Is adjunctive behavior a third class of behavior?

Research during the past 20 years has revealed that intermittent food presentation to a variety of organisms results in an inordinately excessive consumption of water as well as other behaviors including attack, pica, escape, and alcohol consumption. Such behavior has not been thought to be either respondent or operant behavior, but instead has been regarded as a new behavioral category termed "adjunctive behavior." This paper reexamines the rejection of adjunctive behavior as either operant or respondent behavior and concludes that the rejection was premature. This paper also reexamines the argument that there is a unique class of adjunctive behavior and concludes that there is not. It is recommended that given the growing difficulties in maintaining an operant-respondent dichotomy, rather than admitting adjunctive behavior into either of these categories, research efforts should de-emphasize preconceived and outdated notions about categories of behavior and how such behaviors should behave and instead focus on mapping functional relationships between behavior and various procedures for presenting non-contingent stimuli.