Agreed. Another way of saying this is that animals can learn vocabulary, but only humans as far as we know have the ability to assemble that vocabulary into consistent systematic patterns to convey complex ideas. We can create new sentences that we’ve never seen before to express entirely new ideas. In linguistics this property of language is sometimes called generative grammar and it appears to be uniquely possessed by humans, and some believe that it’s innate to the human brain.
This is unrelated, however, to the fact that many animals like the apes in question and the dogs in the previous examples are clearly intelligent and sentient. It’s more an argument about what we really mean by “language”.
I’d say the difference is that between two humans who share a common language and can converse at length and with nuance, and between two humans who have only utterly dissimilar languages and can only communicate through pantomime and example. The difference is that Koko is never going to learn English and Dr. Monkeybrains is never going to learn Gorilla… but Chen and Enzo can and will learn each others’ languages.
My dog has several simple ways he communicates to me and understands a good sized vocabulary of words he responds to. He gets my attenetion by barking at me or licking my face. I will verbaly give him options, when I hit the right one he will turn in a circle. I think the circle means lets go.
The work with Koko is cleverly described as ape using sign language to communicate. That is not untrue for all senses of the term ‘sign language’ because Koko did communicate with signs, but that is only using the words from sign language, not using signs to communicate with language.
Another one is the parrots - I have had several parrots and currently my green-cheek conure is the family flock chatterbox. He has something like 20-30 words and stock phrases he uses to communicate with us, the humans in the household. He can ask for food/water, express alarm, has a label for each human, will praise you when you do something he wants, and so on. But it’s all very set collections of sounds, he can’t break a phrase down into words and reformulate it into a new sentence. He doesn’t understand language the way we do.
So yes, he communicates but he doesn’t know the English language.
A bird like Alex Pepperridge might have a more sophisticated manner of communicating using English words and phrases, he could have some concept of things like “number” and “color”, but even at his most proficient he wasn’t speaking English as a human does.
We all think our pets are communicating with us, and in a certain, limited sense, they are. A wagging tail on a dog is probably a good indication that it is happy; on a cat, not so much. I don’t deny that is communication, albeit involuntary.
But it seems unlikely (or at least unproven) that if your parrot makes a sound like “water!” and you give him water, that this is communicating in an intelligent way. How do you know that he didn’t say a phrase meaningless to him, and drink the water anyway, making you think there was a connection?
Does he ever say any words/phrases that don’t seem to make human sense? Do you just forget those, and only remember the ones that sound “intelligent”? That’s the old “remember the hits, forget the misses” trick used by psychic mind-readers.
Dogs, having spent tens of thousands of years longer with humans, may be the better subjects. Mine understand the phonemes of “car ride” to mean fun and a trip to McDonalds.
Alex was fucking brilliant, but Talking Bird, my collie’s name for our cockatiel, loved to see me fall all over myself when I was unemployed and he imitated the phone ringing. I hate that bird,
They’ve tried to suggest that it’s more than that, but only with weak anecdotal examples. There’s an example that purports to show Koko capable of synthesis, when she supposedly wanted to refer to the ring that one of her handlers was wearing, and, lacking a word for it, made the signs for “finger bracelet”. Even if it’s not merely a fanciful interpretation by the handler, which would be my first guess, it’s not in any way comparable to what we mean by actual language, which is wholly centered on complex and systematic relationships with rich expressiveness that we call a grammar. The appropriate word for how animals communicate with us is not “language”, but “code”.
And again, all due respect to the intelligence and sentience of the higher animals, many of which I’m convinced are far, far more intelligent that we give them credit for, or that we are capable of fully recognizing. But language as humans define it is just not one of their capabilities. Doesn’t mean that they don’t have other qualities of intelligence or sentience that are as good or better than ours.
Aside from those times he’s sitting on the perch practicing his words/phrases (because birds do that, they repeat noises and practice their routines), he’ll use the phrase when we’re eating or drinking, he’ll use it when the other birds are eating or drinking, and there’s a high correlation between him doing while sitting on food or water bowl and that bowl being empty (he uses the same word for food or water).
Birds babble. They’ll make nonsense noises to themselves. They’ll also do it in a “conversation”. Basically, you say something. Then he goes “mumble-burble-hrm”. Then you say something. Then he goes “blurble-hurbble-babble”. He’s not saying anything sensible, but there is a sense of turn-taking and not speaking over the other person. So, it’s some sort of interaction but the message is “I’m paying attention to/interacting with you” rather than something more specific.
Parrots (and some other birds) can enunciate words pretty damn clearly when they want to do so. If he’s saying “pretty bird” or “good girl” it’s not ambiguous or fuzzy. Whether or not he understands what he says is what’s debatable. He certainly doesn’t understand his memorized phrases as discreet and separate words joined together.
Another example is “step up”. It’s a standard command taught to parrots, to get them to step up onto a finger or hand so you can carry them, or to tell them to get on a perch. So, we put a finger out and say “step up” and he does - but sometimes when we put our hand out he’ll say “step up” before we do then steps up onto the offered hand. He’ll sometimes say “step up” on his own, then try to climb into your hand, including pulling your hand off the mouse or keyboard before climbing up. It’s pretty clear he associates the phrase and the action.
It’s like asking if a dog understands the command to “sit” - if he sits on command sure he does. The difference is that my parrot doesn’t just understand and obey, he can also say the command in English, too. Again, that doesn’t mean he has the same understanding as we do, but it’s ridiculous to say he has no idea what he’s saying.
Rico the dog had a vocabulary of more than 200 simple words, according to a peer reviewed article published in Science. The authors had taken care to address the Clever Hans effect. Rico (dog) - Wikipedia
Nova had a terrific episode on the human-dog partnership. It covered Rico, as well as the differences between wolves, apes and dogs. When a human points at an object, the dog responds. The other animals just shrug.
Interestingly, humans are fairly fluent in dog-language. Record a dog in Hungary and play it to a British dog owner, and the Brit can guess what the dog is doing. Clever human!
I haven’t seen the program, but the human-dog relationship is historically long, deeply entrenched, and emotionally profound. Dogs do indeed understand pointing gestures, and are among the very few or perhaps only animals that can do so. It’s been said that elephants do as well, but I’m not sure if that’s been positively confirmed. It’s indisputably true for dogs as many of us can personally attest.
Rocco, our African Grey, does a head bob when excited (like an enthusiastic “Yes!”).
He likes drinking polar water from my copper mug. I’ve paired “want a drink of water?” with offering the mug through hundreds of repititions. Now I’ve started asking “want a drink of water?” and only offering the mug (which he can see in my hand) if he does a head bob.
Cute trick. If he has recently had a drink (and presumably isn’t thirsty) he doesn’t head bob!
If I say “step up” as I approach, but before putting my hand in place, Rocco will often lift a foot in preparation of stepping up.
He also, when desiring attention, will say phrases that elicit laughter from people. Anytime we are in an adjacent room with company, he seems to know that saying, “razzberry sound, What’s that smell?” Will make everyone laugh and draws them to him.
I’m pretty sure that my first dog also had a vocabulary of about 200 words, but unfortunately most of them were synonyms for “walk”. We had to go through that many because he’d go crazy whenever he heard the word, even if we weren’t talking to him, or even thinking about dog-walking (in fact, if we were, we would have remembered not to use the word).
Only vaguely apropos of later posts, I am training my new puppy, named in Egyptian, with commands given in Egyptian.
Partly for snark/swank value and partly to prevent common words from triggering behaviors. My last big Dane loved to jump up on my shoulders, and if I was placed and braced, he would do so with almost exquisite gentleness. But if the command word - “hug” - was said in conversation, I might end up getting knocked on my ass by his desire to say I Wuff Oo.
“Yinkweh” - hug or embrace - is not a frequently heard word these days.
Perhaps we’re looking at this from the wrong direction … instead of asking why animals can’t learn even the very basics of human language … we should ask how humans learn language. Human languages are amazingly complicated; tone, cadence, voice, pronouns, participles, inflections. Even more amazing is the average two-year-old has mastered all this enough to lie.
Just a note about computer chess machines. They are strictly tactical after the opening, crunching all the possible moves by force. So giving the computer enough time per move it would have beat humans long ago. Where Deep Blue made it’s mark was beating the best human where both had the same amount of time. What computers don’t do is analyze positions, they never ask themselves if they want to play an open game or a hedgehog or a gambit. They only randomly pick a sound opening and crunch numbers after … computers don’t play artistically.
Are you sure about this? I haven’t read up on human-computer chess recently but last I heard it was the increase in brute force that has fueled computers’ improvements. Sure, the programmers have tried to make the computers’ computations more efficient, and have tried to shore up some of the weaknesses they have had, but I think brute force is still key. Computers also had endgame tables that could be used to steer the game toward victory, even when it didn’t look like it. Humans described themselves as thinking they were winning, right up until they were losing.
I don’t think computer strategy would ever be considered elegant or insightful the way a human player’s might.
Deep Blue was a “proof of concept” machine, demonstrating that this brute force method would work when applied to nuclear explosions and air foil modeling. I forget off-hand where IBM built their final machine, Los Alamos or Lawrence-Livermore.
I can see why you’re questioning the statement as it wasn’t very well worded. I certainly didn’t mean to imply that brute-force look-aheads are not done, as they obviously are and they’re a very important element of computer chess play. I mentioned Big Blue because it was a particularly famous chess-playing system due to the match with Kasparov, but ironically Big Blue used very fast processors augmented with special hardware that allowed it to exploit brute force to a much greater extent than earlier chess programs – but also to a much greater extent than the modern programs that came after and play equally well or better. Brute force alone would be ineffective beyond a low level of play and there are huge and fundamentally important components in all chess programs that work as I described, and I’m not talking about the relatively mundane opening book and end-game tables.
No matter how great the computer power, realistically it has to be directed efficiently by employing heuristics that evaluate board positions and decide which paths are the ones worth exploring in greater depth and hence expending computer cycles on. Deep Blue had thousands of different evaluation criteria and it quantified them by analyzing thousands of master chess games. When you have a system that can increase its skill by such training, you’re in a real sense embodying knowledge. Grandmasters also worked directly to add to to the heuristics in Deep Blue, the whole of which constituted deep chess knowledge without which brute force alone would have played a relatively poor game.
Modern chess programs that typically run on PCs don’t have access to specialized or superfast hardware and the best ones manage to play at world class levels by relying even more on heuristics to exploit embedded knowledge and less on brute force.
Whether a computer’s chess strategy can be considered “elegant” or “insightful” is pretty much subjective. Kasparov famously believed that Deep Blue exhibited just such characteristics and accused IBM of cheating by planting a human chess master behind the scenes.
I also have some personal observations on that. Back in the days when the PDP-10 MacHack program became the first really good chess program and famously beat the philosopher Hubert Dreyfus who was a skeptic of machine intelligence, I used to enjoy playing it myself and I still can using a PDP-10 simulator on my PC. It’s still superior to some of the weak programs around today. Back in the day, I had one of those early portable chess machines, but I much preferred to play MacHack because even if I lost to both, the games were drastically different. The lower-level chess machine often just seemed aimless and waiting for me to make a mistake, so it was sometimes tedious. MacHack seemed to have clear goals that it pursued aggressively, so it was more fun to play, and when I lost, it was often in spectacular fashion!
The Gardners, the people who did the original Washoe the chimp experiments, learned sign language from books, and then taught it to all the people who worked for them. They claimed to be using American Sign Language, but they never learned it from a native user. At one point they had a whole fleet of chimps, and all they in-house trained hearing people were “eliciting” lots of signs from the chimps according to their written data, which consisted of notebooks in which the people were supposed to write down every sign their assigned chimps made.
At one point, they hired a guy named Ted Supalla. Ted was Deaf, a native signer, the son of Deaf parents, and working on a Ph.D in linguistics. At first the Gardners were excited to have him, but their relationship broke down when Ted realized that the chimp trainers wrote down virtually anything as a sign. Ted’s chimps consistently “underperformed,” because he didn’t interpret virtually any hand motion as a sign. Once, someone’s chimp escaped the grounds. After it was caught, Ted sarcastically suggested “Why don’t you write down that the chimp made the sign for ‘escape.’” He got fired.
