Reafference learning in the presence of exafference

Abstract

Reafference learning has been demonstrated most clearly in the case of position-eonstancy adaptation in which the only stimulus change is caused by the subject's own movement. The present study used the more ecologically representative training situation in which only part of the stimulus change is caused by the subject (reafference), while part of it is caused by an independent source (exafference). The exafference varied the space relation between subject movement and optical movement or the time relation between those two. In both cases, reafference learning was not affected by the exafference, and the subject's varied training experience resulted in a fixed expected optical movement and a fixed expected time lag.In reafference learning, a person learns to expect a certain stimulus change whenever he makes a certain movement (Held, 1961;Welch, 1978). For example, if the person is trained with a diagonal optical movement paired with vertical head turning, he/she comes to require a portion of that diagonal optical movement for an object to appear stationary during that head movement. This specific phenomenon is called position constancy adaptation and is the best exemplification of reafference learning (Hay, 1968(Hay, , 1974Shebilske, 1977). However, the ecological representativeness of the usual laboratory procedure is doubtful. In the real world, other things besides the observer move, and hence there will be a less than perfect correlation between subject movement and optical movement. Externally caused components of stimulus change are called exafference, in contrast with the subject-caused component called reafference (Holst, 1954). The question of the present study is: What happens to reafference learning when exafference is present, when there is an imperfect correlation between the person's movement and stimulus change?The strategy of this study is to see how two of the known properties of position constancy adaptation change when exafference is present during the training situation. Hay (1968) made such a preliminary investigation. Hay and Goldsmith (1973) have found that the adaptation state has two components: an expected optical motion (EOM) and an expected time lag (ETL) between the head movement and the This research was supported by National Science Foundation Grant GB-33449 to the University of Wisconsin-Milwaukee, with the author as principal investigator. A qualitative summary of the results was given in Hay (1974) and was referred to in Welch (1978). optical movement. We shall therefore introduce exafference into both of the corresponding components of the training situation: the optical-motion component (Experiment 1) and the time-lag component (Experiment 2). EXPERIMENT 1A luminous spot in an otherwise dark room is made to undergo horizontal oscillation when the subject nods his head up and down. The resulting diagonal optical motion (OM) relative to the head constitutes the training reafference. Exafference is added to this in the form of a horizontal, vertical, or combined horizontal-and-...

A conditioned weight illusion: Reafference learning without a correlation store

“…For reviews of the controversy, see: Harris (1965), Kornheiser (1976), andWelch (1978). We agree with Hay, for instance, that eye/hand coordination differs significantly from visual position constancy, VPC, the phenomenon that provides perhaps the "best exemplification of reafference learning" (Hay, 1981). Hay's VPC research employs a method similar to that conceived by Wallach and Kravitz (1965).…”

“…However, Hay continues, paradoxically, to advance his data as evidence for Held's hypothesis. Hay (1981) writes, "In reafference learning, a person learns to expect a certain stimulus movement (Held, 1961;Welch, 1978)." This statement is consistent with Held's hypothesis but inconsistent with Hay's data (cf.…”