This article examines how early twentieth-century comparative psychology treated behaviour as a primary site for studying regulation, dysfunction, and adaptation, using experimental neurosis as a historical case study. Beginning with Pavlov's observation that dogs exposed to irresolvable conflict developed persistent behavioural and autonomic disturbances, the paradigm expanded through mid-century work by Liddell, Gantt, and Masserman, who introduced new species, longitudinal designs, and ecologically grounded conditions revealing dysregulation as a multilevel, temporally extended phenomenon. In this tradition, behaviour was not construed as mechanistic output but as an autonomous level of organisation whose structured patterns, i.e., variability, context-sensitivity, stability, and breakdown, made internal regulatory dynamics experimentally accessible in ways physiology alone could not. Hybrid methodologies integrating laboratory control with naturalistic observation showed that pathological breakdown provides access to organisational features obscured during normal function. By mid-century, however, this behaviourally rich framework was displaced by reductionist approaches based on standardised assays and physiological endpoints. Recent historiography documents this shift as an epistemological transformation in which behaviour was redefined from interface to index, from organised system to dependent variable. These tensions persist in contemporary neuroscience and translational research, where behaviour is routinely subordinated to mechanism despite gaps between circuit-level manipulation and behavioural explanation. This historical analysis supports the claim that mechanistic approaches which bypass behaviour incur explanatory loss, eliminating the domain in which neurophysiological processes acquire functional meaning. Progress in understanding psychopathology requires restoring behaviour to ontological and epistemological centrality as the irreducible interface through which adaptive and pathological regulation become scientifically accessible.
Rodrigo J. De Marco (Wed,) studied this question.