Sir

I take exception to Clive Wynne's Review of my book Alex & Me: How a Scientist and a Parrot Discovered a Hidden World of Animal Intelligence — and Formed a Deep Bond in the Process (Nature 455, 864–865; 2008). As a memoir for a general audience, it does not contain in-depth technical detail. But Wynne's questioning of the underlying science is answered by my publications in the peer-reviewed literature — one of which won the Frank A. Beach comparative psychology award (I. M. Pepperberg and J. D. Gordon J. Comp. Psychol. 119, 197–209; 2005).

Detailed controls were included against inadvertent cueing during 30 years' study of the cognitive and communicative abilities of Alex, the African grey parrot; during testing, I never knew what Alex would be asked to identify in any trial. Because tests covered several different topics, neither Alex nor evaluators could target a particular subset of labels. Individuals never tested Alex on the labels they taught.

Wynne writes that the human ear is easily tricked, suggesting that Alex's responses were often not actual human labels. But incoherent responses were discarded as erroneous, and inter-observer reliability trials (with individuals from a different university listening to a video sound track) gave agreement above 90%. Sonagraphic analysis showed that Alex's vocal patterns closely matched his human trainers' and were not “jumbled speech sounds”.

Alex's idiosyncratic labels are overlooked in Wynne's review. The bird would use these irrespective of how a colleague or stranger might cue him otherwise. For example, toy cars and squares were always classed respectively as “truck” and “four-corner”.

Replication failure was not a problem. Published data on my younger subjects detail a level of replication. Birds were presented with a barrage of training techniques, mostly different from those used with Alex, to determine which aspects of Alex's training engendered his success. These birds, with a quarter of Alex's exposure to the single, successful method, could not, by definition, achieve as much as Alex during the same time frame.

Wynne has no basis for implying that my methods might be flawed — other than a possible inherent scepticism, which, ironically, is a common bias discussed at length in the book.