LECTURE 22
The Mean Value Theorem
We’ve just seen how derivatives can be used to find and classify the local maxima and minima of a function.
In fact, with just a little more work we can actually obtain a sketch of what the graph of a function has to
look like without actually plotting the graph point by point.
Before doing so, however, I need to state the fundamental theorems upon which this sort of analysis is
based.
We’ll start with Rolle’s theorem.
Theorem 22.1 (Rolle’s Theorem). Let f be a function that satisfies the following three conditions
(i) f (x) is continuous on the closed interval [a, b]
(ii) f (x) is differentiable on the open interval (a, b)
(iii) f (a) = f (b)
Then there is a point c ∈ (a, b) such that f (c) = 0.
Proof. There are three possibilities to consider.
• f (x) = K, a constant.
In this case, f (x) = 0 identically, and so for any number c ∈ [a, b] we will have f (c) = 0
• f (x) > f (a) at some point x ∈ (a, b).
By the Extreme Value Theorem, we know that f attains a maximum value somewhere in [a, b].
However, since f (a) < f (x), and f (b) < f (x) (by condition (iii)) this maximum will not occur at
the endpoints of the interval. So it must be a local maximum inside the interval; say this maximum
occurs at x = c. But then since f has a local maximum at x = c we must have f (c) = 0.
• f (x) < f (a) at some point x ∈ (a, b).
The Extreme Value Theorem, tells us that f attains a minimum value somewhere in [a, b], and
since there exists x ∈ (a, b) such that f (x) < f (a) = f (b), this minimal value doesn’t occur at the
endpoints a or b. Thus, there must be a local minimum inside (a, b), say at the point x = c. But
then since x = c is a local minimum of f , we have f (c) = 0.
Example 22.2. Prove that x3 + x − 1 has exactly one real root.
• First we show that x3 + x − 1 has at least one real root. To see this, we note f (x) = x3 + x − 1 is
continuous and
f (0) = −1
f (1) = 1
By the Intermediate Value Theorem, we know that since 0 lies between f (0) and f (1), somewhere
in the inteval (0, 1) there’s a number c such that f (c) = 0. That number c will of course also satisfy
c3 + c − 1 = 0.
96
22. THE MEAN VALUE THEOREM 97
Now suppose we had two numbers c1 and c2 such that f (ci ) = 0. Then Rolle’s Theorem tells
us that there is a point d between c1 an c2 such that f (d) = 0. However,
f (x) = 3x2 + 1 ≥ 1 for all x
We conclude that we cannot have two such roots c1 and c2 without contradicting Rolle’s Theorem.
So the statement is proved.
Theorem 22.3 (The Mean Value Theorem). Let f be a function that satisfies
• f is continuous on the closed interval [a, b]
• f is differentiable on the open interval (a, b)
Then there is a number c ∈ (a, b) such that
f (b) − f (a)
f (c) =
b−a
Proof.
Consider the function
f (b) − f (a)
h (x) = f (x) − f (a) − (x − a)
b−a
(This happens to be the function that measures that measures the vertical distance between the graph of
f (x) and the secant line through (a, f (a)) and (b, f (b)).) Since f is continuous and differentiable and since
f (b) − f (a)
g (x) = − (x − a)
b−a
is continuous and differentiable, so will be h (x) = f (x) + g (x). We also have
f (b) − f (a)
h (a) = f (a) − f (a) − (a − a) = 0
b−a
f (b) − f (a)
h (b) = f (b) − f (a) − (b − a) = 0
b−a
So we can apply Rolle’s Theorem to conclude that somewhere between x = a and x = b there is a point
x = c where h (c) = 0. But then
f (b) − f (a) f (b) − f (a)
0 = h (c) = f (x) − = f (c) −
b−a x=c b−a
Hence there is a point c ∈ (a, b) where
f (b) − f (a)
f (c) =
b−a
Corollary 22.4. Suppose f (x) = 0 for all x in an interval [a, b], then f (x) is a constant function on
(a, b).
Proof. Let x1 , x2 be any two points inside the interval [a, b] chosen so that a < x1 < x2 < b. By the Mean
Value Theorem we know that there exists c ∈ (x1 , x2 ) so that
f (x2 ) − f (x1 )
f (c) =
x2 − x1
But f (x) = 0 everywhere in [a, b]. Hence,
f (x2 ) − f (x1 )
0= =⇒ f (x2 ) = f (x1 )
x2 − x1
for all points x1 , x2 ∈ (a, b). We conclude that f is constant on (a, b).
22. THE MEAN VALUE THEOREM 98
Example 22.5. Prove the identity
π
tan−1 (x) + cot−1 (x) =
2
Set
f (x) = tan−1 (x) + cot−1 (x)
Using the identities
d 1
tan−1 (x) =
dx 1 + x2
d 1
cot−1 (x) = −
dx 1 + x2
we find
1 1
f (x) = − =0
1 + x2 1 + x2
Therefore, by the Corollary above, f (x) must be a constant.
tan−1 (x) + cot−1 (x) = C
To figure out the constant we can put x = 1. Now when
sin (θ)
1 = tan (θ) = =⇒ θ = 45◦ ∼ π/4 radians
cos (θ)
so tan−1 (1) = π/4. Similarly,
cos (θ)
1 = cot (θ) = =⇒ θ = 45◦ ∼ π/4 radians
sin (θ)
Hence,
π π π
C = tan−1 (1) + cot−1 (1) = + =
4 4 2
Thus, the constant C is π/2. The stated identity now follows.
