I've read this from wikipedia:
... This relationship holds only for the unit of tissue; for nervous tissue the unit is the nerve cell, for skeletal muscle the unit is the individual muscle fiber and for the heart the unit is the entire auricles or the entire ventricles.
https://en.wikipedia.org/wiki/All-or-none_law
I know about all-or-none principle but i don't understand this statement. What it is trying to imply? I am not sure my question title is reasonable because of my less understanding.
The "all-or-none" principle refers to the roughly accurate concept that neurons fire all-or-none action potentials, individual muscle fibers contract fully or not at all, etc.
If you have a population of, for example, 5 neurons, the population response is no longer all or none, even if the individual units follow the principle. You could have 0, 1, 2, 3, 4, or 5 cells firing an action potential together (the definition of "together" is somewhat vague; depending on the location/function of a neuron you could consider spikes within 1 ms to be synchronous or you could consider spikes within 100 ms to be synchronous, for example). The more units you have, whether they are neurons, muscle fibers, etc, the more "smooth" the possible outputs are.
The reasons for different numbers of cells being recruited could be differences in thresholds, but it could also be from different inputs to each of the units, different firing histories of each of the units, etc. This arrangement allows for more "analog" responses both within the nervous system and at muscles, because the strength of a signal or the strength of a contraction can be regulated by how many neurons participate, or how many muscle fibers are contracted.
One everyday example of the importance of this system is your grip intensity when holding an object. If your grip was all-or-none, you would not be able to handle a disposable coffee cup without crushing it. By using a population response rather than a single neuron connected to all the fibers in a single muscle, the nervous system can control the strength of a muscle contraction. The same principle applies within the CNS, where population activity can encode seemingly continuous representations of features like colors, frequencies, sizes, etc, despite an underlying code that is all-or-none at the single unit level.
