This book investigates children's security of attachment in infancy and its relationship to their cognitive development in the preschool years, presenting evidence that caregivers' proclivity to treat their infants as mental agents and to attribute intentionally to their behavior is critical to their child's cognitive development. The book critiques attachment theory, examines correlates and consequences of attachment security, and details four of the author's studies as well as experiments by Fernyhough (1994). Study 1 found that securely attached toddlers achieved a higher level of search than insecurely attached toddlers, thus providing evidence for greater self-efficacy. Study 2 found vocabulary differences between securely and insecurely attached children on referential language acquisition measures. Using mother as a secure base and punctuating exploration with references to mother may initiate an interaction process whereby objects and surroundings direct communication. Study 3 found that securely attached 2.5-year-olds benefited more from an experimenter's play suggestions than did insecurely attached children, thereby suggesting securely attached children's greater social flexibility. Study 4 found that mothers of securely attached children gave positive rather than negative feedback on a collaborative box construction task, intervened physically only when requested, and used feedback to alter subsequent comments. This difference reflected mothers' proclivity to construe their children as mental agents. Fernyhough's experiments found that securely attached 4-year-olds were more likely than insecurely attached peers to pass an unexpected transfer task and that securely attached 5-year-olds performed better on tasks requiring an understanding of someone else's knowledge of a picture or of how emotions are determined by beliefs. (Contains approximately 280 references.) (KB)