Pavlovian analysis of avoidance conditioning in the goldfish (Carassius auratus).

Abstract

Goldfish were trained in a shuttlebox with light as CS and brief shock as US. Performance was measured in terms of "initial response" to the CS (at least one crossing of the hurdle on any trial) and-where the CS was not terminated by the initial response-in terms of "multiple response" to the CS (more than one crossing on any trial). The level of initial responding was as high in classically conditioned animals (shocked on every trial) as in avoidance animals, whether or not the CS was terminated by response, but lower in control animals yoked with the avoidance animals, and lower also in punished animals (shocked only if they responded). Multiple responding was negligible in avoidance animals, but common in classically conditioned animals and in punished animals. The results can be accounted for in purely Pavlovian terms, no reference to instrumental learning being required.

Instrumental avoidance acquisition by goldfish in a Y-maze using explicit and response-contingent cues

“…Although other investigators have found higher responding during the CS-US interval when the US could be avoided, substantial response levels occurred when the US was not avoidable (Scobie & Fallon, 1974;Zerbolio & Wickstra, 1978a, 1978b. In passive avoidance paradigms, although goldfish learned to inhibit shuttling during CS presentations compared with yoked controls (Woodard & Bitterman, 1973), additional work showed comparable response inhibition with no CS at all (Zerbolio & Wickstra, 1978c). With successive presentation of two stimuli, both Woodard and Bitterman (1971) and Zerbolio and Wickstra (1979a) showed that goldfish learned to respond to the US paired signal but not to the unpaired signal.…”

“…just prior to shock were compared with responses in CS priods (Zerbolio & Wickstra, 1979a). In addition, response-contingent CS termination has not been found to play an important role in avoidance acquisition in goldfish (Scobie & Fallon, 1974;Woodard & Bitterman, 1973;Zerbolio & Wickstra, 1978a, 1978b, 1978c, 1979a.…”

“…Thus, change/no-change served as the differential cue indicating whether or not further environmental modification, and response, was required (Bolles, 1970). Feedback explanations to account for conditioned avoidance in the shuttlebox have been offered before (Woodard & Bitterman, 1973) but have assumed intrinsic feedback from the shuttling response itself rather than feedback from a differential external cue. The intrinsic feedback hypothesis cannot account for the present data because such feedback would occur on every trial with an initial response and would not explain how or why the goldfish swims directly to one cued alley over another.…”

“…But the only response consequence in that study was US omission when the goldfish swam to the S+ alley. One of the variables traditionally thought to affect avoidance is cue termination (Scobie & Fallon, 1974;Woodard & Bitterman, 1973;Zerbolio & Wickstra, 1978a, 1978b. The first experiment examined the role of response-contingent S+ and/or S-termination along with US omission on goldfish Y -maze discriminated avoidance acquisition.…”

Classical conditioning of an “avoidance” response in goldfish using a linear presentation procedure

“…Woodard & Bitterman, 1973). Zerbolio (1981) used a control group in which the light colors of the CS+ and CSwere made the same so they became indistinguishable.…”