Experiments into the influence of linguistic (in)directness on perceived face-threat in Twitter complaints

Abstract To date, there has been little attention for the factors that influence the perception of online complaints. We present two experiments in which we test the impact of the degree of linguistic (in)directness and the formal realization of complaint components on complaint perception. Our experimental stimuli are designed on the basis of French-language authentic Twitter complaints which have been coded in terms of the presence of four constitutive complaint components: the complainable, the negative evaluation of the complainable, the person/company responsible for the complainable, and a wish for compensation. In our experiments, participants are asked to read Twitter complaints, and they are invited to assess them in terms of perceived strength, dissatisfaction, (im)politeness, and offensiveness. Our results indicate that not only the number but also the type of component that is formally realized shape complaint perception. We also find a positive correlation between perceived complaint strength and impoliteness. In addition, different formal realizations of the negative evaluation of the complainable have a different effect on complaint perception; in particular, negative emoji make the complaints softer and more polite. We also discuss methodological issues that have arisen while designing the experiments and that have to do with the operationalization of face-threat.