Why is it not necessary to express a variable as a function of time in order to determine the derivative with respect to time?

In Chapter 4 of Tom M. Apostol's book Calculus Volume I Second Edition, he discusses the chain rule and its applications. Everything seems fine until related rates.

Why is it "not necessary to express [a variable] as a function of [time] in order to determine the derivative" with respect to time?

• Can you please give an example? The question is a bit unclear without context. – Franklin Pezzuti Dyer Jul 15 '17 at 18:19

For context, the question Apostol asks is, "(Regarding an expanding balloon) How fast is the radius of the balloon increasing when the radius is $5$ cm?
What Apostol is saying is that even though the radius, $r(t)$, is a function of time, we don't need to explicitly need to express $r(t)$ as a function, e.g. $r(t) = 2t$, in related rates problems. We're only interested in the quantity, $\frac{d}{dt}r(t)\biggr|_{t=5}$.
$$\dfrac{dV}{dt}\biggr|_{t=5} = 4\pi(r(5))^2\frac{d}{dt}r(t)\biggr|_{t=5}$$