Intermittent Mechanism

Ian here—

In my 2013 “Comparative Media Poetics: Cinema and Videogames” course, I devoted a week to the genres of mystery and suspense. In this first class of this week, we discussed theory. Students read a portion of David Bordwell’s Narration in the Fiction Film, in which he discusses the concept of communicativeness of narration, and the specific ways communicativeness is clamped down in the detective genre. We also discussed the ways in which mysteries play with time, using the formalist conceits of fabula and syuzhet that Bordwell draws from. This dovetailed with our second reading, Jesper Juul’s 2004 article “Introduction to Game Time,” in which he expresses skepticism that videogames could ever pull off a flashback-based story structure.

The screening for this week included the entirety of Alfred Hitchcock’s Stage Fright (1950), as well as selected chapters of Heavy Rain (Quantic Dream, 2010), which I had students play live, and discuss. We re-convened during the next class session for a discussion on unreliable narration and the relative “fairness” of twists. Readings included Kristin Thompson’s chapter on Stage Fright in Breaking the Glass Armor, and Emily Short’s writings on Heavy Rain. Spoilers on the (potentially unfair) twists of both texts below.

False Flashback Fake-out

The word “infamous” doesn’t come up much in discussions of Alfred Hitchcock’s celebrated career. But if anything merits such a distinction, it is the flashback in his 1950 UK film Stage Fright.

Stage Fright begins in medias res, with its protagonist Eve Gill driving her friend Johnny Cooper, who is seemingly concerned about someone being in pursuit. A mere minute into this action (not including the film’s opening credits), we suddenly shift to a flashback, as Johnny narrates the story of being caught up in the actress Charlotte Inwood’s murder of her husband, and now being doggedly pursued as a false suspect by the police.

The rest of the film follows Eve’s attempts to clear Johnny’s name, hiding him from the police and using her skills as a actress to become a mole, insinuating herself into the lives of both Ms. Inwood and the police inspector so as to conduct reconnaissance. Inwood is, of course, revealed to be a scheming femme fatale, and the film plays as a taught little thriller with Hitchcock’s usual comedic touches … until the whole edifice crashes down, in the last few minutes. Because, as it turns out, Johnny actually did commit the murder he is under suspicion for.

Kristin Thompson writes that, at this particular moment “What has seemed to be a suspense plot turns out to be a murder mystery.”[i] This generic revelation is an odd one, for two reasons. First, since the mystery is effectively solved the moment that audiences learn it existed, we’re robbed of our fair chance at guessing. Secondly, looking back on the film’s presentation of events, it means that Johnny is not the only one who has lied. The film’s narration has lied to us, as well, right alongside Johnny.

What does this mean? Well, first of all, let’s look at how the flashback begins:

Thompson offers the following description of this transition, the assumptions it encourages viewers to make, and the ways in which these assumptions are in fact unwarranted:

As he speaks of being in his kitchen at five o’clock, a dissolve leads into a direct presentation of the situation; at the end of the dissolve, while he is speaking of the doorbell ringing, his words fade out quickly. There follows the sound of the doorbell itself. This suggests that the flashback will be a direct replacement for Johnny’s words, a straightforward version of the events which will stick more or less to Johnny’s vantage point. But an examination of the flashback shows that this equation of the flashback with his words is only one factor of several. In fact, the sequence goes beyond Johnny’s telling in several ways…[ii]

What are some of the ways in which this sequence “goes beyond Johnny’s telling“? Well, for one, it includes details that he himself couldn’t possibly have been privy to. We can start with small things, like things Johnny couldn’t have seen firsthand:

Here, Johnny just barely escapes after Charlotte Inwood’s wardrobe assistant Nellie discovers Mr. Inwood’s body. The high-angle shot of Johnny rushing down the staircase and out the door reveals that Nellie catches a glimpse of him as he leaves. But how would Johnny know she saw him, given that his back is turned to her? (This shot is repeated in a sort of flashback-within-a-flashback, where Johnny remembers his escape and the loose ends he left, making things more complicated still!) Then, there’s this:

Hitchcock shows off some expert staging of this comedic chase element, as the horse-drawn cart blocks the pursuing police. But, again, we as viewers see more than Johnny possibly could have personally witnessed.

The implication is clear: the narration is going above and beyond Johnny’s own first-person description. In doing so, it is effectively acting as an independent agent, one that independently verifies Johnny’s story.

Jumping to the end of the flashback:

At this point, I asked a very broad question: Is there anything especially odd about these last two minutes or so of the flashback? If no one bit on this question, I planned to get more specific: Anything to do with point of view? Finally, if students are were confused, the clincher: Who is seeing the police arrive in the auditorium?

The answer is Eve. Once Eve makes her way into this flashback, we begin to see shots that eyeline matches clearly indicate to be her POV.

Put simply: This is weird! Why would Eve’s POV shots be in a flashback that supposedly represents Johnny’s story? Again, the film’s narration has shifted from simply relating Johnny’s story to embellishing it, pulling in the viewpoints of more characters. In doing so, it seemingly corroborates Johnny’s story. This is a problem, since at the end of the film so much of what transpires during it will be revealed to be Johnny’s lie.

I will give Thompson the final word here:

If the sequence were too obviously a subjective memory, the spectator might be more inclined to take it as a questionable version. But this is not the case. As the scene is constructed, there can be no question of its being only Johnny’s lie. The filmic narration participates quite definitely in the lie as well.[iii]

Why don’t you tell us what you really think?

By 2010, game developer Quantic Dream had already released two games: Omikron: The Nomad Soul (1999, Dreamcast and PC) and Indigo Prophecy (2005, Playstation 2, Xbox, and PC). But Heavy Rain was different: it was the first game the studio released under a new exclusive publishing deal with Sony Computer Entertainment. Sony, building a stable of “highbrow” games exclusive to its PlayStation 3 console—which at this point included Flower (thatgamecompany, 2009), and would go on to include The ICO & Shadow of the Colossus Collection (Team ICO, 2011), Journey (thatgamecompany, 2012), and The Last of Us (Naughty Dog, 2013)—aggressively marketed the game as an “interactive drama,” pushing the storytelling capabilities of videogames forward in a manner that had never before been attempted. Its unconventional control scheme would allow players to control everything from small gestures during conversation to life-and-death decisions. Players were also promised that if one of the game’s four protagonists died, the game would continue on, its story having changed, but not ended.

The story players were interacting with? The tale of a child’s abduction at the hands of the Origami Killer, a serial killer known for drowning children in rainwater over a period of days. As the characters we control (everyone from a reporter to an FBI agent to the boy’s grieving father) conduct their own independent investigations, the clock ticks down to get the boy back alive.

Anyway, the killer is totally one of the characters you’re playing as. Specifically, it’s Scott Shelby, the private investigator who works with previous victims’ families throughout the game, collecting evidence. Players assume that Shelby is collecting this evidence for an investigation (as that is what he tells the grieving parents he interviews), but actually he is collecting it to destroy it, cleaning up lingering loose ends that might implicate him.

Heavy Rain‘s form, up to and including its twist, provides a great opportunity to interrogate many of the claims made by game theorists that we’ve read so far in the class. Steven Poole argues that dramatic irony is impossible in videogames, because no knowledge differential exists between viewer (player) and protagonist (player-character).[iv] But Heavy Rain‘s use of multiple protagonists allows for just this sort of gap. Jesper Juul insists that flashbacks can’t work in videogames, because of the “time machine problem” in which failure in the past might render the present impossible.[v] But Heavy Rain has a playable flashback—one that, interestingly, uses the player’s inevitable failure to achieve a certain task as one of the primary motivations for that character’s behavior in the future.

Heavy Rain seems to radically redraw the assumed limitations of the videogame medium.But the question is: Does it do a good job of it?

The consensus, generally, is “no.”

Failures, though, can be fascinating. Even before we started playing the game, I prompted students to pay close attention to the way the game narrates information about the characters to us. The game attempts to preserve some of the epistemological tricks of narration that one finds in the mystery genre elsewhere. However, it ends up being enormously unfair.

In a typical detective-based mystery, the most common way narration restricts the viewer’s knowledge is by restricting it to what the investigator knows: we learn what the detective learns, when he or she learns it. In this situation, communicativeness (to use Bordwell’s term) is largely shaped by the demands of “fair play.” The viewer or reader of a mystery must be given a chance to discover the culprit on their own, before the narrative reveals it. But, ideally, they shouldn’t be able to discover it significantly before the detective does. Basically, then:

• Spectators/readers must have access to all of the clues the detective has access to (this gets to issues of knowledge)
• But they needn’t have access to the deductions the detective makes about these clues (it’s their job to do this on their own)

So a mystery should provide us with the basic knowledge about the facts of the case the detective has, but doesn’t need to actually communicate the significance of these facts.

There’s plenty of room in here for certain kinds of fair plot twists. But twists need to arise out of information that both the spectator/reader and the detective have access to, which the spectator/reader misinterprets, but the detective (presumably) does not. Looking back on the chain of events, we should be able to see where we were mislead, understanding the significance of things we didn’t notice before. If the writer did their job, we should not find any place that the mystery’s narration overtly lied to us (even if the story’s characters did). In order to not lie, the mystery’s narration probably had to be ambiguous in certain respects. Looking back, we should be able to understand that this ambiguity gave us enough rope to hang ourselves by artfully dolling out information, being discrete in its communicativeness. Good twists emerge not from lies but from misdirection, letting our own assumptions do the “heavy lifting” of misleading us.

It is possible to imagine the twist of Heavy Rain working successfully. The dual figure that we initially think is the detective gathering evidence to sift through it, but who is actually the killer gathering evidence to destroy it, is an interesting conceit. But I think that few people would argue that it actually works in Heavy Rain. The question is: Why?

In class discussion, I focused on two primary culprits here. The first, which is dealt with by Emily Short in the blog post we read of hers, has to do with promises of authorship. Short writes:

The twist ending, the discovery that Shelby is the Origami Killer—that felt like a betrayal.

Not because it was unexpected, not because I’d been successfully gulled into caring about Shelby—I could live with that, and movies play those tricks all the time—but because that twist negated the meaning of every truly interesting choice I’d made in the game up to that point. All that time I thought I was at least getting to craft one character, I was being played.[vi]

My students generally agreed with this. As one of them put it, it’s not just that Scott Shelby lies. It’s that the whole conceit of “you controlling him” was a lie. The feeling of agency granted to players is ripped away post facto, as they realize they never actually controlling who they thought that they were controlling, and therefore weren’t in a position to understand their action’s meanings.

Now, it should be noted that this effect can actually be done quite well. For instance, Adam Cadre’s interactive fiction piece 9:05 (2000—playable in your browser here) does something very similar, playing with the gap between the intentions of oblivious players and more nefarious actions of the player-character. The result is a delightful romp. But 9:05 is a five-minute IF piece, and Heavy Rain is a multi-hour game, with characters created through motion capture and voice acting. It encourages more emotional investment in its characters. This can backfire, prompting feelings of betrayal when stunts like these are pulled. (This problem was exacerbated by the the game’s marketing, which stressed players’ story-authoring power.)

However, Heavy Rain also suffers from a more serious problem. Like Stage Fright, its narration lies, quite baldly.

For one particularly offending moment, take a look at the following clip of the detective-killer, Scott Shelby. In it, you’ll see the game’s granular and context-based control scheme. When Shelby sits at his desk, the control the game offers over his actions feels almost like puppetry (a feeling enhanced by the PlayStation Move motion controls, which I’m using as I play in this clip). Once Shelby stands, we are given the option to listen in on his thoughts. (You see and hear this happening starting at the 2 min 27 sec mark.) Here, the player can construct their own first-person narration. There’s just one problem: it is unreliable first-person narration:

Did you catch the offending line? It occurs at the 3 min 16 sec mark, after I’ve had Shelby lie down on the bed and get up again. When triggered with the prompt “fatigue,” Shelby thinks: “I needed the rest. I haven’t been sleeping well since the murders started up again.”

Critic Daniel Weissenberger deservedly ripped into this particular line:

The killer would never think of something he was doing in such passive terms. It’s a line that exists only to create the impression in the player that Scott can’t be the killer. The fact that David Cage resorted to using such a baldy deceptive trick puts him on par with Tom Savage for lying to his audience.[vii]

Students who noticed these types of moments argued that Heavy Rain‘s crime was far more egregious than Stage Fright‘s lying flashback. In Stage Fright, one can at least point out that Johnny is lying to Eve. If one wanted to, one could argue that the narration’s is “in cahoots” with Johnny, motivating its lying. (Thompson, in fact, makes more or less this claim.) But no such fig-leaf exists in Heavy Rain, because no one in the diegesis is being lied to. As one of my students wrote in a discussion board post, “No amount of replaying explains why Shelby either believes himself to be tracking down the killer or why the game lies to the player when it tells you Shelby’s thoughts.” If there were another character that Shelby was lying to, then maybe the game would be motivated in lying to the player, much like in Stage Fright. But their isn’t. And the idea that Shelby himself is somehow guarding his thoughts, as a way of misleading the player, is a non-starter. As my student pointed out, the game carries the “implicit assumption that Shelby isn’t aware that the player exists.”

During class discussion, students continued on this point. One student pointed out that the game’s unusual control options engender the feeling of not so much “being the character,” but rather “being the character’s brain.” The intimate-yet-distanced feeling of the puppet-like commands the player gives out, coupled with the direct access to the character’s thoughts, created a fascinatingly unique mode of player-avatar relation. That is, it did, until the twist shattered it all. This strange and fascinating promise of being the character’s “brain” turned out to be, like the game’s narration, just another lie.

[i] Thompson, Kristin. “Duplicitous Narration in Stage Fright.” In Breaking the Glass Armor: Neoformalist Film Analysis. Princeton: Princeton University Press, 1988. Pg 141. (This was one of the course readings for this week.)

[ii] Thompson, “Duplicitous Narration in Stage Fright,” pg 146

[ii] Thompson, “Duplicitous Narration in Stage Fright,” pg 148

[iv] Poole, Steven. “Electric Sheep.” In Trigger Happy: Vidoeogames and the Entertainment Revolution. New York: Arcade Publishing, 2000. Pg 81. (I had assigned this as reading earlier in the course.)

[v] Juul, Jesper. “Introduction to Game Time.” In First Person: New Media as Story, Performance, and Game. Ed. Noah Wardrip-Fruin and Pat Harrigan. Cambridge, MA: MIT Press, 2004. (This is another assigned reading for this week.)

[vi] Short, Emily. “Heavy Rain’s Storytelling Gaps.” Gamasutra.  (More assigned reading for this week.)

[vii] Weissenberger, Daniel. “Heavy Rain: To the bitter end (Part 4).” Gamecritics.com. (I deliberately don’t give my students this reading, as I want them to catch this moment on their own. And, of course, I want to look clever if they don’t catch it, and I have to point it out myself.)