Wednesday, January 17, 2007

Re: My bird ... has learned to communicate using human language


As is my long-standing policy,

[Left: N'Kisi with owner-trainer, Aimée Morgana, BBC]

I am responding to your private message on an creation/evolution topic (namely whether a bird can talk in the same sense that humans do) to my blog CED, after removing your personal identifying information. Please don't interpret this as an invitation to debate-it isn't. After more than a decade (1994-2005) debating creation/evolution on the Internet, I ceased debating in order to write my blog and books.

----- Original Message -----
From: AN
To: Stephen E. Jones
Sent: Tuesday, January 16, 2007 10:23 PM
Subject: Parrot Speech

>Dear Dr. Jones,

Thanks, but I am just plain Mr. Jones, BSc. (Biology).

>I was scanning information and came upon some comments you made in response to ape communicating through human language.

Presumably you mean "Re: Apes and Language," Stephen E. Jones, 29 Jul 1999, in which I quoted from an article that questioned of apes, "... whether they're understanding what they're doing ..." and pointed out, "if you look at their production of language, you'll find it's vastly different from the manner in which ... a child uses language" such that, "Were a four-year-old child to use language in the way a chimpanzee uses it, we would consider that child disturbed" (my emphasis):

"One bonobo (pygmy chimpanzee), Kanzi, can use the system to say: `I want a cup of coffee, please'. Another, Panbanisha, is said to know around 3,000 words. ... But Dr Tom Sambrook, of the Scottish Primate Research Group, told BBC television of his doubts that the apes' achievements signified all they appeared to. `They can use language effectively to make requests', he said. `But whether they're understanding what they're doing is a much more difficult mystery to disentangle.' ... `But if you look at their production of language, you'll find it's vastly different from the manner in which, for example, a child uses language. Dr Sambrook quoted an earlier researcher's verdict: `Were a four-year-old child to use language in the way a chimpanzee uses it, we would consider that child disturbed.'" (Kirby, A., "Chimps' language skills in doubt," BBC, July 26, 1999).

and/or Re: Parrot cummunication [not my spelling], Stephen E. Jones, Mar 12 2000, in which I wrote, "My problem was the claim that animals can talk" and "A parrot might actually learn that the mimicked words `polly wants a cracker' ... in order to get food. But that does not mean that the parrot knows what the words actually mean. If it doesn't, then it is not talking as humans talk to each other" (my emphasis):

"I have no problem with `interspecies communication'. My problem was the claim that animals can talk. My cat meows for his food and I can understand what he is communicating. My wife claims he has different meows for different things, but I can't pick it. A parrot might actually learn that the mimicked words `polly wants a cracker' caused food to be offered and then says those words in order to get food. But that does not mean that the parrot knows what the words actually mean. If it doesn't, then it is not talking as humans talk to each other."

Earlier I had commented ("Re: Some Parrots Have Ability to Talk With Humans, etc," Stephen E. Jones, Mar 09 2000) on a no longer webbed article about Alex, another claimed "talking parrot," pointing out that "I saw a parrot sing `Happy Birthday' in an opera-singer voice at the Singapore bird park but I no one claimed that it knew what it was singing" (my emphasis) :

"This is a big problem for those who claim that chimps and gorillas can talk. If the claim is that chimps can really use sign language because they are closest to humans, then what is the explanation for a parrot who talk as well, if not better? I saw a parrot sing "Happy Birthday" in an opera-singer voice at the Singapore bird park but I no one claimed that it knew what it was singing. Parrots are just very clever mimics and human beings are very good at training them and reading into their pets' behaviour their own human feelings. Maybe this exposes as an anthropomorphic delusion the whole field of talking apes?"

>Are you interested in parrot communication?

Not very. As the above indicates, I am sceptical (to put it mildly) that "parrot communication" is the same thing as "talking as humans talk to each other" because the parrot does not know "what the words actually mean", as in that example of "a parrot sing `Happy Birthday' in an opera-singer voice at the Singapore bird park."

>I am an independent investigator working with a free-speaking macaw. My bird is not trained and has learned to communicate using human language (English). This experiment is unlike others because the bird communicates voluntarily.

I would say that your bird that "has learned to communicate using human language (English)"-like sounds, not "using human language" in the sense of "talking as humans talk to each other."

[Right: Blue-and-yellow Macaw (Ara ararauna), Jurong Bird Park, Singapore, Wikipedia]

That parrot at Singapore's Jurong bird park in 1996 was also a macaw (perhaps the very same one in the photo) and it had "learned to communicate using human language (English)"-like sounds in singing "Happy Birthday," was so convincing that if anyone heard it without seeing, or knowing, that it was being sung by a parrot, they would probably assume it was a human being thinkingly singing "Happy Birthday" to another human, instead of a bird unthinkingly making a learned sequence of human language-like sounds in response to its keeper's trained cues.

According to Wikipedia, macaws "are monogamous and ... In captivity unmated macaws will bond primarily with one person - their keeper" with whom "Pet macaws thrive on frequent interaction":

"Macaws ... Birds in captivity Macaws eat nuts and fruit. They also gnaw and chew on various objects. They show a large amount of intelligence in their behaviour and require constant intellectual stimulation to satisfy their innate curiosity. Bonding: Macaws have been said to live for up to 100 years; however, an average of 50 years is probably more accurate. The larger macaws may live up to 65 years. They are monogamous and mate for life. In captivity unmated macaws will bond primarily with one person ? their keeper. Pet macaws thrive on frequent interaction, and a lack of this can lead to their mental and physical suffering. Other sub-bondings also take place and most macaws that are subjected to non-aggressive behavior will trust most humans, and can be handled even by strangers if someone familiar is also alongside. Captive pet macaws sometimes display difficult behavior, the most common being biting, screaming, and feather-plucking. Feather-plucking does not normally occur in the wild, strongly suggesting that it is the result of a neurosis related to life in captivity. Most pet macaws had ancestors living in the wild just two or three generations ago, and are not truly domesticated by any reasonable definition. (This is unlike, for example, dogs; some estimates put the domestication of dogs as far back as 40,000 years ago.) All species of macaws have very powerful, large beaks and are capable of causing considerable harm to both children and adults. They tend to be extremely loud: their voices are designed to carry over long distances. This makes macaws very demanding birds to keep as a household pet. Additional complications arise from the intelligence levels of macaws and their negative responses to stimuli people generally use on domestic pets."

This, combined with "the intelligence levels of macaws" and their ability (like many, if not most parrots') to mimic sounds (including human language sounds), adequately explains why they can give a very convincing imitation of their keeper's human language.

>Unlike N'kisi, I have had no luck with mental communication, and I am extremely skeptical of psychic communicators.

Here is an article on N'Kisi's claimed "mental communication"

[Left: "African gray parrot," Wikipedia]

"N'Kisi may look like an ordinary Congo African gray parrot, but she's the subject of a series of telepathy experiments by a former Cambridge University researcher who says the results are "astounding." "The parrot seems to be able to pick up her owner's thoughts with an amazing degree of accuracy," says Rupert Sheldrake, a former Royal Society researcher at Cambridge and author of Dogs That Know When Their Owners Are Coming Home and Other Unexplained Powers of Animals. N'Kisi's owner, Aimee Morgana of Manhattan, read Sheldrake's 1999 book and contacted him through his Web site ( Morgana, 42, a production designer ... thought Sheldrake might be interested in her parrot. "I thought, 'Gee, if he thinks a dog waiting by the door is interesting, wait till he hears about this.' " She says she first noticed N'Kisi's psychic abilities when she saw an explicit picture in the Village Voice personals. "I was thinking, 'Wow, that's a pretty naturalistic work.' " Then, she says, N'Kisi spoke from the parrot's cage across the room: "Oh, look at the pretty naked body." Sheldrake was interested. He explored N'Kisi's psychic abilities using a double-blind test. He asked Morgana to look at photographs in one room while the parrot was in a cage in another. One camera videotaped Morgana looking at photographs, another camera about 55 feet away videotaped the parrot, who made comments that seemed to correspond to many of the photos Morgana was looking at. In one taped session, for example, Morgana is examining a photo of a woman embracing a man when N'Kisi, who was upstairs in her cage and could not see the photograph Morgana was holding in her hand, calls out: "Can I give you a hug?" According to Sheldrake, N'Kisi made 123 comments during the test sessions, and 32 of those were "direct hits" corresponding to the images Morgana was looking at. The chances of that occurring, Sheldrake says, are less than 1 in a billion. Telepathy is made possible, he says, by the emotional bonds between people and animals. "In the case of N'Kisi, there's a very strong connection between her and Aimee." Morgana spends roughly six hours a day teaching N'Kisi vocabulary words by using a children's touch-tone telephone and other toys, and has transcribed the parrot's vocabulary of 560 words into an electronic log. Sheldrake presented preliminary findings on N'Kisi in November at Cambridge University's Department of Clinical Veterinary Medicine. "The general reception here is he made a well-organized and sensible presentation and reported on work that looks to be competent. He seems to have done a good job," says Donald Broom, a professor of animal welfare at the university. "That's not to say everybody will be completely convinced by it. It's reasonable to be more critical when it's a more unusual result, shall we say." So it's no surprise that the N'Kisi Project, as it's known on Sheldrake's Web site, has been met with skepticism among some scientists." (McKelvey, T., " 'Psychic' parrot expected to ruffle scientific feathers, USA Today, February 12, 2001).

Yet, to me (and I am sure most outside observers) this seems to be just an extension of the same Clever Hans selection effect of humans anthropomorphically reading into animal behaviour, human thoughts and words.

Note that "N'Kisi made 123 comments during the test sessions, and" only "32 of those were `direct hits' corresponding to the images Morgana was looking at" and that with unconsciously giving it the benefit of the doubt (since Morgana and Sheldrake wanted it to be true).

Also note that "Morgana spends roughly six hours a day teaching N'Kisi vocabulary words" and yet it only has a "vocabulary of 560 words" (my emphasis). Yet according to Wikipedia, "a 5-year-old native English speaker" about the same age as N'Kisi was, "knows about 4,000 to 5,000 word families" and "would add roughly 1000 new word families every year to their vocabulary" (my emphasis):

"Vocabulary ... Jean Aitchison gives the capacity of the vocabulary of college graduates with Bachelor of Education degrees as an estimate of at least 50,000, where a word is defined as a lexeme or dictionary entry, i.e., sing, sings, sang, sung count as one entry sing ... According to Robert Waring, a 5-year-old native English speaker knows about 4,000 to 5,000 word families. They would add roughly 1000 new word families every year to their vocabulary. A university graduate will have a vocabulary of around 20,000 word families, in which tolerate, tolerance, intolerable, and toleration are considered as one word family..."

So clearly on the evidence, what these birds are doing is not the same thing as human language acquisition, therefore it is not human language they are speaking, but just imitating human sounds (as they would imitate other birds' sounds) that they have associated with their trainer's audio- visual and behavioural cues, without truly understanding what those sounds mean.

It is the same with my golden retriever dog. She `knows' words like "sit," "walk," "picnic" (for "eat your food"), "swim," etc. No doubt if I was willing to spend "six hours a day" I could teach her to bark the sounds and then I could claim she was `talking'! But I do not imagine that she understands what the words mean. To her they are just sounds that trigger an association with a stimulus.

And so I assume it is the same with your macaw and the other parrots (and apes).

>Yet, I do share a common problem with other audible communicative studies (also signs used by apes): the insensitivity and lack of receptivity of people to sounds they do not immediately understand.
>Please comment about the previous topics. Thank you.

See above. The bottom line is that I, like leading linguist Noam Chomsky, am "highly skeptical" of all these claims of birds (and other animals) speaking (or signing) human language, regarding them all as "merely operant conditioning":

"Alex is an African grey parrot. Since 1977 he has been the subject of a running experiment under animal psychologist Irene Pepperberg, initially at the University of Arizona and currently at Brandeis University. Alex had a vocabulary of around 100 words as of 2000, but is exceptional in that he appears to have understanding of what he says. For example, when Alex is shown an object and is asked about its shape, color, or material, he can label it correctly. If asked the difference between two objects, he will also answer that, but if there is no difference between the objects, he will say `none.' When he is tired of being tested, he will say `I'm gonna go away,' and if the researcher displays annoyance, Alex tries to defuse it with the phrase, `I'm sorry.' If he says `Wanna banana', but is offered a nut instead, he will stare in silence, ask for the banana again, or take the nut and throw it at the researcher. When asked how many objects of a particular color or a particular material are on a tray, he gives the correct answer approximately 80% of the time. ... Although Alex shows understanding of what he says, skeptics question whether he is using language. ... Parrots are highly social birds, and it seems likely that when humans are their companions, they attempt to use the communication system of those humans (language). ... Some in the scientific community, notably Noam Chomsky, are highly skeptical of Pepperberg's findings, pointing to Alex's alleged use of language as merely operant conditioning."


By the way, I don't in any way mean to belittle the love you obviously have for your macaw (nor its love for you). I love my dog and my cat very much (although only the dog really reciprocates!) and when my former cat was run over by a car a few years ago and I had to tell the veterinarian to put him down to end his suffering, I was devastated and cried uncontrollably. But they don't need to be or so something that they are not (e.g. be human, or speak human language) for me to love them as they are.

Stephen E. Jones, BSc (Biol).

Genesis 50:22-26. 22Joseph stayed in Egypt, along with all his father's family. He lived a hundred and ten years 23and saw the third generation of Ephraim's children. Also the children of Makir son of Manasseh were placed at birth on Joseph's knees. 24Then Joseph said to his brothers, "I am about to die. But God will surely come to your aid and take you up out of this land to the land he promised on oath to Abraham, Isaac and Jacob." 25And Joseph made the sons of Israel swear an oath and said, "God will surely come to your aid, and then you must carry my bones up from this place." 26So Joseph died at the age of a hundred and ten. And after they embalmed him, he was placed in a coffin in Egypt.


Anonymous said...

Can't resist adding my two cents here, Mr Jones. :-)

I actually own a personal copy of Irene Pepperberg's book "the Alex Studies", whick I highly recommend you check out as a classic in it's field and a textbook description of how to do this kind of research RIGHT.

I hope nobody will be disappointed if I don't set out all of the reasons why I am convinced her research has accomplished EXACTLY what it appears, and she claims, that it has; but there are literally an essay's worth of what about's, couldn't she's, and couldn't the parrot's that you shoud insist be accounted for before accepting Alex's abillities at 'face value', and I'm not that patient! :-D Therefore, anything I might state could by it's nature only be 'ad hominim'. Trust original research from the horse's mouth, nothing less(certainly NOT press reports! :-P *BLYECH* )

Suffice to say that in context of precautions taxen by Dr. Pepperberg against researcher and and expectation cuing(and the fact that ocassionally the RESEARCHER testing Alex will err(app. 1 in 20, particularly around the time of student exams! :-D), in which case Alex will, unlike when HE is in error, typically REPEAT the correct answer, a response, obviously, contrary to that encouraged by their testing procedures!), and his cross-test, cross-subject, 80% or better success rate(correct first-response, by the way), I simply do NOT believe that Alex could answer questions like "how may three-corner paper" and other such(again, yes they've assured against the allternative explanations popping into your mind;don't trust me, READ HER RESEARCH)without know what 'paper', 'how many', etc. 'mean'. What's he doing when he dissrupts a tes session by pulling on a researchers sweater while saying 'wool' or a tray's green felt liner while saying 'green', if not 'commenting', similiarly to a you-know-what. ;-) And try explaing his success at summing the number of beans hidden under two cups to correctly answer 'how many'(and again, in at least one case, tapping one cup and saying'two' and the other and saying 'three'(bothe correctly, by the way)before being asked the question(he said five, incidently)

And he plays with, practices, and recombines words and phenomes in ways STRONGLY annalogous to a you-know-what's babbling.

Obviously, it isn't fully abstract(Pepperberg herself uses the term 'peri-referrential' as opposed to our fully 'referrential' language), stupendously complex, collonies-of-unexposed-deaf-children-will-SPONTANEOUSLY-CREATE-THEIR-OWN, 'self-created' HUMAN language.
But referrential, abstract, two-way communication?
By the numbers, absolutely.

The question creationists SHOULD ask is why there is such an ENORMOUS gap between the exposure and funding of ape, cetacean, and even corvid research, and that for her; she's accomplished everything and agruably more than Sue Savage-Rumbaugh(though she plays by the rules of good science, so CLAIMS less, and all far FAR better documented and more convincingly demonstrated, yet she's only kept affloat by her FOUNDATION;no funding, and no serious scientists following up on her groundbreaking work. It seems incomprehensible-unless you invoke an agenda, perhaps?

You know, Ape researchers, well, the less objective ones(unlike, say, Sally Boysen)and Linguests often sound when talking about each other, DO often sound more like when Catholics and Protestants have at it that proper scientists. Both become dogmattic and dismissive in arguing that THEIR view understands that man isn't a SPECIAL creation, just different, while the other's, even if unconsciously, doesn't. It can sound VERY unscientific at times. :-D
And this is from an Evangelical Creationist-raised, currently undecided, by the way.

One more thing;I practically idolize Dr. Pepperberg as one of my scientific heroes! :-)


Joel L. Lewis

Stephen E. Jones said...


>I actually own a personal copy of Irene Pepperberg's book "the Alex Studies", whick I highly recommend you check out as a classic in it's field and a textbook description of how to do this kind of research RIGHT.

I don't need to. Pepperberg admits that she would "never claim that Alex has full-blown language" but merely what she calls, "basic building blocks that are language-like behaviors":

"Researchers such as [Steven] Pinker and I get along well because I never claim that Alex has full-blown language; I never would. I'm not going to be able to put Alex on a `T' stand and have you interview him the way you interview me. But Alex has basic building blocks that are language-like behaviors - and also elements of phenomena like consciousness and awareness. Is Alex conscious? Personally, I believe so. Can I prove it? No." (Pepperberg, I., "That Damn Bird: A Talk with Irene Pepperberg," Edge, September 23, 2003)

My main point was, as per my post, that all claimed talking parrots, including Alex, cannot "talk *in the same sense that humans do*", i.e. it does not "know ... what the words *actually mean*"; "it is not talking *as humans talk to each other*".

So Pepperberg and I are in agreement Alex the parrot has, at best, merely "language-like behaviors", not *true* "full-blown language".

>One more thing;I practically idolize Dr. Pepperberg as one of my scientific heroes! :-)

I rest my case!

Stephen E. Jones

Anonymous said...

Not to debate, but simply to clarrify tems;

Are you then in agreement with Dr. Pepperberg that Alex can be demonstrated to understand 'language' at a 'peri-referrential'(her term) level-that is, not 'fully' abstract and arbitrary(i.e. 'KEY in the data' vs. 'KEY to the house' both use the same word), but at the level of basic categories; 'color', 'shape', 'number', etc. Do you also feel his behaviour also suggests acquision of at least simple 'slot-based' rules of grammer (e.g. 'wanna x' and 'wanna go y' and spontaneous combinig of known words in specificic orders suck as 'blue paper' or 'three-corner wood', etc.)? What do you make of his 'creative' coining of 'Banerry', and his sound play? Don't Pepperberg's interpretaions of her data, while explicitly NOT fully human-level language, nonetheless explicitly claim that Alex understands that COLOR 'means' color and MATTER means matter, stands for it in his mind;you are in agreement with this?

As I said, I'm really not looking to draw into a debate here; just looking for clear definnitions of terms. Animal researchers and linguists often talk right past each other,I suspect, and I'm never sure precisely what each side is referring to by 'language'. Both sides, also, sometimes seem to treat discarding one interpretation as consinging data to it's polar opposite, but a word need not have full linguistic abstraction to be far above a parrot mimmicking 'Happy Birthday'-think road signs(Stop, Yield, etc.)for example, or better yet, road signs absent the writing. So where, precisely, are you personally drawing the line? Do you feel with Pepperberg that 'shape' and 'color' 'stand for' those cattegories in Alex's mind, i.e. 'have meaning', and that when Alex employs those terms, most specifically as wnswers during test sessions, at some level he 'knows' what information/concepts the words he is using convey? If not, what does Pepperberg's data really demonstrate instead? If so, then what do you make of her chances if she ever gets around to attempting to teach one or more of her subjects to say, for example, that a nut is 'in the blue cup, on to of the red box', or at least a pidgined but grammatically preservative equivelant? By either light, what, in your opinion, would success in that task signify. He certainly seems willing to comment on things, e.g. 'wool' while pulling on somebody's shirt, 'tray' while tapping a testing tray with his beek, etc., and frequently-compate to Sue-Savage Rumbaugh's sparse results in that area. What's your take on that?

Oh yes; LOL, I knew you'd get me for my last statement!:-D I highly regard Pepperberg because she's remained hardnosed in a field of research that that seems to destroy objectivity-see especially the initially carefull and skeptical Sue-Savage Rumbaugh, and refuses to resort to the cheap shots as some of her colleauges( Again, see SSR's treatment of Pepperberg's work in her article in 'Rational Animals'), despite having more cause and more results to warrant it than perhaps any of her colleauges.

One more thing-I'm posting so long after my first because I thought that my last one failed to post. Glad to see all that work and time typing it wasn't wasted!:-)


Joel L. Lewis

Stephen E. Jones said...


>Not to debate, but simply to clarrify tems; [...]

Aa per my policy stated on my blog's front page: "Comments .. any response by me will usually be only once to each individual on that comment, and then I will let him/her have the last word"; you have now had the last word.

Stephen E. Jones