This is an approximation, with some extra breathing room and extra kitchen-sinkery, of a talk I gave at the International Narrative Conference 2016 in Amsterdam. It was part of a panel session, "Engagement with Experimental Narrative: Contributions from the Cognitive Humanities."

As you see, I plan to talk today about social cognition, suspense, and experimental film. But more broadly, I’d like to lay out some concrete, specific ways that cognitive approaches to narrative can offer useful insights into experimental works that ALSO connect up with, let’s say, less experimental narrative arts, as well as with the sorts of really common everyday things people do and narratives they create that proliferate in face-to-face conversation. I’m also going to try to do all this without getting too jargony or ponderous. Let’s see if I can.

The experimental film I’m going to talk about is Michael Snow's Wavelength, It's a classic so-called "structural film" that consists of a 43-minute fixed-frame zoom shot. One room, no panning, just 43 minutes of very... slow... zoom. It features almost no narrative action.

But what makes it interesting—or one thing that makes it interesting!—is that it nonetheless does for many viewers prompt arcs of experience that include suspense. (That said, this depends a LOT on the viewer and the conditions of viewing; for lots of people the arc of experience is boredom, frustration, or bewilderment. More on that later.)

I want to talk about this in the context of a question that I’m going to a bit flippantly call the question of "how to be sophisticated." For instance, Brian Richardson and others have raised the point that it won’t do—but is perhaps tempting!—to ignore so-called "unnatural narratives." These works aren't much of a problem for cognitive narratology in the Fludernik-ish sense. But cognitive approaches to narrative that draw from recent work in cognitive psychology, neuroscience, and so on have focused a lot on explaining how stories please us and make sense to us by recapitulating, or giving us opportunities to practice, various "everyday" skills and activities. The emphasis is often on fiction as a space for working out the same kinds of things people do in real life. This approach has yielded a lot of really interesting work, but we also know that’s not the only kind of story-telling there is.

Meanwhile, when cognitive science tries to use literature as data about how people think, it doesn’t always go very well. On the one hand, we get studies claiming to show that reading literature makes us more empathic—continuing a long and perhaps not so venerable tradition of work telling us that X, Y, or Z will make us smarter, happier, better at STEM subjects, or otherwise more awesome capitalist tools.

Then on the other side of that rock we have the exciting opportunity to run against the hard place of making over-zealous generalizations by focusing on highly educated, industrialized, postmodern, Western humanity. This is a general problem for a lot of research in cognitive psychology, where as you know there is a pattern that can be described with not too much caricature as "let's look at what fifty Princeton undergraduates do and presume that it is characteristic of the entire human race."

Including "unnatural narratives" and other kinds of experimentation in our investigations is important if we want to avoid flattening and overgeneralization. But it’s true too that avant-garde arts are pretty much the WEIRDest of the WEIRD.

So: How can we be sophisticated about this state of affairs? And, also, can we identify how various artistic sophistications work with respect to other kinds of communicative and creative activity?

Well! Here’s what I think.

As you may have gathered, in this session we think a good place to start is with audience engagement. What useful things can we say about being the audience for something (maybe something narrative!) and engaging with it?

It would be nice if we could say something about some cognitive underpinnings of all this stuff that reflects their shared status as cognitive artifacts, and if that something could prove analytically useful for talking about idiosyncrasies, innovations, and disruptions.

If we can do this, we might be on pretty solid ground in terms of being able to say something interesting and maybe even insightful about the kinds of things we care about as humanists. And maybe at the same time we might even avoid falling into either the trap of reducing complex and multivalenced things to, like, taking your vitamins, nor of treating genuinely WEIRD stuff as representative.

I want to show you some examples involving triangles of joint attention: the deep and important experience of looking (or attending) together. (This is not by any means the only place to start, but it’s what I want to talk to you about today.)

We start out with dyadic relations: the relationship between ourselves and others, or between ourselves and things. But to enter the world of reference and of shared intentions you need joint attention.

In joint attention, two or more people know that they are jointly attending to something, and they ALSO know that there is a mutual understanding that they are engaging with each other in this way.

Human beings are, neurotypically, highly equipped for classic joint attention, these scenes where everyone is co-present and so is the thing they're jointly attending to. People are very highly attuned to the attentions of others, and our communicative abilities, including language and gesture, are supported by and keyed to the basic scene of joint attention. Getting the hang of joint attention is a crucial part of language acquisition. It's a foundation of our ability to teach, learn, communicate, and cooperate. It structures the way we navigate the social world and the world of language.

Another nifty thing about people is that we can abstract away from the here and now. So there are lots of characteristically human situations that don't involve classic joint attention, but which build on it in various ways. Personal letters, telephone calls, movable type, email, and many other technologies create common cultural scenes of what Mark Turner and Francis Steen (2013) call "blended classic joint attention".

This weather reporter is of course looking at a camera and pointing at a green screen, and when we watch the weather report we know perfectly well that she’s not really talking to us, but everyone involved understands that we're to think of this as a CJA scene. And there are loads of linguistic constructions used in conversation that invoke similar extensions of the joint attentional scene—because people everywhere talk about hypotheticals, imagined things, the past, the future, and so on.

In discourse and art, these triangles proliferate; they come into play at different levels of a narrative situation (whatever the medium).

Incidentally, this slide took me longer to produce than anything else in this whole talk, so I hope you like it.

Joint attention gives us a way to talk about quite a few interesting narrative things. For instance, our readiness to attend to attention—to think in an attention-oriented way—provides important scaffolding that can organize narrative structure, and which can be disrupted or exposed one way or another to create various effects.

Narrative texts of all kinds rely on our inclination to treat even bare descriptions as representations of the intentional perceptions of another person. The experimental mode on display here (from the Calypso chapter of Ulysses) derives much of its characteristically modern, famously difficult structure by presenting versions of this dynamic in which aspects of the subject (rather than the object) of conceptualization are left highly implicit over extended stretches of narration. [I talk about this sort of thing at length in a paper called "Joint Attention, To the Lighthouse, and Modernist Representations of Intersubjectivity."]

Thinking about narratives together with joint attention can also remind us of the fact that books and films are things that readers and viewers share with and show to each other.

It's tempting, especially when we take the narratological view, to think first of the relationship between authors and readers (and how it is attenuated and mediated and implied) but the scene of reading together may be a better place to start.

It's also worthwhile to look at how and when art, literature, and film represent and thematize scenes of joint attention. One example I’ve written about is Woolf’s To the Lighthouse, which is riddled with scenes of looking together. If you want to hear more about that, I'd be happy to talk your ear off later.

But I promised that I was going to tell you something about suspense, specifically in film. Suspense IS very often discussed in terms of audience engagement, but not typically in terms of joint attention. It’s also not typically considered in the context of weakly narrative or otherwise very experimental film. So let's take a look.

Much has been written about suspense as a function of what viewers know and what they merely anticipate or expect. The psychological, philosophical, and cognitive literature on suspense (e.g. Gerrig 1989, Brewer 1996, Yanal 1996, Carroll 2001) has for that reason been extra interested in what is sometimes called the "paradox of suspense" or the "anomalous experience of suspense": if suspense is about uncertainty, why on earth should people report feeling suspense when they re-read or re-watch something, or when genre conventions make it trivial to predict the outcome?

I want to talk about something a little different, what we might call the poetics of suspense. How can suspense be invoked, enforced, triggered, enhanced? And what can we say about the kind of suspense generated in a film like Wavelength, which consists of a forty-three minute zoom shot in a single room, during which character actions occur only briefly and peripherally? What is the common thread with the effects associated with traditional suspense narratives like those found in Hitchcock? One ingredient is the element of temporality. You have to WAIT for resolution. But resolution of what?

One option is to fire up a scenario of joint attention.

First I'm going to take you very quickly through a classic example from Hitchcock's Notorious. Here a mother-son pair, the Sebastians, have been bringing Alicia (Ingrid Bergman) to the brink of death via slow poisoning. In the scene in question, the movements of a coffee cup serve as the focal point of two and a half minutes of mounting tension, culminating at last in Alicia's sudden awareness of the cup's significance, just in time for her to succumb to the poison and collapse.

The tight focus on the coffee cup both establishes its significance—we never see the poison or any discussion of how it will be administered—and makes it the focal point of a prolonged sequence of building tension.

Selective framing reinforces the overwhelming salience of the coffee cup and its status as a vessel of suspense. It looms in the foreground of every shot of Alicia until she drinks from it. But she never looks at it.

(What's going on here is that the Sebastians are waiting on tenterhooks for her to drink it, while various distractions and disruptions threaten to blow up their scheme. At one point their less in-the-know co-conspirator Dr. Anderson almost drinks it himself. It's quite a sequence. Todd Oakley and I discuss both this scene and elements of Wavelength, together with a couple of other movies, at length in the paper cited above.)

Everyone's doing a lot of looking. But not at the goddamn cup.

Throughout this scene, the camera returns obsessively to the cup, tracking it from Madame Sebastian to Alicia, keeping it in the foreground of each shot on Alicia, watching it travel from the table to her lips and back. Meanwhile, the characters are all infuriatingly unwilling or unable to even glance at it.

Finally, though, all the points of the joint attentional triangle are connected in a flurry of activity: the Sebastians both switch from covert to overt attention to the cup; Alicia responds by staring at the cup in dawning awareness; and finally she looks with equal alarm at the Sebastians.

So that's (a very rushed glimpse of) how a classic suspense scene leverages our desire to complete a joint attentional triangle to give visceral tension to a sequence and to generate suspense. Now on to Wavelength, which sets up a similar dynamic, but in a different way. In 43 minutes, it offers just four snippets of human activity, each very brief. At the beginning, three people move a bookshelf into the apartment, then leave.

Two minutes of traffic noise. Then two women enter and chat briefly as “Strawberry Fields Forever” plays on a transistor radio. They leave.

8 minutes more... slowly zooming, light/exposure shifting, traffic sounds supplanted by monotonic hum.

At 16 minutes, a clattering sound from offscreen. A man enters, then collapses on the floor. You might like to see more! You might like the camera to acknowledge the interest of this scene. But like Alicia, it will not turn its attention to this item of interest. Its focus is elsewhere.

The zoom continues; very soon he's no longer in frame at all.

In the final event, the female resident of the apartment speaks on the phone about the dead stranger in her apartment, whose body now lies off camera. (The hum continues at the same time. It's getting hard to hear.) She says she'll go downstairs, and exits. This is about halfway through the running time.

A progression of different sound wavelengths ensues, moving from low to high frequency, and the bank of windows comes closer and closer, leaving more of the apartment outside the viewing range. Gradually, the camera's ultimate object of interest comes into focus—a photograph hanging on the wall there. The final quarter of the film pulls into the photo.

Like so.

Suspense is not just about what we know or don’t know about outcomes in story worlds. That special feeling of concern and anxiety can be (and very often is) fostered by the construction of unconsummated triangles of joint attention. (There are lots of other ways to do it, too.)

Manipulations of joint attentional dynamics involving the camera are a major locus of filmic experimentation and innovation, much as manipulations of other elements of the triangle are characteristic of high modernist experimentation in prose. (The history of continuity editing in Hollywood film, for instance, is also a history of experimentation and innovation around joint attention on screen.)

In Wavelength, the camera manifestly and inexorably declines to share your ordinary human categories of interest, and there is nothing you can do about it. This kind of tension is key to cinematic suspense—and frustration! (This film, you may have gathered, is very big into the aestheticization of frustration.)

The effect is of course very much predicated on audience engagement—Wavelength illustrates the degree to which suspense only emerges when we enter into the conditions of the work, when we invest it with our own protracted attention. If you bounce, there’s no suspense. The tension is broken.

Thanks very much to Merja Polvinen and Dan Irving for organizing this panel, to my collaborator Todd Oakley for his contributions to work that I drew on for the centerpiece of this talk, and to you for your kind attention.