“The Learning by White Rats of an Inclined Plane Maze”, 1925 (; backlinks):
In the present rat experiment the purpose was to isolate the proprio-ceptive (and intero-ceptive) functions by setting a problem that could be learned only by these. Visual cues were systematically cancelled, auditory clues were nil, olfactory were found by special tests to be inoperative, and tactile were inferentially inoperative (a note on this infra). The one type of stimulus offered as cue for mastering the maze was inclination of the maze from the horizontal, the problem being for the animal to learn to run up-hill instead of down-hill or on the same level.
3 criteria of learning were used: (1) the form of the learning curve plotted for errors, one curve for each of 9 animals, (2) whether the errorless runs exceeded the number that chance alone would have afforded (the maze was a 3-alternative set of paths), and (3) whether in the 200 trials of each animal, pooled but divided into quarters of 50 each, any progressive learning was evidenced by a steady increase in number of errorless runs.
It was found that under (1) 5 animals showed some increase in ability to follow the correct path by its inclination; under (2) 5 animals gave a larger than chance-determined number of errorless runs; and under (3) 3 animals showed some learning. It is concluded that the experiment did not satisfactorily isolate the influence of the proprioceptors, because the semicircular machinery could still be influential; although the author believes that vision played no part, due to the maze being reversible and changeable in its parts, so that the inclination could be kept constant; and because on an incline visual perception would give only an inclination parallel with the animal’s body.
It is suggested that extirpation of the canals does not promise a method of control because such an operation would in itself seriously interfere with locomotion; similarly, anesthetization of proprioceptors or cautery of the post-Rolandic area would seriously interfere with locomotion, at least temporarily.
…What is the particular class of receptors by which the rat learns to react to the inclination of its pathway? With the rotations of maze and of direction of incline all exteroceptive clues were eliminated. (It may seem, at first, as if vision might have aided the animal in the form of a seeing of the higher parts of the maze. This does not apply, however, for the vision involved is that of the rat itself in the rat’s position along the floor of the entrance alley. Particularly in the trials when the true or upward-inclined path lay straight ahead and the animal was already proceeding straight up the incline as it climbed the entrance alley, its body and head position were such as to make the true exit path to be seen as level.)
Differential analysis of the roles played by the intero-ceptive and the proprio-ceptive functions in the learning is at the present stage impossible. Operative elimination of the vestibule and semicircular canals is being considered, but the effects of such excisions upon the animal’s locomotion are so generally abnormal that evidence pro or con is hardly to be expected. Similarly, it goes without saying that any elimination or impairment of the kinesthetic functions will render the subject unable to run at all.