Memory and Attention

Conception of working memory

The most influential view of human cognition is based around the premise that the cognitive architecture is divided into separate modules, such as verbal and spatial components (e.g., Baddeley, 2000 ; Wickens, 1992). However, we conducted a series of experiments showing that spatial and verbal short-term memory systems are functionally equivalent, at least for processing sequential information. Both show similar serial position curves and exhibit similar patterns of errors (Guérard & Tremblay, 2008). The processing of verbal and that of spatial information are sensitive to interference effects (e.g., Tremblay, Nicholls, Parmentier & Jones, 2005), can benefit from Hebb repetitions (Couture & Tremblay, 2006) and can lead to distinctiveness effects (Guérard, Neath, Surprenant & Tremblay, 2010). Based on recent empirical work, key processes in cognition such as the processing of sequential information in memory (e.g., Tremblay et al., 2006) and the deployment of attention over time (e.g., Tremblay, Vachon, & Jones, 2005) seem best represented by a non-modular view. According to the latter view, the multiplication of modules is superfluous and explanations based on the nature of processes rather than the type of information can account for a variety of phenomena in cognition.

Task switching and serial memory

The cost of switching between relatively simple cognitive tasks has been investigated extensively over the last two decades. Accumulating evidence suggests that switching back and forth between tasks leads to slower response times and to more errors than performing a single task repeatedly. From an ecological perspective, one might wonder to what extent results from task switching studies can be generalized to other different, and arguably more complex, tasks. We conducted a series of experiments revealing no cost of switching between serial memory tasks (and, in some cases, a beneficial effect of switching). Such a finding challenges the widely accepted assumption that task-alternation comes with a considerable cost in performance regardless of the cognitive tasks undertaken.

Processing of Spatial and Serial Memory

Over the last four decades or so, most of the research on serial memory has been devoted to the study of verbal memory and less is known about serial memory for lists of spatial items. We conducted a series of experiments in order to examine whether there is some form of visual-spatial rehearsal and to establish whether eye movement was a good candidate for such a mechanism (see Figure 1). We are also interested in mapping out the influence of perceptual organization on serial memory. For instance, we showed that spatially distributing auditory items make them more difficult to remember than the same items presented from a single location (see Figure 2).

Sequence learning

Sequence learning is essential in cognition and underpins activities such as language and skill acquisition. One classical demonstration of sequence learning is that of the Hebb repetition effect whereby serial recall on a repeated list improves over repetitions relative to random lists. A series of experiments has revealed that the learning mechanism is not uniquely verbal and transcends the content of what is to be learned (Couture & Tremblay, 2006), that it can lead to anticipatory behaviour – as evidence by eye movements on target loci prior to their occurrence (Tremblay & Saint-Aubin, 2009), and that error learning can be mistaken for the absence of learning (Lafond, Tremblay & Parmentier, 2010). Indeed, incorrect responses are reproduced increasingly over repetitions (Couture, Lafond & Tremblay, 2008), showing that individuals involuntarily “learn” and repeat their errors. Future research will focus on the robustness of sequence learning when faced with potential vulnerabilities in the organization of the repeated sequence. Overall, our findings place key constraints on modelling the Hebb effect, and provide insightful information about human memory and learning.