Horses are likely to be fooled by a common illusion that makes a line appear longer than it is, according to researchers.
Visual illusions are often used by researchers to investigate the perceptual mechanisms underlying vision among animals. The assumption is that, if a species perceives the illusion like humans do, they probably share the same perceptual mechanisms.
Researchers at the University of Padova in Italy set out to investigate whether horses are susceptible to the Müller-Lyer illusion. It is a common size illusion in which two same-sized lines appear to be different lengths because of the spatial arrangements of arrowheads presented at the two ends of the lines.
Christian Agrillo and his fellow researchers noted that susceptibility to the Müller-Lyer illusion has been reported in both Old and New World monkeys, suggesting that inappropriate size scaling may characterize the visual perception of all simian primates. To date, it is largely unknown how other mammals perceive this illusion.
The only non-primate mammal studied so far is the dog, which does not perceive size illusions the same way as primates.
Like most carnivores, dogs are primarily nocturnal or active in twilight, and their surrounding space is largely sensed through hearing and smell. Size illusions might therefore arise from the acquisition of accurate depth perception and the evolution of a sophisticated system of three-dimensional vision that occurred during primate evolution.
The study team set up an experiment involving 10 horses housed at a private riding stable to determine the effects of the Muller-Lyer illusion, using the animals’ preference for the larger portion of any carrot offered.
In a series of control trials, the horses were offered two carrot sticks of different sizes presented to them on a vertical blackboard, which was wheeled towards them on a trolley. Each of the carrot sticks had been cut to be 1cm square, but one was 20cm long and the other 13.5cm long.
The control trials also got the horses used to the presence of the arrowheads, made out of white plastic sticks, which were placed in various configurations during the control trials.
As soon as the horse chose one of the carrot sticks to eat, the display trolley was withdrawn so the horse could not eat the other portion, too.
Not surprisingly, the horses showed a strong preference for the longer carrot sticks.
In the test trials, the horses were presented with two identical 20cm long carrot sticks together with arrowheads arranged in a way meant to elicit the Müller-Lyer illusion in human observers.
When presented with the illusion, the horses showed a strong preference for the carrot stick that humans perceive as longer, even though both carrot offerings were the same size and length.
Further control trials excluded the possibility that their choices were based on the total size of the carrot stick and the arrowheads together.
Horses, they said, showed a human-like perception of this illusion, meaning that they may display similar perceptual mechanisms underlying the size estimation of objects.
The fact that horses selected the larger portion in the test trials also indicated that horses did not rely on smell to solve the task, because the smell coming from two same-sized sticks was presumably identical.
“The horse has a fairly well-developed visual system, but it also strongly relies on other sensory modalities to acquire information from the external world,” the study team noted.
“The visual systems of humans and horses differ significantly in many respects, including a greater development of binocular vision in humans and inferior visual capabilities in horses compared to humans in terms of acuity, accommodation, and daily colour vision.
“Our finding that horses are sensitive to the Müller-Lyer illusion is relevant to this issue as it indicates that, despite the wide range of neuroanatomical differences between the visual systems of humans and horses, there are interesting parallels in the perceptual mechanisms underlying representation of the visual scene.”
It is widely agreed that, in humans, the illusion operates on the basis of a misapplication of size constancy scaling.
There is good evidence that horses, like humans and non-human primates, are sensitive to depth information contained in static pictures, while dogs use mainly other clues to estimate distance and show little sensitivity if any to pictorial depth cues.
“At the present state of the research, it is not possible to understand if the response to size illusory patterns was inherited in primates and ungulates from a common ancestor or it represents a by-product of habitat-driven evolution.”
Clearly, they said, data from many more species need to be collected to unravel questions arising from whether species are susceptible to the illusion.
The study team comprised Anansi Cappellato, Maria Elena Miletto Petrazzini, Angelo Bisazza, Marco Dadda and Agrillo, all with the University of Padova.
Cappellato, A.; Miletto Petrazzini, M.E.; Bisazza, A.; Dadda, M.; Agrillo, C. Susceptibility to Size Visual Illusions in a Non-Primate Mammal (Equus caballus). Animals 2020, 10, 1673. https://doi.org/10.3390/ani10091673. The study, published under a Creative Commons License, can be read here.