questionable reputation, who would think that pigeons have it over
rats when it comes to the cleanliness quotient?
Greg Jensen, working with
professor Allen Neuringer on his psychology thesis, is using these
feathered subjects to study the effects of number of responses
on ability to emit
random responses—is it easier to be random on two choices
than eight because it is a “simpler” decision?
The pigeons are put in a box (operant chamber) that has had
its back walls removed and replaced with a computer
screen equipped as a touchscreen. They peck at squares that appear
on the screen in
a circular arrangement around a center point. The computers that are
attached to these screens are programmed to reward behavior
that meets a randomness
contingency designed by Jensen.
In setting up
this experiment, he has been truly inventive by using available technology
to do more with less. The computers are several years old, the
operant boxes are more than a decade old, and the language being
used to code the boxes
is an archaic version of the programming language TrueBasic. But
this anachronistic equipment works because it supports the key software
for the experiment,
preventing Jensen from having to write new programs and allowing
him to spend more time
on housekeeping—for the pigeons, that is.
He admits that animal subjects “are not renowned for their kind, clean
treatment of electronics, though the pigeons have been quite a bit cleaner
than the rats I used in earlier years.” Between the feathers and the
dander, it is necessary to clean out the circuitry on a regular basis, and
it is often “a grueling hunt for the problem in which any of a dozen
individual things might be wrong.” If time allows, Jensen hopes
to run a version of this experiment with humans standing in for the
. . .
and one would hope that they will be lower-maintenance subjects.