The Action-Potential Magnitude Effect: Duration, Conductance History, Synaptic Transformation, and Phase-Wave Differentials in Self-Aware Networks

Action potentials are commonly described as all-or-none events. This description is indispensable but incomplete if it is interpreted to mean that every successful spike is physically or informationally identical. Spike height, width, rise and repolarization kinetics, afterpotentials, burst position, axonal propagation, and terminal waveform can vary with cell type, compartment, temperature, ion-channel state, neuromodulation, and recent input history. These variations can, in bounded preparations, alter presynaptic calcium entry and neurotransmitter release. The unresolved computational problem is how a state-dependent cellular event becomes a receiver-specific consequence without confusing waveform, calcium, release, postsynaptic response, and population power as if they were one variable. The proposed solution is a conditional mediation chain: receptor- and synapse-specific input history changes membrane and channel state; that state shapes a cellular event; axons and boutons transform the event; calcium and release translate it probabilistically; and a receiver responds according to its own immediate state and learned organization. Self-Aware Networks (SAN) calls the computational use of this multidimensional, transformed, receiver-relative event the Action-Potential Magnitude Effect (APME). Here magnitude names the consequential event vector and its transformations, not peak voltage, duration, calcium, vesicle count, postsynaptic current, or population power alone. SAN further hypothesizes that receiver-relative differences can contribute to phase-wave differentials (PWDs), Neural Tuning, and distributed neural rendering. Recognition, novelty, surprise, whisker contact, umami, color, and multimodal objects are candidate experimental domains, not universal causes of spike broadening. This paper formalizes the APME chain, separates population-level 1/f power from single-spike waveform measurements, defines typed variables across cellular and network scales, and proposes experiments that compare waveform-aware models against spike-count, rate, timing, phase, and stimulus-history alternatives. It also specifies falsifiers. APME is supported only if waveform variables add reproducible, receiver-specific predictive and causal value after the relevant covariates are controlled. The proposal therefore preserves the all-or-none principle while making its empirical and computational boundary more precise.