to substantiate recommendations is still in the future. We know that the needed social companionship depends on species, age, sex, and rearing history and ranges from sensory contact to direct association with multiple other animals. We know that the utility of devices designed to provide opportunities for foraging and exploration depends on the same variables. We know that housing requirements depend on bodily dimensions, normal postures, and locomotor patterns (e.g., arm swinging versus bipedal leaping versus quadrupedal striding or climbing). We know that some animals rarely tolerate close human contacts, whereas others seem to respond positively to familiar humans. Quantitative specification of minimums for each characteristic might be considered desirable by some, but it is important that standards be validated against performance; that is, do the specified characteristics actually improve the psychological well-being of the affected animals?
Nevertheless, a comprehensive program to improve the psychological well-being of nonhuman primates will attend to each of the variables and include a means to test and assess the influence of each. The benefits of providing a cage companion, an enrichment device, a cage of design and dimensions that appeal to a human's aesthetic sense, and a sensitive caregiver to interact with the animals should be validated and documented. A performance standard should be used to show whether the provided features increase the diversity and amount of normal behavior and decrease the frequency and duration of behavior that results in self-injury or other undesirable consequences. Social companions must be conducive to positive affiliation rather than be a source of stress and a cause of avoidance; an inappropriate companion can be worse than no companion at all. Enrichment devices do their job when they provide otherwise-absent opportunities to engage in species-typical foraging and exploratory activities, but a device that requires excessive time for foraging at the expense of social or other species-typical activities is not an enrichment. Likewise, enrichment devices can be used by multiple animals, but they should not become a source of competitive conflicts. Cage dimensions and furnishings are suitable if they allow for expression of normal postures and locomotor expression, as opposed to open unused spaces and furnishings that do not allow for normal postural positions. Human interactions should provide for activities that the animals appreciate, rather than simply provoke animal activity.
The four bulleted items listed above are related. Whereas spatial requirements are based on individual needs, social housing need not mean that spatial requirements for one animal are multiplied by the number of animals housed. If housing provides sufficient space for one animal to express a normal locomotor pattern, a second animal has access to the same space for locomotor expression minus only the volume of space actually occupied by the first animal. For example, if housing needs to be 2 m high to provide sufficient vertical space for an animal's postural and locomotor needs, two animals do not need vertical space 4 m high. Floor areas likewise need not be simple multiples; in fact, for arboreal