Differentiability vs Continuity: The Key Difference

What Is Differentiability vs Continuity?

Differentiability vs Continuity is the comparison between two properties of a function at a point. A function $f$ is continuous at $x = a$ when its graph has no break there, written precisely as

$$\lim_{x \to a} f(x) = f(a).$$

A function $f$ is differentiable at $x = a$ when the derivative exists there as a single finite number, defined by the limit

$$f'(a) = \lim_{h \to 0} \frac{f(a+h) - f(a)}{h}.$$

The whole relationship runs in one direction only. Differentiability is stronger, so it forces continuity, but continuity is weaker, so it does not force differentiability.

$$\text{differentiable at } a ;\Rightarrow; \text{continuous at } a, \qquad \text{but not the reverse.}$$

That single arrow is the entire topic. This article is a focused contrast: it proves the arrow, shows exactly why it cannot be turned around, and sorts every function into one of three boxes. For the definition of the slope-limit itself, see definition of the derivative; for the broader pairing, see continuity and differentiability.

Does Differentiability Imply Continuity?

Yes, and the proof is short. This is the one solid half of Differentiability vs Continuity, and it comes straight from the limit definition.

Suppose $f'(a)$ exists. Look at the gap $f(x) - f(a)$ as $x$ approaches $a$, and rewrite it by multiplying and dividing by $(x - a)$:

$$\lim_{x \to a} \big[f(x) - f(a)\big] = \lim_{x \to a} \frac{f(x) - f(a)}{x - a} \cdot (x - a).$$

The first factor tends to $f'(a)$ because that derivative is assumed to exist, and the second factor tends to $0$. A finite number times $0$ is $0$:

$$\lim_{x \to a} \big[f(x) - f(a)\big] = f'(a) \cdot 0 = 0.$$

So $\lim_{x \to a} f(x) = f(a)$, which is exactly the definition of continuity at $a$. A function with a slope at a point cannot have a break at that point, because a break would make the difference quotient blow up rather than settle on a finite value.

The useful mirror of this statement is its contrapositive: if $f$ is not continuous at $a$, then $f$ cannot be differentiable at $a$. Any jump, hole, or asymptote instantly rules out a derivative. That is why a discontinuity is the quickest disqualifier of all.

Why Does The Converse Fail?

Continuity is a promise about the function's height; differentiability is a promise about its slope. Fixing the height at a point says nothing about whether the two sides arrive at the same slope. The derivative is itself a limit, and that limit can fail even when the function value behaves perfectly.

The test is the two one-sided derivatives. The derivative at $a$ exists only when the left-hand and right-hand difference quotients agree on one finite value:

$$f'_-(a)=\lim_{h\to0^-}\frac{f(a+h)-f(a)}{h},\qquad f'_+(a)=\lim_{h\to0^+}\frac{f(a+h)-f(a)}{h}.$$

If these two disagree, or if either one races off to infinity, the derivative does not exist even though the graph never broke. Continuity survives; differentiability does not.

What Are The Four Ways Differentiability Can Fail?

A continuous function can fail to be differentiable at a point in three geometric ways, and there is a fourth failure that comes from losing continuity itself. Each one is a concrete case of Differentiability vs Continuity where the weaker property holds and the stronger one does not.

1. A corner (the two slopes are finite but different).

The absolute-value function $f(x) = |x|$ is continuous everywhere, including at $x = 0$. Check its one-sided derivatives there:
$$f'_-(0)=\lim_{h\to0^-}\frac{|h|-0}{h}=\lim_{h\to0^-}\frac{-h}{h}=-1,\qquad f'_+(0)=\lim_{h\to0^+}\frac{|h|}{h}=+1.$$

The left slope is $-1$ and the right slope is $+1$. Both are finite, but they disagree, so $f'(0)$ does not exist. For the graph behind this, see absolute value functions.

2. A cusp (the two slopes race to opposite infinities).

The function $f(x) = x^{2/3}$ is continuous everywhere, since $x^{2/3} = \left(x^{1/3}\right)^2$ is defined for every real $x$ and tends to $0$ at the origin. Its difference quotient at $0$ is
$$\frac{h^{2/3}-0}{h}=h^{-1/3},\qquad \text{so } f'_+(0)=\lim_{h\to0^+}h^{-1/3}=+\infty,\quad f'_-(0)=\lim_{h\to0^-}h^{-1/3}=-\infty.$$

The two sides shoot off in opposite directions, meeting in an infinitely sharp point. No finite slope exists.

3. A vertical tangent (both slopes race to the same infinity).

The cube-root function $f(x) = x^{1/3}$ is continuous everywhere. Its difference quotient at $0$ is

$$\frac{h^{1/3}-0}{h}=h^{-2/3},\qquad \text{so } f'_-(0)=f'_+(0)=\lim_{h\to0}h^{-2/3}=+\infty.$$

Here both sides agree in direction, but they agree on $+\infty$, not on a number. The graph passes smoothly through the origin with a vertical tangent line, and a vertical line has no finite slope, so the function is still not differentiable there. This is the case most often mistaken for a cusp; the tell is whether the two sides share one infinity (vertical tangent) or split into two (cusp).

4. A jump (continuity itself is lost).

The three cases above keep continuity and lose only the slope. The fourth failure loses continuity first. Take the step function

$$f(x) = \begin{cases} 0, & x < 0 \ 1, & x \ge 0 \end{cases}.$$

At $x = 0$ the left limit is $0$ and the right limit is $1$, so $f$ is not continuous there. By the contrapositive proved above, it cannot be differentiable either. This is a jump discontinuity, one member of the wider family of types of discontinuity.

How Do The Three States Compare? (Reference Table)

Every function at a point lands in exactly one of three boxes. This table is the compact answer to Differentiability vs Continuity.

Table: The three states, with a representative example and what the graph does at the point.

Notice there is no fourth row for "differentiable but not continuous." That box is empty on purpose, because the implication in the second section forbids it. Differentiability always drags continuity along with it.

What About A Function That Is Continuous Everywhere But Differentiable Nowhere?

The corner, cusp, and vertical-tangent examples each fail at a single point. It is fair to ask how bad the gap between the two properties can get. The honest answer is: as bad as possible.

There exists a function that is continuous at every real number yet differentiable at no real number. The Weierstrass function, one standard form of which is

$$W(x) = \sum_{n=0}^{\infty} a^n \cos\left(b^n \pi x\right), \qquad 0 < a < 1, ; b \text{ an odd integer}, ; ab > 1 + \tfrac{3}{2}\pi,$$

is unbroken everywhere but has a corner-like roughness at every single point, so no tangent exists anywhere on it. Its graph looks jagged at every scale you zoom into, never smoothing out. This is not a trick of notation; it is a genuine continuous curve with no slope at any point, and it settled a long-standing belief that continuous functions had to be smooth except at scattered points. You will not compute with $W(x)$ in a first course, but knowing it exists is the cleanest proof that continuity is genuinely weaker than differentiability.

Why Does Differentiability Vs Continuity Work This Way?

The asymmetry is not an accident of the definitions. It falls out of what each limit is actually asking.

• Continuity asks one question: does the height settle? It only compares $f(x)$ near $a$ with the single value $f(a)$. A graph that arrives at the same height from both sides passes, no matter how sharply it turns to get there.

• Differentiability asks a harder question: does the slope settle? The difference quotient is a ratio, and it compares the two sides' rates of approach. A corner, a cusp, or a vertical tangent lets the height match while the slopes disagree or explode, so the harder question fails while the easy one passes.

• The stronger property cannot skip the weaker one. A settled slope forces a settled height, because a runaway height would wreck the ratio. That is the whole content of the one-way arrow, seen from the geometry rather than the algebra.

Put simply, "no break" is easy to satisfy and "one clean tangent" is hard, so the hard condition implies the easy one and never the reverse.

Who Discovered That Continuity Does Not Imply Differentiability?

For most of the 1800s, mathematicians assumed a continuous curve had to be smooth almost everywhere. One example demolished that belief.

Two named figures anchor the story:

• Bernard Bolzano (1781–1848, Bohemia) built an early nowhere-differentiable continuous function around 1830, though it stayed unpublished for a century.

• Karl Weierstrass (1815–1897, Prussia) gave the famous published example in 1872 and, with it, much of the rigorous limit-based foundation of calculus.

Where Does Differentiability Vs Continuity Show Up In The Real World?

The gap between "unbroken" and "smooth" is not just a classroom curiosity. It changes how real systems behave and how engineers model them.

• Physics of impacts: a bouncing ball's height is a continuous function of time, but at the instant it strikes the ground the velocity reverses, so the height graph has a corner and the velocity is undefined at that instant.

• Signal processing: a clipped or rectified audio signal stays continuous but develops corners, and the derivative spikes there, which is exactly where distortion and high-frequency content enter.

• Economics and taxation: a piecewise tax or pricing schedule is continuous across a bracket threshold but has a kink, so the marginal rate jumps at the boundary even though the total stays connected.

• Machine learning: the popular absolute-error and hinge loss functions are continuous but have a corner, so smooth gradient descent breaks at that point and practitioners use subgradients instead.

• Modelling rough nature: coastlines and the paths of particles in Brownian motion are modelled by continuous but nowhere-differentiable curves, the real-world cousins of the Weierstrass function.

Across all of these, knowing where a quantity stays continuous but loses its slope tells you exactly where the smooth-calculus toolkit stops working.

What Are The Most Common Mistakes With Differentiability Vs Continuity?

These three errors account for most lost marks on this topic, and each matches a question real students ask on Google's People Also Ask, careers360's NCERT threads, and math help forums.

Assuming that continuity implies differentiability.

Where it slips in:

A student proves a function is continuous at a point and then writes "therefore it is differentiable there," treating the two words as interchangeable.

Don't do this:

Do not run the implication backwards. Continuity is the weaker condition, so it never guarantees a derivative.

The correct way:

Only the forward arrow is valid: differentiable $\Rightarrow$ continuous. To claim differentiability, check the two one-sided derivatives agree on one finite value, as with $\lvert x\rvert$ at $0$, where they give $-1$ and $+1$ and the derivative fails.

Thinking a smooth-looking graph is always differentiable.

Where it slips in:

A student sees a curve drawn without an obvious break and assumes a slope exists everywhere on it, or sketches a cusp as if it were rounded.

Don't do this:

Do not trust the picture alone. A cusp such as $x^{2/3}$ looks like a soft point but has infinite one-sided slopes, and the Weierstrass function looks continuous yet has no slope anywhere.

The correct way:

Test the limit, not the drawing. Compute the difference quotient near the point and see whether it settles on a finite number.

Confusing a cusp with a vertical tangent.

Where it slips in:

A student sees the derivative heading to infinity and labels every such point a cusp, or every such point a vertical tangent.

Don't do this:

Do not stop at "the slope is infinite." The two cases differ in the signs of the one-sided limits.

The correct way:

Compare the two sides. If they head to opposite infinities ($+\infty$ and $-\infty$), it is a cusp, as with $x^{2/3}$. If they head to the same infinity, it is a vertical tangent, as with $x^{1/3}$. Neither is differentiable, but naming them correctly needs the sign check.

Practice Problems On Differentiability Vs Continuity

Work each one, then check against the answer. Answers are verified.

1. Is $f(x) = \lvert x - 3\rvert$ differentiable at $x = 3$?
   (Answer: no. It is continuous there, but $f'-(3) = -1$ and $f'+(3) = +1$ disagree, so it has a corner.)

2. A function is differentiable at $x = 2$. Must it be continuous at $x = 2$?
   (Answer: yes. Differentiability always implies continuity.)

3. A function is continuous at $x = 0$. Must it be differentiable at $x = 0$?
   (Answer: no. Continuity does not imply differentiability; $\lvert x\rvert$ is the counterexample.)

4. Classify the non-differentiability of $f(x) = x^{1/3}$ at $x = 0$.
   (Answer: a vertical tangent, since both one-sided derivatives tend to $+\infty$.)

5. Classify the non-differentiability of $f(x) = x^{2/3}$ at $x = 0$.
   (Answer: a cusp, since the one-sided derivatives tend to $+\infty$ and $-\infty$.)

6. The function $f(x) = 2$ for $x < 1$ and $f(x) = 5$ for $x \ge 1$ fails to be differentiable at $x = 1$. Why?
   (Answer: it is not continuous there, so by the contrapositive it cannot be differentiable.)

Where Should You Go Next After Differentiability Vs Continuity?

The one-directional arrow opens several natural doors.

1. Continuity of a function. Firm up the weaker property first, including the limit test and the kinds of breaks that end it.

2. Differentiability of a function. Go deeper on the stronger property and the one-sided-derivative checks used throughout this article.

3. Definition of the derivative. Return to the limit that powers every argument here, from the implication to each failure mode.

If your child is meeting this contrast for the first time, a live Bhanzu trainer teaches it from the graph up, so the arrow and its three failure modes feel like one idea, in the Bhanzu math program.