Body Maps of Moral Concerns
Mohammad Atari, Aida Mostafazadeh Davani, and Morteza Dehghani
Does a violation of your moral expectations feel like a punch in the gut or a broken heart? Participants read vignettes about violations of moral concerns, such as care, fairness, loyalty, authority, and purity (e.g., “You see a woman clearly avoiding sitting next to an obese woman on the bus”), and evaluated how morally wrong the action depicted was and how strongly they emotionally responded to it. For each vignette, participants created a map of body parts where they felt an emotional response by coloring areas on a body silhouette to indicate where they felt activity became stronger or faster and where they felt activity became slower or weaker. Atari and colleagues also assessed participants’ political orientation and the relevance they assigned to moral concerns. Results indicated that different moral violations elicited different body maps of activation and deactivation—violations of care, fairness, loyalty, and authority were associated with activation in the chest area, and violations of purity were associated with higher activation in the abdomen. The individual levels of concern about moral values also predicted the body maps associated with each concern. Interestingly, political liberals and conservatives seemed to feel moral violations, especially loyalty and purity violations, in different parts of their bodies. Similar findings were also obtained in a study with a larger nationally representative sample of Americans. Overall, the authors concluded that moral values and political orientation influenced where and how moral violations were felt in the body.
Young Children Use Probability to Infer Happiness and the Quality of Outcomes
Tiffany Doan, Ori Friedman, and Stephanie Denison
Children’s inferences about emotions, such as happiness, and the quality of outcomes may depend on the initial probability of the outcome, this research suggests. In a series of experiments, Doan and colleagues showed participants a story in which a girl pulled the handle of a gumball machine and received two “yummy” red gumballs and two “yucky” black gumballs. Although the girl always received the same gumballs, the gumball machine had either mostly yummy or mostly yucky gumballs, and participants rated how happy the girl felt when she received her gumballs. Results showed that 5- and 6-year-olds, but not 4-year-olds, rated the girl as less happy when the gumball machine had mostly yummy gumballs, and therefore the probability of a better outcome was higher. When the girl was removed from the story and participants were asked to rate how good the outcome was rather than to infer happiness, all age ranges used the initial probability to rate the outcome quality (i.e., the outcome was better when the machine had mostly yucky gumballs). These findings suggest a developmental gap in which 4-year-olds use probability to assess outcome quality but not to infer happiness. In another experiment, Doan and colleagues showed that adults use probability to infer both outcome quality and happiness, indicating that probability is as linked to quality as it is to happiness inferences. The authors suggest that children and adults may use initial probability to set a standard against which the outcome is compared, allowing for probability-based inferences of other people’s happiness.
Implicit Gender Bias in Linguistic Descriptions for Expected Events: The Cases of the 2016 U.S. and 2017 United Kingdom Elections
Titus von der Malsburg, Till Poppels, and Roger P. Levy
To analyze gender biases in descriptions of expected events, von der Malsburg and colleagues tested participants’ beliefs and linguistic biases during electoral campaigns with women candidates. During the 2016 U.S. presidential campaign, a woman was expected to win for the first time ever, and during the 2017 United Kingdom general election, a woman was expected to be reelected. In several rounds of data collection throughout the campaigns, participants reported how likely they thought it was that each of the candidates would win the election. In addition, they completed two measures of linguistic bias (text completion and self-paced reading) at each round of data collection, as well as after the election. In the text-completion task, participants completed a text fragment referencing the next president using a pronoun of their choice. In the self-paced reading task, participants used a self-paced moving window to read sentences that included pronomial references (he/she/they) to the future president. In the U.S., even though participants believed that a woman was likely to be elected, they was the most-used pronoun, she was produced less often than he, they was produced more often than she, and she took longer to read than he or they. In the U.K., more participants produced she than he, but she did not confer any reading-time advantages over he. These findings indicate a bias against she rather than a biased preference for he, which seems to be caused by a biased mapping of election-outcome expectation to language. These findings suggest that the language system may itself be a source of implicit biases in addition to previously known biases, such as stereotypes.