Derivative Foundations

One line: The derivative measures the instantaneous rate of change of a function — the slope of its graph at a single point. Key result:

f^{'} (x) = lim_{h \to 0} \frac{f ( x + h ) - f ( x )}{h}

, the limit of average rates of change as the interval shrinks to zero. Where it applies: velocity from position, marginal cost, optimisation, tangent lines, slopes of curves — anywhere "how fast is it changing?" is asked. Who/when: Developed independently by Isaac Newton (c. 1666) and Gottfried Leibniz (c. 1675); placed on rigorous limit foundations by Cauchy and Weierstrass in the 1800s.

The derivative of a function $f$ at a point $x$ is the limit of its average rate of change over an interval as that interval shrinks to zero length. Geometrically, it is the slope of the tangent line to the graph of $f$ at $x$ ; physically, it is an instantaneous rate (such as velocity). It matters because almost every quantitative science models change, and the derivative is the precise mathematical language of change. A surprising fact: although the derivative is built from a quotient that becomes the indeterminate form $0/0$ , this very limit yields exact, finite answers — the foundation of all of differential calculus.

The Core Insight

You cannot directly measure a rate at an instant because rate normally needs two points (rise over run), and an instant is one point. The derivative resolves this by computing the average rate over an interval of width $h$ , then letting $h$ approach zero. The average rates "home in" on a single limiting value — the instantaneous rate. The slope between two nearby points becomes the slope of the tangent at one point.

Formal Statement

Average Rate of Change

For a function $f$ over the interval from $x$ to $x + h$ (with $h \neq = 0$ ), the average rate of change is the slope of the secant line through the two points $(x, f (x))$ and $(x + h, f (x + h))$ :

\frac{Δ y}{Δ x} = \frac{f ( x + h ) - f ( x )}{h}

⟳This is simply "rise over run" — the change in output

f (x + h) - f (x)

divided by the change in input

(x + h) - x = h

Symbols: $f$ is the function; $x$ the base point; $h$ the interval width (a small nonzero number); $Δ y = f (x + h) - f (x)$ the output change; $Δ x = h$ the input change.

The Derivative

The derivative of $f$ at $x$ , written $f^{'} (x)$ or $\frac{df}{d x}$ , is the limit of the average rate of change as $h \to 0$ , provided this limit exists:

f^{'} (x) = h \to 0 lim \frac{f ( x + h ) - f ( x )}{h}

⟳Take equation [1] and let the interval width

h

shrink toward zero; the secant slopes approach the tangent slope.

When this limit exists, $f$ is said to be differentiable at $x$ . An equivalent form fixing a point $a$ :

f^{'} (a) = x \to a lim \frac{f ( x ) - f ( a )}{x - a}

⟳Substitute

x = a + h

into [2], so

h = x - a

and

h \to 0

becomes

x \to a

Symbols: $f^{'} (x)$ the derivative (a new function of $x$ ); $lim_{h \to 0}$ the limit operation; $a$ a fixed point of evaluation.

Differentiability Implies Continuity

If $f$ is differentiable at $x = a$ , then $f$ is continuous at $a$ . The converse is false: continuity does not guarantee differentiability.

f^{'} (a) exists ⟹ x \to a lim f (x) = f (a)

⟳Write

f (x) - f (a) = \frac{f ( x ) - f ( a )}{x - a} \cdot (x - a)

; as

x \to a

the first factor tends to

f^{'} (a)

and the second to

0

, so the product tends to

0

Why It Holds / Derivation

Deriving the Derivative of $f(x)=x^2$ from First Principles

We compute $f^{'} (x)$ for $f (x) = x^{2}$ directly from definition [2].

Step 1 — Form the difference quotient. Substitute into [1]:

\frac{f ( x + h ) - f ( x )}{h} = \frac{( x + h ) ^{2} - x ^{2}}{h}

⟳Replace

f

by the squaring rule at

x + h

and at

x

Step 2 — Expand the numerator. Since

(x + h)^{2} = x^{2} + 2 x h + h^{2}

\frac{x ^{2} + 2 x h + h ^{2} - x ^{2}}{h} = \frac{2 x h + h ^{2}}{h}

⟳The

x^{2}

terms cancel, leaving only terms containing

h

Step 3 — Simplify by cancelling $h$ (valid because

h \neq = 0

inside the limit):

\frac{2 x h + h ^{2}}{h} = 2 x + h

⟳Factor

h

from the numerator, then divide top and bottom by

h

Step 4 — Take the limit. As

h \to 0

, the term

h

vanishes:

f^{'} (x) = h \to 0 lim (2 x + h) = 2 x

⟳

2 x

is constant with respect to

h

, and

h \to 0

, so the sum tends to

2 x

Thus $\frac{d}{d x} (x^{2}) = 2 x$ . The indeterminate $0/0$ never actually occurred because we simplified before taking the limit. $■$

Gottfried Wilhelm Leibniz | 1646–1716

A German polymath — philosopher, logician, and mathematician — Leibniz developed calculus independently of Newton in the mid-1670s. His genius lay in notation: he introduced $\frac{d y}{d x}$ for the derivative and $\int$ for the integral, symbols so well-designed they remain in universal use today. Leibniz thought of $d x$ and $d y$ as "infinitesimal" (infinitely small) differences, and his fraction-like notation makes rules such as the chain rule look almost self-evident. A bitter priority dispute with Newton's supporters clouded his later years, but mathematics ultimately adopted Leibniz's symbolic language.

Building Intuition

Imagine zooming a microscope into the graph of a smooth curve at one point. The more you magnify, the straighter the curve looks — eventually indistinguishable from a straight line. That line is the tangent, and its slope is the derivative. This is local linearity: differentiable functions look linear when viewed closely enough.

Another picture: a car's speedometer. Average speed needs two odometer readings and a clock interval. But the speedometer shows speed right now — the derivative of position. It works by effectively measuring change over a tiny time interval and reporting the limiting rate.

Secant lines (grey) approaching the tangent line (gold) as $h\to 0$ on the curve $y=x^2$

Worked Examples

Instantaneous Velocity of a Falling Stone

Given: A stone's height (in metres) after

t

seconds is

s (t) = 20 - 5 t^{2}

. Find its instantaneous velocity at

t = 2 s

. Step 1 — Set up the difference quotient using definition [2] with

f = s

\frac{s ( 2 + h ) - s ( 2 )}{h} = \frac{( 20 - 5 ( 2 + h ) ^{2} ) - ( 20 - 5 ( 2 ) ^{2} )}{h}

⟳Substitute

t = 2

and

t = 2 + h

into the height formula.

Step 2 — Expand:

(2 + h)^{2} = 4 + 4 h + h^{2}

, so

20 - 5 (4 + 4 h + h^{2}) = - 20 h - 5 h^{2}

after subtracting

20 - 20 = 0

\frac{- 20 h - 5 h ^{2}}{h} = - 20 - 5 h

⟳

s (2) = 20 - 20 = 0

; cancel

h

from the numerator.

Step 3 — Take the limit as

h \to 0

v (2) = h \to 0 lim (- 20 - 5 h) = - 20

⟳The

- 5 h

term vanishes as

h \to 0

Answer:

v (2) = - 20 m/s

— the stone moves downward (negative) at

20

metres per second.

Slope of the Tangent to a Cubic

Given:

f (x) = x^{3}

. Find the slope of the tangent at

x = 2

. Step 1 — Difference quotient:

\frac{( 2 + h ) ^{3} - 2 ^{3}}{h}

⟳Apply definition [2] at the point

a = 2

Step 2 — Expand

(2 + h)^{3} = 8 + 12 h + 6 h^{2} + h^{3}

, subtract

8

\frac{12 h + 6 h ^{2} + h ^{3}}{h} = 12 + 6 h + h^{2}

⟳Use

(a + b)^{3} = a^{3} + 3 a^{2} b + 3 a b^{2} + b^{3}

with

a = 2, b = h

; then cancel

h

Step 3 — Limit as

h \to 0

f^{'} (2) = h \to 0 lim (12 + 6 h + h^{2}) = 12

⟳Both

6 h

and

h^{2}

tend to

0

Answer:

f^{'} (2) = 12

— the tangent line at

(2, 8)

has slope

12

(dimensionless slope). This matches the general rule

\frac{d}{d x} x^{3} = 3 x^{2}

, since

3 (2)^{2} = 12

Common Misconceptions

"The derivative is the fraction $0/0$ ." It is not a fraction at the instant; it is the limit of fractions. We never divide by zero — within the limit

h \neq = 0

, allowing legitimate cancellation of

h

. Only after simplifying do we let

h \to 0

. " $d y / d x$ is a ratio you can split." The Leibniz symbol

\frac{d y}{d x}

is a single symbol meaning "the derivative," not literally

d y

divided by

d x

. (In some later contexts the pieces gain meaning, but at this level treat it as one indivisible notation.) "Every continuous function is differentiable." False — continuity is necessary but not sufficient (see [4] and the counterexample below).

A Continuous Function That Is Not Differentiable

Consider $f (x) = ∣ x ∣$ (absolute value) at $x = 0$ . It is continuous everywhere, but the difference quotient gives different limits from each side:

h \to 0^{+} lim \frac{∣ h ∣ - 0}{h} = + 1, h \to 0^{-} lim \frac{∣ h ∣ - 0}{h} = - 1

⟳For

h > 0

∣ h ∣ = h

so the quotient is

+ 1

; for

h < 0

∣ h ∣ = - h

so the quotient is

- 1

Since the left and right limits disagree, the two-sided limit [2] does not exist: the graph has a sharp corner at $0$ , and no single tangent slope exists. Continuity alone is not enough.

Connections

The derivative sits at the heart of a web of ideas. It is defined using Limits and Continuity — without limits, the "shrinking interval" is meaningless. Geometrically it produces the Tangent Line and connects to Slope of a Line generalised to curves. The mechanical task of computing derivatives without first principles is governed by Differentiation Rules (power, product, quotient, chain). Its inverse operation is studied in Antiderivatives and Integration, and the deep link between the two is the Fundamental Theorem of Calculus. In physics, the derivative of position is Velocity and Acceleration.

Real-World Applications

Physics — kinematics: Velocity is the derivative of position; acceleration is the derivative of velocity. Newton's laws are stated as derivatives.
Economics — marginal analysis: Marginal cost and marginal revenue are derivatives of total cost and revenue functions, guiding profit-maximising decisions.
Engineering — control systems: PID controllers use the derivative term to anticipate and dampen rapid changes in a system's error signal.
Medicine — pharmacokinetics: The rate at which drug concentration in blood changes (its derivative) determines dosing schedules.
Machine learning — gradient descent: Training neural networks minimises a loss function by stepping in the direction of its negative derivative (gradient).
Biology — population dynamics: Growth rate $d N / d t$ of a population is modelled as a derivative in logistic and exponential models.
Climate science: The rate of change of global temperature or CO₂ concentration over time is a derivative driving policy projections.

Historical Development

c. 250 BCE — Archimedes uses tangent and area methods (the "method of exhaustion"), foreshadowing limits without formalising them.
1629 — Pierre de Fermat develops adequality, a method for finding maxima/minima and tangents that anticipates the derivative.
c. 1666 — Isaac Newton formulates "fluxions" (rates of change of "fluents") during his plague-year work at Woolsthorpe.
c. 1675 — Gottfried Leibniz independently develops differential calculus with the notation $d y / d x$ and $\int$ .
1684 — Leibniz publishes the first calculus paper, Nova Methodus, introducing differentials publicly.
1687 — Newton's Principia Mathematica applies fluxional ideas to physics, though stated geometrically.
1696 — L'Hôpital publishes the first calculus textbook, spreading Leibniz's methods across Europe.
1734 — Bishop Berkeley attacks infinitesimals as "ghosts of departed quantities," exposing the lack of rigour.
1821 — Augustin-Louis Cauchy defines the derivative rigorously via limits, and Karl Weierstrass (1860s) supplies the precise $ε$ – $δ$ foundation, ending centuries of ambiguity.

Subtleties & Limits

The notion of "instantaneous rate" raised genuine philosophical alarm. Berkeley's critique — that infinitesimals are treated as nonzero (to divide) yet zero (to discard) — was not fully answered for over a century, until limits replaced infinitesimals. Remarkably, the infinitesimal viewpoint was later rehabilitated rigorously by Abraham Robinson's non-standard analysis (1960s), which constructs a logically sound number system containing genuine infinitesimals — vindicating Leibniz's intuition. A further subtlety: functions can be continuous everywhere yet differentiable nowhere (Weierstrass's 1872 function), a "fractal" curve with corners at every scale — showing how strong the smoothness demanded by differentiability really is.

Derivative Foundations

The Core Insight

Formal Statement

Average Rate of Change

The Derivative

Differentiability Implies Continuity

Why It Holds / Derivation

Deriving the Derivative of $f(x)=x^2$ from First Principles

Gottfried Wilhelm Leibniz | 1646–1716

Building Intuition

Secant lines (grey) approaching the tangent line (gold) as $h\to 0$ on the curve $y=x^2$

Worked Examples

Instantaneous Velocity of a Falling Stone

Slope of the Tangent to a Cubic

Common Misconceptions

A Continuous Function That Is Not Differentiable

Connections

Real-World Applications

Historical Development

Subtleties & Limits

See Also

Before this

Where this leads