Inhibition Accompanies Reference-frame Selection

Spatial relational terms are ambiguous because they can be defined by different and sometimes conflicting frames of reference. Previous research has suggested that multiple reference frames are simultaneously active before a reference frame is selected. Two experiments examined the on-line selection of a reference frame to determine whether it is assisted by inhibition. These experiments used a negative-priming paradigm in which access to a reference frame was assessed on trial n when that reference frame was either available but not selected or not available on trial n – 1. Significant negative priming was observed; it operated along the axis of the reference frame, encompassing both endpoints. In addition, reference-frame selection seems to be independent of object selection. We cast these findings within the view of negative priming as an inhibitory mechanism, and discuss their implications for the use of spatial relations. Spatial relations, such as “above,” are ambiguous because they can be defined according to various and sometimes conflicting sources of information. For example, Figure 1 shows a reclining person looking at an overturned chair. The person is trying to determine where “above” is; the different possibilities are represented by the three boxes. Specifically, Box 1 is above according to an absolute reference frame whose vertical axis is oriented with respect to gravity. Box 2 is above according to an intrinsic reference frame whose vertical axis is oriented with respect to the predefined sides of the chair. Box 3 is above according to a relative reference frame whose vertical axis is oriented with respect to the head and feet of the person viewing the scene (see Levinson, 1996, for formal definitions). Reference frames are applied to a reference object (i.e., the chair) and assist in the search for a located object (i.e., the box). Previous research has shown that multiple reference frames are simultaneously active prior to selection. For example, in a previous study (Carlson-Radvansky & Irwin, 1994), the intrinsic reference frame was dissociated from coincident relative and absolute reference frames by rotating the reference object 90°, as in Figure 1. This resulted in conflicting assignments to vertical spatial relations such as “above.” Significant competition among the dissociated reference frames was observed: Use of one reference frame was more difficult when another indicated a conflicting direction. Another study (Carlson-Radvansky & Logan, 1997) showed that the activation of multiple reference frames is automatic. Despite instructions to use only a specific reference frame, performance was influenced by activation in the noninstructed reference frame. REFERENCE-FRAME SELECTION The simultaneous activation of multiple reference frames is usually followed by the selection of a single reference frame. For example, in one study (Carlson-Radvansky & Irwin, 1993, Experiment 3), participants selected a single reference frame from among two competing ones on 93% of the trials. The current experiments explored whether reference-frame selection involves the inhibition of the nonselected reference frames. This idea is borrowed from studies in visual attention that have shown that selection is most efficient when accompanied by an inhibitory mechanism that reduces the availability of the nonintended items (for reviews, see Fox, 1995; May, Kane, & Hasher, 1995). If referenceframe selection involves the inhibition of the other active reference frames, then some cost should be observed in an attempt to use a reference frame on trial n that was not selected on trial n – 1. Such a cost is commonly referred to as negative priming (Neill, 1977; Tipper, 1985). The current experiments assessed negative priming by comparing responses to experimental and control pairs of prime and probe trials that differed only with respect to the objects used on the prime trials. Consider the pairs of trials shown in Figure 2a. In each box, the small square is the located object and the larger object is the reference object. The task is to judge the acceptability of “above” as a description of the relation between the objects. The experimental prime trial uses a reference object (pocket watch) for which three reference frames are available: coincident relative and absolute reference frames 386 Copyright © 1998 American Psychological Society VOL. 9, NO. 5, SEPTEMBER 1998 Address correspondence to Laura Carlson-Radvansky, Department of Psychology, University of Notre Dame, Notre Dame, IN 46556; e-mail: [email protected]. Fig. 1. Ambiguity of spatial relations. Box 1 is above the chair according to the absolute reference frame based on gravity, Box 2 is above the chair according to the intrinsic reference frame based on the chair, and Box 3 is above the chair according to the relative reference frame based on the reclining viewer.