This dissertation concerns the attachment motive. It presents a theory and empirical data testing hypotheses about the way people obtain support (when desired) from reliable attachment figures. In order to obtain support, a person first needs to identify potentially supportive partners. To do so, the person needs relevant knowledge for predicting whether a potential partner can be counted on to be reliably supportive. The present dissertation examines two relevant types of knowledge structures and provides ways to assess them. The first knowledge structure provides indicators of a potentially supportive partner that should serve as good predictors. In Study 1, I applied the prototype methodology to derive a prototype of indicators that another person can be expected “to be there for me at future times of stress.” Using a false recognition paradigm, I demonstrated that this knowledge affects information processing and constitutes a prototype. I call this kind of knowledge “indicator-knowledge.”

In Study 2, I used these indicators to construct a new type of personality measure that assesses individual differences in knowledge. The measure is called the “Knowledge of Indicators (KNOWI) Scale”. It assesses a person's knowledge of indicators that lead to an expectation that a partner will be there when needed. The KNOWI applies signal detection methods to assess each participant's ability to discriminate good from poor indicators. The resulting index is called the “KNOWI-accuracy index.” In Study 3 I showed that the measure predicts performance on a laboratory task: Participants interacted with a confederate who described a personal problem; in describing the problem, the confederate revealed subtle cues that another person could not be counted on for support. Participants with a high KNOWI-accuracy index recognized, recalled, and correctly interpreted the cues more readily. In Study 4, I demonstrated that the KNOWI Scale possesses convergent and discriminant validity.

In addition to the KNOWI-accuracy index, the KNOWI Scale provides a second index, the KNOWI-readiness index, which is the sum of a respondent's ratings across all indicators. I hypothesized that an elevated motive for support leads to a higher KNOWI-readiness index. I call this general effect the ELMO (“elevated motive”) effect. Studies 5 and 6 confirmed this hypothesis: KNOWI-readiness correlated with other measures of motivation to connect with others. For generality, Study 7 replicated the ELMO effect using a different motive (the intimacy motive). Then Study 8 used the two KNOWI indices together to test the hypothesis that performance on a relevant laboratory task is a joint function of knowledge and motivation. The results confirmed this hypothesis by showing that superior knowledge led to superior performance only when motivation was high (as assessed in two different ways). Study 9, an experience sampling study, extended the findings to show how the KNOWI Scale predicts behavior in everyday life. Motivated and knowledgeable people more often sought support from partners at times of distress.

Studies 10 and 11 focused on a second knowledge structure, a structure containing script-knowledge. The relevant script was the secure base script, which concerns the sequence of events that unfold as a person in distress receives comfort from an attachment figure. Participants were shown animated film clips in which geometric figures depicted the secure base script; the participants were then asked to describe the animations. Participants who recognized the script were more likely to (a) see a common theme in the various animations, (b) later recall details of the animations, (c) include more elements in describing the animations, (d) report details with fewer distortions, and (e) create better organized stories. Thus, recognizing the secure base script enables a person to process information concerning secure base support and to correctly interpret supportive efforts of relationship partners. Finally, I showed that the two knowledge structures—indicator-knowledge and script-knowledge—are themselves correlated and that attachment avoidance is negatively related to both.