Date of Award


Document Type


Degree Name

Master of Arts (MA)




Potter (1975, 1976) has shown that visual masks can exert effects both at the level of the icon and at higher levels of processing. Her methodology involved presenting pictures in a sequence at very fast rates, in order to mimic naturalistic saccadic viewing. The present study used similar methodology in order to investigate whether a picture which violates an expectation about pictures in that series exerts more of a masking effect than a picture which coincides with the expectation.

Each of 38 subjects was shown 36 sequences of eight pictures. In nine trials, every picture in the trial was drawn from the same category. In 27 of the trials, seven pictures were from the same category, and one picture (the mask) was from a different category. The mask appeared in nine trials in the fourth position of the serial presentation, in nine trials in the fifth position, and in nine trials in the sixth position. Before each trial, the category of the majority of the pictures was annonuced. After each trial, subjects were given a forced-choice recognition task. Recognition of pictures presented in the series adjacent to non-conceptually related items was compared to recognition for items adjacent to conceptually related items.

Mask items tended to be remembered more frequently than conceptually related items located in the same serial position. In four of six comparisons, items adjacent to masks were recognized less frequently than control items. In one of six comparisons, items were remembered more frequently.

Thus, in a quickly presented series of conceptually related pictures, a non-conceptually related mask exerts more of a masking effect on adjacent items than does a conceptually related item. In some mask positions, however, this finding was not obtained. Possible explanations for the discrepancy are discussed.