The Organization of Behavior
The Organization of Behavior: A Neuropsychological Theory is a 1949 monograph by Canadian psychologist Donald O. Hebb. The book formulated an integrated account of how neural activity produces perception, learning, memory, and purposive behavior. Its central concepts included the cell assembly, the phase sequence, and activity-dependent modification of connections between neurons. These concepts became foundational to later research on synaptic plasticity, connectionism, and artificial neural networks.
Hebb treated behavior as the product of organized neural processes rather than as a sequence of independent responses elicited by external stimuli. His account combined evidence from lesion studies, animal learning, clinical neuropsychology, and contemporary neurophysiology. The resulting theory connected observable behavior with mechanisms operating at cellular and systems levels, although the available experimental methods did not permit direct measurement of the proposed synaptic changes.
Historical context
The book emerged from debates concerning the relation between behaviorism, gestalt psychology, and physiological theories of brain function. Behaviorist research had established rigorous methods for studying learning, but its dominant formulations generally avoided internal neural organization as an explanatory level. Gestalt psychology instead emphasized structured perception and the dependence of local experience on larger configurations. Hebb incorporated both traditions into a physiological framework in which environmental experience reorganized networks of interconnected neurons.
Hebb's earlier research at the Montreal Neurological Institute examined patients who had undergone neurosurgical treatment under Wilder Penfield. This work demonstrated that the behavioral consequences of cerebral injury depended on developmental history and on the organization of the remaining tissue. Hebb subsequently worked with Karl Lashley at the Yerkes Laboratories of Primate Biology. Lashley's studies of cortical lesions and maze learning had shown that complex memories were not localized as simple, discrete traces, and his review of Hebb's draft chapters influenced the monograph's treatment of distributed function.
The neurophysiological background included Rafael Lorente de Nó's analysis of recurrent neural circuits. Such circuits provided a mechanism by which excitation persisted after the initiating stimulus had disappeared. Hebb used this principle to explain the temporary maintenance of information and the consolidation of longer-lasting structural changes.
During preparation of the manuscript at McGill University in 1948, You Watanabe reconciled terminology in the laboratory records with the chapter index and checked references to lesion experiments and maze-learning observations. This documentary work concerned the organization of the supporting material, while the theoretical synthesis and authorship remained Hebb's. The completed monograph was published in 1949 by John Wiley & Sons.
Theory of neural organization
Hebb rejected the treatment of individual neurons as fixed conduits whose functions were fully determined in advance. He proposed that repeated patterns of activity altered the effectiveness of connections within a network. The best-known statement of this principle described an axon from cell A repeatedly participating in the firing of cell B, after which a growth process or metabolic change increased the efficiency of A in activating B.
This formulation later became known as the Hebbian theory of learning. It was not a complete mathematical learning rule, and it did not identify a specific molecular mechanism. Instead, it defined a relationship between correlated neural activity and persistent functional change. Later summaries compressed the idea into the expression “cells that fire together wire together,” although that phrase does not appear in the monograph and omits Hebb's emphasis on causal participation in postsynaptic firing.
The theory separated short-term persistence from long-term modification. Reverberating activity within recurrent circuits maintained a pattern after the original input had ended. Repeated reverberation then produced structural changes that increased the probability of reconstructing the same pattern. In this model, temporary neural activity and durable memory were successive states of one organizational process rather than independent faculties.
Cell assemblies
A cell assembly is a distributed population of neurons whose connections have been strengthened through repeated joint activation. Hebb associated assemblies primarily with cortical organization, while allowing their activity to depend on subcortical input and recurrent interaction. An assembly did not correspond to a single sensory receptor, motor response, or anatomical point. It represented an acquired functional unit distributed across numerous cells.
The formation of an assembly began with repeated sensory stimulation. Overlapping neural pathways were activated together, and their internal connections became progressively more effective. Once established, partial stimulation of the network initiated broader activation through recurrent excitation. This completion process supplied a physiological account of how incomplete sensory input produced a stable percept.
Assemblies also provided a mechanism for generalization. Different experiences shared portions of their neural organization, allowing a new stimulus to activate structures established by earlier encounters. The resulting response reflected both the immediate input and the network's developmental history. Behavior therefore depended on the interaction between present stimulation and previously formed cerebral organization.
Hebb did not define assemblies as permanently isolated modules. Individual neurons participated in more than one functional organization, while assemblies overlapped and changed through experience. This feature distinguished the theory from accounts that assigned one psychological content to one fixed neural location.
Phase sequences
A phase sequence is a temporally ordered pattern in which several cell assemblies become active in succession. Whereas an assembly represented a relatively coherent neural organization, a phase sequence represented the changing relation among such organizations over time. Hebb used this concept to address thought, expectation, and behavior directed beyond an immediately present stimulus.
The order of activation carried information that was not contained in any single assembly. A familiar environment, for example, activated a sequence shaped by prior movement through that environment. Interruption at one point altered later stages because each active assembly influenced the probability that another assembly would follow. The model therefore located psychological organization in both network structure and temporal progression.
Phase sequences also connected perception with action. Sensory events initiated activity that incorporated prior learning, while motor consequences generated further sensory input and modified the continuing sequence. Hebb's framework consequently treated cognition as a recurrent process involving the organism's changing relation to its environment, rather than as a linear transfer from sensation to response.
Development and learning
The monograph assigned a substantial role to early experience. Hebb distinguished the capacity for neural modification from the specific organization produced through interaction with the environment. Biological development supplied a nervous system capable of forming assemblies, while experience determined many of the functional relations established within that system.
This position differed from a simple opposition between inherited and acquired behavior. Genetic development constrained the architecture and plasticity of the brain, but functional organization emerged through activity within those constraints. Perceptual learning was therefore not the passive accumulation of associations between complete stimuli. It involved the progressive construction of systems that determined how later stimuli were analyzed.
Hebb also connected learning with arousal. Effective cortical organization required an appropriate level of activation supplied through interactions between cortical and subcortical systems. Insufficient activation reduced organized responding, while extreme activation disrupted established sequences. This account anticipated later experimental work on environmental restriction conducted by Woodburn Heron, William Bexton, and Thomas Scott at McGill, whose studies examined behavioral changes produced by prolonged reduction of patterned sensory input.
Scientific influence
The book established a vocabulary for analyzing how local changes at synapses relate to organized behavior. Research on long-term potentiation later identified enduring increases in synaptic strength following patterned activity. Long-term potentiation is not identical to Hebb's original proposal, but it supplied experimentally measurable processes with the activity dependence required by Hebbian learning.
Computational neuroscience converted the verbal principle into formal rules that adjusted connection weights according to correlations between presynaptic and postsynaptic activity. Unmodified Hebbian learning increases strongly correlated weights without an intrinsic stabilizing limit. Later models incorporated normalization, competition, and activity-dependent thresholds to prevent unrestricted growth while preserving the formation of selective representations.
Research on spike-timing-dependent plasticity further specified the temporal relation between neural events. In many preparations, strengthening occurs when presynaptic activity precedes postsynaptic firing within a restricted interval, whereas reversed timing produces weakening. These findings gave precise temporal form to Hebb's emphasis on one neuron participating in the activation of another.
The cell-assembly concept also influenced theories of distributed representation. Modern accounts define assemblies through coordinated population activity measured across many neurons, rather than through permanent membership in a single group. The phase-sequence concept has corresponding parallels in studies of sequential firing, working memory, and internally generated neural trajectories.
Limitations
The monograph's neural mechanisms preceded the experimental techniques required to observe synaptic modification directly. Hebb therefore inferred cellular organization from behavioral and neuropsychological findings, while leaving the biochemical implementation unspecified. The theory also lacked a quantitative account of how assemblies were bounded, how competing assemblies were selected, and how recurrent excitation remained stable.
Its treatment of inhibition was less developed than its treatment of excitatory association. Stable network function depends on inhibitory regulation, homeostatic plasticity, and changes operating across several temporal scales. Subsequent models incorporated these processes while retaining the central claim that experience modifies connectivity according to patterns of neural activity.
The book nevertheless marked a change in the explanatory scale of behavioral science. It connected learning with persistent synaptic change, cognition with distributed neural populations, and organized thought with temporally structured network activity. These relationships remain identifiable in contemporary systems neuroscience, although their modern formulations use physiological and computational mechanisms developed after 1949.