Dog learns over 1,000 words
nytimes.com
nytimes.com
There's just so much sloppy research out there.
The article mentions the fact that humans can use context to help remember words, but the dog has to use rote repetition. It's interesting that the reliance on context that helps us learn stuff in general could mean that dogs may actually be "better" at building a vocabulary of 1000+ unrelated names.
I have a border collie who's remarkably intelligent. But the words she knows are just the beginning. She's incredibly perceptive. Gestures, facial expressions, and vocal tone are all things she pays attention to.
Determining how intelligent an animal is by how much it can learn the words of our language is a remarkably arrogant thing. We've clearly underestimated the intelligence of every animal we've ever dealt with. We only ever revise our estimates of animal intelligence upwards.
I wonder how much of that a dog or similar animal picks up on regularly? Could it be that they actually pick up on the majority of our communicated meaning?
What's more, I wonder if we are we actually poorer communicators than animals, because they understand our messages better than we understand theirs?
Yes, but to be fair, that's common knowledge. Who hasn't run this experiment: tell a dog "you're so stupid and worthless!" in the happiest, most playful talking-to-a-dog tone of voice you can think of while scritching it behind the ears and see how it reacts?
The idea that a dog has a very good command of spoken vocabulary in addition to picking up nonverbal cues is what's new and interesting.
In the past, we would have said that this was an excellent trainer and the dog just does it for the food. In fact, this has been our default for hundreds of years.
For all we know, many dogs know or understand thousands of words, gestures, expressions, and even smells. This border collie probably understands even more than the words associated with the toys its supposed to pick up.
I call this a Jurassic Park problem. We ask bad questions and so we get the wrong answers. We continue to ask if any dinosaurs are missing instead of asking how many dinosaurs there are.
It's of no small amazement to me that we're so quick to treat revelations like this as a surprise. They're not.
I'm all in favor of the results of these experiments being spread as widely as possible in the hopes that it creates doubt about the way we treat all animals. We might treat them better.
I love dogs, I think my dog is super-smart, funny, and very sweet. I've heard her particular breed referred to as "the dog with the human brain" on some dog TV show before, and I'm sure she's not even as smart as a collie. That said, I can't help but feel like your comment is a bit... overwrought. Dogs clearly understand parts of human language and judging them by their ability to understand it seems like the best way to determine their intelligence. I would hardly call it "remarkably arrogant". You have to consider the fact that humans have likely had spoken language for as long as we've had domesticated canines.
In the end I can't help but feel like over-sentimental and hyperbolic (would you say we've "underestimated the intelligence" of, say, jellyfish?) declarations like yours really set us up only for disappointment. Our dogs are great, and probably do actually love us, but they're not furry little geniuses held back only by their lack of proper speaking ability.
All that said, they're still way better than cats.
You begin with the assumption (based on nothing other than a general feeling of human specialness) that "For all we know, dogs responses to human emotions, gestures, and nonverbal cues are a completely autonomic behavior". Considering that we discover all kinds of animals are smarter than we thought with each passing experiment, the burden of proof should lie with you. The limitations appear to be our experiments.
We do not start with this assumption when dealing with human beings. The safe assumption is that animals have their own desires, goals, ambitions, and purposes for existence. Since we have significant trouble understanding them or communicating with them (even though we know that within their own species they communicate perfectly well with each other), we might start from a less biased viewpoint.
As for jellyfish, I don't know. But I do know that we've approached it from the wrong angle. Asking how many words a jellyfish can respond to and then determining whether or not it is intelligent is a surefire way to create a lousy experiment.
BTW, Europeans used to do the same sorts of experiments with people from other countries and civilizations. They knew that white Anglo Saxons were the best and brightest so they looked for evidence to prove it. Africans clearly hadn't invented guns or the printing press so they determined that they had a lower level of intelligence. It's just as ridiculous a methodology when you're talking about studying animals.
Keep in mind that humans have lived in a very close, symbiotic relation with dogs for a very long time. It's at least 20k years, with some indications the relationship could go back 100k years.
It may have even been the advantage dogs gave us that let us win out over rivals such as neanderthals.
Edit: typos
There are a large amount of interesting things she responds to other than simple commands but I can't say she has a human type of intelligence or even close but there is definite problem solving ability (such as opening cabinet doors, the refrigerator, and even pushing the sliding glass door open by standing up on it and moving her paws).
More to the point, our understanding of intelligence was rudimentary until just a short time ago and is still developing.
We're like medieval mechanics, presented with an F-22. At first, things just seemed magic, and we couldn't even theorize how all of it worked. Now, we've started to develop some theories, and we can even make some toy examples that mimic certain aspects of functionality, but we're still far from being able to build even a crappy knockoff from scratch.
She'll do practically anything for cheese. It took about a whole slice of american cheese to teach her to ring a bell on command; then we tied it to the door. We put her leash on and made her ring the bell to open the door - she figured it out in exactly one try.
When we show her a new thing and get some food out, she KNOWS she is going to learn something new. She'll quickly poke it, shove it, pick it up, look at you, and repeat. Just in case that's what we wanted her to do...
Dogs are amazing.
(Obligatory picture: http://www.flickr.com/photos/joshu/4557297986 )
By the way, I didn't know this before, but dogs can be rewarded with toys instead of treats - I taught him a few other tricks this way.
Ours mostly is food-driven. She'll fetch if there's food involved, but not otherwise.
That said, German Shepherds are an interesting case because, as I understand it, their breed standard is actually a behavioral standard: they have to be able to perform a certain level of obedience, tracking, and bite work (think big leather gloves) to get papers.
Animals are just doing tricks. Humans who learn to play a sport aren't.
Animals are just doing it for the food. You go to work every day, why?
It's worth noting that this has been an idea that has less evidence to back it up all the time. This idea will eventually be disproven, but in the meantime, there are an awful lot of people emotionally invested in this specialness being true.
Many years ago my parents' cat started trying to open doors without any training from us. She clearly understood the process from her efforts, but couldn't grip the round handles. Given a house with cranked handles though and any door that opened out rather than in would have been insecure to her.
With only reinforcement of her existing behaviours we've got her to learn: sit, spin, lay, roll (over), pretty (where they sit with their ass on the floor but upright with her front paws off the floor, my mother-in-law's dog helped greatly with teaching by example), tall (where she stands by herself on her back legs) dance (she'll do a 360 while on her back legs) and speak. [Ed: I forgot, we taught her to 'stand', IE move from lay down to sit, or sit to on all fours, because she'd continuously lay when told to sit.] These are all performed easily with verbal command only.
So far we've only trained using single syllable words (dogs don't understand the actual words, their brains are merely receiving a signal and performing an associated action) to keep the input the same. My wife is Canadian and I'm British, so how we pronounce things can be very different sometimes, but a syllable is nearly universal in the Anglosphere... basically we're making our dog multi-region compatible as just in the people she meets there's easily 4 or more dialects that can be giving her commands.
Potential tricks we can teach based off her existing behaviour are: Play dead, jump, growl and as she's got hound in her I'm hoping to one day get a howl command trained. We've noticed hints of 'paw'. She already knows fetch with a thrown object, so it would only be a matter of association training to get her to actually bring things (although for my in-laws their dog just brought the paper when my father-in-law jokingly told the dog to bring it without it being previously trained and their dog was picked up at 8-weeks so I doubt it had previously been trained).
My dog is a Dachshund and Jack Russell Terrier cross and certainly isn't the easiest dog to train, but simple operant conditioning is working great. By making her wait before she's allowed to eat she's actually stopped scavenging when we take her for walks (try ripping a piece of KFC out of a dogs throat, she never bit at me but it's damn hard trying to pry a dogs jaws open and stick fingers down its mouth to stop it choking on a piece of meat with bone in it), she's also stopped stealing food off of the coffee table (she got in from her walk one day and stole a large day old pizza crust off the table and had eaten 4/5ths of it before I finally managed to catch her).
By the best estimates, I have another 14 years with this dog and so far I've mostly been working at removing bad behaviours she had learnt. For all the time I taught tricks, I spent easily twice as long making bad behaviours go extinct. I don't intend to stop training my dog because the more time I spend training her, the more relaxed she is at home (Jack Russell and relaxed are commonly considered antonyms).
I'll have to record it sometime, but the thing I don't think people understand about dogs is that they're consciously trying to comprehend you. When teaching her to sit pretty by verbal command only she would hit this point where she hits the 'zone' her eyes become locked and you can see the gears crunching and then she performed the trick with no signalling. She also tried brute-forcing her tricks; literally she'll perform every trick she's learnt and then give up and start doing 'speak' to every command. You go back 5 minutes later and she'll hit the 'zone'.
And that differs from understanding exactly how?
Example joke: "Ever notice how we park on driveways, drive on parkways, pay tolls to go on freeways and it takes longer to get where you want on an expressway." All illustrate how our usage and the meanings of our words have been modified over time.
A driveway is a section of personal roadway that leads to a personal garage (you would never have left one of the old 70% wooden cars outside with zero rust proofing on the metal, ever). A parkway was supposed to be a scenic roadway to link urban and suburban parks with pleasure roads where people could park freely and enjoy the area, now it's commonly a synonym for any general highway. Freeways were actually speed-limit-free highways, the 'free' never had anything to do with cost until the very late 20th century. An expressway was designed as a high-speed arterial road, which may have a limited number of driveways.
If the dog can tell the difference between being told to fetch the paper vs fetch my shoes, then he understands.
Beyond that, the statement...
> their brains are merely receiving a signal and performing an associated action
Equally applies to humans; our brains just have a more complex form of association. There is no inherent meaning to any of the noises we make that we call words other than they're associated to something. Your associations to those noises is far more complex than the dogs, but you can't call yours understanding and not his; his understanding is simpler, but if he performs the correct trick, then he understands the word in the same sense you do, he associated some kind of meaning to that word, just like you do.
You're right, just because the dog doesn't understand 50+ words for 'drink' (water, coke, pepsi, coffee, tea, etc etc.) doesn't mean it isn't understanding in the same comprehension as if you got a 3 year old to bring you a 'drink'.
I don't know what I was thinking, I must have been in a state of sleep deprived idiocy. The simple fact you can teach a dog in human linguistics to sit when you say 'sit' signifies its understanding. You wouldn't get this response in an animal that all its communication skills are genetically encoded.
I remember reading about how Native Americans used to hunt Wolves with pitfall traps and how wolves would teach their young how to avoid the traps, like they would hunt a deer, they can also do the same with poisoned meats. These clearly show wolves are capable of understanding abstract concepts. Unless sink holes are a natural predator of wolves I doubt there's a genetic cue for "you see that flat piece of ground, you'll fall through that into a big pit" or "you see that steak, that's poisoned, but the one sat on the side of that metal contraption in their yard isn't!"
I was thinking of understanding along the lines that we can understand multiple meanings and abstract ideas. However, 1:1 learning is still understanding, and clearly Canis Lupis is capable of learning abstract ideas so a dog certainly has the ability to (even if we have to rebreed it into the species).
If you had a golf ball, tennis ball and cricket ball; all small enough for a dog to fetch. Would it know which "ball" to fetch if you had a club/racket/bat in your hand?
If yes, then two forms of learning are identical. If not, well ..