Border(collie)-line Intelligence
Researchers claim to have discovered the “g” factor in dogs
A few weeks ago I came across an article about a group of researchers claiming to have discovered a general intelligence, g factor, in dogs (link: http://www.iflscience.com/plants-and-animals/scientists-create-dog-iq-test-help-understand-human-intelligence), and I thought to myself something along the lines of “Here we go again.” The study undertaken by a group of intelligence researchers in the UK used two measures, “point-following” (follow where experimenter is pointing) and “quantity-discrimination” (which bowl has more food), to measure the dog’s intelligence. They found that dogs who scored high on one test tended to score high on the other tests as well, leading them to conclude that there exists an underlying “general” intelligence factor in dogs. They extrapolated their findings to say that canine intelligence follows a similar structure to human intelligence, ergo we can use dogs IQ as a model for humans. They say the canine model of intelligence is advantageous because it lacks confounding ecological factors that one finds in studying humans, like the effects of socio-economic status and education.
Of course, drawing an analogy between animals and humans is problematic. Although canines and humans are both mammalian and share distant evolutionary forbears, it is irresponsible to claim that human intelligence can be studied using a dog analogy. These dogs were tested according to tasks that are an inherent part of their genetic makeup — dogs were selected and groomed thousands of years ago to follow pointing commands and assist in hunting, so it is only natural that they should excel in such tasks. On the other hand, human intelligence is multi-dimensional and goes far beyond these simple application type tests. Also, to argue that ecological factors do not factor into canine intelligence is misguiding. Of course some dogs are raised in better households than others and exposed to more love, training, and better feeding and exercise. Now that is a parallel we can draw between man and pet.
Certainly to claim that dog intelligence can provide an exemplary model of human intelligence is either flat out invalid or far oversimplified. As Dr. Franz de Waal argued at the Hebb Lecture on Friday March 18th, intelligence testing needs to be species relevant! Dr. de Waal studies chimpanzee learning and intelligence and argues that chimpanzees are incredibly intelligent creatures, but their intelligence is uniquely chimp. In the same way, dog intelligence tests can only be generalized to other dogs, not human beings! Dr. de Waal also pointed out that intelligence testing in animals needs to be unaffected by experimenter intervention. He referred to this as the “Clever Hans” effect after a horse that was able to do complex arithmetic…but only when his owner was around to signal him to stop counting. Therefore, humans need to be aware of the effect we have on animal behaviour. This strikes me as a particularly poignant statement when you consider that one of the measurements used in this study was a pointing exercise where the experimenter was having a direct effect on the dog’s behaviour. Who knows what the experimenters body language and facial expression was communicating to the dog at the time?
Furthermore, Dr. de Waal warned against judging animals on their behaviour relevant to human intervention, but instead should be measuring their intelligence relevant to their fellow species mates. Perhaps this has less to do with dogs as they are ecologically nested in a human world, but for chimps, learning from human behaviour is almost irrelevant to them but the ability to mimic and learn from other apes is crucial.
Overall, Dr. de Waal’s fantastic lecture really brought forward the question of whether we can truly measure “intelligence” in other animals, and whether the results of our tests are meaningful (valid and reliable). To paraphrase the insightful words of Franz de Waal, “How are we supposed to know or measure what animals are thinking, when I’m not even sure what you are thinking. And even if you tell me what you’re thinking, I’m not sure if I can ever truly understand it because how do I know that what I feel reflects what you feel?” The wise words of Dr. de Waal should make many researchers stop and reconsider the clinical, sterile methods we use to test intelligence in both humans and animals today.
References:
Arden, R., Adams, M. J., & Arden, R. (March 01, 2016). A general intelligence factor in dogs. Intelligence, 55, 79–85.
