One turn at a time: Behavioral and ERP evidence for two types of rotations in the classical mental rotation task

Abstract

We perform mental rotations in many everyday situations, such as reading a map or following furniture assembling instructions. In a classical mental rotation task, participants are asked to judge whether a rotated stimulus is presented in its mirrored form or its canonical form. Previous results have indicated a degree effect: RT is longer as the angle of rotation increases, and this effect is traditionally explained by arguing that this judgment requires rotating the stimulus back to its upright form. Importantly, in half of the trials, the stimuli are rotated on both the page plane and mirror plane. Namely, we argue that in previous research the task actually involved two different rotation processes. To provide a clear dissociation between these two rotations, we collected EEG data and used the Contralateral Delay Activity (CDA) as an indicator of visual working memory (VWM) load. The results of Experiment 1 suggested different VWM involvement according to the degrees rotations when the item was not mirrored, such that the CDA amplitude generally increased as the degree of rotation was higher. Mirrored trials were all at ceiling in terms of CDA, regardless of their rotation degree. Experiment 2 showed increased CDA amplitude uniquely related to the flip rotation. Thus, we provided ERP evidence that the canonical mental rotation task involves two types of rotations that can be dissociated based on the load they imposed on VWM.

The time course of planar and non-planar rotations in a letter rotation task

“…This latter observation is in line with previous studies in which stimuli were presented on the screen for a limited period of time (500 ms; e.g., Núñez-Peña and Aznar-Casanova, 2009). Considering the relatively short letter representation time in the present study, this could be accounted for by posing a different representation between the two sets of stimuli (Ankaoua & Luria, 2022). Participants might have had difficulties accessing the perceptual information after stimulus offset in canonical trials, whereas they were more able to rely on its internal representation in mirror letter trials.…”

“…Differences between the MR processes elicited by mirror-reversed and canonical letters have also been reported in event-related potentials (ERP) studies (Ankaoua & Luria, 2022;Hamm, Johnson & Corballis, 2004;Núñez-Peña & Aznar-Casanova, 2009;Quan et al, 2017). The rotation-related negativity (RRN) is an ERP component conceived as the psychophysiological correlate of the spatial transformation process in MRTs (for a review see Heil, 2002).…”

“…Alternatively, some researchers have postulated the presence of an additional cognitive sub-process during the MR of mirror-reversed letters (Alivisatos & Petrides, 1997;Ankaoua & Luria, 2022;Corballs & McMaster, 1996). According to this hypothesis, to fully canonicalize mirror-reversed letters, participants rotate these letters out of the plane (non-planar rotation) after their rotation in the plane (planar rotation), (Corballis & McMaster, 1996;Quan et al, 2017).…”

“…According to this hypothesis, to fully canonicalize mirror-reversed letters, participants rotate these letters out of the plane (non-planar rotation) after their rotation in the plane (planar rotation), (Corballis & McMaster, 1996;Quan et al, 2017). It was argued that the presence of this additional process -labelled 'flip-over' -resulted in higher working memory (WM) loads on mirror-reversed compared to canonical letter trials which delayed RTs (Ankaoua & Luria, 2022).…”

“…Differences between mirror-reversed and canonical trials (reflecting the nonplanar rotation) are more likely to arise in the later phase of MR (Quan et al, 2017), occurring increasingly later as a function of the increasing rotation angles (Núñez-Peña and Aznar-Casanova, 2009). These observations have led researchers to suggest that the out-of-plane (non-planar) rotation occurs after planar rotation (Ankaoua & Luria, 2022;Núñez-Peña and Aznar-Casanova, 2009;Quan et al, 2017), allowing participants to fully canonicalize mirror-reversed letters after the planar rotation. However, because the difference between the RRN amplitudes elicited on canonical and mirror-reversed trials was absent for larger rotation angles (Núñez-Peña and Aznar-Casanova, 2009), it was further suggested that the non-planar rotation occurs sequentially after the planar rotation for smaller angles but in parallel for larger angles (Núñez-Peña and Aznar-Casanova, 2009;Quan et al, 2017).…”