courteous canines logo
Kate Connick
Dog News
My Dogs
Contact Us
Useful Links
Site Map


Book Review
The Intelligence of Dogs by Stanley Coren
Review by: Kate Connick, May 2003

Stanley Coren is a frustrating author. On the one hand, he comes across as a well-informed, erudite individual with an approachable, avuncular charm. There's no doubt that he can engage and entertain the reader in the spirit of any great storyteller. On the other hand, this is something of his Achilles heel, as he is inclined to forfeit focus and relevance as he strives to amuse. Ironically, I don't think this bothers the readers who enjoy him most.

The Intelligence of Dogs is a book that ostensibly examines whether dogs have intelligence at all, how that intelligence might be conceptualized, the way in which it differs among breeds and individuals, and the ramifications for the dog owner.

Coren meanders for six chapters before actually presenting his model of dog intelligence. Mind you, the meandering is amusing and informative in its own way, as he explores the topics of canine evolution, communication, and public image, relating them tangentially to the topic at hand.

Still, he invests too much time on these peripheral topics, at the expense of his central thesis. Over three pages alone are devoted to listing all of the commands his own dogs know. It's cute but hardly necessary. We all know that dogs can learn a variety of commands. He gets sidetracked telling stories and relating trivia.

Similarly, I found his chapter on evolution to be puzzlingly misplaced. To assert that all canines possess a shared form of intelligence that helps them function as social predators and scavengers is one thing. To belabor a discussion of the origin of dog is another, unless one ties it directly to the evolution of canine intelligence proper - which Coren does not do.

On the other hand, Coren raises issues that he never fully explores, like the idea of neoteny. If some breeds are more juvenile than others, what are the ramifications for intelligence? Can one predict intelligence based on a dog's appearance, if the two are interrelated? He mentions how neoteny is reflected in behavior by rendering dogs more playful and less fearful, for example, but he never dives in and attempts to discuss how it might relate to intelligence, per se. He doesn't even attempt to field the most obvious question: Are dogs as intelligent as their wild ancestors or counterparts?

Coren does make the point that the estimation of dog intelligence is molded by popular opinion and varies with the zeitgeist. Whether influenced by philosophical and religious leaders or by media stars (like Lassie), our acceptance and perception of dog intelligence has fluctuated throughout history. Little of what we know seems based on real, hard science. Ironically, while I believe that Coren fancies himself as a scientist of sorts, much of his own "proof" of dog intelligence is subjective and speculative, based on personal anecdotes or tales told to him by others. His presentation is thus not especially cogent.

I found it awkward that Coren's initial attempt to conceptualize dog intelligence discusses it vis-a-vis Howard Gardner's model of human intelligence. It strikes me - especially when we begin an exploration of the topic by examining evolution - that intelligence relates more to one's mental facility within one's own niche. Applying a human model of intelligence to the dog thus seems misguided at its very core.

Dogs, for example, don't appear to have intelligence in the musical arena. Um, why would or why should they? What biological advantage would that afford a dog? It seems to me that the ability to use one's mental faculties to locate and dispatch prey, secure breeding opportunities, and ensure personal safety are better barometers of animal intelligence than their ability to appreciate harmony. Perhaps I'm missing something, but if intelligence isn't functional, then I don't see the point in discussing it at all. If it doesn't have any practical value, who cares?

A full 117 pages into the book, Coren finally gets around to proposing a model of canine intelligence (which does appear to relate essentially to the animal's ability to function for itself and to be functional in service to mankind). He conceptualizes manifest (total measurable) intelligence as a combination of three separate and often conflicting factors:

  • Adaptive intelligence: learning rate, memory, problem-solving ability
  • Instinctive intelligence: genetically-dictated areas of behavioral specialization
  • Working/obedience intelligence: trainability for traditional obedience exercises

The remainder of the book looks at these three forms of intelligence. I do think that Coren would have done better to severely edit out much of the first half of the book and explore his model of manifest intelligence with greater depth and consistency. As much as I'd like to buy into his paradigm, I do have problems with it.

Coren makes the mistake of often equating intelligence only with obedience intelligence. This results in him making remarks that are highly inconsistent and thereby undermining his content. For example, he starts the book by presenting Dandie Dinmont terriers as the epitome of stupidity, given their fearlessness in confronting potentially dangerous game. He concludes that, combined with the breed's dearth of obedience titles, this is reflective of "an intellectual flaw." Yet, according to his own model, these very animals have a high degree of instinctive intelligence - an inherent drive to eradicate vermin. They are specialized at what they do, and at that they excel. If the very qualities of instinctive intelligence that are so highly prized in the breed conflict with the development of obedience intelligence, it is unclear why the breed should be considered intellectually impaired. Clearly Coren favors obedience intelligence, but his own model suggests that there's more to intelligence proper than just following commands.

This, of course, begs the question of whether instinctive intelligence and obedience intelligence are really forms of "intelligence" at all. I'm inclined to believe that obedience intelligence is more a matter of personality traits, something that Coren essentially admits. Instincts, on the other hand, are drives which don't require much in the way of conscious thought to express - something that Coren does not adequately address. I don't believe that either of those categories of "intelligence" is intelligence as we think of it. Neither requires the brainwork that adaptive intelligence does. It strikes me that genuine intelligence is adaptive intelligence and nothing more. The other qualities might be drives, aptitudes, or personality traits, but they don't appear to reflect thinking in the conventional sense.

For example, Coren himself quotes an expert who asserts that a certain amount of stupidity is an asset in an obedience dog, as the dog won't vary the routine nor get bored. Coren seems to concur, suggesting that what I would consider truly intelligent dogs (dogs with high adaptive intelligence) often make poor obedience prospects. Yet he persists in using the term "intelligent" interchangeably with obedience intelligence itself. The contradiction is frustrating.

His ordered list of intelligent breeds - the one that ruffled so many feathers among dog people - is based on obedience intelligence and characterized by his trademark junk science. He surveyed 200 obedience judges about what breeds were most trainable, and his list is the result of the survey. It merely formalizes the collective, subjective experiences of AKC obedience judges. It doesn't mean anything more than that.

At least Coren attempted to devise an IQ test for adaptive intelligence. I'm inclined to doubt that his Cosmo-style test measures what he intends. The test favors energetic, food-focused dogs. It's an attempt to quantify his assumptions, but it seems rather goofy if one really looks at it. I would have preferred discussion of potentially objective ways of determining learning rate, retention, and problem solving.

One of his tests purports to measure social learning. You stare directly at your dog's face while he's about eight feet away. Then you smile. The preferred response is for the dog to come to the owner with its tail wagging. Why? What does this have to do with social learning? It appears to favor the cloyingly affectionate, needy, emotionally dependent breeds. I don't understand why he thinks this has anything to do with social learning. More importantly, he doesn't explain why it does.

This is further complicated by his language comprehension test. Again, the dog is eight or so feet away, and in a tone that you ordinarily use to call your dog, you call out a meaningless word, another meaningless word, and then if the dog is still there, call it to come proper. The ideal response is for the dog to ignore the meaningless words but come on a correct recall command. But I'm puzzled. Surely the dog that is best at social learning - if it comes when you stare and smile - will come just as readily when you call out nonsense words in a beckoning tone. Wouldn't it have made more sense to use a sit or down as the test command, rather than a recall, to avoid muddying the distinction between social learning and language comprehension factors?

Another test that is intended to measure problem solving involves tossing a towel over a dog's head to see how long it takes for the dog to remove it. It seems to me that one factor is missing. Motivation. A relatively passive dog might simply accept having a towel placed on its head unless it was focused on something else. My guess is that a dog that is staring at a squirrel might frantically remove a towel more rapidly than a dog that is attempting to compliantly humor its owner when nothing else is going on in the environment.

My point in all of this is that what you think you're measuring on a test isn't necessarily what you're actually measuring. Aside from Coren telling us the tests are supposed to measure this or that, the reader has no reason to believe that the tests actually measure what they pretend to. It is incumbent upon a test-creator to convince others that the test measures what it is supposed to, and Coren simply doesn't do that. Saying that, "All the tests are based on formal laboratory and field testing procedures," simply isn't enough.

Coren drops the ball altogether on instinctive intelligence. At least he attempts to quantify obedience intelligence and adaptive intelligence, but he entirely avoids any attempt to measure instinctive intelligence. I found this peculiar. He includes a chapter that tells us that dogs were selectively bred to perform different tasks, but being able to complete a task isn't necessarily the same as having an instinctive drive to do it. He mentions, for example, dogs that haul carts like Bernese Mountain dogs. There isn't a specific instinct that makes a dog ideal for that task; it's related to size and, perhaps, stable temperament. He mentions entertainment dogs, including dog actors. Again, there is no instinctive specialization that gives a dog a knack for that. It's not the same thing as an inherent need to collect or drive livestock, chase and kill moving things, or find game.

The really odd thing is that measures of instinctive intelligence do exist. All Coren would have had to do is share an explanation of drive theory and the Volhards' test which attempts to measure prey, pack, and defensive drives. Yet he didn't even bother. I found that puzzling.

In the end, I'll give Coren credit for trying, but that's not enough. I don't think he's a stupid guy, and I believe that he could have done much better with this. I'm not convinced that the reader ends up with an improved understanding or appreciation of dog intelligence after reading this book. Moreover, Coren himself doesn't come across as convincing. He leaves too many things unexplored and contradicts himself too much to present a cogent model of dog intelligence. And yet, the readers who enjoy his personality and chatty, anecdotal, informal, quasi-scientific approach probably won't care.

The Intelligence of Dogs, by Stanley Coren.
Published by The Free Press, 1994. ISBN: 0029066832

Buy this book now at!

Buy this book now from Barnes & Noble!


Back to Bookshelf

Home | About | Services | Kate Connick | FAQs | Articles | Store | News | My Dogs
Contact | E-cards | Links | Awards | Webrings | SITE MAP
©2003 Kate Connick