This paper is a pre-print of my MPhil research work, where I tested a hypothesis about Noun use by political Conservatives by analysing Language and Social Network data from Twitter, and contrasting this with findings from my two more traditional survey studies.

Download paper here

View supplementary materials, data & code

Abstract

Political conflict and misunderstanding pose dangers to society. Psychological science has the potential to reduce this by providing unique insights into cognitive patterns associated with different political ideologies. A recent study (Cichocka et al., 2016) found that a seemingly unrelated behaviour – the use of Nouns when completing sentences describing individuals – correlated with political Social Conservatism. The following three-part study further investigated this association, its manifestations in spontaneous communication, and potential underlying mechanisms. Study 1 replicated the association between Noun use and Social Conservatism in a UK-based survey, using a broader set of political measures. It also found that the trait Essentialism, or the tendency to perceive individuals through stable underlying essences, may mediate the effect, alongside differences in education level. Study 2 extended the investigation to spontaneous communication, making use of network and linguistic data from the social networking site Twitter. This study also differentiated between Common Nouns and Proper Nouns, and found that Conservatives, on average, used more Proper Nouns per tweet than Liberals, while the effects for Common Nouns were inconsistent across various measures of Noun use. Study 3 aimed to validate the novel finding with Proper Nouns in a controlled questionnaire setting. The association between Conservatism and Proper Noun use was not confirmed, but the original effect with Common Nouns was replicated, and potentially mediated by sex and/or Essentialism. These findings suggest that the psychological characteristics differentiating Liberals from Conservatives, perhaps including the trait Essentialism, may affect their use of Nouns, although the direction of influence and potential demographic confounds need to be investigated experimentally. These results also suggest that in spontaneous, contextualised communication, such as that on Twitter, these differences may manifest themselves in the use of Proper rather than Common Nouns. Finally, the three studies demonstrate the benefits of combining short and simple, yet highly controlled questionnaires with more exploratory and larger-scale, web-based observations of spontaneous language use for the advancement of knowledge in Political Psychology and Psycholinguistics.