Every term used on this site, defined in plain language. Tap any highlighted word on any page to see its definition without leaving the page.
The tiny span of time in which a dog connects what just happened to what happens next. If your dog does something and a consequence follows within one to three seconds, the dog links the two. If the consequence comes later — even thirty seconds later — the dog either connects it to the wrong thing or does not connect it at all. This window governs everything: every reward, every correction, every marker must land inside it. It is the reason you cannot punish a dog for something he did an hour ago, and the reason markers exist — they capture the exact moment of a correct behavior so the dog knows what he did right, even if the treat takes another second to arrive.
Whatever happens right before a behavior. A hand signal, a verbal command, a sound, the sight of another dog — anything that triggers or sets up the next action. In training, you control the antecedent to produce the behavior you want. "Sit" (the antecedent) leads to the dog sitting (the behavior) leads to "yes" and a treat (the consequence).
Projecting human thoughts, emotions, or motives onto a dog. "He knows what he did." "She is being spiteful." "He is bored." These are human feelings applied to an animal that does not experience the world that way. Dogs do not have morality, spite, or boredom in the way we understand those things. Every time you assume a human motive for a dog's behavior, you move further from understanding what is actually happening and further from being able to fix it.
Behaviors a dog performs to make pressure go away. Cowering, tucking the tail, rolling over, avoiding eye contact, slinking low to the ground — these are not guilt. The dog is not sorry. He has learned through experience that when a certain situation arises (you come home and something is destroyed), pressure follows. The cowering is his attempt to turn that pressure off. It works because we usually soften when we see it. That is conditioning — not conscience.
Anything the dog would prefer to avoid. This does not mean painful or cruel. A sharp "aat" sound is an aversive. Removing your attention is an aversive. A quick leash pop is an aversive. The word simply describes something the dog does not enjoy. Aversives are used as information — they tell the dog "that was not what I wanted" — not as expressions of anger.
Changing how a dog responds to something in its environment. This goes beyond teaching tricks. A dog that lunges at other dogs, a dog that bites out of fear, a dog that destroys the house when left alone — changing these deep patterns is behavior modification. It uses the same tools as basic training (markers, reinforcement, conditioning) but applies them to emotional and behavioral problems.
A type of conditioning where two things happen together so many times that one starts to mean the other. A bell rings, then food appears. After enough repetitions, the bell alone makes the dog expect food. The dog didn't have to do anything to earn it — the connection just formed on its own through repetition.
A signal — verbal or physical — that tells the dog what you want it to do. "Sit," "down," "here," a hand signal, a whistle. The word itself does not matter. You could use any sound in any language. What matters is that the dog has been conditioned to connect that specific signal to a specific behavior, and that the signal is delivered the same way every time.
The process of teaching through repetition. You repeat something enough times in the same way, and the dog forms a connection between two things. That connection becomes automatic — the dog doesn't have to think about it.
A marker that means "that is correct, keep doing it." It does not end the behavior — it sustains it. Most people use the word "good" in a calm, low tone. The dog hears it and stays in position because he knows he is on the right track and more reward may follow. The energy behind it is quiet and steady.
An action taken to interrupt or redirect an unwanted behavior. A correction can be a verbal sound, a light leash pop, a body block, or removing a reward. It is not revenge. It is not anger. It is delivered calmly and immediately, and it is always followed by an opportunity for the dog to try the correct behavior and be rewarded for it. A correction without a path to success is just confusion.
A marker that means "that was wrong, try something else." Most people use "no" or a short sound like "aat." It is not yelling. It is not anger. It is delivered flat and neutral — just information. The dog hears it, stops what it is doing, and tries a different behavior.
Replacing one response with a different one. The dog jumps when you come home — instead of telling him "no," you ask for a sit and reward the sit. Over time, the dog's automatic reaction to your arrival changes from jumping to sitting. You are not suppressing the old behavior. You are building a new one that takes its place.
A gap in time needed for a dog to stop connecting one event to another. In practice, this is most relevant to crate training: if a dog barks in the crate, goes quiet, and you open the crate within two minutes, the dog still believes the barking caused the door to open. You must wait at least two minutes of sustained silence so the dog no longer links the barking to the result. Only then does opening the crate reward the quiet.
A built-in urge that pushes a dog to act. Chasing a squirrel is prey drive. Wanting to be near the pack is social drive. Tugging on a toy is play drive. Drives are not taught — they come with the dog. Training works by channeling drives toward behaviors you want, not by trying to eliminate them.
How long a dog holds a behavior before being released. When you ask a dog to sit and they stay seated for ten seconds before you say "yes," that ten seconds is duration. Duration is built gradually — you start with one second and increase over time, using a continuation marker ("good") to tell the dog they are on the right track. Duration is not a separate command. It is a property of every position command, sustained by the continuation marker and ended by the terminal marker or release word.
The state where your dog sees you as the most interesting and rewarding thing in any environment. A dog with strong engagement looks at you when other dogs appear, comes to you instead of chasing a squirrel, and stays close because your side is genuinely where the best things happen. Engagement is not obedience through fear. It is a relationship where the dog chooses you because you have consistently been worth choosing. This is the single most important concept in this entire system.
The process of teaching a dog that a behavior applies everywhere, not just where it was first learned. A dog that sits perfectly in your kitchen but ignores you at the park hasn't failed — the sit was never trained outside the kitchen. The dog genuinely doesn't know that "sit" means the same thing in a new place with new smells, sounds, and distractions. Generalization is built through progression: once a behavior is reliable in one setting, you practice it in a slightly harder one, then a slightly harder one after that, changing one variable at a time — location, distraction level, distance, intensity — until the behavior works regardless of where the dog is or what's going on around them.
Deep associations formed during a dog's earliest developmental period (roughly the first few weeks of life). These connections are more powerful and more permanent than anything learned later. Positive imprinting is used intentionally — such as exposing detection dogs to specific scents from birth. Negative imprinting can happen accidentally — a traumatic experience during this window can create lifelong behavioral responses that are extremely difficult to change.
A command that means "disengage from whatever has your attention right now." It is one of the most important safety commands. A dog fixating on another dog, sniffing something dangerous on the ground, or staring at a squirrel — "leave it" breaks that focus and redirects the dog back to you.
Using food or a toy to physically guide a dog into a position. Holding a treat above a dog's nose so he looks up and his rear drops into a sit — that is luring. It is a starting point, not a destination. If you lure forever, the dog only follows the food. The goal is to get the behavior with the lure, then remove the lure piece by piece until the dog responds to the command alone.
Controlling a situation so an unwanted behavior cannot happen, without teaching the dog anything new. Holding a tight leash when a dog lunges, crating a puppy who is unsupervised, putting a dog in another room when guests arrive — these are all management. Management is not training. It does not change future behavior. It prevents damage in the moment. It is what you do when the dog is over threshold or when you cannot actively train. Good training plans use management to prevent failure between training sessions.
A signal — a sound or a word — that tells the dog the exact moment it did the right thing. Think of it like the shutter click on a camera: it captures the precise instant of the correct behavior. The marker itself is not the reward. It is a promise that the reward is coming. It only works if the dog has been taught what it means through conditioning.
A training method built around markers. Every correct behavior is captured with a marker sound at the exact moment it happens, then followed by a reward. The marker bridges the gap between the behavior and the reward so the dog knows precisely what it did right — even if the treat takes a second to deliver.
Does not mean "bad." It means "removing something." When you see the word "negative" in a training term, it simply means something was taken out of the situation — food was pulled away, leash pressure was released, attention was withdrawn.
Removing something the dog wants in order to make a behavior happen less often. The dog jumps on you, you turn away and ignore him. Something was removed (your attention) and the behavior (jumping) decreases. Walking away from a dog who is not listening is negative punishment.
Removing something unpleasant in order to make a behavior happen more often. You apply light, steady pressure on the leash. The dog steps toward you. You release the pressure. The dog learns that moving toward you makes the uncomfortable thing stop. The relief itself is the reward.
A dog that waits for guidance rather than experimenting. You hold food in your hand, and this dog stares at you and does nothing. He is not stubborn — he is waiting to be shown. These dogs need physical guidance: gently positioning them into a sit, luring them into a down, then marking and rewarding once they are there. The learning still happens, it just starts differently.
A type of conditioning where the dog figures out that doing something specific leads to a result. The dog tries different things — sitting, pawing, laying down — and discovers which action gets the reward. The dog is actively choosing to do something because of what happened last time they did it.
A dog that naturally tries different things to figure out what gets rewarded. You hold food in your hand, and this dog will paw at it, nudge it, sit, lay down — experimenting until something works. These dogs are active problem solvers. Training them involves waiting for them to offer the behavior you want, then marking and rewarding it.
A body movement the dog uses to know what you want — a hand motion, a head nod, a step in a certain direction. Dogs read physical cues before they process verbal commands, because dogs are body language creatures. Most dogs learn the physical cue first and treat the spoken word as background noise until it is taught separately. A physical cue must be deliberately faded — made smaller over time — so the dog eventually responds to the verbal command alone.
Does not mean "good." It means "adding something." When you see the word "positive" in a training term, it simply means something was put into the situation that was not there before — a treat, a sound, a leash correction. Whether the dog likes it or not is a separate question.
Adding something unpleasant in order to make a behavior happen less often. The dog lunges, you give a quick pop on the leash. Something was added (the pop) and the behavior (lunging) decreases. This does not require force or anger — it is information delivered calmly.
Adding something the dog wants in order to make a behavior happen more often. Dog sits, you give a treat. The treat was added, the sitting increases. This is the most intuitive form of training — do the thing, get the reward.
A reinforcer that works on its own without any training. Food is the most common one. The dog does not need to learn that food is good — it already knows. A primary reinforcer is what gives a marker its power, because the marker gets paired with it over and over.
The process of making training gradually harder, one small step at a time. All training follows this pattern: start where the dog can succeed, build reliability, then increase one variable. The variables are distance (how far), duration (how long), distraction (how much is competing), and intensity (how exciting or scary the distraction is). Change only one at a time. If the dog fails, you moved too fast — go back to the last point where the dog was succeeding and stay there longer. Failure is not a setback. It is the dog telling you where their current limit is.
Anything that makes a behavior less likely to happen again. This is a science term, not what most people picture. It does not mean anger. It does not mean pain or cruelty. Removing a treat from your hand is punishment. A calm, flat "no" is punishment. If the dog does something less often because of what followed, that consequence was punishment — even if it looked gentle.
A command that means "stop barking." It is taught by first teaching the dog to bark on command ("speak"), then introducing the silence after the bark, marking the silence, and rewarding it. You cannot teach a dog what "quiet" means if the dog does not first understand what barking on command means — because you need to create the contrast.
The act of calling your dog back to you — and the dog actually coming. A true recall means the dog responds to one word, immediately, regardless of what else is going on. It is the most important command a dog can know. A dog with reliable recall cannot bite someone, cannot run into traffic, cannot start a fight with another dog.
Interrupting an unwanted behavior and guiding the dog toward a desired one. A redirect is not just stopping the bad thing — it is giving the dog something correct to do instead. A dog fixating on another dog is redirected into a sit. A dog about to jump is redirected into a sit. The redirect creates an opportunity for the dog to earn a reward, which over time builds a new automatic response to the trigger. A redirect without a path to success is just an interruption.
Anything that makes a behavior more likely to happen again. If your dog sits and something good follows, he is more likely to sit again next time. That good thing — whatever it was — is reinforcement. It can be food, play, praise, or anything the dog values.
The specific thing used as reinforcement. Food is a reinforcer. A toy is a reinforcer. Praise, petting, and play can all be reinforcers. The dog decides what is valuable — not you. If the dog does not care about what you are offering, it is not a reinforcer for that dog.
A word that tells the dog the exercise is over and they are free to move. Some people use "OK," some use "free," some use "break." It does not matter what the word is. What matters is that the dog only hears it when you intend to end the exercise, and that it is consistent every time.
A marker that means "you are done, that was correct, reward is coming." It ends the behavior. Most people use the word "yes." When the dog hears it, he knows the task is finished and he is free to move. The energy behind it is clear and upbeat.
The point at which a dog stops being able to think and starts reacting. Below threshold, the dog can hear you, respond to commands, and make choices. Above threshold, the dog is running on adrenaline, fear, or excitement — their brain is offline and no training can reach them. Every dog has a threshold, and it shifts depending on the situation. Finding and staying below it is the foundation of working with any distraction, fear, or reactivity issue. If the dog cannot respond to you, you are past threshold — back up until they can.
A pattern of rewarding where the dog does not get a treat every single time, but cannot predict which time will pay off. First you reward every correct response. Then most of them. Then some. Then randomly. The unpredictability makes the dog try harder, not less — the same reason people keep pulling a slot machine lever. Once a behavior is learned, variable rewards maintain it better than constant rewards.