344
N K Humphrey, G R Keeble
We thought that if the two chambers differed in their sensory characteristics the
monkeys might exhibit a decisive preference, choosing to spend extended periods in
one side rather than the other. But we found it was not so. Whatever the
characteristics of the two chambers the monkeys vacillated endlessly: they would sit
for say 5 s in one side, then up and move to the other side for say 10 s, then back
to the first side, and so on ..., alternating perhaps ten times a minute. This was true
both when the two chambers were very different (so that the monkeys might have
been expected to stay on one side because they preferred it) and when they were
exactly the same (so that the monkeys might have been expected to stay on one side
because there was no point in moving). If the monkeys did exhibit a preference
it was evident only insofar as the average time for which they would stay put in one
chamber depended on the illumination and noise levels; but this time -the average
‘bout’ length—was never anything but brief.
To an observer, watching on closed-circuit television, the monkeys’ behaviour was
intriguing and not easily explicable. There seemed to be nothing stereotyped or
mechanical in the way they moved between the chambers, nor were they wandering
about aimlessly and accidentally entering the tunnel; rather, every move appeared to
be intentional—the monkeys would be sitting apparently contentedly, then suddenly
they would become alert, glance around, and take off quickly through the tunnel to
the other side.
We soon lost interest in studying preference as such and began to ask questions
about the detailed timing of the monkeys’ decisions to change sides. The bout
lengths were by no means constant: on some occasions the monkeys would leave the
chamber within a second or two of entering, on others they would sit for 5, 10, 20, '
or even 40 s. What was the statistical distribution of these bout lengths? What, if
any, simple parameter defined the distribution? How did this parameter—which
might be said to describe the ‘tempo’ of the monkeys’ movements—change as a
function of illumination and noise level?
Being no longer primarily concerned with preference, we kept the situation simple
for the monkeys (and ourselves) by making the two chambers identical. Every one
of the monkeys’ moves could then be interpreted as a decision purely to change sides,
uncomplicated by any possible ‘liking’ for one chamber over the other. We looked
separately at the effects of colour and noise. In experiment 1 both chambers were
illuminated with either red light or blue light, while the noise level was set at 68 dB;
in experiment 2 both chambers were illuminated with white light, while the noise
level was set at either 60 dB or 90 dB.
2 Experimental
2.1 Apparatus and procedure
The apparatus (ground plan shown in figure 2) was designed to create two spatially
identical chambers, there being nothing in the internal layout which would enable the
monkey to tell them apart (and so perhaps show a ‘position preference’). Each
chamber measured 45 cm wide × 80 cm deep × 60 cm high; the tunnel had a lower,
25 cm, ceiling, so that while the monkey could pass through easily on all fours it could
not sit down or turn around once it was inside. The position of the tunnel ensured
that the monkey could not see the space it was entering until it was almost through.
Noise, from a white-noise generator, was provided by loudspeakers in the roof of
each chamber. When the two loudspeakers were set to the same level the sound
density was approximately uniform throughout the testing apparatus. The noise level
was measured with a Dawe sound-level meter with ‘А-weighting’. Some indication of
the subjective loudness of the levels used in the experiments is given by the Dawe
handbook: 60 dB—a quiet office; 68 dB—a large shop; 90 dB—inside a tube train.