Functional Analysis and Design Plan for a Comics Thesaurus
Ryer Banta, Sarah Barrett,
Andrew Brink, and Megan Willan
May 13, 2011
The design of our thesaurus is guided by a functional purpose, which allows for maximum use and benefit for readers’ advisory and related purposes. We envision that the key users of our thesaurus will be librarians, comic novices, and long-time fans looking to expand their reading habits and knowledge of the form. The structure given to the comics domain will illuminate not only direct relationships that may be clear to more experienced users, but also relationships that are less obvious, such as similarities between works from different artistic, genre, and narrative traditions that none-the-less share a related story type, which we identify as a trope in our thesaurus. The design and structure of the thesaurus will allow for meaningful, guided browsing and discovery.
As articulated by Aitchison, Gilchrist, and Bawden, (2000) our (super)thesaurus is organized so that a priori relationships between concepts within the comic book domain are made explicit in order to facilitate navigation within the domain by indexers and searchers alike. In addressing the question of how our indexing language facilitates this navigation—i.e., how it achieves our finding, identifying, and selecting objectives—we turn to Shera and Egan (1956), who identified eight ideals of what subject indexing could—and possibly should—do if it could be developed without regard to limitations of money and personnel (p. 10). Below, we map our requirements to Shera and Egan’s principles, acknowledging that some of our requirements could span several of their principles.
Objective 1: To provide access by subject to all relevant material.
Our thesaurus fulfills this objective by functioning as an expert fan, guiding another enthusiast through the comics canon with the purpose of moving that reader into comics from other media, between readerships within comics, and to enable reader sensemaking by giving structure to the history and tradition of comics. A comprehensive analysis of our domain helped us identify that the existing structure of the comics domain terms and concepts does little to aid users and readers in connecting works in meaningful ways. Rather, the structure, as is, does much to partition works into general, vague categories—i.e., “fiction,” “non-fiction,” and “superhero.” Broad eras, location of creation, format specifications, production details, and simplified genre are too general to hold much meaning or utility. Thus, our indexing language represents concepts that make visible the hidden conventions and connections between art, genre, and narrative tropes so that readers, scholars, and researchers can easily navigate the domain.
As outlined in our domain analysis, comics take four commonly cited forms: comic strips, comic books, manga, and graphic novels. Because of the brevity and self-contained nature of the comic strip form, it falls outside of our domain. Thus, our thesuarus provides access by subject to all relevant comic book, manga, and graphic novel material.
Objective 2: To provide subject access to materials through all suitable principles of subject organization, e.g., matter, process, applications, etc.
As we analyzed our domain, we found that there exist elements organizing and connecting diverse comic books, artists, styles, and traditions: narrative tropes, useful specific genres, and artistic styles. These three elements act as the organizing principles that inform the creation of our preferred terms. Narrative tropes and genres are closely related, but have been separated and conceptualized as distinct elements of organization to allow maximum useful combination of possibilities. This thesaurus allows novel compound facets across these broad organizing principles. Each of these three organizing elements could be visualized as cogs on a combination lock. Each cog has several levels from generality to specificity. The user is able to use terms alone, or in combination with others.
Access via tropes: The identification and indexing of tropes–defined here in the loosest possible sense as those repeatable units of meaning which make narrative work—is a central goal of our thesaurus. Most of the work that has been done on the identification of tropes in popular culture comes from the fans themselves, and in particular, TV Tropes.org. In adapting many of these fan-created concepts for use in our thesaurus, we invoke ideas adapted from Barthes (1974)—i.e., hermeneutic, proairetic, semic, symbolic, and referential—to facet them into meaningful categories.
Access via genre: Our indexing language distills singular distinctions from common compound genres, such as “science fiction” and “superhero”, allowing such distinctions to be cobbled back together to provide improved access. Our genre facets represent effect (i.e., horror, cozy), plot aspect (i.e., apocalyptic, quest for justice), plausibility (i.e., science hero, masked hero), physical location (i.e., urban, rural), and time period (i.e., future, past) to provide useful levels of access.
Access via the visual: Artistic or visual classification of comics has received limited scholarship, either academic or fan-based, compared to other areas within our domain. This is despite the fact that the visual is a hugely important aspect of comics, in and of themselves, but also enormously important in reader appeal. Even non-visual or inexpert readers have visual styles they find enjoyable or repellent. As detailed in our domain analysis, our preferred terms are informed by McCloud’s (2006) four tribes of artists, which classifies creators’ basic approaches to their work. We have collected industry, scholarly, and fan terms and identifications, and classified these techniques and schools under the most relevant tribe. Preferred terms run the gamut from historically acknowledged schools (like the “Ligne Claire” style made famous by Tintin) to fan-created terms for specific visual phenomena (like Jack Kirby’s famous psychedelic method for representing energy, termed “Kirby Dots”). Though terms must be represented in a hierarchical and linear fashion in a thesaurus, we find that plotting these ideas along McCloud’s axes is an effective way to represent concepts that will provide easy subject access.
Objective 3: To bring together references to materials which treat of substantially the same subject regardless of disparities in terminology, disparities which may have resulted from national differences, differences among groups of subject specialists, and/or from the changing nature of the concepts with the discipline itself.
For this thesaurus, there is a desire to take a deep view of comic history that accommodates an international perspective. This thesaurus incorporates established historical categories common within the established literature and discourse of the domain. Core experienced users will recognize the terms and divisions. At the same time, upon analysis, we have discovered that historical category terms are predominately culturally limited in scope and too broad to hold much meaning or utility. Each of the three major facets of our thesaurus–i.e., visual, narrative tropes, and genre–seeks to address an area of weakness in current historical terms and categories used to organize comics.
Objective 4: To show affiliations among subject fields, affiliations which may depend upon similarities of matter studied, of method, or of point of view, or upon us or application of knowledge.
As addressed in Objective 1 above, it is our purpose to help break down silos and promote navigation between forms based on the elements of a comic that make it fun to read.
Objective 5: To provide entry to any subject field at any level of analysis, from the most general to the most specific.
It is especially important to us to provide access to a subject field at various levels of specificity because of our focus on readers’ advisory. Readers’ advisory (RA) is the process of actively assisting a reader (e.g., through an interview process) or passively (e.g., through tools like genre lists) to guide readers to books they will enjoy. Within this broad goal, some practitioners of RA strive to help readers navigate from the kind of reading they are familiar with to works from a different author, genre, or style by identifying the aspects of a book that work for readers and recognizing those features in new and sometimes unexpected places. Our thesaurus draws heavily on this goal, and the principle of providing access to a body of literature based on the various aspects of that literature may provide a point of interest for readers. In order for our tool to be an effective RA tool, the entire hierarchy of our terms must be useful and descriptive. We will limit our specificity, however, because the aim of our thesaurus is to bring together works from across genres and reading tribes, rather than to isolate a single work.
Additionally, this iteration of the thesaurus is limited in its total number of terms, and to cope with that, we will focus on establishing a broad structure before moving on to extremely specific terms.
We had a new challenge in approaching the branch of our thesaurus concerned with visual art. Most approaches to readers’ advisory in comics ignore the visual aspect completely. This seems to be partially because of a lack of vocabulary, but also another manifestation of the deformed book theory. While comics are books (or at least look increasingly like them), they generally are treated like purely textual, traditional books, and any aspects that do not fit into that framework are ignored. As stated earlier, the visual is a hugely important aspect of comics, in and of themselves, but it is also enormously important in reader appeal. Even non-visual or inexpert readers will have visual styles that they find enjoyable or repellent. To cope with this, we have created four umbrella terms: Classic, Animist, Iconoclast, and Formalist. These terms should be easily intelligible and will enable exploration deeper into each of the categories. The terms will also have example images, in order to communicate meaning to visually inexperienced users.
Objective 6: To provide entry through any vocabulary common to any considerable group of users, specialized or lay.
As stated in the introduction, the design of our thesaurus is guided by a functional purpose, which allows for maximum use and benefit for readers’ advisory and related purposes. The structure of our thesaurus will allow users of various levels of knowledge to enter the domain, browse, and retrieve at levels matched to their need and interest. We hope that novel combinations will draw out hidden, subtle connections between works that otherwise seem unrelated. Our concept is that users, regardless of skill level, value elements that appeal to them—within the story, the characters, the tropes, and the art. Many of these appeal factors are not explicitly recognized by readers, but we hope to name and define them in such a way that there is a recognition, an “ah-ha” moment. Additionally, these appeal factors transcend national and stylistic borders, but users at any level of experience may find themselves wary of venturing into uncharted territory. We hope that the design and function of our thesaurus allows for an engaging, fun, rewarding, and helpful way to navigate this complex domain.
In all three branches of our thesaurus, we have drawn terms from a wide variety of sources. As mentioned in Objective 2, we are collecting industry, scholarly, and fan terms, establishing relationships and equivalents where possible. In the visual art branch, these terms run the gamut from historically acknowledged schools (like the “Ligne Claire” style made famous by Tintin) to fan-created terms for specific visual phenomena (like Jack Kirby’s famous psychedelic method for representing energy, termed “Kirby Dots”). Our thesaurus also features a large number of non-preferred terms to cope with the immense variation in vocabulary. Not all of these non-preferred terms have a direct equivalent; they instead will often have a number of related terms that provide more information than the non-preferred term.
Objective 7: To provide a formal description of the subject content of any bibliographic unit in the most precise, or specific, terms possible, whether the description be in the form of a word or brief phrase or in the it form of a class number or symbol.
Subject content is not the focus of this thesaurus. Precision may not be fully achieved in the first version, but the structure is there to continue to develop more specific indexing. Ultimately we do strive to represent faithfully the most detailed descriptions useful for the purpose of making a reading selection. However, there is a level of precision that does not serve our users, and overburdens indexers. Some of that precision could have negative effects such as spoiling the details of the plot, details that readers would prefer to experience in the context of the story. Endings and major character changes may give away too much information about the story. Despite the fact that these tropes could greatly affect how much a reader would enjoy the work, this level of precision could detract from the reading experience.
One area where precision is greatly facilitated by the structure of our thesaurus is the genre section. Genre lists created for comics often simply transfer the genres used for fiction novels. These divisions do not work for comics, which have their own standard plots, typical characters, and constellation of story types. Sometimes the genre ‘superheroes’ is added in an attempt to mitigate this issue, but the typography of genres used for books does not map well to the kinds of stories commonly written in comics. This leaves readers with lists that apply to so many titles as to be useless in helping readers navigate the literature or categories that do not accurately fit the literature. Novice readers find themselves lost in a large and confusing body of literature, and experienced readers view clumsy attempts as ineffective and amateur. Breaking genre down into its components will allow it to be recombined into more meaningful sets that accurately represent the literature of the domain and assist both novice readers and experts looking for new reading avenues.
Objective 8: To provide means for the user to make selection from among all items in any particular category, according to any chosen set of criteria such as: most thorough, most recent, most elementary, etc.
Selection is one of the main goals of this thesaurus. Our defining selection principle focuses on ‘what makes this comic a good read’, mirroring what aspects of a comic an expert fan might use to select a title. This illustrates our core focus on readers’ advisory. Readers’ advisory serves to assist in aiding in the selection of a title most appropriate for the interests of an individual from all possible works. Unlike most non-fiction and scholarly literature, the subject of a comic can be incidental to its ability to satisfy the tastes of the recreational reader, and is therefor unhelpful in selection.
Comics also have an additional aspect missing in traditional fiction: art. Traditional cataloging fails to provide information about the illustration used in comics, which may be a major selection criterion for readers. We have created descriptive categories for comic art to allow readers to select a work that has a visual style they would enjoy without the need to refer to the physical object. When used with catalogs that do not incorporate images, this requires a certain previous knowledge on the part of users who must become familiar with styles to be able to visualize the concept, but search methods like pearl growing from a known title to others would quickly allow a user to become familiar with the vocabulary.
Readers often select by genre, but a single genre label alone is often too general to be useful in selection. The genre of mystery must be subdivided by secondary elements that often mirror other major genre categories, like historical fiction or romance. By faceting genre terms, we hope to enhance the ability of users to select a particular work by using the building blocks that create a genre to accurately and exhaustively describe titles. In this way, works that may fall within the same genre but offer a significantly different reading experience can be differentiated.
There are elements of a work that are more focused than a genre, or which are not traditionally tied to a single genre. These commonalities, or tropes, may also be important to helping a user select between the works they would enjoy.
Design Plan for Constructing Indexing Language
Creating Art Terms:
In creating the art terms we recognized the need to match our terms to real world examples. We, the four editors, compiled approximately one hundred comic art examples from the domain. Each of us focused on finding a wide variety of examples from the broad areas. Specifically, one of us focused on superhero comics, one on Manga, one on independent comics, and another on early comics. We brought all of our examples together and analyzed the art styles in each example. Our goal was to find commonalities between specific examples that transcend the four traditional domain areas.
The first division was to place examples into one of the four tribes or classes identified by McCloud. As we suspected before doing this exercise, the majority of examples fell into the animist class. We did find many examples that fit into the other three classes as well. Our next task was to further divide the four broadest classes. Collectively, we analyzed the art in each example and made groupings that we all could agree on. In each of the tribes we found a variety of art styles. We analyzed elements such as line quality, shading style, relationship to reality, tone, character rendering, background detail, emphasis on storytelling, focus on shape, and dominance of meaning. This analysis became more clear as we categorized more examples, but we realized that this process would need to be clear and operationalized to be of use for indexers.
We have found that the most useful way to analyze comic art is to first determine which of the four tribes best describes the style. The next step an indexer should take is to look at the overall tone. We have conceptualized tone as being a range from looking better than reality, faithful to reality, or worse than reality. Once tone has been determined, the indexer can determine if the predominant mode is expressionism or realism. The final area of analysis is to plot the character and background matrix, which can appear in four fundamental flavors.
Creating Trope Terms:
The four editors of our thesaurus collectively compiled the most useful, domain-specific trope terms. The difficulty in this task was the open, collaborative, mutable nature of the TV Tropes site. The strength of TV Trope’s open nature is that if carefully surveyed and considered, trends and popular concepts emerge. These trends and concepts come straight from the fanbase experts, so our selected terms will resonate with them as users. TV Tropes is a vast resource, so for the scope of this project it was necessary to optimize our term search and compilation. We consulted the index of indexes in hopes that patterns and significant terms and concepts would emerge. We divided the work alphabetically and, as suspected, patterns and important organizing concepts coalesced as we worked through the terms. As we consulted the index of indexes, we also looked at the broader categories that TV Tropes is organized by, to get an idea of the organization scheme and to minimize missing important terms. Now that we have individually compiled terms from our sections, our next task is to compare notes, choose the most salient trope terms, and determine relationships between selections.
Creating Genre Terms:
To facet genre terms, we first surveyed commonly used genre terms in both comics and other readers’ advisory literature. We broke these genres down into their component parts. In thinking about how we should order the genre facets we looked to Ranganathan’s PMEST scheme. Adapting PMEST, our goal was to order the facets so that a descriptive sentence could be applied. Our facet order goes: effect, plot aspect, plausibility, physical location, and time period. A sample sentence applying this order would read:
This comic is a COZY (effect) SURVIVAL (plot aspect) tale about a MASKED HERO (plausibility) in an URBAN (physical location) world during RECENT PAST (time period).
Design Plan Action Items and Timeline
Domain Analysis – Completed
- Discuss initial conceptions of domain and develop utility of language – Completed
- Research areas suggested by Hjorland – Completed
- Match Hjorland areas to our domain – Completed
- Divide areas of research appropriate to expertise and interest – Completed
- History of comics form, format, publishers, fan-perspective – Completed
- Art terms for comics and general art terms – Completed
- Special classifications and indexing specialties – Completed
- Genre research – Completed
- Reader’s advisory research – Completed
- TV Tropes research – Completed
- Bibliometrics – sales and popularity – Completed
- International trends – Completed
- Publisher image and imprints – Completed
- Synthesize research – Completed
- Development of functional requirements and purpose – Completed
- Log un-resolved issues, problems, and concerns – In Progress (ongoing)
- Draw on Domain Analysis – Completed
- Plan further writing and editing – Completed
- Deadline for component submission – Completed ( Thursday, 5/12)
- Work on editing – In Progress (Thursday, 5/12 – Friday, 5/13)
- Compile candidate terms – In Progress (ongoing individually; due Sunday, 5/15, then finalize)
- Assemble citations and definitions- In Progress (individually; due Sunday, 5/15)
- Arrange terms in hierarchical order – In Progress (Sunday, 5/15)
- Discuss other arrangements – In Progress (Sunday, 5/15)
- Narrow term selection to scope (100 Preferred Terms) – In Progress (Sunday, 5/15)
- Data entry of terms into Data Harmony software – In Progress (after finalization on 5/15)
- Divide work load – In Progress (to do on 5/15)
- Set submission deadlines- In Progress (to do on 5/15)
- Perform editing and quality control measures – In Progress (from 5/15 to completion)
- Develop entries for non-preferred terms – In Progress (secondary after 100 Preferred Terms)
- Develop and implement notation system – In Progress (to do on 5/15)
- Create Thesaurus Introduction – In Progress
- Integrate elements from Domain Analysis and Functional Requirements – In Progress
- Add necessary writing – In Progress
- Divide parts to be written – In Progress (to do on 5/15)
- Deadline for component submission – In Progress (to do on 5/15)
- Perform edits – In Progress
- Collective Editing – In Progress
Aitchison, J., Gilchrist, A., & Bawden, D. (2000). Thesaurus construction and use: A practical manual. Chicago: Fitzroy Dearborn Publishers.
Barthes, R. (1974). S Z (1st ed.). New York: Hill and Wang.
McCloud, S. (2006). Making comics?: storytelling secrets of comics, manga and graphic novels (1st ed.). New York: Harper.
Ranganathan, S. R. (1967). Prolegomena to Library Classification. (Madras: Asia Publishing House).
Shera, J. H. and Egan, M. (1956). Classified catalog: basic principles and practices.
Chicago: American Library Association.