Temporal dynamics of access to amodal representations of category-level conceptual information

Abstract

Categories describe semantic divisions between classes of objects and category-based models are widely used for investigation of the conceptual system. One critical issue in this endeavour is the isolation of conceptual from perceptual contributions to category-differences. An unambiguous way to address this confound is combining multiple input-modalities. To this end, we showed participants person/place stimuli using name and picture modalities. Using multivariate methods, we searched for category-sensitive neural patterns shared across input-modalities and thus independent from perceptual properties. The millisecond temporal resolution of magnetoencephalography (MEG) allowed us to consider the precise timing of conceptual access and, by confronting latencies between the two modalities (“time generalization”), how latencies of processing depends on the input-modality. Our results identified category-sensitive conceptual representations common between modalities at three stages and that conceptual access for words was delayed by about 90 msec with respect to pictures. We also show that for pictures, the first conceptual pattern of activity (shared between both words and pictures) occurs as early as 110 msec. Collectively, our results indicated that conceptual access at the category-level is a multistage process and that different delays in access across these two input-modalities determine when these representations are activated.

Dynamic activity patterns in the anterior temporal lobe represents object semantics

“…To address the question of whether PRC semantic effects were larger than in other ATL regions, RSA effects were extracted from each region between 200 and 400 ms. This particular time window was chosen due to prior research implicating this time frame (Bankson et al, 2018 ; Chan et al, 2011 ; Y. Chen et al, 2016 ; Clarke et al, 2015 , 2018 ; Kreiman et al, 2000 ; Leonardelli et al, 2019 ; Mollo et al, 2017 ; Rogers et al, 2019 ; Rupp et al, 2017 ; Schendan & Ganis, 2012 ; Schendan & Maher, 2009 ; Sudre et al, 2012 ) and to avoid circular inference. PRC semantic effects were significantly greater than all ATL sub-regions except the temporal pole ( Figure 3C ); fusiform (p = 0.0078, PRC greater in 8/8 participants), inferior temporal (p = 0.016, PRC greater in 7/8 participants), middle temporal (p = 0.0391, PRC greater in 6/8 participants) but not the temporal pole (p = 0.078, 6/7 participants).…”

“…This research looked to build-upon previous studies relating semantic-feature similarity effects to neural similarity, that suggested the perirhinal cortex uniquely represented semantic item information (Bruffaerts et al, 2013 ; Clarke & Tyler, 2014 ; Kivisaari et al, 2019 ; Martin et al, 2018 ), but perhaps in cohort with the temporal pole (Martin et al, 2018 ). Further, EEG and MEG studies point to semantic-feature effects for individual items beyond around 200 ms (Bankson et al, 2018 ; Clarke et al, 2015 , 2018 ; Leonardelli et al, 2019 ; Mollo et al, 2017 ; Schendan & Maher, 2009 ; Sudre et al, 2012 ) but lacked detailed spatial specificity. It should be stressed that these studies, along with the current study, probe the semantics of objects at a basic-level (e.g., basketball, lawnmower) rather than a superordinate category level (e.g., tool), which may be associated with more posterior regions of the VVP (Bi et al, 2016 ; Connolly et al, 2012 ; Devereux et al, 2018 ; Konkle & Caramazza, 2013 ; Peelen & Downing, 2017 ; Tyler et al, 2013b ) at earlier points in time (Clarke et al, 2015 ; Mace et al, 2009 ).…”

“…RDMs between items were calculated using cosine distance for every time-point between −100 and 1000 ms based on a spatiotemporal pattern defined by the number of electrodes in the ROI and data from ±50 ms either side of the current time point (Y. Chen et al, 2016 ; Tyler et al, 2013a ). A further analysis was conducted which targeted a time window between 200 and 400 ms, which is highly associated with accessing conceptual knowledge for object concepts (Bankson et al, 2018 ; Chan et al, 2011 ; Y. Chen et al, 2016 ; Clarke et al, 2015 , 2018 ; Kreiman et al, 2000 ; Leonardelli et al, 2019 ; Mollo et al, 2017 ; Rogers et al, 2019 ; Rupp et al, 2017 ; Schendan & Ganis, 2012 ; Schendan & Maher, 2009 ; Sudre et al, 2012 ).…”

“…A further analysis was conducted which targeted a time window between 200 and 400 ms, which is highly associated with accessing conceptual knowledge for object concepts (Bankson et al, 2018;Chan et al, 2011;Y. Chen et al, 2016;Clarke et al, , 2018Kreiman et al, 2000;Leonardelli et al, 2019;Mollo et al, 2017;Rogers et al, 2019;Rupp et al, 2017;Schendan & Ganis, 2012;Schendan & Maher, 2009;Sudre et al, 2012).…”

“…Research based on time-sensitive techniques has further pointed toward the access of basic-level conceptual knowledge for visual objects after around 200 to 400 ms (Bankson et al, 2018 ; Chan et al, 2011 ; Y. Chen et al, 2016 ; Clarke et al, 2015 , 2018 ; Kreiman et al, 2000 ; Leonardelli et al, 2019 ; Mollo et al, 2017 ; Rogers et al, 2019 ; Rupp et al, 2017 ; Schendan & Ganis, 2012 ; Schendan & Maher, 2009 ; Sudre et al, 2012 ). The utilization of RSA with EEG, MEG and human intracranial recordings is becoming increasingly popular, highlighting the relationship between dynamic neural activity with low to high-level visual properties (Carlson et al, 2013 ; Cichy et al, 2016 , 2014 ; Contini et al, 2017 ; Kaneshiro et al, 2015 ; Seeliger et al, 2018 ) and basic-level conceptual representations (Bankson et al, 2018 ; Y. Chen et al, 2016 ; Clarke et al, 2018 ).…”