Who Gives a Dog a Command and How It Is Given Matters
Dogs respond differently to hand signals and voice commands from a stranger.
Posted July 5, 2017
It was about six weeks into a beginners level dog obedience class when one of the students showed up without her dog. Her dog had had some kind of medical procedure which prevented it from being in class, but the dog's teen-aged owner didn't want to fall behind. So the girl showed up to observe the new exercises that were being taught so that she could return home and continue training her dog. One of the major functions of any beginning level dog obedience class is to teach the owners how to train and handle their dogs, so one of the senior instructors, Barbara, volunteered to let the girl "train" one of her dogs. The dog that she gave her was a handsome merle (mottled black white and gray) Shetland sheepdog named Charlie. He held several advanced dog obedience titles, and was a gentle, patient, and biddable dog who obviously knew all of the material that was being covered in this basic class.
The girl moved out onto the floor with Charlie who stood next to her. She gave the command "Sit!" but the well-trained dog just continued standing. When the girl saw that he wasn't responding she gave the hand signal to sit and Charlie immediately sat. A few minutes later the class was told to down their dogs and the girl gave Charlie the command "Down!" to no avail. With the dog still sitting there she now gave the hand signal for the dog to lie down and he immediately responded. I watched for a while longer, and it soon became clear that Charlie was responding considerably less reliably to the girl's voice commands than to her hand signals. I knew that Charlie did not have any hearing difficulties, and so I was somewhat puzzled by the fact that he seemed so indifferent to the girl's verbal instructions. I knew that his owner, Barbara, often competed with him in the obedience ring using only verbal instructions and he performed quite well.
The answer to what was going on in this situation has just appeared in a recent report published in the journal Applied Animal Behaviour Science. The research was done by a team of Italian investigators headed by Anna Scandurra from the Department of Biology at the University of Naples. This is the same laboratory which, in a previous investigation, showed that, in general, dogs tend to respond more reliably to hand signals and gestures than to verbal commands (click here for more about that). In this recent study they decided to see whether there was a difference in the effectiveness of hand signals and gestures depending upon whether these commands were given by the dog's owners or by an unfamiliar person.
The researchers wanted to use well-trained dogs so they recruited 22 dogs (six Golden retrievers and 16 Labrador retrievers) and their owners from the Italian School of Water Rescue Dogs. The dogs and their handlers were already qualified as water rescue teams. All of these dogs have been trained to obey a variety of commands such as: "sit", "lie down", "stay", "come", "fetch", and "turn". As part of their certification the dogs are trained to respond to both voice and hand signals. Their basic training is done on dry land and they have to show that they respond to commands reliably even from a distance or from behind a wire mesh barrier before they are accepted for the further training needed for water rescue duties.
In this experiment the dogs would be asked to respond to commands using the sane words and gestures that were normally used during their training. In the first test each dog was given a set of commands delivered by their owner silently, using only hand signals. The second test was the same except that the commands were delivered to the dogs using voice only, while the owners stood still with their hands hanging loosely at their sides.
The next stage of testing involved a clever manipulation. In this test the hand signals and the voice commands were simultaneously given but they were not the same in their meaning. The idea was to pit them against each other to see which command the dog responded to. Thus the verbal command for "lie down" could be pitted against a hand signal instructing the dog to "sit". Similarly the verbal command to "sit" could be set against the hand signal for the dog to "lie down".
In young children, aged 2 to 4 years of age, when you use both a gesture and a verbal signal it is the gesture which seems to be more important. This has also been shown by having conflicting verbal commands and signals set against one another. For example suppose that there are three items on a table, perhaps a cup, a doll, and a ball. A person now points at the cup and says "Get the ball". In this instance the child is most likely to respond to the pointing gesture and retrieve the cup rather than responding to the spoken instruction. Since a lot of data has indicated that the mental abilities of a dog is fairly similar to that of a human 2 to 3-year-old (click here for an example of that) it is not surprising to find that the dogs responded a similar way paying more attention to the hand signals than to the verbal commands. This finding confirms the earlier study from this same laboratory.
However, the novel part of this recent research is that after the dogs were tested having the owners giving the hand signals and verbal commands the dogs were retested—this time with the commands being given by a total stranger. Here the results were rather interesting. When it came to responding to verbal commands the dogs were around 25% less likely to respond to the stranger than they were to their owner. However the dogs responded equally well to the hand signals and gestures from both their owner or the stranger. When it came to the incompatible commands, the dogs responded around 2 1/2 times more frequently to the hand signal as compared to the verbal command when these were given by its owner. For the stranger, when the commands were incongruent the hand signal was preferred overwhelmingly, and the dogs responded to the around nine times more frequently to the hand movement than they responded to the verbal instruction.
What is going on here? Part of the problem is that human voices differ from one another in a lot of ways. Each voice has a specific pitch, a different set of harmonics, different timings when producing words or parts of words, and these may be modified by accents, emphasis and even emotional tone. So command words spoken by one individual may sound quite different to a dog than command words spoken by another person. In other words the commands spoken by a stranger may not sound enough like those given by the dog's owner to be easily recognized for what they were. Contrast this to the fact that hand gestures are quite simple, and one person can easily reproduce basic hand gestures used by another individual. Suppose that a dog is taught that the command to "come" is a simple quick motion where the trainer touches his left shoulder with his right hand. That gesture will look quite similar whether it is given by the dog's owner or by a stranger. This means that there are no dialects or accents in hand signals which might puzzle or confuse a dog.
Such data suggests that if you are dealing with an unfamiliar dog that you know has some training, you would be better off using hand signals (or perhaps a combination of voice and hand signals) if you would like the dog to reliably comply with your commands. By the time the hour-long class session had completed, the young teenager handling Charlie had figured this out for herself and was now giving both voice and hand signals simultaneously and he was responding perfectly.
Stanley Coren is the author of many books including: Gods, Ghosts and Black Dogs; The Wisdom of Dogs; Do Dogs Dream? Born to Bark; The Modern Dog; Why Do Dogs Have Wet Noses? The Pawprints of History; How Dogs Think; How To Speak Dog; Why We Love the Dogs We Do; What Do Dogs Know? The Intelligence of Dogs; Why Does My Dog Act That Way? Understanding Dogs for Dummies; Sleep Thieves; The Left-hander Syndrome
Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission
Anna Scandurra, Alessandra Alterisio, Lieta Marinelli, Paolo Mongillo,Gün Refik Semin, Biagio D’Aniello (2017). Effectiveness of verbal and gestural signals and familiarity with signal-senders on the performance of working dogs. Applied Animal Behaviour Science, 191, 78-83.