The aim of this dissertation was twofold. First, to explore how counterfactuals embedded in narratives modulate updating processes as a consequence of their dual meaning. In two behavioral experiments (Chapter 1), participants responded to test probes following a factual or counterfactual text. The experiments confirmed that the realistic meaning of counterfactuals (an implicit negation) prevented readers to updating the situation model, whereas the alternative meaning of counterfactuals was activated to some degree exactly like a factual meaning. A new experiment with EEG (Chapter 2) clearly showed specific brain signatures of the dual meaning and updating processes in counterfactuals: N400 (in the ERP analyses) and gamma band power modulation (in the time-frequency analyses).
The second focus of this dissertation was to test whether counterfactuals activate embodied representations. A set of 3 behavioral experiments with an Action-Sentence Compatibility Effect paradigm (Chapter 3) revealed an interaction between the meaning of transfer sentences and the direction of finger motions, demonstrating that action-related sentences (either counterfactual or factual) and motor actions share processes in the brain. Finally, in the last experiment (Chapter 4), in an event-related fMRI study participants read factual and counterfactual sentences describing actions with different degrees of effort. The results showed that high-effort sentences elicited more activation than low-effort sentences in the left inferior parietal lobule and other motor areas, independently of their counterfactual or factual format. In addition, counterfactuals showed specific activation in frontal regions related to the additional processes demanded by their dual meaning.