What’s Happening in Research
- The Great Manipulators
In everyday conversation, the word “manipulate” often conveys a taste of the conniving, a sense that someone is trying to control the outcome of a situation. Contrary to our contemporary usage, modern dictionaries define the primary meaning of manipulate as “to move, arrange, operate or control…by hands or by mechanical means especially in a skillful manner.” Its secondary definition is more in line with our daily use of the word; the verb is explained as “to influence or manage shrewdly.”
The parrots of The Alex Foundation show behaviors that express both definitions of the word. They can manipulate objects AND people in a very remarkable manner. Dr. Pepperberg’s paper, “‘Insightful’ string-pulling in Grey parrots (Psittacus erithacus) is affected by vocal competence” (Animal Cognition, Springer-Verlag, 2004), explores the ability of African Greys to obtain objects through the actions of physical manipulation and vocal communication with humans.
In 2003, Dr. Pepperberg began a study of African Grey parrots with a method that had been used to assess “insight” in other avian species, such as ravens. Birds were evaluated on their ability to retrieve food suspended from a string or cord, which was hung below a perch. To obtain this food, the bird, who had no previous training at this task, would stare for a few moments at the apparatus. The bird would then spontaneously reach down, pull up a loop of string onto the perch, secure the string by stepping on it, and then repeat this sequence numerous times. If the bird was successful in the attempt to obtain the food, an “intentional means” was indicated, in that the bird was performing behavior intentionally to accomplish the end behavior of attaining food. This “intentional means-end behavior” suggested a higher-order ability to reason, an ability that is not common among species or even within a given species.
Four parrots participated in Dr. Pepperberg’s work. The birds—Kyaaro, Wart, Alex and Griffin--had varying backgrounds of training, backgrounds that would have telling results on the outcome of the study. The string-pulling abilities of the birds were tested during different time periods with Kyaaro and Alex being tested first in 1995. (Kyaaro, who developed an avian form of Attention-Deficit-Hyperactivity Disorder, has been retired and is now living happily with Maggie Wright of the Grey Play Round Table and her two female Greys.) Alex would be tested again when Griffin and Wart had their first exposure to string-pulling in 2003.
--Kyaaro, almost five years old in 1995, had had four years of training, limited mostly to unsuccessful video and audio input. (Training a parrot through video or audio exposure to words and phrases repeated over and over, does not allow them to learn actual names or labels of objects, as does the model-rival technique). Consequently, Kyaaro’s vocabulary consisted of only a few object and number labels.
--Wart, four and a half years old in 2003, had been trained for three and a half years, mostly in animal-computer interaction. He could use a few labels, such as “want some,” a phase he would utter when a trainer had something he wanted. However, he did not specify the object he wanted the trainer to retrieve.
---Alex, 27 years old in 2003, had received 26 years of human modeling training. At the time of testing, he possessed over 50 object labels, seven colors, five shapes, numbers up to “six,” three categories and many functional phrases. These functional phrases consisted of “I want (object),” “Wanna go (location),” “Come here,” and the command, “Go pick up (whatever!).” Prior to the 2003 study, he had experienced one trial session of string-pulling eight years earlier, in 1995.
--Griffin, eight years old in 2003, had been trained with two methods--the same method used to teach Alex, and also the video training method used with Kyaaro. The training by video experiments caused Griffin the same lack of success as Kyaaro. As a result, Griffin did not have as large of a vocabulary as Alex. But because he also had quite a bit of model-rival training, Griffin did use some of the same labels, commands, and functional phrases as Alex.
Kyaaro, Wart, Alex and Griffin were tested individually to establish their “string-pulling” abilities. Instead of strings, chains of plastic links were used to suspend the target objects. The trials of testing occurred months apart to avoid “massed-trial learning,” in other words, to prevent the birds learning from simply being exposed repeatedly to the same challenge. In Alex’s case, eight years had passed since his initial exposure to string-pulling.
For three of the birds, almonds were offered as the desired object, interchanged occasionally with a piece of chalk. The chalk was used to determine if the type of reward affected the bird’s behavior, because in a bird’s world, chalk is apparently is not as highly coveted as those delicious almonds! Kyaaro’s favorite object was a bell, so for his testing, the bell became the object to obtain. Each bird was placed on a T-stand perch, and after the targeted item had been suspended, the trainers would simply point to the object. If the bird did not seem interested, the trainer would ask “Pick up nut/bell/chalk/treat.”
The results and implications of the study were, to say the least, jaw-dropping. African Grey parrots are astoundingly intelligent, but this study showed just how cleverly they adapted those smarts to the situation at hand.
Both Kyaaro and Wart immediately performed the actions of pulling, stepping and repeating these steps to attain the much-wanted treat. Without hesitation, they executed the actions consistently and with perseverance. After obtaining the almond or bell, the birds enjoyed them as their “treats.” If given a choice between the chalk and the other treat, they always chose the favored treat. During subsequent trials, when chalk was the only object available to obtain, the birds would still retrieve the chalk they saw hanging below them, but would drop it once they had it on their perch. It’s almost as if they enjoyed the challenge of retrieving the chalk as much as they enjoyed the ultimate treat of an almond or bell.
The behavior of Alex and Griffin came from the opposite end of the spectrum—neither of the birds made any attempt to retrieve the nuts. When confronted by the almond hanging below, both Alex (as he had done in 1995) and Griffin looked at the nuts, looked at the trainer, and said, “Want nut.” When the trainer said, “Go pick up nut,” they calmly replied, “Want nut.” This verbal interplay went back and forth during each trial. But Alex took things one step further. During one session, Alex’s demands of “Want nut!” got louder and more intense with the trainer’s failure to comply with his requests.
Dr. Pepperberg states, “The most noteworthy result of the…experiment was that the two parrots with limited vocabularies immediately acted out the correct physical task to obtain their treats, whereas the parrots that had received considerably more effective training in referential English speech attempted instead to manipulate their trainers.”
After their first lack of cooperation in the 2003 string-pulling trials, Griffin and Alex were allowed to observe one of Wart’s successful sessions. But even this did not encourage them to engage in getting those nuts on their own. They ignored the almonds and they ignored the chalk. In fact, in this session, they would not even ask the trainer to get the nut for them. And one has to ask “Why?”
Dr. Pepperberg addresses this, writing, “Determining their actual reasoning is not possible…” But she does offer some potential explanations--one theory is that the birds had learned the trainers would not assist them in this situation, so making a request of the trainers would be useless.
The study does imply an astounding phenomenon which expands the concept of “intentional means-end behavior.” It is this—if a bird’s higher cognitive abilities are established by the ability to manipulate a device to obtain food, what are the implications of a bird’s request to have a human attain it for him, rather than work for it himself?
The behavior of Alex and Griffin, in demanding their trainers fetch their food, could indicate an even higher order of intelligence, in that they know how to manipulate another individual to access their wants. Before Dr. Pepperberg’s 2003 study, there was no prior evidence to suggest that a bird would respond first by the manipulation of others, rather than use a mechanical apparatus.
Dr. Pepperberg explains, “They (Alex and Griffin) were requesting that the trainers do something for them in a very specific and direct manner. They were not treating humans as a physical object to be used (such as a stepping stone to reach something desired) but were engaging in deliberate communication as a problem-solving strategy, a fairly advanced stage of development, even for human infants.”
Those detractors who deny the possibility of inter-species communication will find any reason to reject the validity of the birds’ behavior. But there is one aspect even the disbelievers cannot deny—these birds are great manipulators, whether they are string-pulling to obtain the objects of their desire, or if they are pulling strings to manipulate humans to satisfy their wants.
© 2005 by The Alex Foundation. All rights reserved. The Alex Foundation Newsletter or parts thereof cannot be reproduced in any manner without permission of The Alex Foundation. The Alex Foundation, a non-profit 501c.(3) organization, is based in Waltham, MA.