Cueing vs. Reinforcement

         Many working dog handlers do not understand the difference between queuing and reinforcing a dog. Reinforcing a dog is when the dog is rewarded for finishing the task that he is either trained to do or is in training to do. Cueing a dog is the equivalent of giving the dog a hint about where to go or what to do. There is a huge difference. Cueing a dog decreases the dog’s reliability, whereas rewarding the dog increases the dog’s reliability. The reward should come when the complete problem is finished, not in stages throughout the problem. Some people will argue this point, insisting that the dog has to be rewarded at each stage. This is true if the dog is in training and not a qualified search dog. To properly train the dog, it must be done in stages. Note that I have simplified the stages of air searching for the sake of this article, and it can play out differently depending on the dog and the discipline. For example, the first stage of training is the initial runaway. When the dog runs and finds the person he saw run away, he is rewarded when he goes in to the person. That is the end of that stage. The next stage is when the dog comes back to give his alert (which may have been taught prior to this stage). The dog runs, finds the hidden person, and returns, giving his “I found him” signal. That is the end of that stage, and the dog is rewarded. The next stage is when the dog completes the scenario, finds the person, returns and gives the signal, and then leads the handler to the person. That is the end of the scenario, and he is rewarded. Sometimes in training, the dog will come back, give his “I found him” signal, and immediately lead the handler to the hidden person. In that case, he is rewarded when the handler is right next to the hidden person. 

         Often the confusion on the part of the handler comes when a dog is in training. A handler wants his dog to succeed and will lead the dog from behind, so to speak, by giving the dog cues.

An example of this is when the dog finds the hidden person, comes back, and gives his alert, and because the handler knows where the person is hidden, the handler then points in the direction where the person is hidden as he follows the dog. Because dogs learn by observation as much as by example, the dog will learn that the person he has found is in the direction where the handler pointed and assume that that is the direction that the handler wants him to go. 

The handler will argue the dog indicated that the person is in that direction, and they are only reinforcing what the dog indicated. However, take a step back and think of the dog’s initial training. When teaching the dog to find, in the very beginning, most handlers will get the dog excited by watching the volunteer run away, teasing the dog, then they release the dog and point in the direction that the volunteer ran, following the dog telling the dog to “find ‘em.” Keep in mind that a dog has a much wider peripheral vision than people and can often see the handler pointing even though they appear to be looking straight ahead. Also consider that many dogs will quickly glance back at the handler to see if they are following. Thus, the dog learns at the beginning of training that the hidden person is in the direction that the handler points. Now let’s consider a real search, the handler does not know where the person is hidden and does not know exactly where the dog came from when he returned to the handler with his alert. The dog may have traveled through the woods or field in a winding path due to terrain features back to the handler. In this case the missing person is not in the direction that the dog used to approach the handler. Now, based on his early training, which he has not forgotten, by pointing the handler is telling the dog to go in a different direction. It depends on the dog what he will do. Will he disregard the pointing cue or obey it.  Keep in mind that many dogs are taught to respond to silent signals, cueing falls into that category. 

A person can cue a dog in many different ways. Remember, a dog’s main means of communication is body language. A handler can cue a dog with a look, by turning his head in the direction where the person is hidden, a shift in the body, or any other subtle movement or even sound that is repetitious. Lest you doubt that dogs are aware of such minor changes in body language, I want to refer to Clever Hans. 

Clever Hans was a horse who could perform arithmetic and other intellectual tasks. Upon investigation by psychologist Oska Pfungst, it was found that Hans was reacting to the involuntary cues of his owner or the person who asked the question; both knew the answer to the question. Hans was definitely intelligent because he was able to interpret the involuntary cues in the person’s body language. This discovery became known as the Hans effect and is found in other animals. 

There are a few ways to be sure that your dog is not responding to hidden cues. First, if your dog comes back and gives you the “I found him” signal, do not follow your dog immediately. Act as if you did not see the signal and start to walk in a different direction. A well-trained dog will insist that you follow him. This is a good test of your dog’s training since, in reality, you may miss the dog’s signal on a real search, and the dog must not let you go in the wrong direction. Another way to be sure that you are not giving cues to your dog is to have someone watch you work with your dog when you know where the person is hidden and when you do not. This will have to be done enough times so that the observer can study you and the dog to see if you have repetitious behavior(s) that the dog reacts to. Once you are aware of any cues you give, you can practice avoiding them. 

Handlers subconsciously cue their dog in all kinds of detection work, including but not limited to: drugs, cadavers, bombs, and tracking/trailing and airscenting. In some cases, when a scent dog is following a track, the handler will redirect the dog to where he thinks the track is. As a result of past cueing, the dog may come off of the track and go where the handler directs. Then the handler will report that his dog lost the scent. If the dog works on a lead, the cue can be a subtle shift in the position of the lead by the handler. 

In the case of drug, bomb or cadaver dogs, if the handler knows where the source is hidden, the handler’s body might tense when they reach the hidden source. The dog can pick up on this. If the handler is tense as the dog searches, he may let out a deep breath when they reach the hidden source. This could also be a cue, as well if the handler hesitates, even for a second. This is a common cue since the handler will pause slightly to give the dog a chance to hit on the source or to see if the dog will give the signal. 

The bottom line is that a handler must test his dog in all situations, which includes making the dog think that you do not understand his signal so that he has to give it again. Because much of the initial training for all types of scent work involves scenarios where the handler knows where the scent is, a handler can become confident, often out of habit, of thinking that they know where the scent is. A handler should never try to tell a dog where to put his nose. If we knew where scent is located, we would not need the dogs. Even in a situation where scent was planted for the dog to find, it will not be isolated in the area where it was planted. A good lesson is to use a small smoke bomb after the exercise is completed to see where scent travels. 

A handler should only decide where to start their dog when considering wind, weather, and terrain features, and it is necessary to start the dog to his best advantage to find the scent. This is not the same as a handler who believes he knows where the scent is instead of or better than the dog. 

Because dogs are so intelligent, it can make it more difficult to train them. Often people do not realize or understand this concept. Even in modern times, the Hans effect has been observed in drug dogs and other working dogs. When a dog is extremely eager to please his hander, he will look for every signal that can bring the coveted reward. The dog and handler are a team; each must do his part to make the search successful, and each has a role to play. Both the dog and handler want to succeed. Often a handler will cue his dog because he does not trust his dog completely. Because dogs are so in tune with their handlers, if a handler does not trust his dog, the dog may pick up on that, and it will show in training. The only way a handler can trust his dog completely is to train in all circumstances enough times to see that the dog is reliable. 

Recent research has shown that dogs neurologically process facial expressions and body language the same way people do. (See Journal Reference 2). This further supports the Hans effect and explains why dogs are so in tune with our body language, therefore, very aware of clues.  

The last point that is critical is a new discovery about how dogs link sight and scent, quoted from a study conducted at Cornell University and published in The Journal of Neuroscience: “Cornell University researchers have provided the first documentation that dogs’ sense of smell is integrated with their vision and other unique parts of the brain, shedding new light on how dogs experience and navigate the world.” Please read the full article (Journal Reference 1) about this since we have yet to determine how it influences the way we train scent-detection dogs. But it does tie in with the fact that cueing a dog which is often visual, could be linked with scent or looking for scent. Keep in mind that cueing a dog can also involve sound. For example, if a person holds their breath as they approach the hidden source and then lets out a deep breath, the dog can hear that. The bottom line is that cues are given by movement, sight, and sound and, quite possibly, by scent. As our emotions change, our body gives off different scents. Being aware of the difference between reinforcement and how we cue our dogs will help make them the best working dogs possible. 

Journal Reference:

1. Erica F. Andrews, Raluca Pascalau, Alexandra Horowitz, Gillian M. Lawrence, Philippa J. Johnson. Extensive Connections of the Canine Olfactory Pathway Revealed by Tractography and Dissection. The Journal of Neuroscience, 2022; JN-RM-2355-21 DOI: 10.1523/JNEUROSCI.2355-21.2022
   
2. Magdalena Boch, Isabella C. Wagner, Sabrina Karl, Ludwig Huber, Claus Lamm. Functionally analogous body- and animacy-responsive areas are present in the dog (Canis familiaris) and human occipito-temporal lobe. Communications Biology, 2023; 6 (1) DOI: 10.1038/s42003-023-05014-7
   

---

Citations:

University of Vienna. “Dogs and humans process body postures similarly in their brains.” ScienceDaily. ScienceDaily, 27 June 2023. <www.sciencedaily.com/releases/2023/06/230627225203.htm>.

Cornell University. “New links found between dogs’ smell and vision.” ScienceDaily. ScienceDaily, 18 July 2022. <www.sciencedaily.com/releases/2022/07/220718181030.htm>.