Spectral dependence as a framework for neural coordination.

Theories of large-scale neural coordination frequently assign distinct cognitive functions to discrete, independent frequency bands of oscillatory activity. In the hippocampus, memory encoding and recall are attributed to specific non-overlapping gamma frequencies relayed through separate anatomical pathways. These "spectral parcellation" models assume frequency bands operate as functionally meaningful units that must be selectively generated, maintained, and decoded. However, this framework is inconsistent with biophysical constraints governing synaptic integration, current flow, and excitatory-inhibitory balance. An alternative framework treats oscillatory structure as an emergent consequence of multiscale circuit dynamics under shared constraints. Slow, large-amplitude oscillations such as theta (4-12 Hz) organize inter-regional coordination while higher frequencies reflect local dissipation of synaptic energy through progressively smaller, faster circuit motifs. Cross-frequency coupling arises naturally from this energy redistribution rather than from dedicated multiplexing mechanisms. This "energy cascade" framework makes explicit, falsifiable predictions about how spectral structure scales with behavioral state, metabolic demand, and experimental perturbation. We review evidence from hippocampal theta-gamma interactions demonstrating that: (1) gamma power covaries with theta power, (2) gamma properties shift continuously rather than discretely with circuit state, and (3) perturbations propagate hierarchically across frequencies rather than selectively disrupting isolated bands. These findings support constraint-based models over frequency-parcellation accounts and suggest that oscillations reflect how circuits dissipate energy under physical constraints rather than serving as discrete communication channels.