i read recently that some people believe dogs to have a capacity of up to 500 words. even if that’s not true, though, please answer the question.
do they understand those words if spoken by anybody? my dog goes bonkers if i say ‘leash’ or ‘walk’. but would he do so if ted koppel came into my house and said leash?
that seems to be the true test of a vocabulary; whether an animal learns the phonemes.
Heh, I recall reading that Germany once offered up some bomb-or-drug-sniffing dogs to the United States, but the project failed because the dogs only understood German.
Sofa King, you’ve got to be kidding. ONLY understood German! How uncouth.
Seriously, though, the local police dogs are trained to only respond to German commands. They figure it’s safer.
–Tim
“ja, du bist ein guten hund. ja, du bist. ja, du bist! ach, du bist ein schoenen kleinen hund.”
The dogs don’t understand any human language; they just learn that they’ll be rewarded or punished based on how they respond to certain sounds (words).
You must not spend much time with dogs. My dog (and he’s far from unique) understands quite a few sentences, as well as a large number of words. For example, “What do you want?” causes him to show me what he wants, and it’s not always the same thing. Granted, it’s usually food or to be let outside, but he clearly knows what I’m asking and responds appropriately.
As far as I’m concerned, this is just an oversimplified definition of language. Vastly oversimplified: dogs’ understanding of ‘human’ language is not binary; they’re capable of much more subtle understanding than “simply punishment vs. reward.”
If a dog knows that the word “sit” means to sit, and “speak” means to vocalize, I’d say they’re understanding the language.
As a former dog trainer, I can tell you that every dog, at the beginning of its training experience, has what I always called a “Helen Keller moment.” It’s the moment at which the dog suddenly understands that you’re training it; that there is a correlation between what you’re saying and what they’re doing. There’s a huge vertical rise in the training graph at that moment, and from then on you can teach most dogs a new command in a matter of minutes, if you know how to communicate cause and effect nonverbally.
I have always defined training, for my clients, as developing a common vocabulary with your dog.
I never counted any one dog’s vocabulary, but I never reached a dog’s capacity. And I’m fairly certain I’ve trained some dogs past the 500-command mark. I’ve had dogs who were capable of learning a new command almost every day for as long as I owned them. There were some commands I rarely went back to, but those I did rarely needed more than a reminder; they didn’t need to learn it all over again.
So I have no empirical proof that the 500-command limit is bunk, but I’m certain at least that it will vary from dog to dog; I’m certain it’s not a universally insurmountable limit.
lissener,
fabulous! you are just the type of person i want to talk to. now, when you taught a dog a command (say “Go get your rawhide”), could the owner simply utter that same phrase and the dog got it? did the dog understand the words independent of speaker?
or did the owner have to do the same command with the dog a few times before the dog picked up?
thanks,
jb
When I was growing up we had a golden retriever named Einstein. He understood many full sentences no matter who in the family said them. He probably knew 20 “go get the…” sentences. He could definitely tell the difference between “Einstein go get the stick” and “Einstein go get the sock”. If that isn’t understanding language then I don’t know what is.
How far could you take breeding a dog for language ability? Could you get them to understand very complex sentences, or with conditional statements, for example? (If the ball is black, bring it to me; If it is white, drop it in the hat. Things like that.) Has anyone ever tried this?
IMHO, the dog learns the phrase itself, regardless of speaker. I can name a bunch of examples, but the most telling is this: a dog usually (of course not always) needs to be taught not to accept commands from others, only it’s owner. A stray or a rescue still knows it’s name and whatever commands it was taught, although the owner is nowhere near.
In basic obedience, my trainer showed Cali (my dog) the hand movement, said the word “down”, made her lay down - she understood. Then I stood up and did the same - with the same results.
Now if a guest is over visiting and happens to use the word “outside” or “rabbit” in a sentence, Cali picks her ears up and starts dancing around. This is just using the word in a sentence, no particular emphasis. She knows the difference between OK, and Ohno, or OhBoy, or just O.
What is a word, but a collection of sounds to which we assign meanings? So IMO the dog can learn the words and assign some meaning to them, regardless of who speaks the word.
As with any linguistic communication, the entire command “Go get your rawhide” will be understood when the dog understands its components. If your dog know what “Go get” means, and what “rawhide” means, he may know what “Go get your rawhide” means the first time you say it. This will be the case, of course, if he is very familiar with both terms: if you frequently use “Go get” commands, for example, and if you always name the rawhide when giving it to him.
So you have to break down the teaching of the command: give him the rawhide, call it a rawhide. While he’s chewing it, tell him it’s a rawhide. Don’t confuse him with too much language at first: just say “rawhide,” not “That’s what’s called a rawhide.”
Then teach him “Go get.” Put the raw hide somewhere else; in another room or something. Look him in the eye, say “Go get the rawhide.” Say it clearly and distinctly. (It’s very important not to say it twice. Then the command becomes “Go get your rawhide go get your rawhide.” He’ll eventually get it, but it muddies your accumulating common vocabulary.) Then take him into the other room, and he’ll probably pounce on it. Then of course praise him; make him your hero.
Then leave the rawhide behind and go do the whole thing again. After just a few times, he’ll be straining to go get the rawhide. You can say “Go get the rawhide,” and just let him go, and he’ll go get it. No repeat the command every time he does, and eventually he’ll know exactly what you mean whenever you give him that command.
If you teach him “Go get your leash,” “Go get a stick,” etc., in the same way, eventually he may–some dogs do, some don’t–understand the more abstract “Go get” and be able to differentiate it linguistically from the object he’s to go get. At that point, you’ll be able to teach him to “Go get” things much more quickly.
What you have to do, most of the time, is figure out a way to get a dog to perform an action without knowing its command, then help him to associate this action with a particular command.
but as soon as the dog has grasped the concepts of ‘go get’ and ‘rawhide’, and is familiar with both words, can anyone say it to him and have him understand?
i guess i am trying to compare and contrast dogs and speech-recognition software (in one very specific way). the computer has to be trained with one voice, and if somebody else walked up to it and said “open folder prgrams in folder windows” or something, it wouldn’t have the foggiest. the second person would have to train the computer with his voice.
so is the dog recognizing the string of phenomes “goe get rah hide”, or just the commander’s specific uttering of said phonemes?
From my experience, I’d say that dogs can understand words regardless of the speaker* (as long as the accent was not TOO different) but unless trained otherwise, dogs will probably only respond to their owners.
It’s arrogant/ignorant to think that only humans have language capabilities.
Hey JB,
The dog is hearing us with his ears! In a natural conversion of sounds waves thru the delicate instruments of the ear, to his little brain.
The computer on the other hand is digitally recording an analog sound wave (massively huge file), analyzing it, comparing it to prerecorded patterns and then attempts to convert that sound into a command or a typewritten letter, or whatever the puter is doing.
LOL, software recognition and a dog’s hearing are way not in the same ballpark.
i know, i know. that’s why i said “in one very specific way”.
fwiw, it seems as if the answer to the op is yes, they learn them for everybody. (special thanks to aenea and lissener)
how about APB9999’s question? anybody?
I don’t remember the exact number but I thought seeing eyedogs learned several thousand different tasks and commands when broken down. I know this isn’t on topic to where th OP has gone but it is just another thing to think about.
HUGS!
Sqrl
My experience is that the dog does indeed understand the words, regardless of who says them. If the word “walk” is
used around my dog, his ears perk up. If his name is said in conversation, he will look at the person who said it.
When it comes to commands, he may or may not obey another speaker – he will pay little attention to a neighbourhood kid who tells him to sit, for instance. But he certainly knows the command.
We’ve tried to preface commands with his name, so that he will understand that the command is addressed to him rather than just part of general conversation.
He knows “Let’s go for X” where X can be “walk” and he goes to the front door; “ride” and he goes to the garage door; “nap” and he goes to the bedroom; or “pee” and he goes to the back door.
There’s still a great deal of contention and controversy regarding canine eugenics. One end of the spectrum of opinion on this matter, generally held by breeders and breed enthusiasts, is that each breed has specific, heritable behaviors. The other end, generally held by geneticists and biologists, is that dogs’ physical traits are highly variable and can be predictably controlled by selective breeding, but only the potential for certain behaviors are heritable, and that these potentials are held in common by all dogs–who are all, after all, members of a single species.
My experience as a trainer, and my further experience as manager of a large pet store for six years and encounters with thousands of individual dogs and their owners, in addition to my extensive reading of genetic theory (some specifically in regard to this issue) has left me firmly in the latter camp. I strongly believe that those behaviors thought to be breed specific are behaviors brought out by owners and trainers who expect and emphasize those traits.
I know of specific experiments of this nature conducted with parrots with spectacular results, but not with dogs. I’ll see what I can find.
Seeing-eye dogs further learn–and this is the really impressive part, which should silence anyone who thinks dogs can’t think–selective disobedience.
Seeing-eye dogs are not taught to blindly obey a command. They are taught to judge their surroundings and situation and make a decision as to the safety of obeying that command. If they’re told to cross a street, they must first judge traffic, distance, visibility, etc., and then decide to disobey if disobedience is indicated. They must disobey an immediate command, in other words, in order to obey the larger command of conveying their master safely among the obstacles they encounter along their way.
I agree with almost everything Dex says. One exception: breed specific traits. I have some experience with both Labs and beagles. Most purebred Labs have an amazing retrieving instinct, to the point where (with no training at all), they will fetch a thrown object, return it to the person who threw it, and drop it at that person’s feet. Other breeds tend to either ignore the object, or retrieve it, but begin a game of chase. Labs also have almost a need to swim. If there’s a lake around, they’re in it. Most other dogs have to be encouraged to swim, at least at first. On the other hand, Labs don’t have much of a protective instinct. They love everyone, and seldom make good guard dogs, unlike, say, German Shepherds, which have that basic behavior pattern built in.
Beagles, on the other hand, aren’t much good as retrievers or swimmers. Show them a rabbit, though, and they know what to do. Yeah, other dogs chase rabbits, but beagles (which are basically happy, lazy dogs) track/chase them with an intensity that’s almost scary. Squirrels might generate some mild interest, but rabbits are the enemy.