Subjectivity in Diplomatic Discourse: a double annotation analysis

Although diplomatic texts are undeniably evaluative, linguists have yet to figure out the main features of diplomatic attitudes. One of the main problems in such an evaluation is the absence of specific criteria and guidelines for corpus analysis. We have aimed to describe the typical problems that arise when assessing appraisal (Martin and White, 2005) in diplomatic speeches. Our approach is based on a double annotation experiment on selected diplomatic speeches. The experiment was completed according to extensive guidelines we developed earlier (Anisimova and Zikánová, 2024), allowing for a finer-grained attitude annotation of the data. The primary source of disagreement stems from text spans selected only by one annotator. This could be explained by the nature of the attitude identification process, the subjectivity factor, and human error. The cases of agreement tend to satisfy both attitude subtype and polarity criteria. The analysis has shown that categories of judgement and attitude polarity are central means of attitude expression in diplomatic communication.