How to Train Scent Discrimination for Obedience Competition

Scent discrimination is one of the "advanced" obedience exercises, but it is actually one of the easiest to train. Why? You are working with the dog's single most acute sense: smell.

It's important to remember that in teaching scent discrimination we are not teaching the dog to scent; that comes naturally. We are teaching two things:

• discrimination—choosing one scent over another
• what to do with the discrimination

The discrimination

Although the word is politically charged, in behavioral terms "discrimination" simply means a choice made on the basis of established criteria. In the case of scent articles, that criterion is the handler's scent ("find the one that smells like me").

In fact, human trainers shed hundreds of thousands of skin cells every day, each of which imparts our scent. So in handling or even having articles with us, we impart scent to them. Therefore, the discrimination is, more accurately, "find the one that smells THE MOST like me." When I refer to the "scented" or "hot" article, I am talking about the article that carries the FRESHEST scent. I do not mean to suggest that the other articles carry no scent at all.

Establishing the concept

It is helpful in establishing a single discrimination to first teach the dog about discriminations more generally so that the dog gets the idea that "choice" is an issue and the "right choice" is rewarded.

Before working on scent discrimination, it can be beneficial to train the dog to discriminate among objects based on such criteria as size, shape, or color. For example, work with a cube, an orb, and a rectangle. Shape the dog to touch the cube to the exclusion of the other two. Or, work with a red ball, a yellow ball, and a white ball. Shape the dog to touch the red ball to the exclusion of the other two.

You can extend the lesson with a "stimulus reversal"—after shaping the dog to touch the red ball, shape the dog to touch the white ball and not the red.

These exercises are good for the dog because they put his mind to work learning to make choices among objects generally. They are good for the trainer because they improve observation and timing skills.

Introducing scent

There are a number of ways to introduce your dog to the concept of making a choice based on scent. Remember that this is something the dog tends to do anyway. You can start directly with your own scent, or you can use something more potent and obvious, such as cheese or vanilla extract. The purpose of the more potent starter is not to teach your dog to use its nose, but to give you the ability to perceive more accurately what your dog is doing, and to respond to it.

Take an object, any object, and impart the scent to it. If you are using vanilla extract, put the extract on a cotton swab and wipe the swab on the article. If you are using cheese, rub the cheese directly on the article, trying to impart scent without leaving any actual cheese on the article. Of course you do not need to use any "trick scents," and instead can go directly to using your own scent if that is your preference.

Establishing the scent articles

There is no requirement under AKC regulations that you use dumbbell-type articles when training scent discrimination. The only requirement is that there be articles of two types: metal and leather. For the purpose of this article and for reasons that I will discuss below, I do assume that you use dumbbell-type articles.

If you have not shaped a retrieve at all, start with the metal scent discrimination article. Metal is typically the most difficult material for a dog; by starting with metal everything else becomes easier. If you start with the easiest material and work up to metal, you will be working against yourself.

Note that this article does not address shaping the retrieve beyond stating that you should have the retrieve fully shaped before you introduce scent discrimination to the retrieve articles.

Error-free learning

In this first approach to shaping discrimination, you set it up so that every choice the dog makes is correct. Scent six identical objects with the same scent. Reinforce for every touch of every scented object. Do this for three sessions, ten repetitions (trials) a session.

Then, remove one scented object and replace it with an unscented one. Clicker train for every correct touch. If the dog looks toward the unscented object, do not respond, but click for any responses that focus away from the unscented object. Work at this level until the dog is touching only the scented objects and ignoring the unscented one completely.

Next, remove a second scented object and replace it with a second unscented one. The pace should go more quickly, as you have already taught the dog that he is to ignore the unscented article.

Proceed until you have one scented article and five unscented ones and the dog consistently touches only the scented article, ignoring the others completely.

As you go along, mix up the placement of the articles so that the dog does not get the idea that placement is part of the discrimination "set."

Match to sample

Another way to shape discrimination is to give the dog the sample scent and then ask him to find the object that matches the sample. This is where vanilla extract comes in very handy, as those samples are pungent, making the match relatively easy. If you use a white object, the vanilla leaves a slightly visible stain, making it easier for you to recognize that the dog is going to the correct object.

Start with one scented object to make the connection between sample and match, then add one unscented object to introduce the discrimination.

Scenting the article (your scent)

The mere act of carrying articles in a bag from one place to another will cause your scent to be imparted to the articles you use for training. Again, what you are really asking the dog to do is find the "hottest" article, the one that has the freshest or most intense scent.

There are various ways to make sure that there is a clear difference between the "hot" and "cold" articles when you practice with your dog. You can put the "unscented" articles in a refrigerator before practice; the cold will essentially neutralize scent and will retard incidental scenting. You can put the "hot" article in a Ziploc bag, which will intensify your scent.

In actually scenting the article, it is sufficient simply to grasp the bit in your fist and hold it for a couple of seconds. There is no need to spin the bit in your hand.

"Practice the discrimination on various surfaces, including grass, dirt, and concrete. You can never know for sure that your dog will work to the same level on all surfaces unless you practice on all of them."

Click for responses

As you add unscented articles, you may see your dog responding uncertainly. When this happens, you have to think in terms of having a conversation with your dog. Your dog is asking, "What do you want here?"

If your dog even looks in the direction of the scented article, or rolls his eyes in that direction, or moves his head slightly in that direction, click. He's asking a question and you should answer with the click ("click for action").

Confirm your answer by treating in the direction of the scented article ("treat for position"). You don't have to lure the dog to the article, just treat in that direction to keep the dog oriented that way. In most cases, the dog will "take the hint" and continue to work toward the scented article.

Nose to ground

The dog is not required to work the pile with its nose to the ground, but it is preferred. A dog can successfully "air scent" the pile, but if you have ever had a dog do that, you know that it is a rather exciting time for the handler. When introducing the scent, click for touches and work the articles to the ground as quickly as possible.

Extending the work

Assuming you have placed the articles close together and the dog is indicating to the correctly scented article pretty regularly, it is time to mix it up. Spread the articles and then hide the scented one under a chair or in some tall grass, or put the scented article under an unscented one. You are trying to build the dog's tenacity.

It is a good idea to have a training partner pick up the articles and place them on the ground. In a competition, a judge would use tongs, but tongs are not required 100% of the time. If the article should fall to the ground, the judge or steward might have to pick it up by hand. Therefore, make sure that your dog will not only distinguish between your "hot" and "cold" scents, but will also distinguish your scent from that of other people.

Finally, practice the discrimination on various surfaces, including grass, dirt, and concrete. You can never know for sure that your dog will work to the same level on all surfaces unless you practice on all of them.

Introducing the retrieve element (adduction)

As noted earlier, you should have trained the retrieve (pick up, carry, hold, and deliver) separately and it should be fully established before you look to put it together with the scent discrimination.

Let's return to my earlier statement that you should use dumbbell-type scent discrimination articles. The reason is this: your dog is now conditioned to retrieve a dumbbell from the ground. In most cases, your dog will figure out for himself that he is supposed to pick up the scent discrimination article, if only because it looks pretty much like the retrieve dumbbell you worked with before.

In other words, in many cases you won't even have to tell your dog to "get it"; he will "blend" the retrieve into the scent discrimination for himself (an example of "adduction").

If your dog doesn't "blend" the discrimination and the retrieve for himself, then your job is still pretty simple. Just tell your dog "get it" when he is at the article with his nose down. Assuming that you have shaped and generalized the retrieve, this step should create no difficulty at all.

Completing the exercise (the recall)

From there, just as with the retrieve, the exercise finishes with the recall (this is why the AKC guidelines for judges refer to the recall as a "foundation exercise"). The recall, front, and finish should pose no new issues, though there is one variant in that the dog will have to do the exercise twice in the ring, once for metal and once for leather. How you keep your dog working after the first article is for you to decide.

In my own training and showing I used two finishes: "swing" (to the left) and "around" (to the right). In practice, if I sent the dog to the right he knew that he was going to do another repetition. When I sent him to the left, he knew that the session was over. This readily translated to the ring: I sent the dog to the right after the first article and to the left after the second. In this way I maintained a steady working attitude throughout the entire exercise.

Conclusion

Scent discrimination is not difficult for the dog; it is harder for humans because we cannot readily perceive what the dog is doing. Therefore, we tend to "trick it up" more than we need to. But if there is any working situation in which you can trust your dog to be able to figure it out, scent discrimination is it. Your job is simply to provide the necessary leadership so that your dog knows what to do once he has identified the correct article. Otherwise, trust your dog's nose and stay out of the way as much as possible.

About the author

Morgan Spector is the author of Clicker Training for Obedience and a member of the ClickerExpo Faculty.