On Sept. 6, 2007, Alex, the famed African Grey parrot, died unexpectedly of a heart arrhythmia in the lab of animal psychologist Irene Pepperberg, PhD. The bird's death marked the end of 30 years of research for Pepperberg—and 30 years of friendship.

During those three decades, Alex and Pepperberg, a psychology professor at Brandeis University in Waltham, Mass., made huge leaps in understanding avian cognition. Alex showed that parrots could learn and apply number, color, material and shape labels to objects. He understood concepts of same-different and relative size. He could recognize some Arabic numerals and understood their connection to number values. He had a basic understanding of the concept of zero.

He also had a temper, didn't like sharing attention with other parrots, threw food at his trainers and would refuse to participate in experiments if they were too boring: a scholarly, feathery prima donna.

Pepperberg, who continues her avian cognition studies with two younger parrots, Griffin and Arthur (aka Wart), recently released "Alex & Me" (HarperCollins, 2008). The book chronicles her experiences training, working with and mourning Alex.

How unique was Alex? Was he particularly smart to be able to learn as much as he did?

I think Alex was of average intelligence. He was one of eight or nine birds in a cage and the fellow who was running the store was the one who picked him out for me. What made Alex unique was the way we trained him. For the first 15 years of his life, he was an only bird. In addition to formal training, we treated him the way you'd treat a toddler. When he wanted food, he might say, "Want bean," so we'd respond, "Here's your bean. This is a green bean. The color is green!" We'd label the colors and shapes and the material of things. And he'd sit there and use creative sound play and come up with new labels. If we could connect something to a particular label, we'd give him that object. After he learned "gray," he said "grain, chain, cane," and we gave him those things. He came up with "banacker," so we figured "banana cracker" and gave him dried banana chips. He didn't like those, though, so that fell out of his repertoire.

What goals did you have for Alex's ability to communicate?

No matter how much training we gave Alex, there was no way he and I could have the same kind of conversation you and I are having. What I tried to do—Ouch, Griffin, don't bite me! I've got a very annoyed little bird here. He's upset that I'm not paying 100 percent attention to him—What I tried to do is look at the world from the eyes of a parrot, in terms of how he might interpret the questions and how he might respond to the experimental design. Our purpose was to teach him enough referential speech so that we could investigate his intelligence using this communication code. I wasn't particularly interested in whether he could create a sentence. Toward the end of his life, he was using parts of labels that he already knew in order to produce new labels. When we first started training him and we introduced a new label, he'd produce the vowels first because they were more like parrot whistles and pure tones, and then only later would he put the consonants in, which were difficult for him. Imagine saying p without lips. But later, he changed his strategy. For example, when he was learning the word "seven," he started out by saying "ssss," then "one." Ssss-[pause]-one. Then he went to sss-none, and then se-ben. That shows a phonological awareness of English speech that nobody would ever expect a non-human to have. It's a very advanced way of looking at vocalizations.

How much has science's understanding of bird cognition advanced since you started your work?

It's changed immensely. If you look at a textbook from the '60s, it's basically showing the apes right at the top, then you get down to rats, then pigeons, and the only thing below pigeons are goldfish. People had the attitude that because birds didn't have any apparent cortical area, they weren't capable of intelligent behavior. When you look at studies on avian cognition since then, for example on foraging behavior or caching behavior, or you look at the research into using birdsong in different contexts or you look at the work done on episodic memory and tool use, it's just amazing [what birds can do]. And we've learned more about the avian brain: It is cortical, extremely cortical. Scientists basically relabeled the avian brain because they found that it's derived from the same [proto-cerebral] structures as are mammalian brains.

How can you tell the difference from when your birds give an incorrect answer and when they're just fooling around?

When they're being difficult, they turn their backs to you or they start preening. Alex might throw everything on the tray onto the floor. Or he'd not only give us a wrong answer, he'd give us sequentially all the wrong answers and repeat them, so that if you statistically looked at the data, he was carefully avoiding the correct answer. We knew that wasn't just a random mistake.

Does your work have any implications outside of animal cognition?

It has implications for adaptations to childhood disabilities. For the birds with whom we have no communication initially, we demonstrate to them what we want them to learn with an interactive two-person modeling technique. This makes it much easier for them to learn. My colleague in California, Diane Sherman, adapted our training technique to work with children with communicative disorders. Particularly autistic children [working with this technique] have ... made significant improvements in their communication abilities. That's something that I'm hoping to explore more down the road.

Was it difficult not to view Alex's behavior anthropomorphically?

I made a conscious effort not to. While we were working together, I tried to interpret everything in a very scientific manner. It was only after his death that this big barrier that I had put up between us, in terms of my scientific objectivity, started crumbling because there wasn't going to be any more science.