Semantics in the motor system: motor-cortical Beta oscillations reflect semantic knowledge of end-postures for object use

Abstract

In the present EEG study we investigated whether semantic knowledge for object use is represented in motor-related brain areas. Subjects were required to perform actions with everyday objects and to maintain either a meaningful or a meaningless end posture with the object. Analysis of the EEG data focused on the beta-frequency band, as previous studies have indicated that the maintenance of a posture is reflected in stronger beta-oscillations. Time frequency analysis indicated that the execution of actions resulting in a meaningless compared to a meaningful end posture was accompanied by a stronger beta-desynchronization towards the end of the movement and a stronger subsequent beta-rebound after posture-onset. The effect in the beta-frequency band was localized to premotor, parietal and medial frontal areas and could not be attributed to differences in timing or movement complexity between meaningful and meaningless actions. Together these findings directly show that the motor system is differentially activated during the execution and maintenance of semantically correct or incorrect end postures. This suggests that semantic object knowledge is indeed represented in motor-related brain areas, organized around specific end postures associated with the use of objects.

Keywords: EEG; beta-rebound; end-postures; motor system; objects; semantics.

Figure 1

Overview of the experimental setup. The subject was seated behind a table on which 15 different objects were placed. A buttonbox was placed on the subject's lap from which all movements were initiated. A picture on the screen instructed subjects by which grip they were required to grasp the object. Grips could result in either a meaningful end posture or a meaningless end posture when the object was moved towards the correct goal location.

Figure 2

Picture stimuli used in the experiment. For each object, one picture indicated a grip that would result in a meaningful end posture (left column). By crossing the grips between objects, for each object a grip was specified that would result in a meaningless end posture when the object was brought to the typical goal location (right column).

Figure 3

Overview of experimental procedure. (A) Each trial started with a central fixation cross which was presented for 3000–4000 ms, while the subject was holding the hand at the button box. (B) A picture on the screen indicated by which grip subjects were required to grasp an object. (C) Upon initiating the movement the picture was removed from the screen and the subject grasped the indicated object with the pre-specified grip. (D) When the subject had reached the final end posture with the object they were required to maintain this posture for 4 s. (E) An instruction cue on the screen instructed the subject to put the object back on the table and to return to the starting position, upon which the next trial was initiated.

Figure 4

Beta-desynchronization during movement preparation and execution. EEG data time-locked to the onset of the picture instructing the subject by which grip they were required to grasp the object. (A) Time frequency representations (TFRs) for actions resulting in a meaningful end posture (left graph), a meaningless end posture (middle graph) and the difference between meaningless and meaningful action conditions (right graph). The time-interval and frequency range used for statistical analysis are marked in white. (B) Relative beta-power (16–24 Hz) for actions resulting in a meaningful end posture (blue lines) or a meaningless end posture (red lines). (C) Topographical maps of the beta-power (16–24 Hz) for meaningful actions (upper panel), meaningless actions (middle panel) and the difference between meaningless and meaningful action conditions (lower panel). Electrodes that were used for statistical analysis and for plotting the TFRs are marked in black.

Figure 5

Beta-rebound during posture maintenance. EEG data time-locked to the onset of the end posture. (A) Time frequency representations for actions resulting in a meaningful end posture (left graph), a meaningless end posture (middle graph) and the difference between meaningless and meaningful action conditions (right graph). The time-interval and frequency range used for statistical analysis are marked in white. (B) Relative beta-power (16–24 Hz) for actions resulting in a meaningful end posture (blue lines) or a meaningless end posture (red lines). (C) Topographical maps of the beta-power (16–24 Hz) for meaningful actions (upper panel), meaningless actions (middle panel) and the difference between meaningless and meaningful action conditions (lower panel). Electrodes that were used for statistical analysis and for plotting the TFRs are marked in black.

Figure 6

Source analysis of the beta-rebound. Source reconstructions accounting for the stronger beta-rebound for meaningless compared to meaningful actions from 1200 to 4000 ms relative to the onset of the end posture. The strongest beta modulations were observed in the superior parietal cortex [BA 6 (A)] and in the superior frontal gyrus [BA 9 (B)]. Source activation is projected on a standard brain.

Figure 7

Beta-rebound during posture maintenance after exclusion of objects. EEG data time-locked to the onset of the end posture after exclusion of objects contributing to slower responding in Meaningless action conditions and objects contributing to faster responding in Meaningful action conditions. (A) Time frequency representations for actions resulting in a meaningful end posture (left graph), a meaningless end posture (middle graph) and the difference between meaningless and meaningful action conditions (right graph). The time-interval and frequency range used for statistical analysis are marked in white. (B) Relative beta-power (16–24 Hz) for actions resulting in a meaningful end posture (blue lines) or a meaningless end posture (red lines).