Undesirable Punishment of Escape Behavior: Further Comments

Abstract

by the fact that both rhe bar and one side of the high-voltage winding of the transformer are grounded. Kleist and Biederman propose that the shock source be isolated from ground. A simpler but slightly more dangerous solution is to isolate the bar from ground. One of us has used Grason-Stadler equipment to shape rats to avoid shock in the barpress situation (see Keehn & Webster, 1968). Although direct shaping of the avoidance response appears to be one way of overcoming the difficulty of establishing bar-press avoidance behavior in rats (Meyer, Cho, & Wesemann, 1960), it is sometimes hard to shape animals to the point of actually depressing the bar. Occasionally, a rat will repeatedly touch and withdraw from the bar without ever actually depressing it. These Ss behave toward the bar as though it were live and it may be that they are punished for pressing during escape trials. It is possible that these animals may brush the upper parts of their bodies very lightly against the bar or wall and that such brushing may produce a considerable shock at the point of contact. Measurements of the electrical resistance of the rat indicate that resistance is usuallv at least three times greater between two rear paws than between a rear paw and a front paw or a rear paw and the nose when slight pressures are applied to the surfaces by the meter probes. Most of this resistance will be in the skin rather than other tissue. Dissipated energy rather than passing current is probably the important characteristic of electric shock. With a constant-current shock generator the energy dissipated in the organlsm is proportional to the resistance of the tissue. Thus, even if an animal is cornplet~ng a circuit between the grid floor and the bar, most of the effect will be at rhe rear paws. However, if the bar were to be touched very lightly the resistance at the front paw or snout could be much higher than the resistance at the rear paws, producing a greater dissipation of energy at the front end of the animal's body. If the resistance is too high, of course, the constant current characteristics of the generator are affected and the above argument no longer applies. That undesirable punishment may occur only with a limited number of Ss does not mean that Kleisc and Biederman's point is unimportant. In programs where all rather than a respectable number of Ss must be trained, it is essential that technical problems of the present sort be eliminated.