T-maze learning with one trial per day using brain stimulation reinforcement.

Abstract

Electrical stimulation of the brain is an effective reinforcer in simple learning situations when trials are separated by as much as 15 min. In this experiment 13 rats with electrodes in lateral hypothalamus were trained on a position discrimination with trials separated by 24 hr., and subsequently on reversal problems. 11 Ss learned the original discrimination; 6 of 10 /Ss tested completed 1 reversal; and all 6 tested Ss completed a 2nd reversal. 1 S was tested further and completed 7 reversals. Clearly, in the successful cases, reinforcing properties of electrical stimulation were sufficient to induce learning of a discrimination problem even though trials were separated by 24 hr.

Reward value of brain stimulation is inversely related to uncertainty about its onset.

“…Gallistel (1969) has shown that the avidity with which rats seek ESB is a function of the recency of the last ESB. Kornblith and Olds (1968) and Kent and Grossman (1969) found that individual rats differ widely in the ability to tolerate, i.e., perform consistently, despite long temporal interruptions between ESBs. That not all rats prefer unsignaled ESB to no ESB, then, is merely a reflection of the difference among rats (or electrode placements) in their tolerance for intervals averaging 1 min.…”

Intracranial reinforcement: Performance decrements as a function of stimulus parameters

“…Evidence contradictory to Deutsch's ideas of intracranial reinforcement is plentiful: (a) discrete trial performance decrements may not be present (Wasden, Reid, & Porter, 1965;Wasden, 1966;Scott, 1967;Wasden & Reid, 1968), nor do all investigators fmd it necessary to start their Ss performing with "free trains" of ICS (Kornblith & Olds, 1968), (b) training for ICS on lengthy temporal schedules of reinforcement has been accomplished (Pliskoff, Wright, & Hawkins, 1965), and (c) extinction may not be rapid nor time-dependent, especially when the contingencies of reinforcement are similar to the circumstances of extinction with conven-Psychon. Sci., 1969, Vol.…”

Self-Stimulation

“…Kornblith and Olds (1968) showed that one could nonetheless get rats to learn a Τ maze for BSR, giving only 1 trial per day. Performance at the start of each day's session (after about 24 hours without stimulation) is slow.…”