Just Don't Call Them Furbabies

More exciting research with dogs was released last week. This new study provides more evidence that dogs tend to follow human cues very closely. I wrote about an earlier study and the always fascinating Belayaev Foxes over here and here

The researchers performed a test with dogs and human-raised wolves and 10 month old toddlers. They first performed the test with dogs and toddlers. As has been repeatedly reported in the news, the dogs frequently made the same "mistake" as the toddlers: relying too heavily on nonverbal cues to figure out where an object is hidden. This suggests that dogs reading human "body language" similar to the way human children do. 

But the study gets really interesting when the researchers discuss how they performed the tests again, this time with different dogs and some "extensively socialized, hand-reared wolves." I found the differences between the dogs and the wolves much more interesting than the similarities with the children. All of the dogs by the way, were recruited from local families and were family dogs, not dogs that were specially trained or obtained from a breeding program.

The tests consisted of hiding food behind one of 2 screens:

Subjects witnessed the hiding events in one of the three conditions: Social-Communicative, Non-Communicative, and Non-Social contexts. The task had the same structure in all experimental conditions: Hiding the target object four times behind the screen A (A-trials) and then hiding the object three times behind the other screen (B-trials).
(Differential Sensitivity to Human Communication in Dogs, Wolves, and Human Infants Jozsef Topal, Gyorgy Gergely, Agnes Erdohegyi, Gergely Csibra, Adam Miklosi. Published 4 September 2009, Science 325, 1269 (2009) DOI: 10.1126/science.1176960)

In the Social-Communicative context, the person hiding the treat addressed the subject by name and told her to "watch". After getting eye contact, the person then either hid the object behind one screen or the other. The procedure (which I can't completely reproduce here) describes pretty clear body language, such as using the food to get the subject's focus and moving the object from screen A to B in the B trials.

In the Non-Communicative context, the person hiding the treat provides no (deliberate) communicative signals. No name, no eye contact, no attention with the food.

In the Non-Social context, the food is moved with a transparent line and there is no human to read (or not) involved at all.

I'm leaving out quite a bit. The study and the excellent supplemental materials explain how the dogs were chosen, the efforts made to make sure the dogs and the wolves were as comfortable as possible (the wolves were tested outdoors, for example) and even include a few downloadable videos of the people hiding the toys behind the screens to show how cues were used or eliminated.

The dogs frequently made "perseverative" errors in the social-communicative context. Similar to the toddlers, they expected the food to be hidden behind screen A based on reading the human's cues. The wolves made this error much less frequently, even though they had been raised by humans. In the other 2 contexts the error rates balance out a bit. The dogs seem to have done a little better than the wolves in the Non-Communicative context.

Did you notice how I bolded eye contact? Yeah, I've got a thing about that. See how important it was in how the dogs responded?

Dogs are not wolves and wolves are not dogs. Yes, they are very close genetically, but even when wolves are raised by humans their behavior is fundamentally different from dogs. This behavioral difference is the difference between wolves and dogs. Why are we trying to apply how wolves act in packs to dogs? (I won't even get into how the description of how wolves act in packs is usually wrong too.)

Last, I'll quote from the study again:

Unlike great apes, dogs exhibit some understanding of human referential intentions expressed in communicative gestures, such as pointing, as shown by their success in solving the so-called object choice tasks.

This isn't new, there have been other studies comparing the dog's ability to interpret pointing to both apes and children. But it's been demonstrated many times now. Dogs are incredibly unique in the way that they relate to us. Enjoy it.