Calculus › Section E · Integrals › Topic E.1

Antiderivatives and indefinite integrals

Every derivative fact can be run backwards: if a curve's slope is known, the curve can be rebuilt up to a vertical shift. Antiderivatives are that reversal, and they are the doorway into integral calculus.

Running the machine backwards

Sections C and D taught you to take a function and find its slope function. Now run the machine backwards: given a slope at every point, what curve produced it? Given a car's velocity record, reconstruct its position; given marginal cost, rebuild the cost function. The answers come from antidifferentiation, and because derivatives lose constants, the reconstruction always carries an unresolved shift, written \(C\). A starting value, such as the car's position at time zero, is exactly the extra fact that fixes it.

Antiderivatives and the +C family

An antiderivative of \(f\) is any function \(F\) with \(F' = f\). Since \(\frac{d}{dx}x^3 = 3x^2\), the function \(x^3\) works for \(3x^2\), and so do \(x^3 + 7\) and \(x^3 - \sqrt{5}\), because constants differentiate to zero. On an interval there are no others: by the D.2 corollary that equal derivatives mean a constant difference, the complete family of antiderivatives is \(F(x) + C\). Geometrically it is one fixed curve shape slid up and down, every member carrying exactly the same slope at every \(x\).

Indefinite integral notation

The whole family gets one symbol:

\int f(x)\,dx = F(x) + C

read as the indefinite integral of \(f\). The elongated S is the integral sign, \(f(x)\) is the integrand, and \(dx\) marks the variable being integrated, the same variable the derivative was taken against. Mind the contrast with what is coming in E.3: a definite integral with endpoints is a number, a signed area; an indefinite integral is a family of functions. The notation is deliberately shared, and the reason for that is the fundamental theorem, waiting in E.4.

The reverse power rule

Differentiating \(x^{n+1}\) gives \((n+1)x^n\), so dividing by \(n + 1\) first repairs the factor: for \(n \ne -1\),

\int x^n\,dx = \frac{x^{n+1}}{n + 1} + C

Raise the exponent by one, divide by the new exponent. It works for negative powers too: rewrite \(\frac{1}{x^2}\) as \(x^{-2}\), and \(\int x^{-2}\,dx = \frac{x^{-1}}{-1} + C = -\frac{1}{x} + C\). The gap at \(n = -1\) is real: to differentiate to \(x^{-1}\), a power would have to be \(x^0 = 1\), whose derivative is 0. That case belongs to the natural logarithm and waits for Section F. Because the sum and constant multiple rules for derivatives run backwards too, integrals split over addition and constants factor out, so polynomials integrate term by term.

Worked example: a polynomial, with the check

Integrate \(\int (6x^2 - 4x + 5)\,dx\). Term by term with the reverse power rule: \(6x^2 \to 6 \cdot \frac{x^3}{3} = 2x^3\), then \(-4x \to -4 \cdot \frac{x^2}{2} = -2x^2\), and \(5 \to 5x\). So the integral is \(2x^3 - 2x^2 + 5x + C\). Now the habit that makes this bulletproof: differentiate the answer. \(\frac{d}{dx}(2x^3 - 2x^2 + 5x + C) = 6x^2 - 4x + 5\), exactly the integrand, and the \(C\) vanishes as it must. Every antiderivative you ever write can be verified this way in seconds; do it.

The trig and exponential staples

Running Section C's rules backwards gives the small table worth memorising: \(\int \cos x\,dx = \sin x + C\), since differentiating sine yields cosine; \(\int \sin x\,dx = -\cos x + C\), the minus carrying over from cosine's derivative; \(\int \sec^2 x\,dx = \tan x + C\); and \(\int e^x\,dx = e^x + C\), the exponential being its own antiderivative as well as its own derivative. Each line is nothing but a differentiation rule read from right to left, so the way to grow this table forever is simply to keep differentiating new functions and flipping each result.

Worked example: pinning down one curve

A curve has slope function \(f'(x) = 3x^2 - 2\) and passes through the point \((1, 0)\). Which curve is it? First integrate: \(f(x) = x^3 - 2x + C\), the entire family of curves with that slope behaviour. The point selects the member: substituting, \(0 = 1 - 2 + C\), so \(C = 1\) and the answer is \(f(x) = x^3 - 2x + 1\), called the particular solution. Check: its derivative is \(3x^2 - 2\) and \(f(1) = 0\). This two-step, general family then one constant from one point, is the pattern for every initial-value problem you will meet, from here through the differential equations of Section G.

Worked example: rebuilding motion

A ball is thrown straight up from ground level at \(20\) m/s, ignoring air resistance, so its acceleration is the constant \(a(t) = -9.8\) m/s². Integrate once for velocity: \(v(t) = -9.8t + C\), and the launch condition \(v(0) = 20\) sets \(C = 20\). Integrate again for position: \(s(t) = -4.9t^2 + 20t + C'\), and starting at ground level, \(s(0) = 0\), sets \(C' = 0\). The result, \(s(t) = -4.9t^2 + 20t\), is exactly the height law we differentiated in C.2, now derived from the physics instead of handed to us. Each integration stage spent one measurement to fix one constant.

Where the constants come from

Notice the bookkeeping rule that just ruled the motion example: one integration introduces one unknown constant, and one independent fact, an initial position, an initial velocity, a point on a curve, is needed to determine it. Two integrations, two constants, two facts. Whenever you antidifferentiate, count your constants and count your given facts: with fewer facts than constants, the answer is still a family, not one function.

Common mistakes to avoid

The recurring errors: dropping the \(C\), which silently loses a whole family of answers; misreading the reverse power rule as \(x^{n+1}\) times \(n+1\), when the new exponent divides; trying it at \(n = -1\), where \(\frac{x^0}{0}\) is nonsense and the logarithm will take over in Section F; treating \(\int f(x)\,dx\) as a number, the indefinite integral is a family; forgetting that an initial condition exists exactly to fix the constant, so use it; and not checking by differentiating, the one-second verification that catches nearly every slip on this list.

Key takeaways

An antiderivative of \(f\) is any \(F\) with \(F' = f\), and they all differ by constants, so \(\int f(x)\,dx = F(x) + C\) names the whole family. The reverse power rule raises the exponent and divides by it, with \(n = -1\) reserved for the logarithm; the trig and exponential rules are Section C's table read backwards. Each integration hides one constant, fixed by one initial condition, which is how slope becomes curve and acceleration becomes position. Always verify by differentiating. Next: how adding up thin strips, areas and sums, connects to this, culminating in the fundamental theorem in E.4.

Flashcards

Flashcard

Antiderivative

Any F with F′ = f

Differentiation run backwards.

Test your understanding

Question 1 of 5

What is ∫ 6x² dx?

Question 2 of 5

Evaluate ∫ (1/x³) dx.

Question 3 of 5

What is ∫ sin x dx?

Question 4 of 5

A curve has slope 4x at every point and passes through (1, 5). What is f(x)?

Question 5 of 5

A ball is thrown straight up from ground level at 24 m/s, with constant acceleration −9.8 m/s² (no air resistance). What is its height s(t) in metres?

Standards: 6

← Previous topicNext topic →