FIGURE 1.

Set-up of the experiment in the communicative context condition. Two participants – one in the MRI scanner, one in the behavioral experiment room – describe photographs to each other. (In the non-communicative context, there was only one – MRI – participant.) Trial structure and task were identical in both conditions. Green verbs at the start of a trial indicated that a (color-coded) production photograph would follow; gray verbs indicated a (grayscale) comprehension photograph would follow. Verbs were presented to participants in Dutch (English translation is shown in the figure). Production photographs were color-coded to guide participants’ production: participants were instructed to name the green figure before the red figure, leading them to produce an active or a passive sentence. When participant A in the communicative context produces a description, participant B listens to the description, and vice versa. Mismatches in the communicative context were created by showing a different photograph to speaker and listener (in the non-communicative context, a non-matching sentence recording was played to the participant). In both contexts, the listener needs to press a button when a mismatch is noticed. Feedback screens were only present in the communicative context: they reflect the percentage of hits minus false alarms and misses by both participants. Feedback was only presented within a filler block.