Robotics Reference
In-Depth Information
The scanning process was linked with the tortoise's steering mecha-
nism in such a way that the photo-electric cell was always looking in the
direction of movement, so that when a light was detected in any direc-
tion the tortoise could respond without too much manoeuvring. The
photoelectric-cell and motors were connected in such a way that when
an adequate source of light was detected, the tortoise would turn towards
it and approach it. When the tortoise was moving towards an attractive
light source and met an obstacle on the way, the attraction of the light
would be reduced so the tortoise would lose interest in the light until
after it had dealt with the obstacle. In addition to avoiding bright lights,
the tortoise also disliked steep gradients and all but lightweight physical
obstacles. And while it would go around heavy objects, it could also push
lightweight obstacles out of the way by a process of repeated butting and
withdrawing.
Walter's tortoises possessed some measure of self-recognition. They
were each fitted with a flash-lamp bulb on the head, which was switched
on whenever the photo-electric cell received an adequate light signal.
When a tortoise encountered a mirror or some other reflective surface, its
photo-cell could “see” a reflection of this light signal, making the tortoise
head towards its own reflection. But as it did so this light source appeared
brighter, causing the tortoise to move away from it. As the tortoise got
further away the intensity of the reflected light was reduced below the
tortoise's “I don't like it” threshold and, once again, the tortoise started
to head towards the reflected light. If such behaviour was observed in
an animal it would be accepted as being caused by some degree of self-
awareness. In this respect Walter's tortoises appeared to be exhibiting an
intelligence superior to that of those animals who treat their own reflec-
tion as though it is another animal.
Thus Walter had produced a kind of “social behaviour” in which the
tortoises “danced” around each other, in movements of attraction and
repulsion, which Walter likened to mating. This prompted him to write
I noticed uncertain, randomic, free-will or independent character-
istics [which constituted] aspects of animal behavioral and human
psychology. . . Despite being crude they conveyed the impression
of having goals, independence and spontaneity. [1]
Walter considered his tortoise to be a new “species” of animal and gave
the species the Latin name
Machina Speculatrix
because of their explora-
tory, speculative behavior. In a second series of experiments, Walter cre-
