I S K O

1. Introduction

The Bliss Bibliographic Classification, Second Edition (BC2), is the radically revised and expanded version of → Henry Evelyn Bliss’s original Bibliographic Classification (BC1) published between 1940 and 1953. The Second edition, which began publication in 1977, maintains the general structure and order of BC1, but validates that structure through the use of → facet analysis to determine the detailed internal composition of individual disciplinary classes. The revision has been a radical one in the expectation that it would render BC2 usable in the medium term. In the process of revision much use has been made of the experience of users of BC1, and of the work of the Classification Research Group, several of whose members have been significant contributors.

The classification has been adopted by a modest number of special and academic libraries in the United Kingdom, and it also serves as an exemplar of a faceted → knowledge organization system, and a free-to-access source of terminology in a variety of disciplines. Alongside the → Colon Classification (Ranganathan 1960), BC2 is one of only two widely known faceted general classification schemes in existence and it faces many of the same challenges.

Part of the purpose of the current paper is to document the story of the Second edition, and to create a good account for the historical record (particularly for the use of the Bliss Classification Association), bearing in mind that much of the history of the revision is contained in minor publications or unpublished material (or in personal recollection) none of which are reliably accessible sources in the long term.

[top of entry]

2. History of the scheme

The Bibliographic Classification (BC1) of Henry Evelyn Bliss (Bliss 1940/53) was the last of the traditional general → library classifications to appear and was hailed by many as the best constructed and most scholarly of them all, “the fine flower of the enumerative period” (Langridge 1973, 93). Despite that label, it contained many facilities and devices for synthetic classification, what Bliss called “composite specification” or “composite classification” (Bliss 1929, 154-5; Bliss 1940/53, vol. 1, 19), that is the representation of complex → subject content of documents by the building of compound classmarks. In that respect BC1 stands as a bridge between the enumerative schemes of the Nineteenth century, and the fully faceted schemes of the later Twentieth century. BC1 was adopted by many academic and special libraries, particularly in the United Kingdom and the British Commonwealth, although it was not implemented in any institutions in the United States, other than Bliss’s own library at the City College, New York, where it had originally been developed and tested, and the Claremont Theological Seminary (Thomas 1998, 81).

In 1955, only two years after the publication of the final volume of the Bibliographic Classification, Bliss died, and continued maintenance of the scheme appeared to be a serious problem. Bliss’s death, however, was the prelude to some significant changes in the management and editorship of BC1, and the British Bliss Classification Association, which had been very active in promoting the classification, now took on the major responsibility for the scheme. Initially, this involved the production of the annual Bulletin of the Bliss Classification Association, which had been started in 1954 under Bliss’s editorship, and which published general news about the Classification, together with additions and amendments to the schedules. Dr. John Campbell assumed the role of editor, but during that time the H. W. Wilson Company remained as the Bulletin’s publisher, and they retained the copyright in the scheme.

The 1966 issue of the Bliss Classification Bulletin was, however, the last to be published by H. W. Wilson, and in the same year BC1 went out of print. The 1966 Bulletin (2) recorded that “the Company feels that their responsibility towards the BC may justly be transferred entirely to the British Committee, both as to the production of the Bulletin, and of a future new edition of the BC”. H. W. Wilson generously transferred the copyright to the British Committee, who acknowledged that Wilson had published the classification “in a spirit of duty to the library profession and in doing so sustained a considerable financial loss” (Mills 1966, 2). In 1967, at the annual general meeting (AGM) of the British Committee a new organization was formed, the Bliss Classification Association (BCA), under the chairmanship of Jack Mills, reader at the then North Western Polytechnic, (later Polytechnic of North London, PNL) School of Librarianship.

[top of entry]

2.1 The revision programme

The possibility of a second edition of the Bibliographic Classification had been raised as early as 1954, by Bliss himself writing in the Bulletin (1954, 2-3): “when, after several years, a sufficient accumulation [i.e., of amendments] is on hand, a complete new edition, we hope, may be convenient and economical”. After the formation of the BCA the decision was quickly made to proceed with this new edition, and an appeal was made to the membership of the BCA in 1967, and again in 1969, to create a fund to support the work. This resource was to be supplemented by the Polytechnic in the form of a paid research assistantship, which enabled the drawing up of a programme of work.

Most of the planned work was carried out by the editor-in-chief Jack Mills and his assistants, but was contributed to by colleagues at the School of Librarianship, and, on a pro bono basis, by members of the Bliss Classification Association, users of BC1, and others interested in classification more generally, notably members of the United Kingdom Classification Research Group. Academic subject specialists were also invited to comment on matters such as optimum linear order, expected collocations of topics, and citation order. Many people gave advice on the general structure of subjects, read draft schedules, and provided feedback, and the Bulletin for 1969 (2) records that “a number of classes, some of them extensive ones, have already had, or are having, a great deal of work done on them by individual libraries”. Libraries that had made their own additions or modifications were encouraged to send copies of these to the Polytechnic, and in 1983 a general invitation for participation was published in the academic press (Mills 1983). Some individuals made a more significant contribution in terms of their input, drawing up whole subject classes; particular names include Ken Bell, lecturer at PNL (Philosophy), → Douglas Foskett, librarian at the Institute of Education and later at the University of London Library, with Joy Foskett (Education), Colin Ball (The Arts), and Chris Preddle, sometime secretary of the BCA and librarian at Dr. Barnardo’s (Social Welfare).

Most of the in-house work in collecting and analysing terminology, consulting external sources, and drafting preliminary schedules, was the responsibility of research staff at PNL specifically appointed to the revision project. Staff also had the task of publicising the project and some short journal articles describing the new scheme appeared (Broughton 1974; Mills 1970; 1976b). The original research assistant appointed in 1969 was Valerie Lang, a graduate of the School of Librarianship at University College London, who had previously worked with Kenneth Vernon on the London Classification of Business Studies (Vernon and Lang 1970), but who left in 1972 to take up a permanent library position. She was replaced by Vanda Broughton, another UCL graduate, initially awarded a three year Research Fellowship, but who continued in post until 1990 supported by various research grants from bodies including the Department of Health and Social Services, the King’s Fund Centre, the British Library, the British National Bibliography, and the John C. Cohen Foundation.

[top of entry]

2.2 Radical nature of the revision

In addition to the numerous small-scale revisions and additions which Bliss had doubtless imagined would form the basis of the second edition, from 1963 onwards the Bulletin had carried more substantial new schedules for subjects as diverse as electronics (BCB 1964, 10-12), nuclear reactor engineering (BCB 1965, 6-9), sound reproduction and recording (BCB 1965, 10-11), astronautics (BCB 1965, 12-14), food preservation (BCB 1966, 4), gardening and fruit growing (BCB 1966, 9-11), and printing (BCB 1967, 13-16). These were to be the last “new” schedules published in the Bulletin, following a decision recorded in the Foreword to the 1969 issue to focus on the second edition [1]. They were rather different in nature to the earlier material and incorporated much of the theory that had developed in the 1950s and 1960s, firstly following the ideas on faceted classification of → S. R. Ranganathan, and then the particular version of faceted classification which emerged from the Classification Research Group. The Introduction to BC2 (Mills and Broughton 1977, 11 Section 2.41) makes explicit the difference from what has gone before:

Conceptually, they [the new schedules] were completely consistent with the theories of bibliographic classification advanced by Bliss, but may be said to have introduced a more rigorous analysis and a more consistent pattern than the original BC sometimes showed.

This emphasis on revising the internal order and structure of schedules in BC1, in addition to expanding and updating the vocabulary, clearly had some implications for the direction which the revision would take, and the consequent relationship between BC1 and BC2. A major decision to be made therefore was the extent to which the revision should be a radical one.

Curwen’s paper on revision policies in classification schemes identifies some of the criteria involved in establishing the degree of revision which is either necessary or acceptable from the users’ point of view, and examines the approach taken by several schemes including BC2. He summarizes the practising librarian’s point of view neatly when he says (Curwen 1978, 20):

there is a conflict between the static nature of the collection and the dynamic nature of the new acquisitions which may not fit into the established pattern. Librarians want their collections to be up to date and the classification schemes which organize and display those collections correspondingly so, but at the same time they do not want to have to alter the work which has already been done.

As a consequence, all of the bibliographic schemes have been criticised for changes made to their schedules. Curwen identifies the key problems for revisers as being vocabulary and structure, the first to be accommodated without unnecessarily disturbing the second. The complexity of subject content in modern literature challenges both, and in the case of BC2, because of the introduction of facet principles to drive the structure, it was unlikely the disturbance to the order would be minimal.

A memorandum was prepared for consideration by the BCA, which was also published in the Bulletin for 1969, and reproduced in the Introduction and Auxiliary Schedules in 1977. It lays out the argument for making the revision a radical one, considering two major aspects of the work: firstly, the need to update and expand the vocabulary; and secondly, the need to improve structure and consistency. Given that BC1 met many of the desirable criteria for a modern classification scheme, the memorandum lists the following seven factors to be borne in mind (order follows that in the Introduction (Mills and Broughton 1977, 12):

1. BC will have gone some 20 years without a comprehensive revision.
2. We cannot expect BC to be revised as frequently as DC or LC. So this one must look correspondingly further ahead.
3. If a library has to reclassify a given class to any degree it is often the case that a complete revision would not take significantly longer.
4. A radical revision of the kind being prepared will produce a far more enduringly satisfying scheme than a piecemeal, patching-up operation.
5. Moreover, a revised scheme with a clear facet structure in each class and clear rules for consistent classifying is far easier to maintain and revise than an enumerated one not enjoying these features.
6. We assume that present users of BC already use a wide variety of the alternatives offered by the scheme as well as amendments and expansions of their own. As a result, whatever form the revision takes it will inevitably involve many users in varying degrees of reclassification. So we might as well make the best possible job of it.
7. A number of positive steps will be taken to minimise the work involved in reclassification. The scope and definition of classes will be indicated with maximum clarity and the crucial problem of compound classes will be met by clear and explicit rules for practical classification.

Generally speaking, these were practical considerations, aimed at ensuring an up-to-date vocabulary and order of classes that would be sustainable in the medium term without further major changes, balanced against the least possible disturbance to existing users. It is noteworthy that some of these factors relate to the extent of the re-classification that would accompany different degrees of revision. Without the benefit of hindsight, it was not anticipated that operating conditions in libraries would change significantly over time, and that in-house classification and re-classification would be in a state of decline in the Twenty-first century. The consequent rise of the Library of Congress Classification (LCC) and the Dewey Decimal Classification (DDC) as the dominant systems in recent times ensured that for most libraries re-classification to LCC or DDC would be an easier and more cost-effective undertaking than re-classification to a radically revised BC, whatever its functional or theoretical merits.

[top of entry]

2.3 The role of the Classification Research Group

A particularly powerful influence on BC2 lay in the opportunity which the revision project provided to create a “new” general classification scheme, one which was less US-centric in its approach, and a potential vehicle for recently developed thinking on classification. Such thinking was primarily driven by the faceted classification methods of S. R. Ranganathan which had been the main focus of the Classification Research Group’s work in the preceding twenty years, and which had evolved into a distinctly British version. The role which the CRG played in the creation of BC2 cannot be underestimated. In 1955 they had published in the Library Association Record what is commonly referred to as the CRG “manifesto”, a long statement entitled “The need for faceted classification as the basis for all methods of information retrieval”, and BC2 was to become a test bed for that mission. The CRG’s early expectation had been that the eventual outcome of their deliberations would be a new general scheme of classification, and a number of publications documented that objective (Austin 1969; Classification Research Group 1964a; 1964b; Foskett 1970). Work towards this new general scheme was in part funded by a grant from NATO (Gnoli et al. 2024, 326), an element of the more general effort to improve scientific indexing and dissemination initiated by the Royal Society Scientific Information Conference of 1948, which itself had led to setting up of the CRG. This expectation was thwarted when the work carried out under the NATO funding failed to materialise as a classification scheme, although it resulted in the creation of the → PRECIS indexing system for the British National Bibliography under the direction of Derek Austin, CRG member, and the original research lead for the NATO grant (Austin 1969; 1984).

With the advent of the Bibliographic Classification revision project, the CRG members were naturally interested in this new opportunity for the implementation of their research, and their role in the development of BC2 is openly acknowledged in the Introduction (Mills and Broughton 1977, vi):

Of inestimable value too have been the theoretical and practical contributions of our friends and colleagues in the Classification Research Group which since 1952 has played a significant part in developing the ideas of Ranganathan into the versatile and supple instrument which has made the new BC possible.

Drafts of schedules and discussion papers relating to the BC2 revision were regularly the topic for discussion at CRG meetings from the 1970s onward. Alongside those individuals already mentioned, contributors to the conversation also included Jean Aitchison, → Eric Coates, Robert Fairthorne, Jason Farradane, Barbara Kyle, Derek Langridge and Leonard Will. CRG members who edited or co-edited a volume of the scheme were Eric Coates (Chemistry), and Douglas Foskett, who was instrumental in preparing the second edition of Class J, as well as producing a further revision of that class in 1990 (Foskett 1989). Foskett (1991) also discussed the role played by BC2 in providing a model for a new general classification, using examples of the scheme’s structural analysis.

[top of entry]

3. Format and scale

The Association’s original intention was to present the new edition as a two-volume work, similar in scale and detail to BC1. This was to contain an introduction to the scheme, and an alphabetical index, created by chain indexing procedure. Correspondence with the publisher Butterworths in October 1971 reveals that this had been amended to a three-volume work, in crown octavo format. The new schedules were to be typed at the School of Librarianship, using an IBM golf-ball typewriter which allowed for the easy use of a variety of fonts and styles, in a two column format for the schedules, and three columns for the index. This “camera-ready copy” would be delivered to the publisher who would photographically reduce it to create the published version.

The scale of the work was not originally intended to exceed that of Dewey in terms of the number of classes, and it was stated that it could not rival UDC, whose Master Reference File now contains about 65,000 classes. Nevertheless, the Bliss Classification Bulletin for 1969 (Mills 1969, 7), outlining the general principles and policies for revision states that “we assume that the new BC should at least equal, and frequently exceed, the detail provided by DC and LC”. To this end the cumulations of the British National Bibliography were intended to supply the necessary literary warrant, both in the required level of detail, and the identification of new topics. Despite these early intentions, the size of the vocabulary in published volumes has tended to rise over time, and it is likely that the number of classes now far exceeds that of UDC. By the 1990s Thomas (1995, 108) noted that “the degree of specificity of simple terms exceeds that of the Dewey Decimal Classification and is usually greater than that within the Library of Congress Classification”, but that “the high level of detail in BC2 has varied somewhat from published class to class”. As a rough guide, the average number of classes per page on the most recently published volumes for Fine Arts and Chemistry comes in at 54 and 64 respectively, giving an approximate overall total of 7,500 and 6,400 classes for these two disciplines. Although these are both large fields, and the scale is not likely to be replicated across all 27 proposed volumes, a figure of between 100,000 and 150,000 classes (very much in excess of UDC) does not seem untenable.

It had been rather optimistically stated in the 1969 Bulletin that the work would be completed in three years, and a contract with Butterworths publishers was signed in 1972 for delivery by the end of October 1973. It soon became apparent that this could not be achieved, and the 1973 Bulletin (2) records that the Committee intended to deliver the finished schedules by September 30, 1974, which Butterworths had agreed to, “being realistic about the time needed for a large work of this sort”.

As a consequence of further delays, in 1974 it was proposed that the classification be published in parts in order that the schedules for at least some subjects could be made available more quickly. The 1974 Bulletin (2) records that “Mr Mills thought that ideally one should not issue parts before the whole work, as revision of one class often affected others. But he thought publication in parts a workable idea” [2]. At this stage it was envisaged that there would be a separate introductory volume, plus a volume covering the whole scheme in an abridged version. The membership of the BCA was overall in favour of the plan, and the contract with Butterworths was renegotiated. By the time of the Association’s AGM in November 1974, two classes, J Education and P Religion, were considered ready to go to the publisher, but were not then delivered until August 1976, together with the Introduction and Auxiliary Schedules, followed by class Q Social Welfare in December of that year. All four were published in 1977, in the format which had been finally decided on, and which was to remain in place for subsequent print volumes. In these slimmer volumes the crown octavo format was abandoned in favour of A4, the camera-ready copy initially supplied as A3 sheets, to be photographically reduced by the publisher, and for later volumes as PDFs. In practice, the published text was printed onto a larger page size (22x28 cm for the original Butterworths volumes with slight variations of 21.5x27.6 cm and 21.5x28.5 cm from the European publishing houses). In 2002, BC2’s publishers Bowker-Saur discontinued their print operation, and responsibility for the classification was passed to K. G. Saur in Munich (Bliss Classification Association 2002, 3), and subsequently to de Gruyter.

From 1981 onwards production of the camera-ready copy was mechanised, beginning in 1981 with the computerisation of the A/Z index to class H, the schedules for which were still produced manually on the IBM machine. Subsequently, a set of programmes was created to generate the camera ready copy of the schedules from a coded version, but from the early 2000s these were proving unusable to the point that “it had become impossible to produce further copy to send to the publishers” (Curwen 2002, 2), so a new software package was commissioned from Paul Coates which remains in use. Completed schedules maintained as text files are encoded using a form of mark-up to indicate, for example, the position in the linear sequence, and hierarchical status of individual classes, as well as other structural characteristics. This source code is used to automatically generate the camera-ready copy of schedules (which had previously needed to be typed), and the alphabetical index, the latter still requiring some degree of manual editing.

For publication purposes, the software is able to generate a visually structured display, calculating the hierarchy from the indent codes, such as 07), 08), etc., and identifying notes, node labels and so on, from the “non-class” symbol @. Other symbols are used to determine the typographic format, and to control entry (or otherwise) in the alphabetical index or thesaurus. In practice, this relatively simple markup is able to produce a high-quality print version of the scheme.

In 1987 the BCA was in discussion with a technology company, Computercraft, about the use of the BC software in conjunction with a thesaurus management system. This was in the context of the development of a thesaurus for the Royal Institute of International Affairs, based on the BC2 class R Politics and Administration. A letter dated June 1987 from Jean Aitchison, the consultant on the project, to James Cowan of Computercraft, describes previous work such as the use of the BNF STRIDE software to manage the DHSS-DATA Thesaurus, which was derived manually from the Bliss classification structure for health sciences. Aitchison says that the system “can handle the classmarks, but it cannot interact with the classification”, and describes various features that need to be incorporated to make the software more usable and more attractive to the market.

In 2004 the BCA’s own software was enhanced to include a thesaurus generation function (Broughton 2008), with very reasonable results, although it has not yet been used to produce a publicly accessible thesaurus. Most thesaurus features can be derived automatically from the original markup, such as the broader and narrower term relationships, and preferred and non-preferred terms, while associative terms still need to be determined manually. (For more detail see also below Section 10.)

In 2018/19 a project was put in place to create an XML version of the classification in the expectation that this would create a more accessible and editable form of the schedules. Consultant Richard Light has done some encouraging preliminary work, demonstrating that the BC2 structure and syntax can be expressed fairly readily in XML. Work is ongoing in this area examining various methods of converting both text and the coded format to an editable version; a prototype editor is being used for testing, but this is not yet publicly available.

[top of entry]

4. Published volumes

Publication of the scheme began in 1977, and to date 15 volumes have been formally published, originally by Butterworths, then Bowker-Saur, Saur, and now de Gruyter.

	Introduction and Auxiliary Schedules	1977
A/AL	Philosophy and Logic	1991
AM/AX	Mathematics, Probability, Statistics	1993
AY/B	General Science and Physics	2000
C	Chemistry	2012
H	Physical Anthropology, Human Biology, Health Sciences	1980
I	Psychology and Psychiatry	1978
J	Education (3rd ed.)	1990
K	Society (includes Social Sciences, Sociology and Social Anthropology)	1984
P	Religion, Occult, Morals and Ethics	1977
Q	Social Welfare and Criminology (3rd ed.)	1994
R	Politics and Public Administration	1996
S	Law	1996
T	Economics and Management of Economic Enterprises	1987
W	The Arts	2007

Versions of most of these classes are available on the Bliss Classification Association website (www.blissclassification.org.uk); only Classes H, K, and T are not accessible in electronic format. Draft schedules are also available there for another twelve classes, mainly as PDFs, occasionally in Word format, or as scanned typescript. A number of published and unpublished classes also have links to the source code used to produce the PDFs and final copy.

2/9	Generalia, Phenomena, Knowledge, Information Science and Technology	PDF
D/DF	Astronomy	Word
DG/DY	Earth Sciences	PDF
E/GQ	Biological sciences	PDF
G/GY	Zoology	scanned typescript
GR/GZ	Applied Biological Sciences, Agriculture and Ecology	scanned outline schedule
L/O	History (including Area Studies, Travel and Topography, and Biography)	PDF
LA	Archaeology	PDF
U/V	Technology and Useful Arts (including Household Management and Services)	scanned typescript
WV/X	Music	PDF
X/Y	Language and Literature	PDF
ZA/ZW	Museology	PDF

[top of entry]

5. Arrangements for maintenance

A major disadvantage of BC2 is that, as with its predecessor, there are no formal provisions for maintenance, and ongoing work is dependent on volunteer effort. A powerful argument for the radical nature of the revision was to mitigate the need for short term revision and updating, but fifty years from its inception there are obvious shortcomings in the scheme, not least that some parts remain unfinished. Work has been undertaken to make the scheme as freely accessible as possible in the hope that users, either of the scheme or of the terminologies, would contribute suggestions for updates and amendments, but to date there is little sign of this.

[top of entry]

6. Principles underlying the classification

The classification is based on various theoretical and structural principles, some of which are inherited from BC1 and some of which are part of the more general theory of faceted classification. There is limited discussion of any philosophical or conceptual framework of facet analysis in the LIS literature, but Hjørland has proposed that the fundamental philosophical approach of facet analysis (and hence of BC2) is rationalist, that is, based entirely on logic and reasoning (Hjørland and Albrechtsen 1999, 134; Hjørland 2013; Hjørland and Barros 2024). Descriptions of the standard methodology of facet analysis would tend to support this idea, and certainly, the principles of classical logic are applied consistently in BC2 (Mills 2004). Conversely, some elements of the process (user consultation and mapping to the published literature of a field) can make a strong case for the role played by pragmatism (Dousa and Ibekwe-SanJuan 2014; Gnoli 2017). Interestingly, one of de Grolier’s observations in his unpublished review of BC2 (see Section 11) was that it was “in the English pragmatic tradition” (Mills 1991, 17), although we cannot be completely certain what he might have meant by that. A partial insight is given in his discussion of the CRG work in his 1960 study of general categories in which he states that the CRG classification schemes “are truly specialized, very “pragmatic” classification systems” (Grolier 1960, 99); by this he means the priority given to → user interest in deciding on the arrangement of such schemes, and the difficulty that approach creates for the construction of a general scheme. This aspect of pragmatism is demonstrated in the BC2 context by the use of “utility” or “purpose” in determining citation order, as described by Vickery and Mills (see below Section 7.2.1).

Two main characteristics of faceted classification schemes find their historical roots in philosophy (Mazzocchi and Gnoli 2010): the use of categorical analysis, which can be traced back to Aristotle (Moss 1964) or alternatively to Indian philosophy (Adhikary and Nandi 2003); and the use of level theory as an ordering tool for disciplines, subject fields, or phenomena, drawing on → Comte, Spencer, Needham and Feibleman (Gnoli 2008). Discussion of these philosophical roots does not feature in the BC2 literature, other than a fleeting reference in the Introduction 6.213.31 to Comte, who was a significant influence on Bliss particularly in respect of the “hierarchy of the sciences” (Shera 1965, 83) which principle drove Bliss’s original main class order, substantially retained in BC2. Otherwise, the effects of both categorical analysis and level theory in the BC2 context are explored in more detail below (Sections 6.2 and 7.1.1).

It must be acknowledged that the influence of the CRG on the creation of BC2 is a powerful one, and that many of the features of BC2 occur as a result of the CRG work on a proposed new general classification (Austin 1969; Foskett 1970). A complete discussion of the structure and theory of BC2 is provided in the Introduction to the scheme (Mills and Broughton 1977), to which readers are referred. A briefer description of the scheme is provided by Thomas (1992). A summary of the main features is given here.

[top of entry]

6.1 Relationship with BC1

The reasons for revision, enumerated at Section 2.2 above, indicate some of the intended differences in the second edition. Provision needed to be made for developments in individual subjects, and for the emergence of new subjects, together with the introduction of the larger and more up-to-date vocabulary associated with these changes.

BC1 is widely acknowledged for the sound theoretical basis provided by Bliss, and which emerged from his detailed analysis of the organization of knowledge as evidenced in his writings. The general broad structure of the original classification is maintained, with some modifications as discussed below.

Differences within main classes are largely occasioned by the more rigorous application of facet analysis to their vocabulary, and the imposition of a strict citation order, which had in the original scheme been more tenuous in its nature. It is, however, often the case that the BC1 order reflects the order obtained through the process of facet analysis, and there is less disruption to the overall order of a class than might have been expected. In The Organization of Knowledge (1929, 154-50) Bliss seems to have anticipated something of the nature of faceted classification in his discussion of what he calls “cross-classification” or “duplex arrangements of vertical series, coordinate with regard to one principle or interest, crossed by horizontal series representing other aspects. […] This form of classification, graphically set forth, is often termed tabulation”. A more potentially complex arrangement also appears to be in his mind when he states:

A system of entities and relations may […] be surveyed from different aspects and traversed by diverse interests and purposes. There may accordingly be many classifications crossing or branching in many ways. A one-dimensional serial classification has no structural reach. […] Cross-classifications and two-dimensional graphs are also inadequate; for they become congested with ramifying details that would really require three dimensions for their structural representation, or even four dimensions.

Bliss made various provisions for compounding, but the capacity for this in BC1 is much more limited than in BC2. BC1’s methods of compounding are through the use of auxiliary tables, both general, for commonly occurring concepts, and those special to a particular class, embedded in that class. Many of the special auxiliary tables are retained, albeit revised, in BC2; although they may appear redundant in terms of the general provision for compounding, in many cases they effect some economy in the length of classmarks. Bliss also included a general indicator for compounding between classes. An equivalent is available in BC2 for combination between main classes, in the form of the letter subdivisions of 9 in the Common Auxiliary Table for Form.

Many of the alternatives included in BC1 are retained in BC2, with the additional provision of alternative citation orders where the requirements of different user groups may vary. In the interests of standardization there is always a “preferred” treatment, usually reflecting standard citation order.

Bliss based his classification on three important principles: it should have an underlying theoretical classification of knowledge; it should reflect the way that users studied and applied knowledge (consensus); it should reflect the way that knowledge was published (→ literary warrant). Bliss is also generally credited with the introduction of three other important aspects of classification: gradation in speciality, hospitality, and flexibility. All of these are acknowledged in the revision, and some of Bliss’s principles are directly related to theoretical features of BC2. In his paper on the classification of a subject field, given to the Dorking Conference in 1957, Mills emphasised that the determination of the most significant aspects of a subject — the way in which it is studied, and the way in which it is practised and applied — correspond to Bliss’s ideas of educational and scientific consensus (Mills 1957, 34).

[top of entry]

6.2 Main class order

A significant feature of BC1 was its excellent main class order, at which Bliss had laboured for so many years, and which is widely acknowledged to be the best of all the general classification schemes. Bliss attempted to achieve a perfectly modulated sequence of classes and gave much thought to the relationships between concepts in the various disciplines. His main classes also reflect that careful consideration of the modulation from one subject to another. This order of main classes is for the most part preserved in BC2, although the notational allocation varies somewhat to reflect modern perceptions of the relative documentary size of disciplines.

Bliss’s analysis of the universe of knowledge, and his consequent main class order, is largely derived from a part philosophical and part pragmatic consideration of individual disciplines and their structure. As Langridge says (1976, 5) Bliss’s premise is that “the order of the sciences is the order of things, and the order of things is the order of their complexity”. Classification theorists in the 1960s sought an intellectually more rigorous basis for this important aspect of classification design; in particular, the Classification Research Group, in their work on the viability of a new general classification scheme, examined various scientific theories for determining the relative order of subject domains.

The theory of → integrative levels, in the form proposed by Needham (1937), and Feibleman (1954), was seen as providing a legitimate theoretical foundation for a linear sequence (Foskett 1961; 1978; 1991), and the order in BC1 was observed to approximate very closely to this. The most general levels of matter, life, mind and society correspond neatly to Bliss’s groupings of physical, biological, human and social sciences, and Bliss’s principle of gradation by speciality is consistent with the general tenets of integrative level theory. A developmental order is also evident in BC, moving generally from the abstract to the concrete (implicitly reflecting → Ranganathan’s organizing principle), and which exhibits increasing complexity in the different disciplines, and the principle of dependence of each on preceding disciplines. This is articulated in The Organization of Knowledge and the System of the Sciences (Bliss 1929, 300):

Intelligence accepts the truth that in all probability there is some real transition from sphere to sphere, from the physical to the vital and from the vital to the mental, as there plainly is from the natural to the human and from the individual to the social, and as we may rationally believe there is from the human to the divine and universal.

Spiteri (1995) provides a useful introduction in her analysis of the Classification Research Group’s discussions of integrative level theory as an organizing principle in classification.

[top of entry]

6.3 Modifications of main class order: classes, disciplines, and phenomena

The Introduction to BC2 makes careful distinctions between the three concepts: class, → discipline, and phenomenon. It concedes (1977, 5.54) that: “the term “main class” is an ill-defined one and there is some confusion between it and the idea of discipline”. At 5.542, “a further confusion results from use of the term “disciplines” in referring to what are essentially sub-disciplines”. In BC2, what Bliss would have considered as a discipline (physics, chemistry, biology, and so on) are regarded as sub-disciplines, that is the subdivision of a fundamental discipline, such as science, by particular phenomena (energy, matter, living creatures).

The main class order of BC2 was influenced to an extent by this notion of fundamental disciplines as proposed by Langridge (1976). It claims that there exists a small number of different disciplinary methodological approaches as distinct from the more numerous fields of study in the traditional understanding of disciplines as concerned with specific phenomena. These approaches include philosophy, history, → science, and religion, each with its own particular underlying conceptual framework, theory, and methodologies. These should form the main structure of a classification, with each discipline then subdivided by phenomena. This viewpoint is formally acknowledged in the Introduction to BC2 (1977, 5.53): “[a] stricter and probably more accurate way of regarding disciplines is to see them as reflecting different ’forms of knowledge’, or ways of looking at the phenomena of the world. The concepts and methods of enquiry of the scientist, the philosopher, the historian, the artist and so on are very different, although the phenomena they consider may be to some extent the same”.

Langridge is sceptical about the scientific approach to knowledge organization on the basis that it is only one view of the universe of knowledge, and not necessarily a legitimate one, with its assumptions that there is an objective external world, and that the scientific paradigm is the means to express it; this view is, of course, at odds with Bliss’s conception of the natural order of things. Otherwise, a division along these disciplinary lines is not greatly at variance with the BC order, except insofar as it has been proposed to move class P, Religion, to the end of the sequence (class Z) on the basis that books in the realm of religion have a different conceptual perspective to those in the realm of, for example, science.

The current editor’s view is that this approach to religion is not useful from a bibliographic perspective. A distinction needs to be made between books about religion (which constitute the majority of items in a religion collection and are generally scientific in nature) and books of religion (which may well have a different world view). A case might be made for the separation of religious books, but it would cut across the traditional disciplines (e.g., religious treatments of evolution, or monetary lending, or abortion) and create a new set of ordering problems. The conventional classification solution using a common subdivision (“history of”, “philosophy of”) accommodates such material satisfactorily without making the fundamental discipline the principal means of arrangement.

Some allowance ought also to be made for Bliss’s distinctive treatment of religion in BC1, placing it among the social sciences where it is an example of a sphere of activity in human society, alongside law, economics, politics and so on. This view conforms more readily to the modern academic discipline of “religious studies”, dating from about the 1960s, in contrast to traditional “theology” where a lens of religious belief, commonly of one persuasion only, is applied to scholarship.

[top of entry]

6.4 Phenomena

A distinctive feature of BC2 as compared to BC1 is provision for the treatment of entities, or → phenomena, as separate from their usual disciplinary, or sub-disciplinary, context. Acknowledging that phenomena are essentially scattered in the general classification scheme with its initial division into disciplines, BC2 (1977, 5.552) asserts that “it should be recognised that there is, theoretically, a quite different way of organising a general classification. This would be to make the first division of the field of knowledge into phenomena (from subatomic particles to planetary bodies and stars, from single cells to particular organisms and particular societies, and so on) and to subordinate to each phenomenon the disciplinary aspects from which it may be treated”. Such an alternative approach to classification is accommodated using the anterior numeral classes 4/9.

The Introduction goes on to state (1977, 5.553) that “[s]uch an arrangement would run counter to the way we usually study things and the way most information is marketed, which reflects the division of labour by discipline. There are relatively few persons, if any, specialising in a given phenomenon from all its aspects. Indeed, such a specialised study would require a training which is at present hard to envisage”. Forty years later this is increasingly not the case with, for example, many higher education programmes focusing on a phenomenon rather than disciplinary conventions; women’s studies, in its infancy in 1977, is the classic example, but in the UK degrees as diverse as mediaeval studies, early childhood studies, Celtic studies, addiction studies, Islamic studies and refugee studies, proliferate. Beghtol (1998, 6) in an evaluation of different schemes’ treatment of multidisciplinarity, confirms this, saying that the “growth rate of multidisciplinarity is now remarkably fast, and BC2 appears to offer more complete treatment of these works than the other systems discussed”. Interestingly, Gnoli (2005, 20), writing a little later, in a critique of the phenomena classes in BC2, observes that “focusing on phenomena […] is not very fashionable in present-day classification research: one can find it in past decades in work by CRG members and by Dahlberg, while today people appear to be more concerned with relativistic issues such as locality, cultural biases, social use of information, etc.”

In 1977 it was possible to state (1977, 6.214.1) that “the phenomena in the BC constitute a new feature, which has not been attempted before in a general classification, although studies undertaken for the Classification Research Group in London looked closely at the problems involved” (Austin 1969). In fact, in 1906 → James Duff Brown had attempted something comparable in his Subject Classification where the categorical tables are used to represent aspects of a subject, and the list of subjects has some bizarre collocations such as the inclusion in Physics of church bells (under Sound) and fire engines and chimneys (under Heat).

To date there is very little in the way of developed BC2 schedules for phenomena, and only three classes are allocated in the outline provided in the Introduction (1977, 202).

		(Phenomena) * For multidisciplinary treatment of particular phenomena * Alternative (preferred) is to collocate with most appropriate disciplinary treatment, applying the principle of unique definition (mainly for phenomena appearing in the natural sciences and purpose (mainly for phenomena in social science and humanities)
4		Attributes (e.g. structure, order, symmetry, colour)
5		Activities and processes (e.g. organizing, planning, change, adaptation)
6		Entities (e.g. particles, atoms, molecules, minerals, organisms, communities, institutions, artefacts)

This outline of the phenomena classes goes on to list the “universe of knowledge” and communication and information. It should be observed that the notational allocation here is compromised by the later development of a draft schedule (www.blissclassification.org.uk/Class2-9/2-9_sched.pdf) where the phenomena proper are limited to class 3.

It is clear that it was never the intention to adopt a totally phenomenon based approach in BC2, which aims only to provide a location for documents where concepts are treated from a multi- or non-disciplinary perspective; the three possible means of doing this (1977, 6.221) all envisage at least some documents being allocated to the discipline based classes. While the question of whether provision should be made for “a classification of all or some phenomena in which the phenomenon was cited before the discipline” is touched on (1977, 6.214.56), it is dealt with rather summarily, as it “seems clear that BC would be inconsistent in view of its general policy not to provide such an alternative so it will in fact provide one”.

The Introduction includes a lengthy discussion (1977, 6.214.5) of the difficulties associated with classifying phenomena in a scheme which would still be predominantly discipline based. These include the process of extracting the phenomena from lists in main classes, and the associated problem of deciding on a preferred order of concepts where those concepts occur in more than one disciplinary class and are differently ordered there (animals in zoology and animal husbandry are an obvious example). (Although it is not raised here, there is an associated difficulty with the use of multiple notations for the same concept in different disciplines: Gnoli 2005, 17) Interestingly, Riggs (1987, 42) in his review of class K, Society, comments on the choice of main class name, which suggests that the focus is on the phenomenon of society, rather than the discipline of sociology. This is mirrored in other revised classes, for example, Religion rather than Theology, but it not carried through consistently.

There is also the matter of an alternative provision for multidisciplinary works where a library does not wish to use the phenomena classes. Here, multidisciplinarity is treated as a kind of common subdivision, and filed immediately after the main (disciplinary) class heading (1977, 6.214.55):

It is proposed, therefore, to file the multi-disciplinary treatments immediately after the disciplinary ones, using the numeral -1 to notate it exactly (the common subdivisions in Schedule 1 begin with the number -2):

GYH J   Horse (Zoological studies, general)
GYH J1  Horse (multi-disciplinary studies)
GYH J2H Horse (Zoological studies) – Pictorial material

This solution is achieved through the notion of the “place of unique definition” originally proposed by Farradane (1950, 87), which rules that “for any given concept, the minimum information needed to define it is to be taken as the clue to its location” (1977, 6.214.61) or alternatively, “that class which provides a given concept with its most fundamental defining characteristics, e.g. Zoology provides the place of unique definition for the concept Horse” (1977, 107). It should be acknowledged that Farradane’s system of relational indexing did adopt a non-disciplinary approach, with the objective that concepts, or isolates, as Farradane called them, were context free and must have only a single location. The problems of ordering concepts in an entity-based classification were explored by the CRG, and in particular by Austin (Austin 1969, 156-7; 1976). Reference here should also be made to the → Integrative Levels Classification (ILC), devised by Gnoli and others, which follows a similar line, and makes phenomena the primary approach to the whole system (Gnoli 2011; Szostak 2007). A significant factor here is the emergence, and increasing importance of interdisciplinarity in academic research, which feature was not a consideration in the initial stages of the BC2 revision.

[top of entry]

7. The structure of BC2

The Introduction and Auxiliary Schedules (1977, 38) asserts that:

A bibliographic classification consists of:

5.621	A vocabulary of terms organised into broad facets;
5.622	Within each facet, terms organised into sub-facets, or arrays;
5.623	A citation order for the formation of all compound classes (i.e., classes of more than one term reflecting the intersection of concepts from different facets or from different arrays within the same facet);
5.624	A filing order for the resulting classes (simple and compound);
5.625	A notation to maintain the filing order;
5.626	An alphabetical index of terms to locate them in the classification.

This broadly corresponds to the chronological sequence followed in constructing the classification, beginning with the organization of the vocabulary into facets.

[top of entry]

7.1 Facet analysis in BC2

Vickery (1960, 12) states that the “essence of facet analysis is the sorting of terms in a given field of knowledge into homogeneous, mutually exclusive facets, each derived from single characteristic of division”.

Its faceted structure is certainly the most distinctive feature of BC2, and the one that usually distinguishes it from other western schemes of classification in general use. Alongside the Colon Classification, it is a general scheme constructed ab initio by facet analysis. Most schemes now exhibit some elements of faceting, and the UDC is in a process of gradual conversion to a fully faceted form, but still has many examples of enumeration in its schedules. An exception should be made for the Integrative Levels Classification, but this is intended primarily as a source of metadata for post-coordinate indexing of research material rather than a general organizational tool applicable to physical document collections.

Ranganathan is generally credited with devising the method of facet analysis, although examples of similar analytical approaches to knowledge organization can readily be found in earlier systems, in the UDC, Brown’s Subject Classification (1906), and most noticeably in the work of Kaiser (1911), who identifies categories such as place, process, and concrete in document content (Dousa 2013). There is also evidence that facet analysis emerged independently in other disciplines, particularly in the social sciences (Beghtol 1995). The approach adopted in BC2 is rooted firmly in the Ranganathan tradition but is tempered by the modifications of that tradition made by the UK Classification Research Group throughout the latter part of the Twentieth century, and which employs thirteen fundamental categories as opposed to Ranganathan’s five. It may be useful here to make the distinction between category and facet.

[top of entry]

7.1.1 Fundamental categories

Ranganathan understood a category to be “a form or class of concepts, varying from subject to subject, into which isolates can be grouped”, and a facet to be “the subclasses of a basic class corresponding to a single fundamental category” (Campbell 1957, 13). The glossary in BC2 does not define category (and tends to employ facet for preference) but it may be broadly understood as an abstract and generic concept, one which may be employed in any domain for the purpose of analysing and grouping the vocabulary. Terms or concepts are assigned to categories on the basis of their functional (or sometimes linguistic) nature, and in this they have something in common with role indicators in indexing. Ranganathan’s categories of personality, matter, energy, space and time were considered by the CRG to be limited, and they discerned a larger number. Vickery (1960) suggests a list of thirteen: substance (product); organ; constituent; structure; shape; property; object of action (patient, raw material); action; operation; process; agent; space; and time. BC2 uses a variant, and slightly reduced, version of this set, in naming nine major categories:

end product (often called thing or entity); its types; its parts; its materials; its properties; its processes; operations on it; agents of action; place; time.

In practice this list is itself subject to amendment as the needs of different domains determine, and categories such as patient and by-product are also to be found, as are some humanities and arts related categories, such as genre and form. In 1993 a new fundamental category, that of relation, was used in the analysis of the Mathematics class. It was used to accommodate concepts such as functions, equations, homomorphisms, and so on, and lies conceptually somewhere between processes and properties. There is a body of research around the nature of fundamental categories, and their range and number. Much of the groundwork here is in the faceted classifications developed in the mid-Twentieth century, and which are discussed in some detail in de Grolier’s book A Study of General Categories Applicable to Classification and Coding in Documentation (Grolier 1962).

[top of entry]

7.1.2 Facets proper

The notion of a → facet is very variously understood (Hudon 2020), often being limited to an aspect, or a property of some entity or activity, and that is the way it is commonly considered in web information work (La Barre 2006). In other situations, facets may be equated with, for example, the fields in a bibliographic record, such as author or date of publication (Broughton 2023, 427). In the BC2 context a facet is quite explicitly one conceptual dimension of a subject field.

When a given category is used to accommodate concepts in a particular subject field or domain, the result is a facet. In BC2 (1977, 38) “[a] facet may be defined as the total set of subclasses produced when a class is divided by a single broad principle”. This dividing principle is the fundamental category, so that, for example, the application of part to the vocabulary of medicine will generate the facet of “parts, organs and systems”, or the application of entity to chemistry results in the facet of “substances”.

This can be seen within the context of a particular main class, for example:

Medicine
   Entity, end-product (persons, human beings)
   Parts (anatomy, parts, organs and systems of the body)
   Processes (physiology, diseases and pathology)
   Operations (treatment, preventive medicine, curative medicine, clinical medicine, therapy)
   Agents (medical materials, drugs, technology, medical personnel, organizations)
   Place
   Time

[top of entry]

7.1.3 Order of classes within facets

There is no particular theory that determines the sequence of classes within a facet, and various pragmatic and common-sense solutions are normally adopted, such as order of size, order of development, spatial contiguity, and so on. Ranganathan’s “canon of helpful sequence” (Ranganathan 1967, 183) provides eight principles for such order: later-in-time, later-in-evolution, spatial contiguity, quantitative measure, increasing complexity, canonical sequence, literary warrant, and alphabetical sequence. Vickery (1959, 30) suggests nine possible principles, based on Richardson (1930): logical, geometrical, chronological, genetic, historical, evolutionary, dynamic, alphabetical, and mathematical. For example, in the field of reproductive medicine the developmental sequence “zygote – embryo – foetus” is observed for in utero structures. In some cases there is no particular natural sequence, and the order is arbitrary; for instance, in parts, organs and systems of the body, a head-to-foot order of arrangement is imposed, both for the organs themselves, and within pervasive systems such as bones and nerves.

[top of entry]

7.1.4 Relationships within facets

Since all the concepts within a facet are of the same categorical status i.e., all properties, or all processes, the relationships between concepts are necessarily restricted to those of hierarchy, superordination and subordination. Hierarchical relations are shown in the linear order by indentation within each facet (1977, 6.343), and because the BC2 notation (see below) is not expressive of hierarchy this symbolism is strictly observed throughout. For example, in (the draft schedule for) Music:

WWS W	WIND INSTRUMENTS
Z	Woodwinds
WWT	Flute
WWT W	Piccolo
WWT X	Recorder
XS	Descant
XT	Treble
XV	Tenor
XW	Bass
WWU B	Oboe
C	Oboe d'amore
D	Cor anglais
K	Clarinet (B flat)
L	Clarinet in A
M	Bass clarinet
N	Bassett horn
O	Bassoon

Every indentation implies that the indented class is, in some way, a subclass of the one immediately above it. However, the relation of a subclass to its containing class does not always reflect the conventional relationship of a species to its genus (1977, 7.391.3); for example, the whole/part or system/sub-system relationship is also common.

[top of entry]

7.1.5 Arrays (subfacets)

The number of concepts or classes in a facet may be very considerable; the number of substances in chemistry, for example, runs into many millions. It is therefore necessary to provide some further organizing principle within the facet. The result of such subdivision is known as an array (sometimes referred to as a subfacet), and the means by which it is created is called the principle of → division, or characteristic of division (in the vocabulary world this is usually referred to as a node label). A facet of persons, for example, may be divided by principles of division such as “persons by age”, “persons by social class”, “persons by religion”, and so on, to give arrays of the kind:

Persons	Persons
(by age)	(by religion)
Infants	Atheists
Children	Hindus
Adolescents	Jews
Adults	Muslims
Young adults	Pagans
Middle aged persons	Sikhs
Older persons

An important idea in the construction of arrays is that of mutual exclusivity, which is frequently misunderstood. It states that the concepts within an array must be mutually exclusive and means that (in the examples above) a person cannot be both a child and a middle-aged person, a Jew and a Sikh. If this is not the case, then the array is not properly constructed and should be re-examined. It is also important that the principle of division is exhausted, that is to say that all its possible values are represented. Mutual exclusivity does not mean that the arrays are mutually exclusive, or that concepts from different arrays cannot be combined to give, for example, Hindu children, or older Muslims. Similarly, the complex subjects found in documents may involve more than one concept from the same array (for example, “music education for children and adolescents” or “interfaith activities of Christians and Muslims”).

The same observations about the order of concepts discussed under facets also apply to arrays.

[top of entry]

7.1.6 Relationship between facets: combination of concepts

The second most important feature of the faceted classification, after its logical structure, is the facility to combine concepts to express any complex subject content of documents. In BC2, in principle, any number of concepts can be combined, but this must be done in conformity with various rules. The central rule of combination is that of citation order. Citation order also determines the sequence of facets in the linear order within the classification, using the principle of inversion (see Section 7.3).

[top of entry]

7.2 Citation order

Citation order, in earlier times often referred to as “combination order” (Vickery 1959, 88; 1968, 57), is “the order in which the elements of a compound class are cited when we state the class as a string of terms” (1977, 5.731). Hence, a compound subject described as a string of concepts might be:

Primary school children – reading – phonics

Type of person taught (educand) – subject taught – teaching method

Citation order can be seen in any classification where combination in supported, and in compound subject headings, such as LCSH. The application of rules comparable to citation order can also be observed in the display of any physical collection of items, such as objects in a museum, or foods in a supermarket.

The first cited element will naturally determine where in the physical collection a document is located, and the use of an appropriate citation order is particularly important in the physical document collection since it determines the overall sequence of classes in a subject field and which aspects of a subject are distributed or scattered. In the example above, for instance, material on a particular type of educand is kept together, but material on the subjects taught, or the techniques used, will be scattered. The use of a consistent citation order is also vital since it allows the user to predict where material about a compound subject will be located and enables the retrieval of aspects which are distributed.

[top of entry]

7.2.1 Citation order between facets

The order in which terms from different facets should be combined is mainly determined by established faceted classification theory. Ranganathan was the first to formulate this, stating that order of combination of categories should be according to his famous PMEST formula: Personality – matter – energy – space – time. This follows the rule of decreasing concreteness i.e., that the most specific categories are cited first, working down to the most abstract. This generates a sequence which is intuitive for most library users, with the most concrete, or specific, concepts kept together (for example, animals in zoology) and the more abstract elements scattered.

Vickery (1959, 32) states that the idea of what he calls “utility” “is far more helpful than that of concreteness” in determining the primary facet. This concept he credits to Mills (1957, 34) who describes it as reflecting “the purpose of the classification” and that “material should be collected on that aspect of the subject which is most important to the user”. It is otherwise identified as the end-product, or the purpose of the subject (Mills 2004, 557). This principle is generally employed in BC2 alongside the combination order established by the CRG. The CRG combination order is an elaboration of Ranganathan’s categories, namely:

Thing – kind – part – material – property – process – operation – agent – space – time

This observes the Ranganathan rule of decreasing concreteness, but also employs the idea of dependence (closely related to that of the end purpose); e.g., an operation must be performed on some entity, and the agent of an operation must have an action to be agent of. So, in a subject stated as the chain of terms:

Hospitals – Wards – Floors – Finishes

The CRG order is commonly referred to as standard citation order, which term is generally accredited to Vickery (1960). This order is used as the default order throughout BC2, although there are some local variations according to the needs of the subject.

[top of entry]

7.2.2 Citation order between arrays

As well as combining concepts from different facets, it is also necessary to combine concepts from different arrays in the same facet (as in the examples given above). There is no general principle for this analogous to those for facet citation order (1977, 5.735.1), since the idea of dependence is not relevant. However, there are some individual principles which can be applied. The general principle of purpose (used to determine the primary facet) can often help in deciding on priority, as in the example of “High rise flats” where the function of the building (flats) is cited before simple, or more general properties (high rise), i.e. “Flats – high rise”.

Because there is no comprehensive theoretical principle yet developed, it would seem accurate to say that on the whole deciding citation order in array is largely a pragmatic business (1977, 5.735.5).

[top of entry]

7.3 Filing order

Filing order is the linear order, or sequence, of classes in the classification scheme. Filing order in BC2 is controlled by three main principles: general-before-special, gradation in speciality, and increasing concreteness. Filing order is also reflective of general philosophical notions, such as those employed to determine the order of phenomena and of main classes (Section 6.2 above), and it also demonstrates relationships between classes such as super- and subordination.

General-before-special requires that a class which completely contains another class files before it. Thus, the class of “animals” precedes that of “mammals” and “mammals” precedes “horses”, “dancing” precedes “ballroom dancing”, and “crime” precedes “murder”. Such an order appears to be intuitive for users.

Gradation in speciality is a more complex notion and related to the concept of integrative levels (see Section 6.2). It requires that a subject which is dependent to a degree on another subject should file after it. For example, “biology” makes use of chemical laws and concepts to explain some of its data, although the reverse is not the case. Therefore “chemistry” precedes “biology” in the sequence.

Increasing concreteness is a rule which is implicit in the schedules. It states that classes in the linear sequence should proceed from abstract to concrete; so abstract topics such as “philosophy” or “theory” come before concrete concepts such as “diesel engines” or “bungalows”. The principle of increasing concreteness is particularly important in determining the filing order of facets.

[top of entry]

7.3.1 Filing order of facets

Applying the principle of increasing concreteness to the facet filing order, facets are filed in the reverse of citation order. That is, the last cited facet (usually time) files first in the schedule, followed by the penultimate facet (usually space), and so on, with the primary facet (personality, entity, etc.) filing last; the sequence thus moves from the most abstract to the most concrete. The resulting schedule is said to be inverted.

For example, in the class Literature the first cited facet would be Language, the second cited would be Period, the third literary Form, and so on. If we file these in the order of citation the resulting schedule looks like this:

Literature
    (language)
        English
        French
        German
        (etc.)
    (period)
        Mediaeval
        Sixteenth century
        Seventeenth century
        (etc.)
    (form)
        Poetry
        Drama
        Letters
        (etc.)

However, when concepts are compounded to represent the subjects of documents, the following (example) filing sequence would result:

Literature
    English literature
        Sixteenth century English literature
            Sixteenth century English poetry
    French literature
        French medieval literature
    Mediaeval literature
    Poetry
        English poetry

This breaks the rule of general-before-special, with, for example, French mediaeval literature preceding Mediaeval literature, and sixteenth century English poetry preceding English poetry. Such a sequence would be illogical, and unhelpful and confusing for readers. If, however, the filing order of facets is reversed, or inverted, a much more satisfactory order is achieved. This does not conflict with any of the filing principles and creates a more intuitive collocation of topics.

[top of entry]

7.4 Notation

The → notation in BC2 uses both capital letters of the Roman alphabet and Arabic numerals. Numerals have a lower filing value than letters and precede them in the sequence. Other symbols such as punctuation marks or non-Roman letters are not used. At the top level of the classification, numerals are used for the Generalia classes including phenomena, and letters for main classes. Otherwise, broadly speaking, letter classmarks predominate, and numerals are used to indicate subsidiary and common classes. In the classification display a space is inserted after three characters, but this is simply for ease of reading; it has no semantic significance and can be omitted in labelling documents or records.

The primary function of the notation is to maintain the filing sequence and the order of classes as they have been determined by the methodological principles of BC2. A significant feature of the BC2 notation is that it is non-expressive either of hierarchy or of the structure of built classmarks. One of Bliss’s objectives was to provide short, economical classmarks, and this policy has been followed through in BC2. This has the advantage that overall, the classmarks are short, and therefore more easily understood and remembered by users; additionally, since the classmarks are not required to express anything of the semantic nature of the classes, short classmarks may be allocated to important subjects, that is, those with a substantial volume of literature. The lack of expressivity of the notation can be counted a disadvantage, particularly in the digital age, since the classmarks cannot be readily interpreted by software. The classmarks are not suitable as a basis for machine retrieval since the notational coding for any concept is not necessarily consistent, and it may not be recognised within a built classmark. This has been a problem in the development of the digital version of BC2 where the natural affinity of the faceted scheme with complex search and navigation is compromised by the non-expressive notational coding. The problem is alleviated to some extent by the use of hierarchy codes in the local BC2 markup.

[top of entry]

7.4.1 Retroactive notation

The notation is fully faceted and synthetic and is used to support the synthesis or building of classmarks to represent compound subjects. Because of the principle of inversion, in order to apply the citation order, the concepts in a compound are combined in the reverse order of that in which they appear in the schedule (retroactive synthesis). The notation is allocated in such a way that in each facet notational space is left to allow for the addition to any class of all preceding facets. For example, in class J:

J	Education
JA	Principles of education
JB	Educational administration
JK	Curriculum
JL	Schools by characteristics other than stage of education
JM	Primary schools
JMN	Preparatory schools

The first enumerated subclass of JM Primary schools JMN Preparatory schools is assigned a letter (-N) later in filing value than those introducing preceding classes (-A for Principles of education, -B for Administration... -L for Curriculum) so compounds can be formed by direct addition of these to JM (e.g., JMK Primary school curriculum) without clashing with the enumerated subclasses.

[top of entry]

7.4.2 Intercalators

If this principle were strictly followed throughout the scheme, then number building would be a very straightforward matter, since the notation for each concept in a string need only be combined in the reverse of alphabetical order. In practice, because of the pressure on the notational allocation, it is not possible to sustain a pure retroactive method, and building is sometimes supported by the use of intercalators (a notational symbol which allows the insertion at a required point of one or more qualifying facets). In BC the letter A is often used to replace retroactive synthesis, when the reservation of a large number of letters to accommodate preceding facets and arrays proves inconvenient. For example:

QF	Social security
QFA	* Add to QFA letters A/E following Q in QA/QE      e.g. QFC Social security in UK or favoured country

While this device has maintained the policy of shorter classmarks, in practice it is found confusing by classifiers, and no doubt reinforces the belief that BC2 is complicated to apply.

[top of entry]

7.4.3 Notational replication and specifiers

It is often the case that concepts in one part of a schedule have relevance in another, and since the essence of a faceted scheme is that it should not unnecessarily enumerate, or duplicate, classes, terms from one facet or array are “borrowed” to generate classes in another. A typical (and frequent) example is where geographic terms are used to create national or cultural concepts in a different part of the schedule, but many other examples occur. For instance, in Architecture, terms from the facet of building materials are used to create the classes “Buildings by material” e.g., Timber as a building material WHO MT, Timber buildings WHT VMT. Such a term (i.e. timber) is defined as a specifier, that is, a term which defines a species of something by a characteristic occurring elsewhere in a different relationship. The significance here is in the different relationship; timber as a material is not the same as a building made of timber (nor is Japanese music the same as Music in Japan).

[top of entry]

7.5 Pre-combination within the schedules

A particular feature of BC2 particularly in its later volumes, is the provision of pre-combined or pre-coordinated classes for many compound subjects. This practice has two objectives: firstly, to provide guidance in the business of synthesising classmarks; and secondly to ensure that concepts that are semantically compound but terminologically “simple”, are represented in the schedule (and therefore in the index) and are consequently discoverable. For example, the concept “Arthritis” is a compound one represented by adding the notations for “joints” and “inflammation”. While there is no theoretical reason why the classifier should not construct this for her/himself, it greatly reduces the work of classifying documents if these pre-coordinated classmarks are provided, especially in a field like medicine where technical terminology proliferates. The logical outcome of this feature is that it greatly enlarges the schedules, but it accommodates a much greater range of vocabulary, and it allows for the construction of a much more sophisticated and complex classification than would be the case for a bare facet structure.

[top of entry]

7.6 The alphabetical index

An alphabetical index is provided in each published class, although there is no integrated index to the whole scheme. The index has two main functions: to indicate the locations (and therefore classmarks) of subjects; and to bring together distributed examples of those subjects i.e. to show where a subject occurs in different contexts. For example, in class C Chemistry:

Reactivation
    : Analysis   C9Q 7QL
    : Organic chemistry   CUL BSG
    : Physical chemistry   CBS G

Unlike some classification schemes (Dewey, for example, maintains the Relative Index as a separate database), the index to BC2 is derived directly from the source code for the classification. All elementary classes in the schedule are entered into the index, but compound classes are generally not, nor are classes generated by the use of intercalators. While it would be an ideal solution to include these, the resulting huge increase in the size of the index prevents such an exercise. The same policy is applied to compounds given as examples of combination using intercalators or specifiers. However, distinctive terms provided for by pre-combination as described above (Section 7.5) are usually indexed.

Although the index is generated from the source code, the output requires some manual editing.

[top of entry]

8. Reception of the scheme

A useful overview and critical analysis of commentary on the new scheme in the years immediately after first publication, authored by D. J. Campbell, is to be found in the Bliss Classification Bulletin 1980. As well as reviews proper, this survey includes material from journal articles, and in such cases the full text of the content is frequently included.

Some of these items are more general discussions of BC2 and its features, rather than evaluations of specific publications, and formal reviews of the published classes were unfortunately rather few and far between. This phenomenon caused some disquiet to the BCA Committee, but perhaps reflected the lack of persons able to provide a critical assessment rather than a failure to distribute volumes for review. Campbell (1980, 23) records that “although review copies were sent to nearly fifty library, information science and other journals, mainly but not only in the English-speaking world, few seem to have thought it worthy of review”. He also notes (1980, 29) that pressure had to be brought on the editor of the Library Association Record to review the new scheme, after which a positive critique did in fact appear (Evans 1979). Reviews by practitioners include one for class I Psychiatry from McKeown (1979), medical sub-librarian at Queen’s University of Belfast, as well as those for Anthropology and Medicine by Bury, librarian at the Haddon Library (1980; 1982; 1984). Most of the early reviewers, however, were LIS academics (Roberts 1977; Bakewell 1978; Ramsden 1978; Evans 1979) rather than practitioners, and in some cases the comments are taken from articles on more general classification topics (Coates 1978, 292-3; Curwen 1978, 29-31).

Roberts’ review in Catalogue & Index was responded to by a letter in the following edition of Catalogue & Index from Ross Trotter, head of the Dewey Section at the British Library, which is critical of a great number of aspects of BC2; of these, Campbell (1980, 23) lists “its layout and printing, its multi-volume presentation, […] the faceting, the hierarchy, the treatment of interdisciplinary works, the handling of alternatives, the index, and the outline of the whole scheme” as all coming in for criticism. This in its turn elicited a response from David Evans at the Leeds School of Librarianship, which was more even handed in its assessment of BC2 and commends “its resolution of so many theoretical difficulties” (Evans 1980, 8).

Two unsigned reviews, one general, and one for Class I, appeared in the journal Documentation on Books published in the Netherlands, about which no further bibliographic detail is provided; both are brief and essentially descriptive with no particular positive or negative stance. A more considered opinion is the review in Bulletin des Bibliotheques de France (Laureilhe 1978). Here there are some criticisms, but overall the tone is positive.

A similarly small number of reviews can be located post 1980, including those of Danesi (1981), Wood (1983), Bury (1984), Hudon (1988) and Viti (2010), with only two discoverable for volumes published in the 1990s and 2000s (Perkins 2011; Power 1995). A useful practical evaluation of various positive and negative responses to BC2 is provided by Attar (2002) in her examination of applications of the scheme to the literature collection at the Senate House Library in London, and of responses to a survey circulated to Cambridge libraries in 2000. A review of an associated publication, Maltby and Gill’s A Case for Bliss (1979) concludes that it makes “a pre-packaged introduction to BC”, and “it contains much that is useful” (Howorth 1980), but that it does not attempt to make a case for Bliss as compared with the other general schemes.

Overall, the new classification was received positively, and considered to be technically much superior to other schemes. Thomas (1992, 16) describes it as “a modern, sophisticated, detailed, flexible tool for librarians to assess for total or partial use in collection arrangement, catalogs, indexes, and databases”. Roberts (1977, 5) speaks of &ldsquo;the magnitude of an accomplishment achieved with the minimum of resources. Without doubt BC2 appears as a major bibliographic classification, soundly based on literary warrant and reflecting modern classificatory principles and practices to an extent which renders it technically superior, at all levels, to its main competitors, the Dewey and Library of Congress schemes and, probably, in most respects to the Universal Decimal Classification”. Standard works on classification in the period following publication are also positive; → Foskett (1996, 295-302) devotes a chapter to BC, most of which is a substantial description of the new edition and commends it (1996, 301) as “this modern scheme with its superior main class order and detailed analytico-synthetic schedules’, which “library schools wishing to demonstrate analytico-synthetic classification would find […] by far the best model”. Similarly, Marcella and Newton (1994, 109) allude to “the many excellent qualities it displays, particularly its structure, rigid adherence to and observance of faceted principles, inherent logic and the quality with which it is explained”.

Shortcomings are listed by Curwen (1978) as including: the poor quality of the physical presentation and printing, with its “minute” type size, lack of sharpness, the lack of room for annotating the schedule; errors and omissions in the various indexes; the failure to provide an index to the auxiliary schedules; complexity in the instructions for class-mark synthesis, particularly where intercalators are used; and an excessive number of misprints and errors in indexes, cross-references, and instructions. These problems are echoed in the review by Bakewell (1978) who felt that BC2 had failed to address criticisms concerning presentation and accuracy levelled at the original scheme.

Generally, BC2 was not criticised on the basis of its content, with the exception of the University of Ibadan Library, a long-time user of BC1, which had decided not to adopt BC2 ostensibly because of disagreement with one section in a draft of the Linguistics class (Ukoh 1975). These criticisms were publicly responded to (Mills 1976a), but no further discussion occurred, and it seems far more likely that the decision to switch to the Library of Congress classification was motivated by cost and convenience.

In this vein, Roberts’ critique (1977) raises some broader questions concerning BC2’s position in the classification landscape, and how it might respond to the “centralising tendencies” of modern cataloguing practice; could it be accepted by the relevant agencies, or had it arrived too late? The letter in Catalogue & Index from Ross Trotter, mentioned above, is critical of BC2 at almost every level, but is particularly telling on the administrative front; how could BC2 compete with Dewey’s “monolithic unity, its extensive and international machinery of revision” (Trotter 1977, 7-8). He also questions the relationship with BC2 of BC1 libraries, and the viability of their changing to what is “virtually a new scheme”. In fairness, the revisers of BC2 were not oblivious to this challenge. Mills (1972, 27) writes that the “law of the market which predicts that nothing will keep a good product down does not operate automatically in the world of library classification, where the forces of inertia, the cost-benefits of centralised indexing and the difficulties of re-classification tend strongly to perpetuate the system in occupation, even if a better one comes along”.

In general terms the new classification was applauded, but many commendations came from established supporters of BC2. It must also be conceded that much of this congratulation was for the revision as an exercise in building a faceted classification and the development of associated theory rather than as a practical tool for the organization of a library. Laureilhe’s (1978, 50) review of 1978 represents this position nicely: “this is a very scholarly classification. […] It represents a huge effort of coherent organization and is interesting for classificationists to study”. However, it undoubtedly was also successful as a workable system. Bakewell (1978, 44) says that “the result is far too good to be regarded simply as an excellent teaching tool”, and Coates (1978, 292) commends BC2 for “closing the gap between classification theory and practice” in that it demonstrated the theory “in real life”.

[top of entry]

9. Influences and collaborations

BC2 has been influential both on other classifications and on thesauri, for several of which it has provided the terminological foundation. This borrowing of the BC2 vocabularies and theoretical structure has been a feature since the start of the BC2 revision programme and is characteristic of the general open-source ethos of the scheme.

One of the most significant of such exchanges was with the Universal Decimal Classification, whereby UDC made use of the BC2 work in its plans to create a more fully faceted structure in that scheme. This project was detailed in several papers by McIlwaine (1995; 1996; 1997), and McIlwaine and Williamson (1993; 1994; 1995). McIlwaine’s 1997 paper considered examples of collaboration and cooperation between editors of all the principal general classification schemes; of general interest is the description of the routine exchange of drafts and working papers between the schemes, and the particular collaboration between UDC and Dewey with the aim of achieving more consistency between the two in aspects such as place subdivisions. There is an account (McIlwaine 1997, 88) of the more formal arrangement between the editors of Bliss and of UDC:

An agreement has been reached between the two organizations responsible for the schemes which permits the UDC to use the structure of the revised Bliss scheme as the basis of new revision. It has been agreed that the revisers of the Universal Decimal Classification may develop and convert such classes as seem appropriate (subject to the Bliss Classification Association's approval) into UDC schedule. […] This is a reciprocal arrangement so that, should they wish, the editors of the Bliss revision may similarly borrow the subject structure or terminology from the UDC. This arrangement originated in the concept that an economic and more useful approach to the development of classifications schemes in the present environment would be to use the same basic semantic network as a standard from which indexing languages of any kind (including schemes of classification) may be derived. The end user will see a retrieval tool in the form to which he is accustomed, but the basic framework is identical, whether a decimal notation or an alphabetical notation, or verbal headings, or even a combination is used to represent the concepts identified.

Recent UDC revisions have consequently made use of the BC2 published schedules and drafts. The two most striking examples being those for class H Health Sciences, and class P Religion, which are to some considerable extent a mapping or conversion exercise of the kind described by McIlwaine above. Health sciences has been by far the larger of the two projects and has consumed a great deal of time, partly because of the size of the vocabulary, but also because the underlying structure of the two schemes is so different and representing the BC2 vocabulary in a UDC format was not easy. The UDC has no equivalent of the structural signposts in BC2, such as the characteristics of division, or node labels, which while not classes in themselves act as containers or collecting points for arrays of terms. Trying to map this structure to the more hierarchical UDC proved difficult, as did representation of such classes in the CDS-ISIS database which holds the classification data for UDC.

The revision of class 2 Religion in UDC (Broughton 2000) faced some of the same difficulties, although in this smaller class it was easier to introduce some innovative techniques for addressing them. A major feature of the new class 2 was the use of an extended special auxiliary to support number building for complex subjects instead of the more usual UDC colon device. Numbers so produced are much briefer without significant loss of the semantic representation inherent in colon addition of full main table numbers. The special auxiliary closely follows the structure of the BC2 schedule, bringing a fully faceted dimension to UDC, and the structural signposts were more easily converted to containing classes in this humanities discipline. A particular problem that is not as yet fully resolved is the phenomenon of differential classes within individual named religions; for example, named sacred texts which are specific to the religion, and do not have a generic counterpart in the auxiliary schedule, and therefore not in the database. A feature in parallel between the two classification schemes is the provision of much of this religion-specific vocabulary in the form of expanded schedules, or examples of combination in UDC parlance. A great deal of this collaborative UDC/BC2 work in the area of religion was utilised in a research project funded by the Arts and Humanities Research Board at University College London, designed to test the feasibility of a faceted approach to humanities classification. The Faceted Analytical Theory in Managing Knowledge Structures (FATKS) project (Broughton and Slavic 2007) demonstrated the theoretical principles of building a faceted scheme on the BC2 model, its representation in a database format, and the potential of such a faceted structure to support automatic classification.

The chain of influence from BC2 to UDC was further extended when Dewey’s religion class of 2012 (Mitchell et al. 2012) incorporated some of the same ideas. Broughton (2017, 487) writes of it:

Intended to be used in conjunction with the Abridged Edition was the single volume 200 Religion Class, which was an updated extract from the 23rd full edition. This volume was prepared in collaboration with the former Editor-in-Chief of the Universal Decimal Classification (McIlwaine and Mitchell 2006), and displays the influence of the 2000 revised religion class of that scheme (Broughton 2000), which in its turn was based on the Bliss Bibliographic Classification 2nd edition (Mills and Broughton 1977). The aim of all these revisions was to better and more equitably represent non-Christian religions, and to eliminate Christian bias. Dewey’s 200 Religion Class achieved this structurally, and in terms of the greater vocabulary and number of classes subordinated to each faith, but the notational provision, which cannot easily be changed in such a widely used scheme, still creates an impression of Christian dominance. It is supported by the online Dewey Religion Browser https://www.oclc.org/dewey/ features/religion/browser.en.html which provides an expandable summary of the new arrangement.

Otherwise, BC2 has been employed both as a model for the construction of faceted systems (Broughton 2017b; Broughton and Slavic 2007), and as a basis for the creation of faceted interfaces for digital organization and retrieval (Broughton 2001; 2002a; 2002b; Broughton and Lane 2002; 2004).

[top of entry]

10. BC2 as a source for a thesaurus

Facet analysis as a means of structuring a classification and associated thesaurus has been employed since the late 1960s with the creation of English Electric’s Thesaurofacet (Aitchison, Gomershall and Ireland 1969). Aitchison (1986, 160) considers it “a powerful tool [ ] which engenders a more precise understanding of the terms and how they are interrelated.” In her proposal for constructing an integrated social sciences thesaurus, Aitchison (1981, 77) states that, specifically, BC2 meets most of the structural requirements that allow it to serve “as the master in a compatibility matrix for comparing terms from thesauri as well as notation from classification systems in the sub-fields of the Integrated Thesaurus”.

The facet analytical technique could now be described as standard in thesaurus construction and is detailed in a number of sources (Aitchison, Gilchrist and Bawden 2000; Broughton 2006b; Shearer 2004). Elsewhere Aitchison and Dextre Clarke (2004, 12) confirm that “[t]he fully faceted 2nd edition of the Bliss Bibliographic Classification was an important inspiration for relationships and order of concepts in some of these thesauri”, that is, schemes constructed by Aitchison and others on the faceted model, such as the UNESCO Thesaurus (Aitchison 1977), the BSI ROOT Thesaurus (BSI 1988), and the International Thesaurus of Refugee Terms (Aitchison 1996). The use of an existing faceted classification as a basis for extracting a thesaurus is first described by Aitchison in 1986, and in a paper outlining the creation of the DHSS-DATA Thesaurus (Aitchison, Brewin and Cotton 1985) and ECOT (Educational Courses and Occupations Thesaurus) (Aitchison 1984) from vocabularies in classes H Health sciences, J Education and Q Social Welfare of BC2, BC2 having “a number of features which make it thesaurus-compatible and an attractive source for thesaurus compilers” (Aitchison 1986, 161). Aitchison (1986, 162-165) explains, with a number of practical examples, how the standard thesaural relationships are all implicit in the well-structured faceted scheme, and that it contains scope notes and definitions that also may be of use in the thesaurus. Nevertheless, “[t]he substantial benefits of BC2 as a thesaurus source have to be offset by some problems which must be recognised and dealt with. These problems include gaps in coverage, the repetition of identical concepts in several places in the schedules, imprecise terminology and the occurrence of compound terms which break the thesaurus construction standard rules” (Aitchison 1986, 166). Despite these difficulties Aitchison went on to devise other thesauri based on the BC2 vocabularies, including that for the Royal Institute of International Affairs at Chatham House (Aitchison 1992), generated from class R Politics and Public Administration.

Such problems are acknowledged in other sources (Aitchison 2004; Broughton 2008), and largely arise from the fact that the BC2 schedules were not originally devised with a thesaural version in mind, and so the formatting of captions and class names often fails to meet the conventions of thesaurus construction particularly with respect to compound terms. As Broughton has it (2008, 207):

The BC2 terminology accommodates conceptual compounds, but is weak regarding the principles of how they are to be displayed. Often this is because the meaning of a term is qualified by preceding terms, or is otherwise inferred from contextual clues, and the term, although structurally compound, is not displayed as such, or is only partially represented. This usually occurs, either because a class is ‘added’ from another schedule or part of the schedule, or because subdivisions of a class are expressed conceptually rather than as explicit terms. In the majority of cases, a ‘good’ thesaurus term can be created by restoring the ‘missing’ parts of the class.

Aitchison’s 2004 paper was written as part of a BC2 project to develop a thesaurus format for the scheme. It aimed to lay down some guidelines for formatting in BC2 that would help to avoid these barriers to the extraction of thesauri directly from the schedules. The draft schedule for Music was chosen as a source, since a workshop held earlier in 2004 for music librarians in Cambridge (Aitchison and Curwen 2004) had indicated that there was a ready market for a thesaurus in this discipline. However, Aitchison (2004, 20) observed that “the problems and possibilities revealed in this exercise echo those reported nearly 20 years ago concerning the experience of using Bliss schedules as a source for the DHSS-Data Thesaurus, although the circumstances of this project and some of the technical problems are different”.

After the workshop in 2004 a suite of programs was devised by Paul Coates to support the automatic generation of a thesaurus from the existing source code for the classification. Preliminary results were quite encouraging, but since all the earlier schedules displayed the problems of formatting and style described above, the output needed a considerable degree of manual editing to meet the thesaurus conventions. Revision work since then has incorporated the suggestions made by Aitchison, although to date no new class has been published with an accompanying thesaurus.

[top of entry]

11. BC2 and ontology engineering

BC2 has not so far been used directly in research into ontology engineering, but it exhibits several features elsewhere addressed in the research literature: a sophisticated semantic structure at a high level of organization; a multi-dimensional approach to concept organization; a range of concept interrelationships beyond those used in the standard thesaurus; a coding system that is machine processable and has the capacity to map to other mark-up languages. It is worth noting that most of the work described below references BC2, both as an example of facet methodology, and as a conceptual model.

An → ontology is essentially a terminological tool for a given domain that, when it is made computer processable, supports intelligent search through automatic inference (Szeredi et al. 2014, 66). The purpose of the ontology is to provide background knowledge of the domain and its semantics by identifying its concepts, their classes or categories, attributes, and interrelationships. Therefore, just as it provides an excellent conceptual basis for a thesaurus, there is potential for a KOS like BC2 to support the creation of ontologies.

Since ontology first became a topic of interest in retrieval the relevance of classificatory structures has been noted. Soergel (1999, 1120) asserts that ontologies are necessarily classifications, those deemed necessary in “fields such as AI, natural language processing, and software engineering”, while also observing that “the vast body of knowledge on classification structure and on ways to display classifications developed around library classification and in information science more generally, and the huge intellectual capital embodied in many classification schemes thesauri is largely ignored”, despite the potential usefulness of such knowledge.

There are a number of studies that compare various kinds of knowledge organization systems, but it is generally agreed that the ontology represents the most developed kind of tool, characterized by the highest level of semantic richness. Strictly speaking, ontology lacks agreed definition, and the term has been claimed for various vocabulary tools, some structured at a quite simple level, but ideally the ontology should demonstrate an internal logic and a variety of relationships if it is to support automatic reasoning in a retrieval system. “Ontologies describe objects, classes, attributes, and relations using controlled vocabularies, but are much more than controlled vocabularies because ontologies accommodate many different types of relationships including hierarchical, associative, and equivalencies” (Adamich 2015, 12).

Because of the semantic sophistication of faceted classifications, a number of projects and studies have demonstrated their use as models for ontology building (Prieto-Diaz 2003; Giunchiglia and Zaihrayeu 2007; Madalli et al. 2015). Like other faceted systems BC2 has many more potential relationships than those usually occurring in online search tools. As well as the familiar broader and narrower relationships, and relationships of semantic equivalence inherent in hierarchies and to be found in the thesaurus (the intra-facet relationships), relationships between terms across facets (extra-facet relationships) are also apparent. Examples might include operation/agent relationships (teaching and teachers), operation/product (building and houses), entity/part (vehicle and wheels), all of these coming under the general head associative terms in the thesaurus.

The difficulty for BC2, and, indeed, for other classification systems, is how these relationships can be modelled in a machine processable language. → SKOS (Simple Knowledge Organization System) has been developed as a web language that can represent classifications, thesauri, and the like, but it tends to operate within the confines of a single facet or hierarchy, enabling the representation of broader and narrower (sub-class and super-class) and equivalence relationships. Changes were made to the original version of SKOS to allow it to represent the arrays in a faceted classification or thesaurus through the use of collection to contain the members of a set (SKOS Primer 2009, https://www.w3.org/TR/skos-primer/#seccollections). Their example demonstrates the structural feature of characteristics, or principles, of division (node labels in thesauri). SKOS knows that these are not part of the hierarchy, and can accurately identify the broader terms, as also happens with the BC2 markup language. However, the primer does not make provision for other features, suggesting that to do so would rob SKOS of its “simple” label, and it is left to the user to design their own extensions.

Binding et al. (2021, 942) note in their study of the conversion of the Integrative Levels Classification, “such complex classification schemes [i.e. faceted schemes] have features which are not straightforward to represent in SKOS and which extend beyond the functionality of the SKOS standard”. Similarly, Prasad and Madalli (2009) when considering SKOS in relation to the Colon Classification, observe that SKOS:collection is not able to represent facets, nor synthetic combination of concepts.

Binding et al. built their own extensions to SKOS, and the work done by this group has considerable implications for BC2; as they say (2021, 942):

An approach for representing faceted compound descriptions in the ILC and other faceted classification schemes is proposed, building on previous discussion on extensions to SKOS for coordination. SPARQL examples illustrate how the new extension property could be used to create faceted, compound descriptors when indexing or cataloguing. Basic query patterns are provided that might underpin future search systems. Crucially, the patterns for compound descriptions include concepts and properties, in order to allow relationships between concepts according to the facet syntax.

ILC has an advantage over BC2 in that it includes a generic category of facets, where the relations between facets are made explicit and encoded (e.g. 3 “by agent”, 5 “with transformation“), whereas in BC2 the inter-facet relationships are implicit and hence not easily identified as such. Prasad et al. (2017) also introduce a model, DERA, in which relations are an intrinsic element of a four part data structure: Domain [D]; Entity [E]; Relation [R]; and attribute [A]. The DERA formalization of a domain can be directly translated into description logics which then enable the automatic reasoning required in semantic search and retrieval. A very similar approach is taken by Giunchiglia et al. (2012), using the acronym CERA (with C for class), although this lacks the translation.

Our work on the FATKS project (Broughton and Slavic 2007), which used the BC2 subject vocabularies and structure to design a search and retrieval tool for the humanities, demonstrated that it is possible to model many of the features of a faceted scheme. The FATKS system proved effective at the proof-of-concept stage in both these functions, and it supported a degree of automatic classification through the use of facet status coding, similar to ILC, and an imposed standard citation order for managing complex document content. However, we constructed our own database, and this was before the development of SKOS, so it is unknown whether the data would map easily. BC2 currently has an XML version in development, and it is hoped that this will be the first stage in building a bridge between the conceptual structure and a proper knowledge representation language. BC2’s coded classification data mapped easily into XML and already supports a number of basic ontology features.

[top of entry]

12. Users

It has always been difficult to determine a precise figure for the number of BC2 users at any given period, since, as Attar (2002, 48) puts it, “membership of the Bliss Classification Association includes institutions which do not use Bliss, and […] not all Bliss users are members of the Association”. The situation is also complicated by the “interregnum” post 1967, when BC2 was incomplete and when there was no single authoritative version of BC. While new adopters of BC2 in this period are fairly straightforward to identify it is not so clear to what extent and in what way existing users changed from BC1 to BC2. Although no BC2 schedules were published until 1977 it is apparent that some libraries used draft versions both to test and to apply BC2 before that date, even while there were occasional new adoptions of BC1 and rather more moves away from BC1 to other schemes. A fashion for reclassification to LCC or DDC, and, in the UK, multiple mergers of institutions and the closure of small information services, generated a situation of much instability where classification usage was concerned (Broughton 2006a, 399-401; Broughton 2017a, 483-4).

The Bliss Classification Bulletin for 1988 (Curwen 1988, 11-17) reports on a survey sent out with the 1987 Bulletin which reveals that thirty libraries in the UK had adopted BC2, three of them reclassifying from BC1, while thirty-nine were still using BC1. In another case BC2 was being used to organize a database, and in a second as a tool for vocabulary design. Seventeen former users of BC1 had changed to other classification schemes, or, in some cases, ceased to exist. This survey probably presents the most accurate and complete picture of concurrent usage of the two editions, as well as the Abridged “Schools” version, and the special schedule for Hospitals. It includes a group of 21 unconfirmed users, and there is only a limited record of BC use outside the UK. Overall, 90 libraries are mentioned as current or previous users of BC1 or BC2, and, before the great take-up of LCC by academic libraries in the late 1990s and beyond, BC may have been the most widely used scheme in the UK after Dewey. However, it is only a snapshot in time. As noted below, a number of libraries implemented BC2 after this date, and in 1995 Thomas (1995, 115) referred to “the present 50 or so users in Britain”, although this seems an optimistic number. Both Attar (2002, 48), and the Bliss Classification Bulletin (2001, 5) refer to a later survey put out on the library and information science mailing list LIS-LINK in 2001 in an attempt to establish the numbers of users not in membership of the Bliss Classification Association, but any results failed to be recorded.

As might be expected, initial take up of the new scheme was limited, as the published schedules only gradually became available, and the Tavistock and Portman Institute in London was the only major BC1 library to reclassify to BC2 (Gill 1979) which it did with the benefit of a grant from the British Library. Staff there had been instrumental in developing the new class I, Psychiatry, and a monograph, The Case for Bliss (Maltby and Gill 1979), was published in the same period. A halfway house is described by Brewin (1985, 3) at the UK Department of Health and Social Security, where original Bliss numbers were linked to the BC2 derived DHSS-Data Thesaurus as part of “a major automation project using STATUS, the powerful text-retrieval package developed by the A.E.R.E. [Atomic Energy Research Establishment] at Harwell”.

Smaller academic institutions reclassifying to BC2 included St. Martin’s College Lancaster, a former BC1 user, and Chester College which converted to BC2 from Dewey (Stoddard 1979). Just as a number of libraries in the social service and voluntary sectors chose BC1 because of its unique main class for social welfare, so similar libraries adopted BC2. These included Dr. Barnardo's (Preddle 1982), the Office of Population Censuses and Surveys (Horsey 1983), the Central Council for Education and Training and Social work (Cowley 1987), and the National Institute for Social Work (Ryan and Upton 1991). Additionally, the Bliss Classification Bulletin for 1978 records the adoption of BC2 by Beaconsfield Social Services and the National Council for One Parent Families. As more classes were published throughout the 1990s the list extended to libraries at the National Society for the Prevention of Cruelty to Children (Mitchell and Power 1996), the Merseyside Probation Service (Wright 1995), Alcohol Concern (Johnson 1997), and the National Autistic Society (Curwen 1998, 2). The Supreme Court Library was also applying BC2 to part of its collections (Best 1979) as part of the revision process for class S, Law. Libraries taking up the scheme post 2000 included the library of the General Teaching Council for England in 2002 (Hasleton 2002, 17), the Empire Museum, Bristol (Lane 2004, 4), and, in 2004, St. Peter’s College, the first Oxford institution to adopt BC2 (Mills 2005, 9). It was reported (Curwen 2004, 31) that three other Oxford Colleges and the Pitt-Rivers Museum would like to reclassify to BC2 but lacked the funds for such a project; although it cannot be certain, these were possibly Exeter, Merton, and University College which had all been using BC1. Greater Manchester joined Merseyside, West Midlands and West Yorkshire in the growing number of regional Probation Services libraries using BC2, also in 2004 (Curwen 2004, 31).

BC1 was in use in several libraries in Cambridge and some institutions there began to implement BC2. Russell and Wilkinson (1979, 12) suggest that as early as 1979 four libraries were known to be using, or about to use the scheme; these were certainly King’s, Fitzwilliam, and the Haddon Library (which had used BC1), but it is not clear which was the fourth. At the Haddon, Faculty Library for Archaeology and Anthropology, reclassification for the Anthropology collection was completed in 1988, and the Library was actively involved in the revision of the schedule for Archaeology (Russell and Lane 1999), then as yet unpublished. Queens’ College began to implement the scheme in 1988 (Sargent 1990), and King’s began to reclassify its specialist music collection, the Rowe Library, to BC2 (Russell 1997, 17). In the mid-1990s BC2 use received a boost when two other Cambridge college libraries adopted it (Attar 2000a; 2000b). These converts to BC2 were Sidney Sussex (Lane 1999; Lane and Butler 1998), and the Quincentenary Library at Jesus (Watson 1997) which began its reclassification in 1996. A user and support group, CAMBUG, was set up in 1995 (Lane 1995). An inaugural meeting on 23 March was attended by “ten librarians from nine college and departmental libraries”, suggesting that there was Cambridge interest in the scheme beyond its six actual user libraries. One advantage of the larger user base in Cambridge was the availability of online BC2 data, since “[a]ll Bliss libraries contribute[d] to the Cambridge Union catalogue” (Attar 2000b, 12), the first time that anyone had been able to access records with BC2 classmarks attached. When the new Voyager library management system was introduced at the University Library, BC2 was authorized as “a recognized” classification (MacLeod 2003, 32). The scheme gained some further recognition with the addition of BC2 to the list of classification source codes for use in MARC21 at the Library of Congress (Mills 2003, 9). Among other Cambridge libraries which were using BC1, Homerton reclassified to Dewey, and the Radzinowicz Library in the Department of Criminology (Kraup 1999) retained their modified form of the scheme which is still used today. In her study of the application of BC2 in Cambridge libraries Attar (2000a, 37) suggests that the scheme was chosen by librarians who “were attracted to a faceted, flexible system with a cross-disciplinary aspect which they felt suited the Cambridge teaching system”. In the same paper, disadvantages were seen as the lack of published schedules (2000, 37), and the disparity that then sometimes occurred between draft schedules and the published versions (2000, 39), lack of reader familiarity (2000, 47), and the need for intellectual effort and commitment to classification (2000, 47). More details of the survey carried out are given in an article in the Bliss Classification Bulletin (Attar 2000b), where it is stated that the scheme was viewed sympathetically, and that “the objections concern practical implementation rather than properties of the scheme” (Attar 2000b, 16).

In 1991, the Bliss Classification Bulletin no. 33 reported that:

Committee members had been in communication with Eric de Grolier throughout the year concerning the choice of a new classification scheme for the open access collection of the Bibliotheque de France. M. de Grolier was keen to recommend BC2 as a source. The Classification Research Group would be inviting him to a meeting early in 1991, and BCA members would be able to attend.

The de Groliers were known advocates of BC1 and had tried to persuade Bliss to collaborate with them even before the publication of his original scheme (Goforth 1980, 32). But Bliss had declined their offer to translate the classification into French and promote it as an international tool, fearing that they would take credit for any development. This later offer suggested that their admiration for the scheme was genuine, and in February 1991 de Grolier visited the UK for a meeting (Classification Research Group/Bliss Classification Association 1991, 534; Foskett 1991, 90). A detailed review of BC2 had been prepared by de Grolier [3], and many of the points raised there were addressed in a response by Jack Mills in the 1991 Bulletin (Mills 1991, 17-21). Ultimately Dewey was the chosen scheme rather than BC2 doubtless on account of the latter’s incompleteness, and the “slowness of production” (Mills 1991, 21).

The use of the Bibliographic Classification in larger libraries was more widely on the decline, and it became gradually dislodged from its earlier university library base. A major trend in the early twenty-first century for reclassification, mainly on pragmatic grounds, to either the Library of Congress Classification or Dewey Decimal Classification (Broughton 2012, 399), saw BC disappear either wholly or in part from various institutions, including Birkbeck College, which in 1996 began to reclassify to Dewey. Birkbeck does, however, retain its pre-1996 stock in the original Bliss. The University of London Library at Senate House also abandoned Bliss in favour of LCC, although, as at Birkbeck, older stock remains organized by BC1. BC1 is still retained as the principal scheme at Lancaster University. Re-classification to the second edition, without the benefits of bought-in records and cataloguing services, was never feasible for these larger institutions, even had they been eager to do so. Among Commonwealth nations who had used the original scheme none considered reclassification to BC2, and BC1 was over time replaced by Library of Congress at Tasmania (Eaves et al. 1974), Ibadan (Shoyinka 1975; Ukoh 1975; Fasanya 1977), and Malawi (MacLeod 1984), which had previously used BC1 and Dewey.

Rationalisation of cataloguing practice and the systems used, particularly where amalgamation of institutions and their information services occurred, saw off BC in some smaller units, such as St. Martin’s College Lancaster which converted to Dewey when it was incorporated into a larger HE provider (Weaver and Stanning 2007). Use of BC in the Institutes of Education, which had been a stronghold of BC1, also disappeared as these were absorbed into universities. Several probation services which had taken up BC2 in the 1970s suffered a similar fate.

An accurate assessment of the number of BC2 users in 2024 remains difficult, but because of these losses, it is likely to be quite small. Alongside the thirty BC2 users listed in the 1987 survey, twelve others later adopted the scheme, giving a potential total of forty-two in about 2005. It is certainly the case, however, that by 2005 some BC2 converts had already stopped using the scheme. Six institutional users in the voluntary and charities sector have now ceased to exist, and an attempt to trace others through an examination of websites and OPACs was unproductive, as it is hard to determine whether smaller organizations maintain a library or information service, and if so, how it is organized. The National Society for the Prevention of Cruelty to Children, however, does make its library catalogue visible, and is evidently still using BC2, so it is to be hoped that it is not alone. If so, in combination with the six Oxford and Cambridge libraries, and the Tavistock Institute, the number may be as high as a dozen, but is unlikely to be more than that.

[top of entry]

13. Conclusion

BC2 is widely regarded for its intellectual rigour and for the application of the faceted methodology that make it an exemplar of a faceted classification scheme. At present it remains unfinished, but work is in hand to try and make available the whole classification in a free-to-access form through the publication of both the finished volumes and draft schedules on the Bliss Classification Association website (https://blissclassification.org.uk/bcsched.shtml). Only three classes, H Health sciences, K Society, and T Economics are not accessible there. Most classes exist as PDFs or Word documents, but many are accompanied by their source code, and current work includes the conversion of the code to an XML format that will be supported by an editing tool. Recent revision of the main classes has concentrated on the biological sciences, E/G. Other classes that still require major attention are History, Technology, and Music, and, of course, many early published volumes now need updating.

As of 2024, work continues, but resources, both financial and personnel, are limited. A scheme that at its beginnings was strongly supported by volunteer help finds that less available today, and the situation is complicated by the lack of knowledge of classification building techniques among the profession as a whole. Prospects for adoption by new organizations seem remote in the face of competition from the large systems, Dewey and Library of Congress Classification, and when the trend is towards outsourcing of cataloguing. Satija and Singh (2013, 275) in their study of Colon Classification note a number of features that it shares with BC2 in respect of future development: minority use in the home country; lack of a national committee or substantial institutional backing; poor planning in the revision programme. They conclude that “survival of the system in its present body seems uncertain due to long neglect”. However, they also note the enduring contribution of Ranganathan to library and information science, to the specific principles of knowledge organization, and that “advances in classification theory and practice in the online environment very much depend on facet analysis”. We hope the same is true of BC2, and that the creation of a digital format will encourage the take-up of both the vocabularies and the grammar of BC2 for re-use in a wide variety of contexts.

[top of entry]

Published schedules (including draft schedules)

Introductory Volume

Mills, J. and Vanda Broughton, with the assistance of Valerie Lang. 1977. Introduction and Auxiliary Schedules. London: Butterworths. (A free to access online version is available on the Bliss Classification Association website at http://www.blissclassification.org.uk/Class1/introduction.pdf)

2/9

Class 2/9 Generalia, Phenomena, Knowledge, Information Science and Technology [Draft Schedule]. https://blissclassification.org.uk/Class2-9/2-9_contents.shtml

A/AL

Mills, J. and Ken Bell. 1991. Class A/AL Philosophy and Logic. 1991. London: Butterworths

AM/AX

Mills, J. and Vanda Broughton with the assistance of Nat Lievesley. 1993. Class AM/AX Mathematics, Statistics, and Probability. London: Bowker Saur.

AY/B

Mills, J. and Vanda Broughton. 2000. Class AY/B General Science and Physics. London: Bowker Saur.

C

Mills, J. Vanda Broughton and Eric Coates. 2012. Class C Chemistry. Berlin: de Gruyter.

D/DF

Class D/DF Astronomy [Draft Schedule]. 1972. https://blissclassification.org.uk/bcsched.shtml

DG/DY

Class DG/DY Earth Sciences [Draft Schedule]. 2010. https://blissclassification.org.uk/ClassD/classDG-DY-earth-sciences.pdf

E/F

Class E/F Biological Sciences [Incomplete draft schedule]. 1982. https://www.blissclassification.org.uk/ClassE/E_sched.pdf

G/GY

Class G/GY Zoology [Draft Schedule]. 1982.

H

Mills, J. and Vanda Broughton, with the assistance of Valerie Lang. 1980. Class H Anthropology, Human Biology, Health Sciences. London: Butterworths.

I

Mills, J. and Vanda Broughton, with the assistance of Valerie Lang. 1978. Class I Psychology and Psychiatry. London: Butterworths.

J

Mills, J. and Vanda Broughton. 1978. Class J Education. London; Butterworths.

Foskett, D. J. and Joy Foskett. 1990. Class J Education, Revised Second Edition. London: Bowker Saur.

K

Mills, J. and Vanda Broughton, with the assistance of Valerie Lang. 1984. Class K Society. London: Butterworths.

L/O

Class L/O History [Draft Schedule]. 1997-2002. https://blissclassification.org.uk/bcsched.shtml

LA

Class LA Archaeology [Draft schedule]. 2001. https://blissclassification.org.uk/ClassLA/classLA-archaeology.pdf

P

Mills, J. and Vanda Broughton, with the assistance of Valerie Lang. 1977. Class P Religion, the Occult, Morals and Ethics. London: Butterworths

Q

Mills, J. and Vanda Broughton, with the assistance of Valerie Lang. 1977. Class Q Social Welfare. London: Butterworths.

Preddle, Chris. 1994. Class Q Social Welfare and Criminology, Revised Second Edition. London: Bowker Saur.

R

Mills, J. and Vanda Broughton, with the assistance of Colin Neilson. 1996. Class R Politics and Public Administration. London: Bowker Saur.

S

Mills, J. and Vanda Broughton, with the assistance of Colin Neilson. 1996. Class S Law. London: Bowker Saur.

T

Mills, J, and Vanda Broughton, with the assistance of Valerie Lang. 1987. Class T Economics and Management of Economic Enterprises. London: Butterworths.

U/V

Class U/V Technology and Useful Arts [Draft Schedule]. 1998-. https://blissclassification.org.uk/ClassU-V/ClassU-V-draft.pdf

W

Mills, J. and Colin Ball. 2007. The Arts. Munich: Saur.

WV/WX

Class WV/WX Music [Draft Schedule]. 1972-3, 2000. https://blissclassification.org.uk/ClassWV-WX/classWV-WX-music-draft.pdf

X/Y

Class X/Y Language and Literature [Draft Schedule]. 1998, 2001. https://blissclassification.org.uk/ClassX-Z/classX-Z-languages-and-literatures.pdf

ZA/ZW

Class ZA/ZW Museum Science, Museum Studies, Museology [Draft Schedule]. 2003. https://blissclassification.org.uk/ClassZ/classZ-museology.pdf

[top of entry]

References and secondary sources

References to the editorial content of the Bliss Classification Bulletin (which are numerous and often anonymous) are not included here but are indicated in the text by the issue and page numbers only; authored contributions to the Bulletin are referenced below.

Similarly, references to the content of the Introduction and Auxiliary Schedules are, for the most part, indicated in the text by their section and/or page numbers after the date 1977.

Adamich, Tom. 2015. “Ontologies Part 1: What Are They?” Technicalities 35, no. 2: 12-15.

Adhikary, Madhabmohan and Amitava Nandi. 2003. “Ideas of Ranganathan's Classification Theory Pervaded by Oriental Philosophy”. SRELS Journal of Information Management 40: 275-84.

Aitchison, Jean. 1977. Unesco Thesaurus: A Structured List of Descriptors for Indexing and Retrieving Literature in the Fields of Education, Science, Social Science, Culture and Communication. Paris: Unesco.

Aitchison, Jean. 1981. “Integration of Thesauri in the Social Sciences”. International Classification 8, no. 2: 75-85.

Aitchison, Jean. 1984. ECOT Thesaurus: Educational Courses and Occupations Thesaurus, Prepared for ECCTIS and the Department of Education and Science. Working edition. Milton Keynes: Open University, ECCTIS.

Aitchison, Jean. 1986. “A Classification as a Source for a Thesaurus: The Bibliographic Classification of H. E. Bliss as a Source of Thesaurus Terms and Structure”. Journal of Documentation 42, no. 3: 160-81.

Aitchison, Jean. 1987. Unpublished letter to James Cowan, Computercraft. [Bliss Classification Association papers]

Aitchison, Jean. 1996. International Thesaurus of Refugee Terminology, 2nd edition compiled under the auspices of the International Refugee Documentation Network. New York; Geneva: United Nations High Commissioner for Refugees.

Aitchison, Jean. 2004. “Thesauri from BC2: Problems and Possibilities Revealed in an Experimental Thesaurus Derived from the Bliss Music Schedule”. Bliss Classification Bulletin 46: 20-6. https:// blissclassification.org.uk/B46.pdf.

Aitchison, Jean, P. R. Brewin and J. E. Cotton. 1985. DHSS-DATA Thesaurus: the Thesaurus of the Department of Health and Social Security Library, Working edition. London: DHSS.

Aitchison, Jean and Anthony Curwen. 2004. “Music, Music, Music: the Cambridge Workshop and After”. Bliss Classification Bulletin 46: 27. https://www.blissclassification.org.uk/B46.pdf.

Aitchison. Jean and Stella Dextre Clarke. 2004. “The Thesaurus: A Historical Viewpoint, with a Look to the Future”. Cataloging and Classification Quarterly 37, nos. 3/4: 5-21

Aitchison, Jean, in association with N. Gallimore, S. Boyde et al. 1992. Royal Institute of International Affairs Library Thesaurus. London: RIIA.

Aitchison, Jean, Alan Gilchrist and David Bawden. 2000. Thesaurus Construction: a Practical Manual, 4th ed. London: Aslib.

Aitchison, Jean, Alan Gomershall and Ralph Ireland. 1969. Thesaurofacet: a Thesaurus and Faceted Classification for Engineering and Related Subjects. Whetstone: English Electric Company.

Attar, Karen E. 2000a. “The Application of the Bliss Bibliographic Classification in Cambridge College Libraries”. New Review of Academic Librarianship 6: 35-49.

Attar, Karen E. 2000b. “Blissful Perceptions: BC2 or Not in Cambridge”. Bliss Classification Bulletin 42: 12-15. https://blissclassification.org.uk/B42.pdf.

Attar, Karen E. 2002. “The Practice of Bliss”. Cataloging and Classification Quarterly 34, no. 4: 47-65.

Austin, Derek. 1969. “Prospects for a New General Classification”. Journal of Librarianship 1, no. 3: 149-69.

Austin, Derek. 1976. “The CRG Research into a Freely Faceted Scheme”. In Classification in the 1970s: a Second Look, edited by Arthur Maltby. London: Bingley, 158-94.

Austin Derek. 1984. PRECIS: A Manual of Concept Analysis and Subject Indexing, 2nd ed. London: The British Library Bibliographic Services Division.

Austin, Derek and Peter Butcher. 1969. PRECIS: a Rotated Subject Index. London: The British Library Bibliographic Services Division.

Bakewell, K. G. B. 1978. “Bliss Bibliographic Classification, Second Edition [Review]”. International Classification 5, no. 2: 43-4.

Beghtol, Clare. 1995, “’Facets’ as Interdisciplinary Undiscovered Public Knowledge: S.R. Ranganathan in India and L. Guttman in Israel”. Journal of Documentation 51, no. 3: 194-224. http://dx.doi.org/10.1108/Eb026948.

Beghtol, Clare. 1998. “Knowledge Domains: Multidisciplinary and Bibliographic Classification Systems”. Knowledge Organization 25, nos. 1/2: 1-12.

Best, Kenneth W. 1979. “The Supreme Court Library”. Law Librarian 10, no. 1: 13-16.

Binding, Ceri, Claudio Gnoli, and Douglas Tudhope. 2021. "Migrating a Complex Classification Scheme to the Semantic Web: Expressing the Integrative Levels Classification Using SKOS RDF". Journal of Documentation 77, no. 4: 926-45. https://doi.org/10.1108/JD-10-2020-0167.

Bliss, Henry Evelyn. 1929. The Organization of Knowledge and the System of the Sciences. New York: Henry Holt.

Bliss, Henry Evelyn. 1940/1953. A Bibliographic Classification, Extended by Systematic Auxiliary Schedules for Composite Specification and Notation. New York: Wilson.

Bliss, Henry Evelyn. 1954. “Foreword”. Bliss Classification Bulletin 1, no. 1: 1-3. https://blissclassification.org.uk/B01(I-1)-1954.pdf.

Bliss Classification Association. “Bibliographic Classification”. https://blissclassification.org.uk/.

Bliss Classification Association. 2002. “Minutes of the 2001 AGM”. Bliss Classification Bulletin 44:3-5. https://www.blissclassification.org.uk/B44.pdf.

Brewin, Paul. 1985. “Use of STATUS for Housekeeping at the DHSS”. Vine 15, no. 1: 3-9.

British Standards Institution. 1988. ROOT Thesaurus, 3rd ed. Milton Keynes: British Standards Institution.

Broughton, Vanda. 1974. “The New Bliss Classification”. Catalogue & Index, no. 33 (Spring): 6.

Broughton, Vanda. 2000 “A New Classification for the Literature of Religion”. International Cataloguing and Bibliographic Control 29, no. 4: 59–61.

Broughton, Vanda. 2001. “Faceted Classification as a Basis for Knowledge Organization in a Digital Environment; The Bliss Bibliographic Classification and the Creation of Multi-Dimensional Knowledge Structures”. New Review of Hypermedia and Multimedia 7, no. 1: 67-102.

Broughton, Vanda. 2002a. “Organizing a National Humanities Portal; a Model for the Classification and Subject Management of Digital Resources”. Information Research Watch International Jun: 2-4.

Broughton, Vanda. 2002b. “Facet Analytical Theory as a Basis for a Knowledge Organization Tool in a Subject Portal”. In Challenges in Knowledge Representation and Organization for the 21st Century. Integration of Knowledge Across Boundaries. Proceedings of the Seventh International Conference of the International Society for Knowledge Organization, Granada, Spain, 10-13 July 2002, edited by M. J. López-Huertas and F. J. Muñoz-Fernández. Advances in Knowledge Organization, Vol. 8, Würzburg, Ergon, 135-41. Also available at: http://www.ucl.ac.uk/fatks/paper2.htm.

Broughton, Vanda. 2006a. “Classification”. In British Librarianship and Information Work 1991-2000, edited by J. H. Bowman. Aldershot: Ashgate, 467-88.

Broughton, Vanda. 2006b. Essential Thesaurus Construction. London: Facet.

Broughton, Vanda. 2008. “A Faceted Classification as the Basis of a Faceted Terminology”. Axiomathes 18, no. 2: 193-210.

Broughton, Vanda. 2012. “Classification, Knowledge Organization, and Subject Access”. In British Librarianship and Information Work 2006-2010, edited by J. H. Bowman. London: The Editor, 398-411.

Broughton, Vanda. 2017a. “Classification, Knowledge Organization and Subject Access”. In British Librarianship and Information Work 2011-2015, edited by J. H. Bowman. Lulu.com: The Editor, 479-93.

Broughton, Vanda.2017b. “La classification à facettes comme théorie générale pour l’organisation des connaissances”. Les Cahiers du Numérique 13, no. 1:25-48.

Broughton, Vanda. 2023. “Facet Analysis: The Evolution of an Idea”. Cataloging and Classification Quarterly 61, nos. 5/6: 411-38. https://doi.org/10.1080/01639374.2023.2196291.

Broughton, Vanda and Heather Lane. 2000. “Classification Schemes Revisited: Applications to Web Indexing and Searching”. In Internet Searching and Indexing: The Subject Approach, edited by A. R. Thomas and J. R. Shearer. New York: Haworth, 143-55. Also published in Journal Of Internet Cataloguing 2000 2, nos. 3/4: 143-55.

Broughton, Vanda and Heather Lane. 2004. “The Bliss Bibliographic Classification in Action: Moving From a Special to a Universal Faceted Classification via a Digital Platform”. In Knowledge Organization and the Global Information Society. Proceedings of the 8th International ISKO Conference, 13-16 July 2004, London, UK, edited by I. C. McIlwaine. Würzburg: Ergon, 73-8.

Broughton, Vanda and Aida Slavic. 2007. “Building a Faceted Classification for the Humanities: Principles and Procedure”. Journal of Documentation 35, no. 5: 727-54.

Brown, James Duff. 1906. Subject Classification, with Tables, Indexes, etc., for the Sub-Division of Subjects. London: Library Supply.

Bury, Susan. 1980. “Comparison of Classification Schemes for Libraries”. Library Science with a Slant to Documentation 17, no. 3: 73-82.

Bury, Susan. 1982. “Comments on Class H Anthropology”. Bliss Classification Bulletin 7, no. 3: 14-18. https://blissclassification.org.uk/B24(VII-3)-1982.pdf.

Bury, Susan. 1984. “National Library of Medicine Classification Compared with Bliss Class H”. Health Libraries Review 1, no. 4: 179-90.

Campbell, D. J. 1957. “Glossary to Dr. Ranganathan’s Paper”. In Proceedings of the International Study Conference on Classification for Information Retrieval, Held at Beatrice Webb House, Dorking, England, 13-17 May 1957. London: ASLIB, 13-14.

Campbell, D. J. 1980. “A Summary of Reviews of, and Published Comments on, the 2nd Edition of the Bibliographic Classification.” Bliss Classification Bulletin 5, no. 1: 18-30. https://blissclassification.org.uk/B14(V-1)-1971.pdf.

Classification Research Group. 1955. “The Need for a Faceted Classification as the Basis for All Methods of Information Retrieval”. Library Association Record 57, no. 7: 262-8.

Classification Research Group. 1964a. Some Problems of a General Classification Scheme. Report of a Conference Held in London June 1963. London: Library Association.

Classification Research Group. 1964b. “Bulletin No. 8”. Journal of Documentation 20, no. 3: 158-9.

“Classification Research Group/Bliss Classification Association”. 1991. Library Association Record 93, no. 8: 534.

Coates, Eric J. 1978. “Classification in Information Retrieval: the 20 Years Following Dorking”. Journal of Documentation 34, no. 4: 288-98.

Cowley, Debbie. 1987. “The Use of Bliss at CCETSW”. Bliss Classification Bulletin 29: 8-10. https://blissclassification.org.uk/B29-1987.pdf.

Curwen, Anthony G. 1978. “Revision of Classification Schemes: Policies and Practices”. Journal of Librarianship 10, no. 1: 19-38.

Curwen, Anthony G. 1988. “Bliss User Survey 1987-88”. Bliss Classification Bulletin 30: 11-17. https://blissclassification.org.uk/B30-1988.pdf.

Curwen, Anthony G. 1998. “Editorial”. Bliss Classification Bulletin 40: 1-2. https://blissclassification.org.uk/B40.pdf.

Curwen, Anthony G. 2002. “Editorial”. Bliss Classification Bulletin 44:2. https://blissclassification.org.uk/B44.pdf.

Curwen, Anthony G. 2004. “In the News”. Bliss Classification Bulletin 46: 31. https://blissclassification.org.uk/B46.pdf.

Danesi, Daniele. 1981. “Il Ritorno di Bliss [The Return of Bliss]”. Associazione Italiana Biblioteche. Bollettino d’Informazioni 21, no. 1: 31-44.

Dousa, Thomas M. 2013. Julius Otto Kaiser and his Method of Systematic Indexing: an Early Indexing System in its Historical Context. PhD diss., University of Illinois. http://hdl.handle.net/2142/46755.

Dousa, Thomas M. and Fidelia Ibekwe-SanJuan. 2014. “Epistemological and Methodological Eclecticism in the Construction of Knowledge Organization Systems (KOSs); The Case of Analytico-synthetic KOSs”. In Knowledge Organization in the 21st Century: Between Historical Patterns and Future Prospects, Proceedings of the Thirteenth International ISKO Conference, May 2014, Krakow, Poland, edited by W. Babik. Würzburg: Ergon, 152-9.

Eaves, D., R. Goodram and M. Howard. 1974. “The University of Tasmania's Reclassification Programme: The First Year”. Australian Academic Research Libraries 5, no. 3: 101-12.

Evans, B. W. 1979. “Bliss Bibliographic Classification, Second Edition [Review]”. Library Association Record 81, no. 9: 441.

Evans, David. 1978. “Response to Ross Trotter”. Catalogue & Index, no. 48: 8.

Farradane, Jason E. L. 1950. “A Scientific Theory of Classification and Indexing and its Scientific Application.” Journal of Documentation 6, no. 2: 83-99.

Fasayana, J. O. 1977. “Ibadan Adopts LC.” International Library Review 9, no. 3: 279-87.

Feibleman, James K. 1954. “Theory of Integrative Levels”. British Journal for the Philosophy of Science 5, no. 17: 59-66.

Foskett, A.C. 1996. The Subject Approach to Information, 5th ed. London: Library Association.

Foskett, D. J. 1961. “Classification and Integrative Levels”. In The Sayers Memorial Volume: Essays in Librarianship in Memory of William Charles Berwick Sayers, edited by D.J. Foskett and B. I. Palmer. London: Library Association, 136-50.

Foskett, D. J. 1970. Classification for a General Index Language: A Review of Recent Research by the Classification Research Group. London: Library Association.

Foskett, D. J. 1978. “The Theory of Integrative Levels and its Relevance to the Design of Information Systems”. Aslib Proceedings 30, no. 6: 202-8.

Foskett, D. J. 1989. “Classifications and Information Languages: Thoughts on Revising a Bibliographical Classification Scheme”. International Forum on Information and Documentation 14, no. 1: 3-7.

Foskett, D. J. 1991. “Concerning General and Special Classifications”. International Classification 18, no. 2: 87-91.

Gill, Lindy. 1979. “The Tavistock Reclassification Project”. In The Case for Bliss: Modem Classification Practice and Principles in the Context of the Bibliographic Classification, by Arthur Maltby and Lindy Gill. London: Bingley, 1979, 109-20.

Giunchiglia, Fausto and Ilya Zaihrayeu. 2007. “Lightweight Ontologies”. Technical Report DIT-07-071 University of Trento Department of Information and communication technology, https://www.researchgate.net/publication/30531092_Lightweight_Ontologies.

Giunchiglia, Fausto, Biswanath Dutta, Vincenzo Maltese, and Feroz Farazi. 2012. “A Facet-Based Methodology for the Construction of a Large-Scale Geospatial Ontology”. Journal of Data Semantics 1: 57–73. https://doi.org/10.1007/s13740-012-0005-x.

Gnoli, Claudio. 2005. “BC2 Classes for Phenomena. An Application of the Theory of Integrative Levels”. Bliss Classification Bulletin 47: 17-21. https://blissclassification.org.uk/B47.pdf. Also at: https://repository.arizona.edu/handle/10150/105307.

Gnoli, Claudio. 2008. “Categories and Facets in Integrative Levels”. Axiomathes 18, no. 2: 177-92.

Gnoli, Claudio. 2011. “Vickery’s Late Ideas on Classification by Phenomena and Activities”. In Facets of Knowledge Organization: Proceedings of the ISKO UK Second Biennial Conference, 4th-5th July 2011, London, edited by Alan Gilchrist and Judi Vernau. Bingley: Emerald-Aslib, 11-24.

Gnoli Claudio. 2017. “Syntax of Facets and Sources of Foci: a Review of Alternatives”. In Faceted Classification Today: Theory, Technology and End Users: Proceedings of the International UDC Seminar, 14-15 September 2017, London, United Kingdom, edited by Aida Slavic and Claudio Gnoli. Würzburg: Ergon, 243-256.

Gnoli, Claudio, Richard P. Smiraglia and Rick Szostak. 2024. “Phenomenon-based Classification: An Annual Review of Information Science and Technology (ARIST) paper”. Journal of the Association for Information Science and Technology 75, no. 3: 324-343. Also available in ISKO Encyclopedia of Knowledge Organization, edited by Birger Hjørland and Claudio Gnoli, https://www.isko.org/cyclo/phenomenon.

Goforth, Allene M. 1980. “Henry Evelyn Bliss (1870-1955): A Biography”. MLIS diss., University of North Carolina Chapel Hill.

Grolier, Eric de. 1962. A Study of General Categories Applicable to Classification and Coding in Documentation. Paris: Unesco.

Hasleton, Angela. 2002. “Membership News”. Bliss Classification Bulletin 44: 17-19. https://blissclassification.org.uk/B44.pdf.

Hjørland, Birger. 2013. “Facet Analysis: the Logical Approach to Knowledge Organization”. Information Processing and Management 49, no. 2: 545-57.

Hjørland, Birger and Hanne Albrechtsen. 1999. “An Analysis of Some Trends in Classification Research”. Knowledge Organization 26, no. 3: 131-9.

Hjørland, Birger and Thiago Henrique Bragato Barros. 2024. “Domain Analysis Versus Facet Analysis.” Knowledge Organization 51, no. 1: 19-25.

Horsey, Chris. 1983 “The Use of Bliss at OPCS”. Bliss Classification Bulletin 7, no. 4: 10-12, 21. https://blissclassification.org.uk/B25(VII-4)-1983.pdf.

Howorth, Graham B. 1980. “The Case for Bliss [Review]”. Bliss Classification Bulletin 5, no. 1: 30-1. https://blissclassification.org.uk/B14(V-1)-1971.pdf.

Hudon, Michele. 1988. “Bliss Bibliographic Classification: Class T Economics and Management of Economic Enterprises. 2nd ed [Review]”. Canadian Journal of Information Science 13, nos. 1/2: 76-77.

Hudon, Michele. 2020. “Facet”. Knowledge Organization 47: 320-33. Also available in ISKO Encyclopedia of Knowledge Organization, edited by Birger Hjørland and Claudio Gnoli. https://www.isko.org/cyclo/facet.

Johnson, Roy. 1997. “Alcohol Concern’s Library and Information Service – a History”. Bliss Classification Bulletin 39:12-14. https://blissclassification.org.uk/B39.pdf.

Kraup, Helen. 1999. “Radzinowicz Library”. Bliss Classification Bulletin 41: 15-16.

Kaiser J. 1911. Systematic Indexing. London: Pitman.

La Barre, Kathryn. 2006. “The Use of Faceted Analytico-Synthetic Theory in the Practice of Website Construction and Design”.  PhD diss., Indiana University. Retrieved from https://netfiles.uiuc.edu/klabarre/www/LaBarre_FAST.pdf.

Lane, Heather. 1995. “News from Cambridge”. Bliss Classification Bulletin 37: 28. https://blissclassification.org.uk/B37.pdf.

Lane, Heather. 1999. “Managing Organizational Change: Perspectives on Conversion”. Bulletin of the Association of British Theological and Philosophical Libraries 6, no. 2: 5-10.

Lane, Heather. 2004. “Minutes of the AGM: item 6.4”. Bliss Classification Bulletin 46: 3-5. https://blissclassification.org.uk/B46.pdf

Lane, Heather and Sarah Butler. 1998. “Notes on the Implementation of Classes DG/DY (Earth Sciences) Draft Schedule in the Library of Sidney Sussex College, Cambridge”. Bliss Classification Bulletin 40: 18-22. https://blissclassification.org.uk/B40.pdf.

Langridge, D. W. 1973. Approach to Classification for Students of Librarianship. London: Bingley.

Langridge, D. W. 1976. Classification and Indexing in the Humanities. London: Butterworths.

Laureilhe, Marie-Thérèse. 1978. “Review of Bliss Bibliographic Classification, Second Edition”. Bulletin des Bibliothèques de France 23, no. 1: 49-50.

MacLeod, Marion. 2003. “A Blissful Voyage”. Bliss Classification Bulletin 45: 32-3. https://blissclassification.org.uk/B45.pdf.

MacLeod, Roderick A. 1984. “Reclassification at Chancellor College Library: Background and 1984 Project”. Malawi Library Association Bulletin 4, no. 1: 60-3.

Madalli, Devika P., B. Preedip Balaji and Amit Kumar Sarangi. 2015. “Faceted Ontological Representation for a Music Domain”. Knowledge Organization 42, no. 1: 8-24.

Maltby, Arthur and Lindy Gill. 1979. The Case for Bliss: Modern Classification Practice and Principles In the Context of the Bliss Bibliographic Classification. London: Bingley.

Marcella, Rita and Robert Newton. 1994. A New Manual of Classification. Aldershot: Gower.

Mazzocchi, Fulvio and Claudio Gnoli. 2010. “S.R. Ranganathan’s PMEST Categories: Analyzing Their Philosophical Background and Cognitive Function”. Information Studies 16: 133-148.

McIlwaine, Ia C. 1995. “Preparing Traditional Classifications for the Future: Universal Decimal Classification”. Cataloging and Classification Quarterly 21, no. 2: 49-58.

McIlwaine, Ia C. 1996. “New Wine in Old Bottles: Problems of Maintaining Classification Schemes”. In Knowledge Organization and Change: Proceedings of the 4th International ISKO Conference. Washington. 15-18 July 1996, edited by Rebecca Green. Advances in Knowledge Organization no. 5, Frankfurt/Main: Indeks Verlag, 122-36.

McIlwaine, Ia C. 1997. “Classification Schemes: Consultation with Users and Cooperation Between Editors”. Cataloging and Classification Quarterly 24, nos. 1/2: 81-95.

McIlwaine, Ia C. and Joan S. Mitchell. 2006. “The New Ecumenism: Exploration of a DDC/UDC View of Religion”. Extensions and Corrections to the UDC 28: 9-16. Reprint of the paper published in Knowledge Organization for a Global Learning Society: Proceedings of the Ninth International ISKO conference: 4-7 July 2006, Vienna, edited by Gerhard Budin, Christian Swertz and Konstantin Mitgutsch. Würzburg: Ergon Verlag, 323-30.

McIlwaine, Ia C. and Nancy J. Williamson. 1993. “Future Revision of UDC: Progress Report on a Feasibility Study for Restructuring”. Extensions and Corrections to the UDC 15: 11-17.

McIlwaine, Ia C. and Nancy J. Williamson. 1994. “A Feasibility Study on the Restructuring of the Universal Decimal Classification into a Full-Faceted Classification System”. In Proceedings of the Third International Society for Knowledge Organization (ISKO) Conference: Knowledge Organization and Quality Management, Copenhagen, Denmark, 20-24 June 94, edited by Hanne Albrechtsen and Susan Oernager. Frankfurt/Main: Indeks Verlag, 406-13.

McIlwaine, Ia C. and Nancy J. Williamson. 1995. “Restructuring of Class 61 – Medical Sciences”. Extensions and Corrections to the UDC 16: 11-66.

McKeown, Ann H. P. 1979. “Bliss Bibliographic Classification, Second Edition. Class I Psychology and Psychiatry [Review]”. Catalogue & Index, no. 53: 5

Mills, Jack. 1957. “The Classification of a Subject Field”. In Proceedings of the First International Conference on Classification for Retrieval. London: Aslib, 29-42.

Mills, Jack. 1966. “Foreword”. Bliss Classification Bulletin 3, no. 3: 2-3. https://blissclassification.org.uk/B09(III-3)-1966.pdf.

Mills, Jack.1969. “Preparation of a New Edition of the Bliss Classification”. Bliss Classification Bulletin 4, no. 3:7-20. https://blissclassification.org.uk/B12(IV-3)-1969.pdf.

Mills, Jack. 1970. “A New Edition of Bliss”. Catalogue and Index no. 17: 8-9.

Mills, Jack. 1972. “The Bibliographic Classification.” In Classification in the 1970s: a Discussion of Development and Prospects for the Major Schemes, edited by A. Maltby. London: Bingley: London, 25-52.

Mills, Jack. 1976a. “A Comment on the Article by R. A. Ukoh “Library Classification and Change: the Example of Bliss”. (Libri 1975 25(3): 168-73)”. Libri 26, no. 2: 156-7.

Mills, Jack. 1976b. “The New Bliss Classification”. Catalogue & Index, no. 40: 3-6.

Mills, Jack. 1983. “Attention Please for BC2!” International Classification 10, no. 1: 24-8.

Mills, Jack.1991. “BC2 as a General Classification: a Reply to Some Criticisms”. Bliss Classification Bulletin 33: 17-21. https://blissclassification.org.uk/B33-1991.pdf.

Mills, Jack.2003. “Progress of BC2 in 2002: Editor’s Report to the AGM”. Bliss Classification Bulletin 45: 8-9. https://blissclassification.org.uk/B45.pdf.

Mills, Jack. 2004. “Faceted Classification and Logical Division in Information Retrieval”. Library Trends 52, no. 3: 541-70.

Mills, Jack. 2005. “Progress of BC2 in 2004: Editor’s Report to the AGM”. Bliss Classification Bulletin 47: 8-9. https:// blissclassification.org.uk/B47.pdf.

Mills, Jack and Vanda Broughton. 1977- . Bliss Bibliographic Classification, 2nd ed. London: Butterworths.

Mitchell, Joan S. et al. (Eds.). 2012. Dewey Decimal Classification. 200 Religion Class Dublin, Ohio: OCLC.

Mitchell, Sue and Gerry Power. 1996. “Bliss Bibliographic Classification at NSPCC Library”. Bliss Classification Bulletin 38: 8-12. https:// blissclassification.org.uk/B38.pdf.

Moss, Wilfred Raymond. 1964. “Categories and Relations: Origins of Two Classification Theories”. American Documentation 15, no. 4: 296-301.

Needham, Joseph. 1937. Integrative Levels: A Revaluation of the Idea of Progress. Oxford: Clarendon Press.

Perkins, Alex. 2011. “Classifying Art with Bliss”. Art Libraries Journal 36, no. 4:18-21.

Power, Gerry. 1995. “New Bliss Class Q”. Assignation 13, no. 1: 46-7.

Prasad, A. and Devika Madalli. 2009. “Classificatory Ontologies”. In Classification at a Crossroads: Multiple Directions to Usability. Proceedings of the UDC Seminar 2009. Available in Extensions and Corrections to the UDC 31: 223-32.

Prasad, A., Fausto Giunchiglia, and Devika Madalli. 2017. “DERA: from Document-centric to Entity-centric Knowledge Modelling”. In Faceted Classification Today: Theory, Technology and End Users. Proceedings of the International UDC Seminar 2017, edited by Alida Slavic and Claudio Gnoli. Würzburg: Ergon.

Preddle, Chris. 1982. “The Use of Bliss at Dr. Bamardo's Library”. Bliss Classification Bulletin 7, no. 3: 10-12. https://blissclassification.org.uk/B24(VII-3)-1982.pdf.

Prieto-Diaz, Ruben. 2003. “A Faceted Approach to Building Ontologies”. In Proceedings of the 2003 IEEE International Conference on Information Reuse and Integration, 27 29 October 2003 Las Vegas. New York: IEEE, 458-65.

Ramsden, Michael J. 1978. “A New Life for Bliss”. Australian Academic and Research Libraries 9, no. 4: 210-4.

Ranganathan, S. R. 1960. Colon Classification. Basic Classification, 6th ed. New York: Asia Publishing House.

Ranganathan, S. R. 1967. Prolegomena to Library Classification, 3rd ed. London: Asia Publishing House.

Richardson, Ernest Cushing. 1930. Classification, Theoretical and Practical, 3rd ed. New York: H. W. Wilson.

Riggs, Fred W. 1987. “Class K Society [Review]”. International Classification 14, no. 1: 42-4.

Roberts, Norman. 1977. “Bliss Bibliographic Classification, Second Edition [Review].” Catalogue & Index, no. 46: 5-7

Royal Society of London. 1948. Royal Society Scientific Information Conference, 21 June – 2 July 1948. Report and papers submitted. London: The Royal Society.

Russell, Elizabeth. 1997. “Cambridge Postscript from a Peripatetic Classifier”. Bliss Classification Bulletin 39: 17-18. https://blissclassification.org.uk/B39.pdf.

Russell, Elizabeth and Heather Lane. 1999. “Working Towards an Archaeology Schedule at the Haddon Library”. Bliss Classification Bulletin 41:17. https://blissclassification.org.uk/B41.pdf.

Russell, Elizabeth and Julia Wilkinson. 1979. “Bliss is’t to be in King’s”. Bliss Classification Bulletin 4, no. 4: 9-12. https://blissclassification.org.uk/B21(VI-4)-1979.pdf

Ryan, Giustina and Angela Upton. 1991. “The National Institute for Social Work”. Bliss Classification Bulletin 33: 10-13. https://blissclassification.org.uk/B33-1991.pdf.

Sargent, Clare. 1990. “Classifying the Undergraduate Collection at Queens’ College, Cambridge”. Bliss Classification Bulletin 32: 10-12. https://blissclassification.org.uk/B32-1990.pdf.

Satija, Mohinder, and Jagtar Singh. 2013. “Colon Classification: A Requiem”. DESIDOC Journal of Library & Information Technology 33, no. 4: 265-76. https://doi.org/10.14429/djlit.33.4.4881.

Shearer, James R. 2004. “A Practical Exercise in Building a Thesaurus”. Cataloging and Classification Quarterly 37, nos. 3/4: 35-56.

Shera, Jesse. 1965. Libraries and the Organization of Knowledge. Hamden, CT.: Archon Books.

Shoyinka, Patricia. 1975. “Adoption of the Library of Congress Classification at the University of Ibadan: Decisions and Practices”. Nigerian Librarian 11, nos. 1/2: 65-85.

SKOS Simple Knowledge Organization System Primer. 2009. http://www.w3.org/TR/skos-primer.

Soergel, Dagobert. 1999. “The Rise of Ontologies or the Reinvention of Classification”. Journal of the American Society for Information Science 50, no. 12: 1119-1120

Spiteri, Louise F. 1995. “The Classification Research Group and the Theory of Integrative Levels”. Katharine Sharp Review 1 (Summer). http://alexia.lis.uiuc.edu/review.old/summer1995/spiteri.html.

Stoddard, Hilda. 1979. “Reclassification by Bliss”. Catalogue & Index, no. 55: 4-5.

Szeredi, Peter, Gergely Lukacsy and Tamas Benko. 2014. The Semantic Web Explained. Cambridge: University Press.

Szostak, Rick. 2007. “Interdisciplinarity and the Classification of Scholarly Documents by Phenomena, Theories, and Methods”. In Interdisciplinarity and Transdisciplinarity in the Organization of Scientific Knowledge: Actas del VIII Congreso ISKO-España: León, 18, 19 y 20 de Abril de 2007, edited by Blanca Rodríguez Bravo and M.a Luisa Alvite Díez. León: Universidad de León Secretariado de Publicaciones, 469-77.

Thomas, Alan R. 1992. “Bliss Bibliographic Classification, 2nd Edition: Principal Features and Applications”. Cataloging and Classification Quarterly 15, no. 4: 3-17.

Thomas, Alan R. 1995. “Bliss Classification Update”. Cataloging and Classification Quarterly 19, nos. 3/4: 105-17.

Thomas, Alan R. 1998. “Bibliographical Classification: The Ideas and Achievements of Henry Evelyn. Bliss”. Cataloging and Classification Quarterly 25, no. 1: 51-102.

Trotter, R. 1977. “Letter.” Catalogue & Index, no. 47: 7-8.

Ukoh, R. A. 1975. “Library Classification and Change: the Example of Bliss”. Libri 25, no. 3: 168-73.

Vernon, Kenneth and Valerie Lang. 1970. London Classification of Business Studies. London: London Business School.

Vickery, Brian Campbell. 1959. Classification and Indexing in Science, 2nd ed. London: Butterworths.

Vickery, Brian Campbell. 1960. Faceted Classification: a Guide to the Construction and Use of Special Schemes. London: Aslib.

Viti, Elisabetta. 2010. “Class T “Economics” in Bliss Bibliografic Classification”. JLIS.it: Italian Journal of Library and Information Science 1, no. 2: 331-56. http://archive.org/details/jlis_it-34/page/352/mode/2up.

Watson, Rhona. 1997. “Quincentenary Library, Jesus College, Cambridge”. Bliss Classification Bulletin 39: 14-16. https://blissclassification.org.uk/B39.pdf.

Weaver, Margaret and Michael Stanning. 2007. “Reclassification Project at St. Martin's College: A Case Study”. Library Review 56, no. 1: 61-72.

Wood, Neville. 1983. “Class H of the Revised Bliss Bibliographic Classification: a Review Article”. Journal of Librarianship 15, no. 1: 66-77.

Wright, Kathleen. 1995. “Using Q in MPS”. Bliss Classification Bulletin 37: 24-7. https://blissclassification.org.uk/B37.pdf.