Do Horses Recognize Humans?
Q: Does my horse recognize me from other people?

A: We all get a sense that our horses recognize us by our appearance or the sound of our voice, and that they can distinguish us from strangers or less familiar people. Certainly we know horses learn associations between a person coming around an expected time and their getting fed, turned out, or exercised. It’s difficult from this simple scenario to discriminate between a horse learning by reinforcement and a horse actually recognizing a specific person providing the reinforcement.

One very old study showed that horses depended on facial features as well as clothing to recognize individuals. A number of more recent studies have shown that horses seem to be able to tell when the audio recording of the voice and the sight of familiar handlers match up; that is, compared to when the voice recording is from a different person than the one a horse is shown. This is known as “cross-modal recognition,” because the horses were asked to combine multiple sensory cues. By the way, this also seems to be true when researchers asked the question of whether horses recognize and discriminate between familiar horse herdmates. Horses seem to detect when the recorded horse vocalizations played for them and the visual appearance of the horse actually presented match up.

It’s also notable that in these studies, the humans or horses used as cues in the tests were kind of paraded past the subject horses and taken behind a barrier out of sight. Then researchers played the audio voice or vocalizations. So the horses were not seeing and hearing the cues at the same time. Yet they could still seem to discriminate when the audio and visual cues matched up. In humans, it’s understood that individual recognition of other people depends on the immediate audio and visual cues but also draws on past experience or knowledge of that individual. So these studies suggest that horses, too, are relying on both immediate cues and some representation in their memory.

A number of studies show that, in some manner, horses respond to what a person is paying attention to. You know, like when I point at something and you are inclined to look where I’m pointing. One study showed that horses obeyed known voice commands similarly from their longtime handler and from a new unfamiliar handler as long as both directed their attention toward the horses. Yet these horses always obeyed commands from that longtime handler even when that handler was not physically directing his attention to the horses, like when he was facing away from the horse or his eyes were closed. So here, familiarity with the handler seemed to be the best way to assure a horse obeyed a command.

These studies I’ve briefly summarized all had something interesting in common. The researchers counted the kinds of behaviors that showed the horses were reacting to the audio and visual cues provided—for example the amount of time the horse spent looking in a certain direction. And in all the studies, the horses spent more time reacting to the cues that were most incongruent, such as the mismatched audio and visual presentations of people or the known command given by an unfamiliar person. Researchers explained this increase was due to “violation of expectations,” so they speculated that the horses were trying to figure out why the different cues presented to them were not adding up to what they expected.

I think both our experiences and the science tell us that horses probably best recognize individual people by using all sorts of cues together: voice, physical characteristics, typical postures and movements, and the sequence of things happening the way they’ve learned to expect them to happen based on simple conditioning and reinforcement.