Invited commentary on animal models in psychiatry: animal models of non-conventional human behavior.

Conventional behavior, of which linguistic behavior is the principal variety, is identified as responses having formal properties that are not determined by the natural properties of stimulus objects, but instead by properties attributed to those objects under the auspices of particular groups. Given the ubiquity of this type of behavior in the repertoires of human beings and its complete absence in those of non-humans, the argument is made that animal models of human disorders, in which disturbances of conventional behaviors constitute defining features, are not sufficiently analogous to these conditions in humans to be pursued with good result. Because conventional behavior of the linguistic type is ubiquitous in the repertoires of normally developed human adults, it is suggested that the behavior of pre-verbal infants and/or non-verbal persons is preferable to that of adults as the phenomenal source for the construction of animal models of human psychological events. The observation and measurement of psychological events is held to be complicated by a number of their characteristics, including their complexity by virtue of whole organism participation, their essential complementarily with stimulus events, and the corrigibility of both form and function over their repeated occurrences, among others. The implications of these features for modeling enterprises are discussed.