Attribution of mental states to self and others, i.e., mentalizing, is central to human life. Current measures are lacking in ability to directly gauge the extent of individuals engage in spontaneous mentalizing. Focusing on natural language use as an expression of inner psychological processes, we developed the Mental-Physical Verb Norms (MPVN). These norms are participant-derived ratings of the extent to which common verbs reflect mental (opposite physical) activities and occurrences, covering ~80% of all verbs appearing within a given English text. Content validity was assessed against existing expert-compiled dictionaries of mental states and cognitive processes, as well as against normative ratings of verb concreteness. Criterion Validity was assessed through natural text analysis of internet comments relating to mental health vs. physical health. Results showcase the unique contribution of the MPVN ratings as a measure of the degree to which individuals adopt the intentional stance in describing targets, by describing both self and others in mental, opposite physical terms. We discuss potential uses for future research across various psychological and neurocognitive disciplines.

Dehumanization is frequently cited as a precursor to mass violence, but quantitative support for this notion is scarce. The present work provides such support by examining the dehumanization of Jews in Nazi propaganda. Our linguistic analysis suggests that Jews were progressively denied the capacity for fundamentally human mental experiences leading up to the Holocaust. Given that the recognition of another’s mental experience promotes moral concern, these results are consistent with the theory that dehumanization facilitates violence by disengaging moral concern. However, after the onset of the Holocaust, our results suggest that Jews were attributed a greater capacity for agentic mental states. We speculate this may reflect a process of demonization in which Nazi propagandists portrayed the Jews as highly capable of planning and intentionality while nonetheless possessing a subhuman moral character. These suggestive results paint a nuanced portrait of the temporal dynamics of dehumanization during the Holocaust and provide impetus for further empirical scrutiny of dehumanization in ecologically valid contexts.

No abstract

Dehumanization is frequently cited as a precursor to mass violence, but quantitative support for this notion is scarce. The present work provides evidence for this link by examining the dehumanization of Jews in Nazi propaganda. Our linguistic analysis of Nazi propaganda suggests that Jews were progressively denied the capacity for fundamentally human mental experiences in the lead up to the Holocaust. These results are consistent with the notion that, given that the recognition of another’s mental experience promotes moral concern, dehumanization facilitates violence by disengaging moral concern. However, after the onset of the Holocaust, our results suggest that Jews were attributed a greater capacity for agentic mental states. We speculate this may reflect a process of demonization in which Nazi propagandists portrayed the Jews as highly capable of planning and intentionality while nonetheless possessing a subhuman moral character. These suggestive results paint a nuanced portrait of the temporal dynamics of dehumanization during the Holocaust and provide impetus for further empirical scrutiny of dehumanization in ecologically-valid contexts.

This paper describes our approach to theCLPsych 2021 Shared Task, in which weaimed to predict suicide attempts based onTwitter feed data. We addressed this challengeby emphasizing reliance on prior domainknowledge. We engineered novel theory drivenfeatures, and integrated prior knowledgewith empirical evidence in a principledmanner using Bayesian modeling. Whilethis theory-guided approach increases bias andlowers accuracy on the training set, it was successfulin preventing over-fitting. The modelsprovided reasonable classification accuracy onunseen test data (0.68 ≤ AUC ≤ 0.84). Ourapproach may be particularly useful in predictiontasks trained on a relatively small data set.