[David Bailey]

I get the impression that lots of sceptics - here and in other places - have made unsuccessful attacks of Rupert Sheldrake's work - described at his site. He clearly knows his statistics and has a solid background in conventional research. Rupert picks experiments where the effect is pretty strong - so strong that people have already noticed it in an informal way. The chances that he makes fundamental statistical blunders seems to me to be really remote.

In the case of the N'Kisi experiments, it is surely valid to ask the question "When the parrot vocalises an actual word which is not 'camera', does he get a hit more times than chance would predict?

Suppose you had a greedy parrot that said 'peanut' frequently, are we really saying that such a parrot could never be tested for telepathy!

What would be unacceptable would be to trawl the data for arbitrary manipulations that might satisfy a statistical test. Since I am pretty confident that Rupert knows his statistics, accusing him of doing that would be tantamount to accusing him of fraud.

David Smith - I have seen remarkable film of another parrot that could respond to request such as "Pick up the red square". The idea that their language is devoid of understanding seems to have been blown away.

David