Difference Quotients for Cube Roots Away from Zero

Why Cube Roots Need a Special Difference Quotient

For a polynomial, expanding the difference between two nearby function values often makes the difference quotient straightforward to simplify. A cube root does not expand in the same way. Instead, its defining relationship to cubes gives us a useful factorization.

This tutorial studies \(f(x)=\sqrt[3]{x}\), the real cube-root function. At any nonzero input, factoring a difference of cubes turns the quotient into one whose limit can be evaluated directly. At zero, that same factorization does not produce a finite limiting value: the quotient becomes unbounded.

The distinction matters because a derivative at an input is a finite real limit. A quotient that grows without bound does not give a derivative, even if it has a clear pattern of growth on both sides.

The Difference Quotient and the Factoring Identity

For \(f(x)=\sqrt[3]{x}\), the difference quotient at \(a\) is

The identity we will use is the difference-of-cubes factorization. For any real \(u\) and \(v\),

To see why it is true, multiply out the right-hand side:

In the difference quotient, take \(u=\sqrt[3]{a+h}\) and \(v=\sqrt[3]{a}\). Then \(u^3=a+h\), \(v^3=a\), and therefore \(u^3-v^3=h\). The identity gives a factorization of the numerator that contains the denominator \(h\).

Deriving the Formula at a Nonzero Input

Suppose \(a\ne 0\). For \(h\ne 0\), use the identity with \(u=\sqrt[3]{a+h}\) and \(v=\sqrt[3]{a}\):

Dividing this equality by \(h\) in the difference quotient, or equivalently using the displayed factorization of \(h\), gives

This equality is valid for \(h\ne 0\). In particular, the factorization has not asked us to set \(h=0\) in the original quotient. As \(h\to 0\), continuity of the cube-root function gives \(\sqrt[3]{a+h}\to\sqrt[3]{a}\). Thus the denominator on the right tends to \(3(\sqrt[3]{a})^2\).

Because \(a\ne 0\), its cube root is nonzero, so \(3(\sqrt[3]{a})^2>0\). The denominator therefore tends to a nonzero number, and the reciprocal tends to its reciprocal. This proves the derivative formula:

Since \((\sqrt[3]{a})^2=a^{2/3}\), the same value can also be written \(1/(3a^{2/3})\), where \(a^{2/3}\) is understood as the square of the real cube root of \(a\). This interpretation works for negative \(a\) as well: the cube root is real, and its square is positive.

Worked Examples at Nonzero Inputs

Why Zero Must Be Analyzed Separately

The nonzero-input argument depends on the limiting denominator \(3(\sqrt[3]{a})^2\) being nonzero. At \(a=0\), that expression is zero, so the theorem's reciprocal-limit step does not apply. We must return to the difference quotient and analyze it directly.

Since \(f(0)=0\), for \(h\ne0\) the quotient at zero is

The last equality follows from \(h=(\sqrt[3]{h})^3\). As \(h\to0\) through nonzero values, \((\sqrt[3]{h})^2\to0\) while remaining positive. Its reciprocal grows without bound in both directions of approach. More precisely, for any \(M>0\), if \(0<|h|<1/M^{3/2}\), then \(|\sqrt[3]{h}|<1/\sqrt{M}\), so \(1/(\sqrt[3]{h})^2>M\). Therefore the quotient tends to \(+\infty\), not to a finite real number. Consequently \(f'(0)\) does not exist as a real derivative.

A Common Mistake: Using the Formula at Zero

A tempting but invalid step is to take the nonzero-input formula and substitute \(a=0\):

then conclude that the derivative at zero is zero, or that the formula gives a derivative there. The displayed expression is undefined: it divides by zero. The formula was proved only when \(a\ne0\), precisely because that condition ensured its limiting denominator was nonzero.

The correction is to use the definition at zero rather than extend a formula beyond the inputs for which it was derived. The quotient there is \(1/(\sqrt[3]{h})^2\), and it grows without bound as \(h\to0\). This is not a finite derivative. The earlier material on vertical tangents distinguishes this type of unbounded behavior from an ordinary derivative value; it does not turn an infinite limit into a real derivative.

What the Factorization Establishes

The difference-of-cubes method answers two related questions. First, it proves differentiability at every nonzero input by reducing the difference quotient to a reciprocal with a nonzero limiting denominator. Second, it identifies why zero cannot be included in that argument: the limiting denominator becomes zero, and direct analysis shows the quotient is unbounded.

The nonzero derivative values are all positive, including at negative inputs, because the denominator is three times a nonzero square. As the input approaches zero through nonzero values, the formula's denominator becomes small, consistent with the quotient becoming unbounded at zero. This observation is useful for interpreting the computation, but the existence or nonexistence of a derivative at any particular input is established by the corresponding difference-quotient limit.