Basic question about neurons

Richard Norman rsnorman at
Sun Nov 25 11:45:26 EST 2001

On Sun, 25 Nov 2001 13:03:20 -0000, "Theophilus Samuels"
<theophilus.samuels at> wrote:
>chris ackerman <cma1114 at> wrote in message
>news:lumL7.44668$Ze5.25808688 at
>> I am a neophyte so forgive the simpleness of my question, but why is it
>> that I always read descriptions and see pictures of neurons as having
>> many dendrites and few axon terminals? If this is representative of most
>> neurons in the brain, how is it possible to have so many inputs and so
>> few outputs?
>  The question you ask is interesting. Physically speaking, neuronal input
>far exceeds its output (i.e. number of dendrites (+ other axons synapsing on
>the neuron) >>> single axon). However, looking a little deeper, this is not
>entirely the case when considering input and output with respect to the
>number of electrical signals recieved and transmitted per unit time.
>  Dendrites are almost never myelinated, and they respond with graded
>depolarisation or hyperpolarisation that decrements spatially and temporally
>(some may actually generate action potentials). Firstly, if we consider the
>time it takes for an AP to fire and the refractory period (say 5-10 ms) then
>theoretically the output (electrical signals or action potentials) of an
>axon ranges in the hundreds per second. Alright, considering that the number
>of synaptic inputs to a single axon can range into the 100,000s, the output
>number is still relatively small (although it is still 100 times more than
>you might have first expected). Secondly, even though it is highly
>debatable, as the AP is an all-or-nothing response, one might argue that
>they are all that matters since they're the absolute and final product of
>the culminated input signals. An analogy would be as follows: consider an
>election process. Many thousands of individuals will vote for a single
>candidate to win and represent the majority. It is this individual that
>represents the collective voice of the many and is therefore the most
>important consequence of the entire original electorial process. It is this
>person that is the first representative met by other elected persons and
>influences other decisions. So, if we now revert back to the neuron, we can
>see that the really important consequence of all input signals is the final
>action potential. And therefore, even though we might think that the output
>profile of a neuron is severely limited, in reality it must be adequate
>(perhaps even more than adequate!), otherwise we wouldn't be here to ask
>such questions.
It doesn't make too much sense to count number of synaptic inputs per
second against number of action potentials per second as a measure of
inputs vs outputs for a cell.  I believe the original question was
really comparing degree of convergence with degree of divergence in
the wiring pattern of the nervous system -- a purely spatial question.

You also don't realize the possibility that a tremendous number of
neurons in the vertebrate CNS are "local" and don't send axons out any
distance, if they have axons at all.  Neurons can work perfectly well
without making action potentials at all -- graded potentials in axons
can cause graded release of transmitter.  And there are many synapses
in "microcircuits" with a small cluster of cell processes (dendrites?
axons? -- whatever, it doesn't really matter) forming mutually
reciprocal synapses on each other.

But it still always comes down to the fact that the number of
post-synaptic endings (cell inputs) must equal the number of
pre-synaptic endings (cell outputs).

More information about the Neur-sci mailing list