Dogs Have an Amazing Sense of Smell, But Do They Understand Our Words?
Does Your Dog Really Know What "Sit!" Means?
“Most scientists who talk about dogs have their scientist hat and their dumb hat,” says Gustavo Aguirre, a veterinary ophthalmologist ... “And whenever they start talking about dogs, they put on their dumb hat. They say things that as scientists they have to know can’t possibly be right.” (“The Truth About Dogs,” Stephen Budiansky, The Atlantic Monthly, July, 1999.)
Bacon Strips v. Beggin’ Strips
One of the most common myths about dogs is that they have linguistic abilities that, in some ways, make them “smarter” than three-year olds.
In his 1993 book The Intelligence of Dogs Stanley Coren writes: “When a child responds correctly to the request ‘Give me your hand,’ we grant it some linguistic ability; obviously then, a dog’s appropriate response to ‘Give me a paw,’ represents equivalent language ability.”
Not necessarily. For one thing, if you’ve trained a dog to “give paw” by putting your hand toward the dog’s front paws while she’s seated and then giving her the command, at some point all you need to do is put your hand out and the dog will give paw, with or without the vocal cue. She’ll also give you her paw if you say, “Give me bah,” or any other one-syllable word that ends in “ah.” This indicates that context and previous learning experiences are what enable dogs to respond to the command, not an understanding of the words involved. On the other hand, even newborns can detect the subtle differences between the sounds “pah” and “bah” or “dah” and “gah” while dogs tend to respond to all four sounds equally.
Meanwhile, if you held your hand out to a child and said, “Give me your gand,” she would either giggle or argue with you: “No, not gand! Hand!” So the child is not only able to discriminate between very subtle differences in sound, she’s capable of understanding the meanings of words. And most importantly, she can respond verbally. A dog can’t do either of those things.
Another difference is that the child also knows what a hand is. She could point out, “My hand, your hand,” or point to a hand in a picture book or maybe even look up at the image of hand on a Walk/Don’tWalk sign and say, “Look! Hand!”
You could also ask the child to go touch the dog’s “paws,” which she would be able to do quite easily (if the dog lets her). But if you asked the dog to go touch the girl’s “feet” he would have no idea what you were talking about. Even if the dog had previously been trained to touch someone’s feet with his paw or his nose, he would only be exhibiting a learned behavior without any sort of real understanding of what the words you’re saying actually mean.
Another thing to consider is that at some point the child would also be able to make sense of the phrase, “Will you come give me a hand with these dishes?” This shows an understanding of a word used in a different context.
In her book, Inside of a Dog, Alexandra Horowitz writes, “One component in understanding a word is the ability to distinguish it from other words. Try asking your dog on one morning to go for a walk; on the next, ask if your dog wants to snow forty locks in the same voice. If everything remains the same, you’ll probably get the same, affirmative answer.” In other words, the dog would respond happily to the idea of snowing forty locks.
Meanwhile a child would, at some point, be able to distinguish between homophones like do and dew, pause and paws, and prince and prints, etc., while a dog wouldn't. He’d be like the dog in the TV ads who can’t distinguish between “beggin’ strips” and “bacon strips.”
The final problem with Coren’s idea is that he seems to be basing the comparative linguistic abilities of toddlers and dogs on old research pertaining to how and when children learn to use and understand certain aspects of language.
When Do Children Start Developing Linguistic Abilities?
Children start developing some linguistic abilities while still in the womb. For instance, a fetus in a multi-lingual home is able to recognize the differences in rhythms (prosody) between the two languages spoken by her parents or other relatives. Children also begin learning and memorizing vowels in their native language while still in the womb.
During the first two months, infant vocalizations are mainly expressions of discomfort (crying and fussing). From 2 - 4 months, they begin making “comfort sounds,” typically in response to pleasurable interactions with their parents or nannies. The earliest comfort sounds may be grunts or sighs, with later versions being more vowel-like “coos.”
At 4 - 7 months, babies typically engage in “vocal play,” and shortly after that, they begin to babble. No other animal does anything like babbling.
At about ten months, infants start to utter recognizable words that, for the most part, involve naming things. For instance, a child will say “duck” while he hits a toy duck off the edge of the bath, or “papa” when she hears the front door opening.
There is also a spurt of vocabulary acquisition during the second year, which is when word combinations also begin to appear. Novel combinations appear sporadically as early as 14 months. By 25 months, almost all children are able to combine words. Remember, these kids are actually speaking one or two words in combination with others, not just responding to auditory cues, which is what dogs are actually capable of doing.
So while most dogs may know more about the way a household works—at least from their perspective—when it comes to linguistic skills dogs aren’t remotely smarter than toddlers.
Responding to Auditory Cues vs. Linguistic Ability
It may seem like fine point, but there is actually a vast difference between understanding the meanings of words, and having the ability to respond correctly to spoken cues. Two border collies, Rico and Chaser, know the names of all their toys. (Rico had over 200, Chaser has over 1000.) Chaser is also able to distinguish between commands related to picking up the toy and putting it somewhere else.
Surely this means that the dogs must have some linguistic ability? Right?
Wrong. According to linguists there is no such thing as “some linguistic ability.” You either have the whole package, or you don’t.
Humans are easily fooled by dogs. Even some of the smartest people believe their dogs respond to the “meanings” of words and phrases, when they don’t.
In a 2016 article written by Julie Hecht for The BARk magazine, dog trainer Patricia McConnell provides insight into how we think our dogs know what we’re saying, but our words may have a completely different “meaning” to the dog.
McConnell initially thought [that her dog] Willie knew the name of her partner, Jim. 'To teach Willie, I would say, ‘Where’s Jim?’ and Jim would call Willie over. When Willie consistently went to Jim, I’d say it as Jim was driving up, and Willie would run to the window. One day, Jim was sitting on the couch, and I said, ‘Where’s Jim?’ and Willie ran to the window, all excited. This difference in definitions is more common than people realize — dogs don’t have the exact same concept of words that we do.
I would argue that dogs don’t have any concept of words at all. None.
Note that Ricoh and Chaser exhibited an ability to differentiate between what are essentially prey objects, they didn’t develop a “vocabulary” for electronics, furniture, or even food items.
One might argue, “My dog knows what I mean when I say get off the bed or get off the couch.” And there’s an easy way to determine if that’s true. When your dog is on the bed, simply point to the floor and say, “Off the couch!” The chances are pretty good that your dog will get off the bed even though you’ve told him to get off the couch! The words are irrelevant to the dog.
Some owners might say, “My dog knows what sit means! He has to have some linguistic ability!” But Ian Dunbar points out something most dog owners already know but may not pay attention to when discussing these issues, which is that quite often if a dog is already sitting, and you tell him to sit, he’ll lie down. Again, the dog is not responding to your words, but patterns of behavior.
Birds Do It, Bees Do It—Even Educated Fleas Do It
It seems to me that some of the confusion about the nature of language comes from the fact that there’s supposedly a language of birds, a language of bees, along with the fact that humans and animals both exhibit body language, and all sorts of other so-called types of language that don’t use words, syntax or grammar, it can be easy for the layman or even a scientist to be fooled into thinking dogs have at least a rudimentary ability to understand words and phrases. (Dr. Coren believes that canine barks, yips, and growls exhibit a rudimentary grammar.) 
Meanwhile, a linguist—someone who actually studies the properties and evolution of language—would tell you that there is no such thing as a rudimentary linguistic ability. It comes as a package. You either have the whole kit and kaboodle, or you don’t. If you don’t—if you can learn the names of all your toys but don’t exhibit any kind of linguistic understanding beyond that—then it’s not really language. It’s pretty amazing, but it's not language.
“Life Is an Adventure—Where Will Your Dog Take You?”
1) Dr. Coren has an interesting, though I think seriously flawed, opinion about a dog’s yips, barks, and growls. He thinks there’s a kind of canine grammar at work in how dogs use different combinations of vocalizations. For example, Coren says that a growl followed by a bark means the dog wants to engage you in play, whereas a bark followed by a growl means he wants you to back off or leave him alone.
He then makes the point that while a yip and a bark can be combined and that a bark and a growl can also be strung together, a yip and a growl can’t. A yip is submissive and a growl is aggressive. This proves, he says, that dogs use something like syntax when they ‘speak’ because they seem able to combine different vocalizations to give different shades of meaning.
There is a simpler explanation. The reason that certain types of vocal sounds can be combined is simply that they arise from similar emotions, so that the sounds a dog makes when feeling them are not only vocally, but emotionally compatible.
For instance, fear (“I’m scared!”) and feelings of social resistance (“Keep away!”) can coexist quite easily, which is why a yip and a bark can be combined (“I’m scared so keep away!”). Social resistance (“Keep away!”) and aggression (“I’ll hurt you!”) are also compatible, which is why a bark and a growl can be put together (“Keep away or I’ll hurt you!”).
However, it’s rare for a dog to have submissive feelings (“Please don’t hurt me!”) while experiencing aggressive ones (“I’ll hurt you!”). This is why you almost never hear a yip and growl together. It isn’t syntax or grammar; it’s emotion, pure and simple.
I think it’s important to note that before Helen Keller learned sign language, she emitted all kinds of non-verbal sounds; grunts, groans, etc., none of which had any meaning for her. Being deaf, she couldn’t possibly have known she was making them. Being blind she couldn’t see what effect, if any, they may have had on others.
“Before my teacher came to me,” Keller wrote in 1908, “I lived in a world that was a no-world. I cannot hope to describe adequately that unconscious time of nothingness. Since I had no power of thought, I did not compare one mental state with another.”
Surely, this must be what a dog’s experience of life is like. There is no thought process going on in his mind. He isn’t even able to compare one emotional state with another. How could he, without the use of language to help him form his thoughts? (Try thinking without words sometime and see how far you get.)