My approach is as follows:
a = dv/dt
= (dv/dx) * (dx/dt)
= (dv/dx) * v
Putting v = 0:
a = (dv/dx) * 0 = 0 m s^(-2)
But, what I don't understand is this:
If v=0, then dx/dt must also be 0. Consequently, dx must also be 0 at that particular instant. But, we are writing acceleration as (dv/dx) *...