Introduction
Dog training can be likened to a sport, where understanding both the rules and the skill of playing are key. Success in any sport comes from frequent and consistent practice. A common issue I observe in dog training is the lack of sufficient practice. Many individuals may read articles, watch videos, or even attend training sessions, but far too few continue to train their dogs consistently outside of these structured environments. Believing one can become proficient in dog training without regular practice is a misconception. While some people might naturally possess better timing, coordination, and athletic abilities, which could ease the training process, the necessity for extensive practice remains unchanged.
In my YouTube videos, I often discuss a step-by-step approach to training any dog. For a deep dive into this process, I recommend watching this particular video.
Overview of Training Guidelines
In this article, I aim to outline the rules of the game, or more accurately, 'the guidelines,' of dog training. Focusing on obedience training and behavioral issues, these guidelines can help navigate the most common challenges. We'll explore the three main events:
The reward event
The reinforcement event
The correction event
Understanding when these events start and end, along with the exceptions to the rules, can considerably simplify dog training. We'll begin by examining the start and conclusion of each event before delving into the exceptions to the rules, offering clear direction for addressing most training and obedience scenarios.
The Reward Event
The journey begins with the reward event, a notion inspired by Michael Ellis. His discussions during an online course at Leerburg Online University introduced the term 'reward event.' Ellis highlighted the term to encourage a deeper understanding among trainers, particularly addressing the common misconception regarding dogs' ability to associate actions with outcomes within a very short timeframe.
Scientific studies have supported the idea that we have about one second, give or take, to effectively link a dog's behavior with a consequence. This is precisely why we employ markers—to accurately capture the moment a dog performs correctly or incorrectly, thus leading into the event predicted by the marker. Despite this roughly one-second window for association, the event ensuing the marker—be it a reward or a different outcome—can extend well beyond a second, providing ample time to deliver the reward. This approach underlines the importance of timing in marking a behavior while also allowing flexibility in how the subsequent event unfolds.
The Reinforcement Event
In our dog training approach, the reinforcement event is categorized into commands and teachable moments, each designed for distinct situations and outcomes.
Commands
Commands involve direct instructions we give our dogs, with the expectation of immediate action. This process begins with the issuance of a command, then followed by a physical cue—like a lure or leash pressure—to guide the dog towards the desired behavior. The event reaches its conclusion when the dog successfully performs the action, often leading into a reward event to reinforce the behavior positively.
Teachable Moments
Teachable moments arise from situations where a dog engages in behavior that we want to prevent—such as violating a boundary or breaking a stay command. These moments provide an opportunity to reinforce preferred behaviors without punishment. For example, if a dog jumps on the sofa or enters a restricted area like the kitchen, we can utilize these instances to instill the preferred behavior:
Identify the unwanted behavior—e.g., the dog jumps on the sofa.
Choose to issue a direct command, such as “off,” or use a marker word, like “wrong,” to indicate the preferred behavior to the dog. The preference for using “wrong” stems from its simplicity and helps eliminate the risk of mistakenly issuing an incorrect command during the crucial one-second window—for example, saying “down” instead of “off."
This cue marks the beginning of the reinforcement event, aiming to modify the dog's behavior. The event concludes once the dog complies, demonstrating their understanding and adjustment, such as getting off the sofa or returning to the stay position.
After compliance, we offer praise to acknowledge the right behavior, though not a tangible reward, focusing on positive reinforcement through acknowledgment rather than physical rewards.
Guidelines for Teachable Moments
For teachable moments, I adhere to a couple of fundamental guidelines:
Negative Reinforcement: I utilize negative reinforcement to encourage the behaviors I want to see. It's important to note that negative reinforcement is not a form of punishment but rather a strategy to reinforce desired actions. Typically, I employ leash pressure to guide the dog towards the desired behavior.
Praise Without Physical Reward: Once the dog complies with the command, I offer praise but refrain from providing a physical reward. The rationale behind this is that during teachable moments, the dog is essentially triggering the event, making our response reactive rather than proactive. If a dog learns that initiating a specific behavior results in a reward, they're likely to repeat this behavior to obtain the same outcome, which contradicts the training objective.
Commands vs. Teachable Moments
This distinction between commands and teachable moments underscores the comprehensive and nuanced approach required in effective dog training. Commands demand immediate compliance with specific instructions, whereas teachable moments are about guiding and adjusting undesired behaviors, thereby setting clear boundaries and expectations for our dogs. It's helpful to understand that the reinforcement event begins the moment we desire our dogs to perform an action and concludes only after that action has been successfully completed. A common mistake is to give up mid-way through the reinforcement event. It's imperative to recognize that perseverance is key; relinquishing effort before the behavior is fully completed undermines the training process and fails to establish the desired outcomes.
The Correction Event
The final aspect of our training methodology is what I term the 'correction event,' which typically incorporates the use of positive punishment to address behaviors we wish to discourage. This is often executed through a quick leash correction or a stim from a remote training collar. The correction event is initiated the moment the dog engages in an undesired action and concludes immediately after the correction is administered. This phase of training is the most direct and unambiguous, as it operates under a clear and simple principle: to promptly address and rectify undesirable behaviors with minimal exceptions to the rule.
Flexibility in Training
Returning to the concept of the reward event, it's important to understand that while this event traditionally concludes when the dog receives their reward, there are instances where it can be interrupted for either a reinforcement or correction event. To illustrate, let’s consider the 'stay' command. Imagine your dog is performing well, so you use your continuation marker, signaling that a reward is forthcoming, yet they are not to leave their position. As you approach to deliver the reward, your dog prematurely exits the stay position. At this moment, the appropriate response hinges on your dog's current stage in their training journey. If they're still grasping the basics, you might initiate a teachable moment by using your chosen indicator word, like 'wrong,' to communicate the error, guiding them back into the stay position without dispensing the reward. Conversely, if your dog has progressed beyond the foundational stages and understands the expected behavior, breaking the stay may warrant a correction event, applying a suitable form of positive punishment to address the lapse.
From this example, it's clear that the dog did not receive the reward initially anticipated from the reward event, as it was superseded by either a reinforcement or correction event, depending on the dog's response and stage of training. This demonstrates the importance of flexibility in training, adapting our responses based on the dog's behavior to guide their learning process effectively.
Transitioning from Teachable Moments to Correction Events
When do we transition from a teachable moment to a correction event? While you may establish your own criteria, my approach involves shifting to the correction event once the dog responds appropriately to the marker word 'wrong' for the behavior we're addressing. For instance, consider the scenario where a dog jumps on the sofa. If I say 'wrong' and the dog immediately jumps off without requiring leash guidance, this indicates an understanding of the instruction within the context of this specific behavior—be it staying off the sofa or respecting other boundaries. This reaction signals to me that the dog has had sufficient opportunity to grasp this rule. Therefore, if the dog repeats the behavior, instead of saying 'wrong,' I'll use 'no,' my marker for initiating positive punishment. Following this change, if the dog jumps on the sofa again, I say 'no,' thereby entering the correction event. The event concludes once the correction has been administered.
Exceptions to the Rules for Reinforcement Events
Let us now examine the exceptions to the rules for the reinforcement event, which apply to commands but not to teachable moments. Even though a teachable moment falls under the reinforcement event, it is treated differently. When the dog triggers a teachable moment, I will use the leash cue to get the dog to perform the desired behavior. I would not correct a dog for not complying during a teachable moment, as I will make the dog do it regardless. Additionally, I would not reward a dog in the middle of a teachable moment since that would violate the principles I have previously outlined.
Remember, as mentioned earlier, in the context of a teachable moment, if desired, one may use a command instead of a marker word like “wrong,” which I use as a marker predicting negative reinforcement. For example, when the dog jumps on the sofa, we can say “off” instead of “wrong”; for entering the kitchen, we can say “exit” instead of “wrong”; or when the dog breaks a stay, we can use the commanded position the dog just broke. The rationale behind this approach is to communicate and demonstrate to our dogs what we want them to do when they make a mistake, without initially punishing them.
Interrupting Reinforcement Events
A reinforcement event as it relates to a command may be interrupted by either a reward event or a correction event. Interrupting with a reward event is more common at the beginning of a dog’s training program due to the learning process. For a dog with low perseverance, one that may give up easily, I would use what is known as assisted shaping. This involves rewarding minor progressions or efforts towards the final behavior, thereby increasing the dog’s tenacity and perseverance. However, it is much less common when the dog already knows how to perform all the behaviors on command or with a physical cue. Consider the following example:
Suppose you are training the “come when called” command, also known as recall, and you define recall as the dog sitting directly in front of you, which is referred to as the sit-front position. To complete the reinforcement event during a recall, the dog must sit in front of you. Imagine you aim to accelerate the speed of your dog’s recall. As your dog begins running towards you, you use your terminal marker, indicating release plus reward. By using this marker, you have released your dog from the event. Following the terminal marker, you throw your dog’s favorite toy behind you, prompting the dog to run past you to retrieve the reward. This illustrates how a reinforcement event can be interrupted for a reward event. However, one should not do this too frequently, as it might turn the recall command into the dog running past you rather than coming directly to you.
We can also interrupt a reinforcement event with a correction event. For the dogs I train, I prefer the dog to be at least 80% proficient in the command within that environment, since dogs need to be generalized. This means a dog may perform a command perfectly in one environment but still need more training to perform at the same level in a different environment. If the dog meets this proficiency level and does not comply with the command, I would say “no” and enter the correction event. It is also essential to ensure the dog has heard the command, which is why some trainers say the dog’s name before issuing a command. This process occurs rapidly; within the first second after issuing the command, one should ascertain whether the dog will attempt the behavior. If the dog does not, we then enter the correction event.
Executing the Correction Event
For clarity, this would be the process: let us say you tell your dog to sit, and within the first second, you know your dog is choosing not to perform the requested command. Once you observe this, you would use your marker that predicts positive punishment; for me, that is the word “no.” Immediately after saying your marker word, you would correct your dog with a leash pop on the training collar or a stim on the remote collar as an example. Once the dog receives the correction, that would conclude the correction event. If, after the correction, the dog is still not sitting, you would reissue the sit command, and if needed, help the dog perform it with the leash cue. The reason for this is that we do not want to correct multiple times for the same mistake. This means I would not say sit, then no, and then correct, say sit again, then no again, and correct again, and so on. After the first correction, we must assume that the dog will be mildly stressed and, because of this, may have a harder time performing the behavior. Therefore, the sequence goes: no, then correct, then re-command, then assist.
Avoiding Reinforcement Event Interruption Errors
In addition, we do not interrupt a reinforcement event with another reinforcement event. This is a common mistake many people make, and I admit to having done it myself. This often occurs when individuals see me use the marker “wrong” in my videos but do not fully understand its application. Let me provide an example: they issue a command to the dog, such as “down,” and the dog does not comply. Then, they say “wrong” and use leash pressure to make the dog perform the action. This might seem logical at first glance. However, the issue lies in the fact that both the command and the word “wrong” instruct the dog to perform an action. If we start saying “down wrong,” our command effectively becomes “down wrong,” much like we do not want our command to become “down down.” Instead, when we issue a command, we should already know whether we are going to enter a reinforcement event and assist the dog in performing the action or if we will enter a correction event upon non-compliance.
As mentioned earlier, this applies to commands but not to teachable moments. If I tell a dog “wrong,” I will show them what I want them to do. I will not say “wrong” and then correct the dog if they do not comply; instead, I will simply use the leash cue to get the desired behavior. However, I have corrected dogs in the middle of a teachable moment if they resist the leash pressure cue, assuming the dog is well trained on leash pressure. This means that if I direct the dog where I want them to go—whether back in a stay or out of a restricted area such as the kitchen—by applying the leash cue and the dog resists, I will say “no,” quickly pop the leash, and then return to the leash pressure. This often results in instant compliance at that point, allowing us to finish the teachable moment.
It is important to reserve this for non-compliance rather than mistakes. In instances of mistakes, such as issuing a command and the dog performing a different action, I reissue the command or assist the dog in performing the correct action with a physical cue. This differs from repeatedly issuing a command due to non-compliance. For instance, if you command your dog to sit and the dog spins instead, you can either continue with the same reinforcement event and cue the dog into a sit or pause briefly and reissue the command. Since the dog performed an action (spinning) after the first command, reissuing the command offers another opportunity for compliance.
Understanding the Correction Event
For the correction event, we almost invariably follow through with the correction, and this is where many individuals err. We understand that the correction event concludes the moment the dog receives the correction, and we have defined a correction as positive punishment. However, just as our dogs can become classically conditioned to a sound that predicts a reward, a dog can also be classically conditioned to a sound that predicts positive punishment. This is known as a conditioned punisher. Much like a dog that is classically conditioned to a word predicting a reward exhibits a change in behavior upon hearing the word—such as licking their lips, salivating, opening their mouth, moving their paws in excitement, or wagging their tail—we also observe a change in behavior when a dog hears the word predicting a correction. The four most common responses are instant compliance, where the dog immediately performs the desired action upon hearing the marker; evasion, where the dog attempts to avoid the correction; freezing in place; and running to the handler. There may be variations of these responses, but these are the most frequently observed.
Follow-Through in Correction Events
The main point here is that regardless of the dog’s response to the marker predicting the correction, the correction must still be administered. Recall from earlier that the correction event concludes when the dog receives the correction, not when the dog complies or attempts to escape upon hearing the marker, but when the correction is delivered.
I have taught this concept to numerous students, and I consistently pose the same question after imparting this knowledge. Nearly all students respond incorrectly. Let us see how you fare. Imagine you are training your dog not to jump on the sofa, beginning with a teachable moment. For several days, each time your dog jumps on the sofa, you say “wrong” and then guide your dog off the sofa, followed by praise. On the fourth day, your dog jumps on the sofa, you say “wrong,” and your dog immediately jumps off the sofa without any assistance. You recognize this as a milestone indicating that your dog has learned the rule, and the next time the dog jumps on the sofa, it will not be a reinforcement event but rather a correction event.
Later, the dog jumps on the sofa, but instead of saying “wrong,” you say “no” to initiate the correction event. Upon hearing “no,” your dog instantly jumps off the sofa. What is your next course of action?
If you’re like most people you would respond that you would praise the dog for performing the desired action or perhaps even offer a reward—both common answers. However, these responses are incorrect. Instead, you must follow through with the correction event by administering the correction. You might wonder, “But did the dog not do what we wanted?” Yes, the dog got off the sofa, which was the desired outcome, but the dog still triggered the correction event. Similar to all our markers, we use them to denote a specific moment in time when our dog’s behavior is correct or incorrect. We said “no” when the dog was on the sofa, and if the dog is classically conditioned to the word “no,” it understands that a correction is imminent. If the dog learns that there is no way to avoid the correction once the event is triggered, it will quickly learn to avoid any behavior that triggers this event.
Interestingly, the higher our consistency in following through with the correction event—meaning correcting 100% of the time after saying “no” compared to correcting only 40% of the time—you will achieve greater reliability in all your commands and rules while administering fewer corrections. The reason is that inconsistency in rule enforcement leads to prolonged learning periods for the dog, resulting in more mistakes, extended training durations, and consequently, more corrections in the long run.
Recap and Application
Now you should have a comprehensive understanding of these events and their application in real-world scenarios with your dog. To recap, let us review the primary obedience and behavioral issues.
Commands
It is essential to achieve 80% proficiency in the dog before correcting non-compliance. How is this accomplished? The process is straightforward: teach the dog a physical cue that ensures the desired behavior. Consider the “sit” command as an example. You instruct your dog to sit by lifting your hand with a treat, functioning as a food lure. The dog's nose lifts, and the hindquarters lower, resulting in the dog sitting. Upon sitting, you mark the behavior and initiate a reward event. Once the dog reliably performs the behavior with the physical cue, you introduce the command. The command precedes the physical cue to prevent overshadowing. Overshadowing occurs when two stimuli are presented simultaneously, and the dog focuses on the more relevant stimulus, ignoring the other. Thus, if you say “sit” while luring the dog into the sit, the dog will focus on the food lure and disregard the command. To avoid this, issue the command, pause briefly, and then present the physical cue. Commands can be considered questions, and the physical cue is the answer. You ask the dog the question, and then you provide the answer. Eventually, the dog will respond to the question before the answer is provided. Once the dog anticipates the answer 80% of the time, you may correct for non-compliance as previously outlined.
The Stay Command
Next, we address the stay command. Initially, when the dog breaks the stay command, we employ a teachable moment. We either restate the commanded position or use our marker word predicting negative reinforcement. This process continues until the dog returns to the stay position independently after hearing the command or marker. Subsequently, we transition to the correction event. When the dog breaks the position again, the correction event is triggered, ensuring reliability for the stay command.
Behavioral Issues
For behaviors that we wish to prevent, such as boundary violations, we follow the same procedure as for the stay command. The dog violates the boundary, and we begin with a teachable moment, transitioning to the correction event once the dog responds effectively to the teachable moment.
For dangerous or destructive behaviors, it is advisable to proceed directly to a correction event, provided the behavior is not rooted in fear or aggression. An example is digging. The moment the dog begins to dig, administer a correction to stop the unwanted behavior. It is important to note that the longer a behavior has been practiced, the more corrections will be required. If the dog is corrected the first time it attempts digging, it is likely not to repeat the behavior. However, if the dog has been digging for months and you decide to stop it, multiple corrections may be necessary before the behavior ceases.
Finally, we have fear and aggression issues, which are considerably more complex. I will dedicate an entire article to this subject alone.
Conclusion
We have now covered the reward, reinforcement, and correction events, including their initiation, conclusion, and exceptions to the rules. We have demonstrated how these principles can guide us in addressing obedience and behavioral issues. This structured approach will enhance your ability to train and communicate with your dog, leading to the desired results. As you apply these principles, remember that every moment with your dog is an opportunity to learn and grow together. Stay patient, stay consistent, and celebrate the small victories along the way. Your dedication will lead to a well-behaved companion and a bond that enriches both your lives. Believe in the process, and you'll witness the transformative power of training.
コメント