1. Introduction
[1.1] In the last few years, the pairings of Jane and Maura on Rizzoli & Isles (TNT, 2010–16) and Clarke and Lexa on The 100 (CW, 2014–) have contended for favorite F/F couple or favorite television couple on sites such as AfterEllen and Zimbio (note 1). Yet they have also been commonly cited as examples of queerbaiting, a charge that has grown louder against various shows and has been addressed on both mainstream and LGBT-oriented sites (Bridges 2016; Gennis 2014; K. 2013; Langfelder 2016; Rose 2013).
[1.2] I use the term queerbaiting to refer to situations where those officially associated with a media text court viewers interested in LGBT narratives—or become aware of such viewers—and encourage their interest in the media text without the text ever definitively confirming the nonheterosexuality of the relevant characters. However, I approach its application to Clarke/Lexa, a pairing of two canonically queer women, not as an incorrect or overly broad use of an already defined term, but as evidence of a conceptual motility that invites further consideration.
[1.3] While there has been much discussion about the significance of producer/viewer engagement to fan cultures (Shefrin 2004; Jenkins 2006; Stein 2015), as well as fan-queered readings and writings of mainstream media (Bacon-Smith 1992; Hellekson and Busse 2006; Jenkins 1992), I identify queerbaiting as a phenomenon emerging from the increased significance of media paratexts (Gray 2010), especially promotional content and producer commentary, and what I call queer contextuality, referring to how both the current and previous landscapes of LGBT media narratives inform evaluations of particular texts. The analytical model I propose is applicable not just to both Rizzoli & Isles (R&I) and The 100; it outlines a more general space of textual and paratextual interpretation that can account for the historical contingencies productive of both queer subtext and queerbaiting (note 2).
[1.4] In what they both do and do not challenge as queer enough, queerbaiting discourses can be read for broader tensions between more and less progressive energies in fan cultures. Still, as viewer critique, they represent a politically significant moment in a longer tradition of fan activism for LGBT story lines as early as Star Trek (NBC, 1966–67) (Tulloch and Jenkins 1995) and in media content worlds more generally (Brough and Shresthova 2012). Furthermore, queerbaiting discourses address the character of producer/viewer interactions, and for queer female narratives in particular, serve as a touchstone for greater mainstream recognition of the subordinate status of sexual minorities and women in both representational and real-life contexts.
2. Reading intertextually: Paratextuality and queer contextuality
[2.1] The complexity of reading texts has been a cornerstone of both fan studies and queer media studies: despite authorial intentions, multiple readings are available, and the character of televisual texts makes them particularly open to interpretation. Following Stuart Hall's (1973) Encoding/Decoding, Fiske's (1987) influential model distinguished between preferred readings, which are intended by the producer, and alternative or resistant readings, where audiences read against the grain.
[2.2] While queer readings have not traditionally been the preferred readings of mainstream media, the subordinate status of such noncanonical interpretations has been challenged from two angles. First, in his analysis of popular media, Doty (1993) rejected a hierarchical ordering of readings, arguing against considering queer interpretations as simply "sub-textual, sub-cultural, alternative readings, or pathetic and delusional attempts to see something that isn't there" (xii). Instead, he maintained, they should be seen as always present elements of putatively "straight" narratives of popular culture. Because of this, Doty's work is part of queer studies approaches pointing to how the queer is constitutive of, rather than simply marginal to, the heteronormative (Butler 1990, 1993). Second, despite the availability of queer interpretative challenges, scholars have argued for the importance of viewers having canonically LGBT representations rather than being confined to excavating queerly resistant readings (Jenkins 1995; Gross 2001). As I will demonstrate, both of these threads entwine in discourses on queerbaiting.
[2.3] In terms of the mechanics of reading texts, Fiske (1987) identified three codes that structure meaning—reality, representation, and ideology. "Reality" is constructed through shared sociocultural understandings of what is signified by physical appearance and by verbal and nonverbal communication; visual and auditory aspects conveyed by technical elements (e.g., camera work, lighting, sound, and music) compose the representational level; and a third, more complex level of ideological coding (for concepts such as "individualism," "capitalism," and "patriarchy") is organized from the other two types of code. Textual polysemy often involves interpretative contradictions, where one reading implies narrative facts that are incompatible with those of another reading. However, contradictory readings emerge not simply through properties of a text itself, but through additional content that viewers bring to their readings; this is the phenomenon of intertextuality.
[2.4] In media studies, intertextuality often refers to viewers drawing on other texts of the same kind; reading the rebooted Battlestar Galactica (SyFy, 2004–9), for example, by referencing the original series, other sci-fi shows (such as Star Trek), and so on. However, as Gray (2010) discussed, media texts are also associated with various paratexts both from official sources, including prequel and sequel films, studio promotional materials, tie-in toys and games, and deleted scenes and producer commentary on DVD and Blu-ray releases, and from fan-produced texts.
[2.5] Gray did not consider in any detail producers' commentary about their texts, a genre that I call producer paratext. Its impact has increased with the growth of social media and the move of entertainment conventions such as Comic-Con into mainstream culture. However, Gray noted that official paratexts do not always communicate a uniform message, in part because promotional content is the purview of marketers rather than producers. Another kind of official paratext (not mentioned by Gray) is the gag reels commonly included on DVD sets, which constitute a hybrid product of the marketing and creative teams; although the reels are put together by the studio, actors are aware that what they are filmed doing on set may end up on those reels, and this content is therefore also a channel through which producers communicate with viewers. Such a range of sources for official paratexts makes them—like texts—polysemous.
[2.6] Media paratexts make up a transmedia environment in which both producer and consumer participate (Jenkins 2006), examples of which have been identified for various shows, such as Heroes (NBC, 2006–10) (Mann 2009), Lost (ABC, 2004–10) (Mann 2010), and Glee (Fox, 2009–15) (Marwick, Gray, and Ananny 2014; Stork 2014). Producer involvement with viewers derives in part from the greater precariousness of television production in a postnetwork era of audience fragmentation (Lotz 2007), and the proliferation of "quality" scripted programming (Carr 2014; Leopold 2013) is also an impetus for producers to heighten the appeal and profile of their shows through paratextual engagement.
[2.7] Queerbaiting discourses highlight the fact that such engagement around queer content or readings has implications specific to viewer experiences of LGBT media narratives more generally, which make up what I have labeled queer contextuality. There are two elements in counterpoint. On the one hand, there are now unprecedented numbers of LGBT characters across network, cable, and streaming sites (GLAAD 2015), reflecting the pursuit of niche audiences. In the 1990s, gay and lesbian narrative elements were initially targeted at affluent, predominantly white, and progressive straight viewers attracted to programming that they saw as hip and edgy (Becker 2006), but the industry also began explicitly courting LGBT viewers. For example, Sender (2007) discussed how the cable channel Bravo began "dualcasting" in the early 2000s, seeking to attract a mix of straight women and gay men, and Himberg (2014) argued that the use of lesbian content reflects current network strategies of "multicasting" to appeal to an even broader mix of audiences. On the other hand, there are still few shows with queer female characters as leads, and there remains a cumulative representational deficit across the time period that many viewers have been consuming media. In addition, a disproportionate number of queer characters continue to meet untimely ends (Riese 2016), a residue of the Hays Code's stipulation that homosexuality could be depicted only in an unappealing or negative fashion (Russo 1987). Queer contextuality, then, informs how viewers assess (1) the validity of reading queerness in a text, (2) the political and economic feasibility (particularly in regard to studio and network financial considerations) of having a canonical LGBT narrative, and (3) the quality of the canonical LGBT narratives that are produced.
[2.8] Queerbaiting is the outcome of increased paratextual discourse about LGBT content at a specific moment of queer contextuality. We might be tempted to reserve the term, in the sense that I use it, for media texts that fail to have canonically queer characters despite textual and paratextual content that suggests the possibility. However, doing so would not explain why The 100, with a canonical F/F couple, has also been widely cited as an example of queerbaiting. The crucial element is not a lack of canonicity, but how satisfactorily queerness plays out in the canonical text relative to viewer expectations that emerge from the reading of multiple texts and paratexts and that take account of queer contextuality. That is, queerbaiting's referents expand because the text–paratext–queer contextuality matrix changes over time, although its structure remains the same.
[2.9] I represent these elements in figures 1 through 4, which provide a diagrammatic way to map out different media texts relative to each other, and to situate queerbaiting in relation to other interpretative possibilities. The figures do not imply that a text can be definitively classified as queerbaiting or not; indeed, what counts as queerbaiting is often contested (note 3). What the model presents is a schema for how a particular text comes to be seen as queerbaiting by at least some viewers. The vertical axis represents viewer expectations of canonical queer content, while the horizontal axis represents textual content on a spectrum from subtextual to canonical (figure 1).
Figure 1. Text–paratext–queer contextuality matrix, basic elements A. [View larger image.]
[2.10] Intertextual readings of textual content and paratextual content inform viewer expectations of canonical queer content (figure 2). In turn, intertextual readings are informed by queer contextuality, and queer contextuality also informs viewer expectations of canonical queer content independent of any one media text, as well as assessments of the quality of canonical queer content (figure 3).
Figure 2. Text–paratext–queer contextuality matrix, basic elements B. [View larger image.]
Figure 3. Text–paratext–queer contextuality matrix, basic elements C. [View larger image.]
[2.11] Figure 4 additionally labels three different spaces: the realm of queerbaiting (shaded red), where viewers have high expectations for quality canonical queer content, but consider the quality of the actual queer content to be low; desired queer representation (shaded green), where the quality of canonical queer content matches viewers' high expectations; and queer subtext (shaded blue), where queer content is only suggested, not canon, but viewer expectations of having canonical queer content are low. I return to these diagrams after examining the examples from R&I and The 100, but first, provide a snapshot of the shift in viewer practices from reading subtextually to identifying queerbaiting that underscores the historicity of queerbaiting's emergence.
Figure 4. Text–paratext–queer contextuality matrix, with queer subtext, queerbaiting, and desired queer representation areas situated. [View larger image.]
3. From subtextuality to queerbaiting
[3.1] While media producers have never prevented viewers from reading against the grain, the norm used to be not to acknowledge, let alone encourage, queer interpretations of canonically heterosexual characters and narratives. Breaking from this, Xena: Warrior Princess (USA Networks and syndication, 1995–2001) was one of the earliest mainstream programs where content strongly suggestive of same-sex romantic or sexual interest, between Xena and her companion Gabrielle, was both intentionally included and publicly acknowledged by producers (Maris 2016). Although it aired well before the concept of queerbaiting that I use here emerged, Xena was also read intertextually, which included taking account of the canonical text, official paratexts, and queer contextuality.
[3.2] In a 1996 interview, executive producer Liz Friedman noted, "We opened up [a season 2 episode] with the two of them fishing naked, and we're about to have a Halloween episode that will certainly have some nice moments for our queer fans, a little lesbian vampire show" (Friedman 1996). Lucy Lawless (Xena) also commented suggestively in another interview that Xena's fantasy vacation would be "a biennial sailing trip to Lesbos" (B. 2003). Such comments occurred in the context of both Xena and Gabrielle having occasional male love interests, and a sizable audience segment who preferred Xena to be romantically paired with the god Ares. While the protagonist of Ellen DeGeneres's sitcom Ellen (ABC, 1994–98) had come out in a fourth-season episode, the show ended the next season, and there were no other US shows with queer female leads. Despite dissatisfaction about the continuous "will they/won't they" tease, then, a large proportion of Xena/Gabrielle fans accepted that subtext was as far as Xena the show could go, and fanfiction would be the domain for explicitly romantic narratives (Hanmer 2014).
[3.3] A decade or two later, a number of shows are being referenced as examples of queerbaiting, although the concept has only begun to attract in-depth academic analyses, primarily of male/male pairings. As Collier (2015) and Fathallah (2015) discuss, the canonical text of Sherlock (BBC and PBS, 2010–), which has a large "Johnlock" (John Watson/Sherlock Holmes) fandom, has included various nods to Watson and Holmes as a couple, including other characters assuming that they are together romantically and interactions between the two suggestive of attraction or flirtation. One episode shows Holmes about to kiss another male character, Moriarty, before revealing that the moment was being imagined by a female fan. Collier (2015) identifies the show as an example of queerbaiting primarily because of such textual elements, even as show producers have sought to shut down Johnlock readings. In contrast, while Fathallah (2015) recognizes that various scenes could be seen as queerbaiting, she argues that the text still provides "queer disruption" by offering comments by Watson or Holmes that "can be read as ruptures in the performance of heterosexual masculinity" (491), as well as through the campy villainry of Moriarty, whose initial reading by Sherlock as gay and later narrative establishment as straight provide "a 'hauntology' of queerness" (492).
[3.4] Collier (2015) also discusses Supernatural (WB and CW, 2005–), many fans of which favor the M/M pairing of Dean and Castiel (Destiel), as an example of queerbaiting. As in Sherlock, there are many textual elements that invite viewers to read the characters' relationship as romantic, including each one's importance to the other, scenes of emotional intensity and physical intimacy, and the failure of both to deny that they are a couple when other characters assume it. Furthermore, the Supernatural creative team has supported fans' queer readings in producer paratexts on social media, in interviews, and at fan conventions, albeit only to a degree. Collier describes how producers of both shows seek to assert ultimate control of the narrative, whether through episode content or through public interactions with fans.
[3.5] Collier's work aside, there has been little discussion of producer paratexts and queerbaiting; in fact, besides Gray (2010), most discussions about the significance of media paratexts concern those that are fan-produced, although not all explicitly theorize them as paratextual (those that do include Fathallah 2016; Leavenworth 2015). Nordin (2015) notes, without exploring further, that fan discourses of queerbaiting sometimes reference official paratexts such as trailers and promos, which are read as a promise of what the text could deliver.
[3.6] Cavalcante's (2013) analysis of TransAmerica (2005) does not address queerbaiting, but is pertinent in discussing how paratexts do "double work." He examines how the film's reviews, posters, and DVD commentary sought to corral readings about LGBT subjectivity, a phenomenon that has also been noted for other films, such as Fight Club (1999) (Brookey and Westerfelhaus 2002) and Brokeback Mountain (2005) (Cooper and Pease 2008). However, Cavalcante argues that despite an intent for "paratextual domestication" to "subordinate transgender themes and advance certain reductive tropes and stereotypes," the paratexts also "invite wider cultural conversations about sexuality and gender, and include transgender individuals within the national family" (88).
[3.7] Neither textual nor paratextual content alone can determine the meanings of media texts, but these accounts illustrate the significance of paratexts for reading texts. That is, the queerly polysemous is not simply a matter of textual character—the queerness that Doty (1993) and others have argued inheres in popular texts—or of reception—the ability of viewers (including scholars) to read queerly. Rather, text, paratext, and queer contextuality are all at play, and it is their specific configurations that determine how queerbaiting discourses arise for any particular show, as I illustrate in analyses of R&I and The 100.
4. Methodology
[4.1] I discuss textual and paratextual material in R&I and The 100 that has been particularly important for romantic readings of Jane/Maura and Clarke/Lexa. For the textual content of R&I, Fiske (1987) provides a useful model for outlining relevant codings of the characters' appearance and communication, technical filmic conventions, and ideological codings of family and intimacy. Televisual semiotic conventions are also relevant to Clarke and Lexa's scenes on The 100, but since their relationship was canonical and viewers therefore did not need to read a romantic bond through the usual "subtextual" methods, I primarily outline the key narrative developments.
[4.2] With both shows having a large amount of paratextual content, I focus on material that is most pertinent to queerbaiting discourses. These include producer paratexts from show runners, writers, and actors that are representative of their comments about the relationships in question (note 4). For R&I, I also discuss promotional paratexts for upcoming seasons and segments from the DVD blooper reels.
5. Textual content: Rizzoli & Isles
[5.1] Rizzoli & Isles, coproduced by Warner Horizon Studios, Hurdler Productions, and Ostar Entertainment, premiered on Time Warner's TNT network in 2010, and wrapped its final, seventh season in 2016. Jane Rizzoli, a homicide detective, and Maura Isles, chief medical examiner for Massachusetts, work together in a Boston police department. Jane's mother and brothers are often present, while Maura is not usually shown to have her own familial unit independent of Jane and her family. Already friends when the show begins, both women have had several male love interests since season 1, but there is a variety of textual content supporting a romantic reading of their relationship.
[5.2] The contrast between the two characters' appearance and demeanor evokes a butch/femme dynamic, especially combined with Jane's repeatedly protecting Maura from danger. Jane, a tomboy from childhood, favors pants, T-shirt, and blazer, while Maura is typically clad in conventionally feminine fashion. Furthermore, the show has acknowledged both that Jane might be seen as a lesbian (in 1.02, an angry interviewee calls her a "a skinny greaseball dyke detective") and that people may assume or believe Jane and Maura to be a couple (1.06, 2.03, 5.04). The two have discussed the possibility of being attracted to women (1.06), and Jane has asked Maura, somewhat jokingly, to confirm that she is not attracted to her (2.03). They have frequently expressed physical and verbal affection and had emotionally revealing conversations, on two occasions while lying on a bed together (1.01, 3.08).
[5.3] As for technical elements, multiple scenes have been framed, lit, and scored to code emotional and physical intimacy between Jane and Maura. For example, in 1.09, as the two sit inclined toward each other on a park bench having an emotionally intense conversation, there are close-ups of their faces, as well as of Jane's hand caressing Maura's knee, and the score marks the interaction as emotional and intimate.
[5.4] In terms of ideological codings for family and intimacy, Jane's mother, who lives in Maura's guesthouse, interacts with Maura like she is another daughter. Jane is also often over at Maura's, although Jane's family members are often present as well. These scenes of domesticity therefore have some similarity to depictions of married couples, although the family dynamic also undercuts a romantic connection that is more readily available in scenes with just Jane and Maura. A significant narrative parallel occurs in a plotline involving serial killer Charles Hoyt, whose modus operandi is to restrain the man of a heterosexual couple and force him to watch Hoyt assault and murder the woman before also being killed. Hoyt's treatment of Jane and Maura significantly mirrors this; he threatens to rape Maura when she interviews him in 1.08, while his main goal regarding Jane is to kill her, and after he escapes in 2.10 and captures the two women, he attempts to kill Maura while forcing Jane, who is restrained, to watch.
6. Textual content: The 100
[6.1] The 100, coproduced by CBS Television Studios, Warner Bros. Television, and Alloy Entertainment, premiered on the CW network in 2014, with season 3 airing in 2016. In a postnuclear apocalyptic world, Clarke Griffin, the lead character, heads a group, the "Arkers," who establish a settlement called Arkadia, while Lexa is the commander of the "Grounders," with whom the Arkers find themselves in conflict at the beginning of the show. The narrative suggests that Clarke would be categorized as bisexual and Lexa as a lesbian in modern identity parlance, although these terms are not used onscreen.
[6.2] Clarke and Lexa meet in the middle of season 2 while negotiating a truce between their two forces. The two grow closer after agreeing to fight a common enemy, during which Lexa tells Clarke about her last lover, a woman who is now dead. Over several episodes, Clarke and Lexa share scenes of emotional intensity and physical proximity. Ultimately Lexa initiates a kiss, which Clarke reciprocates briefly before apologizing that she is not yet ready to be with anyone. They part ways at the end of season 2 after Lexa makes a tactical decision to abandon Clarke and their military alliance. As season 3 opens, Lexa has Clarke brought to the city of Polis, from which Lexa now rules, for Clarke's protection. Initially furious at Lexa, Clarke rebuilds her personal and professional bonds with her over the next episodes as they work together again.
[6.3] In 3.07, Lexa invites Clarke to stay in Polis as her guest despite a new injunction that no Arkers be permitted within Grounder territory, but Clarke reluctantly decides that she must return to Arkadia to deal with problems there. When she goes to say goodbye to Lexa, they kiss and make love for the first time. After Clarke returns to her room, one of Lexa's advisers, opposed to her closeness to Clarke, shoots at Clarke, but when Lexa rushes into the room, his bullet strikes her instead, and Lexa dies with Clarke trying in vain to save her.
7. Paratextual content: Rizzoli & Isles
[7.1] As I noted earlier, paratexts are subject to multiple interpretations since they come from more than one source; additionally, the same person may say different things at different times or be intentionally ambiguous. Therefore, I cite the content below not as evidence of the "true" intentions of producers, but as key paratexts that have been read alongside textual content in discussions about queerbaiting.
[7.2] R&I producers have commented that they did not intend to depict sexual tension between Jane and Maura, and stated or strongly implied that the pairing would never become romantic, but they have also revealed an intent to encourage the hopes of Jane/Maura (or "Rizzles") fans. Janet Tamaro, show runner for seasons 1 through 4, noted that "the lesbian theory endlessly amuses me, and it amuses the cast," but "Rizzoli and Isles have been heterosexual from the first episode" (Hochman 2011). Executive producer Jan Nash stated that the two are no more than "each other's best friends" (Bendix 2014), and was quoted in TV Guide as saying that Jane/Maura was "not something that will be in the cards," in a paragraph-long quotation captioned "No Rizzles…ever." However, that part of his comment was soon deleted, and the subheading changed to "No Rizzles this season?" (Eng 2015) (note 5). Although neither the change to the text nor the reasons for it were acknowledged, a likely reason is that the original article too categorically shot down any chance of a Jane/Maura relationship, so the text was edited to make it more ambiguous.
[7.3] Angie Harmon (Jane) and Sasha Alexander (Maura) have also stated that they see the relationship between their characters as only friendship, although they recognize their onscreen chemistry and understand why viewers imagine them as a couple (Snarker 2012). Still, Alexander has suggested that Maura might be open to same-sex encounters, commenting that "if Jane were open to it, I think Maura would absolutely experiment because she's just a little bit more open-minded" (Anderson-Minshall 2012). Furthermore, Harmon acknowledged that the show has intentionally included lesbian innuendo: "Sometimes we'll do a take for that demo…I'll brush by [Maura's] blouse or maybe linger for a moment" (Hochman 2013).
[7.4] The awareness of "that demo" (i.e., the lesbian demographic) is also evident in gag reels, which show Harmon and Alexander playing with explicitly romantic tropes. An outtake of the park bench scene I described earlier zooms in on Harmon and Alexander caressing each other's thighs, and then shows Harmon leaning in as if Jane is about to kiss Maura, before she is stopped by a giggling Alexander. The season 1 gag reel also includes an outtake of another emotionally charged scene, in which Harmon ad libs, "Kiss me," and Alexander leans toward her, eyes closed and lips puckered, responding, "Just kiss me." Outtakes like these offer almost-textual possibilities for the episodes, since they are so close to what is seen on the show. Thus, while being dissonant with actual episodes, these paratexts are nods to fan wishes in teasing what canon could in theory look like. In addition, in the season 3 gag reel, Harmon fakes a kiss with Alexander, and then, striking a couple of suggestive poses with her, turns to the camera and asks sultrily, "Like that? Like this?," again pointing to the actors' awareness of viewer desire to see Jane and Maura together.
[7.5] Paratexts from TNT's marketing have also been suggestive, hinting repeatedly—though not exclusively—at a romantic interest between Jane and Maura. In a season 2 promotion (https://www.youtube.com/watch?v=lNfU0wez4mY), Harmon comments that "[Maura]'s funny, and annoying, and they [Jane and Maura] fight," and the clip then cuts to Maura saying, "That's a sign of sexual frustration!" Another season 2 promo (https://www.youtube.com/watch?v=czKCM3GSsHw) shows Jane and Maura at a speed-dating event, having no luck connecting with a string of men and finally ending up at a table together. Jane remarks, "We should do this again sometime," Maura agrees, and the two toast to that as the voice-over says, "Rizzoli and Isles—a perfect match." In a season 3 promo (https://www.youtube.com/watch?v=CZweX1vZNo4), a tight shot of Jane and Maura huddled together is followed by a voice-over comment by Jane that "opposites attract" during a slow-motion shot of the two staring at each other while walking, and then more clips of Jane and Maura interacting affectionately. A season 4 poster shows the two handcuffed together with the tagline "Bound for Life." And for season 5, one promo (https://www.youtube.com/watch?v=ZIXfDHHltrE) intercuts the text "Behind every great woman is another great woman" with footage extracted from 1.06, when Maura tells Jane that various women thought that Jane was hot in her (fake lesbian) dating profile, and Jane responds, "I might flip for that!"; from 2.03, when Jane asks Maura, "You don't want to sleep with me…do you?"; from 5.04, when Maura tells some men that Jane is a lesbian to stop them from hitting on her; and from 2.13, when Jane and Maura snuggle as Jane discourages a man from pursuing Maura by saying, "Yes, we are still…together."
8. Paratextual content: The 100
[8.1] Whereas the paratexts for R&I pertinent to queerbaiting discourses suggested a romantic relationship that remained noncanonical, those for The 100 involved producers courting viewers invested in seeing a F/F relationship, and Clarke/Lexa fans in particular, through several strategies (Ryan 2016b; http://wedeservedbetter.com/).
[8.2] From late season 2 onward, producers and writers highlighted the distinctiveness of The 100's treatment of same-sex relationships. After Clarke and Lexa first kissed in season 2, The 100 was written up by several outlets for its "brave," "progressive," and "groundbreaking" approach to sexuality (Murphy 2015; Peitzman 2016; MacDonald 2016), and in retweeting the links, show runner Jason Rothenberg also reiterated these descriptors. Javier Grillo-Marxuach, one of the show's writers (credited with the episode in which Lexa was killed), also tweeted before season 3 began that viewers looking for a "truly progressive approach to LGBT relationships" should try watching The 100.
[8.3] Producers were also encouraging when responding to fan questions about Lexa and her relationship with Clarke. Since Alicia Debnam-Carey (Lexa) was not a series regular, there were fears that Lexa might be killed off even in season 2, and many fans who tweeted the creative team about this noted that death was a common fate for television lesbians. One early reassurance came in March 2015, after Lexa's betrayal of her alliance with Clarke had aired; Layne Morgan, a personal assistant on the show, tweeted that while "half the Internet" was crying that "Clexa [Clarke/Lexa] is dead," she was thinking, "Clexa has begun," and Kim Shumway, a show writer, replied "Yes" to this tweet. Another example came in December 2015, when a fan asked whether there was "hope for Clexa," and Rothenberg tweeted back, "Always."
[8.4] Twitter communication from the creative team also strongly suggested that Lexa would live through season 3. In January 2016, Rothenberg tweeted that they were shooting the season 3 finale in downtown Vancouver; "Come say hi!" Answering the invitation, fans took and circulated photographs of Eliza Taylor (Clarke) and Debnam-Carey filming together. Another cast member, Henry Ian Cusick, also tweeted that he was with both Taylor and Debnam-Carey on set during the finale's filming. In addition, Rothenberg did advance promotion for episode 3.07 on the lesbian entertainment site AfterEllen, which was granted an advance screener and an exclusive interview (Valerie Anne 2016); the coverage was clearly intended to generate excitement rather than apprehension.
[8.5] More generally, the producers have expressed support for LGBT issues and viewers. For instance, Rothenberg tweeted a behind-the-scenes photo of Taylor and Debnam-Carey with rainbow-colored candy (with his comment being "and yes, they chose RAINBOWS"), and writers retweeted a ClexaForMe Twitter tag that trended in December 2015, which was used to share positive comments about the Clarke/Lexa story line, as well as the importance of media representations showing that LGBT characters could be normal, strong, and happy, and of LGBT issues more broadly.
9. Discussion
[9.1] Intertextuality is central to the interpretation of media texts, which, as Gray (2010) commented, involves a "complex hall of paratextual and intertextual mirrors through which meaning and reception must pass" (119). The particular development of queerbaiting is due to changes in the (para)texts that are read intertextually, and the queer contextuality of when they are read.
[9.2] Figure 5 maps out R&I and The 100, as well as Xena: Warrior Princess, in the representation of interpretative space I presented in section 2, showing how it accommodates both current differences between shows and changes over time. R&I is similar in various respects to previously discussed examples of queerbaiting, such as Sherlock and Supernatural: heightened expectations for a Jane/Maura pairing developed from intertextual reading of texts and paratexts, but the consistent lack of definitive canonicity located the show for many viewers in the realm of queerbaiting. For The 100 prior to Lexa's death in 3.07, viewer expectations for canonical queer content were also high, because of both textual and paratextual content, and since viewers considered its content high quality, the show fell in the realm of desired queer representation. After Lexa's death, viewers assessed the show's content poorly, shifting the show into the domain of queerbaiting. For Xena, viewer expectations of Xena/Gabrielle becoming canon were relatively low, partly because there were fewer official paratexts suggesting this possibility, and partly because industry constraints of the time pointed to the low likelihood of such an outcome. The Xena/Gabrielle relationship therefore fell, like same-sex pairings for many shows, into a realm of queerly subtextual readings, although, were this incarnation of Xena to air today, it would likely be accused of queerbaiting (note 6).
Figure 5. Text–paratext–queer contextuality matrix, with examples of show mappings. XWP indicates Xena: Warrior Princess; R&I, Rizzoli & Isles; and W13, Warehouse 13. The 100 is divided into the show before and after episode 3.07. [View larger image.]
[9.3] The complexity of reading meaning in texts has long been acknowledged; it is complex also in paratexts. As I noted earlier, Cavalcante (2013) argues that official paratexts could not fully circumscribe TransAmerica's capacity to trouble normative discourses of gender and sexuality. In the case of queerbaiting, it is not so much that paratexts do double work, but that paratextual claims to authoritative textual meaning cannot escape becoming part of the matrix of interpretation, being read intertextually with the text (and other paratexts) rather than superseding them. Therefore, once a show is seen as queerbaiting for not making a queer narrative canon, producer denials that such a narrative will ever occur cannot extract the text from the queerbaiting discourse. In any case, the many sources and widely varying contents of paratexts make it impossible to pin down any single "official" message. This is especially true of R&I.
[9.4] For R&I, the most relevant official paratexts were promotional content and interviews about the Jane/Maura relationship, while The 100 exemplifies paratextual engagement through social media. However, both these sets of paratexts are indicative and constitutive of the normalization of LGBT visibility in mainstream media. In 2009, it was noteworthy that actor Stephanie March coyly declined to shoot down the possibility that her Law and Order: SVU (NBC, 1999–) character was in love with another woman (Warn 2009), although neither did the creative team explicitly court this pairing's fandom through extensive paratextual commentary. In comparison, multiple R&I promotional paratexts suggested a romantic connection between Jane and Maura, and producers and actors were asked repeatedly about it, while The 100, like various other shows now airing, has explicitly sought LGBT viewers by highlighting the Clarke/Lexa relationship and the show's approach to sexuality more generally.
[9.5] R&I and The 100 also underscore two crucial sides of queer contextuality: what recent and current LGBT narratives suggest is now viable on mainstream media, and how past and existing deficiencies constitute a context of representational deficits that viewers see as requiring improvement. R&I aired on TNT, a basic cable network, and although it is no longer unheard of for cable shows to have queer lead characters, TNT has had no such programming. Since R&I was consistently among TNT's highest-rated scripted programming (Futon Critic 2010), many viewers recognized that neither the show runners nor the network had any financial incentive to change the depiction of Jane and Maura's relationship. In any case, R&I was susceptible to queerbaiting charges whether a canonical Jane/Maura relationship was economically feasible or not. If it was not feasible, then the suggestions were fully misleading; if it was, then the producers were guilty of teasing a narrative they failed to deliver.
[9.6] For The 100, queer contextuality is multilayered. At one level, the show exemplifies recent improvements in queer female representation on US television, particularly with its female lead having a multiseason female love interest. However, as many have noted, Lexa's fate was strikingly similar to that of another lesbian character, Tara, on a 2002 Buffy the Vampire Slayer (WB and UPN, 1997–2003) episode: both were killed right after reuniting romantically with another woman and having their first onscreen love scenes, and each died of a bullet meant for another character rather than for a meaningful or heroic reason. It is true that, within the story world of The 100, Lexa's killer did not object to her love for Clarke because of Clarke's gender, but nonetheless the narrative link between lesbian sex and happiness on the one hand and death on the other reproduced a longstanding trope of LGBT representation. An unprecedented level of viewer discussion produced many accounts situating Lexa's death in the longer history of "Bury Your Gays," including a collaborative cataloguing of lesbian character deaths that began with 65 and now stands at 181 (as of June 12, 2017) (Riese 2016).
[9.7] Developments in official paratexts and queer contextuality are deeply intertwined: the official encouragement of queer fandoms demonstrated by R&I and The 100 has occurred in current contexts of LGBT media representation, while paratextual content is interpreted in the light of queer contextuality. However, viewer response to Lexa's death on The 100 may constitute something of a watershed for producer engagement with viewers about LGBT story lines or readings, with calls by viewers and critics not just for improvements in the treatment of LGBT narratives, but also for new discussions on the ethics of how producers appeal to fans, particularly those who are sexual or racial minorities (Roth 2016; Ryan 2016b). Queer contextuality, then, reflects and constitutes not only changes in media representation, but also recent increases in attention to sociopolitical inequality. The queerbaiting discourse has also broadened into commentary linking a recent string of deaths of female and minority characters to the underrepresentation of these groups among content producers (Ryan 2016a), while a viewer initiative to raise money for the Trevor Project, which serves LGBT youth in crisis, in honor of Lexa connects those criticizing The 100 for queerbaiting with other fan endeavors for social justice beyond the media domain (Hinck 2012; Jenkins 2012; Lopez 2011).
[9.8] Such phenomena do not mean that queerbaiting critiques are informed only by progressive energies; the fact that in such critiques a text is deemed queer enough only when two characters are in a canonically romantic relationship tends to obscure other nonnormative modes of being and relating. For all the male love interests on R&I , Jane and Maura's emotional primacy to each other constitutes a homosocial intimacy that the dominant binaries of heterosexual/homosexual and romantic/platonic evacuate from the actual diversity of human relations (Sedgwick 1990). Their relationship and those of many other same-sex pairs who are not romantically involved in their canon texts can be read as instantiating intimacy outside of the normatively kinned family unit or couple, in forms such as romantic friendships (Faderman 1981), chosen family among lesbians and gays (Weston 1991), and other queered familial configurations (Halberstam 2011). Conversely, canonical LGBT characters and relationships can be problematically normative, as on shows including Ellen (ABC, 1994–98) (Dow 2001), Will and Grace (NBC, 1998–2006) (Battles and Hilton-Morrow 2002), and The L Word (Showtime, 2004–9) (Chambers 2006), and fans are often invested in LGBT narratives that reinforce mainstream discourses about gender and relationships (Ng 2008; Russo 2013; Scodari 2003).
[9.9] However, the purpose of my analysis has not been to evaluate the legitimacy of queerbaiting claims either on the basis of the availability of queer readings of the texts as they stand or on the basis of textual and paratextual "evidence." Rather, the examples I discussed highlight the necessity of considering multiple phenomena to account for the development of queerbaiting discourses, which in turn highlight changes in producer/viewer interactions around queer readings that both reflect and constitute conditions of LGBT representation and industry production more broadly.
10. Conclusion
[10.1] While initial accusations of queerbaiting were aimed only at texts that failed to deliver unequivocally canonical LGBT narratives, changes in queer contextuality and the increased prominence of official paratexts that comment on or allude to queer readings have expanded queerbaiting's scope. These developments highlight the historical contingency of queerbaiting as emergent from intertextual readings of textual content and producer/viewer interactions during a particular period of LGBT media representation—queer contextuality—that inform viewer expectations, which now extend to canonical LGBT narratives.
[10.2] R&I epitomizes how many shows become embroiled in discussions of queerbaiting; despite denials of the possibility by a show runner and some actors, multiple official paratexts teased a Jane/Maura romance that was reinforced by some textual content, even as such a reading was also textually contradicted. For The 100, with Clarke and Lexa a canonical pairing, the key articulations between text and paratext are the creative team's courtship and encouragement of LGBT viewers and Clarke/Lexa fans, counterposed to the narrative characteristics of Lexa's death. Although the specific textual and paratextual elements are different, the same general matrix of text–paratext–queer contextuality that I presented is active in both shows.
[10.3] Viewer demands for particular kinds of canonical narratives reflect a shift away from satisfaction with solely subtextual readings, long a creative mainstay for queer fandoms. However, alternative readings are not simply supplanted by canon, particularly since the scope of fan works always exceeds the quantity and variety of stories and possibilities that a canonical text can depict. As Figure 3 above diagrams, productive interpretative space outside of queerbaiting discourses remains for queer subtext that "hold[s] value and pleasure outside the mass media's limited repertoire" (Russo 2013, 458), even though that may not be the domain in which fans would prefer to see the forms of canonical representation they are pushing for.
[10.4] Another alternative for viewers dissatisfied with mainstream LGBT narratives is independent productions, including Web-based series that depict a greater range of gender, sexual, and racial diversity (Christian 2012, 2014). Still, even in an era of declining linear television audiences, viewership for US network and cable programming indicates the enduring economic and representational significance of popular media, for which they can be expected to attract continued fan attention and critique.
[10.5] Queerbaiting discourse constitutes increased viewer vocalness about both representation and how producers engage with viewers over it, exposing one of the front lines in conflicts over media texts and their meanings, as well as highlighting how current conditions of LGBT representation—which are significantly broader, along several dimensions, than they were a decade or two ago—remain shadowed by deficits both historical and contemporary.