Technique for Studying Adaptation to Disarranged Hand-Eye Coordination

Errors in localization are produced by any optical device that intercepts the line of sight and displaces the retinal image. The viewer's error in reaching for or pointing to an object tends to equal the angular displacement of the retinal image. However, according to many reports this error in localizacion is reduced during prolonged viewing. Helmholtz (2, Sect. 29) was one of rhe first to report that an S shifted his localization after viewing his hand through a prism. In general, Ss have adapted to the disarranged condition provided by mirrors or prisms and more recently by a dosed circuit television system ( 3 ) . Many of these disarrangement experiments, e.g., with mirror drawing, are designed to study the learning of altered skills. Consequently, some variant of that classical paradigm of learning-trial, error, and correction-is often used to explain the observed reduction of error. However, in at least one related experiment ( 1 ) adaptation to the disarrangement of a spatial coordination has been observed although S had no information concerning his actual or potencial errors in localization. This result casts doubt on the adequacy of the trial-and-error explanation; a doubt that is neither raised nor eliminated by the resulrs of most experiments on disarranged coordination simply because their designs incorporate the assumption thac recognition of error is essential to adaptation. In these experiments S, watching his clumsy attempts to mark targets in repeated trials, sees the errors he makes, the same errors measured by E. Clearly the applicability of a trial-and-error theory must be checked under condirions wherein S's recognition of error is eliminated. Measurable errors of hand-eye coordination are obtained by instructing S to make either his hand or his markings coincide with a visible object (target). E can then measure the deviation from coincidence of hand and objecc. S recognizes his errors only when he can see these deviations from coincidence. If he sees neither his hand nor his markings he cannot recognize his errors. To this end he can be instructed to make the felt position of his hand coincide with the seen position of an object. This method of localization, described in detail below, allows E to measure errors in localization which S cannoc view. The adaptation, discussed originally by Helmholtz, occurs only when S does view his hand under the disarranged condicion. This viewing, called here the exposure condition, need include no visible objecc (set of contours) other than the hand. S may see only his hand in an otherwise dark field.