The graph Frank uses, which I stole from the uncanny valley wikipedia page, and the accompanying Star Wars explanation Frank provides illustrates the problem quite nicely – on the far left side, you have R2-D2 and C3PO. On the far right side, Han Solo. The uncanny valley is, of course, Jar-Jar Binks. As a representation draws closer to reality, we are less inclined to accept the representation as a fictional construct that stands-in for real life, and we become more and more distracted by everything that looks wrong about it. The result is a strange but undeniable phenomenon where Mr. Incredible appears more persuasively realistic than the computer generated image of a young Jeff Bridges in Tron Legacy.

I love this idea, and I think it’s applicable for representations of reality outside of the visual. In particular, I find it a useful explanation for a problem of plausibility in narrative, especially as it relates to coincidence and character networks. In my proposed Uncanny Valley of Narrative Plausibility, a movement toward reality maps onto the increasing role of chance in narrative, and the closer one moves toward the valley, the higher the likelihood that meetings between characters or important turning points in the plot appear to happen by accident. The idea is the same as that of the visual uncanny valley: there comes a point where we find plausibility in narrative less persuasive (and maybe just less interesting) than circumstances more patently contrived. Let’s look at some examples.

On the far right side of the graph, of course, we have reality, where chance meet-ups in a bar generally lead to nothing, and that one, totally unlikely time I ran into my college roommate in a New York State rest stop even though neither of us lives in New York is just that: totally unlikely. The role of randomness in our lives is so prevalent that we try desperately to pretend it doesn’t exist by believing in fate, and we so love coincidences that we see them all the time. Most of the time, though, the guy standing in front of you in the line for curly fries will not turn out to be your brother.

On the far left side of the graph are television shows like police procedurals and some soap operas. Coincidence is so prevalent and unavoidable on these shows that it doesn’t even register as coincidence – every character is someone else’s former lover, and every scrap of paper is a relevant credit card receipt for a rare color of automotive paint that just happens to be the exact same paint found at the crime scene. Everything is a clue, every pregnancy test comes back positive, and we hardly even notice how bizarrely significant everything is, because it’s a story. We want there to be clues everywhere! Clues are far more interesting than boring crumpled scraps of paper that mean nothing, and we watch shows like this because we like it when interesting, unrealistic things happen all the time.

In the middle area, things get more complicated, and we come to the reason I stumbled onto this idea in the first place. My two examples here are Game of Thrones and Treme, in part because they both air on HBO Sunday nights, and their proximity invites comparison. Mostly, though, I’m drawing on Treme because I think it’s actually fairly unusual to fall into the Narrative Plausibility version of the uncanny valley, and trying to figure out what’s weird about Treme is what first led me merrily skipping down this path.

Poor Tyrion, victim of narrative happenstance

So first, the not weird – as you move away from the police procedural end of the spectrum, things get bigger, and messier, and often darker. Game of Thrones is a good example, though you could just as easily use any number of critically-acclaimed hourlong dramas (certainly The Wire, but also Friday Night Lights, or The Good Wife, or Justified), and this point about coincidence is easiest to see in shows that have big, intersecting character networks. Game of Thrones, like the novel it’s based on, follows many different character groups (several feuding families, the Night’s Watch, the Dothraki horse people), and as events force groups to split apart and characters to splinter away from their families, the narrative increasingly resembles a map full of potential plot connections passing each other in the night. Inevitably, though, encounters happen, creating a cascade of new narrative possibilities. In one early example, Catelyn Stark is traveling across the country on her way home from a trip to the capitol, and happens to stop for a meal at the same inn where Tyrion Lannister, the man she suspects of attempting to murder her son, has also stopped for the day. When he recognizes her, she rallies support from the tavern full of people and carries him off to be tried in her sister’s court.

From one perspective, this whole plot seems to result from one chance meeting. It feels plausible because these things happen in real life (I, after all, ran into my college roommate in a rest stop in New York), and with so many significant characters all running around Westeros, it feels entirely appropriate that they should happen to show up at the same inn one night. On the other hand, of course, this whole plot is the epitome of narrative contrivance, every bit as unlikely as the tell-tale credit card receipt. One episode after Catelyn finds evidence to suspect Tyrion, they show up at the same tavern, at the same time, miles away from anywhere? What’s more, the tavern is full of men who just happen to be wearing the sigils of several houses that owe allegiance to Catelyn’s family? And then this incredibly unlikely encounter leads to a fight to the death in a terrifying mountain court and Tyrion gaining the support of a band of wildings? The event itself, and then just as important, the impressive chain of subsequent events caused by the meeting, is unbelievably unlikely, but it hits such an ideal narrative sweet spot. It feels random and plausible, but it’s also meaningful and significant, and we buy it right away because it’s fiction, and it’s doing exactly what we like fiction to do. Causes have powerful, immediate, interesting effects, and we can assume that boring stuff happens in the background while also only paying attention to the interesting stuff.

Janette and Delmond

There’s an episode in this season of Treme with a scene not unlike the Catelyn/Tyrion tavern showdown. Two former New Orleans residents meet up in a bar in NYC to watch a Saints game, and Treme‘s audience already knows both of them. Janette used to own a restaurant in New Orleans and has moved to NYC to restart her career, is a former lover of Davis the DJ, and has run into several other characters while out and about during parades or in clubs. Delmond is an accomplished jazz trumpeter who has similarly moved to New York for his career, and whose father is one of the New Orleans Indian chiefs. In the bar, they meet and realize they vaguely recognize each other, and chat for a little while. Janette goes to one of Delmond’s gigs. A few episodes later, they have dinner. And then… nothing. They make no useful career connections. They like each other, but do not become best friends. They do not sleep together. The season isn’t over yet, but somehow I doubt they’ll convince each other to move back to their homelands. The best I’m hoping for right now is a slightly more fully-fleshed metaphor about how Janette’s new discovery of soulful yet refined food is similar to Delmond’s New Orleans jazz fusion.

No question, this scene from Treme is farther to the right of the reality spectrum than the one from Game of Thrones. As the episode makes clear, this is a bar full of New Orleans ex-pats who are all there for a Saints game, so the same time/same place thing actually makes a fair amount of sense. And really, the likelihood that these two would meet again without building a relationship that has a significant impact on either of their lives is also completely within the boundaries of normal life. It is… real. Weirdly real. Uncomfortably, oddly real. And instead of thinking, wow, Treme is truly dedicated to its verisimilitude, all you think is, “why should I care about this scene? What is its purpose?” Down there, deep in the uncanny valley of narrative plausibility, all you can see is artifice and missed opportunities, and you lose track of how good the acting can be, or the show’s political message, or the fact that you actually do like watching it.

So this is my proposed Uncanny Valley of Narrative Plausibility – a piece of storytelling so actually possible, it draws more attention to its flaws than its good attributes. I’m not sure how many other examples I could come up with, largely because there aren’t that many shows which try to move farther to the right side of the graph. In spite of all of this, I do like Treme. It worked its way into this unpleasant place by trying to do something experimental with television storytelling, by de-emphasizing plot and pushing against the ways we usually depict community.

I just wish it looked a little more like a story, and less like real life.

One of my favorite pieces on serial television is Sean O’Sullivan’s “Reconnoitering the Rim,” an article about Deadwood and third seasons of television. In it, he divides narrative into the possible, the necessary, and the possible disguised as the necessary, and describes the last category as a special problem for television. “Out of the array of possible stories and interests presented in the start-up operations of a narrative,” he writes, “some get selected and acquire the force of necessity without having ever really been necessary all along. This force of necessity accrues from the existence of a terminus, which asks that the possible acquire some sort of shape over the course of the regular production of episodes.” Or in other words: an ending looms inevitably on the horizon of any show, and its existence means that a couple of the nearly infinite potential plotlines are going to become weighty, conflict-producing devices full of momentum and vigor, and some are not. It could be any of them – a betrayal, the arrival of bank robbers, a massive fire, a skilled poker player – but at some point, one of those possibilities takes the lead and pulls the narrative forward. Once you’ve reached the end, you look back and believe that it had to be the bank robbers, but that’s only because you’ve reached an end point, and all of those possibilities had to get shut down for an ending to form and ossify.

The thing that I’ve found so amazing about Terriers is its ability to live almost entirely inside that magical, unfixed narrative place Sean O’Sullivan describes. It’s hard to know at any moment whether an episode is going to be a one-off procedural plot or is going to take a sudden left turn into some sticky multi-episode morass, and little plots that you think are complete tend to come back and haunt you; it’s a return of the possible not unlike Freud’s return of the repressed. (See, for instance, Britt’s drug dealer connection, the idea to steal Jason’s wallet, the sudden violent resurrection of Lindus’ real estate fraud, and the question of who actually killed Hank’s friend). Even the dog, that gorgeous goofy bulldog who seems like he’s going to be Hank and Britt’s sidekick in the show’s pilot, disappears into narrative conclusion before magically resurfacing as a key player for both plot and character development.

So much of this first season of Terriers, in fact, has been a dramatization of an unwillingness to declare an ending. Hank’s relationship with Gretchen, of course, is plagued by his inability to let her go, and his potential relapse looms continually on the horizon. Britt gets in trouble when guys from his criminal past show up to threaten his current domestic bliss. It seems as though Hank and Britt have dealt with Lindus, and of course that little problem comes back like a cat with eight lives left. Nothing stays buried, including Hank’s failed rape investigation, Katie’s affair, or Jason’s molest-y past. You drive a dead guy off a cliff, wipe your hands and walk away, and then sure enough, there you are two days later clambering down that cliff ‘cause you forgot something in his pocket. It’s the exact opposite of the procedural’s famous short-term memory problem.

It is a lovely meeting of content and form, then, that the show itself can never seem to decide when a character is really gone, or when an episode’s plot has actually concretized into resolution – except that with sudden explosive force, endings seem to come flying up out of nowhere to smack you in the face with their cruel finality. “Fustercluck” is the best example of this, although I think last week’s brutal final act is almost as good. This is why I will be particularly bereaved if, after the conclusion of tomorrow’s season finale, Terriers never comes back. It’s a show that seems like it could always come back, and it would return sneakily, from an unexpected direction.

There are obviously examples of shows that get bad and for whatever reason, do not recover. We all know the tragic stories, the sad shambling corpses of formerly entertaining programs lingering on long past their prime like miserable shark-jumping zombies. Gilmore Girls season seven. Prison Break, Heroes, Alias, Entourage. There are a number of reasons things can go wrong, including changes in the creative staff, pressure from networks, a resistance to imaginative or risky storytelling, a concept that’s meant to be small saddled with the burden of far too much time (oh, Prison Break, you poor bastard). But I would argue that some shows can and do get better, even in the face of some dismally low points.

Friday Night Lights – This is obviously the premiere example of how rough things can get on a show and still come back for an amazing third and fourth season. It’s also a good example of how quickly terrible subplots can completely derail the rest of a show (see also: the Coma Baby plot of Veronica Mars season two). The Landry/Tyra murder plot is so, so awful and was so thoroughly panned as soon as it happened, FNL spent much of the rest of the season trying to get through that damn subplot as quickly as possible and then force everyone involved to forget it ever happened. Not only did the show manage to exit out of that dark hole of implausible violence as gracefully as one could imagine, the show has since had the excellent judgment to avoid anything similarly out of character.

Battlestar Galactica – Sure, sure, it’s great when you can get a science fiction show to speak to topical issues of morality and terrorism in a way that forces people to talk about the intellectual potential of pulp genres. But for the most part, Battlestar’s New Caprica episodes were just treading water until the characters could get back into space (and back into shape, in the infamous case of Tubby Apollo). Even worse, although the explicit references to insurgency, colonialism and prisoners of war brought the show attention for being so politically relevant, it was some of Battlestar’s most heavy-handed thematic work. Those New Caprica episodes were about as subtle as poking out an eye (whoops, sorry Colonel Tigh), but once the show got back into space and Lee Adama lost all that weight, things were back on track.

The West Wing – This one is a complicated example, but worth thinking through. The show suffered one of the worst, most irrevocable changes a show can experience – the departure of its idiosyncratic, driving creative force – and that kind of departure is almost a self-fulfilling prophecy. It’s a change any fervent fan will declare to be the End of the World, and when a show then immediately proves to be much worse than it used to be, it’s easy to write off the show entirely. I understand the argument, and I also believe that post-Sorkin West Wing never reached the same heights as it did in the Sorkin years, but I also think season seven of that show was a vast improvement. It could never go back to being a Sorkin show, but it did grow into its new identity as the Santos-Vinick race overtook the final Bartlet years. It would never be as fizzy or fascinatingly idealistic as the first few seasons, but it was still miles better than the dark days of Leo’s heart attack and the overt Macbeth references, and it was entertaining television.

Dollhouse – A different kind of improvement narrative from the previous examples, but one that probably happens more frequently. Shows begin, and they’re bad. Gradually, with practice and hindsight and feedback, they get better, and the change can be so drastic that the show is nearly unrecognizable. Dollhouse falls in this category, though like so many shows, the change came too late. I’d also list Cougar Town here, as well as Parks and Recreation, Community, Fringe, Sons of Anarchy, and of course, the troublesome Chuck.

I’m not trying to argue that Chuck may not be in trouble – from what I’ve seen so far this season, things look dubious. But the beauty of television’s episodic structure is that new beginnings and fresh starts happen all the time, and no matter how serialized or intricate a show may be, the very concept of an episode promises that things can change. It’s a whole new show every week, with different writers and directors, different guest stars and returning characters, new plots and character arcs. It seems to me this is a reason fans hold onto television shows even after they’re long dead (oh Smallville, you keep on keepin’ on), because the distinct separation of each piece of narrative means it’s easier to believe that the start of the next episode is also the start of a different, better version of the same show you’ve been watching for so long.

I don’t want to chide my husband if he doesn’t want to watch Chuck any more. Maybe it won’t get better, and he’ll have saved all of that time for Boardwalk Empire or The Walking Dead or (one day, because he loves me) Veronica Mars. But I do want to explain why I’ll keep watching, and why that choice makes sense to me.

It wasn’t until some point in college when television’s ability to capture my entire brain overtook my easy daily slide into a text, and it is especially thanks to one summer in Boston that the most basic advantages of television as a narrative medium became clear to me. The enemy of absorption is distraction, and I finally realized while trying to finish Adam Bede next to a wall actually vibrating with the impact of a reggae bass line that the simple presence of sound coming out of a speaker – my headphone speakers, not the neighbors’ blasting behemoths – made a marathon viewing of Buffy the Vampire Slayer far more attractive than poor George. I watched every single episode of Buffy that summer, thanks to equal parts fascination with the show and desperation to focus on something other than that awful, perpetual noise. Television was far more reliable than an audio book, which has none of television’s multi-layered aural aesthetic, and leaves too much empty space in the background. Music made an effective sound block, but it left my brain running in circles, always testing how much of the Constant Bass Line from Hell was slipping through. Buffy was my best friend, and once I spent time powering through many episodes a day, I began to appreciate the way the repeating rhythms of an episode format were laying bare the skeletons of plotlines, making it easier to see character development and audience suspense stretched out over the frame of a twenty-three episode season. I went to Buffy for the improved escape of multi-sensory entertainment, and almost in spite of myself, stumbled into absorption with television as television and not just a story that kept going and kept me going.

There inevitably comes a time when life interjects into one’s seven season marathon, and you have to figure out how to do things that you wish you could do quietly, like read or write, inside a space that is cluttered with bass lines and stomping and overheard conversations. At the moment I’m exploring the potentials of a noise machine. I wish I were watching TV.

It makes sense that a serialized story pushes viewers away from a standard week-to-week viewing pattern, as continuing plotlines create desire for the next installment and make it pleasurable to view several episodes at once. Maybe a viewer will opt out of a show for a few weeks and then catch up all at once on DVR or hulu, or maybe he or she will wait until a season is completed and watch it all on DVD. More than just establishing a desire for multiple episodes at a time, though, the ability to view television outside of weekly installment structure allows viewers to hold off altogether until they can be sure that a show has ended. When season six of Lost appears on DVD, there will be nothing between you and a legal, uninterrupted, start-to-finish viewing experience. Aside from the promise of continuous narrative, alternate viewing technologies allow audiences to watch completed shows with a determinate length, as well as know in advance the circumstances of their endings. A viewer sitting down with Deadwood right now can choose to learn that its abrupt cancellation will likely lead to a frustrating resolution, or that despite its beginning appearance as a show centered on high school, Buffy the Vampire Slayer will continue long enough for its characters to grow out of their awkward teenage years. The presence or absence of these two simple pieces of knowledge – how long a story will be, and how it will end – make serialized television stressful and potentially frustrating in the moment, and profoundly satisfying in retrospect. This, in large part, is why television serialization continues to be a topic of interest for television creators and viewers. It is the form of narrative that best strains the gap between traditional television viewing and alternate methods (how long will it be?), and it reflects viewers’ anxieties away from the fiction and onto the creation process (how will it end?).

As a way of coping with the inevitable worry and potential dissatisfaction that comes from unknown endings, serialized television shows have also retained a much stronger episode structure than earlier forms of long-term storytelling, like the serialized novel. Without the benefit of a known end point, serial television shows often provide a strong impression of the episode as a whole unit, either by coupling the long-arc stories with episodic plotlines in shows like The X-Files or House, or by establishing the episode as a unit that is whole outside of, or in spite of, incomplete plot. Shows like Mad Men or The Wire work with thematic or aesthetic motifs to displace any anxiety about large-scale endings onto episodes that are meant to be complete and satisfying as individual pieces as well as seamless parts of a longer story. This, too, is a reason for our continued consideration of serialized television as a form – its strong episode structure, even in shows that play with the limits of how porous an episode’s boundaries can be, make it a different and more complicated type of serialization than simple long-form storytelling. Even in the most serialized shows, the divisions are still more powerful than the continuous line.

It’s hard to think about anything other than Dickens, frankly – a trap this person clearly fell victim to while registering his or her car. Occasionally I get flashes of TV or new media -related ideas, among them, “Were Dickens’ sketches the YouTube videos of his day?”, “What is it we like about really long stories if it’s not plot?”, and, “If his insistence on repeated public readings of the most violent scene in his corpus, the scene in Oliver Twist when Sikes murders Nancy, did indeed contribute to Dickens’ premature death (as was argued by a lecturer this morning), do we need to worry about the health of such TV violence aficionados as the Davids Chase, Milch, or Simon?” (Answers: Yes, hmm, and I hope not.)

At the moment, though, television looks like a far-off vision of the future when seen from a world emphasizing daily Victorian teas and frequent discussion of the New Poor Law. Better luck tomorrow.

]]> http://telephonoscope.com/2010/08/04/tales-from-dickens-universe-part-2/feed/ 0 http://telephonoscope.com/2010/08/03/tales-from-dickens-universe-part-1/ http://telephonoscope.com/2010/08/03/tales-from-dickens-universe-part-1/#comments Tue, 03 Aug 2010 19:23:00 +0000 kvanaren http://www.telephonoscope.com/?p=1389 I’ve written in the past about a peculiarly strong affiliation TV writers have claimed with nineteenth-century novels, and especially with Charles Dickens, and the many qualities of his work that are useful for people who think about television today – Dickens’ serial publishing, his focus on urban spaces, his melodrama, his intricately woven plots. It’s something I have continued to muse about over the past several months, and it’s a topic I feel especially drawn to expound on this week. Because I am at Dickens Universe.

Yes, Dickens Universe, a week-long Dickens-themed conference/workshop/summer camp/party held at UC Santa Cruz every year, and featuring lectures from Dickens scholars, seminars for graduate students and members of the general public, workshops on writing, pedagogy, and presentation skills, and nightly parties with themed cocktails that coordinate with the current year’s primary text. (This year: Oliver Twist and Sketches by Boz. Last night’s drink: Nancy’s Heart of Goldschlager cocktails.) It’s an unusual space for academics, something that combines graduate student development opportunities with a forum for peer feedback, and then adds in the nearly unheard of element of presenting one’s ideas to an audience outside of the academy. It’s pretty great, really, and not just because each day’s schedule includes two coffee breaks, a Victorian tea, post-prandial potations (yes, really), and the aforementioned nightly party.

I came to Dickens Universe well aware that ol’ Charlie has been actively re-appropriated in the world of television as a father of intellectually respectable mass entertainment of a form not unlike Lost or Deadwood or Damages or [insert multi-plot serialized show here]. I was also aware that from what I’ve found, most references to television’s nineteenth-century analogues have been whittled down to just one authorial figure, a jovial Dickens perched in the background of today’s television landscape. Wilkie Collins, Anthony Trollope, Elizabeth Gaskell, or Thomas Hardy are nowhere to be seen, much less Balzac or Flaubert. What I have been surprised to discover is that at least colloquially, many Dickens scholars have performed the same kind of singular appropriation, only in reverse.

Where interviews and critical pieces about television reference Dickens over and over, Dickens scholars reference one show – The Wire – with similar fervor. I’ll admit, some of this is at least prompted by me. “I work on television,” I say, and the near-unanimous response is “The Wire!” But I hasten to add that it would certainly be here whether or not I were here, frequently bringing up TV. On the Universe’s first full day, graduate students and faculty got into small groups to brainstorm teaching ideas about Oliver Twist, and when we reported back to the big group, we ended with a giant list of possible avenues for further discussion. We had the novel as a form, affect theory, Dickens as a social reformer, caricatures and characterization, thingness in Oliver Twist, Oliver as the novel’s vacant center, negative depictions of marriage, etc. etc. etc., and as a suggestion from one of the groups, The Wire. In connection with Oliver Twist, they mentioned that season four might be particularly relevant.

It is particularly relevant, of course, but so would a discussion of melodrama, serialization, violence, audience, and any number of other things about television more generally. Right now, though, I find the selection of that singular touchstone show to be sort of satisfying. Television seems to have picked Dickens, and in turn, Dickens scholars have picked The Wire. Even if it’s somewhat unfair on both sides, the symmetry is too pleasing to pass up.

The Whedon/Abrams panel was probably the panel highlight of my Comic-Con experience, and it particularly stood out after the long string of much less exciting movie release panels that we had to sit through in order to guarantee we’d have seats. This is of course a severe reduction, but my impression of the big name movie panels we saw (including RED, Battle: Los Angeles, Salt, MegaMind, and Tron Legacy) is that most actors are pretty boring (except for Will Ferrell and Tina Fey for MegaMind), most of the screened questions from the audience are repetitive (“What kind of fight training did you have to do?” “How are you dealing with this as an adaptation?”), and although the idea of using Comic-Con as a platform for previewing new things is a good one, everyone has a higher level discussion when we see enough from a preview to actually talk about it. (“So, we really didn’t see any glimpse of what the evil aliens will look like in this film, but… could you talk about them?”)

Seriously, look at Mary-Louise Parker's face as she listens to Bruce Willis.

Which is why it was such a relief, and so completely awesome, to have Joss Whedon and JJ Abrams come out and be able to talk about well-known work, about narrative and storytelling, about inspiration and the production process, and about their own experiences as fans. It was fascinating to hear both of them talk about the differences between making movies and making television, and to hear Joss Whedon say that although movies are hugely satisfying because they have firm and final endings, he feels that long form storytelling is more rewarding and much harder. The moderator, Doc Jensen, also asked about serialization on television, and both men admitted to understanding the financial motivation for non-serialized shows but having absolutely no creative interest in them. “Stories imply time,” said JJ Abrams. “Stories imply inevitability and some kind of progress.” “I don’t think the networks will ever, ever ask for that,” added Whedon. “The networks will never admit that people want that, because they do see the cash cow of ‘The Mentalist! Let’s all do The Mentalist!’ And when Lost first hit and was blowing up huge…they were still like, ‘We don’t want that. That successful, Emmy-winning thing? No, we don’t want that.’…And it’s very weird, because ultimately, the serial is always going to be the thing people remember. What do they remember about Cheers? It’s Sam and Diane, not a great joke from Cheers.”

What came out of their discussion of serialization, including Joss Whedon’s trials with the cancellations of Firefly and Dollhouse and the unrepeatable structure of a show like Lost, is a huge disconnect between the stories people like Abrams and Whedon want to tell on TV and the capacity for network television to produce those shows. Recently, Abrams has been more successful than Whedon in creating television for the networks, but as he mentions in his descriptions of both Fringe and his new show UnderCovers, it’s because he’s been careful to balance the standalone aspect of an episode and his own dominant interest in longer story arcs. One of the most interesting moments for me was when Joss Whedon admitted that in making Dollhouse, he just had not realized how much the networks have changed, and had not come to terms with the idea that he has a cable rather than network mentality. “I definitely was trying to fit a square peg into a round hole.”

Whedon did talk about preparing to do The Avengers, and insisted on describing to JJ how fully he loved Star Trek. “I have had actual moments of sheer fucking panic because I love Star Trek so much.” In turn, Abrams talked about the process of rewriting, and when Whedon mentioned that he doesn’t really write second drafts, Abrams shot back “You bastard!” But however much fun they had joking around and teasing each other, it was hard not to come away with a sense of how tricky the television landscape is right now, and how swiftly the networks are changing. I can only hope that when Joss Whedon is done with that silly Avengers movie project, there will be a new Whedon show on cable.

I can also hope that next year, after we figuring out some of the basic Comic-Con ins and outs this time around, we get to attend more panels like this one.

Such was the case for “Fly,” an episode that abandoned every cast member except for the two leads, and stuck them in an underground lab while they tried to kill a fly for nearly the whole hour. Alan Sepinwall writes that episodes like these are called “bottle episodes”: episodes produced with the goal of costing as little as possible in order to reserve money for guest appearances or awesome explosions in the season finale. “Fly” uses only two of the show’s actors, and it takes place entirely on sets already built for the show almost all of which are indoors, which cuts out the expense of shooting on location. It would make sense for an episode like this to be a departure from the show’s typical tone or style of storytelling, and in some ways that was the case. There were no gorgeous wide shots of the desert or sudden gross-out images of a decapitated head attached to a turtle. It was not a depiction of Walt being torn between two different lives, as so much of Breaking Bad has been – none of the usual scenes where a cell phone call interrupts an important life event, or Walt is called away from his meth deal by a family emergency. Still, “Fly” was overwhelmingly invested in many of the things most characteristic to Breaking Bad. It was an episode that allowed both Walt and Jesse ample time to consider the narratives they have constructed about their lives and to think back through their past actions and work out what went wrong. And somehow, even in the restrictive space of the lab, “Fly” manages to ratchet the tension so effectively that by the end, I was actually covering my eyes while Jesse teetered on the unstable top step of a ladder.

“Fly” is like nothing so much as a Beckett play, and so it’s no wonder many people hated it. There’s no action, nothing happens, why is he so obsessed with that stupid fly, what are they even talking about right now, etc. etc. etc. Episodes like these, those that stand out from the rest of a show, tend to be controversial. (Think of how many people hated “Exposé,” myself included). In many instances, it’s because an audience member may not care for whatever it is that specific episode is doing – if you really hate Beckett, you probably hated “Fly,” and that’s okay. If you cannot stand musicals, then Buffy’s musical episode will leave you cold, and nothing Joss Whedon did would ever make you feel otherwise. I do think, though, that there’s something deeper happening when an audience resists one particularly unusual episode of a television show. Those episodes can feel a little like a violation, or a betrayal of a contract. We know by now that an episode of Buffy is supposed to look and sound like one thing in particular, and we’ve been watching for the past six seasons because we like whatever that is. The idea that any random episode could just go off the rails from what we anticipate is unnerving, because it upsets the norm and hints at an instability we didn’t previously realize existed.

This is exactly the reason that shows can jump the shark, but it’s also the reason that television can be so fascinating over the long term. Even the shows most inextricably rooted to a formula have the potential for a single episode to derail the status quo. In any week, the show you’ve been loving as Happy Days may choose to send the Cunninghams to Hollywood, thereby giving Fonzi a chance to waterski over a marine predator. But on the other hand, every new episode is also an opportunity to be a costume drama, or a farce, or some bizarre metafictional exploration of minor characters. Or a Beckett play – but just for an hour.

It’s just one of those vagaries of the television production world, the bi-annual presence of the Nielsen ratings sweeps, but I cannot get over how odd it is that we now have a season where television fictions all rise up simultaneously into frenzies of melodrama. For most of these shows, the bulk of the winter is a slow burn, where characters change in tiny, easily reversible stages, and the startling events that threaten to explode prematurely quickly die back down. On Lost, characters have been marching determinedly around that silly island, forging and breaking allegiances, pointing to creepy kids standing in the jungle, but never making much progress toward resolution. On shows like Bones, Booth and Brennan moved inexorably closer to a romantic relationship and immediately backed away before it could overtake the familiar episodic patterns. CSI, Law and Order, and NCIS continue to chug on as they always have, although Law and Order: SVU has increasingly begun to go off the rails into strangely burdensome emotional stakes – an attempt, no doubt, to wrest popularity back to NBC’s still-floundering 10pm timeslot. Rick Castle will never actually get together with Kate Beckett, even though her apartment did blow up a few weeks ago, and Dr. House is still a jerk.

But every May, just because it’s May, the months-long slow burn erupts into a full on conflagration, and the aim of the game is to present as convincing an argument as possible that the rules of the show you’ve been comfortably watching aren’t set in stone, despite what you may have thought. The characters you assumed were immutable and eternal will die in dramatic car crashes, or they will finally marry each other, or if it’s a J.J. Abrams show, the organization the protagonist assumed she was working for has all along been just a part of another, much more secret organization, and it’s actually evil. You watch, and you keep watching, because the show needs to keep alive the possibility that what you’re watching is progressing rather than repeating, and these fin de siècle gestures at the end of every season are crucial to that belief.

For a lot of shows now, particularly on cable and premium channels, that belief in progress and the possibility for real change is one that’s well founded. On Mad Men, Sons of Anarchy, United States of Tara, Treme and the like, characters have memories, and retain the experiences of previous episodes and seasons, so when a character dies or someone gets married, it tends to stick in a way that deaths and marriages often don’t on network shows. But you’ll notice, none of the shows I just named are reaching the ends of their seasons right now – two aren’t even on. Cable and premium channels work on a different audience model, and aren’t nearly as beholden to the Nielsen sweeps as the networks still are, and so they don’t participate in the annual month of May eruptions. I’m not suggesting that Bones and Booth don’t remember that they just kissed a few weeks ago, but that events like those, and particularly, events that crop up as a result of these May shenanigans, tend to be erasable. Characters die, and they do tend to stay dead, but the consequences of those deaths dissolve pretty quickly, leaving everyone about as cheerful as before come next November. When was the last time you heard anyone mention poor, disfigured, tragically dead George on Grey’s Anatomy? How about Edgar on 24? (Confession: I haven’t been watching 24 in a while, so maybe Edgar’s death is being mourned more fully than I’m supposing). How about that life-threatening brain tumor Allison had on Medium last season?

It’s May, the season of deaths, weddings, and babies on TV. Enjoy them now, because in most cases, they won’t last.