The Work in Question

Wayne de Fremery and Michael Buckland

Wayne de Fremery (wayne.defremery@dominican.edu) is Professor of Information Science and Entrepreneurship at Dominican University of California where he directs the Françoise O. Lepage Center for Global Innovation. Michael Buckland (buckland@berkeley.edu), is Emeritus Professor, School of Information, University of California, Berkeley.

The International Conference on Cataloging Principles (Paris, 1961) led to wide acceptance of Seymour Lubetzky’s distinction between books and works, where books denoted particular physical objects and works concerned conceptual abstractions associated with the creative labor of particular authors. Lubetzky’s formulation of works is included in many of the world’s cataloging frameworks, including the Functional Requirements for Bibliographic Records (FRBR). Several conceptual and practical difficulties arise from the widespread adoption of Lubetzkian works in practices associated with knowledge organization. However, FRBR and other knowledge organization frameworks that utilize works as central entities could be made more useable and useful if work, as an organizing principle, were de-emphasized and seen as one among many concepts used for aggregating sets and supersets of objects according to how likely they are to be useful to users of knowledge organization tools like catalogs.

The International Conference on Cataloguing Principles held in Paris in 1961 was influential in advancing standardization in terminology and rules for descriptive cataloging. The word work has played a central but problematic role as part of a now more standard global cataloging terminology and in the design of bibliographic systems. Delegates, following the usual practice at the time, used the term work as a count noun to denote any individual physical instance of a book. This was consistent with the definition of work in the vocabulary prepared for the conference: “Any expression of thought in language or symbols or other medium for record or communication.”1 However, one US delegate, Seymour Lubetzky, urged a different and more limited meaning, using work to denote a literary creation which might have multiple expressions and physical versions. Lubetzky asserted this usage in a working paper he prepared for the conference entitled “The Function of the Main Entry in the Alphabetical Catalogue—One Approach.”2

The Draft Statement of Principles prepared for the conference followed previous custom in stating that a library catalog had two objectives. The first objective was to be an efficient instrument for ascertaining whether the library contains a copy of a particular book. The second objective was to ascertain “which works by a particular author and which editions of a particular work are in the library.”3 For this second objective the definition of work clearly mattered. In a paper prepared for the conference, Lubetzky explained his position that books and other library materials were not themselves works but were representations of an author’s creative achievement, which he called a work; that these representations could take different forms and use differing names and titles; and that, therefore, the library catalog should not only list each particular book but also “identify the author and the work represented by the item or publication and to relate the various works of the author and the various editions and translations of the work.”4 In other words, the catalog should “enable a user of the catalogue ... to determine with certainty whether or not the library has a particular work, under whatever name or title, and to select the edition or translation which will best serve his purpose.”5 Lubetzky’s position was also evident in a discussion of draft principle 9.12 concerning publications by corporate authors when he argued, without success, that the phrase “content of the work” should be changed to “the work represented by the publication.”6

Lubetzky formulated his usage by contrasting book and work, where book denoted a particular physical object characterized by a text and work meant a literary creative effort made manifest in one or more books. (Any literary creation not made manifest was not of concern in this context.) Lubetzky and his UCLA colleague, Robert M. Hayes, used their considerable prestige to advance this view. Others, notably Richard Smiraglia, also adopted this view.7 Eventually, Lubetzky’s notion of a work became accepted as a foundational component for library cataloging through the Functional Requirements for Bibliographic Records (FRBR) model advanced by the International Federation of Library Associations.8

The FRBR model is composed of three groups: Group 1 is concerned with books and works in Lubetzky’s sense; Group 2 with authors and others responsible for Group 1 entities; and Group 3 is concerned with topics (concepts, objects, events, places). Here, we are primarily concerned with Group 1 and Group 2 since, as we describe, Group 2 formulates Group 1. According to FRBR, an author’s work is realized through one or more media forms (“expressions”); an expression is embodied in one or more manifestations (typically an edition); and a manifestation is exemplified by one or more instances (“items”) as shown in figure 1.

The Work in Question: Some Conceptual and Practical Difficulties

In the FRBR framework, a work is defined as the outcome of a creative effort. This focus on outcomes and creative effort creates several conceptual and practical problems. We review some of these problems and propose a path forward.

Unimportant, Unknown, Contested, and Difficult to Conceptualize Authors

Even though it remains central to cataloging objectives as they have been institutionalized by the adoption of frameworks such as FRBR, authorship is not always of interest to information seekers or users of catalogs.

Even when authorship is of interest to information seekers, catalogers—while expert at describing the material features of documents—are only infrequently qualified to resolve questions that may arise about authorship when authorship is contested or unknown.9 Frequently, of course, authorship is unknown or contested by domain experts, as well as by authors themselves, as in copyright disputes or cases of plagiarism. The FRBR model would have catalogers be the arbiters of any such disputes, at least as far as how a bibliographic record is described.

In addition to being practically fraught in many cases, the attribution of authorship can also be understood to be conceptually complex, as bibliographers, literary scholars, and philosophers have long understood. Without rehearsing what Roland Barthes meant when he announced the death of “the author,”10 how Michel Foucault conceived of his “author function,”11 or the complex role played by authors in what Jerome McGann call the “socialization of texts,”12 it is easy to acknowledge that authorship as a concept is complex and that this complexity is not taken into account by FRBR despite its centrality to the formulation of works.

The Tenuous Categorical Boundaries of Works and Their Practical Implications

The categorical boundaries of FRBR works must be defined tenuously because, as a concept, authorship can be debated and differently understood and, as a practical, socially accepted and verifiable attribution, authorship is not always possible to record, as its designers acknowledge.13 The Expedition of Humphrey Clinker, a novel by Tobias Smollett, helps to illuminate some of the practical implications of a FRBR work’s tenuous categorical boundaries. As O’Neill and Vizine-Goetz help us to understand,14 catalogs will tell us that there are 110 different English editions of Smollett’s work but what counts as the work is less clear. Facsimiles and reprints are included. So, in general, are different editions unless, perhaps, they are so heavily annotated or illustrated as to have a changed character. For Lubetzky and for FRBR a work is by definition the outcome of creative labor. Practical difficulties arise immediately when one tries to distinguish where one work ends and another begins, even if we overlook the fact that most books described by library catalogs are the result of creative efforts that include scribes, publishers, and other copyists in addition to authors.

Translations and abridged editions of Humphrey Clinker will frequently be included in catalogs as a work by Smollett, but summaries will not be. Revisions by the author are included, but not adaptations by others. Patrick Wilson considered a translation to be a new and different work,15 even though the FRBR rules and Tillett’s “Family of works” table suggest a literal translation should be considered part of the same work while a free translation should not be, even if a free translation might well express the author’s intended meaning better.16 A Turkish translation of a German translation of Humphrey Clinker would leave any earnest individual cataloger looking for a good rule to follow and individuals in different cataloging environments to make different choices. Similar situations arise with adaptations and adaptations of adaptations. In theory a plagiarized text, as mentioned earlier, should be, by definition, part of a FRBR work for the text plagiarized. But which work and whose could only be discerned if two different works were found to be the same. And, if two works were discovered to be, in fact, one, catalogers would be confronted by the riddle of how to categorize and make discoverable the “plagiarized work,” an inherent oxymoron in frameworks that utilize the Lubetzkian work. Our point is not that plagiarism is a crucial problem for catalogers but rather that the tenuousness of the categorical boundaries of the Lubetzkian work make it difficult to apply the rules of frameworks such as FRBR consistently.

Tillett’s “Family of works” table expresses pragmatic judgements about when a book should be considered part of a work. Her judgements are sensible and defensible, but they are nevertheless arbitrary and open to unavoidable difficulties of interpretation: How free does a translation need to be to constitute a different work? And how should freeness be assessed? Opinions can differ concerning the significance of any added annotation and so whether an annotated edition is part of the same work or should be treated as a new and different work. Similarly, popular textbooks commonly transition through successive revised editions with responsibility gradually moving from one author to another. When does it become a new and different work? How should one decide? Literary scholars, musicologists, and art historians will debate what counts as a distinct intellectual or artistic creation. One scholar will establish authorship of a literary work, only to have another raise questions, and both are liable to change their opinions over time. Acknowledging that questions such as these are frequently difficult questions to answer, even for domain experts, we may doubt that now standard terms appearing in frameworks such as FRBR, especially the term work, “free us from the baggage of past terms that were ambiguous.”17 Holden provides a useful discussion of applying Lubetzky’s work concept to music, serials, and aggregate works.18

Singular, Mutually Exclusive, and without Context

The difficulties created by conceptually tenuous notions of authorship fundamental to definitions of works in FRBR and similar frameworks are compounded by the apparently unquestioned assumption that the creations of authors are singularly novel, can be easily separated from their contexts, and contain no portion of other works. As we have been emphasizing, common sense—as well as common understanding in literary studies19—suggests that few if any literary texts arrive ab ovo from the minds of creators. But even if we assume, as a common practical matter, authorship can be attributed, such attributions can serve to dislodge whatever is taken as a work from its contexts. According to FRBR, for example, an individual pamphlet would be a work if it were the result of an identifiable creative effort. But if it were one part of an ongoing debate, for example, information about the context of the larger debate, which would give meaning to the pamphlet, would not be reflected. Conversely, it is possible for a single text to present more than a single work by a single author. Indeed, as frequently happens, hierarchical descriptive frameworks that have Lubetzkian works as the largest superset leave catalogers struggling to describe a single bibliographic item that includes portions of two or more works since that they are not well-qualified to disambiguate the works. Using works as a descriptive category assumes that what the descriptions are helping to organize and make findable are singular and mutually exclusive, and that the usefulness of the descriptions themselves would not be enhanced by additional contextual information, when common sense and every day experience suggests otherwise.

FRBR Inside-out, Upside-down, and Backward

In addition to practical difficulties in the interpretation of individual cases, there are other conceptual issues. An examination of these issues presents opportunities to reconsider how catalogs might function better as epistemological tools.

Sets and the Work as an Epistemological Tool

So, what is a FRBR work? How does it exist? Although they do not make any such assertion or formulate it as such, for Lubetzky and in FRBR, a work is an epistemological tool. It is an abstract concept used as an organizing device for defining arbitrary sets of objects and their relationships with one another: sets of one or more expressions; sets of one or more manifestations; sets of one or more items. In what way, if any, is it anything more than that? There is a tradition in library and information science of treating abstract tools as if they had substance, a tradition denounced by Frohmann.20 To point out that the work is an epistemological tool and not any particular physical object (or group of them) is not to deny the force that the abstraction has as an epistemological and organizational tool. It is obvious that the idea of a work has been widely adopted and is useful. The question becomes whether the work is the best epistemological tool for achieving the objectives of the library catalog, particularly now that library catalogs are being asked to be more than efficient instruments for ascertaining whether a library contains a copy of a particular book and which versions of which books by particular authors might be available.

The work as an epistemological tool is used as an abstract organizing device for physical items. Particular books are grouped by being ascribed to a shared creative origin. They are contextualized and organized by socially accepted beliefs and practices associated with authorial creation. These shared beliefs and practices among catalogers enable the creation of descriptions that can conveniently and usefully organize sets of objects. These sets serve the historic objectives of library catalogues by enabling answers to traditional questions concerning what is in a collection. It is important to notice that, as we have shown, the power of the work as an epistemological tool for creating and organizing sets of objects is not, in fact, dependent upon any relationship with verifiable factual historical events associated with the creation of physical objects like books. Although creative effort formulates the concept of a work, the concept cannot encompass the historical realities it is used to index. The real power of Lubetzky’s work is not drawn from any self-evident relationship between physical texts and how they might have been created, but rather from its power to contextualize, and thereby organize, a set of items by formulating them in relation to an arbitrarily defined notion of creative effort adopted as a social norm by catalogers. Indeed, as the creators of FRBR themselves recognize, “the concept of what constitutes a work and where the line of demarcation lies between one work and another may in fact be viewed differently from one culture to another.”21 The concept of what constitutes a work, as well as how works might be demarcated, are culturally formulated. Recognizing that the work gains its power to organize from these socially sustained conceptual relationships enforced by catalogers and not necessarily any historically-grounded truth reveals how the work performs as an epistemological tool. So we are also presented the opportunity to reconsider frameworks such as FRBR from several perspectives. Given that how works are formulated conceptually and distinguished from one another is culturally formulated we can consider methods for documenting how catalogers in their cultural and historical contexts have formulated works rather than assuming, as a matter of practice, that a work is a work no matter who catalogs it and in what sociohistorical context. While the creators of widely used models such as FRBR acknowledge that cultural perspectives may affect how works are formulated in catalogs, the models themselves have no mechanism for capturing how. We can consider how other abstractions, if they were to be socially adopted as a standard, might be used for contextualizing items and formulating sets that productively help readers looking to make use of a textual resource.22 We can similarly reconsider how FRBRs hierarchal organization might be productively reorganized.

FRBR Upside-down

The FRBR diagram could as easily be inverted or read bottom-up as a hierarchical, set-theoretic, tree structure in which one or more items constitute a set named manifestation; one or more manifestations constitute a set named expression; and one or more expressions constitute a superset that is named a work. Thus manifestation, expression, and work are progressively larger supersets of items. Viewed this way, bottom up, a work is defined as and by whatever set of items form the starting point. It need no longer be defined by an attributed creative origin. This does not remove the difficulty of deciding what to include, but it does avoid the difficulties created by assuming that a work refers to anything other than an abstraction formulated differently by people working in specific sociohistorical contexts. It is simpler and for that reason preferable according to the principle of Occam’s razor by which a simpler explanation is to be preferred to a more complex one. In a manner similar to how textual bibliographers are guided toward consensus beliefs about certain works by cataloging the differences among copies of a work, users of a catalog could be guided by specific observations about specific objects organized into increasingly abstract conceptual groups rather than the other way around. Instead of a work flowing down through expressions, manifestations, and items, the reverse would be any set of related items that can be aggregated by manifestation, by expression, and, finally and abstractly as a single superset of all the items included. In this way, a bibliographical framework which turned FRBR upside-down would be usable for organizing any affinity group of items, for any set of interest to a reader. The significant difference would be that the cataloging effort would be directed toward readers rather than sustaining an abstraction formulated by the cataloging community. It would be more flexible and so more powerful.

Usefully Similar

FRBR is useful because it offers aggregation at the manifestation and expression levels, but, as we have indicated, FRBR’s Group 1 structure could presumably be applied to any set of documents. de Fremery and Buckland consider the usefulness of situationally “usefully similar” gatherings.23 With this approach the FRBR structure would help to coordinate cataloging practices by creating usefully similar groupings of documents for bibliographical purposes. FRBR Group 1 items are similar because they are the product of the same creative effort, but, while authorship is one important attribute for organizing usefully similar documents, bibliographical description can (and does) easily account for the many other ways that documents could be considered usefully similar to one another in ways not reducible to the traditional metadata, notably author, topic, title, genre or format. “Usefully similar” could include writings from a particular point of view, with a distinctive style, using analogous symbolism, or a similar plot or methodology. See, for example, Jarmo Saarti’s description of the variety of ways that different literary texts have been formulated as similar to one another, which include traditional metadata categories but also categories such as “recreational” and “serious fiction.”24 Netflix famously organizes its media content into quirky categories of usefully similar movies and television programs, where similarity is formulated by categories such as “action with a side of romance,” “lavish reality lifestyles,” “short-ass movies”25 and utility is understood as “making users’ viewing experience more enjoyable”26 and, of course, what serves Netflix’s bottom line. Although less entertaining, the Library of Congress’s Genre/Form Terms (LCGFT) manual provides a similar means of articulating useful similarities among objects, as do Library of Congress subject headings (Group 3 in the FRBR framework). Our point is that any aggregating principle could complement Lubetzsky’s work as an epistemological tool, and many have. “Usefully similar” provides an expansive basis for considering relationships among books and other media, as well as means of organizing them.

In brief, while the structure of FRBR categories is useful the categories themselves need not be formulated in relation to the concept work. A more reader-oriented library service could be focused on how usefully similar items might be found and be found to be useful by users of catalogs. This differs from a more exclusive focus on authorial creativity but, importantly, the models are not mutually exclusive. The distinction to be found is that a focus on what is usefully similar attempts to empathize with users and what they might consider usefully similar to a document they seek, this instead of requiring users to navigate a genealogical hierarchy based upon abstract assumptions about creative origins to find what they need. Where Netflix organizes its materials with the explicit aim of making its users’ experience more enjoyable to better serve its business objectives, we might redouble our efforts to organize our catalogs so that users’ reading experiences are more enjoyable to serve our aims of making desired information discoverable. See figure 2.

Inside-out and Backward

Formulated by traditional beliefs about literary production (Lubetzky majored in German and French) and in support of traditional cataloging objectives, FRBR and similar frameworks are organized to emphasize authors.27 But this emphasis is backward if one wishes to have the catalog focused on serving readers. A catalog designed for readers would try to start with how readers might find documents usefully similar to what they have in mind or in hand. Suppose that instead of organizing a collection to support the discovery of Dashiell Hammett and his work The Maltese Falcon, a reader could be led toward the resources usefully similar to what they have in mind, perhaps a resource about falcons, or news from Malta. In this case, Dashiell Hammett is not irrelevant because a reader may indeed have The Maltese Falcon in mind because they just finished Hammett’s book The Glass Key. The distinction is that in one case the catalog is organized to enable the discovery of a literary work while the other is organized to enable the discovery of something usefully similar to what is of interest to a reader. For example, novels that feature the same characters but are written by different authors or books owned by a historically important figure. For a user looking to be briefly distracted, “short-ass movies” could be put into relation with “short-ass fiction,” for which we have a host of less colloquial terms (Micro fiction, Microfiction, Short-short stories, Sudden fiction, Very short fiction) in the LCGFT manual under “Flash fiction.”28 Citing Bartlett and Hughes (2011) and Vernitski (2007), Rafferty (2015) describes a variety of ways that literary texts have been organized by categories of similarity formulated by concepts associated with intertextuality, where intertextuality after Genette (1997) is meant to mean “a relationship of co-presence between two text or among several text” and “the actual presence of one text within another.”29 By putting the notion of the work in question, it becomes possible to reconsider the categorical structures that frameworks such as FRBR enforce and the kinds of discovery they facilitate. We can ask if we might better support readers’ ability to make the best use of any set of media objects by composing catalogs to reveal objects that are similar to what they have in mind rather than authors they may not care to know.

Toward More Useful, Reader-Oriented Catalogs

The Library Reference Model and BIBFRAME

As part of efforts to create more useful, reader-oriented catalogs, in 2017 FRBR was consolidated and harmonized with related models, notably the Functional Requirements for Authority Data (FRAD) and the Functional Requirements for Subject Authority Data (FRSAD), to form the Library Reference Model (LRM). The LRM incorporates the FRBR model beneath a new top-level entity named “res,” Latin for thing, which can be any “entity in the universe of discourse.”30 As the authors of the LRM framework describe, “in terms of general approach and methodology, the modeling processes that resulted in the IFLA LRM model adopted the approach taken in the original FRBR study,”31 namely “an entity analysis technique that begins by isolating the entities that are the key objects of interest to users of bibliographic records.”32 One irony of the LRM’s formulation is that despite its authors’ stated emphasis on users of bibliographic records, the model, by fully integrating FRBR and its methodologies, retains FRBR’s emphasis on authors and the presumption that, except when concerned with the most abstract “things,” users of bibliographic records are wishing to find, identify, select, obtain, and explore resources33 as they might be organized by the concept of works. As incorporated into LRM, FRBR brings its power as an epistemological tool for organizing and creating sets of objects, but also its conceptual and practical weaknesses, primary among them the assumption that creators of bibliographic resources are ordinarily and primarily “key objects of interest to users of bibliographical records”34 as the creators of the LRM, borrowing from FRBR, contend.

The Library of Congress’ Bibliographic Framework (BIBFRAME) data model diverges fundamentally from LRM and FRBR because, although the top-level entity is named work, it is understood as a “conceptual essence of a cataloging resource”35 including “authors, languages, and what it is about (subjects).”36 This is distinct from the work as the result of creative effort. BIBFRAME usefully relaxes the commitment to an idealized “creative effort” as an epistemological formulation for describing and organizing bibliographical objects that, contrary to the assumptions of traditional formulations of cataloging objectives, may or may not reside in a library collection. Indeed, BIBFRAME was designed to “integrate with and engage the wider information community while also serving the very specific needs of its maintenance community—libraries and similar memory organizations.”37 It does so without jettisoning useful epistemological tools for organizing objects by networking descriptions in such a way that, in theory, any particular attribute of one of its classes (works, instances, items) can be shown in relation to any other. In other words, a catalog formulated according to BIBFRAME enables a user to find, identify, select, obtain, and explore resources in a bigger, but less well-defined bibliographic universe, according to the useful and powerful logic of networked associations.

While powerful, a weakness of the BIBFRAME model is that the framework is formulated to describe relationships between resources rather than how any particular resource is likely to be usefully similar to a resource that a user would wish to find, or, having performed a search, come to learn that they want. While it can powerfully present a variety of relationships between resources, as well as organize resources according to such relationships, the strength of networked relations among objects described by BIBFRAME as they might be measured by various network centralities become a surrogate for likely utility for a user.

Bibliographical Control

BIBFRAME, LRM, FRBR and other frameworks enable and engender different kinds of bibliographical control. In his essay on bibliographic control called Two Kinds of Power, Patrick Wilson distinguishes two interdependent kinds of bibliographical control: exploitative control, the ability to make the best use of a body of writings for any particular end, and descriptive control, “an ability to line up a population of writings in any arbitrary order, to make the population march to one’s command.”38 Simplifying, exploitative control is what is desired by a user, the ability to use the best bibliographical resource while pursuing some end. The ability to “exploit” the best resources is facilitated by descriptive control, i.e., descriptive efforts that enable “a population of writings” to be organized and reorganized. In theory and in practice, the ability to identify and make use of appropriate bibliographical resources for particular ends while drifting through the expanse of what Wilson describes as the bibliographical universe depends on descriptions of what can be found in the bibliographical universe.

The best use of a body of writings implies judicious selection using whatever criteria would make the selected set of references march on command and be best for the reader’s purpose. Authorship, as we have described, can be, but is not necessarily, helpful to users of a catalog when determining what might be the best textual means for the ends that they purse. Authorship is helpful not necessarily because it describes any verifiable historical reality but because it provides a means of lining up “writings” in an arbitrary order, which is to say that it begins to provide a form of descriptive control that can be exploited. The various kinds of works in question here describe function similarly.

Viewed in retrospect, FRBR, LRM, BIBFRAME, and related frameworks represent the latest evolutionary steps building on the staples used by Gesner, Schrettinger, Panizzi, Dewey, and so many others: author, title, topic, genre, and format. The historic approach has been to standardize, to generalize, and to try to be reader-friendly. But this approach can never fully satisfy readers because readers want, in Wilson’s words, the ability to line up a population of writings in any arbitrary order. Readers’ interests are not limited to or defined by author, title, topic, genre, or format as these have been formulated by catalogers but by a far wider variety of attributes. A reader who can describe or identify a book that they desire for any reason (its style, its points of view, its historical associations, its high-quality laid paper, the stitch of its sewn binding, etc.) can be expected to want other similar writings. So a very different approach is needed. Not only has technology been transformed but also handcrafted bibliographic descriptions are now richly augmented by access to full-text, paratext (blurbs, reviews, publicity), related writings, and more. The options have become more extensive and more flexible. Statistical analyses and language models of various size, along with descriptive categories of all kinds from industry and academia already enable recommender services to line up media objects to march to various commands in ways that were not previously feasible and that far exceed the power of bibliographic models still firmly anchored by the abstraction work and associated concepts of authorship. A different approach rooted directly in readers’ interests deserves attention. Changes may be difficult to accommodate and the ideal never perfectly attained, but the ability of the structure of the FRBR Group 1 model to manage populations of writings could be very useful if and only if it ceases to be limited to Lubetzky’s sense of a work.

Summary and Ways Forward

Traditional western cataloging practice is to arrange edition-level entries by author and then by title. However, a text may exist in dozens, even hundreds of different editions. Lubetzky proposed the aggregation of all editions for the same creative effort, for which he used the term work even though work also had (and still has) other meanings. His proposal was implemented in Group 1 of the Functional Requirements for Bibliographic Records (FRBR) developed by the International Federation of Library Associations. FRBR went further, specifying four levels of aggregation: work, expression, manifestation, item. The FRBR work, now incorporated into newer frameworks such as LRM, remains problematic. It provides a form of bibliographical control but one formulated by traditional beliefs and forms of descriptive practice that only sometimes enable users of catalogs to make writings line up and march according to their commands. Frameworks such as BIBFRAME productively loosen the definition of “work” so that a broader set of objects can be organized and described with more precision by putting descriptions into networked relationships. Users of systems that make use of BIBFRAME, at least in theory, can have objects in the bibliographical universe march to their command according to the rules of networked descriptions. While a powerful form of control, networked relations among objects become a surrogate for likely user utility, which is not the same as something usefully similar to the best textual means for a user’s particular end. While acknowledging the power of FRBR and other knowledge organization frameworks that utilize works as central entities, we propose that these frameworks could be made more useable and useful if work were supplemented by conceptual entities that organize and formulate sets and supersets of objects according to how likely they are to be usefully similar to objects of interest to users of knowledge organization tools like catalogs.

One potentially useful way forward toward a more user-oriented descriptive framework would be to allow users to know and make use of information about the people creating the catalogs and the epistemological formulations used to organize their searches. As we have noted, none of the available frameworks have a place for describing catalogers and how they have done their cataloging in distinct places and sociocultural contexts. Rather than assuming that users of catalogs should adopt a categorical structure formulated by librarians and implemented by catalogers within broad parameters but differently according to the circumstances of their descriptive practice and circumstance, information about catalogers and their circumstances, as well as the categorical formulations with which they work, can be made explicit. It can be formulated as information that would allow users to understand if the category of what they desire is usefully similar to categories of things librarians have formulated and often assume to be universally useful as epistemological tools. In short, one way forward, which can be tested through a variety of empirical means, would be to let users put the work and other epistemological assertions of catalogs in question by making how they have been formulated part of the information they can use to gain bibliographic control.

References and Notes

1. International Federation of Library Associations, Report/International Conference on Cataloguing Principles, Paris 9th–18th October 1961, Arthur Hugh Chaplin and Dorothy Anderson, eds. (London: Organizing Committee of the International Federation of Library Associations, 1963), 112.
2. IFLA, International Conference on Cataloguing Principles, 139–43.
3. IFLA, International Conference on Cataloguing Principles, 25.
4. IFLA, International Conference on Cataloguing Principles, 139.
5. IFLA, International Conference on Cataloguing Principles, 139.
6. IFLA, International Conference on Cataloguing Principles, 47.
7. See Richard P. Smiraglia, The Nature of “A Work”: Implications for the Organization of Knowledge (Lanham, Maryland and London: Scarecrow Press, 2001).
8. IFLA Study Group on the Functional Requirements for Bibliographic Records, Functional Requirements for Bibliographic Records: Final Report (International Federation of Library Associations and Institutions, 1997/2009), https://www.ifla.org/publications/functional-requirements-for-bibliographic-records. Several authors provide historical context for FRBRs widespread adoption, as well as useful critiques. Smiraglia and Holden, for example, both provide a clear historical account of the notion of a work as advanced by Lubetzky and adopted in the FRBR model. Svenonius provides conceptual background. Maxwell and Tillett provide introductions to FRBR. Coyle offers an insightful critique. See the following: Richard Smiraglia, The Nature of “A Work”: Implications for the Organization of Knowledge (Lanham, Maryland and London: Scarecrow Press, 2001); Richard Smiraglia, “Work,” in Knowledge Organization 46, no. 4 (2019): 308–19, which is also available in Encyclopedia of Knowledge Organization, edited by Birger Hjørland and Claudio Gnoli, International Society for Knowledge Organization, http://www.isko.org/cyclo/work; Chris Holden, “The Bibliographic Work: History, Theory, and Practice,” Cataloging & Classification Quarterly 59, no. 2–3 (2021): 77–96, also in Gretchen L. Hoffman and Karen Snow, eds., Cataloging and Classification: Back to Basics (London: Routledge, 2021); Elaine Svenonius, The Intellectual Foundation of Information Organization (Cambridge and London: MIT Press, 2000); Robert Maxwell, FRBR: A Guide for the Perplexed (Chicago: American Library Association, 2008); Barbara Tillett, What is FRBR? A Conceptual Model for the Bibliographic Universe (Washington, DC: Library of Congress, 2004), https://www.loc.gov/cds/downloads/FRBR.PDF; Karen Coyle, FRBR, Before and After: A Look at Our Bibliographic Models (Chicago: American Library Association, 2015).
9. Arnold S. Wajenberg, “A Cataloger’s View of Authorship,” in Elaine Svenonius, ed., Conceptual Foundations of Descriptive Cataloging (San Diego: Academic Press, 1989), 21–27.
10. See Roland Barthes, “The Death of the Author,” trans., Stephen Heath, in Image, Music, Text, 142–48 (London: Fontana, 1977).
11. See Michel Foucault, “What is an Author,” trans., Donald Bouchard and Sherry Simon, in The Norton Anthology of Theory and Criticism, Vincent B Leitch, ed., 3rd ed. (New York and London: W.W. Norton & Company, 2018), 1394–1408.
12. See Jerome J. McGann, “The Socialization of Texts,” in The Textual Condition (Princeton: Princeton University Press, 1991), 69–87.
13. IFLA Study Group on the Functional Requirements for Bibliographic Records, Functional Requirements for Bibliographic Records: Final Report, 17.
14. Edward T. O’Neill and Diane Vizine-Goetz, “Bibliographic Relationships: Implications for the Function of the Catalog,” in Elaine Svenonius, ed., Conceptual Foundations of Descriptive Cataloging (San Diego: Academic Press, 1989) 167–79. In particular, see pages 170–71.
15. Patrick Wilson, “The Second Objective,” in Elaine Svenonius, ed., Conceptual Foundations of Descriptive Cataloging (San Diego: Academic Press, 1989), 10.
16. Barbara Tillett, “Bibliographic Relationships,” in Carol A. Bean and Rebecca Green, eds., Relationships in the Organization of Knowledge (Dordrecht: Kluwer Academic Publishers, 2001), 23. Table reprinted in Barbara Tillett, What is FRBR?, 4.
17. Barbara Tillett, “The FRBR Model (Functional Requirements for Bibliographic Records)” (Presentation at the ALCTS Institute on Metadata and AACR2, San Jose, CA, United States, April 4–5, 2003), https://www.loc.gov/catdir/cpso/frbreng.pdf.
18. Holden, “The Bibliographic Work.”
19. See also the widely cited work of Julia Kristeva on intertextuality.
20. Bernd Frohmann, Deflating Information: From Science Studies to Documentation (Toronto: University of Toronto Press, 2004).
21. IFLA Study Group on the Functional Requirements for Bibliographic Records, Functional Requirements for Bibliographic Records: Final Report, 17.
22. Wayne de Fremery and Michael Buckland, “Context, Relevance, and Labor,” Journal of the Association for Information Science and Technology 73, no. 9 (2022): 1268–78.
23. Wayne de Fremery and Michael Buckland, “Copy Theory,” Journal of the Association for Information Science and Technology 73, no. 3 (2022): 407–18.
24. Jarmo Saarti, “Fictional Literature Classification and Indexing,” in Encyclopedia of Knowledge Organization, edited by Birger Hjørland and Claudio Gnoli, International Society for Knowledge Organization, March 26, 2019, http://www.isko.org/cyclo/fictional.
25. Jessica Derschowitz, “The Story Behind Netflix’s Secret Category Codes: From Short-Ass Movies to Highbrow Horror, There’s a Shortcut to Finding Exactly What You Want,” Tudum, March 18, 2023, accessed July 14, 2023, https://www.netflix.com/tudum/articles/netflix-secret-codes-guide.
26. Derschowitz, “The Story Behind Netflix’s Secret Category Codes.”
27. “Although in the FRBR model the entities of all three groups are defined, the main focus is on the first group.” Marcia Lei Zeng, Maja Žumer, Athena Salaba, eds., Functional Requirements for Subject Authority Data (FRSAD): A Conceptual Model (International Federation of Library Associations and Institutions, 2010), 6, https://www.ifla.org/wp-content/uploads/2019/05/assets/classification-and-indexing/functional-requirements-for-subject-authority-data/frsad-final-report.pdf.
28. Library of Congress, “Library of Congress Genre/Form Terms for Library and Archival Materials,” s.v. “Flash Fiction,” accessed June 13, 2023, https://www.loc.gov/aba/publications/FreeLCGFT/GENRE.pdf.
29. Gérard Genette, Palimpsests: Literature in the Second Degree, trans. Channa Newman and Claude Doubinsky (Lincoln: University of Nebraska Press, 1997), 1–2, cited in Pauline Rafferty, “FRBR, Information, and Intertextuality,” Library Trends 63, no. 3 (2015): 504. See also Sarah Bartlett and Bill Hughes, “Intertextuality and the Semantic Web: Jane Eyre as a Test Case for Modelling Literary Relationships with Linked Data,” Serials 24, no. 2 (2011): 160–65, and Anat Vernitski, “Developing an Intertextuality-Oriented Fiction Classification,” Journal of Librarianship and Information Science 39, no. 1 (2007): 41–52, https://doi.org/10.1177/096100060707481.
30. Pat Riva, Patrick Le Boeuf, and Maja Žumer, eds., IFLA Library Reference Model: A Conceptual Model for Bibliographic Information (NP: International Federation of Library Associations and Institutions, 2017), 20, https://repository.ifla.org/bitstream/123456789/40/1/ifla-lrm-august-2017_rev201712.pdf.
31. Riva, Le Boeuf, and Žumer, eds., IFLA Library Reference Model, 9.
32. Riva, Le Boeuf, and Žumer, eds., IFLA Library Reference Model, 9.
33. See the LRM’s “User Tasks Summary,” in Riva, Le Boeuf, and Žumer, eds., IFLA Library Reference Model, 15.
34. Pat Riva, Patrick Le Boeuf, and Maja Žumer, eds., IFLA Library Reference Model: A Conceptual Model for Bibliographic Information (NP: International Federation of Library Associations and Institutions, 2017), 9, https://repository.ifla.org/bitstream/123456789/40/1/ifla-lrm-august-2017_rev201712.pdf, citing IFLA Study Group on the Functional Requirements for Bibliographic Records, Functional Requirements for Bibliographic Records: Final Report, 3.
35. Library of Congress, “Bibframe Ontology,” s.v., “work,” accessed February 20, 2023, https://id.loc.gov/ontologies/bibframe.html#c_Work.
36. Library of Congress, “Overview of the Bibframe 2.0 Model,” accessed February 20, 2023, https://www.loc.gov/bibframe/docs/bibframe2-model.html.
37. Library of Congress, “Bibliographic Framework as a Web of Data: Linked Data Model and Supporting Services,” 3, accessed February 20, 2023, https://www.loc.gov/bibframe/pdf/marcld-report-11-21-2012.pdf.
38. Patrick Wilson, Two Kinds of Power: An Essay on Bibliographic Control (Berkeley: University of California Press, 1968), 25.