Well, let me add my anecdata. Two dogs, Mike and Mac, lived and grew old together. Mac died. Six months went by. Then, at a family gathering, Mike was sitting with a small child named Jackson. “Look at those two,” said a relative, affectionately. “Mac and Jack! Mac and Jack!” Note here that the relative had temporarily mixed up his living and dead dogs, and that he was using a loud, singsong tone of voice. And also that Jackson was usually called Jackson, not Jack.
Mike jumped up (an unusual burst of activity for him) and started looking around wildly. Then he began whimpering. To me, it looked as if he was looking for his dead friend, then getting distressed when he couldn’t find him.
Inflection was important here–since his death, Mac was often talked about in Mike’s hearing, with no reaction from Mike. But when he heard Mac’s name called out, it was obvious that he had that particular pattern of sound associated with Mac himself, just as he knew to come when he heard Mike.
Blake, let me start off by saying that overall I agree with you: there is no evidence that most animals understand their name as an identity marker, they are only responding to conditioning.
But I must nitpick that you are conflating classical (Pavlovian) and operant conditioning. Classical conditioning consists of stimulus-stimulus pairing to produce a reflex response to a conditioned stimulus. Pavlov paired a bell ringing with food (stimulus/stimulus) to produce drooling in response to a bell ringing. Operant conditioning, on the other hand, is based on the ABC model: antecedent, behavior, consequence. Organisms learn new behavior through consequences. This is what is happening with a dog learning to respond to its name. Antecedent: owner calls name. Behavior: dog comes over. Consequence: dog receives affection.
I only know of a few animals that I believe past the language test, and those are the bonobos trained by Sue Savage Rumbaugh. Check out this video of Kanzi, who responds to novel commands, such as, “Put the pine needles in the refrigerator.” This video shows Savage Rumbaugh wearing a welding mask to avoid giving unconscious cues to Kanzi. I have also seen other videos where Kanzi is wearing headphones listening to commands given by a researcher in another room. Research shows that Kanzi and his sister Panbanisha can follow novel commands as well as a four year old child.
As for Kanzi knowing his name, the most amazing video I ever saw of Kanzi was as follows (you’ll have to go to UC San Diego and ask the anthropology department for a copy, I watched it in college): Kanzi was sitting with Savage Rumbaugh and another “control” bonobo that did not have any language training (we’ll call him Mike). Savage Rumbaugh said, “Kanzi, tickle Mike.” Kanzi obligingly tickled Mike’s belly with his hand. Savage Rumbaugh then said, “Mike, tickle Kanzi.” There was a pause as Mike looked from the researcher to Kanzi and back again; he had heard his name, but clearly he did not understand. Then, amazingly, Kanzi took Mike’s hand and placed it on his own belly. To me, this shows strong evidence that not only does Kanzi understand English language construction (he comprehended who was the actor in each sentence as well as the verb), but he knew that “Kanzi” referred to himself and “Mike” referred to the other bonobo. What’s more, he tried to correct Mike when Mike did not respond!
I think the problem here is that Blake said up thread that he thought humans couldn’t pick names out of an unfamiliar language spoken naturally. This simply isn’t true - it can be a very challenging task, but it depends on a variety of factors (how distinct are the phonemes, rapidy of speech etc.). It would be almost impossible to pick some names out of some sentences, while others would be relatively easy. Thinking it is an impossible task for humans is why he dismissed it as being impossible for dogs.
However, this doesn’t really have any bearing on whether dogs can understand language. Just because some dogs can pick their name (a distinct sound) out of a sentence (lots of sounds run together) doesn’t mean they are parsing them grammatically.
The dog Rico is impressive, but with such a small set of commands it’s a big stretch to say he has any understanding of grammer. Blake’s sock example above would be a good test to perform, what contexts can Rico apply these skills to?
The capacity for animal learning is a very interesting and quite open subject. However, I do think that many people don’t appreciate quite how big the gulf between human and animal learning is. Some animals exhibit clever tool use in the wild. However, tasks that humans could learn in minutes such as cracking nuts and fishing for grubs often take primates and birds months and years to become competant at.
Blake is taking quite an extreme position against anecdotal data. I’m fairly confident from what people are saying in this thread that some dogs are quite good at recognising their names in a sentence. However, I would treat with a degree of scepticism how good they actually are at this. A proper scientific test would isolate variables and take measures to eliminate false positives. Does the dog only respond to it’s owners voice? Is it responding to any other cues? Does it respond to a variety of voices? To recordings? What range of intonations will it accept?
I don’t think anyone is suggesting this though. Dogs can be conditioned to respond to commands, but they don’t always obey them. Pavlov’s dogs may have salivated at the sound of a bell, but I bet there was less drool to collect if they had just eaten a rack of ribs.
I am surprised no one has mentioned this recent paper, in which researchers fairly carefully characterized the language capabilities of a border collie and concluded it “understood” that words referred to specific objects and classes of object, among other things, demonstrating behavior that cannot be explained by simple conditioning. This dog was named “Chaser”.
A link to the study itself is at the end of that article.
Having found this myself a few weeks ago, it makes the arrogant certainty of certain people in this thread rather amusing. It’s a peer-reviewed journal article, Blake, “Border collie comprehends object names as verbal referents”, what do you say about that?
Here is a video I found on YouTube, which shows some of the tests. There are too many cuts to get a good idea of the dog’s true abilities, but regardless it is still very impressive.
I found the PDF of the actual paper available for free from the author’s website and read it last night. This seems to be exactly what you wanted, Blake, I hope you stop by to comment on it. The link is:
Experiment 2 is relevant to this thread. Here are some excerpts:
This seems pretty clear to me. Chaser is an otherwise unremarkable border collie obtained from an ordinary breeder.
An equally interesting read is Experiment 3, in which they conclude Chaser learned the names for categories of objects that she already knew specific names for (e.g. “ball”, “frisbee”, etc.) and could retrieve one of those objects based on the category name as well.
I think one issue here may be that dog’s vary enormously in capability. I’ve certainly owned or known dog’s that don’t respond in the ways indicated in this thread (including “smart” breeds such as border collie). But I’ve also seen the behaviour of our last dog Dusty who repeatedly
Responded to his name when used in a different context and neutral tone by people who were oblivious to his presence in the room (his sister also did this)
Responded appropriately to commands that he had not been trained to follow and consistently respond to conversational topics that involved him, often involving only words that he wasn’t trained to respond to.
Enact what seemed to be systematic and considered planned behaviour towards a goal.
It was difficult to get a firm handle on what he could do because he showed no interest in language that wasn’t directly relevant to him (and was notorious for his tendency to treat commands as suggestions :D) But where it was relevant to him his behaviour was entirely consistent with his understanding a substantial vocabulary of words.
Like some of the early replies mentioned the problem is what you OP are seeking to get answered. Many, many animals can obviously respond to a sound.
The general problem in identifiying if animals are self-aware or just aware, if they can solve problem or if they just learn from prior mistakes/repetition is an endless issue among zoologists.
If you want to know if they can identify with their name you must establish several of things:
They can respond to a specific sound.
They are aware of themselves.
They can connect their self-awarness to a sound and identify with it (how would you prove this?)
edit: Some apes I think have pointed to themselves, saying their names. But how can you show that they don’t consider this to be a game?
I live alone, except for my cats, and I made the remark recently, “I hope this weather warms up soon.” Just talking to myself, really.
But Hope, who had been napping in front of the fireplace, raised her head and meowed. She heard her name in my remark, and responded.
It’s an anecdote, and those are never data. But Hope did respond to a sentence that contained her name. Not sure what this anecdote is worth, but there you go.
You do not know that is a fact; you know, I assume, that many studies indicated otherwise.
These studies may be credible; we don’t know, because we haven’t read them, because you haven’t linked to them. What you have done is made a condescending as well as inaccurate statement designed to get someone’s back up.
I do not think that the most effective means for sharing information.
I remember a study with tamarin monkeys that suggested they have some of the same cognitive building blocks for language that humans do. A series of linguistic tests that have been given to human infants, children, and adults was given to the tamarins and they displayed a similar ability recognize sounds that formed words out of a stream of words with no pauses between them.
In other words humans can recognize what makes up individual words and what makes up syllables within a word by recognizing repeating patterns in a language - even without a pause between the words and even not knowing the language we’re listening to. It is theorized that some non-human primates possess the same ability.
That means they most probably can hear their own name in a sentence unrelated to them. Anecdotally I know my dogs can. I would expect non-human primates to be even better at it.
Not sure if this is what they were referring to, but both dolphinsand parrots have what could be considered names.
Parrots assign these names to their young by making specific, distinct vocalizations that the young learn to mimic. This identifies them to the parent from a distance and allows them to recognize a parent calling to them.
In the case of dolphins the young create their own signature whistles innately which are then mimicked by the parents or others close in their social circles when they become separated.
In both cases there is a two-way communication going on that is the equivalent of having a name and recognizing it when it is called.
