Do soothing vocal cues enhance horses' ability to learn a frightening task?

When working with horses, it is frequently asserted that horses have an inherent understanding of harsh voice cues that would be used as reprimands versus soothing voice cues that may be used as positive reinforcers or calming modifiers. If horses are unable to understand this difference while their handlers assume they can, it may potentially lead to unfair or inappropriate training. A total of 107 horses from 2 different horse facilities in the United States and 7 different horse facilities in Europe were randomly assigned to either soothing voice treatment (SV; n = 58) or harsh voice treatment (HV; n = 49). The learning task involved horses of various breeds and ages learning to cross a tarpaulin. Methodology was standardized across locations. SV involved handlers saying good horse in a soft soothing manner whenever horses made forward progress toward the tarpaulin. HV involved saying quit it in a loud harsh manner whenever horses made forward progress toward the tarpaulin. Praat software was used to assess similarities in vocal spectrograms and acoustic parameters of different handlers and treatments. Mean pitch for SV and HV was 236.2 +/- 2.2 Hz and 322.1 +/- 8.9 Hz, respectively, both well within the equine hearing range and different at P < 0.001. Average intensity (loudness) for SV and HV was 51.2 +/- 1.7 dB and 61.7 +/- 1.2, respectively, different at P < 0.001. Contrary to our hypotheses, risk of failing the task (>10 minutes to cross the tarpaulin for the first time) was not different between treatments (22.4% failures on SV; 24.5% failures on HV; P = 0.41). Also, for those horses who did cross the tarpaulin, the total time to achieve the calmness criterion (crossing with little or no obvious anxiety) did not differ between treatments (139.9 +/- 50.4 for HV vs. 241.6 +/- 40.3 for SV; P = 0.25). There was no difference between the average heart rate (HR; n = 70 horses) of horses that crossed (82.9 +/- 7.0 beats/minute) versus those that failed (77.4 +/- 6.7; P = 0.43). Also, there was no difference between the average HR of HV horses (85.7 +/- 3.9 beats/minute) versus SV horses (77.9 +/- 3.7 beats/minute; P = 0.16). Furthermore, there was no difference between the maximum HRs, with HV horses registering a mean of 143.4 +/- 11.25 beats/minute and SV horses registering a mean of 166.1 +/- 9.5 beats/minute; P = 0.20. In the context of this study, soothing vocal cues did not enhance horses' ability to perform a novel potentially frightening task. (C) 2015 Elsevier Inc. All rights reserved.