Psychopathology and social behaviour

Paranoia and other symptoms of clinical psychosis also occur in non-clinical samples. I am working with Vaughan Bell to conduct large-scale behavioural experiments to ask how paranoia affects social behaviour and the interpretation of others’ behaviour and intentions.

The aim of this research agenda is (i) to quantify how deviations from “normal” psychology systematically affect social cognition and behaviour; (ii) to provide an explanation for part of the enormous variation in social behaviour that can be observed in humans; and (iii) eventually, to try to understand why disorders of the social brain (such as autistic spectrum disorders and psychotic spectrum disorders) persist – albeit at low levels – in natural populations.

Our work in this area has produced some of the following key findings:

• Paranoia exerts a marked but specific effect on social cognition: We used a Dictator Game paradigm to explore social cognition in paranoia. The Dictator Game has the useful feature that the intentions underpinning different donations are somewhat ambiguous: selfish allocations could stem from greed or from malevolent intentions (i.e. the desire to prevent the partner from getting anything). Here, we use a large-N online study to show that variation in paranoid ideation predicts increased attribution of harmful intent but does not affect attributions of self-interest to dictators in this task.
• Reduced pro-sociality in paranoia is not only due to reduced trust: Previous studies found that paranoid ideation is associated with reduced cooperative behaviour and explained this in terms of distrust of the partner. Here, we challenge this hypothesis by showing that paranoia predicts reduced pro-sociality even when trust is not a relevant concern. This suggests alternative explanations could also be important, including the idea that paranoia is associated with finding mutually cooperative interactions less subjectively rewarding.
• Experimentally-induced social threat prompts paranoid thinking: Large-scale epidemiological studies have suggested that social threat poses a risk for developing psychosis, of which paranoia is a key symptom. However these studies report associations and cannot determine causal pathways. We conducted a large-N, pre-registered set of experiments where we exposed participants to social threat in different forms – either interacting with a higher status partner or interacting with a political opponent. Exposure to these mild social stressors increases people’s tendency to attribute harmful intent to their partners in situations where the partner’s true intentions are ambiguous. Thus, we show experimentally that paranoid thinking is context-specific and responds specifically to social threat. Importantly, we also show that higher levels of pre-existing paranoia do not result in dysregulated responses to social threats. Instead, higher levels of paranoia mean that people have a lower threshold for detecting threat rather than an impaired reactivity to it.