On the independence of naming and locating masked targets in visual search.

Twelve Ss were required to name or locate masked letters in arrays containing noise elements. Accuracy in locating was lower when target letters were presented among Vs than when presented among triangles or with no noise. Naming was as accurate when Vs and triangles were presented with the target as when no noise was presented. Since noise did not affect naming in the same way it affected locating, it was concluded that naming is independent of locating. RECENTLY, BARON (1973) AND SNYDER (1972) SUGGESTED that the target in visual search tasks must be located before it can be named; thus naming is thought to be contingent on locating. The experiment reported here tested two predictions derived from this hypothesis. The first prediction was based on the contention that the encoding of a character's location must occur prior to the encoding of its identity. If this is true, locating responses should be more accurate than naming responses. This prediction is readily confirmed in studies where single letters are presented (Dick & Dick, 1969; Logan, in press), but when noise items are simultaneously present in the array, a different picture emerges. When the target and noise are easy to discriminate, locating is more accurate than naming, but when the discrimination is difficult, locating and naming are equally accurate (Logan, in press). In terms of a strong interpretation of the first prediction (i.e., that locating must be more accurate than naming in oil conditions), these latter results speak against the notion of contingent processing. However, a weaker interpretation is possible: the prediction may merely serve to differentiate conditions where processing is contingent from those where it is not. Because of this ambiguity, the second prediction was derived. The notion of a contingency between locating and naming implies that manipulations which impair locating accuracy should produce a similar •This research was supported in part by Grant no. A0127 from the National Research Council of Canada to Albert S. Bregman. The author was supported by a postgraduate scholarship from the National Research Council of Canada. The experiment was reported at the annual meeting of the Canadian Psychological Association, June 1974, Windsor, Ontario. The author is grateful to Jane Zbrodoff and Barry O'Neil for their help in the information analyses. tRequests for reprints should be sent to Gordon D. Logan, Department of Psychology, Queen's University, Kingston, Ontario, Canada.