The all-or-nothing principle indicates that a nerve cell fires at maximum potential or not at all, based on a threshold on the stimulus.
Is this a statement which is always true, or only mostly-always true?
In electrical engineering, a discipline I have studied, we have a similar rule, a logic signal is responded to as though it is either a high or a low. However, in the middle are what are known as "metastable states," where the rules of thumb do not apply. In fact, you can even theoretically lock up a logic circuit by presenting a steady stimulus in between two states. Needless to say, designers of such circuits take great effort to ensure their logic is never exposed to such stimulus.
Do we have any evidence of neurons exhibiting behavior which defies the all-or-nothing law in some circumstances, or if not, what sorts of metastable behavior are seen when the stimulus is exceedingly close to that of the threshold?