Obedience
Below is an excerpt from the U.S. Military Dog Training Handbook by Department of Defense (DoD)
SIT AND DOWN
Teaching Phase
SIT AND DOWN
Teaching Phase
Sit and Down are best taught by luring the dog into position with soft,
appetizing food. This food must be something like small pieces of meat or specialty
dog food that the dog is eager for and that it swallows quickly without chewing.
Crunchy treats do not work well. The dog is first taught to eat from the hand and then
taught to maintain soft contact with the hand and follow the hand until it is allowed to
eat. Then the handler uses the closed hand to lure the dog into position (square, erect
sit; or sphinx-like down position with both elbows in contact with the ground), and
then loosens the hand so that the dog can lick and nibble the food out of it while
holding position. Then, before the dog breaks position the handler releases the dog by
saying "OK" and enticing the dog out of position. Through a number of steps, this is
developed into sit and down on command with stay (for 5 or 6 seconds), using the
dog‘s food motivation only.
The same method can be used with the ball, but because the dog‘s level of
excitement will be much higher than in the case of food, the technique requires more
skill and experience. In addition, prior to luring the dog into position using the ball,
the animal must be taught to release it cleanly on command ("out"), to refrain from
biting the reward until given permission, and to respect (i.e. not bite) the trainer‘s
hands. The handler must be able to hold the ball/kong in his/her hand an inch or two
from the dog‘s head without having the dog take the ball until it is given permission.
In addition, the dog must learn to follow the hand closely, the way a dog would
naturally follow a hand in which food is held, but without snapping at or biting the
hand.
"Sit" and "Down" commands are given as the dog is lured into position. "Good" is used when the dog is holding position well, before the dog is fed. The dog
is fed in position, without being released. "OK" and enticement are used to release the
dog from position. "No" is used to mark errors, and to tell the dog that it will not be
rewarded. For instance, the dog is in down position, the trainer moves the hand
towards its nose to feed it, the dog begins to crawl towards the hand to eat, the trainer
says "No" and stands upright, withdrawing the hand and the food until the dog restabilizes
in down position. "Stay" may be used to steady the dog in position, once
longer sits and downs are introduced.
The "yes" release marker is not employed until the dog is proficient at sit
and down, stays in position until it hears the "OK" cue, and has had many rewards in
position. "Yes" is introduced by having the dog sit or lie down, saying "yes," enticing
the dog out of position (so that it releases), and then feeding the dog. "Yes" may be
used with food reward or ball reward interchangeably.
Training Phase
The first training of sit and down begins with the stay component rather
than the actual sitting or downing motion. That is to say, when we begin to prepare
the dog for the experience of being forced to sit or lie down, we apply the force to
make the dog stay in sit or down position, rather than sit or lie down in the first place.
The handler uses a handful of food to lure the dog into position and then rewards the
animal. Then he/she tells the dog to "stay" and waits for a mistake. In fact, the
handler does whatever is necessary to cause the dog to break the stay -- stands upright,
holds the stay an unusually long time, etc. When the dog attempts to break
(prematurely release from) the stay the handler applies a very quick but rather gentle
pop on the leash (up and away from the handler in the case of the sit, and directly
backwards along the dog‘s spine in the case of the down), sufficient to stop the dog
from breaking, and then the handler quickly brings the food hand back and feeds the
dog and repeats the exercise. When this is done skillfully it is not clear what keeps the
dog in position, the collar correction or the dog‘s fixation on the food-bearing hand.
With time the dog comes to associate the leash correction with sit and down position.
Now we can begin to use collar corrections to enforce the "sit" and "down" motions,
in addition to the "stay."
Alternate mode of correction for the down. For the down, especially, it is
very advantageous for the handler to use a ―social‖ correction (4.2.7.3.3)—a slapping
correction rather than a collar correction-- because a slap is normally faster and if
done well a slap is more effective in pressing the dog into the down position. But we
cannot just suddenly slap a dog with leash or hand and expect it to understand. The
animal must learn the meaning of this correction and connect it to the previously
understood skill. The handler stands with food in hand and signals the dog into the
down with a long hand movement towards the ground and past the dog‘s nose, with
food in the hand. This is merely an exaggeration of the movement the handler
normally uses to lure/signal the dog into down position with food. Once the dog is
down, it is fed in position, then given the "sit" command, enticed up to the sitting
position, and then signaled back into the down. This sequence of sit-down-sit-down is
repeated several times. After this repetition, the dog will anticipate the next "down" command and it will be waiting eagerly to lie down. As the handler again makes the
long downward hand gesture combined with the command "down," he/she clips or
cuffs the dog rather gently on the muzzle with the ends of the fingers. The dog will
notice the contact, perhaps blink or flinch away, and then quickly lie down because
both force of habit and the near proximity of the food will guide it into this well rehearsed
behavior. We must be clear —We are not forcing the dog to lie down. The
animal is lying down voluntarily to obtain the food. Before it even feels the contact of
fingers on its‘ muzzle it is already beginning the down motion. But we are preparing
the dog (training phase) for the experience of being forced to lie down (proofing
phase).
Proofing Phase
During proofing of the sit and down skills we begin to challenge the dog‘s
understanding, with more distracting surroundings and longer stays, less frequent
food or ball reinforcement, and more praise reinforcement instead. When the dog performs correctly it is rewarded and encouraged, and when it refuses commands or
becomes distracted, the corrections that were introduced and ―attached‖ to the
exercises during training are used in a stronger form, to ensure compliance. For
instance, if the dog refuses to sit it is given a quick, popping collar correction
upwards. If it refuses to lie down the handler slaps it lightly but sharply on top of the
neck or skull with the flattened hand or with a loop of the leash.
As a rule, early in the proofing process, to keep the animal motivated and
reduce stress, we tend to give the dog rewards after a correction. For instance, if the
trainer gives a "sit" command, but the dog is distracted by another dog nearby and
therefore does not sit, the handler delivers a quick, popping correction on the
leash/collar, the dog sits, and then the handler rewards (either with "yes" and release,
or "good," "yes," and release, see 4.2.4.1.4, this Chapter). Rewards after corrections
help to reduce stress, and help the dog keep trying, even under a little bit of
pressure.
Later in proofing we raise our standard. If the dog must be corrected to
secure compliance, then the animal does not receive any reward beyond a bit of praise
and petting. Then the dog is released and immediately asked to repeat the exercise. If
this repetition is correct, then the dog is rewarded.
We should always keep in mind the difference between a mistake on the
dog‘s part, where the animal is trying to do as the trainer asks but just makes an error,
and disobedience or refusal. As a rule we do not often correct simple errors, instead
we punish them with the word "No" and we withhold reward (omission). Sharp
corrections (positive punishment and negative reinforcement) are normally reserved
for disobedience or refusal.
During proofing of sit and down, the motivation is normally supplied by
ball. The dog is initially taught using food if possible and then once it understands
the exercises the ball is introduced. Introduction of the ball will result in the dog
becoming much more excited than it did when working for food. Ball motivation is
thus a good way to challenge the dog so that it makes a few mistakes, and also it
helps the dog to shake off any discouragement or stress it feels as psychological
pressure gradually becomes a part of training.
COMMUNICATION DURING SIT AND DOWN
While initially teaching the sit and down with food, verbal cues are of
relatively little importance. The handler‘s gestures as he/she lures with food are most
important. However, we normally give the "sit" or "down" command as we lure the
dog into position, praise the dog with "good" before and during reward, and release
with the cue "OK." Later, as we move into training and proofing stages and begin to
use ball reward, these verbal cues become very important, and we also begin to make
use of the markers, "yes," and "no."
Reward in position for sit and down. Our first concern with sit and down is
establishing stability —making the dog understand that its job is to stay still without
fidgeting or creeping. The best way to do this is to make sure that the animal receives
its reward while it is still holding the sit or down. This is called a "reward in position."
For both sit and down, reward in position is performed as
follows —The handler, with the ball in pocket, gives the "sit" or "down" command.
When the dog moves swiftly and correctly into the appropriate position, then the
handler marks this behavior with the "good" cue (intermediate marker). This cue tells
the dog that it performed well and earned reward, but that it must not break position
yet. Then the handler gets the ball out of his/her pocket and holds the ball very
closely in front of the dog‘s nose. Some care and a bit of training is required so that
the dog does not creep or break position as the ball is brought out, and does not try to
take the ball before given permission. After a moment in which the handler makes
sure the dog is steady, he/she gives the "yes" cue (terminal bridge) which is the dog‘s
authorization to take the ball. If this technique is done well, the dog takes the ball
while still in position and then releases from the position. Use of "good" and "yes" cues, combined with reward in position, achieves the twin goals of making sure we
mark and reward the correct movement into sit or down position, yet also keep the
position stable and prevent creeping, fidgeting, or breaking towards the handler and
reward. Reward in position (using “good’ then “yes”-release) teaches the dog to
hold a position.
"Yes" -release for sit and down. Once we have a stable sit or down, our
next concern is making sure that the dog understands that the correct response to the
commands "sit" or "down" is a rapid, crisp movement. To do this we must reward the
sit or down movement, rather than the stay. What is important here is that we pick out
and mark the critical aspect of the skill—in the case of the sit, the moment the dog
fully sits, and in the case of the down, the moment that the dog gets both elbows in
contact with the ground. Using the down as our example, the handler does this by
giving the "yes" cue the instant both elbows are in contact. When it hears "yes," the
dog will release from position and wait to be rewarded. The handler then breaks
position and retrieves the ball from his/her pocket and gives it to the dog. The handler
must not be in a hurry to get the ball out, and it is extremely important that the
handler not break his/her position until after he/she has said "yes." The critical aspect
of timing here is when the handler says "yes." The "yes" must be timely, but the
delivery of the ball can be and should be done deliberately and without hurry.
Remember that the function and value of the "yes" terminal marker is that it bridges
delays to reward. Therefore, if the "yes" is properly conditioned and well-timed, there
is no need to hurry in delivering the ball. Reward at completion of a movement (with
“yes”-release) teaches the dog speed in assuming a position.
"No" marker. "Good" and "yes" are not the only markers we can use, nor is the power of marking technique limited to rewards. We can also mark behaviors we want to punish, by using the word "no." "No" is used much like "yes," in the sense that it is used to mark a behavior, and bridge a delay. In this case a delay to punishment. Earlier we covered two kinds of punishing response contingencies, positive punishment (giving the dog something that it dislikes) and negative punishment (or omission, taking away from the dog something that it likes). "No" can serve to signal both of these response contingencies. "No" is also used like "yes" in the sense that it tends to be a terminal cue—when the dog does something that earns a "no," the animal often has to start the whole exercise again and therefore it can break when it hears "no."
We can use the "no" to gently punish the dog (through omission) for overeager or careless mistakes. For instance, if we are working the dog through a sit-down-sit sequence, and the dog does not wait for the sit command but pops up without permission, then the handler gives the "no" and makes the dog go back into the down again and wait for the command before rising up into sit. If the animal rises to the sit correctly, it can be rewarded in position with "good" and "yes," or it can be rewarded with "yes" and allowed to break immediately. Here the "no" gives us the ability to improve performance without killing the enthusiasm of an eager dog.
We can also use the "no" more forcefully as a predictor of physical (positive) punishment. Let us say the handler leaves the dog on a down-stay and steps a few feet away, and the dog breaks position and moves toward the handler without permission. If the handler then simply corrects the dog, the animal will be corrected in the act of approaching the handler, which can be hopelessly confusing for an eager dog. What is needed is a tool to tell the dog exactly what critical behavior earned it the punishment. Therefore, the handler gives the "no" immediately when the dog‘s elbows lift from the ground. Then the handler calmly approaches the dog and administers a correction of appropriate strength for the dog, normally by popping the leash two or three times, then takes the dog back to the exact place where the dog was lying and commands it to lie down. Then the handler steps away and, if the dog holds position correctly, the handler performs a reward in position first "good" to reinforce the act of holding position, then approach and placement of the ball directly in front of the dog‘s nose, and "yes" to release the dog into the ball.
ADVANCED SIT AND DOWN EXERCISES
To meet certification standards, the dog must eventually learn to transition from sit to down and back up to sit again at heel position, and while at EOL. For these exercises, the dog must not only understand sit and down, but also how to move from one position to the other without creeping forward or changing its alignment. This is where "good" and "yes" and "no" cues come into their own, because they give the handler the power to teach the dog to understand the difference between a perfectly correct down (in which the dog does not creep forward) and an incorrect down (in which the dog lies down every bit as fast, but creeps forward as it does so).
For sit and down EOL, "good" is used to let the dog know immediately when it has performed a correct sit or down, and then the handler approaches and delivers reward in position (by putting the ball close to the dog‘s head and releasing the dog into it with "yes"). If the dog creeps as it transitions from sit to down or vice versa, then the handler marks the mistake with "No," and either makes the dog repeat the skill or calmly delivers a correction and then makes the dog repeat the skill.
For sit and down at heel, "good" is used in the same way to mark a correct transition. Reward in position is performed by taking the ball from the pocket and holding it directly in front of the dog‘s nose at heel position, and then releasing the dog into it with "yes." If the dog creeps forward or slews sideways at heel while performing the transition, the handler marks the error with "no," (perhaps followed by a correction) and sends the dog back to correct heel (see 4.3.3.1) in the original posture (sit or down). Then the handler makes the dog repeat the exercise, rewarding the dog in position if it is executed correctly.
For real-world operations/utilization, the down has greater importance than the sit. The down is the dog‘s most stable position. When given the ―down‖ command the dog must drop immediately no matter where it is and what speed it is traveling, and then lie still and silent until given another command or released, even under intense distraction (gunshots, decoys carrying bite equipment and cracking whips, etc.). Accordingly, substantial psychological pressure must often be applied to achieve this level of obedience. In order for such treatment to be fair and effective, corrections used for the down must be thoroughly trained (see 4.2.2.2), and "good," "yes," and "no" must be properly applied so that the dog understands what it is being corrected for and can adjust its behavior to avoid corrections.
HEELING (MARCHING)
Heeling is an attention-based exercise in which the dog walks at its handler‘s left side (personnel who carry weapons on the left side often teach the dog to heel on the right) with the shoulder even with the handler‘s knee, keeping pace and position no matter what the handler‘s pace or direction, and sitting automatically when the handler halts. The primary functions of heeling are to refine the dog‘s obedience to its handler, and to provide the ability to transport the dog under close control through hazardous or distracting circumstances with both hands free. A well-trained dog concentrates completely on its handler while heeling, and heeling is therefore very fatiguing and not an appropriate way to transport the dog long distances. For mere transportation, where all that is important is that the dog remains under control, does not pull against the leash, and does not use any more energy than necessary, a "walk easy" skill is used instead. Walk easy (the command is normally "easy") is much less strict and demanding than heeling.
It is traditional in DoD to use the verbal command "heel," and also slap the left hip with the left hand, and to repeat these verbal and gestural commands at each change of pace or direction. However, if we view heeling as a tactical tool rather than a parade ground skill, then we must realize that, a) slapping the hip is unnecessary and inadvisable, because the left hand should be left free (for weapon-handling, for instance), and b) repeated commands are also unnecessary and inadvisable, because the verbal command "heel" is all that is necessary to tell the dog that it should place itself at heel position and remain there, no matter what the handler‘s movement or direction, until released. Accordingly, in the discussion of heeling that follows, the left-hip slap and repeated commands are not used. The CST method for teaching heeling described below normally results in a dog that positions itself for marching close to the handler's left knee and hip, with the handler's left hand hanging or swinging outside of the dog's head. In effect, the dog positions itself "between" the handler's left hand and hip and looks directly up at the handler's face, rather than positioning itself "outside of" the handler's hand and arm and looking around the handler's elbow at the handler. This close positioning is in most situations advantageous because it gives better and closer control of the dog and results in fewer training problems and less stress for the dog. Weapons retention is not normally an issue for law enforcement applications because the dog works on the handler's non-weapon-carrying side and the handler need not carry the left arm pinched close to the body (a left-handed handler normally trains the dog to heel on the right side and so can carry the right arm loosely while holding the left arm tight to the body for weapons retention).
The CST method of teaching heeling concentrates on teaching the dog to understand the exact position that it must maintain at heel, how to move its body in order to reach that position, and on making sure that the dog is highly motivated for the work. This is the best way to prepare the dog for the physical corrections that may later be necessary to render heeling ―fail safe for real-world tactical scenarios, in which handler safety depends on his/her control of the dog.
Teaching Phase
The finish. The dog first learns to heel not by walking at the handler‘s left side, but instead by learning to ―finish.‖ ―Finish‖ is an expression used by competitive obedience trainers to describe a skill in which the dog moves from position in front of the handler to heel position. In DoD, the dog does not normally walk around behind the handler, passing to the handler‘s right—instead the dog passes by the handler‘s left hand, turns in place and sits at heel (called the "military" finish).
To teach the finish the dog is lured with food or the ball (in the left hand) from position in front, past the handler‘s left side, and behind the handler about 2 or 3 feet (the handler normally takes one long step back with his/her left foot). Then the handler turns the dog in towards himself/herself (the dog turns counter-clockwise), and leads the dog forward into heel position and asks the dog to "sit." When using food, the handler then allows the dog to eat from the left hand while in heel position. When using the ball, the handler holds the ball in the left hand just in front of the dog‘s nose, gives the "yes," and flicks the ball into the dog‘s mouth. Initially the handler leads the dog through the entire path, covering nearly as much ground as the dog does. With time the handler moves less and less, taking a smaller step backward and shortening the circle he/she makes with the left hand and the reward. Eventually the handler does not step backward with the left foot, simply commanding the dog to "heel" and making a long gesture with the left hand to send the dog past the left side and behind, and turn the dog and bring it back up into heel position. The amount of room the dog is given to accomplish the movement is gradually decreased, and when the entire teaching sequence is performed well, the dog begins "flipping" or "swinging" to heel position (rather than walking behind the handler, turning around, and walking up into heel position). A wall or barrier is often used initially, to help the dog reach a straight heel position at the conclusion of the movement. If the dog sits crookedly or too far forward, the handler gives the "no" cue and makes the animal repeat the whole exercise correctly before rewarding.
In the next step the dog‘s attention is shifted from the left hand to the handler‘s armpit. When the dog reaches heel position, the handler marks this behavior with "good," and then carefully moves the left hand and the ball up above the dog‘s head to a position just in front of/under the armpit. Then he/she gives the "yes" and drops the ball into the dog‘s mouth.
After a few repetitions, the handler actually places the ball in his/her armpit prior to rewarding the dog. The movement begins as before. The dog is led through the finish with the ball in the left hand, concluding with the dog at heel position and the ball held directly in front of the dog‘s nose or on the left side of the dog‘s head. The handler then marks the correct completion of the exercise with "good," and then raises the left hand above the dog‘s head, transfers the ball to the right hand, and places the ball in the left armpit, clamping it there with the bicep. Then the handler drops both hands to natural positions, with the left as always hanging just outside of the dog‘s head. The dog should stare straight up towards the ball from heel position. The handler rewards the dog by saying "yes" (the dog will normally release and rear straight up towards the ball) and delivering the ball directly into the dog‘s mouth by unclamping the bicep so that the ball drops free.
Reward in position at heel. Eventually, the handler begins leaving the ball in his/her pocket until the dog correctly finishes. We now expect the dog to complete the finish in order to obtain the ball reward, but without seeing or following the ball reward. The handler makes the same motion with the left hand, and even cups the hand as though he/she is holding a ball, but the ball remains in the right pocket (at this point the leash is normally transferred to the left hand for the first time and held in a manner similar to holding the ball). When the dog reaches correct heel position, the handler marks with "good," and reaches with his/her right hand into the right pocket, withdraws the ball, reaches across the body and places the ball in the left armpit, and then says "yes" and drops the ball. This is a type of reward in position technique (see 4.2.4.1.4.1).
Another technique for reward in position at heel. In a variation on this procedure, the handler does not place the ball in his/her armpit after the "good" cue, but instead places the right hand with the ball directly in front of the dog‘s head, or to the left of the dog‘s head (to bend the dog‘s head away from the handler and straighten the animal‘s spine) and then gives the "yes" release and rewards in position. Once the dog can correctly complete this skill, it has learned a specific position and a specific movement that will become the basis of heeling/marching.
Eventually the gesture of the left hand (the prompt) that is used to send the dog to heel is faded out, so that the dog swings into correct heel position and into attention on the word "heel" alone.
Now we are ready to teach heeling proper, actual marching with the dog at heel position. Initially the handler keeps the ball in the left armpit to give the dog a focal point. Holding a very, very short leash in the left hand (from 3 to 8 inches, but without any tension on the leash between hand and the dog‘s collar), with the hand outside of and just behind the dog‘s head, the handler gives the command "heel" and shuffles very carefully forward a few feet. When the dog moves well, between the handler‘s left hand and hip while looking straight up at the ball held in the handler‘s armpit, then the handler marks this behavior with "good," and comes to a halt carefully so that the dog does not lose position (initial heeling is often performed along a wall or fence). Once the dog is in sit-halt position, then the handler gives the "yes" release and drops the ball into the dog‘s mouth. It is important to remember that in the initial stages of marching, the dog is rewarded after only a few steps, and normally always in the sit halt. That is, when the dog moves well, the handler marks this with "good" and then comes to a halt and performs reward in position.
With further practice the trainer can leave the ball in the right pocket, asking the dog to move at heel while looking up at the handler‘s face or armpit rather than the ball. Once the dog moves well, then the handler marks this behavior with "good," comes to a halt, takes the ball out of the pocket with the right hand and then performs reward in position in one of the two ways described in 4.3.3.5 and 4.3.3.6. Place the ball in the left armpit and drop it to the dog after saying "yes," or transfer it to the left hand and hold it near the dog‘s head while it is sitting in heel position, and then flicking it into the dog‘s mouth after saying "yes." Gradually the dog is taught to heel for longer periods with the ball in sight and without the ball in sight, depending on the circumstances.
Eventually, the heeling pattern is made longer, with turns and halts and changes of pace, and the dog is made to work for extended periods for the "good" and the "yes" and the reward. If, while moving, the dog loses position by running wide or forging forward or swinging out into a crabbing motion, then the handler marks this error with "No," halts, and re-commands the dog to finish to heel. Once the dog is back in correct position, the handler encourages with "good" and resumes heeling again. If the dog this time maintains correct position it is given "good," sit-halt, and reward in position as in 4.3.3.5 and 4.3.3.6.
Up until this point, the "yes" has normally been given only after the "good," as a way of releasing the dog into the ball. This practice has been advantageous because it served to make sure that the dog always got its reward while holding the desired position. Such rewards in place are optimal for making sure that reward anticipation does not interfere with steadiness in the sit and down and correct position while heeling. However, they do not make full use of the power of the well-conditioned "yes," to instantaneously identify to the dog, and reward, very specific aspects of performance. But now we are ready to begin using the full potential of the "yes."
Once the dog shows that it understands how to finish quickly and efficiently to heel position, maintain focus on the handler by looking up towards the handler‘s face while maintaining correct heel position, move at heel without losing position, and correct itself back to the proper position when it happens to lose position, then we are ready for the final step, in which the handler begins rewarding directly out of heel with the "yes" marker. At any moment that the handler judges the dog should be rewarded, either after the finish, while heeling, or after the dog corrects itself back into position in response to the "No," then the handler rewards the dog by saying "yes." The dog will release from heel and show that it expects reward, and the handler can then withdraw the ball from the pocket and give it to the dog.
Use of the "yes" marker in this way enables the handler to bring to the dog‘s attention and selectively reward very fine-grained aspects of performance, such as small differences in speed of movement, angle, or posture. Competitive trainers find this useful because they are interested in polish and speed and precision, because these are the things that win trophies. However, MWD trainers should also be interested in polish and speed and precision, because these are the hallmarks of a dog that fully understands commands and skills: Only when the dog has full understanding is it fair and effective to apply pressure to the dog in order to make sure that it always performs correctly, even under real-world conditions where failure to perform is dangerous for dog and handler and those personnel that depend upon the MWD team.
Note that the "heel" command is given only in order to finish the dog, and when the handler first goes into motion. The handler may re-command "heel" after a "no" also. In finished form the command for finish and for heeling is the verbal command "heel" only. There is no gesture of hand or body.
Training Phase
We begin training the heeling skill when the handler no longer has to hold the ball in the left hand to signal/lure the dog to heel position (see 4.3.3.2). Now, instead of the ball, the handler holds the leash with no more than 12-18 inches of slack. After saying "heel," the handler delivers a slight popping correction on the leash. As always during the training, the leash input does not force the dog to finish. The dog is already on its way to heel position, because of habit and its desire for the ball. This technique merely serves to "connect" heeling and the leash input, and bring the finish under the control of the leash input. Similarly, slight pops can also be given as the dog arrives at heel position to encourage it to stop in the correct place, as it sits, and while it is sitting to encourage it to keep its eyes focused up at the handler. All of these movements are thus brought under the control of leash inputs, so that subsequently the dog will not be confused if it receives a correction (see proofing heeling) for failing to complete any of them quickly enough or correctly.
Proofing Phase
Once the finish and heeling, the halt, and attention to the handler have all been trained/‖connected‖ to leash inputs, then we are ready to begin making the "heel" command an obligation for the dog rather than request. This is done by sharpening the leash inputs so that the dog begins to deliberately avoid errors in order to avoid the corrections. Plenty of ball reward is provided during proofing to keep the dog‘s drive up and ensure that it still enjoys heeling work. Most corrections are preceded by the marker "no" (given exactly when the dog goes wrong), to make absolutely sure that the dog associates the correction with what it did wrong rather than some other behavior. The handler must realize that if the "no" is used properly to mark undesirable behaviors, the correction need not be delivered immediately. In fact, in many cases the technique works better if the handler is very deliberate and takes his/her time delivering the correction in a calm and measured manner (as long as the "no" has been delivered at exactly the right moment).
Corrections for attention — to bring the dog‘s focus to its handler — are made by lifting the leash-hand quickly directly toward the handler‘s face. Once the dog comes into proper attention, then the animal receives "good" and reward in position, or simply the "yes." Almost all other corrections are made directly towards the dog‘s tail, jerking backwards with the leash held in the left hand, directly behind the dog‘s head, over its back, with only a few inches of leash held between the hand and the dog‘s collar. In this way the hand hangs in a natural position just behind and outside of the dog‘s head, and any input with the leash serves to stop the dog from coming too far forward past the handler‘s knee. When a dog is taught to heel using the ball (and generating prey/retrieving drive), most of the errors are related to the dog wanting to move too far forward while heeling and come across the handler‘s body, in anticipation of receiving the ball. The leash corrections just described are ideal for counteracting this tendency and encouraging attention, when combined with proper rewards in position as described above (in 4.3.3.5 and 4.3.3.6).
Once left turns are begun, the same correction will help to prevent the dog from bumping or riding against the handler. When right turns are begun, if necessary the left hand holding the leash is brought from behind the dog‘s head across the handler‘s front and up, to generate a correction that brings the dog‘s head towards the handler‘s centerline and up towards his/her face.
During the proofing stage, the handler reduces the frequency of ball reward (substituting praise and petting) and asks the dog to heel for longer and longer periods, and with greater and greater distractions. The eventual goal is tactical heeling capability — meaning that the dog finishes to heel from any distance on verbal command, and remains at heel after one command no matter what the handler‘s pace or direction, off-leash, and under intense distraction (such as gunfire and stimulation from decoys dressed in bite gear cracking whips, etc.).
"No" marker. "Good" and "yes" are not the only markers we can use, nor is the power of marking technique limited to rewards. We can also mark behaviors we want to punish, by using the word "no." "No" is used much like "yes," in the sense that it is used to mark a behavior, and bridge a delay. In this case a delay to punishment. Earlier we covered two kinds of punishing response contingencies, positive punishment (giving the dog something that it dislikes) and negative punishment (or omission, taking away from the dog something that it likes). "No" can serve to signal both of these response contingencies. "No" is also used like "yes" in the sense that it tends to be a terminal cue—when the dog does something that earns a "no," the animal often has to start the whole exercise again and therefore it can break when it hears "no."
We can use the "no" to gently punish the dog (through omission) for overeager or careless mistakes. For instance, if we are working the dog through a sit-down-sit sequence, and the dog does not wait for the sit command but pops up without permission, then the handler gives the "no" and makes the dog go back into the down again and wait for the command before rising up into sit. If the animal rises to the sit correctly, it can be rewarded in position with "good" and "yes," or it can be rewarded with "yes" and allowed to break immediately. Here the "no" gives us the ability to improve performance without killing the enthusiasm of an eager dog.
We can also use the "no" more forcefully as a predictor of physical (positive) punishment. Let us say the handler leaves the dog on a down-stay and steps a few feet away, and the dog breaks position and moves toward the handler without permission. If the handler then simply corrects the dog, the animal will be corrected in the act of approaching the handler, which can be hopelessly confusing for an eager dog. What is needed is a tool to tell the dog exactly what critical behavior earned it the punishment. Therefore, the handler gives the "no" immediately when the dog‘s elbows lift from the ground. Then the handler calmly approaches the dog and administers a correction of appropriate strength for the dog, normally by popping the leash two or three times, then takes the dog back to the exact place where the dog was lying and commands it to lie down. Then the handler steps away and, if the dog holds position correctly, the handler performs a reward in position first "good" to reinforce the act of holding position, then approach and placement of the ball directly in front of the dog‘s nose, and "yes" to release the dog into the ball.
ADVANCED SIT AND DOWN EXERCISES
To meet certification standards, the dog must eventually learn to transition from sit to down and back up to sit again at heel position, and while at EOL. For these exercises, the dog must not only understand sit and down, but also how to move from one position to the other without creeping forward or changing its alignment. This is where "good" and "yes" and "no" cues come into their own, because they give the handler the power to teach the dog to understand the difference between a perfectly correct down (in which the dog does not creep forward) and an incorrect down (in which the dog lies down every bit as fast, but creeps forward as it does so).
For sit and down EOL, "good" is used to let the dog know immediately when it has performed a correct sit or down, and then the handler approaches and delivers reward in position (by putting the ball close to the dog‘s head and releasing the dog into it with "yes"). If the dog creeps as it transitions from sit to down or vice versa, then the handler marks the mistake with "No," and either makes the dog repeat the skill or calmly delivers a correction and then makes the dog repeat the skill.
For sit and down at heel, "good" is used in the same way to mark a correct transition. Reward in position is performed by taking the ball from the pocket and holding it directly in front of the dog‘s nose at heel position, and then releasing the dog into it with "yes." If the dog creeps forward or slews sideways at heel while performing the transition, the handler marks the error with "no," (perhaps followed by a correction) and sends the dog back to correct heel (see 4.3.3.1) in the original posture (sit or down). Then the handler makes the dog repeat the exercise, rewarding the dog in position if it is executed correctly.
For real-world operations/utilization, the down has greater importance than the sit. The down is the dog‘s most stable position. When given the ―down‖ command the dog must drop immediately no matter where it is and what speed it is traveling, and then lie still and silent until given another command or released, even under intense distraction (gunshots, decoys carrying bite equipment and cracking whips, etc.). Accordingly, substantial psychological pressure must often be applied to achieve this level of obedience. In order for such treatment to be fair and effective, corrections used for the down must be thoroughly trained (see 4.2.2.2), and "good," "yes," and "no" must be properly applied so that the dog understands what it is being corrected for and can adjust its behavior to avoid corrections.
HEELING (MARCHING)
Heeling is an attention-based exercise in which the dog walks at its handler‘s left side (personnel who carry weapons on the left side often teach the dog to heel on the right) with the shoulder even with the handler‘s knee, keeping pace and position no matter what the handler‘s pace or direction, and sitting automatically when the handler halts. The primary functions of heeling are to refine the dog‘s obedience to its handler, and to provide the ability to transport the dog under close control through hazardous or distracting circumstances with both hands free. A well-trained dog concentrates completely on its handler while heeling, and heeling is therefore very fatiguing and not an appropriate way to transport the dog long distances. For mere transportation, where all that is important is that the dog remains under control, does not pull against the leash, and does not use any more energy than necessary, a "walk easy" skill is used instead. Walk easy (the command is normally "easy") is much less strict and demanding than heeling.
It is traditional in DoD to use the verbal command "heel," and also slap the left hip with the left hand, and to repeat these verbal and gestural commands at each change of pace or direction. However, if we view heeling as a tactical tool rather than a parade ground skill, then we must realize that, a) slapping the hip is unnecessary and inadvisable, because the left hand should be left free (for weapon-handling, for instance), and b) repeated commands are also unnecessary and inadvisable, because the verbal command "heel" is all that is necessary to tell the dog that it should place itself at heel position and remain there, no matter what the handler‘s movement or direction, until released. Accordingly, in the discussion of heeling that follows, the left-hip slap and repeated commands are not used. The CST method for teaching heeling described below normally results in a dog that positions itself for marching close to the handler's left knee and hip, with the handler's left hand hanging or swinging outside of the dog's head. In effect, the dog positions itself "between" the handler's left hand and hip and looks directly up at the handler's face, rather than positioning itself "outside of" the handler's hand and arm and looking around the handler's elbow at the handler. This close positioning is in most situations advantageous because it gives better and closer control of the dog and results in fewer training problems and less stress for the dog. Weapons retention is not normally an issue for law enforcement applications because the dog works on the handler's non-weapon-carrying side and the handler need not carry the left arm pinched close to the body (a left-handed handler normally trains the dog to heel on the right side and so can carry the right arm loosely while holding the left arm tight to the body for weapons retention).
The CST method of teaching heeling concentrates on teaching the dog to understand the exact position that it must maintain at heel, how to move its body in order to reach that position, and on making sure that the dog is highly motivated for the work. This is the best way to prepare the dog for the physical corrections that may later be necessary to render heeling ―fail safe for real-world tactical scenarios, in which handler safety depends on his/her control of the dog.
Teaching Phase
The finish. The dog first learns to heel not by walking at the handler‘s left side, but instead by learning to ―finish.‖ ―Finish‖ is an expression used by competitive obedience trainers to describe a skill in which the dog moves from position in front of the handler to heel position. In DoD, the dog does not normally walk around behind the handler, passing to the handler‘s right—instead the dog passes by the handler‘s left hand, turns in place and sits at heel (called the "military" finish).
To teach the finish the dog is lured with food or the ball (in the left hand) from position in front, past the handler‘s left side, and behind the handler about 2 or 3 feet (the handler normally takes one long step back with his/her left foot). Then the handler turns the dog in towards himself/herself (the dog turns counter-clockwise), and leads the dog forward into heel position and asks the dog to "sit." When using food, the handler then allows the dog to eat from the left hand while in heel position. When using the ball, the handler holds the ball in the left hand just in front of the dog‘s nose, gives the "yes," and flicks the ball into the dog‘s mouth. Initially the handler leads the dog through the entire path, covering nearly as much ground as the dog does. With time the handler moves less and less, taking a smaller step backward and shortening the circle he/she makes with the left hand and the reward. Eventually the handler does not step backward with the left foot, simply commanding the dog to "heel" and making a long gesture with the left hand to send the dog past the left side and behind, and turn the dog and bring it back up into heel position. The amount of room the dog is given to accomplish the movement is gradually decreased, and when the entire teaching sequence is performed well, the dog begins "flipping" or "swinging" to heel position (rather than walking behind the handler, turning around, and walking up into heel position). A wall or barrier is often used initially, to help the dog reach a straight heel position at the conclusion of the movement. If the dog sits crookedly or too far forward, the handler gives the "no" cue and makes the animal repeat the whole exercise correctly before rewarding.
In the next step the dog‘s attention is shifted from the left hand to the handler‘s armpit. When the dog reaches heel position, the handler marks this behavior with "good," and then carefully moves the left hand and the ball up above the dog‘s head to a position just in front of/under the armpit. Then he/she gives the "yes" and drops the ball into the dog‘s mouth.
After a few repetitions, the handler actually places the ball in his/her armpit prior to rewarding the dog. The movement begins as before. The dog is led through the finish with the ball in the left hand, concluding with the dog at heel position and the ball held directly in front of the dog‘s nose or on the left side of the dog‘s head. The handler then marks the correct completion of the exercise with "good," and then raises the left hand above the dog‘s head, transfers the ball to the right hand, and places the ball in the left armpit, clamping it there with the bicep. Then the handler drops both hands to natural positions, with the left as always hanging just outside of the dog‘s head. The dog should stare straight up towards the ball from heel position. The handler rewards the dog by saying "yes" (the dog will normally release and rear straight up towards the ball) and delivering the ball directly into the dog‘s mouth by unclamping the bicep so that the ball drops free.
Reward in position at heel. Eventually, the handler begins leaving the ball in his/her pocket until the dog correctly finishes. We now expect the dog to complete the finish in order to obtain the ball reward, but without seeing or following the ball reward. The handler makes the same motion with the left hand, and even cups the hand as though he/she is holding a ball, but the ball remains in the right pocket (at this point the leash is normally transferred to the left hand for the first time and held in a manner similar to holding the ball). When the dog reaches correct heel position, the handler marks with "good," and reaches with his/her right hand into the right pocket, withdraws the ball, reaches across the body and places the ball in the left armpit, and then says "yes" and drops the ball. This is a type of reward in position technique (see 4.2.4.1.4.1).
Another technique for reward in position at heel. In a variation on this procedure, the handler does not place the ball in his/her armpit after the "good" cue, but instead places the right hand with the ball directly in front of the dog‘s head, or to the left of the dog‘s head (to bend the dog‘s head away from the handler and straighten the animal‘s spine) and then gives the "yes" release and rewards in position. Once the dog can correctly complete this skill, it has learned a specific position and a specific movement that will become the basis of heeling/marching.
Eventually the gesture of the left hand (the prompt) that is used to send the dog to heel is faded out, so that the dog swings into correct heel position and into attention on the word "heel" alone.
Now we are ready to teach heeling proper, actual marching with the dog at heel position. Initially the handler keeps the ball in the left armpit to give the dog a focal point. Holding a very, very short leash in the left hand (from 3 to 8 inches, but without any tension on the leash between hand and the dog‘s collar), with the hand outside of and just behind the dog‘s head, the handler gives the command "heel" and shuffles very carefully forward a few feet. When the dog moves well, between the handler‘s left hand and hip while looking straight up at the ball held in the handler‘s armpit, then the handler marks this behavior with "good," and comes to a halt carefully so that the dog does not lose position (initial heeling is often performed along a wall or fence). Once the dog is in sit-halt position, then the handler gives the "yes" release and drops the ball into the dog‘s mouth. It is important to remember that in the initial stages of marching, the dog is rewarded after only a few steps, and normally always in the sit halt. That is, when the dog moves well, the handler marks this with "good" and then comes to a halt and performs reward in position.
With further practice the trainer can leave the ball in the right pocket, asking the dog to move at heel while looking up at the handler‘s face or armpit rather than the ball. Once the dog moves well, then the handler marks this behavior with "good," comes to a halt, takes the ball out of the pocket with the right hand and then performs reward in position in one of the two ways described in 4.3.3.5 and 4.3.3.6. Place the ball in the left armpit and drop it to the dog after saying "yes," or transfer it to the left hand and hold it near the dog‘s head while it is sitting in heel position, and then flicking it into the dog‘s mouth after saying "yes." Gradually the dog is taught to heel for longer periods with the ball in sight and without the ball in sight, depending on the circumstances.
Eventually, the heeling pattern is made longer, with turns and halts and changes of pace, and the dog is made to work for extended periods for the "good" and the "yes" and the reward. If, while moving, the dog loses position by running wide or forging forward or swinging out into a crabbing motion, then the handler marks this error with "No," halts, and re-commands the dog to finish to heel. Once the dog is back in correct position, the handler encourages with "good" and resumes heeling again. If the dog this time maintains correct position it is given "good," sit-halt, and reward in position as in 4.3.3.5 and 4.3.3.6.
Up until this point, the "yes" has normally been given only after the "good," as a way of releasing the dog into the ball. This practice has been advantageous because it served to make sure that the dog always got its reward while holding the desired position. Such rewards in place are optimal for making sure that reward anticipation does not interfere with steadiness in the sit and down and correct position while heeling. However, they do not make full use of the power of the well-conditioned "yes," to instantaneously identify to the dog, and reward, very specific aspects of performance. But now we are ready to begin using the full potential of the "yes."
Once the dog shows that it understands how to finish quickly and efficiently to heel position, maintain focus on the handler by looking up towards the handler‘s face while maintaining correct heel position, move at heel without losing position, and correct itself back to the proper position when it happens to lose position, then we are ready for the final step, in which the handler begins rewarding directly out of heel with the "yes" marker. At any moment that the handler judges the dog should be rewarded, either after the finish, while heeling, or after the dog corrects itself back into position in response to the "No," then the handler rewards the dog by saying "yes." The dog will release from heel and show that it expects reward, and the handler can then withdraw the ball from the pocket and give it to the dog.
Use of the "yes" marker in this way enables the handler to bring to the dog‘s attention and selectively reward very fine-grained aspects of performance, such as small differences in speed of movement, angle, or posture. Competitive trainers find this useful because they are interested in polish and speed and precision, because these are the things that win trophies. However, MWD trainers should also be interested in polish and speed and precision, because these are the hallmarks of a dog that fully understands commands and skills: Only when the dog has full understanding is it fair and effective to apply pressure to the dog in order to make sure that it always performs correctly, even under real-world conditions where failure to perform is dangerous for dog and handler and those personnel that depend upon the MWD team.
Note that the "heel" command is given only in order to finish the dog, and when the handler first goes into motion. The handler may re-command "heel" after a "no" also. In finished form the command for finish and for heeling is the verbal command "heel" only. There is no gesture of hand or body.
Training Phase
We begin training the heeling skill when the handler no longer has to hold the ball in the left hand to signal/lure the dog to heel position (see 4.3.3.2). Now, instead of the ball, the handler holds the leash with no more than 12-18 inches of slack. After saying "heel," the handler delivers a slight popping correction on the leash. As always during the training, the leash input does not force the dog to finish. The dog is already on its way to heel position, because of habit and its desire for the ball. This technique merely serves to "connect" heeling and the leash input, and bring the finish under the control of the leash input. Similarly, slight pops can also be given as the dog arrives at heel position to encourage it to stop in the correct place, as it sits, and while it is sitting to encourage it to keep its eyes focused up at the handler. All of these movements are thus brought under the control of leash inputs, so that subsequently the dog will not be confused if it receives a correction (see proofing heeling) for failing to complete any of them quickly enough or correctly.
Proofing Phase
Once the finish and heeling, the halt, and attention to the handler have all been trained/‖connected‖ to leash inputs, then we are ready to begin making the "heel" command an obligation for the dog rather than request. This is done by sharpening the leash inputs so that the dog begins to deliberately avoid errors in order to avoid the corrections. Plenty of ball reward is provided during proofing to keep the dog‘s drive up and ensure that it still enjoys heeling work. Most corrections are preceded by the marker "no" (given exactly when the dog goes wrong), to make absolutely sure that the dog associates the correction with what it did wrong rather than some other behavior. The handler must realize that if the "no" is used properly to mark undesirable behaviors, the correction need not be delivered immediately. In fact, in many cases the technique works better if the handler is very deliberate and takes his/her time delivering the correction in a calm and measured manner (as long as the "no" has been delivered at exactly the right moment).
Corrections for attention — to bring the dog‘s focus to its handler — are made by lifting the leash-hand quickly directly toward the handler‘s face. Once the dog comes into proper attention, then the animal receives "good" and reward in position, or simply the "yes." Almost all other corrections are made directly towards the dog‘s tail, jerking backwards with the leash held in the left hand, directly behind the dog‘s head, over its back, with only a few inches of leash held between the hand and the dog‘s collar. In this way the hand hangs in a natural position just behind and outside of the dog‘s head, and any input with the leash serves to stop the dog from coming too far forward past the handler‘s knee. When a dog is taught to heel using the ball (and generating prey/retrieving drive), most of the errors are related to the dog wanting to move too far forward while heeling and come across the handler‘s body, in anticipation of receiving the ball. The leash corrections just described are ideal for counteracting this tendency and encouraging attention, when combined with proper rewards in position as described above (in 4.3.3.5 and 4.3.3.6).
Once left turns are begun, the same correction will help to prevent the dog from bumping or riding against the handler. When right turns are begun, if necessary the left hand holding the leash is brought from behind the dog‘s head across the handler‘s front and up, to generate a correction that brings the dog‘s head towards the handler‘s centerline and up towards his/her face.
During the proofing stage, the handler reduces the frequency of ball reward (substituting praise and petting) and asks the dog to heel for longer and longer periods, and with greater and greater distractions. The eventual goal is tactical heeling capability — meaning that the dog finishes to heel from any distance on verbal command, and remains at heel after one command no matter what the handler‘s pace or direction, off-leash, and under intense distraction (such as gunfire and stimulation from decoys dressed in bite gear cracking whips, etc.).
Comments
Post a Comment