The relationship pet owners have with their animals was examined in a series of studies. In Study 1 survey questionnaires were used to investigate the demographic variables related to pet ownership in 312 New Zealand families of 8-12 year oIds. Almost 90% of families owned at least one pet, and over half of these families included a child who was the sole owner of a pet. Parental employment level, living locality, and sibling status (number and position) were related to pet ownership. Parents acquired pets for their children mainly to teach responsibility and care, or because their child had asked for the pet, and these reasons were related to sibling status. Parents who had owned pets before their children were born were more likely to own pets and attribute "family member" status to pets. In Study 2 the intergenerational continuity of attitudes to pets throughout three generations was examined by survey questionnaires. Results indicated that intergenerational continuity of attitudes regarding pet ownership, attachment to pets, and species of pets exist through family generations. The main study, Study 3, examined the language used to talk to dogs, and compared it to the language used in addressing infants. Results indicated few differences between speech to dogs and speech to children. Based on these findings, and Hummert and Ryan's model of patronising speech, a model of speech to dependents was developed, incorporating the dimensions of care, control, and communication. Implications for future research include the distinction between child-owned and family-owned pets, and comparison of parent and child beliefs about the role of the family pet. The model of simplified speech to dependents also requires further testing of speech to dogs and to other recipients of simplified speech.