You'll watch say ten cows come down to that [water] bowl. The boss cow gets to drink first. Of that ten, there'll be two that'll say, 'Ah, the hell with this', and they'll go downThese were the words of a cattle rancher from western Canada recorded during interviews that tried to capture farmers' and ranchers' views about animal welfare (Spooner, Schuppli, & Fraser, 2012). The quote indicates a little of the rancher's understanding of cattle: that they are individuals with very different personalities, that they have relationships with each other -sometimes hierarchical -based on individual recognition, and that they form specific affiliations with other members of the herd.This kind of "everyday understanding" is possessed by a great many farmers and ranchers whose work revolves around animals. It is based on what scientists might term "narrative data" (Fraser, 2009): descriptive observations of how specific individuals interact with other specific individuals, how they deal with the environment in their individual ways, how they respond to different human beings, and how all this changes (or remains consistent) in different individuals over time with maturation and learning. This type of information gives farmers and ranchers an understanding of cattle that is similar in many respects to the understanding that observant dog-owners have of their dogs, or that attentive zoo keepers develop of the animals in their charge.Some scientists develop a similar understanding of animals through research. As one example, Goodall's (1971) narrative accounts of the lives of chimpanzees led her to recognize emotions and cognitions in the animals, and a similar approach has been used on other species (e.g., Moss, 1988;Smuts, 1999). However, the everyday understanding of animals possessed by farmers and ranchers is quite different from the scientific understanding that comes from much of the formal research on animal behavior, partly because of the dominant research paradigm that behavioral scientists have traditionally used: studying animals under experimental conditions rather than observing their normal lives; treating animals as exemplars of a species, not as unique individuals; and relying on quantitative data while