Larger capacity for unconscious versus conscious episodic memory

Episodic memory is the memory for experienced events. A peak competence of episodic memory is the mental combination of events to infer commonalities. Inferring commonalities may proceed with and without consciousness of events. Yet, what distinguishes conscious from unconscious inference? This question inspired nine experiments that featured strongly and weakly masked cartoon clips presented for unconscious and conscious inference. Each clip featured a scene with a visually impenetrable hiding place. Five animals crossed the scene one-by-one consecutively. One animal trajectory represented one event. The animals moved through the hiding place, where they might linger or not. Participants' task was to observe the animals' entrances and exits to maintain a mental record of which animals hid simultaneously. We manipulated information load to explore capacity limits. Memory of inferences was tested immediately, 3.5 or 6 minutes following encoding. Participants retrieved inferences well when encoding was conscious. When encoding was unconscious, participants needed to respond intuitively. Only habitually intuitive decision-makers exhibited a significant delayed retrieval of inferences drawn unconsciously. Their unconscious retrieval performance did not drop significantly with increasing information load, while conscious retrieval performance dropped significantly. A working memory network, including hippocampus, was activated during both conscious and unconscious inference and correlated with retrieval success. An episodic retrieval network, including hippocampus, was activated during both conscious and unconscious retrieval of inferences and correlated with retrieval success. Only conscious encoding/retrieval recruited additional brain regions outside these networks. Hence, levels of consciousness influenced the memories' behavioral impact, memory capacity, and the neural representational code.Behavioral Data Raw behavioral data for retrieval- and stimulus awareness tasks are provided as CSV files. Variable descriptions are provided in text files. fMRI Functional and structural volumes are provided as raw data in NIfTI-format, and still need to be pre-processed before they can be used in statistical models that prerequisite standardized (i.e. slice-time corrected, realigned, coregistrered, normalized and potentially segmented or smoothed) data. Structural volumes have been skull-stripped to ensure anonymity of participants. Scanner parameters are provided in ScannerParams.PDF Session structure: functional volumes were collected in 12 sessions per Subject. Every session encompasses an encoding, pause, and retrieval part. The MR Scanner was continuously acquiring images throughout sessions, resulting in images of the different parts (encoding, pause, retrieval) being grouped together within each session folder. This means that onsets and movement regressors for pause and retrieval parts need to be adjusted if statistical models are to only contain these parts (isolated from the volumes outside these parts). This was done in the related publication (CURRENT-BIOLOGY-D-20-01249), where encoding and retrieval parts are modelled separately. Adjusted onsets, as well as scan indices for encoding and retrieval parts are included in this repository in the MATLAB datastructure model_data.mat. Encoding and retrieval are named "tasks" in this datastructure, not "parts". Detailed variable descriptions and data access documentation is provided in the text file "model_data_description.TXT" Statistical masks: due to a rapid TR of 1220 ms, combined with a 64-channel head coil, functional images exhibited a drop in signal intensity around the center of the brain. This is a general physical drawback of multichannel acquisitions. The signal collapses towards the center of the brain. The implicit statistical masking of SPM under default parameters may thus lead to the exclusion of large parts of these regions of interest from the analysis. Therefore, a custom explicit inclusive statistical mask has been constructed for the related publication, based on SPM's implicit mask, but replacing missing areas with normalized automated anatomical labeling (AAL) areas. Additionally, white matter was excluded from this custom explicit mask. To this aim, the average normalized structural (T1) image was segmented per group and experiment (experiment with strong masking; experiment with weak masking) using SPM's built-in segmentation function. The resulting white matter tissue map was then inverted and multiplied with the custom explicit inclusive statistical mask to create the final custom explicit mask. The resulting masks (spatially normalized to SPM's MNI305 T1 template) are included in this repository. Retrieval of weakly masked clips (conscious encoding) was on average more easy for participants than retrieval of strongly masked clips, leading to participants giving only correct responses in the following sessions of the weak masking condition (Note: VPxxxx are subject IDs): VP6048 Session 6 Session 7 Session 8 Session 11 Session 12 VP6191 Session 4 Session 9 Session 11 VP6207 Session 1 Session 8 Session 11 Session 12 VP6302 Session 3 VP6349 Session 1 VP6457 Session 3 Session 9 Session 11 VP6581 Session 2 Session 5 Session 6 Session 8 Session 11 VP6650 Session 4 Session 12 VP6685 Session 5 Session 7 Session 11 VP6690 Session 6 Session 7 Session 10 Session 12 VP7010 Session 6 Session 11 Session 12 VP7018 Session 1 Session 8 VP7040 Session 4 Session 7 VP7111 Session 3 Session 6 Session 7 Session 9 Session 12 VP7206 Session 3 Session 5 Session 8 Session 11 VP7452 Session 8 VP7469 Session 4 VP7474 Session 1 Session 3 Session 6 Session 7 Session 8 Session 9 Session 10 Session 12 Therefore, for these subjects & sessions, no onsets for regressor "R_Experiment_incorrect" can be provided in the table model_data.weak_masking.retrieval(subject).onsets{session}. For second level brain-behavior correlations, additional regressor-files (CSV) are provided containing reaction times and retrieval accuracies (as used in the related publication), aggregated per subject. File names are: RT_differences_for_fMRI_analyses.CSV and Accuracies_for_fMRI_analyses.CSV Funding provided by: Gottfried und Julia Bangerter-Rhyner-StiftungCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100005688Award Number: Center for Cognition, Learning, and Memory (CCLM) Ph.D. stipend to S. WuethrichSee related publication https://doi.org/10.1016/j.cub.2021.06.01