Abstract
Comic Vine (CV) is a semantic platform focused on documenting published comics. Developers Dave Snider, Ethan Lance, and Tony Guerrero launched the site on December 2006 as part of their first series of proprietary wiki platforms covering various entertainment industries. More than a wiki, CV was at launch a news and review website covering comics and offering discussion forums for users. This article examines how a semantic platform has developed to offer descriptive features catering to one industry (comics), using a proprietary architecture. Using the walkthrough methodological approach, I find that such practices, while not adhering to open-web standards, contribute to architectural design diversity (ADD). Standards in semantic data often push toward common grounds and exchange parameters. The ADD concept presented in this article focuses on highlighting divergent technical schemes in the computing sciences that do not rely on a few standards. I draw mainly on approaches and theories from information studies, grounded in contextual insights from communication studies and human-computer interaction.
Introduction
Comic Vine (CV) is a freely available online platform focused on cataloging published print comics. Promoters Dave Snider, Ethan Lance, and Tony Guerrero launched the site on December 6, 2006, as part of their first series of proprietary wiki platforms covering various entertainment industries (Guerrero, 2022). More than a wiki, CV was, at launch, a news and review website covering comics and offering discussion forums for comic readers. While many entities have owned the site, CV has been closely associated with gaming website GameSpot, as one of its online outlets since its purchase in 2016. CV is not a professional-level cataloging and indexing resource aimed at institutional users such as librarians, researchers, publishers, commercial users, enterprises, and industrial users. Its developers and promoters appear to gear the platform toward casual comic readers and collectors. Yet, this rich and data-extensive platform could fulfill the information needs (Wilson, 1981) of professional users. One of the objectives of this article is to explore how CV’s unique semantic architecture adheres to the information needs of comic readers, collectors, and fans. Semantic architecture, as used here, draws on accepted definitions of architectures as the representation of the structure of a system (Gerber et al., 2008), such as a semantic platform. An important challenge and potential criticism of CV, explored in this article, is that its atypical and proprietary semantic architecture does not adhere to existing standards related to semantic technologies. I refer to this phenomenon as architectural design diversity (ADD).
CV is not the only English and French-language comic-related semantic platform. English and French-language comic markets are part of the top-five production spaces along with Japanese, Korean, and Chinese (Groensteen, 2010). One feature of CV unique in English and French-language platforms is its semantic architecture and high degree of links between related elements appearing in any entry accessed by users. This effective knowledge graph is unique to CV in the English-language comic world, although French-language platform BDGest (Table 1) has similar properties. Another feature that separates CV from other semantic media, such as BDGest, is the extent of its qualitative description of comic characters, creators, and comic book series beyond simple indexing. This feature makes CV closer to semantic technologies such as Wikipedia (Table 1) or the comic-focused Toonopedia (Table 1). Few comic-related semantic platforms combine narrative, extensive qualitative entry descriptions with bibliographic and cataloging data, in graph-like structures. Toonopedia, for example, which explores comic characters in an encyclopedia-like format, is not a reliable cataloging database as it lacks bibliographic data. Seemingly, CV can attract a wide audience of casual users interested in current discussions related to comics, such as news and reviews, while providing extensive semantic data. As Kwon et al. (2020) mention, an important feature of semantic platforms is relation of data with one another through links. With CV, users can search different aspects of comics, be they through characters, creators, or serial data about a series publication. When viewed together in one page, all these elements are pulled together within one page (wiki) for the user. CV covers comic books, a cultural realm with artistic (Beaty, 2012) and literary (Meskin, 2009) values that are often unrecognised and contested.
Links and Databases.
This article adds another scholarly evaluation of CV such as Custer et al. (2016), examining how a semantic platform has developed to offer descriptive features catering to one industry, using a proprietary architecture. I argue that such practices, while not adhering to open-web standards, contribute and are beneficial to ADD. ADD is the practice of using non-standard descriptive features that cater to the information needs of one industry or community of practice. Information needs are about what information people lack, attempt to seek, and make sense of (Wilson, 1981). ADD is not a rejection of the standards that underlay and provide access to the internet. Standards such as Internet protocol suite (TCP/IP) allow various manufacturers and their customers to access shared technological realms (Newman, 2018); however, as argued by both Zittrain (2008) and Wu (2010), single standards can limit innovation and development. Thus, in the semantic media space, CV may contribute to the generativity and the development of alternatives that may one day challenge existing standards and improve them.
I begin by providing a history of documenting comics, followed by a description of the semantic contexts of comic archiving online, and sections on user-generating semantic practises and participatory comic cultures. Following this literature review, I use the walkthrough method (Light et al., 2018) to explore CV’s semantic platform. Based on these findings, I theorise the significance of ADD using CV as a case study. Based on Gillespie’s definition of media platforms (2010) and work by Iliadis et al. (2023), in this article, a semantic platform is defined as the semantic-based computing spaces where administrators, markup languages, and consumer products meet to produce “facts” for online audiences. I follow Iliadis’ (2022) definition of semantic media as computing resources used for orchestrating and conveying facts, information, meaning, and knowledge to users of popular internet products through such things as metadata languages, computational ontologies, knowledge graphs, and web schemas.
When it comes to comics (and thus CV), the many terms describing people who consume comics, such as fans or readers, may be banded around. While this is not a user study, it is important to clarify their meaning of various terms. In this article, I adopt the common taxonomy of comic readers, comic collectors, comic fans, and comic speculators (Beaty, 2012; Gordon, 2012; Woo, 2012) as such; readers consume comics casually, but collectors select comics carefully before adding them to their collections. Collectors curate their collections to develop them.
The term “fan” is often used as a larger class containing collectors. However, based on current practises of comic fans on different media, such as web comics, fans may follow a comic with great enthusiasm and have affective investment, but they may not necessarily collect “copies” of a work. For example, being a fan of a web comic does not require any collection of a physical “copy” of the comic. Fans of web comics can copy their pages digitally on computing devices, but this “copy” is not something they usually trade or even preserve as an artifact. Finally, the Marxist economic theory can help us explain speculators as consumers who perceive exchange values in comics that they own or want, beyond their utility. Comics become goods that are traded and sold instead of read or collected for personal use.
Literature Review
The study and the documentation of comics have often relied on semiotic approaches (Groensteen, 1999, 2011) instead of semantic ones. Semantics covers the meaning and indexing of linguistic expressions while semiotics investigates the representation of signs (Føllesdal, 1997). Groensteen contends that many elements that constitute comics are meaningful signs (1999). He describes panels, comic bubbles, and written texts as narrative devices imbedded within a visual medium (Groensteen, 1999). This native comic syntax, through juxtaposed images creating meaning, panels, and the space between them (known as closure), was also the topic of research of McCloud (1993/1994) in his visual text on comics. McCloud’s work helped helm comics as potent examples of visual rhetoric. Comics’ core form, sequential art, is unique in that it can be as much art as literature (Eisner, 2008). Eisner’s work focused more on the narrative and literary value of comics where he explores drawings, captions, and colours as potential signs. These semiotic analyses of comics rely on the indexing of comics beyond the documentation of bibliographic data, contents to determine themes, and subjects and topics contained within them.
However, the semiotic study of comics has not precluded the study of its semantic subtext. Literature on the representation of structured comics data via techniques of semantic platformization (e.g., building metadata about comics that can then be browsed and consumed on internet platforms for semantic media and comics) is sparse. Guérin et al. (2013) presented their research on a custom-made platform the team developed at Université de la Rochelle, in France. The platform, eBDtheque (Table 1), is a database of comics that does not just list comic titles, publication dates, creators, and other bibliographical data but also provides data and annotates captions within the comics, making them searchable by users. The algorithm embedded in the platform also catalogs elements such as panels and bubbles. The research team used the database on a limited European comics’ corpus with a small data set.
Hibbett (2019) attempted to create a transmedia corpus across comics and several other texts of Marvel character Doctor Doom using mainly the Marvel Chronology Project (MCP) (Table 1), an online database specializing on cataloging Marvel comics. As part of his investigation, he also looked at the Grand Comics Database (GCD) (Table 1). While the MCP limits data users can retrieve to titles where characters appear, the GCD is more versatile, as observed by Rhode and Bottorff (2001) who chronicled the early years of that generalist comics’ database. Monaco (2017) described some of the challenges of the GCD’s cataloging formats.
Wiart (2019) has investigated the place of platforms such as BDGest as part of comprehensive studies on literary social networks in the French-speaking world. While his exploration is descriptive, he does mention how literary platforms for books and comics, such as BDGest, supplement the work of libraries and librarians, offering them opportunities to provide more insight about materials than bibliographic data (Wiart, 2019). He notes how these platforms, seldom created for professional users such as librarians, still perform valuable functions in terms of informing the public about literary works (Wiart, 2019). Wiart does not pay any particular attention to the specific characteristics of comics, versus other literary forms.
A History of Documenting Comics
The push to catalog comics is not new. Comics are serial media with collectible value, meaning that readers and collectors were naturally inclined to know about past knowledge related to the contents and the creators that created the comics they read. Publishers and resellers had a vested commercial interest to catalog comics because their serial nature encouraged multiple purchases of series and sometimes even duplicate copies of the same work due to perceived investment and market values. Next, much like in library science, organizational data also focus on the document itself (Serchay, 1998). Version of a comic book, such as a first edition from 1944 or a reprint from 1972, matters to the printed comics’ community of practice. Thus, the version of the work matters as much as its title, its contents, or its contributing creators.
Even though comic books exhibited the same documentation features as books, their cataloging was not necessarily universal. In the United States, many early comic books were not originally documented and archived by the Library of Congress, as like other books and printed materials were (The Library of Congress, 2017). This lack of documentation forced the comics-related community of practice, which includes publishers, creators, resellers, distributors, fandom, readers, and collectors to fill knowledge gaps about comic books. Much data about comics could have been lost forever if it were not for the contributions of other actors who documented comics for various purposes. The cataloging was not always performed by documentation professionals such as librarians or archivists (Scott, 1998). Much of the work was done mostly by fans, collectors, amateurs, and early comic book resellers (Serchay, 1998). This practice continues today, although one of the most important cataloging practices was often performed by documentation professionals working at publishers who may have been formally trained or not in library science, as I will explore further in the following sections.
Comics-related librarianship and its focus on collection acquisition, youth literature, and the burgeoning “graphic novel” market forced a shift away from periodicals to the cataloging of books. The librarians cataloging comics did so to serve their patrons with curated lists, ignoring the wider availability of comics (Steele, 2005). This practice reflects the needs of a community of practice that differs from CV users.
Cataloging of comics continues to be performed through the archiving and private librarianship practises of comic publishers. Most major publishers invested in serious record-keeping practices that helped them curate their intellectual property. The distribution of such data is proprietary, exclusive, and often prohibitive with analog records and comics. Such bibliographic data are often featured in publishers’ previews sent to news sites but also used by distributors and digital comics’ platforms such as ComiXology (Murel, 2020) which sells digital versions of printed comics through apps.
Another cataloging source for comics is comic book price guides (Wyburn & Roach, 2012) published since the 1970s and consumed by readers, collectors, speculators, as well as resellers and collectible store owners. Through printed booklets, online price guides, and apps that document and determine comics’ valuation, they entered a commodification phase where their main use value was their exchange value. CV does not promote, nor document the exchange value of comics, focusing on characters, creators, and stories. The work of dedicated comic fans publishing various fanzines since the 1960s has helped provide cataloging data on comics (Highsmith, 1987). Their work sought to bridge the gap of metadata on comics through the cataloging of comics through publishing records, sample images, such as covers, indicia, subscriptions, advertisements, and announced comics. Much data missing and access to the printed “databases” generated by fans were transient until such data started to be published online.
User-Generated Cataloging for Comics
Building semantic data sets about comics is not new and may explain why the practice occurs frequently in the contemporary information economy. As early as the 1980s, some publishers and many comic book stores offered comic readers custom-printed cardboard cataloging cards that they could use to catalog their collections. These cards served a similar function as book catalogs in libraries but encouraged comic book consumers to provide the needed labour to document their collection. Data recorded and added to these individualised knowledge graphs included the publisher’s name, the comic’s title, issue number, and sometimes the names of the authors and artists. Store-printed cataloging cards could be blank, allowing the collector to insert the series and basic biographical data. They would often have numbers going from 1 to 100 and space to add a few more hundreds of series with more issues. Store-based cataloging cards served the dual purpose of being branded business cards on one side and a catalog on the back. The design of the card was gamified tools to encourage collectors to list their possessions and complete holes in their collections, possibly at the back issue bin of the very same store.
The second type of cataloging cards, published at least as far back as 2017 (Babos, 2017), were often printed by larger publishers as checklists for events that jumped from one series to another, or one title to another. Thus, the checklist helped the comic reader know what comics to buy in comic crossover or multipart series, easily. These cards were not business cards but more promotional items often handed to comic stores so they could be given away freely to customers or printed as advertisements within comics. They had better production values and were often the same size or compatible with a trading card checklist. With such cards, the cataloging was partly done by the publisher, as the title of the series or crossover events were listed. All collectors had to do was to check empty boxes to list what they had.
Even with early cataloging software helping collectors organise their comic collections in the 1990s (Culbertson & Jackson, 2016), collectors were still responsible for providing missing information about comic books, either by manually copying it from other sources or by performing the cataloging themselves. Even in electronic formats, such cataloging was not shareable within the community of practice. While early cataloging software was probably not focused on enabling semantic and sharing processes, cataloging practices of comics were well-established with comic readers and collectors. Alternatively, the GCD took shape in 1994 and was based on the Amateur Press Alliance for Indexing, a custom standard created in 1978 to index comics and related literature (Klein, 1997).
Participatory Culture and Comics’ Semantics
Using Jenkins’s participatory culture theory (2009), we can construe these early forms of auto-documentation performed by comic readers and collectors as being earlier practises that supported the current comic-related crowdsourced cataloging efforts of semantic data, as performed by semantic media such as CV. Jenkins describes participatory culture as being the opposite of consumer culture (2009). Culture is no longer created by producers alone but also by consumers producing contents consumed by other consumers (Jenkins, 2009). The interaction with culture becomes bidirectional as consumers can respond to producers and even modify cultural goods produced by the latter.
While Jenkins’s participatory culture is mostly used with actual cultural contents such as comics, their cataloging involves the production of semantic data by comic readers and collectors. Interestingly, combining cataloging of semantic data and participatory culture may seem incompatible, as the cultural goods created from original producer-generated contents are not modified by comic collectors. One may argue that all that they do is generate missing data that help to catalog comics. In this perspective, collectors, readers, and fans do not produce new contents from the comics. Participatory culture, to some extent, involves the photocopying, the reproduction, or even the resampling of existing bits of culture that are then reinterpreted and redistributed to the public without the express authorization of content producers.
Thus, could it be that cataloging existing data does not infringe in any way on contents produced by comic publishers? The collectors do not craft new stories from original ones generated by publishers. Moreover, they do not repurpose original contents. They merely document stories and share noncopyrightable semantic data. But comics and its fandom are a bit more complex than that. Observing the phenomenon known as closure, for example, we can understand the cataloging of comics as more than the recording of bibliographic metrics. Closure, as mentioned earlier and defined by McCloud, is the space between two panels that force the comic reader to add meaning, to generate a form of understanding of what happened between two drawings (1993).
Closure is found within the comic page, but it is also generated by comic readers beyond the comic page, especially when dealing with serially published materials such as comics, that often tell an unending story crafted by creators and their publishers, whose mythology and canon are closely guarded by fans and comic readers with a vested interest in maintaining continuities of meta-narratives. Cook (2013) notes that the canon of a massive serialised collaborative fiction, especially in the context of comics, involves publishers, creators, and the fandom. Closure, when seen in these regards, is also the documentation and often debates among fans about what is “real” or not in the lives of comic-based fictional characters. To some extent, the cataloging of comics reinforces this practice that comic fans have of determining what is real, regardless of what has been printed or not.
This meta-narrative is documented as thoroughly as bibliographic data and, at the same time, through cataloging. Thus, Batman’s frequent meetings with the international “Batmen of all Nations,” a single story published in Detective Comics #215 (Moldoff et al., 1955) in 1955 was no longer canon in current continuity until it was brought back. In 2007, writer Grant Morrison restored this story and the international Batmen that had ceased to exist for decades (Morrison & Williams, 2007). The Batmen of all Nations were a group of international vigilantes inspired by Batman that met in one story but were erased from DC Comics’ canon in 1984, until Morrison’s story in Batman #667.
The bibliographic and semantic data concerning the “Batmen of all Nations” never ceased to exist and instead supported the greater meta-narrative about the many international Batmen. Semantic media such as CV have detailed descriptions of the story even when it was not canon. The documentation of this story, performed by fans, went beyond the needed bibliographic data required to document the original story and could have served as a resource for future DC Comics creators, such as Morrison, looking to reuse these non-canon characters and stories in newer stories.
Semantic data take a novel form when it is used and invested to such extent by readers, collectors, and fans beyond the original designs of many comic creators and their publishers. Another similar form of closure and semantic data are the “No-Prizes.” The no prizes are literally non-awards that Marvel Comics editors frequently awarded to readers who can explain continuity errors related to a Marvel Comics event, place, or character. No-Prizes were published in letter columns by Marvel Comics editors, responding to letters from fans. The letter column itself had been a performative space where fans could offer feedback and discuss with creators and editors (Gordon, 2012), thereby building the initial kernel of printed comics’ fandom, participatory culture. Yockey (2017) mentions that the No-Prize awards (which remits nothing but reputation and recognition to fans) help fans assert their own ascendency over editors over the shared fiction created. Again, Cook (2013) argues that the No-Prizes are participatory meaning-making practices involving producers and consumers in the production of fictional worlds.
Regardless of the composition guidelines found on CV (L.A.M.P, 2011), narrative and qualitative interpretation of documented comics will occur. This interpretation is very much from the perspective of fans explaining to their peers what happened in a comic book. Much effort is spent by community managers at CV to encourage neutral wikis that inform visitors without personal biases. Composition instructions such as those for writing a character’s origin, creation, evolution, and major story arcs are mixed with the website’s style sheet, instructing contributors about how to use tags and add images or links. It requires a degree of digital literacy and thematic expertise for a contributor to navigate both aspects of the wiki. Because contributors do not have to consider bots and site-wide clashes between editors such as on Wikipedia, CV probably offers a more familiar and easier user experience.
Methods
I draw mainly on approaches and theories from information studies (cataloging, information needs, communities of practice) grounded in contextual insights from human-computer interaction. Specifically, I borrow from existing literature on librarianship related to comics and graphic novels (Highsmith, 1992). To understand the context of semantic platforms such as CV, I briefly explore the history of comic-related metadata by looking at five documentation sectors (the Library of Congress, comic-related librarianship, the archives and librarianship works of publishers, comic book price guides, and fanzines).
Then, I use the walkthrough method to explore CV’s semantic architecture (Light et al., 2018). Walkthroughs are a step-by-step approach to investigating apps and other technological artifacts while browsing and taking notes about them (Light et al., 2018). With the walkthrough, I landed on CV’s home page and took notes of the environment before searching for specific terms to explore interior pages on the site. I performed several searches to obtain different results as needed. For example, I searched for comic issues with known multiple covers as this was needed in some of my analyses. The queries I used were predetermined and not random. I knew what results I was looking to unearth from CV, being familiar with the platform. In that sense, it was not a grounded study. However, beyond the demonstrative intent, the search terms used had no other values or connotations related to gender, social class, culture, and related descriptors.
A Walkthrough of CV’s Semantic Architecture
Upon landing on CV’s home page, one sees a featured article taken from the blog and news and review parts of the site as seen in Figure 1. Underneath are listings of the most popular issues. Links to most popular issues take readers to specific entries from a comic book series. This link reinforces the preponderance of serial comic books, as opposed to other comics as issues are installments of larger series. Within the listing of an individually spotlighted issue, there is an entire story summary, an index, as well as bibliographical data that appear to be secondary to the descriptive contents about the story and the comic as seen in Figure 2. A top banner allows users to navigate horizontally through a list of previous and subsequent issues, represented by short title listing with issue numbers and small cover thumbnails.
Comic Vine’s home page.
Sample issue-level Comic Vine entry.
Below the top banner is the story summary on the left and a picture of the cover of the issue on the right. Indexed data, such as the creators (i.e., Jack Kirby, Alex Toth, Stan Lee), the characters appearing in the story (i.e., Archie, Casper, Superman), teams (as superhero teams such as the Justice League or the Avengers), the locations (Wakanda, New York, Pluto), the concepts, special objects (i.e., Wonder Woman’s lasso of truth, Green Lantern’s ring), and story arcs the installment is a part of, are available below the story summary (i.e., branded storylines such as Onslaught, Crisis on Infinite Earths). Branded stories are events publicised by publishers as such, so even when appearing in several comic book series, they are part of the same “event.” These events and branded storylines are the ones for which publishers may print cataloging cards for. Visitors to the page are encouraged to log in to post their own reviews of the issue below the story summary and the indexed data.
Of interest to the comic book medium is the place of the cover and other visual elements pertaining to sequential art. As seen in Figure 3, the cover is posted on the left side of the page with the width taking about two fifths of the page. Often comics will have multiple variant covers. Only one appears at the top while all other variants are below the bibliographic data. Sometimes, publishers do label some covers as the primary and others as variants. There is no established policy in the North American nor the European comic book industry on labeling multiple covers for one comic book. Sometimes, they are retail incentives, sometimes they are labeled A, B, C, and so on. It is unclear what policy editors and fans at CV use to determine which cover to select from those available as the primary one featured above the bibliographic data.
Sample issue-level Comic Vine entry with multiple covers.
Below this cover, as mentioned earlier, the bibliographical data about the comic follow. In regular bibliographical standards, creators and editors would be listed there, as opposed to CV’s index section, below the story summary. Here, the issue details cover the title of the story, volume of the series, issue number, and the cover date of the comic, which does not necessarily correspond with the actual street date. For decades, the cover date of comics was a few months after their availability in stores. The gap has shortened recently. The last element is the date the comic book appeared in stores. It appears from the sample observed that the store date is used with newsstand and retail stores, as well as specialty comic book stores who operate in the “direct market” where products are not usually returnable. This information is not always available and may even be problematic to obtain as different types of stores may get copies of books at different times.
Finally, below this bibliographical section as seen in Figure 4, there is a list of the most active contributors to the wiki page, including numbered contributions. Of course, usernames are listed instead of users’ real identities. “Deactivated users” are also listed, although their usernames are not available nor displayed. Clicking on the username of a contributor takes the visitor to that person’s personal page where contributions to the wiki are listed. These contributions include forum posts, contribution counts to the wiki, comic book reviews, blog posts, list of comic books added to CV, a wiki history, and basic social network features, such as a list of users followed and followers. This individualised wiki profile is the core networking tool used to create a sense of community on CV while encouraging users to add more data to the platform. As I will investigate in the following sections, there are kernels of gamification in the bare bone social media platform.
Sample Comic Vine spotlight/event entry.
Before we go more deeply into the gamification features of CV, it must be mentioned that the architecture of the platform has a responsive design, that is, the width of the page automatically readjusts itself depending on the width of the screen the visitor uses to browse the site. On larger screens, as noted previously, the right-ended column containing the comic book covers, bibliographic data, and a listing of the top contributors to the wiki occupy about two-fifth of the standard desktop page. When browsing a wiki on a tablet or a smaller device, such as a smartphone, this column appears below the left-handed one where the story summary and the indexed data appear, as well as user reviews of the comic book. On a technical level, the placement of visual and bibliographic data below the story summary and the indexed data signifies that the story elements are primary to the former. When viewed through semiotic lenses, this indicates that similar to many comics, the visual part may be viewed as being a lesser contribution than the narrative parts. But it is not just the visual data, in the form of the comic book covers, that appear secondary. The bibliographic data also appear as secondary.
CV’s separation of indexed data which support the narrative description of the comic books from bibliographic data is interesting. In standard book catalogs, bibliographic and indexed data appear together. To some extent, for the librarian, the bibliographic data appear to be primary to find a particular version of a book, while the indexed data will not usually change much between editions of the same work. But CV is not fulfilling pure cataloging needs of users looking for a comic book to borrow from a library, nor the needs of a librarian ordering a copy of such book for her library. CV fills in comic readers, collectors, and fans about the contents of a serialised comic book as part of a meta-narrative.
The treatment of the narrative aspects of the comics is interesting as CV could have displayed sample pages, panels, or strips of the comic books it features in its wiki. The only major visual element presented in its documentation of a comic book is the work’s cover. It is true that taking sample elements from the inside of the comic books would be more labour intensive (which could discourage users to contribute to the wiki) and perhaps a bit difficult with rarer works. This would require more server space to host images and would have to be well managed to dampen copyright infringements. Thus, a reasonable fair-use sample of the sequential art within the comic books would have to be sampled and posted within the wiki. This would require greater moderation and editing by administrators and community managers.
Mostly, CV avoids all of this by focusing solely on covers, which usually are not representative of the sequential art within the work. The cover may be drawn by a different artist, drawn in another media or style, and is often a semiotic element acting much like a movie poster or a book cover. It is a symbolic representation of the work that is part of the whole, but not necessarily the story as told with various sequences of images. Much like movie posters and book cover, the comic book cover is a marketing and branding device used to inform audiences about the creative work. As much as it is part of the work, audiences want to watch the film or read the comic book to grasp any meaning from the works.
Thus, the cover is not presented as a primary element of the wiki but serves mainly to identify preceding issues from the same serial works, in the top banner, and to identify the actual comic book issue within a series. However, there are icons next to indexed elements such as characters, creators, teams, and locations, as well as the wiki contributors’ avatars. The closest form of semantic data identifying these images would be alternative (Alt) texts that describe images within an HTML tag. While these Alt tags are used to describe images for visually impaired users, they can also be used as crude forms of semantic data (Gibson, 2007; Huang & Sundaresan, 2000; McCall & Chagnon, 2022). These Alt tags are found on all icons, but they are not found on any of the comic book covers displayed within the wiki. This architectural omission reinforces the secondary use of the comic covers and, to some extent, would benefit the semantic value of the data published at CV.
Contribution Gamification
The Wiki Points system used by CV consists of points attributed by editors to contributors based on additions to the wiki, editing, corrections, and grammar revisions. Every contribution allows contributors to obtain points; however, it is not an automated system. It relies on the good will of editors. Specifically, within 24 hours, editors check contributions. Contributors with contributions over 1,000 points can bypass the moderation step and publish directly on the platform for most edits (Poet, 2013). At 5,000 points, there is no moderation involved. There does not appear to be a set number of points per contribution (Poet, 2013). This Wiki Points system is the core of a gamified structure at CV.
Responding to another user’s question in a forum post, moderator “Morpheus_” explained the mechanics of the gamified system as such: \Since the site operates primarily as the means to present news about comic books, or anything comic related, as well as being an encyclopaedia for everyone to search, then, by default, the wiki is, a, by far, more important aspect to it than a single forum. Both are enjoyable, in different ways (Morpheus_, 2011).
Thus, the Wiki Points system can encourage crowdsourced data generation from CV’s community. In crowdsourced platforms, whether paid, gamified, or volunteered, user involvement is essential as adding contents and verifying them still require humans because of domain expertise and the contextual nature of each platform (Sarasua et al., 2015). Gallus (2016) notes that awards can increase the contributions of users to a wiki.
Many editors and community managers worked directly at CV to fill in data on comics. Their contributions appear next to that of other users as regular contributors, even though they were paid employees. They also gain points. Thus, older editors such as Tony Guerrero who was at one time an employee of the CV has amassed Wiki Points much like other contributors. Enabling Wiki Points allows past or current CV employees to amass prestige and build their reputation. This reputation rests partly in direct exchanges with other users but also through an impressive contribution record that enhances the legitimacy of edits and contribution on the platform and all contributors.
The gamification scheme used by CV does not only foment competition between visiting users for the most amount of contribution, thereby enabling the prestige and reputation on this semantic media. It artificially promotes employees to display openly their contributions. Gaved et al. (2006) describe three types of wiki contributions. They are placeholders which are bare entries created to cover a topic; completers which are placeholder entries further expanded by contributors; and housekeepers that are entries focused on editing and improving existing entries (Gaved et al., 2006). Each of these types are available on CV, but there is no indication as to whether Wiki Points awarded by editors mimic a weight system based on Gaved et al.’s model. Clearly, editors perform much of the housekeeping work on CV, since junior contributors (under 1,000 points) have their work moderated before they are posted.
Discussion: ADD
Following the walkthrough of CV’s semantic architecture, I will discuss what was found and further the ADD framework. Standards in semantic data often push toward common grounds and exchange parameters. There are frequent arguments pushing toward one standard coming from different actors. One of them could be to offer one standard that allows interoperability between systems, to facilitate the documentation of knowledge for developers and their users. This is what proponents of Schema.org, the semantic data scheme underpinning Google’s Knowledge Graph (Table 1), argue (Iliadis et al., 2023; Zaino, 2012) and defend. For other actors, such as IndieWeb (IndieWebCamp, 2022) and the fediverse (King, 2021), concerns about privacy and data commodification prompt their semantic data schemes to allow user data portability.
While CV offers an application programming interface (API) allowing developers to pull semantic data from the platform, it is not well publicised or necessarily compatible with existing semantic schemes from other large players. Such heteroclite architecture, while not favourable to easy exchange between platforms, highlights diversity of design and eschews the one-design-fits-all perspective promoted by dominant standards such as Google’s Knowledge Graph, Wikimedia, and others. Furthermore, I argue that ADD contributes to problem-solving challenges in semantic data schemes, even in the case of proprietary structures because of its ability to serve niche communities of practices, such as comic collectors, adequately. When it comes to semantic media standards, Paulheim (2017) suggests that while the construction of such architecture should benefit from the expertise of its original community of practice, enriching an existing standard can benefit all other knowledge graphs being developed.
The ADD concept introduced in this article focuses on highlighting divergent technical schemes in the computing sciences that do not rely on a few standards as defined by Bowker and Starr (2000). Standards, according to Bowker and Starr, are necessary in the production of material and textual objects; their usage is beyond a particular community of practice; and finally, they ensure interoperability between different information systems (2000). While data-exchange APIs are available, CV’s architecture, layout, classes, and categories are configured for the unique printed comic publishing experience. It is a specialised database that offers answers to comic readers the way no other competing product can with basic search tools. It is so specialised that it is not equipped to handle neighbouring comic formats such as comic strips and web comics. Yet, its format is suitable for media such as published games and serial television series and film.
As an exchange platform, the semantic data cataloged by CV encourage third-party developers to adopt compatible architectures but, at the same time, does not allow for mass data dump and export into other schemes such as Google’s Knowledge Graph and MediaWiki (Table 1). The information and data contained at CV are useful, primarily for its users and particularly for print comic readers. While such “limited” semantic architecture may be criticised for its lack of sharing and openness, it does serve the specialised needs of its community of practice better than competing products from Wikimedia (Table 1) and competing platforms such as the Internet Movie Database (IMDb) (Table 1). When presenting serial media from animation, television, and film, CV’s format offers better user experience than IMDb and a less-generic layout than Wikimedia. CV also compares well against competing printed comics-related semantic media platforms such as StachMyComics.com (Table 1), the GCD, Comic Book Database (Table 1), Lambiek’s Comiclopedia (Table 1), Don Markstein Toonopedia (Table 1), BDGest, or Cover Browser (Table 1).
While CV brands itself as a wiki, its interface and architecture are unlike that of the WikiMedia-based sites. It was developed to feature serial contents—comic books—and not much else. Only web comics and comic strips republished as comics can be featured on the platform. Competing semantic media such as Wikipedia, Comiclopedia, or Toonopedia do not have such constraints. Web comics and comic strips are also often serial works but of a different kind. In both, there is no standard reading unit, beyond the single page (Saint-Louis, 2021). Perhaps CV could detail individual web comic pages and strips, but the published unit would be much smaller than the usual 32-page comic book or the 50 plus graphic novel.
There is no direct way to import data from other semantic sources into CV, even though its database structure is MySQL, a common solution with Apache-based web sites. A developer wanting to import a MySQL data dump from CV would have to rework some semantic classes to use them for another platform such as Google’s Knowledge Graph. CV, being comic book-centric is probably different enough from the librarians’ machine-readable cataloging records (MARC)’ standard (Culbertson & Jackson, 2016; Holloway et al., 2021). This kind of ADD is good because it benefits comic book readers, collectors, and fans first. These form the community of practice that consumes and contributes contents to CV. Schema.org uses many of the same categories to describe comic issues, but its classes go beyond the usual comic-related semantic data by focusing on metadata useful for its own standard. From a practical side, Schema.org’s scheme is less user-friendly than that of CV.
Rafaeli and Ariel (2008) offer a synthesis of many of the challenges that wikis have with contributor participation, such as what motivates people, status from posting often, the types of users who participate, and the skewed weight of contributions from a minority of users versus all visitors using the platform. One major concern that they mention is that many contributor-driven online communities cannot survive due to a lack of contributions from motivated users (Rafaeli & Ariel, 2008). They also argue that there needs to be a value-added benefit and strong thematic fit for users to remain on semantic platforms (Rafaeli & Ariel, 2008). Anything can discourage potential contributors. ADD as presented thus far is focused on semantic media and the differentiation of knowledge graphs as a structural value.
CV is no longer being developed actively with new features even though it still has an active bug report system. It is at best maintained as legacy code attached to GameSpot. Its main value is its aggregation of data that is wide and still used often by comic readers, collectors, and fans. Contributions remain notwithstanding the lack of development of the platform. Its interface is also less complicated than the usual WikiMedia wiki which is used by the parent company Fandom with other properties. As noted earlier, how comics were documented was often contradictory to librarians’ practices because at first, comics were shunned by official sources such as the Library of Congress. Thus, collectors and fans had to fill in knowledge gaps that were not necessarily shared publicly by librarians and archivists at comic publishers.
Conclusion
The viability of CV as a semantic platform featuring ADD is uncertain. The platform has been sold so many times and linked with GameSpot for several years as a subcomponent of the latter, instead of being a peer platform. Again, in October 2022, it was acquired by Fandom, Inc. There has been no visible back-end work done to enhance the serial publishing expertise of CV for GameSpot even though registered user accounts are shared across both platforms. GameSpot still covers games, a media sector that is much more popular than comics.
While it is unclear if CV would exist without its association with GameSpot, its semantic data attract the attention of many non-comic fans who flock to its platform to learn about superheroes and similar comic-centric properties. CV’s semantic architecture was created for comic readers, collectors, and fans and is thus not easily compatible with other semantic web schemes. In future research, it would be insightful to explore how CV contributors navigate the platform and compare their user experience with contributing similar contents on Wikipedia or other competing wikis. This could aid in determining if the ADD of CV is an asset or not. Finally, testing the compatibility of semantic data and standards spearheaded by CV with standards such as Schema.org or Wikidata would be worthwhile endeavours.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
