Most riders have growled at their misbehaving horses, but do the animals get a sense that their owners are displeased?
A new British study has delivered interesting findings on the reactions of horses to human voices at opposite ends of the emotional spectrum – laughter and growling.
Researchers involved in the University of Sussex study used 28 horses from two riding schools in East Sussex, exposing each of them to eight recorded human non-verbal emotional vocalisations.
They comprised two males and two females laughing, and two males and two females growling.
Each horse received either male or female stimuli, but not both, while standing in an outdoor arena. The sound came from a speaker hidden in vegetation 15 metres away.
Amy Victoria Smith and her colleagues, writing in the journal Scientific Reports, found that horses were found to adopt a freeze posture for significantly longer immediately after hearing the growling when compared to the laughter.
This, they said, suggested that negative voices promoted vigilance behaviours and may be perceived as more threatening.
In support of this interpretation, horses held their ears forwards for longer and performed fewer ear movements in response to negative voices, which further suggested increased vigilance.
In addition, horses showed a right-ear/left-hemisphere bias with the laughter when compared to the growling, suggesting that horses perceived laughter as more positive than growling.
“These findings,” they said, “raise interesting questions about the potential for universal discrimination of vocal affect and the role of lifetime learning versus other factors in interspecific communication.”
The researchers said their findings suggest that horses appear to discriminate behaviourally between different types of human vocalisation in ways that suggest they perceive negative human voices as more threatening than positive voices.
“Interestingly, horses did not discriminate behaviourally between male and female voices in the paradigm we used.
“We had predicted that horses would respond more negatively to male voices, and specifically negative male voices, due to their having relatively lower fundamental and formant frequencies than female voices.
“However, there is similar evidence that dogs do not discriminate the sex of a human signaller when hearing emotional vocalisations, despite having the ability to discriminate gender in human voices.
“It is therefore possible that emotional cues are more salient than gender cues in such paradigms and so are responded to preferentially.”
The freeze response seen with the growl forms part of the “fight, flight, or freeze” reaction to a perceived threat.
They said individual horses typically freeze in response to a distant and relatively mild threat, whilst closer and more extreme threats may provoke vocalisations, direct avoidance, and attack behaviours.
“The freeze response therefore appears to be an appropriate reaction in the present paradigm where the stimulus is mildly aversive and distant, i.e. comfortably outside the horse’s flight zone.
Discussing the evidence of a right auditory lateralisation towards the positive vocal signals, they said: “In many species, a right-ear bias indicates that signals are preferentially processed in the left brain hemisphere and are generally associated with the perception of familiar or positive stimuli.”
They noted that auditory laterality in horses had not been studied extensively, although research suggested that incoming signals were processed primarily in the contralateral brain hemisphere.
“Auditory laterality in horses has not been established in relation to emotional contexts, however horses show auditory laterality in social situations and demonstrate both gaze and limb preferences in emotional situations.
“Lateralised ear behaviour may therefore also be interesting in an emotional context.”
The ability of non-human species to discriminate between human vocalisations raised interesting questions about the potential universality of emotion discrimination through auditory signals, they said.
“Determining the particular acoustic parameters that give rise to this discrimination would be a useful avenue for future research.
“For example, negative emotional arousal is expressed through harsh, low-frequency tones across a wide range of species, and so these cues may be readily responded to even in the vocalisations of other species, without explicit prior experience of these species.”
Their results, they said, complemented the current body of research on the ability of dogs to discriminate human vocal emotions, extending this work to another key domesticated species.
While previous research by others had suggested that horses do not discriminate between harsh or soothing human voices, the results of the current research present a different picture.
They noted that previous research in this area had involved training, which they propose could have added a confounding variable that could mask potential differences.
“Spontaneous discrimination paradigms, such as that used in the present study, may be best placed to detect subtle differences in behavioural responses,” they said.
The full study team comprised Amy Victoria Smith, Leanne Proops, Kate Grounds, Jennifer Wathan and Karen McComb, all from the University of Sussex; and Sophie Scott, from University College London.
Domestic horses (Equus caballus) discriminate between negative and positive human nonverbal vocalisations
Amy Victoria Smith, Leanne Proops, Kate Grounds, Jennifer Wathan, Sophie K Scott and Karen McComb
Scientific Reports, volume 8, Article number: 13052 (2018) https://doi.org/10.1038/s41598-018-30777-z