Information processing in neurobiological systems is commonly thought to rely on the assessment of a signal-to-noise ratio as the key mechanism of signal detection; it assumes and requires that both signal and noise are concurrently available. An alternative theory holds that detection proceeds by the system appreciating any instantaneous input by the input's departure from the moving average of past activity. The evidence reviewed here suggests that this latter transduction mechanism provides a unique, formal account of the highly dynamic, neuroadaptative plasticity (i.e., tolerance, dependence, sensitization) that ensues upon mu-opioid receptor activation. The mechanism would appear already to operate with the receptor-G protein coupling that occurs upon agonist binding to mu-opioid receptors, and also with highly integrated responses such as whole-organism analgesia. The mechanism may perhaps operate ubiquitously with further neuronal and non-neuronal, cell surface, and intracellular-signaling systems, and may govern the experience-dependent regulation of synaptic strength. The transduction mechanism defines a continuously evolving process; the process's most peculiar feature is that it makes any input generate not one but two outcomes that are paradoxical, or opposite in sign.