Dogs Have Different Retrieval Strategies

My Standard Poodle Raja loves to play ball. It is his favorite activity in the world, something that he could do all day if he found someone with the strength and endurance to keep throwing the ball for him. In Raja’s case, the ball is a football-shaped yellow rubber ball that looks like it was constructed out of Swiss cheese – most of the ball consists of air, enclosed by soft rubber struts. As Raja catches the ball, it squishes in his mouth, allowing him to chew it several times as he brings it back to me.

Sometimes I initiate the game by saying, “Where’s your ball? Go find your ball!” Raja then runs around the house looking for the ball, which could have been left anywhere. After a couple of minutes of searching, he triumphantly shows up with it, his tail wagging furiously. So from this, I know that he recognizes the word “Ball.” What I don’t know is how he searches for the ball. Does he look in each room systematically, or does he go to where he thinks the ball might have been left? I confess that I have never paid attention.

Now, however, there is an article in the April 2008 issue of Animal Cognition that suggests that dogs have different strategies for how they search for objects (Juliane Kaminski, Julia Fischer and Josep Call, 2008, Prospective object search in dogs: mixed evidence for knowledge of What and Where, Animal Cognition 11, 367-371, DOI 10.1007/s10071-007-0124-1).

The study involved two Border Collies, Rico and Betty, who each know the names of some 200 objects (see the previous post about Rico in this blog). The dogs’ searching strategies were tested in two experiments.

In one experiment, five or six objects were placed in each of two rooms in an apartment, and each dog was asked to retrieve a specific object, which could be in either room. The dogs were free to go into each room and search, until they found the object that they were asked to get. When they returned, they received either a treat or some playtime. Then they were asked to get the next object in each group, until all of the objects had been retrieved.

The first time the dogs entered each room, they could see all of the objects there. The question was, the next time they were asked to get a particular object, would they go right away into the correct room and bring it, or would they randomly roam through both rooms until they found the requested object?

Both dogs did very well at retrieving the correct objects. Rico brought the right object in 46 out of 48 times, and Betsy brought the right object in 38 out of 40 times.

However, the dogs had different retrieval strategies. Rico tended to go straight away to the room where the requested object was located, while Betsy tended to go first to the room where the object was NOT located.

Because Betsy had a different search behavior, the experimenters wanted to find out more about how she searched, So the second experiment revolved around Betsy. In this experiment, four rooms were used. The dog was in one room, while the experimenters placed a different object into each of two other rooms, while the fourth room remained empty. Then one of the experimenters took Betsy on a leash through each room each time before she was asked to retrieve an object. In the first set of 12 requests, Betsy was taken through the rooms in the same, consistent order, first room 1, then room 2, and then room 3. In the next set of 12 requests, the experimenter took Betsy through the rooms in a random order, with the sequence of rooms changing each time. In all of her walks through the rooms on a leash, Betsy could see each of the objects that had been placed there, one of which she would be requested to bring after her walk.

As before, Betsy tended to start her search in a room where the requested object was not located. However, the experimenters observed that Betsy had a preferred search pattern. She tended to start looking each time in one particular room (for example, the kitchen) and then moved on to another room (for example, the hallway), rather than going straight to the room that had the requested object.

These experiments show that the two Border Collies had two different search strategies. Rico went directly to the object. Betsy was much more systematic, starting her search in a particular place and systematically searching room after room until she found the object.

Why did Betsy start looking in a room that did not have the object? We can speculate that if the object were prey, it would be unlikely to stay in one place, and a systematic search would be more efficient than going straight to where the object had been when last seen.

Alternatively, we can speculate that Rico is much more bold and cut-to-the-chase, while Betsy is more ploddingly systematic.

---Con Slobodchikoff

Comments

[magnoliasouth]

Do you think that Betsy (also named Betty in the article, if I understand this right) was taught differently and this is the result of her teaching?

For example, during training the owner may have moved objects from original locations before sending her to retrieve.

Another example could also be that the owner taught her by moving with her from room to room while looking for objects.

Just a thought.

[David Rentz]

Hi Con

Most interesting. We have neighbours with problem dogs and they won't do anything to constrain them. They follow none of the leash laws at all. They have chased our Cassowaries away. They have had hundreds of dollars worth of fines but they persist in letting their Great Dane and Dalmatian run wild in the neighbourhood. Awful.

Dave Rentz