1.4.1 Differentiation from first principles
Suppose that is any function. Let denote any value in the domain of such that is differentiable at (that is, such that the graph of has a gradient at the point , as illustrated in Figure 17. Now consider a second point on the graph, with coordinates , where is a positive or negative number, but not zero.
The gradient of the line that passes through the two points is
and this expression can be simplified slightly to give
This expression is known as the difference quotient for the function at the value . As the second point gets closer and closer to the first point , that is, as the value of gets closer and closer to zero, the value of the difference quotient gets closer and closer to the gradient of the graph at the point . That is, it gets closer and closer to . (Remember that the value of can never actually be 0.)
So, to find a formula for , you need to consider what happens to the difference quotient for at , as gets closer and closer to zero, taking either positive or negative values as it does so. In other words, you need to find, in terms of , the limit of the difference quotient as tends to zero. You saw this procedure carried out for the function in the last section, and in the next example you’ll see it carried out for the function .
Example 1 Differentiating from first principles
Differentiate from first principles the function .
Solution
Write down the difference quotient and use the fact that .
The difference quotient for the function at is
Simplify the difference quotient. Start by multiplying out the term in the numerator.
Multiplying out gives
So
Work out what happens to the value of the difference quotient as gets closer and closer to zero.
The second term in the final expression above contains the factor , and the third term is , so as gets closer and closer to zero, both of these terms get closer and closer to zero. So the value of the whole expression gets closer and closer to the value of the first term, . That is, the formula for the derivative of the function is
The formula for the derivative of found in Example 1 tells you that, for example, the gradient of the graph of at the point with -coordinate is
and the gradient of this graph at the point with -coordinate is
You can see from Figure 18 that the gradients of the tangents to the graph at these two points do seem to be the numbers calculated using the formula.
In the next activity, you’re asked to use differentiation from first principles to find the derivative of the function .
Activity 5 Differentiating from first principles
a.Multiply out the expression . To do this, start by writing . Then replace the expression by the expansion of that was found in the solution to Example 1, remembering to enclose it in brackets. Finally, multiply out the brackets and collect like terms.
b.Hence differentiate from first principles the function .
c.What is the gradient of the graph of the function at the point with -coordinate ?
Answer
a.Multiplying out gives
b.The difference quotient for the function at is
By part (a),
Each of the terms in the final expression above, except the first term, contains the factor . Hence, as gets closer and closer to zero, each of the terms except the first term gets closer and closer to zero. So the value of the whole expression gets closer and closer to the value of the first term, . That is, the formula for the derivative is
c.By the formula found in part (b), the gradient of the graph of the function at the point with -coordinate is
(The graph of is shown below.)
In each of the three examples of differentiation from first principles that you’ve seen so far, the function was differentiable at every value of . If a function isn’t differentiable at some values of , then you can still differentiate it from first principles, but the process won’t work for the values of at which it’s not differentiable. For these values of the value of the difference quotient won’t get closer and closer to a particular number as gets closer and closer to zero (where can be either positive or negative).
For example, if the graph of a function has a sharp corner at , then the difference quotient for at will get closer and closer to a particular value when is positive, but will get closer and closer to a different value when is negative. As another example, if the graph of the function has a vertical tangent at , then the magnitude of the difference quotient for at will just keep getting larger and larger, without getting closer and closer to any particular value. In general, saying that a function is differentiable at a particular value of is the same as saying that the difference quotient for at tends to a limit as tends to zero. It must tend to the same limit for positive values of as for negative values.
The method of differentiation from first principles is summarised in the box below. The notation ‘’ means ‘the limit as tends to zero of’.
Differentiation from first principles
For any function , the derivative of is given by the equation
for each value of in the domain of for which this limit exists.
Differentiation from first principles can be used to find formulas for the derivatives of many of the functions that you’ll need to work with. However, it’s a laborious process, so usually we don’t do it! Instead, in this course you’ll get to know the formulas for the derivatives of a range of standard functions, such as , , , , and so on. You’ll also learn about ways in which you can combine these formulas to obtain formulas for the derivatives of other, related functions. For example, if you know the formulas for the derivatives of and , then you can combine them to obtain a formula for the derivative of . In these ways you’ll be able to find formulas for the derivatives of most of the functions that you’ll need to work with.
Of course, the idea of differentiation from first principles is still needed, to find the derivatives of the standard functions, to check that the rules for combining them are valid, and to differentiate functions that aren’t standard functions or combinations of standard functions.
Formulas for derivatives are powerful mathematical tools in many different situations, in both pure and applied mathematics. At this level of study, not only will you learn how to find such formulas, but you’ll also be introduced to a few of the ways in which they can be used. You’ll see many more uses of them if you go on to study further modules in mathematics, or some modules in areas such as science and economics.
The invention of calculus
The history of calculus goes back to the second half of the seventeenth century, when Isaac Newton in England and Gottfried Wilhelm Leibniz in what is now Germany both independently developed the basic ideas. Newton’s ideas were rooted in the applications of mathematics, while Leibniz’s were rooted in pure mathematics.
Newton developed the ideas of calculus starting in about 1665. He called his ideas the ‘method of fluxions’ and wrote a treatise about them in 1671, which was not published in his lifetime, although its contents circulated in manuscript form, and a publication containing the method appeared in 1704. Leibniz then independently developed similar ideas, starting in about 1674. A manuscript that he wrote in 1675 includes the notation used in integral calculus to this day, as well as a standard rule for combining derivatives, the product rule. Leibniz’s notation for differential calculus is also still used today, as you’ll see shortly. Leibniz first published work on calculus in 1684.
The two men continued to develop their ideas for the next few years. However by the early 1700s Leibniz was being accused by Newton’s associates of having plagiarised Newton’s work. The allegation was that Leibniz had seen some of Newton’s unpublished papers and had merely invented a new notation for Newton’s ideas. The ensuing bitter argument led to a Royal Society investigation, which upheld the charge. However, the investigation was largely carried out by Newton’s friends, and Newton, who was President of the Royal Society, secretly guided its report. Investigations by modern historians have shown that the accusation against Leibniz was unjust. Newton and Leibniz arrived at equivalent results following different paths of discovery.
Isaac Newton
Isaac Newton was born in Lincolnshire, and studied and worked at the University of Cambridge. He was one of the world’s greatest physicists, mathematicians and astronomers, and is remembered in particular for his work on classical mechanics. He did much of his initial work on calculus at his family home in Lincolnshire, while Cambridge University was closed due to an outbreak of plague. In his later life Newton largely abandoned physics and mathematics, and wrote theological tracts before becoming Master of the Mint, a highly-paid government official, in London. He also worked on alchemy throughout his life. He was knighted in 1705, but for political reasons rather than for his scientific work or public service.
Gottfried Wilhelm Leibniz
Gottfried Wilhelm Leibniz was born in Leipzig, and attended university there and in Altdorf. He was a universal thinker who graduated in philosophy and law, and was self-taught in mathematics. He went on to work intensively on mathematics in Paris, before accepting the position of Counsellor and librarian at the court in Hanover, where he remained for the rest of his life. While there, he worked on many different projects, making important contributions to mathematics, philosophy, theology and history. Some of Leibniz’s projects were related to his salaried position, but his employers also allowed him to work on other projects of his choosing. He was interested in formalising calculations, and constructed the first mechanical calculator that could add, subtract, multiply and divide.
Before we begin the process of building up a collection of formulas for the derivatives of standard functions, and techniques for combining them, it’s useful for you to learn an alternative notation for derivatives.




