This article is from the WeChat official account:neural reality (ID: neureality)< span class = "text-remarks">, author: Catherine Offord, translation: Mollie, proofreading: Du Yu, editor: Orange Soda, “real nervous” cognitive science translation team produced, the title figure from the visual China
At a barbecue party in the suburbs of Melbourne, Australia, Sarah Bissoyer (Sarah Byosiere) had the “illusion of seeing the dog Picture” idea. It was 2015. She just got a master’s degree in canine cognition in the United States and came to La Trobe University (La Trobe University) Exchange visits. At the barbecue party, she met a group of psychologists who studied how the human brain perceives optical illusions. When chatting with them, Bissoyer suddenly realized: The same method may open a window for us, let us understand how dogs understand the world they are in, and how their perception is different from ours. “We asked such a crazy question: can we show illusion images to dogs, and do they produce illusions?” she recalled.
This idea, which at first sounded like a gimmick, quickly turned into a formal proposal. Philippe Chouinard explained that (Philippe Chouinard), psychologists often use optical illusion images to study how the human brain extracts the external world. What shortcuts have been taken when information. Chouinard is an expert in illusion and one of the La Trobe University psychologists at that barbecue party. He said that shortcuts are usually a reliable way for the brain to quickly obtain information, but sometimes shortcuts can also make mistakes-for example, causing us to “see” something that does not exist, or feel that two identical objects are of different sizes. If we can find out when and how this type of error occurred, these findings can provide some clues to help us understand humans or other animals (such as home Canine, Canis familiaris) how the cognitive function of is realized.
In the Ebbinghaus-Tichenner illusion, the two circles at the center have the same diameter. In the eyes of humans, the circle surrounded by a smaller circle looks bigger. According to a study, this illusion is just the opposite in the eyes of dogs. THE SCIENTIST
Bissoyer decided to delve into this issue in her PhD research. She came to La Trobe University and started work under the guidance of canine researcher Pauline Bennett (Pauline Bennett). She designed a device that allows dogs to convey what they perceive. The research team built a small laboratory in which the touch screen displays various optical illusion images, and dogs can interact with the screen with their noses. Bissoyer said that the first few animals that participated were the Lagotto Romagnolo dogs brought by volunteers. (Lagotto Romagnolo dogs), let It took them several months to get acquainted with the instrument. But their preliminary results published in 2016 were eye-opening.
The research team used several different versions of the Ebbinghaus-Titchener illusion (Ebbinghaus-Titchener illusion). In this illusion diagram, there are two circles of the same size, but because of the other circles arranged around them, they are of different sizes in the eyes of humans. Each dog has been trained-the researcher exchanged a large number of dog biscuits for their correct answer, so that they learned to tap the corresponding position on the screen with their nose, and choose the larger one from the two circles. The team also used a series of pictures of the control group. The two central circles in the picture and the overall visual stimulus are both large and small.different. They use these pictures to collect data to test whether dogs are more inclined to choose one of the pictures.
What excites Bissoyer is that experimental data show that dogs are indeed susceptible to the Ebbinghaus illusion. Faced with two circles of equal size in the illusion picture, they always choose one instead of the other. However, dogs do not fall into the trap like humans, thinking that the circle surrounded by small circles is bigger-their choice is the opposite of ours.
As the first step in training dogs to interact with optical illusions, researchers need to make them “touch an object with their nose (the one used in this experiment) It’s the red ball on the stick)” is associated with the “reward”. SARAH BYOSIERE
Bissoyer is now the director of the Canine Thinking Center at Hunter College in New York. She said that although humans and dogs have maintained a close relationship for thousands of years, these results fully demonstrate that science has little understanding of canine perception. She noticed that most studies on dog cognition (It is estimated that 3/4 of the published studies) rely on visual tasks. There are many tasks adapted from primate research. However, “we don’t necessarily know whether dogs understand the world in the same way as ours,” Bissoyer said, so researchers are sometimes not sure what secrets of canine cognition are revealed by these experimental results. Understanding how canine visual perception is different from ours, this “may have a big impact on the way we interpret the results.”
1. Visual tricks designed for dogs
Researchers can test their sensitivity to illusions by training animals to interact with screens or other devices, just like Bissoyer did in her 2016 study. Alternatively, the researcher can also choose a method called “spontaneous preference experiment”(spontaneous preference experiments) method, using plates and foods of different sizes to show dogs the same optical illusion images. In this way, the dog will be interested in food without any training.
In the Delboeuf illusion, the two circles at the center have the same diameter. In human eyes, the circle surrounded by the smaller circle looks bigger. Two studies have found that dogs do not seem to perceive any difference. THE SCIENTIST
A few years ago, Christian Agrillo (Christian Agrillo) with him at the University of Padua(University of Padua, located in northern Italy) colleagues tried the second method. They tried to arrange dog treats in a round shape to test the sensitivity of different breeds of dogs to the “Deboff size illusion.” In the classic Deboff size illusion, there are two circles of the same size, and their relative sizes are distorted by the two rings around them. Usually, humans will feel that the circle surrounded by small circles is larger than the circle surrounded by large circles-people think this effect comes from a visual quirk that leads us to round the size of the previous circle up, because it It is almost the same size as the circle, but round the size of the next circle down because it is much smaller than the circle.
In Agliro’s experiment, the researchers provided 13 dogs with two plates containing food, one meter apart. In the control group, two plates of the same size contained different amounts of snacks, and the dog had to choose between the two. In the experimental group, dogs need to choose between foods of the same size but on plates of different sizes. The research team believes that in both cases, dogs should want to choose the one that looks more. becauseTherefore, the team puts forward the hypothesis: if dogs perceive this illusion as humans, they will choose small plates because there are more snacks on them.
A few years ago, Christian Agliro and colleagues from the University of Padua, Italy conducted an experiment with this device and Suppose the dog chooses the food that looks more. They found that although people tend to feel that there is more food on a small plate, dogs do not seem to perceive a difference in the amount of food between the two plates-this result may indicate that the dog’s perception is not interfered by such illusions. KAILEY WHITMAN
But they didn’t. In the control experiment, the dogs did move to the plate with more snacks. But in the experimental group, when they chose the same amount of food on different-sized plates, Agliro said that “their performance is basically random.” He added that they did not know whether this meant that the dog was not affected by this optical illusion, or it was just that these experimental conditions could not reasonably detect this. No matter which plate they chose, the dogs that participated in the experiment were rewarded with food, so they might not have any motivation to choose the food that seemed a little more.
At the same time, Bissoyer and her colleagues published their research results. They used trained dogs and displayed images of Deboff’s size illusion on the touch screen. The research results seem to prove that dogs will not be affected by this optical illusion. Recently, Bissouye also reported that different versions of Ponzo’s illusion experiments were conducted with trained dogs and all had no effect (null results) . For humans, the Ponzo illusion surrounds the same shapes with lines and grids, which can distort our perception of the size of these shapes.
Ponzo illusions use lines and grids to distort the relative size of two shapes. Generally speaking, people will feel that the rectangle on the left in the picture looks bigger. Several studies done by a research team show that dogs do not see the difference. THE SCIENTIST
Although “the lack of evidence does not mean that there is no evidence,” Agliro said, the current research results show that dogs may not be deceived by these illusion games-this is how researchers design canines in the future One thing to keep in mind when doing cognitive experiments.
Second, interpret the difficult problem of optical illusion experiment
Researchers have found that in the face of certain types of optical illusions, dogs and humans seem to react similarly. However, several studies have also emphasized the need to be very cautious in conducting and interpreting canine perception experiments. A few years ago, researchers from the University of Lincoln in the United Kingdom did a study to make dogs and the Mu’s illusion displayed on the touch screen (Muller-Lyer illusion)Interact. There are two lines of the same length on the Miao’s illusion diagram, which seem to be different because of the direction of the arrows at both ends. The research team found that dogs trained to choose longer line segments always choose the line with the arrow inward, just like humans would make. This may mean that dogs have the same perception of this illusion as humans.
In Miao’s illusion, the length of two equal line segments is distorted by the arrows at both ends. People usually think the line segment above looks longer. A dog trained to choose a longer line segment will also choose the upper line, but this result can be interpreted as the dog chose an overall larger visual stimulus, rather than the human being that the horizontal line looks longer. . THE SCIENTIST
However, through additional control groups and rigorous data analysis, the researchers found another way to interpret the results of the experiment: dogs did not choose the inward line of the arrow based on the length of the line segment they saw. The choice is the biggest visual stimulus overall.
Researchers try to rule out this different situation in the optical illusion experiment. For example, in the Ebbinghaus illusion experiment conducted by Bissoyer and his colleagues, they used many pictures with circles of different sizes. This can rule out the possibility that dogs make choices based on the overall size of the image (not the size of the center circle), after all, this will also cause them The performance is opposite to that of humans.
In behavioral choice experiments, confounding factors have always been a problem, especially when the researcher does not know what the dog sees or notices, Emory University (Emory University) neuroscientist Gregory Berns (Gregory Berns) said. “When we design experiments, we will unconsciously design from a human perspective.” Burns has studied canine cognition for nearly ten years, and he was the first group to train dogs to sit in MRI(One of the researchers of MRI scanner). In studies that evaluate canine behavior, “If a dog does not do what you want it to do, we usually cannot determine whether it does not understand what you want it to do, or it does understand but is unwilling to do so, or if there is something else. motivation”. Burns added that research usually evaluates the behavior of dogs at the group level to find out the subtle trends in their perception, but this may cause us to ignore the perception and relationship between dogs.Significant differences in cognition.
In 2016, Sarah Bisoy and her colleagues at La Trobe University in Australia published an experiment in which the picture above was used Device. They trained the dogs to choose black circles that looked larger, and found that they all chose black circles surrounded by larger blue circles. This effect is exactly the opposite of what is seen in humans, indicating that the visual perception of dogs and humans may differ in some aspects of this illusion. KAILEY WHITMAN
However, these scholars who use behavioral experiments to study illusions retort that even in the face of these challenges, this method can at least provide a preliminary plan for exploring “how dogs see the world”. As Bissoyer explained, these studies “give us a preliminary understanding of how dogs perceive their environment visually. It can help us understand the visual similarities and differences between humans and dogs”, and have the opportunity to design in the future Better research.
She and another researcher began to consider explaining how dogs respond to certain illusions from a perceptual perspective. She said that for those cases where dogs react differently from people (that is, dogs choose visual stimuli that are opposite to humans, or show no preference) It is possible that the dog’s visual system is giving feedback to different stimuli. For example, it is well known that humans are particularly good at seeing overall patterns in images containing sub-elements. In contrast, Bissoyer noticed that canine perception may be better at perceiving local visual stimuli in images-this phenomenon may explain why dogs respond differently to Ebbinghaus and Ponzo illusions than humans, because of these two illusions. All images need to be viewed together to produce the expected illusion effect.
When looking at such a Navon figure (Navon figures), you can look at the overall image (here E and R), and you can also look at the partial details ( Here is the arrangement of N and S). Some studies have shown that humans are better at perceiving overall stimuli than dogs. THE SCIENTIST
Bissoyer added that this difference between species can reflect the different evolutionary pressures faced by dogs and humans. Burns said that there is indeed some evidence in the scientific literature that dogs have less visual preference for overall stimuli than humans—although there are still few studies on the subject, he cautioned.
Chouinard found another way to understand the difference in perception between humans and dogs: that is, how likely an animal is to see similar stimuli as exactly the same without being able to perceive the subtle differences between them. Chouinard added that the results of his and Bissoyer’s research can show that dogs are less likely to perceive differences in stimuli than humans.
Those experiments in which people and dogs reacted the same way also require rigorous explanations. Elias Garcia-Pelegrin (Elias Garcia-Pelegrin) is a PhD student in psychology at the University of Cambridge, studying the recognition of crows Knowing ability. He explained that if an animal is affected by a certain optical illusion just like a human, the reason may be the common neural mechanism of the two—but the same may not be the case. “Just because they reacted similarly does not mean that their cognitive processes are the same.”
three, let’s do some magic
Behavioural experiments to study the visual illusion of dogs are not the only way to explore the perception of dogs. For example, the research done by Burns is a functional MRI (fMRI) research, which requires dogs to passively look at the images on the screen. And perform brain imaging on it. In this way, researchers can observe neural signals associated with different stimuli. The preliminary draft recently released by Burns’s group shows that according to fMRI experiments, dogs cannot easily extrapolate from three-dimensional objects to two-dimensional pictures of these objects, and vice versa. This discovery can beIt can affect how scientists design vision-based tasks to measure the perception of dogs.
Gregory Burns, a neuroscientist at Emory University, uses fMRI to study dog perception. Emory University
Other researchers are exploring how to show different types of illusions to animals. For example, Garcia-Pellegran recently co-authored a forward-looking article. The article suggested that magic performances may be a powerful tool for studying animal perception abilities—even simple magic tricks, such as making a thing seem to disappear. “Magicians will use carefully prepared tricks to take advantage of the blind spots of our attention and perception,” he said, “If the attention and perception mechanisms of other animals are the same as humans(analogous) or homologous (homologous), then magic should also have an effect on them.”
Garcia-Pellegran also admits that magic and other dynamic optical illusions are obviously more complicated than static two-dimensional images. He is trying to conjure the crow he is studying, and make the object “disappear”. When observing a magic show, what is called is not only visual perception, but also other cognitive factors, such as “violation of expectations” when unexpected things happen. >(violation of expectation), and “object permanence” (object permanence). Bisoyer mentioned that it is very difficult to design control conditions for magic experiments. However, she added that her team hired aA magician came to the laboratory to explore this type of research.
Currently, Bissoyer and her team are experimenting with another trick to deceive dogs. It was a magic trick that was very popular on the Internet a few years ago, in which the dog owner wanted to lift a large blanket in front of him, like a screen, and raise and lower the blanket several times under the dog’s gaze. Then, after raising the blanket once, the dog owner hid behind the wall and put down the blanket again, making the dog feel that its owner has disappeared out of thin air.
Recently, Bissoy’s team launched a civilian science project called “Furry Challenge(What the Fluff!?)” Study how animals react to this illusion. “We let our owners do this magic on their dogs,” she said. “We will analyze their videos to see if we can draw conclusions about the permanence of the objects in this magic and the violation of expectations.” Bisoyeh also Continue to study trained animals and try to answer basic questions about dog cognition-it is these questions that originally sparked her interest in Australia, she said. “I think the best inspiration is sometimes born at a barbecue party.”
This article is from the WeChat official account:neural reality (ID: neureality)< span class = "text-remarks">, author: Catherine Offord, translation: Mollie, proofreading: Du Yu, editor: Orange Soda