Tangent and Normal Lines: Reading a Curve at a Point

A tangent line describes a curve’s direction at a chosen point. A normal line passes through the same point perpendicular to that tangent. For a differentiable graph y = f(x), the derivative supplies the tangent slope; when that slope is nonzero, its negative reciprocal supplies the normal slope. [2], [3]

1. Two lines, two different directions

1. Tangent: use the derivative to find the curve’s direction at the point. [2]
2. Normal: keep the same point, but choose the direction perpendicular to the tangent. [3]
3. Important exception: a horizontal tangent has a vertical normal, so the negative-reciprocal calculation cannot be used as an ordinary finite slope. [3]

The essential information is a point and a direction. A tangent must pass through the chosen point on the curve and have the derivative’s slope there. The normal shares that point, but its direction is perpendicular. These are local constructions: finding them at one point does not determine their directions elsewhere on the curve. Consequently, a tangent-line problem should begin by identifying exactly where the line is required, rather than differentiating and treating the resulting expression as a single fixed slope. [2], [3]

2. Turn the derivative into line equations

Suppose the graph is y = f(x), the chosen input is a, and the point is P = (a, f(a)). Evaluate the derivative at that input: m = f′(a). The tangent equation is y − f(a) = m(x − a). This point-slope form keeps the two required ingredients visible: substituting x = a gives y = f(a), and the coefficient m supplies the tangent’s slope. Notice the distinction between f′(x), which can vary along the curve, and f′(a), which is the particular number needed here. [2]

For a finite, nonzero tangent slope m, perpendicularity gives m × m_normal = −1. Thus the normal equation is y − f(a) = −(1/m)(x − a). An equivalent form is (x − a) + m(y − f(a)) = 0. This second form is especially useful because it avoids dividing by m: if m = 0, it immediately becomes x = a, the vertical normal. The table compares the two constructions without hiding that exception. [3]

3. Worked example: a parabola at (1, 1)

Consider the illustrative curve y = x² at x = 1. The point is (1, 1). To calculate the derivative directly, the difference quotient is ((x + h)² − x²)/h = 2x + h for h ≠ 0. Letting h approach zero gives f′(x) = 2x, so f′(1) = 2. Applying the tangent rule gives y − 1 = 2(x − 1), or y = 2x − 1. Both the point and the evaluated derivative are necessary; using 2x as though it were a fixed line slope would not produce the required straight line. [2]

The normal slope is −1/2. Its equation is y − 1 = −(1/2)(x − 1), which simplifies to y = −x/2 + 3/2. Two checks catch common arithmetic mistakes: substituting x = 1 gives y = 1 in both lines, and multiplying their slopes gives 2 × (−1/2) = −1. The first check confirms the shared point; the second confirms perpendicularity. Merely changing 2 to −2 would reverse the sign but would not produce a perpendicular line. [2], [3]

This example also shows precisely how the tangent differs from the curve nearby. Write x = 1 + h. The parabola gives y = 1 + 2h + h², whereas the tangent gives y = 1 + 2h. Their vertical difference is exactly h². Thus, for this example, the tangent retains the constant and linear parts of the curve’s change while leaving out the quadratic part. It is a useful local straight-line description, not a replacement for the parabola everywhere. [2]

4. Vertical lines and parametric curves

Slope notation is not always the most convenient language. For a parametric curve x = x(t), y = y(t), let the chosen parameter be T and the point be (X, Y). The tangent direction is (x′(T), y′(T)), provided this vector is nonzero. A division-free tangent equation is x′(T)(y − Y) = y′(T)(x − X). The normal equation is x′(T)(x − X) + y′(T)(y − Y) = 0. These equations express direction and perpendicularity without requiring every line to have a finite slope. [1], [3]

For example, if x′(T) = 0 but y′(T) ≠ 0, the tangent equation gives x = X, while the normal equation gives y = Y: a vertical tangent and horizontal normal. If both derivatives vanish, however, the direction-vector method supplies no direction. That does not automatically prove that the geometric curve has no tangent. Another representation, such as an implicit equation, may still establish one. Distinguishing failure of a formula from failure of the geometric object prevents an important overstatement. [1], [3]

5. How the idea extends to vector curves

For a vector-valued curve r(t), the derivative r′(t) gives a tangent vector whenever it exists and is nonzero. The tangent line passes through the curve’s point parallel to that vector. Dividing by its length gives the unit tangent, T(t) = r′(t)/‖r′(t)‖. This normalization preserves direction while making the vector’s length one. It separates the direction information from the magnitude of the derivative. [1]

The principal unit normal uses a further derivative: N(t) = T′(t)/‖T′(t)‖, where T′ exists and is nonzero. Why is it perpendicular? Because T always has length one, its derivative is orthogonal to it. This construction selects a normal from the change in tangent direction, rather than simply applying a negative-reciprocal slope formula. For ordinary plane-graph exercises, the perpendicular-line equations are sufficient; for vector curves, keeping tangent vectors and unit normal vectors distinct makes the terminology more precise. [1]

Explore next

1. How can tangent lines be used to approximate function values near a known point?
2. How do you find tangent and normal lines to an implicitly defined curve?
3. What distinguishes a corner, a cusp, and a vertical tangent?
4. How does the principal unit normal describe a changing tangent direction?

Sources

Open a source or expand its cited passages.

Relevance scores are model estimates of how well each source addresses this question, not probabilities or fact-check ratings. Reasoned guidance is identified on hover.

1. 96Tangent - Wikipediaen.wikipedia.orgProvides normal-line equations, parametric formulas, and the zero-derivative qualification.›
2. 95Tangent Line -- from Wolfram MathWorldmathworld.wolfram.comProvides the central derivative-based definition of a tangent line.›
3. 94Calculus III - Tangent, Normal and Binormal Vectorstutorial.math.lamar.eduDirectly explains nonzero tangent vectors, unit tangents, and principal unit normals.›