Linearization
Linearization is the process of approximating a complicated function near a specific point by replacing it with the equation of its tangent line at that point. Because a tangent line perfectly matches both the value and the slope of a function at the point of tangency, it serves as an excellent stand-in for the function when you only need to evaluate it for inputs close to that point. This idea turns difficult calculations — like finding √4.02 or sin(0.03) — into simple arithmetic. The central insight is that a smooth curve 'looks straight' when you zoom in close enough. Formally, if f is differentiable at x = a, then near x = a the function behaves almost identically to the linear function L(x) = f(a) + f'(a)(x − a). This linear function L is called the linearization of f at a, and the expression f(x) ≈ L(x) is called a linear approximation or tangent-line approximation. The error in the approximation shrinks rapidly as x approaches a. Linearization appears throughout science and engineering. Physicists replace sin θ with θ for small angles, making pendulum equations solvable. Engineers linearize control systems around an operating point. Economists approximate marginal cost curves. Mastering linearization is also essential preparation for more advanced topics such as Taylor series, Newton's method, and differentials.
- Identify the function and the center point
- Compute f(a) and f′(a)
- Write the linearization formula
- Substitute the target input
- Assess the quality of the approximation
Identify the function and the center point
Choose the function f(x) you want to approximate and the center point x = a. The center should be a value at which you can compute f(a) exactly, and it should be close to the input value x you actually care about. For example, to estimate √4.02, choose f(x) = √x and a = 4, because f(4) = 2 is known exactly.
Compute f(a) and f′(a)
Evaluate the function and its derivative at the center. These two numbers completely determine the tangent line. Using the example: f(x) = x^{1/2}, so f′(x) = \frac{1}{2}x^{-1/2}. At a = 4: f(4) = 2 and f′(4) = \frac{1}{2}·4^{-1/2} = \frac{1}{4}.
Write the linearization formula
Substitute into L(x) = f(a) + f′(a)(x − a). For our example: L(x) = 2 + \frac{1}{4}(x − 4). This is the equation of the tangent line to y = √x at the point (4, 2). It is a simple linear expression that replaces the square-root function near x = 4.
Substitute the target input
Plug in the value you want to approximate. To find √4.02, compute L(4.02) = 2 + \frac{1}{4}(4.02 − 4) = 2 + \frac{0.02}{4} = 2 + 0.005 = 2.005. The actual value is √4.02 ≈ 2.00499..., so the error is less than 0.00001 — extremely accurate for one line of arithmetic.
Assess the quality of the approximation
The linearization is most accurate when x is close to a. The second derivative f″(x) controls the curvature of f, and curvature measures how fast the tangent line drifts away from the curve. When |f″| is large near a, the approximation deteriorates more quickly. A rough error bound comes from the second-order Taylor remainder: |error| ≤ \frac{M}{2}(x-a)^2, where M is an upper bound on |f″| between a and x.
The Key Equation
- L(x) — Linearization (same as f): The linear (degree-1) approximating function whose graph is the tangent line to f at x = a.
- f(a) — Function value at center (same as f): The exact value of the original function evaluated at the chosen center point a; this gives the y-intercept of the tangent line.
- f'(a) — Derivative at center (same as f per unit x): The instantaneous rate of change of f at x = a; this is the slope of the tangent line.
- (x - a) — Displacement from center (same as x): The horizontal distance between the target input x and the center point a; the approximation is reliable when this quantity is small.
Worked Example
Problem. Use linearization to approximate ln(1.05). Choose the center a = 1, where the natural logarithm is easy to evaluate exactly.
Answer. ln(1.05) ≈ 0.05
Origins · 1 minBrook Taylor and how this was discovered
Brook Taylor introduced the general formula for expanding functions as infinite polynomial series, now called Taylor series, in his 1715 work 'Methodus Incrementorum Directa et Inversa'. Linearization is precisely the first-order (degree-1) truncation of a Taylor series, making Taylor the direct intellectual ancestor of the concept. His work unified and extended earlier results by Newton and Leibniz on approximating functions with polynomials. Taylor was also a talented musician and painter, and he published a book on the mathematical theory of perspective in art — one of the first rigorous treatments of vanishing points and projective geometry.
Related Concepts
- The Tangent Line and the Derivative — Linearization is literally the equation of the tangent line, so a solid understanding of what the tangent line is and how the derivative defines its slope is a direct prerequisite.
- Differentials (dy and dx) — Differentials express the same approximation idea in Leibniz notation: dy = f′(a) dx. Understanding differentials reinforces the geometric meaning of linearization.
- Taylor Series — A Taylor series is a higher-order generalization of linearization. The linearization L(x) is the first-order Taylor polynomial, so this concept is the natural next step.
- Newton's Method — Newton's method for finding roots of equations uses repeated linearization: it replaces f with its tangent line at a guess, finds the root of that line, and repeats.