Nominalization and Lowered Perception of Agency in Discourse: An Empirical Study

Could the way an event is encoded in language affect people’s perception of the responsibility attributed to the agent of the event? This article aimed to empirically investigate the effect of nominalization on the participants’ perception of the responsibility attributed to an agent in discourse. To answer the research question, two tasks were designed. In the first task, 331 participants read two scenarios, namely “An Accident in the Restaurant” and “Environmental Pollution”, in two nominalized and non-nominalized versions. Then, they answered two questions regarding the responsibility of the agent in bringing about the accident and the amount of damage on a Likert scale. The results showed that the nominalized version of the Restaurant scenario had caused participants to rate the responsibility of the agent in the accident significantly lower than in the non-nominalized version. No effect of nominalization was found for either question in the Environment scenario, nor for the question asking about the amount of damage in the Restaurant scenario. In order to test the reliability of the observed effect, a second task was designed in which an accident happened in an interview session. One hundred forty-six new participants participated in the second task. The results of the second task showed the same patterns as in the first task. Participants had rated the responsibility of the agent in the nominalized version of the scenario significantly lower compared to their responsibility in the non-nominalized version. The findings confirm the role of nominalization in concealing agency in discourse.