Unit 4, lesson 35

The Mean Value Theorem

Why a smooth function must at some point have an instantaneous rate of change equal to its average rate, which hypotheses that needs, and how it proves that a zero derivative means a constant function.

  • Theory part 69
  • Practice part 70
  • Mean Value and Monotonicity

Theory, part 69

Practice, part 70

Suppose you drive 150 kilometers in 2 hours. Your average speed is 75 kilometers per hour, and at some instant your speedometer must have read exactly 75. You may have gone slower and faster, but you cannot have been below 75 the whole time or above it the whole time. The mean value theorem turns this everyday reasoning into a precise statement, and much of the rest of calculus leans on it.

Key terms

Rolle’s theorem

Definition Let a<ba < b and let ff be a function with these three properties.

  1. It is continuous on the closed interval [a,b][a, b].
  2. It is differentiable on the open interval (a,b)(a, b).
  3. Its values at the ends agree, f(a)=f(b)f(a) = f(b).

Then there is at least one number cc in (a,b)(a, b) with f′(c)=0f'(c) = 0.

The proof uses the extreme value theorem. If ff is constant, every cc in (a,b)(a, b) works. Otherwise some value of ff on [a,b][a, b] is above or below the common end value f(a)=f(b)f(a) = f(b). Then the global maximum or the global minimum, which exist by continuity, is attained at an interior point cc, and Fermat’s theorem gives f′(c)=0f'(c) = 0 because ff is differentiable at cc.

In plain words If a smooth curve starts and ends at the same height, somewhere in between it has a horizontal tangent.

Example Let f(x)=x2−4xf(x) = x^2 - 4x on [0,4][0, 4]. It is a polynomial, so it is continuous and differentiable, and f(0)=f(4)=0f(0) = f(4) = 0. Solving f′(c)=2c−4=0f'(c) = 2c - 4 = 0 gives c=2c = 2, which lies in (0,4)(0, 4).

Mean value theorem

Definition Let a<ba < b. If ff is continuous on [a,b][a, b] and differentiable on (a,b)(a, b), then there is at least one number cc in (a,b)(a, b) with

f′(c)=f(b)−f(a)b−a.f'(c) = \frac{f(b) - f(a)}{b - a}.

It follows from Rolle’s theorem applied to ff minus its secant line,

g(x)=f(x)−f(a)−f(b)−f(a)b−a(x−a),g(x) = f(x) - f(a) - \frac{f(b) - f(a)}{b - a}(x - a),

which is continuous on [a,b][a, b], differentiable on (a,b)(a, b) and has g(a)=g(b)=0g(a) = g(b) = 0.

In plain words Somewhere between aa and bb the tangent line is parallel to the secant line joining the two ends. The instantaneous rate equals the average rate at least once.

Example Let f(x)=x3f(x) = x^3 on [0,2][0, 2]. The average rate is

f(2)−f(0)2−0=8−02−0=4.\frac{f(2) - f(0)}{2 - 0} = \frac{8 - 0}{2 - 0} = 4.

Solving 3c2=43c^2 = 4 gives c=23≈1.155c = \frac{2}{\sqrt{3}} \approx 1.155, which is in (0,2)(0, 2). The other solution, −23-\frac{2}{\sqrt{3}}, is outside the interval.

The hypotheses

Definition The conditions continuity on [a,b][a, b] and differentiability on (a,b)(a, b) are part of both theorems. Differentiability is not needed at the endpoints, and continuity is needed at every point including the endpoints. If either condition fails, the conclusion can fail. The conditions are sufficient, not necessary, so a function can break them and still have a suitable cc.

In plain words One corner or one jump is enough to break the theorem. It is not enough to check only that the function is defined.

Example The function f(x)=∣x∣f(x) = \lvert x \rvert on [−1,1][-1, 1] has f(−1)=f(1)=1f(-1) = f(1) = 1, yet f′(x)f'(x) is −1-1 or 11 and never 00. Rolle’s theorem does not apply because ff is not differentiable at 00. The function g(x)=xg(x) = x for 0≤x<10 \le x < 1 with g(1)=0g(1) = 0 has g(0)=g(1)g(0) = g(1) and g′(x)=1g'(x) = 1 on (0,1)(0, 1). Here continuity at the endpoint 11 fails.

Cauchy’s mean value theorem

Definition Let a<ba < b, and let ff and gg be continuous on [a,b][a, b] and differentiable on (a,b)(a, b). Then there is a number cc in (a,b)(a, b) with

(f(b)−f(a))g′(c)=(g(b)−g(a))f′(c).\big(f(b) - f(a)\big) g'(c) = \big(g(b) - g(a)\big) f'(c).

If in addition g′(x)≠0g'(x) \neq 0 for all xx in (a,b)(a, b), then g(b)≠g(a)g(b) \neq g(a) and

f(b)−f(a)g(b)−g(a)=f′(c)g′(c).\frac{f(b) - f(a)}{g(b) - g(a)} = \frac{f'(c)}{g'(c)}.

The proof applies Rolle’s theorem to

h(x)=(f(b)−f(a))g(x)−(g(b)−g(a))f(x),h(x) = \big(f(b) - f(a)\big) g(x) - \big(g(b) - g(a)\big) f(x),

which has h(a)=h(b)=f(b)g(a)−f(a)g(b)h(a) = h(b) = f(b) g(a) - f(a) g(b). If g′≠0g' \neq 0 on (a,b)(a, b), Rolle’s theorem applied to gg itself shows g(b)≠g(a)g(b) \neq g(a), so the division is allowed. Taking g(x)=xg(x) = x gives back the ordinary mean value theorem.

In plain words Think of (g(t),f(t))(g(t), f(t)) as a point moving along a curve. Somewhere the direction of motion is parallel to the chord joining the start and the end. This version is the key step in proving l’Hôpital’s rule.

Example Take f(x)=x2f(x) = x^2 and g(x)=x3g(x) = x^3 on [0,1][0, 1]. The equation is

(1−0)⋅3c2=(1−0)⋅2c,(1 - 0) \cdot 3c^2 = (1 - 0) \cdot 2c,

so c(3c−2)=0c(3c - 2) = 0, and the solution in (0,1)(0, 1) is c=23c = \frac{2}{3}.

Constant functions

Definition If f′(x)=0f'(x) = 0 for every xx in an open interval II, then ff is constant on II. To prove it, take any x1<x2x_1 < x_2 in II. The function is differentiable, hence continuous, on [x1,x2][x_1, x_2], so the mean value theorem gives a cc between them with

f(x2)−f(x1)=f′(c)(x2−x1)=0.f(x_2) - f(x_1) = f'(c)(x_2 - x_1) = 0.

The same holds on any interval if ff is continuous on it and f′=0f' = 0 at its interior points.

In plain words A function whose slope is zero everywhere on an interval cannot change. The word interval matters. On a domain with a gap, the function can take different constant values on the separate pieces.

Example The function f(x)=arctan⁡x+arctan⁡(1x)f(x) = \arctan x + \arctan\left(\frac{1}{x}\right) has

f′(x)=11+x2−1x2⋅11+1/x2=11+x2−1x2+1=0f'(x) = \frac{1}{1 + x^2} - \frac{1}{x^2} \cdot \frac{1}{1 + 1/x^2} = \frac{1}{1 + x^2} - \frac{1}{x^2 + 1} = 0

for every x≠0x \neq 0. It is constant on (0,∞)(0, \infty), where f(1)=π4+π4=π2f(1) = \frac{\pi}{4} + \frac{\pi}{4} = \frac{\pi}{2}, and constant on (−∞,0)(-\infty, 0), where f(−1)=−π2f(-1) = -\frac{\pi}{2}. It is not constant on its whole domain.

Corollaries

Definition Let II be an open interval.

  1. If f′(x)=g′(x)f'(x) = g'(x) for all xx in II, then f−gf - g is constant on II, so f(x)=g(x)+Cf(x) = g(x) + C for some constant CC.
  2. If ff is differentiable on II and ∣f′(x)∣≤M\lvert f'(x) \rvert \le M for all xx in II, then ∣f(x2)−f(x1)∣≤M∣x2−x1∣\lvert f(x_2) - f(x_1) \rvert \le M \lvert x_2 - x_1 \rvert for all x1,x2x_1, x_2 in II.
  3. If ff is differentiable on II and f′(x)≥mf'(x) \ge m for all xx in II, then f(x2)−f(x1)≥m(x2−x1)f(x_2) - f(x_1) \ge m(x_2 - x_1) whenever x1<x2x_1 < x_2 in II.

Each follows by applying the mean value theorem on [x1,x2][x_1, x_2].

In plain words A bound on the derivative gives a bound on how much the function can change. Two functions with the same derivative differ by a constant, which is why antiderivatives come with a +C+ C.

Example Suppose ff is differentiable on an open interval containing [1,4][1, 4], with f(1)=10f(1) = 10 and f′(x)≥2f'(x) \ge 2 there. Then f(4)−f(1)≥2⋅3f(4) - f(1) \ge 2 \cdot 3, so f(4)≥16f(4) \ge 16. Also, since ∣cos⁡x∣≤1\lvert \cos x \rvert \le 1, the second corollary with I=RI = \mathbb{R} gives

∣sin⁡x2−sin⁡x1∣≤∣x2−x1∣for all real x1,x2.\lvert \sin x_2 - \sin x_1 \rvert \le \lvert x_2 - x_1 \rvert \qquad \text{for all real } x_1, x_2.

Check yourself

  1. Find the value cc promised by the mean value theorem for f(x)=x2+2xf(x) = x^2 + 2x on [0,3][0, 3].
Show the answer

Since f(0)=0f(0) = 0 and f(3)=15f(3) = 15, the average rate is

15−03−0=5.\frac{15 - 0}{3 - 0} = 5.

The derivative is f′(x)=2x+2f'(x) = 2x + 2, and solving 2c+2=52c + 2 = 5 gives c=32c = \frac{3}{2}, which is in (0,3)(0, 3).

  1. Does Rolle’s theorem apply to f(x)=1x2f(x) = \frac{1}{x^2} on [−1,1][-1, 1]?
Show the answer

No. Although f(−1)=f(1)=1f(-1) = f(1) = 1, the function is not defined at 00, so it is not continuous on [−1,1][-1, 1]. In fact f′(x)=−2x3f'(x) = -\frac{2}{x^3} is never 00.

  1. Show that x3+x−1=0x^3 + x - 1 = 0 has exactly one real solution.
Show the answer

Let f(x)=x3+x−1f(x) = x^3 + x - 1. Since f(0)=−1<0<1=f(1)f(0) = -1 < 0 < 1 = f(1), the intermediate value theorem gives at least one root. If there were two roots r<sr < s, Rolle’s theorem would give cc in (r,s)(r, s) with f′(c)=0f'(c) = 0, but f′(x)=3x2+1>0f'(x) = 3x^2 + 1 > 0 for every xx. So there is exactly one.

How this lesson connects

It also uses Rolle theorem.