BS?

February 14, 2007 6:21 PM

animal telepathy

is this bullshit? http://www.sheldrake.org/nkisi/index.html

Their methods sound pretty solid and their results are impressive. Are they simply liars? Is there something else at play here they aren't being upfront about?

I wouldn't mind believing this was possible, but id like some help scrutinizing this before i go telling my friends about "this cool parrot"

posted by Tryptophan-5ht to science & nature (13 comments total)

Info for skeptics.
posted by cerebus19 at 6:28 PM on February 14, 2007 [1 favorite has favorites]

Corrected Link.
posted by Frank Grimes at 6:33 PM on February 14, 2007

I love: 'psychic' parrot.
Good link Cerebus
posted by strangelove at 6:34 PM on February 14, 2007

"Hits" and "misses" were scored when at least two out of three transcribers verified that N'kisi had said one of the 19 key words used in selecting the images, such as "flower".

If the transcribers knew the keywords, they could be hearing them in random parrot gibberish. Like when people hear ghosts in Electronic Voice Phenomena, or when people have 'talking' dogs. The dogs aren't actually saying "Hellooooo" or "I love Momma", they're barking/howling/whining in a way that sounds like those words to us. If you have a French background, the "I love Momma" dogs can sometimes be heard saying "barbe à papa" if you flip your French/English mental switch.

Talking pets are still ubercute though.
posted by CKmtl at 6:44 PM on February 14, 2007

A lot of these kinds of animal communications studies suffer seriously from the fact that they're not double-blind. That was the problem with the chimpanzee sign-language studies. Washo would be shown a picture and would sign. Some human would view film of it and interpret the sign to decide if it was a semantically meaningful response.

The problem was that the human doing that interpretation knew what the picture was. The right to do that study would have been a human who didn't know what the picture was who tried to interpret the hand gestures -- and then someone else would take that interpretation and compare it to the picture to see if it made sense.

Same thing here. The problem is that the people deciding whether what the parrot says is relevant know what the pictures contain.

There are other problems as well. It's not as easy as you might think to determine whether the number of "hits" exceeds the bounds of chance.
posted by Steven C. Den Beste at 6:46 PM on February 14, 2007 [2 favorites has favorites]

It's just shitty science reporting.
posted by pealco at 6:48 PM on February 14, 2007

In addition to what Steven C Den Beste said ...

I don't know if it was Koko or Washo, but I heard a story that when they had a Deaf native signer watch the films, she said she didn't see any signing. (As I said when I posted this a few days ago, there could be cultural bias there as well, but at least it's an interesting factoid.)
posted by spaceman_spiff at 6:52 PM on February 14, 2007

steven - the parrot feed was transcribed by three people who didn't know what was on the cards.

i think the bigger indictment would be that they discarded 60 trials (because the bird didn't say anything or it didn't say anything on the list) while their conclusions rest on 71 trials. ouch.
posted by Tryptophan-5ht at 6:55 PM on February 14, 2007

There was a different language study done by a different researcher who did not try to use American Sign Language. Instead, he created a synthetic language which was based on pointing to symbols printed on a board. The synthetic language had a regular syntax and there were no words in it which were more than one part of speech.

He did his study working with Bonobos chimps, not regular chimps. There are ways in which the behavior of Bonobos are more like humans than the behavior of regular chimps.

When he talked to his Bonobos, he constructed sentences by pointing in sequence to symbols on his board. They talked back by doing the same thing. I watched film of his experiment and I was completely convinced that they understood what was going on and were truly communicating with language.

For one thing, using the symbol board instead of ASL removed a huge source of ambiguity and possibility for researcher self-deception. When a Bonobos pointed to a symbol it was completely unambiguous. For another, the language structure and vocabulary were well designed to remove puns and ambiguity. And because the language structure was well handled, random strings of symbols would have been syntactically invalid. The Bonobos didn't make those mistakes.

What they were saying when they pointed to symbols on the board made sense in context. They even used them to talk to each other, and what they said to each other also made sense.

Unlike the ASL experiments, which IMHO were complete crap, I thought that Bonobos experiment was well conceived and rigorous, and I found it completely convincing.

(Some quick googling shows that the synthetic language was known as "Yerkish".)
posted by Steven C. Den Beste at 7:49 PM on February 14, 2007 [1 favorite has favorites]

I wouldn't mind believing this was possible

Neither would Rupert Sheldrake. He's got a long history of believing that all kinds of odd things are possible, and carefully designing experiments so as not to disturb that belief.

I would want to see any Sheldrake result scrutinized with an extreme scrute, and replicated independently several times by skilled experimenters with no woo-woo axe to grind, before even beginning to contemplate taking it seriously.
posted by flabdablet at 9:22 PM on February 14, 2007

Until a skeptical scientist reproduces the experiment and results without discovering a normal (not paranormal) explanation, and until another skeptical scientist or ten say "No fucking way!" but then come up with similar results, I would just laugh it off as more loony crap from the type of folk who believe in horoscopes and magic crystals and angels and elves.

But if skeptical scientists do reproduce the results, and it starts to look as if the world might have millions of flying creatures with big claws and sharp beaks that can read your mind, tear your eyes out, and carry your eyeballs back to feed their young, I would remember to think only good thoughts about parrots.

Me, I love parrots.
posted by pracowity at 3:11 AM on February 15, 2007 [1 favorite has favorites]

Could be the observer-expectancy effect, as seen in Clever Hans.

Parrots are pretty smart though.
posted by media_itoku at 6:09 AM on February 15, 2007

This parrot is bogus but Alex is not. Of course, there is no "telepathy" at play with Alex. You can read about Alex in "Animals in Translation" by Temple Grandin and in various Scientific American articles.
posted by avagoyle at 2:19 PM on February 15, 2007