A talk from the PsychoSlav 2022 conference, presenting four SPR experiments testing whether comprehenders manage to arrive to a coherent and globally correct representations of garden-path sentences, that are faithful to the input