Modularity in short-term memory : a comparison between olfactory, visual and auditory stimuli
Johnson, Andrew James
This thesis is concerned with memory modularity. Recognition and serial position recall for sequences of odours, unfamiliar-faces and pure-tones is examined and compared. In the first set of experiments, 2-alternative-forced-choice (2AFC) recognition produced qualitatively equivalent serial position functions across the three stimulus types. Recency in the absence of primacy was apparent for the backward testing procedure, in contrast, the forward testing procedure produced an absence of serial position effects. This item-based task, therefore, provides some evidence for amodal processing. A second set of experiments employed a single-probe serial position recall paradigm and produced qualitatively different functions for the three stimulus types. Unfamiliar-faces showed both primacy and recency, pure-tones showed recency only and, in contrast, odours showed neither recency nor primacy. These functions were consistent across sequence length, i.e. 4-6 items. Primacy for unfamiliar-faces was immune to articulator suppression suggesting a non-verbally based representation. A third set of Experiments employed a modified serial order reconstruction paradigm in which test-items were presented sequentially rather than simultaneously. Unfamiliar-face and pure-tone stimuli demonstrated both primacy and recency, in contrast, odours demonstrated recency only. The absence of cross-stimuli functional equivalence for both order-based tasks contrasts with that for the item-based task, suggesting that item- and order-based memorial processes impact differentially upon arguments for modularity (see Guerard and Tremblay, in press). A fourth set of experiments investigated the role of verbal coding in modified-reconstruction. The functions produced for both odour and unfamiliar-face stimuli were shown not to be characteristic of verbal memory. A final study in this set employed nonwords presented both visually and aurally. Results showed that, regardless of modality of input, the functions replicated those reported above for nonverbal visual and auditory stimuli. The findings demonstrate that the functions for modified-reconstruction are not underpinned by the processes of verbalisation. In summary, the order-based tasks are consistent with modularity, where the item-based tasks support amodal processing. The findings suggest that for order memory at least, olfactory memory is qualitatively different to that of visual and auditory memory.