Photo by Sean Carroll
It is a truth universally acknowledged—at least by those of the feline persuasion—that an empty box on the floor must be in want of a cat. Ditto for laundry baskets, suitcases, sinks, and even cat carriers (when not used as transport to the vet). This behavior is generally attributed to the fact that cats feel safer when squeezed into small spaces, but it might also be able to tell us something about feline visual perception. That’s the rationale behind a study in the journal Applied Animal Behaviour Science with a colorful title: “If I fits I sits: A citizen science investigation into illusory contour susceptibility in domestic cats (Felis silvers catus).”
The paper was inspired in part by a 2017 viral Twitter hashtag, #CatSquares, in which users posted pictures of their cats sitting inside squares marked out on the floor with tape—kind of a virtual box. The following year, lead author Gabriella Smith, a graduate student at Hunter College (CUNY) in New York City, attended a lecture by co-author Sarah-Elizabeth Byosiere, who heads the Thinking Dog Center at Hunter. Byosiere studies canine behavior and cognition, and she spoke about dogs’ susceptibility to visual illusions. While playing with her roommate’s cat later that evening, Smith recalled the Twitter hashtag and wondered if she could find a visual illusion that looked like a square to test on cats.
Smith found it in the work of the late Italian psychologist and artist Gaetano Kanizsa, who was interested in illusory (subjective) contours that visually evoke the sense of an edge in the brain even if there isn’t really a line or edge there. The Kanizsa square consists of four objects shaped like Pac-Man, oriented with the “mouth” facing inward to form the four corners of a square. Even better, there was a 1988 study that used the Kanizsa square to investigate the susceptibility of two young female cats to illusory contours. The study concluded that, yes, cats are susceptible to the Kanizsa square illusion, suggesting that they perceive subjective contours much like humans.
But the 1988 study was conducted in the laboratory and “primed” the two feline subjects via standard operant conditioning methods. Smith wanted to design a similar study that increased the sample size and observed the cats’ behavior in their natural environment—which is less stressful for cats than a lab environment—with no advance priming. A “citizen science” project involving cat owners recruited on Twitter seemed like just the ticket, especially given the ongoing COVID-19 pandemic. People were spending a lot more time at home with their pets, and they were likely to have more time to conduct the trials.
The only supplies participating owners needed were a printer with black ink, printer paper, scissors, tape, and a ruler, plus sunglasses and a digital camera or smartphone to record their cats’ behavior. Smith serially sent participants six randomized daily stimuli to print out and set up on the floor, per instructions, while the cat was not in the room. The stimuli included a simple square, the Kanizsa square illusion, and a Kanizsa control in which the Pac-Man mouths faced outward instead of inward. All dimensions were such that a cat could comfortably sit or stand inside with all its limbs without being able to sprawl.
The cats would be permitted into the room, and the owners would don the sunglasses and avoid interacting with their pets so as not to give the beasts any cues. The humans would videotape the cats’ behavior with the pairs of stimuli and upload the videos to a shared Dropbox for the project. If the cat sat or stood with all its legs inside the contours of a stimulus within the first five minutes, the owners would stop recording and make note of the chosen shape. If the cat didn’t select one of the stimuli in the first five minutes, the trial would end.
Although some 500 pet cats and their owners expressed interest, only 30 completed all six of the study’s trials over the course of the two-month study last summer. Of those, nine of the cats selected at least one of the stimuli by sitting within its contours (illusory or otherwise) for at least three seconds—a pretty good duration given the notorious fickleness of cats. As for preferences, cats selected the Kanizsa illusion just as often as the square; they selected both of those more often than the control stimulus. In other words, the cats treated the illusory square the same way they treated the real square.
“It’s the presence of the contours, either in the Kanizsa square or in the real square, that causes cats to sit inside, rather than the presence of shapes on the floor,” Smith told Ars. “Brains are very sensitive to contours that differ in luminance. Vision has evolved to answer questions having to do with boundaries and contours.”
The study comes with the usual caveats, notably the final small sample size (the result of participant attrition, a common challenge with citizen science projects). Smith and her co-authors also suggest replicating the study in a more controlled setting, despite the advantages gained from conducting the trials in the comfort of the cats’ own homes. “For the sake of cats, the home was really ideal, but otherwise, for the sake of science, it is best to do things in controlled settings [like a lab],” said Smith.
Smith and Byosiere are also keen to adapt some of the latter’s work with dogs and visual illusions to the study of cat behavior and cognition. "Cat cognition research is certainly lacking in comparison to domestic dogs," the authors concluded. “Although the reason for this is unclear, the use of citizen science as a precursor to in-lab investigations of cat cognition could greatly help bridge this divide.”