PROBING COLLINEARITY BIAS IN THE CLOSE INDUCTION FIELD

The substrates of the Poggendorff illusion can be evaluated using as f ew as two line segments, a test segment whose collinearity is judged, and an induction segment which serves to bias that judgment. Previous research from this laboratory has shown that a very short induction se gment can produce a substantial bias of perceived collinearity when it is centered at the tip of the test segment, and there is some evidenc e that the symmetry of this configuration affects the strength of this bias. Four experiments were conducted to clarify the issue of symmetr y in the stimulation of the ''half-field'' zones which lie on each sid e of the tip. That Exp.1a showed displacements which moved the inducti on segment to an eccentric position, i.e., occupying a single half-fie ld, reduced collinearity error. With additional displacement which pro duced a gap there was a rebound of induction effect. Exp. 1b indicated that an eccentric induction segment does not produce the usual streng th or pattern of bias as a function of its orientation relative to the test segment. Exp. 2a suggested that the strength of induction effect is a function of symmetry within the half-fields rather than the leng ths of the segments per se. Finally, in Exp. 2b the strength of induct ion was a joint function of the position of a short segment within one half-field and the length of another segment which stimulated the oth er half-field. The effect showed a complex oscillation with changes in position. We discuss the induction mechanism as beginning with ''cont our filters'' which register the induction segment(s) and which combin e their responses on the basis of ''tandem activation'' of the close h alf-fields.