‘Bloody Fucking Amazing!’ – Re-Evaluating Emotionality in US Hollywood Screenplays

Linguistic expressions of emotion have been described as ‘an, if not the most, essential feature of telecinematic language at large, used for example to create drama and conflict’ ( Werner , 2021, p. 45; cf. Kozinski, 2011; Bednarek, 2018 ). Indeed, we know that emotional lexis is elevated in telecinematic discourse compared to non-fictional monitor corpora of spoken (American) English (Berber Sardinha and Veirano Pinto, 2017). Recent studies on the topic have relied on large-scale corpora of US American feature films or television series to describe the lexico-grammatical shape of film and television language in general. Note that these studies do not explicitly focus their description on emotional lexis alone so that the quantitative trends they posit are not analysed and discussed in much detail. More importantly perhaps, the degree of variability of emotional lexis across film genres, time periods and character types still remains unclear. This study provides such an insight into the inter- and intrageneric variation of emotional lexis of US American Hollywood screenplays across the past five decades. Specifically designed Python scripts are used, allowing for the automatic segmentation of functional text spans, based on their consistent levels of textual indentation, for example, screen directions, character tags and character contributions. This procedure gives rise to the Augsburg Screenplay Corpus (ASC, 852,659 tokens), containing 100 different US American Hollywood screenplays, equally stratified across four main film genres (Action/Thriller, Comedy, Drama, Horror) and five consecutive time periods (1970s, 1980s, 1990s, 2000s, 2010s). The study employs concordance and collocation analysis, which are combined with the semantic key domain method (Rayson and Smith, 2006) to probe emotional lexis in the corpora. Results show that (a) the frequency of emotional lexis varies significantly across the four main film genres, (b) the types of emotional expressions used across film genres are notably different, (c) the general level of emotionality in US Hollywood screenplays is rather stable across time and (d) intrageneric dispersion of emotional lexis between films and film characters (also between character gender) is considerable. More specifically, the excessive use of expressive or emotional language is typically associated with a small number of film characters. This suggests that screenwriters use these ‘surge features’ (Taavitsainen, 1999 ) deliberately as implicit cues for characterisation (Culpeper, 2001), reserving them for certain characters while limiting their use for others.