“Further Analysis of the Conditioned Reflex Method in Relation to the Experimental Neurosis”, Howard S. Liddell, George F. Sutherland, Richard Parmenter, Quin F. Curtis, O. D. Anderson1937-04-21 (psychiatry, animal psych)⁠:

Our experiments have proceeded upon the hypothesis that the experimental neurosis so frequently develops in the course of conditioned reflex experiments because the conditioned animal cannot, through procrastination or evasion, avoid making difficult decisions.

Systematic variations in the amount of neuromuscular freedom permitted the sheep and pig during conditioning to food and shock have been related to the nature and predictability of the animal’s conditioned responses as well as to features of behavior indicative of ‘tension states’ before the onset of the experimental neuroses.

As motor outlets (such as are employed in locomotion, opening of the food box, etc.) are blocked signs of nervous tension appear while the conditioned responses become stereotyped and predictable.