Responding to the frame: classification, material boundaries, and expressiveness in personal digital bibliography
July 19, 2013, 13:30 | Long Paper, Embassy Regents D
A systematic bibliography is a thematic collection of references to intellectual works, such as the subject guides produced by academic librarians. Free of the constraints of physical libraries, these metadata representations are able to describe and relate any resource in the bibliographic universe at the will of the bibliographer, to fulfill an endless variety of goals and express any number of associated opinions. A consumer health bibliography might provocatively endorse certain elements of alternative medicine; a guide to anthropology might focus on “scientific” methods as opposed to ethnographic ones, as espoused by the department at a particular institution.
Accordingly, the ultimate collection, comments Roger Chartier, is not one of actual books but of citations, a “library without walls,” as in Conrad Gesner’s monumental sixteenth-century compendium (Chartier, 1994). Instead of shelves, Gesner’s intricate classification scheme endows the bibliography’s contents with form and structure. Although expressed in words, a classification of this sort is meant to establish relationships between abstract concepts, for which the identifying terms are merely replaceable labels. A category labeled “Geometry” indicates the idea of geometry, not the word. Current standards, such as the 2005 NISO standard for controlled vocabularies and the 2010 IFLA recommendations for the representing subjects in the library catalog (FRSAD) continue this mode. (Jonathan Furner [2012] contributes a useful critique of the FRSAD approach.) Even library classifications are designed as abstractions, independent of the material world, to be potentially expressed in any shelving configuration; in these classifications, as well, the concept that designates a class is conceived as a purely mental construct. Its proper identification is a notation or code (e.g., GN 790 in the Library of Congress Classification), for which any label is just a form of convenient documentation.
The interpretive nature of these concepts, of their system of relations in a classification scheme, and of their application to any set of physically or virtually collected works, has been widely accepted within information studies (as in, for example, work by Birger Hjorland, Jens-Erik Mai, and Hope Olson). A subject class like GN 790 (Anthropology — Prehistoric archaeology — Megaliths) embodies a remarkable complex of judgments: that the significance of megaliths has to do with anthropology, that a subject is both an identifiable and important characteristic of documents, that the set of documents assigned to GN 790 are equally about “megaliths,” that “aboutness” is a decision easily and concretely made, as only some examples of the complex of appraisals that GN 790 expresses.
The recognition of classification as a mode of interpretation, and an accompanying sense that a citation collection or bibliography is itself a form of creative expression, is similar to the recognition of textual scholars that the application of markup languages such as TEI to digital texts is equally an interpretive act, and that critical editions expressed as digital collections are interpretive efforts. In textual studies, such realizations have led scholars such as Jerome McGann (2001), Johanna Drucker (2006), and Bonnie Mak (2011) to examine both the ways in which the material qualities of printed documents contribute to meaning and the means in which these material effects are transformed through digitization. Mak, for example, traces the semantic contributions of the page in manuscript, print, and digitized versions of a fifteenth-century Latin text. While the page signifies differently across these versions, the page as a material and conceptual construct always “matters.”
In contrast, due to the longstanding emphasis on the nature of classification as abstraction, the contribution of any material component to the meaning of a systematic bibliography would initially seem quite strange. Classifications are most often studied as self-contained systems of their own, and not as applied to works in practice. Even when classifications have been studied as infrastructural components of document collections, the relation between class and text is seen as a dialectic of ideas, and not as a means of constituting a thing with a specific material presence. The translation of paper-based systems to digital environments has only confirmed this rejection of presentation as having anything at all to do with the meaning of either classification schemes or the systematic bibliographies structured by them.
This paper describes how a user study focused on the authoring of personal digital collections, or personal systematic bibliographies, demonstrated a link between the material instantiation of a bibliography and the expressiveness of its structuring. Findings from this study suggest that meaning making for both writers and readers of personal digital bibliographies inheres to some degree within the abilities of the bibliographic environment to enact framing devices that clarify the system of relations between items as the primary focus of the text. Just as Gesner’s famous bibliography achieved notoriety not merely for its comprehensive selection of works but for its instantiation of relations between its contents, so does the expressiveness of personal digital collections result from the interplay between structure and contents, and not merely from the contents alone. In this paper, I contend that the ability to articulate and perceive this set of relations requires a frame with material presence, and not just mental presence.
The study described here is part of a larger project to investigate personal digital collections, or personal systematic bibliographies, as a form of creative expression. On the Web, social media services that enable users to aggregate content through the creation of personal digital collections have proliferated. Through communities such as Pinterest and GoodReads, users select and share resources linked (or cited) from elsewhere. Personal collections are enabled for retail sites, such as Amazon, and content providers, such as YouTube and Spotify. Cultural heritage institutions, including libraries and museums, have also encouraged users to create personal collections, through services like MyMet from the Metropolitan Museum of Art in New York. Marty and Kazmer (2011) contend that such personal bibliographies facilitate the co-construction of knowledge between institutions and their user communities. As opposed to Gesner’s opus, however, most Pinterest boards and YouTube playlists are trivial and bland. The project’s initial investigation identified, out of the masses, a few salient examples that did constitute compelling expression and defined characteristics that contributed to these examples’ expressiveness: an original purpose for creating the bibliography, a distinct authorial voice in the presentation of the bibliography, and a sense of emotional intimacy underpinning the bibliography’s contents (Feinberg 2010, 2011).
The second stage of the project began to consider the design of environments to author such collections: could we encourage writers of personal bibliographies to generate more expressive examples? (One can see this as similar to an investigation focused around describing a markup language to more effectively characterize the poetic nature of texts as opposed to their informative nature, a complaint against TEI articulated by Jerome McGann.) A first experiment examined whether exposure to bibliographies that embodied all three of the identified expressive characteristics would affect the process or product of collection design (Feinberg, et al, 2012). Twelve participants created bibliographies by selecting items from specially constructed digital video libraries of source content. After creating one bibliography, participants interacted with expressive examples that were also created from the that video library. Participants then created a second bibliography using a different library of video source material. This simple intervention did not work; participants’ bibliographies and design processes did not change after experiencing the examples. Participants did, however fluently read and describe the distinctive characteristics of the expressive bibliographies. Participants especially noticed the use of descriptive infrastructure (titles, annotations, and so on) in the examples to convey meaning, yet participants made little use of such descriptive elements themselves.
This paper focuses on a second experiment to encourage descriptive infrastructure as a means of facilitating expressiveness in participant bibliographies. We examined whether a structured task that asked participants to separately consider the mechanisms of resource selection, description (titles, labels, annotations), and arrangement (systems of ordering, or relations between items) would enhance expressiveness. To focus on task, and not interface, we switched from a digital environment to a simple physical one, using small libraries of print books as source materials and bulletin boards as document substrates (with preprinted paper slips for citations, along with index cards and Post-its for descriptive elements). To our surprise, while task structure did not influence designs, all the bibliographies were more expressive, compared to the examples, than the previous experiment, and 19 of 24 participants implemented classificatory structure in their work. Many of these structures were reinforced via complex visual arrangements. This paper describes how what initially appeared to be small, inconsequential changes in the material conditions of our experiment resulted in markedly different authoring processes, as well as significantly more complex products. I discuss what the framing devices illuminated via this experiment signify for the development of digital environments to support the authoring of personal expressive bibliographies. As Johanna Drucker (2005) delineates the areas on a page where graphic elements contribute to its meaning, I will describe how material elements of personal digital bibliographies may contribute to (or detract from) their textuality.