Something mechanical in something living

The mechanical comic vs. BigDog
February 22, 2016

By Gregg Murray

Why do people find Army robot BigDog creepy but C-3PO funny? It’s not just because BigDog lugs around equipment for killing people and C-3PO delivers whimsical one-liners. Okay, that may be part of it.

In the 1970s, roboticist Masahiro Mori, building upon a fascinating essay of Sigmund Freud, suggested that as robots become more human-like they induce an increasingly eerie response from human beings. This is the infamous “uncanny valley.” Freud thought that uncanny automata teetered between familiar and unfamiliar — the heimlich and unheimlich — thus causing the effect. In his paper, Mori presented a graph that dips into a “valley” as this uncanny effect becomes stronger.

A perceptual mismatch

In April 2015 a team of umlauted computer scientists (Jari Kätsyri, Klaus Förger, Meeri Mäkäräinen and Tapio Takala) writing in open-access Frontiers In Psychology journal evaluated competing hypotheses of the uncanny valley based on empirical evidence.

One theory is that its effect is caused by difficulty in categorizing a “humanoid” as either human or droid. This idea gained wide acceptance. But what they found better evidence for is that the uncanny effect is caused by inconsistent human likeness.

Uncanny bots are perceived as bots, and uncanny humans are perceived as humans, yet their crossover characteristic — or characteristics — prevent a continuous pattern of perception.

It’s not because we can’t tell the difference. We are never fooled into thinking, not for a nanosecond, that C-3PO is a human, even as he exhibits certain human characteristics. We know BigDog is a not human, but the way he’s walking is not robotic. It is too human to be robotic.

It appears, as Kätsyri et al would assert, that there is a “perceptual mismatch between artificial and human features.” Something doesn’t quite compute, but we still know what we’re looking at. That’s uncanny.

When a human being makes a physical gaffe, it can produce a comic moment. Henri Bergson’s “Laughter: An Essay on the Meaning of the Comic,” published in 1900, looks at why we laugh at physical comedy, especially in the plays of Molière.

His unique study tells us something about the early twentieth century’s affection for automata, but it may also help us understand the uncanny valley.

Laughable situations, from a man stumbling in the street to marionettes and puppets in the theater, are related to automatism, Bergson asserts. And it’s not uncanny.

The mechanical comic

A man, running along the street, stumbles and falls; the passers-by burst out laughing. They would not laugh at him, I imagine, could they suppose that the whim had suddenly seized him to sit down on the ground. They laugh because his sitting down is involuntary.

Bergson thinks that the comic is “something mechanical in something living.” Much of Molière’s political satire highlights the mechanical, insensitive responses of the aristocracy to issues facing what would come to be known as the tiers-état, or Third Estate.

Those unfamiliar with Molière may simply consult the infamous anecdote about a “great princess” now attributed, if apocryphally, to Marie Antoinette. When told that French peasants were hungry, Antoinette suggested that they eat cake. The entire episode is perhaps from Jean-Jacques Rousseau’s imagination, but it has a humorous effect because hers is an automatic response.

This is obviously a different effect from the eeriness of the uncanny. Here, there is no temporary confusion about the identification of a non-species characteristic. She is a human being who isn’t able to act like a human being for some reason. She is so out of touch that it’s funny. Bergson thinks that such humor acts as a social corrective.

Why BigDog isn’t funny

Bergson’s observations are relevant to categories of contemporary American comedy. The archetype of the Yes Man, for instance, represents a kind of robotic moron. The Dumb Blonde as well. This latter figure is particularly indicative of American culture because its real-world analogue is imagined by the continued presence of a heteropatriarchal society that secretly desires her attributes. The “Dumb Blonde” doesn’t exist. She is willed into being to criticize a set of behaviors. The intent, I would assert, is to demean women.

Regardless, from the Yes Man to the Dumb Blonde, Bergson would argue that such caricatures are subjects of ridicule because being human ought to involve something more than these mechanical displays. We correct their behavior through our laughter.

But BigDog does indeed exhibit qualities — namely the way it walks — that do not produce the effect of physical comedy. Its human-like movements carry no corrective to what walking ought to look like. Something is out of place, some “perceptual mismatch” between its clearly robotic body and its all-too-human movements. Unlike Bergson’s stumbling man, there is nothing funny, nothing enlightening, about the eerie moment when BigDog buckles to the earth floor.

Gregg Murray is Assistant Professor of English at Georgia State University and the editor of Muse /A Journal. He is a practicing poet, literary critic, and scholar. Amy Kurzweil is a cartoonist, writer, and teacher living in Brooklyn. Her graphic memoir, Flying Couch, will be published in October 2016 with Catapult/Black Balloon. Preorder now. Enter the Promo code: UNDERTHECOVERS for 25% off, now until April 15.


Abstract for A Review Of Empirical Evidence On Different Uncanny Valley Hypotheses: Support For Perceptual Mismatch As One Road To The Valley Of Eeriness

The uncanny valley hypothesis, proposed already in the 1970s, suggests that almost but not fully humanlike artificial characters will trigger a profound sense of unease. This hypothesis has become widely acknowledged both in the popular media and scientific research. Surprisingly, empirical evidence for the hypothesis has remained inconsistent. In the present article, we reinterpret the original uncanny valley hypothesis and review empirical evidence for different theoretically motivated uncanny valley hypotheses. The uncanny valley could be understood as the naïve claim that any kind of human-likeness manipulation will lead to experienced negative affinity at close-to-realistic levels. More recent hypotheses have suggested that the uncanny valley would be caused by artificial–human categorization difficulty or by a perceptual mismatch between artificial and human features. Original formulation also suggested that movement would modulate the uncanny valley. The reviewed empirical literature failed to provide consistent support for the naïve uncanny valley hypothesis or the modulatory effects of movement. Results on the categorization difficulty hypothesis were still too scarce to allow drawing firm conclusions. In contrast, good support was found for the perceptual mismatch hypothesis. Taken together, the present review findings suggest that the uncanny valley exists only under specific conditions. More research is still needed to pinpoint the exact conditions under which the uncanny valley phenomenon manifests itself.