Chapter VII · Additional theorems in the differential and integral calculus
Maxima and minima: the first non-vanishing derivative
When , Taylor's theorem still decides between maximum, minimum and neither.
Why this matters
The second-derivative test of §§123–124 says: if and there is a maximum, and if a minimum. But it says nothing when as well, and that happens for very simple functions: , and all have , and they behave in three different ways. §153 closes this gap with the first application of Taylor’s theorem. Hardy himself remarks that the result matters more for completeness than for practice; its value is that it shows clearly why the second-derivative test works.
The idea
Near a point where the first derivatives vanish, Taylor’s theorem says that is almost exactly its leading term . So close to , the graph looks like a multiple of , and everything depends on two facts about that term:
- if is even, on both sides, so has the sign of on both sides: a minimum if it is positive, a maximum if it is negative;
- if is odd, changes sign with , so is above on one side and below on the other: neither.
Try this. For each function in the menu, read the table: which derivative is the first that is not zero? Predict maximum, minimum or neither, then check the graph. Then drag the zoom down to .
- φ(x) = x⁴
- leading term φ⁽⁴⁾(0)x⁴/4!
| k | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
|---|---|---|---|---|---|---|---|---|
| φ⁽ᵏ⁾(0) | 0 | 0 | 0 | 24 | 0 | 0 | 0 | 0 |
What to notice
As you zoom in, the blue curve and the dashed leading term become indistinguishable: near the function is, in effect, its leading term, up to a relative error that tends to . The parity of gives the shape, the sign gives the direction. For every derivative at vanishes, so the test says nothing, yet there is a minimum.
Precise statements
Recall Hardy’s definition (§123): is a maximum if for all in some interval , and a minimum similarly. These are what we now call strict local maxima and minima.
Suppose that has derivatives at , that
If is even, is a maximum when and a minimum when . If is odd, is neither a maximum nor a minimum.
In Hardy’s words: if there is to be a maximum or minimum, the first derivative that does not vanish must be of even order. The case is the familiar second-derivative test; the theorem says the same argument works one level further up whenever .
Proofs
Use Taylor’s theorem in the form of §151, which needs only the derivatives at the one point . All terms but the last vanish, and for small the last one is dominated by , whose sign we know.
Expand about
Since , the theorem of §151 gives
For small the bracket has the sign of
Because , there is an with whenever . Then is non-zero and has the same sign as .
even
Then for every , so for the difference has the sign of on both sides of : negative everywhere near (a maximum) or positive everywhere (a minimum).
odd
Then has the sign of , so takes one sign for small positive and the opposite sign for small negative . The value is exceeded on one side and not reached on the other: neither a maximum nor a minimum.
Two limitations are worth stating plainly. The test needs some non-zero derivative: if all of them vanish at , as for at , it gives no information at all. And the theorem only describes near ; a local maximum need not be the largest value of on its domain.
Worked examples and exercises
Verify the test for at , where is a positive integer.
Hint
Solution
The -th derivative is for , which is at for , while . So the first non-vanishing derivative has order and is positive. The test predicts a minimum when is even and neither when is odd. This agrees with the graphs: are with equality only at , while change sign at .
Test , where and are positive integers and , for maxima and minima at and .
Hint
Solution
Write with . By Leibniz’s rule, every derivative of order at contains a positive power of in each term and so vanishes, while .
So at : if is odd, neither. If is even, a minimum when and a maximum when . Since exactly when is odd and , the case ” even, odd, ” gives a maximum at , and every other case with even gives a minimum.
At the roles swap: if is odd, neither; if is even, a minimum when and a maximum when . For example (in the widget, with , , , ) has a maximum at (since ) and neither at .
Test , , and , , for maxima and minima at .
Hint
Solution
Let . Its derivatives at of order up to vanish (those of order cancel with the polynomial, and even-order derivatives of at are zero). The first non-zero one is , of odd order. So none of , has a maximum or minimum at .
Let . Its first non-zero derivative at is , of even order. So has a maximum at when is odd (, ) and a minimum when is even (). Check two of these in the widget.
Show that , with , has and , but that is not a maximum or a minimum in Hardy’s strict sense. What does this say about the test?
Hint
Solution
, so and . For , , so and : . But for every integer , at points as close to as we like, so is not greater or less than all nearby values. The test cannot be applied: does not exist ( oscillates near ), and there is no first non-vanishing derivative. The theorem needs that derivative to exist.
Check yourself
At we have and . Then is
The first non-vanishing derivative has odd order , so near the difference changes sign with . The sign of an odd-order derivative only says on which side is larger. No further derivatives are needed: the first non-zero one already decides.
If and , then has neither a maximum nor a minimum at .
False. has and a minimum at . When the second-derivative test is merely silent; look further up, to the first derivative that is not zero.
The first non-vanishing derivative of at is . What is ?
Even order, negative value: on both sides. A maximum.
If every derivative of at is zero, then is constant near .
False. (with value at ) has all derivatives zero at and is positive everywhere else; it has a minimum at that the test cannot detect. This is the same function that defeated Taylor’s series in the last lesson.
Hardy’s language
Summary
- At a stationary point, look for the first derivative that is not zero, say .
- Near , with , so it has the sign of .
- even: maximum if , minimum if . odd: neither.
- The test needs a non-vanishing derivative; for at it is silent.