Pepperberg

Acquisition of the same/different

concept by an African Grey parrot
(Psittacus erithacus)
Learning with respect to categories of color, shape, and material.
Background: General
Studies have shown that at least one avian subject, an African grey
parrot, can exhibit capacities once thought to belong exclusively to
humans and possibly non human primates only.
Wolfgang Kohler, is well
known for his studies
on insight learning
using chimpanzees.
Insight learning occurs
when one suddenly
realizes how to solve a
problem.
Background: General
Kohler showed the power of Kohler found that the chimpanzees
insight learning by placing a spent most of their time
banana above the reach of unproductively rather than slowly
chimpanzees and watching working towards a solution. They
how they attempted to reach would run around, jump, and be
the food. In the room there generally upset about their inability
were several boxes, none of to reach the snack until, all of a
which was high enough to sudden, they would pile the boxes
enable the chimpanzees to on top of each other, climb up, and
reach the banana. grab the bananas.
Background: General
In order to identify items as same or
Chimps and gorillas have been different, animals must recognize
found to be able to inter- the category shared by items, and
communicate. Studies have realize that sameness can be applied
found that non-human to other, unfamiliar objects; this
primates have the ability to requires the skill knows as a
express abstract ideas. The symbolic representation.
cognitive capacity of Evidence has been found that many
conceptual conceptualisation bird species have the ability to
has been thought to persist in recognize same/different items, this
other species. being an advantageous benefit for
the survival of birds.
Learning theory: Operant conditioning
Skinner’s Rats

Operant conditioning is a method of learning that occurs through

rewards and punishments for behavior.

The work of Skinner was rooted in a view that classical conditioning was
far too simplistic to be a complete explanation of complex human
behavior. He believed that the best way to understand behavior is to
look at the causes of an action and its consequences. He called this
approach operant conditioning.
Learning theory: Operant conditioning
Skinner’s Rats
Skinner (1948) studied operant conditioning by conducting experiments
using animals which he placed in a ‘Skinner Box’.
Skinner showed how positive
reinforcement worked by
placing a hungry rat in his
Skinner box. The box contained
a lever on the side, and as the
rat moved about the box, it
would accidentally knock the
lever and Immediately deliver a
food pellet.
Learning theory: Social learning theory
Bandura Aggression

In social learning theory, Albert Bandura agrees with the behaviorist

learning theories of classical conditioning and operant conditioning.
However, he adds two important ideas:

1. Mediating processes occur between stimuli & responses.

2. Behavior is learned from the environment through the

process of observational learning.
Learning theory: Social learning theory
Bandura Aggression

Children observe the people around them behaving in various ways. This
is illustrated during Bandura’s famous Bobo doll experiment.

Individuals that are observed are called models. In society, children are
surrounded by many influential models, such as parents within the
family, characters on children’s TV, friends within their peer group and
teachers at school. These models provide examples of behavior to
observe and imitate, e.g., masculine and feminine, pro and anti-social,
etc..
Aim:
To find whether a parrot is able to use vocal labels to demonstrate
the symbolic comprehension of the concepts ‘same’ and ‘different’.

Symbolic comprehension: The ability to use abstract

representations to show understanding.
Subject: Alex
This parrot (Alex) has learned to use English vocalizations to
identify, request, refuse, or comment upon more than 80 different
objects of various colors, shapes, and materials.

He has demonstrated a rudimentary capacity for categorization can

identify quantity for collections of up to 6 objects and has functional
use of phrases such as "come here," "no," and "I want X" and
"Wanna go Y, " where X and Y are appropriate objects or location
labels.
Same/Different:
One cognitive capacity in particular, comprehension of
same/different, has been singled out as a concept not typically
attributable to non primates.

There is significant debate in the literature concerning exactly what

has been demonstrated in the various animal studies purporting to
examine same/different.
Same/Different:
Zentall and his colleagues have demonstrated that pigeons (Columba
livia) can show a significant degree of same/different concept transfer,
including use of symbols to represent same versus different.

Starlings (Sturnus vulgaris), cowbirds (Molothrus ater), and

mockingbirds (Mimus polyglottos) appear capable of classifying novel
series of tones as ascending or descending- i.e., "same" or "different"
from ascending or descending reference series-but only for sequences
that fall within the range of frequencies used in training.
Same/Different:

The present study was therefore undertaken to see if an avian subject,

the aforementioned African Grey parrot, could use vocal labels to
demonstrate symbolic comprehension of the concepts of same and
different.
Participant:
The subject, an African Grey parrot named Alex, had been the focus of
a study on interspecies communication and cognitive abilities since
June 1977.

He was allowed free access (contingent upon his vocal requests; e.g.,
"Wanna go gym") to all parts of the laboratory for the 8 h/day that
trainers were present; hence, trials occurred at various locations.
Water and a standard seed mix for psittacids were available
continuously; fresh fruits, vegetables, specialty nuts and toys were
provided at the bird's vocal request (e.g., "I want cork.")
Procedure:
Alex was to be presented with two objects that could differ with
respect to three categories: color, shape, or material (e.g., a yellow
rawhide pentagon and a gray wooden pentagon; a green wooden
triangle and a blue wooden triangle).

He would then be queried "What's same?" or "What's different?" The

correct response would be the label of the appropriate category, not
the specific color, shape, or material marker, that represented the
correct response (e.g., "color," not "yellow").
Procedure: Training sessions
The primary technique, called the model/rival, or M/R, approach, has
humans demonstrate to the parrot the types of interactive responses
desired.

One human acts as the trainer of a second human by presenting

objects, asking questions about these objects, giving praise and
reward for correct answers, and showing disapproval of incorrect
answers (errors similar to those being made by the bird at the time;
e.g., "wood" for "green wood").
Procedure: Training sessions

The second human acts both as a model for the bird's responses and
as a rival for the trainer's attention. Roles of model/rival and trainer
are frequently reversed, and the parrot is given the opportunity to
participate in these vocal exchanges.

Training to acquire "color" and "shape" as labels separate from

the questions "What color?" and "What shape?" (vocalizations
acquired incidentally from their use by trainers in a different
study) took 4 months; training to acquire "mah-mah," a totally
novel vocalization, took 9 months.
Procedure: Training sessions
To lessen the possibility of trainer-induced cuing trials were
conducted by secondary trainers who never trained Alex on
same/different.
• A secondary trainer would present to the bird, in a variable but
previously determined order, the objects to be identified.

• The parrot was thus shown an exemplar or number of exemplars,

asked "What's this?" "What color?" "What shape?" "How many?"
"What's same?" or "What's different?" and was required to
formulate a vocal English response from the 80 possible
vocalizations in his repertoire.
Procedure: Data collection
"What shape?" and "What color?" were used during data collection
for categories, or as a second question when Alex responded to
"What's this?" by identifying correctly only the material of a colored
or shaped object.

The principal trainer repeated out loud what she heard the bird say. If
that was indeed the correct response (e.g., the appropriate category
label), the parrot was rewarded by praise and the object(s).
Procedure: Data collection
If the identification was incorrect or indistinct, the examiner removed
the object(s), turned his/her head (a momentary time-out), and
emphatically said "No!"

The examiner then implemented a correction procedure in that the

misnamed object or collection was then immediately and repeatedly
presented until a correct identification was made; errors were
recorded.
Results:
Familiar objects: objects seen prior to training.

Alex's score was 99/129

= 76.6% for all trials
(first trials plus
correction procedures).
69/99 = 69.7% on first-
trial-only performance.
Results:
Novel objects: objects never seen before.

Alex's score was

96/113 = 85% on all
trials.
79/96 = 82.3% on
first-trial
performance.
Conclusion:
• Parrots are able to demonstrate the understanding of the
symbolic concept of ‘same’ and ‘different’

• Parrots are able to respond to verbal questions to vocalize

categorical labels

• Alex was able to score higher with the novel objects compared
to the familiar ones. This doesn’t make sense, but Pepperberg
said that ‘The results here were not surprising’. He argues that
because they were new objects and they were the reward, Alex
would be more naturally inclined to pay closer attention to
them.
Evaluation:

GRAVE
Evaluation:

GRAVE Generalisability

This study has low generalisability as the sample size is only one and
specific to only one breed of parrot.
Note: It has subsequently been found that other breeds have these
abilities also.
Evaluation:

GRAVE Reliability

This study’s reliability is very good as they used a standardised measure

which was designed to ensure that Alex didn’t see anything that he had
already seen and in combination. All the situations were novel, so Alex
could not rely upon memory. This is an excellent example of the care and
effort required by the researchers.
Evaluation:

GRAVE Application

Supports the idea that you could train animals to behave in certain ways
in a certain context. This could be as mundane as training useful
responses to comands. Potentially this could lead to behavioural learning
strategies for when animals are ready to be released into the wild.
Evaluation:

GRAVE Validity

Validity is good in this study as it was a case study and focused on one
participant only.
Lab study, so there was a high standardisation of tests which leads to
higher reliability.
But, would birds be exposed to these shapes in nature?
Evaluation:

GRAVE Ethics

Ethics are high as there was only one participant, rewards given, no
deprivation and appropriate caging was provided. These are all
considerations of the ethics for working with animals.
Note: If you believe that the ethics outlined for working with animals is
not comprehensive enough, you could argue that the ethics are not so
great.
Evaluation: Other considerations

Operant conditioning :
Some in the scientific community are skeptical of Pepperberg's findings,
pointing to Alex's communications as operant conditioning. Critics point
to the case of Clever Hans, a horse who could apparently count, but who
was actually understanding subtle cues from the questioner. In another
case, Nim Chimpsky, a chimpanzee, was thought to be using language,
but there is some debate over whether he simply imitated his teacher.
Evaluation: Other considerations
Operant conditioning :
Herbert Terrace, who worked with Nim Chimpsky, says he thinks Alex
performed by rote rather than by using language; he calls Alex's
responses "a complex discriminating performance", adding that in every
situation, "there is an external stimulus that guides his response."

However, supporters of Alex say he was able to talk to and perform for
anyone involved in the project as well as complete strangers who
recorded findings, though such interactions do not involve the strict
conditions required to exclude rote and operant responding.
 More resources for CIE A/S
Psychology (9990) can be found at:
www.mrgregoryonline.com
Research methods Biological
Issues and debates Cognitive
Approaches Learning
Statistics Social