Cues and Signals in Clicker Training
Cues in clicker training and any dog training are the signals that you will use to inform your dog of what exercise you are doing and what behaviors you expect. The ultimate goal of any dog training is to have a repertoire of reliably performed behaviors on command and to use those in various different situations (for more information about different situations and scenarios, please see the Criteria page ).
Although there are many theories about how to structure the communication system that leads to a reliable response from our dogs, there are always signals of some kind that are involved in this communication. We will discuss these signals in a little more depth, here below.
What are cues (commands)?
We will discuss the commonly named “commands” in dog societies, but first, what actually are cues and commands?
Cues and commands (either verbal or physical) are a type of stimulus. A stimulus can be an action, a signal or anything that precedes a certain dog behavior.
For example: Your dog sees a squirrel and gets frantic on a lead. In this case we have the stimulus (the squirrel) that is followed by the dog’s reaction to it (the behavior).
There are many different stimuli that are provided by the environment and we may have little to no control over them, those stimuli are generally known as triggers.
The next group of stimuli are classically conditioned (learned) stimuli that we use in our daily communication attempts with our dogs. These are the verbal or physical cues (commands). These are usually words or gestures that once were meaningless to our dogs but that now have a meaning to them. Once we have a dog that is reliably performing the same actions by responding to cues (commands) that is known as “behaviors under stimulus control” and that is the goal of any animal training.
There are other groups, but we generally stick to these two in order to simplify things as much as possible. For our purposes here, we will mainly focus on the second group of stimuli, the cues.
What types of cues are used in dog training?
In general we use two different types in dog training:
- Verbal cues
- Physical cues
The one that you will use depends on your personal preferences or the sport that you are involved in with your dog. In real life scenarios, I prefer verbal over physical cues. The issue with physical cues is that a dog physically needs to face you (or be watching you) in order to perform the exercise or receive the message, which may become an issue when you are faced with some situations in which your dog needs to respond instantly.
On the other hand, training your dog to respond to physical cues (gestures) is easier than with verbal ones. There are many reasons for this, but regardless of which method you are planning to use, you have to be clear with your dog about the “language” that you will use, whether physical or verbal. And here is where many dog owners face the first issue.
Problems with over-cueing
Although this is part of the luring and free shaping page of the website, the reason that I’m mentioning it here also, is because even though we may intend to use a luring concept, we may end up with a creating a physical cue instead, if we “over-cue” our dog.
For example, if you are making an up-sweeping motion with your hand in order for your dog to sit, keep this as your ‘signal’ and isolate it from the rest of your body movements. Many people tend to also bend down, or nod their head when doing this hand signal, and if you do this for a few repetitions, then the whole sequence (including the bending or the head-bobbing) becomes the cue, not just the up-sweeping hand motion.
The issue with this is that eventually, you may stop bending down, or another member of your family may try exercising with the dog and perhaps that person doesn’t bend, now suddenly, the dog doesn’t respond to the cue.
Unfortunately, most dog owners mislabel this as “disobedience” when in fact the blame is on the handler’s side as they are changing the appearance of the cue. Dogs are very visually oriented, and they are always trying to read our gestures. If you are to use body gestures as a cue for certain exercises, make sure that you are consistent. Avoid any unnecessary body movements. The more unnecessary body gestures you use, the broader the “command” for that particular behavior is, and the more mistakes you can run into.
Think of it as a security password. If you use a three digit password, it is easy to remember the number sequence, if you use five, six or more digits, the chances for making mistakes drastically increases.
Where to start
This is the part that baffles almost every dog owner that is not familiar with the clicker training concept; the topic of when to actually name the behavior.
In old dog training circles, the training started straight based on command, which means that you would start the training sequence as follows; (we will use the “sit” command for this example):
Traditional way of dog training:
- “Sit” (command)
- Pull the leash upwards (negative reinforcement)
- The dog sits (behavior)
- Relieve the leash pressure (reward)
In a clicker training system, first you lure the dog into the desired behavior, or reward the appearance of the behavior itself (as described in the luring and free shaping part of the website). Once the behavior is fluent (after a certain number of successful repetitions) you then name it. It goes something like this:
Clicker training systems for training the “sit” exercise:
- Luring or free shaping technique
- Dog offers behavior
- Click/mark the behavior
- Reward the dog
Once the dog is fluent with the action:
- Give the verbal cue (“Sit”)
- Dog offers the position (behavior)
It is difficult to master this technique simply because we are so verbally oriented and it is sometimes difficult for us to understand that our spoken language doesn’t mean anything to our dogs.
Until we give a meaning to each of the commands that we use in training, our dogs are simply reacting to other cues or stimuli, they do not actually understand the command that we have spoken.
Many people might ask, “How is it possible that in other types of training, dogs understand the commands first?” and the simply answer is…it’s not.
In compulsion-based methods, dogs start by being physically manipulated into the positions, therefore there is no “trial and error” period and no “problem-solving” activity. They have not understood that “sit” means to put their hind end down on the ground; they were simply not given any other choice, as they were physically put into that position.
In clicker training concepts, the process of learning includes trial and error, as well as problem solving activities. If you were to use the command “sit” too early, you would end up with the same results that most dog owners end up with; a dog that sits every now and then and with a poor response to it. This is because, as mentioned above, until the dog knows what the action is that is wanted; it is useless for us to label that action.
When this happens, and our dog no longer responds reliably to a verbal command, we call this a “poisoned” cue.
A poisoned cue is a command that wasn’t trained properly and therefore dogs don’t perform it reliably. For example: if you called your dog to “come” and he doesn’t (simply because he doesn’t understand the meaning of the cue), you are actually “training” him that the word “come” doesn’t mean anything. It just becomes another “sound” he hears you say, not a “cue” that he is paying attention for. Remember, as much as “come” has a definition in English, this doesn’t mean anything to our dog and it is no different than “boo” or “blah”, it is just another sound.
Problems can start, for example, if you start training your dog a recall using the same command “come”. By this point, the dog has already learned that “come” means keep doing whatever you are doing, or he may even connect this with an invitation for a play session, as perhaps, in the past you have followed it up by chasing after him.
Now you are faced with a problem. By trying to train your dog with a cue that means something different to him, than what you intend it to, this will delay your training efforts, or even jeopardize them completely.
If you have used a certain command and there were no results or the results were inconsistent, we call this a “poisoned cue”. To avoid this, name the behaviors after they are already established. If you already used the word without results, previously, then choose another word (cue/command) and start fresh. In the end, whichever words or language we use is indifferent to our dogs because they don’t understand the English language. They won’t be debating the definition of a word; they will simply associate it to the meaning (action) that we have trained them for.
This is the way to name behaviors (assign cues) in a clicker training system, once you have the behavior on cue then you start adding criteria in order to get reliable results in various different situations.
The use of additional cues and signals in clicker training is an ongoing subject of discussion and a polemic among dog trainers and experts worldwide. We all agree on the benefits of using clicker training principles (regardless of if you use a clicker, verbal marker, whistle, etc.) for marking wanted behaviors, however where the discussions come in is when signals are used to mark the moment that the dog failed to perform or missed, some dog trainers tend to use signals to inform their dogs that they have failed, still others say that these such signals are not necessary or even that they produce unnecessary stress, etc.
Which additional cues and signals to use?
In broad terms there are two types of additional cues in clicker training that are used, a marker to indicate what is good and on track for a reward and a marker that is used to indicate that the given action or behavior is unwanted.
In these unwanted situations, markers like “wrong”, an “uh-uh” sound or sometimes a “no” command are often used. Now this is not to be confused with the cue used for correction (which I will describe below), it is merely a signal to the dog that the behavior he is performing will not bring a reward.
In the case of wanting to let your dog know that he is on the right track, some dog trainers use an additional cue such as “good” that serves as a type of a signal that the performed behavior is desirable and for the dog to keep performing it as he is on track for the reward.
The meaning of the “good” marker
The word “good” can be used in both clicker training and marker training. This word will be used as a guiding marker. Normally, we use it for endurance and duration, for example, when we want our dog to remain in a certain position or to reinforce that position.
Ultimately, our dog connects that the word “good” will eventually be followed by the release marker (the clicker or verbal marker), which means a reward. The “good” marker will become a signal to our dog that we are happy with his performance and to keep doing it, in order to be released/rewarded.
Sometimes you can even reward your dog on the guiding marker (“good”), for example: You are training your dog endurance while sitting, sit your dog, take a few steps backward praise him with good, wait a few seconds repeat the marker “good” and approach your dog to reward him while still in the sit position, and then take a few steps back again, repeat “good” again (assuming that your dog is maintaining the sit position) and then after a few seconds release your dog with the clicker or marker that you are using and reward him.
In the beginning your dog will probably break the position in order to approach you or to access the reward, you simply withhold the reward and repeat the action, dogs normally grasp the idea fast and from that point, you can phase out the frequency of the rewards while using “good” and your marker itself.
Signals for corrections (conditioned punisher)
In compulsion based training systems dog trainers use specific signals to predict corrections, and this is something totally different than withholding a reward in the clicker training system.
In much the same way that clicker training uses to train the dog that the clicker is always followed with a reward, compulsion based training uses a specific verbal marker that is always followed by an aversive action (leash pop, e-collar correction, etc.). Therefore the dog learns to connect that signal with the unpleasant action.
The reason that some dog trainers use signals in the form of conditioned punishers is to “inform” the dog of when they are doing something wrong so that the dog can correct his behavior himself in order to avoid the correction. One of the most used words for a conditioned punisher is “No”.
It is important to know this, as many of dog owners today adopt their dogs at various ages and may not be aware of the dog’s previous “training” history. If you then use the word “No” as a cue for withholding a reward, but your dog has a negative experience with that command, he will react negatively and will probably stop performing future actions as he may start avoiding any work in order to avoid possible corrections.
This may completely jeopardize your training process.
TIP: If you are working with a dog that you are not sure of their past or if you are a so-called “balanced dog trainer” (meaning a trainer who uses both, the clicker training principles as well as corrections), make sure that you are using a different verbal signal for withholding a reward than the signal that was used for corrections.
If you are dealing with a dog that you have just recently been introduced to, you can observe the dog’s body language when you use the “No” or “Wrong” command (over 90% of all dog owners use these words in conjunction with corrections); if you see any of the signs of stress, fear, or nervousness (lowering body posture, ears down, tail between the legs, turning head left or right avoiding eye contact, licking lips, body shake or yawning, etc.) it is best to use some other commands from that point as the dog already has an “emotional attachment” to the “No” or “Wrong”.
The proper way to inform your dog that he is wrong
In clicker training the whole concept is based on two simple principles: mark and reward the behaviors that you like, and withhold (don’t reward) the behaviors that you are trying to eliminate. Now as mentioned earlier, some dog trainers use additional cues for the latter and some don’t.
I personally use them as I think that it is important for our dogs to understand when they are performing something that is not wanted or part of the desirable repertoire. However, in saying that, there are rules about when and how to use this signal.
Body language speaks more than a thousand words
As much as we as a species are totally dependent on our verbal communication and language, and our body language is more of an unregulated display of body gestures that follows the contexts of our verbal stories; in a dog’s world, body language is the way of communication.
It is important to always keep this in mind, no matter how much our dog appears to be good at understanding our commands, his primary way of receiving input is by observing us. This is why so many people struggle to have dogs trained for verbal cues. In clicker training, the timing of marking the moment when the dog is “wrong” is as equally important as marking it when he is right.
Our dogs don’t know what the exercise should look like; they don’t know the concept of it or what the final goal is. They are observing us in order to predict their upcoming reward. It is because of this that we need to be aware of our body gestures.
Dogs taught with clicker training concepts are training through a “trial and error” system, and in every step that the dog takes towards the final goal, we mark and reward him. Normally, our body gestures are more or less always the same when we do that. We click (or use a verbal marker), reach for the reward (or have the treat already in our hand) and then move our hand in order to deliver the reward. It is that hand movement that our dog is so attached to.
Often we move differently at moments when our dog is performing something undesirable or while he is going through the trial and error phase. It is important to remember that if you are doing any body movements (especially similar ones all the time around your dog), your dog may be confused as you are sending him mixed signals. For example: if your dog always gets rewarded in a position facing you and just up above his head, then if you do that similar movement when your dog is doing wrong, then you are sending confusing signals to your dog, as he already associates that movement with receiving his reward.
In order to avoid this issue there are two options:
- Avoid using body language, in order to prevent your dog from getting attached to the wrong signals
- Use specific and different body language (signals) for when the dog is right and for when the dog is wrong
Using signals vs. not using signals in clicker training
This is a list of pros and cons for additional cues that tell our dog what we don’t want:
Some good things about not using “wrong” signals and cues include:
- Your dog doesn’t get attached to unnecessary cues
- Your dog has the freedom of experimenting in order to deduce the next step, etc.
Some of the issues that you may face without these signals are:
- Your dog can get frustrated if it takes too long to figure out the next step (criteria is too high, etc.)
- Most humans instinctively tend to produce some type of body (or verbal) gestures when their dog is working, so there is a chance that the dog will get attached to an unintentional cue anyway.
- Your dog may have issues understanding the unwanted behavior (or a part of that behavior) as there are no clear markers when the behavior happened.
The benefits of using a “wrong” marker are a great help in everyday dog training. The most important thing is for the dog to understand the purpose of it.
You can mark behaviors that are unwanted and thereby greatly speed up the training process; as well it is a clear message that the performed behavior is unwanted.
Maybe one of the most important benefits of additional cues is that you are capable of using the correct timing in order to inform your dog in the moment.
How to properly use the “wrong” command in clicker training
In some cases, it is very difficult to isolate the wrong behaviors down the road. Often, dogs will start slacking in their work, and you will sometimes see that they tend to “skip” some parts of an exercise in the behavior chain. For example:
You are training your dog sit and stay; you sit your dog then move to your position, but in the meantime your dog has broken his position, ran over to you and performed his sit again on his own, but in front of you, not where you left him. Now if you were to withhold a reward without using any verbal or physical signals to inform the dog that he did a wrong behavior, it is up to him to conclude which behavior sequence was unwanted; the breaking of the sit position, the action of running towards you, or the finishing position (sitting in front of you).
Keep in mind that it is always up to the animal (dog) to interpret what behavior brings or doesn’t bring the reward, so if you don’t have a good communication language, you will have difficulties pointing and isolating a specific behavior to your dog. Often after a few bad repetitions, the dog stops performing the whole exercise in general as a result of frustration.
In order to properly address this situation above (and other situations as well), you need to have a specific language developed and it is a simple one; Use “wrong” or whichever verbal cue you want to use, and a distinctive physical reaction that is unique (this will be a physical cue for not delivering a reward). The moment that your dog has broken the sit position, use the verbal marker “wrong” and do not continue with the exercise in any way (do not go to your place which may allow your dog to perform any other behavior like coming back to you or a finish “sit” position) at that moment the whole exercise is cancelled and you go back to the beginning.
TIP: Pay attention to your body position and gestures in these moments. If you are cancelling the exercise, and delivering your “wrong” marker, be sure not to maintain your “rewarding” posture and body language, as this would be contradictory information for your dog. He will be hearing “wrong” but seeing “right”. Break your position and use different body language so that the message is clear to your dog.
When to start using the “wrong” marker in clicker training
Like many things, there is a controversy about when and how to introduce the “wrong” marker to our dogs. From my personal experience, I wouldn’t recommend that it be used from the start of dog training. A dog needs to first go through his learning “trial and error” period to develop and understand how the whole concept works in general. If I was to add the “wrong” from the beginning, it would just serve as more confusion and perhaps it may limit my dog’s potential for exploring with trial and error.
Once my dog has a basic understanding of “sit” and “down”, and we are performing some easy free shaping exercises, I will start adding in the “wrong” cue for those exercises. Since the dog is already familiar with the behavior itself, or understands the concept of it, it would be easy for him to understand that every time he hears the “wrong” cue, the reward is gone. But since he knows the behavior already, he would know what to do instead, in order to access the reward. This is the best and the less stressful way for a dog to learn the “wrong” command.
The more that your dog becomes fluent, the more often you would be able to use this cue in order to help your dog recognize the unwanted behaviors.
This may not be necessary for some simple exercises, but as you advance further, additional cues can be a great tool in your toolbox.
Return from Cues to Clicker Training
Return from Cues to Training Your Dog and You