Individuals with generalized anxiety disorder (GAD) symptoms often struggle with heightened sensitivity and arousal in response to perceived threats. Moreover, interpersonal dysfunction in GAD has become increasingly a focus of empirical investigation and treatment, given the possibility that responses to social interactions may contribute to GAD symptom maintenance. Laboratory studies and cross-sectional trait assessments of interpersonal problems comprise most of our understanding of interpersonal dysfunction in GAD. However, how GAD symptoms interact with perceived interpersonal threats to predict affective responses (increased arousal, lower valence) within daily life remains poorly understood. Therefore, the purpose of the present study was to examine effects of in vivo social perceptions on state affect, and how GAD symptoms may moderate those relationships. Participants (N = 161) completed baseline measures of trait GAD and depression symptoms (as a covariate). Then participants completed 30 social interaction surveys over the subsequent 10 days. In each survey, participants rated interaction partners’ dominant, cold, and immoral behavior (each conceptualized as interpersonal threats) as well as their own arousal and valence in response to the behavior. Multilevel modeling analyses of between- and within-person effects revealed that mean perceptions of cold and immoral behavior predicted higher arousal and lower valence as hypothesized, whereas mean perceived dominance unexpectedly predicted only lower valence. All within-person fluctuations in social perceptions predicted both higher arousal and lower valence. Regarding the moderating effects, GAD symptoms unexpectedly buffered the effect of average perceived cold behavior on valence and strengthened the effect of average perceived immoral behavior on valence. These results provide a deeper understanding of how social perceptions may contribute to affect in naturalistic interactions, and add to the literature on interpersonal correlates of GAD symptoms.

The central argument of this dissertation is that virtue ethics is overly individualistic. In response, I develop and defend a more relational, ecological account - what I call extended virtues. First, following Andy Clark, Kim Sterelny, and others, I argue that cognition and emotion can be embedded in, scaffolded by, or even extended to include various environmental resources. These arguments undermine default internalism about cognitive and affective processes. Next, I show how recent work in social and personality psychology similarly undermines individualism about the bearers of these cognitive and affective processes. Taken together, these arguments have significant but heretofore underappreciated implications for virtue ethics. After reviewing the literature which attempts to spell out the ethical implications of embedded, scaffolded, and extended cognition, I conclude that a more substantive engagement with virtue ethics is needed. I then show how plausible, mainstream theories of virtue assume default internalism and individualism, and are thus subject to charges of empirical inadequacy. Finally, I formulate my account of extended virtues in response to these shortcomings. I begin by making three explicit arguments for why an account of extended virtues is needed. I then develop two further arguments - the process argument and the bearer argument - which yield the conclusion that the processes relevant to, and the bearers of, moral and intellectual virtues can be embedded, scaffolded, or extended. After providing examples and filling in details about the hypotheses of embedded, scaffolded, and extended virtue, I propose that virtues are less like dispositions and more like relations. I conclude by suggesting that ecological metaphors such as stewardship are more fitting than traditional views of morality as inner strength.