Some people get carried away with their anthropomorphism. Just the other day, I saw a lady pull into the grocery-store parking lot in a very feminine VW Beetle indeed: above its “eyes”—which is to say, its headlights—she’d applied a set of long black plastic lashes, which conveyed the impression that it had been stolen from the lower-end fleet of the Kardashians. The incident reminded me of a curious recent study that revealed just how humanlike we think cars are beneath their hoods. We’re also largely oblivious to the many subtle attributions that we’re routinely making to these mindless, mobile cauldrons of steel.

Before we get ahead of ourselves, let’s look more closely at this notion of cars with “faces”. It’s not that vehicles actually have faces, of course, but instead that we so readily perceive them as so. Cars, Pixar’s animated franchise, simply played off this well-known effect. You can look at it as a processing error, a mental misfire, or a cognitive fluke to see faces in inanimate objects this way. But some theorists, most notably the cognitive scientist Stewart Guthrie, claim that these glitches are the result of our brains working exactly the way they’re supposed to, reflecting an “error-management” device that has rendered us extra-vigilant to cues of agency in the environment—so vigilant, in fact, that we detect faces and bodies that aren’t there, from the Virgin Mary in burnt toast to alien faces on the surface of the moon to long-lashed coquettes in Volkswagens.

It may seem odd, but in fact it makes perfect adaptive sense. In an environment teeming with countless other living organisms (“agents” in technical speak)—many of which, including other human beings, could have proven deadly if overlooked in the harsh landscape—a trigger-happy perceptual bias that only occasionally had our ancestors confusing boulders for predators came with negligible risk. Making mistakes in the opposite direction, however, may well have proven fatal.

In contemporary environments, our eyes traverse a sea of human-made modern objects that are also capable of generating false positives. Most of us, fortunately, don’t actually think that unplugged electrical outlets are “surprised,” for instance, but it’s still easy enough to see such a facial expression scattered about our walls. And when it comes to cars, unless we’re disturbed, we don’t really believe that such objects have minds and personalities, yet we can���t help but regard them this way nonetheless. One reason for this, according to the anthropologist Sonja Windhager and her colleagues in a 2012 article from Evolution and Human Behavior, is our shared morphology with cars. If you look at a car from the front, they point out, you’ll see a bilaterally symmetrical “face” with a windshield for a forehead, headlights for eyes, side-view mirrors for ears, a grille for a nose and the additional air-intake feature for a mouth.

In fact, these researchers showed that our tendency to see humanlike characteristics in the front-ends of automobiles is so sensitive that simply by altering the physical shape of these “facial” features—elongating the windshield, for instance, or narrowing the headlights—observers’ views of the car’s temperament and personality change as well. More to the point, they change in predicable ways; just as we evolved to make reliable and automatic inferences about hidden characteristics of a human being based on the biometric aspects of their faces, our perception of cars is similarly affected. Compared to those of adults, for example, children’s faces have telltale features such as a large forehead, large eyes, thin and arched brows, a shorter nose, and a smaller mid- and lower face: an admixture of cues that prompt a normal caretaking response. Manipulate the front-end appearance of cars’ “faces” in a similar way, and you’ll perceive an automobile with a dependent, childlike innocence about it.

How did Windhager and her co-authors test their hypothesis that the same cognitive system involved in our everyday processing of faces is at play in our implicit reasoning about cars? They showed a set of standardized, black-and-white mock-up images of the front ends of 46 different car models—from the Ford Galaxy to the BMW 645ci to the Kia Picanto to the Mini Cooper to the Maybach—to two groups of subjects in a random order, and simply had each participant rate every car on 19 different traits. These trait scales covered the subject’s perception of the target vehicle’s “age”, “sex”, “attitudes”, “personality” and “emotions”. It does sound strange to ask someone how “neurotic” or “aroused” a car appears, but the point here was that if the subjects in the study were in general agreement about which cars corresponded with which traits, and in a way that conformed to the known biometric markers of human face-processing, it would be evidence that the automobiles’ different “faces” led to this consensus.

First, however, the scientists had to rule out the obvious possibility that any convergent ratings wouldn’t merely be the result of social learning. The advertising slogans are inescapable—if you have a radio, TV, or computer, your brain has bathed in them. Marketing campaigns, not the cars’ appearance, may influence responses. In addition, there are cultural stereotypes about driver attributes that could also confuse things: more women drive minivans, for instance, and men drive trucks.

To address these potential biases, Windhager recruited two very different groups of subjects. The first group, from Vienna, Austria, consisted of folks like most of you reading this now. That’s to say, they came in or around a reasonably modern city; people who’ve likely been subjected to decades of crafty car adverts. The other group, by contrast, consisted of rural villagers from Ethiopia; they hailed from places where, according to the authors, “street transportation is dominated by small trucks, off-road vehicles, some taxi buses, horses and carts, donkeys and people carrying their goods.” In other words, there’s not much choice in the auto market in towns like Dukem and Laga Tafo, Ethiopia, and car advertising simply doesn’t exist there. Moreover, there’s not enough automobile diversity to develop stereotypes. The car models these rural villagers saw in the study, therefore, were completely new to them. (In fact, only a few of these subjects had ever even used a computer mouse, and so the study needed to be administered to them as an old school paper-and-pencil test.)

As Windhager and her colleagues predicted, there was a striking similarity in the patterns of car ratings made by the Austrians and Ethiopians. “Even though street scenery could not be more different,” write the authors, “we found a high cross-cultural consistency in child-adult, female-male, and submissive-dominant attributions to cars.” Moreover, the ratings on the trait dimensions matched known human face-processing patterns. For example, cars with more slit-like, laterally extended headlights (i.e., “eyes”) evoked more adult, male, dominant ratings than those with round headlights closer to the center of the vehicle.

Car manufacturers might want to take note of this work, since the aesthetic airs given by their various designs may or may not be what they’re going for. Male drivers may unknowingly rule out the Nissan Micra, for example, because it’s got the most “feminine” look. To be the owner of the most “childish” vehicle (the Toyota Aygo) probably doesn’t feel so great either. Among the “saddest” car faces is the melancholy visage of the Kia Picanto, while the Honda Civic rates among the most “neurotic”.

There were a small number of cultural differences. Overall, the Ethiopians saw all the cars as more “arrogant” than did the Austrians, for instance. And as a group, the latter had more extreme opinions of the different car models than did the former. But these differences were astonishingly minor in the scheme of things. “To repeat,” the authors hammer in, “the [main] pattern looked the same for Austrian and Ethiopian ratings.”

The whole research program may sound a bit silly to you. But in fact, there are practical applications to Windhager’s research program that make it more important than it appears at first glance. Since so much of human decision-making is influenced by automatic inferences that operate below the level of our conscious awareness, having an understanding of our species’ intuitive perceptions of cars—these ubiquitous, dangerous inanimate objects in modern society that are all around us—could ultimately save lives. “Do we change lanes sooner when an adult, dominant car appears in the rear-view mirror?” the authors ponder. “Do children perceive cars as agents with eyes and therefore assume they see them anyway when they try to cross the street?” These questions and others are fodder for follow-up research in this area.

Windhager and her team are on it. Until we know more, just beware of any run-ins with the Chrysler Crossfire… that’s one angry son of a bitch, apparently.

Jesse Bering, an Associate Professor at the Centre for Science Communication at the University of Otago, New Zealand, is the author of Perv (2013), Why Is the Penis Shaped Like That? (2012) and The Belief Instinct (2011). Follow him on Twitter @JesseBering.