Neuropsychologla, 1974, Vol. 12, pp. 407 to 408. Pergamon Press. Printed in England.

NOTE
NON-EQUIVALENCE OF VERTICAL AND HORIZONTAL PLANES
IN STRIPE DISCRIMINATION BY MONKEYS
R. M. RmL~V
Institute of Psychiatry, London SE5 8AF, England

(Received 15 October 1973)
Abstract--Four monkeys were tested for direct and reversal transfer of a task presented in
the vertical and horizontal planes. Transfer was found within but not across planes.

INTRODUCTION
MONKEYS have been trained to discriminate between stripes oriented at 90 ° to each other when the stripes
were presented in the vertical and horizontal planes [1, 2]. But the plane of presentation of the stimuli is
not necessarily irrelevant to the discrimination. The methods of both direct--and reversal--transfer training
have been used here to determine whether the vertical and horizontal planes are perceptually equivalent.

METHOD
The subjects were four naive monkeys (Maeaca mulatta) weighing between 3.2 kg and 2.7 kg. Testing was
carried out in a modified WGTA, 40 trials being given 5 days each week for peanut reward. The left/right
position of the rewarded test object was determined according to a pscudorandom schedule [3].
The test objects were black-painted wooden cubes (edge 4.5 cm) with parallel grooves (3.5 era × 0.5 cm
× 0.5 cm) milled into one face, leaving an intact border of 0.5 cm on all edges of that face. The cubes were
attached to the lids of the test boxes so that, for each pair of test objects, the grooves were either on the top
or the front face of the cube, and the grooves of one cube were at 90 ° to the grooves of the other. Two complete sets of test objects were made so that replicas could be interchanged during training.
Table 1. Transfer of training between vertical and horizontal plane
Tasks

Animals
A1

1

2

3

Vertical plane

Horizontal plane

Vertical plane

perpendicular
300

parallel
620
(5)
perpendicular
240
(4)
perpendicular
40
(3)
parallel
40
(2)

parallel
180
(6)
perpendicular
120
(8)
parallel
120
(10)
perpendicular
310
(6); 

A2

parallel
90

A3

perpendicular
90

A4

parallel
190

"Parallel" and "perpendicular" refer to the orientation of grooves
to the front top edge of the rewarded stimulus. Figures denote the
learning score to criterion of 90 per cent correct in 100 trials excluding trials in criterion. Figures in brackets denote errors during the
first 10 trials of each transfer test.
407

408

No'rE

The tasks were given in the order shown in Table 1. In Tasks 1 and 3 the grooves were on the front faces
of the cubes, i.e. vertical. In Task 2 the grooves were on the top face of the cubes, i.e. horizontal. The
terms parallel and perpendicular refer to the orientation of the grooves to the front top edge of the rewarded
stimulus, the unrewarded stimulus having the opposite orientation in each case. Task 3 was a direct reversal
of Task 1 for each animal. Task 2 could be considered to be, in a different plane, a reversal of Task 1 for
animals A1 and A2 and a non-reversal of Task 1 for animals A3 and A4.
RESULTS
Table I shows the number of trials to criterion of 90 per cent correct responses in 100 trials, together with
the errors made during the first 10 trials of each transfer test. The scores on tasks which were a reversal of
the immediately preceding task were compared with the scores on tasks which were a non-reversal of the
preceding task for each animal. Neither trials to criterion (t = 2-73, 0.10 > P > 0.05, df = 3), nor errors during the first 10 trials (t = 0-61, P > 0.20, df = 3) reveal a sitmificant difference. Comparing Task 3 with Task
2, for each animal, trials to criterion fail to reveal a significant difference but errors in the first 10 trials of
Task 3 are significantly higher than in Task 2 (t = 3.29, P < 0"05, df = 3).
DISCUSSION
The results suggest that, for the monkey, the vertical and horizontal visual planes are not equivalent since
transfer was not demonstrated from a task learnt in one plane to the same task tested in the other plane. The
significantly larger number of initial errors in Task 3 than in Task 2 can be attributed to the fact that Task 3
was a reversal (in the same plane) of Task 1. There would seem to be some transfer between tasks in the same
plane even when a task in the other plane is interposed between them. Thus failure to demonstrate transfer
from a task in one plane to a task in another plane cannot be ascribed to a failure of these animals to show
any transfer at all. Further, while it might be argued that transfer did not occur across planes because the
animals failed to shift attention between the vertical and horizontal faces of the cube (the faces being in a
different position in space) the fact that some negative transfer was shown in the initial errors in Task 3
suggests that, at least for this task, shift of attention was rapid. If the vertical and horizontal planes were
to be perceptually equivalent under more natural conditions then a certain degree of change of position
would have to be tolerated since it is not possible for an object to change its orientation in space without
most of that object also changing its position.
Assuming that human adults would transfer readily between planes, it would be interesting to determine
when, both phylogenetieally and ontogenetically, such a capacity develops.

Acknowledgment is made to Dr. G. Ettlinger for advice throughout and to The Wellcome Trust for financial
support.
REFERENCES
1. Tlorm, T. J. Reversal and nonreversal shifts in monkeys. J. comp. Physiol. PsychoL 58, 324-326, 1964.
2. McGomGt~, B. O. and J o ~ , B. T. (Personal communication).
3. GEt~EmaA~, L. W. Chance orders of alternating stimuli in visual discrimination experiments. J. genet.
Psychol. 42, 206-208, 1933.
R6sum6---On a test6 quatre singes sur un transfert direct et de renversement d'une 6preuve
pr6sent6e selon les plans vertical et horizontal. Le transfert existait dans un m6me plan mais
non d'un plan/l l'autre.
Ztt~mmen/assung--Bei 4 Affen wurde direktes und reversales Transfer-Training 10¢i vertikal
und horizontal presentierten Ebenen getestet.