information sci- This notion of a scientific epistemology is not new. ence (Zins 2007b). Information science, then, would be There has been, for some time now, an interest in its de- closely related to epistemology, and scientific epistemology velopment. Perhaps most notably, this came in the form of would be the “six knowledge fields” mentioned above. the “replacement naturalism” of W.V.O Quine (1969), an Further still, the classification of the sciences would attempt to supplant the conceptual analysis of knowledge be a relevant consideration, just as the classification of 98 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations atoms is critical to quantum mechanics. In general, scien- Perhaps invariably in human activities (e.g., warfare), a tific epistemology would be concerned with any and all dramatic improvement in strategy is followed by a dra- attempts to account for regularities relating to what we matic improvement in results. If any such enhancement might refer to as intelligence (data, information, message, is made to the scientific process, it would surely have a and knowledge). spectacular effect on the pace and depth of discovery. In this view, classical epistemology (the concern with One might argue that the current strategy has worked just issues such as the justification of knowledge) constitutes fine, producing a perfectly acceptable pace. But our an- the philosophic area of this science. Just as the philoso- cestors would have said the same of the traditional cav- phy of mind can be viewed as a subdiscipline of psy- alry; after all, it was so much faster than marching. They chology, so classical epistemology might best be thought would have given little consideration to the possibility of of as the philosophic portion of scientific epistemology. flying at 1000 miles per hour. A mob action is often ef- (Perhaps needless to say, we could retain the present fective, but an organized army is vastly more so. meaning of “epistemology,” and refer to “scientific epis- The scientific enterprise is not a war; this metaphor temology” as something such as “knowledge science.”) has unappealing connotations. However, science is a criti- The idea is not that epistemology would develop as an cal part of the effort to build civilization, and this has empirical discipline, as was the hope of replacement natu- been very much like a war. It has involved enemies of a ralists. Rather, that it will develop by some combination of sort (poverty, illiteracy, actual war) and it has produced an methods into a body of established knowledge, i.e., a sci- enormous number of killed and wounded. Now, the ad- ence. (Notice that disciplines such as mathematics are, or vancement of civilization is dependent on progress in can be viewed as, sciences, despite no use of empirical science (among other things), and scientific epistemology methods.) The applicable methods, in this case, might be might well produce a second scientific revolution, an- largely those already in use in the information sciences (ar- other acceleration in the pace of discovery (something of chival research, content analysis, etc.). However, we now a parallel to the second industrial revolution or the in- have other methods emerging, most notably the analysis of formation revolution). If so, it would surely have a dra- symmetry so common in physics. The attempt to unify the matic effect on the advancement of civilization in gen- concept representations of science is an application of this eral. It would serve to bring this war-like process to a method, and, as we see, it appears to be fruitful. In any close much more quickly, thereby minimizing further ca- case, a scientific epistemology would provide the basis for sualties. (The notion that the development of civilization a systematic approach to discovery; rather than each inves- is an open-ended process is an unsubstantiated, if very tigator simply following his or her own muse, diverging in common supposition. We have no evidence in support of whatever direction seems interesting, we could proceed in it. On the contrary, the development of all natural sys- a more organized manner; scientific epistemology would tems seems to involve a type of maturation process, fol- provide, in effect, a greatly enhanced perspective from lowed by a period of relative stability.) which to choose research topics. And while we are exploring such far-reaching possibili- Up to this point in time, scientific research as a whole ties, we might want to briefly consider certain ultimate has not had the benefit of the sort of perspective that implications. The physics community has been, for some would be provided by a true science of knowledge; as a time, very much engaged in the effort to produce a “the- result, it has had no detailed, well-considered strategy. We ory of everything” (ToE), a single conceptual model have all been moving in the same general direction, into from which the greater body of laws can be derived. This the unknown, but our actions have not had the same ef- effort is motivated by many similar, though more limited, fect that they would have, if the effort had been more successes (e.g., the reduction of Kepler’s laws and Gali- highly organized. If this perspective had been in place, leo's theories of motion to Newtonian mechanics). This we would not have overlooked such gaping deficits as the effort has not succeeded, primarily because general rela- failure to provide a characterization for the large-scale tivity and quantum mechanics are not easily reconciled. structure of civilization. If this sort of methodical ap- But this failure is perhaps simply a matter of perspective. proach is possible, science would take on the character of The idea of producing a single theory is arguably the an organized campaign. In this case, the historical process problem. Theorists would be more successful, perhaps, if would have been one that began with uncoordinated, they were to pursue a single system of fundamental scattered activity (thousands of years ago), developed physical theories. This would be an effort to transform a into a massive, yet still uncoordinated, mob action (over set (or “heap”) of laws to a system of laws. We might, the 19th and 20th centuries) and would now be consum- then, want to consider a corresponding effort, one flow- mated in the transition to a tactically sophisticated offen- ing from a science at the other end of the spectrum of sive. disciplines, the pursuit of a system of theories which Knowl. Org. 40(2013)No.2 99 M. G. Channon. The Unification of Concept Representations characterizes all patterns amongst the laws of science. Childe, Gordon V. 1983. Man makes himself. [s.l.]: Plume. The ToE for physics is an attempt to unify, simplify and Chisholm, Roderick M. 1982. The foundations of knowing. consolidate the laws of physics, and a great deal of pro- Minneapolis: University of Minnesota Press. gress has been made toward this end. The epistemic Cleve, James Van. 1985. Epistemic supervenience and the equivalent would be a similar attempt as concerns the circle of belief. The monist 68: 90-104. conceptualization of laws in general. If successful, this Corballis, Michael C. 2003. From hand to mouth: The origins effort would produce a theorist’s dream machine, a set of of language. New Jersey: Princeton University Press. constraints on theoretical models in all disciplines and in- Dalley, Stephanie and Oleson, John Peter. 2003. Sennach- dications of otherwise unsuspected features and regulari- erib, Archimedes, and the water screw: The context of ties of the cosmos. This might turn out to be a flight of invention in the ancient world. Technology and culture 44: fancy, but the potential return is so great and the invest- 1-26. ment of resources so small that the ride would be well Davies, Paul. 2004. Cosmic blueprint: New discoveries in natures worth the risk. (Notice that the effort would not require ability to order the universe. Philadelphia; London: Temp- anything such as multi-billion dollar particle accelerators leton Press. or space-based telescopes.) Yet the results would be rele- Deutsch, David. 2011. The beginning of infinity: Explanations vant throughout the full extent of intellectual activity. that transform the world. [New York]: Viking. Durant, Will. 1939. The story of civilization. New York: Si- References mon and Schuster. Durkheim, Emile. 1915. The elementary forms of the religious Almeder, Robert F. 1999. Harmless naturalism: The limits of life, a study in religious sociology. London; New York: G. science and the nature of philosophy. Chicago: Open Court Allen & Unwin; Macmillan. Publishing. Fairlie, John A. 1923. The Separation of Powers. Michigan Bertalanffy, Ludwig Von. 1968. General systems theory. New law review 21: 393–436. York: George Braziller. Feldman, Richard. 2012. Naturalized epistemology. In Boesch, Christophe and Boesch, Hedwige. 1984. Mental Zalta, Edward N., The Stanford encyclopedia of philosophy. map in wild chimpanzees: An analysis of hammer Available http://plato.stanford.edu/archives/sum2012/ transports for nut cracking. Primates 25: 160-70, doi:10. entries/epistemology-naturalized/. 1007/BF02382388. Foley, Richard. 1994. Quine and naturalized epistemol- Bonjour, Laurence. 1994. Against naturalized epistemology. Midwest studies in philosophy 19: 243–60, doi:10. ogy. Midwest studies in philosophy 19: 283-300, doi:10. 1111/j.1475-4975.1994.tb00288.x. 1111/j.1475-4975.1994.tb00290.x. Frisch, John A. 1940. Did the Peckhams witness the inven- Borner, Katy, Chen, Chaomei and Boyack, Kevin W. tion of a tool by ammophila urnaria?” American midland 2003. Visualizing knowledge domains. Annual review of naturalist 24: 345–50. information science and technology 37: 179-255. Fumerton, Richard. 1994. Skepticism and naturalistic Bostrom, Nick. 2002. Anthropic bias: Observation selection effects epistemology. Midwest studies in philosophy 19: 321–40, in science and philosophy. New York; London: Routledge. doi:10.1111/j.1475-4975.1994.tb00292.x. Bouwens, Rychard J., Illingworth, Garth D., Franx, Marijn Geyl, Pieter. 2001. History of the Dutch-speaking peoples and Ford, Holland. 2008. z ~ 7–10 galaxies in the 1555-1648. London: Phoenix Press (UK). HUDF and GOODS fields: UV luminosity functions. Gilmour, Iain and Sephton, Mark A. 2004. An introduction The astrophysical journal: 230-50, doi:10.1086/590103. to astrobiology. Cambridge: Cambridge University Press. Brisch, Nicole. 2008. Religion and power: Divine kingship in Goetz, Philip W. 1987. Propaedia: Outline of knowledge and the ancient world and beyond. Chicago: The Oriental Insti- guide to the Britannica. Chicago: Encyclopaedia Britan- tute of the University of Chicago. nica, Inc. Brockmann, H. Jane. 1985. Tool use in digger wasps Golding, M. P. 1967. History of philosophy of law. In (hymenoptera: sphecinae). Psyche: A journal of entomology Encyclopedia of Philosophy. New York; London: Macmil- 92: 309–29, doi:10.1155/1985/73184. lan Publishing Co. and The Free Press, pp. 254-64. Brocks, Jochen J., Logan Graham A., Buick, Roger and Goldman, Alvin I. 1991. Liaisons: Philosophy meets the cogni- Summons, Roger E. 1999. Archean molecular fossils tive and social sciences. Massachusetts: The MIT Press. and the early rise of eukaryotes. Science 285: 1033–6, Haack, Susan. 1995. Evidence and inquiry: Towards reconstruc- doi:10.1126/science.285.5430.1033. tion in epistemology. Osxford; Massachusetts: Wiley- Channon, Martin. 2011. Forum: The philosophy of clas- Blackwell. sification: The Stowe Table as the definitive periodic Harman, Gilbert. 1974. Thought. New Jersey: Princeton system. Knowledge organization 38: 321–7. University Press. 100 Knowl. Org. 40(2013)No.2 M. G. Channon. The Unification of Concept Representations

Harrison, Edward. 2000. Cosmology: The science of the uni- Kukla, André. 2010. Extraterrestrials: A philosophical perspec- verse. Cambridge: Cambridge University Press. tive. [Lanham, Maryland]: Rowman & Littlefield. Haskell, Edward. 1972. Full circle: The moral force of unified Lakatos, Imre. 1970. Falsification and the methodology of science. New York: Gordon & Breach Publishing scientific research programs. In Lakatos, Imre and Mus- Group. grave, Alan, eds., Criticism and the growth of knowledge: Pro- Henriques, Gregg R. 2008. Special section: The problem ceedings of the international colloquium in the philosophy of sci- of psychology and the integration of human knowl- ence, London, 1965, Vol. 4. Cambridge: Cambridge Uni- edge contrasting Wilson’s consilience with the tree of versity Press. knowledge system. Theory & psychology 18: 731–55, Laszlo, Ervin. 1996. The systems view of the world: A holistic doi:10.1177/0959354308097255. vision for our time. Cresskill, New Jersey: Hampton Press. Hoffer, Thomas B, Hess, Mary Hess, Welch Jr., Vincent Lemarchand, Guillermo A. 2006. The Life-time of tech- and Williams, Kimberly. 2007. Doctorate recipients from nological civilizations; A description of the terrestrial United States universities: Summary report 2006 (survey of technological adolescence case. In Burdyuzha, Vladi- earned doctorates). Chicago: National Opinion Research mir, ed., The future of life and the future of our civilization, Center (University of Chicago). Available http://www. Dordrecht : Springer, pp. 457–67. norc.org/SED.htm. Lu, Peter J. 2004. Early precision compound machine Hopkins, Andrew M. and Beacom, John F. 2006. On the from ancient China. Science 304: 1638, doi:10.1126/ normalization of the cosmic star formation history. The science.1096588. astrophysical journal 651: 142–54, doi:10.1086/506610. Lycan, William G. 1988. Judgement and justification. Cam- James, William. 1902. The varieties of religious experience. bridge: Cambridge University Press. New York: Modern Library. Maffie, James. 1990. Recent work on naturalized episte- Jenkins, Iredell. 1973. Art for art’s sake. In Wiener, Philip. mology. American philosophical quarterly 27: 281–93. P., ed., Dictionary of the history of ideas; Studies of selected Mannucci, Filippo, Buttery, Helen, Maiolino, Roberto, pivotal ideas. New York: Charles Scribner’s Sons. Marconi, Alessandro and Pozzetti, Lucia. 2006. Evi- Kibunjia, Mzalendo. 1994. Pliocene archaeological occur- dence for strong evolution of the cosmic star formation density at rences in the Lake Turkana Basin. Journal of human evo- high redshift. Available http://arxiv.org/abs/astro-ph/ lution 27: 159–71, doi:10.1006/jhev.1994.1040. 0607143, doi:10.1051/0004-6361:20065993. Kim, Jaegwon. 1988. What is naturalized epistemology? Montesquieu, Charles de Secondat, Baron de. 1748. De In Tomberlin, James E., ed., Philosophical perspectives: l'esprit des loix, ou du rapport que les loix doivent avoir avec la Epistemology. Asascadero, CA: Ridgeview Pub Co, pp. constitution de chaque gouvernement, mœurs, climat, religion, 381–406. commerce, etc. (sic) ; à quoi l'auteur a ajouté des recherches sur Kimbel, William H., Walter, Robert, C., Johanson, Don- les lois romaines touchant les successions, sur les lois françaises et ald C., Reed, Kaye, Aronson, James L., Assefa, Ze- sur les lois féodales . Genève: Barrillot & Fils. lalem, Marean, Curtis W., Eck, Gerald Gilbert, Bobe, Ota, Kazuaki, Iye, Masanori, Kashikawa, Nobunari, Shi- Rene, Hovers, Erella, Rak, Yoel, Vondra, Carl, Ye- masaku, Kazuhiro, Kobayashi, Masakazu, Totani, To- mane, Tesfaye, York, Derek, Chen, Yang, Evensen, monori, Nagashima, Masahiro, Morokuma, Tomoki, Norman, M., Smith, Patrick E.. 1996. Late pliocene Furusawa, Hisanori, Hattori, Takashi, Matsuda, Yuichi, homo and oldowan tools from the Hadar Formation Hashimoto, Tetsuya and Ouchi, Masami. 2008. Reio- (Kada Hadar Member), Ethiopia. Journal of human evo- nization and galaxy evolutionpProbed by z = 7 Lyα lution 31: 549–61, doi:10.1006/jhev.1996.0079. emitters. The astrophysical journal 677: 12–26, doi:10. Kitcher, Philip. 1999. Unification as a regulative ideal. Per- 1086/529006. spectives on science 7: 337–48, doi:10.1162/posc.1999.7.3. Owen, Peter D. 1987. The art of painting. In Goetz, Phi- 337. lip W. Goetz, ed., The encyclopedia Britannica. Chicago: Knoll, Andrew H. and Carroll, Sean B. 1999. Early animal Encyclopaedia Britannica. evolution: Emerging views from comparative biology Peckham, George W. and Peckham, Elizabeth G. 1898. and geology. Science 284: 2129–37, doi:10.1126/science. On the instincts and habits of the solitary wasps. Madison: 284.5423.2129. Wisconsin Geological and Natural History Survey. Kornblith, Hilary. 1994. Naturalizing epistemology. Massa- Putnam, Hilary. 1982. Why reason can’t be naturalized. chusetts: The MIT Press. Synthese 52: 3–23, doi:10.1007/BF00485252. Kramer, Samuel Noah. 1988. History begins at Sumer: Thir- Quigley, Carroll. 1961. The evolution of civilizations: An in- ty-nine firsts in recorded history. Philadelphia: University of troduction to historical analysis. New York: Macmillan. Pennsylvania Press. Quine, Willard van Orman. 1969. Ontological relativity and other essays. New York: Columbia University Press. Knowl. Org. 40(2013)No.2 101 M. G. Channon. The Unification of Concept Representations

Quine, Willard van Orman. 1990. Norms and aims. In http://www.usccb.org/beliefs-and-teachings/what-we- The Pursuit of Truth. Cambridge: Harvard University believe/catechism/catechism-of-the-catholic- Press. church/epub/index.cfm#. Schopf, J. William. 1999. Cradle of life: The discovery of Verma, Aprajita, Lehnert, Matthew D., Schreiber, Natas- earth’s earliest fossils. Princeton: Princeton University cha M. Förster, Bremer, Malcolm N. and Douglas, Press. Laura. 2007. Lyman-break galaxies at z ~ 5 – I. First Schrecker, Paul. 1948. Work and history: An essay on the significant stellar mass assembly in galaxies that are structure of civilization. Princeton: Princeton University not simply z ~ 3 LBGs at higher redshift. Monthly no- Press. tices of the royal astronomical society 377: 1024–42, doi: Semaw, Sileshi, Renne, Paul, Harris, John W. K. , Feibel, 10.1111/j.1365-2966.2007.11455.x. Craig S., Bernor, Bernor, Raymond, L., Fesseha, Ne- Whewell, William. 1847. The philosophy of the inductive sci- breed and Mowbray, Ken. 1997. 2.5-million-year-old ences founded upon their history. London: J.W. Parker. stone tools from Gona, Ethiopia. Nature 385: 333– Williston, Samuel Wendell. 1892. Notes on the habits of 336, doi:10.1038/385333a0. ammophila. Entomology News 3: 85–6. Shuster, Gabriela and Sherman, Paul W. 1998. Tool use Wilson, Edward O. 1998. Consilience: The unity of knowledge. by naked mole-rats. Animal Cognition 1: 71–4, doi:10. New York: Alfred A. Knopf. 1007/s100710050009. Wood, Bernard. 1997. The oldest whodunnit in the Silk, Joseph. 1999. Microwave background and structure world. Nature 385: 292–3, doi:10.1038/385292a0. formation. In Dekel, Avishai and Ostriker, Jeremiah P., Yüksel, Hasan, Kistler, Matthew D., Beacom, John F. and eds., Formation of Structure in the universe, Cambridge: Hopkins, Andrew M. 2008. Revealing the high-redshift Cambridge University Press, pp. 98–131. Available star formation rate with gamma-ray bursts. The astro- http://adsabs.harvard.edu/abs/1999fsu..conf...98S. physical journal 683: L5–L8, doi:10.1086/591449. Steup, Matthias. 1995. An introduction to contemporary episte- Zeilik, Michael and Gregory, Stephen A. 1997. Introductory mology. Upper Saddle River, N.J.: Prentice Hall. astronomy and astrophysics. Singapore ; London : Stich, Stephen P. and Nisbett, Richard E. 1980. Justifica- Brooks/Cole / Thomson Learning tion and the psychology of human reasoning. Philoso- Zins, Chaim. 2007a. Conceptions of information science. phy of Science 47: 188–202. Journal of the American Society for Information Science and Tilton, George Robert. 1988. Age of the solar system. In Technology 58: 335–50, doi:10.1002/asi.20507. Kerridge, John F. and Matthews, Mildred S., eds., Meteor- Zins, Chaim. 2007b. Conceptual approaches for defining ites and the early solar system. Tucson, Ariz.: University of data, information, and knowledge: research articles. Arizona Press, pp. 259–275. Available http://adsabs. Journal of the American Society for Information Science and harvard.edu/abs/1988mess.book..259T. Technology 58: 479–93, doi:10.1002/asi.v58:4. United States. Catholic Church. 1997. Catechism of the Catho- lic Church. New York: Image Doubleday. Available:

102 Knowl. Org. 40(2013)No.2 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification

Reader-Interest Classification: Concept and Terminology Historical Overview

Daniel Martínez-Ávila* and Rosa San Segundo**

*/** Universidad Carlos III de Madrid, C/Madrid 126 28903 Getafe (Madrid), Spain, * , **

Daniel Martínez-Ávila is an assistant professor at Carlos III University of Madrid. He has a PhD in library and information science and a background in computer science. His research interests include classification systems, the application of poststructural and critical theories to information organization, and free software philosophy.

Rosa San Segundo has a PhD in information science and is a professor of classification and knowledge organization. She has been Secretary and Head of the Department of Information and Library Sciences at Carlos III University. She is President of the Spanish Chapter of ISKO, and a member of the UDC Consortium Adviso- ry Board.

Martínez-Ávila, Daniel, and San Segundo, Rosa. Reader-Interest Classification: Concept and Terminology Historical Overview. Knowledge Organization. 40(2), 102-114. 69 references.

ABSTRACT: During the last century, the concept of reader-interest classifications and its related terminology have shown a well-established presence and commonly-agreed characteristics in the literature and other classification discourses. During the period 1952-1995, it was not unusual to find works, projects, and discourses using a common core of characteristics and terms to refer to a recognizable type of projects involving alternative classifications to the DDC and other traditional practices in libraries. However, although similar projects and characteristics are being used until the present day, such as those of implementation of BISAC in public libraries, the use of reader-interest classification-related terms and references have drastically declined since 1995. The present work attempts to overview the concept and terminology of reader-interest classifications in a historical perspective emphasizing the transformation of the concept and its remaining characteristics in time.

Received 24-10-2012; Revised 31-12-2012; Accepted 12-14-2013

1.0 Introduction The present study attempts to serve as an overview and analyze the history and transformation of the concept of Following a social view of concepts, “as socially negoti- reader-interest classification and its terminology, from the ated meanings that should be identified by studying dis- emergence of the concept to the present day. courses rather than by studying individual users or a priori principles” (Hjørland 2009, 1530), reader-interest clas- 2.0 Reader-interest classification definition sifications might be considered a well-accepted concept during the last century according to the discourses of The term ‘reader-interest classification’ has been used to classifications, showing an agreed-upon core of character- describe various approaches to library classification. In istics in the literature and a recognizable, although some- general, this umbrella term—and several of its near- how shifting, related terminology along the time. However, synonyms—refer to alternatives to traditional library sys- despite the presence of some of their characteristics and tems such as the Dewey Decimal Classification (DDC), the criticisms in some current classifications, the usage of Universal Decimal Classification (UDC), and the Library some of its terminology and terms found in the reader- of Congress Classification (LCC). According to its advo- interest classification literature of the past seems to have cates, reader-interest classification provides a more suit- declined or even disappeared from the current discourses. able arrangement for the reader because it gathers related Knowl. Org. 40(2013)No.2 103 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification terms scattered across the system and is more intuitive to Paul Dunkin (1969, 124) expressed the reader-interest clas- use. sification philosophy as follows: “RIC [reader-interest clas- Within the literature, there are a great variety of terms sifications] centers not on shelving books nor on logic, but that refer to this concept with minor variations; among on people and the fields of interest related to the everyday them are ‘alternative arrangement,’ ‘user-orientated ar- needs of the people.” In another definition, Sharon Baker rangement,’ ‘categorized arrangement,’ ‘verbal arrange- (1988a, 3) highlights the influence of commercial culture ment,’ ‘bookstore arrangement,’ ‘stock categorization’ (in- on this system, describing ‘reader interest classification’ as a cluding the different regional spelling variations), ‘reader- “natural language classification similar to what the major centred classification’ (used more recently by some Aus- bookstore chains use.” More recently, “reader interest clas- tralian authors), ‘two-tier arrangement,’ ‘integrated stock,’ sification” was defined in the Harrod’s Librarian’s Glossary of ‘intensive use of paperbacks,’ and ‘subject departmenta- Terms Used in Librarianship, Documentation and the Book Crafts lism.’ This variety of terminology and the regional differ- as a “simple and broad classification intended to reflect the ences of use were pointed out by Sapiie (1995, 144): special interests of readers rather than the subject contents of books as such” (Prytherch 1990, 515). One of the most There are nearly as many variations on the names comprehensive and recent definitions of the concept, this given to reader-interest classification as there are in- time under the name ‘Reader’s Interest Classification,’ was stances of its use. Librarians in the U.S. have given by Mohinder Satija (2004, 182) in A Dictionary of brought the plan into the modern age with the up- Knowledge Organization: to-date terms of merchandising, marketing and bookstore arrangement. In Britain, such terms as A classification designed to serve the immediate reader interest categories, categories or categoriza- needs of the targeted users. Such systems violate tion are favored, but broad interest groups or user the filiatory sequence to bring together disparaged orientation are also used. Librarians create centers subjects needed by a user group. These are useful in of interest in France and special interest comers in mission oriented or multidisciplinary subjects. In a Japan. In Germany, immediate concern areas or al- commerce college, e.g., it may be more pragmatic to ternative arrangement is used; in The Netherlands, place commercial law with commerce at 380. It is broad subject arrangement, reader interest catego- true to say that reader’s interest classification ries or topics of current interest; and in South Af- adopted so far are not always satisfactory and rica, the plan is called reader’s interest classification. sometimes correspond to ephemeral vogues. It reflects a middle level of ambition in knowledge or- In practice, many of these terms are used indistinctly in ganisation. It is a compromise between ad hoc clas- the literature or are studied together in a variety of case sification and rigorously scientific classification. studies. However, ‘reader-interest classification’ seems to be the most representative of these terms. Examples of Although the original reader-interest classification of the the relevance of the term ‘reader-interest classification’ Detroit Public Library was later presented as a “reader- may be found in its use as the authorized term in the Li- interest book arrangement,” too (Detroit Public Library brary Literature & Information Science Full Text bibliographic 1955), it should be noted that ‘arrangement’—one of the database thesaurus and also in its use by Ruth Rutzen, main aspects of the concept—is not always mentioned in Home Reading Services Director at Detroit Public Li- these definitions, as authors generally accept its impor- brary (in the first acknowledged case of a library applying tance; that is, it is assumed that the only way that readers a classification of these characteristics). According to, can see their interests reflected in the library collection is Rutzen (1952, 478): through physical display and organization. Indeed, one of the most commonly expressed and well-accepted princi- the term reader interest classification is not a new ples of reader-interest classifications is that the arrange- library term. It has a familiar sound to those who ment of books by categories has to be simple and self- have followed the literature on adult education in explanatory, minimizing the need to use the catalog or libraries. What is the purpose of the reader interest staff to find a specific book or subject (McCarthy 1982; classification? It is yet another effort to make our Sapiie 1995). What is more, this arrangement has to be service more meaningful and pertinent to the inter- organized according to the reader’s perspective and not ests and needs of the general reader. What is it? It that of the book, the library, or the librarian. Outside of is a plan to arrange books on the shelf in terms of the discipline of knowledge organization, reader-interest use and interest by the potential reader rather than classifications were thus employed by the user-orientation strictly by subject content. movement in library and information science. 104 Knowl. Org. 40(2013)No.2 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification

3.0 Reader-interest classifications as part of called users’ standards. Such a contrast is neatly illustrated the user-orientation movement by comments made by Alan Sykes (1982, 383), librarian at Camden Library, after attending a “reader-interest classi- User orientation in libraries was defined by Den Reader fication course” in Surrey, another library adopting a as “any action by library staff which helps to make (and reader-interest classification: keep) the library relevant, busy, pleasant to use, and en- courages self help” (Reader 1982, 35). Reader added that: Above all, especially in the context of this course, we had to try to arrange stock in such a way that it at the same time an attempt is made to keep a bal- reflected the mind of the reader and not the mind ance between the commonly-agreed areas of ser- of the traditional ivory-tower, Dewey-obsessed, li- vice (recreation/information/education). The guid- brarian. Since most readers were browsers, the best ing light is always to make stock as accessible as arrangement was probably by broad subject areas - possible to readers, and to ask whether any of Li- in short, a popular arrangement. brarianship’s ‘sacred cows’ are sacred simply to the profession, and are of no practical help to readers. For Douglas Betts, principal librarian at Surrey County Libraries, there was a danger of developing reader-interest There are three important concepts within this definition categories which accounted for the needs of the librarian that relate to reader-interest classifications: self-help, areas instead of the user, or in other words, of following the of service (or interest), and the questioning of librarian- same process applied in traditional classifications: “lists ship’s ‘sacred cows,’ i.e. traditional methods of classifica- (sometimes helpful) of favoured topics and fiction genres tion and arrangement in libraries. Within the literature on appear in some surveys, although the categories tend to be reader-interest classification, this last point has been of the librarian’s, not the reader’s” (Betts 1982, 65). However, particular interest to several authors. For Ainley and Tot- Betts regarded the categorization system as only a small terdell, for instance, the rejection of traditional methods part of a larger plan towards a more user-orientated ser- of classification and arrangement was one of the main vice—one which combined the physical re-presentation of raisons d’être behind reader-interest classifications in librar- public libraries and a systematic demand-related approach ies (Ainley and Totterdell 1982). to stock logistics and books selection (Betts 1982). In general, it was claimed that traditional classification Opposition to the use of the DDC and UDC has ex- systems were not designed with the user in mind, but tended to the recent implementations of bookstore rather according to the interests of the collection or the schemes in libraries (Martínez-Ávila, Olson, and Kipp staff or in the name of some academic dogma which was 2012a; Martínez-Ávila, Kipp, and Olson 2012b). However, of no use to the readers—i.e., according to one of Li- it is arguable that, despite two decades of the reader- brarianship’s ‘sacred cows.’ Elsewhere, Totterdell (1978, interest movement, satisfactory alternatives have yet to be 13) also stated that “librarians may fear that the commu- developed. As pointed out by Birger Hjørland, part of the nity’s ideas of what the library’s role should be may not problem is one of paradigms; that the ‘bibliographic para- coincide with theirs,” making the opposition between li- digm’ is sometimes wrongly assimilated to the ‘positivist brarians and “the community” even more evident. Ac- view.’ Such a perspective might also have ramifications for cording to Den Reader (1982, 41): the debate concerning cases of reader-interest classification and the philosophy that informs them. Hjørland (2007, 2) Librarians know how good their libraries are, but writes: does the public? ... Without orientation towards its users, the library is in danger of dying, and where It seems as if the term ‘the bibliographical para- does that leave those who argue only about the ne- digm’ has only been used negatively as a contrast of cessity of maintaining standards? Standards, yes, something better. In this context it has been sug- but in a users’ library, not a librarians’ library! gested that it is a part of ‘the systems-oriented perspective’ (or ‘physical paradigm’) in library and in- As Reader pointed out, standards and standardization formation science, which in the received view, is were not considered to be contrary to the concept of opposed to user-oriented paradigms. user-orientation; it was only the philosophy behind those traditional standards that was rejected. In the context of reader-interest classification, the system- Within the literature on reader-interest classification, it oriented perspective would be represented by traditional was quite common to contrast the traditional standards schemes and views such as Dewey; in the user-oriented held by library/librarians (such as Dewey) with what were paradigms, on the other hand, the “something better” Knowl. Org. 40(2013)No.2 105 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification would correspond to reader-interest classifications. In a non-inquirer reader—one who never approaches the li- different context, another example of assimilation of the brarian for help and does not want to be talked to. This ‘bibliographic paradigm’ to positivism and contrast of kind of reader is usually characterized as a browser since something else can be found in San Segundo (2004). sometimes his/her autonomy of use vs. library training However, it cannot be assumed that a proper loca- does not allow the reader any other main way to access to tional view—the way facets are displayed within the sys- the information. Because of this, ‘browser’ is one of the tem—depends on objective definitions of potential use, most common near-synonyms of ‘non-purposive reader’ meaning, and mental activity. Even when users (in oppo- to be found within the context of reader-interest classifica- sition to librarians or information scientists) are claimed tion—a term that emphasizes how access is acquired rather to be the guiding force in the development of the system, than what type of access is preferred. it should also be recognized that there may be different As suggested above, the purposive reader is usually de- groups of users who would benefit from different sets of fined as a reader who is more familiar with the library and facet development. Despite, or maybe because of, the ar- how the traditional library classification works. This is not guably unfair assimilation implied in the use of the singu- usually the case with the non-purposive reader—as Ørvig lar term “reader” or “user” (similiar to the discussion of states, “the planless reader and the browser, who in fact the singular term “public” in other knowledge organiza- constitute great percentage of the public library clientele, tion systems discussed in, for instance, Olson 1997 and have little or no use for the perfect catalogue” (Ørvig Olson and Schlegl 2001), the existence of several groups 1955, 224). Traditional library classifications and catalogs of users is usually recognized among reader-interest clas- are therefore considered to be more adequate for pur- sification practitioners. Indeed, the inclusion of those us- posive readers (who are in the minority), while non- ers with a high-grade of expertise (as well as classifica- purposive users (who are in the majority) are assumed to tionists) is considered a handicap for the overall outcome have access to the materials by browsing. The distinction of the system. In this vein, another key opposition in the between purposive and non-purposive readers is thus reader-interest classification literature is the division be- sometimes defined by the relationship between access tween purposive and non-purposive reader—a distinction methods and the knowledge of the system. usually made to justify the convenience of this type of On the other hand, the traditional approach is consid- classification for the (non-purposive) public. Such an ap- ered to fail to meet the needs of the real public, with proach is reflected in various studies, such as Jones some authors arguing that the majority of users in most (1971), Donbroski (1980), Ainley and Totterdell (1982), reader-interest cases are browsers or non-purposive read- Betts (1982), and Morson and Perry (1982). ers (Ainley and Totterdell 1982, 121). A reader interest perspective thus justifies a non-purposive friendly scheme 4.0 Definition of purposive reader and which is able to meet the majority’s needs, as opposed to non-purposive reader the traditional practices represented by the DDC that only meet the librarian’s design needs. Authors in favor As in the case of the term ‘reader-interest classification,’ of reader-interest classification generally assume that tra- the terms ‘purposive reader’ and ‘non-purposive reader’ ditional library classifications, such as Dewey, fit the pur- were not always used exclusively within the literature. Near posive readers’ information-seeking patterns because they synonyms of ‘non-purposive reader’ found in the literature provide a specific location for a specific title. Once the include ‘non-specific reader,’ ‘casual reader,’ ‘plan-less user’s information need is translated into a query through reader,’ and ‘browser.’ Some near synonyms of ‘purposive the catalog, the exact location of the title is provided by reader’ in the literature include ‘specific reader,’ ‘systematic the call number and by following a systematic arrange- reader,’ ‘subject-orientated reader,’ and, in one case, there ment of the books on the shelves. are some more general and ambiguous terms such as Yet, the assumption that the needs of the purposive ‘scholar’ and ‘student’ (Ørvig 1955, 224). reader are best met by Dewey or any other traditional li- In general, a purposive reader is defined as a searcher brary system also supposes that the needs of the non- who accesses a book in the shelves, knows exactly where it purposive reader are best met by alternative classification is located, after using an indexing tool that he or she knows systems. These alternative classifications were given the well (such as a library classification system). A non- term reader-interest classification following the assimilation purposive reader, on the other hand, is usually defined as a of non-purposive reader’s interests to reader interests (as a searcher with no specific title in mind and who does not singular and homogeneous group) and the assumption that necessarily have any training on the library classification this group was in the majority. One of the problems here system or library use. Sometimes, the definition of non- would be the attempt to characterize the binary pur- purposive readers also includes what Mary Ørvig called the posive/non-purposive reader as a dichotomy. Theoretically, 106 Knowl. Org. 40(2013)No.2 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification any kind of reader can access information equally by which it was referred in the reader-interest classification browsing or by direct access, even if the majority of access literature. Outside of this moment, the concept of sub- or preference is realized by only one method. However, ject departmentalism changed so rapidly that it could not here it should also be noted that if the main characteristic really be considered a reader-interest classification or of the distinction between purposive and non-purposive even related to any type of library classification. It could readers is the level of knowledge of the system—and not be argued that the only point in common between this the method of access—then these groups would in effect concept and the reader-interest classification movement be mutually exclusive. But even in this case, the distinction would be how both apply a different scheme for the user- should be made between Dewey users and non-Dewey us- oriented divisions. However, as with the case of the ‘two- ers, instead of browsers and Dewey users—it is reasonable tier arrangement,’ the idea of splitting the collection into to think that every reader is able to browse without training several departments without any “user-oriented” section although not every reader is able to use Dewey without remains beyond the parameters of this study. training. More recently, outside the context of the reader- It is commonly agreed that reader-interest classifica- interest classification, other authors have studied the dis- tions—as alternatives to the established and so-called tinction between purposive readers and non-purposive “non-friendly” standards such as Dewey—were first used readers such as Ágústa Pálsdóttir, who has distinguished in the late 1930s at the Detroit Public Library in the between purposive information seeking and information United States. Reader-interest classification thus first encountering (Pálsdóttir 2010). came to light some 60 years after the publication of the first edition of the Dewey Decimal Classification (then called 5.0 History and origins of reader-interest A Classification and Subject Index for Cataloguing and Arranging classifications Books and Pamphlets of a Library) and only few years after the publication of the second edition of its European ad- Thus, the origins of reader-interest classifications are aptation, called for the first time “Classification Décimale linked to the inadequacy of traditional library classifica- Universelle”—Universal Decimal Classification. Reader- tion systems to primarily cater to that majority of non- interest classifications therefore arrived just as the UDC purposive readers, “the reader.” However, the origin of was being promoted as a universal tool for classification, the reader-interest classification concept in the timeline is and the DDC in the United States came to be regarded as not completely obvious according to this criterion, since, the most adequate standard for libraries. as pointed out before, the term reader-interest classifica- This second edition of the UDC, published in the years tion covers a wide range of concepts that display some 1927-1933, emphasized the “universal” nature and purpose common characteristics not all of them related to the of the library classification as a one-fits-all scheme—to the classification system itself. While these concepts were extent that its competitors sometimes felt that their share usually practiced by a single—but geographically dis- of the market was under threat. On the other hand, since persed—movement, it is somewhat difficult to establish a the first edition of the UDC was originally intended as a unique time line for reader-interest classification as a tool for Otlet and La Fontaine’s “Universal Bibliographic whole. Indeed, the variety of concepts within the term, Repertory,” it was commonly and mistakenly believed or make it difficult to trace their emergence—that is from promoted by some librarians that the UDC could be only previous forms to a stage that could be “considered” to useful for the organization of bibliographic materials, be a reader-interest classification (according to its charac- while the DDC would be more useful for libraries. teristics). As such, it is not only a challenge to identify the As Indian librarian and theoretician Ranganathan origin of reader-interest classifications as a movement, pointed out, every adaptation that departed (or evolved) but also to identify the origin of some of the individual from the “one good custom” that is DDC was also terms covered by the umbrella term. branded as “heresy” (Ranganathan 1967, 528). Indeed, at One example of this problem is the concept ‘subject the time that these observations were made, the division departmentalism’—the organization of materials sepa- between “General classification vs. Specialist classifica- rated by subject areas—that was at one point included in tion,” i.e., the division between classifications for libraries reader-interest classification movement. According to Mi- and classifications for knowledge and the sciences, be- chael A. Overington, the first experiments with subject came a subject of dispute. The UDC, Bliss Classification, departmentalism in public libraries were made in Chicago and Ranganathan’s Colon Classification (along with every in 1893 but were first fully developed in Cleveland in other classification that was either more faceted or con- 1925 and in Los Angeles in 1926 (Overington 1979). sidered more complex than the DDC) were accused by Considered as a whole, this concept might only be con- advocates of the DDC of not being adequate for public sidered a reader-interest classification at the exact point in libraries—in other words, of being specialist classifica- Knowl. Org. 40(2013)No.2 107 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification tions. That being said, the division between specialist and troit Public Library, came up with the idea of a classifica- general classification were very different from later argu- tion scheme that offered an alternative to the classifica- ments that established a connection between general clas- tion of subjects represented in the DDC. In his proposal, sifications (DDC) and purposive readers, and between Ulvering stated, “For some time I have wondered specialist classifications (adaptations or alternative classi- whether our popular book lending service as organized fications) and casual readers. on traditional lines is pointed directly enough toward our In fact, while library classification practitioners seemed service objectives; that is, whether the organization of to be divided between DDC adepts and “changers,” it is our circulating units is adapted to the function we are try- also true that library classifications seemed to change ing to fulfill” (Rutzen 1952, 479). Others would later echo along either theoretical or practical lines in libraries (al- this as “[to] classify not by subject but by patrons’ reading though it has to be acknowledged that both views contain inclinations” (Woodford 1965, 119). theoretical and practical aspects). A few years later, when Ulverling became a full librar- It might also be argued that library classification prac- ian at Detroit, his idea was partially adopted in a pilot ex- titioners were divided according to their attitude towards periment at the Main Library in 1941, in what was called change. Those in favor of DDC, naturally enough, mili- the “Browsers’ Alcove” in the Open Shelf Room at the tated against the faceted nature and complexity of the Main Library. The adoption of this experiment is consid- new schemes (such as UDC, Bliss or Colon Classification). ered to be the first time that reader-interest classification “Changers,” on the other hand, sought to modify the was applied. The original experiment at Detroit was ap- DDC but never actually made claims against the universal plied to a collection of about three thousand books and stance of classifications. According to Ranganathan, dis- was composed by the following headings: Background agreement between the two groups centered on concerns Reading (Classics, Art, Music, Belles Lettres), Everyman’s over the inadequate foundation of old schemes, the nota- Affairs (Current National Problems), World Today-World tional plane, minority interests (and how multiple facet Tomorrow (The International Scene), Personal Living- orders could be useful in different contexts), overly hasty Home and Family (Family Relations, Maintenance of solutions, and the non-recognition of class number as a House and home), Work and Play (Crafts and Hobbies), proper name (Ranganathan 1967, 529). Adventure (Mostly Travel, geographical and scientific ex- Ironically, some of the arguments employed by those ploration), Bright Side (The Light, the Gay, the Humor- in favor of DDC were also used by a more recent group ous), Industrial Era (Men, Machines, Mass Production and of “changers”—pro-reader interest classifications— Its Effects), Human Experience (Biography, and Some against the DDC. This group was associated with the Types of Travel and History), Other Places (Travel), and practical transformation of library classifications—most Exploring Science (Application of Modern Science). systems, after all, were developed on the basis of obser- The public responded positively to this test, and the vation and daily experience of library users’ needs. entire bookmobile service was rearranged according to All cases were also later influenced by previous experi- this system in 1945. From December 1948, the system ences of other libraries and other reader-interest classifica- was extended to both old and new branches in Detroit, tion cases (such knowledge mainly being transferred via and, by 1952, the newly established branches of Edison, mouth-to-mouth, professional courses, or articles in pro- Wilder, and Hubbard had been categorized following the fessional journals). However, if reader interest classifica- scheme. By 1952, some of the headings included in the tions followed a genealogy or transformation process, we final scheme were: Background Reading (includes Phi- have to decide upon the moment in which this concept losophy, History, Great books, Belles Lettres, and classics emerged as the predecessor of future moments, i.e., we in all fields, “arranged in one alphabet by author”), Cur- have to decide upon the first properly named example of rent Affairs (Foreign Relations, National Problems, Labor reader interest classification. And in the case of the reader- and Capital, March of Science), Family Life (Marriage interest classifications, this point is commonly agreed to be and Family, Child Care, Health, Country Living), Your the Detroit Public Library in 1941 (Rutzen 1952, 479; Ør- Home (Planning, Decoration, Management), The Arts vig 1955, 224; Ainley and Totterdell 1982, 9). (Painting and Sculpture, Theater, Poetry), and People and Places (Our Neighbors, In Many Lands) (Rutzen 1952, 6.0 The first reader-interest classification: 480). Once this first attempt was considered successful, Detroit Public Library the scheme was expanded to 12 headings and applied to other branches. Four collections which moved into new The roots of the first reader-interest classification, im- buildings—the Elisabeth Knapp, Sherwood Forest, Lin- plemented at the Detroit Public Library, date to 1936, coln, Jefferson and Jessie Chase branches—were com- when Ralph A. Ulverling, then associate librarian at De- pletely classified according to the system, and several old 108 Knowl. Org. 40(2013)No.2 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification branches were partially reclassified along the same lines 67). Its legacy however was significant: not only were simi- (Rutzen 1952, 479). lar schemes adopted in other library systems in the United By 1955, the final version of the scheme was a combi- States (such as the Boston Public Library), but they were nation of 14 “subject sections” (categories of interest for also presented to a European audience (Ørvig 1955). the browser) and “information sections” with fields such While it is commonly accepted that reader-interest clas- as “content,” “alternatives,” and “purpose.” Each section sification is a unique concept, Rutzen claimed that some of was subdivided into subheadings and represented by an al- the ideas for the Detroit Public Library project came from pha-numerical notation. Subject sections were intended to library and information science theorists such as Douglas serve readers with specific needs, while information sec- Waples (1937), Lowell Martin (1940), and John Chancellor tions contained factual material and textbooks for answer- (who in 1930 briefly outlined a scheme for shelving books ing specific questions. This version of the scheme was by reader interest) (Rutzen 1952, 479; Rutzen 1959, 55). proudly made public by the Detroit Public Library (1955). The project also drew upon experiences of librarians in the In the introduction, and subsequent editions and publica- United States which could be considered as the direct pre- tions (e.g. Rutzen 1959, 55), the system was presented as cursors to the first reader-interest classification. Among follows: these librarians was Ralph Munn who established the Pub- lic Affairs Room: “It is our hope that we can present books This is a book arrangement planned to fit the needs dealing with public questions so effectively that even the and uses of the greatest number of people. It recog- uninterested reader will be attracted to them” (Munn cited nizes the variety of reasons prompting people to by Rutzen 1952, 479). There were also librarians such as come to the library. It is not a classification of the Helen D. Marvin from West Park Branch in Cleveland fields of knowledge but a shelving arrangement Public Library who talked about the attention paid to the based on broad areas of interest which relate them- reader-interest area Home and Family Living. In words of selves to the everyday needs of people. These broad Rutzen, “Our experiment seems not to be so different in areas have been designated as interest categories. ideas as in the extent to which we are applying them.” The They are subdivided by a varying number of sub- practical influence of reader-interest classification led to its headings, depending on the type of category and the transmission from library to library and still remains size of the collection. Some categories are browsing among its agreed characteristics. sections for the general reader; others are subject groupings aimed at a particular use by the reader. 7.0 Transformation of reader-interest classifications during the 1970s: At this time, there was no theoretical division between the user-centered revolution peak casual readers and purposive readers, although Dewey was retained in most old branches that adopted the new Following the experience at the Detroit Public Library, system and was eventually dropped in smaller collections. reader-interest classifications became particularly popular There was thus a well-established belief that there was with many public libraries during the late 1970s. In part, only one kind of reader and that browsing would be im- this can be explained by the fact that the user-orientation proved by meeting all readers’ specific needs. On the movement reached its peak at this time—reflected by the other hand, there were multiple reasons why Dewey was fact that it started to gain acceptance within library and retained in many of the collections: first, catalog cards information science and information organization around were produced at the central library, carrying the Dewey 1970, as pointed out by Hjørland (2007, 3) and Nahl number; second, all experiments could be undone if re- (1996; 2003). sults were not considered positive (a “certain safeguard if Another factor in their popularity might also have our experiment should prove not to be workable” been a general desire at that time to experiment with (Rutzen 1952, 481)); third, it was difficult to have the col- change; the oil crisis of 1973 and the subsequent reces- lections rearranged manually. This was also a problem for sion meant that libraries were forced to seek out alterna- the different versions of the scheme, and something that tive sources of funding and to develop innovative pro- somehow contributed to the idea that reader-interest jects and commercial-oriented practices in order to classifications were only adequate for smaller collections. maximize their resources. As a consequence, some librar- Eventually the scheme in Detroit was dropped due to ies started to look towards bookstores and commercial the lack of universalization and centralization, which practices for solutions and were thus influenced by some meant a waste of resources (Ainley and Totterdell 1982, 9); of their information organization practices (and in some this was reportedly related to the cumbersome and expen- cases these libraries ended up adopting some of these sive need to reclassify and relocate materials (Hyman 1982, practices). Knowl. Org. 40(2013)No.2 109 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification

The relationship between the user-centered movement reader-interest classifications to be the best of alterna- and the development of library standards has been recently tives. In defending Dewey, Trotter (1984, 1) stated that: studied by Gretchen L. Hoffman (2009, 633). According to Hoffman, library and information science standards started The recent obsession with ‘reader interest treat- to move toward the user-centered paradigm in research ment’ is at base nothing more than broad enumera- and practice with contributions from researchers such as tive classification taken to extremes. All this ap- Paisley (1968), Allen (1969), and Zweizig (1976) shifting proach does is to set up a small number of very the paradigm within library and information science from broad disciplines, and then to ignore, more or less, systems and standards to users. any principles of subdivision within them. I per- This interest in user-centered practices and theories sonally feel that this is something of a cop-out, also affected the way that libraries were physically organ- leaving the reader with most of the work of locat- ized and a wide range of practices and arrangements ing the sort of material he or she requires. A regu- were embraced under the reader-interest or user-centered lar classification, with principles of division and umbrella. Among the main studies conducted on library subdivision, and backed by a good alphabetical in- classification from a user-centered perspective were dex, is to my mind far superior. works by Groombridge (1964), Luckham (1971), and Taylor and Johnson (1973) (see Ainley and Totterdell In addition, Trotter (1984, 3) also claimed that the only 1982). As pointed out by Ainley and Totterdell, most of possible challenger to Dewey might have been Bliss Classi- these works—and those similar to them—mainly focused fication 2 (BC2) but that this had failed: “I also feel that on non-fiction collections. even with these faults Dewey offers a better deal than any One of the first proponents of reader-interest classifi- alternative—and that includes reader interest arrangement. cation in the UK is considered to be Archibald William The only possible challenge is Bliss 2 but I doubt it.” McClellan, Chief Librarian of Tottenham. From 1949 to On the other hand, Arthur Maltby, who was more 1970, McClellan studied a wide range of concepts related critical of the DDC, also expressed his disappointment in to the social and pragmatic role of libraries within society the BC2. Concerning the reader-interest classifications, and the way libraries could serve the community. One of Maltby (1984, 5) stated, “One public library alternative his central concerns was the reader, and McClellan com- has been reader interest arrangement, which has had a mitted himself to finding the best way to arrange the col- mixed public reception and varying degrees of success … lection in order to meet the reader’s interests (McClellan there are some Scottish examples of the phenomenon, 1973), with a particular focus on the integration of stock for instance in Falkirk District (Bo’ness), Glasgow (Cas- and the division of the collection (or two-tier arrange- tlemilk) and Renfrew District (Ferguslie Park). Headings ment). In subsequent years, the experiment conducted by in the last name include ‘Sexy Books.’” On the other Tottenham library with reader-interest arrangements was hand, Maltby had also praised Ainley and Totterdell’s an important source of inspiration for many other librar- work on reader-interest classifications—a recognition that ies all across the country. surprised both of these authors because of his work on classification. 8.0 Decay of non-fiction reader-interest Some commentators have however argued that the ret- classifications during the 1980s and 1990s rospective classification and the degree of training that systems such as BC2 require could impede their adop- During the 1980s, there was a loss of interest among tion. As Hjørland (2007, 8) pointed out: public libraries in applying reader-interest classifications to non-fiction. Despite this, the philosophy behind the It is somewhat ironic than the most used tool for reader-interest classification movement was still consid- classification in libraries today is the DDC first pub- ered relevant and debate continued as to the validity and lished 1876. More that hundred years of research adequacy of Dewey. One example of this debate can be and the development of other kinds of knowledge found on Arthur Maltby and Ross Trotters’ interesting organizing systems has not resulted in making DDC discussion in the Catalogue and Index journal of 1984 on obsolete. For example, the BC2 is generally consid- the adequacy of Dewey and on reader-interest classifica- ered theoretically more advanced, but has difficul- tions as possible solutions to problems faced by contem- ties being used in practice. The main reason may be porary libraries (“Dewey as an Asset” (Trotter 1984) and that most of the English-language books bought by “Dewey Decimal Classification: a Liability?” (Maltby 1984)). a given library are pre-classified with the DDC by While both authors recognized that Dewey might not the Library of Congress. Another reason may be be totally adequate for libraries, neither did they consider that they are not considered user-friendly because 110 Knowl. Org. 40(2013)No.2 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification

users have to learn certain principles. It is, however, rowed’ from a bookstore model because it is thought thought provoking that classification systems de- people prefer to browse and choose books by genre veloped later and generally thought more advanced rather than alphabetically by author.” In practice, the are not able to compete efficiently. concept has been noticeable popular in the UK and the Eastern States of Australia since the late 1980’s (Maker In some ways, Hjørland’s observations on the BC2 might 2008b). A quick survey of the literature reflects the de- also apply to reader-interest classification, given that the velopment of this popularity; initially, fiction categoriza- problem of reclassification concerns all systems. Al- tion was either secondary or non-existent in the reader- though the main problem with the adoption of systems interest classification experiments (Sawbridge and Favret such as BC2 might be one of document reclassification, 1982; Wijland 1985) or was given the same importance as the argument of user-friendliness seems to have been the non-fiction (Augenanger 1981; Venter 1984; Sivulich decisive factor among libraries in either adopting new 1989). The central role given to fiction categorization can schemes or continuing with some sort of old system (in- be seen in the works of Harrel (1985), Baker and Shep- deed, in reader-interest classifications, user-friendliness herd (1987), Borden (1987), Langhorne (1987), Baker was often the only factor involved in the decision). (1988b), Kellum (1989), McGrady (1990), Scott (1995), Although it suffered a decline during the 1980s, non- Saricks (1997, 2006), and Scilken (1998). fiction reader-interest classification was still practiced in It is perhaps worth emphasizing that the increased some libraries during the first half of the 1990s. Among popularity of the concept during the 1980s and 1990s co- them were the De Beauvoir Junior School Library in the incided with the declining interest in reader-interest clas- UK (Bridgwater 1990) and the Glasgow City libraries sification of non-fiction. The concept of reader-interest (Miller 1992) while other instances can be found in Jac- classification kept on developing at its own pace during quelyn Sapiie’s work (one of the very last bibliographic these decades. As Sapiie (1995, 143) pointed out, “recent retrospectives on reader-interest classifications found in literature since 1980 indicates that there is some variety the literature) (Sapiie 1995). Although De Beauvoir Junior of the schemes in use today in the United States, Britain School Library pointed out that the previous experience and many other countries as an alternative to the major of Brent and Camden libraries in the 1980s had influ- classification systems.” However, it should be noticed enced their own experiment in reader-interest classifica- that the term reader-interest classification was also aban- tion, their interest in the advantages of this system were doned for a more commercially orientated vocabulary as practically confined to the children’s section, leaving the new terms and ideas from bookstores and commercial adult section with a shadow of a system that had been spheres became increasingly popular in the United States. designed to provide for the whole collection: “Readers One example of this trend can be found in Mary Jo will be aware that such schemes have been very success- Langhorne’s work, where bookstore approaches in librar- ful, particularly in the organisation of children’s libraries ies were compared with previous reader-interest experi- (though many adult users also express relief at the intro- ments in the UK in the areas of fiction and non-fiction duction of an easier system)” (Bridgwater 1990, 53). categorization, labeling and signage, visibility, and physi- The De Beauvoir and Glasgow City libraries might be cal location and display (Langhorne 1987). Although considered two of the very last non-fiction reader- bookstore techniques seemed to be a more commercial interest classifications cases per se reported in the litera- and more appealing concept in the 1990s than ‘reader- ture, the former denominated under the system of broad interest classification’ or ‘alternative arrangement,’ it was categories (or subject categories), and the latter as an al- virtually alike in all but name and the new techniques did ternative arrangement. In some ways both cases were the not add anything new to the previous concepts. last of an era in which the concept of reader-interest Relabeling of the reader-interest concept as a “book- classification captured the attention of many librarians. store (bookshop) approach” did not only take place in the United States in the 1990s but also in countries that 9.0 Rise of the reader-interest classifications of followed the British tradition during the 2000s, such as fiction and bookstore practices since the 1980s the United Kingdom itself and Australia. While these approaches had many characteristics in common with the Coinciding with the popularity shifting from non-fiction reader-interest classifications (which had been very popu- reader-interest classification projects to fiction reader- lar during the previous decades in some of these coun- interest classification projects, a new terminology related tries), rarely did they include the terms ‘bookstore classi- to bookstore and commercial practices emerged in library fications’ or ‘classifications.’ classifications. As Richard Maker (2008a, 171) points out, For instance, the 2002 Audit Commission’s “Building the concept of fiction categorization was initially “‘bor- Better Library Services” report in the UK indicated several Knowl. Org. 40(2013)No.2 111 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification aspects that libraries could improve if they adopted book- in 2007, when Perry Branch Library in Maricopa County stores practices (Audit Commission 2002). Although the (Arizona) was presented in the literature as the first case Audit Commission’s report did not mention library classifi- of Dewey-BISAC switching in US public libraries. cation systems, it applied arguments that were almost iden- On May 30, 2007, The Arizona Republic announced that tical to those employed more than 20 years earlier during Perry Branch Library located in the town of Gilbert in the reader-interest experiments. While the term ‘reader- the Maricopa County Library District would be “the first interest’ was completely omitted from the report, refer- public library in the nation whose entire collection was ences to having “the user at the center” are clearly reminis- categorized without the Dewey Decimal Classification cent of the user-centered movement days. In fact, the Au- System” (Wingett 2007). Despite it not being exactly true dit Commission report borrowed several ideas from the that it was the first library in the US to drop Dewey, Perry user-centered movement in order to describe their own vi- Branch Library did mark a milestone in the field as the sion of the ideal library (Audit Commission 2002, 25): first public library in the US to adopt the book industry standard BISAC instead of Dewey as the classification Services need to be designed with the user at the system for organizing the collection. centre—built around a realisation that people use Perry Branch Library opened in June 2007, and, for their services out of choice—and a clear under- the organization of its 24,000-item collection, 50 BISAC standing of the services and experience people headings were used instead of Dewey. This idea had been want: 1. Libraries need to provide the books and in- previously devised by director Harry Courtright in 2005 formation services people want—or people will and was implemented in 2007 by adult services coordina- have no reason to come. 2. These services need to tor Marshall Shore with the opening of the Perry Branch, be easily accessible, in terms of opening times and although Nanci Hill, Head of Readers’ Services at the location—or many potential users may be put off Nevins Memorial Library in Methuen, Massachusetts, using them. 3. They need to provide a positive and stated that the beginning of the concept was a pilot plan welcoming experience for the user, in terms of the in two libraries in Delaware County (Pa.) in 1988 (Hill environment and how easy services are to use—or 2010). According to Amy Wang (2009) of The Arizona people will choose to go elsewhere. 4. And, people Republic, the conversion plan for the system in the Perry need to be aware of the full range of services on Branch took nearly five years, although county officials offer and how they can get them. say that by 2009 it only took from one to two months to make a library Dewey-less. Other interesting concepts used in the Audit Report were After this case, several other libraries in the United the terms ‘reader development’ and ‘reader development States and abroad have looked to Maricopa as a source of schemes.’ These terms basically coincided with the practi- inspiration for the remodeling and new opening projects cal applications of reader-interest classifications to the of their systems. As well, they started to consider BISAC collection development and stock control. This latter as- and other bookstore-like techniques as a good alternative pect was also pointed out by many reader-interest classifi- to the traditional practices in knowledge organization. De- cation advocates, such as Ainley and Totterdell (1982), as spite the omission of every reference to the old and failed being one of the main advantages of reader-interest clas- reader-interest classification projects in the discourses of sifications in the past. It could be argued that these ad- adoption of BISAC in libraries, with the exception of a vantages seem to have remained valid over time. brief mention in an editorial of School Library Journal (Kenney 2007), these cases of DDC-BISAC switching for 10.0 Current cases of implementation of BISAC in non-fiction have also been studied as new cases of reader- public libraries as new cases of reader-interest interest classifications (Martínez-Ávila 2012). classifications? 11.0 Conclusions Regarding non-fiction, one further step in the concept of reader-interest classification might arguably be the recent In accepting that the concept and terminology of reader- experiments of US public libraries adopting the scheme interest classifications has transformed and changed over of the American book industry, BISAC, for the classifica- time, despite its variations, it still shows an agreed-upon tion of the collection. Since the second half of the core of characteristics shaping its meaning and allowing 2000’s, several public libraries in the United States have its conceptualization. As pointed out by Martínez-Ávila been experimenting with BISAC as an alternative classifi- (2012), some of these characteristics of reader-interest cation system to Dewey for non-fiction. These kinds of classifications include: dropping traditional practices like experiments gained major attention by the media mainly Dewey; using alphabetical/natural language categories for 112 Knowl. Org. 40(2013)No.2 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification the physical arrangement of materials; organizing and gov.uk/SiteCollectionDocuments/AuditCommission representing knowledge according to topics of interest Reports/NationalStudies/ACKLibraries.pdf (Accessed instead of academic disciplines; carrying out the classifi- 20 August, 2010). cation process at the book selection service instead of Augenanger, Petra. 1981. Der ‘nahbereich’ aus sicht des the cataloging service and all that difference implies; and benutzers [The ‘immediate concern’ area from the accepting the influence of bookstore practices in such di- user’s point of view]. Buch und bibliotek 33 Jan.: 46-8. verse aspects as guiding, signage, and display. Among the Baker, Sharon L. 1988a. Designing libraries to meet the alleged advantages of reader-interest classifications pro- needs of browsers. The unabashed librarian 67: 3-5. jects are: meeting the user’s needs (by gathering together Baker, Sharon L. 1988b. Fiction classification schemes: an materials of interest that had been previously dispersed experiment to increase use. Public libraries 26 no. 3: 75-7. by Dewey), reversibility of the experiments, the possibil- Baker, Sharon L. and Shepherd, Gay W. 1987. Fiction ity of putting different changes and innovations into ef- classification schemes: the principles behind them and fect at one time (although sometimes these are impossi- their success. Reference quarterly 27: 245-51. ble to separate or even identify clearly and not always re- Betts, Douglas. 1982. Reader interest categories in Surrey. lated to the classification scheme), the cost saving (includ- In Ainley, Patricia and Totterdell, Barry, eds., Alternative ing aspects such as the acquisition of cheap formats), arrangement: new approaches to public libraries stock. London: greater shelving flexibility, improved adequacy for brows- Association of Assistant Librarians, pp. 60-77. ing, increase in circulation, and a good method of identi- Borden, William Alanson. 1987. On classifying fiction. fication in the most demanded areas for stock control The unabashed librarian 63: 25-6. purposes. Finally, some of the detected shortcomings of Bridgwater, Sue. 1990. Out of the doldrums and into the reader-interest classifications include: problems of reclas- curriculum: De Beauvoir Junior School Library. The sification according to a non-standardized scheme, inade- school librarian 38: 53-4. quacy of reader-interest classifications for all sized librar- Detroit Public Library. 1955. The reader interest book ar- ies (specifically for bigger libraries), the creation of other rangement in the Detroit Public Library. Detroit: Home “distributed relatives” when rearranging knowledge ac- Reading Services, Detroit Public Library. cording to topics of interest, the possibility of low quality Donbroski, Lyn. 1980. Life without Dewey. Catalogue & in the nature of the categories reflecting readers’ inter- index 57 Summer: 3-6. ests, and a lack of centralization and universalization, Dunkin, Paul S. 1969. Cataloging USA. Chicago: American which is widely regarded as the main cause that led the Library Association. first reader-interest classification cases to fail. As for the Groombridge, Brian. 1964. The Londoner and his library. terminology, historically, there has not been a single, uni- London: Research Institute for Consumer Affairs. fied terminology on the concept of reader-interest classi- Harrel, Gael. 1985. The classification and organization of fication in the literature. During the different stages of adult fiction in large American public libraries. Public the concept its terminology was changed from the origi- libraries 24 no. 1: 13-4. nal term “reader-interest classification” used in the case Hill, Nanci Milone. 2010. Dewey or don’t we? Public librar- of the Detroit Public Library (and first expressed in the ies 49 no. 4: 14-20. literature in 1952) to the more commercial oriented ter- Hjørland, Birger. 2007. Arguments for ‘the bibliographi- minology of the 1990s and after. However, although the cal paradigm.’ Some thoughts inspired by the New terminology and some of the reader-interest classification English Edition of the UDC. Information Research 12(4) features have varied according to the literature during the paper colis06. Available at http://InformationR.net/ period of 1952-1995, the concept has also kept a core of ir/12-4/colis06.html (Accessed 11 March, 2011). common and agreed-upon characteristics, alleged advan- Hjørland, Birger. 2009. Concept theory. Journal of the tages and shortcomings throughout this time, i.e. there is American Society for Information Science and Technology 60: a continuity in the discursive formations that allow 1519-36. reader-interest classification to be discussed as a concept. Hoffman, Gretchen L. 2009. Meeting users’ needs in cataloging: what is the right thing to do? Cataloging & References classification quarterly 47 no. 7: 631-41. Hyman, Richard Joseph. 1982. Shelf access in libraries. Chi- Ainley, Patricia and Totterdell, Barry. 1982. Alternative ar- cago: American Library Association. rangement: new approaches to public library stock. London: Jones, Ken H. 1971. Towards a re-interpretation of pub- Association of Assistant Librarians. lic library purpose. New library world 73 no. 3: 76-82. Audit Commission. 2002. Building better library services. Audit Kellum, Debra. 1989. Fiction separation. The unabashed li- Commission. Available http://www.audit-commission. brarian 71: 3-4. Knowl. Org. 40(2013)No.2 113 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification

Kenney, Brian. 2007. Desert storm. School library journal 53 mation science, Second Edition. New York: Marcel Dek- no. 8: 9. ker; Inc., pp. 3028-42. Langhorne, Mary Jo. 1987. Marketing books in the school Olson, Hope A. 1997. Thinking professionals: teaching library. School library journal 33 no. 5: 31-33. critical cataloging. Technical services quarterly 15 no. 1/2: Luckham, Bryan. 1971. The library in society: a study of the pub- 51-66. lic library in an urban setting. London: Library Association. Olson, Hope A. and Schlegl, Rose. 2001. Standardization, Maker, Richard. 2008a. Finding what you’re looking for a objectivity, and user focus: a meta-analysis of subject reader-centred approach to the classification of adult access critiques. Cataloging & classification quarterly 32 no. fiction in public libraries. The Australian library journal 2: 61-80. 57: 169-77. Ørvig, Mary. 1955. The reader interest arrangement, an Maker, Richard. 2008b. Reader centred classification of American shelving system with future. Libri 5: 223-32. adult fiction in public libraries. Australasian public librar- Overington, Michael A. 1969. The Subject departmentalised ies and information services 21: 168-71. public library. London: Library Association. Maltby, Arthur. 1984. Dewey Decimal Classification: a li- Paisley, Williams. 1968. Information needs and sses. An- ability. Catalogue and index 72: 4-5. nual review of information science and technology 3: 1-30. Martin, Lowell A. 1940. The Purpose and Administrative Pálsdóttir, Ágústa. 2010. The connection between pur- Organization of Branch Systems in Large Urban Li- posive information seeking and information encoun- braries. M.A. thesis, University of Chicago. tering: a study of Icelanders’ health and lifestyle in- Martínez-Ávila, Daniel. 2012. DDC-BISAC switching as a formation seeking. Journal of documentation 66: 224-44. new case of reader-interest classification. Ph.D. dissertation, Prytherch, Ray. 1990. Harrod’s librarian’s glossary of terms Universidad Carlos III de Madrid. used in librarianship, documentation and the book crafts, 7th ed. Martínez-Ávila, Daniel, Olson, Hope and Kipp, Margaret. Worcester, U.K.: Gower. 2012a. New roles and global agents in information or- Ranganathan, Shiyali Ramamrita. 1967. Prolegomena to li- ganization in Spanish libraries. Knowledge organization 39: brary classification. Assisted by M.A. Gopinath 3rd edi- 125-36. tion. Asia Publishing House. Martínez-Ávila, Daniel, Kipp, Margaret and Olson, Hope. Reader, Den. 1982. User orientation in a Heartfordshire 2012b. DDC or BISAC: the changing balance between Branch. In Ainley, Patricia and Totterdell, Barry, eds., corporations and public institutions. Knowledge organiza- Alternative arrangement: new approaches to public libraries tion 39: 309-19. stock. London: Association of Assistant Librarians, pp. McCarthy, Austin. 1982. Burning issues: stock appeal in 101-18. Sunderland. In Ainley, Patricia and Totterdell, Barry, Rutzen, Ruth. 1952. Shelving for readers. Library journal eds., Alternative arrangement: new approaches to public librar- March 15: 478-82. ies stock. London: Association of Assistant Librarians, Rutzen, Ruth. 1959. A classification for the reader. In pp. 88-100. Eaton, Thelma and Strout, Donald E., eds., The role of McClellan, Archibald William. 1973. The reader, the library classification in the modern American library. Champaign, Il- and the book: selected papers 1949-1970. London: Clive linois: The Illini Union Bookstore. Bingley. Sapiie, Jacquelyn. 1995. Reader-interest classification: the McGrady, Amanda. 1990. Putting fiction back together. user-friendly schemes. Cataloging & classification quarterly The unabashed librarian 76: 6. 19 no. 3/4: 143-55. Miller, Andrew. 1992. Alternative arrangement in Glas- Saricks, Joyce. 1997. A-leaving Cheyenne. Public libraries gow City Libraries. Public library journal 7 no. 5: 131-3. May/June: 154-5. Morson, Ian and Perry, Mike. 1982. Two-tier and total: Saricks, Joyce. 2006. Thinking outside the genre and stock arrangement in Brent. In Ainley, Patricia and Dewey boxes. Booklist 102 no. 13: 64. Totterdell, Barry, eds., Alternative arrangement: new ap- San Segundo, Rosa. 2004. A new conception of represen- proaches to public libraries stock. London: Association of tation of knowledge. Knowledge organization 31: 106-11. Assistant Librarians, pp. 101-18. Satija, Mohinder Partap. 2004. A dictionary of knowledge or- Nahl, Diane. 1996. The user-centered revolution: 1970- ganization. Amritsar: Guru Nanak Dev University. 1995. In Kent, Allen, ed., Encyclopedia of microcomputers. Sawbridge, Lynn and Favret, Leo. 1982. The mechanics New York: Marcel Dekker, Inc., pp. 143-99. Available and the magic of declassification. Library association re- http://www2.hawaii.edu/~nahl/articles/user/user1to cord 84: 385-6. end_toc.html (Accessed 5 March, 2011). Scilken, Marvin H. 1998. Shelving genre fiction. The un- Nahl, Diane. 2003. The user-centered revolution. In abashed librarian 109: 9. Drake, Miriam A., ed., Encyclopedia of library and infor- 114 Knowl. Org. 40(2013)No.2 D. Martínez-Ávila, and R. San Segundo. Reader-Interest Classification

Scott, Paul. 1995. Wheels within wheels genre fiction in 09/09/09/20090909deweyless0909.html (Accessed 21 public libraries. Australasian public libraries and information February, 2010). services 8: 191. Waples, Douglas. 1937. People and print. Chicago: Univer- Sivulich, Kenneth G. 1989. Merchandising your library. sity of Chicago. Public libraries 28 no. 2: 97-100. Wijland, Hans Ban. 1985. Toegankelijker opstelling van Sykes, Alan. 1982. Categorization or how Dewey gains a informatie: di Duitse Dreigeteilte Bibliothek en de prefix and loses sanctity. Library association record 84: Britse alternative arrangement [A more accessible ar- 383-4. rangement of information: the West German Tripar- Taylor, John N. and Johnson, Ian M. 1973. Public libraries tite Library and the British alternative arrangement]. and their use. London: H.M.S.O. Bibliotheek en samenleving 13(7/8) July/August: 133-238. Totterdell, Barry. 1978. Public library purpose: a reader. Lon- Wingett, Yvonne. 2007. Gilbert Library to be first to drop don: Clive Bingley. Dewey Decimal. The Arizona republic May 30. Available Trotter, Ross. 1984. Dewey as an asset Catalogue and index http://www.azcentral.com/arizonarepublic/local/articl 72 Spring: 1-4. es/0530nodewey0530.html (Accessed 24 February, Venter, Trude. 1984. ‘n rangskikkingsmetode om die ge- 2010). gruik van nie-fiksie in openbare biblioteke te bevorder Woodford, Frank B. 1965. Parnassus on Main Street: a history [An arrangement system to stimulate the use of non- of the Detroit Public Library. Detroit, MI: Wayne State fiction in public libraries]. South African journal for li- University Press. brarianship and information science 52(4) December: 110-2. Zweizig, Douglas L. 1976. With our eye on the user: Wang, Amy B. 2009. Sun City Library embraces Dewey- needed research for information and referral in the less world. The Arizona republic Sept. 9. Available http:// public library. Drexel library quarterly 12 no. 1/2: 48-58. www.azcentral.com/community/westvalley/articles/20

Knowl. Org. 40(2013)No.2 115 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments

Knowledge Organization and Representation in Digital Environments: Relations Between Ontology and Knowledge Organization†

Carlos Henrique Marcondes

Fluminense Federal University, Department of Information Science, R. Lara Vilela, 126, 24210-590, Niterói, Rio de Janeiro, Brazil, phone 55 21 26299758,

Carlos Henrique Marcondes holds a PhD in information science and is associate professor at Fluminense Fed- eral University, Department of Information Science, Niterói, Rio de Janeiro, Brazil. His research interests include knowledge organization and representation in digital environments, use of semantic web technologies, information retrieval, and foundational issues.

Marcondes, Carlos Henrique. Knowledge Organization and Representation in Digital Environments: Relations Between Ontology and Knowledge Organization. Knowledge Organization. 40(2), 115-122. 47 references.

ABSTRACT: Knowledge organization faces the challenge of contributing to the management of the amount of knowledge produced and available in the Web environment. Computational ontologies are new artifacts for knowledge recording and processing and also one of the foundations of the semantic web; they pose new challenges to knowledge organization in clarifying its interdisciplinary relations and specific role within knowledge management disciplines. What are its relations to ontology? A draft of these relations is presented, obtained from authors who discuss foundational issues, with the aim of identifying the actual role of knowledge organization in the Web environment. While ontology discusses the ultimate nature of being, knowledge organization emphasizes additional practical issues unfolding all possible manifestations of that which is. A primary question is: how to seek information, how to be informed?

Received 15-12-2012; Revised 13-1-2013; Accepted 16-1-2013

† This paper is a revised and extended version of a previous one presented in the Brazilian ISKO 2011 Conference

1.0 Introduction tent in an innovative mode, different from current computational processing paradigm is the Semantic Web pro- General rules that are always applicable for the choice of posal. The Semantic Web will bring structure to the names of subjects can no more be given than rules with- meaningful content of Web pages, creating an environ- out exception in grammar. Usage in both cases is the su- ment where software agents roaming from page to page preme arbiter—the usage, in the present case, not of the can readily carry out sophisticated tasks for users. (Bern- cataloger, but of the public in speaking of subjects (Cut- ers-Lee, Hendler, and Lassila 2001, 2). ter 1904, 69). Nowadays, knowledge organization (KO) The importance of the Semantic Web vision to KO is faces the challenge of contributing to the management emphasized by Hjörland (2007, 371): “and of course in of growing stocks of knowledge records produced and particular the new concept considered by many the most available on the Web in order to enable their reuse and important frontier in knowledge organization: ‘the se- appropriation, as required by the emergence of the in- mantic web.’” The Semantic Web proposal is ambitious formation society. The Web poses new challenges to KO. with respect to KO. Berners-Lee, Hendler, and Lassila A comprehensive proposal aimed at organizing Web con- (2001, 2) write: 116 Knowl. Org. 40(2013)No.2 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments

The Semantic Web is not “merely” the tool for Faceted Classification Theory, and disciplines such as conditioning individual tasks that we have discussed formal and computational ontology. Texts by authors so far. In addition, if properly designed, The Se- with a focus on the foundations of these disciplines such mantic Web can assist the evolution of human as Guarino, Smith, and Guizzardi are analysed. Authors’ knowledge as a whole. claims related to foundational questions are contrasted as to the discipline objectives, the need, and the grade of Based on the vision of the Semantic Web, different theo- formalization, the rules for building taxonomies, the pro- retical and practical proposals have posed the question of posal of high level/foundational ontologies, and the as- how to assign semantics to Web content to allow programs sortment of properties. The aim is to identify the KO to process it more efficiently and thus help us with the specific role within the disciplines which deal with the enormous task of organizing and provide access to it. management of humanity’s stock of knowledge, mainly Within this context, ontologies are new computational arti- in digital environments inspired by the Semantic Web facts that can provide computational semantics to web proposal. The work is organized as follows. After this in- content, allowing programs, in addition to data processing, troduction, Section 2 discusses the aims of formal ontol- make inferences about this content. Different communities ogy, and Section 3 the aims of KO. Section 4 presents a have developed computational ontologies with varying de- discussion of possible methodological intersections and grees of success. However, with the exception of some complementary approaches between the two disciplines. cases, we are far from the vision proposed by the Semantic Finally, Section 5 presents conclusions. Web (Berners-Lee, Hendler, and Lassila 2001). Research in ontology has an interdisciplinary nature, in- 2.0 What are the aims of ontology? volving contributions from philosophy, logic, computer science, among other disciplines, and, increasingly, contri- For some time, issues related to the modeling of specific butions from KO, too, aimed at successfully facing the domains in computer environments have required inquir- trends previously posed. The relationship between these ies on the ultimate nature of the entities in these do- disciplines and the KO legacy due to the development of mains. This is a requirement to ensure correspondence bibliographic classifications is recognized in the current between these models and reality, thus enabling com- ontology literature (Smith and Welty 2001). Several authors puters to reason correctly on these models. To cope with in information science (Wersig and Neveling 1975; Sarace- these issues, contributions are needed from formal ontol- vic 1995) have stressed its interdisciplinary relationships ogy. The discipline applies principles of philosophical on- along the history of the discipline in an attempt to clarify tology in the sense of Husserl’s Logical Investigations to the its own scope. modelling of computer systems. Authors such as Today, the emergence of the Semantic Web proposal Guarino (1997, 1) define formal ontology as dealing with and of ontologies as new artifacts to record humanity’s “formal distinctions between the elements of a domain, knowledge pose new challenges to the comprehension of independently of their actual reality;” Guarino (1995, 5) KO relationships and their roles within the different cogni- claims that it deals with a priori distinctions: tive and knowledge management sciences. What should the relationships between disciplines of ontology and KO be? – among the entities of the world (physical ob- How important is ontology to KO and vice-versa? As jects, events, regions, quantities of matter; Epistemology discusses the problem of how we know and – among the meta-level categories used to model ontology discusses and formally describes the things we the world (concepts, properties, qualities, states, know, what does KO discusses? What is its role among the roles, parts. spectrum of disciplines engaged with the problem of large scale knowledge management and appropriation, especially Discussing the scope of formal ontology, Guarino (1995, in today’s Web environment? There is a need to clarify the 2) claims that “As such, formal ontology is a recent ex- nature of the relations with those disciplines in order to pression of traditional ontology, intended as the branch develop fruitful relations with them. How can such arti- of philosophy which deals with the a priori nature of real- facts, considered as bases of the Semantic Web, be useful ity,” i.e., a nature which is always present in reality, inde- to KO? What may be the contribution of ontology to KO? pendent of the domain considered. Stressing this particu- What may be the specific contribution of KO to the de- lar aspect of domain independence, Guarino and Guiz- velopment of computational ontologies? To what extent zardi (2006, 117) claim that: “in particular, so-called for- are both disciplines complementary? mal ontology is completely neutral for what concerns its This work develops an analysis and sketches the rela- domain of application.” tionships between KO, with an emphasis on Ranganathan’s Knowl. Org. 40(2013)No.2 117 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments

Formal ontology claims itself to be an autonomous or focus: it seeks, not predication or explanation, knowledge discipline having relationships with computer but rather taxonomy. science especially with artificial intelligence, information science, logic, and philosophical ontology. Relevant con- McGuinness (2003) defines a computational ontology as tributions have been made by researchers around forums an artifact where there are formally declared controlled such as IAOA (The International Association for Ontol- vocabularies, relations between classes, hierarchical class- ogy and its Applications), events such as FOIS (Interna- subclass relations, properties of classes, value restrictions tional Conference on Formal Ontology in Information at the class level, inclusion of individuals to classes, dis- Systems)—which is in its 7th edition in 2012—and the jointedness between classes, arbitrary logical relations be- Journal of Applied Ontology. tween terms, and inverse and part-whole relations. As From a historical perspective, disciplines such as artifi- stated before (Smith 2006), the backbone of every ontol- cial intelligence have been changing their focus as ogy is a taxonomic structure, in other words, class- Guarino (1995, 625) stresses: subclass and class-instance relations. Thus, according to those authors, foundational issues related to the applica- AI researchers seem to have been much more inter- tion of ontology to modeling systems and knowledge ested in the nature of reasoning rather than in the representation in digital environments are: nature of the real world. Recently, however, the potential value of task-independent knowledge bases – What exists? (or ‘ontologies’) suitable to large-scale integration – Which are the entities that exist in a specific domain? has been underlined in many ways. – How are they? – What are their differences? According to the author, the rise of computational on- – What are their similarities? tology, although it has roots in AI, shows a sharp differ- – Which entity is similar to which entity? ence between task-independent knowledge bases or on- – What are their properties? tologies, and reasoning processes, which were the focus of early AI. That seems a fundamental change in per- Ontological analysis aims at, as stated by Guarino and spective as it poses the question of how to develop a rep- Welty (2009, 8), identifying all logical consequences of resentation of (or to model, according to Le Moigne the choices made when modelling a domain, and enabling 1990, a specific domain which is independent of any par- inferences which are logically and ontologically correct. ticular computational application). Some important methodological contributions of onto- Barry Smith and Christopher Welty (2001, 4) claim logical analysis to domain modelling are: that: “Philosophical ontology is the science of what is, of the kinds and structures of objects, properties, events, – identifying metaproperties of the properties occurring processes and relations in every area of reality”; in addi- in a domain, such as those which ensure Identity, De- tion, Smith (2006, 2) claims that: pendence, and Integrity criteria to individuals involved in those properties (Guarino 1997); It seeks not explanation but rather a description of – identifying and analyzing roles as features which can reality in terms of a classification of entities that is add semantics and precision to modelling (Masolo et exhaustive in the sense that it can serve as an an- al. 2004); and, swer to such questions as: What classes of entities – identifying types of formal-ontological relations are needed for a complete description and explana- (Smith and Grenon 2004). tion of all the goings-on in the universe? Or: What classes of entities are needed to give an account of According to different authors (Guarino 1998, Guizzardi what makes true all truths? Or: What classes of en- 2005), formal ontology should provide the bases to the tities are needed to facilitate the making of predic- development of the so-called application ontologies, tions about the future? . computational artifacts (Guarino 1998) that model a specific application and are tied to computational systems Smith (2006, 5) makes a significant distinction between which this author calls “ontology-driven systems.” Guari- philosophical ontology and science as he says that: no (1998) proposes that ontologies may be developed in different levels of generality: top-level or foundational Philosophical ontology is a descriptive enterprise. It ontologies that provide very general concepts like space, is distinguished from the special sciences not only time, matter, object, event, action, etc., which are inde- in its radical generality but also in its primary goal pendent of a particular domain; domain ontologies and 118 Knowl. Org. 40(2013)No.2 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments task ontologies of specialized concepts provided by top- had always been a preoccupation of previous classifica- level ontologies providing concepts to a generic domain tion systems, although systems such as Dewey and even or task like medicine and diagnosis; and application on- UDC are both flexible enough to cope with the emer- tologies that provide concepts which are specializations gence of new disciplines or subjects; indeed, Foskett of the former to a particular application. (1996) proposes hospitality as a quality criteria for classi- The efficiency of such systems depends on the accu- fication systems. racy of these application ontologies. The importance of At the same time, the Analytico-Synthetic Classifica- accurate knowledge representations in digital environ- tion must 2) account for representing compound subjects ments is stressed by Davis, Shrobe, and Szolovits (1993, as they appear in knowledge records—books, documents, 19): “Imperfect surrogates mean incorrect inferences …. scientific articles—in contrast to the previous system, If the world model is somehow wrong (and it must be) which Ranganathan calls “enumerative” due to the fact some conclusions will be incorrect, no matter how care- that they simply enumerate current knowledge and define fully drawn.” All of these proposals constitute a compre- a unique, rigid position for a document within it. The abi- hensive methodological framework for the development lity to cope with compound subjects reveals a central is- of domain systems. sue both in theory and practice of KOS, which is a preoccupation with users. Actually, one difference between 3.0 What are the aims of knowledge organization? faceted classification systems and previous ones is their focus on user needs and retrieval. Most past classification schemas are just evolutionary DDC and LCC belong to the type of system that was proposals of previous ones. The informal natures of past designed for shelving purposes, while UDC, from the be- precoodinated classification schemas are emphasized by ginning, was developed as a bibliographic system de- Vickery (2008b): signed for retrieval purposes and consequently belong to the same—more advanced—type of system as the BC In an enumerative, precoordinated classification, the system (Bliss Bibliographic Classification) and the CC hierarchical links ostensibly represent the generic re- (Colon Classification) (Ranganathan 1968). lation between a class and its subclasses, but in prac- When developing information systems, information tice they may also be used for the class-membership science always considered users’ information needs relation. The nature of the link becomes somewhat (Buckland 1991). Modern KO theories always consider indeterminate when, for example, a part or attribute users’ points of view when developing KOS, and domain is shown as a subclass of an entity. analysis has been proposed as a solid base to the development of KOS (Hjörland and Albrechtsen 1995). Do- Consequently, KOS operations depended on human in- main analysis is an approach strongly driven by users tervention demanding implicit knowledge. The lack of needs. The concept of relevance, so important to infor- principles and formalism in such systems needed to be mation science (Saracevic 1975), was always considered as complemented with the expertise and experience of in- a parameter in the Ranganathan formulations: “the issue formation professionals when faced with concrete prob- of relevance appears frequently in the theories of both lems of classification and seeking books in a library cata- Ranganathan and the CRG” (Spiteri 1998, 5). This is log. In such scenarios, there was a clear need of method- what Ranganathan (1967) aimed when he proposes the ologies and principles to build classification schemas. Canon of Relevance, the Canon of Relevant Sequence, This situation led S. R. Ranganathan to write his Prole- and the Canon of Helpful Sequence. The concept of gomena to Library Classification. What Ranganathan pro- relevance poses the question: for whom is the classifica- poses in the Prolegomena is not a classification system in it- tion scheme is being developed? self but principles for the development of such systems. The third of the Laws from Ranganathan states for According to Ranganathan (1967, 30): “As a result classi- every reader his book. Attending to user needs related to fication was taught in about 30 schools of these coun- the recovery of compound subjects thus enforces the tries, on the basis of Postulates and Guiding Principles.” requisites he recommends for a classification scheme. The principles proposed are organized throughout the Ranganathan (1967, 88) developed the concept of facet, Prolegomena in Normative Principles or Laws, Canons, and defined as “any component—be it a basic subject or an Postulates. isolate—of a compound subject.” When developing spe- When exposing the Analytico-Synthetic Classification cific KOS to a domain, categories of the basic taxonomic Theory, Ranganathan (1967) poses two requisites: 1) to structure are deployed according to PMEST categories— propose classification systems which could cope with the Personality, Matter, Energy, Space, and Time—being an dynamism of contemporary knowledge, an issue which expression of these facets in a domain; the objective is to Knowl. Org. 40(2013)No.2 119 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments represent a domain with its same characteristics—their preferred order to present the values in each facet, i.e., facets—that will enable their combination to represent the concepts in an array. compoud subjects. Facets thus enable users to access diferent expressions of this entity, as they are discussed in 4.0 Discussion knowledge records—books, documents. Facet analysis aims at making different possible access points explicit, Traditional classification systems have always being con- that is, it makes explicit the links between a classification cerned with a disciplinary approach. The pitfalls of such system and the knowledge records to which it serves as a an approach are beginning to be questioned. Gnoli (2008, finding aid. 178) writes: Thus facet analysis aims to answer the question: under what aspect does that which exists in the world manifest? Many scholars in bibliographic classification have To hit this objective, Ranganathan proposes an analysis observed that the disciplinary approach is not the which makes a division characteristic explicit to the next only possible one, and that together with benefits level of the taxonomic structure and the display of all the (like reflecting the most frequent approach of re- values in which this characteristic may be expressed in a searchers) it also brings limitations, especially for in- given domain, addressing the issue of under which dif- terdisciplinary and innovative research. Indeed, dis- ferent aspects the characteristic may be expressed. To de- ciplines act as a canonical grid. scribe this process, Ranganathan uses expressions as “unitary group,” “unitary class,” and “individualization” This questioning leads to the examination of other ap- (Ranganathan 1967, 57-59). proaches as feasible for developing a KOS (Gnoli and Ranganathan proposes an analysis process which ex- Poli 2004, 157): “Such a possibility of shifting from dis- hausts the possibilities of values to that characteristic in a ciplines to phenomena as the base unit for the structure domain, a process he names “Complete assortment of a of classification has been remarked in recent decades by Universe” by the application of a “Scheme of assort- several researchers in classification.” ment” (Ranganathan 1967, 58). Examples include a clas- Ontology comes into play as a viable strategy with sification schema to boys and their characteristics. An- which, for example, to construct robust domain models. swering such a question, the Theory of Faceted Classifi- An ontological grounded knowledge of the objects of cation enables the generation/classification of a specific the domain should make their codification simpler, mode expression of a compound subject (by concatenating the transparent, and natural. Indeed, ontology can give values of different facets) in the classification device and greater robustness to models by furnishing criteria and at the same time, to define an access point to further in- categories by which to organize and construct them formation recovery. (Gnoli and Poli 2004). Gnoli (2009, 2) stresses the need The importance of the facet approach is that it that a KOS should have a solid ontological basis: “it is stresses that several characteristics may be considered si- believed that as more accurately is reality reflected in a multaneously in the development of taxonomic struc- KOS, as more effective will it be even for practical appli- tures for a domain thus resulting in several aspects or hi- cations.” erarchies by which knowledge records may be accessed. On the other hand, computer science has long faced This is due to the fact that faceted analysis aims not one the problem of representing domains in a digital envi- canonical taxonomy but a retrieval device to attend to the ronment. That means achieving accuracy, formality, logi- variety of users needs: “different descriptions are correct: cal and unambiguous semantics in developing knowledge each of them expresses a facet of the object. Yet they are representation schemes in the absence of human assis- all descriptions of the same object. Hence, one of the tance. Since the 1970s, the computer industry realized the main tasks of information science is to find ways to inte- increasing cost of developing computer systems which grate different descriptions of the same object” (Gnoli do not correspond to users needs. The faults of com- and Poli 2004, 152). puter system developers in clearly understanding user’s Actually, faceted analysis is useless if it is limited to needs concerning the system let to the development of just one facet, thus resulting in just a single hierarchy, not methodologies aimed at accurately capturing the systems’ different from the hierarchical enumerative classification requirements. An essential requirement of such method- systems. The usefulness of facet analysis as a retrieval ologies was that they may be a communication tool be- mechanism is that several aspects can be revealed thus re- tween system developers and users. Mylopoulos (1992, sulting in multiple access points to a knowledge records 20) defines conceptual modeling as “the activity of for- collection. In addition to each facet being derived from a mally describing the physical and social world around us class, faceted analysis also proposes principles for the for the purpose of understanding and communication.” 120 Knowl. Org. 40(2013)No.2 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments

Pioneers in the development of such methodologies of the domain have been fixed—the subsequent step is were Ed Yourdon and Larry Constantine (1975), Chris to devise their dimensions of analysis. Here is where fac- Gane and Trish Sarson (1977), Tom de Marco (1979), and eted analysis can best play its role. Maintaining our refer- others. These methodologies, also called structured analy- ence domain of biology, two series of facets follow. The sis or system engineering, clearly divide the system devel- first series is centered on the governing concept of or- opment process into two distinct phases: requirement ganism as an individual whole and lists the “viewpoints” analysis, which produces a logical model of the system to from which organisms so taken can be seen. (Poli and be developed, and software implementation, which con- Obrst 2010, 15). cerns the development of programs, user interface de- The second series of facets list all the other view- sign, testing, tuning, and installation of the application in points, those not focused on the organism as a whole. a computer production environment. These may comprise, for instance, genetics (focus on the While conceptual methodologies as structured analysis genes), ethology (focus on some population of organ- has its focus on processes, another pioneer, Peter Chen isms), and ecology (focus on an entire ecosystem). But, (1976) proposed the Entity-Relationship (E-R) model, a again, this is not the entire story. A substantial number of methodology which has a focus on entity, represented as other facets can and should be developed, concerning, an aggregate of attributes, i.e., data, and their relations with for instance, the growth and development of organisms, other entities. The E-R methodology is aimed at designing their reproduction, or their alimentation. For each of databases. Since then, conceptual modelling has been an these facets, appropriate ontologies can be developed. important focus of research in computer science. The (Poli and Obrst 2010). primary product of such methodologies is what was called This position suggests that faceted analysis may be a the conceptual model, usually a graphical diagram. modeling phase to be developed after ontological analysis Researchers as Guarino and Guizzardi (2006), Guiz- and the definition of the domain core categories com- zardi (2005), and many others have emphasized the need prising the domain ontology. As a methodological phase, that conceptual modeling should have solid ontological faceted analysis may thus indicate possible access points bases. Nowadays, many meetings and workshops focus and issues related to the interface design phase. on the convergence of ontology, conceptual modeling, and software engineering. Computer science considers 5.0 Concluding remarks conceptual modeling, including increasingly ontological analysis (Guizzardi 2005), as an essential phase of system Formal ontology aims at defining what exists, here and development. There is also an increased use of concep- now, and it looks for ontological foundations of what ex- tual models in KO, such as the FRBR model (IFLA 1998) ists. Moreover computational ontology and KO both de- and the CIDOC-CRM, as guides to the development of a velop methodologies to model specific domains; domain KOS in digital environments. The development of a KOS modeling is a basic, common activity to both disciplines. in digital environments is a motivating factor to the adop- However the faceted analysis phase has a pragmatic ap- tion of conceptual modeling, incorporating the advances proach to domain modeling, aiming at developing an effi- on ontological analysis and the formalism provided by cient KOS to providing access to knowledge records. computer ontologies (Giunchiglia et al. 2009) in the cur- KO cannot ignore the knowledge provided by ontol- rent KOS development. ogy, as it reveals the ultimate nature of what does exists. The role of faceted classification as bases for a KOS If knowledge domains were not represented in an onto- on the Web has been emphasized by many authors in- logically consistent way in digital environments, as cluding Denton (2009), Gnoli and Hong (2006), Priss stressed by Gnoli, computational inferences based on (2008), Putkey (2011), Uddin (2007), and Vickery (2008a). them will lead to inconsistencies. As knowledge is pro- This fact points toward a rapprochement between KO duced and recorded according to its nature, the proper- and conceptual modelling and the need to integrate fac- ties and different aspects of how things exist are viewed eted analysis within conceptual modelling methodologies or thought by users; so it may be accessed and organized (Prieto-Díaz 2003). A proposal in this direction is Poli accordingly. Ontology provides the methodological tools and Obrst (2010), who suggest a framework for domain for modeling domains in an ontologically consistent way. modelling comprising foundational ontologies, cross- In conclusion, KO—faceted analysis—aims at identi- domain ontologies, domain specific ontologies, and fac- fying all possible aspects of a phenomenon which may be eted ontologies within a domain, reflecting the various of interest in order to preview users’ information needs. aspects of interest in a domain. It always works with users’ needs related to different as- Once the basic ontological structure of a domain has pects of a domain. The needs of contemporary culture been established—that is to say, once the levels of reality imposes to KO that it must now develop KOS in digital Knowl. Org. 40(2013)No.2 121 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments environments. Limitations of computers’ capacities for Foskett, Antony Charles. 1996. The subject approach to infor- logical and semantic processing imposes the requisite of mation. London: Library Association Publishing. developing computational representations that are logi- Gane, Chris and Sarson, Trish. 1977. Structured system cally and ontologically coherent Those requisites are analysis: Tools and techniques. [Texas]: McDonnell Doug- much more rigid in the present digital environment than las Systems Integration Company. when old KOS were operated by information profession- Giunchiglia, Fausto, Dutta, Biswanath and Maltese, Vin- als and several assumptions were implicit based on em- cenzo. 2009. Faceted lightweight ontologies. In Concep- pirical experience of their operators. tual modeling foundations and applications, lecture notes in Furthermore, KO, in face of the exponential growth computer science 5600: 36-51. Available http://link. of knowledge that characterizes contemporary society springer.com/search?facet-author= should record it in formats that increasingly enable se- Gnoli, Claudio. 2008. Categories and Facets in Integrative mantic processing and inferences by programs. Some ini- Levels. Axiomathes 8: 177–92. tiatives towards this direction are the FRBR Core vocabu- Gnoli, Claudio. 2009. The ontological approach to knowledge or- lary/namespace coded in RDF and identified by a URI ganization. Invited paper sent for discussion at the 2nd (Newman and Davis 2005), and SKOS (2004), a model Seminar on Ontology Research, Rio de Janeiro, 21-22 for describing KOS as thesauri and taxonomies, both to September 2009. be used in semantic Semantic Web applications. From a Gnoli, Claudio and Mei, Hong. 2006. Freely faceted clas- wide perspective, KO is user oriented, takes into account sification for web-based information retrieval. New re- the users’ mental models and needs in seeking informa- view of hypermedia and multimedia 12: 63-81. Available tion to make sense of their praxis, and has, as the objec- http://dx.doi.org/10.1080/13614560600758944. tive, the development of information access and recovery Gnoli, Claudio and Poli, Roberto. 2004. Levels of reality systems aimed at answering the question of how to and levels of representation. Knowledge organization 31: search for information and how be informed. 151-60. Guarino, Nicola. 1995. Formal ontology, conceptual analy- References sis and knowledge representation. International Journal of Human Computer 43: 625-40. Available http://nemo.nic. Berners-Lee, Tim, Hendler, James and Lassila, Ora. 2001. uoregon.edu/wiki/images/7/79/Guarino_IJHCS1995_ The semantic web. Scientific American. Available http:// Formal_Onto_conceptual_analysis.pdf. www.scian.com/2001/0501issue/0501berners-lee.html. Guarino, Nicola. 1997. Some organizing principles for a unified Buckland, Michael. 1991. Information as thing. Journal of top-level ontology. Padova: National Research Council. the American Society of Information Science. 42: 351-60. Guarino, Nicola. 1998. Formal ontology in information Available http://www.sims.berkley.edu/~buckland/ systems. In Guarino, Nicola, ed., Proceedings of FOIS’98, thing.html. Trento, Italy, 6-8 June 1998. Amsterdam: IOS Press, pp. 3- Chen, Peter Pin-Shan. 1976. The entity-relationship mo- 15. del—Toward a unified view of data. ACM transactions Guarino, Nicola and Guizzardi, Giancarlo. 2006. In the on database systems 1: 9-36. defense of ontological foundations for conceptual CIDOC-CRM. 2013. The CIDOC conceptual reference model. modelling. Scandinavian journal of information systems 18: Available http://www.cidoc-crm.org/ 115-26. Cutter, Charles A. 1904. Rules for a Dictionary Catalog. Guarino, Nicola and Welty, Christopher A. 2009. An Washington: Government Printing Office. overview of OntoClean. In: International handbook on in- Davis, Randal, Shrobe, Howard and Szolovits, Peter. formation systems. [Berlin]: Springer, pp. 201-20. Avail- 1993. What is a knowledge representation? AI Maga- able http://wiki.loa-cnr.it/Papers/GuarinoWeltyOnto zine 14: 17-33. Available http://groups.csail.mit.edu/ Cleanv3.pdf. medg/ftp/psz/k-rep.htm. Guizzardi, Giancarlo. 2005. Ontological foundations for struc- Denton, Willian. 2009. How to make a faceted classification tural conceptual models. Enschede, The Nederlands: and put it on the web. [Toronto]: Miskatonic University CTIT, Telematica Institut. Press. Available http://www.miskatonic.org/library/ Hjørland, Birger. Semantics and knowledge organization. facet-web-howto.html. 2007. Annual review of information science and technology 41: IFLA Study Group on the Functional Requirements for 367-406. Bibliographic Records. 1998. FRBR – Functional Re- Hjørland, Birger and Albrechtsen, Hanne. 1995. Toward a quirements for Bibliographic Records: final report. new horizon in information science: Domain analysis. 1998. Munchen: K.G. Saur. Journal of the American Society for Information Science 46: 400-25. 122 Knowl. Org. 40(2013)No.2 C. H. Marcondes. Knowledge Organization and Representation in Digital Environments

De Marco, Tom. 1979. Structured Analysis and system specifi- Ranganathan, Shiyali Ramamrita. 1968. Choice of a scheme cation. Englewood Cliffs, N.J.: Prentice Hall. for classification. Available http://www.isibang.ac.in/ Masolo, Claudio, Vieu, Laure, Bottazzi, Emanuele, Cate- ~library/portal/Pages/chp5.pdf. nacci, Carola, Ferrario, Roberta, Gangemi, Aldo and Saracevic, Tefko. 1995. A natureza interdisciplinar da ciên- Guarino, Nicola. 2004. Social roles and their descrip- cia da informação. Ciência da Informação 24 n1. Available tions. In Dubois, Didier, Welty, Christopher and Wil- http://revista.ibict.br/ciinf/index.php/ciinf/article/ liams, Mary-Anne, eds., Proceedings of 9th. International view/530/482 Conference on the Principles of Knowledge Representation and Saracevic, Tefko. 1975. RELEVANCE: a review of and a Reasoning (KR, 2004), Whistler, Canada, June 2-5, 2004, framework for thinking the notion in Information Sci- [Palo Alto, California]: AAAI Press, pp. 267-77. ence. Journal of the American Society for Information Science McGuinness, Deborah L. 2003. Ontologies come to age. 26: 321-43. In Fensel, Dieter, Hendler, Jim, Lieberman, Henry, SKOS. 2004. Simple knowledge organization system. 2004. and Wahlster, Wolfgang, eds., Spinning the semantic web: Available http://www.w3.org/2004/02/skos/. Bringing the World Wide Web to its full potential. Massachu- Smith Barry. Ontology and information systems. 2006. Avail- setts: MIT Press. Available http://www.ksl.stanford. able http://ontology.buffalo.edu/ontology%28PIC% edu/people/dlm/papers/ontologies-come-of-age-mit- 29.pdf. press-(with-citation).htm. Smith, Barry. and Grenon, Pierre. 2004. The Cornucopia Le Moigne, Jean-Luis. 1990. A teoria do sistema geral. Lis- of formal-ontological relations. Dialectica 58: 279-96. boa: Instituto Piaget. Smith, Barry and Welty, Christopher. 2001. Ontology: Mylopoulos, John. 1992. Conceptual modelling and Telos. Towards a new synthesis. In Welty, Christopher and In Loucopoulos, Pericles and Zicari, Roberto, eds., Smith, Barry, eds., Formal ontology in information systems. Conceptual modelling, databases, and CASE: An integrated Ongunquit, Maine: ACM Press. Available portal.acm. view of information systems development. New York: Wiley, org/citation.cfm?doid=505168.505201. pp. 49–68. Spiteri. Louise. 1998. A simplified model of facet analy- Newman, Richard and Davis, Ian. 2005. Expression of sis. Canadian journal of information and library science 23: 1- core FRBR concepts in RDF. Available http://vocab. 30. Available http://www.iainstitute.org/en/learn/ org/frbr/core.html. research/a_simplified_model_for_facet_analysis.php. Poli, Roberto and Obrst, Leo. 2010. The interplay be- Uddin, Mohammad Nasir and Janecek, Paul. 2007. Fac- tween ontology as categorial analysis and ontology as eted classification in web information architecture: a technology. In: Theory and applications of ontology: Com- framework for using semantic web tools. The electronic puter applications. [Berlin]: Springer, pp.1-26. library 25 n2: 15. Prieto-Diaz, Rubén. 2003. A faceted approach to building Vickery, Brian. 2008a. Faceted classiﬁcation for the web. ontologies. In IRI 2003. IEEE International Conference on Axiomathes 18: 145–60, DOI 10.1007/s10516-007- Information Reuse and Integration, 2003, pp. 458-65. 9025-9. Priss, Uta. 2008. Facet-like structures in computer sci- Vickery. Brian. 2008b. On knowledge organization. 2008b. ence. Axiomathes 18: 243–55, DOI 10.1007/s10516- Available http://classic-web.archive.org/web/2008040 007-9023-y. 4103206/www.lucis.me.uk/knowlorg.htm. Putkey, Theresa. 2011. Using SKOS to express faceted Wersig, Gernot and Neveling, Ulrich. 1975. The phe- classification on the semantic web. Library philosophy nomena of interest to information science. The infor- and practice. Available http://www.webpages.uidaho. mation scientist 9: 127-40. edu/~mbolin/putkey.htm. Yourdon, Edward and Constantine, Larry L. 1975. Struc- Ranganathan, Shiyali Ramamrita. 1967. Prolegomena to a li- tured design. New York: Yourdon Press. brary classification. Bombay: Asia Publishing House.

Knowl. Org. 40(2013)No.2 123 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts

Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts†

Teija Oikarinen* and Terttu Kortelainen**

*Archaeology, Faculty of Humanities, P.O. Box 1000, 90014 University of Oulu, Finland, **Information Studies, Faculty of Humanities, P.O. Box 1000, 90014 University of Oulu, Finland, < [email protected]>

Teija Oikarinen, MA archeology, holds a doctoral student position in the doctoral programme of Memornet, co- ordinated by the University of Tampere and financed by the Finnish Academy. The Memornet trains researchers from various fields of study to be capable of identifying changes in memory functions and to react to these. Her ongoing dissertation work focuses on digitalization of archeological data as local and global process, specifically on archeological excavation and the relation between archeological practices and data at the turning point of the digitalized eScience era. She has worked as a field manager in archeological excavations, as well.

Terttu Kortelainen, Ph.D., is University Lecturer of Information Studies at the University of Oulu, Finland. Her research interests are in informetric research, social media, and the evaluation of libraries. She has also su- pervised research projects focusing on usability of web services and evaluation of libraries. Her publications consist of four study books, a doctoral thesis, articles on informetrics, and articles on the study projects of the department. She is a member of the Advisory Committee of the Finnish Social Science Data Archive and the publication board of the Finnish Information Studies publication series.

Oikarinen, Teija, and Kortelainen, Terttu. Challenges of Diversity, Consistency, and Globality in Index- ing of Local Archeological Artifacts. Knowledge Organization. 40(2), 123-135. 55 references.

ABSTRACT: We consider documents produced in archeological post-excavation analysis and re-raise a question of archeological cataloguing, which is a specific case in the context of global progress of digitalization in archeology. The catalogue of archeological artifacts from the excavation of the city of Jakobstad, Finland was analyzed through a content analysis. Quantitative analysis was conducted using SPSS statistical package, and the results are presented in figures and tables. The analysis was based on a qualitative definition of variables describing the archeological artifacts. The analysis shows that the catalogue of artifacts is mainly systematic, but the results also reveal non-uniformity in cataloguing. In the free description column, several categorizations were found that could be used in developing the structure of an archeological catalogue. Traditional cataloguing methods are still practiced in archeology, but these do not fulfill requirements of the future use of data. In this case, a vocabulary and a tool for cataloguing archeological artifacts would contribute to the development of cataloguing and future access of data. These devices should be flexible and support uniqueness of the artifacts. There exist tools and vocabularies for archeological cataloguing and these could be localized to fulfill the needs for the future digitalization of archeological data.

Received 11 December 2012; Revised 4 January 2013; Accepted 8 January 2013

† The authors thank the Finnish Cultural Foundation and the Memornet Doctoral Programme for supporting this work. The authors thank the editor and two anonymous referees for their insightful comments and suggestions as well.

1.0 Introduction facts are classified and described on the basis of their archeological identification and interpretation such as use In archeology, field studies such as excavations are the or function of objects and their dating. The catalogues most visible part of research. However, it is a short pe- and reports produced in the post-excavation stage are a riod compared to the post-excavation stage in which arti- very relevant part of an archeological research. The qual- 124 Knowl. Org. 40(2013)No.2 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts ity of these documents has crucial impact on the future vocabularies for artifacts and materials (e.g., Material The- access and comprehension of the archeological data from saurus 1997; Object Thesaurus 1999), and procedures for cata- the excavations. loguing such as Artifact Cataloging System (2007) and Artefact The purpose of archeological catalogues of artifacts is Catalog Codes (2012) for certain periods and groups of arti- to support future research and reinterpretation of the facts. There exist also efforts to combine vocabularies data. They include several characteristics through which from different disciplines offered by web related technolo- the artifacts are described. It is crucial that the descriptive gies such as Getty Vocabularies Web Services (2013) that spe- terms are used in a uniform way; otherwise, the catalogue cialize in art, architecture and material culture. They offer is impractical (Crook et al. 2002). The premise for future vocabularies as a set of standard terminologies which are research is that, although cataloguing is implemented in a clustered with local vocabularies, e.g., local variant of certain place and time, the terminology used should be terms, such as collection-specific terms, to support end- applicable also in information retrieval and understanding users (Harpring 2010). These vocabularies are, however, (Taylor 2004). This is complicated by the heterogeneous not applicable in all local collections. nature of archeological data (Huvila 2006; Kintigh 2006a; Classification models for museum collections have Kintigh 2006b; Kintigh and Altschul, 2010; Snow et al. been developed. A general taxonomy model of museum 2006; Richards 2002; Richards 2009; Richards and Hard- artifacts utilizes facets in classification: these are context man, 2008). facets (creator, style and period, geographical location), Discussion of the quality of archeological catalogues physical property facets (object type, material and tech- is hoped for, and they are required to be comprehensive, nique), and motif facets (subject presented) (Ménard et accurate, and logged in a format that is efficient and al. 2010). The aim of this model is to organize data in a flexible to use (Crook et al. 2002). However, the nature definite hierarchical structure and to offer a cognitive and quality of cataloguing (i.e., as a process of knowledge economy to increase effectiveness in use. However, it is organization) and catalogues (as a knowledge organiza- essential to understand that archeological cataloguing for tion system) have rarely been the interest of archeolo- research and dissemination purposes differs from that of gists, while in the field of knowledge organization (KO), museums which have a non-research emphasis, i.e., exhi- this is an essential question (Hjørland 2008). bitions (e.g., Xia 2006) and different research emphasis. This paper combines the approaches of knowledge or- However, these goals are impossible to achieve without ganization and archeology. It aims to analyse how archeo- developing archeological cataloguing as a KOS. logical data is organized in the catalogues of archeological Deokattey et al. (2010) describe vocabularies and classi- artifacts. Previously, Gnaden and Haldaway (2000) have fication schemas as conventional tools for developing studied observer-related variations in artifact recording and knowledge organization, and ontologies as tools for repre- analyzed their effects statistically. Archeological cataloguing senting and defining concepts and their relations system- as classification and as a method for archeological analysis atically. A controlled vocabulary controls the variations of have been discussed by Dunnell (1986), Read (2009), and the use of synonyms and near-synonyms, homographs, Rouse (1960), for example. In any case, these earlier studies and, e.g., grammatical variations by establishing a single have not focused on the aim of this article. In this study, form of the term (Noruzi 2006). Different kinds of con- archeological cataloguing is seen as a stage in which arti- trolled vocabularies such as classification schemes, thesauri, facts are described in digital format, as a database. and taxonomies have had an import role in the organiza- In general, knowledge organization systems (KOS) con- tion and retrieval of information in different environments sist of classification and categorization schemes that or- (Mai 2008). Both thesauri and ontologies are basically rep- ganize content and control variant versions of key infor- resentational vocabularies for a specific domain, but the mation. They may also include vocabularies, such as main difference is “the use of descriptors and concepts to thesauri, semantic networks, and ontologies (Souza et al. map a given domain” (Deokattey et al. 2010, 174). Both 2012; Hodge 2000). To solve the problems of attainability controlled vocabularies and free index terms are used to and accessibility of cultural heritage data at global level, create ontologies, but ontologies are more flexible than different kinds of knowledge organization systems have thesauri (Deokattey et al. 2010). been developed for organizing of data. For example, the The main problem in culture heritage data sources, CIDOC-CRM (2011) is a standard conceptual reference their interoperability in semantic level and publishing model to combine and share data of museums, libraries, them as linked data (i.e., in their semantic aggregation) is and archives and to map separate archeological records, their heterogeneity (Mäkelä et al. 2011). For example, ref- glossaries, and thesauri with computer-based semantic web erence vocabularies used for this purpose have not been technologies (e.g., Doerr, Ore and Stead 2004; Doerr, mapped to each other. Ad hoc fixes in “inadequate data- Schaller and Theodoridou 2004). There exist established bases” are created during cataloguing to describe objects, Knowl. Org. 40(2013)No.2 125 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts they can differ from general semantics used in the disci- museums’ collection management including guidelines pline, and one item in a database can describe several ob- for cataloguing archeological data resources using (Muse- jects creating compound textual explanations. (Mäkelä et ovirasto 2013; NBA 2013) Spectrum 4.0 standard (KDK al. 2011, 1, 3-8). 2011; KDK 2012), indicating that there is an immediate Knowledge organization has its focus on the nature and need to research the quality of archeological cataloguing. the quality of knowledge organizing processes and knowl- The cataloguing and description in archeology to some edge organizing systems which are used to organize docu- degree resemble that in libraries. They share the same ments, document representations, works, and concepts. purpose, to support the future use of the items. The dif- The research has focused on studying activities such as ference between them is that in information science this document description, indexing, and classification con- procedure is called cataloguing (i.e., making a catalogue) ducted in memory institutions by different kinds of infor- and indexing (i.e., describing the content of the publica- mation specialists and computer based retrieval technolo- tion), whereas in general, in archeology, the previous term gies (Hjørland 2008). Knowledge organization combines covers both functions. In the next section, we have a look theoretical and practical contributions from different scien- at factors affecting the quality of cataloguing. tific disciplines to design systems to process information (Gonzalez 2007 cited in Bonome 2012). They consist of 2.0 Factors affecting the quality of cataloguing three main elements: knowledge, human beings, and auto- mated systems, which are in dynamic and complex interac- The main purpose of indexing is to describe the content tion (Bonome 2012). The main interest is in the design of of an item in a form suitable for inclusion in some type efficient processes for the wide-scale knowledge represent- of database, from which it can be found on the basis of ing, processing, and sharing (Bonome 2012) and promot- its content when needed in future. (Lancaster 1991). In- ing the retrievability of information (Souza et al. 2012). dexing of the content consists of two phases, (1) concep- Organizing knowledge includes three basic elements: tual analysis and (2) translation. Conceptual analysis in- the object that carries the content, i.e., knowledge, the cludes the decision of what an item is about. Often this is context constituting the frame of reference, and the aim done from the viewpoint of the interests of a future user which is supported by organizing knowledge (Bonome audience. Translation means the description of the result 2012). Knowledge can be understood as a representation of the conceptual analysis through index terms originat- of information assimilated by a person, scientific or sub- ing from an indexer’s head or a controlled vocabulary. jective, interpretative knowledge. The environment in (Lancaster 1991). Lancaster (1991) refers to indexing of which knowledge is generated can be described as con- published and printed documents, but simultaneously al- text represented usually by organizations which influence so to the indexing of non-print documents, such as au- human decision making. (Bonome 2012). In this case, the dio-visual, visual and sound media, and realia. Realia re- object is an archeological artifact, the context an archeo- fers to virtual or other objects serving as illustration logical excavation, and reporting, including the catalogu- (Smith 1997), e.g., to archeological artifacts which could ing process, aiming to contribute archeological research be visualized by using images or video clips. in the future. The expected reliable contribution of an ar- The following factors, based on Lancaster (1991) may cheological catalogue in future research sets requirements have an effect on the quality of indexing. Indexer factors, for the archeological catalogue. This is one specific proc- i.e., characteristics of the person conducting the indexing, ess in creating KOS for archeology. include subject knowledge, experience, concentration, There are no universal criteria or schema for archeo- comprehension of the item, and knowledge of user logical cataloguing, and, with regard to this study, no ex- needs. Vocabulary factors refer to the characteristics of act instructions for the cataloguing of artifacts from his- the vocabulary used in indexing, and include specificity, torical eras. The National Board of Antiquities (NBA) ambiguity, or imprecision, quality of entry vocabulary, gives orders how to produce archeological reports includ- quality of structure, and availability of related aids. Doc- ing archeological catalogue of artifacts in Finland. The ument factors, in the case of archeology, refer to difficul- structure of an artifact catalogue is required to be sys- ties in the interpretation of the item. Process factors in- tematic and to use correct terminology (Museovirasto clude type of indexing, rules and instructions, haste 2010). However, there exist no local archeological vo- originating from required productivity and exhaustivity cabularies and cataloguing tools for Finnish archeologists, of indexing. Environmental factors include, for example, and free describing or indexing are the mostl-used meth- heating or cooling, lightning, and noise in the environ- ods. The parallel Ontology for Museum Domain (2011) does ment where indexing is conducted (Lancaster 1991). not fulfill the requirements of archeology. The current An indexing failure may originate from misinterpreta- Museum 2015 project aims to create an architecture for tion of an item in question, from a difficulty in choosing 126 Knowl. Org. 40(2013)No.2 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts

the most specific term to represent it, or from the use of 4. How do the characteristics in the free description col- an inappropriate term (Lancaster 1991.) In Lancaster’s umn depend on other information in the artifact cata- (1991, 76) words, indexers could be expected to “perform logue?; and, more effectively when they are given precise rules and in- 5. How should archeological cataloguing be developed? structions”. In the field of archeology, the absence of standard guidelines causes problems (Crook et al. 2002.) 4.0 Material and methods Random errors are caused by errors in judgement of fluctuations in observation conditions. Systematic errors The research material of this study consists of urban ar- occur when an incorrect attribute is consistently applied cheological material from Lassfolk, Jakobstad (in Finnish to a class of artifacts. Illegitimate errors are genuine, ac- Pietarsaari) in Finland. The site and its archeological lay- cidental mistakes such as transcription errors (Gnaden ers were threatened by a construction project, and, in and Holdaway, 2000; see also Crook et al. 2002). 2007 and 2008, the Lassfolk area was a research object of archeological excavations organized by the Department 3.0 Research questions of Monuments and Sites, NBA. (Oikarinen 2008; 2009.) Excavation area C was chosen for the analysis because it The aim of this article is to study what kind of data has included the most of artifacts with free description and it been produced when cataloguing the archeological arti- was the most rich in artifacts covering 910 subnumbers facts. The overall goal is to study the quality of an ar- of artifacts. The whole excavation was much larger, but cheological catalogue of artifacts to understand how it this amount is seen enough for analysis to produce quali- could be developed when traditional human-based activi- tative results about the content. ties for information retrieval, integration and analysis of The method applied in the study is content analysis and, data are becoming assisted or even replaced by computer- more precisely, the content decomposition method (Tuomi based technologies (Hjørland 2008). The content of the and Sarajärvi 2009). Content analysis is defined as “distin- artifact catalogue is a database-like structure of informa- guishing characteristics” or as “a research technique for the tion consisting of stable columns for variables and rows objective, systematic, and quantitative description of the for documented artifacts called subnumbers (Figure 1). manifest content of communication” (Berelson 1971, 18). Subnumbers or sub-units represent artifacts documented Content analysis means qualitative analysis whereas de- from the same context (i.e., strata where the artifacts were composition means the quantitative decomposition of found) of the grid coordinate system in the excavation. content (Tuomi and Sarajärvi 2009). In this study, the more The research questions in this study are: precise term “content decomposition method” is used, although qualitative analysis is used to define and classify the 1. Which general concept variables are distinguished from variables in the catalogue of artifacts. The quantitative the catalogue of archeological artifacts? Into which analysis was conducted, applying SPSS statistical package variables have the characteristics of the subnumbers to generate a general view of the results of the qualitative been divided in the column for free description? classification and description process. So far, the method 2. What is the quantitative distribution of these vari- has not been applied to archeological catalogues. ables? In the catalogue, variables for stable columns were 3. Is the cataloguing done systematically? What kinds of main number (i.e., identification number for archeological properties of the artifacts are catalogued and not cata- artifact collection in certain excavation), subnumber (i.e., logued? Is there any possibility to define the concepts is different for each row in catalogue), coordinates (x, y, z, not catalogued? i.e., location), context (i.e., stratigraphic context of the ar-

Main Sub- Unit X- Y- Z- Main Type Measure Material Description Total Weight(g) Else number number (context) coordinate coordinate coordinate material (item) (cm) Bottle 1 green, 1 KM2008051 13 CSY2 478 476 Glass 2 11,7 glass transparent Window KM2008051 14 CSY2 478 476 Glass Green 1 1,4 glass KM2008051 15 CSY2 478 476 Metal Iron Key Big 1 10,9x3,2x0,9 46,2 Stone Brownish KM2008051 16 CSY3 488 476 Pottery Vessel 1 20 ware glazing

Figure 1. A part of a catalogue of artifacts translated from Finnish (Oikarinen 2009, Appendix 8). Knowl. Org. 40(2013)No.2 127 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts tifact as codes), main material, material, type (i.e., artifact study, the variables were coded as numbers: color relating type), total (amount of artifact pieces), weight (as grams), to the number of colors mentioned (1 = no color men- and measures (verbal or numerical), as well as description tioned, and 2 = one color mentioned etc.). The number of for free expression and a column else (i.e., notes). A sub- forms was coded through the same kind of system (1 = no number means one row, i.e., unit in the catalogue. The log- form mentioned, etc.). The use, surface finish, number, ic to catalogue the artifacts from one excavation area is to dating, size, and burnt were coded in values according to list them according to their material alphabetically, and in mentions of them, like 1 = yes or 2 = no. The number of the order of their context number and x- and y- properties was counted as a sum for each subnumber as 1 coordinates. (Figure 1.) One subnumber can consist of = not mentioned, 2 = one property mentioned, etc. In ad- one or numerous pieces of artifacts. Consequently, the dition, there was an option to create more variables and as- sample is enough to permit making general assumptions sign them more values, for example, in form. The data was about cataloguing practises by methods used in analysis. described by statistical tables and diagrams. By using decomposition analysis, any document can be divided into different variables and values of variables, 5.0 Results which can be studied and counted. The examined catalogue included artifacts described by the following vari- The archeological artifacts from the city of Jakobstad are ables: color (for example red), use (for example, a precise typical urban artifacts from Finland dating from the 17th concept like knife from which it is also possible to con- to the19th century. Their most general main material clude its function), surface finish (for example glazed), group was pottery (49.3% of subnumbers), then glass form (for example, flat), number (adjustment—for exam- (29.5%) and metal (16.8%). There were also small quanti- ple, one handle), dating, size (in words like “small”), and ties of bones, leather, bark and stone, wood artifacts, and burnt. In addition to these, there were variables, descrip- wool (Figure 2). The interest in this study was focused on tionElse and datingElse. In the research material of this free expression in the description column, which includes

Figure 2. Relative proportions of subnumbers in percentages in the main material. 128 Knowl. Org. 40(2013)No.2 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts the following variables: color, use, form, surface finish, for this is that in the cataloguing phase they had not been burnt, dating, and size. These variables were compared to restored and were unidentifiable. In this sample, all wool, information in other columns through cross-tabulation. bark and stone, wood, and bone artifacts had been de- The main materials of the subnumbers were cross- scribed. However, not much can be concluded on the basis tabulated with the characteristics describing them. Nine of these very sporadic (2-6) subnumbers (Table 1). hundred ten subnumbers had received a total of 1,496 de- The descriptions of the subnumbers by classifying vari- scriptions. A substantial number (82.3%) of the subnum- ables in the free description column were studied through bers had been described (Table 1). Two hundred sixty-five cross-tabulation to see how many variables were used to glass artifacts had received 268 descriptions (98.9%), for describe each subnumber. The differences were clear: leather, the figures were respectively 18/19 (94.7%), and 17.7% of subnumbers had received no classifying vari- ceramics respectively, 377/499 (83.9%). This indicates that ables, 37.9% had received one, and 21.5% two classifying glass and ceramics are easy to describe or identify. Glass variables. Some subnumbers were described by three occurs in different colours and ceramics have a lot of char- (11.5%) or four (10.9%) variables. There were only a few acteristics to describe such as decoration or glaze, and subnumbers in which five (0.22%) or seven (0.11%) prop- moreover recognizable pieces, such as a piece of a handle. erties were mentioned (Figure 3). This result refers to non- Table 1 indicates that 17.7% of all subnumbers had not uniformity and omission problems in indexing (Bernier been described. The lowest percent of described subnum- 1980) but also to the diversity of archeological material. bers was in metals. About half of the metal subnumbers The most common defined attribute in this research had been described, 73/153 (47.7%). An obvious reason material is color, although it was not mentioned in 50.3%

The number of variables describing the sub- Sub-numbers Nr of described sub- % of described subnumbers total numbers (1-7) numbers Main material 0 1 2 3 4 5 7 A B B/A Ceramics 72 106 108 65 95 2 1 449 377 83.9 Glass 3 171 59 31 4 0 0 268 265 98.9 Bone 0 6 0 0 0 0 0 6 6 100 Metal 80 48 20 4 1 0 0 153 73 47.7 Leather 1 8 6 4 0 0 0 19 18 94.7 Wood 0 2 3 1 0 0 0 6 6 100 Shell 5 0 0 0 0 0 0 5 0 0 Bark and stone 0 2 0 0 0 0 0 2 2 100 Wool 0 2 0 0 0 0 0 2 2 100 Subnumbers 161 345 196 105 100 2 1 910 749 82.3 total Table 1. Totals of subnumbers in different materials and the number of characteristics describing them.

Figure 3. Relative proportions of subnumbers according to the number of variables applied in the free description of subnumbers. Knowl. Org. 40(2013)No.2 129 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts

Figure 4. Number of colors mentioned per subnumber according to main materials. of the subnumbers (Figure 4 and Table 2). Most often, information. However, it produces non-uniformity, as all there was a description of one (32.6%) or two (12.6%) subnumbers are not treated in an equal way. It is debat- colors. The most colorful materials according to the re- able if the information is needed if the type of an artifact sults are ceramic and glass. This outcome can be dis- has been defined correctly such as red ware. Another op- cussed, what it indicates, because all the artifacts have tion would be a column for color also to trace after inter- some kinds of colors. This result, too, indicates non- preted colors for materials or to instruct more precisely uniformity and omission problems in the indexing of the what kind of other colors should be described. artifacts. Color is a very subjective quality and easy to describe, Number of colors men- Frequency of % but, without standards, it is very difficult to distinguish tioned subnumbers between different colors or shades. The most used color 0 458 50.3 standard is the Munsell Colour Chart (Munsell 1907), but 1 297 32.6 using it is very time-consuming and interpretative in na- 2 111 12.2 ture (Goodwin 2000). It is debatable if the subjective col- 3 32 3.5 or definition is informative for the user of the catalogue. 4 10 1.1 When the color is typical for material and obvious for the 5 1 0.1 6 1 0.1 cataloguer, it is not mentioned in the free description, Total 910 100.0 such as red or white in ceramics. This is an established practice and a time-efficient way to catalogue when using Table 2. Totals of colors mentioned in the description column established names for artifacts and their materials. Usually by proportions of subnumbers. only special colors were mentioned, or typical but totally different variations in the material, such as green or Colors are described most often in glasses (97.4.% or transparent window glass. This is a tradition, and it opti- 261/268 described subnumbers with 1-4 colors) and ce- mizes consumption of time not to catalogue redundant ramics (41.2% or 185/449 subnumbers described with 1- 130 Knowl. Org. 40(2013)No.2 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts

6 colors). This is mainly a representation of color of the of artifacts, there are words like piece, handle, or piece of material in glasses but decoration or glaze in ceramics. bottom in ceramics, which are diagnostic properties and Also sporadic numbers of bone, metal, and leather were can be used to identify the use of an artifact, but they can described with one color in this sample. Surface finishing also describe forms, or the amount of “pieces” of arti- was described or identified in 31.1% of the subnumbers facts or that the artifacts are broken. of artifacts or pieces of them (Table 3). Most of them The results indicate confusion in the use of two con- were ceramics (55.5% or 249/449 of subnumbers), bone cepts, material and type. For example, in the material col- (50% or 3/6 of subnumbers), wood artifacts (50% or 3/6 umn there was the concept “yarn,” even though it is type of subnumbers), and leather (36.8% or 7/19 of subnum- and should be recognized as material, i.e., as wool or so- bers) in this material. In almost 70% of the total amount mething else. The confusion concerns especially fragile of subnumbers, the surface finishing was not described. organic or metal artifacts which have to be collected with In general, there are many kinds of surface finishing in soil samples to preserve them after removal from their archaeological artifacts such as stitching in leather, or original context, and they need processing, which is done stamps, marks of manufacturing or decoration in glasses by a professional conservator. The cataloguing, however, and pipes. is usually done before this phase. There are also materials and types which are rarer and difficult to recognize or de- Variables fine without conservation treatment and consulting with described in the specialists. This confusion refers to cases difficult to subnumbers interpret and that, according to Lancaster (1991), belongs Yes No to problems caused by document factors. Moreover, the Fre- Fre- Total of quency quency description column can contain information about use % % subnum- of sub- of sub- (i.e., indicating function and type). It can specify it or it bers numbers numbers can name it, even if the type, as a hyper concept has not Use 365 40.1 545 59.9 910 been filled in. Sometimes the reader can deduce the use Surface 283 31.1 627 68.9 910 (i.e., a type) of the artifact by comparing the information finish contents in the type and description columns. Burnt 26 2.9 884 97.1 910 The content of type column (Table 5) shows that al- Dating 4 0.4 906 99.6 910 most one third of subnumbers lack the definition of type Size 12 1.3 898 98.7 910 (31.9%). There are mentions such as “an artifact” in the Table 3. Totals and relative proportions of the variables in the type column in 5.3% of artifacts, which refers to an uni- description column. dentified type of artifact. Such ad hoc fixes during cataloguing are mentioned by Mäkelä et al. (2011). Most defi- Number of forms Frequency of % nitions of type refer to pottery originating from vessels mentioned subnumbers (34.5%). These results imply that the vessels are easy to 0 841 92.4 identify and cataloguers are familiar with them. Table 5 is 1 32 3.5 a concrete example of a case where there exist no vo- 2 20 2.2 cabularies to define the type and, consequently, the field 3 15 1.6 is left empty or the terminology is sometimes haphazard. 4 2 0.2

Total 910 100.0 Frequency of sub- % Table 4. Frequencies of the values of the variable form in the numbers description column. Artifact type: not defined 290 31.9 Artifact type: vessel 314 34.5 The concepts form and use are problematic variables to Artifact type: vessel+? 1 0.1 define when aiming at objectivity. It is practical to charac- Artifact type: artifact 48 5.3 Artifact type: miscellaneous terize the variable use as the original purpose of the item. 257 28,2 artifact types Interpreting the life-time functionality as use of an ar- Total 910 100.0 cheological artifact is not always straightforward (Bahn and Renfrew, 1997). The variable use was identified in Table 5. Distribution of subnumbers by artifact types mentioned 40.1% subnumbers (Table 3). This means that in about in the type column. 60% of subnumbers, its function was not defined, whereas the classified variable form was not defined in 92.4% The absence of information concerning material or type of subnumbers. (Table 4.) In many cases in the catalogue refers to problems in the interpretation of the artifact, Knowl. Org. 40(2013)No.2 131 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts i.e., to document or indexer factors (Lancaster 1991, 79). the catalogue contains random and illegitimate errors Dating is hardly ever mentioned in the description col- (Gnaden and Holdaway, 2000). All the factors which may umn (0.4% of subnumbers) (Table 3). have effect on the quality of content, i.e., indexer, vo- The problems of non-uniformity in the description of cabulary, document, process, and environmental factors the subnumbers are emphasized in the description of (Lancaster 1991, 79) were recognized in this study. types, uses, forms, and colors of the artifacts. The analy- The results imply to non-uniformity: the characteris- sis of uses, forms, and main materials showed that the tics of some archeological artifacts had been described by variable form is mainly described for wood, leather, and several variables, while others were not described at all. metal. For example, from metal artifacts, it is difficult to Descriptions were mostly focused on inadequately identi- recognize the type and, instead of this, there are adjec- fied artifacts. Artifacts of metal and leather, for instance, tives describing them, which were recognized as forms. may be difficult to recognize before conservation treat- These refer to problems in interpretation and naming of ment. Therefore, these may be described through subjec- the artifacts, i.e., document and vocabulary factors (Lan- tive adjectives. Instead, in the cases of well-known arti- caster 1991). 64.4% of ceramic artifacts were described fact types, free descriptions either are sometimes missing, by their uses (289/449 of subnumbers). A large number or they are very carefully catalogued depending on time of leather artifacts were described with terminology re- schedule of the project. Redundant information, such as ferring to their use (14/19 of subnumbers). general color of the artifacts, is often left out of the cata- Occasionally (1.3% of subnumbers) in the description logue. There are also mentions of details difficult to gen- column, there was a verbal description of the size of an eralize as variables, for example, artifacts that are broken artifact, like “a big key” or measures of some special part but described as well-preserved. In general the condition of an artifact (Table 3). Also there were only sporadic of artifacts is not reported. mentions of burnt in subnumbers (2.9%), which merely The understanding of the columns material and type means that burnt has been recognized in these cases (Ta- were sometimes mixed, or they had not been identified, ble 3). The percentage of measured artifacts of all sub- or the data entry was empty. To identify uncommon arti- numbers is only 9.7% in the measure column of the artifacts as types, profound knowledge is needed or an excel- fact catalogue (Table 6). That means that, in this project, lent collection to compare them with. This appears also there was no time to measure more artifacts with infor- in the use of special vocabulary in the naming of manu- mative properties and dimensions. facturing techniques or decoration in the description column. Facts, which are familiar to a cataloguer, are easy to Frequency of sub- % name, such as different kinds of colors or surface finish. numbers It would be possible to classify the latter into at least two Yes 88 9.7 more categories, namely manufacture and decoration. Measure No 822 90.3 The same concerns the concepts use and form, which Total 910 100.0 could be divided into more precise sub-variables. These Table 6. Totals and relative proportions of values in the measure variables were easily mixed implying that the terminology column. used in the catalogue is difficult to categorize. Moreover, the catalogued information may be mislead- 6.0 Discussion ing in some cases: for example, the description column included a mention of three pieces of clay pipes with de- The creation of an archeological catalogue is a time- coration, although besides them there were also three un- consuming and challenging phase after the field study, decorated pieces. The total number of pieces (in this case and the catalogue is needed as an access point to the arti- six) is calculated in the total column. A reader has to be fact collection of the archeological site, also in the future. aware of such cataloguing practices to interpret the cata- This emphasizes the need to study its quality. logue correctly. The study focused on the description of archeological All the practices and problems mentioned above reflect artifacts in the column for free description in the excava- “constraints of a scholarly domain” originating from its tion catalogue. The most frequently occurring variables discourse, history, schools of thought, paradigms, research describing the studied 910 subnumbers were color, use, fronts, activities, etc. (Mai 2008, 23). Reasons for the prob- surface finish, and form, whereas burnt, dating, and size lems derive from conventions of archeological cataloguing, occurred infrequently. Although the information content circumstances, and lack of appropriate tools. When cata- in the catalogue was mainly systematic, the results reveal loguing artifacts, archeologists still create classifications us- problems challenging its use and correct understanding in ing generalized terms or terms created by themselves. This future research. Systematic errors were not identified, but can be seen as subjective knowledge originating from cul- 132 Knowl. Org. 40(2013)No.2 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts ture, for example, or as assimilated knowledge that a per- The achieved semantic indexing of archeological docu- son gives an expression (Bonome 2012). Their non- ments requires incorporation of existing ontologies and standardized use makes information content incomparable controlled vocabularies as conceptual framework (Vla- between documents and collections, and difficult to search. chidis et al. 2010). At the same time, the usefulness and re- In general, the purpose of the catalogue is to aid in identi- alizability of vocabularies have been discussed critically fying objects and providing access to artifact collections, from the viewpoints of digital data preservation and inte- which is enhanced by information retrieval by index term gration related to terminological specialities and different of free description retrieval. languages (Eiteljorg 2011). However, controlled terminol- Detailed descriptions may reveal errors in general level ogies exposed in the Web, which are also applicable by and categorizations (Crook et al. 2002). Nevertheless, free ex- combinable in other programmatic solutions may be useful pressions and comments may also include valuable ar- for creating knowledge organization systems, such as the cheological information. Therefore, there is a need to recently proposed Terminology Web Services for archeol- preserve unique local vocabulary and cultural history, for ogy (Binding and Tudhope 2010). example, in the naming of the artifacts after their original Flexibility contributes to future classification (Bowker names or atypical uses or reuses. The problem of hetero- and Star 1999), and the challenge is to be aware of the ef- geneity related to archeological documents and collec- fects of pre-established knowledge, technologies, and vo- tions is worldwide (Snow et al. 2006). In rescue excava- cabularies in decisions made during cataloguing such as in tions, there are usually no possibilities to correct the cata- artifact categorizations. Lock (2003, 82) has warned of logue after the ending of the projects, although treatment this issue at general level: “The intentions of an analysis and identification of artifacts still continue. Users do not should not be determined by what the technology will do know that the interpretation of artifacts is sometimes but by the archeological questions being asked.” This is preliminary. These problems refer to indexer and process important because archeological research is interpretative factors mentioned by Lancaster (1991). in nature. To be applied to archeological cataloguing, the These problems could be neutralized by using vocabu- components or vocabularies in the user interface should laries to identify and name different materials and artifact not restrict too much free description, but, on the other types, and to describe them correctly. They could act as hand, should allow the analysis of the data for specific entries for index terms in information retrieval. A proper research questions. Also balancing the needs between terminology in naming could contribute also in aggregat- global standardisation and local free description is ing and linking (i.e., as cross-references) between differ- needed. ent documents in digital formats. Also metadata relating The decomposition of archeological catalogue in this the document should be standardized. If the access to ac- study produced new categorizations for archeological ar- tual artifact collection is not possible to gain, the most re- tifacts. These are: decoration, surface finish, marks of liable way to use catalogued data is with digital images of manufacturing, color, condition, dating, diagnostic prop- artifacts. erties (contributing to the identification of the artifact), Moreover, from a database-like structure, the artifacts and special properties or functions, such as reuse. These are difficult to combine with for example the context de- could be used together with the existing categorizations scriptions in the report and other documented data con- coordinates, i.e., location, context (i.e., stratigraphic concerning the same excavation although this is needed. For text of the artifact), main material, material, type, (free) example, because the context of the artefacts (i.e., strata) description, total (i.e., amount of pieces), measure (meas- is presented as codes, to understand the content of the ured dimensions), weight and else (i.e., notes) in develop- context column other archeological documents are ing systems or meta-structure for documents. needed. If the catalogue has not been created systemati- As a result of this study, it is possible to extract rec- cally, the amounts of different kinds of materials, artifact ommendations: 1) vocabularies would improve the quality types, their weight and division in documented contexts of the catalogue; 2) an effective cataloguing tool with (i.e., in strata type) etc. can not be analyzed statistically. guidelines, vocabularies, and a possibility to combine in- In archeology heterogeneous—or diverse—information formation from different types of digital documents and is the fact that we have to accept (e.g., Snow et al. 1996), comparative archeological collections would improve the but new technologies may allow production of locally es- interpretation (i.e., research and analysis process); 3) the tablished content which is retrievable and also understand- existence of combined (linked or integrated) archeologi- able at the global level. For example, new artificial intelli- cal data resources such as artifact catalogues, images, gence and Semantic Web technologies such as natural lan- maps, contextual and textual information, and/or access guage processing (NLP) have already been experimented to them could help archeological research and analysis. to create ontologies in archeology, with promising results. According to literature, controlled vocabularies increase Knowl. Org. 40(2013)No.2 133 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts possibilities to the use of latest technologies in making archeological artifacts. This refers to broader issues in con- the disparate catalogues interoperable and usable in com- ceptualizations and reporting in archeological discipline re- parative archeological studies (Vlachidis et al. 2010). This flecting also to the heterogeneous nature of archeological study also revealed the need for localized tools and more data (e.g., Huvila 2006; Snow et al. 2006). specific guidelines in cataloguing. Archeological catalogues reflect Xia’s view (2006, 271) There are several general requirements for an archeo- of the difficulty of using “patterned descriptions to pre- logical catalogue in digital format: 1) it should describe the cisely elucidate variations of individual objects.” This is catalogued artifacts systematically and understandably to solved by using individual descriptions ending up in non- the user who usually is an archeologist; 2) it should pre- systemicity in the catalogues. Solution to this and the serve the local nuances of artifacts, and describe them at a enormous quantity of archeological data could be elec- consistent level; 3) it should work as a database which is tronic publishing (Xia 2006) combined with shared vision analyzable for research purposes; and 4) it should be com- of knowledge organization systems discussed by for ex- parable and combinable with other relevant archeological ample Bonome (2012), Souza et al. (2012), and da Silva and data sets. The cataloguing should be flexible and controlled Ribeiro (2012). by cataloguers even if cataloguing tools would be available. At the same time, the introduction of user-generated In this specific local context, the development also in- free-form tags, or folksonomies seems to be removing hi- cludes the challenge of unifying the goals of museum erarchy from the scheme of knowledge organization and archeological cataloguing principles as a coherent through facilitating knowledge discovery and web indexing whole serving both disciplines (NBA 2013), although this (Noruzi 2006), which could also contribute to content de- is beyond the scope of this paper. scription or creating metadata for archeological data. This is an example of a KOS which integrates knowledge con- 7.0 Conclusions tributed by individuals from the social networks (Bonome 2012), e.g., of archeologists and other relevant specialists. Keeping in mind the unique nature of archeological data Further requirements are set by Mai (2008) who calls and the need for unique descriptions, this study reveals for a more domain-centered approach in the design of some major problems in cataloguing large collections of controlled vocabularies where knowledge and expertise of archeological artifacts. The lack of standards and uniform indexing should be complemented with that of informa- practices in content descriptions results in subjectivity and tion behavior to match the actors’ information needs. A incorrect or irrelevant vocabulary, which can lead to a dan- comprehensive archeological KOS should be based on the ger of misinterpreting the content and to incomparability cognitive work analysis of archeologists (Mai 2008), which between documents and catalogues. This can also have an combines the cataloguing of the artifacts with their context effect on the reliability of archeological field work reports. information and all the other information (i.e., varying data The study proves a need for localized vocabulary develop- resources) produced in the excavation. ing towards ontologies in archeology, and contributes to the development of tools and practices in the cataloguing References of archeological artifacts and their future use in digital environments. Artefact catalog codes–The London Museum of Archaeology. 2012. Due to the increasing possibilities and utilization of London Chapter, Ontario Archaeological Society. Avail- Web-related technologies and information technologies to- able http://www.ssc.uwo.ca/assoc/oas/misc/catcodes. day, knowledge organization systems may extend from lo- html. cal to almost global scale, and they spread over the capacity Artifact cataloging system. 2007. Society for Historical Ar- of organizations to set standards or rules to be imple- chaeology (SHA). The University of Montana. Avail- mented in them. The Web increases possibilities for inter- able http://www.sha.org/research/artifact_cataloging action and has an impact on knowledge creation by offer- _system.cfm. ing new channels for interaction and for creating social Bahn, Paul and Renfrew, Colin. 1996. Archaeology: theories, networks, new communication tools, and ways how to methods and practice. London: Thames and Hudson Ltd. process and retrieve information. It blurs traditional organ- Berelson, Bernard. 1971. Content analysis in communication izational boundaries as a context of interaction and self- research. New York: Hafner Publishing Company. recognition (Bonome 2012). Bernier, Charles L. 1980. Subject indexes. In Kent, Allen, These changes also concern the setting of requirements Lancour, Harold and Daily, Jay E., eds., Encyclopedia of for future archeological knowledge organization systems, library and information science, 29. New York: Marcel and the process of archeological cataloguing. This study Dekker Inc, pp. 191-205. revealed problems mainly in the naming and describing of 134 Knowl. Org. 40(2013)No.2 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts

Binding, Ceri and Tudhope, Douglas. 2010. Terminology Goodwin, Charles. 2000. Practices of color classification. web services. Knowledge organization 37: 280-6. Mind, culture, and activity 7: 19-36. Available http://www. Bonome, María G. 2012. Analysis of knowledge organi- sscnet.ucla.edu/clic/cgoodwin/96col_class. pdf zation systems as complex systems: A new approach Harpring, Patricia. 2010. Introduction to controlled vocabularies. to deal with changes in the web. Knowledge organization Featuring the Getty vocabularies. Available http://www. 39: 104-9. getty.edu/research/tools/vocabularies/intro_to_vocabs. Bowker, Geoffrey C. and Star Leigh, Susan. 1999. Sorting pdf things out: Classification and its consequences. Cambridge Hjørland, Birger. 2008. What is knowledge organization (Mass.); London: MIT Press. (KO)? Knowledge organization 35: 86-101. CIDOC-CRM. 2011. CIDOC Conceptual Reference Hodge, Gail. 2000. Systems of knowledge organization for digital Model. International Council for Museums. Available libraries: Beyond traditional authority files. Washington D.C: http://www.cidoc-crm.org/. The Digital Library Federation Council on Library and Crook, Penny, Lawrence, Susan and Gibbs, Martin. 2002. Information Resources. Available http://www.clir.org/ The role of artefact catalogues in Australian historical pubs/reports/pub91/reports/pub91/pub91.pdf archaeology: A framework for discussion. Australasian Huvila, Isto. 2006. The ecology of information work: A case historical archaeology 20: 26-38. Available http://asha study of bridging archaeological work and virtual reality based docs.org/aha/20/20_04_Crook.pdf knowledge organization. Åbo: Åbo Akademis förlag - Åbo Deokattey, Sangeeta, Neelameghan, Arashanipalai and Akademi University Press. Kumar, Vijai. 2010. A method for developing domain Kansallinen digitaalinen kirjasto (KDK) [National Digital ontology: A case study for a multidiciplinary subject. Library]. 2011. Kansallisen digitaalisen kirjaston kokonais- Knowledge organization 37: 174-84. arkkitehtuuri: Liite B standardisalkku. Kansallinen digi- Doerr, Martin, Ore, Christian-Emil and Stead, Stephen. taalinen kirjasto. Available http://www.kdk.fi/images/ 2004. The CIDOC reference model – A New standard stories/tiedostot/kdk%20standardisalkku%202011-08 for knowledge sharing. In J. Grundy, John C. Grundy, -29.pdf. Hartmann, Sven, Laender, Alberto H.F., Maciaszek, Kansallinen digitaalinen kirjasto (KDK) [National Digital Leszek A. and Roddick John F, eds., Proceedings, tutorials, Library]. 2012. Kansallinen digitaalinen kirjasto –hanke. posters, panels and industrial contributions at the 26th Interna- Väliraportti ajalta huhtikuu 2011 – kesäkuu 2012. Kansal- tional Conference on Conceptual Modeling - ER 2007 Auck- linen digitaalinen kirjasto. Available http://www.kdk. land, New Zealand. CRPIT, 83, pp. 51-56. Available fi/images/tiedostot/KDK-vliraportti_27092012-1.pdf. http://crpit.com/confpapers/CRPITV83Doerr.pdf. Kintigh, Keith. 2006a. The challenge of archaeological Doerr, Martin, Schaller, Kurt and Theodoridou, Maria. data integration. In Union Internationale des Sciences 2004b. Integration of complementary archaeological Préhistoriques et Protohistoriques, Lisbon, 4-9 September sources. In Niccolucci, Franco and Hermon, Sorin, 2006. Available http://archaeoinformatics.org/articles/ eds., Proceedings of the International Conference on Computer Kintigh2006UISPP.pdf. Applications and Quantitative Methods in Archaeology Con- Kintigh, Keith. 2006b. The promise and challenge of ar- ference, CAA 2004, 13-17 April, 2004, Prato, Italy. Bu- chaeological data integration. American antiquity 71: 567- dapest: Archeolingua. Available http://www.ics.forth. 78. Available http://www.public.asu.edu/~kintigh/Kin gr/isl/publications/paperlink/doerr3_caa2004.pdf. tigh2006CyberinfrastructureAmAnt.pdf Dunnel, Robert C. 1986. Methodological issues in Ameri- Kintigh, Keith and Altschul, Jeffrey H. 2010. Sustaining canist artifact classification. Advances in archaeological the digital archaeological record. Heritage management 3: method and theory 9: 149-207. 264-74. Available http://www.digitalantiquity.org/wp- Eiteljorg, Harrison. 2011. What are our critical data- uploads/2011/01/20110127-Kintigh-Altschul-Forum- preservation needs? In Kansa, Eric C. et al. eds., Ar- on-Sustaining-the-Digital-Archaeological- chaeology 2.0. new approaches to communication and collabora- Record-from-Heritage-Management.pdf tion. Los Angeles: Cotsen Institute of Archaeology Lancaster, Frederick Wilfrid. 1991. Indexing and abstracting Press, pp. 251–64. Available http://www.escholarship. theory and practise. Champaign, Ill: University of Illinois, org/uc/item/1r6137tb. Graduate School of Library and Information Science. Getty Vocabularies. 2013. Getty Research Institute. J. Paul Lock, Gary. 2003. Using computers in archaeology. London; Getty Trust. Available http://www.getty.edu/research/ New York: Routledge. tools/vocabularies/index.html. Mäkelä, Eetu, Hyvönen, Eero and Ruotsalo, Tuukka. Gnaden, Denis and Holdaway, Simon. 2000. Understand- 2011. How to deal with massively heterogeneous cul- ing observer variation when recording stone artifacts. tural heritage data – lessons learned in CultureSampo. American antiquity 65: 739-47. Knowl. Org. 40(2013)No.2 135 T. Oikarinen and T. Kortelainen. Challenges of Diversity, Consistency, and Globality in Indexing of Local Archeological Artifacts

Semantic web 0: 1-24 Available http://www.semantic- Richards, Julian D. 2002. Digital preservation and access. web-journal.net/sites/default/files/swj160_0.pdf. European journal of archaeology 5: 343-66. Available http:// Mai, Jens-Erik. 2008. Actors, domains, and constraints in eja.sagepub.com/cgi/content/abstract/5/3/343. the design and construction of controlled vocabular- Richards, Julian D. 2009. From anarchy to good practice: ies. Knowledge organization 35: 16-29. The evolution of standards in archaeological comput- Material Thesaurus. 1997. British Museum. Trustees of the ing. Arheologia e calcolatori 20: 27-35. Available http:// British Museum. Available http://www.collectionslink. eprints.whiterose.ac.uk/10707/1/Arch_%26_Calcolatori org.uk/assets/thesaurus_bmm/matintro.htm?phpMyAd _-_standards_2009_richards.pdf. min=OYNyINPdn3sQmoXugKH1gcCLSW0. Richards, Julian D. and Hardman, Catherine. 2008. Step- Ménard, Elaine, Mas, Sabine and Alberts, Inge. 2010. Fac- ping back from the trench Eege. An archaeological eted classification for museum artefacts. Aslib proceedings: perspective on the development of standards for re- New information perspective 62: 523-32. Available http:// cording and publication. In Greengrass, Mark and www.emeraldinsight.com/journals.htm?articleid=18767 Hughes, Lorna eds., The virtual representation of the past. 53&show=pdf Farnham, Surrey and Burlington, USA: Ashgate Pub- Munsell, Albert H. 1907. A color notation. Boston: Geo. H. lishing Company, pp. 101-12. Available http://eprints. Ellis co. Available http://www.gutenberg.org/etext/ whiterose.ac.uk/7795/1/richardsjd4.pdf 26054 Rouse, Irving. 1960. The classification of artifacts in ar- Museovirasto [National Board of Antiquities]. 2010. Oh- chaeology. American antiquity 25: 313-23. jeita historiallisen ajan kaivauskertomuksen laatijalle. Raken- da Silva, Armando Malheiro and Ribeiro, Fernanda. 2012. nushistorian osasto. Available http://www.nba.fi/fi/File/ Documentation / information and their paradigms: 1028/tutkimuskertomusohje-rho-2010.pdf Characterization and importance in research, educa- Museovirasto [National Board of Antiquities]. 2013. Mu- tion and professional practise. Knowledge organization 39: seoiden kokoelmahallinnan kokonaisarkkitehtuuri 1.0. Museovi- 111-24. raston ohjeita 2. Museo 2015. Museovirasto. Available http:// Smith, Bryan. 1997. Virtual realia. The Internet TESL jour- www.nba.fi/fi/File/1806/kokonaisarkkitehtuuri.pdf. nal III (7). Available http://iteslj.org/Articles/Smith- National Board of Antiquities [NBA]. 2013. The museum Realia.html. 2015 oroject. Available http://www.nba.fi/en/develop Snow, Dean. R., Gahegan, Mark, Giles, C. Lee, Hirth, ment/museum_2015 Kenneth G., Milner, George R., Mitra, Prasenjit and Noruzi, Alireza. 2006. Folksonomies: (Un)controlled vo- Wang, James Z. 2006. Cybertools and archaeology. Sci- cabulary? Knowledge organization 33: 199-203. ence 311: 958-9. Available http://clgiles.ist.psu.edu/ Object Thesaurus. 1999. British Museum. Trustees of the Bri- papers/Science-cyber-historical_2006.pdf. tish Museum. Available http://www.collectionslink. Souza, Renato Rocha, Tudhope, Douglas and Almeida, org.uk/assets/thesaurus_bmon/Objintro.htm?phpMy Maurício Barcellos. 2012. Towards a taxonomy of Admin=OYNyINPdn3sQmoXugKH1gcCLSW0. KOS: Dimensions for classifying knowledge organiza- Oikarinen, Teija. 2008. Pietarsaari, 2/2/7-10, Lassfolkin tion systems. Knowledge organization 39: 179-92. kortteli (PSL-07). Kaupunkiarkeologinen koekaivaus Taylor, Arlene G. 2004. The organization of information. 6.9.-28.9.2007. (Unpublished excavation report) Westport, Conneticut; London: Libraries Unlimited. Museovirasto: Rakennushistorian osasto. Tuomi, Jouni and Sarajärvi, Anneli. 2009. Laadullinen tut- Oikarinen, Teija. 2009. Pietarsaari, 2/2/7-10, Lassfolkin kort- kimus ja sisällönanalyysi. Helsinki: Kustannusosakeyhtiö teli (PSL-08). Kaupunkiarkeologinen pelastuskaivaus 19.5.- Tammi. 8.8.2008. (Unpublished excavation report) Museovi- Vlachidis, Andreas, Binding, Ceri, Tudhope, Douglas and rasto: Rakennushistorian osasto. May, Keith. 2010. Excavating grey literature: A case Ontology for Museum Domain (MAO). 2011. Semantic Com- study on the rich indexing of archaeological documents puting Research Group. Available http://www.seco. via natural language-processing techniques and knowl- tkk.fi/ontologies/mao/. edge-based resources. Aslib proceedings 62: 466-75. Read, Dwight W. 2009. Artifact classification: A conceptual Xia, Jingfeng. 2006. Electronic publishing in archaeology. and methodological approach. Walnut Creek, Calif.: Left Journal of scholarly publishing 37: 270-87. Available http:// Coast Press. arizona.openrepository.com/arizona/bitstream/10150/ 105391/1/Archaeology.pdf

136 Knowl. Org. 40(2013)No.2 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

Knowledge Creation and Commercialization Activities in Polish Public HEUs in the Area of Technical and Engineering Sciences

Urszula Sienkiewicz* and Izabela Kijeńska-Dąbrowska**

*Kredyt Bank S.A., ul. Kasprzaka 2/8, 01-211 Warsaw, Poland, **The Information Processing Institute, Al. Niepodleglosci 188b, 00-608 Warsaw, Poland,

Urszula Sienkiewicz is a graduate of the Warsaw University of Technology, Faculty of Mathematics. From 2003-2012 she was an assistant at the Institute of Statistics and Demography at the Warsaw School of Eco- nomics, and from 2011-2012 a statistician at the Information Processing Institute. Currently she is an expert in credit risk analysis at a Polish bank. Her research interest is event history analysis.

Izabela Z. Kijeńska-Dąbrowska, PhD, is assistant professor at the Information Processing Institute, Warsaw, Po- land. She completed her dissertation at the Warsaw School of Economics in 2010. Her recent research focuses on the Polish public research and development system. Her research interests concern aspects of regional and global innovation systems, development of new technologies, knowledge based economy; economic and technical problems of the knowledge and technology transfer from research institutions to industrial sectors.

Sienkiewicz, Urszula, and Kijeńska-Dąbrowska, Izabela. Knowledge Creation and Commercialization Ac- tivities in Polish Public HEUs in the Area of Technical and Engineering Sciences. Knowledge Organiza- tion. 40(2), 136-146. 17 references.

ABSTRACT: History of knowledge organization within higher education units (HEU) changes with respect to the idea of measuring activities of academia. The visible evolution of HEU's role in the economy is indicated in the performance of particular entities. Apart from the education activity, the production of new knowledge and publication of research results are no longer the sole aspects of HEU performance. The knowledge organization structure requires entrepreneurial behaviour from academia. In this paper, activities related to the commercialization of research results performed within HEU are analysed. The study concerns units in the area of technical and engineering studies and covers different aspects of research and development (R&D) performance. There is a visible relation between the level of research/teaching team quality and publication activity and their economic influence. Statistical analyses conducted try to detect relations and/or influence of publications activity and researchers’ level of education on commercialization benefits from research projects performance. HEU with a relatively high commercialization performance are those with the highest level of publication activity. At the same time, entities with a high number of well-experienced researchers are those with significant benefits from research projects. These results are important for the idea of measuring modern HEU performance with respect to traditional knowledge organization in academia.

Received 21 November 2012; Revised and Accepted 30 January 2013

1.0 Introduction The idea for these studies arises from the great pressure that is placed on research units and researchers to foster The commercialization of research and development development of new ideas. These new ideas, in particularly “R&D” results is a complex and dynamic process which in the area of technical and engineering sciences, are to be starts from researcher’s idea and develops through a very implemented in the real economy in the form of innova- complex structure of environmental factors. In this paper, tion. Recent publications (Hage and Meeus 2009; Viale and we try to explore the basic factors of knowledge organiza- Etzkowitz 2010) underline the aspects of quality of re- tion having an influence on research commercialization. search conducted within research entities that influence the Knowl. Org. 40(2013)No.2 137 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs developmental pace of modern economies. The necessity performance. The Research Unit Questionnaire is a report of a complex approach toward the interdisciplinarity activi- which units are required to submit annually in order to ap- ties of HEU, in particular technical universities and poly- ply for grants for basic statutory science funds. The data technics, is also evident in analyses of knowledge organiza- used in the study concerns units in the area of technical tion (Silva and Ribeiro 2012). and engineering studies and covers different aspects of re- There are a number of studies in literature showing the search and development performance. relationship between research commercialization performance and other characteristics of research units at the insti- 3.0 Analysis methods and variables tutional (Owen-Smith and Powell 2003; Foltz 2007) and individual level (Meyer 2006; Agrawal and Anderson 2002; 3.1 Dependent variables O’Shea et al. 2005). The relationship at the institutional level shows that factors such as publishing activity, number In order to analyze the commercialization of research re- of researchers, and research funding are related to patent- sults, we analysed patents and benefits from research per- ing performance (Wong and Singh 2010). In the literature, formance in two ways. Firstly, analysis concerns only the there are also findings on factors having an impact on pub- dichotomous information, e.g. whether the unit is patent- lication activity such as internationalization (Abramo, ing or not and whether the unit is commercializing or D'Angelo, and Solazzi 2011) and individual institutions, in- not. Then the relationship between the quantity of individual disciplinary areas within each institution and indi- formation about the level of publications and patents as vidual organization units (faculties, departments, etc.) well as the level of publication and benefits is explored. within each area (Abramo, D'Angelo, and Pugini 2008), to- To test the relationship in question, nonparametric test gether with the size of faculty’s research budget or research of distribution equality, logistic regression, and correla- infrastructure (Baskurt 2011). The main emphasis is placed tion analysis methods were used. on publishing activity. It draws a distinction between quantity of publications and quality of publications and their 3.1.1 Patent performance distinct influence on commercialization activity (Wong and Singh 2010). The recent research (Moneda Corrochano, Our indicator of entity patent output in the database is López-Huertas, and Jiménez-Contreras 2012) gives also based on information about the number of patents issued evidence that changes of the number and quality of papers in 2005 year. The information is divided into the number might also be influenced by the number of sophisticated of domestic patents and foreign patents. According to for- co-authors undersigned. While the total value of publica- eign patents, more than 94% of entities did not have any tion productivity might decrease, its quality rises with an patent granted abroad. Therefore all the conclusions in this increase in papers published in ISI journals. There is also paper involve only the patents granted in Poland. evidence for differences in influence on patenting and revenue from sale of R&D. As Geuna and Nesta (2006) 3.1.2 Commercialization performance show, university licensing is not profitable for most universities, although some do succeed in attracting substantial The second dependent variable considered was informa- additional revenues. tion about commercialization expressed by revenue from In this paper, we examine the impact of the quality of R&D results sales. The revenue considered is in total val- the scientific research and sources of funds for research ues and then divided into revenues from domestic and on the tangible results of scientific activity, i.e. the num- foreign sources. ber of patents generated by research and revenues from the sale of R&D results. For the explanatory variables, we 3.2 Explanatory variables use information about publication activity, number of unit employees, funds for research, and the quality of the As explanatory variables, information about publication university to which the unit belongs as stated by the activity, number of unit employees, quantity of funds for Hirsch index. research activity, and quality of the university to which the unit belongs, as stated by the Hirsch index, are used. 2.0 Data origin and representativeness 3.2.1 Publication activity Data comes from the Information Processing Institute and Research Unit Questionnaire (detailed information on the The publication activity was divided into three stages. website http://nauka-polska.pl/shtml/ankieta/ankieta_jed First the number of all papers published was analysed. _informacje.shtml), which collects data about research unit Then it was divided into two categories—papers pub- 138 Knowl. Org. 40(2013)No.2 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs lished in scientific domestic journals and papers pub- which can be interpreted as percentage increase in units lished in scientific foreign journals. probability to patenting or commercializing produced by an increase in a characteristic’s value by particular unit. To 3.2.2 Number of researchers show the highest influence, the Gini statistic is calculated; this expresses the discriminatory power of variable. The This characteristic of research unit was divided into the Gini statistic is calculated as the area under ROC (receiver number of employees with PhDs or higher scientific de- operating curve) minus 0.5. The ROC curve is created for grees and the number of employees without scientific ordered variable values and the corresponding percentage degrees. of units which experienced the event under consideration. The higher the Gini is, the higher is the discriminatory 3.2.3 Funds for research power of the variable. This means that the high percentage of units did not experience the activity under considera- In the database, there is information about sources of tion, which corresponds to the lower values of the vari- funds for research in a total amount which is the total able. For instance, the increase of one research employee output from R&D activity (including output from the sale with a PhD or higher degree increases the chance for pat- of R&D results) and information about statutory budget. enting by 1.2%. According to the patenting activity, the statistically sig- 3.2.4 Quality of university nificant differences between distributions of the characteristics occur for publications activity, number of employees, The quality of university is represented by the Hirsch In- and funds for research. Based on the odds ratios, all the dex for universities during period 2000 – 2008. The h index significant characteristics have positive influence. Referring is defined as number of papers with citations higher or to total patenting activity, there is a significant influence of equal to h (Hirsch 2005). For university comparison pur- all type of publications. A slightly different situation can be poses, we used a modification of the Hirsch Index pro- observed for domestic patenting activity. The statistically posed by Molinari and Molinari (2008). This modification significant influence is only for total publications number takes into account number of publications of the univer- and domestic publications number. There is no significant sity. influence for scientific publications. The Hirsch index was calculated by Kierzek (2008) An entirely different situation emerges for commerciali- based on the databases from Institute of Scientific Infor- zation activity. There is a common characteristic of univer- mation (Philadelphia, USA): Science Citation Index Ex- sities which has statistically significant influence for all panded (SCI-EXPANDED. 1973 – present), Social Sci- three types of commercialization—the Hirsch index and ences Citation Index (SSCI, 1973 – present), and Arts & modification of the Hirsch index. In the case of revenue Humanities Citation Index (A&HCI, 1975 – present) from abroad, there are several statistically significant vari- (www.ibch.poznan.pl/PI/Sprawy_Nauki/). ables, such as the number of employees and the number of publications. However, taking into account publications 4.0 Results in contrast to patent activity, the signification is given for the number of scientific publications. Together with Descriptive statistics show that distribution of patenting Hirsch index signification, this is a clear illustration of the and commercialization activity are both very right skewed strong influence of university quality on commercialization (the skewness coefficient is positive and high), which activity. means that the level of patenting and commercializing re- Table 5 and Table 6 present the Pearson correlation co- search result in Poland is lower than the average. Moreover, efficients between explanatory variables and patenting and right skewness is a characteristic mark for all units under commercialization. They are calculated only for those units consideration. with patenting and commercialization activity. This statis- The analysis of the influence of a unit’s characteristics tics show there is no linear relation between explanatory on the propensity to patenting and commercializing is conducted and dependent variables. In order to explore ex- shown in Table 3 and Table 4. Firstly, the distinction be- istence of any relationship, the one way analysis of vari- tween distribution of the characteristics among those who ance is conducted (Table 7 and Table 8). The explanatory patent and those who are not using Kolmogorov-Smirnov variables are divided into four groups using quartiles val- statistic, which measures the highest difference between ues. Then the influence of every single unit’s characteristic cumulative distribution functions, is tested. Then the level on the group means is tested. In case the assumption of of influence on the propensity to patenting and commer- variance homogeneity is not proved, the Welch analysis of cializing is explored. This level is expressed by odds ratios variance was conducted (in Table 7 and Table 8, such vari- Knowl. Org. 40(2013)No.2 139 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

Coeff of Variation Lower Upper Variable Name Mean (%) Quartile Median Quartile Skewness Maximum All patents (#) 1,79 153,46 0,00 1,00 3,00 2,35 13,00 Domestic patent (#) 1,70 157,42 0,00 1,00 2,00 2,50 13,00 Foreign patents (#) 0,10 458,64 0,00 0,00 0,00 5,16 3,00 All publication (#) 159,44 77,87 76,00 136,00 212,00 1,88 716,00 Domestic publications (#) 125,03 86,48 51,00 106,00 167,00 2,15 637,00 Scientific publications (#) 34,65 169,78 4,00 12,00 38,00 2,94 303,00 R&D employees (#) 116,89 54,81 76,00 103,00 144,00 1,49 404,25 Employees without sci, degree (#) 31,41 86,13 11,00 28,00 43,00 1,78 157,00 Employees with PhD or higher (#) 85,48 56,21 53,00 76,00 110,00 1,09 264,25 Total output from R&D activity 4411,20 105,53 882,98 2980,87 6125,24 1,69 21030,55 (ths. PLN) Statutory Budget (ths. PLN) 3002,93 110,40 709,94 1793,05 4028,10 1,90 17537,62 Sale of R&D Results (ths. PLN) 439,58 218,05 0,00 21,76 505,31 4,12 7172,64 Sale of R&D Results, Domestic 264,33 258,30 0,00 9,19 209,79 4,59 4984,40 (ths. PLN) Sale of R&D Results, Aborad (ths. 175,26 350,94 0,00 0,00 37,12 7,70 6368,80 PLN) Citations in 2004 year (#) 2168,22 103,27 480,00 2142,00 3193,00 3,17 15853,00 Citations in 2005 year (#) 2739,76 103,26 607,00 2711,00 3928,00 3,01 19183,00 Hirsch index 30,68 38,80 22,50 31,00 34,50 1,11 77,00 Modified Hirsch index 1,38 16,87 1,21 1,34 1,62 0,32 2,09

Table 1. Descriptive statistics ables are marked with “w”). The results of the analysis of most of the significant variables, the influence direction is variance show, for instance in case of number of scientific not constant, in particular according to characteristics de- publications, there is statistically significant dependency be- scribing university quality such as citation number and tween those publications and number of patents. In other Hirsch index. words, we can say that the number of scientific publications differentiates the average number of patents. 5.0 Conclusions The statistics show that not all the variables which had an influence on patenting or commercialization activity The analyses conducted show that there are different fac- have an influence on the level of such activity. In the case tors influencing patenting performance and research result of patenting activity, those variables are domestic publica- sale performance. As far as patenting is concerned, the tions and total science funds. Moreover we can observe analysis indicates that this activity is influenced by the that the direction of this influence is not always constant. number of publications, most of all the domestic publica- For instance, the increase of total number of R&D em- tions; the number of employees with scientific degrees, ployees causes an increase of the average number of pat- PhD or higher; and funding for science (including both to- ents, but only for the three first groups. In the last group, tal and statutory funds). In the case of revenue from the which is the group of units with the highest number of sale, the greater importance is found for the characteristics employees, the average number of patents is lower than in defining the quality of the scientific environment of the the former groups. This situation occurs for both total individual and quality of the entity, as expressed in the number of patents and number of domestic patents. number of scientific publications published in interna- According to revenue from the sale of R&D results, tional scientific journals. Such characteristics as number of there are only several variables for which we saw statisti- publications, including in particular the number of domes- cally significant results and what is more, for the majority tic publications, does not skew the sale of research results of this the significance is at the 0.1 level. Additionally, for in any of the three considered types of revenue (total, do- 140 Knowl. Org. 40(2013)No.2 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

Pearson correlations correlations Pearson Table 2. Table Knowl. Org. 40(2013)No.2 141 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

Patents, Total Patents, Domestic Variable name K-S Odds Ratio Gini K-S Odds Ratio Gini All publication (#) 0,353 * 1,006468 * 0,397 0,336 * 1,004583 * 0,369 Domestic publications (#) 0,271 ** 1,006408 * 0,334 0,254 ** 1,004253 ** 0,306 Scientific publications (#) 0,298 * 1,005696 ** 0,301 0,283 1,005155 0,283 R&D employees (#) 0,264 ** 1,006660 ** 0,269 0,252 ** 1,004708 ** 0,240 Employees without sci. degree (#) 0,186 0,999205 0,030 0,197 0,996535 0,057 Employees with PhD or higher (#) 0,268 ** 1,012457 * 0,325 0,259 ** 1,009812 ** 0,298 Total output from R&D activity (ths. PLN) 0,376 * 1,000103 * 0,391 0,364 * 1,000077 ** 0,363 Statutory Budget (ths. PLN) 0,338 * 1,000131 * 0,357 0,321 * 1,000099 ** 0,332 Sale of R&D Results, Domestic (ths. PLN) 0,297 1,000206 0,239 0,280 1,000168 0,241 Sale of R&D Results, Abroad (ths. PLN) 0,189 1,001133 0,161 0,167 1,000034 0,129 Citations in 2004 year (#) 0,191 1,000080 0,181 0,175 1,000064 0,153 Citations in 2005 year (#) 0,191 1,000057 0,180 0,175 1,000043 0,151 Hirsch index 0,191 1,015921 0,169 0,175 1,011511 0,140 Modified Hirsch index 0,097 2,062918 0,088 0,090 1,692073 0,068 Table 3. Kolmogorov-Smirnov statistics and logistic regression results – Patents

Revenue from R&D, Total Revenue from R&D, Domestic Revenue from R&D, Abroad Variable name K-S Odds Ratio Gini K-S Odds Ratio Gini K-S Odds Ratio Gini All patents (#) 0,188 1,192238 0,228 0,186 1,138484 0,215 0,174 1,086160 0,165 Domestic patent 0,179 1,214503 0,227 0,173 1,152158 0,217 0,149 1,070405 0,137 (#) Foreign patents (#) 0,030 1,044110 0,029 0,015 0,877415 0,015 0,131 2,187855 0,129

All publication (#) 0,245 1,001893 0,217 0,206 1,001665 0,179 0,285 ** 1,003718 ** 0,334 Domestic publica- 0,129 1,000564 0,086 0,186 1,001598 0,174 0,172 1,002536 0,134 tions (#) Scientific publica- 0,164 1,007735 0,139 0,087 1,001811 0,075 0,369 * 1,007474 ** 0,386 tions (#) R&D employees 0,226 1,003923 0,214 0,181 1,003103 0,157 0,292 ** 1,005664 ** 0,285 (#) Employees without 0,096 1,001145 0,034 0,101 1,001897 0,031 0,080 0,999391 -0,004 sci. degree (#) Employees with 0,294 * 1,006595 ** 0,248 0,205 1,004886 0,172 0,346 * 1,010473 * 0,347 PhD or higher (#) Total output from R&D activity (ths. 0,312 * 1,000073 ** 0,290 0,214 1,000045 0,198 0,527 * 1,000193 * 0,583 PLN) Statutory Budget 0,241 1,000041 0,189 0,195 1,000012 0,163 0,481 * 1,000180 * 0,475 (ths. PLN) Sale of R&D Re- sults, Domestic 0,910 * 1,492514 * 0,910 0,489 * 1,001188 * 0,539 (ths. PLN) Sale of R&D Re- sults, Abroad (ths. 0,449 1,177181 0,449 0,282 * 1,001279 ** 0,258 PLN) Citations in 2004 0,213 1,000195 0,143 0,209 1,000107 0,119 0,203 1,000070 0,148 year (#) Citations in 2005 0,290 ** 1,000199 ** 0,229 0,277 1,000114 0,209 0,226 1,000071 0,226 year (#) Hirsch index 0,389 * 1,048751 ** 0,261 0,267 ** 1,030471 ** 0,173 0,365 * 1,040889 ** 0,335 Modified Hirsch 0,409 * 18,709834 * 0,366 0,285 ** 6,349258 ** 0,237 0,397 * 22,180335 * 0,424 index Table 4. Kolmogorov-Smirnov statistics and logistic regression results – Revenue from R&D 142 Knowl. Org. 40(2013)No.2 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

patents producing units among correlations 5. Pearson Table Knowl. Org. 40(2013)No.2 143 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

Pearson correlations among units having revenue from sale of R&D of sale from revenue having units among correlations 6. Pearson Table 144 Knowl. Org. 40(2013)No.2 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

Patents, Total Patents, Domestic

Variable name Group means Group means F Statistic F Statistic 1 2 3 4 1 2 3 4 Domestic patent (#) 55,71 W * 1,18 2,05 3,58 8,27 27,44 w * 1,00 2,00 3,10 7,69

All publication (#) 3,36 W ** 2,21 3,00 4,05 4,50 3,04 w ** 2,11 3,00 3,75 4,35

Domestic publications (#) 0,48 2,95 3,81 3,88 3,06 0,37 2,89 3,67 3,63 2,94

Scientific publications (#) 2,69 W ** 2,81 2,94 2,37 5,72 2,84 w ** 2,57 2,88 2,26 5,56

R&D employees (#) 5,62 W * 1,90 3,50 4,37 4,06 5,53 w * 1,80 3,35 4,16 3,94

Employees without sci, de- 0,97 3,89 2,70 3,00 4,11 0,94 3,53 2,89 2,63 4,11 gree (#) Employees with PhD or 5,68 W * 1,95 3,44 3,89 4,56 3,20 ** 1,85 3,44 3,33 4,67 higher (#) Total output from R&D activ- 1,31 2,37 3,68 3,42 4,28 1,50 2,26 3,61 3,11 4,22 ity (ths, PLN) Statutory Budget (ths, PLN) 2,20 ** 2,47 2,95 3,53 4,83 2,59 ** 2,37 2,67 3,37 4,78

Sale of R&D Results, Do- 0,54 3,15 4,33 3,05 3,61 0,74 2,88 4,36 3,00 3,50 mestic (ths, PLN) Sale of R&D Results, Abroad 0,30 3,37 . 2,88 3,83 0,20 3,27 . 2,71 3,56 (ths, PLN) Citations in 2004 year (#) 0,21 3,29 3,47 3,93 3,08 0,15 3,18 3,11 3,73 3,25

Citations in 2005 year (#) 0,82 3,29 2,94 4,41 3,08 0,51 3,18 2,76 4,00 3,25

Hirsch index 0,12 3,29 3,75 3,56 3,15 0,02 3,18 3,31 3,39 3,33

Modified Hirsch index 0,73 3,36 3,39 4,50 2,82 0,84 3,04 3,39 4,42 2,75

* - 0.01

** - 0.1

w - Welch Anova

Table 7. Analysis of variance results for patenting units mestic, and abroad). In addition, studies have proven that, be less affected by the quality of the research unit and even if there exists a factor which skews the fact of patent- more affected by the quality of individuals—the research- ing or sale, such as the number of publications, there might ers. On the other side, the process can be considered con- not be a conclusion that the increase in the number of pub- trary, starting from the demand for specific technology. An lications will increase the number of patents or increase the entity requiring particular technology needs to select the revenue from the sale. Most such variables do not have an research units, as guided by certain criteria. This can en- impact on the level of the studied phenomenon, or the in- hance research units with better quality and more experi- fluence is not linear. Finally, it is vital to understand the enced researchers as favoured by enterprises. Finally, it processes of knowledge creation and commercialization of causes stronger relationships between research unit quality research results with respect to the overall activity of HEU. and commercialization of R&D results. However, answer- Knowledge creation in the academic environment might ei- ing these research questions requires analysis at the indi- ther arise from an idea of the researcher or evolve from the vidual level—the researcher (demand), as well as at recipi- particular market (economy) need. ents of technology and knowledge (supply). The results of the analysis indicate that none of the in- vestigated mechanisms have an influence on the commer- References cialization of R&D results process. This might be caused by the dual nature of the research commercialization proc- Abramo, Giovanni, D’Angelo, Ciriaco Andrea and Pugini, ess. From one perspective, this process starts with the re- Fabio. 2008. The measurment of Italian universities’ searcher's idea and then develops through research and re- research productivity by a non parametric-bibliometric sults in patent or in direct commercial application. In this methodology. Scientometrics 76: 225-44. case, we can expect that the number of patents or the Abramo, Giovanni, D’Angelo, Ciriaco Andrea, C.A., So- amount of revenue from the sale of research results may lazzi Marco. 2011. The relationship between scientists’ Knowl. Org. 40(2013)No.2 145 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

for revenue fromR&D. sale of Analysis of variance results of variance results Analysis Table 8. Table 146 Knowl. Org. 40(2013)No.2 U. Sienkiewicz, and I. Kijeńska-Dąbrowska. Knowledge Creation and Commercialization Activities in Polish Public HEUs

research performance and the degree of internation- Molinari, Jean-François and Molinari, Alain. 2008. A new alization of their research. Scientometrics 86: 629-43. methodology for ranking scientific institutions. Scien- Agrawal, Ajay and Anderson, Rebecca. 2002. Putting pat- tometrics. 75: 163-74. ents in context: Exploring knowledge transfer from Moneda Corrochano, Mercedes, López-Huertas María MIT. Management science 4: p. 44-60. José and Jiménez-Contreras Evaristo. 2012. Spanish Baskurt, Oguz K. 2011. Time series analysis of publica- research in knowledge organization (2002-2010). tion counts of a university: what are the implications? Knowledge organization 40: 28-41. Scientometrics 86: 645-56. O’Shea, Rory P., Allen, Thomas J., Chevalier, Arnaud and Foltz, Jeremy D., Barham, Bradford L. and Kim, Kwan- Roche, Frank. 2005. Entrepreneurial orientation, tech- soo. 2007. Synergies of trade-offs in university life sci- nology transfer and spinoff performance of U.S. uni- ences research. American journal of agricultural economics versities. Research Policy 34: 443-57. 89: 353-67. Owen-Smith, Jason and Powell, Walter W. 2003. The ex- Geuna, Aldo and Nesta, Lionel J.J. 2006. University patent- panding role of university patenting in the life sci- ing and its effects on academic research: The emerging ences: Assesing the importance of experience and European evidence. Research policy 35: 790-807. connectivity. Research policy 32: 1695-711. Hage, Jerald and Meeus, Marius. 2009. Innovation, science Silva, Armando Malheiro and Ribeiro, Fernanda. 2012. and institutional change, a research handbook. New York: Documentation / information and their paradigms: Oxford University Press. Characterization and importance in research, educa- Hirsch, Jorge E. 2005. An index to quantify an individuals tion, and professional practice. Knowledge organization scientific research output. PNAS 102: 16569-72. 39: 111-24. Kierzek, Ryszard. 2008. Polska nauka w indeksie Hirsch. Viale, Riccardo and Etzkowitz, Henry. 2010. The capitali- Sprawy Nauki 137 (29). zation of knowledge: A triple helix of university-industry- Meyer Martin. 2006. Are patenting scientists the better government. Cheltenham, UK: Edward Elgar. scholars? An exploratory comparison of inventor- Wong, Poh Kam and Singh, Annette. 2010. University authors with their non-inventing peers in nano-science patenting activities and their link to the quantity of and technology. Research policy 35: 1646-62. scientific publications. Scientometrics 83: 271-94.

Knowl. Org. 40(2013)No.2 147 ISKO News

ISKO News Edited by Hanne Albrechtsen Communications Editor

Conference ISKO-France, 10-1 October 2013, Paris Collaborative information retrieval: Specification of information retrieval problem Contexts, languages and cultures in knowledge Representation of actors in collaborative informa- organization tion retrieval Communication in collaborative information retrieval The main focus of the French Chapter of ISKO since its User-centered design and use: creation in 1996 has been on the issues of knowledge or- Decision support information system ganization. This topic has been treated in terms of organ- Customizing responses izational structures, technical tools for mediation and also User modeling forms and mechanisms for knowledge sharing. These issues are at the core of the process of production, structure Instruction to authors and access to information. In continuation of these reflections, the 9th edition of ISKO-France conference aims to – Format: word or rtf address issues related to the relevance of information – Articles could be in English or in French through contextualization in the different phases of these – Page size : 1500 words for an extended abstract (paper processes. The conference is in collaboration with DI- title, authors, authors’ emails, conference themes) CEN-IDF, CNAM, Université Marne-la-Vallée – Proposal should be sent to philippe@kislin .fr The democratization of information and communication technology, especially the Internet, enables a greater Important dates: number of producers and distributors to provide information, as it offers a greater number of users to have access. – 10 April 2013 : Deadline for extended abstract However, the production, provision and access to informa- – 30 April 2013 : Notification of acceptance tion are related to the use of this information, making the – 30 June 2013 : Deadline for the final version of ac- issues related to the relevance of information essential. cepted papers Are invited, original work on theoretical, methodologi- – 25 September : Deadline for registration cal and application aspects, focusing on the multiple dimensions of production, organization and exploitation of knowledge. Publication of ISO 25964-2, Topics (not exhaustive) the standard for interoperability of thesauri

History and epistemology of knowledge organization: Published on 4 March 2013, ISO 25964-2 deals with inter- Universal classifications, historical and anthropo- operability between thesauri and other vocabularies, espe- logical approaches cially the principles and practice of mapping between Classifications and Web Knowledge them. Types of vocabulary to be considered include classi- Contextualization of information: fication schemes (including those for records manage- Contextualization and annotation ment), taxonomies, subject heading schemes, name author- Usage oriented annotation ity lists, ontologies, terminologies, and synonym rings. Contextualization and folksonomy Interoperability is defined as the ability of two or more Contextualization and information visualization systems or components to exchange information and to Competitive intelligence, Territorial intelligence use the information that has been exchanged. Plainly this Information and contextualization in decision- ability is fundamental to the workings of the information making networks we rely on so much, and even more to the de- Specification of a decisional problem velopment of the Semantic Web (SW). Two levels of in- From decisional problem to information problem teroperability apply to thesauri: 148 Knowl. Org. 40(2013)No.2 ISKO News a. presenting data in a standard way to enable import and now freely available from the ISO 25964 Secretariat site use in other systems (dealt with in ISO 25964 Part 1, at http://www.niso.org/schemas/iso25964/. Although which was published in 2011) not formally part of either standard, this table is depend- b. complementary use of vocabularies, such as providing ent on both of them. It is an example of how the user mappings between the terms/concepts of one thesau- community continues to develop practical tools to sup- rus and those of another (covered in ISO 25964 Part 2) port the emerging Semantic Web.

Unless mappings are prepared with care as recommended Getting hold of a copy in ISO 25964-2, any SW inferences based on them are liable to deliver misleading conclusions to the unsuspecting Like any other ISO standard, either or both parts of ISO Web surfer. 25964 can be purchased from the national standards bo- The full title of Part 2 is Information and documen- dy in your country e.g. AFNOR, BSI, DIN or NISO. Al- tation - Thesauri and interoperability with other vo- ternatively you can order it directly from ISO in Switzer- cabularies - Part 2: Interoperability with other vo- land at http://www.iso.org/iso/store.htm. You can also cabularies. view the documents at some public/academic libraries. Important topics covered in the standard include Journal editors may request a review copy from ISO’s structural models for mapping, guidelines on mapping Head of Marketing, Roger Frost . types, and how to handle pre-coordination (which occurs especially in classification schemes, taxonomies and sub- Project Organization ject heading schemes). The overall focus of ISO 25964 is on thesauri, and Development of the standard was managed by a Working with the exception of terminologies, comparable stan- Group known as ISO TC46/SC9/WG8, which has par- dards do not exist for the other vocabulary types. Rather ticipants from 17 countries: Belgium, Bulgaria, Canada, than try to standardize them, Part 2 deals only with inter- China, Denmark, Finland, France, Germany, New Zea- operability between these and thesauri. Thus the clause land, Russia, South Africa, South Korea, Spain, Sweden, on each vocabulary type presents: UK, Ukraine, and USA. The Group is chaired by Stella Dextre Clarke of the UK, and its Secretariat is provided – Key characteristics of the vocabulary (descriptive, not by NISO (USA). Active members have included: normative) – Semantic components/relationships (descriptive, not Sylvie Dalbin (FR) Jutta Lindenthal (DE) normative) Johan De Smedt (BE) Marianne Lykke (DK) – Where applicable, recommendations for mapping be- F. Javier García Marco (ES) Esther Scheven (DE) tween the vocabulary and a thesaurus (normative). Michèle Hudon (CA) Douglas Tudhope (GB) Daniel Kless (DE) Leonard Will (GB) In the case of ontologies, terminologies, and synonym Traugott Koch (DE) Marcia Zeng (US) rings, mapping to or from a thesaurus is not often useful. Other forms of complementary use are recommended. This is especially true of ontologies, which in the context of the Semantic Web can be deployed in combination with thesauri. ISO 25964-2 clarifies in some detail the differences between thesauri and ontologies, with examples to illustrate the potential for interoperable function.

Practical implementation and continuing work

What about SKOS, the complementary W3C standard for publishing Simple Knowledge Organization Systems on the Web? Happily the development teams responsible for SKOS and ISO 25964 (respectively) have maintained a close working relationship throughout, leading to good compatibility between the standards. Jointly they have gone on to establish a table showing correspondence between ISO 25964 and SKOS/SKOS XL data models, Knowl. Org. 40(2013)No.2 149 Book Reviews

Book Reviews

Edited by Joseph T. Tennis

Book Review Editor

The Disorder of Things: Metaphysical Foundations of the Disunity ated with meaning and purpose. He emphasizes the im- of Science by John Dupré. Massachusetts; London: Harvard portance of classification as a lens through which an or- University Press, 1993, 308p. ISBN0-674-21261-4 (Hb) ganism is described completely and argues that the current classification methods used within the biology do- Human Nature and the Limits of Science by John Dupré. Ox- main are ineffective in presenting the diversity of mean- ford; New York: Oxford University Press, 2001, 201p. ing associated with an organism. He proposes an alter- ISBN 0-19-926550-X (Pb) nate method of classification in which cross-classification is used to identify complex relationships. He presents evi- John Dupré emphasizes the pluralistic and social aspects dence to support such a classification from within the bi- of knowledge in his two books The Disorder of Thing and ology domain and argues that there is no place for a uni- Human Nature and The Limits of Science. In The Disorder of que privileged scheme of classification that assigns every- Things, Dupré presents two ideas that for him are closely thing to a class defined by common possession of the related: a rejection of the notion that science can represent appropriate essence. Here Dupré’s discourse regarding a single unified project, and a metaphysical assertion that classification is all too familiar to those in the knowledge the contents of the world are diverse such that each sub- organization domain. The challenges so clearly identified ject has its own characteristic behavior and interaction. He in this chapter are the very challenges that knowledge or- begins by presenting a comprehensive discussion of the ganization scholars have earnestly examined through the three philosophical doctrines associated with the metaphor years and are the basis of the philosophical underpin- of an ordered universe: determinism, reductionism, and nings associated with faceted classification. essentialism. Dupré interweaves his notion of a universe In Part II of the book, discussion turns to pluralism. lacking order throughout this discussion, skillfully present- Here Dupré argues against the belief that the world is ing abundant evidence drawn from the biology domain to composed of a single substance that is common to all validate his thesis. He concludes that the contents of the organisms. His argument centers on the concepts of ma- world are diverse such that each subject has its own charac- terialism and reductionism, which he identifies as the ba- teristic behavior and interaction. He rejects the notion of sis for the theory of scientific unity. He concludes this “natural kinds” and the unique orderly organization of section by refuting these concepts, stating that biology things that exist and embraces a pluralistic stance. The Dis- consists of many complex and interdependent entities order of Things is organized in four parts: I: Natural Kinds that cannot be presented through a unified method. He and Essentialism; II: Reductionism; III: The Limits of draws upon examples from the sciences of ecology and Causality; and IV: Some Consequences of Disorder. genetics to support his argument against scientific unity. Dupré’s discussion in Part I of the book focuses on classi- In parts III and IV of the book, Dupré presents his ar- fication. Discussions in the remaining three parts of the gument against casual order defending his belief that the book provide interesting reading related to pluralism. Be- occurrence of casual order is not as prominent as alleged. low is a summary of the key concepts. Here, Dupré describes a pluralistic epistemology, as is In Part I of the book Dupré introduces his theory of presented by the later philosophy of Wittgenstein, in pluralism through a discussion of ordinary language and which science can be seen in terms of the family resem- classification. Dupré presents classification in the tradi- blance concept. He concludes the book by stating that tional sense, an orderly, unique and perhaps hierarchical there can be nothing unique about science because there arrangement of things. He argues, as many knowledge is nothing common to the various domains of science organization scholars have before him, that the process and science is a human product evaluated only in terms of classification is difficult due to the complexity of lan- of its contributions to the success of humanity. guage and the challenges associated with representing the In his later work Human Nature and The Limits of Science, many social and cultural aspects as well as aspects associ- Dupré reinforces the importance of pluralism through 150 Knowl. Org. 40(2013)No.2 Book Reviews his discussion of human nature. Dupré begins the book Relevance to the Knowledge Organization Domain by presenting the bond that exists between humans and science, a bond that humans find essential in providing The Disorder of Things and Human Nature and The Limits of explanations to natural occurrences and ultimately en- Science make interesting reading for the knowledge organi- hancing human understanding. However, Dupré dis- zation audience. In both books, Dupré highlights one of agrees with such devotion to science, stating that science the basic tenets of the knowledge organization domain, alone cannot answer the most complex questions that pluralism, as the lens through which knowledge is viewed humans can ask and specifically, questions regarding hu- and through which all aspects of life including the social man nature and behavior. Instead Dupré advocates the and cultural are embraced. In The Disorder of Things Dupré combination of empirical knowledge stemming from the challenges the concept of “natural kinds,” discovering that sciences with wisdom and insight into human nature, an orderly, unique and hierarchical arrangement of knowl- stemming from humanistic studies. Dupré believes that edge is not easily accomplished due to the complexity nat- only through such a pluralistic approach can humans un- urally inherent in knowledge. Through several examples derstand nature. As in The Disorder of Things, Dupré is op- drawn from the biology domain, Dupré presents a valid posed to the mechanistic and reductionist view of the defense for the pluralistic representation of knowledge. world in which the world is understood through a de- This theme has been discussed by various knowledge or- tailed analysis of how its components work disregarding ganization scholars but recently, the work of Smiraglia, van the contextual or environmental influences. Dupré be- den Huevel and Dousa (2011) best highlights the impor- lieves an understanding of the world stems from a thor- tance of a pluralistic view of knowledge by introducing the ough investigation of not only how things work but also concept of a “multiverse of knowledge,” where knowledge what they do and why. This type of understanding incorpo- embodies physical, conceptual and social elements. Also rates the interaction between humans and their environ- relevant to the knowledge organization audience is Dupré’s ments including the social context in which humans exist. discussion of the challenges associated with classification, a Dupré opposes the mechanistic and reductionist view prevailing discourse in the knowledge organization domain that presents humans as machines with distinguishable over the years. In Human Nature and The Limits of Science, mechanical sub-units designed to respond to particular the social aspect of knowledge is highlighted reaffirming features of the environment but ignoring the diversity of one of the basic tenets within the knowledge organization human behavior. Dupré believes that it is only the plural- domain and one that has been widely explored by various istic approach that provides the key to understanding the knowledge organization scholars including Hjørland and genuine autonomy of much of human behavior, for Pedersen (2005), Mai (1998, 2009), Campbell (2006), genuine autonomy can only exist in the interaction be- Kwasnik (1999), and Mills (2004). tween humans and society. Human Nature and the Limits of Science contains seven chapters however, of particular in- References terest to the knowledge organization audience is Chapter; 2. In this chapter, Dupré illustrates the problematic and Campbell, D. Grant. 2006. A phenomenological frame- controversial grounding assumptions of evolutionary work for the relationship between the semantic web psychology, the belief that natural selection is an engine and user-centered tagging systems. In: Furner, Jona- that directs changes in the frequency of genes towards than and Tennis, Joseph T., eds. Information realities: sha- adaptive end. His arguments opposing this belief center ping the digital future for all: proceedings of the 17th American around human language and his belief that humans are Society for Information Science and Technology SIG/CR Clas- ontologically dependent on their social context and that sification Research Workshop, 3-8 November 2006, Austin, aspects of the mind depend ontologically on the com- Texas. Available https://journals.lib.washington.edu/ munity in which they are embedded. Social context and index.php/acro/article/viewFile/12489/10987. community is also an important construct within the Hjørland, Birger and Karsten Nissne Pedersen. 2005. A knowledge organization domain where the meaning of substantive theory of classification for information re- words is examined from the community in which the trieval. Journal of documentation 61: 582-97. words and the language are used (Mai 2004) and knowl- Kwasnik, Barbara H. 1999. The role of classification in edge is a reflection of communities and the society in knowledge representation and discovery. Library trends which they belong (Hjørland and Pedersen 2005). Simi- 48 no. 1: 22-47. larly, tagging is yet another process where members of a Mai, Jens-Erik. 1998. Organization of knowledge: An in- community share their reflections and provide a glimpse terpretive approach. In: Toms, Elaine G., Campbell, D. of life within a community (Campbell 2006). Grant and Dunn, Judy, eds. Information science at the dawn of the next millennium: proceedings of the 26th Annual Con- Knowl. Org. 40(2013)No.2 151 Book Reviews

ference of the Canadian Association for Information Science, 3- tures in universes of knowledge. In: Slavic, Aida and 5 June 1998, Ottawa, Ontario, Canadian Association Civallero, Edgardo, eds. Classification and ontology: formal for Information Science, pp. 231- 42. approaches and access to knowledge: proceeding of the Internati- Mai, Jens-Erik. 2004. Classification in context: Relativity, onal UDC Seminar, 19-20 September 2011, The Hague, reality, and representation. Knowledge organization 31: 39- Ergon-Verlag, Würzburg, pp. 25-41. 48. Mai, Jens-Erik. 2009. Classification in a social world: bias and trust. Journal of documentation 66: 627-42. Elizabeth Milonas Mills, Jack. 2004. Faceted classification and logical division Palmer School of Library and Information Science in information retrieval. Library trends 52 no. 3: 541-70. Long Island University, CW Post Smiraglia, Richard P., Charles van den Heuvel, and Tho- Brookville, New York 11548 mas M. Dousa. Interactions between elementary struc- [email protected]

Knowl. Org. 40(2013)No.2

KNOWLEDGE ORGANIZATION KO

Publisher References should be listed alphabetically by author at the end of the article. Author names should be given as found in the sources (not ab- ERGON-Verlag GmbH, Keesburgstr. 11, D-97074 Würzburg breviated). Journal titles should not be abbreviated. Multiple citations to Phone: +49 (0)931 280084; FAX +49 (0)931 282872 works by the same author should be listed chronologically and should E-mail: [email protected]; http://www.ergon-verlag.de each include the author’s name. Articles appearing in the same year should have the following format: “Jones 2005a, Jones 2005b, etc.” Is- Editor-in-chief (Editorial office) sue numbers are given only when a journal volume is not through- paginated. Dr. Richard P. SMIRAGLIA (Editor-in-Chief), School of Information Examples: Studies, University of Wisconsin, Milwaukee, Northwest Quad Building Dahlberg, Ingetraut. 1978. A referent-oriented, analytical concept B, 2025 E Newport St., Milwaukee, WI 53211 USA. E-mail: smi- theory for INTERCONCEPT. International classification 5: 142-51. [email protected] Howarth, Lynne C. 2003. Designing a common namespace for searching metadata-enabled knowledge repositories: an international Instructions for Authors perspective. Cataloging & classification quarterly 37n1/2: 173-85. Pogorelec, Andrej and Šauperl, Alenka. 2006. The alternative model Manuscripts should be submitted electronically (in Word format) in Eng- of classification of belles-lettres in libraries. Knowledge organization 33: lish only via email to the editor-in chief and should be accompanied by 204-14. an indicative abstract of 150 to 200 words. Manuscripts of articles Schallier, Wouter. 2004. On the razor’s edge: between local and should fall within the range 6,000-10,000 words. Longer manuscripts overall needs in knowledge organization. In McIlwaine, Ia C. ed., Knowl- will be considered on consultation with the editor-in-chief. edge organization and the global information society: Proceedings of the Eighth In- A separate title page should include the article title and the author’s ternational ISKO Conference 13-16 July 2004 London, UK. Advances in name, postal address, and E-mail address, if available. Only the title of knowledge organization 9. Würzburg: Ergon Verlag, pp. 269-74. the article should appear on the first page of the text. Smiraglia, Richard P. 2001. The nature of ‘a work’: implications for the or- To protect anonymity, the author’s name should not appear on the ganization of knowledge. Lanham, Md.: Scarecrow. manuscript, and all references in the body of the text and in footnotes Smiraglia, Richard P. 2005. Instantiation: Toward a theory. In Vaughan, that might identify the author to the reviewer should be removed and Liwen, ed. Data, information, and knowledge in a networked world; Annual confer- cited on a separate page. ence of the Canadian Association for Information Science … London, Ontario, June Criteria for acceptance will be appropriateness to the field of the jour- 2-4 2005. Available http://www.cais-acsi.ca/2005proceedings.htm. nal (see Scope and Aims), taking into account the merit of the contents Illustrations should be kept to a necessary minimum and should be and presentation. The manuscript should be concise and should con- embedded within the document. Photographs (including color and half- form as much as possible to professional standards of English usage tone) should be scanned with a minimum resolution of 600 dpi and and grammar. Manuscripts are received with the understanding that they saved as .jpg files. Tables and figures should be embedded within the have not been previously published, are not being submitted for publi- document. Tables should contain a number and title at the bottom, and cation elsewhere, and that if the work received official sponsorship, it all columns and rows should have headings. All illustrations should be has been duly released for publication. Submissions are refereed, and cited in the text as Figure 1, Figure 2, etc. or Table 1, Table 2, etc. authors will usually be notified within 6 to 10 weeks. The entire manuscript should be double-spaced, including notes and The text should be structured by numbered subheadings. It should references. contain an introduction, giving an overview and stating the purpose, a Upon acceptance of a manuscript for publication, authors must main body, describing in sufficient detail the materials or methods used provide a wallet-size photo and a one-paragraph biographical sketch and the results or systems developed, and a conclusion or summary. (fewer than 100 words). The photograph should be scanned with a Footnotes are accepted only in rare cases and should be limited in minimum resolution of 600 dpi and saved as a .jpg file. number; all narration should be included in the text of the article. Para- graphs should include a topic sentence and some developed narrative; a typical paragraph has several sentences. Italics may not be used for em- Advertising phasis. Em-dashes should not be used as substitutes for commas. Italics should not be used for emphasis. Em-dashes should be used Responsible for advertising: ERGON-Verlag GmbH, Keesburgstr. 11, as substitutes for commas. Paragraphs should include a topic sentence 97074 Würzburg (Germany). and some developed narrative. A typical paragraph has several sen- © 2013 by ERGON-Verlag GmbH. tences. All Rights reserved. Reference citations within the text should have the following form: (Au- thor year). For example, (Jones 1990). Specific page numbers are required for quoted material, e.g. (Jones 1990, 100). A citation with two KO is published bi-monthly by ERGON-Verlag GmbH. authors would read (Jones and Smith, 1990); three or more authors – The price for the print version is € 229,00/ann. including air- would be: (Jones et al., 1990). When the author is mentioned in the text, mail delivery. only the date and optional page number should appear in parenthesis – – The price for the print version plus access to the online version e.g. According to Jones (1990), … (PDF) is € 258,00/ann. including airmail delivery. Knowl. Org. 40(2013)No.2

KO KNOWLEDGE ORGANIZATION

Scope Aims

The more scientific data is generated in the impetuous present times, Thus, KNOWLEDGE ORGANIZATION is a forum for all those in- the more ordering energy needs to be expended to control these data in terested in the organization of knowledge on a universal or a domain- a retrievable fashion. With the abundance of knowledge now available specific scale, using concept-analytical or concept-synthetical approaches, as well as quantitative and qualitative methodologies. the questions of new solutions to the ordering problem and thus of im- KNOWLEDGE ORGANIZATION also addresses the intellectual proved classification systems, methods and procedures have acquired and automatic compilation and use of classification systems and thesauri unforeseen significance. For many years now they have been the focus in all fields of knowledge, with special attention being given to the prob- of interest of information scientists the world over. lems of terminology. Until recently, the special literature relevant to classification was KNOWLEDGE ORGANIZATION publishes original articles, published in piecemeal fashion, scattered over the numerous technical reports on conferences and similar communications, as well as book reviews, letters to the editor, and an extensive annotated bibliography of journals serving the experts of the various fields such as: recent classification and indexing literature.

KNOWLEDGE ORGANIZATION should therefore be available philosophy and science of science at every university and research library of every country, at every infor- science policy and science organization mation center, at colleges and schools of library and information sci- mathematics, statistics and computer science ence, in the hands of everybody interested in the fields mentioned library and information science above and thus also at every office for updating information on any topic related to the problems of order in our information-flooded times. archivistics and museology KNOWLEDGE ORGANIZATION was founded in 1973 by an journalism and communication science international group of scholars with a consulting board of editors repre- industrial products and commodity science senting the world’s regions, the special classification fields, and the sub- terminology, lexicography and linguistics ject areas involved. From 1974-1980 it was published by K.G. Saur Ver- lag, München. Back issues of 1978-1992 are available from ERGON- Beginning in 1974, KNOWLEDGE ORGANIZATION (formerly Verlag, too. INTERNATIONAL CLASSIFICATION) has been serving as a As of 1989, KNOWLEDGE ORGANIZATION has become the official organ of the INTERNATIONAL SOCIETY FOR KNOWL- common platform for the discussion of both theoretical background EDGE ORGANIZATION (ISKO) and is included for every ISKO- questions and practical application problems in many areas of concern. member, personal or institutional in the membership fee (US $ 55/ In each issue experts from many countries comment on questions of an US $ 110). adequate structuring and construction of ordering systems and on the Rates: From 2013 on for 6 issues/ann. (including indexes) problems of their use in opening the information contents of new litera- € 229,00 (forwarding costs included) for the print version resp. € 258,00 ture, of data collections and survey, of tabular works and of other ob- for the print version plus access to the online version (PDF). Member- ship rates see above. jects of scientific interest. Their contributions have been concerned with ERGON-Verlag GmbH, Keesburgstr. 11, D-97074 Würzburg; Phone: +49 (0)931 280084; FAX +49 (0)931 282872; E-mail: ser- (1) clarifying the theoretical foundations (general ordering theory/ [email protected]; http://www.ergon-verlag.de science, theoretical bases of classification, data analysis and re- Founded under the title International Classification in 1974 by Dr. duction) Ingetraut Dahlberg, the founding president of ISKO. Dr. Dahlberg (2) describing practical operations connected with indexing/classifi- served as the journal’s editor from 1974 to 1997, and as its publisher cation, as well as applications of classification systems and (Indeks Verlag of Frankfurt) from 1981 to 1997. thesauri, manual and machine indexing The contents of the journal are indexed and abstracted in Social Sci- (3) tracing the history of classification knowledge and methodology ences Citation Index, Web of Science, Information Science Abstracts, INSPEC, (4) discussing questions of education and training in classification Library and Information Science Abstracts (LISA), Library, Information Science & Technology Abstracts (EBSCO), Library Literature and Information Science (5) concerning themselves with the problems of terminology in gen- (Wilson), PASCAL, Referativnyi Zhurnal Informatika, and Sociological Ab- eral and with respect to special fields. stracts.