Only a handful of species appear capable of recognising themselves in a mirror. So, should we count horses among them?
Researchers from the University of Pisa in Italy explored the issue in a pilot study involving four horses, Gina, Calippo, Julia and Betsie.
While the horses performed some interesting behaviors at the mirrors, the evidence fell short of satisfactory proof that they could recognize themselves.
Researchers Paolo Baragli , Elisa Demuru, Chiara Scopa and Elisabetta Palagi, writing in the open-access peer-reviewed journal PLOS ONE, said only humans and a few other species, such as great apes, dolphins, elephants and magpies, have been found capable of recognizing themselves in a mirror.
To do so required complex cognitive, social and emotional skills, they said.
The study team noted that horse domestication went back 6000 years. Today, they were at the heart of a global industry and a special bond linked horses and humans.
Some recent findings even suggested that physiological variables, such as heart-rate variability, of humans and horses can show a sort of coupling process affected by the nature and time of the contact between horse and rider.
Recent findings indicated that horses were able to solve three-choice tasks in a flexible way, to use their long-term memory for concepts and categories, to integrate different sensory systems to individually recognize other horses and humans, and to combine different facial cues from other horses to gather information on the environment.
Horses have also been shown to communicate their emotions, understand facial expressions of both horses and humans, and have demonstrated the ability to reconcile after conflicts.
“Taken together, these findings are indicative that horses, like other highly cognitive social animals, show some degree of awareness, which implies the ability to assess and deduce the significance of a situation according to both the social environment and the self,” the study team said.
They said mirror self-recognition was not important as such, but its significance lay in what it may unveil about the sense of self experienced by animals in relation to their social environment. In short, it provided information about the cognitive and emotional skills necessary to develop complex social relationships and to engage in behaviors relying on different levels of empathy.
The study team described their experiment in which the four horses, all living at pasture in a balanced environment, were placed in front of a mirror under a series of scenarios to allow their behaviors to be videoed and analyzed.
The study involved what is known as a mirror mark test, in which a mark is placed on the animal’s body in a place it cannot see to assess its behavior toward it when it sees it in the mirror. If an animal considers the image as its own, it will use its reflection in the mirror to detect the mark and will try to inspect, touch, explore or scrape it.
Baragli and his colleagues devised a protocol that included two tactile touch controls.
For the experiment, at the Italian Horse Protection rescue center in Montaione, the horses were exposed to the mirror over six days under the different scenarios within a familiar L-shaped covered enclosure.
In the key scenarios, the mirror, which measured 150cm by 220cm, was first covered and then it was uncovered. This enabled the researchers to see which behaviors related directly to the presence of the mirror reflection or the presence of the apparatus in the enclosure.
In another scenario, each horse was released into the mirror enclosure with a cross formed on each cheek using clear ultrasound gel. This was the “sham” scenario.
In another crucial scenario, gel crosses were again placed on each jowl, but one was colored using eyeshadow powder, either yellow or blue. The following day, this scenario was repeated but with the colored cross on the other side.
Video analysis of behavior was performed independently by two observers.
The study team found significant differences in behavior between when the mirror was covered and the mirror was uncovered.
“There was no remarkable behavior in front of the covered mirror in any of the tested horses,” they reported.
“On the contrary, under the open-mirror condition the behavior of the horses strongly differed. All of the tested horses spent a significantly higher amount of time in front of the open compared to the covered mirror.
“Three horses out of four engaged in a significantly longer exploring activity under the open mirror than the closed mirror. Furthermore, all horses looked behind the mirror only when it was uncovered.”
When the mirror was exposed, three out of four horses also performed mouth movements when in front of the reflecting surface, with one horse reaching statistical significance.
Three of the horses – Betsie, Gina and Julia – performed scraping behavior towards both cheeks with a significantly different frequency across the four conditions − when the mirror was covered, when it was exposed, when the sham crosses were being used, and when the colored mark was on the right check.
They showed a significant increase in the scraping activity towards both cheeks when the right cheek bore a colored cross compared to the sham scenario.
One out of four horses performed scraping behavior at higher frequency when the mark was on the left check compared to the sham scenario.
Discussing their findings, the study team said the horses showed some behavioral patterns in response to the presence of the reflecting surface which were not performed in the other conditions.
While some behaviors were ambiguous, at least some of them indicated that horses could understand that the image in the mirror was not a real animal.
To successfully pass the mark test, the authors said a sequence of behavioral steps towards the mirror had to have been performed by the horses.
Firstly, there had to be a social response (often aggressive) toward the image. Only one of the tested horses had clearly shown aggressive behaviors toward the image in the mirror.
The second step was related to physical inspection of the mirror, including looking behind it.
“Compared to the covered mirror, in the open mirror condition the four horses spent much more time in the mirror area and three of them explored the mirror for a longer time.
“Looking behind the mirror has been reported in several primate and non-primate species and it has been interpreted as the attempt to check for the actual absence/presence of [another horse] behind the reflecting surface.
“All of our tested horses looked behind the mirror several times, while none of them did it in the covered mirror condition.
During the third step, animals should perform repetitive mirror-testing behaviors – that is, inspecting body parts only visible through the mirror.
“Our horses showed ambiguous behaviors and three of them (Calippo, Gina and Julia) opened their mouths and protruded their tongues more frequently in the open mirror than in the covered mirror condition, but only one (Calippo) reached statistical significance.”
Then, if they passed the first three steps, animals needed to use the mirror to touch an otherwise imperceptible mark on their own body.
The horses, they said, could not be considered to have passed the mark test, as they did not show the requested sequence of behaviors in the first three steps that were considered mandatory by some scholars.
“However, this procedure is still highly debated and it has not been applied in some studies,” they noted.
That said, the horses did show a specific sequence of behaviors with the open mirror compared to the covered mirror, indicating motivation to explore the image in the mirror.
Discussing the scraping behavior, the researchers said the presence of both sham and colored marks actually affected the frequency of scraping towards both cheeks, indicating that horses were able to perceive the mark.
In one horse, Betsie, the amount of scraping was always significantly higher in the presence of the colored mark compared with the sham mark, while in other two horses, Julia and Gina, such a difference was present only when the colored mark was on the right side.
In conclusion, they said the horses did not match the complete expected behavioral steps to fit the mirror self-recognition paradigm.
The self-directed behavior towards the colored mark showed by three of the horses was not sufficient per se to affirm that horses were capable of mirror self-recognition, they said.
Baragli P, Demuru E, Scopa C, Palagi E (2017) Are horses capable of mirror self-recognition? A pilot study. PLoS ONE 12(5): e0176717. https://doi.org/10.1371/journal.pone.0176717