Understanding the Derivative
Start a path from thisLegalThis path builds one idea: a derivative measures how fast something is changing at a single instant, and that same number is the slope of a tangent line on a graph. It starts with ordinary average rates of change, shows why "change right now" needs a new idea, defines the derivative from that idea, ties it to geometry, and ends with how to read and use derivatives. Computational rules and techniques are left for later branches.
Calculus begins with a question you can already ask in algebra: how much does one quantity change when another changes? This chapter sets up the vocabulary of change, ratio, and slope. The derivative is a refined version of these ideas, so it only makes sense once they feel natural.
A rate of change compares two changes: how much an output moves relative to how much an input moves. If a car's position changes by 60 miles while time changes by 1 hour, the rate is 60 miles per hour. The word "per" is the signal that you are dividing one change by another. Every derivative is a rate of this kind, so it always carries units of "output per input."
Over an interval from x = a to x = b, the average rate of change is the total change in output divided by the total change in input:
b−af(b)−f(a)
This number describes the whole interval as if the change had happened at one steady pace. A trip that averages 60 mph may include stops and bursts of speed, and the average hides all of that. This limitation is what motivates the derivative.
On a graph, the average rate of change is the slope of the straight line through two points on the curve. That line is called a secant line. Steep secants mean fast average change, flat ones mean slow change, and downward ones mean decrease. This links the numerical idea of a rate to the geometric idea of steepness, and the link is the one the derivative will keep.
Mastery Check: A plant is 10 cm tall on day 2 and 16 cm tall on day 5. What is its average growth rate, what are the units, and what does that number look like as a line on a graph of height against day?
Averages describe intervals, but we often care about a single moment, like the speedometer reading at one instant. This chapter shows why that is a real puzzle and how shrinking an interval points toward an answer. It is the conceptual turning point of the path.
At a single instant, no time passes and no distance is covered. Speed as distance divided by time would then be 0 divided by 0, which has no meaning. Yet a speedometer clearly shows a definite value, so there must be a sensible way to talk about change at a moment. The derivative is the resolution of this tension.
Instead of asking about one instant directly, compute average rates over smaller and smaller intervals that start at the moment of interest. Each interval gives a legitimate number, because the time elapsed is small but not zero. If those numbers settle toward a single value as the interval shrinks, that value is the natural meaning of "speed right now." The idea is to watch where the averages are heading, not to divide by zero.
A limit is the value that a quantity gets arbitrarily close to as some input approaches a target, whether or not the quantity ever reaches it. It gives a precise way to say "what the averages are heading toward." You need only an intuitive grasp of limits here: closeness, trend, and stability. The formal machinery can wait for its own branch.
Mastery Check: Suppose a falling ball's average speeds over the next 1 second, 0.1 seconds, and 0.01 seconds are 14, 10.2, and 9.82 m/s. What would you say its speed is at that moment, and why isn't that a division by zero?
With the "squeeze" idea in hand, we can write it down exactly. This chapter turns the intuition into a definition and explains the notation. The aim is to read the formula as a story about shrinking intervals, not to memorize it.
Take the average rate of change from x to x + h, where h is the width of the interval:
hf(x+h)−f(x)
This is the difference quotient. It is an ordinary average rate, and h is the knob that controls how small the interval is. Everything in the definition of the derivative is built from this expression.
The derivative of f at x is what the difference quotient approaches as h shrinks toward zero:
f′(x)=limh→0hf(x+h)−f(x)
The result is a single number measuring the instantaneous rate of change of f at that input. When the limit exists, the function is said to be differentiable there. The definition is the squeeze from the previous chapter with the interval width named h.
The derivative at one point is a number, such as "the speed at t = 3 is 29 m/s." Doing the same calculation at every input produces a new function, f′, that reports the rate of change everywhere. That is why the derivative is described as a function derived from another. It tells you how the original function behaves across its whole domain, not only at one location.
The symbols f′(x), dy/dx, and df/dx all express the same idea. The Leibniz form dy/dx is read as "a tiny change in y per tiny change in x," which keeps the ratio meaning visible. Treat each as shorthand for "the instantaneous rate of change of the output with respect to the input." Knowing this prevents the notation from feeling like arbitrary code.
Mastery Check: In your own words, what is h doing in the difference quotient, and why does the derivative at a point come out as one number while the derivative of the whole function is another function?
The derivative has a visual meaning that makes it easy to picture. This chapter connects the limiting process to a line touching a curve, so you can estimate and reason about derivatives just by looking at a graph.
Fix a point on a curve and draw secant lines to a second point that moves closer and closer. As the second point approaches the first, the secant lines rotate and settle toward a single limiting line, the tangent line. Its slope is the limit of the secant slopes, which is exactly the derivative. So "instantaneous rate of change" and "slope of the tangent line" are two descriptions of the same quantity.
The derivative's sign and size describe the graph's behavior. A positive value means the curve is rising at that point, a negative value means it is falling, and a value near zero means it is nearly flat. A larger magnitude means a steeper curve. Reading these features lets you sketch the story of a function's changing behavior without computing anything.
If you zoom far into a smooth curve at one point, it looks almost like a straight line, and that line is the tangent. This is called local linearity. The derivative is the slope of the straight line that best imitates the curve near that point. It explains why the derivative is useful for approximation: close to a point, a complicated function behaves like a simple linear one.
Mastery Check: Picture a hill's profile as a curve. At the steepest uphill stretch, then at the very top, what would you say about the tangent line and the derivative at each spot?
Once the derivative is understood as a rate and a slope, the next step is to read it fluently in real situations. This chapter develops habits of interpretation, shows where the idea can break down, and points to directions beyond this foundation.
A derivative's units are always output units per input unit, such as meters per second, dollars per item, or degrees per hour. Asking "per what?" turns a bare number into a meaningful statement. A derivative of 5 for a company's cost function means that, at the current production level, one more unit adds about 5 dollars. This habit lets you carry the concept across physics, economics, biology, and beyond.
When the derivative is positive, the quantity is increasing, and when it is negative, it is decreasing. A derivative of zero means the quantity is momentarily not changing, as at a peak, a valley, or a pause. A shift from positive to negative signals that growth has turned into decline. This is the basis for finding the best or worst values of a quantity.
Not every point has a derivative. At a sharp corner, the slopes approached from the left and right disagree, and at a vertical jump or break no single tangent exists. Seeing these cases as failures of the squeeze, where the averages do not settle to one value, reinforces what the derivative actually requires. The idea fits only where a curve is locally smooth.
With the core meaning secure, several directions open up. Differentiation rules turn the limit definition into fast calculation. The second derivative measures how the rate of change itself changes, which relates to curvature and acceleration. Optimization uses zero derivatives to find the best outcomes. The integral reverses the process by accumulating rates back into totals, and multivariable calculus extends slopes to surfaces with many inputs. Each of these builds directly on the picture of instantaneous change developed here.
Mastery Check: A town's population function has a derivative of −200 people per year at this moment. Explain in plain language what that says, and what would have to happen to the derivative for the population to stop shrinking.