It's Smart to be a Dummy

Event Scenes

It’s Smart to be a Dummy

A Note on the Crash Test Dummy

“What interests us is solely the relation between mechanization and death. […] Both are involved in the mass production of meat.”

– Siegfried Gideon

The ad says it all: “Sensitive, Slouching, Bald. (How Much More Lifelike Can They Get?)”1 Crash test dummies sit in a cut-away Lexus as if X-rayed, revealing both the skeletal automobile and the spectral interior dummy of its passengers. The rhetoric of lifelikeness develops force from the double sense of sensitive: the instrumental sensitivity of the dummy’s role in the spectacle of the simulated automobile crash; the photographic sensitivity of the dummy body, able to image this interior space of testing and safety. Lifelikeness extends to the wired physiognomy, rendering carefully posed limbs slouching, electrode-implanted artificial skin bald. Sensitive, Slouching, Bald: a generic description of a range of possible drivers, enabling the transformation from crash test dummy to human driver. When the ad queries How Much More Lifelike Can They Get?, the question reads as a challenge, as perhaps the test itself, determining the quanta of life that dummies Get. The transformation requires that we take the question seriously: dummies are lifelike enough for us to believe in safety and keep on driving. The ad continues: “In the realm of safety crash-testing, man is not the measure of all things. Dummies are. So it follows quite logically, the better the dummy, the better the car.” If safety is one feature of an automobile, here it is a real qualitative essence. Just as safety forms the ontology of our being in the driver’s seat, the dummy is the essence of the human in this realm of safety crash testing, a realm marked by the assured logic of measurement, cause and effect, a realm of ends. In the purity of the crash test is the appearance of a sacramental writing circuit, the dummy’s animation become a vehicle guaranteeing the safety of all drivers, its sensitivity inscribing the test in a range of measurement devices.

What is safe? Motown claims to make good on the safety promised by ubiquitous crash test dummies, yet the test drive continues. As a metonymies for safety, advertising dummies increasingly cross-over to articulate what it is to be human – no longer a stand-in, but a double. Richard Wolkomir writes in Smithsonian: “It is OK to anthropomorphize crash-test dummies because anthropomorphism is what they do: they are our stunt doubles. They repeatedly crash into walls to show what would happen to us.”2 This thumbs up to anthropomorphization is tied to the predictive power of the dummy, its prophetic immediacy in the crash a supra-temporal projection of the truth of human safety. The celebrity status of the dummy marks an autonomy of safety as a discourse, the living safety zone as a virtuality, freed of bodily media. From an initial use of living humans and fresh cadavers, test scores have risen through progressive refinement of subjects. “[L]oaded with sensors and accelerometers,” and a zombie-like refusal to stay dead on the charts, by far the biggest hits have been dummies.3 The crash test dummy is a mode of processing humans into meat and has an unavoidable impact on our survival. Wolkomir writes: “Dummies are simulacra of us, their proportions exact, their transducers analogous to our nerves. They are, as one engineer puts it, injury-measuring devices. They are just machines. But they look so much like us it is upsetting to see them take our hits.” At the center of contemporary safety is this mimetic and its link-up to human emotion and upset, predicated on a visceral reaction to dummy injuries. It may be better to revise the question of what is safe, to instead ask how safety is produced: what are the techniques that invest the crash test dummy with representative power? While the development of dummy technology is tightly coupled to statistical analyses of human accidents, it is not immediately clear what computer-measured “boppings in the chest with a 51.5-pound pendulum” or “head hits with about 2,000 pounds of pressure on the forehead” signify when performed on an automata. The terrible range of simulated accidents pivots on the installation of an organic converter: from data to human safety. The result is a nation-wide demarcation of public safety in terms of mass and meat. Who will survive, who will be roadkill?

What is safe? An etymology of health (of ~salvus~) as a sort of normative organicity, safety is a promise of life. A promise, perhaps, that in as much as we are safe, we survive. Being safe – safety as being. Safety’s ontology is the implied promise of the survival of *something* and simultaneously an epistemology that tests for the conditions of this survival. This simultaneity sets out the unstable conditions for safety testing. The test produces a zone of simulation whose artificiality will be the condition of the real it simulates. One should speed-read Nietzsche’s “calculability,” Freud’s “reality testing,” or Ronell’s “test drive,” of which she asks: “Why is our security – whether or not you are prepared to admit this – based on testability?”4 Safety is a station identification of a public survival, its promise issued from the emergency broadcasting system. The test asks: can there be enough knowledge, or knowledge necessary enough, to end the contingency of the accident, the catastrophe of the real? The answer is yes, if you believe crash test dummies.

“In the beginning there was Parachute Man.”5 Dummies have been in on the test results since the genesis of crash produced life, the test pilot Parachute Man landing in the driver’s seat in 1946, but their ascendancy over other test media has been “Darwinian.” The exegesis of the National Highway Traffic Safety Act of 1966 localized the safe into zones and vectors of force articulated into a range of legality on the surface of the dummy body, and led to the development of specific safety technologies (seat belts; air bags). L. M. Patrick’s 1973 essay “Dynamic Response of Humans and Dummies” marks an early threshold of human-dummy replacement, a historic test inscribing the dummy’s relative advantage, a diagram of the writing circuit. Patrick describes testers inscribing “points and/or lines marked on the body components to permit the angular and translational motion of the component to be recorded.”6 “Translational” is both the spatial skid of a body in space and the semantic transfer between signifying systems, a conversion from the force of accident into readability. The marks both prepare the dummy for receiving the crash and are re-marked with the event of the crash itself. The already-written dummy as crash receiving apparatus allows the calculus of “angular and translational motion” to record what is human. Instant replay reads the impact: life as articulated force, the mediality of the dummy body making visible silent statistical patterns.

Attempts to crash pass the driver’s test using live subjects consistently fail due to what the literature of crash-testing calls a “subjective non-injury (reversible injury) impact level.”7 In other words, you will not be able to report on your injury if you are dead; your body will be unreadable, tenderized meat, your voice will be silenced; but, on the other hand, if you are not fatally injured, your injury will not be valid for the test, and thus equally unreadable. The “subjective non-injury (reversible injury) impact level” defines the living subject in the prose of the crash test. Subjectivity is this inner limit of non-injury: that’s life. In turn, the use of corpses as test subjects reverses this principle in producing only irreversible injuries. All injuries are already fatal to corpses. The crash opens the crypt beyond the pleasure principle loop into the realm of the living dead. If live subjects can not speak except from within an unreadable and unsafe safety, the corpse as test case speaks only on the condition of an artificial safety: we already know it’s already dead. The periodic return of the dead in the history of crash testing marks a belief in this technical readability. While the wounds of the living are never fatal, the wounds of the dead always are – the subjects remain dead for all possible tests – and thus have a certain consumability for testing purposes. The dead are clearly subject to a fatal crash, their animation keyed to their guaranteed fatality. This cannibalization returns as indigestion in scandals over the use of cadavers in crash testing. The Los Angeles Times reports: “Chastised by the Vatican and under fire from the government, Heidelberg University promised […] to prove it got relatives consent to use cadavers in car crash tests partially financed by Washington.”8 This German connection re-runs a Frankenstein dream latent in the use of animated cadavers. The dead live on the far side of animation, their mobility co-opting Resurrection theology. The Vatican’s furor taps into this grave robbery. The sacrifice of the living/dead body-fetish is the material guarantee of life in the beyond. The ability of the corpse to take a hit makes it articulate, the ventriloquism the Vatican fears mouthed by the experimenters. “You can’t use a department-store mannequin and run it in these tests. It doesn’t tell you anything,” says one researcher.9 Ears pressed close to the corpse, the researchers listen for whispered test results. Yet cadavers lose something in so-called “dynamic performance.” Bodily media decay test results into indeterminacy.10 The terrible presence of death marks the corpse as unsafe, already dead. The cadaver can only embody those who do not survive. The non-reversibility of the crash-injured corpse is a sign of its death – the wound that never heals, the voice that is always a recording. The cadaver embodies our mortal injuries, and in this invalidates its test for public consumption: the statistical non-survivors of crashes are already corpses, leaving a bad taste in the mouth.

The living can not crash as such. Within the impact level the subject survives to report back, but, for precisely this reason, what is produced is a non-injury or a reversible injury. The articulated human voice emerges from the untestable inside of safety. Only the catastrophe of irreversible damage produces the media converting noise into data, a surface of life as readability, suspended between the happy speech of the survivor and the battered remains of meat. Unable to pass the test because unable to take the test, to live is never to be safe. To be live, it must be Memorex.

The dummy is the king of the road. L. M. Patrick writes: “While there is merit in trying to duplicate the human insofar as possible by the dummy, it may be necessary to sacrifice some of the degrees of freedom for repeatability.” Sacrifice occurs precisely in the gap of safety – repeatability a continuous feed type, a Heidelberg press, a meat grinder, spitting out the dummies that sell us safety. A long run is a guarantee of success. The aporia between freedom of live subjectivity and inert dead matter is bridged in the acceleration to repeatability. The ability of the dummy’s limbs to pop into place is the guarantee of a drive without end or obstacle, of an infinitely repeatable testing without proving a thing, creating the real as a safety zone. The dummy is an ideal media for recording human sensation. This ensures our ability to mourn their loss: no one remembers the dummy.

Crash Injuries is the medical textbook used both by the crash testers and by J. G. Ballard in his vehicular fantasies of accident and arousal. “That is the ultimate book – all those comparisons of facial damage in rollover, comparing 52 Buicks with 55 Buicks – bizarre connections.”11 Pumped up on medical documentation of crash injuries, Ballard tails the official discourse of testing. “Sequences showing auto-crash victims brought about a marked acceleration of pulse and respiratory rates. Many volunteers become convinced that the fatalities were still living.”12 The textbook pre-exists the crash: the analogical correspondence-effect of images to inscribed body produces an index of testing truth, hallucinated verification on the dummy surface. Crash Injuries is a technology for the medicalization of the crash. The image-shift from book to organic virtuality is the optical authorization of the writing circuit. For Ballard, “[t]he intimate time and space of a single human being [are] fossilized”13 in the aesthetics of the automobile surface; “one’s dealing with fundamental entities like one’s own musculature, one’s own sort of highly conventionalized response to one’s own body, one’s tenancy in time and space.” Ballard’s characters pursue this phenomenological recovery with an auto-erotic mania combining nostalgia for the body in a time of the virtual traveler with a realization of this nostalgia in the crash. In zero landscapes of concrete highways, they enact their obsession with the repeatability of famous crash deaths: Albert Camus, Jayne Mansfield, JFK. The somatic dummy arising in the crash is the progeny of erotic investment in the vehicle; “like bees in the plant world, men have always been the sex organs of the technological world”14 “Auto-amputation” is the eroticized impact, cathexis in the “the latent sexual content of the automobile crash.” In such erotics, one is in the position of already being a dummy.

Automobiles are apparatuses for producing meat; they are intended to crash – it’s built into the design. We drive to produce ourselves in the crash. There is a purgative, blanching pleasure in driving by the wreck. Between the driver and the corpse is the dummy: survival requires that other drivers become corpses, absorbed by the survivor as the knowledge of the dummy, the knowledge of having survived to drive again. Statistical safety is in being public, that is, being a dummy, where the alternative is to be a corpse. The survivor is the one who can drive away.

The American picaresque completes itself in the automobile accident (The Great Gatsby, Thelma and Louise). The 70s thematize the crash. Evel Kneivel, Vanishing Point, The Dukes of Hazzard: all exist for the production of automobile crashes. The contemporary is marked by a virtualization of the crash and an autonomy of the dummy. “Dummies Get Smarter for Car-Crash Tests” reads The New York Times front page, describing the increased precision of testing.15 They reproduce: male, female, child, and “the dummy tummy” for pregnant female dummies. The article drives home the vitality of crash-testing for automobile safety. The dummy body’s increased sensitivity is figured as intelligent human thought. In the words of the Lexus ad: “they are extremely sophisticated machines that can do more than simply take a hit. In fact, they’re designed to act human. As such, they can think… And they can feel.” Parenthetically, it adds that thinking and feeling result from their being “(equipped with computers.)” The day of the living dummies: the crash test is a state-sponsored bio-mechanical laboratory for artificial life and intelligence to replace the indeterminate non-repeatability of the human. Replaying the Turing Test of artificial intelligence, safety is readable when the dummy’s injuries are indistinguishable from our own. The transparency of the dummy is illuminated and propelled by AI, wetware, and a range of intra-corporeal prostheses. Computer-driven calculation digitizes dummy representation, and in this coupling, the interior of the computer becomes a privileged zone, a black box or Turing machine converting life into readability. From now on all crash tests will occur on microprocessors, in the stacked space of data registers. The new LS-DYNA3D computer coding protocol is designed specifically for dummy modelling: it “is easy to use, robust, and validated.”16 Biological ramparts: dummy representation is already a more complex machine, a combinatory device suturing animal and technology. “Their world is Darwinian. The dummy’s ecological niche is car smackups. Data from the crashes may lead to new designs. Accident statistics can change. Federal rules are altered. In response to these factors, dummies mutate. It is survival of the most sensitive, the dummy whose instruments and physical reactions best mimic a humans. And so dummies are evolving to be more like us.”17 The analogical system links the smudged makeup of the dummy’s inscribed skin to the stream of raw data running from the dummy’s organs. The computer records 4,000 times during the two-fifths-of-a-second “measurement interlude” – that is, during the crash.

Suzanne Hilton’s It’s Smart to Use a Dummy provides a valuable history of dummy-human replacement, from the waxwork to the hot nuclear test site dummy. She writes: “The dummy […] cannot indicate yet whether a real person would have suffered brain damage from such an accident.”18 The dummy gets as lifelike, or as much life, as it can get. The future tense projected by dummy testing – the salvation implicit in the ~salvus~ of safety – is the possibility of some other test, a supra-temporal moment in which the test will be passed and safety achieved. This projection removes the test from real time into dummy time. The crash includes the future as a moment within its data. What is safe? The GM engineer is asked “did the dummies survive?” He replies: “I don’t know… I won’t be able to tell until I look at the numbers.”19


1. Lexus Advertisement, The New Yorker, December 6, 1993, 10.

2. Richard Wolkomir, “Sitting in our stead: crash dummies take the hard knocks for all of us,” Smithsonian, 26:4, 7/95, 34.

3. New Yorker 10-11.

4. Avital Ronell, “The Test Drive,” Deconstruction is/in America, Ed. Anselm Haverkamp (New York: New York University Press, 1995), 205.

5. Wolkomir 34.

6. L. M. Patrick, “Dynamic Response of Humans and Dummies,” Human Impact Response: Measurement and Simulation (New York: Plenum Press, 1973), 19.

7. R. G. Snyder, W. M. Crosby, C.C. Sno w, J. W. Young, and P. Hanson, “Seat Belt Injuries in Impact,” The Prevention of Highway Injury (Ann Arbor, MI: University of Michigan Press, 1967), 199.

8. “Auto Safety Crash-Testing Ignites F uror,” Los Angeles Times, 11/25/93.

9. “Perspectives,” Newsweek, Dec ember 6, 1993, 17.

10. J. D. States, “Trauma Evaluation Needs,” Human Impact Response: Measurement and Simulation (New York and London: Plenum Press, 1973), 9.

11. “Interview with JGB by Andrea Jun o and Vale,” Re/Search, No. 8/9, 1988, 10.

12. J. G. Ballard, The Atrocity Ex hibition (San Francisco, CA: Re/Search Publications, 1990), 97.

13. J. G. Ballard, Crash (New York: Vintage Books, 1985), 12.

14. Marshall McLuhan, Understandin g Media (New York: Signet, 1964), 196.

15. “Dummies Get Smarter for Car-Cras h Tests,” New York Times, 1/23/94.

16. See the LS-DYNA3D Web page at

17. Wolkomir 32.

18. Suzanne Hilton, Its Smart to U se a Dummy (Philadelphia, PA: Westminster Press, 1971), 86.

19. The New York Times 19.